SAP NetWeaver

Master Data Management (MDM)

MDM Import Manager

Reference Guide

Release: MDM 7.1 SP08

October 6, 2011

SAP AG • Dietmar-Hopp-Allee 16 • 69190 Walldorf Germany • +49 1805 34 34 34 • fax +49 1805 34 34 20
Copyright © 2002-2011 SAP AG. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or for any purpose
without the express permission of SAP AG. The information contained herein may be
changed without prior notice.

Some software products marketed by SAP AG and its distributors contain proprietary
software components of other software vendors.

Microsoft, Windows, Excel and Access are registered trademarks of Microsoft Corporation.

IBM, DB2 and DB2 Universal Database are trademarks or registered trademarks of IBM
Corporation in the United States and/or other countries.

Oracle is a registered trademark of Oracle Corporation.

HTML, XML, XHTML and W3C are trademarks or registered trademarks of W3C®, World
Wide Web Consortium, Massachusetts Institute of Technology.

Java is a registered trademark of Sun Microsystems, Inc. JavaScript is a registered

trademark of Sun Microsystems, Inc., used under license for technology invented and
implemented by Netscape.

The Sentry Spelling-Checker Engine Copyright © 1994-2004 Wintertree Software Inc. All
rights reserved.

QuarkXPress technology used under license from Quark, Inc. © 1986-2004 Quark
Technology Partnership. All rights reserved. Quark, QuarkXPress and QuarkXPress
Passport are trademarks of Quark, Inc. and all applicable affiliated companies, Reg. U.S.
Pat. & Tm. Off. and in many other counties.

SAP, R/3, mySAP, mySAP.com, xApps, xApp, SAP NetWeaver and other SAP products and
services mentioned herein as well as their respective logos are trademarks or registered
trademarks of SAP AG in Germany and in several other countries all over the world. All other
product and service names mentioned are the trademarks of their respective companies.
Data contained in this document serves informational purposes only. National product
specifications may vary.

These materials are subject to change without notice. These materials are provided by SAP
AG and its affiliated companies (“SAP Group”) for informational purposes only, without
representation or warranty of any kind, and SAP Group shall not be liable for errors or
omissions with respect to the materials. The only warranties for SAP Group products and
services are those that are set forth in the express warranty statements accompanying such
products and services, if any. Nothing herein should be construed as constituting an
additional warranty.

09 08 07 987654321

SAP AG
Dietmar-Hopp-Allee 16
69190 Walldorf
Germany
+49 1805 34 34 34 tel
+49 1805 34 34 20 fax
info@sap.com email
http://www.sap.com web

Printed in the United States of America.

ii MDM Import Manager Reference Guide

Contents
Introduction .............................................................................. 13
Welcome .......................................................................................... 15
MDM Import Manager Overview ...................................................... 16
Data Import Fantasy..................................................................... 17
Data Import Reality ...................................................................... 18
Challenge: Combine Information from Multiple Tables ......................... 19
Challenge: Create a Pivot Table .......................................................... 19
Challenge: Create a Reverse Pivot Table ............................................ 20
Challenge: Map Fields ......................................................................... 20
Challenge: Map Values ........................................................................ 20
Challenge: Convert Data Types ........................................................... 21
Challenge: Create Measurement Values.............................................. 21
Challenge: Convert Data Values .......................................................... 22
Challenge: Split a Single Field to Expand Hierarchy ............................ 23
Challenge: Split a Single Field into Multiple Fields ............................... 23
Challenge: Split a Single Field into Multiple Data Values ..................... 24
Challenge: Combine Two Fields to Build a Hierarchy........................... 25
Challenge: Combine Fields to Create Value Combinations .................. 25
Challenge: Combine Fields to Reconstitute Hierarchy ......................... 26
Challenge: Combine Fields to Merge Values ....................................... 27
Challenge: Combine Fields to Create a Multi-Valued Field .................. 27
Challenge: Import Attribute Data as Name/Value Pairs ........................ 28
Challenge: Match Records................................................................... 28
Key Advantages ........................................................................... 29
About This Guide ............................................................................. 31
What You Will Find in This Guide ................................................ 31
What You Will Not Find in This Guide .......................................... 31
Assumptions ................................................................................ 31
Screen Captures .......................................................................... 32

Part 1: Basic Concepts............................................................ 33

What is Parametric Import?.............................................................. 35
The Current Tables ...................................................................... 36
Non-Main Table Imports ...................................................................... 37
Relationship Imports ............................................................................ 37
Table Joins and Lookups ............................................................. 38
Pivoting and Reverse Pivoting ..................................................... 38
Partitioning ................................................................................... 38
Value Mapping ............................................................................. 39
Value Conversion ......................................................................... 39
Record Matching .......................................................................... 39
Import Maps ................................................................................. 39
Record Filtering............................................................................ 40

MDM Import Manager Reference Guide iii

Part 2: Guided Tour ................................................................. 41
Starting Import Manager ................................................................... 43
Connecting Import Manager to an MDM Repository..................... 43
Setting Up Unencrypted Repository Connections ................................. 43
Setting Up Secure Repository Connections .......................................... 44
Connecting Import Manager to Source Data ................................ 45
Source Properties ................................................................................ 45
Getting the Next Source Data File ................................................ 46
Connecting to a New Data Source ............................................... 46
Data Staging ................................................................................. 47
Excel-Related Import Challenges ................................................. 48
MDM Import Manager Main Window ................................................ 49
Switching Layout .......................................................................... 49
Selecting the Current Tables ........................................................ 50

Part 3: Import Workflow .......................................................... 51

Working with the MDM Import Manager ........................................... 53
MDM Import Manager at a Glance ............................................... 54
Source Hierarchy Pane ........................................................................ 55
Destination Hierarchy Pane .................................................................. 55
Source Preview Tab ............................................................................. 56
Partition Field/Value Tab ...................................................................... 57
Map Fields/Values Tab......................................................................... 58
Match Records Tab ..............................................................................59
Import Status Tab................................................................................. 60
Status Bar ............................................................................................ 61
Configuration Options ....................................................................... 62
Import Maps...................................................................................... 69
Import Map Operations ..................................................................... 70
Opening and Closing Maps .......................................................... 70
Saving Maps ................................................................................. 72
Importing and Exporting Map Files ............................................... 73
Updating Maps for MDM 7.1 ......................................................... 74

Part 4: Browsing Source Data ................................................ 77

Overview .......................................................................................... 79
The Source Preview Tab .................................................................. 80
The Source Hierarchy Tree .............................................................. 81
Tables, Fields, and Data Values ................................................... 82
Table Joins and Lookups .............................................................. 82
Field Types in the Source Hierarchy Tree .................................... 83
XML Source Files and the Source Hierarchy Tree ....................... 84
Source Hierarchy Operations ........................................................... 85
Displaying Distinct Values ............................................................ 86
Removing Distinct Values ............................................................. 88
Setting The Case Sensitivity Option ............................................. 88

iv MDM Import Manager Reference Guide

 Adding an Empty Field ................................................................. 89
Adding a Clone Field .................................................................... 89
Renaming a Field ......................................................................... 90
Deleting a Field ............................................................................ 91
Joining Tables .............................................................................. 91
Looking Up a Field ....................................................................... 93
Splitting a Field ............................................................................ 94
Building a Hierarchy ..................................................................... 95
Creating a Compound Source Field ............................................. 96
Creating a Pivot Table.................................................................. 96
Creating a Reverse Pivot Table ................................................... 96

Part 5: Table Joins and Lookups ........................................... 97

Overview .......................................................................................... 99
Multiple Tables of Source Data .................................................. 100
Table Joins and Lookups ........................................................... 101
Joining Tables ............................................................................ 103
Looking Up a Field ..................................................................... 104

Part 6: Pivoting/Reverse Pivoting ........................................ 107

Overview ........................................................................................ 109
Pivoting ...................................................................................... 110
Challenge: Single Pair of Metadata/Data Fields ................................. 111
Challenge: Metadata Field Combination ............................................ 112
Challenge: Data Field Combination.................................................... 113
Challenge: Single Name/Value Field Pair .......................................... 114
Challenge: Multiple Name/Value Field Pairs ...................................... 115
Challenge: Multi-Valued Information .................................................. 116
Challenge: No Key Fields .................................................................. 117
Reverse Pivoting ........................................................................ 118
Challenge: Single Set of Data Fields.................................................. 119
Challenge: Metadata Field Combinations........................................... 120
Challenge: Multiple Sets of Data Fields ............................................. 121
Pivoting and Reverse Pivoting Operations ..................................... 122
Creating a Pivot Table................................................................ 123
Example: Single Pair of Metadata/Data Fields ................................... 127
Example: Metadata Field Combination .............................................. 128
Example: Data Field Combination...................................................... 129
Example: Single Name/Value Field Pair............................................. 130
Example: Multiple Name/Value Field Pairs ........................................ 131
Example: [Natural Order] Pivots......................................................... 132
Creating a Reverse Pivot Table ................................................. 133
Example: Single Set of Data Fields .................................................... 135
Example: Metadata Field Combinations ............................................. 136
Example: Multiple Sets of Data Fields................................................ 137
Imagining Horizontal Pivots........................................................ 138
Pivoting vs. Adding Horizontal Pivots ................................................. 138
Reverse Pivoting vs. Removing Horizontal Pivots .............................. 140

MDM Import Manager Reference Guide v

Part 7: Splitting Fields ........................................................... 143
Overview ........................................................................................ 145
Splitting Fields ................................................................................ 146
Splitting into Hierarchy ................................................................ 146
Splitting into Multiple Fields ........................................................ 148
Setting the Split Delimiters.......................................................... 150

Part 8: Classification Data Import ........................................ 153

Overview ........................................................................................ 155
Building a Hierarchy ....................................................................... 156
Parent/Child Value Pairs w/ Node Name .................................... 156
Parent/Child Value Pairs w/o Node Name .................................. 157
Hierarchy Codes w/ Node Name ................................................ 157
The Create Hierarchy Field Command ........................................ 158
Taxonomy Import............................................................................ 161
Importing the Hierarchy .............................................................. 162
Importing the Attributes............................................................... 163
Importing the Attribute Text Values ............................................ 165
Importing Attribute Links ............................................................. 166
Attribute Data Import as Name/Value Pairs .................................... 167
The Attribute Display Configuration Option ................................ 168
The [Attribute/Value] Hierarchy Field ......................................... 168
The Attribute/Value Hierarchy..................................................... 169

Part 9: Partitioning Fields/Values ........................................ 171

Overview ........................................................................................ 173
Partitioning and Field Mapping ................................................... 173
Source Field Partitioning .................................................................... 174
Destination Field Partitioning .............................................................. 176
Source and Destination Field Partitioning ........................................... 178
Field Value Partitioning ...................................................................... 179
Distinct Value Lists and Value Combinations ............................. 180
Record Subsets .......................................................................... 182
Partitions in the Source and Destination Hierarchies .................. 183
The Partition Field/Value Tab ..................................................... 184
Field and Value Partitioning ........................................................ 185
Inheritance .................................................................................. 185
Multiple Partitions ....................................................................... 186
Nested and Combined Partitions ................................................ 187
Nested vs. Combined ......................................................................... 189
To Combine or Not to Combine .......................................................... 191
Merged Values ................................................................................... 192
Mathematical Analysis ................................................................ 193
Infinite Restructuring Flexibility ................................................... 194
Splitting vs. Partitioning: A Comparison ...................................... 194

vi MDM Import Manager Reference Guide

 Partitioning Operations .................................................................. 196
Partitioning by a Single Field or Attribute ................................... 197
Partitioning by Multiple Items ..................................................... 198
Combining Partitions .................................................................. 199
Setting the Partition Combine Delimiter ..................................... 200
Splitting a Partition ..................................................................... 201
Removing Partitions ................................................................... 202
Reordering Partitions ................................................................. 202
Restoring Inheritance ................................................................. 203

Part 10: Mapping Fields/Values ............................................ 205

Overview ........................................................................................ 207
The Map Fields/Values Tab ....................................................... 208
The Field Mapping Grids.................................................................... 209
The Value Conversion and Mapping Grids......................................... 210
Value Mapping ........................................................................... 212
Eliminating Value Differences ............................................................ 213
Collapsing Source Value Variations ................................................... 214
Mapping to Multi-Valued Destination Fields ....................................... 215
Mapping to Hierarchy Destination Fields ............................................ 216
Mapping to Object Lookup Destination Fields .................................... 217
Mapping to Main Table Lookup Destination Fields ............................. 218
When Value Mapping Is Required ..................................................... 220
Value Conversion ....................................................................... 221
Automatic Trimming of Leading and Trailing Spaces ......................... 221
Automatic Data Type Conversion ...................................................... 222
Measurement Value Creation ............................................................ 222
Source Field-Level Settings ....................................................... 224
Destination Field-Level Settings ................................................. 225
Hierarchy Values ........................................................................ 226
Hierarchy Representation in Source Data .......................................... 227
Importing Into a Hierarchy Lookup Field ............................................ 228
Importing Into the Hierarchy Lookup Table ........................................ 228
NULL Values .............................................................................. 230
Mapping To and From NULL Values .................................................. 230
Converting To and From NULL Values .............................................. 231
Missing Source Data Fields ............................................................... 231
The NULL Handling Options .............................................................. 231
The Flat NULL Handling Option ......................................................... 232
The Hierarchy NULL Handling Option ................................................ 233
The NULL Interpretation Option ......................................................... 236
The Case Sensitivity Option ....................................................... 237
Field Mapping Delimiters............................................................ 238
Value Conversion Filters ............................................................ 238
Lookups Consisting of Multiple Display Fields ........................... 241
Compound Lookup Fields .................................................................. 241
Mapping From a Single Corresponding Source Field ......................... 242
Mapping From Multiple Corresponding Source Fields ........................ 242
Promoted Lookup Table Display Fields .............................................. 243

MDM Import Manager Reference Guide vii

 Map-Within-a-Map (Create Compound Field Command) ..................... 243
Qualified Lookup Fields and Lookups-Within-Lookups ....................... 244
Value Mapping and Lookup Table Display Fields ............................... 246
Multi-Valued Destination Fields .................................................. 250
Field Mapping to Multi-Valued Fields .................................................. 250
Field Mapping to Coupled Numeric Attributes..................................... 250
The Multi-Valued Update Option ........................................................ 251
The Qualified Update Option .............................................................. 253
Product Applications and Automotive Parts Data................................ 257
The Tuple Update Option ................................................................... 259
Multilingual Destination Fields .................................................... 265
Object Lookup Destination Fields ............................................... 265
The Data Group Display Option ......................................................... 265
Writeable Once Destination Fields ............................................. 266
GM Time Fields .......................................................................... 267
The MDIS Unmapped Value Handling Option ............................ 268
Field Mapping Operations .............................................................. 269
Source Field Mapping................................................................. 270
Mapping Fields ........................................................................... 271
Field Mapping Grid Icons ................................................................... 271
Source Fields ..................................................................................... 271
Destination Fields............................................................................... 271
Unmapping Fields....................................................................... 273
Automapping Source Fields........................................................ 274
Finding Matching Destination Fields ........................................... 274
The Find Matching Field Options ................................................ 275
Creating a Compound Source Field ........................................... 275
Setting the Field Mapping Delimiters .......................................... 276
Setting The Case Sensitivity Option ........................................... 278
Setting the Flat NULL Handling Option....................................... 279
Setting the Hierarchy NULL Handling Option ............................. 280
Applying Field-Level Value Conversion Filters ........................... 281
Filtering Unlinked Attributes ........................................................ 283
Adding a New Destination Attribute ............................................ 284
Setting the NULL Interpretation Option....................................... 285
Setting the Multi-Valued Update Option ..................................... 286
Setting the Qualified Update Option ........................................... 287
Setting the Tuple Update Option ................................................ 289
Setting the Data Group Display Option ....................................... 290
Setting the MDIS Unmapped Value Handling Option ................. 291
Value Conversion and Mapping Operations ................................... 294
Source Value Conversion and Mapping ..................................... 295
Source Values Grid ............................................................................ 296
Destination Values Grid...................................................................... 297
Editing Source Values ................................................................ 297
Single-Valued Destination Field ......................................................... 297
Multi-Valued Destination Field ............................................................ 298
Mapping Source Values.............................................................. 299
Mapping a Source Value .................................................................... 303

viii MDM Import Manager Reference Guide

 Unmapping Values ............................................................................ 304
Adding New Destination Values ......................................................... 304
Adding New Destination Hierarchy..................................................... 306
Removing New Values ...................................................................... 309
Automapping Source Values ............................................................. 309
Mapping to a Compound Lookup Field .............................................. 310
Mapping to a Coupled Numeric Attribute............................................ 312
Mapping to an Object Lookup Destination Field......................... 312
Using the Data Groups Hierarchy ...................................................... 313
Adding a New Data Group ................................................................. 313
Applying Value-Level Value Conversion Filters ......................... 314
Restoring Converted Values .............................................................. 314
Setting to Original Values .................................................................. 314
Setting Values to NULL ..................................................................... 315
Clearing Values ................................................................................. 315
Applying Operators ............................................................................ 316
Adjusting Measurement Units ............................................................ 317
Normalizing Values ............................................................................ 318
Changing the Case of Values ............................................................ 318
Replacing Values............................................................................... 318
Accepting Truncated Values .............................................................. 320
Expanding Year Ranges .................................................................... 321
Expanding Month Qualifications......................................................... 322
Finding Unmapped Source Values ............................................ 323
Reordering Mapped Destination Values..................................... 324
Finding Mapped Destination Values .......................................... 325
Selecting Source Values in a Hierarchy ..................................... 325
Handling Missing Source Data ................................................... 327
Mismatches Between Number of Header and Record Fields ............. 327

Part 11: Matching Records ................................................... 329

Overview ........................................................................................ 331
The Match Records Tab............................................................. 332
The Value Matching Lists................................................................... 333
The Default Import Actions Grid ......................................................... 334
The Record Matching Grids ............................................................... 336
Record Matching Concepts ........................................................ 338
Value Matches ................................................................................... 338
Record Matches ................................................................................ 339
Match Level ....................................................................................... 340
Match Type........................................................................................ 340
Example: Single Matching Field ......................................................... 342
Example: Single Matching Field Combination .................................... 342
Example: Multiple Matching Fields ..................................................... 343
Example: Multiple Matching Fields with Field Combination ................ 344
Value Mapping and Record Matching: A Comparison ............... 345
Record Matching and Lookup Table Display Fields............................ 345
Import Actions ............................................................................ 346
Import Action Inheritance ................................................................... 347
Source and Destination Record Import Actions .................................. 348
Active Records and Inheritance ......................................................... 349

MDM Import Manager Reference Guide ix

 Source Record Merging .............................................................. 350
The Key Field ..................................................................................... 350
The Merge Source Records Option .................................................... 351
The Merge Qualified Links Option ...................................................... 352
To Merge or Not to Merge .................................................................. 354
Merging Qualified Links and the Qualified Update Option................... 355
Record Filtering .......................................................................... 356
Full Refresh vs. Net Change Synchronizations ................................... 356
Filtered Match Level and Match Types ............................................... 357
Record Filtering Behavior ................................................................... 358
The Virtual [Match Level/Type] Field.......................................... 359
Updating the Record Matching Field .......................................... 359
Normalized Record Matching Field............................................. 360
Mask Updates and Record Replacement ................................... 361
Key Mappings and Record Replacement ................................... 362
Checkout and Record Replacement ........................................... 362
Checkout and Record Matching ................................................. 363
Record Matching on Non-Sortable Fields ................................... 363
Record Matching Operations .......................................................... 364
Adding a Single Matching Field .................................................. 365
Adding Multiple Matching Fields ................................................. 365
Combining Matching Fields ........................................................ 366
Splitting a Matching Field Combination ...................................... 366
Removing Matching Fields ......................................................... 366
Setting the Key Field .................................................................. 367
Setting the Active Records ......................................................... 367
Setting the Record Filtering Criteria............................................ 368
Setting a Default Import Action ................................................... 370
Setting a Source Record Import Action ...................................... 371
Setting a Destination Record Import Action ................................ 372

Part 12: Importing Data ......................................................... 373

Import Execution............................................................................. 375
Monitoring Import Status............................................................. 375
Performing the Import ................................................................. 377
Refreshing the Record Matching ................................................ 377
Import, Checkout, and Workflow..................................................... 378
The Auto-Join Checked Out Records Option ............................... 378
The Checkout/Workflow Option .................................................. 379
Checkout and Workflow Failure Conditions ................................ 380
A Note about Import and Validations .......................................... 381

Part 13: Importing Relationships ......................................... 383

Importing Parent/Child Relationship Links ...................................... 385
Relationship Configuration Options ............................................ 386
Aggregation Option ............................................................................ 387

x MDM Import Manager Reference Guide

 Update Options.................................................................................. 387
Relationship File Format ............................................................ 388

Part 14: Multilingual Support ................................................ 391

Multilingual Basics ......................................................................... 393
Language Layers ....................................................................... 393
Language Inheritance ................................................................ 394
Multilingual Operations .................................................................. 396
Connecting to the MDM Import Manager ................................... 396
Field Mapping to Multilingual Destination Fields ........................ 397
The Expand Multilingual Fields Option ................................................ 397
As Multiple Fields Setting ................................................................... 397
As Field and Language Setting ........................................................... 398
Value Mapping to Multilingual Destination Values ..................... 399
Importing Multilingual Data into Lookup Tables ......................... 400
Import into Flat Tables ....................................................................... 402
Import into Hierarchy Tables .............................................................. 402
Record Matching and Multilingual Destination Fields ................. 403
Source Record Merging and Multilingual Data ........................... 403

Part 15: Remote Systems and MDM .................................... 405

Introduction .................................................................................... 407
What is a Remote System?........................................................ 407
Key Mapping .............................................................................. 407
[Remote System] and [Remote Key] Fields ............................... 408
Key Mapping Behavior ............................................................... 409
Key Mapping Reassignment ...................................................... 410
What is a Port? .......................................................................... 411
Remote System Operations ........................................................... 412
Connecting to a Remote System ............................................... 412
Connecting to a Port .................................................................. 413
Field Mapping ............................................................................ 413
Value Mapping ........................................................................... 414
Record Matching ........................................................................ 414
Lookup Table Imports Using the [Remote Key] Field ......................... 414
[Remote Key] Values and Lookup Field Automapping ............... 415
[Remote Key] Values and Record Replacement ........................ 416

Part 16: Reference ................................................................. 417

Command Reference ..................................................................... 419
File Menu ................................................................................... 419
View Menu ................................................................................. 419
Source Menu .............................................................................. 420
Fields Menu ............................................................................... 420
Values Menu .............................................................................. 422

MDM Import Manager Reference Guide xi

 Records Menu ............................................................................ 422
Configuration Menu .................................................................... 423
Help Menu .................................................................................. 423
Partition Field/Value Tab Buttons ............................................... 423
Map Fields/Values Tab Buttons .................................................. 424
Match Records Tab Buttons ....................................................... 425
Miscellaneous Context Menus .................................................... 425
Display and Format Conventions.................................................... 426
Glossary ......................................................................................... 429
Index............................................................................................... 430

xii MDM Import Manager Reference Guide

INTRODUCTION

This part of the reference guide provides an overview of the MDM

Import Manager, including its distinctive features, terminology, and
overall design philosophy.

13
Welcome
Congratulations on choosing SAP NetWeaver MDM. The MDM Import
Manager is one of the powerful suite of MDM tools for next-generation
master data management and single-source, cross-media catalog
publishing.
Most MDM customers have worked to painstakingly create an MDM
repository that contains from tens of thousands to millions of products.
The major impediment to creating these large repositories has been the
complexity of aggregating – and normalizing – rich electronic master
data.
Because of the shortcomings of existing import tools, such master data
aggregation has required that source data be meticulously pre-
processed into a format that can be imported, a labor-intensive and
manual process that is very slow and very, very expensive.
By contrast, the MDM Import Manager is a GUI tool that eliminates most
of this pre-processing and provides a WYSIWYG interface for cleansing,
normalizing, and restructuring master data as part of the import process
itself. Moreover, it supports the import of electronic master data from
any tabular or XML format, with a consistent user interface and identical
functional capabilities regardless of the source application file format.
Just as MDM’s patent-pending indexing and parametric search
technology broke through the SQL performance bottleneck to deliver
performance that is 100–1000 times faster than SQL alone, the MDM
Import Manager radically simplifies and streamlines the import process
to break through the “master data bottleneck,” dramatically reducing the
time and expense associated with master data aggregation while
simultaneously increasing accuracy and throughput by up to two or
more orders of magnitude.

MDM Import Manager Reference Guide 15

MDM Import Manager Overview
Rich master data is the essential lifeblood of a truly usable MDM
repository, which must contain not only basic transactional data about
each product (e.g. a part number, price, and short description), but also
product attribute information, long descriptions, images, PDFs, and
even more-detailed descriptive information, such as marketing text,
features and bullets, and so on.
Moreover, in a master data management system such as MDM that
supports single-source cross-media publishing, the master data must be
properly structured to be fully searchable in a published electronic
catalog, and also sufficiently rich and of sufficiently high-quality to
satisfy the requirement for product families and structured table layouts
in a published paper catalog. Finally, for the repository to be of any use,
the master data must be comprehensive and up-to-date.
Unfortunately, most of what passes for master data is really just raw
transactional data; truly rich master data usually does not exist, and
often must be painstakingly created from scratch – and at great
expense – for each MDM repository. And even if the raw master data
exists in electronic format, it often resides in hundreds if not thousands
of data sources in a variety of incompatible formats and structures that
are completely different from that of the MDM repository.
Aggregating raw product data into something that deserves to be called
rich master data has historically been an expensive and almost
insurmountable challenge, one that requires a tremendous amount of
manual preprocessing (which must be repeated for each source data
file), as well as direct user intervention for exception handling on a
record-by-record basis when the records are actually imported.
All this is history with the MDM Import Manager, which solves the import
challenge once and for all. It is a powerful GUI tool that allows you to
import data from virtually any flat or relational electronic source file (e.g.
Access, Excel, delimited text, SQL, XML, and any ODBC-compliant
source), and to completely restructure, cleanse, normalize, and
rationalize raw data into rich master data as part of the import process
itself.
The MDM Import Manager features an intuitive and highly efficient user
interface coupled with innovative, patent-pending functionality that
reduce the time and effort required to import raw electronic data by a
factor of 100–1000 times. Instead of performing the import a record at a
time, as with other import tools, it allows you to preprocess the source
data a field at a time, dramatically reducing from hours to literally
minutes the amount of time spent processing each source data file.
Perhaps most importantly, the MDM Import Manager produces
cleansed, normalized, and rationalized master data that is 100%
accurate.

16 MDM Import Manager Reference Guide

DATA IMPORT FANTASY
Historically, most import tools that have been used to aggregate master
data have operated on one single record at a time. Each field in the
source (import) table is first mapped to a corresponding field in a
destination (repository) table, and then each record is imported.
The problem is that for a successful data import using such tools, the
source data must have the same overall table and field structure as the
destination data, and the following additional conditions must all be met:
• Single import table. All the source data fields to be imported must
be from a single table.
• Values isolated in distinct fields. All the source data values
represented by the source data must be stored in distinct fields
rather than embedded within one or more text description fields. In
addition, each field must contain a single data value per record rather
than multiple delimited values that represent either a hierarchy or
multiple fields of information, or multiple values for the same field.
• Fields and field names match. There must be a direct one-to-one
relationship between individual source fields and individual
destination fields, and moreover, corresponding fields must have the
same names.
NOTE ►► The relationship between individual source and destination
fields identifies a field-oriented one-to-one relationship rather than the
record-oriented one-to-one relationship typical of a relational DBMS.
This is a new way of looking at data values field-by-field from a vertical
point of view rather than row-by row from a horizontal point of view.

• Values match. When mapping to destination fields whose values are

from a fixed domain of valid values (lookup fields and text attributes),
there must be a direct mapping between the values within the
individual source and destination fields that were previously mapped.
Not only that, the source data values themselves must precisely
match the corresponding destination data values (which must
already exist) to avoid polluting the repository with illegal and
inconsistent values from outside the valid value domain.
• Data types match. When mapping to destination fields without a
value domain, the data types of the source and destination fields
must be the same, and when mapping to numeric measurements, all
source data values must have the same physical unit of measure.
• Records match. Source records must be easily matched against
existing destination records, to avoid duplication of data. Based on
the value of a single key field, new records must all be created and
existing records must all be either all updated or all replaced.

MDM Import Manager Reference Guide 17

In the unlikely event that all of the above conditions were met – a data
import fantasy – the import could proceed automatically without user
intervention. The field mappings and the value mappings could both be
done automatically because of the obvious mapping relationship
between corresponding source and destination fields and between
matching source and destination values, no data values would need to
be converted, no exceptions would need to be handled, and the import
could proceed uninterrupted against existing records because new
records would all be created and existing records either all updated or
all replaced.

DATA IMPORT REALITY

Unfortunately, all of the requirements for a clean import are rarely if ever
met, and so the master data aggregator must perform a series of labor-
intensive pre-processing steps that can take hours or days, to
sufficiently manipulate, massage, and transform the source data into a
format that can be properly handled by the import tool. Moreover, during
record-at-a-time imports, the import tool must flag each discrepancy or
unexpected value for exception handling, which means that the import
can never proceed smoothly without interruption.
Specific, real-world import challenges include the following:
• Combine information from multiple tables
• Create a pivot table
• Create a reverse pivot table
• Map fields
• Map values
• Convert data types
• Create measurement values
• Convert data values
• Split a single field to expand hierarchy
• Split a single field into multiple fields
• Split a single field into multiple data values
• Combine two fields to build a hierarchy
• XXX PARSE Parse fields
• Combine fields to create value combinations
• Combine fields to reconstitute hierarchy
• Combine fields to merge values
• Combine fields to create a multi-valued field
• Import attribute data as name/value pairs
• Match records

18 MDM Import Manager Reference Guide

Each of these challenges – and the MDM Import Manager Innovations
designed to address them – are described in the following sections.

Challenge: Combine Information from Multiple Tables

Import fantasy: Source data values all from a single import table
Import reality: Values in multiple import tables
MDM feature: Table joins and lookups

The source data fields to be imported may not all belong to a single
source table. In this case, there must be a way to first join tables
together and to then look up fields from each lookup source table into
the primary source table, to create virtual extended source records
(containing both primary table fields and lookup fields) for import.
Alternatively, source tables must be “flattened” into a single table as a
pre-processing step in an external application.
INNOVATION ►► Source data can be combined from multiple tables
using the table join and lookup features of the MDM Import Manager,
eliminating the need to pre-flatten source data in an external
application.

Challenge: Create a Pivot Table

Import fantasy: Source data is structured as distinct fields in import table
Import reality: Data is structured as name/value pairs
MDM feature: Pivot table creation

When field and attribute data is stacked “vertically” in the source table
as name/value pairs and needs to be rearranged into a “horizontal”
table structure from which it can be imported into the fields and
attributes of a regular table, the import tool must provide a mechanism
to “rotate” the source data.
INNOVATION ►► The records and fields of a source table can be
completely restructured using the pivot mechanism of the MDM Import
Manager.

NOTE ►► Attribute data that is stacked vertically as name/value pairs

can also be imported directly without pivots by first partitioning the
name and value source fields, and then mapping the resulting
name/value hierarchy to a special destination field containing a
name/value hierarchy of attributes and attribute data values.

MDM Import Manager Reference Guide 19

Challenge: Create a Reverse Pivot Table
Import fantasy: Source data is stacked for import into qualified table
Import reality: Data is structured as rectangular table of records and fields
MDM feature: Reverse pivot table creation
When qualified data is arranged “horizontally” in the source table and
needs to be rearranged into a “vertical” table structure from which it can
be imported into the fields and qualifiers of a qualified table, the import
tool must provide a mechanism to “rotate” the source data.
INNOVATION ►► The records and fields of a source table can be
completely restructured using the reverse pivot mechanism of the
MDM Import Manager.

Challenge: Map Fields

Import fantasy: Source and destination field names match
Import reality: Field names do not match
MDM feature: Field-level mapping

Corresponding source and destination fields may not have the same
name. In this case, field mapping must be done manually before
proceeding with the import.
INNOVATION ►► Fields can be mapped using the MDM Import
Manager, to indicate not only the correspondence between source and
destination fields, but also which source fields should be imported.

Challenge: Map Values

Import fantasy: Source and destination field values match
Import reality: Values do not match
MDM feature: Value-level mapping

When mapping to destination fields whose values are from a fixed

domain of valid values (lookup fields and text attributes), corresponding
source and destination values may not precisely match, and some
source values may not already exist as destination values (e.g. source
values USA, UK, and FRA correspond to destination values United
States, England, and France, respectively, while source value MEX
does not yet have a corresponding destination value). In this case, a
record-at-a-time import tool must flag each discrepancy for exception
handling during the import process itself, and the user must manually
and tediously identify the corresponding destination value or indicate
that a new value should be created to correspond to the source value.
Alternatively, the source data must be meticulously cleansed and
scrubbed as a pre-processing step in an external application.

20 MDM Import Manager Reference Guide

 INNOVATION ►► Source fields that contain text values and are
mapped to lookup fields or text attributes can be mapped not only at
the field level, but also at the value level against the domain of legal
destination values, eliminating the need to pre-cleanse source data in
an external application.

INNOVATION ►► For each field, the MDM Import Manager collapses

the set of all values across the entire set of source records down to the
set of distinct values, whose number will usually be dramatically
smaller than the total number of records, often by several orders of
magnitude.

DATA INTEGRITY ►► Value-level mapping of source values forces a

user to normalize source data, implicitly defining a precise set of
transformation rules that: (1) eliminate differences between source and
destination values; (2) collapse multiple source value variations into a
single destination value; (3) properly map multiple delimited source
values into a multi-valued MDM field; and (4) explicitly add new values
to the set of destination values. The import cannot occur until each
distinct source data value is mapped and all discrepancies in the
source data resolved.

Challenge: Convert Data Types

Import fantasy: Source and destination field data types match
Import reality: Data types do not match
MDM feature: Automatic data type conversion

When mapping to destination fields without a value domain, the data

types of the corresponding source and destination fields may be
different, such as when numeric, currency, date, time, or Boolean
source data are each stored as text values within text fields. In this
case, the source data values must be converted to the proper data type
as part of the import process.
INNOVATION ►► The data types of source data values are
automatically converted by the MDM Import Manager based on the
data type of the destination field to which the source field is mapped,
even when stored in the source data as text values. If necessary,
individual values can be manually edited to override errors or
omissions in the source data.

DATA INTEGRITY ►► The import cannot occur until the data type of
each distinct source data value is properly converted.

Challenge: Create Measurement Values

MDM Import Manager Reference Guide 21

Import fantasy: Source values all have the appropriate unit of measure
Import reality: Units of measure missing and inconsistent
MDM feature: Automatic measurement value creation

When mapping to destination fields that represent physical

measurements (measurement fields and numeric attributes), the units of
measure in the source data may be: (1) missing entirely, such as when
all the values in a source field have the same implicit unit; (2)
inconsistent, such as when different text strings are used to represent
the same physical unit (e.g. inches, inch, in, or ‘"’); or (3) different for
each source data value. In each case, the unit of measure strings must
be normalized and appended to each numeric source data value to
create measurement values as part of the import process.
NOTE ►► The syntactical challenge of creating measurements is
further compounded because the source data values may be stored
either: (1) in a single numeric field; (2) in a single text field; or (3) in a
pair of fields (one containing the numeric value and the other
containing the corresponding unit of measure string); additionally, each
field may contain: (4) a single value; or (5) multiple delimited values.
Each of these distinct cases must be understood and properly handled
by the import tool.

INNOVATION ►► Unit of measure strings are automatically

normalized and appended to each numeric source data value based
on the physical dimension of the destination measurement field or
numeric attribute to which the source field is mapped. If necessary,
each numeric measurement value can be converted to a target unit, or
manually edited to override errors or omissions in the source data.

DATA INTEGRITY ►► MDM’s measurement data type guarantees

unit of measure consistency and permits each individual measurement
value to have its own distinct unit. Moreover, unit of measure strings
are automatically normalized against MDM’s built-in dictionary of over
2000 synonyms for over 750 units in over 70 physical dimensions. The
import cannot occur until each numeric source data value has an
associated unit of measure.

Challenge: Convert Data Values

Import fantasy: Source values properly formatted for import
Import reality: Values need to be adjusted, corrected, or reformatted
MDM feature: Value conversion grids and filters

With or without value mapping, data type conversion, or measurement

value creation, individual source data values may need to be manually
adjusted, corrected, or reformatted.

22 MDM Import Manager Reference Guide

 INNOVATION ►► Not only can each individual source value be
manually edited, built-in and custom-designed value conversion filters
can automatically convert or reformat a single value or the entire set of
distinct values with a single command. Built-in filters include: (1)
conversion from one unit of measure to another, such as from inches
to millimeters; (2) case change, such as from UPPER CASE to Mixed
Case; or (3) month and year range expansion for efficient import of
automotive data.

DATA INTEGRITY ►► Value conversion filters automate repetitive

and error-prone transformations, eliminating manual typing and the
possibility of user error. They also allow powerful reformatting
algorithms to be applied to an entire group of values.

Challenge: Split a Single Field to Expand Hierarchy

Import fantasy: Source data values isolated in distinct fields
Import reality: Delimited values represent a hierarchy
MDM feature: Split into hierarchy

A single field may contain a set of delimited values that implicitly

represent the full path of a leaf-node value in a hierarchy (e.g. “Tools >
Power Tools > Drills > Accessories” represents the Accessories
category for power drills in a four-level hierarchy). The delimited values
must be expanded into a hierarchy as part of the import.
INNOVATION ►► The values in a source field can be expanded into
a hierarchy based on the specified delimiter using the split-into-
hierarchy feature of the MDM Import Manager, so that you can
precisely map to existing destination hierarchy – and even
automatically create new destination hierarchy – as part of the value-
mapping process.

DATA INTEGRITY ►► When expanding into a hierarchy, the MDM

Import Manager automatically synthesizes the delimited values across
the entire set of source records into a single integrated hierarchy.

DATA INTEGRITY ►► You can specify how to handle trailing NULLs

in delimited values based on whether you are importing into a
hierarchy lookup field or into the hierarchy table itself. However,
regardless of the hierarchy NULL-handling setting, internal NULLs are
always preserved as nodes because ignoring them would collapse
hierarchy structure.

Challenge: Split a Single Field into Multiple Fields

Import fantasy: Source data values isolated in distinct fields

MDM Import Manager Reference Guide 23

Import reality: Delimited source field value represents multiple fields of data
MDM feature: Split into multiple fields

A single field may contain a set of delimited values that actually belong
in distinct fields (e.g. “Printers; Laser; 1200” represents a product with
Category=Printers; Printer Type=Laser; and DPI=1200 dpi). The
delimited values must be split into multiple fields as part of the import.
NOTE ►► Sometimes multiple delimited values do not belong in
distinct fields. Rather, they represent multiple values for a single field
(e.g. “110V; 115V; 220V; 230V”), in which case no splitting is
necessary and the delimited field should instead be mapped to an
MDM multi-valued field or attribute.

INNOVATION ►► The values in a source field can be split into

multiple fields based on the specified delimiter using the split-into-
multiple-fields feature of the MDM Import Manager, so that you can
map, convert, and create measurements from the individual values in
each split field.

Challenge: Split a Single Field into Multiple Data Values

Import fantasy: Source data values isolated in distinct fields
Import reality: Delimited source field value represents multiple data values
MDM feature: Multi-valued delimiter

Sometimes multiple delimited values do not belong in distinct fields.

Rather, they represent multiple values for a single field (e.g. “110V;
115V; 220V; 230V” represents a product that is available for all four
voltages). In this case, no splitting is necessary and the delimited field
value must be split into multiple values as part of the import, and the
delimited field mapped to an MDM multi-valued field or attribute.
INNOVATION ►► The value in a source field can be split into multiple
values based on the specified multi-valued delimiter, so that you can
map, convert, and create measurements from each of the individual
values.

24 MDM Import Manager Reference Guide

Challenge: Combine Two Fields to Build a Hierarchy
Import fantasy: Direct mapping between individual source and destination fields
Import reality: Two fields represent a single hierarchy of values
MDM feature: Create hierarchy field

Hierarchy in the source data may have been flattened into a pair of
fields in the source table, with one field for the node and one for its
parent node (e.g. “Printers”; “Printers; Laser”; and “Printers; Dot Matrix”
represent the Printers node and its two children). The parent/child value
pairs must be assembled to build this hierarchy as part of the import.
INNOVATION ►► MDM allows you to build a hierarchy from two
fields whose values identify a set of parent/child hierarchy
relationships.

Challenge: Combine Fields to Create Value Combinations

Import fantasy: Direct mapping between individual source and destination fields
Import reality: Mapping between field combinations rather than individual fields
MDM feature: Partitions

It often may not be possible to directly map individual source fields to

individual destination fields – or to map, convert, or create
measurements from the underlying values within those individual source
and destination fields – because there may not be a direct one-to-one
relationship between individual source and destination fields.
Instead, the appropriate field mapping relationship may exist between
the combination of two or more source fields and/or the combination of
two or more destination fields and/or attributes (such as when the
CATEGORY and SUBCATEGORY source fields together correspond to
the Category destination field, or alternatively, the CATEGORY source
field corresponds to the Category destination field together with the
Printer Type and DPI attributes). In these cases, fields must be
combined to allow the mapping of field combinations rather than
individual fields; meanwhile, combining fields into field combinations
simultaneously combines the underlying values into value combinations,
to allow the value-level mapping, conversion, and measurement
creation of value combinations rather than individual values.
NOTE ►► Field and value combinations represent one-to-many,
many-to-one, and many-to-many relationships between the fields
rather than the rows of two tables.

MDM Import Manager Reference Guide 25

 INNOVATION ►► MDM’s generalized partitioning mechanism allows
you to combine two or more fields into field combinations (and
underlying values into a corresponding set of value combinations), so
that you can map, convert, and create measurements from value
combinations rather than individual values. Multiple partitions can be
nested (creating hierarchy among value combinations) or combined
(merging value combinations).

INNOVATION ►► For a field combination created by partitioning, the

set of all possible value combinations is collapsed down to the set of
those that actually exist in the underlying records.

DATA INTEGRITY ►► Although the set of distinct value combinations

will be larger than each individual set of distinct values, each particular
value combination will then correspond to a smaller, more specific and
more precisely defined subset of underlying records, and thereby
permit an accurate value-level mapping that would not otherwise be
possible.

Challenge: Combine Fields to Reconstitute Hierarchy

Import fantasy: Direct mapping between individual source and destination fields
Import reality: Multiple fields represent a single hierarchy of values
MDM feature: Nested partitions

Hierarchy in the source data may have been flattened into multiple fields
in the source table, such as when a category hierarchy is stored as
three distinct source fields, with the first two representing internal
category nodes and the third representing the leaf node category.
Again, fields must be combined to reconstitute this hierarchy as part of
the import.
INNOVATION ►► Nested partitions allow you to create field
combinations from multiple fields in the source data to create or
reconstitute hierarchy represented by the underlying value
combinations, so that you can precisely map to existing destination
hierarchy – and even automatically create new destination hierarchy –
as part of the value-mapping process.

DATA INTEGRITY ►► You can specify how to handle trailing NULLs

in delimited values based on whether you are importing into a
hierarchy lookup field or into the hierarchy table itself. However,
regardless of the hierarchy NULL-handling setting, internal NULLs are
always preserved as nodes because ignoring them would collapse
hierarchy structure.

26 MDM Import Manager Reference Guide

Challenge: Combine Fields to Merge Values
Import fantasy: Direct mapping between individual source and destination fields
Import reality: Multiple fields represent a single merged value
MDM feature: Combined partitions

Values from two or more fields in each source record may need to be
merged into a single value for import, such as when a first name and
last name are stored in two distinct source fields and must be combined
into a single name field within the repository, or a numeric value and an
associated unit of measure are stored in two distinct source fields but
are properly stored in a single measurement field in the repository.
Again, fields must be combined to merge values as part of the import.
INNOVATION ►► Combined partitions allow you to create field
combinations from multiple fields in the source data to concatenate the
underlying value combinations into a single merged value for each
record, so that you can map, convert, and create measurements from
the merged values.

DATA INTEGRITY ►► You can preserve or collapse empty positions

in the single merged value (which correspond to NULLs in the source
data) based on whether position matters in the merged value.

Challenge: Combine Fields to Create a Multi-Valued Field

Import fantasy: Direct mapping between individual source and destination fields
Import reality: Multiple fields represent the multiple values of a multi-valued field
MDM feature: Combined partitions

Instead of being merged, values from two or more fields in each source
record may need to be concatenated into a set of delimited values that
belong in a single multi-valued field, such as when two size values are
stored in two distinct source fields and must be stored in a multi-valued
measurement field within the repository. Again, fields must be combined
to concatenate values as part of the import.
INNOVATION ►► Combined partitions allow you to create field
combinations from multiple fields in the source data to concatenate the
underlying value combinations into a set of delimited values for each
record, so that you can map the field combination and the underlying
set of delimited values to a multi-valued field.

DATA INTEGRITY ►► You can preserve or collapse empty positions

in the set of delimited values (which correspond to NULLs in the
source data) based on whether position matters.

MDM Import Manager Reference Guide 27

Challenge: Import Attribute Data as Name/Value Pairs
Import fantasy: Source attribute data is structured as distinct fields in import table
Import reality: Attribute data is structured as name/value pairs
MDM feature: Partitions; special destination field containing attribute name/values
When attribute data is stacked “vertically” in the source table as
name/value pairs and the number of attributes represented by the
source data is too large to use the pivot and map mechanism, the
attribute data needs to be imported into the attributes directly.
INNOVATION ►► Attribute data that is stacked vertically as
name/value pairs can be imported directly by first partitioning the name
and value source fields, and then mapping the resulting name/value
hierarchy to a special destination field containing a name/value
hierarchy of attributes and attribute data values.

Challenge: Match Records

Import fantasy: Source records easily matched against destination records
Import reality: Duplicate matching records and matching conflicts
MDM feature: Matching against multiple key fields and field combinations

Matching source records against existing destination records is

extremely difficult. Source records may correspond to a single existing
destination record or to multiple existing destination records. The match
may be based on the value of a unique key field, multiple key fields
(only some of which may match), or key field combinations. Finally, for
each source record, the user must be able to indicate that the source
record should be: (1) skipped; (2) created; (3) used to update existing
destination records; or (4) used to replace existing destination records.
INNOVATION ►► Record matching can be done simultaneously
against multiple key fields and/or key field combinations to locate
existing records that correspond to the source records. A default import
action (skip, add, update, replace) can then be applied to each
matching group and then overridden for each source record.

DATA INTEGRITY ►► Using multiple key fields for record matching

does an extremely thorough job of locating existing repository records
that match each source record, which highlights: (1) potential data
duplication among existing records, when there are multiple matching
records; and (2) potential errors in the incoming import data, when
there are conflicts between the various key fields.

28 MDM Import Manager Reference Guide

KEY ADVANTAGES
Advantages of using the MDM Import Manager include the following:
• Infinite flexibility. The MDM Import Manager can handle any tabular
electronic source data, enabling you to arbitrarily restructure and
rationalize it against destination data that has an entirely different
structure. Moreover, the underlying engine can not only import but
also transform data for destinations outside the repository.
• No programming. Despite its flexibility, the MDM Import Manager
requires no programming whatsoever, with a WYSIWYG interface
that allows you to implicitly define a set of intricate transformation
rules that can be saved and then reused on a recurring basis, either
through the GUI or programmatically through an API.
• End-to-end application. The MDM Import Manager includes
comprehensive functionality to interactively browse, sort, restructure,
and import source data that completely eliminates the need to use an
external application for any step in the import process.
• No pre- or post-processing. Not only is the pre-processing typically
required by other import tools completely eliminated, no post-
processing is required after the source data is within the repository,
since it is completely transformed as part of the import process.
• Consistent view of source data. Regardless of the application file
format of the source data (Access, Excel, delimited text, SQL, XML),
the MDM Import Manager offers precisely the same user interface,
functionality, and overall user experience for performing the import.
• Full relational support. The MDM Import Manager can combine
related information from multiple tables, so that relational source data
(even multiple sheets within an Excel file) do not need to be pre-
flattened into a single table in an external application prior to import.
• Single interface. An “at-a-glance” interface simultaneously provides
both a high-level “birds-eye” view of the schema structure of the
source data (including tables, fields, and table relationships) and a
low-level detailed view of the data itself within each of the tables.
• Field-oriented processing. User processing within the MDM Import
Manager is performed field-by-field prior to import rather than record-
by-record during import, reducing the magnitude of the processing
effort from the number of records to the number of fields.
• Distinct value lists. For each field, the MDM Import Manager
collapses the set of all values across the entire set of source records
down to the set of distinct values, whose number will usually be
dramatically smaller than the total number of records, often by
several orders of magnitude.

MDM Import Manager Reference Guide 29

• Value conversion and mapping. Data types are automatically
converted and unit of measure strings automatically normalized and
appended to numeric source data values based on the data type and
physical dimension of the destination field, even when stored in the
source data as text. In addition, you can apply value conversion
filters both at the field level and at the individual value level, and map
source data values to the domain of legal destination values.
• 100% accuracy. Unlike other import applications that claim up to
96% accuracy for importing source data, the approach used by the
MDM Import Manager supports and even guarantees 100% accuracy
of the imported data, because the precise transformations must all
be defined before you perform the import.
DATA INTEGRITY ►► Each source data field is handled in
sequence, with a beginning, a middle, and an end to the process of
validating, correcting, and/or converting source data values. All
exceptions are clearly visible in the set of distinct values for each field,
and each can be handled accordingly. With or without value mapping,
data type conversion, or measurement value creation, each distinct
source value can be manually adjusted or corrected to completely
eliminate source data inconsistencies.

30 MDM Import Manager Reference Guide

About This Guide
This reference guide focuses on providing the most useful information in
the fewest pages. However, making the most of the MDM Import
Manager requires an understanding of the concepts that make it so
powerful; therefore we provide ample conceptual background as well.
Our ultimate goal is to help you learn to use the MDM Import Manager
as it was designed to be used, for maximum ease of use and power in
aggregating your master data.

WHAT YOU WILL FIND IN THIS GUIDE

• Concepts required to make best use of the MDM Import Manager
• Step-by-step procedures for preparing source data for import
• Step-by-step procedures for importing data and tracking status
• Operational tips and techniques

WHAT YOU WILL NOT FIND IN THIS GUIDE

• How to create an MDM repository, and other DBMS administration
topics (described in the MDM Console Reference Guide).
• A description of the distinctive features of the MDM user interface
(described in the MDM Data Manager Reference Guide)
• How to manage and edit records in the MDM repository (also
described in the MDM Data Manager Reference Guide).
• MDIS documentation (described in the MDM Console Reference
Guide).

ASSUMPTIONS
We have made certain assumptions in this reference guide, to better
focus on the information you will find most useful.
• The data to be imported resides in Access, Excel, SQL Server, or
Oracle databases that have been set up as ODBC data sources.
• You are familiar with the MDM Client software, or at least the MDM
Data Manager Reference Guide. In particular, you are familiar with
the different types of controls in the MDM user interface.
NOTE ►► Since each MDM repository is unique, we cannot choose
examples of table names, field names, attributes, and so on, that
would apply to all users. For the sake of example, we call the main
table that contains product records “Products” and the taxonomy table
that contains product categories “Categories.”

MDM Import Manager Reference Guide 31

SCREEN CAPTURES
All MDM 7.1 applications now follow the standard SAP color scheme, as
illustrated in Figure 1. However, screen captures included in this
reference guide may continue to show the old color scheme if the
functionality illustrated by the screen capture has not changed. In cases
where functionality has changed or new features have been added, the
screen captures have been updated to the new color scheme.

Figure 1. MDM 7.1 color scheme

32 MDM Import Manager Reference Guide

PART 1: BASIC CONCEPTS

This part of the reference guide explains essential concepts that will
enable you to make the best use of the MDM Import Manager. It
explains in conceptual terms the processes for which the remainder of
this reference guide provide the step-by-step procedures.

33
What is Parametric Import?
The MDM Import Manager is a GUI “cockpit” that defines a very specific
sequence of steps to prepare source data for import. But rather than
perform the import a record at a time as with other import tools, the
MDM Import Manager offers a comprehensive set of data
transformation capabilities that allow you to pre-process the source data
a field at a time prior to import.
Moreover, for each field, the MDM Import Manager collapses the set of
all values across the entire set of source records down to the set of
distinct values, whose number will usually be dramatically smaller than
the total number of records, often by several orders of magnitude.
Field-at-a-time rather than record-at-a-time processing and the distinct
value lists of the MDM Import Manager are together the lynchpin of its
™
patent-pending Parametric Import technology, which supports the
import of electronic source data from any tabular or XML format.
Parametric Import is a new and radically more efficient approach to
importing and transforming data that is conceptually similar to
parametric search, as follows:
• Parametric search lists the complete set of distinct values for each
field or attribute in a repository. Selecting a value during a drill-down
search selects all of the underlying records that contain that value.
• Parametric import lists the complete set of distinct values for each
field in the source data. Changing a value as part of the import (by
mapping or converting it) changes it in all of the underlying records
that contain that value.
INNOVATION ►► In conjunction with field-at-a-time rather than
record-at-a-time processing, distinct value lists make import exception
handling much more efficient because:
(1) the number of distinct values for each field is dramatically smaller
than the total number of values;
(2) each distinct value represents and is a proxy for not only all
instances of the value across the entire set of underlying records,
but also the subset of records that contain the value;
(3) examining and then transforming each source data value must be
done only once for each distinct value rather than once for each
instance of the value; and as a result,
(4) the maximum number of exceptions is the sum rather than the
product of the number of distinct values in each mapped field.

INNOVATION ►► Distinct value lists also support partitioning,

another patent-pending innovation of the MDM Import Manager that
allows you combine fields to create field and value combinations, for
additional transformation flexibility.

MDM Import Manager Reference Guide 35

For example, suppose there are 5000 source records, each with a value
for a Country field that represents the country of origin, and moreover,
there are twelve distinct values for Country across the 5000 records. If
all of the source values for Country were in upper case and must be
converted to mixed case for import, using the set of distinct values
would require you to convert each of twelve values once rather than
individually convert its value in each of 5000 source records. Similarly, if
all of the source values for Country were three-letter abbreviations that
must be mapped to the standard country names within the repository,
using the set of distinct values would require you to map each of twelve
values once rather than individually replace its value in each of 5000
source records. The benefit for the Country field is clear: twelve
transformations rather than 5000.
INNOVATION ►► Parametric import replaces the traditional concept
of exception handling with a very specific and predictable sequence of
steps to prepare source data for a clean and exception-free import.

THE CURRENT TABLES

The MDM Import Manager imports records from the current source table
into the current destination table. The current source and destination
tables are displayed in the pair of drop-down list controls at the far left of
the toolbar (Figure 2).

Figure 2. Current source and destination tables shown at left of toolbar

NOTE ►► The MDM Import Manager imports records from a single

table into a single table.

When you open the drop-down list of source tables (Figure 3), the table
list includes the None choice separated by a gray line from all the tables
in the source file, in alphabetical order.

Figure 3. Drop-down table list open for source table selection

NOTE ►► For XML source files, the source table list displays the
nested elements defined in the XML schema.

36 MDM Import Manager Reference Guide

When you open the drop-down list of destination tables (Figure 4), the
table list includes None, the main table of the MDM repository, the
normal subtables in alphabetical order, the object tables, and the
parent/child relationships, with each set separated by a gray line.

Figure 4. Drop-down table list open for destination table selection

NOTE ►► When you first start the MDM Import Manager, it sets both
the current source table and the current destination table to None. You
must select a current source table and a current destination table
before you can perform most of the MDM Import Manager operations.

Non-Main Table Imports

The MDM Import Manager can be used to import not only to the main
table but also to the subtables of an MDM repository. However, there
are several differences when importing to a subtable with respect to
which fields require or permit optional value mapping and which fields
must be used for record matching.
NOTE ►► See “When Value Mapping Is Required” on page 218,
“Value Mapping and Lookup Table Display Fields” on page 246, and
“Record Matching and Lookup Table Display Fields” on page 345.

NOTE ►► You cannot import to the Masks or Named Searches tables.

Relationship Imports
You can also use the MDM Import Manager to import into parent/child
product relationships; as above, there are differences with respect to
field mapping and record matching when importing the links that define
the related records of these relationships (see “Importing Parent/Child
Relationship Links” on page 385).

MDM Import Manager Reference Guide 37

TABLE JOINS AND LOOKUPS
The MDM Import Manager eliminates the need to flatten relational
source data in an external application prior to import. It automatically
detects join relationships between tables defined in the data source, and
also lets you manually create additional relationships. You can then look
up fields from lookup tables into the primary source table to create a
virtual “lookup-extended” source table for import whose virtual records
consist of both the fields of the primary source table and the fields
looked up from each of the lookup tables.
Table joins and lookups are described in more detail in “Table Joins and
Lookups” on page 101.

PIVOTING AND REVERSE PIVOTING

The pivoting and reverse pivoting mechanisms of the MDM Import
Manager allow you to “rotate” source data, creating an entirely new
table from the source table, to completely transform the data in ways
that the other restructuring mechanisms cannot address.
In particular, pivoting is often required when field and attribute data is
stacked “vertically” in the source table and needs to be rearranged into
a “horizontal” table structure from which it can be imported into the fields
and attributes of a regular table. Conversely, reverse pivoting is often
required when qualified data is arranged “horizontally” in the source
table and needs to be rearranged into a “vertical” table structure from
which it can be imported into the fields and qualifiers of a qualified table.
Pivoting is described in more detail in “Pivoting” on page 110; reverse
pivoting is described in more detail in “Reverse Pivoting on page 117.

PARTITIONING
The MDM Import Manager extends the MDM Client partitioning
mechanism to combine fields into field combinations and thereby create
value combinations, allowing you to deal with the value combinations as
a proxy for the corresponding subsets of underlying records.
Partitioning can be used to achieve several very different objectives: (1)
establish a one-to-one relationship between source and destination
fields and field combinations that might not otherwise exist between
individual source and destination fields; (2) create or reconstitute
hierarchy represented by the source data; and (3) concatenate values
stored in two or more distinct source fields into a single merged value
for each record.
Partitioning is described in more detail in “Partitioning and Field
Mapping” on page 173.

38 MDM Import Manager Reference Guide

VALUE MAPPING
The value mapping mechanism provided by the MDM Import Manager
offers an entirely new approach to handling exceptions in source data,
which can be pre-processed a field-at-a-time prior to import rather than
processed a record-at-a-time during import. All exception handling can
be handled up-front and the actual import can then proceed without
interruption regardless of the size of the record set.
Value mapping also provides a powerful way to map from source to
destination hierarchy, and to automatically create destination hierarchy
(internal and leaf nodes) as part of the import.
Value mapping is described in more detail in “Value Mapping” on page
212.

VALUE CONVERSION
Even source data values that do not need to be mapped may need to
be converted, either because the values are inconsistent with repository
values, because their data types do not correspond, or because
numeric values are missing a physical unit of measure. Such data
conversion is handled automatically by the MDM Import Manager, and
can be overridden on a value-by-value basis.
Value conversion is described in more detail in “Value Conversion” on
page 221.

RECORD MATCHING
The preceding operations transform records for import, which can then
be matched against existing records to determine whether a new record
should be created, or an existing record should be updated or replaced.
Record matching can be done against a single key field, multiple key
fields, key field combinations, and even against non-key fields to detect
differences between source and destination field values.
Record matching is described in more detail in “Record Matching
Concepts” on page 338.

IMPORT MAPS
All of the partitions, mappings, and other conversions can be saved in
an import map within the MDM repository. A map is useful if you want to
save your work before you have performed the import, if you want to
subsequently apply the same transformations for future imports of
additional records or updates from the same data source, or if you want
to later use the map in conjunction with MDIS processing.
Import maps are described in more detail in “Import Map” on page 69.

MDM Import Manager Reference Guide 39

RECORD FILTERING
Record filtering expands the concept of import by complementing the
source-centric set of import records with a destination-centric set of
records that includes not only matched records but also unmatched
records in the destination set.
Record filtering is described in more detail in “Record Filtering” on page
356.

40 MDM Import Manager Reference Guide

PART 2: GUIDED TOUR

This part of the reference guide explains how to start and stop the MDM
Import Manager. It assumes that you are familiar with the MDM Client
software, or at least have familiarized yourself with the different types of
controls in the MDM user interface using the MDM Data Manager
Reference Guide.

41
Starting Import Manager
When you start MDM Import Manager, a connection wizard connects
Import Manager to the MDM repository you want to import to, and then
connects Import Manager to the source data you want to import.
A repository must be running in order for Import Manager to connect to
it.
In order to connect to the source data, the source file or database must
be available through the Windows file system, an MDM repository port,
or the DBMS.

CONNECTING IMPORT MANAGER TO AN MDM REPOSITORY

The first step in starting Import Manager is connecting to a repository.

 To connect Import Manager to a repository:

1. Double-click the MDM Import Manager icon.
2. In the Connect to MDM Repository screen, select the desired MDM
repository, the language layer to work in, and enter your user name
and password.
3. Click Next to connect to the source data you want to import (see
“Connecting Import Manager to Source Data” on page 45).

NOTE ►► If an MDM repository does not appear in the list, you must
add it by clicking “…” (browse) to set up the connection (see the
following sections for more information).

NOTE ►► If a repository’s TCP/IP port number changes, you must re-

add the repository as the old entry will load whatever repository is
loaded on the old TCP/IP port.

MULTILINGUAL ►► More information about language layers and

multilingual repositories is provided in “Part 14: Multilingual Support.”

Setting Up Unencrypted Repository Connections

 To set up an unencrypted connection to an MDM repository:
1. In the Connect to MDM Repository screen, click “…” (browse).
2. In the Choose Repository dialog, select the Master Data Server on
which the repository is running.
If the Master Data Server has not been previously connected to by
Import Manager, type the name or IP address of its host, or click “…”
(browse) to select it from a list.

MDM Import Manager Reference Guide 43

 NOTE ►► If the Master Data Server is configured to listen on non-
"MDM default" ports, you must type in the port number after the Master
Data Server name, using the format ServerName:PortNumber (for
example, ServerXYZ:54321). Otherwise, Import Manager will be
unable to connect to the Master Data Server.

3. Select the repository (you may have to click the “refresh” button next to
the Repository field to load the repository names).
4. Click OK to add the repository to the Connect to MDM Repository list.

Setting Up Secure Repository Connections

Secure connections are only possible for repositories on SSL-enabled
Master Data Servers (see “Network and Communication Security” in the
MDM 7.1 Security Guide for more information).

 To set up a secure connection to an MDM repository:

1. In the Connect to MDM Repository screen, click “…” (browse).
2. In the Choose Repository dialog:
a. Select the SSL-enabled Master Data Server
b. Click Secure Connection
c. Enter the paths to the client Key File and SSL Library
3. Click the “refresh” button next to the Repository field and choose the
repository you want to connect securely to.
4. Click OK to add the repository to the Connect to MDM Repository list.
5. In the Connect to Repository dialog, a lock icon similar to the one
shown below indicates that communications with the selected
repository will be established on a secure connection.

Figure 5. The lock icon on a Connect to MDM Repository dialog

44 MDM Import Manager Reference Guide

CONNECTING IMPORT MANAGER TO SOURCE DATA
Once Import Manager connects to a repository, the next step is to
connect it to the source data you want to import.

 To connect Import Manager to source data:

1. In the Connect to Source screen, choose the source Type and the
Remote System the data originated from.
2. Select the source properties that correspond to the selected Type (see
below for more information about each property).
3. Click OK to open Import Manager.

Source Properties
Each source Type requires a different set of source properties, as
indicated by the bullets in Table 1.

Table 1. Source Properties

Delimited Text

XML Schema
SQL Server
Fixed Text
Access

Oracle
Type

ODBC
Excel

XML
Port

Property Description

Remote System ● ● ● ● ● ● ● ● ● Name of the remote system

Port ● Name of the port

XML Schema ● Name of the XML schema

File Name ● ● ● ● ● ● Name of the source data file

DBMS Server ● ● Name of the DBMS server

Database ● ● Name of the source database

User ● ● Username of the connection

Password ● ● Password of the connection

Columns ● Columns of the fixed text

Delimiter ● Delimiter of the delimited text

Read-Only ● ● Open source as read-only?

Root ● Root element of XML schema

NOTE ►► A maximum field name length of 64 characters is imposed

on Access and Excel import files; all other types have no length limits.

NOTE ►► For Type=Delimited Text, use \t to specify a tab delimiter.

NOTE ►► For Type=Fixed Text, the Columns values are entered as a

semi-colon delimited list (e.g. 4;7;9:) with a colon used to separate
column header widths and column data widths (e.g. 25;50;10:4;7;9)

MDM Import Manager Reference Guide 45

 NOTE ►► For Type=Oracle, a 32-bit Oracle client is required (64-bit
clients are not supported).

NOTE ►► For Type=Port, selecting a Remote System limits the set of

ports but is not strictly required, since port names should be unique.

PORTS ►► The Port list includes an entry for each inbound port
defined in the MDM Console for the selected Remote System; those
that do not have files waiting or are locked by another user are
disabled in the drop-down list. When you select a port, the MDM Import
Manager automatically selects the first source data file that is waiting
to be processed; you can then use the Get Next Source File command
to select each succeeding file.

XML SCHEMAS ►► The XML Schema list includes an entry for each
XML schema defined in the MDM Console.

TIP ►► For Type=XML and Type=XML Schema, you can select

multiple XML source files, each having the same structure.

TIP ►► To reference a SQL Server named instance in the DBMS

Server property, append a backslash (\) and the instance name after
the machine name.

GETTING THE NEXT SOURCE DATA FILE

For source Type=Port, MDM Import Manager automatically selects the
first source data file waiting to be processed. You can use the Get Next
Source File command to get the next source data file waiting at the
current port without closing your MDM Import Manager session.

 To get the next source file from the selected port:

1. From within the MDM Import Manager, choose File > Get Next Source
File from the main menu.
2. The MDM Import Manager selects the next source data file waiting at
the current port.
3. The existing MDM Import Manager session starts fresh with new
source data as if you exited and restarted the MDM Import Manager.

CONNECTING TO A NEW DATA SOURCE

You can use the Connect to Source command to connect to an entirely
new data source without closing your MDM Import Manager session, as
described in this section.

46 MDM Import Manager Reference Guide

 To connect to a new data source without closing your existing MDM
Import Manager session:
1. From within the MDM Import Manager, choose File > Connect to
Source from the main menu.
2. The MDM Import Manager opens the Connect to Source dialog.
3. Choose a new Type and Remote System and other source properties.
4. The existing MDM Import Manager session starts fresh with new
source data as if you exited and restarted the MDM Import Manager.

DATA STAGING
Prior to MDM 7.1, Import Manager used MS Access to stage data from
MS Access, MS Excel, text, and XML source files. As part of the staging
process, Import Manager converted source data from its original
structure into MS Access tables. These MS Access tables, and not the
original source files, were then displayed in the Import Manager.
For XML and XML Schema source files, the conversion flattened the
nested hierarchical data structures contained in the original XML file into
a set of Access tables. These tables required joins in order to
reconstruct the data relationships in the original file, and the end result
was that the “source” tables displayed in Import Manager bore little
resemblance to the nested structure of the original XML source file.
In MDM 7.1, Import Manager no longer uses MS Access to stage data
from text and XML source files. As a result, the previous limitations on
field name lengths (64 characters) is removed for text and XML files. In
addition, Import Manager now retains and displays the nested structure
of source XML files, which should result in a more intuitive Import
Manager experience (see “XML Source Files and the Source Hierarchy
Tree” on page 84 for more information).
NOTE ►► Import Manager still uses MS Access to stage data from
MS Access and MS Excel source files.

MDM Import Manager Reference Guide 47

EXCEL-RELATED IMPORT CHALLENGES
Because an Excel spreadsheet does not have an associated schema
that defines the data type of each column, the Jet driver used to connect
to Excel has several limitations that must be addressed by the MDM
Import Manager:
• Empty columns. The Jet driver may incorrectly conclude that a
spreadsheet column is empty.
• Data types. If the column is not empty, Jet may incorrectly determine
the data type of the values in the column.
• Mixed data types. If the column contains both numeric and text
values, Jet will return NULL for values of the minority data type.

The mixed data type problem is overcome internally by the MDM Import
Manager in the way it connects to the Jet driver, and does not require
any changes to your system.
However, MDM Import Manager can overcome the other problems only
by changing the value of a Windows registry setting, which it offers to do
for you automatically the first time you connect to an Excel data source.

NOTE ►► If you do not allow the MDM Import Manager to change the
registry setting, some Excel source data values may be ignored or
truncated, or their data type improperly converted.

48 MDM Import Manager Reference Guide

MDM Import Manager Main Window
The MDM Import Manager main window includes three primary panes,
with several tabs in the bottom pane. Panes may be resized by dragging
the splitters that separate them (Figure 6).

Figure 6. MDM Import Manager main window (default layout)

TIP ►► To cycle among the panes in the main window, press F6.

TIP ►► To resize panes, place the mouse pointer over a splitter until
the pointer changes to a vertical or horizontal pipe with arrows, then
click and hold down the mouse button while you drag the splitter in the
direction of the arrows.

SWITCHING LAYOUT
You can rearrange the panes in the main window by using the Switch
Layout command..

 To switch the layout of the main window:

♦ Choose View > Switch Layout from the main menu.

MDM Import Manager Reference Guide 49

SELECTING THE CURRENT TABLES
When you first start the MDM Import Manager, it sets both the current
source table and the current destination table to None. You must select
a current source table and a current destination table before you can
perform most of the MDM Import Manager operations.

 To specify the current source table:

♦ Click on the drop-down table list (Figure 7), and select the source table
whose records you want to import.
♦ Alternatively, choose View > Source Table from the main menu and
choose from the cascading menu of tables.

Figure 7. Drop-down table list open for source table selection

 To specify the current destination table:

♦ Click on the drop-down table list (Figure 8), and select the destination
MDM repository table into which you want to import.
♦ Alternatively, choose View > Destination Table from the main menu
and choose from the cascading menu of tables.

Figure 8. Drop-down table list open for destination table selection

50 MDM Import Manager Reference Guide

PART 3: IMPORT WORKFLOW

This part of the reference guide explains how the import workflow and
the panes and tabs of the MDM Import Manager main window.

51
Working with the MDM Import Manager
The MDM Import Manager allows you to import data into an MDM
repository from any tabular or XML format, with the same user interface
and identical functional capabilities regardless of the source file format.
Importing source data is akin to stir frying – almost all of the work is in
the preparation. In the case of import, preparation of a source data file
includes some or all of the following steps:
1. Connect to the source data file or database and to the destination
MDM repository, to establish programmatic access from the Import
Manager to the source and destination schemas and data.
2. Browse source tables, fields, and records, examine the set of
distinct data values for each field, to become familiar with the
source data and its structure.
3. Join tables and then look up fields from secondary tables into the
primary source table, to combine information from multiple tables
and create virtual extended source records for import.
5. Split fields, to break delimited field values into a hierarchy or into
multiple fields, for subsequent field-level processing on individual
field values in each new field.
6. Partition fields, to restructure source data by combining fields into
field combinations, for subsequent field-level processing on value
combinations rather than individual field values.
7. Map fields, the most traditional step in the import process, to
identify the destination field for each source field, and also the type
of mapping and conversion required for the underlying data values.
8. Map and convert values, to transform and normalize source data by
eliminating differences between source and destination values,
converting data types, and creating measurement values.
9. Match records, to specify key fields for identifying existing records
that match the source records, and whether new records should be
created and existing records should be updated or replaced.

Each step is a part of the process of implicitly defining a precise set of

transformation rules prior to import that allows you to completely
restructure, cleanse, normalize, and rationalize raw product data into
rich master data as part of the import process itself.

MDM Import Manager Reference Guide 53

MDM IMPORT MANAGER AT A GLANCE

1 2

3 4 5 6 7

8
Figure 9. MDM Import Manager main window

The main window of the MDM Import Manager consists of the panes
and tabs shown in the numbered callouts of Figure 9, listed below, and
described in the following sections:
1. Source Hierarchy pane
2. Destination Hierarchy pane
3. Source Preview tab
4. Partition Field/Value tab
5. Map Fields/Values tab
6. Match Records tab
7. Import Status tab
8. Status bar

The tabs in the bottom pane of the main window are arranged in
“workflow” order, from left to right. Various functions are performed in
the Source Hierarchy pane, but most of your work in preparing source
data for import is done in these tabs. Once you have completed
partitioning, field mapping, value conversion and mapping, and record
matching in these tabs, all that’s left to do to import the data is…press
the Execute Import toolbar button.

54 MDM Import Manager Reference Guide

Source Hierarchy Pane
The Source Hierarchy pane (top-left pane) contains a tree representing
the hierarchy of tables, fields, and data values of the source file whose
records you intend to import (Figure 10). Use the tree in the Source
Hierarchy pane to view the source table and field structure, to view
distinct data values for each field, to define and view join and lookup
relationships between tables, and to perform various source operations.

Figure 10. Source Hierarchy pane

Destination Hierarchy Pane

The Destination Hierarchy pane (top-right pane) contains a tree
representing the hierarchy of tables, fields, attributes, qualifiers, and
data values of the destination MDM repository into which you intend to
import the source records (Figure 11). Use the tree in the Destination
Hierarchy pane to view the destination repository table and field
structure and to view distinct data values for each field, attribute, and
qualifier.

Figure 11. Destination Hierarchy pane

MDM Import Manager Reference Guide 55

Source Preview Tab
The Source Preview tab (tab in bottom pane) contains two subpanes
(Figure 12). The left Tables subpane lists the source tables; the right
Records subpane contains a list view of the selected source table’s
records in a row/column grid, with a row for each record and a column
for each of the fields. Use the Source Preview tab to interactively
browse the records in each of the source tables and sort by any of the
columns in ascending or descending order.

Figure 12. Source Preview tab

56 MDM Import Manager Reference Guide

Partition Field/Value Tab
The Partition Field/Value tab (tab in bottom pane) lists, on the left, the
source fields that are available to partition the selected source field or
value in the Source Hierarchy tree, and on the right, the destination
fields and attributes that are available to partition the selected
destination field or attribute in the destination hierarchy (Figure 14). Use
the Partition Field/Value tab to create the partitions that combine fields
into field combinations, for subsequent processing of value
combinations rather than individual field values.

Figure 14. Partition Field/Value tab

NOTE ►► You cannot partition a source field or value until you have
selected the current source table. You cannot partition a destination
field or value until you have selected the current destination table.

MDM Import Manager Reference Guide 57

Map Fields/Values Tab
The Map Fields/Values tab (tab in bottom pane) contains two pairs of
grids (Figure 15). The top pair of Field Mapping grids contain the source
fields on the left and the destination fields, attributes, and qualifiers on
the right; the bottom pair of Value Conversion and Mapping grids
contain the set of distinct data values corresponding to the current
source field on the left and corresponding to the current destination field
on the right. Use the Field Mapping grids to map source fields to
destination fields, attributes, and qualifiers. Then, for each pair of
mapped source and destination fields, use the Value Conversion and
Mapping grids to: (1) map values (when mapping to lookup fields and
text attributes); (2) create measurement values (when mapping to
measurement fields or numeric attributes); and (3) manually adjust,
correct, and reformat individual data values.

Figure 15. Map Fields/Values tab

NOTE ►► You cannot map fields until you have selected the current
source and destination tables. You cannot map and convert values
until you have mapped the corresponding field.

58 MDM Import Manager Reference Guide

Match Records Tab
The Match Records tab (tab in bottom pane) contains several controls
(Figure 16). Use the dual-list Value Matching control to specify the
destination fields to use as matching fields. Use the Default Import
Actions grid to specify the default import action for each group of source
records. And use the Source Records grid and Matching Destination
Records grid to override the default import action for each individual
source record, and for each source record, each of its matching
destination records.

Figure 16. Match Records tab

NOTE ►► You cannot perform record matching until you have
selected the current source and destination tables and mapped the
fields that you intend to use as matching fields.

MDM Import Manager Reference Guide 59

Import Status Tab
The Import Status tab (tab in bottom pane) summarizes the status of the
current import (Figure 18). Use the Import Status tab to view the status
of: (1) table mapping; (2) table joins; (3) field lookups; (4) field mapping;
(5) field mapping; (6) field and value mapping; (7) matching fields; (8)
record matching; and (9) import actions. The Import Status tab also
identifies the next required action items in the import workflow.

Figure 18. Import Status tab

60 MDM Import Manager Reference Guide

Status Bar
The Status bar displays different information for each tab, as
summarized in Table 2.

Table 2. Status Bar Information

Tab Position Status Bar Items
1  “p records”
Source Preview 2  “q fields”
3  “r tables”
Partition Field/Value n/a  none
 “r selected”
1*  “r leaf nodes selected”
 “s overmapped”
 “m values”
 “m leaf node values”
 “m of n values converted”
Map Fields/Values
 “m of n values mapped”
2*
 “m of n leaf node values converted”
 “m of n leaf node values mapped”
 “m of n values mapped (p added)”
 “m of n leaf node values mapped (p added)”
3  “x of y fields mapped”
 “r source selected”
1
Match Records  “r destination selected”
2  “m of n records active”
Import Status n/a  none
* Only leaf nodes considered when the mapped destination field is a hierarchy lookup field.

NOTE ►► The MDM Import Manager displays certain information

listed above only if the value is: (1) greater than one (e.g. “r selected”);
or (2) nonzero (e.g. “s overmapped”). Moreover, it sometimes uses the
same embossed position on the status bar to display different
information, depending upon the circumstances (e.g. “m values”, “m of
n values converted” and “m of n values mapped”).

MDM Import Manager Reference Guide 61

Configuration Options
Configuration options affect how the MDM Import Manager interprets
and processes source data.
You can use the Options command to open the Configuration Options
dialog and change the option settings for the MDM Import Manager. The
dialog contains a grid that enables you to specify values for each MDM
Import Manager option.
Options for the MDM Import Manager are listed in Table 3.

Table 3. Configuration Options

Option Description
The default setting for how implicit and explicit NULLs
are handled in source data when you map a field that
Default flat NULL handling contains multiple delimited values:
 Preserve NULLs
 Ignore NULLs
The default setting for how implicit and explicit NULLs
are handled in source data when you map a field that
contains hierarchy values (see “The Hierarchy NULL
Handling Option” on page 233):
 Preserve NULLs
Default hierarchy NULL  Ignore Non-Leaf Trailing NULLs
handling  > Preserve Leaf NULLs as NULLs
 > Preserve Leaf NULLs as Internal Leaf Nodes
 Ignore All Trailing NULLs
Ignore Non-Leaf Trailing NULLs has been split into two
subsettings for flexible handling of the leaf trailing
NULLs.
The default setting for how case differences are
handled when determining the set of distinct values for
a source field:
 Case insensitive
Default case sensitivity
 Case Sensitive
For SQL Server or Oracle data sources, case sensitivity
of the source field is determined by database collation
and cannot be changed by the MDM Import Manager.
The value delimiters are the strings used by various
Default value delimiters operations both to identify delimited source data values
and to separate merged data values.
The default delimiter string used by the Split into
Split Hierarchy [SH] Hierarchy command to identify individual delimited
values within a single field (“\”).
The default delimiter string used by the Split into
Split Multiple Fields [SM] Multiple Fields command to identify individual delimited
values within a single field (“;”).
The default delimiter string used by the Combine
Partition Combine [PC] Partitions command to separate each pair of individual
merged values (“;”).

62 MDM Import Manager Reference Guide

 Option Description
The default delimiter string used to identify individual
delimited values within a single field when mapping to a
Multi-Valued [MV] multi-valued field or a compound lookup field (“;”).
Accepts “\t” for Tab and “\n” for Return.
The default delimiter string used to identify coupled
Coupled Numeric [CN] numeric value pairs when mapping to a coupled
numeric attribute (“@”).
The default setting for whether NULLs in source data
are ignored or used to overwrite existing values:
Default NULL interpretation
 Ignore NULLs
 Overwrite existing values with NULLs
The default setting for whether incoming source data
values either are appended to or completely replace the
set of existing destination data values of a multi-valued
Default multi-valued update field when updating existing records:
 Append
 Replace
The default setting for whether incoming source
subrecords either are appended to, completely replace,
or update the set of existing qualified links of a qualified
lookup field when updating existing records:
 Append
 Replace
Default qualified update
 Update
See also the Default Matching Qualifiers option below.
Note that the Append and Replace options use matching
qualifiers for qualified link aggregation only, while the
Update option uses them both for qualified link
aggregation and for matching.
The default setting for which qualifiers to use as
matching qualifiers when matching existing qualified
links when updating existing records:
 None
Default matching qualifiers
 All
You can set specific subsets of matching qualifiers only
at the field level, since the set of qualifiers is different
for each qualified lookup field.
Default tuple update Tuple updating is performed based on these settings.
The default setting for which tuple member fields to use
as matching fields for the tuple’s records:
 None
Matching tuple fields  All
Because each tuple field can have a different set of
member fields, you can set specific subsets of matching
tuple member fields only at the field level.
The default setting for handling incoming tuple records
which do not already exist for the matching record:
New records option
 Skip
 Create

MDM Import Manager Reference Guide 63

 Option Description
The default setting for handling incoming tuple records
which already exist for the matching record:
 Skip
Existing records option  Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
 Delete
The default setting for whether the data groups
hierarchy is hidden or displayed in the Destination
Default data group display Values grid for object lookup destination fields:
 Hide data group
 Display data group
The Automap and Find next matching destination field
commands match source and destination field names
Find matching field
based on these settings (see “The Find Matching Field
Options” on page 275).
The type of comparison to perform of the source field
name against the destination field name:
Operator
 Contains
 Exact
Whether the comparison should be against normalized
values or alternatively, whether or not to match case:
Type  Normalized
 Case insensitive
 Case sensitive
Whether to automatically automap fields when loading
Automap fields when loading
an import map into Import Manager. Does not apply to
map
MDIS. (No/Yes).
Value automap Value automap is performed based on these settings.
How to match hierarchy values when using Automap in
the Value Conversion and Mapping grids:
 Full path
Hierarchy
 Leaf nodes only
The Full path setting ignores the root node on the
source and/or the destination hierarchies.
Whether to normalize source and destination values
Normalized
when attempting to automap (No/Yes).
Automatically map NULL Whether to automatically map the NULL source value to
value the NULL destination value (No/Yes).
How to handle missing units when importing a
taxonomy classification (see “Taxonomy Import" on
page 161):
Automatically add missing units
 None
 When all missing
 When any missing
Record Matching Record matching is performed based on these settings.
Whether to perform record matching when you select
Perform record matching
each matching field (No/Yes).

64 MDM Import Manager Reference Guide

 Option Description
How to match against the value of a normalized record
matching field (see “Normalized Record Matching Field”
Normalized record matching on page 360):
field
 Normalized match against normalized value
 Exact match against non-normalized value
Import The actual import is performed based on these settings.
How to handle import into Writeable Once fields (see
“Writeable Once Destination Fields” on page 266):
Writeable once field update
 Ignore Conflicting Value
 Reject Record on Conflicting Value
Whether to collapse and merge duplicate source
Merge source records using records for import as a single virtual source record (see
key field “The Merge Source Records Option” on page 351)
(No/Yes).
Whether to collapse and merge duplicate subrecords as
Merge qualified links using
a single qualified link (see “The Merge Qualified Links
matching qualifiers Option” on page 352) (No/Yes).
Synchronize Family Whether to synchronize the entire Family Hierarchy
Hierarchy after import after import (No/Yes).
Whether to update the value of a record matching field
Do not update record
(see “Updating the Record Matching Field” on page
matching field value
356) (No/Yes).
Whether to auto-join checked out records during import
Auto-Join checked out
(see “The Auto-Join Checked Out Records Option” on
records
page 378) (No/Yes).
Whether or not to check out the imported records or
launch a workflow job containing the imported records
(see “The Checkout/Workflow Option” on page 379):
 None
Checkout/Workflow
 Checkout
 workflowname
 :
 workflowname
Whether or not – and how – to reassign existing key
mappings during the import (see “Key Mapping
Reassign existing key Reassignment” on page 410):
mappings  No
 Into original records only
 Into original and checked out records
Whether or not to delete duplicate qualified links:
Eliminate duplicate qualified
links  Do not eliminate duplicates
 Eliminate duplicates (cached qualifiers only)
The MDM Import Manager can automatically link
Autolink attributes unlinked attributes to a category as part of the import
based on these settings.
Whether to automatically link unlinked attributes to the
When mapping to an category value of a source record when source record
unlinked attribute fields containing non-NULL values are mapped to those
attributes (No/Yes).

MDM Import Manager Reference Guide 65

 Option Description
Whether to automatically link unlinked attributes to a
When reassigning an
category when the category value is changed for an
existing record to
existing record and data values would otherwise be
another category
discarded (No/Yes).
The MDM Import Manager can automatically map
unmapped values (see “The MDIS Unmapped Value
Default MDIS handling
Handling Option” on page 268) and set the Default
Import Action for MDIS based on these settings.
Whether to first attempt to automap to existing
Automap unmapped values destination values source data values that do not exist
in the map file (No/Yes)
How to handle still unmapped source data values
(either because they do not exist or because
automapping was turned off):
 Ignore
Unmapped value handling
 Add
 Fail
You can set the Map option only at the field level, since
the set of destination values is different for each field.
The numeric format options allow you to specify the
Numeric format
format of incoming numeric data values.
Decimal symbol The decimal symbol for source data values (“.”).
1000s separator The thousands separator for source data values (“,”).
The date options allow you to specify the format of
Dates
incoming date values.
The format of incoming dates:

Date format  YYYYMMDD

 MM/DD/YYYY
 DD.MM.YYYY
NULL value string The string to interpret as a NULL date value (000000).
The Relationships options allow you to specify how the
Relationships MDM Import Manager handles the import of relationship
links.
How to group the source records when replacing
existing relationship links:
Aggregate by
 Parent
 Child
How to handle the set of existing relationship links:

Links update  Append

 Replace
 Update
How to handle new links when updating existing links:
New links
(update only)  Skip
 Create

66 MDM Import Manager Reference Guide

 Option Description
How to handle existing links when updating existing
links:
 Skip
Existing Links
 Update (Mapped Qualifiers)
(update only)
 Update (All Mapped Fields)
 Replace
 Delete
Whether or not – and how – to expand multilingual text
fields into multiple fields in the Destination Fields grid of
the Map Fields/Values tab (see “Field Mapping to
Expand Multilingual Text Fields Multilingual Destination Fields” on page 397):
 No
 As Multiple Fields
 As Field and Language
How to display attributes and attribute values in the
Destination Fields and Destination Values grids of the
Map Fields/Values tab (see “Attribute Data Import as
Attribute display Name/Value Pairs” on page 167):
 As multiple attributes
 As attribute/value hierarchy
The type of comparison to perform of the source field
value against the destination field value when
Attribute/value hierarchy automapping to the Attribute/Value hierarchy:
automap operator
 Contains
 Exact
Whether to use predefined priority values (Highest,
Display priority field for High, Normal, Low, Lowest) or allow import of custom
Attribute taxonomy links table integer values:
using  Predefined values
 Custom values
Whether or not to disable the automatic mapping of
Disable automatic hierarchy values by remote key for source fields mapped to
value mapping according to keymappable hierarchy lookup fields :
remote key  Yes
 No

MDM Import Manager Reference Guide 67

 To open the Configuration Options dialog that allows you to view and
edit options settings:
1. Choose Configuration > Options from the main menu.
2. MDM opens the Configuration Options dialog shown in Figure 19.

Figure 19. Configuration Options dialog

TIP ►► Click the “+” icon to expand any node in the tree.

3. Click in the Value column for the option setting you want to change.
4. If the Value cell is a drop-down list, select the desired option. If the
Value cell is an edit field, double-click inside the field and replace the
existing value with a new value.
5. Click OK to save any new values and close the dialog.
MULTILINGUAL ►► More information about languages and
multilingual repositories is provided in “Part 14: Multilingual Support.”

REMOTE SYSTEMS AND MDM ►► More information about remote

systems and MDM is provided in “Part 15: Remote Systems and
MDM.”

68 MDM Import Manager Reference Guide

Import Maps
All of the partitions, mappings, and other conversions can be saved in
an import map within the MDM repository. A map is useful if you want to
save your work before you have performed the import, if you want to
subsequently apply the same transformations for future imports of
additional records or updates from the same data source, or if you want
to later use the map in conjunction with MDIS processing.
Specifically, saving a map takes a snapshot of the state of the current
Import Manager session so that it can later be restored when you load
the map during a future session, as summarized in Table 4.

Table 4. Import Map Components

Setting Description
All of the global default configuration option settings are saved in the map, along
Configuration
with overridden field-level settings. When you load the map, all the settings are
Options
restored.
All of the joins and lookup fields are saved in the map, including whether the
Joins and
lookup fields were looked up using [All]. When you load the map, saved joins
Lookups
and lookups that do not exist in the source file are ignored.
All of the field and value level partitions are saved in the map. When you load
Partitions the map, saved partitions of fields or values that do not exist in the source file
are ignored.
For each field, all of the field mappings are saved as source/destination field
Field
name pairs. When you load the map, saved mappings for which either field
Mappings
name does not exist in the source file are ignored.
For each field, value conversions are saved as follows: (1) at the field level, the
Value filters are saved; and (2) at the value level, all of the manually overridden value
Conversions conversions are saved as source/destination value pairs. When you load the
map, saved values that do not exist in the source file are ignored.
For each field, all of the value mappings are saved as source/destination value
Value
pairs. When you load the map, saved mappings for which either value does not
Mappings
exist in the source file are ignored.
Record
The record matching fields and the matching options are saved in the map.
Matching
Import
The default import action for each class and type of match are saved in the map.
Actions

CAUTION ►► When you open an existing map against a new source

file and resave it, the MDM Import Manager discards old map data that
does not exist in the new source file and replaces it with current data.
This means, for example, you will lose the mapping and/or cloning of a
field that does not exist, and/or the mapping of values for fields and/or
values that do not exist. To open and resave a map, be sure to open it
against the original source file against which the map was created.

MDM Import Manager Reference Guide 69

Import Map Operations
The following sections describe the various import map operations, and
are summarized in Table 5.
Table 5. Import Map Operations
Operation Description

Open Map Opens the selected map.

Close Map Closes the current map.

Save Map Saves the current map.

Save Map As Saves the current map to a new name.

Save Map Update Saves an update of the current map.

Import Map Imports a map file into the repository.

Export Map Exports a map file from the repository.

OPENING AND CLOSING MAPS

When you open a map, MDM performs a series of tests to ensure that
the import map is compatible with the source file and destination
repository currently connected to the Import Manager.
These tests and their error conditions are described in Table 6.

Table 6. Tests Performed When Opening an Import Map

Test Error Condition Outcome

Source tables in map are Source table name in map is Import Manager
present in connected source. not found in connected source. cannot open map.

Destination tables in map are

Destination table in map is not Import Manager
present in connected MDM
found in connected repository. cannot open map.
repository.
Import Manager
Mapped source fields are Source field name in map is
opens map without
present in connected source. not found in connected source.
the missing field.
Mapped destination fields are Import Manager
Destination field in map is not
present in connected MDM opens map without
found in connected repository
repository. the missing field.

70 MDM Import Manager Reference Guide

When you open an import map, the MDM Import Manager loads the
configuration option settings that were stored in the map, overriding the
settings currently in effect. When you later close the map or exit the
MDM Import Manager, MDM asks whether or not you want to save the
map’s settings as the new default settings, as shown in Figure 20.

Figure 20. Configuration options save confirmation dialog

Only one import map can be open at a time in the Import Manager. If
you try to open a new map without closing the current one, the Import
Manager prompts you to save changes to the current map before it
opens the new map. If there are no changes to save, the Import
Manager automatically closes the old map and opens the new one.

 To open an import map:

1. Choose File > Open from the main menu to open the Open Map
dialog, shown in Figure 21 below.

Figure 21. Open dialog

2. Select a map from the Available maps list and click OK.
3. When opening the map, Import Manager verifies that the map is
compatible with the current repository and alerts you if it finds any
problems.

 To close the currently open map:

♦ Choose File > Close from the main menu.

MDM Import Manager Reference Guide 71

SAVING MAPS
You can save import maps as described in this section.

 To save the current MDM Import Manager state to the current map:
1. Choose File > Save from the main menu.
2. MDM saves the current MDM Import Manager state to the current map.

 To save the current MDM Import Manager state to a new map:

1. Choose File > Save As from the main menu.
2. MDM opens the Save Map As dialog shown in Figure 22.

Figure 22. Save As dialog

3. Select or type a name for the new map and click OK to close the dialog
4. MDM saves the current state in a map within the MDM repository.
TIP ►► The File > Save As command saves to a new map. You can
also use the File > Save command to save to the current map if one
has already been opened or saved.

 To save updated header info to a map:

1. Open the map.
2. Choose File > Save Update from the main menu.
3. MDM saves updated header info to the current map.

72 MDM Import Manager Reference Guide

IMPORTING AND EXPORTING MAP FILES
Sometimes you may want to transfer maps among MDM repositories.
You can export a map from a repository by saving it as a map file on
your file system. You can then import the map file into the target
repository, as described in this section.

 To export a map from the repository into a map file:

1. Choose File > Export from the main menu.
2. MDM opens the Select Map dialog. shown in Figure 23.

Figure 23. Select Map dialog

3. Select a map from the list of maps and click OK to close the dialog.
4. MDM opens the Windows file Open dialog for you to name the map
file. Navigate to the appropriate folder, select or type a file name, and
click Save.
5. MDM exports the map into a map file.

 To import a map file into a map within the repository:

1. Choose File > Import from the main menu.
2. MDM opens the Windows file Open dialog.
3. Browse to the map file and click Open.
4. MDM opens the Save As dialog shown in Figure 22 above.
5. Select or type a name for the new map and click OK to close the dialog
6. MDM saves the map file into a map within the MDM repository.

NOTE ►► After importing a map file from another repository, it is a

good idea to open the map in Import Manager and verify that (a) the
map is compatible with the new repository; and (b) the field mappings
are correct (see “Opening and Closing Maps” on page 70 for more
information).

MDM Import Manager Reference Guide 73

UPDATING MAPS FOR MDM 7.1
All import maps created in MDM 5.5 must be updated by MDM before
they can be used with the MDM 7.1 versions of Import Manager and
MDIS. Until MDM updates them, you cannot open outdated maps in the
Import Manager or use them with MDIS.
Whenever you import an old map or connect MDM 7.1 Import Manager
to an MDM repository containing outdated maps, MDM offers to
automatically update the outdated import maps. If MDM cannot
completely update a map, you must update it manually in Import
Manager before it can be used.
NOTE ►► Outdated maps are grayed-out and labeled “[Outdated]” in
the Import Manager’s Open Map dialog.

NOTE ►► You cannot manually update a map until MDM has

attempted to update it automatically

CAUTION ►► Updating maps is irreversible. You may want to archive

a repository before updating its maps.

You can update outdate maps as described below.

 To automatically update all outdated maps in an MDM repository:

1. Connect Import Manager to the MDM repository.
2. Import Manager opens the dialog shown in Figure 24.

Figure 24. Outdated Maps dialog

3. Click Yes on the dialog to update the maps.

NOTE ►► If you click No, you will be unable to open the outdated
maps in Import Manager or use them with MDIS, and Import Manager
will reopen the dialog the next time you connect to the MDM repository.

4. MDM updates the old maps and writes the results to the
UpdateMapLog xml file in the Import Manager’s installation directory.

74 MDM Import Manager Reference Guide

 To manually update an outdated map in an MDM repository:
1. In Import Manager, connect to a source file that corresponds to the
outdated map.
2. Choose File > Open from the main menu to open the Open Map
dialog.
3. Select the map from the Open Map dialog and click OK.
NOTE ►► Maps in need of manual repair have [Requires Manual
Update] appended to their names.

4. Correct the problems noted for the map in the UpdateMapLog file.
5. Save the map.

MDM Import Manager Reference Guide 75

PART 4: BROWSING SOURCE DATA

This part of the reference guide explains how to familiarize yourself with
the source data to determine what processing will be necessary to
perform the import.

77
Overview
Data that needs to be imported usually comes from a foreign source
and you’ve probably never seen it before.
Consequently, the first step in preparing source data for import is simply
to familiarize yourself with its table and field structure, as well as the
underlying source records and data values. Becoming familiar with
source data allows you to determine what MDM Import Manager
processing will be necessary before you can perform the import.
The MDM Import Manager allows you to browse and become familiar
with source data in several different ways:
• Source Preview tab. For browsing all of the records and fields of
each table in a traditional row/column grid.
• Source Hierarchy pane. For browsing the overall table and field
structure of the source file – and the underlying set of distinct data
values for each field – in a multi-level tree.

These two complementary methods of browsing source data are

described in the following sections.

MDM Import Manager Reference Guide 79

The Source Preview Tab
The Source Preview tab displays a grid containing the records of each
source table or nested XML element, with a row for each record and a
column for each of the fields, as shown in Figure 25.

Figure 25. The Source Preview tab

NOTE ►► The Tables subpane displays nested XML elements in

their natural hierarchy, rather than in a list.

 To browse the records of any of the source tables:

1. If necessary, click on the Source Preview tab to make it the active tab.
2. In the Tables subpane, select the source table or nested XML element
whose records you want to browse.
3. In the Records subpane, interactively browse the records of the
selected table and sort by any of the columns in ascending or
descending order.
NOTE ►► When you look up fields based on join relationships
between tables, the MDM Import Manager appends [Lookup] to the
name of the table in the Tables subpane, and adds each lookup field to
the grid in the Records subpane as “field <table>” (where ‘field’ is the
name of the lookup field and ‘table’ is the name of the lookup table).
(See “Table Joins and Lookups” on page 101 for more information
about table joins and lookups.)

80 MDM Import Manager Reference Guide

The Source Hierarchy Tree
The Source Hierarchy tree provides a high-level view of the table and
field structure of the source file, the join relationships between tables,
the lookup fields, and – if you have used the Display Distinct Values
command to display the distinct data values for one or more fields – the
set of values for each field, as shown in Figure 26.
The Source Hierarchy tree also provides a place from which to perform
other source operations, such as creating join and lookup relationships,
adding, cloning, and splitting fields, and partitioning fields and values.

Source file

Tables

Fields

Data values

Matching fields

Lookup table

Lookup field
Figure 26. Nodes of the Source Hierarchy tree

The types of nodes of the Source Hierarchy tree are described in the
following sections, and summarized in Table 7.

MDM Import Manager Reference Guide 81

TABLES, FIELDS, AND DATA VALUES
At a minimum, the Source Hierarchy tree contains two levels that
include the tables and fields of the source file whose records you intend
to import.
If you display the list of distinct data values for any field in the current
source table, the Source Hierarchy tree contains an additional level that
includes the data values.
TIP ►► You can use the Display Distinct Values command to display
the list of distinct data values for any field in the current source table
(see “Displaying Distinct Values” on page 86 for more information).

TABLE JOINS AND LOOKUPS

If the source file contains join relationships between source tables, or if
you create your own join relationships, the Source Hierarchy tree
identifies the joined matching fields, and also includes the joined lookup
tables and corresponding matching fields as children of the
corresponding table and matching field nodes, respectively.
TIP ►► You can use the Join command to create join relationships
that don’t already exist in the source data file (see “Joining Tables” on
page 103 for more information).

NOTE ►► The MDM Import Manager appends [Join] to the name of

both matching fields that participates in a join relationship.

Finally, if you have looked up fields from joined lookup tables to

combine all of the fields for import into a single source table, the Source
Hierarchy tree also includes the lookup fields as children of the joined
lookup table nodes.
TIP ►► You can use the Lookup command to look up fields from
joined lookup tables into the current source table (see “Looking Up a
Field” on page 104 for more information).

NOTE ►► Joins are transitive: if Table A is joined to Table B and

Table B is joined to Table C, Table C is effectively joined to Table A.
The MDM Import Manager automatically detects this condition and
allows “transitive lookups” from Table C into Table A.

NOTE ►► See “Table Joins and Lookups” on page 101 for more
information about how tables and fields associated with joins and
lookups are reflected in the Source Hierarchy tree.

82 MDM Import Manager Reference Guide

Table 7. Source Hierarchy Tree Nodes
Level Node Type Description
Root File/Database The name of the source file or database.
1 Tables The list of tables, in alphabetical order.
The list of fields of each table, in alphabetical order (as
2 Fields children of each table node). (For matching fields, [Join] is
appended to the name of the field.)
The distinct data values of each field (as children of each
3 Data Values
field node, if displayed).
Each table node has as its last children (after the field
2 Lookup Tables
nodes) any lookup tables to which it is joined.
Each field node has as its first children (before the value
3 Matching Fields
nodes) any matching fields to which it is joined.
Each lookup table node has as its first children any lookup
3 Lookup Fields
fields that have been looked up from it.

FIELD TYPES IN THE SOURCE HIERARCHY TREE

The Source Hierarchy tree includes fields from the original data source,
as well as additional fields that you create during the import process.
The different types of fields as well as the operations you can perform
on them are summarized in Table 8.

Table 8. Field Types and Permitted Operations

Field Type Clone Delete Rename Split Partition
Original Source Field • • •
Lookup Field • • •
Added Field • • • • •
Cloned Field • • • • •
Partitioned Field (1) •
Split Hierarchy Field (1) • • (2)
Split Multiple Field • • • • •
Created Hierarchy Field [delimited] • • •
Created Hierarchy Field [hierarchy] (1) • • (2)
Compound Field • •
‘•’ = permitted;
1
Clone before partitioning or splitting, then apply the operation to the original and the clone.
2
Split original into multiple rather than into hierarchy, and then partition the split fields.

MDM Import Manager Reference Guide 83

XML SOURCE FILES AND THE SOURCE HIERARCHY TREE
Prior to MDM 7.1, Import Manager flattened all record data from XML
source files before importing the data into MDM (see “Data Staging” on
page 47). Specifically, the hierarchical nesting that occurrs naturally in
XML files was replaced by “artificial” tables which then had to be
“joined” in the Import Manager to restore the original data relationships.
Starting in MDM 7.1, Import Manager no longer flattens XML record
data. Instead, Import Manager now preserves the original hierarchical
structure of XML source files and reflects this structure in the Source
Hierarchy tree, as shown in Figure 27.

Figure 27. XML Structure in the Source Hierarchy Tree

Ways in which the Source Hierarchy tree now reflects an XML file’s
structure include:
• Top-level node is the source XML file
• Tables in the tree represent nested XML structures
• All tables are nested under the root element
• Fields in the tree represent data-storing XML elements
• Joins and _ID fields are no longer added or required

With these changes, users no longer have to manually recreate the

relationships implicit in the XML schema, as they are preserved by
MDM and accurately reflected in the source hierarchy tree.
NOTE ►► To ensure that Import Manager correctly interprets the
structure of an XML file, specify its corresponding XML schema file in
the Connect to Source dialog (see “Starting Import Manager” on page
43 for more information).

84 MDM Import Manager Reference Guide

Source Hierarchy Operations
The following sections describe the various operations you can perform
on nodes of the Source Hierarchy tree, as summarized in Table 9.

Table 9. Source Hierarchy Operations

Operation Description
Display
Displays the distinct values for the selected field.
Distinct Values

Add Field Adds an empty field to the hierarchy.

Clone Field Clones the selected field and its values.

Rename Field Renames the selected field.

Delete Field Removes the selected field from the hierarchy.

Join Joins the selected field to a field in another table.

Unjoin Unjoins the selected field.

Lookup Looks up a field from another joined table.

Unlookup Removes the selected lookup field from the hierarchy.

Split into Splits the data values within the selected field into a
Hierarchy hierarchy value.
Split into Splits the data values within the selected field into
Multiple Fields multiple fields.
Creates a new field based on the selected fields and
Create Hierarchy Field
then splits it into a hierarchy.
Create Combines fields that have been mapped to the
Compound Field multiple display fields of a destination lookup field.
Create
Creates a new table by pivoting on the selected fields.
Pivot Table
Create Reverse Creates a new table by reverse pivoting on the
Pivot Table selected fields.

NOTE ►► As of MDM 7.1, operations that previously required an

updateable data source are now performed virtually. As a result, all
source hierarchy operations are now available to all data types.

MDM Import Manager Reference Guide 85

DISPLAYING DISTINCT VALUES
In conjunction with field-at-a-time rather than record-at-a-time
processing, distinct value lists make import exception handling much
more efficient because:
• The number of distinct values for each field is dramatically smaller
than the total number of values.
• Each distinct value represents and is a proxy for not only all
instances of the value across the entire set of underlying records, but
also the subset of records that contain the value.
• Examining and then transforming each source data value must be
done only once for each distinct value rather than once for each
instance of the value; and as a result.
• The maximum number of exceptions is the sum rather than the
product of the number of distinct values in each mapped field.

When you use the Display Distinct Values command to display the list of
distinct values for a field in the Source Hierarchy tree, the MDM Import
Manager adds the values to the tree as children of the field node, as
shown in Figure 28.

Distinct values

Figure 28. Display Distinct Values command

NOTE ►► The distinct value list is like the lookup values of the
corresponding lookup table if such a table actually existed.

NOTE ►► You can only display data values in the Source Hierarchy
tree for fields of the current source table.

Displaying values as nodes in the tree is an optional convenience that

allows you to efficiently familiarize yourself with the set of distinct values
for a field (as opposed to browsing multiple occurrences of each value
in the Source Preview grid).

86 MDM Import Manager Reference Guide

 NOTE ►► The set of distinct values is meaningful primarily for fields
that have a relatively small set of distinct values relative to the total
number of records, not for an index or description field, where the
values may be unique for each record. The MDM Import Manager
warns you before displaying the values if the number of distinct values
is greater than 100, as shown in Figure 29.

Figure 29. Confirm Distinct Value Display dialog

Value nodes also provide a foundation for partitioning fields and values.
(See “Partitioning and Field Mapping” on page 173 for more information
about partitioning.)
NOTE ►► If you haven’t already used the Display Distinct Values
command to manually add the values as tree nodes, the values and
value combinations are automatically added when you partition a field.
By contrast, you must use the Display Distinct Values command to
manually add the value nodes prior to partitioning an individual field
value.

NOTE ►► The set of distinct values that are added as nodes to the
tree correspond to the set of distinct values that appear in the Value
Conversion and Mapping grids of the Map Fields/ Values tab (see “The
Value Conversion and Mapping Grids” on page 210).

TIP ►► You can also use the Display Distinct Values command in the
Destination Hierarchy tree to display the set of values for destination
fields, attributes, and qualifiers of the current destination table. If the
field is a hierarchy field, the entire tree structure is added to the
Destination Hierarchy tree.

 To display the distinct data values for a field:

1. In the Source Hierarchy tree, select the field node whose distinct data
values you want to display.
2. Right-click on the node and choose Display Distinct Values from the
context menu, or choose Source > Display Distinct Values from the
main menu.
3. MDM adds the set of distinct data values as children of the selected
field node (Figure 28 above).

MDM Import Manager Reference Guide 87

REMOVING DISTINCT VALUES
You can use the Remove Distinct Values command to remove the set of
distinct values from the hierarchy tree, as described in this section.
NOTE ►► You cannot use the Remove Distinct Values command if
the field or any of its values have been partitioned.

 To remove the distinct data values for a field:

1. In the Source Hierarchy tree, select the field node whose distinct data
values you want to remove.
2. Right-click on the node and choose Remove Distinct Values from the
context menu, or choose Source > Remove Distinct Values from the
main menu.
3. MDM remove the set of distinct data values as children of the selected
field node.

SETTING THE CASE SENSITIVITY OPTION

You can specify whether to ignore or recognize case differences in the
source data when the MDM Import Manager determines the set of
distinct values for a field.
NOTE ►► The procedure for using the Set Case Sensitivity command
is included here for reference. (See “The Case Sensitivity Option” on
page 237 for complete information about case sensitivity.)

 To set the Case Sensitivity option for one or more fields:

1. In the Source Hierarchy tree, select the field nodes for which you want
to set the Case Sensitivity option.
2. Right-click on one of the nodes and choose Set Case Sensitivity from
the context menu, or choose Fields > Set Case Sensitivity from the
main menu.
TIP ►► Alternatively, you can right-click on the field in the Source
Fields grid and choose Set Case Sensitivity from the context menu.

3. MDM displays a cascading menu of three choices:

 Case Insensitive
 Case Sensitive
 Inherit Default
4. To set the Case Sensitivity option and override the global default
setting, choose either Case Insensitive or Case Sensitive, or
alternatively, to restore inheritance, choose Inherit Default.

88 MDM Import Manager Reference Guide

 NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

ADDING AN EMPTY FIELD

You can add to the source data an empty field with the single NULL
value in its list of distinct values. For example, if the source data is
missing a field for which all of the values would be the same (such as
when data from a particular manufacturer is missing the manufacturer
field), you can add a field, and then use the Map Fields/Values tab to
map the single NULL value to the corresponding destination value (see
“Missing Source Data Fields” on page 231 for more information).
When you use the Add Field command, the MDM Import Manager
creates an empty field and adds it to the Source Hierarchy tree. Added
fields are highlighted in italics, as shown in Figure 30.

Figure 30. Adding an empty field

TIP ►► You can delete an added field using the Delete Field
command.

 To add an empty field to the Source Hierarchy tree:

1. In the Source Hierarchy tree, select the field node after which you want
to add an empty field.
2. Right-click on the node and choose Add Field from the context menu,
or choose Source > Add Field from the main menu.
3. MDM adds a new field node named “New Field ” as a sibling of and just
after the original node (Figure 30 above).
4. Press F2 to edit the name of the new field.
5. MDM highlights the name of the new field for editing.
6. Type the name you want for the new field and press Enter.

ADDING A CLONE FIELD

If you need to perform two different operations on a single source field,
such as: (1) mapping it to two different destination fields; or (2) mapping

MDM Import Manager Reference Guide 89

it to one destination field by itself and also partitioning it by another
source field, you can clone the field to create a copy of the original field.
When you use the Clone Field command to create a clone of a source
field, MDM creates a duplicate of the original field that contains the
same data value for each source record. You can then perform
subsequent processing on the clone field just as if it were an original
field in the source data. Clones in the Source Hierarchy tree are
highlighted in italics, as shown in Figure 31.

Figure 31. Adding a clone field

TIP ►► You can delete a clone field using the Delete Field command.

 To create a clone of a field:

1. In the Source Hierarchy tree, select the field node you want to clone.
2. Right-click on the node and choose Clone Field from the context menu,
or choose Source > Clone Field from the main menu.
3. MDM adds a new field node named “field <Clone>” as a sibling of and
just after the original node (Figure 31 above).
4. Press F2 to edit the name of the clone.
5. MDM highlights the name of the clone for editing.
6. Type the name you want for the clone and press Enter.

RENAMING A FIELD
Fields that you add to the Source Hierarchy tree are automatically
named based on the original field name and the command that you use
to create the field. You can then rename the field as described in this
section.
The commands that add fields to the Source Hierarchy tree and the
corresponding default fields names are summarized in Table 10.

90 MDM Import Manager Reference Guide

Table 10. Default Names for Fields Added to the Source Hierarchy Tree
Command Field Name
1,2
Add Field New Field (n)
Clone Field1,2 field <Clone> (n)
Split into Hierarchy1,2 field <Split Hierarchy> (n)
1,3
Split into Multiple Fields field <Split Multiple> <m> (n)
1
“field” is the name of the original field
2
“(n)” does not appear unless n >= 2
3
‘m’ ranges from 1 to the number of new fields

NOTE ►► The maximum field name length is 64 characters for import

maps which use Access or Excel source types.

 To rename a field in the Source Hierarchy tree:

1. In the Source Hierarchy tree, select the field node you want to rename.
2. Right-click on the node and choose Rename Field from the context menu,
or press F2, or choose Source > Rename Field from the main menu.
3. MDM highlights the field name for editing. Type the new name you
want for the field and press Enter.

DELETING A FIELD
You can delete a field that you previously added to the Source
Hierarchy tree as described in this section.

 To permanently delete a field from the Source Hierarchy tree:

1. In the Source Hierarchy tree, select the field node you want to delete.
2. Right-click on the node and choose Delete Field from the context menu,
or press Del, or choose Source > Delete Field from the main menu.
3. MDM prompts you to confirm that you really want to delete the field.
Click OK to remove the field from the Source Hierarchy tree.

JOINING TABLES
For data sources that do not store relationships between tables or do
not make them available through a programmatic interface, or
alternatively, to create additional relationships that are not explicitly
defined in the data source, the MDM Import Manager allows you to
manually define join relationships from a primary table to a lookup table.

MDM Import Manager Reference Guide 91

 NOTE ►► The procedures for using the Join and Unjoin commands
are included here for reference. (See “Table Joins and Lookups” on
page 101 and “Joining Tables” on page 103 for complete information
about joining tables.)

 To manually join a primary table to a lookup table:

1. In the Source Hierarchy tree, select the primary matching field node in
the primary table from which you want to establish the join relationship.
NOTE ►► The primary table does not need to be the current source
table.

2. Right-click on the node and choose Join from the context menu, or
choose Source > Join from the main menu, and select first the lookup
table and then the lookup matching field from the two-level cascading
menu of source tables and fields for each table.
NOTE ►► You can only join fields with the same data type; other
fields are disabled in the cascading menu of fields.

NOTE ►► You can also drag-and-drop the primary matching field

from the primary table onto the lookup matching field in the lookup
table to create the join relationship.

3. MDM creates the new join relationship and reflects it in the Source
Hierarchy tree.
NOTE ►► See “The Source Hierarchy Tree” on page 81 for more
information about the nodes of the Source Hierarchy tree.

 To remove a join relationship between two tables:

1. In the Source Hierarchy tree, select either the primary matching field
node in the lookup table (i.e. the [To: …] node) or the lookup matching
field in the primary table (i.e. the [From: …] node) that corresponds to
the join relationship you want to remove.
2. Right-click on the node and choose Unjoin from the context menu, or
choose Source > Unjoin from the main menu.
3. MDM eliminates the join relationship and removes the matching field
nodes, the lookup table node, and any lookup field nodes from the
Source Hierarchy tree.

92 MDM Import Manager Reference Guide

LOOKING UP A FIELD
Once the join relationships have been established, you can look up
fields from a lookup table into a primary table, adding the field into the
list of primary table fields, and in effect, building an extended virtual
record consisting of fields from the primary table and lookup fields from
one or more lookup tables.
NOTE ►► The procedures for using the Lookup and Unlookup
commands are included here for reference. (See “Table Joins and
Lookups” on page 101 and “Looking Up a Field” on page 104 for
complete information about looking up fields.)

 To look up a field from a lookup table into a primary table:

1. In the Source Hierarchy tree, select the lookup table node in the
primary table into which you want to look up a field.
NOTE ►► The primary table does not need to be the current source
table.

2. Right-click on the node and choose Lookup from the context menu, or
choose Source > Lookup from the main menu, and select the lookup
field from the cascading menu of fields in the lookup table.
TIP ►► You can also drag-and-drop the lookup field from the lookup
table onto the lookup table in the primary table to create the lookup.

TIP ►► You can select [All] from the cascading menu to lookup all of
the fields of the lookup table.

3. MDM creates the new lookup field and adds it to the Source Hierarchy
tree.
NOTE ►► See “The Source Hierarchy Tree” on page 81 for more
information about the nodes of the Source Hierarchy tree.

 To remove a lookup field from a primary table:

1. In the Source Hierarchy tree, select the lookup field node you want to
remove in the primary table.
2. Right-click on the node and choose Unlookup from the context menu,
or choose Source > Unlookup from the main menu.
3. MDM eliminates the lookup field and removes its node from the Source
Hierarchy tree.
NOTE ►► When you remove a join relationship, the MDM Import
Manager automatically removes the lookup fields as well.

MDM Import Manager Reference Guide 93

SPLITTING A FIELD
You can split a field containing multiple delimited values into: (1) a
single field containing the values expanded into a hierarchy; or (2)
multiple fields, each containing a single value.
NOTE ►► The procedures for using the Split into Hierarchy and Split
into Multiple Fields commands are included here for reference. (See
“Splitting Fields” on page 146 for complete information about splitting
fields.)

NOTE ►► The Split into Hierarchy and Split into Multiple Fields
commands described in this section identify the individual delimited
values within a single field using the current Split Hierarchy or Split
Multiple Fields delimiter strings, respectively, which must be properly
set prior to performing the commands (see “Setting the Split
Delimiters” on page 150 for more information).

 To split a field with multiple delimited values into a hierarchy:

1. In the Source Hierarchy tree, select the field node whose multiple
delimited values you want to expand into a hierarchy.
2. Right-click on the node and choose Split into Hierarchy from the
context menu, or choose Source > Split into Hierarchy from the main
menu.
3. MDM adds a new field node named “field <Split Hierarchy>” as a
sibling of and just after the original node.
4. Press F2 to edit the name of the split field.
5. MDM highlights the name of the split field for editing.
6. Type the name you want for the split field and press Enter.

 To split a field with multiple delimited values into multiple fields:

1. In the Source Hierarchy tree, select the field node whose multiple
delimited values you want to split into multiple fields.
2. Right-click on the node and choose Split into Multiple Fields from the
context menu, or choose Source > Split into Multiple Fields from the
main menu.
3. MDM adds one or more new field nodes named “field <Split Multiple>
<m>” (where ‘m’ ranges from 1 to the number of new fields) as a
sibling of and just after the original node.
4. Press F2 to edit the name of each split field.
5. MDM highlights the name of the split field for editing.
6. Type the name you want for the split field and press Enter.

94 MDM Import Manager Reference Guide

BUILDING A HIERARCHY
You can build a hierarchy from a set of fields identifying parent/child
relationships, as described in this section.
NOTE ►► The procedure for using the Create Hierarchy Field
command is included here for reference. (See “Building a Hierarchy”
on page 156 for complete information about building a hierarchy.)

 To combine two or three fields containing hierarchy data into a

hierarchy:
1. In the Source Hierarchy tree, select the two or three field nodes you
want to use build into a hierarchy:
 Parent/child value pairs with node name – three
 Parent/child value pairs w/o node name – two
 Hierarchy codes with node name – two
2. Right-click on one of the nodes and choose Create Hierarchy Field from
the context menu, or choose Source > Create Hierarchy Field.
3. MDM opens the Create Hierarchy Field dialog shown in Figure 32.

Figure 32. Create Hierarchy Field dialog

4. In the Hierarchy Field Name edit control, type the name for the new
field.
5. In the Parent Field, Child Field, and Node Name Field drop-down
controls, identify the field containing the parent value, the child value,
and the node name value respectively.
TIP ►► For two-node selection of parent/child value pairs without a
node name, set the Node Name Field to None. MDM uses the values of
Child Field as the name of each node.

TIP ►► For two-node selection of hierarchy codes with a node name,

set the Parent Field to None. MDM automatically decodes the values of
Child Field to determine the parent/child hierarchy relationships.

6. Click OK to close the Create Hierarchy Field dialog.

7. The MDM Import Manager creates a hierarchy field named “field <Split
Hierarchy>” as a sibling of and just after the first new node (Figure 106
above).

MDM Import Manager Reference Guide 95

 8. Press F2 to edit the name of the hierarchy field.
9. MDM highlights the name of the hierarchy field for editing.
10. Type the name you want for the hierarchy field and press Enter.
NOTE ►► The Create Hierarchy Field command uses the current Split
Hierarchy delimiter string to delimit the nodes in the new hierarchy field
and then to split the individual delimited values within the new field
(see “Setting the Split Delimiters” on page 150 for more information).

CREATING A COMPOUND SOURCE FIELD

Please see "Creating a Compound Source Field" on page 275 for
complete information about compound source fields and how to create
them

CREATING A PIVOT TABLE

Please see “Pivoting” on page 110 and “Creating a Pivot Table” on page
123 for complete information about pivot tables and how to create them.

CREATING A REVERSE PIVOT TABLE

Please see “Reverse Pivoting” on page 117 and “Creating a Reverse
Pivot Table” on page 133 for complete information about reverse pivot
tables and how to create them.

96 MDM Import Manager Reference Guide

PART 5: TABLE JOINS AND LOOKUPS

This part of the reference guide explains how to combine information

from multiple source tables into a single table for import by creating join
relationships between the tables, and then looking up fields from one
table into another, so that all of the fields exist in the records of a single
table.

97
Overview
The MDM Import Manager imports records from a single table at a time
into a single table at a time.
However, when the source data is relational, and the data to be
imported exists as fields in multiple source tables, you must first create
join relationships between the tables, and then look up fields from one
table into another, so that all of the fields exist in the records of a single
table. In effect, table joins and lookups allow you to combine information
from multiple source tables into a single virtual table for import.
NOTE ►► When you have a single-table data source or an XML data
source, no table joins or lookups are necessary, and this step can be
skipped entirely. Table joins and lookups are necessary only with multi-
table data sources and only when the data to be imported exists as
fields in multiple tables.

The benefit of using table joins and lookups to create a virtual “lookup-
extended” source table consisting of fields from more than one table is
that the source data does not need to be “pre-flattened” (i.e. normalized)
into a single table within an external application prior to import, nor do
additional tables need to be created in the original data source. Instead,
the flattening can be done “on-the-fly” within the virtual workspace
created by the MDM Import Manager, as part of the import process
itself.
NOTE ►► The virtual lookup-extended source table contains all of the
records and all of the fields of the primary table, plus each of the
lookup fields from each of the lookup tables.

Each of the operations that support table joins and lookups is described
in the following sections, and summarized in Table 11.

Table 11. Table Join and Lookup Operations

Operation Description

Join Joins the selected field to a field in another table.

Unjoin Unjoins the selected field.

Lookup Looks up a field from another joined table.

Unlookup Removes the selected lookup field from the hierarchy.

MDM Import Manager Reference Guide 99

MULTIPLE TABLES OF SOURCE DATA
By way of example, consider the Products, Manufacturers, and
Countries tables shown in Figure 33, three tables that contain related
information on products.
Each product record in the Products table contains a MFG code but no
manufacturer name, which is contained in the Manufacturers table for
each manufacturer record.
Similarly, each manufacturer record in the Manufacturers table contains
a CNTRY code but no country name, which is contained in the
Countries table for each country record.
Finally, the Products table contains no information whatsoever on the
country of origin.
Products:
NAME … MFG Countries:
Belt … 2011 Manufacturers: CODE NAME
End Brush … 1212 CODE NAME CNTRY E789 ENG
Flap Disc … 2011 1212 Weiler U123 F456 FRA
Grinder … 2011 2011 Radnor E789 U123 USA
Grinder … 2011 3478 CGW U123
Nut-Driver … 5124 5124 Morgan F456
Ratchet … 3478

Figure 33. Multiple tables containing related information

If you want to import records from the Products table along with the
manufacturer name and country name for each product record, then you
must combine the information from the three tables using table joins and
lookups, as described in the following sections and as illustrated in
Figure 34.
Products: Manufacturers: Countries:
NAME … MFG NAME NAME
Belt … 2011 Radnor ENG
End Brush … 1212 Weiler USA
Flap Disc … 2011 Radnor ENG
Grinder … 2011 Radnor ENG
Grinder … 2011 Radnor ENG
Nut-Driver … 5124 Morgan FRA
Ratchet … 3478 CGW USA

Figure 34. Information combined from multiple tables

100 MDM Import Manager Reference Guide

 TABLE JOINS AND LOOKUPS
In simple terms, a join combines information from the records in two
tables by using the values in a pair of matching fields to match records
in the two tables.
A lookup is a type of join that uses the value of the primary matching
field for each record in a primary table to match the corresponding
record with the same value for the lookup matching field in a lookup
table, and then “looks up” additional information from one or more
lookup fields in the lookup table record into the primary table record.
These table join and lookup terms – and how each type of table and
field associated with joins and lookups is reflected in the Source
Hierarchy tree – are shown in Figure 35 and summarized in Table 12.

Primary
matching
fields
Lookup
matching
fields

Lookup table
Lookup fields
Transitive
lookup table

Figure 35. Table joins and lookups in the Source Hierarchy tree

MDM Import Manager Reference Guide 101

Table 12. Table Join and Lookup Terms and the Source Hierarchy Tree
Term Description / Display in Source Hierarchy Tree
Join Combines information from records in multiple tables.
Matching Fields The pair of fields whose values are used to match records.
Lookup Uses a join to look up information from one table into another.
Primary Table The table into which fields are looked up.
The table from which fields are looked up.
Lookup Table
 Added as the last child of the primary table
The matching field in the primary table.

Primary Matching Field  [Join] appended to the name of the primary matching field.
 Also added as the first child of the lookup matching field, as
[From: primary table.primary matching field]
The matching field in the lookup table.

Lookup Matching Field  [Join] appended to the name of the lookup matching field.
 Also added as the first child of the primary matching field, as
[To: lookup table.lookup matching field]
Fields looked up from the lookup table into the primary table.
Lookup Fields
 Added as children of the lookup table node

NOTE ►► Joins are transitive: if Table A is joined to Table B and

Table B is joined to Table C, Table C is effectively joined to Table A.
The MDM Import Manager automatically detects this condition,
displays “transitive join” Table C as the child not only of Table B but
also of the joined lookup Table B node that is a child of Table A, and
allows “transitive lookups” from Table C into Table A.

TECHNICAL NOTE ►► In a typical relational DBMS, joins are usually

supported by explicit primary key and foreign key matching fields within
each of the two tables, and lookups occur as a result of explicit queries
or views. By contrast, MDM automatically creates implicit primary key
and foreign key matching fields when you create a field in the
repository of type Lookup, and the value of the lookup field is the
value(s) of the display field(s) of the lookup table record.

TECHNICAL NOTE ►► In a typical DBMS application, most joins are

equi-joins. An equi-join is symmetrical, in that both tables are equal
with no “from” or “to” table; combining information from the two tables
creates a “result” table that includes the specified fields from either of
the tables, but might not include all of the records of either table. By
contrast, the MDM Import Manager uses left-outer joins to creates the
lookup-extended table. A left-outer join treats the “from” table as the
primary table and the “to” table as the lookup table, and the result table
includes all of the primary table fields and each of the lookup fields, but
most importantly, includes all of the records of the primary table.

102 MDM Import Manager Reference Guide

In the MDM Import Manager, you can use the Join command (or drag-
and-drop) to join a primary source table to one or more lookup source
tables, and then use the Lookup command (or drag-and-drop) to look up
fields from the lookup tables into the primary table, as described in the
following sections.

JOINING TABLES
For data sources that do not store relationships between tables or do
not make them available through a programmatic interface, or
alternatively, to create additional relationships that are not explicitly
defined in the data source, the MDM Import Manager allows you to
manually define join relationships from a primary table to a lookup table.
NOTE ►► For data sources that store “primary key / foreign key”
and/or “main table / related table” relationships and that make this
information available through a programmatic interface, the MDM
Import Manager automatically detects these join relationships when
you connect to the data source and reflects them in the Source
Hierarchy tree. For example, Microsoft Access stores them and makes
them available through the ADO interface used by the MDM Import
Manager to connect to Access.

When you use the Join command to manually join a primary table to a
lookup table, as shown in Figure 36, MDM does the following:
• Creates a link between the primary matching field in the primary
table and the lookup matching field in the lookup table
• Appends [Join] to the name of each matching field in the tree.
• Adds the corresponding matching field of the other table as the first
child of each matching field.
• Adds the lookup table as the last child of the primary table.

Figure 36. Source Hierarchy tree before and after Join

NOTE ►► The Join command joins two tables in the Source

Hierarchy tree and in the MDM Import Manager’s virtual workspace,
but does not change the structure of the underlying data source.

MDM Import Manager Reference Guide 103

 To manually join a primary table to a lookup table:
1. In the Source Hierarchy tree, select the primary matching field node in
the primary table from which you want to establish the join relationship.
NOTE ►► The primary table does not need to be the current source
table.

2. Right-click on the node and choose Join from the context menu, or
choose Source > Join from the main menu, and select first the lookup
table and then the lookup matching field from the two-level cascading
menu of source tables and fields for each table.
NOTE ►► You can only join fields with the same data type; other
fields are disabled in the cascading menu of fields.

NOTE ►► You can also drag-and-drop the primary matching field

from the primary table onto the lookup matching field in the lookup
table to create the join relationship.

3. MDM creates the new join relationship and reflects it in the Source
Hierarchy tree.
NOTE ►► See “The Source Hierarchy Tree” on page 81 for more
information about the nodes of the Source Hierarchy tree.

 To remove a join relationship between two tables:

1. In the Source Hierarchy tree, select either the primary matching field
node in the lookup table (i.e. the [To: …] node) or the lookup matching
field in the primary table (i.e. the [From: …] node) that corresponds to
the join relationship you want to remove.
2. Right-click on the node and choose Unjoin from the context menu, or
choose Source > Unjoin from the main menu.
3. MDM eliminates the join relationship and removes the matching field
nodes, the lookup table node, and any lookup field nodes from the
Source Hierarchy tree.

LOOKING UP A FIELD
Once the join relationships have been established, you can look up
fields from a lookup table into a primary table, adding the field into the
list of primary table fields, and in effect, building an extended virtual
record consisting of fields from the primary table and lookup fields from
one or more lookup tables.

104 MDM Import Manager Reference Guide

When you use the Lookup command to look up a field from a lookup
table into the primary table, as shown in Figure 37, the MDM Import
Manager adds the lookup field as a child of the lookup table. You can
then perform subsequent processing on the lookup field just as if it were
an original field in the source data.

Figure 37. Source Hierarchy tree before and after Lookup

NOTE ►► The Lookup command adds a lookup field to the Source

Hierarchy tree and to the MDM Import Manager’s virtual workspace,
but does not change the structure of the underlying data source.

NOTE ►► The lookup table can be nested more than one level below
the primary table through the effect of “transitive” joins, as described in
the previous section.

 To look up a field from a lookup table into a primary table:

1. In the Source Hierarchy tree, select the lookup table node in the
primary table into which you want to look up a field.
NOTE ►► The primary table does not need to be the current source
table.

2. Right-click on the node and choose Lookup from the context menu, or
choose Source > Lookup from the main menu, and select the lookup
field from the cascading menu of fields in the lookup table.
TIP ►► You can also drag-and-drop the lookup field from the lookup
table onto the lookup table in the primary table to create the lookup.

TIP ►► You can select [All] from the cascading menu to lookup all of
the fields of the lookup table.

3. MDM creates the new lookup field and adds it to the Source Hierarchy
tree.
NOTE ►► See “The Source Hierarchy Tree” on page 81 for more
information about the nodes of the Source Hierarchy tree.

MDM Import Manager Reference Guide 105

 To remove a lookup field from a primary table:
1. In the Source Hierarchy tree, select the lookup field node you want to
remove in the primary table.
2. Right-click on the node and choose Unlookup from the context menu,
or choose Source > Unlookup from the main menu.
3. MDM eliminates the lookup field and removes its node from the Source
Hierarchy tree.
NOTE ►► When you remove a join relationship, the MDM Import
Manager automatically removes the lookup fields as well.

106 MDM Import Manager Reference Guide

PART 6: PIVOTING/REVERSE PIVOTING

This part of the reference guide explains the pivoting operations of the
MDM Import Manager, and how to use them to “rotate” source data prior
to subsequent restructuring operations.

107
Overview
Pivoting and reverse pivoting are perhaps the most powerful
restructuring mechanisms available within the MDM Import Manager.
They “rotate” source data, creating an entirely new table from the
source table, to completely transform the data in ways that the other
restructuring mechanisms cannot address.
In particular, the pivot and reverse pivot operations behave as follows:
• Pivot. Translates the distinct data values in one or more fields into
field names, increasing the number of fields and reducing the
number of rows in the pivot table.
• Reverse pivot. Translates one or more field names into the data
values within a single field, reducing the number of fields and
increasing the number of rows in the reverse pivot table.

Figure 38 illustrates the symmetry of the pivoting and reverse pivoting

operations, each of which reverses the action of the other.

Field Field Field Field X Y Z

1 X a 1 a b c
1 Y b 2 d e f
1 Z c
Pivot ►
2 X d
◄ Reverse Pivot
2 Y e
2 Z f

Figure 38. Pivot vs. reverse pivot

Pivoting and reverse pivoting are fundamentally different from – and if

they are necessary, must precede – the splitting and partitioning
operations described in subsequent parts of this guide, since they
create an entirely new table from the source table on which all of those
other restructuring operations should be performed.
NOTE ►► Splitting expands a single field of values into multiple
fields, each containing a single value extracted from the original field.
Conversely, partitioning combines multiple fields and the values within
them into field and value combinations. In both cases, the operations
are within and between fields of the original source table, leaving the
set of original source records untouched.

TIP ►► When pivoting or reverse pivoting is necessary, the basic

sequence of steps is to: (1) select the source table you want to
restructure; (2) perform the pivot or reverse pivot; and (3) change the
current source table to the newly created pivot or reverse pivot table.

MDM Import Manager Reference Guide 109

PIVOTING
Pivoting creates a new table by translating the distinct data values in
one or more source fields into field names, increasing the number of
fields and reducing the number of rows in the pivot table.
Specifically, pivoting: (1) treats as metadata the values within one or
more source fields and uses them as the name of new fields in the pivot
table; (2) rearranges the corresponding data values from one or more
other source fields within the new fields, and (3) collapses duplicates of
each product record in the source table into a single record.
Figure 39 illustrates a simple pivot operation on a single pair of fields.

Original table: Pivot table:

Key Metadata Data Key X Y Z
1 X a 1 a b c
1 Y b 2 d e f
1 Z c
Pivot ►
2 X d
2 Y e
2 Z f

Figure 39. Pivot translates metadata values into field names

NOTE ►► A pivot in the MDM Import Manager is like adding one or

more horizontal pivots in MDM Publisher, which takes the distinct data
values from the pivot columns and uses them to label the base tables
which have been arranged side-by-side.

Pivoting is very flexible. You can pivot by a single pair of source fields,
one containing the metadata values and the other the data values. You
can pivot by multiple metadata field combinations and/or multiple data
field combinations. Finally, you can pivot simultaneously by multiple
pairs of corresponding fields of metadata and data.
In each case, you must identify the source fields that participate in the
pivot, which ones contain metadata and which ones contain data, which
ones must be combined, the one-to-one correspondence between fields
and/or field combinations, and the key field that the MDM Import
Manager should use to identify duplicate records.
Various pivoting challenges are illustrated in the following sections.
NOTE ►► Pivoting is often required when field and attribute data is
stacked “vertically” in the source table as name/value pairs and needs
to be rearranged into a “horizontal” table structure from which it can be
imported into the fields and attributes of a regular table, as described in
subsequent parts of this guide.

110 MDM Import Manager Reference Guide

Challenge: Single Pair of Metadata/Data Fields
Most frequently, the fields by which to pivot consist of an individual field
of metadata and an individual field of data, which together comprise a
single pair of corresponding source fields.
Consider the table shown in Figure 40, which contains two fields, QTY
and PRICE, that are used to store quantity pricing information.

PART NO QTY PRICE

101 50 $70
101 250 $60
101 500 $50
102 50 $80
102 250 $70
102 500 $60
103 50 $90
103 250 $75
103 500 $60

Figure 40. Single pair of corresponding metadata and data fields

NOTE ►► PART NO is used as the key field to identify the product

records that have been duplicated for each quantity.

In the figure above, the set of distinct values of QTY are the metadata
values that represent the field names for each corresponding PRICE
data value. Figure 41 illustrates the table that results from pivoting by
QTY against PRICE.

PART NO 50 250 500

101 $70 $60 $50
102 $80 $70 $60
103 $90 $75 $60

Figure 41. Pivot table

MDM Import Manager Reference Guide 111

Challenge: Metadata Field Combination
Sometimes, the fields by which to pivot consist of multiple metadata
fields that correspond to a single data field in the source table.
Consider the table shown in Figure 42, which contains three fields,
DIST, QTY, and PRICE, that are used to store distributor-specific
quantity pricing information.

PART NO DIST QTY PRICE

101 McMaster 50 $70
101 McMaster 250 $60
101 McMaster 500 $50
101 Grainger 50 $68
101 Grainger 250 $58
101 Grainger 500 $48
102 McMaster 50 $80
102 McMaster 250 $70
102 McMaster 500 $60
102 Grainger 50 $78
102 Grainger 250 $68
102 Grainger 500 $58
103 McMaster 50 $90
103 McMaster 250 $75
103 McMaster 500 $60
103 Grainger 50 $88
103 Grainger 250 $73
103 Grainger 500 $58

Figure 42. Metadata field combination

In the figure above, the set of distinct value combinations of DIST and
QTY are the metadata values that represent the field names for each
corresponding PRICE data value. Figure 43 illustrates the table that
results from pivoting by DIST and QTY against PRICE.

PART NO M/50 M/250 M/500 G/50 G/250 G/500

101 $70 $60 $50 $68 $58 $48
102 $80 $70 $60 $78 $68 $58
103 $90 $75 $60 $88 $73 $58

Figure 43. Pivot table

NOTE ►► Each field name in the pivot table is a metadata value

combination (abbreviated in the table above for space considerations).

112 MDM Import Manager Reference Guide

Challenge: Data Field Combination
Alternatively, the fields by which to pivot consist of a single metadata
field in the source table that corresponds to multiple data fields, each of
which contains a different type of data.
Consider the table shown in Figure 44, which contains three fields,
DIST, PART NO, and PRICE, that are used to store distributor-specific
part numbers and distributor-specific pricing information (two pieces of
data for each distributor).

SKU DIST PART NO PRICE

101 McMaster M101 $70
101 Grainger G101 $68
101 Applied A101 $50
102 McMaster M102 $80
102 Grainger G102 $78
102 Applied A102 $80
103 McMaster M103 $90
103 Grainger G103 $88
103 Applied A103 $73

Figure 44. Data field combination

NOTE ►► In this case, SKU is used as the key field to identify the
product records that have been duplicated for each distributor.

In the figure above, the set of distinct values of DIST are the metadata
values that represent the field names for each corresponding PART NO
and PRICE data value. Figure 45 illustrates the table that results from
pivoting by DIST against both PART NO and PRICE.

SKU P.NO: M P.NO: G P.NO: A PRICE: M PRICE: G PRICE: A

101 M101 G101 A101 $70 $68 $50
102 M102 G102 A102 $80 $78 $80
103 M103 G103 A103 $90 $88 $73

Figure 45. Pivot table

NOTE ►► When multiple data fields correspond to the metadata field

(so that multiple data values correspond to each metadata value), the
pivot table contains a complete set of columns for each data field.

MDM Import Manager Reference Guide 113

Challenge: Single Name/Value Field Pair
A very common real-world challenge is when the pair of fields by which
to pivot represent both field and attribute name/value pairs, where each
attribute applies to just some of the product records, and each product
record is duplicated for each name/value pair.
NOTE ►► This name/value structure for attribute information is a
common alternative to very wide (and sparse) source records with a
large number of fields, most of which would contain NULL values.

Consider the table shown in Figure 46, which contains two fields, NAME
and VALUE, that are used to store field and attribute information for
each source record.

PART NO NAME VALUE

101 Width 4 mm
101 Diameter 2 mm
101 Weight 8 oz
101 Material Steel
102 HP 4 hp
102 RPM 600 rpm
102 Weight 5 lb
103 Width 8 mm
103 Diameter 4 mm
103 Weight 1 lb
103 Material Copper

Figure 46. Single pair of name/value fields

In the figure above, the set of distinct values of NAME represents the
field and attribute names for each corresponding data value. Figure 47
illustrates the table that results from pivoting by NAME against VALUE.

PART NO Width Diameter Weight Material HP RPM

101 4 mm 2 mm 8 oz Steel
102 5 lb 4 hp 600 rpm
103 8 mm 4 mm 1 lb Copper

Figure 47. Pivot table

NOTE ►► The new table is now structured so that each field and
attribute can be individually split, partitioned, and/or mapped as
described in subsequent parts of this guide.

114 MDM Import Manager Reference Guide

 NOTE ►► Attribute data that is stacked vertically as name/value pairs
can also be imported directly without pivots by first partitioning the
name and value source fields, and then mapping the resulting
name/value hierarchy to a special destination field containing a
name/value hierarchy of attributes and attribute data values (see
“Attribute Data Import as Name/Value Pairs” on page 167 for more
information).

Challenge: Multiple Name/Value Field Pairs

Another real-world challenge is when each source record contains
multiple name/value pairs of field and attribute information.
NOTE ►► Multiple name/value field pairs for each source record is a
common alternative to duplicating the product record for each
name/value pair.

Consider the table shown in Figure 48, which contains five pairs of
name/value fields and just a single source record for each product
record.
PART NO NAME1 VAL1 NAME 2 VAL2 NAME 3 VAL3 NAME 4 VAL4 NAME 5 VAL5

101 Width 4 mm Height 6 mm Diameter 2 mm Weight 8 oz Material Steel

102 HP 4 hp RPM 600 rpm Weight 5 lb

103 Height 12 mm Width 8 mm Weight 1 lb Material Copper Diameter 4 mm

Figure 48. Multiple pairs of name/value fields

Figure 49 illustrates the table that results from pivoting by each of the
five NAME fields against each of the five VAL fields.

PART NO Width HP Height RPM Diameter Weight Material

101 4 mm 6 mm 2 mm 8 oz Steel
102 4 hp 600 rpm 5 lb
103 8 mm 12 mm 4 mm 1 lb Copper

Figure 49. Pivot table

NOTE ►► In this example, the pivot does not reduce the number of
rows in the table because the product records had not been duplicated
in the source table.

MDM Import Manager Reference Guide 115

Challenge: Multi-Valued Information
A variation on name/value pairs is when the source data represents
multi-valued field and attribute information. Each product record is once
again duplicated, not only for each name/value pair, but also for each
instance of a field or attribute name, which can occur once for each
value of the multi-valued field.
Consider the table shown in Figure 50, in which the Color attribute
appears multiple times for each product record.

PART NO NAME VALUE

101 Width 4 mm
101 Diameter 2 mm
101 Weight 8 oz
101 Color Red
101 Color Green
101 Color Blue
102 HP 4 hp
102 RPM 600 rpm
102 Weight 5 lb
102 Color Black
102 Color White
103 Width 8 mm
103 Diameter 4 mm
103 Weight 1 lb
103 Color White

Figure 50. Name/value fields representing multi-valued information

Figure 51 illustrates the table that results from pivoting by NAME

against VALUE.

PART NO Width Diameter Weight HP RPM Color

101 4 mm 2 mm 8 oz Red; Green; Blue
102 5 lb 4 hp 600 rpm Black; White
103 8 mm 4 mm 1 lb White

Figure 51. Pivot table with multi-valued field

NOTE ►► All of the multiple Color values for each product record are
automatically combined into a single field in the pivot table: (1) in the
order the values appear in the source table; and (2) using the Multi-
Valued Delimiter to separate individual values.

116 MDM Import Manager Reference Guide

Challenge: No Key Fields
Sometimes, you may need to base a pivot on all of the fields in a table,
such as when a table has only two fields. In these cases, there are no
fields available to use as key fields.
Consider the table shown in Figure 52, which contains two fields, MAKE
and TRANSMISSION.

MAKE TRANSMISSION
Ford Automatic
VW Automatic
VW Manual
Toyota Automatic
Toyota Manual
Porsche Manual

Figure 52. A two-field table

In the figure above, the set of distinct values of the MAKE field are the
metadata values that represent the field names for each corresponding
TRANSMISSION value. Prior to MDM 7.1, however, the lack of a key
field meant you could not pivot on this table. With MDM 7.1, Import
Manager includes a [Natural Order] key field option, which uses the
order in which data appears in the source file as the table’s implied key
values. Import Manager then creates an Implicit Key field on the pivot
table to serve as the table’s key field.
Figure 51 illustrates the table that results from pivoting by MAKE against
TRANSMISSION, using the [Natural Order] option as the key field.

Implicit Key Ford VW Toyota Porsche

1 Automatic Automatic; Manual Automatic; Manual Manual

Figure 53. Pivot table with Implciit Key field added

NOTE ►► The [Natural Order] option is especially useful for creating

pivot tables from XML source files, as in MDM 7.1, Import Manager no
longer adds _ID join fields to XML source hierarchies.

MDM Import Manager Reference Guide 117

REVERSE PIVOTING
Reverse pivoting creates a new table by translating one or more source
field names into data values within a single field, reducing the number of
fields and increasing the number of rows in the reverse pivot table.
Specifically, reverse pivoting: (1) places the names of one or more
source fields into a new field of metadata values in the pivot table; (2)
eliminates those fields and rearranges the data values within them into
another new field of corresponding data values; and (3) duplicates each
product record in the source table for each field that is eliminated.
Figure 54 illustrates a simple reverse pivot operation.

Original table: Reverse pivot table:

Key X Y Z Key Metadata Data
1 a b c 1 X a
2 d e f 1 Y b
1 Z c
Pivot
2 X d
Reverse Pivot ►
2 Y e
2 Z f

Figure 54. Reverse pivot translates field names into metadata values

NOTE ►► A reverse pivot in the MDM Import Manager is like

removing one more horizontal pivots in MDM Publisher.

Like pivoting, reverse pivoting is very flexible. You can reverse pivot by
a single set of source fields, and you can reverse pivot by multiple sets
of source fields, where each set contains a different type of data values.
In each case, you must identify the source fields that participate in the
reverse pivot and the source fields in each set.
Various reverse pivoting challenges are illustrated in the following
sections.
NOTE ►► Reverse pivoting is often required when qualified data is
arranged “horizontally” in the source table and needs to be rearranged
into a “vertical” table structure from which it can be imported into the
fields and qualifiers of a qualified table, as described in subsequent
parts of this guide.

118 MDM Import Manager Reference Guide

Challenge: Single Set of Data Fields
Most frequently, the fields by which to reverse pivot consist of a single
set of data fields, all of which contain the same type of data, and each
field corresponds to a single metadata value.
Consider the table shown in Figure 55, which contains three fields that
are used to store quantity pricing information, where each field
corresponds to a single quantity value.

PART NO Q50 Q250 Q500

101 $70 $60 $50
102 $80 $70 $60
103 $90 $75 $60

Figure 55. Single set of data fields

Figure 56 illustrates the table that results from reverse pivoting by each
of the three quantity pricing fields.

PART NO Names Values

101 Q50 $70
101 Q250 $60
101 Q500 $50
102 Q50 $80
102 Q250 $70
102 Q500 $60
103 Q50 $90
103 Q250 $75
103 Q500 $60

Figure 56. Reverse pivot table

TIP ►► During field and value mapping in a later step of the import
process, the Names field would most likely be mapped to the Quantity
display field of the qualified lookup table (and each Names value
converted or mapped to a quantity value), and the Values field would
most likely be mapped to the Prices qualifier.

MDM Import Manager Reference Guide 119

Challenge: Metadata Field Combinations
Sometimes, each field in the single set of data fields by which to reverse
pivot represents a combination of metadata values, so that when you
perform the reverse pivot, each field value corresponds to a metadata
value combination.
Consider the table shown in Figure 57, which contains four fields that
are used to store distributor-specific quantity pricing information, where
each field corresponds to a single distributor/quantity value combination.

PART NO M100 M500 G100 G500

101 $70 $50 $68 $48
102 $80 $60 $78 $58
103 $90 $60 $88 $58

Figure 57. Each data field corresponds to a metadata value combination

Figure 58 illustrates the table that results from reverse pivoting by each
of the four distributor/quantity pricing fields.

PART NO Names Values

101 M100 $70
101 M500 $50
101 G100 $68
101 G500 $48
102 M100 $80
102 M500 $60
102 G100 $78
102 G500 $58
103 M100 $90
103 M500 $60
103 G100 $88
103 G500 $58

Figure 58. Reverse pivot table

TIP ►► When the Names column represents destination value

combinations, the display fields and/or qualifiers of the destination
qualified lookup table would most likely need to be partitioned prior to
field mapping.

120 MDM Import Manager Reference Guide

Challenge: Multiple Sets of Data Fields
The most complex challenge occurs when the fields by which to reverse
pivot consist of multiple sets of data fields, where the fields within each
set contain the same type of data, each set contains a different type of
data, and each field corresponds to a single metadata value.
Consider the table shown in Figure 59, which contains six fields of
distributor-specific information and two types of data. Three of the fields
contain distributor-specific part numbers and three of them contain
distributor-specific prices.

SKU M P.NO M PRICE G P.NO G PRICE A P.NO A PRICE

101 M101 $70 G101 $68 A101 $50
102 M102 $80 G102 $78 A102 $80
103 M103 $90 G103 $88 A103 $73

Figure 59. Multiple sets of data fields

Figure 60 illustrates the table that results from reverse pivoting by the
three pairs of distributor-specific fields.

SKU Names Values (1) Values (2)

101 M P.NO; M PRICE M101 $70
101 G P.NO; G PRICE G101 $68
101 A P.NO; A PRICE A101 $50
102 M P.NO; M PRICE M102 $80
102 G P.NO; G PRICE G102 $78
102 A P.NO; A PRICE A102 $80
103 M P.NO; M PRICE M103 $90
103 G P.NO; G PRICE G103 $88
103 A P.NO; A PRICE A103 $73

Figure 60. Reverse pivot table

TIP ►► When there are multiple sets of source fields, the relationship
between the fields of source data – and the proper reverse pivot – may
not be immediately obvious. In these cases, it can help to imagine the
source table as having been horizontally pivoted, and each subset of
fields as representing a base table (see “Imagining Horizontal Pivots”
on page 138 for more information).

MDM Import Manager Reference Guide 121

Pivoting and Reverse Pivoting Operations
The following sections describe the pivoting and reverse pivoting
operations you can perform on a source table, including:
• Pivoting by a single pair of source fields.
• Pivoting by multiple metadata fields.
• Pivoting by multiple data fields.
• Pivoting by multiple pairs of source fields.
• Reverse pivoting by a single set of source fields.
• Reverse pivoting by multiple sets of source fields.

The operations are summarized in Table 13.

Table 13. Pivoting and Reverse Pivoting Operations

Operation Description
Create
Creates a new table by pivoting on the selected fields.
Pivot Table
Create Creates a new table by reverse pivoting on the
Reverse Pivot Table selected fields.

NOTE ►► Pivoting and reverse pivoting are fundamentally different

from – and if they are necessary, must precede – the splitting and
partitioning operations described in subsequent parts of this guide,
since they create an entirely new table from the source table on which
all of those other restructuring operations should be performed.

TIP ►► When pivoting or reverse pivoting is necessary, the basic

sequence of steps is to: (1) select the source table you want to
restructure; (2) perform the pivot or reverse pivot; (3) change the
current source table to the newly created pivot or reverse pivot table;
and (4) proceed with the subsequent steps in the import process.

NOTE ►► If the proper pivot or reverse pivot is not immediately

obvious, it can help to imagine the tables as having been horizontally
pivoted (see “Imagining Horizontal Pivots” on page 138 for more
information).

NOTE ►► In prior versions of Import Manager, pivoting and reverse

pivoting operations were permitted only when the data source was
updateable. This limitation is removed as of MDM 7.1.

122 MDM Import Manager Reference Guide

CREATING A PIVOT TABLE
To create a pivot table, you must identify the source fields that
participate in the pivot, which ones contain metadata and which ones
contain data, which ones must be combined, the one-to-one
correspondence between metadata and data fields and/or field
combinations, and the key field or fields.
NOTE ►► You use the same generic procedure to create a pivot
table regardless of the structure of the source table and the
restructuring that must be performed. Specific pivot examples are then
illustrated in the subsections that follow.

 To create a pivot table:

1. In the drop-down list of source tables, make sure the table you want to
pivot is the current source table.
2. In the Source Hierarchy tree, select all the field nodes corresponding to
both the metadata fields and the data fields by which you want to pivot.
3. Right-click on one of the nodes and choose Create Pivot Table from the
context menu, or choose Source > Create Pivot Table from the main
menu, as shown in Figure 61.

Figure 61. Create Pivot Table context menu command

4. MDM opens the Create Pivot Table dialog shown in Figure 62.

MDM Import Manager Reference Guide 123

 Figure 62. Create Pivot Table dialog

5. In the Key Fields dual-list drop-down control, move one or more fields
from the Available Fields list to the Selected Fields list to identify the
key fields on which to perform the pivot, as shown in Figure 63.

Figure 63. Key Fields dual-list drop-down control

NOTE ►► The Key Fields drop-down control contains the source

fields that were not selected as fields by which to pivot.

NOTE ►► For XML, XML Schema, and text sources, the Key Fields
drop-down control includes a [Natural Order] option. This option
cannot be combined with other fields.

6. In the dual-list control of fields, drag-and-drop the data fields from the
Field Values Become Field Names list to the Field Values Become
Field Values list, as shown in Figure 64.

124 MDM Import Manager Reference Guide

 Figure 64. Drag-and-drop between field lists

NOTE ►► The Field Values Become Field Names list initially contains
the entire set of source fields that were selected in the Source
Hierarchy tree as fields by which to pivot.

7. If necessary, select two or more fields in either list that must be

combined into a field combination, and click on the Combine button, or
right-click on one of the fields and choose Combine from the context
menu, as shown in Figure 65.

Figure 65. Combine context menu command

8. If necessary, drag-and-drop fields or field combinations within each list

to create the one-to-one correspondence between metadata fields and
data fields.
NOTE ►► The two lists must contain the same number of distinct
items before you can proceed.
9. Click on the Preview button to display a preview of the first ten records
of the pivot table, as shown in Figure 66.

MDM Import Manager Reference Guide 125

 Figure 66. Pivot preview

NOTE ►► You must perform the Preview before you can proceed.

10. When you have verified that the pivot operation you have defined will
have the desired effect, click OK to close the Create Pivot Table dialog.
11. In the drop-down list of source tables, select the newly created pivot
table on which to perform subsequent import processing.
NOTE ►► Pivoting creates additional fields out of field values. MDM
automatically converts any characters in the field values that are
invalid for field names to valid ones, as follows: period (.) 
underscore (_); exclamation (!)  vertical bar (|); square brackets ([])
 angular brackets (<>); accent grave (`)  single quote ('); and
double quote (")  single quote ('). In addition, MDM removes
preceding and trailing spaces.

NOTE ►► Various pivoting examples are illustrated in the following

subsections.

126 MDM Import Manager Reference Guide

Example: Single Pair of Metadata/Data Fields
Figure 67 shows a source table with a single pair of fields by which to
pivot.

Figure 67. Single pair of fields by which to pivot

Figure 68 shows the corresponding field settings and preview in the

Create Pivot Table dialog.

Figure 68. Field settings in the Create Pivot Table dialog

MDM Import Manager Reference Guide 127

Example: Metadata Field Combination
Figure 69 shows a source table with a combination of metadata fields by
which to pivot.

Figure 69. Metadata field combination

Figure 70 shows the corresponding field settings and preview in the

Create Pivot Table dialog.

Figure 70. Field settings in the Create Pivot Table dialog

NOTE ►► DIST and QTY have been combined in the Field Values
Become Field Names list.

128 MDM Import Manager Reference Guide

Example: Data Field Combination
Figure 71 shows a source table with a combination of data fields by
which to pivot.

Figure 71. Data field combination

Figure 72 shows the corresponding field settings and preview in the

Create Pivot Table dialog.

Figure 72. Field settings in the Create Pivot Table dialog

NOTE ►► PART NO and PRICE have been combined in the Field

Values Become Field Values list.

MDM Import Manager Reference Guide 129

Example: Single Name/Value Field Pair
Figure 73 shows a source table with a single pair of name/value fields
by which to pivot.

Figure 73. Single pair of name/value fields

Figure 74 shows the corresponding field settings and preview in the

Create Pivot Table dialog.

Figure 74. Field settings in the Create Pivot Table dialog

NOTE ►► If multiple instances of the same Name appear for each

product record (which can represent multi-valued information), all of
the values are automatically combined into a single field in the pivot
table: (1) in the order the values appear in the source table; and (2)
using the Multi-Valued Delimiter to separate individual values.

130 MDM Import Manager Reference Guide

Example: Multiple Name/Value Field Pairs
Figure 75 shows a source table with multiple pairs of name/value fields
by which to pivot.

Figure 75. Multiple pairs of name/value fields

Figure 76 shows the corresponding field settings and preview in the

Create Pivot Table dialog.

Figure 76. Field settings in the Create Pivot Table dialog

NOTE ►► Each of the NAME fields have been included in the Field
Values Become Field Names list and each of the VALUE fields have
been included – and arranged in the same order – in the Field Values
Become Field Values list.

MDM Import Manager Reference Guide 131

Example: [Natural Order] Pivots
Figure 75 shows a source table with only two fields.

Figure 77. Multiple pairs of name/value fields

Since both fields are needed for the pivot, [Natural Order] must be
selected as the key value. This results in a new field, Implicit Key,
being added to the table.
Figure 76 shows the corresponding field settings and preview in the
Create Pivot Table dialog.

Figure 78. Field settings in the Create Pivot Table dialog

NOTE ►► The [Natural Order] field is new for MDM 7.1; in previous
versions you could not pivot a two-field table.

132 MDM Import Manager Reference Guide

CREATING A REVERSE PIVOT TABLE
To create a reverse pivot table, you must identify the source fields that
participate in the reverse pivot and the source fields in each set.
NOTE ►► You use the same generic procedure to create a reverse
pivot table regardless of the structure of the source table and the
restructuring that must be performed. Specific reverse pivot examples
are then illustrated in the subsections that follow.

 To create a reverse pivot table:

1. In the drop-down list of source tables, make sure the table you want to
reverse pivot is the current source table.
2. In the Source Hierarchy tree, select all the field nodes corresponding to
the data fields by which you want to reverse pivot.
3. Right-click on one of the nodes and choose Create Reverse Pivot Table
from the context menu, or choose Source > Create Reverse Pivot Table
from the main menu, as shown in Figure 79.

Figure 79. Create Reverse Pivot Table context menu command

MDM Import Manager Reference Guide 133

 4. MDM opens the Create Reverse Pivot Table dialog shown in Figure
80.

Figure 80. Create Reverse Pivot Table dialog

NOTE ►► The Values Fields list contains the entire set of source
fields that were selected in the Source Hierarchy tree as fields by
which to reverse pivot.

5. If the fields by which to reverse pivot include multiple sets of data

fields, combine one field from each set, so that there is one field or
field combination in the list for each metadata value, as shown in
Figure 81.

Figure 81. Combine context menu command

6. Click on the Preview button to display a preview of the first ten records
of the reverse pivot table, as shown in Figure 82.

134 MDM Import Manager Reference Guide

 Figure 82. Reverse pivot preview

NOTE ►► You must perform the Preview before you can proceed.

7. When you have verified that the reverse pivot operation you have
defined will have the desired effect, click OK to close the Create
Reverse Pivot Table dialog.
8. In the drop-down list of source tables, select the newly created reverse
pivot table on which to perform subsequent import processing.

NOTE ►► Various reverse pivoting examples are illustrated in the

following subsections.

Example: Single Set of Data Fields

Figure 83 shows a source table with a single set of data fields by which
to reverse pivot.

Figure 83. Single set of fields by which to reverse pivot

MDM Import Manager Reference Guide 135

Figure 84 shows the corresponding field settings and preview in the
Create Reverse Pivot Table dialog.

Figure 84. Field settings in the Create Reverse Pivot Table dialog

Example: Metadata Field Combinations

Figure 85 shows a source table where each field by which to reverse
pivot represents a combination of metadata values.

Figure 85. Each field represents a combination of metadata values

Figure 86 shows the corresponding field settings and preview in the

Create Reverse Pivot Table dialog.

136 MDM Import Manager Reference Guide

 Figure 86. Field settings in the Create Reverse Pivot Table dialog

Example: Multiple Sets of Data Fields

Figure 87 shows a source table with multiple sets of data fields by which
to reverse pivot.

Figure 87. Multiple sets of fields by which to reverse pivot

Figure 88 shows the corresponding field settings and preview in the

Create Reverse Pivot Table dialog.

MDM Import Manager Reference Guide 137

 Figure 88. Field settings in the Create Reverse Pivot Table dialog

NOTE ►► One field from each set has been combined in the Value
Fields list so that there is one field or field combination for each
metadata value.

IMAGINING HORIZONTAL PIVOTS

Recall that creating a pivot table in the MDM Import Manager is like
adding horizontal pivots in MDM Publisher, and creating a reverse pivot
table is like removing horizontal pivots, a parallel that can be very useful
to help you determine how to properly arrange the fields on which to
pivot or reverse pivot in the applicable dialog.

Pivoting vs. Adding Horizontal Pivots

Pivots tend to be straightforward, but imagining the horizontal pivot that
corresponds to a pivot table is a good warm-up for imagining reverse
pivots, which require that you imagine the horizontal pivots on the
original source table rather than the pivot table.
Figure 89 repeats the pivot table shown in Figure 41, which was based
on a single pair of metadata/data fields.

PART NO 50 250 500

101 $70 $60 $50
102 $80 $70 $60
103 $90 $75 $60

Figure 89. Pivot table with a single pair of metadata/data fields

138 MDM Import Manager Reference Guide

In Figure 90, the pivot table above has been redrawn as a table layout
with a horizontal pivot by QTY and a vertical pivot by PART NO, leaving
just PRICE beneath the horizontal pivot span.

QTY 50 250 500

PART NO PRICE PRICE PRICE
101 $70 $60 $50
102 $80 $70 $60
103 $90 $75 $60

Figure 90. Alternative table layout with a single horizontal pivot

TIP ►► Pivot table correlation: (1) the vertical pivot becomes the Key
Field; (2) the horizontal pivot is placed into the Fields Become Field
Names list; and (3) the field beneath the horizontal pivot span is placed
into the Fields Become Field Values list.

NOTE ►► The table layout above could be made to look like the pivot
table in Figure 89 by setting the column header format in MDM
Publisher to swap horizontal pivot values with column headers and
hide the column names.

Figure 91 repeats the pivot table shown in Figure 43, which was based
on multiple metadata fields.

PART NO M/50 M/250 M/500 G/50 G/250 G/500

101 $70 $60 $50 $68 $58 $48
102 $80 $70 $60 $78 $68 $58
103 $90 $75 $60 $88 $73 $58

Figure 91. Pivot table with multiple metadata fields

In Figure 92, the pivot table above has been redrawn as a table layout
with two horizontal pivots by DIST and QTY and a vertical pivot by
PART NO, leaving just PRICE beneath the horizontal pivot span.

DIST McMaster Grainger

QTY 50 250 500 50 250 500
PART NO PRICE PRICE PRICE PRICE PRICE PRICE
101 $70 $60 $50 $68 $58 $48
102 $80 $70 $60 $78 $68 $58
103 $90 $75 $60 $88 $73 $58

Figure 92. Alternative table layout with multiple horizontal pivots

MDM Import Manager Reference Guide 139

 TIP ►► Pivot table correlation: (1) the vertical pivot becomes the Key
Field; (2) the horizontal pivots are placed into and combined in the
Fields Become Field Names list; and (3) the field beneath the
horizontal pivot span is placed into the Fields Become Field Values list.

Finally, Figure 93 repeats the pivot table shown in Figure 45, which was
based on multiple data fields.

SKU P.NO: M P.NO: G P.NO: A PRICE: M PRICE: G PRICE: A

101 M101 G101 A101 $70 $68 $50
102 M102 G102 A102 $80 $78 $80
103 M103 G103 A103 $90 $88 $73

Figure 93. Pivot table with multiple data fields

In Figure 94, the pivot table above has been redrawn as a table layout
with a horizontal pivot by DIST and a vertical pivot by SKU, leaving both
PART NO and PRICE beneath the horizontal pivot span.

DIST McMaster Grainger Applied

SKU PART NO PRICE PART NO PRICE PART NO PRICE
101 M101 $70 G101 $68 A101 $50
102 M102 $80 G102 $78 A102 $80
103 M103 $90 G103 $88 A103 $73

Figure 94. Alternative table layout with a single horizontal pivot

TIP ►► Pivot table correlation: (1) the vertical pivot becomes the Key
Field; (2) the horizontal pivot is placed into the Fields Become Field
Names list; and (3) the fields beneath the horizontal pivot span are
placed into and combined in the Fields Become Field Values list.

NOTE ►► A pivot table with multiple data fields corresponds to a

table layout with multiple fields beneath the horizontal pivot span.

Reverse Pivoting vs. Removing Horizontal Pivots

Reverse pivots with a single set of data fields tend to be relatively
straightforward. The challenge occurs when the fields by which to
reverse pivot consist of multiple sets of data fields, as illustrated below.
NOTE ►► Imagining reverse pivots requires that you imagine the
horizontal pivots on the original source table rather than the pivot table.

Figure 95 repeats the original source table shown in Figure 59.

140 MDM Import Manager Reference Guide

 SKU M P.NO M PRICE G P.NO G PRICE A P.NO A PRICE
101 M101 $70 G101 $68 A101 $50
102 M102 $80 G102 $78 A102 $80
103 M103 $90 G103 $88 A103 $73

Figure 95. Original source table with multiple sets of data fields

NOTE ►► The original source table above is effectively the same

table as the pivot table in Figure 93, except that the fields above are
arranged in pairs by distributor rather than in triplets by type of data.

In Figure 96, the original source table above has been redrawn as a
table layout with a horizontal pivot by DIST and a vertical pivot by SKU,
leaving both PART NO and PRICE beneath the horizontal pivot span.

DIST McMaster Grainger Applied

SKU PART NO PRICE PART NO PRICE PART NO PRICE
101 M101 $70 G101 $68 A101 $50
102 M102 $80 G102 $78 A102 $80
103 M103 $90 G103 $88 A103 $73

Figure 96. Alternative table layout with a single horizontal pivot

TIP ►► Reverse pivot table correlation: each pair of PART NO and

PRICE fields beneath the horizontal pivot span are combined in the
Value Fields list for each distributor.

NOTE ►► An original source table with multiple sets of data fields

corresponds to a table layout with multiple fields beneath the horizontal
pivot span.

By pulling out the implicit metadata value for the distributor from each
pair of PART NO and PRICE fields in the original source table, how to
proceed with the reverse pivot has become more obvious.

MDM Import Manager Reference Guide 141

PART 7: SPLITTING FIELDS

This part of the reference guide explains how to extract data values that
are buried within a field so that you can perform subsequent field-level
processing, such as when a field contains a set of delimited values or
one or more data values within an unstructured text description.

143
Overview
Sometimes, a single field may contain more than just a single, isolated
field value within each record. In these cases, the value (or multiple
values) may need to be extracted from the field before you can perform
subsequent field-level processing.
Extraction problems – and the corresponding MDM solution – include:
• Delimited values represent a hierarchy. If a single field contains a
set of delimited values that implicitly represent the full path of a leaf-
node value in a hierarchy, you can use the Split into Hierarchy
command to expand the values into a hierarchy based on the
specified delimiter. For example, “Tools > Power Tools > Drills >
Accessories” represents the Accessories category for power drills in
a four-level hierarchy.
• Delimited values belong in multiple fields. If a single field contains
a set of delimited values that actually belong in distinct fields, you
can use the Split into Multiple Fields command to split the values
into multiple fields based on the specified delimiter. For example,
“Printers; Laser; 1200” represents a product with Category=Printers;
Printer Type=Laser; and DPI=1200 dpi.

Each of the operations for extracting values from within a field is

described in the following sections, and summarized in Table 14.

Table 14. Splitting Operations

Operation Description
Split into Splits the data values within the selected field into a
Hierarchy hierarchy value.
Split into Splits the data values within the selected field into
Multiple Fields multiple fields.

TIP ►► You can use the Set Split Delimiter commands to set the
delimiter strings that the MDM Import Manager uses to split values.

MDM Import Manager Reference Guide 145

Splitting Fields
You can split a field containing multiple delimited values into: (1) a
single field containing the values expanded into a hierarchy; or (2)
multiple fields, each containing a single value, as described in the
following sections.
NOTE ►► Several different mechanisms allow you to interpret the
multiple delimited values stored within a single field, each applicable
depending on the level of consistency of the source data: (1) Clone,
when the data values in a single field are either/or; and (2) Split, when
the data values are positional and delimited.

NOTE ►► A symmetry exists between splitting a single field and

partitioning multiple fields (see “Splitting vs. Partitioning: A
Comparison” on page 194 for more information).

SPLITTING INTO HIERARCHY

A source data field may contain delimited values that implicitly represent
the full path of a leaf-node value in a hierarchy, as shown in Figure 97.

Source: Corresponding hierarchy:

CATEGORY ► CATEGORY [hierarchy]
Printers; Laser └ Printers
Printers; Inkjet; B&W ├ Laser
├ Inkjet
Printers; Inkjet; Color │ ├ B&W
Printers; Dot Matrix; 300 DPI │ └ Color
└ Dot Matrix
Printers; Dot Matrix; 600 DPI
├ 300 DPI
Printers; Dot Matrix; 1200 DPI ├ 600 DPI
└ 1200 DPI

Figure 97. Delimited values and corresponding hierarchy

When you use the Split into Hierarchy command to split a field with
multiple delimited values, the MDM Import Manager creates a single
new field that contains the delimited values expanded into a hierarchy.
You can then perform subsequent processing on the split field just as if
it were an original field in the source data. Split fields in the Source
Hierarchy tree are highlighted in italics, as shown in Figure 98.

146 MDM Import Manager Reference Guide

 Figure 98. Splitting a field into a hierarchy

NOTE ►► The Split into Hierarchy command adds a hierarchy field to

the Source Hierarchy tree and to the MDM Import Manager’s virtual
workspace, but does not change the structure of the underlying data
source.

TIP ►► You can delete a split field using the Delete Field command.

DATA INTEGRITY ►► When expanding into a hierarchy, the MDM

Import Manager automatically synthesizes the delimited values across
the entire set of source records into a single integrated hierarchy.

DATA INTEGRITY ►► You can specify how to handle trailing NULLs

in delimited values based on whether you are importing into a
hierarchy lookup field or into the hierarchy table itself. However,
regardless of the hierarchy NULL-handling setting, internal NULLs are
always preserved as nodes because ignoring them would collapse
hierarchy structure.

 To split a field with multiple delimited values into a hierarchy:

1. In the Source Hierarchy tree, select the field node whose multiple
delimited values you want to expand into a hierarchy.
2. Right-click on the node and choose Split into Hierarchy from the
context menu, or choose Source > Split into Hierarchy from the main
menu.
3. MDM adds a new field node named “field <Split Hierarchy>” as a
sibling of and just after the original node (Figure 98 above).
4. Press F2 to edit the name of the split field.
5. MDM highlights the name of the split field for editing.
6. Type the name you want for the split field and press Enter.

MDM Import Manager Reference Guide 147

 NOTE ►► The Split into Hierarchy command identifies the individual
delimited values within a single field using the current Split Hierarchy
delimiter string, which must be properly set prior to performing the
command (see “Setting the Split Delimiters” on page 150 for more
information).

NOTE ►► Regardless of the special NULL handling that occurs later

in the Map Fields/Values tab, the Split into Hierarchy command
displays all express and implied source NULLs as nodes in the Source
Hierarchy tree. The preserve or ignore NULLs behavior must be
properly set prior to mapping the split field, where it is reflected in the
values that are displayed in the Source Values grid. (See “The
Hierarchy NULL Handling Option” on page 233 for more information
about NULL handling in hierarchies).

SPLITTING INTO MULTIPLE FIELDS

A source data field may contain delimited values that that actually
belong in distinct fields, as shown in Figure 99.

Source: Corresponding multiple fields:

CATEGORY ► Split <1> Split <2> Split <3>
Printers; Laser Printers Laser
Printers; Inkjet; B&W Printers Inkjet B&W
Printers; Inkjet; Color Printers Inkjet Color
Printers; Dot Matrix; 300 DPI Printers Dot Matrix 300 DPI
Printers; Dot Matrix; 600 DPI Printers Dot Matrix 600 DPI
Printers; Dot Matrix; 1200 DPI Printers Dot Matrix 1200 DPI

Figure 99. Delimited values and corresponding multiple fields

When you use the Split into Multiple Fields command to split a field with
multiple delimited values, the MDM Import Manager creates n new
fields, (where ‘n’ is the maximum number of delimited values in a single
th
field across the entire set of source records, and the n field contains
th
the n delimited value in each record). You can then perform
subsequent processing on the split fields just as if they were original
fields in the source data. Split fields in the Source Hierarchy tree are
highlighted in italics, as shown in Figure 100.

148 MDM Import Manager Reference Guide

 Figure 100. Splitting a field into multiple fields

NOTE ►► Sometimes multiple delimited values do not belong in

distinct fields. Rather, they represent multiple values for a single field
(e.g. “110V; 115V; 220V; 230V”), in which case no splitting is
necessary and the delimited field should instead be mapped to an
MDM multi-valued field or attribute (see “Mapping to Multi-Valued
Destination Fields” on page 215 for more information).

TIP ►► You can delete split fields using the Delete Field command.

 To split a field with multiple delimited values into multiple fields:

1. In the Source Hierarchy tree, select the field node whose multiple
delimited values you want to split into multiple fields.
2. Right-click on the node and choose Split into Multiple Fields from the
context menu, or choose Source > Split into Multiple Fields from the
main menu.
3. MDM adds one or more new field nodes named “field <Split Multiple>
<m>” (where ‘m’ ranges from 1 to the number of new fields) as a
sibling of and just after the original node (Figure 100 above).
4. Press F2 to edit the name of each split field.
5. MDM highlights the name of the split field for editing.
6. Type the name you want for the split field and press Enter.
NOTE ►► The Split into Multiple Fields command identifies the
individual delimited values within a single field using the current Split
Multiple Fields delimiter string, which must be properly set prior to
performing the command (see “Setting the Split Delimiters” on page
150 for more information).

MDM Import Manager Reference Guide 149

SETTING THE SPLIT DELIMITERS
The current split delimiters determine how the Split into Hierarchy and
Split into Multiple Fields commands identify the individual delimited
values within a single field, and must be properly set prior to performing
the commands.
The two split delimiters include:
• Split Hierarchy delimiter. The Split into Hierarchy command
identifies the individual delimited values using the current Split
Hierarchy delimiter string.
• Split Multiple Fields delimiter. The Split into Multiple Fields
command the individual delimited values using the current Split
Multiple Fields delimiter string.

Specifically, the commands use as the current delimiter either: (1) the
global default delimiter string, if the field is set to inherit; or (2) the field-
specific delimiter string, if it has been manually set for the field.
You can set the global default delimiter string for all fields that inherit the
default using the Configuration Options command, or alternatively,
override the default and set it on a field-by-field basis using the Set Split
Delimiter commands, as described in this section.
NOTE ►► Inherited global default values appear in the Source Fields
grid in gray.

TIP ►► Be careful not to include leading or trailing spaces in the

delimiter string unless your intent is to distinguish between
occurrences of the delimiter with and without surrounding spaces (e.g.
“X;Y” represents a single value but “X ; Y” represents two values).
MDM will automatically trim the leading and trailing spaces from
individual delimited values even if you do not include the spaces.

NOTE ►► See “Configuration Options” on page 62 for more

information about the global default delimiter strings.

 To set the Split Hierarchy or Split Multiple Fields delimiter string for
one or more fields:
1. In the Source Hierarchy tree, select the field nodes for which you want
to set the Split Hierarchy or Split Multiple Fields delimiter string.
2. Right-click on one of the nodes and choose Set Split Delimiter from the
context menu, or choose Fields > Set Value Delimiter from the main
menu, and then choose from the cascading menu:
 Split Hierarchy
 Split Multiple Fields

150 MDM Import Manager Reference Guide

 3. MDM opens a Set Delimiter dialog like the one shown in Figure 101.

Figure 101. Set Delimiter dialog

4. MDM displays the current value for the delimiter string and whether or
not the global default is being inherited.
NOTE ►► No value is displayed in the edit control if you have
selected multiple fields for which the current values are not all the
same. Similarly, a gray check ( ) appears in the Inherit Default
checkbox if some but not all of the selected fields are inheriting the
global default.

5. To set the delimiter string and override the global default value, enter
the new delimiter string into the edit control, or alternatively, to restore
inheritance, check the Inherit Default checkbox.
6. Click OK when you are done to close the dialog.

MDM Import Manager Reference Guide 151

PART 8: CLASSIFICATION DATA IMPORT

This part of the reference guide explains how to import classifications

and attribute data; specifically, how to build a hierarchy from a set of
fields containing flattened hierarchy data, so that you can perform
subsequent field-level processing on the hierarchy; how to import a
complete taxonomy, including its hierarchy, attribute definitions, and
attribute linkages; and how to import attribute data that appears in the
source data as name/value pairs.

153
Overview
Sometimes, source data may comprise a classification scheme
consisting of just a classification hierarchy or a complete taxonomy, or
may comprise the attribute data itself, as follows:
• Classification hierarchy. A set of fields represents a hierarchy that
has been flattened in various ways, and which must be rebuilt into
the hierarchy prior to import into MDM.
• Complete taxonomy. A set of tables represents an entire taxonomy,
consisting of the hierarchy itself, the attribute definitions, and the
node-specific attribute linkages.
• Attribute data. The source table represents attribute data in a pair of
fields that contain attribute names and attribute values as
name/value pairs.

The operation for creating a hierarchy from a set of fields is described in

the following section, and summarized in Table 15.

Table 15. Build Hierarchy Operation

Operation Description
Creates a new field based on the selected fields and
Create Hierarchy Field
then splits it into a hierarchy.

NOTE ►► See “Taxonomy Import” on page 161 for information on the

sequence of operations for importing a complete taxonomy.

NOTE ►► See “Attribute Data Import as Name/Value Pairs” on page

167 for information on importing attribute data as name/value pairs.

MDM Import Manager Reference Guide 155

Building a Hierarchy
Source data fields may represent a hierarchy that has been flattened in
various ways, as summarized in Table 16 and further described in the
following sections.

Table 16. Flattened Hierarchy Representations

Type # Fields Source Fields Description
Parent/Child  Parent node value Pair of fields contain
Value Pairs 3  Child node value parent/child value pairs; third
w/ Node Name  Child node name field contains node name.
Parent/Child Pair of fields contain
 Parent node value
Value Pairs 2 parent/child value pairs; child
 Child node value
w/o Node Name value contains node name.
Single field contains the
Hierarchy Codes  Hierarchy node code
2 encoded hierarchy; second
w/ Node Name  Node name
field contains node name.

NOTE ►► Each record in the data source identifies a distinct internal

or leaf node in the hierarchy.

PARENT/CHILD VALUE PAIRS W/ NODE NAME

A pair of source data fields may contain parent/child value pairs, with a
third field indicating the node name, as shown in Figure 102.

Source: Corresponding hierarchy:

PARENT CATEGORY NODE NAME ► CATEGORY [hierarchy]
001 Printers └ Printers
001 002 Laser ├ Laser
├ Inkjet
001 003 Inkjet │ ├ B&W
003 004 B&W │ └ Color
└ Dot Matrix
003 005 Color
├ 300 DPI
001 006 Dot Matrix ├ 600 DPI
006 007 300 DPI └ 1200 DPI

006 008 600 DPI

006 009 1200 DPI

Figure 102. Parent/child value pairs with a third name field

NOTE ►► In the illustration above, the parent/child values are codes

rather than node names.

156 MDM Import Manager Reference Guide

PARENT/CHILD VALUE PAIRS W/O NODE NAME
A pair of source data fields may contain parent/child value pairs, with
the child value indicating the node name, as shown in Figure 103.

Source: Corresponding hierarchy:

PARENT CATEGORY ► CATEGORY [hierarchy]
Printers └ Printers
Printers Laser ├ Laser
├ Inkjet
Printers Inkjet │ ├ B&W
Inkjet B&W │ └ Color
└ Dot Matrix
Inkjet Color
├ 300 DPI
Printers Dot Matrix ├ 600 DPI
Dot Matrix 300 DPI └ 1200 DPI

Dot Matrix 600 DPI

Dot Matrix 1200 DPI

Figure 103. Parent/child value pairs

HIERARCHY CODES W/ NODE NAME

A single field may contain codes where the parent/child hierarchy
relationship is encoded within the class code itself, with a second field
indicating the node name, as shown in Figure 104.

Source: Corresponding hierarchy:

CODE CATEGORY ► CATEGORY [hierarchy]
17000000 Printers └ Printers
17010000 Laser ├ Laser
├ Inkjet
17020000 Inkjet │ ├ B&W
17020100 B&W │ └ Color
└ Dot Matrix
17020200 Color
├ 300 DPI
17030000 Dot Matrix ├ 600 DPI
17030100 300 DPI └ 1200 DPI

17030200 600 DPI

17030300 1200 DPI

Figure 104. Single coded value with a second name field

MDM Import Manager Reference Guide 157

THE CREATE HIERARCHY FIELD COMMAND
Regardless of how the hierarchy is represented in the source data, you
can use the Create Hierarchy Field command to recreate the hierarchy.
When you use the Create Hierarchy Field command to convert the
parent/child value pairs into a hierarchy, the MDM Import Manager
creates two new fields: (1) a field that contains the full path of the node
indicated by the corresponding record, as a set of delimited values; and
(2) a field with the delimited values of the first new field automatically
expanded by MDM into a hierarchy using the Split into Hierarchy
command, as shown in Figure 105

Source: Two new fields:

PARENT CATEGORY ► HIERARCHY [delimited] HIERARCHY [hierarchy]
Printers Printers └ Printers
Printers Laser Printers\Laser ├ Laser
├ Inkjet
Printers Inkjet Printers\Inkjet │ ├ B&W
Inkjet B&W Printers\Inkjet\B&W │ └ Color
└ Dot Matrix
Inkjet Color Printers\Inkjet\Color
├ 300 DPI
Printers Dot Matrix Printers\Dot Matrix ├ 600 DPI
Dot Matrix 300 DPI Printers\Dot Matrix\300 DPI └ 1200 DPI

Dot Matrix 600 DPI Printers\Dot Matrix\600 DPI

Dot Matrix 1200 DPI Printers\Dot Matrix\1200 DPI

Figure 105. Two new fields of the Create Hierarchy Field command

You can then perform subsequent processing on the split field just as if
it were an original field in the source data. The two new fields are added
to the Source Hierarchy tree, with the second one highlighted in italics
(to indicate that it can be deleted) as shown in Figure 106.

Figure 106. Creating a hierarchy field from a pair of fields

NOTE ►► In prior versions of Import Manager, creating a hierarchy

field was permitted only when the data source was updateable. This
limitation is removed as of MDM 7.1.

158 MDM Import Manager Reference Guide

 TIP ►► You can delete the split hierarchy field using the Delete Field
command.

 To combine two or three fields containing hierarchy data into a

hierarchy:
1. In the Source Hierarchy tree, select the two or three field nodes you
want to use build into a hierarchy:
 Parent/child value pairs with node name – three
 Parent/child value pairs w/o node name – two
 Hierarchy codes with node name – two
2. Right-click on one of the nodes and choose Create Hierarchy Field from
the context menu, or choose Source > Create Hierarchy Field from the
main menu, as shown in Figure 107.

Figure 107. Create Hierarchy Field context menu command

3. MDM opens the Create Hierarchy Field dialog shown in Figure 108.

MDM Import Manager Reference Guide 159

 Figure 108. Create Hierarchy Field dialog

4. In the Hierarchy Field Name edit control, type the name for the new
field.
5. In the Parent Field, Child Field, and Node Name Field drop-down
controls, identify the field containing the parent value, the child value,
and the node name value respectively.
TIP ►► For two-node selection of parent/child value pairs without a
node name, set the Node Name Field to None. MDM uses the values of
Child Field as the name of each node.

TIP ►► For two-node selection of hierarchy codes with a node name,

set the Parent Field to None. MDM automatically decodes the values of
Child Field to determine the parent/child hierarchy relationships.

6. Click OK to close the Create Hierarchy Field dialog.

7. The MDM Import Manager creates a hierarchy field named “field <Split
Hierarchy>” as a sibling of and just after the first new node (Figure 106
above).
8. Press F2 to edit the name of the hierarchy field.
9. MDM highlights the name of the hierarchy field for editing.
10. Type the name you want for the hierarchy field and press Enter.
NOTE ►► The Create Hierarchy Field command uses the current Split
Hierarchy delimiter string to delimit the nodes in the new hierarchy field
and then to split the individual delimited values within the new field
(see “Setting the Split Delimiters” on page 150 for more information).

160 MDM Import Manager Reference Guide

Taxonomy Import
Source data tables may represent an entire taxonomy, consisting of the
hierarchy itself, attribute definitions, and node-specific attribute links.
Importing a complete taxonomy – independently of main table records –
is actually a multi-step operation that imports from multiple source
tables into multiple destination tables, one at a time as follows:
1. Import the hierarchy into the taxonomy table.
2. Import the attributes into the attributes table.
3. Import the attribute text values into the attribute text values table.
4. Import the node-specific attribute links into the attribute links table.
TIP ►► Building the hierarchy itself is described in the previous
section (see “Building a Hierarchy on page 156 for more information on
how to build the hierarchy using the Create Hierarchy Field command).

When you import a taxonomy, you import the hierarchy and the
hierarchy records into the taxonomy table, and then you import the other
elements of the taxonomy into three “virtual” destination tables related
to the taxonomy, as summarized in Table 17.

Table 17. Taxonomy-Related Destination Tables

Table Name Description
taxonomy table The taxonomy table itself.
taxonomy table [Attributes] The related table of attribute definitions.
taxonomy table [Text Values] The related table of attribute text values.
taxonomy table [Links] The related table of node-specific attribute links.

NOTE ►► For each taxonomy table, the drop-down list of destination

tables includes the taxonomy table in the first group of items and the
three taxonomy-related tables in the last group (see “The Current
Tables” on page 36 and “Selecting the Current Tables” on page 50).

TIP ►► In the following sections, set the Expand Multilingual Fields

configuration option to the As Multiple Fields or As Field and Language
settings to import multilingual data.

MDM Import Manager Reference Guide 161

IMPORTING THE HIERARCHY
Assuming you have already created the hierarchy using the Create
Hierarchy Field command, the first step of importing the taxonomy is to
import the hierarchy into the taxonomy table, as described in this
section.

 To import the hierarchy into the taxonomy table:

1. Set the Source Table to the table that contains the hierarchy records
(and the hierarchy field you created using the Create Hierarchy Field
command).
2. Set the Destination Table to the taxonomy table.
3. Make the Map Fields/Values tab the active tab.
4. In the Source Fields and Destination Fields grids, map the hierarchy
field to the display field of the taxonomy table by selecting the fields
and clicking on the Map button.
5. In the Source Values and Destination Values grids, add the hierarchy
to the taxonomy table by selecting: (1) all of the nodes one level under
the root of the source hierarchy; and (2) the applicable node of the
destination hierarchy, clicking on the Add button, and choosing Add
Branch as Child from the cascading menu.
NOTE ►► For (1) The Add button will be enabled only when the
source hierarchy is opened to only one level under the root (no further
levels are seen); then you select all the nodes. Import Manager will
then automap existing nodes and add new nodes in their respective
place of the destination hierarchy

NOTE ►► If you want to save the map to be used by MDIS, then you
should map only the field name and not any value. By using the Set
MDIS Unmapped Value Handling command you will tell what to do with
unmapped values and where to locate the root of the hierarchy of the
source.

6. Map the other fields and values as necessary.

7. Make the Match Records tab the active tab.
8. In the Value Matching grids, perform record matching on the display
field of the taxonomy table.
9. Set the import actions.
10. Perform the import.
11. Select File > Close from the main menu to prepare for the next step.

162 MDM Import Manager Reference Guide

IMPORTING THE ATTRIBUTES
The second step of importing the taxonomy is to import the attribute
definitions into the [Attributes] table, as described in this section.
NOTE ►► The [Attributes] table fields are summarized in Table 18.

 To import the attribute definitions into the [Attributes] table:

1. Set the Source Table to the table that contains the attribute definitions.
2. Set the Destination Table to the [Attributes] table.
3. Make the Map Fields/Values tab the active tab.
4. In the Source Fields and Destination Fields grids, either: (1) map the
attribute name to the Name field; or (2) map the remote key value to
the Alias field.
TIP ►► You can use the Alias field to store remote key values since
MDM supports key mapping for attribute text values but not for the
attributes themselves.

5. In the Source Fields and Destination Fields grids, either: (1) map the
attribute dimension to the Dimension field; or (2) map the attribute
default unit to the Default Unit field.
NOTE ►► If you map: (1) to the Default Unit field, the MDM Import
Manager automatically determines the applicable dimension for each
attribute; (2) to the Dimension field, the MDM Import Manager does not
set a default unit; or (3) to both, the MDM Import Manager ignores the
Dimension mapping.

NOTE ►► To facilitate organization of the units across all of the

dimensions, the Default Unit values are a hierarchy of dimensions and
the units for each dimension.

TIP ►► You can partition the attribute dimension by attribute default

unit in the source data to create a hierarchy of dimensions and units.

NOTE ►► You must map manually dimensions or units one-by-one.

6. Map the other fields and values as necessary.

7. Make the Match Records tab the active tab.
8. In the Value Matching grids, perform record matching on the Name
field or the Alias field, as applicable.
9. Set the import actions and perform the import.
10. MDM imports the attribute definitions.
11. Select File > Close from the main menu to prepare for the next step.

MDM Import Manager Reference Guide 163

Table 18. [Attributes] Table Fields
Field Type Description
Name Text (multilingual) The name.
Alias Text The alias.
Definition Text (multilingual) The definition.
Type Lookup [Flat] The type (Text / Numeric / Coupled Numeric).
Key Mapping Boolean Whether key mapping (No/Yes)?
Multi-Valued Boolean Whether multi-valued (No/Yes)?
Dimension Lookup [Flat] The dimension (list of dimensions).
Default Unit Lookup [Hierarchy] The default unit (hierarchy of units).
Decimal Places Integer The decimal places (3).
Show Fractions Boolean Whether show fractions (No/Yes).
Ratings Lookup [Flat] (m/v) The ratings (Nom/Min/Max/Avg/Typ) (Nom).
Coupled Delimiter Text The coupled delimiter (@).
Coupled Name Text (multilingual) The coupled name.
Coupled Default Unit Lookup [Hierarchy] The coupled default unit.
Coupled Dimension Lookup [Flat] The coupled dimension (list of dimensions).
Coupled Decimal Places Integer The coupled decimal places (3).
Coupled Show Fractions Boolean Whether coupled show fractions (No/Yes).
* Where applicable, the default value appears in bold.

164 MDM Import Manager Reference Guide

IMPORTING THE ATTRIBUTE TEXT VALUES
The third step of importing the taxonomy is to import the attribute text
values into the [Text Values] table, as described in this section.
NOTE ►► The [Text Values] table fields are summarized in Table 19.

 To import the attribute text values into the [Text Values] table:
1. Set the Source Table to the table that contains the attribute text values.
2. Set the Destination Table to the [Text Values] table.
3. Make the Map Fields/Values tab the active tab.
4. In the Source Fields and Destination Fields grids, either: (1) map the
attribute name to the Attribute Name field; or (2) map the remote key
value to the Attribute Alias field.
TIP ►► You can use the Attribute Alias field to store remote key
values since MDM supports key mapping for attribute text values but
not for the attributes themselves.

5. In the Source Fields and Destination Fields grids, either: (1) map the
text value to the Text Value field; or (2) map the text value remote key
value to the [Text Value Remote Key] field.
NOTE ►► MDM supports key mapping for attribute text values using
the [Text Value Remote Key] field.

6. Map the other fields and values as necessary.

7. Make the Match Records tab the active tab.
8. In the Value Matching grids, perform record matching on the Attribute
Name field or the Attribute Alias field, as applicable.
9. Set the import actions and perform the import.
NOTE ►► You cannot create new attributes during this step and so
the None/None row in the Import Actions grid is disabled.

10. MDM matches each attribute and updates its set of text values.
11. Select File > Close from the main menu to prepare for the next step.

Table 19. [Text Values] Table Fields

Field Type Description
Attribute Name Text The name of the attribute to which it applies.
Attribute Alias Text The alias of the attribute to which it applies.
Text Value Text (multilingual) The attribute text value.
[Text Value Remote Key] Key Mapping The remote key value of the text value.

MDM Import Manager Reference Guide 165

IMPORTING ATTRIBUTE LINKS
The fourth and final step of importing the taxonomy is to import the
node-specific attribute links into the [Links] table.
Table 20. [Links] Table Fields
Field Type Description
Attribute Name Text The name of the attribute to which it applies.
Attribute Alias Text The alias of the attribute to which it applies.
Node Name Lookup [Hierarchy] The name of the linked node (taxonomy hierarchy).
The custom integer value or predefined priority value
Priority Lookup [Flat]
(Highest / High / Normal / Low / Lowest).
Link Type Lookup [Flat] The type of link (Ordinary / Master / Variant).

TIP ►► You can configure Import Manager to accept integer values in

the Priority field rather than the predefined values (Highest / High /
Normal / Low / Lowest) by setting the configuration option "Display
priority field for Attribute taxonomy links table using" to "custom
values."

 To import the node-specific attribute links into the [Links] table:

1. Set the Source Table to the table that contains the attribute links.
2. Set the Destination Table to the [Links] table.
3. Make the Map Fields/Values tab the active tab.
4. In the Source Fields and Destination Fields grids, either: (1) map the
attribute name to the Attribute Name field; or (2) map the remote key
value to the Attribute Alias field.
TIP ►► You can use the Attribute Alias field to store remote key
values since MDM supports key mapping for attribute text values but
not for the attributes themselves.

5. Map the other fields and values as necessary.

6. Make the Match Records tab the active tab.
7. In the Value Matching grids, perform record matching on the Attribute
Name field, the Attribute Alias field, and the Node Name field.
TIP ►► You can also combine either the Attribute Name or Attribute
Alias field with the Node Name field.

8. Set the import actions and perform the import.

NOTE ►► You cannot create new attributes during this step and so
the None/None row in the Import Actions grid is disabled.

9. MDM matches each attribute and updates its set of node-specific

attribute links.

166 MDM Import Manager Reference Guide

Attribute Data Import as Name/Value Pairs
Recall from the section on pivoting that field and attribute data may be
stacked “vertically” in the source table, with a pair of source fields
representing name/value pairs and each product record duplicated for
each name/value pair, and specifically, that when source data is
structured this way, you can use a pivot to “rotate” the source data so
that each field or attribute becomes a field in the source table.
NOTE ►► See “Pivoting” on page 114 for more information on
pivoting name/value pairs.

However, attribute data that is stacked vertically as name/value pairs

can also be imported directly without pivots by first partitioning the name
and value source fields, and then mapping the resulting name/value
hierarchy to a special destination field containing a name/value
hierarchy of attribute names and data values.
Consider the table shown in Figure 109, which contains three fields,
PART NO, which stores the part number, and NAME and VALUE, which
store attribute name and value information for each source record.

PART NO NAME VALUE

101 Cable Length 10 in
101 Color Blue
101 Color Red
101 Color Green
102 Cable Length 5 in
102 Color Green
103 Cable Length 5 in
103 Color Blue

Figure 109. Attribute data as name/value pairs

TIP ►► Remember to partition the name field by the value field, to

create a two-level hierarchy of attribute names and values.

This alternative approach allows you to avoid the tedious manual task of
mapping at the field level, by turning mapping into a value mapping
process. More importantly, by taking advantage of the automatic value
mapping and exception handling features of MDIS, this approach also
supports full MDIS automation of attribute data import, which would not
otherwise be possible with a variable set of attributes.
NOTE ►► You must use this alternative approach when the number
of attributes represented by the name/value pairs is too large to pivot
(because it would result in a very wide table with more than 255 fields).

MDM Import Manager Reference Guide 167

THE ATTRIBUTE DISPLAY CONFIGURATION OPTION
In order to import the two-level hierarchy of attribute names and values
that results from partitioning the name and value source fields, the
destination attributes must appear in the Destination Fields grid as a
single combined hierarchy item rather than a list of individual items.
You can use the Attribute Display configuration option to specify how to
display attributes and attribute values in the Destination Fields and
Destination Values grids of the Map Fields/Values tab, respectively, as
summarized in Table 21.

Table 21. Attribute Display Configuration Option

Setting Description
Display each attribute as a separate item with its
As Multiple Attributes
corresponding list of attribute data values.
As Attribute/Value Display a single item for all attributes with a two-
Hierarchy level hierarchy of attributes and data values.

THE [ATTRIBUTE/VALUE] HIERARCHY FIELD

The single attribute/value hierarchy item in the Destination Fields grid of
the Map Fields/Values tab is named [Attribute/Value] Hierarchy and is
of type Attributes, as shown in Figure 110.

Figure 110. [Attribute/Value] Hierarchy field in Destination Fields grid

NOTE ►► If there are multiple taxonomy lookup fields, the name of

the taxonomy lookup field is appended in angular brackets (<>).

NOTE ►► You can add new text attribute text values but you cannot
add new attributes using this approach; to add a new attribute use the
Add attribute button before mapping to the destination field.

NOTE ►► The Set Value Conversion Filter command is disabled at

the field level since: (1) source values can be text; and (2) numeric
source values can be of different measurement types.

168 MDM Import Manager Reference Guide

THE ATTRIBUTE/VALUE HIERARCHY
The attribute/value hierarchy of attribute names and values appears in
the Destination Values grid of the Map Fields/Values tab, as shown in
Figure 111, and subject to the constraints summarized in Table 22.

Figure 111. Attribute/value hierarchy in Destination Values grid

Table 22. Attribute/Value Hierarchy of Names and Values

Item Description
 Attributes displayed as a two-level hierarchy of names and values
Name/Value  Attribute names are at the first level, values at the second level
Hierarchy  Text values of a text attribute appear as its children
 Numeric values of a numeric attribute do not appear in the tree.
 Finds and maps the matching attribute name
Automap Button  Finds and maps the matching text values of a text attribute
 Converts the numeric values of a numeric attribute
 Attribute name can be mapped only to attribute name in a one-to-
one mapping
 Text values can be mapped to text values only if the text attribute
Map Button name is already mapped and only to values of the same attribute
(one-to-one, many-to-one, or one-to-many if multi-valued)
 Numeric values are automatically converted when the numeric
attribute name is mapped
 Text values can be added only if the text attribute name is already
Add Button
mapped and only to the same attribute
 All text attribute values automatically unmapped when the text
attribute name is unmapped
Unmap Button
 All numeric attribute values will be deconverted back to text when
the numeric attribute name is unmapped
Value Conversion  Disabled at the field level
Filters  Enabled or disabled at value level based on selected source row

MDM Import Manager Reference Guide 169

PART 9: PARTITIONING FIELDS/VALUES

This part of the reference guide explains the partitioning mechanism of

the MDM Import Manager, and how to use it to effortlessly import
source data into a destination repository even if the source and
destination structures would otherwise be incompatible.

171
Overview
Traditional import tools typically offer a field mapping capability that
allows you to map an individual source field to an individual destination
field, and then to import the data values from the source field directly
into the destination field.
However, it often may not be possible to directly map individual source
fields to individual destination fields – or to map, convert, or create
measurements from the underlying values within those individual source
and destination fields – because there may not be a direct one-to-one
relationship between individual source and destination fields.
Instead, the appropriate field mapping relationship may exist between
the combination of two or more source fields and/or the combination of
two or more destination fields and/or attributes (such as when the
CATEGORY and SUBCATEGORY source fields together correspond to
the Category destination field, or alternatively, the CATEGORY source
field corresponds to the Category destination field together with the
Printer Type and DPI attributes). In these cases, fields must be
combined to allow the mapping of field combinations rather than
individual fields; meanwhile, combining fields into field combinations
simultaneously combines the underlying values into value combinations,
to allow the value-level mapping, conversion, and measurement
creation of value combinations rather than individual values.

PARTITIONING AND FIELD MAPPING

To deal with the challenge of different source and destination field
structures, MDM’s generalized partitioning mechanism allows you to
combine two or more fields into field combinations and underlying
values into a corresponding set of value combinations.
In each case, you create field combinations by partitioning either an
entire field or just an individual value by another field or attribute. And
multiple partitions can be nested (creating hierarchy among the value
combinations) or combined (merging the value combinations).
Partitioning is one of the most powerful features of the MDM Import
Manager. It is the mechanism that allows you to arbitrarily restructure
electronic source data and rationalize it against destination data that
has an entirely different structure.
NOTE ►► Partitioning is part of the process of rationalizing and
preparing source data for import prior to the next transformation step
comprising field mapping, value mapping, value conversion, and
measurement value creation, all of which are fully described in the next
part of this guide.

MDM Import Manager Reference Guide 173

Source Field Partitioning
Sometimes, multiple source fields may correspond to a single
destination field.
Consider Figure 112, in which the individual source field CATEGORY
(which contains a single Printers category) does not contain sufficient
information to be mapped by itself to the more detailed destination field
Category (which contains three corresponding printer categories: Laser
Printers, Inkjet Printers, and Dot Matrix Printers).

Source: Destination:
CATEGORY ► Category
Printers Laser Printers
Bearings Inkjet Printers
Motors Dot Matrix Printers
Mounted Bearings
Unmounted Bearings
One-Phase Motors
Two-Phase Motors
Three-Phase Motors

Figure 112. Individual source field missing destination information

Now consider Figure 113, which illustrates that the missing detail exists
in the SUBCATEGORY source field, which together with CATEGORY
corresponds to destination field Category. In this case, you must
combine CATEGORY and SUBCATEGORY with partitioning prior to
field mapping, so that they can be field mapped as a combination to
Category, and their underlying values mapped as value combinations
rather than as individual values.

Source: Destination:
CATEGORY SUBCATEGORY ► Category
Printers Laser Laser Printers
Printers Inkjet Inkjet Printers
Printers Dot Matrix Dot Matrix Printers
Bearings Mounted Mounted Bearings
Bearings Unmounted Unmounted Bearings
Motors One-Phase One-Phase Motors
Motors Two-Phase Two-Phase Motors
Motors Three-Phase Three-Phase Motors

Figure 113. Source fields combined for value mapping

174 MDM Import Manager Reference Guide

 NOTE ►► Strictly speaking, if SUBCATEGORY contains only unique
values as in the example above, you can ignore CATEGORY and
simply field map SUBCATEGORY directly to Category. However, if
SUBCATEGORY contains duplicate values, partitioning is essential.

Similarly, in Figure 114, the two source fields FIRST NAME and LAST
NAME together correspond to the single destination field Name, and
must be merged with partitioning prior to field mapping, so that they can
be field mapped as a combination to Name, and their underlying values
mapped as merged value combinations rather than as individual values.

Source: Destination:
FIRST NAME LAST NAME ► Name
Joe Williams Joseph Williams
Jim Williams James Williams
Dave Rodgers David Rodgers
Dave Smith David Smith

Figure 114. Source fields combined for value mapping

Even without value mapping, multiple source fields may need to be

combined with partitioning prior to field mapping. In Figure 115, the
individual numeric source field SOCKET SIZE cannot be mapped by
itself to the measurement destination field Socket Size because it is
missing the unit of measure for each numeric value.

Source: Destination:
SOCKET SIZE ► Socket Size [length]
3 1 in
5 1.5 in
20 2 in
5 10 mm
2

Figure 115. Individual source field missing unit of measure information

Now consider Figure 116, which illustrates that the missing unit of
measure information exists in the SOCKET UNIT source field, which
together with SOCKET SIZE comprises the measurement values that
are stored in destination field Socket Size. In this case, you must
combine SOCKET SIZE and SOCKET UNIT with partitioning prior to
field mapping, so that they can be field mapped as a combination to
Socket Size, and measurement values created from their underlying
merged value combinations.

MDM Import Manager Reference Guide 175

 Source: Destination:
SOCKET SIZE SOCKET UNIT ► Socket Size [length]
3 in 1 in
5 in 1.5 in
20 mm 2 in
5 mm 10 mm
2 in

Figure 116. Source fields combined for measurement value creation

Finally, source fields may need to be combined so that the individual

values within each field can be treated as the set of multiple delimited
values for mapping to a multi-valued field, as shown in Figure 117.

Source: Destination:
SOCKET SIZE1 SOCKET SIZE2 ► Socket Size [length]
1/2 in 3/4 in 1 in; 1 1/2 in
1/4 in 3/8 in 16 mm; 18 mm
20 mm 22 mm 2 in; 2 1/2 in
5 mm 6 mm 10 mm; 12 mm
2 in 2 1/2 in

Figure 117. Source fields combined for mapping to a multi-valued field

Destination Field Partitioning

Just as multiple source fields may correspond to a single destination
field, a single source field may correspond to multiple destination fields
and attributes.
Consider Figure 118, in which the more detailed source field
CATEGORY (which contains seven printer categories) cannot be
mapped to the individual destination field Category (which contains only
a single Printers category) without losing the additional detail about
printers.

176 MDM Import Manager Reference Guide

 Source: Destination:
CATEGORY ► Category
300 DPI Laser Printers Printers
600 DPI Laser Printers
1200 DPI Laser Printers
300 DPI Inkjet Printers
600 DPI Inkjet Printers
300 DPI Dot Matrix Printers
600 DPI Dot Matrix Printers

Figure 118. Destination field missing source information

Now consider Figure 119, which illustrates that the additional printer
detail is stored in the Printer Type and DPI destination attributes, which
together with Category correspond to source field Category. In this
case, you must combine Category, Printer Type, and DPI with
partitioning prior to field mapping, so that CATEGORY can be field
mapped not just to Category by itself but rather to the field and
attributes as a combination, and CATEGORY values mapped to their
underlying value combinations rather than to their individual values.

Source: Destination:
CATEGORY ► Category Printer Type DPI
300 DPI Laser Printers Printers Laser 300
600 DPI Laser Printers Printers Laser 600
1200 DPI Laser Printers Printers Laser 1200
300 DPI Inkjet Printers Printers Inkjet 300
600 DPI Inkjet Printers Printers Inkjet 600
300 DPI Dot Matrix Printers Printers Dot Matrix 300
600 DPI Dot Matrix Printers Printers Dot Matrix 600

Figure 119. Destination field and attributes combined for value mapping

NOTE ►► If the multiple source values represent inconsistencies in

the source data or unnecessary detail that can be eliminated, then
partitioning is not necessary and the multiple source values should be
collapsed during the value-mapping step. (See “Collapsing Source
Value Variations” on page 214 for more information about collapsing
source value variations during value mapping.)

MDM Import Manager Reference Guide 177

Source and Destination Field Partitioning
Sometimes, multiple source fields may correspond to multiple
destination fields and/or attributes, so that both source and destination
fields must be combined with partitioning prior to field mapping.
Consider Figure 120, in which the source field CATEGORY (which
contains two lighting categories) cannot be mapped directly to the
destination field Category (which also contains two lighting categories)
because the categories are broken down completely differently and the
source and destination values simply don’t correspond at all.

Source: Destination:
CATEGORY ► Category
Commercial Lighting ≠ Incandescent Lighting
Residential Lighting ≠ Fluorescent Lighting
Commercial Plumbing Brass Plumbing
Residential Plumbing PVC Plumbing

Figure 120. Source and destination fields broken down differently

Now consider Figure 121, which illustrates that CATEGORY, together

with the additional information stored in the SUBCATEGORY source
field, corresponds to Category, together with the additional information
stored in the Type destination attribute. In this case, you must combine
CATEGORY and SUBCATEGORY and also combine Category and
Type, so that the combination of source fields can be field mapped to
the combination of destination fields, and underlying source value
combinations mapped to destination value combinations.

Source: Destination:
CATEGORY SUBCATEGORY ► Category Type
Commercial Lighting Incandescent Incandescent Lighting Commercial
Commercial Lighting Fluorescent Fluorescent Lighting Commercial
Residential Lighting Incandescent Incandescent Lighting Residential
Residential Lighting Fluorescent Fluorescent Lighting Residential
Commercial Plumbing Brass Brass Plumbing Commercial
Commercial Plumbing PVC PVC Plumbing Commercial
Residential Plumbing Brass Brass Plumbing Residential
Residential Plumbing PVC PVC Plumbing Residential

Figure 121. Source and destination both combined for value mapping

NOTE ►► You can partition as many source fields and destination

fields and attributes as necessary to establish a correspondence
between source and destination field and value combinations.

178 MDM Import Manager Reference Guide

Field Value Partitioning
Sometimes, multiple source or destination fields may need to be
combined, but unlike in the earlier examples, you must partition
individual field values rather than an entire field.
Consider Figure 122, in which the missing CATEGORY info exists not in
one but rather in three additional source fields, each of which
corresponds to a different CATEGORY value, and which collectively
correspond to destination field Category. Here, you must combine value
Printers with PRINTER TYPE, Bearings with BEARING TYPE, and
Motors with MOTOR TYPE to create the proper value combinations.

Source: Destination:
CATEGORY PRINTER TYPE BEARING TYPE MOTOR TYPE ► Category
Printers Laser ├ Printers
Printers Inkjet │ ├ Laser
Printers Dot Matrix │ ├ Inkjet
Bearings Mounted │ └ Dot Matrix
Bearings Unmounted ├ Bearings
Motors One-Phase │ ├ Mounted
Motors Two-Phase │ └ Unmounted
Motors Three-Phase └ Motors
├ One-Phase
├ Two-Phase
└ Three-Phase

Figure 122. Source values combined with partitioning

Alternatively, consider Figure 123, in which you must combine Category

value Printers with Printer Type, Bearings with Bearing Type, and
Motors with Motor Type to create the proper value combinations.

Source: Destination:
CATEGORY ► Category Printer Type Bearing Type Motor Type
Laser Printers Printers Laser
Inkjet Printers Printers Inkjet
Dot Matrix Printers Printers Dot Matrix
Daisywheel Printers Printers Daisywheel
Mounted Bearings Bearings Mounted
Unmounted Bearings Bearings Unmounted
One-Phase Motors Motors One-Phase
Two-Phase Motors Motors Two-Phase
Three-Phase Motors Motors Three-Phase

Figure 123. Destination values combined with partitioning

MDM Import Manager Reference Guide 179

DISTINCT VALUE LISTS AND VALUE COMBINATIONS
Recall that both the Source Hierarchy tree and the Destination
Hierarchy tree allow you to display distinct value lists for each field.
Moreover, each distinct value represents and is a proxy for not only all
instances of the value across the entire set of underlying records, but
also the subset of records that contain the value.
Partitioning relies upon the concept of distinct value lists and extends it
to value combinations. Instead of working with the set of distinct values
in a single field, you work with the set of distinct value combinations
across the multiple fields that have been combined with partitioning.
Partitioning relies upon and extends distinct value lists as follows:
• Field-level partitioning. Partitioning a field by another field partitions
each of the distinct values of the partitioned field as if it were directly
partitioned.
• Value-level partitioning. Partitioning a field value by another field
(either directly or by virtue of partitioning the field) creates the set of
value combinations of that value with each of the distinct values of
the partitioning field that actually exist in the underlying records.
NOTE ►► For a field combination created by partitioning, the set of all
possible value combinations is collapsed down to the set of those that
actually exist in the underlying records.

• Subsets of underlying records. Just as each distinct value

represents and is a proxy for the subset of records that contain the
value, each distinct value combination represents and is a proxy for
the smaller subset of records that contain the value combination, to
allow the value-level mapping, conversion, and measurement
creation of value combinations rather than individual values.
NOTE ►► Although the set of distinct value combinations will be
larger than each individual set of distinct values, each particular value
combination will then correspond to a smaller, more specific and more
precisely defined subset of underlying records, and thereby permit an
accurate value-level mapping that would not otherwise be possible.

TECHNICAL NOTE ►► In mathematical terms, partitioning n fields:

(1) generates the n-dimensional Cartesian product of ordered value
combinations across the n individual value domains; and (2) removes
from the complete set those value combinations that do not exist in the
underlying records. Creating this collapsed set of actual value
combinations is equivalent to partitioning – in the mathematical sense
– the set of underlying records into non-overlapping subsets of
records, where each value combination represents a single subset.

Consider the table of records and field values shown in Figure 124.

180 MDM Import Manager Reference Guide

 Record Field1 Field2
1 a x
2 c x
3 d z
4 b y
5 c x
6 b x
7 c y

Figure 124. Sample table of records and field values

For the table above, Figure 125 below illustrates the following as it
relates to partitioning Field1 by Field2:
• Distinct values. The set of distinct values for Field1 and Field2
(Field1• and Field2•, respectively).
• All possible value combinations. The set of all possible value
combinations created by the Cartesian product of Field1 and Field2
(Field1 x Field2).
• Actual value combinations. The set of actual value combinations
created by partitioning Field1 by Field2 (Field1 • Field2).

Field1• + Field2• ► Field1 x Field2 ► Field1 • Field2

a x a x a x
b y a y b x
c z a z b y
d b x c x
3 distinct
4 distinct values b y c y
values b z d z
c x Partitioning:
c y 7 actual value
c z combinations
d x
d y
d z

Cartesian product:
12 possible value
combinations

Figure 125. Distinct value lists and value combinations

MDM Import Manager Reference Guide 181

RECORD SUBSETS
The partitioning mechanism of the MDM Client has been enhanced and
extended within the MDM Import Manager to address the specific
requirements for importing source data.
Recall that in the MDM Client, the Partition command in Taxonomy
mode splits a category into multiple child categories according to the
values of an attribute, thereby partitioning the set of records in the
category into multiple subsets corresponding to each new child
category.
Similarly, partitioning a family in Family mode splits the family into
multiple child families according to the values of a field or attribute,
thereby partitioning the set of records in the family into multiple subsets
corresponding to each new child family.
In both cases, each subset of records corresponds to a particular value
combination represented by: (1) either the value of the original category
(in Taxonomy mode) or the value or value combination represented by
the original family (in Family mode); and (2) one of the values of the
field or attribute by which you partition.
Just as in the MDM Client, partitioning in the MDM Import Manager
splits a set of records into smaller subsets (hence the term partitioning),
and it does so by creating nodes that correspond to a set of actual value
combinations among the partitioned items.
But whereas the emphasis in the MDM Client is on the nodes as a proxy
for the underlying subset of records, the emphasis in the MDM Import
Manager is instead on the value combinations themselves, to allow the
value-level mapping, conversion, and measurement creation of value
combinations rather than individual values.
NOTE ►► By default, the set of distinct values for a field automatically
partitions the entire set of records into multiple subsets, each
corresponding to a distinct value. Partitioning a field by another field
further partitions each subset into multiple subsets, each
corresponding to an actual value combination.

TECHNICAL NOTE ►► Partitioning has the effect of: (1) horizontally

partitioning a single set of records into multiple subsets; by (2)
vertically combining multiple fields into a single field combination and
the underlying values into value combinations.

182 MDM Import Manager Reference Guide

PARTITIONS IN THE SOURCE AND DESTINATION HIERARCHIES
A partition can be applied in either the Source or Destination Hierarchy
tree to: (1) a field node, for inheritance by each value node beneath it in
the hierarchy that is not directly partitioned; or (2) a value node (either
internal or leaf), overriding the inheritance, if any, from a partitioned
node above it in the hierarchy (see the next section for more information
about inheritance).
NOTE ►► In the Source Hierarchy tree, you partition a source field or
value by another field in the current source table by selecting from the
Available Fields list on the left side of the Partition Field/Value tab.

NOTE ►► In the Destination Hierarchy tree, you partition a source

lookup field or value by another lookup field or attribute in the current
destination table by selecting from the Linked Attributes list on the right
side of the Partition Field/Value tab.

Table 23 summarizes how you can partition fields and values.

Table 23. Partitions in the Source and Destination Hierarchies

Partition… By… Description
Source Hierarchy
Partition each of the values of a field by each of the values of a
Field Field
second field.
Field
Field Partition a single field value by each of the values of a second field.
Value
Destination Hierarchy (NYI)
Lookup Lookup Partition each of the values of a lookup field by each of the values
Field Field of a second lookup field.
Lookup Lookup Partition a single lookup field value by each of the values of a
Field Value Field second lookup field.
Taxonomy Partition each of the values of a taxonomy lookup field by each of
Lookup Attribute the values of an attribute. The partition is applied only to the values
Field of the taxonomy field to which the attribute is linked.
Taxonomy
Linked Partition a single taxonomy lookup field value by each of the values
Lookup
Attribute of an attribute linked to that value.
Field Value
Within the existing partition of a taxonomy lookup field value by an
Linked Linked
attribute, partition each of the values of the attribute by each of the
Attribute Attribute
values of a second attribute linked to that lookup field value.
Linked Within the existing partition of a taxonomy lookup field value by an
Linked
Attribute attribute, partition a single value of the attribute by each of the
Attribute
Value values of a second attribute linked to that lookup field value.

TIP ►► Source partitions are usually applied at the field level, to

partition an entire field by another field. Destination partitions are
usually applied at the value level, to partition an individual category
value by a linked attribute.

MDM Import Manager Reference Guide 183

THE PARTITION FIELD/VALUE TAB
The Partition Field/Value tab displays two pairs of dual-list controls for
multiple-item selection, one for partitioning nodes of the Source
Hierarchy and one for partitioning nodes of the Destination Hierarchy
tree, as shown in Figure 126.

Figure 126. The Partition Field/Value tab

You can add both source and destination partitions using this tab, as
follows:
• Source partitioning. In the Source Hierarchy tree, you can partition
any field or value by any other field in the current source table by
selecting from the Available Fields list on the left side of the tab.
• Destination partitioning. In the Destination Hierarchy tree, you can
partition: (1) any lookup field or value by any other lookup field; or (2)
any taxonomy lookup field or value by any other lookup field or linked
attribute in the current destination table by selecting from the Linked
Attributes list on the right side of the tab.

NOTE ►► The lists on the left are disabled if you have not yet
selected the current source table. The lists on the right are disabled if
you have not yet selected the current destination table.

NOTE ►► When the source is an XML file, the Available Fields list is
limited to nodes which are siblings (on the same level as), or children
of (nested below) the currently selected Source Hierarchy tree node.

NOTE ►► The name of the selected node always appears as the first
item in the Partition By list as “fieldname” (for a field node) or
“fieldname [value]” (for a value node) to allow you to combine nodes by
which you partition with the partitioned node.

184 MDM Import Manager Reference Guide

FIELD AND VALUE PARTITIONING
When you partition a field node (or an internal value node), the MDM
Import Manager applies the partition to each leaf value node beneath it
that inherits the partition as if it were directly partitioned.
When you partition a value node (either directly or by virtue of
partitioning the field), the MDM Import Manager creates the set of actual
value combinations of that value (or value combination if the value node
is already the result of a partition) with each of the distinct values of the
partitioning field, and adds a child node to the hierarchy corresponding
to each value combination, as shown in Figure 127.

Figure 127. Node in Source Hierarchy tree before and after partitioning

NOTE ►► Partitioning a single leaf value node results in a single set

of value combinations. Partitioning a field node (or an internal value
node) results in a set of value combinations for each leaf value node
beneath it that inherits the partition.

NOTE ►► Nodes that are the result of partitioning have a purple icon
( ) to the left of the node name. Nodes that have been directly
partitioned (as opposed to having been partitioned through inheritance)
have a small purple square ( ) to the left of the node icon.

INHERITANCE
By default, when you partition an internal node (i.e. a node that has at
least one child node), each of the child nodes inherits the parent’s
partitions, each of its child nodes inherits the inherited partitions, and so
on, until the partitions are inherited by and applied to a leaf node.
NOTE ►► When a node is inheriting partitioning: (1) the Partition By
list in the Partition Field/Value tab contains the partition items that are
inherited; and (2) the Inherit checkbox is checked in the Partition
Field/Value tab to indicate that the items shown in the Partition By list
are inherited.

MDM Import Manager Reference Guide 185

Inheritance of partitions makes defining and refining the set of value
combinations extremely efficient. When you partition a field node, the
partition is inherited by each of the child value nodes, so that by
partitioning the field node, you can partition each of the values with a
single command.
When you directly partition any internal node or a leaf value node, the
inheritance of partitions from nodes above it in the hierarchy is broken,
and the value node is partitioned only by the fields and/or attributes
specified directly for that node.
In effect, directly partitioning the node breaks the chain of inheritance
from nodes above it in the hierarchy, and the partitions for the node
override any inherited partitions.
In this way, inheritance can be overridden for any node by directly
partitioning it in the hierarchy. Moreover, any child of the directly
partitioned node then inherits the partitions of the node unless the
partitioning information of the child node is itself directly partitioned.
NOTE ►► Nodes that have been directly partitioned (as opposed to
having been partitioned through inheritance) have a small purple
square ( ) to the left of the node icon in the hierarchy tree.

NOTE ►► In addition to the purple square, the Inherit checkbox is not

checked in the Partition Field/Value tab for nodes that have been
directly partitioned. To restore inheritance, recheck the Inherit
checkbox (see “Restoring Inheritance” on page 203 for more
information).

NOTE ►► Directly partitioning an internal node, all of whose children

are also directly partitioned, will not result in any nodes being added to
the hierarchy, since partitioning is only applied to leaf nodes and the
chain of inheritance is broken by each of its children.

MULTIPLE PARTITIONS
The result of a partition is that additional child value nodes are added to
the hierarchy.
When you partition a node by more than one field or attribute, each
partition is applied sequentially. In other words, the first partition is
applied to the leaf value node or to each leaf value node beneath it that
inherits the partition to create additional child value nodes, the second
partition is then applied to the newly created leaf value nodes beneath it
that inherit the partition to create additional child value nodes, and so
on, as shown in Figure 128, Figure 129, and Figure 130.

186 MDM Import Manager Reference Guide

 Figure 128. Multiple partitions before first partition

Figure 129. Multiple partitions after first partition

Figure 130. Multiple partitions after second partition

NESTED AND COMBINED PARTITIONS

By default, each partition that you add for a particular node in the
hierarchy is nested (i.e. applied sequentially) as described in the
previous section. Nested partitions add multiple levels of hierarchy, with
a level of child nodes created for the set of values in each of the nested
partitions.
Partitions can also be combined (i.e. applied as a group). Combined
partitions add a single level of hierarchy, with child nodes created for
each of the merged value combinations of the combined partitions.
A nested partition displays each value of each value combination as a
distinct node in the hierarchy, with the field nodes represented as sibling

MDM Import Manager Reference Guide 187

nodes at the top level of the hierarchy, the values of each field node
represented as sibling nodes at a second level in the hierarchy as
children of the first set of sibling nodes, the values of the first partitioning
field or attribute represented as sibling nodes at a third level in the
hierarchy, and so on for each field or attribute participating in the
partition, as shown in Figure 131.

Figure 131. Nested partitions (Brush Diameter • Brush Diameter Unit)

NOTE ►► Reordering nested partitions changes the hierarchy

structure created by the multiple partitions but results in the same set
of leaf nodes.

By contrast, a combined partition displays each value combination

consisting of the individual values of the field node and each of the
fields or attributes participating in the combined partition as a single
merged value in a single node in a hierarchy, with the entire set of value
combinations represented as sibling nodes at a single level in the
hierarchy, as shown in Figure 132.

Figure 132. Combined partitions (Brush Diameter • Brush Diameter Unit)

NOTE ►► The name of the selected node always appears as the first
item in the Partition By list as “fieldname” (for a field node) or
“fieldname [value]” (for a value node) to allow you to combine nodes by
which you partition with the partitioned node.

188 MDM Import Manager Reference Guide

Consider the following table of original records that contains values for
the fields CITY, STATE, and ZIP:

Original records:
CITY STATE ZIP
Los Angeles CA 90046
Los Angeles CA 90025
Hollywood CA 90069
Hollywood CA 90036
Miami Beach FL 33139
Hollywood FL 33019

Then the nested partition of STATE by CITY would be:

Nested:
STATE • CITY

├ CA
│ ├ Los Angeles
│ └ Hollywood
└ FL
├ Miami Beach
└ Hollywood

… and the combined partition of STATE by CITY would be:

Combined:
STATE • CITY
CA; Los Angeles
CA; Hollywood
FL; Miami Beach
FL; Hollywood

Nested vs. Combined

Not coincidentally, the value combination represented by each bottom-
level leaf node in a nested partition corresponds to the value
combination represented by each top-level sibling node in a combined
partition, and also to precisely the same subset of records, as follows:

MDM Import Manager Reference Guide 189

 Original records: Nested: Combined:
CITY STATE ZIP ► STATE • CITY ► STATE • CITY

Los Angeles CA 90046

CA
► ► CA; Los Angeles
└ Los Angeles
Los Angeles CA 90025

Hollywood CA 90069
CA
► ► CA; Hollywood
└ Hollywood
Hollywood CA 90036

FL
Miami Beach FL 33139 ► ► FL; Miami Beach
└ Miami Beach
FL
Hollywood FL 33019 ► ► FL; Hollywood
└ Hollywood

NOTE ►► Combining partitions results in fewer internal nodes but

exactly the same set of leaf nodes being added to the hierarchy, so
that the hierarchy has less depth and fewer levels than if the partitions
were not combined.

Nested and combined partitions would thus appear to be equivalent and

interchangeable, and in some ways, they are. However, there are subtle
differences between nested and combined partitions, as illustrated
above and further described below:
• Node values. Even though the bottom-level leaf nodes of a nested
partition correspond to the top-level siblings of a combined partition,
and both correspond to the same value combinations and subset of
records, the node values themselves are very different. In a nested
partition, the value of the bottom-level leaf node is the value of the
last partition field, whereas in a combined partition, the value of the
top-level sibling node is the merged value of all the partition fields.
• Hierarchy. A nested partition creates the entire hierarchy
represented by the value combinations of the partition, along with all
associated internal nodes, while a combined partition flattens the
hierarchy and does not create any internal nodes.
• Number of nodes for each field value. A nested partition creates a
single node for each value of the first partitioned field, with a child for
each value of the first field or attribute by which you partition,
“grandchild” nodes for each value of the second partition, and so on
for each partition, whereas a combined partition creates multiple
nodes for each value of the first partitioned field, with a child for each
value combination of each of the partitions, making it more difficult
with a combined partition to further partition a value of the partition.

190 MDM Import Manager Reference Guide

To Combine or Not to Combine
Regardless of whether partitions are nested or combined, partitioning
creates the same set of value combinations, each of which corresponds
to the same subset of records. So the obvious question is when to
combine partitions and when to allow them to remain nested.
The partial answer is that nested and combined partitions achieve the
same effect when: (1) partitioning fields by fields; (2) field mapping to a
destination field that requires value mapping; and (3) value mapping to
existing destination values and/or hierarchy.
NOTE ►► When value mapping to existing destination values, it
doesn’t matter whether the partition is nested or combined, since the
value of the destination field already exists and only the value
combination represented by the node is relevant, not the value of the
node itself.

However, nested and combined partitions can be used to very different

effect when: (1) partitioning individual field values rather than an entire
field; (2) merging field values without value mapping; (3) creating new
destination values; or (4) creating new destination hierarchy; as follows:
• Partition individual field values. In the example above, if the
partition of STATE value CA were to be further partitioned by ZIP,
but value Florida were to remain partitioned only by CITY, the nested
partition of STATE by CITY would be more appropriate than a
combined partition, which would create two nodes containing the
STATE value CA rather than one. In this case, the nested partitions
would be as follows:

Value partition:
STATE • CITY [ • ZIP]

├ CA [ • ZIP]
│ ├ Los Angeles
│ │├ 90046
│ │└ 90025
│ └ Hollywood
│ ├ 90069
│ └ 90036
└ FL
├ Miami Beach
└ Hollywood

• Merge field values without value mapping. If values need to be

merged for value conversion without value mapping, a combined
partition is appropriate, since these merged values are used to
create the destination field values.

MDM Import Manager Reference Guide 191

• Create new destination values. If new destination values need to
be created as part of value mapping, either a nested partition or a
combined partition may be appropriate depending upon the
circumstances, since nested and combined partitions result in
different leaf node values and the value of each node can be used to
be automatically define the new destination field values.
• Create new destination hierarchy. If new destination hierarchy
needs to be created as part of value mapping, a nested partition is
appropriate, since its internal nodes can be used to automatically
define the new destination hierarchy.

Guidelines on when to nest partitions and when to combine them are

summarized in Table 24.

Table 24. Guidelines for Nesting and Combining Partitions

Nest or Combine When…
 Partition is mapped to a destination field that requires value
Doesn’t Matter
mapping to existing destination values.
 Partition is mapped to a destination hierarchy field that requires
the creation of new destination hierarchy as part of the value-
Nest mapping process.
 Individual field values need to be partitioned.
 Partition is mapped to a destination field that requires value
conversion without value mapping and the values of the distinct
Combine
fields participating in the partition must be concatenated into a
single merged value for effective conversion.
 Partition is mapped to a destination field that requires value
Depends
mapping to new destination values.

Merged Values
When you combine two or more partitions, MDM creates the merged
value by: (1) concatenating the individual values of the field node and
each of the fields or attributes participating in the combined partition;
and (2) separating each pair of individual values with the Partition
Combine delimiter string specified for the field corresponding to the first
value of the pair.
NOTE ►► The Combine Partitions command separates each pair of
individual values that are merged using the current Partition Combine
delimiter string of the field corresponding to the first value of the pair,
which must be properly set prior to performing the command (see
“Setting the Partition Combine Delimiter” on page 200 for more
information).

192 MDM Import Manager Reference Guide

Consider Figure 133, which illustrates: (1) three source fields PREFIX,
CODE, and SUFFIX that must be concatenated into a single merged
value to form a Part Number with the format PREFIX-CODE/SUFFIX;
and (2) the values of the combined partition PREFIX • CODE • SUFFIX
if the Partition Combine delimiters are ‘-’ for PREFIX and ‘/’ for CODE.

Original: Combined partition:

PREFIX CODE SUFFIX ► PREFIX • CODE • SUFFIX
113 1234 21 ► 113-1234/21
113 1450 53 ► 113-1450/53
126 3308 43 ► 126-3308/43
109 2049 14 ► 109-2049/14

Figure 133. Merged Values in a Combined Partition

NOTE ►► Setting a different Partition Combine delimiter string for

each pair of values allows you to precisely format the concatenated
string for value mapping or conversion in the next step.

MATHEMATICAL ANALYSIS
In mathematical terms, partitioning n fields: (1) generates the n-
dimensional Cartesian product of ordered value combinations across
the n individual value domains; and (2) removes from the complete set
those value combinations that do not exist in the underlying records.
Creating this collapsed set of actual value combinations is equivalent to
partitioning – in the mathematical sense – the set of underlying records
into non-overlapping subsets of records, where each value combination
represents a single subset.
NOTE ►► You can then further partition either the entire domain of
value combinations or just a subset of the value combinations to create
additional Cartesian products based on the previous results.

Each value of a field’s value domain corresponds to an underlying

subset of records that contain the value, so that there is a correlation
between values and value combinations and corresponding subsets of
underlying records. In particular, each value domain partitions the entire
set of records into a set of non-overlapping subsets such that each
record belongs to exactly one subset, and the entire set of subsets
contains the entire set of records. Each partition of a field by another
field further partitions each subset into smaller subsets. Thus the term
partitioning is used to describe the mechanism of creating value
combinations because of the precise partitioning effect that it has on the
underlying subsets of records.

MDM Import Manager Reference Guide 193

The partitioning mechanism is necessary when the subsets of source
records implicitly created by partitioning the entire set of source records
by the values of just a single source field do not correspond to the
subsets of destination records implicitly created by partitioning the entire
set of destination records by the values of just the single corresponding
destination field. In these cases, the mechanism is used to further
partition, or refine, the subsets of source and/or destination records into
smaller subsets that do correspond to one another. Thus, the effect of
establishing a one-to-one correspondence between source and
destination values and/or value combinations is mathematically
equivalent to establishing a correspondence between the underlying
subsets of source and destination records.

INFINITE RESTRUCTURING FLEXIBILITY

Each of the various flavors of partitioning described above has a slightly
different effect but are still just variants of the same overall partitioning
mechanism, and each can be combined with any other, to achieve any
desired effect.
Using the partitioning mechanism of the MDM Import Manager, nested
and combined partitions can be applied to source fields and/or values
and destination fields and attributes and/or values in an unlimited
number of ways, enabling you to arbitrarily restructure and rationalize
electronic source data against destination data that has an entirely
different structure. Moreover, the underlying engine can not only import
but also transform data for destinations outside the repository.

SPLITTING VS. PARTITIONING: A COMPARISON

In the MDM Client, virtually all of the commands in Taxonomy mode that
allow you to structure and restructure the taxonomy exist as pairs of
symmetrical commands (Partition and Consolidate Children, Merge and
Split, Promote and Demote), where either command in the pair
completely reverse the action of the other.
In the MDM Import Manager, this same symmetry exists between the
effects of: (1) splitting a single field; and (2) partitioning multiple fields,
as illustrated in Figure 134:
• Splitting. Splitting starts with [A] a single field containing multiple
delimited values, and finishes with either [B] a single field containing
a hierarchy (Split into Hierarchy) or [C] multiple fields (Split into
Multiple Fields), each containing a single value.
• Partitioning. By contrast, partitioning starts with [C] multiple fields,
each containing a single value, and finishes with either [B] a single
field containing a hierarchy (Nested Partition) or [A] a single field
containing multiple delimited values (Combined Partition).

194 MDM Import Manager Reference Guide

 [A]
Single Field (delimited values)
Field
X; Y; Z

Split into Hierarchy

▼

[B]
Single Field (hierarchy value)
Split into Field
▲
Multiple Fields Combined
▼ └X Partition
└Y
└Z

▲
Nested Partition

[C]
Multiple Fields (individual values)
Field1 Field2 Field3
X Y Z

Figure 134. Symmetry of splitting and partitioning

MDM Import Manager Reference Guide 195

Partitioning Operations
The following sections describe the various partitioning operations you
can perform on nodes of the Source and Destination Hierarchy trees,
including:
• Partitioning a field or value node by a single field or attribute.
• Partitioning a field or value node by multiple fields and/or attributes.
• Combining separate partitions.
• Splitting partitions that were previously combined.
• Removing partitions.
• Reordering partitions.
• Restoring inheritance of partitioning information.

The partitioning operations are summarized in Table 25.

Table 25. Partitioning Operations

Operation Description
Adds the selected items to the list of partitions for the
Add Partition
selected tree node.
Removes the selected partition items from the list of
Remove Partition
partitions for the selected tree node.
Combines the selected partition items for the selected
Combine Partitions
tree node.
Splits the previously combined partition items for the
Split Partitions
selected tree node.
Reorders the selected partition items for the selected
Reorder Partitions
tree node.
Eliminates custom partitions for the selected tree
Inherit Partitions
node and restores inheritance of partitions.

TIP ►► You can use the Set Combine Delimiter command to set the
delimiter string that the MDM Import Manager uses to separate the
individual values that are merged when combining partitions.

196 MDM Import Manager Reference Guide

PARTITIONING BY A SINGLE FIELD OR ATTRIBUTE
You can partition a source or destination field or value node by a single
field or attribute item as described in this section.

 To partition a field or individual field value by a single field or attribute

item:
1. In the appropriate hierarchy tree, select the field or value node you
want to partition.
2. If necessary, click on the Partition Field/Value tab to make it the active
tab. This tab lists all the items by which you can partition the selected
node in each tree, as shown in Figure 135.

Figure 135. Source Hierarchy field node before partitioning

NOTE ►► MDM displays the selected node as the first item in the
Partition By list.

NOTE ►► In the Linked Attributes list for the Destination Hierarchy,

field names are displayed in normal type and attribute names are
highlighted in italics.

3. In the appropriate Available items list, select the item by which you
want to partition the selected node, and click on the Add button.
4. MDM partitions the selected node by the item, as shown in Figure 136.
Note that in the figure there are now new value nodes underneath the
partitioned node.

MDM Import Manager Reference Guide 197

 Figure 136. Source Hierarchy node after partitioning by a single field

NOTE ►► You can repeat this procedure to successively partition any

field or value node by other items.

NOTE ►► When you directly partition a node, MDM: (1) displays a

small purple square ( ) to the left of the node icon in the hierarchy
tree; and (2) unchecks the Inherit checkbox in the Partition Field/Value
tab. To restore inheritance, recheck the Inherit checkbox (see
“Restoring Inheritance” on page 203).

TIP ►► Nodes are nested in the hierarchy tree in the same order as
the items shown in the Partition By list on the Partition Field/Value tab.
You can change the nesting of the nodes by simply dragging-and-
dropping the list items to rearrange them into the desired order.

PARTITIONING BY MULTIPLE ITEMS

You can partition a field or value node by a combination of multiple field
and/or attribute items as described in this section.

 To partition a field or individual value by multiple field and/or attribute

items:
1. In the appropriate hierarchy tree, select the field or value node you
want to partition.
2. If necessary, click on the Partition Field/Value tab to make it the active
tab. This tab lists all the items by which you can partition the selected
node in each tree, as shown in Figure 135 above.
3. In the appropriate Available items list, select the items by which you
want to partition the selected node.
4. Click on the Add button and choose from the cascading menu:
 Add as Individual
 Add as Combination

198 MDM Import Manager Reference Guide

 5. MDM partitions the selected node by the selected items, either
individually or simultaneously, as shown in Figure 137. Note that in the
figure, there are now new value nodes underneath the partitioned node
for each combination of values of the partitioning fields.

Figure 137. Source Hierarchy node after partitioning by multiple items

NOTE ►► When you directly partition a node, MDM: (1) displays a

small purple square ( ) to the left of the node icon in the hierarchy
tree; and (2) unchecks the Inherit checkbox in the Partition Field/Value
tab. To restore inheritance, recheck the Inherit checkbox (see
“Restoring Inheritance” on page 203).

NOTE ►► When you partition a field or value node by multiple items

at the same time, MDM automatically creates a combined rather than a
nested partition, as described in the next section.

COMBINING PARTITIONS
You can combine two or more partitions that were previously added
separately as described in this section.

 To combine two or more existing partitions for a node:

1. In the appropriate hierarchy tree, select the node whose partitions you
want to combine.
2. If necessary, click on the Partition Field/Value tab to make it the active
tab.
3. In the appropriate Partition By list, select the two or more partition
items you want to combine into a single partition.
4. Click on the Combine button, or right-click on one of the items and
choose Combine Partitions from the context menu.
5. MDM combines the selected partition items.
NOTE ►► See “Splitting a Partition” on page 201 for more information
about how to split combined partitions.

MDM Import Manager Reference Guide 199

 NOTE ►► The Combine Partitions command separates each pair of
individual values that are merged using the current Partition Combine
delimiter string of the field corresponding to the first value of the pair,
which must be properly set prior to performing the command (see
“Setting the Partition Combine Delimiter” on page 200 for more
information).

NOTE ►► Combining partitions for a node that is currently inheriting

partitions from an ancestor in the hierarchy directly partitions the
selected node and breaks the inheritance for that node, so that it no
longer inherits partition changes made to its ancestor.

NOTE ►► When you directly partition a node, MDM: (1) displays a

small purple square ( ) to the left of the node icon in the hierarchy
tree; and (2) unchecks the Inherit checkbox in the Partition Field/Value
tab. To restore inheritance, recheck the Inherit checkbox (see
“Restoring Inheritance” on page 203).

SETTING THE PARTITION COMBINE DELIMITER

The Combine Partitions command separates each pair of individual
merged values with the current Partition Combine delimiter string of the
field corresponding to the first value of the pair, which must be properly
set prior to performing the command.
Specifically, the command use as the current delimiter either: (1) the
global default delimiter string, if the field is set to inherit; or (2) the field-
specific delimiter string, if it has been manually set for the field.
You can set the global default delimiter string for all fields that inherit the
default using the Configuration Options command, or alternatively,
override the default and set it on a field-by-field basis using the Set
Combine Delimiter command, as described in this section.
NOTE ►► Inherited global default values appear in the Source Fields
grid in gray.

NOTE ►► See “Configuration Options” on page 62 for more

information about the global default delimiter strings.

 To set the Partition Combine delimiter string for one or more fields:
1. In the appropriate hierarchy tree, select the field nodes for which you
want to set the Partition Combine delimiter string.
2. Right-click on one of the nodes and choose Partitions > Set Combine
Delimiter from the context menu, or choose Fields > Set Value
Delimiter > Partition Combine from the main menu, to open the Set
Combine Delimiter dialog shown in Figure 138.

200 MDM Import Manager Reference Guide

 Figure 138. Set Combine Delimiter dialog

TIP ►► Alternatively, you can right-click on the item in the Partition By

list whose combine delimiter string you want to set and choose Set
Combine Delimiter from the context menu.

3. MDM displays the current value for the delimiter string and whether or
not the global default is being inherited.
NOTE ►► No value is displayed in the edit control if you have
selected multiple fields for which the current values are not all the
same. Similarly, a gray check ( ) appears in the Inherit Default
checkbox if some but not all of the selected fields are inheriting the
global default.

4. To set the delimiter string and override the global default value, enter
the new delimiter string into the edit control, or alternatively, to restore
inheritance, check the Inherit Default checkbox.
TIP ►► You can restore inheritance for all of the value delimiter
strings for the selected fields by choosing Fields > Set Value Delimiter
> Inherit Defaults from the main menu.

5. Click OK when you are done to close the dialog.

SPLITTING A PARTITION
You can split a partition that was previously combined as described in
this section.

 To split a partition for a hierarchy node:

1. In the appropriate hierarchy tree, select the node whose combined
partition you want to split.
2. If necessary, click on the Partition Field/Value tab to make it the active
tab.
3. In the appropriate Partition By list, select the combined partition item
you want to split into its individual component partitions.
4. Click on the Split button, or right-click on one of the items and choose
Split Partitions from the context menu.
5. MDM splits the selected partition items.
NOTE ►► See “Combining Partitions” on page 199 for more
information about how to combine partitions.

MDM Import Manager Reference Guide 201

 NOTE ►► Splitting a partition for a node that is currently inheriting
partitions from an ancestor in the hierarchy directly partitions the
selected node and breaks the inheritance for that node, so that it no
longer inherits partition changes made to its ancestor.

NOTE ►► When you directly partition a node, MDM: (1) displays a

small purple square ( ) to the left of the node icon in the hierarchy
tree; and (2) unchecks the Inherit checkbox in the Partition Field/Value
tab. To restore inheritance, recheck the Inherit checkbox (see
“Restoring Inheritance” on page 203).

REMOVING PARTITIONS
Removing partitions in the Source or Destination Hierarchy is a matter
of simply selecting the fields and/or attributes by which a node was
partitioned, and removing them from the list of partitions for the node, as
described in this section.

 To remove one or more partitions from a node:

1. In the appropriate hierarchy tree, select the node from which you want
to remove the partitions.
2. If necessary, click on the Partition Field/Value tab to make it the active
tab.
3. In the appropriate Partition By list, select the one or more partition
items you want to remove, and click on the Remove button.
4. MDM removes the partition item(s) from the selected node, and
restructures the hierarchy to eliminate any nodes that had been
created as a result of the partition(s).
NOTE ►► Removing a partition for a node that is currently inheriting
partitions from an ancestor in the hierarchy directly partitions the
selected node and breaks the inheritance for that node, so that it no
longer inherits partition changes made to its ancestor.

NOTE ►► When you directly partition a node, MDM: (1) displays a

small purple square ( ) to the left of the node icon in the hierarchy
tree; and (2) unchecks the Inherit checkbox in the Partition Field/Value
tab. To restore inheritance, recheck the Inherit checkbox (see
“Restoring Inheritance” on page 203).

REORDERING PARTITIONS
You can reorder the fields and attributes by which a node is partitioned
by dragging and dropping the items in the Partition By list of the
Partition Field/Value tab.

202 MDM Import Manager Reference Guide

 NOTE ►► Reordering the partitions for a node that is currently
inheriting partitions from an ancestor in the hierarchy directly partitions
the selected node and breaks the inheritance for that node, so that it
no longer inherits partition changes made to its ancestor.

NOTE ►► When you directly partition a node, MDM: (1) displays a

small purple square ( ) to the left of the node icon in the hierarchy
tree; and (2) unchecks the Inherit checkbox in the Partition Field/Value
tab. To restore inheritance, recheck the Inherit checkbox (see
“Restoring Inheritance” on page 203).

RESTORING INHERITANCE
Directly partitioning an internal node or a leaf node breaks the
inheritance for that node. You can restore the inheritance for that node
as described in this section.

 To restore inheritance for a node:

1. In the appropriate hierarchy tree, select the node for which you want to
restore inheritance.
2. If necessary, click on the Partition Field/Value tab to make it the active
tab.
3. Check the Inherit checkbox.
4. MDM restores the inheritance to the selected node.
NOTE ►► Restoring the inheritance for a directly partitioned node
replaces the partition items for the node with the inherited partitions
from its ancestor in the hierarchy.

MDM Import Manager Reference Guide 203

PART 10: MAPPING FIELDS/VALUES

This part of the reference guide explains the field mapping and value
conversion and mapping mechanisms of the MDM Import Manager,
which allow you to cleanse, normalize, and rationalize source data
against destination data as one of the steps in the import process.

205
Overview
Field mapping is perhaps the most traditional and fundamental step in
the import process, in which most import tools allow you to map an
individual source field to an individual destination field, and then import
data values from the source field directly into the destination field.
The MDM Import Manager takes this basic field mapping capability and
extends it with a powerful value mapping and conversion capability, for
transforming, cleansing, and rationalizing source data as part of the
import process itself. In particular, value-level operations include:
• Value mapping. Source fields that contain text values and are
mapped to lookup fields or text attributes can be mapped at the value
level against the domain of legal destination values, eliminating the
need to pre-cleanse source data in an external application.
• Data type conversion. The data types of source data values are
automatically converted by the MDM Import Manager based on the
data type of the mapped destination field, even when stored in the
source data as text values.
• Measurement value creation. Unit of measure strings are
automatically normalized and appended to each numeric source data
value based on the physical dimension of the mapped destination
measurement field or numeric attribute.
• Data value conversion. Not only can each individual source value
be manually edited, built-in and custom-designed value conversion
filters can automatically convert or reformat a single value or the
entire set of distinct values with a single command.

Moreover, as described in the previous part of this guide, the MDM

partitioning mechanism allows you to combine fields into field
combinations prior to field mapping, so that you can: (1) field map to
and/or from partitioned field combinations rather than individual fields;
and then (2) value map to and/or from the underlying value
combinations rather than individual values.
NOTE ►► If the Partition Field/Value tab is the workhorse that allows
you to restructure source data with partitioning, the Map Fields/Values
tab is the workhorse that allows you to transform source data with field-
level mapping, and value-level mapping, conversion, and
measurement value creation.

INNOVATION ►► Field and value mapping are where field-at-a-time

processing and distinct value lists become enormously efficient. For
each mapped field or field combination, you must map, convert, or
adjust each distinct value or value combination just once to transform it
for all instances across the entire set of source records.

MDM Import Manager Reference Guide 207

THE MAP FIELDS/VALUES TAB
The Map Fields/Values tab contains two pairs of dual-grid controls, as
shown in Figure 139. These grids allow you to perform both field-level
mapping and value-level operations, as follows:
• Field level. For each source field or field combination whose data
values you want to import, you can use the top pair of Field Mapping
grids to map the field to the corresponding destination field.
• Value level. For each mapped source field, you can use the bottom
pair of Value Conversion and Mapping grids to either map, convert,
or create measurements from the underlying field values, dependent
upon the mapped destination field and data type.

Figure 139. Map Fields/Values tab

The two pairs of grids in the Map Fields/Values tab are described in
Table 26.

Table 26. Grids of the Map Fields/Values Tab

Grid Description
Field Mapping Grids

Source Fields Displays the fields of the current source table.

Displays the fields, tuples, attributes, and qualifiers of the current

Destination Fields
destination table.
Value Conversion and Mapping Grids
Displays the set of distinct source values for the current source
Source Values
field, either as a list or as a tree.
Displays the set of distinct destination values for the current
Destination Values
destination field, either as a list or as a tree.

208 MDM Import Manager Reference Guide

The Field Mapping Grids
Field mapping is used to establish the correspondence between source
and destination fields. The sets of source and destination fields
available for mapping are displayed in the Source Fields and
Destination Fields grids, respectively.
• Source Fields grid. Displays the fields of the current source table as
a tree (if XML) or a flat list.
• Destination Fields grid. Displays the fields of the current
destination table as a tree.

Both grids display fields according to their “natural” order. For example,
if a source file is in XML format, Import Manager uses a tree in the
Source Fields grid to depict the nested structure of fields within the
associated XML schema. Likewise, the Destination Fields grid displays
tuples, qualified lookup fields, and other nested objects in their natural
hierarchies, preserving the relationships within them that could
otherwise be lost in a flat list.
The Source Fields and Destination Fields grids are shown in Figure
140.

Figure 140. Field Mapping grids

NOTE ►► The Source Fields and Destination Fields grids are each
disabled if you have not yet selected the current source or destination
table, respectively.

NOTE ►► The use of trees in the Source Fields and Destination

Fields grid is new in MDM 7.1. In previous versions, these grids
contained only flat lists.

For each source field or field combination whose data values you want
to import, you can use the Field Mapping grids to map the field to its
corresponding destination field.
The columns of the Field Mapping grids are described in Table 27 (see
“Field Mapping Operations” on page 269 for more information).

MDM Import Manager Reference Guide 209

The Value Conversion and Mapping Grids
Value conversion and mapping layers upon and augments field-level
mapping and is used to transform source data at the value level.
Specifically, for each mapped source field, the MDM Import Manager
uses the bottom pair of Value Conversion and Mapping grids to display:
(1) the set of distinct values of the field as a list or a tree in the Source
Values grid; and (2) the set of distinct values of the mapped destination
field as a list or a tree in the Destination Values grid, as shown in Figure
141.

Figure 141. Value Conversion and Mapping grids

Table 27. Columns of the Field Mapping Grids

Column Description
Source Fields Grid
Mapped Whether the source field has been mapped ( ).
The display order of the field within the source hierarchy (n).
Position (Pos.)
 Sorting by this column restores the field order of the source hierarchy.
Name The name of the source field.
Type The data type of the source field (e.g. Text, Numeric).
The setting for the Flat NULL Handling option (gray if inheriting default):
Flat
NULLs*  Preserve NULLs
 Ignore NULLs
The setting for the Hierarchy NULL Handling option (gray if inheriting default):
 Preserve NULLs
 Ignore Non-Leaf Trailing NULLs
 > Preserve Leaf NULLs as NULLs
Hierarchy NULLs*
 > Preserve Leaf NULLs as Internal Leaf Nodes
 Ignore All Trailing NULLs
Ignore Non-Leaf Trailing NULLs has been split into two subsettings for flexible
handling of the leaf trailing NULLs.
The setting for the Case Sensitivity option (gray if inheriting default):
 Case Insensitive
Case  Case Sensitive
Sensitivity* For SQL Server or Oracle data sources, case sensitivity of the source field is
determined by database collation and cannot be changed by the MDM Import
Manager.
SH* The Split Hierarchy delimiter string (gray if inheriting default).
SM* The Split Multiple Fields delimiter string (gray if inheriting default).

210 MDM Import Manager Reference Guide

 Column Description
PC* The Partition Combine delimiter string (gray if inheriting default).
MV* The Multi-Valued delimiter string (gray if inheriting default).
CN* The Coupled Numeric delimiter string (gray if inheriting default).
Filters* The field-level filters (None, or semi-colon delimited list).
Destination Field The name of the corresponding mapped destination field.
Destination Fields Grid
Mapped Whether a source field is mapped to the destination field ( )
The display order of the field within the destination hierarchy (n).
Position (Pos.)
 Sorting by this column restores the field order of the destination hierarchy.
The field type:
 F – standard destination field
F/D/Q/A  D – display field of a lookup with multiple display fields
 Q – qualifier
 A – attribute
Name The name of the destination field.
Language The language instance of an expanded multilingual field.
The data type of the destination field (e.g. Integer, Text, Lookup [Flat]) or attribute
Type (e.g. Text, Numeric, Coupled), where a dimension appears as “Measurement
[dimension]” (for fields) or “Numeric [dimension]” (for attributes).
The setting of the NULL Interpretation option:
NULLs*  Ignore
 Overwrite
The setting of the Multi-Valued Update option or the Qualified Update option, as
applicable for the corresponding multi-valued or qualified lookup destination field:
MV/Qual*  Append
 Replace
 Update
The setting of the Data Group Display option:
Data Group*  Hide Data Group
 Display Data Group
The setting of the MDIS Unmapped Value Handling option:
 Ignore
 Add
MDIS*  Map
 Fail
You can set the Map option only at the field level, since the set of destination
values is different for each destination field.
Source Field The name of the mapped source field.

* Hidden by default; unhide to display.

MULTILINGUAL ►► More information about languages and

multilingual repositories is provided in “Part 14: Multilingual Support.”

For each mapped source field, you must use the Value Conversion and
Mapping grids to either map, properly convert, or create measurement
values from each of the distinct values for the field, where the required
mapping or conversion depends upon the mapped destination field type.

MDM Import Manager Reference Guide 211

 NOTE ►► Distinct value lists allow you to map, convert, or adjust
each distinct value just once to transform it for all instances across the
entire set of source records.

NOTE ►► The Source Values grid is disabled if the current source

field has not yet been mapped. The Destination Values grid is disabled
if the current mapped source field does not require value mapping.

The columns of the Value Conversion and Mapping grids are described
in Table 28 (see “Value Conversion and Mapping Operations” on page
294 for more information).
Table 28. Columns of the Value Conversion and Mapping Grids
Column Description
Source Values Grid
1 Whether the source value has been converted ( ) / mapped (
Status/Conv./Mapped
).
Value The original source value.
Converted Value The converted source value.
Destination Value The corresponding mapped destination value(s).
Destination Values Grid
Pos. The ordered position of the mapped destination value.
Mapped Whether a source value is mapped to the destination value ( )
Value The destination value.
Source Value The corresponding mapped source value(s).
1
Name changes depending on the selected field and whether value mapping is required.

NOTE ►► If the cell selection in the Source Values grid is on the

Converted Value column, typing enters the Converted Value edit
control rather than triggering typeahead seek on the sorted column.

VALUE MAPPING
When mapping to destination fields whose values are from a fixed
domain of valid values (lookup fields and text attributes), corresponding
source and destination values may not precisely match, and some
source values may not already exist as destination values.
In these cases, source fields that contain text values and are mapped to
lookup fields and text attributes can be mapped not only at the field
level, but also at the value level against the domain of legal destination
values, eliminating the need to pre-cleanse source data in an external
application.

212 MDM Import Manager Reference Guide

 NOTE ►► For each field, the MDM Import Manager collapses the set
of all values across the entire set of source records down to the set of
distinct values, whose number will usually be dramatically smaller than
the total number of records, often by several orders of magnitude.

Value-level mapping of source values forces a user to normalize source

data, implicitly defining a precise set of transformation rules that: (1)
eliminate differences between source and destination values; (2)
collapse multiple source value variations into a single destination value;
(3) properly map multiple delimited source values into a multi-valued
MDM field; and (4) explicitly add new values to the set of destination
values.
NOTE ►► The value in the Converted Value column is most useful for
adding new destination values or hierarchy, where it is used as the
new item or node value. When mapping to existing destination values,
the Converted Value is just a convenience for Automap, since each
value could be manually mapped to any destination value regardless of
its value.

NOTE ►► Value mapping to the value domain of a lookup field or text

attribute is like record matching to an index field in that both help to
avoid creating duplicates of records or values that already exist.

For each source value, you can use the Value Conversion and Mapping
grids to manually map to the corresponding destination value or values,
or you can automatically map values using the Automap button.
When value mapping is required, each distinct source data value must
be mapped and all discrepancies in the source data resolved in order to
perform the import, as indicated by a colored thunderbolt ( ) in the
Mapped column of the Source Values grid.
NOTE ►► When you map a source field to a destination field that
requires value mapping, the name of the first column changes from
Status to Mapped.

Eliminating Value Differences

Value mapping can be used to eliminate differences between source
and destination values, including the addition of new destination values.
This is necessary to avoid polluting the repository with illegal values,
which must be translated to the valid domain of destination values
during the import process.

MDM Import Manager Reference Guide 213

Consider Figure 142, in which source field STATE has been mapped to
destination field State. The source values of the STATE field are
completely different from the destination values of the State field, which
is itself a lookup into the States table that contains the set of consistent,
normalized two-letter state codes for each of the fifty states. Moreover,
the source value Penna has no corresponding destination value.

Source: Destination:
STATE ► State [lookup]
Calif ≠ CA
New Jersey ≠ NJ
Tenn ≠ TN
Flor ≠ FL
Penna ?

Figure 142. Source and destination value differences

In this case, each source value can be mapped to its corresponding
destination value, and a new destination value PA (shown in blue)
added to correspond to the source value Penna, as shown in Figure
143.

Source: Destination:
STATE ► State [lookup]
Calif ► CA
New Jersey ► NJ
Tenn ► TN
Flor ► FL
Penna ► PA

Figure 143. Value mapping eliminates value differences

NOTE ►► Source values that do not have corresponding destination

values can be added during the value mapping process.

Collapsing Source Value Variations

Value mapping can also be used to collapse multiple source value
variations into a single destination value. This is necessary when source
data originates in a system that permits entry of inconsistent data
values, which must be normalized during the import process.
Consider Figure 144, in which source field STATE has once again been
mapped to destination field State. This time, four source values (Calif,
Ca., California, and Cal) all correspond to a single destination value CA.

214 MDM Import Manager Reference Guide

 Source: Destination:
STATE ► State [lookup]
Calif CA
Ca.
California
Cal

Figure 144. Multiple source value variations

In this case, each source value can be mapped to the same destination
value CA, as shown in Figure 145.

Source: Destination:
STATE ► State [lookup]
Calif ►
Ca. ►
CA
California ►
Cal ►

Figure 145. Value mapping collapses value variations

Mapping to Multi-Valued Destination Fields

When the mapped destination field is an MDM multi-valued field, you
can use value mapping to map each set of delimited values in a single
source field to multiple values of the destination field. This is necessary
for most source data since other systems usually represent multi-valued
information as a delimited string.
Consider Figure 146, in which source field UOM, which contains
multiple delimited values in each record, has been mapped to
destination multi-valued field Unit of Measure.

Source: Destination:
UOM ► Unit of Measure [l]
Each; Carton Box
Box; Cylinder; Each Carton
Carton Each
Each; Package Package
Box

Figure 146. Multiple delimited source values

MDM Import Manager Reference Guide 215

As part of the value-mapping process, the set of delimited values
represented by each UOM source value can be mapped to the one or
more corresponding Unit of Measure destination values (including the
new value Cylinder, shown in blue), as shown as Figure 147.

Source: Destination:
UOM ► Unit of Measure [l]
Each; Carton Box
Box; Cylinder; Each Carton
Carton Each
Each; Package Package
Box Cylinder

Figure 147. Value mapping to multi-valued field

Mapping to Hierarchy Destination Fields

When the mapped destination field is an MDM hierarchy field, you can
use value mapping to map the values in a source field to the values of
the destination field, and also to create new destination hierarchy
NOTE ►► see “Hierarchy Values” on page 226 and “Adding New
Destination Hierarchy“ on page 306 for more information.

Consider Figure 148, in which source field CATEGORY has been

mapped to destination hierarchy lookup field Category.

Source: Destination hierarchy:

CATEGORY ► Category [lookup]
Laser Printers ├ Printers
Inkjet Printers │ ├ Laser
Dot Matrix Printers │ └ Dot Matrix
Mounted Bearings └ Bearings
Unmounted Bearings ├ Mounted
One-Phase Motors └ Unmounted
Two-Phase Motors
Three-Phase Motors

Figure 148. Destination hierarchy lookup field

As part of the value-mapping process, each source field value is

mapped directly to the corresponding value in the destination hierarchy
(including the new value for the Inkjet printers category, and an entirely
new branch for Motors and the three Motors categories, shown in blue).

216 MDM Import Manager Reference Guide

 Source: Destination hierarchy:
CATEGORY ► Category [lookup]

├ Printers
Laser Printers ► │ ├ Laser
Inkjet Printers ► │ ├ Inkjet
Dot Matrix Printers ► │ └ Dot Matrix
├ Bearings
Mounted Bearings ► │ ├ Mounted
Unmounted Bearings ► │ └ Unmounted
└ Motors
One-Phase Motors ► ├ One-Phase
Two-Phase Motors ► ├ Two-Phase
Three-Phase Motors ► └ Three-Phase

Figure 149. Value mapping to destination hierarchy lookup field

Mapping to Object Lookup Destination Fields

An MDM object lookup field (i.e. image, text block, PDF, or binary
object) is like any MDM lookup field, in that the value of the field is a link
to one or more values in the underlying lookup table.
When the mapped destination field is an MDM object lookup field, you
can map the values in a source field to the objects of the underlying
object lookup table, thereby linking the corresponding object to each
source record as part of the import.
The subtle differences among object lookup fields are summarized in
Table 29.

Table 29. Mapping to Object Fields

Object Value Cell in Destination Values Grid
Lists the name of each image in the Images table rather than the image
Image itself; each source value must be mapped to the name of an existing
destination image.
Lists the value of each text block in the Text Blocks table; you can map to
Text Block existing text blocks or add and map to new text blocks added as part of
the import, and also add new data groups prior to mapping.
Lists the name of each PDF in the PDFs table rather than the PDF itself;
PDF each source value must be mapped to the name of an existing
destination PDF.
Lists the name of each Binary Object in the Binary Objects table rather
Binary Object than the object itself; each source value must be mapped to the name of
an existing destination Binary Object.

MDM Import Manager Reference Guide 217

 NOTE ►► To distinguish among objects that have the same name,
you can display the Data Groups hierarchy in the Destination Values
grid so that each object appears as a leaf node beneath its data group.
(See “The Data Group Display Option” on page 265 for more
information.)
NOTE ►► You must use Data Manager to import new image, PDF,
or binaryobjects into the repository, since the MDM Import Manager
imports links to existing objects, not objects themselves.

NOTE ►► When the destination field is an image, PDF, or binary

object lookup field, the Add button and the Add Value commands for
the Destination Values grid are both disabled.

NOTE ►► When the destination field is a text block object lookup

field, the Values column in the Destination Values grid displays only
the first 255 characters of each text block, but allows you to edit new
text blocks of arbitrary length.

Mapping to Main Table Lookup Destination Fields

Import Manager handles main table lookup fields (Lookup [Main])
differently than other MDM lookup fields. Specifically, Import Manager
does not display the complete set of display field values of the records
of the underlying lookup table. Instead, the values it displays for a main
table lookup field are limited by both the key mappings for the lookup
table and the values in the source file.
Also, Import Manager does not automatically display the values of a
Lookup [Main] destination field in the Destination Values grid when you
select the field in the Destination Fields grid. Instead, for a main table
lookup field value to appear in the Destination Values grid, all of the
following conditions must be met:
• The lookup table must have key mapping enabled
• The lookup field must be mapped to a source field
• The source field must contain key values for the lookup table
• The destination value must have a key on the current remote system
• The destination value’s key must match a source field value

NOTE ►► The current remote system is the remote system that was
selected in Import Manager’s Connect to Source dialog (see
“Connecting to a Remote System” on page 412 for more information).

These differences are summarized in Table 30.

218 MDM Import Manager Reference Guide

Table 30. Differences between Lookup [Main] and Other Lookup Fields
Destination Values Lookup [Main] Fields Other Lookup Fields
Must have key mappings? Yes No
Visible only after field is mapped? Yes No
Limited by remote system? Yes No
Limited by source values? Yes No

Consider Figure 150, which displays key mappings for the lookup main
table, Customers, and which also shows customer-related fields in a
source file. Note that the values of source field ID match some of the
remote key values for the Customers main table.

Customers: Source:
Name [DF] [Key Mapping] ID Name
MDM; 301 301 John Smith
John Smith
CRM; 001 303 Susan Gray
Ken Hong CRM; 002 305 Mary Hsu
Susan Gray MDM; 303
Ed Bauer MDM; 307

Figure 150. Main table key mappings and source field values

Now consider Figure 151, in which source field ID is mapped to the

destination field Customer, which is a lookup into the Customers main
table. Note that the only destination values displayed for the Customer
field are those whose key mapping values for the current remote
system, MDM, match values in the mapped Source ID field.

Source: Destination:
ID ► Customer [lookup]
301 John Smith
303 Susan Gray
305

Remote System: MDM

Figure 151. Source values and main table lookup destination values

Once you map the source field to the main table lookup destination field,
Import Manager automaps each source field value to its corresponding
destination value for your convenience. You can also add and map
destination values as needed (shown in blue in Figure 152).

MDM Import Manager Reference Guide 219

 Source: Destination:
ID ► Customer [lookup]
301 ► John Smith
303 ► Susan Gray
305 ► Mary Hsu

Figure 152. Source values and main table lookup destination values

During import, MDM creates a new record on the lookup table for every
new destination value you added and mapped. As shown in Figure 153,
this new record is blank except for the mapped source value, which is
added as the value of the lookup table’s primary display field.

Customers:
Name [DF] Telephone
John Smith 800-555-1212
Susan Gray 212-999-9876
Mary Hsu

Figure 153. New record added to destination table

NOTE ►► Source values are added to text display fields only. If the
lookup table has no text display fields, the new record is entirely blank.

When Value Mapping Is Required

The Map checkbox resides at the bottom of the Map Fields/Values tab,
just between the Value Conversion and Mapping grids. Based upon the
current table and the type of the mapped destination field, it indicates
whether or not the current mapped source field requires value mapping,
and if checked, that you must map source to destination values.

Figure 154. Map checkbox (value mapping required)

NOTE ►► The Map checkbox is: (1) disabled and automatically

checked when value mapping is required, as shown in Figure 154; (2)
disabled and automatically unchecked when value mapping is neither
required nor permitted; and (3) enabled for you to check or uncheck
manually when value mapping is optional.

Whether or not particular destination field, attribute, and qualifier types

require or permit optional value mapping is summarized in Table 31.

220 MDM Import Manager Reference Guide

Table 31. Value Mapping Requirements
Required or Optional* to… Not Required and Not Permitted to…
Current table = Any table Current table = Any table
 Lookup [Flat]  Non-lookup fields
 Lookup [Hierarchy]  Numeric attributes
 Lookup [Taxonomy]  Coupled numeric attributes
 Lookup [Qualified]  Non-lookup qualifiers
 Lookup [Object]  AAA
 Lookup [Main] 
 Booleans 
 Text attributes
 Lookup qualifiers 
 Promoted Text lookup display fields* 
Current table = Hierarchy lookup table  AAA
 AAA
 Primary display field*
Current table = Main / non-lookup subtable
Current table = Any lookup table
 Display fields
 Text display fields*

NOTE ►► Value mapping is optional to fixed-width Text fields that

are: (1) promoted lookup table display fields when the current table is
the main table and the lookup table has multiple display fields; (2) the
primary display field when the current table is a hierarchy lookup table;
and (3) actual display fields when the current table is a flat lookup table
(see “Value Mapping and Lookup Table Display Fields” on page 246).

VALUE CONVERSION
Regardless of whether or not a field requires value mapping, individual
source data values may require: (1) manual adjustment or correction;
(2) data type conversion based on the respective source and destination
field data types; and (3) the assignment of a unit of measure.
You can manually perform value conversion on each source value by
editing the contents of each cell in the Converted Value column of the
Source Values grid, or you can automatically convert one or more
values by applying a value conversion filter at either the individual value
level or at the field level (see “Value Conversion Filters” on page 238 for
more information about value conversion filters).
When value mapping is not required, each value must be properly
converted in order to perform the import, as indicated by a green bullet (
) in the Conv. column of the Source Values grid.

Automatic Trimming of Leading and Trailing Spaces

When mapping source fields containing values with leading or trailing
spaces, it is not necessary to manually trim the leading or trailing
spaces from these values, as Import Manager automatically trims
leading and trailing spaces when it imports values into a repository.

MDM Import Manager Reference Guide 221

Automatic Data Type Conversion
When value mapping is not required, the data types of the
corresponding source and destination fields may be different, such as
when numeric, currency, date, time, or Boolean source data are each
stored as text values within text fields.
In these cases, the data types of source data values are automatically
converted within the Source Values grid based on the data type of the
destination field to which the source field is mapped, even when stored
in the source data as text values.
Original source values appear with their original data type in the Value
column, and the converted values appear in the Converted Value
column with the data type of the mapped destination field and the
proper grid cell type for editing (e.g. text, numeric, measurement,
currency, date, time, Boolean). If necessary, individual values can then
be manually edited to override errors or omissions in the source data.
NOTE ►► The Import Manager can parse and convert only Arabic
numbers when doing automatic value conversion to numeric fields.

Measurement Value Creation

When mapping to destination fields that represent physical
measurements (measurement fields and numeric attributes), the units of
measure in the source data may be: (1) missing entirely, such as when
all the values in a source field have the same implicit unit; (2)
inconsistent, such as when different text strings are used to represent
the same physical unit (e.g. inches, inch, in, or ‘"’); or (3) different for
each source data value.
The syntactical challenge of creating measurements is further
compounded because the source data values may be stored either: (1)
in a single numeric field; (2) in a single text field; or (3) in a pair of fields
(one containing the numeric value and the other containing the
corresponding unit of measure string); additionally, each field may
contain: (4) a single value; or (5) multiple delimited values.
For example, in Figure 155, all the measurement values have the same
implicit unit (which is missing entirely), and the numeric values are
stored in a single numeric field.
Source: Destination:
WIDTH (INCHES) ► Width [length]
3 3 in
2 1 in
1 1.5 in
2½

Figure 155. Measurement values with implicit unit in numeric field

222 MDM Import Manager Reference Guide

In Figure 156, multiple delimited measurement values each have
different units and inconsistent unit strings, and are stored in a single
text field.

Source: Destination:
SOCKET SIZE ► Socket Size [length]
2 “; 2.5 INCH 1 in; 1.5 in
10 millimeters; 12 mm 15 mm; 20 mm
15 mm; 20 millimeters 2.5 in; 3 in
2 1/2 IN; 3 “

Figure 156. Delimited measurement values with inconsistent unit strings

Finally, in Figure 157, the measurement values are stored in a pair of

fields, one containing the numeric value and the other containing the
corresponding unit of measure string.

Source: Destination:
WIDTH WIDTH (UNIT) ► Width [length]
3 inches 3 in
2 “ 1 in
1 inch 1.5 in
2 1/2 IN

Figure 157. Numeric value and unit string stored in two source fields
In each of these cases, unit of measure strings are automatically
normalized and appended to each numeric source data value based on
the physical dimension of the destination measurement field or numeric
attribute to which the source field is mapped. If necessary, each
numeric measurement value can be converted to a target unit, or
manually edited to override errors or omissions in the source data.
DATA INTEGRITY ►► MDM’s measurement data type guarantees
unit of measure consistency and permits each individual measurement
value to have its own distinct unit. Moreover, unit of measure strings
are automatically normalized against MDM’s built-in dictionary of over
2000 synonyms for over 750 units in over 70 physical dimensions. The
import cannot occur until each numeric source data value has an
associated unit of measure.

MDM Import Manager Reference Guide 223

SOURCE FIELD-LEVEL SETTINGS
The steps leading up to field mapping (and then to value conversion and
mapping) require that each of the various field-level settings for source
fields be set at the proper times and in the proper sequence.
Field-level settings for each source field include:
• Flat NULL handling. The Flat NULL Handling option determines
how NULLs are interpreted in multiple delimited values.
• Hierarchy NULL handling. The Hierarchy NULL Handling option
determines how NULLs are interpreted in hierarchy values.
• Case sensitivity. The Case Sensitivity option determines how case
differences are interpreted in source values.
• Value delimiters. The value delimiter strings determine how multiple
delimited values within a single field are identified and generated.
• Value conversion filters. One or more value conversion filters are
applied in sequence to each of the distinct values of a field.

The field-level settings for source fields (and when they must be set) are
summarized in Table 32. Each field-level setting can be set on a field-
by-field basis.
NOTE ►► NULL handling, case sensitivity, and delimiters must be set
before you field map because they affect how original values are
interpreted and how converted values are generated; the NULL
Handling and Case Sensitivity options also affect the set of original
values. By contrast, value conversion filters must be set after you map
because they depend on the destination data type.

Table 32. Field-Level Settings for Source Fields

Field Setting Must Be Set… Description
Flat NULL Handling Option (see page 279)
Flat Specifies whether to preserve or ignore
Prior to Field Mapping
NULL Handling NULLs in multiple delimited values.
Hierarchy NULL Handling Option (see page 280)
Hierarchy Specifies how to preserve or ignore
Prior to Field Mapping
NULL Handling NULLs in hierarchy values.
Case Sensitivity Option (see page 88)
Prior to Display Distinct Values, Specifies whether to ignore or recognize
Case Sensitivity
Split, Join, and Field Mapping case differences in source values.

Split Delimiters (see page 150)

Split Hierarchy Delimits values when splitting a field into
Prior to Split into Hierarchy
Delimiter a hierarchy.
Split Multiple Fields Delimits values when splitting a field into
Prior to Split into Multiple Fields
Delimiter multiple fields.
Partition Combine Delimiter (see page 200)

224 MDM Import Manager Reference Guide

 Field Setting Must Be Set… Description
Partition Combine Delimits merged values when combining
Prior to Combine Partitions
Delimiter partitions.
Field Mapping Delimiters (see page 276)
Multi-Valued Delimits values when mapping to a multi-
Prior to Field Mapping
Delimiter valued field or a compound lookup field.
Coupled Numeric Delimits value pairs when mapping to a
Prior to Field Mapping
Delimiter coupled numeric attribute.
Value Conversion Filters (see page 281 and 314)
Value After Field Mapping (and before Specifies the value conversion filters to
Conversion Filters value-level edits/conversions) be applied to each individual field value.

DESTINATION FIELD-LEVEL SETTINGS

The various field-level settings for each destination field can be set at
any time. Field-level settings for each destination field include:
• NULL interpretation. The NULL Interpretation option determines
how NULL values are interpreted for existing values.
• Multi-valued updates. The Multi-Valued Update option determines
how updates are handled in multi-valued destination fields.
• Qualified updates. The Qualified Update option determines how
updates are handled in qualified lookup destination fields.
• Data group display. The Data Group Display option determines
whether the Data Groups hierarchy is displayed for object lookup
destination fields.
• MDIS handling. The MDIS Unmapped Value Handling option
determines how unmapped values are handled during MDIS
processing.

The field-level settings for destination fields are summarized in Table

33. Each field-level setting can be set on a field-by-field basis.

MDM Import Manager Reference Guide 225

Table 33. Field-Level Settings for Destination Fields
Field Setting Must Be Set… Description
NULL Interpretation Option (see page 285)
Specifies whether NULL source values
NULL Interpretation Any time prior to import are ignored or used to overwrite existing
destination values.
Multi-Valued Update Option (see page 286)
Specifies whether source values are
Multi-Valued Update Any time prior to import appended to or replace existing multi-
values.
Qualified Update Option (see page 287)
Specifies whether source subrecords are
Qualified Update Any time prior to import appended to, replace, or update existing
qualified links.
Data Group Display Option (see page 289)
Specifies whether the Data Groups
Data Group Display Prior to Field Mapping hierarchy is hidden or displayed in the
Destination Values grid for object fields.
MDIS Unmapped Value Handling Option (see page 291)
Specifies how to handle unmapped
MDIS Unmapped
Any time prior to import values when running MDIS with a .map
Value Handling
file.

HIERARCHY VALUES
Recall that MDM supports hierarchical, parent/child relationships among
records with a special hierarchy table type and with lookup fields into
the hierarchy table. By contrast, a typical relational DBMS does not offer
native support for representing these hierarchical relationships.
As a result, importing hierarchy values into MDM encounters a variety of
challenges that depend upon the various ways the hierarchy values are
represented in the source data.
In addition, regardless of how the hierarchy values are represented,
how they are interpreted by MDM depends upon whether you are
importing into a hierarchy lookup field or into the hierarchy table itself.
NOTE ►► Whether you are importing into a hierarchy lookup field or
into the hierarchy table itself specifically affects how NULLs should be
interpreted (see “The Hierarchy NULL Handling Option” on page 233
for more information).

These various challenges and considerations are described more fully

in the following subsections.

226 MDM Import Manager Reference Guide

Hierarchy Representation in Source Data
Since a typical DBMS does not offer native support for representing
hierarchical relationships, hierarchy values can appear in source data in
at least three different ways:
• Delimited within a single field. A single source data field may be
used to represent each hierarchy value as multiple delimited values
(with or without trailing delimiters). You can then use the Split into
Hierarchy command to expand the delimited values into a hierarchy
(see “Splitting into Hierarchy” on page 146 for more information).

Category or Category
Printers Printers \ \
Printers \ Dot Matrix Printers \ Dot Matrix \
Printers \ Dot Matrix \ 300 DPI Printers \ Dot Matrix \ 300 DPI
Printers \ Dot Matrix \ 600 DPI Printers \ Dot Matrix \ 600 DPI
Printers \ Dot Matrix \ 1200 DPI Printers \ Dot Matrix \ 1200 DPI

• Expanded into multiple fields. Multiple source data fields may be

used to represent each hierarchy value, with each field representing
a single level of the hierarchy. You can then use nested partitions to
combine the values from each field into a hierarchy (see “Source
Field Partitioning” on page 174 for more information).

Category 1 Category 2 Category 3

Printers
Printers Dot Matrix
Printers Dot Matrix 300 DPI
Printers Dot Matrix 600 DPI
Printers Dot Matrix 1200 DPI

• Pair of parent/child fields. A pair of source data fields may be used

to represent each hierarchy node value, with one field for the node
and one for its parent node. You can then use the Create Hierarchy
Field command to materialize the hierarchy (see “Building a
Hierarchy” on page 156 for more information).

Parent Category
Printers
Printers Dot Matrix
Dot Matrix 300 DPI
Dot Matrix 600 DPI
Dot Matrix 1200 DPI

NOTE ►► The three source data examples above represent exactly

the same set of internal- and leaf-node hierarchy values.

MDM Import Manager Reference Guide 227

Importing Into a Hierarchy Lookup Field
When you are importing into a hierarchy lookup field, the source data
represents the assignment of lookup values to nodes of the hierarchy.
Since lookup values can be assigned only to leaf nodes in the hierarchy,
the source data should contain only leaf-node references, as illustrated
using multiple fields below:

… Category 1 Category 2 Category 3 …

Printers Dot Matrix 300 DPI
Printers Dot Matrix 600 DPI
Printers Dot Matrix 1200 DPI

NOTE ►► The fields representing the hierarchy node values are

contained in source records that also include other fields for the
records of the table containing the hierarchy lookup field.

NOTE ►► When importing into a hierarchy lookup field, you can: (1)
map to existing leaf nodes in the hierarchy; and (2) add new internal
and leaf nodes to the hierarchy.

NOTE ►► You can optionally map the source field representing the
hierarchy to a destination hierarchy lookup field. If the field is mapped,
then: (1) value mapping is required; (2) you can map only leaf nodes;
and (3) all of the leaf nodes must be mapped to proceed with the
import. This is consistent with the standard requirement that all values
of a lookup field must be mapped.

NOTE ►► Internal-node references in the source data may or may

not be acceptable when importing to the hierarchy lookup field, and
must be properly interpreted using the Hierarchy NULL Handling option
(see “The Hierarchy NULL Handling Option” on page 233 for more
information).

Importing Into the Hierarchy Lookup Table

When you are importing into the hierarchy lookup table itself, the source
data represents the actual internal- and leaf-node records of the
hierarchy table, as illustrated using multiple fields below:

… Category 1 Category 2 Category 3 …

Printers
Printers Dot Matrix
Printers Dot Matrix 300 DPI
Printers Dot Matrix 600 DPI
Printers Dot Matrix 1200 DPI

228 MDM Import Manager Reference Guide

 NOTE ►► The fields representing the hierarchy node values are
contained in source records that also include other fields for the
records of the hierarchy table itself.

NOTE ►► When importing into the hierarchy table itself, you can: (1)
map to existing internal and leaf nodes; and (2) add new internal and
leaf nodes to the hierarchy.

NOTE ►► You must map the source field representing the hierarchy
to the primary display field of the hierarchy table. Once the field is
mapped: (1) value mapping is optional; (2) if you do map values, you
can map both internal and leaf nodes; and (3) all of the internal and
leaf nodes must be mapped to proceed with the import.

NOTE ►► With respect to record matching: (1) if you do map values,

the [Remote Key] field can be used as a matching field only if it is
combined with the primary display field; and (2) if you do not map
values, the [Remote Key] field can be used as the matching field by
itself (see “Lookup Table Imports Using the [Remote Key] Field” on
page 414 for more information).

NOTE ►► Internal node value mapping allows you not only to import
a complete hierarchy consisting of internal and leaf nodes, but also to
populate for internal nodes: (1) the other fields of the hierarchy table
records (i.e. fields other than the primary display field); (2) other
language layers of all of the fields; and (3) the [Remote Key] field.

NOTE ►► Internal-node references in the source data are acceptable

when importing to the hierarchy table, and must be properly interpreted
using the Hierarchy NULL Handling option (see “The Hierarchy NULL
Handling Option” on page 233 for more information).

NOTE ►► The Replace import action is disabled for import into the
hierarchy table itself; the Import Manager and MDIS will substitute
Update (All Mapped Fields) in old maps that have the Replace import
action.

MDM Import Manager Reference Guide 229

NULL VALUES
In a typical DBMS, a field that is empty is said to have a NULL value,
which corresponds to missing or inapplicable data.
In the MDM Import Manager, a NULL value is displayed in trees and
grids as <NULL> to distinguish it from an actual field value, and appears
in the Value Conversion and Mapping grids as follows:
• Source Values grid. A <NULL> appears in the list of distinct source
values for a particular source field only if one or more source data
records contain no value for the field.
• Destination Values grid. For convenience, a <NULL> value always
appears in the list or tree of distinct destination values as a target for
mapping, regardless of whether or not the corresponding destination
field actually includes empty values in any of its records.

NULL values pose a number of challenges when mapping and

converting source data values, as described in the following sections.

Mapping To and From NULL Values

You can map to and from NULL values as follows:
• NULL to NULL. In the most straightforward case, you can simply
value map the <NULL> source value that represents inapplicable or
missing data to the <NULL> destination value. The <NULL> source
value is ignored during the import process.
• NULL to non-NULL. By contrast, if the <NULL> represents a default
value, or you know what the missing data should be, you can value
map the <NULL> to any existing non-NULL destination value, or
even change the converted value for the <NULL> source value and
then Add a new destination data value based on the converted value.
• Non-NULL to NULL. Finally, if one or more non-NULL source data
values need to be ignored as part of the import, you can value map
them to the <NULL> destination value.
NOTE ►► The MDM Import Manager interprets a field value
containing only whitespace (i.e. spaces or tabs) as a NULL value.

NOTE ►► When you map a source value (either NULL or non-NULL)

to the <NULL> destination value, the NULL is either ignored or used to
set existing destination values to NULL based on the setting of the
NULL Interpretation option (see “The NULL Interpretation Option” on
page 236 for more information).

230 MDM Import Manager Reference Guide

Converting To and From NULL Values
You can also convert source data values to and from NULL simply by
editing the corresponding value in the Converted Value column within
the Source Values grid.
By default, <NULL> is converted to <NULL>. You can also set the
converted value to a non-NULL value if the NULL represents a data
value; alternatively, you can set to NULL the converted value for any
other non-NULL value using the Set Value to NULL value conversion
filter to ignore it as part of the import.

Missing Source Data Fields

Sometimes, a field itself will be missing entirely from the source data, so
that in effect, all of its values are NULL. For example, the source data
from a particular manufacturer may not include a manufacturer field.
In these cases: (1) use the Add Field command in the Source Hierarchy
tree to add a new field whose list of values will contain the single
<NULL> value; (2) map the field to appropriate destination field; and (3)
map or convert the <NULL> value to the appropriate destination value.

The NULL Handling Options

In the MDM Import Manager, implicit and explicit NULL values also arise
in additional situations, as follows:
• Multiple delimited values. When a single field contains multiple
delimited values and you map the field to a multi-valued field, empty
positions in the delimited field values represent implicit NULLs.
• Splitting into hierarchy. When you split a field containing multiple
delimited values using the Split into Hierarchy command, empty
positions in the delimited field values again represent implicit NULLs
that become NULL-valued nodes in the resulting hierarchy.
• Nested partitions. When you combine fields with a nested partition,
explicit NULL values in any of the fields again become NULL-valued
nodes in the resulting hierarchy.
• Combined partitions. When you combine fields with a combined
partition, explicit NULL values in any of the fields become empty
positions in the resulting set of delimited values; when you map the
field combination to a multi-valued field, these empty positions again
represent implicit NULLs.

Depending upon the circumstances and how you would like the MDM
Import Manager to interpret the source data, you can specify that these
implicit and explicit NULLs be either preserved or ignored.

MDM Import Manager Reference Guide 231

How to handle NULLs is based on one of the two NULL handling
options (one applicable to handling delimited flat values and one to
handling hierarchy values), as described in the following sections.
NOTE ►► The Map Fields/Values tab populates the Source Values
grid with values based on the current NULL-handling settings, which
affects not only how original values are interpreted and how converted
values are generated, but also the set of original values, and therefore
must be properly set prior to mapping the field.

The Flat NULL Handling Option

The Flat NULL Handling option applies when you have a source field
that contains multiple delimited values or when you combine fields with
a combined partition. It has two settings, as described in Table 34.

Table 34. The Flat NULL Handling Option

Setting Description
Every NULL value is preserved in the list of distinct
Preserve NULLs values for the field, including implicit NULLs that appear
as empty positions in a set of delimited values.
Every NULL value is ignored in the list of distinct values
Ignore NULLs for the field, with empty positions collapsed and trailing
NULLs eliminated in a set of delimited values.

NOTE ►► Unlike the four settings of the Hierarchy NULL Handling

option, which have various subtleties, the two settings of the Flat NULL
Handling option are simple and either preserve or ignore all NULLs.

The default flat NULL-handling setting is to preserve NULLs, which most

accurately reflects the original source data. By contrast, ignoring NULLs
can be used to efficiently collapse inconsistencies in the source data
when they do not reflect real value differences. Whether to preserve or
ignore NULLs depends in large part on whether the position of data
values matters within the multiple delimited values.
For each source field, you can set the Flat NULL Handling option (see
“Setting the Flat NULL Handling Option” on page 279), which depends
upon whether the NULLs represent missing or inapplicable values.
NOTE ►► See “Source Field-Level Settings” on page 224 for more
information about all of the source field-level settings.

Table 35 illustrates the two alternative NULL-handling settings for the

values in two fields combined with partitioning. The same NULL-
handling behavior also applies to the values in a single field that already
contains delimited values and is mapped to a multi-valued field.

232 MDM Import Manager Reference Guide

Table 35. Preserve and Ignore NULL Handling for Delimited Flat Values
Field1 Field2 Preserve NULLs Ignore NULLs
value1 value2 value1; value2 value1; value2
value1 <NULL> value1; value1
<NULL> value2 ; value2 value2
<NULL> <NULL> ; <NULL>

 Delimited flat values occur: (1) when mapping to a multi-valued field or a compound
lookup fields; and (2) with combined partitions.
 The interpretation of implicit NULLs in a string of delimited values depends upon the
value of the Multi-Valued delimiter string.
 The field containing delimited values may itself have been created with a combined
partition, in which case the Multi-Valued delimiter string and Partition Combine delimiter
string should be the same.
 For Ignore NULLs, all implicit NULLs are collapsed and eliminated; when all positions
are empty, the result is <NULL>.
 Ignoring NULLs may shrink the set of distinct values because differences are eliminated
and duplicate values then collapsed.
 Do not ignore NULLs if position matters because different source values with the same
sequence of values in different positions would look like the same value after empty
positions were eliminated, and would then be collapsed into a single distinct value.

The Hierarchy NULL Handling Option

The Hierarchy NULL Handling option applies when you have a source
field that is split into hierarchy or when you combine multiple fields with
a nested partition. It has four settings, as described in Table 36.

Table 36. The Hierarchy NULL Handling Option

Setting Description
Every NULL value is preserved, including implicit NULLs
Preserve NULLs
that appear as NULL node values.
Ignore Non-Leaf Preserves internal NULLs and those trailing NULLs
Trailing NULLs necessary to interpret all values as leaf nodes.
Preserve Leaf NULLs
Trailing NULLs preserved as nodes named NULL.
as NULLs
Preserve Leaf NULLs
Trailing NULLs preserved as internal leaf nodes.
as Internal Leaf Nodes
Ignore All Preserves internal NULLs and ignores all trailing
Trailing NULLs NULLs, so that internal node values are preserved.

NOTE ►► The Ignore Non-Leaf Trailing NULLs setting has two

subsettings whose particular use depends upon whether or not you
want to preserve inner node assignments in your hierarchy.

Consider the hierarchy values represented by the three fields in the

three records below:

MDM Import Manager Reference Guide 233

 … Field 1 Field 2 Field 3 …
A
A B
A B C

In the simplest analysis, ignoring NULLs using the hierarchy NULL-

handling setting boils down to whether the three hierarchy values above
should be: (1) NULL-extended for interpretation as three leaf nodes; or
(2) interpreted instead as two internal nodes and one leaf node.
The three “ignore trailing NULLs” settings are summarized in Table 37.

Table 37. The Three “Ignore Trailing NULLs” Settings

Ignore Non-Leaf Trailing NULLs:
Preserve Leaf NULLs… Ignore All
Item
Trailing NULLs
As NULLs As Internal Leaf Nodes

Used when The hierarchy

A hierarchy lookup field.
importing to… lookup table.

Those trailing NULLs not required for

Ignores… interpretation as a leaf node (based on other All trailing NULLs.
leaf nodes at same level).
Views each A leaf-node A leaf- or inner node An internal- or leaf-
source value assignment in the assignment in the lookup node record in the
as… lookup hierarchy. hierarchy. hierarchy table.

TIP ►► Recall that for proper organization of the records within an

MDM repository, a hierarchy lookup field can normally be assigned
only to the value of a leaf node in the hierarchy. If, however, the
inbound records have inner node assignments in which the lookup field
is assigned to the value of an internal (i.e. non-leaf) node and you wish
to preserve these inner node assignments during inbound processing
of records, you can use the Ignore Non-Leaf Trailing NULLs (Preserve
Leaf NULLs as Internal Leaf Nodes) setting to create internal leaf nodes
instead of nodes named NULL.

The default hierarchy NULL-handling setting is to preserve all NULLs,

which most accurately reflects the original source data. Alternatively,
you can ignore some or all trailing NULLs using the three “ignore trailing
NULLs” settings. In all cases, internal NULLs are always preserved.
For each source field, you can set the Hierarchy NULL Handling option
(see “Setting the Hierarchy NULL Handling Option” on page 280), which
depends upon whether the NULLs represent missing or inapplicable
values, and whether you are importing into a hierarchy lookup field or
into the hierarchy table itself.
NOTE ►► See “Source Field-Level Settings” on page 224 for more
information about all of the source field-level settings.

234 MDM Import Manager Reference Guide

Table 38 illustrates the three alternative NULL-handling settings for the
values of three fields in six records combined with nested partitioning.
The same NULL-handling behavior also applies to the values in a single
field that already contains delimited values and is split into a hierarchy.
NOTE ►► Unlike the Flat NULL Handling option, which looks at each
individual record by itself, the Hierarchy NULL Handling option looks at
the entire set of source records and synthesizes the delimited values
across the entire set into a single integrated hierarchy.

DATA INTEGRITY ►► Regardless of the hierarchy NULL-handling

setting, internal NULLs are always preserved as nodes because
ignoring them would collapse hierarchy structure. Similarly, with the
Ignore Non-Leaf Trailing NULLs settings, leaf-node NULLs are always
preserved and/or automatically added as nodes to the hierarchy when
other nodes at the same level of the same branch have a non-NULL
value either: (1) because without the trailing NULL, the full-path value
would represent an internal rather than a leaf node; or (2) to preserve
the inner node assignment using an internal leaf node.

Table 38. Preserve and Ignore NULL Handling for Hierarchy Values
Ignore Non-Leaf Trailing NULLs:
Preserve Leaf NULLs… Ignore All
Preserve
Field1 Field2 Field3 Trailing
NULLs as Internal
as NULLs NULLs
Leaf Nodes
<NULL> <NULL> <NULL> Field 1 Field 1 Field 1 Field 1

Printers <NULL> <NULL> ├ NULL ├ NULL ├ NULL ├ NULL

│ └ NULL └ Printers └ Printers └ Printers
Printers Dot Matrix <NULL>
│ └ └ NULL ├ Dot Matrix ├ Dot Matrix └ Dot Matrix
Printers Dot Matrix 300 DPI └ Printers │ ├ 300 DPI │ ├ 300 DPI ├ 300 DPI
Printers Dot Matrix 600 DPI ├ Dot Matrix │ ├ 600 DPI │ ├ 600 DPI ├ 600 DPI
Printers Dot Matrix 1200 DPI │ ├ 300 DPI │ ├ 1200 DPI │ ├ 1200 DPI └ 1200 DPI
│ ├ 600 DPI │ └ NULL │ └ [Dot Matrix]
│ ├ 1200 DPI └ NULL └ [Printers]
│ └ NULL
└ NULL
└ NULL
 Hierarchy values occur: (1) when splitting into hierarchy; and (2) with nested partitions.
 The root node of the hierarchy is the field name itself.
 The individual hierarchy values across the set of records are synthesized into a single integrated
hierarchy rather than into a list of distinct values.
 The string “NULL” (as opposed to the <NULL> value) is used as the node value when the source
data has no value for an internal node.
 For the two Ignore Non-Leaf Trailing NULLs settings, inner node assignments are assigned to
either: (1) a leaf node named NULL; or (2) an internal leaf node.
 For all three of the Ignore Trailing NULLs settings, when the entire path of node values is <NULL>,
a single trailing NULL node is preserved under the root.

MDM Import Manager Reference Guide 235

The NULL Interpretation Option
Just as the NULL Handling options determines how NULLs are reflected
in source values, the NULL Interpretation option determines how NULLs
are reflected in destination values.
Consider first that when creating new destination records or replacing
existing destination records, importing NULL source data values into
destination fields can be handled in a single way: the destination field
value corresponding to the NULL source value remains empty.
By contrast, when updating existing destination records as part of the
import – and in particular when updating non-NULL destination fields
that already have one or more data values – importing NULL source
values can be handled in one of two ways.
Specifically, the set of incoming NULL source values can either be: (1)
ignored, so that corresponding destination field values in the existing
destination records remain undisturbed; or (2) used to overwrite the
corresponding destination field values in the existing destination
records, wiping them out so that destination field values becomes
empty, as described in Table 39.

Table 39. The NULL Interpretation Option

Setting Description
NULL source values are ignored, preserving existing
Ignore NULLs
data values in existing destination records.
Overwrite Existing NULL source values are used to overwrite existing
Values with NULLs destination field values, wiping them out.

NOTE ►► Among the pair of update import actions, the NULL

Interpretation option only affects the Update (All Mapped Fields) action,
because the Update (NULL Fields Only) action updates only fields that
do not yet have data values, making the Ignore and Overwrite settings
equivalent. (See “Import Actions” on page 346 for more information
about the import actions of the Record Matching tab.)

The default NULL-interpretation setting is to ignore NULLs. By contrast,

using NULLs to overwrite existing destination field values can be used
to wipe out values that are no longer valid.
For each destination field, you can set the NULL Interpretation option
(see “Setting the NULL Interpretation Option” on page 285) based on
the requirements of the situation.
NOTE ►► See “Destination Field-Level Settings” on page 225 for
more information about all of the destination field-level settings.

236 MDM Import Manager Reference Guide

THE CASE SENSITIVITY OPTION
Depending upon the circumstances and how you would like the MDM
Import Manager to determine the set of distinct values for a field, you
can specify that case differences between different instances of the
same value are either ignored or recognized, as described in Table 40.

Table 40. The Case Sensitivity Option

Setting Description
Case is ignored. Instances of the same value with case
Case Insensitive differences are collapsed into a single value in the list of
distinct values for the field.
Case is recognized. Every distinct value – including
Case Sensitive instances of the same value with case differences – is
preserved in the list of distinct values for the field.

The default Case-Sensitivity setting is to be case insensitive, which

ignores case and conveniently collapses inconsistencies in the source
data when case differences do not reflect real value differences (such
as with the values GENERAL MOTORS and General Motors, which
represent the same value). By contrast, case sensitivity recognizes case
and allows you to distinguish between different source values when
case differences represent value differences (such as with the values 6
mHz and 6 MHz, which truly represent different values).
For each field, you can set the Case Sensitivity option (see “Setting The
Case Sensitivity Option” on page 278), which depends upon whether
case differences represent legitimate value differences.
NOTE ►► See “Source Field-Level Settings” on page 224 for more
information about all of the source field-level settings.

NOTE ►► The Map Fields/Values tab populates the Source Values

grid with values based on the current case-sensitivity setting, which
affects the set of original values, and must be properly set prior to: (1)
displaying values using the Display Distinct Values command; (2)
partitioning the field; (3) splitting the field using the Split into Hierarchy
or Split into Multiple Fields commands; or (4) mapping the field.

NOTE ►► For SQL Server or Oracle data sources, case sensitivity of

the source field is determined by database collation and cannot be
changed by the MDM Import Manager.

MDM Import Manager Reference Guide 237

 DATA INTEGRITY ►► If the Case Sensitivity option is set to Case
Insensitive and you map a source field to a destination measurement
field or attribute whose corresponding dimension has case-sensitive
units (e.g. Frequency, Power, and Pressure, which have units with
both “milli-“ and “mega-“ prefixes), MDM will offer to change the setting
of the option for you automatically.

FIELD MAPPING DELIMITERS

The current field mapping delimiters determine how the MDM Import
Manager interprets source data values as part of the value conversion
and mapping process, and must be properly set prior to mapping a
source field to a destination field.
The two field mapping delimiters include:
• Multi-Valued delimiter. MDM identifies the individual delimited
values within a single field when mapping to a multi-valued field or a
compound lookup field using the Multi-Valued delimiter string.
• Coupled Numeric delimiter. MDM identifies the coupled numeric
value pairs when mapping to a coupled numeric attribute using the
Coupled Numeric delimiter string.

For each field, you can specify the field mapping delimiters (see “Setting
the Field Mapping Delimiters” on page 276).
NOTE ►► See “Source Field-Level Settings” on page 224 for more
information about all of the delimiters and source field-level settings.

VALUE CONVERSION FILTERS

Not only can each source value be manually edited, built-in and custom-
designed value conversion filters can automatically convert or reformat
a single value or the entire set of distinct values with a single command.
For each field, you can apply one or more value conversion filters at the
field level, and for each value, you can apply one or more value
conversion filters at the individual value level (regardless of whether or
not value mapping is required), as follows:
• Field level. When you apply value conversion filters at the field level,
the conversions are applied sequentially to each converted value that
has not been individually edited or converted at the value level,
starting with the original value.
• Value level. When you apply a value conversion filter at the value
level, the conversion is applied to the converted value resulting from
the prior application of field-level filters, value-level filters, and
manual edits.

238 MDM Import Manager Reference Guide

 NOTE ►► When you first map a field, each value in the Converted
Value column appears in gray to indicate that it is inheriting value
conversion filters from the field level. When you apply a manual edit or
a value-conversion filter to an individual value, the Converted Value
then appears in black to indicate that you have overridden inheritance
from the field level. You can use the Restore Converted Value
command to restore inheritance for a value. (See “Source Values Grid”
on page 296 for more information about the Source Values grid.)

NOTE ►► Value conversions are saved in a map file and applied

during future imports as follows: (1) at the field level, the filters
themselves are saved and applied; and (2) at the value level, the final
translation of each original value to a converted value (rather than the
sequence of edits and conversions) is saved and applied.

NOTE ►► No conversion filter is necessary to trim leading and trailing

spaces from source values, as Import Manager does this automatically
when importing values into an MDM repository.

NOTE ►► See “Source Field-Level Settings” on page 224 for more

information about all of the source field-level settings.

Value conversion filters automate repetitive and error-prone

transformations, eliminating manual typing and the possibility of user
error. They also allow powerful reformatting algorithms to be applied to
an entire group of values.
The value conversion filters affect the Converted Values and are
summarized in Table 41.

MDM Import Manager Reference Guide 239

Table 41. Value Conversion Filters
Filter Description
Restore
Restores inheritance of the field-level value conversion filters.
Converted Value*
Set to Original Value* Sets the selected values to their original unconverted values.

Sets the selected values to <NULL>.

Set Value to NULL  When value mapping is not required, a <NULL> value is deemed to be
converted
Clears the selected values.
Clear Value  When value mapping is not required, an empty value is deemed to be
not yet converted
Applies the numeric or string operator to the selected values:
 + – add the specified value
 - – subtract the specified value
 * – multiply by the specified value
 / – divide by the specified value
 Equals – set to the specified value
Apply Operator…  NULL Equals – set to the specified value only if NULL
 Round – round to the specified number of decimal places
 Ceiling – ceiling to the specified number of decimal places
 Truncate – truncate to the specified number of decimal places
 Prepend – prepend the specified string
 Append – append the specified string
 Truncate – truncate the string to the specified width
Sets the unit for all of the selected measurement values, overriding the
Set Measurement Unit existing unit, if any. Unit is chosen from the cascading menu of applicable
units based on the dimension of the mapped destination.
Adds the unit to just those selected measurement values that do not
Add Missing Unit already have a unit. Unit is chosen from the cascading menu of applicable
units based on the dimension of the mapped destination.
Converts the selected measurement values (which must already have a
Convert to Unit unit) to the target unit. Unit is chosen from the cascading menu of
applicable units based on the dimension of the mapped destination.
Normalizes the selected text values. Normalization returns the capitalized
Normalize alphanumeric string from which all punctuation other than the underscore
(_) and other “unwanted” characters have been removed.
Changes the case of the selected text values:
 UPPER CASE
Change Case  lower case
 Sentence case
 Title Case
 tOGGLE cASE

Replace… Finds and replaces strings in the selected text values.

Accept Accepts the selected text values that were truncated because they were
Truncated Value too long for the width of the mapped destination text field.

Expands the year range for the selected text values into multiple year
Expand Year Range values. For example, the single value 1997-94 expands to the four distinct
values 1994, 1995, 1996, and 1997.
Expands the month qualifier for the selected text values into multiple month
Expand Month* values. For example, the value 1999,6-2001,8 expands to June 1999,
NULL, and August 2001.

* Value-level filter only.

240 MDM Import Manager Reference Guide

LOOKUPS CONSISTING OF MULTIPLE DISPLAY FIELDS
Recall that you must value map from a source field to the values of a
destination lookup field, as shown in Figure 158.

Source: Destination:
NAME ► Name [lookup]
Joe Williams ► Joseph Williams
Jim Williams ► James Williams
Dave Rodgers ► David Rodgers
Dave Smith ► David Smith

Figure 158. Simple value mapping to destination lookup field

Compound Lookup Fields

When the underlying lookup table has just a single display field, each
lookup field value is simply the value of that display field.
However, when the lookup table has multiple display fields, the lookup
field is in effect a compound field, and each lookup field value is the
combination of the values of the display fields, as shown in Figure 159.

Destination: Lookup display fields:

Name [lookup] First Name Last Name
Joseph Williams ◄ Joseph Williams
James Williams ◄ James Williams
David Rodgers ◄ David Rodgers
David Smith ◄ David Smith

Figure 159. Corresponding lookup table with multiple display fields

NOTE ►► Normally, when a single source field corresponds to

multiple destination fields, you must combine the destination fields with
partitioning prior to field and value mapping. However, partitioning is
not necessary with a compound lookup field, which automatically
combines its underlying lookup table’s multiple display fields.

NOTE ►► The Add Value commands for the Destination Values grid
are always disabled when the destination is a compound lookup field.

MDM Import Manager Reference Guide 241

Mapping From a Single Corresponding Source Field
When just a single source field corresponds to a compound lookup field,
you have no choice but to field and value map the source field directly to
the compound lookup field. For value mapping purposes, the compound
lookup field then behaves somewhat like an MDM multi-valued field, in
that both consist of multiple values for each record.
Consequently, when you map a single source field to a compound
destination lookup field, MDM uses the Multi-Valued delimiter to
separate each source data value into the multiple delimited values
corresponding to the value combinations of the compound lookup field.
NOTE ►► The edit controls in the Converted Value column of the
Source Values grid are disabled and cannot be opened for editing any
of the individual delimited values.

NOTE ►► The Add button between the Value Mapping grids and the
Add Value commands for the Destination Values grid are also both
disabled.

TIP ►► If the source value combinations represented by the single

source field do not all correspond to existing values of the destination
lookup field, you must use the Split into Multiple Fields command to
split the delimited source values into multiple source fields, and then
map them to the compound lookup field, described in the next section.

NOTE ►► Whereas a multi-valued field stores multiple values in a

single field, a compound lookup field corresponds to a multi-field value
combination across each of the multiple lookup table display fields.

Mapping From Multiple Corresponding Source Fields

When multiple source fields correspond to a compound lookup field –
and each of the source fields corresponds to one of the individual
display fields of the underlying lookup table – the MDM Import Manager
offers a nested, two-step, “map-within-a-map” mechanism that allows
you to more efficiently map source values to the value combinations
represented by the compound lookup field.
Consider Figure 160, in which multiple source fields correspond to a
compound lookup field. Specifically: (1) multiple source fields (FIRST
NAME + LAST NAME) together correspond to a destination lookup field
(Name); (2) the compound lookup field represents the combination of
the underlying lookup table’s multiple display fields (First Name + Last
Name); and (3) each source field corresponds to one of the display
fields (FIRST NAME = First Name and LAST NAME = Last Name).

242 MDM Import Manager Reference Guide

 Source: Destination: Lookup display fields:
FIRST NAME LAST NAME ► Name [lookup] First Name Last Name
Joe Williams Joseph Williams ◄ Joseph Williams
Jim Williams James Williams ◄ James Williams
Dave Rodgers David Rodgers ◄ David Rodgers
Dave Smith David Smith ◄ David Smith

Figure 160. Multiple source fields and multiple display fields

Promoted Lookup Table Display Fields

To facilitate map-within-a-map to a compound lookup field, each of the
individual display fields of its underlying lookup table is automatically
promoted and appears in the Destination Fields grid, just as if it were a
field of the destination table, and its individual values appear in the
Destination Values grid (see “Value Mapping and Lookup Table Display
Fields” on page 246).
NOTE ►► Promoted lookup table display fields appear in the
Destination Fields grid as “display field n [lookup field]” and compound
lookup fields appear as “lookup field [display field 1; display field n]”
(where “display field n” is the name of the nth lookup table display field
and “lookup field” is the name of the compound lookup field).

Map-Within-a-Map (Create Compound Field Command)

Assuming the one-to-one correspondence between source fields and
promoted display fields, you can then perform the map-within-a-map as
follows: (1) map each source field to one of the display fields; (2) for
each source field, convert its values, or optionally map to the values of a
Text display field; (3) use the Create Compound Field command to
create a compound source field for mapping to the compound
destination lookup field; (4) map the compound source field to the
destination lookup field; and (5) automap the compound source field’s
source value combinations to the lookup field values.
NOTE ►► Normally, when multiple source fields correspond to a
single destination field, you must combine the source fields with
partitioning prior to field and value mapping. A combined partition
merges original source data values, which makes value mapping of the
value combinations more difficult. By contrast, the Create Compound
Field command merges converted or mapped values, so that validation
can be performed at the individual field level; moreover, Automap of
the compound source field can then have a 100% success rate (for
value combinations that already exist in the destination field) since the
components of the value combination have already been validated.

MDM Import Manager Reference Guide 243

Qualified Lookup Fields and Lookups-Within-Lookups
The map-within-a-map mechanism is especially useful – and its benefits
best illustrated – when the mapped destination field is a qualified lookup
field where: (1) the number of value combinations represented by the
records of the underlying qualified lookup table is very large; and (2) the
display fields are themselves nested lookups-within-lookups.
NOTE ►► With nested lookups, map-within-a-map corresponds
logically to the multi-level search-within-a-search offered by the MDM
Client, both of which use the nested lookups to restrict the larger set of
lookup table records.

Consider Figure 161, which lists some of the vehicles in the Vehicles
qualified table, which contains forty thousand vehicle specifications.

Vehicles
1996; Toyota; Celica
1997; Toyota; Celica
1998; Toyota; Celica
1999; Toyota; Celica
1996; Ford; Mustang
1997; Ford; Mustang
1997; Ford; Pinto
1998; Ford; Pinto
1998; Chevrolet; Impala
1999; Chevrolet; Impala
2000; Chevrolet; Impala
1997; Chevrolet; Caprice
1998; Chevrolet; Caprice
1999; Chevrolet; Caprice
2000; Chevrolet; Caprice

Figure 161. Records of a qualified table

The records above correspond to the sets of values for each of the
Vehicles table display / nested lookup fields shown in Figure 162.

Year [l] Model [l]

1996 Make [l] Celica
1997 Toyota Mustang
1998 Ford Pinto
1999 Chevrolet Impala
2000 Caprice

Figure 162. Nested lookup values of each qualified table display field

244 MDM Import Manager Reference Guide

Now, suppose the three source data fields YEAR, MAKE, and MODEL
have the sets of unnormalized values shown in Figure 163.

MAKE
YEAR Toyota
2000-1998 Toyota MODEL
97-98 Ford Celica
2000-1997 Ford Motor Mustang
97-96 Chevy Pinto
1996-1998 Chevrolet Impala
1999 Caprice

Figure 163. Source data fields and values

Using a combined partition to merge the values of YEAR, MAKE, and

MODEL as a prelude to value mapping to the lookup values of the
Vehicles table would combine their unmapped values, and in the
process create multiple instances of each unnormalized value across
the set of value combinations, as shown in Figure 164. As a result,
automapping the value combinations would be impossible, and
manually mapping them against the lookup values corresponding to the
forty thousand records of the Vehicles table exceptionally tedious.

YEAR; MAKE; MODEL

2000-1998; Chevy; Impala
97-98; Ford Motor; Pinto
2000-1997; Chevrolet; Caprice
97-96; Ford; Mustang
1996-1998; Toyota; Celica
1999; Toyota; Celica

Figure 164. Combined partition of source data values

This is where map-within-a-map comes into play. Rather than partition,

you can first field and value map each of the source data fields to the
corresponding qualified table lookup fields, each of which contains a
much smaller number of distinct values than the qualified table itself, so
that value mapping is much simpler. For example, YEAR maps to Year,
and the source value 2000-1998 maps to the multiple values 1998,
1999, and 2000, while MAKE maps to Make, and the two source values
Chevy and Chevrolet map to Chevrolet.
NOTE ►► Each promoted lookup table display field that is itself a
nested lookup field allows multi-valued mapping, regardless of whether
it is itself multi-valued.

MDM Import Manager Reference Guide 245

After the source data values have been mapped, you can then use the
Create Compound Field command to combine the normalized mapped
values (rather than the unnormalized original values) into a single
combination value that corresponds to the multiple display fields of each
record of the qualified table.
NOTE ►► Map-within-a-map allows you to normalize the source data
values against the nested lookup values before the compound field
creates multiple instances of each value, and then combine not the
original source values but rather the mapped values, so that the strings
of the combined values precisely correspond to the display fields of the
qualified table, for an Automap that will have a 100% success rate.

In the automotive example above, the benefit of map-within-a-map over

partitioning is more than just the convenience of being able to Automap
rather than manually map source value combinations against a large set
of destination values. Rather, map-within-a-map is essential when some
of the individual source values themselves must be mapped to multiple
destination values, so that map-within-a-map actually expands the
number of source value combinations, sometimes dramatically.
Finally, consider Figure 165, another set of automotive data, in which
not only the YEAR but also the MODEL must be expanded into multiple
values.

YEAR MAKE MODEL

1997-94 Chevy ALL
1996-93 Olds Cutlass
1998-97 Chevy Camaro
2000-1997 Caddy Eldorado
1994 Buick ALL

Figure 165. Map-within-a-map with value expansion

In the first source record, the YEAR would be value mapped to 1994,
1995, 1996, and 1997, and the MODEL mapped to all Chevrolet
models, and the value combination represented by the single record
automatically expanded to perhaps tens or even hundreds of value
combinations for automapping.

Value Mapping and Lookup Table Display Fields

Recall that a lookup table is a valid table, its display field or display field
combination defines the legal values of the corresponding lookup field
into the table, and value mapping to destination lookup fields is
therefore required to enforce source data value validation against the
set of legal destination values represented by the lookup table records.
In addition, value mapping is optional to fixed-width Text fields that are:
(1) promoted lookup table display fields when the current table is the

246 MDM Import Manager Reference Guide

main table and the lookup table has multiple display fields; (2) the
primary display field when the current table is a hierarchy lookup table;
and (3) actual display fields when the current table is the lookup table.
While not absolutely necessary for source data value validation, value
mapping to Text display field values affords less opportunity for data
entry and editing error and supports more consistent normalization than
simple value conversion. In particular, value mapping: (1) saves typing
when normalizing against existing destination values; and (2) requires
explicit action and precise definition when adding new values.
NOTE ►► When the subtable is not a lookup table (such as a
subtable of manufacturer part numbers and part information), the table
is not a valid table, no validity checking against the display field or
display field combination is necessary, and hence value mapping to
the Text display fields is neither required nor permitted.

When the lookup field and corresponding lookup table consist of

multiple display fields, an additional objective is to ensure the validity of
source data values not only against the individual values of each of the
display fields, but also against the value combinations represented by
the display fields taken together, as follows:
• Main table. When the current table is the main table, validate source
values against: (1) the individual display field values by permitting
optional value mapping to each of the promoted Text display fields;
and (2) display field value combinations by requiring value mapping
of the compound source field created with Create Compound Field.
• Lookup table. When the current table is a lookup table, validate
source values against: (1) the individual display field values by
permitting optional value mapping to each of the Text display fields;
and (2) display field value combinations by requiring record matching
against the display fields as a matching field combination.
NOTE ►► When the lookup table is a hierarchy table, value mapping
is optional to its primary display field (which appears as a hierarchy
value) and also to its other display fields (which appear as flat values);
when the hierarchy lookup table is the current table, record matching is
required only to its primary display field and/or the [Remote Key] field
(see “Lookup Table Imports Using the [Remote Key] Field” on page
414 for more information).

For example, consider a main table with a Manufacturer lookup field into
the Manufacturers lookup table, which has two display fields: (1) Name
(a fixed-width Text field); and (2) Code (an Integer field).

In this case, you can optionally value map to the Name field; when the
current table is the main table, value mapping is required from the
compound source field to Manufacturer, and when the current table is

MDM Import Manager Reference Guide 247

the Manufacturers table, record matching is required against Name and
Code as a matching field combination.
If the Manufacturers lookup table is a hierarchy table, then value
mapping would still be optional to the Name field as its primary display
field, and record matching would be required to the Name field alone
rather than with Code and/or to the [Remote Key] field (see “Lookup
Table Imports Using the [Remote Key] Field” on page 414 for more
information).
The value mapping rules are summarized in Table 42.

Table 42. Value Mapping Rules Summary

Current Table Single Lookup Display Field Multiple Lookup Display Fields
Value map required from individual Value map: (1) optional from individual
source field to lookup field (which is source fields to promoted Text lookup table
the lookup table display field). display fields; (2) optional to promoted
Main
hierarchy table primary display field; and
Table (3) required from compound source field to
compound lookup field (which is the lookup
table display field combination).
Value map optional from individual Value map optional from individual source
source field to hierarchy table field to hierarchy table primary display field
Hierarchy display field (record match required (record match required against hierarchy
Lookup Table against hierarchy table display field). table primary display field and/or the
[Remote Key] field).
Value map optional from individual Value map optional from individual source
source field to Text lookup table fields to Text lookup table display fields
Any
display field (record match required (record match required against lookup
Lookup Table against lookup table display field). table display fields as matching field
combination).

NOTE ►► Current table display fields appear in the Destination Fields

grid as “display field [DF]”, promoted lookup table display fields appear
as “display field n [lookup field]”, and compound lookup fields appear
as “lookup field [display field 1; display field n]”; (where “display field” is
the name of a current table display field, “display field n” is the name of
the nth lookup table display field, and “lookup field” is the name of the
compound lookup field).

NOTE ►► See “Record Matching and Lookup Table Display Fields”

on page 345 for more information about record matching.

NOTE ►► Field mapping to promoted lookup table display fields and

creating a compound source field for mapping to the compound lookup
field is possible only if the source data has individual fields that
correspond to the individual display fields of the lookup table.
Otherwise, you must simply field and value map from the applicable
source field to the compound lookup field.

248 MDM Import Manager Reference Guide

 NOTE ►► When importing to the primary display field of a hierarchy,
value mapping allows you to: (1) map source values to the correct
hierarchy value (e.g. source value “C” to destination value “A>B>C”
rather than “D>E>C”); and (2) add new nodes at specific locations in
the hierarchy (without value mapping, new nodes would simply be
added by default as sibling children of the root).

NOTE ►► When the current table is the hierarchy table, value

mapping to the primary display field is optional because: (1) value
mapping would be an unnecessary overhead if you don’t need to
resolve ambiguities or add new nodes (as described above); and (2)
when you value map, you cannot record match on just the [Remote
Key] field, which must instead be combined with the primary display
field; record matching on the [Remote Key] field alone makes it
convenient to populate values of other language layers.

What’s new and ramifications for import are summarized in Table 43.

Table 43. What's New

Item Description
Support for source hierarchy data now includes support for a pair of fields
Create Hierarchy containing parent/child values, to complement previous support for
command delimited values within a single field and values in multiple fields.
The existing NULL Handling option became the Flat NULL Handling
option with two settings for dealing exclusively with delimited flat values,
Hierarchy NULL
and a second Hierarchy option added to deal with hierarchy values. The
Handling option new option has three settings; the third setting is designed for importing
into the hierarchy table itself; which was previously not fully supported.
Value mapping to the primary display field of the hierarchy when the
Internal node value current table is a hierarchy table now permits (and requires) mapping of
internal nodes as well as leaf nodes. Internal node value mapping allows
mapping to primary
you not only to import a complete hierarchy consisting of internal and leaf
display field on nodes, but also to populate for internal nodes: (1) the other fields of the
hierarchy table import hierarchy table records (i.e. fields other than the primary display field); (2)
other language layers of all of the fields; and (3) the [Remote Key] field.

Optional value mapping Value mapping to the primary display field of a hierarchy when the current
table is a hierarchy table is now optional rather than required. Not value
to primary display field
mapping avoids an unnecessary overhead if you don’t need to resolve
on hierarchy table ambiguities or add new nodes, and also allows you to record match on
import the [Remote Key] field by itself (see below).
You can now use the [Remote Key] field as a matching field either by
itself or in conjunction with the display field(s) of a flat table or the primary
[Remote Key] record display field of a hierarchy table for record matching. In previous versions,
matching only the display field(s) could be used. Record matching on the [Remote
Key] field by itself makes it convenient to populate values of other
language layers.
When importing records to a hierarchy table, internal node records were
ignored during import because value mapping was required but did not
permit mapping to internal nodes (error message about no value map
posted to the log file). Now that value mapping is optional and internal
Internal node record
node mapping supported, both internal- and leaf-node records are
matching imported rather than just the leaf-node records (whether or not you value
map) because: (1) if you do not value map, the match occurs through
record matching; and (2) if you do value map, the match occurs through
value mapping.

MDM Import Manager Reference Guide 249

MULTI-VALUED DESTINATION FIELDS
Recall that many of the MDM field and attribute types can be defined as
multi-valued; that is, a single field in a single record can be used to store
multiple data values.

Field Mapping to Multi-Valued Fields

When you map a source field to a multi-valued destination field, the
drop-down edit controls in the Converted Value column of the Source
Values grid are multi-valued, as shown in Figure 166.

Figure 166. Multi-valued drop-down control

You can then map and convert values as follows: (1) whether or not the
mapped destination field requires value mapping, you can use the multi-
valued drop-down control to edit or create one or more values for each
source value; or (2) if the destination field requires value mapping, you
can map each source data value to one or more destination values even
without manually expanding it into multiple values.
TIP ►► You can use the multi-valued drop-down control to either
create or map to multiple values regardless of whether or not the
source data value consists of multiple delimited values.

Field Mapping to Coupled Numeric Attributes

When you a map a source field to a coupled numeric attribute (which is
always multi-valued), the MDM Import Manager automatically breaks
each coupled numeric value pair in the source data into a primary value
and a coupled value based on the setting of the Coupled Numeric
delimiter string.
When you edit a source value, the drop-down edit control in the
Converted Value column of the Source Values grid is a modified version
of the multi-valued drop-down edit control that allows you to edit each
value, as shown in Figure 167.

250 MDM Import Manager Reference Guide

 Figure 167. Coupled numeric drop-down control

When you apply a value conversion filter, the command for each filter
leads to a cascading menu that lets you apply the filter to each value, as
shown in Figure 168.

Figure 168. Cascading menu for a coupled numeric attribute

The Multi-Valued Update Option

When creating new destination records or replacing existing destination
records, importing source data values into multi-valued fields is exactly
the same as importing them into single-valued fields: for each source
record, the one or more values are placed into the empty multi-valued
field of the new destination record.
By contrast, when updating existing destination records as part of the
import – and in particular when updating non-NULL multi-valued fields
that already have one or more data values – the update of multi-valued
fields can be handled in one of two ways.
Specifically, the set of incoming source data values can either: (1) be
appended to; or (2) completely replace the set of existing destination
data values, as described in Table 44.

MDM Import Manager Reference Guide 251

Table 44. The Multi-Valued Update Option
Setting Description
The set of incoming source data values is appended to the set of
Append existing destination values, so that the field in the updated record
contains both the new and the existing data values.
The set of incoming source data values completely replaces the set
Replace of existing destination values already in the multi-valued field, so that
the field in the updated record contains only the new data values.

NOTE ►► Among the pair of update import actions, the Multi-Valued

Update option only affects the Update (All Mapped Fields) action,
because the Update (NULL Fields Only) action updates only fields that
do not yet have data values, making the Append and Replace settings
equivalent. (See “Import Actions” on page 346 for more information
about the import actions of the Record Matching tab.)

For each multi-valued destination field or attribute, you can set the Multi-
Valued Update option (see “Setting the Multi-Valued Update Option” on
page 286) based on the requirements of the situation.
NOTE ►► See “Source Field-Level Settings” on page 224 for more
information about the field-level settings for source fields.

Table 45 illustrates the two alternative Multi-Valued Update settings for

a multi-valued field with different sets of existing data values and the
same set of incoming delimited source values.

Table 45. Multi-Valued Update Examples

Incoming Source Existing Destination Updated Destination Data Values
Data Values Data Values Append Replace
<NULL> Green; Blue Green; Blue
Green Green; Blue Green; Blue
Blue Green; Blue Green; Blue
Green; Blue
Green; Blue Green; Blue Green; Blue
Red Red; Green; Blue Green; Blue
Red; Green; Blue Red; Green; Blue Green; Blue

NOTE ►► Append and Replace differ only when the set of existing
destination values includes values not in the set of incoming source
values.

252 MDM Import Manager Reference Guide

The Qualified Update Option
Qualified lookup fields pose an additional challenge as it relates to multi-
valued destination fields. Specifically, with qualified lookup fields (which
are always multi-valued), each of the multiple values is not just a single
data value, but rather an entire subrecord consisting of a set of values
for both the display fields and the qualifiers of the qualified table record.
NOTE ►► Moreover, the source data file will usually contain not one
record per product but multiple records per product. Each source
record then represents a single value for the multi-valued qualified
lookup field (consisting of a set of values both for the display fields of
the qualified table record and for the qualifiers) and the multiple source
records for each product are treated as multiple instances of the same
product, each with a different “value” for the qualified lookup field.

With qualified lookup fields, the Append and Replace settings of the
Qualified Update option have precisely the same meaning and behavior
as they do for the Multi-Valued Update option with normal multi-valued
fields and attributes. Specifically:
• Append. Appends to the set of existing qualified lookup values
(preserving and appending to the set of existing links to the qualified
table records).
NOTE ►► To determine whether to append a new qualified lookup
value, MDM compares the data values of an incoming source record
against both the values of the display fields of the linked qualified table
record and all of the qualifiers taken together as a single “monolithic”
value combination.

• Replace. Completely replaces the set of existing qualified lookup

values (eliminating the existing links to the qualified table records
and discarding the qualifier values associated with those links).

This monolithic “all qualifiers” approach to comparing incoming and

existing values is sufficient and appropriate in some circumstances
(such as those described in the next section) but it provides no ability to
update individual qualifier values of existing qualified lookup values
rather than replace or append to the set of qualified lookup values.
For those cases where the comparison should instead be made against
some subset of rather than all the qualifier values (to permit selective
update of the qualifier values of existing qualified links), the Qualified
Update option offers – in addition to the Append and Replace settings –
a third setting that is enabled only for qualified lookup fields.
Specifically, the Update setting allows you to specify which particular
qualifiers MDM should use as matching qualifiers for matching
purposes, and then how to update the values of the remaining qualifiers
of existing qualified lookup values, as described in Table 46.

MDM Import Manager Reference Guide 253

Table 46. The Qualified Update Option for Qualified Lookup Fields
Setting Description
Comparing against all of the qualifiers, incoming source data value
Append combinations that do not already exist are appended to the set of
existing qualified lookup values.
With no comparison against the qualifiers whatsoever, the set of
Replace incoming source data value combinations completely replaces the
set of existing qualified lookup values.
Comparing against just the matching qualifiers, incoming source data
value combinations that already exist update the remaining qualifiers
of existing qualified lookup values. Update actions include:
New links (no matching qualified lookup value)
 Skip
 Create
Update
Existing links (matching qualified lookup value)
 Skip
 Update (NULL Qualifiers Only)
 Update (All Mapped Qualifiers)
 Replace
 Delete

You can specify a different Update action for source value combinations
that do match and those that do not match an existing qualified lookup
value based on the matching qualifiers, as described in Table 47.

Table 47. Update Actions for Qualified Lookup Values

Update Action Description
New Links
Skip Skip source value combinations that do not exist.
Create a new link to a qualified lookup table record and
Create
set the associated qualifier values.
Existing Links
Skip Skip source value combinations that already exist.
Update just the NULL remaining qualifiers of each
Update (NULL Qualifiers Only)
existing qualified lookup value.
Update all the remaining qualifiers of each existing
Update (All Mapped Qualifiers)
qualified lookup value.
Replace all the remaining qualifiers of each existing
Replace qualified lookup value by first deleting them and then
creating a new value combination.
Delete this qualified lookup value during import
Delete
processing

NOTE ►► The Update setting provides a “mini” record-match-within-

a-record-match for the subrecords represented by each qualified
lookup value. (See “Import Actions” on page 346 for more information
about the import actions of the Record Matching tab.)

254 MDM Import Manager Reference Guide

For example, assume a main table of product records that includes a
qualified lookup field into the Distributors qualified table, shown in
Figure 169. The Distributors table contains a list of three distributors,
and is used to store distributor-specific part numbers and pricing for
each product. Each qualified table record contains a distributor name in
the Name field; the table also contains qualifiers Dist No and Price
(shown in square brackets), which have values only when a qualified
table record is assigned to the value of a main table qualified lookup
field.

Distributors:
Name [Dist No] [Price]
Grainger
Applied
Wesco

Figure 169. Qualified table records

Now consider Figure 170, which illustrates two incoming source records
(each containing the distributor-specific part number and price for
SKU=113) and the corresponding destination record for the same
product that has existing links to the three distributors and their
associated distributor-specific part number and price values.

Source: Destination:
SKU DISTRIB DIST NO PRICE ► SKU Distrib [l] [Dist No] [Price]
113 Wesco 444-201 $3.50 113 Wesco 444-101 $3.52
113 Applied AP-201 $3.49 Grainger X437-5 $3.55
Applied AP-201 $3.65

Figure 170. Incoming distributor-specific information

Note that the two source records contain part number and price values
for two of the three distributors, for which the destination record already
contains qualified links and qualifier values.
Unfortunately, neither Append nor Replace is flexible enough to handle
this simple import, since neither allows you to simultaneously leave
some existing qualified links untouched, update other existing links with
new source data, and add new qualified links. In particular:
• Append will correctly preserve the link to Grainger, but will create
second, duplicate links to Wesco and Applied rather than update the
values associated with the existing links.
• Replace will update the links to Wesco and Applied with new part
number and price values, but will also wipe out the link and
associated values for Grainger.

MDM Import Manager Reference Guide 255

However, with the Update setting, you can specify matching qualifiers
for comparing against existing links, as well as what Update action to
take for new and existing qualifier value combinations.
Figure 171 illustrates the different effects of the three Qualified Update
settings on the resulting set of qualified links to the product record for
SKU=113 when incoming source fields are mapped to destination field
Distrib and to destination qualifiers Dist No and Price.

Destination: Destination:
SKU Distrib [l] [Dist No] [Price] SKU Distrib [l] [Dist No] [Price]
113 Wesco 444-101 $3.52 113 Wesco 444-201 $3.50
Grainger X437-5 $3.55 Applied AP-201 $3.49
Applied AP-201 $3.65
Replace
Wesco 444-201 $3.50
Applied AP-201 $3.49

Append
Destination:
Destination: SKU Distrib [l] [Dist No] [Price]

SKU Distrib [l] [Dist No] [Price] 113 Wesco 444-101 $3.52
113 Wesco 444-201 $3.50 Grainger X437-5 $3.55
Grainger X437-5 $3.55 Applied AP-201 $3.49
Applied AP-201 $3.49 Wesco 444-201 $3.50

Update Update
Matching Qualifiers = None Matching Qualifiers = Dist No
New Links = Skip or Create New Links = Create
Existing Links = Update or Replace Existing Links = Update or Replace

Figure 171. Qualified Update settings for qualified lookup fields

NOTE ►► As shown above, which qualifiers you specify as matching

qualifiers affects the resulting set of qualified links. Moreover, the more
qualifiers you specify, the more rigorous the match to an existing
qualified link, and the fewer qualifier values you can update.

NOTE ►► The matching qualifiers specified not only for the Update
setting of the Qualified option but also for the Append and Replace
settings are also used by the Merge Qualified Links option when
merging qualified links, and moreover, affect the resulting set of
qualified links (see “Source Record Merging” on page 350).

TIP ►► You can tune the behavior of the Update setting of the
Qualified Update option not only by adjusting the set of matching
qualifiers but also by adjusting the particular Update actions for new
and existing links.

256 MDM Import Manager Reference Guide

Product Applications and Automotive Parts Data
With automotive parts data, the source data typically represents
application records rather than product records, and contains multiple
records per product, where each source record represents a single
application of the product (and is again treated as a product record that
represents a single value for the multi-valued qualified lookup field).
Consider the table of source application records shown in Figure 172.

PART NO YEAR MAKE MODEL CA EQUIP A/C P/B

A2-444 1998 Toyota Celica Yes Yes Yes
A2-444 1998 Toyota Celica No Yes Yes
A2-444 1996 Toyota Celica No Yes Yes
A2-444 1998 Ford Mustang No Yes Yes
A2-444 1997 Ford Mustang No No No
C6-859 1996 Toyota Celica Yes Yes Yes
C6-859 1996 Toyota Celica No Yes Yes
C6-859 1998 Ford Mustang Yes No Yes

Figure 172. Automotive data and source application records

In this case, using either the Replace setting of the Qualified Update
option, or alternatively, the Append setting to match against all of the
qualifiers, is appropriate.
NOTE ►► Recall that qualifiers are part of neither the main table
record nor the qualified table record, but rather are defined by each
association of a main table record with a qualified table record. When
using the Update setting, the qualifiers on which to match depends
upon whether the qualifier values logically represent an extension to:
(1) the main table record (such as with multiple pricing or alternate part
number data, as described in the previous section, in which case the
corresponding qualifiers should be specified as matching qualifiers); or
(2) the qualified lookup table record (such as with application records,
as described in this section, in which case they should not be specified
as matching qualifiers).

This table of application records contains information that corresponds

to the following destination fields and qualifiers:
• 1 field of the Parts main table (PART NO)
• 3 fields of the Vehicles qualified table (YEAR, MAKE, MODEL)
• 3 qualifiers of the Vehicles qualified table (CA EQUIP, A/C, P/B)

The records themselves correspond to the following records of the Parts

and Vehicles tables:
• 2 distinct part records (A2-444; C6-859)

MDM Import Manager Reference Guide 257

• 4 distinct vehicle records (4 for A2-444; 2 for C6-859)
• 8 distinct application records (5 for A2-444; 3 for C6-859)
NOTE ►► The same vehicle can be linked to each part more than
once, with different qualifiers for each link (two links to 1998 Toyota
Celica for A2-444; two links to 1996 Toyota Celica for C6-859).

This single table of source data illustrates many of the challenges of

dealing with complex source data, and a methodology for doing so
using the innovative mechanisms of the MDM Import Manager:
• Field mapping. The PART NO field is field mapped to the Part No.
field of the main table, and the CA EQUIP, A/C, and P/B fields are
field mapped to the qualifiers of the Vehicles table.
• Map-within-a-map. Since nested display fields are automatically
promoted, the YEAR, MAKE, and MODEL fields are field mapped to
the Year, Make, and Model fields of the Vehicles table.
• Nested lookups-within-lookups. Since Year, Make, and Model are
nested lookup fields, YEAR, MAKE, and MODEL are value mapped
to the values of the Year, Make, and Model fields (see page 243).
• Create Compound Field command. Once YEAR, MAKE, and
MODEL are field and value mapped, the Create Compound Field
command creates the qualified lookup field values.
• Qualified lookup fields and Automap. The compound source field
is mapped to the qualified lookup field, and the combined values
automapped to its values; if necessary, new records are added to the
Vehicles table.
• Qualified Update option. The Qualified Update option is set to
Append, so the new application records are appended to the existing
qualified lookup field values.
NOTE ►► To determine whether an incoming application already
exists, it is treated monolithically for the purposes of the comparison as
consisting of all of the display field values of the qualified table record
and all of the qualifier values for that particular application.

258 MDM Import Manager Reference Guide

The Tuple Update Option
As with qualified lookup fields, a tuple field does not represent a single
data value, but rather an entire subrecord (or set of subrecords) that
belongs exclusively to its parent record. The value of this subrecord is
the combined values of the tuple’s member fields, including both its
display and non-display fields.
When importing records into a multi-valued tuple field, Import Manager
first checks the tuple’s matching fields to see whether a matching tuple
record already exists in the repository. Import Manager’s default
behavior is to use all of a tuple’s member fields as matching fields, but
you can choose a subset of matching fields on a tuple-by-tuple basis.
Once Import Manager determines whether or not the source record
represents a new or existing tuple record, it then performs the
corresponding update action specified for the tuple field.
You can use the Tuple Update command to specify a different update
action for source records that do match or do not match an existing
tuple record based on the tuple field’s matching fields, as described in
Table 48.

Table 48. Update Actions for Tuple Fields

Update Action Description
New Records (If no matching tuple record is found)
Skip Do not import the incoming tuple record.
Create Add the incoming tuple record.
Existing Records ( a matching tuple record is found)
Skip Do not import the incoming tuple record.
Update (NULL Fields Only) Update only destination fields which have NULL values.
Update (All Mapped Fields) Update the values of all mapped destination fields.
Delete the entire existing tuple record and replace it
Replace
with the incoming tuple record.
Delete Delete the entire existing tuple record.

NOTE ►► The Tuple Update command provides a “mini” record-

match-within-a-record-match for the tuple records represented by each
tuple field value. (See “Record Matching Concepts” on page 338 for
more information about record matching.)

CAUTION ►► The Replace and Delete actions permanently delete the

entire existing tuple record from the repository, including: (1) the values
of any unmapped destination fields on the matching tuple record; and
(2) any nested tuple records belonging to the matching tuple record.

MDM Import Manager Reference Guide 259

When setting update actions for nested tuple fields, each multi-valued
tuple field can have a update action different than its parent or child
tuple field in the tuple hierarchy. This lets you customize update
behavior on a tuple-by-tuple basis, rather than forcing you to use a
monolithic, one-size-fits-all approach.
NOTE ►► The Tuple Update option is disabled for single-valued tuple
fields nested within in a tuple structure. Instead, Import Manager
applies the update option set for its parent tuple in the tuple structure.

An important factor to consider when choosing update actions for

nested tuple fields is that Import Manager updates tuple records in a
“top-down” manner, starting with the tuple field values belonging to the
current table record and then progressing to each subsequent tuple field
in the tuple hierarchy.
Consider Figure 173, which illustrates a Customers table containing a
nested tuple hierarchy (Contact  Phone), where:
• The Customers table includes the multi-valued Contact tuple field.
• The Contact tuple includes the multi-valued Phone tuple field.

Customers

├ Id
├ Company
├ Contact [tuple]
├ Name
└ Phone [tuple]
├ Area Code
├ Number
└ Extension

Figure 173. A Nested Tuple Hierarchy in a Customers table

Figure 174 shows the same Customers table in tabular format, with
sample records aded.

Customers:
Contact [tuple]
Id Company Phone [tuple]
Name
Area Code Number Ext.
800 5551212 300
John Smith
1 Acme 310 777-1234
Mary Hsu 212 444-1234
2 LittleCo Susan Gray 808 333-3333

Figure 174. Nested Tuples in a Customers Table

260 MDM Import Manager Reference Guide

In Figure 174, notice that the Customers table has a single record for
the company Acme, and that this Acme record has two private
subrecords in its Contact tuple: one for John Smith and one for Mary
Hsu. The John Smith tuple record owns two private Phone tuple
records; the Mary Hsu tuple record owns one.
Figure 175 represents a source record to import into the Acme record’s
Contacts tuple (assume all fields on the source file are mapped to their
counterparts on the Customers table):

Source:
Company Name Area Code Number Ext
Acme John Smith 310 777-1234 5

Figure 175. Source Records to Import

Assume also that the source record is imported using the tuple update
actions specified in Figure 176.

Tuple Matching Fields Update Action

New Record: Create

Contact Name
Existing Record: Replace

Area Code; New Record: Create

Phone
Number Existing Record: Update (NULL Fields Only)

Figure 176. Contact and Phone Tuple Update Actions

To import this source record, Import Manager performs the following

actions, in order:
1. Searches for the Acme record on the Customers table.
2. Performs the default import action specified for the record match
type in Import Manager’s Match Records tab.
3. In the Acme record, searches for Contact tuple records with the
Name value, “John Smith”.
4. Based on the results of the search, peforms the corresponding
tuple update action specified for the Contact tuple.
5. In the resulting “John Smith” Contact tuple records, searches for
Phone tuple records containing the values “310” and “777-1234” in
the Area Code and Number fields.
6. Based on the results of the search, performs the tuple update
action specified for the Phone tuple.

Note that the import/update action performed after each matching step
determines whether or not Import Manager continues to the next step.

MDM Import Manager Reference Guide 261

Based on the tuple hierarchy in the Customers table, the Contact tuple
field always gets updated prior to the Phone tuple. The update actions
selected for the Contact tuple field, therefore, affect not only the Contact
tuple records but also any Phone tuple records they might own.
This is reflected in the results of the import, shown in Figure 177.

Customers:
Contact
Id Company Phone
Name
Area Code Number Ext.
John Smith 310 777-1234 5
1 Acme
Mary Hsu 212 444-1234 301
2 LittleCo Susan Gray 808 333-3333

Figure 177. After Importing Source Records

As a result of the Replace action being the Contact tuple’s update option
for existing records (Figure 176), when Import Manager matches the
source record’s Name field value “John Smith” to the Contact tuple
record with the Name value “John Smith”, it deletes the existing John
Smith tuple record, including all of the Phone tuple records it owned,
and replaces it with the imported record values.
These phone tuple records are deleted even though the Phone tuple
action is set to Update (NULL Fields Only). Because the Phone tuple
field belongs to the Contact tuple, all of the phone records for John
Smith got deleted with the John Smith Contact tuple record before the
Phone tuple’s update options come into play.
Finally, depending on the tuple update options selected, you can update
a nested tuple record without having to map any member fields of
“higher-up” tuples in the tuple hierarchy (although you always have to
map least one field on the current table).
Consider Figure 178, which represents a new source record to import
into the Customers table.

Source:
Company Area Code Number Ext
BigCorp 404 987-1234

Figure 178. Source Records to Import

Assume that the source fields are mapped to their destination

counterparts, and that the source record is imported using the tuple
update actions specified in Figure 179.

262 MDM Import Manager Reference Guide

 Tuple Matching Fields Update Action

New Record: Create

Contact Name
Existing Record: Replace

Area Code; New Record: Create

Phone
Number Existing Record: Update (NULL Fields Only)

Figure 179. Contact and Phone Tuple Update Actions

The result of the import is shown in Figure 180.

Customers:
Contact
Id Company Phone
Name
Area Code Number Ext.
BigCorp <NULL> 404 987-1234

Figure 180. After Importing Source Records

Note that a new Contact tuple record is added to the BigCorp customer
record, even though no Contact tuple member fields were mapped. To
preserve the structure of the tuple hierarchy, Import Manager:
1. Created a “placeholder” Contact tuple record, inserting NULL
values into the unmapped Contact tuple member field, “Name”.
2. Imported the source record values as a Phone tuple record
belonging to the NULL Contact tuple record.

Import Manager creates as many nested placeholder records as are

required to maintain a tuple hierarchy structure. That is, if you map only
the member fields of the sixth tuple in a six-tuple hierarchy, Import
Manager creates and nests the missing five tuple records. You can then
overwrite the NULL values in these records ater via Data Manager or
another import.
Of course, the result of any import to a nested tuple record depends on
the tuple update option settings for both the destination tuple field and
for all of the tuple fields “above” it in the tuple hierarchy.
For example, assume the Contact tuple field’s New Record update
action was set to Skip instead of Create, as shown in Figure 181.

MDM Import Manager Reference Guide 263

 Tuple Matching Fields Update Action

New Record: Skip

Contact Name
Existing Record: Replace

Area Code; New Record: Create

Phone
Number Existing Record: Update (NULL Fields Only)

Figure 181. Contact and Phone Tuple Update Actions

The result of the import is shown in Figure 182.

Customers:
Contact
Id Company Phone
Name
Area Code Number Ext.
BigCorp

Figure 182. After Importing Source Records

Note that now, no new Contact tuple record is added to the BigCorp
customer record, even though the Phone tuple’s New Record update
action remains set to Create. Because Import Manager imports from the
top to the bottom of a tuple hierarchy, when it cannot find a Contact
tuple record to match the source record, it performs the Contact tuple’s
New Record action and “skips” creating the new Contact tuple record. At
this point, any tuple fields below the Contact tuple in the tuple hierarchy
are also skipped.
Consider what happens instead if the Contact tuple field’s New Record
update action remains set to Create but the Phone tuple’s New record
update action is changed to Skip, as shown in Figure 183.

Tuple Matching Fields Update Action

New Record: Create

Contact Name
Existing Record: Replace

Area Code; New Record: Skip

Phone
Number Existing Record: Update (NULL Fields Only)

Figure 183. Contact and Phone Tuple Update Actions

264 MDM Import Manager Reference Guide

The result of the import is shown in Figure 184.

Customers:
Contact
Id Company Phone
Name
Area Code Number Ext.
BigCorp <NULL>

Figure 184. After Importing Source Records

Note that in this scenario, a new NULL Contact tuple record is added
even though no Phone tuple record was created. When importing the
source record values, Import Manager creates the new Contact tuple
record according the Contact tuple’s update actions, and then performs
the Phone tuple update action, which is set to skip the new Phone tuple
record. Because the Phone tuple is lower in the tuple hierarchy, its
update actions do not change the results of the Contact tuple record’s
update action.

MULTILINGUAL DESTINATION FIELDS

When you map a source field to a multilingual destination field, how you
map – and if and how you configure MDM to expand the destination
field – depends on the structure of the source data. (See “Field Mapping
to Multilingual Destination Fields” on page 397 for more information.)

OBJECT LOOKUP DESTINATION FIELDS

When you map a source field to an object lookup destination field, the
Converted Value column of the Source Values grid displays either: (1)
the object name (for images,PDFs, and binary objects); or (2) the object
value itself (for text blocks).
You can then map values in the Source Values grid to the objects of the
underlying object lookup table in the Destination Values grid, linking the
corresponding object to each source record as part of the import.

The Data Group Display Option

To distinguish among images, PDFs, and binary objects that have the
same name, or text blocks that have the same value, you can display
the Data Groups hierarchy in the Destination Values grid so that each
object name appears as a leaf node beneath its data group, as
illustrated in Figure 185 and described in Table 49.

MDM Import Manager Reference Guide 265

 Figure 185. Data Groups hierarchy in Destination Values grid

NOTE ►► Object names or values are displayed in black, while data

group nodes are highlighted in gray.

Table 49. The Data Group Display Option

Setting Description
The object names or values are displayed as a list in the
Hide Data Group Destination Values grid, making it impossible to
distinguish among identical names and values.
The object names or values are displayed as leaf nodes
Display Data Group beneath their data group in the Data Groups hierarchy in
the Destination Values grid.

For each object lookup destination field, you can set the Data” on page
289) based on the requirements of the situation.
NOTE ►► See “Destination Field-Level Settings” on page 225 for
more information about all of the destination field-level settings.

WRITEABLE ONCE DESTINATION FIELDS

Recall that a field can be designated in the MDM Console as Writeable
Once, so that its value can be set only once and never changed. This
can be a problem within the MDM Import Manager if you map a source
field to a Writeable Once destination field to populate missing values,
but some of the values of the destination field are already set.
Naturally, if the destination value for a record is empty, then MDM
properly sets the destination value with the source value without a
problem. Similarly, if the destination value for a record is set but is the
same as the source value, MDM avoids a potential problem by ignoring
the source value, even when the import action for the Record is Update.
The challenge is what to do with the import record if the destination
value is set and is different than the source value.

266 MDM Import Manager Reference Guide

 DATA INTEGRITY ►► The “safest” behavior would be to reject
source records that attempt to reset the value of a Writeable Once
field. However, the MDM Import Manager needs to be able to support
automation and high-volume imports without triggering multiple
exceptions, especially when ignoring conflicting values is not an issue.

You can use the Writeable Once Field Update configuration option to
specify how to handle imports when a source field is mapped to a
Writeable Once destination field and the source value is different than
the destination value, as summarized in Table 50.

Table 50. Writeable Once Field Update Configuration Option

Setting Description
Allow record import and ignore source value if it
Ignore Conflicting Value
conflicts with existing destination value.
Reject Record on Reject the record import if the source value
Conflicting Value conflicts with existing destination value.

NOTE ►► This option is relevant only for records where the

destination value is set and different than the source value. Further, if
the Writeable Once field is used as a record matching field, then this
option is relevant only when the Do Not Update Record Matching Field
configuration option is set to No (see “Record Filtering” on page 356 for
more information).

GM TIME FIELDS
When you map a source field to a destination GM Time field, the date
values must be in one of the following formats:
YYYYMMDD hh:mm:ss
MM/DD/YYYY hh:mm:ss
DD.MM.YYYY hh:mm:ss

NOTE ►►A space is required between the date and time values.

The format used should be the same as specified in the Configuration

Options -> Dates -> Date Format setting.

MDM Import Manager Reference Guide 267

THE MDIS UNMAPPED VALUE HANDLING OPTION
When you run MDIS, it handles the import using an existing map file,
which specifies a mapping from each source data value to each
destination data value.
When subsequent imports include source data values that were not
mapped when you created the map file, you can handle these
unmapped values in one of three ways.
Specifically, values that are unmapped either because they do not exist
or because Automap Unmapped Values was turned off can either be:
(1) ignored; (2) added to the set of destination data values; or (3)
mapped to existing destination data values, as described in Table 51.
Table 51. The MDIS Unmapped Value Handling Option
Setting Description
Still unmapped values are ignored, and a NULL inserted for
Ignore
the value of the destination field in the destination record.
Still unmapped values are added to the set of destination
Add
data values for the destination field.
Still unmapped values are mapped to the specified existing
Map
destination data values for the destination field.
Still unmapped values cause the import of the underlying
Fail
records that contain the value to fail

For each destination field, you can set the MDIS Unmapped Value
Handling option (see “Setting the MDIS Unmapped Value Handling
Option” on page 291) to specify the proper handling of unmapped
values during MDIS processing.
NOTE ►► You can use the global default value of the MDIS
Unmapped Value Handling option to Ignore or Add unmapped values,
but you must set the option on a destination field-by-field basis if you
want to Map to existing destination data values.

NOTE ►► See “Source Field-Level Settings” on page 224 for more

information about the field-level settings for source fields.

NOTE ►► MDIS processing is subject to the following considerations:

(1) all configuration regarding the MDIS import should be created using
the MDM Import Manager and saved in an import map; (2) the
Automap Unmapped Values configuration option allows you to specify
whether or not to attempt to automap unmapped values; (3) the Default
MDIS Unmapped Value Handling configuration option allows you to
specify whether to ignore or add unmapped values, or to map to an
existing destination data value.

268 MDM Import Manager Reference Guide

Field Mapping Operations
The following sections describe the operations you can perform on
fields in the Field Mapping grids, and are summarized in Table 52.
Table 52. Field Mapping Operations
Operation Description
Buttons
Automatically maps each source field to a destination field with
Automap
the same name.

Map Maps the selected source field to the selected destination field.

Unmap Unmaps the selected fields.

Add Adds a new attribute to the destination taxonomy.

Source Fields Grid

Find Next Matching Finds the first or next similarly destination field with a name
Destination Field similar to the selected source field.
Create Combines fields that have been mapped to the multiple display
Compound Field fields of a destination lookup field.
Set Field Mapping Sets the Multi-Valued or Coupled Numeric delimiter for the
Delimiter selected source fields.

Set Case Sensitivity Sets the Case Sensitivity option for the selected source fields.

Set Flat
Sets the Flat NULL Handling option for the selected source fields.
NULL Handling
Set Hierarchy Sets the Hierarchy NULL Handling option for the selected source
NULL Handling fields.
Set Value Sets the field-level value conversion filters for the selected source
Conversion Filter field.
Destination Fields Grid

Filter Attributes Sets the unlinked attribute filter for the destination fields grid.

Set NULL Sets the NULL Interpretation Option for the selected destination
Interpretation fields.
Set Multi-Valued Sets the Multi-Valued Update option for the selected multi-valued
Update destination fields.
Sets the Qualified Update option for the selected qualified lookup
Set Qualified Update
destination fields.
Sets the Qualified Update option for the selected tuple destination
Set Tuple Update
fields.
Set Data Group Sets the Data Group Display option for the selected object lookup
Display destination fields.
Set MDIS Unmapped Sets the MDIS Unmapped Value Handling option for the selected
Value Handling destination field.

MDM Import Manager Reference Guide 269

 NOTE ►► See “The NULL Handling Options” on page 231, “The
Case Sensitivity Option” on page 237, “Value Conversion Filters” on
page 238, “The NULL Interpretation Option” on page 236, “The Multi-
Valued Update Option” on page 251, “The Qualified Update Option” on
page 253, “The Data Group Display Option” on page 265, and “The
MDIS Unmapped Value Handling Option” on page 268 for more
information about these field-level settings.

SOURCE FIELD MAPPING

Mapping fields is extremely straightforward using the Field Mapping
grids, shown in Figure 187.

Figure 187. Field Mapping grids

For each source field or field combination whose data values you want
to import, the field mapping process includes some or all of the following
steps:
1. Select a source field.
2. Set the delimiter values.
3. Set the NULL handling.
4. Select the corresponding destination field, attribute, or qualifier.
5. Examine the set of source values and the set of corresponding
destination values to determine if the fields should truly be mapped.
6. If necessary, return to one of the earlier steps in the import
workflow to further split or partition the field and then attempt to
map again.
7. Map the source field to the destination field.

Each of the steps above is described in the following sections.

270 MDM Import Manager Reference Guide

MAPPING FIELDS
You can map an unmapped source field to an unmapped destination
field as described in this section. Corresponding source and destination
fields do not need to have the same name.

Field Mapping Grid Icons

To make the relationships between fields more understandable, the
Field Mapping grids now display source and destination fields in their
natural hierarchies. To accurately depict these hierarchies, Import
Manager includes both “structural” nodes (such as root nodes and fields
containing nested values) as well as data fields in the Field Mapping
grids. While you can map any data field from the Source Fields grid to
any data field on the Destination Fields grid, you cannot map structural
nodes on either grid.
Import Manager use icons to depict which items on the grids can and
can not be mapped, as described in Table 53.

Table 53. Field Mapping Grid Icons

Icon Description Mappable?
Field or Attribute Yes
Structural node No; map nested fields instead.

Source Fields
Each of the fields of the current source data table – along with each of
the fields that you looked up from other tables – appear in the Source
Fields grid.
NOTE ►► Fields that have been partitioned or any of whose values
have been partitioned appear as “field [Partition]”.

NOTE ►► Lookup fields may appear as “field <table>” (where “table”

is the name of the joined table).

Destination Fields
Each of the fields of the current destination data table – along with any
nested fields – appear in the Destination Fields grid.
NOTE ►► For simplicity, the different “types” of destination fields,
attributes, qualifiers, and nested display fields are each generically
referred to as “fields” unless the context is dealing with a specific type.

The destination field type is identified by the F/D/Q/A column, as

summarized in Table 54.

MDM Import Manager Reference Guide 271

Table 54. Destination Field Types Indicated by the F/D/Q/A Column
F/D/Q/A Type Description
F Field Fields of the current destination table.
The individual display fields of each lookup field whose
D Display Field
value is comprised of multiple display fields.
The qualifiers of each of the qualified fields.

Q Qualifier  With multiple qualified fields, each qualifier appears as

“qualifier <qualified field>” (where “qualified field” is the
name of the qualified field) (NYI)
The attributes of each of the taxonomy fields.

A Attribute  With multiple taxonomy fields, each attribute appears as

“attribute <taxonomy field>” (where “taxonomy field” is
the name of the taxonomy field)

MULTILINGUAL ►► In a multilingual repository, the Destination

Fields grid may contain multiple instances of each multilingual field
(one per language), allowing you to field map each language-specific
source field to the applicable language layer of the multilingual
destination field. More information about languages and multilingual
repositories is provided in “Part 14: Multilingual Support.”

 To map a source field to a destination field:

1. In the Source Fields grid, select the unmapped source field that you
want to map.
2. In the Destination Fields grid, select the destination field to which you
want to map the source field.
TIP ►► If the name of the destination field is the same as the source
field, you can automatically locate the destination field by right-clicking
on the source field in the Source Fields grid and choosing Find Next
Matching Destination Field from the context menu (see “Finding
Matching Destination Fields” on page 274 for more information).

3. Click on the Map button (between the fields grids) to map the selected
source field and the selected destination field.
4. MDM indicates that the source and destination fields have been
mapped by placing a green thunderbolt ( ) in the Mapped column of
both grids, and by placing the names of the source and destination
mapped fields into the corresponding column of the other grid.
NOTE ►► You must map a source field to the qualified field if you
want to import to any of its qualifiers.

NOTE ►► The Map button is only enabled on main table lookup fields
when key mapping is enabled on the lookup table.

272 MDM Import Manager Reference Guide

 NOTE ►► You must map a source field to the taxonomy field in order
to import to any of its attributes unless: (1) the import action for every
new source record is Skip; (2) the import action for every existing
source record is either of the two Update actions; and (3) the taxonomy
field value of every matching destination record is non-NULL.

TIP ►► If the source data does not contain a field that corresponds to
the taxonomy field, you can: (1) use the Add Field command to add a
new field; (2) field map the new field to the taxonomy field; and (3)
value map the <NULL> value of the new field to the corresponding
value of the taxonomy field.

DATA INTEGRITY ►► The MDM Import Manager automatically links

unlinked attributes to a category as part of the import when: (1) you
import into the category records for which the source data field that is
mapped to the attribute has non-NULL values; or (2) the category
value is changed for an existing record (i.e. an existing record is
reassigned to another category) and data values would otherwise be
discarded. (See “Configuration Options” on page 62 for more
information about the Autolink Attributes options.)

UNMAPPING FIELDS
You can unmap one or more previously mapped source fields as
described in this section.

 To unmap one or more fields:

1. In the Source Fields grid, select the mapped source fields that you
want to unmap.
TIP ►► Alternatively, you can select the mapped destination fields in
the Destination Fields grid.

NOTE ►► When you select a mapped source field in the Source

Fields grid, the corresponding mapped destination field is automatically
selected in the Destination Fields grid (but not the other way around).

2. Click on the Unmap button (between the fields grids) to unmap the
source fields from their mapped destination fields.
NOTE ►► The Unmap button is focus-dependent and unmaps the
selected fields from the grid that either: (1) currently has the focus; or
(2) last had the focus if neither fields grid currently has the focus.

3. MDM indicates that the source and destination fields have been
unmapped by removing the green thunderbolt ( ) from the Mapped
column of both grids for the unmapped fields.

MDM Import Manager Reference Guide 273

AUTOMAPPING SOURCE FIELDS
If some or all of the source fields have names that are the same or
similar to the corresponding destination fields, you can use the Automap
button to automatically map them to their matching destination fields, as
described in this section.
TIP ►► The Automap button provides a convenient way to quickly get
started with the field mapping process.

 To automatically map one or more source fields to the first matching

destination fields:
♦ Click on the Automap button (between the fields grids). MDM
automatically maps each unmapped source field to the first destination
field with a name that is the same or similar, if one exists.
NOTE ►► See “The Find Matching Field Options” on page 275 for
more information about how MDM performs the field name
comparisons.

FINDING MATCHING DESTINATION FIELDS

Alternatively, if there may be more than one destination field with a
name similar to each source field and you want to manually choose to
which destination field the source field should be mapped, you can use
the Find Next Matching Destination Field command to quickly locate
each matching destination field in sequence for potential mapping, as
described in this section.

 To find destination fields whose names are similar to a source field:

1. In the Source Fields grid, select the source field for which you want to
find a similarly named destination field.
2. Right-click on the field and choose Find Next Matching Destination
Field from the context menu, or choose Fields > Find Next Matching
Destination Field from the main menu.
3. MDM moves the selected field in the Destination Fields grid from the
current selection to the next similarly named destination field.
TIP ►► You can right-click on the source field and choose Find Next
Matching Destination Field repeatedly (or simply press F3) to find each
similarly named destination field one after another.

274 MDM Import Manager Reference Guide

THE FIND MATCHING FIELD OPTIONS
To accommodate subtle differences between source and destination
field names and increase the likelihood of finding a matching field with
the Automap or the Find Next Matching Destination Field commands, the
MDM Import Manager by default: (1) performs a “contains” comparison
against the destination field name; and (2) normalizes (i.e. removes
punctuation characters and spaces) from both the source and
destination field names before performing the comparison.
You can use the Configuration Options command to change the settings
of the Find Matching Field options, as summarized in Table 55.

Table 55. Find Matching Field Configuration Options

Option Value Description
Contains Performs a contains comparison.
Operator
Exact Performs an exact comparison.
Normalized First normalizes source and destination field names.
Special Case Insensitive Performs a case insensitive comparison.
Case Sensitive Performs a case sensitive comparison.

NOTE ►► See “Configuration Options” on page 62 for more

information about the Find Matching Field options.

CREATING A COMPOUND SOURCE FIELD

Recall that when a destination lookup field represents the combination
of the underlying lookup table’s multiple display fields, each of the
individual display fields is promoted and automatically appears in the
Destination Fields grid as a target for mapping.
To create a compound source field for mapping to a destination lookup
field that consists of multiple display fields: (1) map each source field to
one of the display fields; (2) for each source field, either convert or map
its values to the values of the display field; (3) use the Create Compound
Field command to create a compound field for mapping to the
destination lookup field, as described in this section; (4) map the
compound field to the destination lookup field; and (5) automap the
compound field’s source value combinations to the lookup field values.
NOTE ►► The source value combinations created by the Create
Compound Field command correspond to the destination lookup value
combinations created by the lookup table’s multiple display fields.

NOTE ►► Creating a compound source field is possible only if the

source data has individual fields that correspond to the individual
display fields of the lookup table.

MDM Import Manager Reference Guide 275

 To create a compound source field:
1. Right-click in the Source Fields grid and choose Create Compound
Field from the context menu, or choose Source > Create Compound
Field from the main menu, and select from the cascading menu of
destination lookup fields that consist of multiple display fields.
NOTE ►► Destination lookup fields whose individual display fields
have not all been mapped and their values all converted or mapped
appear in the cascading menu in gray.

2. MDM adds the compound field to the Source Hierarchy tree and to the
list of fields in the Source Fields grid.
NOTE ►► The compound source field appears in the Source Fields
grid as “compound lookup field <source field 1; source field n>” (where
“compound lookup field” is the name of a destination lookup field and
“source field n” is the name of the nth source field).

NOTE ►► Once a compound field is created, the source fields used

to create the compound field cannot be unmapped until the compound
field is deleted.

SETTING THE FIELD MAPPING DELIMITERS

The current field mapping delimiters determine how the MDM Import
Manager interprets source data values as part of the value conversion
and mapping process, and must be properly set prior to mapping a
source field to a destination field.
The two field mapping delimiters include:
• Multi-Valued delimiter. MDM identifies the individual delimited
values within a single field when mapping to a multi-valued field or a
compound lookup field using the current Multi-Valued delimiter string.
• Coupled Numeric delimiter. MDM identifies the coupled numeric
value pairs when mapping to a coupled numeric attribute using the
current Coupled Numeric delimiter string.

Specifically, field mapping uses as the current delimiter either: (1) the
global default delimiter string, if the field is set to inherit; or (2) the field-
specific delimiter string, if it has been manually set for the field.
You can set the global default delimiter string for all fields that inherit the
default using the Configuration Options command, or alternatively,
override the default and set it on a field-by-field basis using the Set Field
Mapping Delimiter commands, as described in this section.
NOTE ►► Inherited global default values appear in the Source Fields
grid in gray.

276 MDM Import Manager Reference Guide

 NOTE ►► A source data field that contains multiple delimited coupled
numeric value pairs requires that you set both delimiters. For example,
“5hp @ 3500rpm; 6hp @ 4500rpm; 7hp @ 5500rpm” would require
that the Coupled Numeric delimiter be set to “@” and the Multi-Valued
delimiter be set to “;”.
TIP ►► Be careful not to include leading or trailing spaces in the
delimiter string unless your intent is to distinguish between
occurrences of the delimiter with and without surrounding spaces (e.g.
“X;Y” represents a single value but “X ; Y” represents two values).
MDM will automatically trim the leading and trailing spaces from
individual delimited values even if you do not include the spaces.

NOTE ►► See “Source Field-Level Settings” on page 224 for more

information about all of the source field-level settings; see
“Configuration Options” on page 62 for more information about the
global default delimiter strings.

 To set the Multi-Valued or Coupled Numeric delimiter string for one

or more fields:
1. In the Source Fields grid, select the fields for which you want to set the
Multi-Valued or Coupled Numeric delimiter string.
2. Right-click on one of the fields and choose Set Field Mapping Delimiter
from the context menu, or choose Fields > Set Value Delimiter from the
main menu, and then choose from the cascading menu:
 Multi-Valued
 Coupled Numeric
3. MDM open a Set Delimiter dialog which displays the current value for
the delimiter string and whether or not the global default is being
inherited.
NOTE ►► No value is displayed in the edit control if you have
selected multiple fields for which the current values are not all the
same. Similarly, a gray check ( ) appears in the Inherit Default
checkbox if some but not all of the selected fields are inheriting the
global default.

4. To set the delimiter string and override the global default value, enter
the new delimiter string into the edit control, or alternatively, to restore
inheritance, check the Inherit Default checkbox.
5. Click OK when you are done to close the dialog.

MDM Import Manager Reference Guide 277

SETTING THE CASE SENSITIVITY OPTION
The current setting of the Case Sensitivity option determines whether
the MDM Import Manager ignores or recognizes case differences in the
source data when it determines the set of distinct values for a field, and
must be properly set prior to: (1) displaying values using the Display
Distinct Values command; (2) partitioning the field; (3) splitting the field
using the Split into Hierarchy or Split into Multiple Fields commands; or
(4) mapping the source field to a destination field.
Specifically, the MDM Import Manager uses as the current case-
sensitivity setting either: (1) the global default setting, if the field is set to
inherit; or (2) the field-specific setting, if it has been manually set for the
field.
You can set the global default case-sensitivity setting for all fields that
inherit the default using the Configuration Options command, or
alternatively, override the default and set it on a field-by-field basis using
the Set Case Sensitivity command, as described in this section.
NOTE ►► Inherited global default values appear in the Source Fields
grid in gray.

NOTE ►► See “The Case Sensitivity Option” on page 237 for more
information about the Case Sensitivity option; see “Source Field-Level
Settings” on page 224 for more information about all of the source
field-level settings; see “Configuration Options” on page 62 for more
information about the global default Case Sensitivity option.

 To set the Case Sensitivity option for one or more fields:

1. In the Source Fields grid, select the fields for which you want to set the
Case Sensitivity option.
2. Right-click on one of the fields and choose Set Case Sensitivity from
the context menu, or choose Fields > Set Case Sensitivity from the
main menu.
3. MDM displays a cascading menu of three choices:
 Case Insensitive
 Case Sensitive
 Inherit Default
4. To set the Case Sensitivity option and override the global default
setting, choose either Case Insensitive or Case Sensitive, or
alternatively, to restore inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

278 MDM Import Manager Reference Guide

SETTING THE FLAT NULL HANDLING OPTION
The current setting of the Flat NULL Handling option determines
whether the MDM Import Manager preserves or ignores NULLs in the
source data when you map a field, and must be properly set prior to
mapping a source field to a destination field.
Specifically, the MDM Import Manager uses as the current flat NULL-
handling setting either: (1) the global default setting, if the field is set to
inherit; or (2) the field-specific setting, if it has been manually set for the
field.
You can set the global default flat NULL-handling setting for all fields
that inherit the default using the Configuration Options command, or
alternatively, override the default and set it on a field-by-field basis using
the Set Flat NULL Handling command, as described in this section.
NOTE ►► Inherited global default values appear in the Source Fields
grid in gray.

NOTE ►► See “The Flat NULL Handling Option” on page 232 for
more information about the Flat NULL Handling option; see “Source
Field-Level Settings” on page 224 for more information about all of the
source field-level settings; see “Configuration Options” on page 62 for
more information about the global default Flat NULL Handling option.

 To set the Flat NULL Handling option for one or more fields:
1. In the Source Fields grid, select the fields for which you want to set the
Flat NULL Handling option.
2. Right-click on one of the fields and choose Set Flat NULL Handling
from the context menu, or choose Fields > Set Flat NULL Handling from
the main menu.
3. MDM displays a cascading menu of three choices:
 Preserve NULLs
 Ignore NULLs
 Inherit Default
4. To set the Flat NULL Handling option and override the global default
setting, choose either Preserve NULLs or Ignore NULLs, or alternatively,
to restore inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

MDM Import Manager Reference Guide 279

SETTING THE HIERARCHY NULL HANDLING OPTION
The current setting of the Hierarchy NULL Handling option determines
how the MDM Import Manager preserves or ignores NULLs in the
source data when you map a field, and must be properly set prior to
mapping a source field to a destination field.
Specifically, the MDM Import Manager uses as the current hierarchy
NULL-handling setting either: (1) the global default setting, if the field is
set to inherit; or (2) the field-specific setting, if it has been manually set
for the field.
You can set the global default hierarchy NULL-handling setting for all
fields that inherit the default using the Configuration Options command,
or alternatively, override the default and set it on a field-by-field basis
using the Set Hierarchy NULL Handling command, as described in this
section.
NOTE ►► Inherited global default values appear in the Source Fields
grid in gray.

NOTE ►► See “The Hierarchy NULL Handling Option” on page 233

for more information about the Hierarchy NULL Handling option; see
“Source Field-Level Settings” on page 224 for more information about
all of the source field-level settings; see “Configuration Options” on
page 62 for more information about the global default Hierarchy NULL
Handling option.

 To set the Hierarchy NULL Handling option for one or more fields:
1. In the Source Fields grid, select the fields for which you want to set the
Hierarchy NULL Handling option.
2. Right-click on one of the fields and choose Set Hierarchy NULL
Handling from the context menu, or choose Fields > Set Hierarchy
NULL Handling from the main menu.
3. MDM displays a cascading menu of three choices:
 Preserve NULLs
 Ignore Non-Leaf Trailing NULLs
 Ignore All Trailing NULLs
 Inherit Default
4. To set the Hierarchy NULL Handling option and override the global
default setting, choose either Preserve NULLs, Ignore Non-Leaf Trailing
NULLs, or Ignore All Trailing NULLs, or alternatively, to restore
inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

280 MDM Import Manager Reference Guide

APPLYING FIELD-LEVEL VALUE CONVERSION FILTERS
When you apply value conversion filters at the field level, the
conversions are applied sequentially to each converted value that has
not been individually edited or converted at the value level, starting with
the original value.
For each field, you can apply one or more value conversion filters at the
field level, as described in this section.
NOTE ►► When you first map a field, each value in the Converted
Value column of the Source Values grid appears in gray to indicate
that it is inheriting value conversion filters from the field level.

NOTE ►► Value conversion filters can also be applied at the

individual value level (see “Applying Value-Level Value Conversion
Filters” on page 314 for more information).

NOTE ►► See “Value Conversion Filters” on page 238 for more

information about the value conversion filters.

 To apply a single value conversion filter at the field level:

1. In the Source Fields grid, select the field to which you want to apply the
value conversion filter.
2. Right-click on the field and choose Set Value Conversion Filter from the
context menu, or choose Fields > Set Value Conversion Filter from the
main menu, and then choose from the cascading menu of value
conversion filters.
NOTE ►► Each time you apply a single field-level value conversion
filter, you replace any previously applied field-level filters for the field.

3. Follow the instructions for the corresponding value conversion filter, in

the subsections of “Applying Value-Level Value Conversion Filters”
starting on page 314.
4. MDM sets the filter and applies it to each Converted Value that it is
inheriting value conversion filters from the field level.

 To cancel field-level value conversion filters for the selected field:

1. In the Source Fields grid, select the field for which you want to cancel
the value conversion filters.
2. Right-click on the field and choose Set Value Conversion Filter from the
context menu, or choose Fields > Set Value Conversion Filter from the
main menu, and then choose None from the cascading menu of value
conversion filters.

MDM Import Manager Reference Guide 281

 To apply multiple value conversion filters at the field level:
1. In the Source Fields grid, select the field to which you want to apply
multiple value conversion filters.
2. Right-click on the field and choose Set Value Conversion Filter from the
context menu, or choose Fields > Set Value Conversion Filter from the
main menu, and then choose Multiple from the cascading menu to
open the Set Value Conversion Filters dialog shown in Figure 188.

Figure 188. Set Value Conversions dialog

3. Click on the Add button and choose from the cascading menu of value
conversion filters.
4. Follow the instructions for the corresponding value conversion filter, in
the subsections of “Applying Value-Level Value Conversion Filters”
starting on page 314.
NOTE ►► Each time you add a value conversion filter, it is added to
the end of the Filters list of previously applied filters for the field.

TIP ►► To remove a filter, select it and click on the Remove button.

TIP ►► To modify a filter, right-click on it and choose Modify from the

context menu.

TIP ►► To reorder a filter, drag-and-drop it in the list.

5. Click OK when you are done to close the Set Value Conversions
dialog.
TIP ►► To quickly remove all field-level value conversion filters for
the selected field, right-click on the field and choose Set Value
Conversion Filter from the context menu, or choose Fields > Set Value
Conversion Filter from the main menu, and then choose None from the
cascading menu of value conversion filters.

282 MDM Import Manager Reference Guide

FILTERING UNLINKED ATTRIBUTES
Before you map a source field to a destination attribute, it would often
be convenient to know precisely which attributes are already linked to a
category. This way you can identify a corresponding linked attribute (if
one exists) rather than inadvertently map to an unlinked attribute, or
worse, map to a new attribute that you create as part of the import.
NOTE ►► The MDM Import Manager automatically links unlinked
attributes to a category into which you import records for which the
source data field that is mapped to the attribute has non-NULL values.
(See “Configuration Options” on page 62 for more information about
the Autolink Attributes options.)

 To highlight the attributes that are linked to one or more categories:

1. Right-click anywhere in the Destination Fields grid and choose Filter
Unlinked Attributes from the context menu, or choose Fields > Filter
Unlinked Attributes from the main menu, to open the Filter Unlinked
Attributes dialog shown in Figure 189.

Resizable
Dialog

Figure 189. Filter Unlinked Attributes dialog

2. Move one or more categories whose linked attributes you want to

highlight from Available Categories to the Selected Categories list.
3. Click OK when you are done to close the Filter Unlinked Attributes
dialog. MDM highlights in gray the attributes in the Destination Fields
grid that are not linked to the categories you selected.
NOTE ►► You can still map to an unlinked attribute even when it is
highlighted in gray.

TIP ►► To cancel or change the selected categories, you can perform

the Filter Unlinked Attributes command again.

TIP ►► To sort the linked attributes as a group before the unlinked

attributes, sort the Destination Fields grid by the F/D/Q/A column.

MDM Import Manager Reference Guide 283

ADDING A NEW DESTINATION ATTRIBUTE
Sometimes, category-specific fields of source data do not have a
corresponding destination attribute. You can create one or more new
attributes as part of the import process as described in this section.

 To create a new destination attribute to map to from a source field:

1. In the Source Fields grid, select the source field for which you want to
create a new corresponding destination attribute.
2. Click on the Add button (between the fields grids) to open the Add
Attribute dialog shown in Figure 190.

Figure 190. Add Attribute dialog

NOTE ►► If the current destination table has multiple taxonomy

fields, Add first brings up a cascading menu of taxonomy field names
for you to select the one to which you want to add an attribute.

3. MDM automatically populates the Name property with the name of the
selected source field.
4. Select the type of attribute from the radio controls, specify the other
properties for the attribute, and click OK when you are done to close
the dialog and create the attribute.

DATA INTEGRITY ►► The MDM Import Manager automatically links

the new attribute to each category into which you import records for
which the source data field that is mapped to the attribute has non-
NULL values. (See “Configuration Options” on page 62 for more
information about the Autolink Attributes options.)

284 MDM Import Manager Reference Guide

SETTING THE NULL INTERPRETATION OPTION
The current setting of the NULL Interpretation option determines
whether NULL source values are ignored or used to overwrite existing
destination data values.
Specifically, the MDM Import Manager uses as the current NULL
Interpretation setting either: (1) the global default setting, if the field is
set to inherit; or (2) the field-specific setting, if it has been manually set
for the field.
You can set the global default NULL Interpretation setting for all fields
that inherit the default using the Configuration Options command, or
alternatively, override the default and set it on a field-by-field basis using
the Set NULL Interpretation command, as described in this section.
NOTE ►► Inherited global default values appear in the Destination
Fields grid in gray.

NOTE ►► See “The NULL Interpretation Option” on page 236 for

more information about the NULL Interpretation option; see
“Destination Field-Level Settings” on page 225 for more information
about all of the destination field-level settings; see “Configuration
Options” on page 62 for more information about the global default
NULL Interpretation option.

 To set the NULL Interpretation option for one or more fields:

1. In the Destination Fields grid, select the object fields for which you
want to set the NULL Interpretation option.
2. Right-click on one of the fields and choose Set NULL Interpretation
from the context menu, or choose Fields > Set NULL Interpretation
from the main menu.
3. MDM displays a cascading menu of two choices:
 Ignore
 Overwrite
4. To set the NULL Interpretation option and override the global default
setting, choose either Ignore or Overwrite, or alternatively, to restore
inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

MDM Import Manager Reference Guide 285

SETTING THE MULTI-VALUED UPDATE OPTION
The current setting of the Multi-Valued Update option determines
whether incoming source data values either are appended to or
completely replace the set of existing destination data values of a multi-
valued destination field.
Specifically, the MDM Import Manager uses as the current Multi-Valued
Update setting either: (1) the global default setting, if the field is set to
inherit; or (2) the field-specific setting, if it has been manually set for the
field.
You can set the global default Multi-Valued Update setting for all fields
that inherit the default using the Configuration Options command, or
alternatively, override the default and set it on a field-by-field basis using
the Set Multi-Valued Update command, as described in this section.
NOTE ►► The Multi-Valued Update option applies only to multi-
valued destination fields.

NOTE ►► Inherited global default values appear in the Destination

Fields grid in gray.

NOTE ►► See “The Multi-Valued Update Option” on page 251 for

more information about the Multi-Valued Update option; see
“Destination Field-Level Settings” on page 225 for more information
about all of the destination field-level settings; see “Configuration
Options” on page 62 for more information about the global default
Multi-Valued Update option.

 To set the Multi-Valued Update option for one or more fields:

1. In the Destination Fields grid, select the multi-valued fields for which
you want to set the Multi-Valued Update option.
2. Right-click on one of the fields and choose Set Multi-Valued Update
from the context menu, or choose Fields > Set Multi-Valued Update
from the main menu.
3. MDM displays a cascading menu of three choices:
 Append
 Replace
 Inherit Default
4. To set the Multi-Valued Update option and override the global default
setting, choose either Append or Replace, or alternatively, to restore
inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

286 MDM Import Manager Reference Guide

SETTING THE QUALIFIED UPDATE OPTION
The current setting of the Qualified Update option determines whether
incoming source data values either are appended to or completely
replace the set of existing destination data values of a qualified lookup
destination field.
Specifically, the MDM Import Manager uses as the current Qualified
Update setting either: (1) the global default setting (in conjunction with
the setting of the Matching Qualifiers option), if the field is set to inherit;
or (2) the field-specific setting, if it has been manually set for the field.
You can set both the global default Multi-Valued Update setting and the
Matching Qualifiers setting (which work together in tandem) for all fields
that inherit the default using the Configuration Options command, or
alternatively, override the default and set it on a field-by-field basis using
the Set Qualified Update command, as described in this section.
NOTE ►► The Qualified Update option applies only to qualified
lookup destination fields.

NOTE ►► Inherited global default values appear in the Destination

Fields grid in gray.

NOTE ►► See “The Qualified Update Option” on page 253 for more
information about the Qualified Update option; see “Destination Field-
Level Settings” on page 225 for more information about all of the
destination field-level settings; see “Configuration Options” on page 62
for more information about the global default Qualified Update and
Matching Qualifiers options.

 To set the Qualified Update option for one or more fields:

1. In the Destination Fields grid, select the qualified lookup fields for
which you want to set the Qualified Update option.
2. Right-click on one of the fields and choose Set Qualified Update from
the context menu, or choose Fields > Set Qualified Update from the
main menu.
3. MDM displays a cascading menu of four choices:
 Append
 Replace
 Update
 Inherit Default
4. To set the Qualified Update option and override the global default
setting, choose either Append, Replace, or Update, or alternatively, to
restore inheritance, choose Inherit Default.
NOTE ►► You can inherit only All or None for Matching Qualifiers
since the set of qualifiers is different for each qualified lookup field.

MDM Import Manager Reference Guide 287

 5. MDM opens the Set Matching Qualifiers dialog shown in Figure 191.

Figure 191. Set Matching Qualifiers dialog

NOTE ►► The New Links and Existing Links drop-down controls are
disabled when you choose Append or Replace; they are enabled only
when you choose Update.

6. In the dual-list control for qualifier selection, move the qualifiers that
you want to update to the Available Qualifiers list and move the
qualifiers you want to match to the Matching Qualifiers list.
7. If you chose Update in step 4, in the New Links drop-down control,
choose from the list of new link options:
 Skip
 Create
8. If you chose Update in step 4, in the Existing Links drop-down control,
choose from the list of existing link options:
 Skip
 Update (NULL Qualifiers Only)
 Update (All Mapped Qualifiers)
 Replace
9. Click OK when you are done to close the Set Matching Qualifiers
dialog.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

NOTE ►► The Append and Replace options use matching qualifiers

for qualified link aggregation only, while the Update option uses them
both for qualified link aggregation and for matching. (See “Merging
Qualified Links and the Qualified Update Option” on page 355 for more
information about qualified link aggregation.

288 MDM Import Manager Reference Guide

SETTING THE TUPLE UPDATE OPTION
The current setting of a tuple field’s Tuple Update option determines
whether incoming source values are added as new tuple records, used
to update an existing tuple record, or skipped entirely.
Specifically, the Tuple Update option allows you to specify:
• Which of the tuple’s member fields (if any) MDM should use as
matching fields; and then
• How to update the values of the remaining member fields for new or
existing tuple records.

You can set Tuple Update options on a field-by-field basis using the
Tuple Update command, as described in this section.
NOTE ►► If you do not explicitly set the Tuple Update options for a
tuple field, Import Manager updates the field using the default tuple
update settings from the Configuration Options dialog (see
“Configuration Options” on page 62 for more information).

NOTE ►► The Tuple Update command is disabled for single-valued

tuple fields nested within a tuple structure. Instead, Import Manager
applies the update option set for its parent tuple in the tuple structure.

NOTE ►► See “The Tuple Update Option” on page 259 for more
information about the behaviors of each Tuple Update option.

 To set the Tuple Update option for one or more fields:

1. In the Destination Fields grid, select the tuple field for which you want
to set the Tuple Update option.
2. Right-click on the field and choose Tuple Update from the context
menu, or choose Fields > Tuple Update from the main menu.
3. MDM opens the Set Matching Tuples dialog shown in Figure 192.

Figure 192. Set Matching Tuples dialog

MDM Import Manager Reference Guide 289

 4. In the dual-list control for field selection, move the tuple member fields
that you want to update to the Available fields list and move the tuple
member fields that you want to match on to the Matching fields list.
NOTE ►► If no fields are in the Matching fields list, then Import
Manager will update all of the tuple‘s records with the incoming source
data, according to the settings in the New records and Existing records
drop-down controls.

5. In the New records drop-down control, choose from the list of options:
 Skip
 Create
6. In the Existing records drop-down control, choose from the list of
options:
 Skip
 Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
 Delete
7. Click OK when you are done to close the Set Matching Tuple Fields
dialog.

SETTING THE DATA GROUP DISPLAY OPTION

The current setting of the Data Group Display option determines
whether the Data Groups hierarchy is hidden or displayed in the
Destination Values grid for object lookup destination fields.
Specifically, Import Manager uses as the current Data Group Display
setting either: (1) the global default setting, if the field is set to inherit; or
(2) the field-specific setting, if it has been manually set for the field.
You can set the global default Data Group Display setting for all fields
that inherit the default using the Configuration Options command, or
alternatively, override the default and set it on a field-by-field basis using
the Set Data Group Display command, as described in this section.
NOTE ►► The Data Group Display option applies only to object
lookup destination fields.

NOTE ►► Inherited global default values appear in the Destination

Fields grid in gray.

NOTE ►► See “The Data Group Display Option” on page 265 for
more information about the Data Group Display option; see
“Destination Field-Level Settings” on page 225 for more information
about all of the destination field-level settings; see “Configuration
Options” on page 62 for more information about the global default Data
Group Display option.

290 MDM Import Manager Reference Guide

 To set the Data Group Display option for one or more fields:
1. In the Destination Fields grid, select the object fields for which you
want to set the Data Group Display option.
2. Right-click on one of the fields and choose Set Data Group Display
from the context menu, or choose Fields > Data Group Display from
the main menu.
3. MDM displays a cascading menu of two choices:
 Hide Data Group
 Display Data Group
4. To set the Data Group Display option and override the global default
setting, choose either Hide Data Group or Display Data Group, or
alternatively, to restore inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

SETTING THE MDIS UNMAPPED VALUE HANDLING OPTION

The current setting of the MDIS Unmapped Value Handling option
determines whether unmapped values are ignored, added to the set of
existing destination values, or mapped to an existing destination value
during MDIS processing.
Specifically, the MDM Import Manager uses as the current MDIS
Unmapped Value Handling setting either: (1) the global default setting, if
the field is set to inherit; or (2) the field-specific setting, if it has been
manually set for the field.
You can set the global default MDIS Unmapped Value Handling setting
for all fields that inherit the default using the Configuration Options
command, or alternatively, override the default and set it on a field-by-
field basis using the Set MDIS Unmapped Value Handling command, as
described in this section.
NOTE ►► Inherited global default values appear in the Destination
Fields grid in gray.

 To set the MDIS Unmapped Value Handling option for one or more
fields:
1. In the Destination Fields grid, select the fields for which you want to set
the MDIS Unmapped Value Handling option.
2. Right-click on one of the fields and choose Set MDIS Unmapped Value
Handling from the context menu, or choose Fields > Set Unmapped
Value Handling from the main menu.

MDM Import Manager Reference Guide 291

 3. MDM displays a cascading menu of four choices:
 Ignore
 Add
 Map
 Fail
 Inherit Default
4. To set the MDIS Unmapped Value Handling option and override the
global default setting, choose either Ignore, Add, Map, or Fail, or
alternatively, to restore inheritance, choose Inherit Default.
NOTE ►► You cannot inherit the Map option since the set of
destination values is different for each destination field.

5. If you chose Map in step 4, MDM opens one of the MDIS Mapped
Unmapped Value dialogs shown in Figure 193 and Figure 194, for
single and multi-valued destination fields, respectively.

Resizable
Dialog

Figure 193. MDIS Map Unmapped Value dialog (single-valued)

Resizable
Dialog

Figure 194. MDIS Map Unmapped Value dialog (multi-valued)

292 MDM Import Manager Reference Guide

 6. The MDIS Map Unmapped Value dialogs are populated with the set of
existing values for the destination field. Choose: (1) one value for a
single-valued destination field; or (2) one or more values for a multi-
valued destination field, and click OK.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current setting unless you have selected
multiple fields for which the settings are not all the same.

MDM Import Manager Reference Guide 293

Value Conversion and Mapping Operations
The following sections describe the various value conversion and
mapping operations you can perform on source and destination values
in the bottom pair of Value Conversion and Mapping grids, and are
summarized in Table 56.

Table 56. Value Conversion and Mapping Operations

Operation Description
Buttons
Automatically maps each source value to destination
Automap
values that are equivalent.
Maps the selected source values to the selected
Map
destination values.

Unmap Unmaps the selected values.

Adds the selected source values as new destination

Add
values.

Remove Removes the selected new destination values.

Source Values Grid

Apply Value Converts the selected source values using the
Conversion Filter specified value conversion filter.
Find Next Unmapped Finds the first or next unconverted or unmapped
Source Value source value.
Reorder Mapped Reorders the mapped destination values for the
Destination Values selected source value.
Find Next Mapped Finds the first or next destination value to which the
Destination Value selected source value is mapped.
Selects the applicable children of the selected internal
Select Entire Branch
node.
Destination Values Grid

Add Value Adds a new destination value.

Remove Value Removes the selected new destination values.

294 MDM Import Manager Reference Guide

 NOTE ►► NULL handling, case sensitivity, and delimiters must be set
before you field map because they affect how original values are
interpreted and how converted values are generated; the NULL
Handling and Case Sensitivity options also affect the set of original
values. By contrast, value conversion filters must be set after you map
because they depend on the destination data type. (See “The NULL
Handling Options” on page 231, “The Case Sensitivity Option” on page
237, and “Value Conversion Filters” on page 238 for more information
about these field-level settings.)

SOURCE VALUE CONVERSION AND MAPPING

Converting and mapping field values is also extremely straightforward
using the Value Conversion and Mapping grids, shown in Figure 195.

Figure 195. Value Conversion and Mapping grids

For each mapped source field, the value conversion and mapping
process depends upon whether or not value mapping is required:
• No value mapping. If value mapping is not required, you must make
sure that every source value is properly converted, as indicated by a
green bullet ( ) in the Conv. column of the Source Values grid.
• Value mapping. If value mapping is required, you must make sure
that every source value is mapped, as indicated by a colored
thunderbolt ( ) in the Mapped column of the Source Values grid

Converting and mapping source values is described in the following

sections; the specific meanings of each of the colored thunderbolt icons
when value mapping is required are summarized in Table 57.

MDM Import Manager Reference Guide 295

Table 57. Mapped Icons in the Source and Destination Values Grids
Values Grid Icon Description
Value is mapped to an existing destination value.
Value is mapped to multiple existing destination values.
Source Value is mapped to a new destination value.
Value is mapped to multiple new destination values.
Value is mapped to multiple destination values, some of which are new.
Single source value is mapped to the value.
Multiple source values are mapped to the value.
Destination
Single source value is mapped to the new value.
Multiple source values are mapped to the new value.

Source Values Grid

Each of the values of the current source field appears in the Source
Values grid and is available for conversion and/or mapping, as follows:
• Value column. Original source values appear with their original data
type in the Value column.
• Converted Value column. Converted values appear in the
Converted Value column with their data types automatically
converted based on the data type of the mapped destination field,
(even when stored in the source data as text values), and the proper
grid cell type for editing (e.g. text, numeric, measurement, currency,
date, time, Boolean). If necessary, individual values can then be
manually edited to override errors or omissions in the source data.
NOTE ►► If the mapped source field or any of its values has been
partitioned, then the merged values (combined partition) or the
hierarchy of values (nested partition) appear in the Source Values grid.

NOTE ►► When you first map a field, each value in the Converted
Value column appears in gray to indicate that it is inheriting value
conversion filters from the field level. When you apply a manual edit or
a value-conversion filter to an individual value, the Converted Value
then appears in black to indicate that you have overridden inheritance
from the field level. You can use the Restore Converted Value
command to restore inheritance for a value.

NOTE ►► When value mapping is required, the value in the

Converted Value column is most useful for adding new destination
values or hierarchy, where it is used as the new item or node value.
When mapping to existing destination values, the Converted Value is
just a convenience for Automap, since each value could be manually
mapped to any destination value regardless of its value.

296 MDM Import Manager Reference Guide

Destination Values Grid
Each of the values of the current destination field appears in the
Destination Values grid and you can map to any of them.
NOTE ►► If the mapped destination field or any of its values has
been partitioned, then the merged value combinations (combined
partition) or a hierarchy of value combinations (nested partition) appear
in the Destination Values grid.

NOTE ►► A hierarchy of values appears in the Destination Values

grid also when the current table is a hierarchy table and the mapped
destination field is the display field of the hierarchy table.

MULTILINGUAL ►► In a multilingual repository, the Destination

Values grid may contain inherited values from language layers other
than the current language, which you can set as part of the value
mapping. More information about languages and multilingual
repositories is provided in “Part 14: Multilingual Support.”

EDITING SOURCE VALUES

Whether or not the mapped destination field requires value mapping,
you can use the edit controls in the Converted Value column of the
Source Values grid to manually edit each source value.
NOTE ►► When the mapped destination field does not require value
mapping, you are done converting values when “n of n values
converted” appears on the status bar.

Single-Valued Destination Field

When a source field is mapped to a single-valued destination field, you
can manually edit each individual source value using MDM’s standard
single-valued edit controls, as described in this section.
NOTE ►► If the destination field requires value mapping, you can
map each source data value even without manually editing it.

 To edit an individual source value (single-valued destination):

1. In the Source Values grid, select the value that you want to edit.
2. Make sure the current cell is on the Converted Value column, as
shown in Figure 196.

MDM Import Manager Reference Guide 297

 Figure 196. Single-valued edit control

3. Type a new value, or press Enter to open the cell and edit or overwrite
the current value.
NOTE ►► The cell type of the Converted Value cells depends upon
the data type of the mapped destination field.

TIP ►► You can also double-click or press F2 to enter the cell and
begin editing.

4. Press Enter when you are done to close the cell and save the value.

Multi-Valued Destination Field

When a source field is mapped to a multi-valued destination field, you
can manually edit each individual source value using MDM’s special
multi-valued drop-down edit controls to edit or create one or more
values for each source value, as described in this section.
NOTE ►► If the destination field requires value mapping, you can
map each source data value to multiple destination values even
without manually expanding it into multiple values.

 To edit an individual source value consisting of one or more delimited

values (multi-valued destination):
1. In the Source Values grid, select the value that you want to edit.
2. Make sure the current cell is on the Converted Value column.
3. Double-click on the cell or click on the down triangle to open the multi-
valued drop-down control, as shown in Figure 197.

298 MDM Import Manager Reference Guide

 Figure 197. Multi-valued drop-down edit control

NOTE ►► The cell type of the individual cells within the drop-down
control depends upon the data type of the mapped destination field.

4. MDM populates each row of the multi-valued drop-down control with

one of the multiple delimited values of the original source value.
5. Edit the values in each row of the drop-down control as follows:
 Use the Up and Down arrow keys to move from value to value.
 Press Enter to open a row and edit or overwrite its current value.
 Move to the empty row beyond the last value to add a value.
 Press Ctrl+Up and Ctrl+Down to reorder the values.
 Press Del to delete a value.
6. Click on the up triangle or press Shift+Enter when you are done to
close the drop-down control and save the values.

MAPPING SOURCE VALUES

The value mapping operations are described in the following sections
and summarized in Table 58.

MDM Import Manager Reference Guide 299

Table 58. Value Mapping Operations
Button Description and Constraints
All cases
 Map the selected source item(s) to the selected destination item(s).
 Mapping multiple source values to a single destination value eliminates duplicate
source values and collapses source value variations into a single, normalized
destination value.
 Mapping a single source value to multiple destination values creates a multi-value (but
is legal only if the destination field is multi-valued); if the source value is delimited, the
individual values are irrelevant for the mapping.

 Neither Source Values grid nor Destination Values grid can be empty
Hierarchy  Flat
 At least one of the selected source nodes must be a leaf node (can map source leaf
nodes only; cannot map an internal node)
Flat  Hierarchy (hierarchy lookup field)
 At least one of the selected destination nodes must be a leaf node (can map to
destination leaf nodes only; cannot map to an internal node)
Flat  Hierarchy (hierarchy table primary display field)
 No constraints (can map to both destination internal and leaf nodes)
Hierarchy  Hierarchy (hierarchy lookup field)
 At least one of the selected source nodes and one of the selected destination nodes
Map must be a leaf node (can map leaf nodes to leaf nodes only; cannot map from or to
internal nodes)
Hierarchy  Hierarchy (hierarchy table primary display field)
 No constraints (can map any leaf or internal node to any leaf or internal node)
Single Selection  Multiple Selections
 The selected source value must not yet be mapped to the selected destination value
 If single-valued destination, the selected source value cannot already be mapped to a
destination value other than the selected destination value (otherwise, would create a
one-to-many multi-valued mapping)
Multiple Selections  Single Selection
 At least one of the selected source values must not yet be mapped to the selected
destination value (otherwise, no new mappings are represented by the selections)
 If single-valued destination, none of the selected source values can already be mapped
to a destination value other than the selected destination value (otherwise, would create
a one-to-many multi-valued mapping)
Single Selection  Multiple Selections
 Destination must be multi-valued
 The selected source value must not yet be mapped to at least one of the selected
destination values (otherwise, no new mappings are represented by the selections)
Multiple Selections  Multiple Selections
 Disabled (too confusing)
All cases
 Unmap the selected item(s) in focus that were previously mapped.
Unmap
 The Unmap applies to whichever values grid last had the focus
 At least one of the selected items in current or previous focus must already be mapped

300 MDM Import Manager Reference Guide

 Button Description and Constraints
Flat / Hierarchy  Flat
 If necessary, add the selected source item(s) [leaf node(s)] after the last destination
item, and then map each source item [leaf node] to each new or existing destination
item.

 Destination Values grid must be flat

Add  At least one of the selected source items must not already be mapped
Hierarchy  Flat
 At least one of the unmapped selected source nodes must be a leaf node (can add
source leaf nodes only; cannot add from an internal node)
Hierarchy  Hierarchy
 Not applicable (see Add as Sibling/Child and Add Branch as Sibling/Child)
Flat / Hierarchy  Hierarchy
 If necessary, add the selected source item(s) [node(s)] as siblings after the selected
destination node; then map each source item [node] to each new or existing destination
leaf node.

Add  Destination Values grid must be a hierarchy

as Sibling  At least one of the selected source items [nodes] must not already be mapped
 Source items must be leaf nodes (hierarchy lookup field)
 Source items can be internal or leaf nodes (hierarchy table primary display field)
Hierarchy  Hierarchy
 All the selected source nodes must be leaf nodes or at the same level
Flat / Hierarchy  Hierarchy
 If necessary, add the selected source item(s) [node(s)] as children of the selected
destination node; then map each source item [node] to each new or existing destination
leaf node.

 Destination Values grid must be a hierarchy

 At least one of the selected source items [nodes] must not already be mapped
 Source items must be leaf nodes (hierarchy lookup field)
Add
as Child  Source items can be internal or leaf nodes (hierarchy table primary display field)
 Cannot add a child to a leaf node that already has records linked
 Cannot add a child to a leaf node that has just been added from a source item [leaf
node] (hierarchy lookup field)
 Can add a child to a leaf node that has just been added from a source item [node]
(hierarchy table primary table display field)
Hierarchy  Hierarchy
All the selected source nodes must be leaf nodes or at the same level
Hierarchy  Hierarchy only
 If necessary, add the selected branches and leaf nodes in their entirety as siblings after
the selected destination node; if necessary, add each of their descendents; then map
each source node to each new or existing destination node.

Add  Source and Destination Values grids must both be a hierarchy

Branch  All the selected source nodes must be at the same level and at least one must be an
internal node
as Sibling
 At least one of the source nodes must not already be mapped
 Each source leaf node is mapped to each new or existing destination leaf node
(hierarchy lookup field)
 Each source internal and leaf node is mapped to each new or existing destination
internal or leaf node (hierarchy table primary display field)

MDM Import Manager Reference Guide 301

 Button Description and Constraints
Hierarchy  Hierarchy only
 If necessary, add the selected branches and leaf nodes in their entirety as children of
the selected destination node; if necessary, add each of their descendents; then map
each source node to each new or existing destination node.

 Source and Destination Values grids must both be a hierarchy

 All the selected source nodes must be at the same level and at least one must be an
internal node
Add  At least one of the source nodes must not already be mapped
Branch  Each source leaf node is mapped to each new or existing destination leaf node
as Child (hierarchy lookup field)
 Each source internal and leaf node is mapped to each new or existing destination
internal or leaf node (hierarchy table primary display field)
 Cannot add a child branch to a leaf node that already has records linked
 Cannot add a child to a leaf node that has just been added from a source item [leaf
node] (hierarchy lookup field)
 Can add a child to a leaf node that has just been added from a source item [node]
(hierarchy table primary table display field)
All cases
 Remove the selected destination item(s) that were previously added.
Remove Flat  Hierarchy & Hierarchy  Hierarchy
 Removing an internal node also removes all of its children.

 At least one of the selected destination items must be new

NOTE ►► You are done mapping values when “n of n values

mapped” appears on the status bar.

NOTE ►► Button actions are selection-dependent, forgiving, and with

the exception of Unmap, focus-independent. “Forgiving” means that
any selected value that can’t be operated upon does not disable the
button; i.e. at least one but not all the selected items need to
participate in order for the button to be enabled.

NOTE ►► Single and multiple selection refers to whether a single

item is selected or multiple items are selected in each values grid.

NOTE ►► Many-to-one value mapping of multiple source values to a

single destination value collapses source value variations into a single,
normalized destination value.

NOTE ►► One-to-many value mapping of a single source value to

multiple destination values populates a multi-valued destination field
with multiple values.

NOTE ►► Delimiters in delimited source values are most useful for

conversion of numeric source values that do not require mapping, but
just a convenience for Automap when value mapping is required.

302 MDM Import Manager Reference Guide

 TIP ►► Since you cannot add or map the source root node value, you
can instead add and map an entire source hierarchy as follows:
(1) close the source tree to expose only the root;
(2) reopen the root node to display just the first level children;
(3) press Ctrl+A to select all the nodes;
(4) use Ctrl to unselect the root node; and
(5) use the appropriate Add Branch choice to add the hierarchy.

TIP ►► The Add Branch as Sibling/Child buttons only add source

internal nodes leading to unmapped source leaf nodes that are added
and mapped as part of the Add. You must use the Add as Sibling/Child
buttons to add source internal nodes without adding leaf nodes.

Mapping a Source Value

You can map a source value to a destination value as described in this
section. Corresponding source and destination values do not need to be
the same.

 To map a source value to a destination value:

1. In the Source Values grid, select the source value that you want to
map.
NOTE ►► When the source field is a hierarchy, you can only map
leaf-node source values to destination values (destination flat or
hierarchy lookup field).

2. In the Destination Values grid, select the destination value to which

you want to map the source value.
NOTE ►► When the source field is mapped to a destination
hierarchy, you can only map to leaf-node destination values
(destination hierarchy lookup field).

TIP ►► When a source field is mapped to a multi-valued destination

field you can map each source value to multiple destination values,
regardless of whether it consists of multiple delimited values. However,
if the destination field is a lookup into a table that has key mapping
enabled, you can only map each source vale to one destination value.

3. Click on the Map button (between the values grids) to map the selected
source value and the selected destination value.
4. If you are mapping to an inherited value in a multilingual repository,
choose from the cascading menu:
 Set Missing Value
 Do Not Set Missing Value

MDM Import Manager Reference Guide 303

 MULTILINGUAL ►► More information about languages and
multilingual repositories is provided in “Part 14: Multilingual Support.”

5. MDM indicates that the source and destination values have been
mapped by placing a green thunderbolt ( ) in the Mapped column of
both grids, and by placing the source and destination mapped values
into the corresponding column of the other grid.
TIP ►► You can map any non-NULL source value to the <NULL>
destination value to discard the source value. Alternatively, you can
map a <NULL> source value to any non-NULL destination value.

Unmapping Values
You can unmap one or more previously mapped source values as
described in this section.

 To unmap one or more source values from their mapped destination

values:
1. In the Source Values grid, select the mapped source values that you
want to unmap.
TIP ►► Alternatively, you can select the mapped destination values in
the Destination Values grid.

NOTE ►► When you select a mapped source value in the Source

Values grid, the corresponding mapped destination values are
automatically selected in the Destination Values grid (but not the other
way around).

2. Click on the Unmap button (between the values grids) to unmap the
source values from their mapped destination values.
NOTE ►► The Unmap button is focus-dependent and unmaps the
selected values from the grid that either: (1) currently has the focus; or
(2) last had the focus if neither values grid currently has the focus.

3. MDM indicates that the source and destination values have been
unmapped by removing the green thunderbolt ( ) from the Mapped
column of both grids for the unmapped values (but only if they are no
longer mapped to or from any other values), and by removing the
source and destination mapped values from the corresponding column
of the other grid.

Adding New Destination Values

You can add one or more new values to the list of destination values in
either of two ways, as described in this section.

304 MDM Import Manager Reference Guide

 To add and map one or more source values to the list of destination
values:
1. In the Source Values grid, select the source values that you want to
add.
2. Click on the Add button (between the values grids) to add the source
values to the list of destination values.
NOTE ►► The Add button is disabled if the display field of the
destination lookup field is compound and the source field is not
compound, or if the single destination display field is not of type Text,
Text Normalized, Text Block, Integer, or Measurement.

NOTE ►► You must use MDM Data Manager to import new images,
PDFs, and binary objects into the repository, since the MDM Import
Manager imports links to these objects, not the objects themselves.
NOTE ►► If the object lookup destination field is a text block field and
the Data Groups hierarchy is displayed in the Destination Values grid,
you can add a new text block value into a specific data group. If the
Data Groups hierarchy is not displayed, the text block is added to a
Data Group named Imported Text Blocks.

3. MDM: (1) adds the Converted Value of each source value that does
not already exist to the list of destination values in blue; (2) maps all of
the selected source values to their corresponding destination values;
(3) indicates that the source and destination values have been mapped
by placing one of the blue thunderbolts ( ) in the Mapped column of
both grids; and (4) places the source and destination mapped values
into the corresponding column of the other grid.
NOTE ►► Values that do not already exist are added and all values
are mapped.

 To add one or more values directly to the list of destination values:

1. In the Destination Values grid, right-click anywhere and choose Add
Value from the context menu, or choose Values > Add Destination
Value from the main menu.
NOTE ►► The Add Value commands are disabled if the display field
of the destination lookup field is compound or if the single destination
display field is not of type Text, Text Normalized, or Text Block.

2. MDM adds a new value named “New Value” just after the last value in
the list of Destination Values, and highlights it for editing (Figure 198).

MDM Import Manager Reference Guide 305

 Figure 198. Adding a new destination value

3. Type the new value you want and press Enter.

4. MDM adds the new value in blue.

Adding New Destination Hierarchy

Recall from the previous parts of this guide that you can create or
reconstitute source hierarchy as part of the import process by splitting or
partitioning source fields.
One of the most powerful features of the MDM Import Manager is the
ability – on the destination side – to not only precisely map to existing
destination hierarchy, but also to automatically create new destination
hierarchy as part of the value-mapping process.
The set of destination values appears as a hierarchy in the Destination
Values grid in the following cases:
• Hierarchy lookup field. When the mapped destination field is a
hierarchy lookup field, the destination values are the hierarchy of the
underlying lookup table.
• Hierarchy table primary display field. When the current table is a
hierarchy lookup table and the mapped destination field is the
primary display field of the hierarchy table, the destination values are
also the hierarchy of the lookup table.
• Nested partition. When you apply a nested partition to the mapped
destination field or any of its values, the resulting hierarchy of
destination value combinations appear as a hierarchy of values in the
Destination Values grid. (NYI)

When the Destination Values grid is a hierarchy of values, the Add

button (between the values grids) pops up a context menu with four
different command choices, as shown in Figure 199.

306 MDM Import Manager Reference Guide

 Figure 199. Pop-Up Add Menu

Subject to the constraints identified in Table 58 on page 300, you can

add new leaf-node destination values and also internal branch nodes to
the destination hierarchy, as summarized in Table 59 and described in
this section.

Table 59. Add to Destination Hierarchy Add Operations

Operation Description
Adds the selected source values as siblings of the selected
Add as Sibling
destination value.
Adds the selected source values as children of the selected
Add as Child
destination value.
Adds the selected internal source values and all of their
Add Branch as Sibling
children as siblings of the selected destination value.
Adds the selected internal source values and all of their
Add Branch as Child
children as children of the selected destination value.

 To add one or more internal- or leaf-node source values as branches

or as individual nodes to the hierarchy of destination values:
1. In the Source Values grid, select the source values you want to add.
NOTE ►► If the Source Values grid is a hierarchy (destination
hierarchy lookup field), the source values must all be leaf nodes or at
the same level. As a convenience, you can select and add both
internal and leaf nodes at the same time with the Add commands.

2. Click on the Add button (between the values grids) and choose from
the cascading menu:
 Add as Sibling
 Add as Child
 Add Branch as Sibling
 Add Branch as Child

MDM Import Manager Reference Guide 307

 NOTE ►► All four commands always appear when the Destination
Values grid is a hierarchy. The Add as Sibling and Add as Child
commands are always enabled, while the Add Branch as Sibling and
Add Branch as Child commands are enabled only when the Source
Values grid is also a hierarchy.

NOTE ►► The Add button is disabled if the display field of the

destination lookup field is compound and the source field is not
compound, or if the single destination display field is not of type Text,
Text Normalized, Integer, or Measurement.

3. MDM adds the Converted Value of each source value that does not
already exist to the hierarchy of destination values in blue, and maps
all of the selected leaf-node source values to their corresponding
destination values, as described in the previous section.
NOTE ►► Only leaf nodes can be mapped to leaf nodes.

NOTE ►► When the current table is a hierarchy table and the

mapped destination field is the primary display field representing the
hierarchy, MDM maps both internal and leaf nodes as part of the Add.

 To add one or more values directly to the hierarchy of destination

values:
1. In the Destination Values grid, select the value either: (1) after which
you want the new value to appear as a sibling; or (2) under which you
want the new value to appear as a child.
2. Right-click on the value and choose Add Hierarchy Value from the
context menu, or choose Values > Add Destination Hierarchy Value
from the main menu, and then choose from the cascading menu:
 As Sibling
 As Child
NOTE ►► The Add Value commands are disabled if the display field
of the destination lookup field is compound or if the single destination
display field is not of type Text, Text Normalized, or Text Block.

3. MDM adds a new value named “New Value” either: (1) as a sibling of
and just after the selected value; or (2) as the last child of the selected
value, and highlights it for editing (Figure 198).

308 MDM Import Manager Reference Guide

 Figure 200. Adding a new destination hierarchy value

4. Type the new value you want and press Enter.

5. MDM adds the new value in blue.

Removing New Values

You can remove one or more previously added destination values as
described in this section.

 To remove one or more values from the set of destination values:

1. In the Destination Values grid, select the previously added values that
you want to remove.
2. Click on the Remove button (between the values grids), or right-click on
one of the values and choose Remove Value from the context menu.
3. MDM removes the values from the set of destination values.
NOTE ►► If you remove an internal node, all of its child nodes are
automatically removed as well.

Automapping Source Values

If some or all of the source values are exactly the same as the
corresponding destination values, you can use the Automap button to
automatically map them to their matching destination values, as
described in this section.
NOTE ►► The Automap button uses the Converted Value in the
Source Values grid when comparing source and destination values.

NOTE ►► When you map a source field containing remote key values
to a destination lookup field for a table that has key mapping enabled,
the MDM Import Manager automatically automaps the source values
based on the remote keys of the lookup table for that remote system.

MDM Import Manager Reference Guide 309

 To automatically map one or more source values to their
corresponding destination values :
♦ Click on the Automap button (between the values grids). MDM
automatically maps each unmapped source value to the set of identical
destination values.
TIP ►► Each time click on the Automap button, MDM attempts to map
additional source values based on the current set of Converted Values.

NOTE ►► When you automap from a source hierarchy to a

destination hierarchy, MDM compares either the full path or just the
leaf node of the source and destination value based on the setting of
the Hierarchy Value Automap option (see “Configuration Options” on
page 62 for more information). The Full Path setting ignores the root
node of both the source and destination hierarchies.

NOTE ►► When the current table is a hierarchy table and the

mapped destination field is the primary display field representing the
hierarchy, MDM automaps both internal and leaf nodes.

NOTE ►► If the source values are delimited strings that represent a

hierarchy and you want to compare the source values against the full
path of the destination values, then: (1) use the Split into Hierarchy
command to create a source hierarchy; and (2) make sure the
Hierarchy Value Automap option is set to Full Path.

NOTE ►► As a convenience, Automap will overmap a source value if

it matches more than one destination value, even if the destination field
is single-valued. You can then use the Find Next Mapped Destination
Value command to find each mapped destination value and unmap all
but one of the multiple mapped destination values (see “Finding
Mapped Destination Values” on page 325 for more information). You
cannot import the source data if any source values remain
overmapped.

Mapping to a Compound Lookup Field

Recall that when the current table is the main table and a destination
lookup field consists of multiple display fields, the MDM Import Manager
automatically promotes the individual lookup table display fields into the
list of main table fields.
NOTE ►► Promoted lookup table display fields appear in the
Destination Fields grid as “display field n [lookup field]” and compound
lookup fields appear as “lookup field [display field 1; display field n]”
(where “display field n” is the name of the nth lookup table display field
and “lookup field” is the name of the compound lookup field).

310 MDM Import Manager Reference Guide

Whether you map directly or indirectly to the compound lookup field
depends upon how many source fields correspond to it:
• Single source field. If the source data has only a single
corresponding field, you must simply field and value map from the
applicable source field to the compound destination lookup field.
MDM uses the Multi-Valued delimiter to separate each source data
value into the multiple delimited values corresponding to the value
combinations of the compound lookup field.
NOTE ►► The edit controls in the Converted Value column of the
Source Values grid are disabled and cannot be opened for editing any
of the individual delimited values.

NOTE ►► The Add button between the Value Mapping grids and the
Add Values commands for the Destination Values grid are also both
disabled.

TIP ►► If the source value combinations represented by the single

source field do not all correspond to existing values of the destination
lookup field, you must use the Split into Multiple Fields command to
split the delimited source values into multiple source fields, and then
map them to the compound lookup field as described in this section.

NOTE ►► Whereas a multi-valued field stores multiple values in a

single field, a compound lookup field corresponds to a multi-field value
combination across each of the multiple lookup table display fields.

• Multiple source fields. If the source data has individual fields that
correspond one-for-one to the individual display fields of the lookup
table, you can field map the individual source fields to the promoted
lookup table display fields and then create a compound source field
for mapping to the compound destination lookup field, as described
in this section.

 To map to a compound lookup field when you have multiple

corresponding source fields:
1. Field map from individual source fields to the promoted lookup table
display fields.
2. Either convert source field values, or optionally value map to the
values of the individual Text display fields.
3. Use the Create Compound Field command to create a source field
combination corresponding to the display field combination.
4. Field map from the compound source field to the compound lookup
field.
5. Value map from the compound field value combinations to the value
combinations represented by the lookup field.

MDM Import Manager Reference Guide 311

Mapping to a Coupled Numeric Attribute
When the mapped destination is a coupled numeric attribute, each
source data value consists of one or more primary/coupled value pairs;
when you apply a value conversion filter, the command for each filter
leads to a cascading menu for the Primary Value and the Coupled Value,
as shown in Figure 201.

Figure 201. Cascading menu for a coupled numeric attribute

NOTE ►► A source data field that contains multiple delimited coupled

numeric value pairs requires that you set both the Coupled Numeric
and Multi-Valued delimiters. For example, “5hp @ 3500rpm; 6hp @
4500rpm; 7hp @ 5500rpm” would require that the Coupled Numeric
delimiter be set to “@” and the Multi-Valued delimiter be set to “;” (see
“Setting the Split Delimiters” on page 150 for more information).

MAPPING TO AN OBJECT LOOKUP DESTINATION FIELD

You can map a source field to an object lookup destination field just like
any other lookup field, and then map source values to destination
values in either of two ways:
• Flat list. MDM displays all of the object lookup values in a list in the
Destination Values grid and you can map to existing destination
values (and add new text values).
NOTE ►► See “Mapping a Source Value” on page 303 and “Adding
New Destination Values” on page 304 for more information on how to
map source values (for all object lookup fields) and add new
destination values (for text block lookup fields).

• Data Groups hierarchy. MDM displays all of the object names or

values as leaf nodes beneath their data group in the Data Groups
hierarchy in the Destination Values grid.
NOTE ►► For each object lookup destination field, you can set the
Data Group Display option (see “The Data Group Display Option” on
page 265) based on the requirements of the situation.

312 MDM Import Manager Reference Guide

Using the Data Groups Hierarchy
You can map a source field to an object lookup destination field using
the Data Groups hierarchy as described in this section.

 To map to object lookup fields using data groups:

1. Make sure the Value Map > Hierarchy configuration option is set to Full
Path, and the Data Group Display option for the object lookup
destination field is set to Display Data Group.
2. Partition the source data group field by the source object field to create
a hierarchy of data groups and objects.
3. Map the partitioned source data group field to the object lookup
destination field.
4. Map each of the source values by mapping the source hierarchy leaf-
node values to corresponding destination hierarchy leaf-node values.
NOTE ►► You must use MDM Data Manager to import new images,
PDFs, and binary objects into the repository, since MDM Import
Manager imports links to these objects, not the objects themselves.

NOTE ►► If the object lookup destination field is a text block field and
the Data Groups hierarchy is displayed in the Destination Values grid,
you can add a new text block value into a specific data group. If the
Data Groups hierarchy is not displayed, the text block is added to a
Data Group named Imported Text Blocks.

Adding a New Data Group

If the object lookup destination field is a text block field and the Data
Groups hierarchy is displayed in the Destination Values grid, you can
add a new data group to the Data Groups hierarchy before adding new
text block values, as described in this section.

 To add a new data group directly to the Data Groups hierarchy:

1. In the Destination Values grid, select the data group either: (1) after
which you want the new data group to appear as a sibling; or (2) under
which you want the new data group to appear as a child.
2. Right-click on the data group and choose Add Data Group from the
context menu, or choose Values > Add Data Group from the main
menu, and then choose from the cascading menu:
 As Sibling
 As Child
3. MDM adds a new data group named “New Value” either: (1) as a
sibling of and just after the selected data group; or (2) as the last child
of the selected data group, and highlights it for editing.
4. Type the new name you want for the data group and press Enter.
5. MDM adds the new data group in light blue.

MDM Import Manager Reference Guide 313

APPLYING VALUE-LEVEL VALUE CONVERSION FILTERS
When you apply a value conversion filter at the value level, the
conversion is applied to the converted value resulting from the prior
application of field-level filters, value-level filters, and manual edits.
For each value, you can apply one or more value conversion filters at
the individual value level as described in the following sections.
NOTE ►► If the Converted Value consists of multiple delimited
values, the value conversion filter is applied to each of the individual
values within the Converted Value.

NOTE ►► Unless noted, each of the value conversion filters can also
be applied at the field level (see “Applying Field-Level Value
Conversion Filters” on page 281 for more information).

Restoring Converted Values

When you apply a manual edit or a value-level conversion filter to an
individual value, the Converted Value changes from gray to black to
indicate that you have overridden inheritance from the field level.
You can use the Restore Converted Value filter to discard manual edits
and value-level conversions and restore inheritance for one or more
Converted Values, as described in this section.
NOTE ►► You can use the Restore Converted Value filter only at the
value level and not at the field level.

 To restore inheritance for one or more values:

1. In the Source Values grid, select the values for which you want to
restore inheritance.
2. Right-click on one of the values and choose Restore Converted Value
from the context menu, or choose Values > Apply Value Conversion
Filter > Restore Converted Value from the main menu.
3. MDM restores inheritance for each value, which once again appears in
gray in the Source Values grid.

Setting to Original Values

When you apply field-level conversion filters, they are automatically
applied to each source value in the Source Values grid that is inheriting
field-level conversions, so that each Converted Value will be different
from its corresponding original Value. Moreover, applying any manual
edits or value-level conversion filters will also change the Converted
Value so that it is different from the original Value.

314 MDM Import Manager Reference Guide

You can use the Set to Original Value filter to set one or more
Converted Values back to their original value – and in effect “undo” the
effects of the field-level value conversion filters and any manual edits or
value-level conversion filters – as described in this section.
NOTE ►► You can use the Set to Original Value filter only at the
value-level and not at the field level.

 To set one or more values back to their original values:

1. In the Source Values grid, select the values that you want to set back
to their original values.
2. Right-click on one of the values and choose Set to Original Value from
the context menu, or choose Values > Apply Value Conversion Filter >
Set to Original Value from the main menu.
3. MDM sets each value back to its original value, which appears in black
in the Source Values grid.

Setting Values to NULL

You cannot directly enter a NULL as the Converted Value. Instead, you
can use the Set Value to NULL filter to set one or more Converted
Values to NULL, as described in this section.
NOTE ►► When a field does not require value mapping, a source
value with a <NULL> Converted Value has a green bullet ( ) in the
Conv. column and is deemed to be properly converted.

 To set one or more values to NULL:

1. In the Source Values grid, select the values that you want to set to
NULL.
2. Right-click on one of the values and choose Set Value to NULL from
the context menu, or choose Values > Apply Value Conversion Filter >
Set Value to NULL from the main menu.
3. MDM sets each value to <NULL>.

Clearing Values
You can use the Clear Value filter to quickly clear one or more
Converted Values, so that you don’t inadvertently forget to manually
enter a value for each source value, as described in this section.
NOTE ►► When a field does not require value mapping, a source
value with an empty Converted Value is deemed to not yet be
converted.

MDM Import Manager Reference Guide 315

 To clear one or more values:
1. In the Source Values grid, select the values that you want to clear.
2. Right-click on one of the values and choose Clear Value from the
context menu, or choose Values > Apply Value Conversion Filter >
Clear Value from the main menu.
TIP ►► You can also press the Del key to clear a single value when
the cell selection is on the Converted Value cell.

3. MDM clears each value.

Applying Operators
You can use the Apply Operator filter to apply a numeric or text operator
to one or more Converted Values, as described in this section.

 To apply a numeric or text operator to one or more values:

1. In the Source Values grid, select the values to which you want to apply
the operator.
2. Right-click on one of the values and choose Apply Operator from the
context menu, or choose Values > Apply Value Conversion Filter >
Apply Operator from the main menu, to open the Apply Operator dialog
shown in Figure 202.

Figure 202. Apply Operator dialog

3. Choose from the drop-down control of operators (only some of which

will appear based on whether the destination is numeric or text):
NOTE ►► If the mapped destination is a coupled numeric attribute,
first choose Primary Value or Coupled Value from the cascading menu.

4. Enter into the Operand2 edit control the value that is to be applied by
or to the operator.
5. Click OK to close the dialog. MDM applies the operator to each value
and replaces it with the new value.

316 MDM Import Manager Reference Guide

Adjusting Measurement Units
Whenever possible, unit of measure strings are automatically
normalized against MDM’s built-in dictionary of over 2000 synonyms for
over 750 units in over 70 physical dimensions and appended to each
numeric source data value, based on the physical dimension of the
mapped destination measurement field or numeric attribute.
However, when the unit is missing entirely or simply unrecognized by
MDM, you can use the three measurement filters to set, add, or convert
the unit for one or more Converted Values, as summarized in Table 60
and described in this section.

Table 60. Measurement Value Conversion Filters

Filter Description
Sets the unit for the source values, overriding any existing unit.
Set Measurement Unit
 Affects all the selected values
Adds the unit to the source values that are missing a unit.
Add Missing Unit
 Only affects selected values without a unit
Converts the source values to the unit.
Convert to Unit
 Only affects selected values with a unit

TIP ►► The Convert to Unit measurement filter is most useful after

you have already applied either Set Measurement Unit or Add Missing
Unit, so that all source values have a unit and can be converted.

 To apply a measurement filter to one or more values:

1. In the Source Values grid, select the values to which you want to apply
the measurement filter.
2. Right-click on one of the values and choose the appropriate filter from
the context menu, or choose Values > Apply Value Conversion Filter
from the main menu, and choose the appropriate filter from the
cascading menu:
 Set Measurement Unit
 Add Missing Unit
 Convert to Unit
3. Choose the unit from the cascading menu of applicable units based on
the dimension of the mapped destination.
NOTE ►► If the mapped destination is a coupled numeric attribute,
first choose Primary Value or Coupled Value from the cascading menu,
and then choose the corresponding unit from the second-level
cascading menu of applicable units.

4. MDM sets, adds, or converts the unit for each value.

MDM Import Manager Reference Guide 317

Normalizing Values
Normalizing a value returns the capitalized alphanumeric string from
which all punctuation – including the underscore (_) – and other
“unwanted” characters have been removed. For example, the
normalized value of “abc_123-456;N/R” is “ABC123456NR”.
You can use the Normalize filter to normalize one or more Converted
Values, as described in this section.

 To normalize one or more values:

1. In the Source Values grid, select the values that you want to normalize.
2. Right-click on one of the values and choose Normalize from the context
menu, or choose Values > Apply Value Conversion Filter > Normalize
from the main menu.
3. MDM normalizes each value.

Changing the Case of Values

You can use the Change Case filter to change the case of one or more
Converted Values, such as from all UPPER CASE to Mixed Case, as
described in this section.

 To change the case of one or more values:

1. In the Source Values grid, select the values whose case you want to
change.
2. Right-click on one of the values and choose Change Case from the
context menu, or choose Values > Apply Value Conversion Filter >
Change Case from the main menu, and choose the new case from the
cascading menu:
 UPPER CASE
 lower case
 Sentence case
 Title Case
 tOGGLE cASE
3. MDM changes the case of each value.

Replacing Values
You can use the Replace filter to “find and replace” each occurrence of a
specified string in one or more Converted Values, as described in this
section.
NOTE ►► If the Converted Value consists of multiple delimited
values, the value conversion filter is applied to each of the individual
values within the Converted Value.

318 MDM Import Manager Reference Guide

 To find and replace a text string in one or more values:
1. In the Source Values grid, select the values in which you want to find
and replace.
2. Right-click on one of the values and choose Replace from the context
menu, or choose Values > Apply Value Conversion Filter > Replace
from the main menu, to open the Find and Replace dialog shown in
Figure 203.

Figure 203. Find and Replace dialog

3. In the Find What text box, type the text you want to match.
4. In the Replace With text box, type the text you want to replace the
found text.
5. Select or clear the option checkboxes:
 Match Case
 Whole Words Only
6. Click Find Next. MDM displays and highlights the first value (or
delimited value) that matches your text in the Value preview box.
7. Find and replace each occurrence of the value as follows:
 Find Next – find the next occurrence without replacing
 Replace – replace and move to the next occurrence
 Replace All – replace all occurrences without confirmation
 Cancel – terminate the operation and close the dialog
NOTE ►► If you are applying the Replace filter at the field-level, the
filter is not interactive and there are only two buttons: OK and Cancel.

8. MDM displays an informational dialog when all occurrences of the text

you want to match have been found.
NOTE ►► If MDM does not find all the occurrences of the text you
type, make sure that you have properly selected all of the applicable
rows in the Source Values grid.

NOTE ►► The Regular expression option is always grayed out and

will be removed from the Find and Replace dialog in an upcoming
release of MDM Import Manager.

MDM Import Manager Reference Guide 319

Accepting Truncated Values
When the mapped destination field is a fixed-width text field, one or
more source values may be too long to fit into the field. In this case, the
MDM Import Manager automatically truncates each source value to fit
into the field, based on its maximum width.
To alert you to this truncation (which may represent the loss of
important information), the MDM Import Manager does the following:
• Warning dialog. The MDM Import Manager displays a warning
dialog when you first map the field to indicate that one or more
source values have been truncated.
• Unconverted value. Source values that were truncated do not have
a green bullet ( ) in the Conv. column, and must be manually edited
before they are deemed to have been properly converted.

In order to properly convert each unconverted source value so that you

can proceed with the import, you must manually edit its Converted
Value. You can accept the truncated value without editing it by entering
the Converted Value cell and simply pressing Enter, which converts the
value and places the green bullet ( ) in the Conv. column.

DATA INTEGRITY ►► MDM forces you to enter the Converted Value

cell for each source value that was truncated to guarantee that you are
aware of the truncation and have had a chance to properly address it.

Alternatively, you can use the Accept Truncated Value filter to

automatically accept each truncated value, as described in this section.

 To accept the truncated value of one or more values:

1. In the Source Values grid, select the values whose truncated values
you want to accept.
2. Right-click on one of the values and choose Change Case from the
context menu, or choose Values > Apply Value Conversion Filter >
Change Case from the main menu.
3. MDM accepts and places the green bullet ( ) in the Conv. column of
each truncated value.

320 MDM Import Manager Reference Guide

Expanding Year Ranges
Often, multiple years may be represented in source data as a year
range, such as when the source value “1995-91” represents the years
1991, 1992, 1993, 1994, and 1995.
You can use the Expand Year Range filter to expand the year ranges in
one or more Converted Values into a set of individual years, for more
convenient value mapping to a year field, as described in this section.
NOTE ►► The Expand Year Range filter is enabled only if the source
field is mapped to a multi-valued field that requires value mapping.

TIP ►► If you have two source fields (e.g. Start Year and End Year),
first partition Start Year by End Year and then combine partitions with
the Combine Delimiter set to “-“.

 To expand the year ranges represented by one or more values:

1. In the Source Values grid, select the values whose year range you
want to expand.
2. Right-click on one of the values and choose Expand Year Range from
the context menu, or choose Values > Apply Value Conversion Filter >
Expand Year Range from the main menu.
3. MDM expands each year range into multiple delimited year values by
populating the multi-valued drop-down control with each individual
value, as shown in Figure 204.

Figure 204. Expanding a year range into multiple delimited year values

MDM Import Manager Reference Guide 321

Expanding Month Qualifications
Often, year ranges in the source data may be qualified with a numeric
value representing the value of a month, such as when the source value
“1999,6-2001,8” represents June 1999 through August 2001.
You can use the Expand Month filter to expand the month qualifications
in one or more Converted Values into a delimited set of qualified values,
for more convenient value mapping to automotive qualifiers, as
described in this section.

 To expand the month qualifiers represented by one or more values:

1. In the Source Values grid, select the values whose month qualifiers
you want to expand.
2. Right-click on one of the values and choose Expand Month from the
context menu, or choose Values > Apply Value Conversion Filter >
Expand Month from the main menu.
3. MDM populates the multi-valued drop-down control with the month
qualifier corresponding to each year in the expanded year range, as
shown in Figure 205.

Figure 205. Expanding month qualifiers into multiple delimited values

322 MDM Import Manager Reference Guide

FINDING UNMAPPED SOURCE VALUES
Quickly jumping to each unmapped source value can be useful when
the Source Values grid contains many values, and especially when it
contains a hierarchy, which cannot be sorted by the Mapped column.
You can use the Find Next Unmapped Source Value command to find
each unmapped source value, as described in this section.

 To find unmapped source values:

1. In the Source Values grid, select any source value.
2. Right-click on the value and choose Find Next Unmapped Source Value
from the context menu, or choose Values > Find Next Unmapped
Source Value from the main menu.
3. MDM moves the selected value in the Source Values grid from the
current selection to the next unmapped source value.
TIP ►► You can right-click on the source value and choose Find Next
Unmapped Source Value repeatedly (or simply press F3) to find each
unmapped value one after another.

NOTE ►► The Find Unmapped Source Value command also finds

unconverted values.

NOTE ►► The Find Unmapped Source Value command is most useful

when the Source Values grid contains a hierarchy; otherwise, you can
simply sort by the Mapped column of the grid.

NOTE ►► When the mapped destination field is a hierarchy lookup

field, the command ignores internal nodes and finds only leaf nodes;
when the current table is a hierarchy table and the mapped destination
field is the primary display field representing the hierarchy, the
command finds both internal and leaf nodes.

TIP ►► You can also use the Select Entire Branch > Unmapped Leaf
Nodes command to select rather than jump to each unmapped leaf
node value (see “Selecting Source Values in a Hierarchy” on page 325
for more information).

MDM Import Manager Reference Guide 323

REORDERING MAPPED DESTINATION VALUES
When you map a source value to the values of a multi-valued
destination field, the Pos. column in the Destination Values grid
identifies the ordered position of each mapped destination value.
By default, the order of the mapped destination values is as follows:
• Automap. The mapped destination values are ordered in the same
order as the delimited source values.
• Manual map. The mapped destination values are ordered in the
sequence in which the values were mapped.

After mapping, you can use the Reorder Mapped Destination Values
command to change the order of the mapped destination values for
each source value.

 To reorder the mapped destination values for a source value:

1. In the Source Values grid, select the value whose mapped destination
values you want to reorder.
2. Right click on the value and choose Reorder Mapped Destination
Values from the context menu, or choose Values > Reorder Mapped
Destination Values from the main menu, to open the Reorder Mapped
Destination Values dialog shown in Figure 206.

Figure 206. Reorder Mapped Destination Values dialog

3. Drag-and-drop the mapped destination values within the dialog to

place them into the order you would like them to be mapped.
4. Click OK when you are done to close the Reorder Mapped Destination
Values dialog and reorder the mapped destination values.
5. MDM changes the Pos. values to reflect the new ordering.

324 MDM Import Manager Reference Guide

FINDING MAPPED DESTINATION VALUES
When a destination field is multi-valued, a source value can be mapped
to multiple destination values.
Quickly jumping to each mapped destination value can be useful when
the Destination Values grid contains many values, and especially when
it contains a hierarchy, which cannot be sorted by the Mapped column
of the grid.
You can use the Find Next Mapped Destination Value command to find
each mapped destination value for a source value, as described in this
section.
TIP ►► Recall that Automap will overmap a source value if it matches
more than one destination value, even if the destination field is single-
valued. You can then use the Find Next Mapped Destination Value
command to find each mapped destination value and unmap all but
one of the multiple mapped destination values.

 To find the mapped destination values for a source value:

1. In the Source Values grid, select the source value for which you want
to find the mapped destination values.
2. Right-click on the value and choose Find Next Mapped Destination
Value from the context menu, or choose Values > Find Next Mapped
Destination Value from the main menu.
3. MDM moves the selected value in the Destination Values grid from the
current selection to the next mapped destination value.
TIP ►► You can right-click on the source field and choose Find Next
Mapped Destination Value repeatedly (or simply press F3) to find each
mapped destination value one after another.

SELECTING SOURCE VALUES IN A HIERARCHY

A Source Values grid that contains a hierarchy rather than a flat list
poses two additional challenges when you attempt to identify and select
unconverted or unmapped leaf node values for further processing (such
as editing the Converted Value or applying a value conversion filter):
• Two kinds of nodes. A hierarchy contains two kinds of nodes: (1)
internal nodes (which can often be ignored); and (2) leaf nodes
(whose values must be converted or mapped to destination values).
• No sorting. The values in the Source Values grid cannot be sorted
by the Conv. or Mapped columns, making it impossible to group all of
the unconverted or unmapped values.

MDM Import Manager Reference Guide 325

You can use the Select Entire Branch commands to conveniently select
all or some of the children of an internal node in the Source Values grid,
as summarized in Table 61 and described in this section.

Table 61. Select Entire Branch Commands

Command Description
Selects the selected internal node and all of its internal and leaf
All Nodes
node children.
All Leaf Nodes Selects all of the leaf node children of the selected internal node.
Selects all of the unmapped leaf node children of the selected
Unmapped Leaf Nodes
internal node.

 To select source hierarchy values:

1. In the Source Values grid, select the internal node whose children you
want to select.
2. Right-click on the node and choose Select Entire Branch from the
context menu, or choose Values > Select Entire Branch from the main
menu, and choose the appropriate selection from the cascading menu:
 All Nodes
 All Leaf Nodes
 Unmapped Leaf Nodes
3. MDM selects the applicable children of the selected node.
TIP ►► You can also use the Find Next Unmapped Source Value
command to jump to rather than select each unmapped leaf node
value (see “Finding Unmapped Source Values” on page 323 for more
information).

TIP ►► To quickly select unmapped values when the Source Values

grid is a list rather than a hierarchy, sort the grid by the Mapped
column to group all of the unmapped values at the bottom of the list,
and then use Shift selection to conveniently select all of them.

326 MDM Import Manager Reference Guide

HANDLING MISSING SOURCE DATA
As described in the previous sections, records of source data may be
missing data values that are required to proceed with the import. These
data values may be missing because they are implied.
When the data is implied, you can automatically create the implied
values using various features of MDM Import Manager, so that you can
populate the applicable destination fields with the correct data values as
part of the import.
The different types of implied data – and the MDM Import Manger
features designed to address them – are summarized in Table 62.

Table 62. Handling Implied Source Data

Implied Data Challenge MDM Feature
An entire field is completely missing from Use the Add Field command to add a new
the source data because all of the records field, map the new field to the applicable
have the same value. For example, there destination field, and map the single NULL
may be no Manufacturer field in the product value to the applicable value of the mapped
data file from the manufacturer itself. destination field.
Some or all of the text values of a source Map the NULL value to the applicable
field are missing and all the missing values destination value.
correspond to the same value.
Some or all of the numeric values of a Convert the NULL value to the applicable
source field are missing and all the missing numeric value.
values correspond to the same value.
All of the numeric values of a source field Use the Set Measurement Unit value
mapped to a measurement are missing a conversion filter to add the unit to every
unit, which is implied by the field name. source data value.
Some of the numeric values of a source Use the Add Missing Unit value conversion
field mapped to a measurement are filter to add the unit to every source data
missing a unit, and all the missing units value.
correspond to the same unit.
All of the coupled values of a source field Use the Apply Operator value conversion
mapped to a coupled numeric attribute are filter with the Equals operator to set the
missing and all the missing values coupled value for every source data value.
correspond to the same value.
Some of the coupled values of a source Use the Apply Operator value conversion
field mapped to a coupled numeric attribute filter with the NULL Equals operator to set
are missing and all the missing values the coupled value for every source data
correspond to the same value. value that is missing.

Mismatches Between Number of Header and Record Fields

In text files which use headers, the number of header fields may not
always match the number of "value" fields in a given record. If an import
record has fewer fields than its header, MDM automatically populates
the missing fields with blank values during the import. If the record has
more fields than the header, MDM issues an error message.

MDM Import Manager Reference Guide 327

PART 11: MATCHING RECORDS

This part of the reference guide explains how to identify one or more
matching fields or matching field combinations to match incoming
source records to existing destination records, and how you can apply
different import actions (create, update, and replace) to groups of
source records and on a record-by-record basis.

329
Overview
Once you have completed the field mapping and value conversion and
mapping steps, record matching shifts the focus from the field-oriented
processing to the more traditional record-oriented processing.
Record matching – often called key field matching – is usually the final
step in the import process, in which most import tools allow you to
identify the single key field to be used to match source records to
existing destination records, so that you can update or replace them
rather than create duplicates in the repository.
NOTE ►► This step is called record matching rather than key field
matching because you can specify non-key fields as matching fields.

NOTE ►► Record matching shifts the focus from the individual fields
(vertical view of the source data) to the individual records (horizontal
view of the source data).

Record matching in the MDM Import Manager takes basic key field
matching and extends it as follows:
• Multiple matching fields. You can specify multiple matching fields,
for a more precise match if no single field is completely reliable as a
key field.
• Matching field combinations. You can also specify matching field
combinations if no individual field identifies each destination record,
such as when MFG and PART NO together comprise the key field.
• Matches converted or mapped source values. Rather than using
the original source value, matching is based upon the normalized,
converted, or mapped source value, providing for a more accurate
and precise match against destination record values.
• Case-by-case operations. The matching process breaks the source
records into different groups based on the type of match – exact,
partial, or conflict – and then allows you to apply a different default
import action – skip, create, update, or replace – to each group.
• Record-by-record operations. You can then override the default
import action for each individual source record and each matching
destination record on a record-by-record basis, for even more
precise import control.
• Analytic capabilities. Finally, you can selectively identify for special
processing groups of records that match particular conditions, such
as “the SKU field matches but the Description field has been
changed” or “the List Price has changed by more than 20%.”

MDM Import Manager Reference Guide 331

THE MATCH RECORDS TAB
The Match Records tab contains several controls, as shown in Figure
207. These controls together allow you to: (1) meticulously match
incoming source records to existing destination records; and then (2)
efficiently and precisely specify the import action for each group of
source records and then for individual records, as follows:
• Value matching. You can use the dual-list Value Matching control to
specify a single matching field, multiple matching fields, or matching
field combinations.
• Default import actions. You can use the Default Import Actions grid
to specify the default import action for each group of source records,
with a different import action for each group.
• Record matching. You can use the pair of Record Matching grids to
override the default import action for each individual source record,
and for each source record, each of its matching destination records.

Figure 207. Match Records tab

The controls of the Match Records tab are described in Table 63.

Table 63. Controls of the Match Records Tab

Control Description
Lists the mapped destination fields, and allows you to select one
Value Matching Lists
or more as matching fields or matching field combinations.
Lists the groups of source records, and allows you to specify the
Import Actions Grid
default import action for each group.
Lists the source records, and for each source record, its
Record Matching
matching destination records, and allows you to override the
Grids
default import action for each individual record.

332 MDM Import Manager Reference Guide

 NOTE ►► Since you cannot import into the original protected version
of a checked out record, the MDM Import Manager displays – and
performs record matching against – the checked out version of each
checked out record, regardless of whether the import user is already a
member of the checkout.

The Value Matching Lists

Value matching is used to identify the one or more fields or field
combinations whose values should be used to match each source
record to one or more existing destination records.
Specifically, the MDM Import Manager uses the Mapped Destination
Fields list on the left to display the destination fields to which you have
mapped a source field, as shown in Figure 208.

Figure 208. Value Matching lists

You can then identify one or more mapped destination fields as

matching fields or matching field combinations by moving them to the
Matching Fields list on the right.
NOTE ►► The Mapped Destination Fields and Matching Destination
Fields lists are both disabled if you have not yet selected the current
source and destination tables.

The Value Matching lists are described in Table 64.

Table 64. The Value Matching Lists

List Description
Mapped 1
Lists the destination fields to which source fields have been mapped.
Destination Fields
Lists the mapped destination fields you have selected as matching
Matching Fields fields, and allows you to combine them into matching field
combinations.
1
For XML sources, limited to destination fields mapped to source fields in the current table.

NOTE ►► Not all mapped destination types can be used for record
matching. Those that appear in the Mapped Destination Fields list and
that you can select as Matching Fields are listed in Table 65.

MDM Import Manager Reference Guide 333

Table 65. Destination Types Available in Mapped Destination Fields List
1, 2, 3, 4, 5, 6, 7, 8, 9
Available as Matching Fields Not Available as Matching Fields
 AutoID  Text Large
 Integer  Lookup [Text Block]
 Real  Lookup [Text HTML]
 Currency  Lookup [Qualified]
 Date  Qualifiers
 Time  Attributes
 GM Time  Boolean
 Timestamp
 Text
 Text Normalized
 Lookup [Flat]
 Lookup [Hierarchy]
 Lookup [Main]
 Lookup [Taxonomy]
1
Cannot be multi-valued
2
Cannot be a promoted display field
3
If hierarchy table, must be the main display field
4
Cannot be a compound field (i.e. lookup to a table with multiple display fields)
5
Can be mapped only from Add, Clone or Split into Multiple Fields extended source fields
6
If mapped from extended source field, must connect to source with read/write permission
7
Must be mapped from a source field which is in the current language layer
8
If not the main table or a hierarchy table, all of the display fields must be selected
9
Record matching on a sortable field for which there is no sort index is extremely slow!

MULTILINGUAL ►► In a multilingual repository, all of the language

layer instances of a multilingual Text field appear in the Mapped
Destinations Fields list and are available as Matching Fields (where
each field appears as “fieldname <language layer>”). More information
about languages and multilingual repositories is provided in “Part 14:
Multilingual Support.”

The Default Import Actions Grid

Value matching breaks the entire set of source records into several
different groups based on the following:
• Number of matching destination records. Each source record can
match: (1) no existing destination records; (2) a single destination
record; or (3) multiple destination records (see “Matc” on page 340
for more information).
• Type of match. There are several different types of possible value
matches between each source record and the corresponding set of
matching destination records (see “Match Type” on page 340 for
more information).

334 MDM Import Manager Reference Guide

The MDM Import Manager indicate the number of records in each group
(and how many of the records in each group are active) within the
Default Import Actions grid, as shown in Figure 209.

Figure 209. Default Import Actions grid

You can then specify the Default Import Action for each group of source
records, and have the action automatically applied to each active record
in the group.
NOTE ►► The Default Import Actions grid is disabled if you have not
yet selected the current source and destination tables and at least one
matching field.

NOTE ►► See “Active Records and Inheritance” on page 349 for

more information about active records.

The columns of the Default Import Actions grid are described in Table
66. All but the Default Import Action column are read-only.

Table 66. Columns of the Default Import Actions Grid

Columns Description
The active records in each group of source records as “m of n”
Active (where ‘m’ is the number of active records in the group and ‘n’ is
the total number of records in the group) (see page 349).
The match level (see page 340):
Match Level  None
 Single
 Multiple
The match type (see page 340):
 None
Match Type  Exact
 Partial
 Conflict
The default import action (see page 346):
 Skip
Default Import Action  Create
 Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace

MDM Import Manager Reference Guide 335

The Record Matching Grids
The MDM Import Manager uses the two Record Matching grids to
display the set of source records, and for each source record, the set of
zero or more matching destination records, as shown in Figure 210.

Figure 210. Record Matching grids

For each source record, you can use: (1) the Source Records grid to
override the Import Action inherited from the Default Import Actions grid;
and (2) the Matching Destination Records grid to override the Import
Action inherited from the current source record.
NOTE ►► The Record Matching grids are disabled if you have not yet
selected the current source and destination tables and at least one
matching field.

The columns of the Record Matching grids include the fields of each
record and additional fields of color-coded record matching information,
as described in Table 66. All but the Import Action column in each grid
are read-only.

Table 67. Columns of the Record Matching Grids

Column Description
Source Records Grid
Record No. The record number of the source record in the original data source.
Active Whether or not the source record is active.
The import action for this record (gray if inheriting Default Import Action):
 Skip
Import Action  Create
 Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
The number of matching destination records:
Match Level  None
 Single
 Multiple

336 MDM Import Manager Reference Guide

 Column Description
The overall type of match:
Match Type  Exact
 Partial
 Conflict
Exact The number of Exact match matching destination records.
Partial1 The number Partial match matching destination records.
Conflict1 The number Conflict match matching destination records.
matching field [Eql]2 The number of Equal value match destination records for matching field.
matching field [Und]1, 2 The number of Undefined value match destination records for matching field.
matching field [Neq]1, 2 The number of Not Equal value match destination records for matching field.
The value of source field (displayed with the gray background if the field has
source field(s)
not been mapped).

Matching Destination Records Grid

Active Whether or not the destination record is active.
The import action for this record (gray if inheriting source Import Action):
 None
Import Action  Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
 Delete
The match type for the destination record with the current source record:
Match Type  Exact
 Partial
 Conflict
Whether the match type for the destination record with the current source
Exact 1, 2
record is Exact.
Whether the match type for the destination record with the current source
Partial1, 2
record is Partial.
Whether the match type for the destination record with the current source
Conflict1, 2
record is Conflict.
The type of value match for the destination record with the current source
record:
matching field [All]1, 2  Equal
 Undefined
 Not Equal
Whether the value match for the destination record with the current source
matching field [Eql]2
record is Equal for matching field.
Whether the value match for the destination record with the current source
matching field [Und]1, 2
record is Undefined for matching field.
Whether the value match for the destination record with the current source
matching field [Neq]1, 2
record is Not Equal for matching field.
destination field(s) The value of destination field.
1
Only present if you have specified multiple matching fields.
2
Hidden by default; unhide to display.

MDM Import Manager Reference Guide 337

RECORD MATCHING CONCEPTS
Record matching layers upon the following concepts in comparing
source and destination records to determine if they match:
• Value matches
• Record matches
• Match class
• Match type

Each of these concepts is described in the following sections.

NOTE ►► Some of the following record matching concepts are
mathematically and logically complex, and are included here only as a
reference for interested users. You do not need to fully understand the
concepts to make effective use of the record matching functions. 

Value Matches
For each matching field or field combination, there are three possible
types of value matches, as follows:
• Equal. Every pair of corresponding source and destination values is
equal (where some but not all of the matching pairs may be NULL).
• Undefined. At least one pair of corresponding source and
destination values consists of a NULL and non-NULL value, and
every other pair is equal.
• Not Equal. At least one pair of corresponding source and destination
values is not equal.
NOTE ►► For an individual matching field, a single pair of source and
destination field values is compared for each source and destination
record. For a matching field combination, a pair of values is compared
for each individual matching field in the combination.

The three types of value matches are summarized in Table 68.

338 MDM Import Manager Reference Guide

Table 68. Value Matches for a Matching Field or Field Combination
Value Match Individual Matching Field Matching Field Combination
The single pair of corresponding Every pair of corresponding source
source and destination values is and destination values is equal, and
equal and non-NULL. at least one pair is non-NULL.
Equal
Examples: Examples:
 X and X  X and X; A and A
 X and X; NULL and NULL
The single pair of corresponding At least one pair of corresponding
source and destination values source and destination values
consists of a NULL and non- consists of a NULL and non-NULL
NULL value or two NULL values. value, and every other pair is equal.
Undefined
Examples: Examples:
 X and NULL  X and X; A and NULL
 NULL and NULL  X and NULL; A and NULL
 X and NULL; NULL and NULL
The single pair of corresponding At least one pair of corresponding
source and destination values is source and destination values is not
not equal. equal.

Examples: Examples:
Not Equal  X and Y  X and X; A and B
 X and Y; A and B
 X and Y; A and A
 X and Y; A and NULL
 X and Y; NULL and NULL

NOTE ►► For the purposes of the value match, NULL and NULL is:
(1) treated as Undefined for an individual matching field; and (2)
ignored for a matching field combination.

NOTE ►► Tuple matching handles NULL values differently than

described in this section. If the NULL Interpretation option for the
tuple’s matching field is set to Overwrite, source NULL values in the
matching field will match any destination NULL value. If the NULL
Interpretation option is set to Ignore, source NULL values will not
match any destination value.

Record Matches
The destination records that match a particular source record depend
upon the value matches for the matching fields or field combinations. An
existing destination record matches a source record as follows:

MDM Import Manager Reference Guide 339

• Single matching field or field combination. The value match for
the single matching field or field combination is Equal
• Multiple matching fields or field combinations. The value match
for at least one of the matching fields or field combinations is Equal.
NOTE ►► If there is not at least one Equal value match among the
one or more matching fields or field combinations (and the underlying
pairs of source and destination values), then the source record simply
does not have a matching destination record in the repository.

Match Level
The match level for each source record indicates the number of
matching destination records. The three match levels include:
• None. The source record matches no destination records.
• Single. The source record matches exactly one destination record.
• Multiple. The source record matches more than one destination.
NOTE ►► When a source record matches multiple destination
records, each matching field or field combination may have multiple
value matches, each corresponding to a different matching record.

NOTE ►► When you use the Filter command, there is a fourth match
level named Filtered that corresponds to both source and destination
records (see “Record Filtering” on page 356 for more information).

Match Type
For each source record, the overall match type is based on: (1) the
combination of match types for each individual matching destination
record; where (2) the match type for each matching destination record is
based on the combination of value matches for each individual matching
field or field combination.
For each matching destination record, the match types include:
• Exact. All the individual value matches are Equal.
• Partial. At least one value match is Equal and at least one
Undefined; no value matches are Not Equal.
• Conflict. At least one value match is Equal and at least one value
match is Not Equal.

For a source record that matches a single destination record, the overall
match type for the source record is the same as the match type for the
single matching destination record.

340 MDM Import Manager Reference Guide

For a source record that matches multiple destination records, the
overall match type of the source record is based on the “best” match
type among each of the individual matching destination records, as
follows:
• Exact. At least one Exact match among the matching destination
records.
• Partial. No Exact matches but at least one Partial match among the
matching destination records.
• Conflict. No Exact or Partial matches; all matching destination
records are Conflict matches.
The match types are summarized in Table 69.

Table 69. Overall Match Type for a Source Record

Match Type Single Matching Destination Multiple Matching Destinations
All the individual value matches At least one Exact match among the
are Equal. matching destination records.

Example: Examples:
Exact
 Equal and Equal  Exact and Exact
 Exact and Partial
 Exact and Conflict
At least one value match is Equal No Exact matches but at least one
and at least one Undefined; no Partial match among the matching
value matches are Not Equal. destination records.
Partial
Example: Examples:
 Equal and Undefined  Partial and Partial
 Partial and Conflict
At least one value match is Equal No Exact or Partial matches; all
and at least one value match is Not matching destination records are
Equal. Conflict matches.
Conflict
Example: Examples:
 Equal and Not Equal  Conflict and Conflict

NOTE ►► The only match type possible with a single matching field
or field combination is Exact, since either a Partial or Conflict match
requires an Equal match combined with at least one other value match
(and without even a single Equal match, then the destination record
simply doesn’t match the source record).

NOTE ►► When you use the Filter command, there are two match
types named Source and Destination that correspond to the Filtered
match level (see “Record Filtering” on page 356 for more information).

MDM Import Manager Reference Guide 341

Example: Single Matching Field
Figure 211 illustrates the various value match and match type
conditions for a single matching field (SKU).
Source Destination Value Match Match Type
SKU SKU SKU [Record]
113-224 Equal Exact
113-224 NULL Undefined None
101-44 Not Equal None
113-224 Undefined None
NULL NULL Undefined None*
101-44 Undefined None

Figure 211. Single matching field (SKU)

Example: Single Matching Field Combination

Figure 212 illustrates the various value match and match type
conditions for a single matching field combination (Mfr. and PartNo.).
Source Destination Value Match Match Type
Mfr PartNo Mfr PartNo Mfr; PartNo [Record]
Baldor 22345 Equal Exact
Baldor NULL Undefined None
Baldor 4403 Not Equal None
NULL 22345 Undefined None
Baldor 22345 NULL NULL Undefined None
NULL 4403 Not Equal None
SKF 22345 Not Equal None
SKF NULL Not Equal None
SKF 4403 Not Equal None
Baldor 22345 Undefined None
Baldor NULL Equal Exact
Baldor 4403 Undefined None
NULL 22345 Undefined None
Baldor NULL NULL NULL Undefined None
NULL 4403 Undefined None
SKF 22345 Not Equal None
SKF NULL Not Equal None
SKF 4403 Not Equal None
Baldor 22345 Undefined None
Baldor NULL Undefined None
Baldor 4403 Not Equal None
NULL 22345 Equal Exact
NULL 22345 NULL NULL Undefined None
NULL 4403 Not Equal None
SKF 22345 Undefined None
SKF NULL Undefined None
SKF 4403 Not Equal None

Figure 212. Single matching field combination (Mfr. & PartNo.)

342 MDM Import Manager Reference Guide

Example: Multiple Matching Fields
Figure 213 illustrates the various value match and match type
conditions for two matching fields (SKU and UPC).
Source Destination Value Match Match Type
SKU UPC SKU UPC SKU UPC [Record]
113-224 22345 Equal Equal Exact
113-224 NULL Equal Undefined Partial
113-224 4403 Equal Not Equal Conflict
NULL 22345 Undefined Equal Partial
113-224 22345 NULL NULL Undefined Undefined None
NULL 4403 Undefined Not Equal None
101-44 22345 Not Equal Equal Conflict
101-44 NULL Not Equal Undefined None
101-44 4403 Not Equal Not Equal None
113-224 22345 Equal Undefined Partial
113-224 NULL Equal Undefined Exact
113-224 4403 Equal Undefined Partial
NULL 22345 Undefined Undefined None
113-224 NULL NULL NULL Undefined Undefined None
NULL 4403 Undefined Undefined None
101-44 22345 Not Equal Undefined None
101-44 NULL Not Equal Undefined None
101-44 4403 Not Equal Undefined None
113-224 22345 Undefined Equal Partial
113-224 NULL Undefined Undefined None
113-224 4403 Undefined Not Equal None
NULL 22345 Undefined Equal Partial
NULL 22345 NULL NULL Undefined Undefined None
NULL 4403 Undefined Not Equal None
101-44 22345 Undefined Equal Partial
101-44 NULL Undefined Undefined None
101-44 4403 Undefined Not Equal None

Figure 213. Multiple matching fields (SKU and UPC)

MDM Import Manager Reference Guide 343

Example: Multiple Matching Fields with Field Combination
Figure 214 illustrates the various value match and match type
conditions for a three matching fields, consisting of an individual
matching field (SKU) and a matching field combination (Mfr. and
PartNo.).
Source Destination Value Match Match Type
SKU Mfr PartNo SKU Mfr PartNo SKU Mfr; PartNo [Record]
113-224 Baldor 22345 Equal Equal Exact
113-224 Baldor NULL Equal Undefined Partial
113-224 Baldor 4403 Equal Not Equal Conflict
113-224 NULL 22345 Equal Undefined Partial
113-224 NULL NULL Equal Undefined Partial
113-224 NULL 4403 Equal Not Equal Conflict
113-224 SKF 22345 Equal Not Equal Conflict
113-224 SKF NULL Equal Not Equal Conflict
113-224 SKF 4403 Equal Not Equal Conflict
NULL Baldor 22345 Undefined Equal Partial
NULL Baldor NULL Undefined Undefined None
NULL Baldor 4403 Undefined Not Equal None
NULL NULL 22345 Undefined Undefined None
113-
Baldor 22345 NULL NULL NULL Undefined Undefined None
224
NULL NULL 4403 Undefined Not Equal None
NULL SKF 22345 Undefined Not Equal None
NULL SKF NULL Undefined Not Equal None
NULL SKF 4403 Undefined Not Equal None
101-44 Baldor 22345 Not Equal Equal Conflict
101-44 Baldor NULL Not Equal Undefined None
101-44 Baldor 4403 Not Equal Not Equal None
101-44 NULL 22345 Not Equal Undefined None
101-44 NULL NULL Not Equal Undefined None
101-44 NULL 4403 Not Equal Not Equal None
101-44 SKF 22345 Not Equal Not Equal None
101-44 SKF NULL Not Equal Not Equal None
101-44 SKF 4403 Not Equal Not Equal None

Figure 214. Multiple matching fields (SKU and Mfr. & PartNo.)

344 MDM Import Manager Reference Guide

VALUE MAPPING AND RECORD MATCHING: A COMPARISON
Recall that value mapping against a destination lookup field validates
source data values against the set of legal destination values, that the
value of the lookup field corresponds to the value of the lookup table
display field or display field combination, and that each lookup field
value corresponds to a record of the underlying lookup table.
Value mapping therefore not only validates source data values against
the set of lookup field values, but also in effect matches them against
existing lookup table records, so that you can map to existing records
rather than create duplicates in the lookup table.
In the same way, record matching uses the values of the matching fields
you specify to match source records against existing destination table
records, so that you can update or replace existing records rather than
create duplicates in the current destination table.
Record matching and value mapping thus complement one another to
match rather than create duplicates of existing destination records.
NOTE ►► When the current table is the main table, record matching
and value mapping prevent duplicates of main table and lookup table
records, respectively. When the current table is a lookup table, they
prevent duplicates of lookup table records, with the match based on:
(1) the mapped value if you are value mapping; and (2) on the record
match if you are not value mapping.

Record Matching and Lookup Table Display Fields

When the current table is a lookup table, record matching is required
against the display field as a matching field. When the lookup table has
multiple display fields, the required record matching is to all of the
display fields as a matching field combination. When the lookup table is
a hierarchy table, the record matching is required only to its primary
display field, regardless of the number of lookup table display fields.
NOTE ►► Value mapping against lookup fields is required to match
lookup table records because record matching is reserved for matching
main table records. By contrast, when the current table is the lookup
table, record matching is available for matching lookup table records,
and so value mapping to the display fields is either optional or not
required, depending upon the field type (see “Value Mapping and
Lookup Table Display Fields” on page 246 for more information).

NOTE ►► When the current table is a lookup table: (1) the Mapped
Destination Fields list by default includes all of the display fields of the
table; (2) each display field appears as “field [DF]” and appears in gray
until it has been mapped; (3) the Add as Individual command is
disabled for display field items; and (4) all of the display fields must be
added together as a matching field combination.

MDM Import Manager Reference Guide 345

Consider a main table with a Manufacturer lookup field into the
Manufacturers lookup table, which has display fields Name and Code.
When the current table is the Manufacturers table, record matching is
required against Name and Code as a matching field combination. If the
Manufacturers lookup table is a hierarchy table, then record matching
would be required to the Name field alone rather than with Code.
NOTE ►► When the current table is a lookup table, the [Remote Key]
field can also be used as a matching field, as follows: (1) if you map
values for the display field(s) of a flat table or the primary display field
of a hierarchy table, the [Remote Key] field can be used as a matching
field only if it is combined with the display field(s) or the primary display
field; and (2) if you do not map values for the display field(s), it can be
used as the matching field by itself (see “Lookup Table Imports Using
the [Remote Key] Field” on page 414 for more information).

IMPORT ACTIONS
The various import actions that you can apply to source and destination
records are described in Table 70.

Table 70. Import Actions for Source and Destination Records

Import Actions Description
Source Records
Skip Skip this source record during import processing.
Create a new destination record based on this source
Create
record.
Update just the NULL fields of the matching existing
Update (NULL Fields Only)
destination record(s) based on this source record.
Update the matching existing destination record(s) with
Update (All Mapped Fields)
all of the mapped fields of this source record.
Replace the matching existing destination record(s) by
Replace first deleting them and then creating a new destination
record based on this source record.
Destination Records
Do not perform any action on this destination record
None
during import processing.
Update just the NULL fields of this destination record
Update (NULL Fields Only)
based on the source record.
Update this destination record with all of the mapped
Update (All Mapped Fields)
fields of the source record.
Replace this destination record by first deleting it and
Replace then creating a new destination record based on the
source record.
Delete Delete this destination record during import processing.

346 MDM Import Manager Reference Guide

When you choose to update or replace existing destination records, the
appropriate import action depends upon how comfortable you are
overwriting existing destination data based on what you believe about
the quality and completeness of the incoming source data, as described
in Table 71.

Table 71. Different Levels of Import Actions

If the source data is… Import Action
Use the Replace import action, to replace matching
Accurate and complete
destination records in their entirety.
Use the Update (All Mapped Fields) import action, to
Accurate and incomplete
preserve unmapped fields but overwrite existing data.
Less accurate than existing Use the Update (NULL Fields Only) import action, to fill
destination data in missing data but not overwrite existing data.

Import Action Inheritance

Import actions are inherited and can be set at several different levels, as
follows:
• Default Import Actions grid. You can set a different Default Import
Action for each combination of Match Level and Match Type. The
Default Import Action is automatically applied to each corresponding
source record in the group for which the import action has not been
manually set at the source record level, and also to each matching
destination record for which the import action has not been manually
set at the destination record level.
• Source Records grid. When you manually set the import action for
an individual source record, it is applied to the source record and to
each matching destination record for which the import action has not
been manually set at the destination record level.
• Matching Destination Records grid. When you manually set the
import action for an individual destination record that matches a
source record, it is applied to the destination record.
NOTE ►► When you set the Default Import Action for a group of
source records, each action in the Import Action column in both the
Source Records and Destination Records grids appears in gray to
indicate that it is inheriting the default. When you manually set the
import action for an individual source or destination record, the action
then appears in black to indicate that you have overridden inheritance
from the Default Import Actions grid or the Source Records grid,
respectively. You can use the Inherit Default command to restore
inheritance for a record.

MDM Import Manager Reference Guide 347

The Default Import Action mechanism allows you to quickly set the
import action for large groups of source records, and then to precisely
override the default on a record-by-record basis.

Source and Destination Record Import Actions

For each combination of Match Level and Match Type, there is a
different set of available import actions in the Default Import Actions
grid, and a corresponding set of available import actions for each source
record in the Source Records grid, and each matching destination
record in the Destination Records grid, as summarized in Table 72.

Table 72. Import Actions by Match Level and Match Type

Match Level / Source Actions Corresponding Destination Actions
Match Type (default in bold) (default in bold)

None Skip n/a

None Create n/a

None
Skip
Single Delete

Exact Update (NULL Fields Only) Update (NULL Fields Only)

Partial Update (All Mapped Fields) Update (All Mapped Fields)
Conflict
Replace Replace

None
Skip
Delete

None
Update (NULL Fields Only)
Update (NULL Fields Only)
Multiple Update (All Mapped Fields)
Delete
Exact
Partial None
Conflict Update (NULL Fields Only)
Update (All Mapped Fields)
Update (All Mapped Fields)
Delete

None
Replace
Replace

Filtered Skip n/a

Source Create n/a

Filtered Skip Skip

Destination Deleted Deleted

NOTE ►► When you use the Filter command, the Source Records
grid includes all of the invalid source records, plus a dummy source
record that corresponds to the leftover destination records, which
appear in the Destination records grid when the dummy is selected.

NOTE ►► The Match Records tab populates the Source Records grid
with all source records – including duplicates – regardless of the Merge
Source Records setting, which only affects how records are imported.

348 MDM Import Manager Reference Guide

 NOTE ►► When a source record is to be imported against multiple
matching destination records, the MDM Import Manager applies the
default destination record action (inherited either from the default
import action or from the source import action) to all of the multiple
matching records; if only some of them should be affected, you must
override the inherited default on a destination record-by-record basis.

NOTE ►► When the same destination record is a match for more

than one source record, each source record may specify a different
action for the destination record. In this case, the destination record
may need to be cloned so that each of the source record actions can
be properly and individually applied to the destination record, as
described in Table 73.

Table 73. Cloning Requirements When Importing Source Records

Cloning Requirements for a Destination Action for Second Source Record
Destination Record None Update Replace Delete
None OK OK OK OK
Destination
Action for Update OK Clone Clone Clone
First Source Replace OK Clone Clone Clone
Record
Delete OK Clone Clone Clone

Active Records and Inheritance

Recall from the previous sections that when you set a Default Import
Action for a group of records, it is automatically applied to and inherited
by each source record in the group, and for each source record, further
applied to and inherited by each matching destination record.
Similarly, when you manually set the Import Action for an individual
source record, it is applied to the source record, and further applied to
and inherited by each matching destination record.
In fact, this inheritance occurs only for those records that are currently
active when you set the import action.
By default, every source record in each Match Level / Match Type group
of source records is active when you first select the one or more
matching fields or matching field combinations.
You can then limit the set of active records within each group based on
the individual value matches and other analytic criteria. This breaks the
import records into smaller, more precise subgroups, and allows you to
apply import actions to the active subgroup of records in each group.
Subgroups of records within each group can be made active based on
each of the three value match types, as described in Table 74.

MDM Import Manager Reference Guide 349

Table 74. Setting Active Records
Value Match Tag Active Records
Equal [Eql] Records for which at least one value match is Equal.
Undefined [Und] Records for which at least one value match is Undefined.
Not Equal [Neq] Records for which all value matches are Not Equal.

NOTE ►► The more stringent requirement (“all” rather than “at least
one”) is necessary in this last case to avoid making it too broad and
causing all records to be active.

SOURCE RECORD MERGING

Sometimes, multiple source records may represent different instances
of the same product record (even if some fields have different values).
For example, recall that with automotive parts data, the source data
typically represents application records rather than product records, and
will contain multiple records per product, where each source record
represents a single application of the product (see “Product Applications
and Automotive Parts Data” on page 257 for more information).
Duplicate source records are more than academic. As part of the source
aggregation process, the MDM Import Manager can detect and either:
(1) preserve these multiple instances of the same record; or (2) merge
them automatically into a single virtual source record prior to importing.
It does this based on the matching field, which is used not only for
identifying matching destination records, but also for distinguishing
between and identifying duplicate source records.
Merging is one of the most powerful aggregation capabilities of the
Import Manager, and the ability to either preserve or collapse duplicates
provides even finer user control over the aggregation process.

The Key Field

If you have specified multiple matching fields or field combinations, each
one might cause different sets of source records to be considered
duplicates for the purposes of the import and source record merging.
When there are multiple matching fields or field combinations, you can
set any single matching field or field combination to be the key field for
the purpose of identifying duplicate source records (see “Setting the Key
Field” on page 367 for more information).
NOTE ►► By default, the first matching field or field combination is
automatically designated as the key field.

350 MDM Import Manager Reference Guide

The Merge Source Records Option
Depending upon how you would like the Import Manager to import the
source data, you can specify that duplicate source records based on the
key field are either preserved or merged, as described in Table 75.

Table 75. The Merge Source Records Option

Setting Description
Duplicate source records are preserved as individual
Off
records for import as a record group.
Duplicate source records are collapsed and merged for
On
import as a single virtual source record.

You can use the Merge Source Records configuration option to specify
whether duplicate source records should be preserved or collapsed (see
“Configuration Options” on page 62 for more information).
Prior to import, the MDM Import Manager groups each distinct set of
duplicate source records, and then processes each group as follows:
1. Records in the group whose import action is Skip are ignored and
removed from the group.
2. The import action of the last remaining record in the group is used
as the import action for the entire group (“last one wins”).
NOTE ►► The "last remaining record" may not necessarily be the last
record to appear in the source file.

3. If either: (a) Merge Source Records is On; or (b) the import action of
the group is an Update action, then the duplicate source records in
the group are merged into a single record as follows:
• Single-valued fields. The values are collapsed, with the last non-
NULL value in the group used as the value of the field in the
merged record (“last one wins”).
• Multi-valued fields. The values are combined, with the set of
distinct values across the entire set of duplicate records in the
group used as the value of the field in the merged record.
NOTE ►► Even when Merge Source Records is Off, records of each
group for which the import action is an Update action are always
merged because multiple sequential updates would lead to the same
last-one-wins behavior anyway; by merging first, a single update can
instead be applied to the destination record after the duplicate source
records have been merged into a single record.

4. If the records were not merged in step (3), identical source records
in the group are collapsed, and each of the remaining individual
records is preserved for import as a group rather than one-by-one.

MDM Import Manager Reference Guide 351

5. Finally, either the group of duplicate source records or the single
merged record is imported, creating one or more new destination
records, or updating or replacing one or more existing destination
records.

Consider the set of four duplicate source records (based on the SKU
key field) and corresponding import actions shown in Figure 215.

SKU Description Color [MV] Price Import Action

#1 203 Widget Blue $3.49 Update (NULL Fields Only)
#2 203 Widget Plus Green $3.51 Replace
#3 203 Blue $3.58 Update (All Mapped Fields)
#4 203 Widget Red $3.69 Skip

Figure 215. Duplicate source records prior to merge

When processed: (1) the last record is ignored because its import action
is Skip; (2) the import action of the group is therefore Update (All
Mapped Fields); (3) the three remaining records are therefore merged,
regardless of the Merge Source Records setting, as shown in Figure
216; (4) the value of the Description field is taken from the second
record because its value in the third record is NULL; (5) the value of the
Color multi-valued field is combined from all three records; and (6) the
value of the Price field is taken from the third record.

SKU Description Color [MV] Price Import Action

203 Widget Plus Blue; Green $3.58 Update (All Mapped Fields)

Figure 216. Duplicate source records after merge

The Merge Qualified Links Option

Just as the Merge Source Records option determines how to handle
duplicate source records, the Merge Qualified Links option determines
how to handle duplicate qualified links based on the matching qualifiers
(see “The Qualified Update” on page 253).
Specifically, the subrecords representing the qualified links are first
combined as a multi-valued field as a result of the source record
merging described in the previous section. Then you can specify that
duplicate subrecords based on the matching qualifiers are either
preserved or merged, as described in Table 76.

352 MDM Import Manager Reference Guide

Table 76. The Merge Qualified Links Option
Setting Description
Duplicate subrecords are preserved for import as a set
Off
of qualified links.
Duplicate subrecords are collapsed and merged for
On
import as a single qualified link.

You can use the Merge Qualified Links configuration option to specify
whether the duplicate links represented by the subrecords should be
preserved or collapsed (see “Configuration Options” on page 62 for
more information).
NOTE ►► If a group of duplicate source records is not being merged
(i.e. Merge Source Records is Off and the import action for the group is
not an Update action), then merging qualified links merges duplicate
qualified links (represented by a group of delimited subrecords) only
within each individual source record, not among and across the entire
group of duplicate source records.

After merging source records, the MDM Import Manager groups each
distinct set of duplicate subrecords, and then processes each group as
follows:

Groups representing the value of a new qualified link:

1. If the import action for new links is Skip, the group is ignored.
2. If Merge Qualified Links is On, then the duplicate subrecords are
merged into a single qualified lookup value.

Groups matching the value of one or more existing qualified links:

3. If the import action for existing links is Skip, the group is ignored.
4. If either: (a) Merge Qualified Links is On; or (b) the import action for
existing links is an Update action, then the duplicate subrecords in
the group are merged into a single subrecord as follows:
• Single-valued qualifiers. The values are collapsed, with the last
non-NULL value in the group used as the value of the qualifier
in the merged subrecord (“last one wins”).
• Multi-valued qualifiers. The values are combined, with the set of
distinct values across the entire set of duplicate subrecords in
the group used as the value of the qualifier in the merged
subrecord.

MDM Import Manager Reference Guide 353

 NOTE ►► Even when Merge Qualified Links is Off, subrecords are
always merged if the import action for existing links is an Update action
because multiple sequential updates would lead to the same last-one-
wins behavior anyway; by merging first, a single update can instead be
applied to the qualified link after the duplicate subrecords have been
merged into a single subrecord.

All groups:
1. If the subrecords were not merged in step (2) or (4), identical
subrecords in the group are collapsed, and each of the remaining
individual subrecords is preserved for import as a group rather than
one-by-one.
2. Either the group of duplicate subrecords or the single merged
subrecord is imported, creating one or more new qualified links, or
updating or replacing one or more existing qualified links.
NOTE ►► When importing application records, each source record is
still treated as a product record, one that represents a single value for
the qualified lookup field that stores the application. When merged, the
group of source records for each product is collapsed into a single
virtual record representing the multiple applications for the product.

To Merge or Not to Merge

The primary effects of merging are to: (1) assemble non-NULL values
from multiple records or links; (2) create multi-valued field values; and
(3) eliminate duplicate or bad data (although “last-one-wins” provides no
guarantee that the correct value will be chosen for single-valued fields).
Merging source records is most useful for creating multi-valued fields,
particularly when aggregating the qualified links represented by each of
multiple source records into the multiple links of a qualified lookup field.
Even when not merging, the grouping that always occurs facilitates the
Create of duplicate source records, and also many-to-many Replace, as
follows: (1) with record merging off, you can replace a group of existing
destination records with a group of incoming source records (based on
the key field); and (2) with record merging on and qualified link merging
off, you can replace a group of existing qualified links with a group of
incoming qualified links (based on the matching qualifiers).
NOTE ►► With or without merging, duplicate source records that are
completely identical across all fields (not just the key field), and
duplicate qualified links that are completely identical across all
qualifiers (not just the matching qualifiers) are always collapsed.

354 MDM Import Manager Reference Guide

 NOTE ►► Merging source records is a distinct concept from matching
and updating existing destination records. When merging records,
each group of duplicate source records is first merged, and then each
single surviving merged record is matched against and its values used
to create a new or update or replace an existing destination record.

MULTILINGUAL ►► In a multilingual repository, the source record

merging logic collapses multiple records that contain different language
layers of data for the same record. More information about languages
and multilingual repositories is provided in “Part 14: Multilingual
Support.”

Merging Qualified Links and the Qualified Update Option

Recall that the Qualified Update option allows you to specify matching
qualifiers, that the Update setting uses them for matching existing
qualified links, and that the Append setting implicitly uses all of the
qualifiers as matching qualifiers. In each case, the setting of the
Qualified Update option is used to determine how incoming source data
is processed against existing qualified lookup values only when the
import action for the group of records is Update (All Mapped Fields).
However, when merging qualified links, MDM again relies on these
matching qualifiers to identify duplicate subrecords for grouping and
merging – even when the import action for the record group is Create,
Replace, or Update (NULL Fields Only) – and the Qualified Update
option and corresponding matching qualifiers must therefore be set
accordingly for proper merging.
NOTE ►► Merging of records or subrecords occurs when either: (1)
the corresponding Merge option is On; or (2) the import action for the
group of existing links is an Update action.

Specifically, when the Qualified Update option is set to Inherit Default,

MDM uses the setting of Matching Qualifiers configuration option to
aggregate import records, as follows:
• None. Use only the non-qualifier display fields.
• All. Use the non-qualifier display fields plus all the qualifiers.

By contrast, when the Set Qualified Update command has been used to
override the global default setting of the Qualified Update option, MDM
uses the non-qualifier display fields plus the matching qualifiers you
specify in the Set Matching Qualifiers dialog.

MDM Import Manager Reference Guide 355

RECORD FILTERING
Record filtering expands the concept of import by complementing the
source-centric set of import records with a destination-centric set of
records that includes not only matched records but also unmatched
records in the filtered destination set.
Specifically, it allows you to specify basic search criteria to filter the
source and destination records in conjunction with various operations
within MDM Import Manager and MDIS, and when performing the import
itself.
Filtering narrows down both the source and destination record sets, but
with slightly different effects, as follows:
• Source. Narrows down the set of source records so that those that
do not meet the filtering criteria can be excluded from the import. The
filter is exclusive, and source records that do not meet the filtering
criteria are deemed invalid.
• Destination. Narrows down the set of destination records, excluding
those that do not meet the filtering criteria from record matching. The
filter is inclusive, and expands the set of records on which you can
perform import operations to those that do meet the filtering criteria.
NOTE ►► Invalid source records that do not meet the filtering criteria
are “filtered out.” By contrast, the leftover destination records that are
unmatched but do meet the filtering criteria are “filtered in.”

Full Refresh vs. Net Change Synchronizations

Typically, the MDM Import Manager is used to perform a “net change”
synchronization, in which a file of source records is imported to update
and replace matching destination records and create new ones, leaving
untouched those destination records that do not match source records.
Consider, by contrast, a “full refresh” requirement, in which a file of
source records must be imported and should update and replace a
corresponding set of destination records in its entirety, whether or not
every destination record matches a source record.
In particular, source records that meet the filtering criteria but do not
exist should be created, those that meet the filtering criteria but do exist
should be used to update or replace existing destination records, and
those that do not meet the filtering criteria should be ignored.
Similarly, destination records that meet the filtering criteria and match
source records should be updated or replaced, and those that meet the
filtering criteria but do not match source records should be deleted. In
effect, all existing destination records that meet the filtering criteria must
be affected by the import, not just those matching source records.

356 MDM Import Manager Reference Guide

Consider the example import illustrated in Figure 217, which consists of
120 source records (only 110 of which meet the filtering criteria), and
100 destination records that meet the filtering criteria (only 80 of which
match source records), for a total of 80+30+20=130 distinct source and
destination records that meet the filtering criteria.

Source: Destination:

Matched (80) ◄ Update/Replace ► Matched (80)

Delete ► Leftover (20)

Unmatched (30) ◄ Create
100 total

Invalid (10) ◄ Skip

120 total

Figure 217. Record filtering example

In the example, 80 source records update or replace 80 matching

destination records, 30 new source records are created, 10 invalid
source records are skipped, and 20 leftover destination records are
deleted.

Filtered Match Level and Match Types

Recall that there are three match levels for each source record and
three match types, for a total of seven match level / match type
combinations:
• None/None
• Single/Exact
• Single/Partial
• Single/Conflict
• Multiple/Exact
• Multiple/Partial
• Multiple/Conflict

MDM Import Manager Reference Guide 357

When you use record filtering, two additional match level / match type
combinations are enabled in the Default Import Actions grid, comprised
of an additional Filtered match level and two additional match types for
source and destination records, as follows:
• Filtered/Source. Invalid source records that do not meet the
filtering criteria.
• Filtered/Destination. Leftover destination records that do meet the
filtering criteria but are not matched by source records.

You can use the Filter command to filter by one or more lookup values
of any lookup field that is used for record matching, and both the source
and destination records will be forced to match those lookup value(s).
NOTE ►► Be careful if you set the import action for invalid source
records to Create, because they may correspond to destination records
that were not matched because MDM used the filtering criteria to limit
the set of destination records that participated in record matching.

Record Filtering Behavior

The record filtering behavior of the MDM Import Manager is summarized
in Table 77.

Table 77. Record Filtering Behavior of the MDM Import Manager

Level Description
The filter command brings up a dialog on the selected
Filter Command lookup field that allows you to select one or more lookup
values by which to filter.
Record matching is performed only against filtered
Record Matching destination records that match the search criteria rather than
the entire repository.
You can add filtering criteria to any lookup field used as a
Matching Fields Grid matching field; MDM adds [Filtered] to the matching field or
field combination.
Two match level / match type combinations correspond to
Default
filtered records: Filtered/Source (invalid source records) and
Import Actions Grid
Filtered/Destination (leftover destination records).
Import actions correspond to Filtered/Source records (Skip /
Filtered
Create) and Filtered/Destination records (Skip / Update
Import Actions
(NULL Fields Only) / Update (All Mapped Fields) / Delete).
Includes all of the invalid source records, plus a dummy
Source Records Grid source record that corresponds to leftover filtered
destination records.
Includes the leftover filtered destination records when the
Destination Records Grid dummy source record is selected in the Source Records
grid.

358 MDM Import Manager Reference Guide

THE VIRTUAL [MATCH LEVEL/TYPE] FIELD
Typically, the match level / match type value combination for each
imported or updated record (including filtered destination records, if you
are using record filtering) is transient, used for the duration of the import
session for the purposes of assigning and performing import actions,
and then discarded by MDM once the import is complete.
Sometimes, however, you may want to actually store each new and
updated record’s match level / match type value combination in the
repository itself, perhaps for user inspection, or so that a workflow can
automatically branch or perform validations based on the value.
To support this requirement, the MDM Import Manager includes for
every source table a virtual field named [Match Level/Type], which
appears in: (1) the list of source fields in the Source Hierarchy; and (2)
the Source Fields grid of the Map Fields/Values tab.
This field always contains nine values, each corresponding to one of the
match level / match type value combinations (see Table 72 on page 348
for more information), which appear in the Value Conversion and
Mapping grid of the Map Fields/Values tab when it is selected in the
Source Fields grid.
You can then map this field to a destination field in the MDM repository,
and then map each of its values to a corresponding value in the
destination field, and MDM will then update the destination value for
each imported record as part of the import.
NOTE ►► The destination field can be of type Text or Lookup, and
you must already have created its nine values prior to the import, since
adding source field values to the list of destination field values is not
permitted.

UPDATING THE RECORD MATCHING FIELD

The values of the record matching field are used to match incoming
source records against existing destination records.
When creating new destination records or replacing existing destination
records, the source record value is always used as the value of the
record matching field, since in each case, the source record results in
an entirely new destination record.
By contrast, when updating existing destination records as part of the
import, the value of the source record may or may not be used to
update the value of the record matching field.
You can use the Do Not Update Record Matching Field Value
configuration option to specify how the set of incoming source values
are treated, as described in Table 78.

MDM Import Manager Reference Guide 359

Table 78. Do Not Update Record Matching Field Value Option
Setting Description
Update the values of the record matching field with the
No
source values.
Do not update the values of the record matching field
Yes
with the source values.

Whether or not the value of the record matching field is updated may
seem academic, since for a matching destination record, the source and
destination values will presumably be the same.
However, there are two circumstances where it does makes a difference
whether or not MDM updates the existing value of record matching field:
• Time Stamp field. If the field is updated, a Time Stamp field that
tracks the record matching field will see the field value as having
been updated; otherwise, the Time Stamp value will not be changed.
• Text Normalized field. If the field is updated, a record matching field
of type Text Normalized may be updated with source values that are
formatted differently than the existing values.
NOTE ►► The next section discusses this and other considerations
for a record matching field of type Text Normalized.

NORMALIZED RECORD MATCHING FIELD

When a record matching field is of type Text Normalized, several
challenges relate to how to match and how to update existing values.
You can use the Normalized Record Matching Field configuration option
to specify whether to perform normalized or exact matches between
source values and the values of the record matching field, as
summarized in Table 79.

Table 79. Normalized Record Matching Field Configuration Option

Setting Description
Normalized Match Against Perform a normalized match of normalized source
Normalized Value values against normalized destination values.
Exact Match Against Perform an exact match of literal source values
Non-Normalized Value against non-normalized destination values.

Normalized matches have the advantage of matching existing values

even when the source value is formatted differently. For example, with a
normalized match, source value 123-ABC will match 123/ABC because
the match will compare normalized values 123ABC to 123ABC.

360 MDM Import Manager Reference Guide

Using a normalized match with a record matching field of type Text
Normalized increases the likelihood that incoming source records can
be properly matched against existing destination records. At the same
time, several considerations arise when using a normalized match:
• Too many matches. The normalized match may result in matches
against too many destination records, some of which should not
match. For example, with a normalized match, 123-ABC will not only
match 123/ABC but also 12-3/ABC, 123 [ABC], and 12.3-ABC.
TIP ►► For each source record that matches multiple destination
records, you can use the Matching Destination Records grid of the
Record Matching tab to identify which of the matching records is the
correct match and which should be skipped during the import.

• Incorrect source formatting. In first example above, if the correct

format is 123/ABC, you would want the source value 123-ABC to
match the destination value 123/ABC but not to overwrite it, since the
existing format was correct.
TIP ►► You can use to the Update Record Matching Field Value
configuration option to determine whether or not the record matching
field value should be updated with incoming source values, as
described in the previous section.

MASK UPDATES AND RECORD REPLACEMENT

When the import replaces existing destination records with incoming
source records, the masks that contain the replaced records must be
updated at the same time.
For each Replace operation, how the masks are updated depends upon
how many source records replace how many matching destination
records, as summarized in Table 80.
NOTE ►► Recall that source records may be grouped rather than
merged, and that each group may match one or more existing
destination records.

MDM Import Manager Reference Guide 361

Table 80. Mask Updates and Record Replacement
Number of Number of
Replacement Matching
Description
Source Destination
Records Records
Every mask containing the 1 matching destination
m 1 record will have it replaced by the m replacement
source records.
Every mask containing the 1 matching destination
1 1 record will have it replaced by the 1 replacement source
record.
Every mask containing any of the n matching
1 n destination records will have them replaced by the 1
replacement source record.
Every mask containing all of the n matching destination
records will have them replaced by the m replacement
source records.
m n
Masks containing some but not all of the n matching
destination records will have them removed from the
mask.

KEY MAPPINGS AND RECORD REPLACEMENT

Existing key mappings are always preserved, even for destination
records set to be replaced with the Replace action, as follows: (1) if a
source field is mapped to the [Remote Key] field, MDM will reject the
source record rather than perform the replace; and (2) if a source field is
not mapped to the [Remote Key] field, MDM will perform the replace
after first transferring the [Remote Key] values of the target replaced
records to the new record. In each case, the Replace action will not
result in the loss of [Remote Key] values.

CHECKOUT AND RECORD REPLACEMENT

When an import is set to check out the records of an import, it is not
possible to replace existing destination records with incoming source
records, and MDM will therefore not allow you to perform the import if
any of the import actions are set to Replace.
Specifically, if the Checkout/Workflow configuration option is set to
Checkout or to a workflow that performs a checkout, you must either: (1)
eliminate the Replace import action(s) by changing them to Update; or
(2) set the configuration option to None or to a non-checkout workflow.
NOTE ►► MDM uses the Import Status tab to inform you that the
import cannot proceed because of Replace with checkout.

362 MDM Import Manager Reference Guide

CHECKOUT AND RECORD MATCHING
Recall that an MDM user within the MDM Data Manager can check out
a record, and that other users – subject to join permissions – can
become members of the checkout. Non-members of the checkout see
the original version of the record; members of the checkout see the
checked out version (unless they use the All Versions command).
Notwithstanding the above, regardless of whether or not an import user
is already a member of the checkout, the MDM Import Manager always
displays – and performs record matching against – the checked out
version of each checked out record.
NOTE ►► See “Import, Checkout, and Workflow” on page 378 for
more information about checkout- and workflow-related import failures.

RECORD MATCHING ON NON-SORTABLE FIELDS

The MDM Console allows you to specify that no sort index be built for a
field that would otherwise be sortable. You can use such fields as
record matching fields; however, the record matching process will be
extremely slow, and the MDM Import Manager displays a warning dialog
when you select such a field for record matching.

MDM Import Manager Reference Guide 363

Record Matching Operations
The record matching operations are described in the following sections,
and summarized in Table 81.

Table 81. Record Matching Operations

Operation Description
Matching Field Lists
Add
Adds the selected fields to the list of matching fields.
Matching Field
Remove Removes the selected fields from the list of matching
Matching Field fields.
Combine Combines the selected matching fields into a
Matching Fields matching field combination.
Split Splits the selected matching field combination into
Matching Fields individual matching fields.
Sets the selected matching field to be the key field
Set Key Field
used to identify duplicate source records.

Set Active Records Sets the value matches that should be active.

Sets the filtering criteria for source and destination

Filter Records
records.
Default Import Actions Grid
Set Default Sets the default import action for a Match Level /
Import Action Match Type group of source records.
Source Records Grid
Set Source Sets the import action for the selected source
Record Import Action records.
Matching Destination Records Grid
Set Destination Sets the import action for the selected matching
Record Import Action destination records.

NOTE ►► Whenever you Add, Remove, Combine, or Split matching

fields as described in the following sections, the MDM Import Manager
performs record matching on the new set of matching fields and field
combinations, and then when record matching is complete, updates:
(1) the Default Import Actions grid to reflect the number of source
records falling into each Match Level / Match Type group; (2) the
Source Records grid to reflect the record matching status of each
source record; and (3) the Matching Destination Records grid to reflect
the record matching status of each matching destination record.

364 MDM Import Manager Reference Guide

ADDING A SINGLE MATCHING FIELD
You can add a single matching field as described in this section.

 To add a single matching field:

1. In the Mapped Destination Fields list, select the field that you want to
make a matching field, and click on the Add button.
2. MDM performs record matching based on the new set of matching
fields and displays a progress dialog to indicate the status of the record
matching, as shown in Figure 218.

Figure 218. Record matching progress dialog

ADDING MULTIPLE MATCHING FIELDS

When the number of source records and/or the number of destination
records is very large, the record matching process can take some time.
If you intend to match on multiple fields – either as individual matching
fields or as a matching field combination – it can be more efficient to
add them all at once as described in this section.

 To add multiple matching fields:

1. In the Mapped Destination Fields list, select the fields that you want to
make matching fields.
2. Click on the Add button and choose from the cascading menu:
 Add as Individual
 Add as Combination
3. MDM performs record matching based on the new set of matching
fields and field combinations and displays a progress dialog like the
one shown in Figure 218 to indicate the status of the record matching.
NOTE ►► When the current table is a lookup table: (1) the Mapped
Destination Fields list by default includes all of the display fields of the
table; (2) each display field appears as “field [DF]” and appears in gray
until it has been mapped; (3) the Add as Individual command is
disabled for display field items; and (4) all of the display fields must be
added together as a matching field combination.

MDM Import Manager Reference Guide 365

COMBINING MATCHING FIELDS
You can combine two or more matching fields that were previously
added separately into a matching field combination as described in this
section.

 To combine two or more existing matching fields:

1. In the Mapped Destination Fields list, select the two or more fields that
you want to combine into matching field combinations.
2. Click on the Combine button, or right-click on one of the fields and
choose Combine Matching Fields from the context menu.
3. MDM performs record matching based on the new set of matching
fields and field combinations and displays a progress dialog like the
one shown in Figure 218 to indicate the status of the record matching.

SPLITTING A MATCHING FIELD COMBINATION

You can split a matching field combination into individual matching fields
as described in this section.

 To split a matching field combination:

1. In the Mapped Destination Fields list, select the matching field
combination that you want to split into individual matching fields.
2. Click on the Split button, or right-click on the field combination and
choose Split Matching Fields from the context menu.
3. MDM performs record matching based on the new set of matching
fields and field combinations and displays a progress dialog like the
one shown in Figure 218 to indicate the status of the record matching.

REMOVING MATCHING FIELDS

Removing matching fields is a matter of simply selecting the fields and
removing them from the Matching Fields list as described in this section.

 To remove one or more matching fields:

1. In the Matching Fields list, select the one or more matching fields you
want to remove, and click on the Remove button.
2. MDM performs record matching based on the remaining set of
matching fields and field combinations and displays a progress dialog
like the one shown in Figure 218 to indicate the status of the record
matching.

366 MDM Import Manager Reference Guide

SETTING THE KEY FIELD
Recall that if you specify multiple matching fields or field combinations,
the first one is, by default, automatically designated as the key field for
the purposes of merging duplicate source records. You can designate a
different matching field or field combination as the key field, as
described in this section.

 To change the key field:

1. In the Matching Fields list, select the matching field or field
combination that you want to be the key field.
2. Right-click on the matching field and choose Key Field from the context
menu, or choose Records > Set Key Field from the main menu.
3. MDM appends [Key] to the name of the selected matching field or field
combination to indicate that it is now the key field.

SETTING THE ACTIVE RECORDS

Recall that by default, every source record in each Match Level / Match
Type group of source records is active when you first select the one or
more matching fields or matching field combinations.
You can then limit the set of active records within each group based on
the individual value matches and other analytic criteria. This breaks the
import records into smaller, more precise subgroups, and allows you to
apply import actions to the active subgroup of records in each group.
Subgroups of records within each group can be made active based on
each of the three value match types as described in this section.

 To change the value match criteria for which records are active:
1. In the Matching Fields list, select the matching field or field
combination for which you want to change the value match criteria.
2. Right-click on the matching field and choose from the context menu, or
choose Records > Set Active Records from the main menu, and choose
from the cascading menu:
 All Matches Active [All]
 Equal Matches Active [Eql]
 Undefined Matches Active [Und]
 Not Equal Matches Active [Neq]
3. MDM appends the selected tag to the name of the selected matching
field or field combination to indicate the value matches that are active.
NOTE ►► [All] both is replaced by and completely replaces [Eql],
[Und], and [Neq], which are themselves added to one another.

MDM Import Manager Reference Guide 367

SETTING THE RECORD FILTERING CRITERIA
Recall that if you select one or more lookup fields to be matching fields,
you can filter the set of source and destination records that participate in
an import by their lookup values.
Filtering by lookup values narrows or expands the set of source and
destination records by:
• Excluding all source records which do not have the filter value.
• Including all destination records which do have the filter value.

Sometimes, however, you may not want to set the filter to a specific
lookup value. For example, if you set your map to filter records by the
Manufacturer lookup value, BMW, you must then change this filter when
you want to import Toyota records, and then save the change to a new
map. You then have to repeat this process for each new Manufacturer
value you want to import.
To avoid such inconveniences, Import Manager now includes a
[Mapped Values] filter option, which automatically sets the filter criteria
to whatever lookup values are mapped in the source file.
NOTE ►► If you are importing via MDIS, the [Mapped Values] filter
option can import whatever unmapped lookup values are in the source
file as well (see “The MDIS Unmapped Value Handling Option” on
page 268 for more information).

Filtering by [Mapped Values] expands the set of source and destination

records by:
• Including all source records with mapped lookup values.
• Including source records with unmapped lookup values (MDIS only).
• Including all destination records that have any mapped source value.

With the [Mapped Values] filter option, you can create flexible, reusable
maps that automatically filter the lookup values in your source records,
freeing you from having to change filters and/or create new maps for
each new lookup value you want to filter by.
You can set the record filtering criteria for source and destination
records as described in this section.

368 MDM Import Manager Reference Guide

 To set the record filtering criteria for a matching field:
1. In the Matching Fields list, select the matching field for which you want
to set the filtering criteria.
2. Right-click on the matching field and choose Filter from the context
menu, or choose Records > Set Filter from the main menu to open the
Filter dialog shown in Figure 219:

Figure 219. Filter dialog

NOTE ►► If the matching lookup field has previously been combined

with another matching lookup field, Filter first brings up a cascading
menu of lookup field names for you to select the one for which you
want to set the filtering criteria.

NOTE ►► The [Mapped Values] filter option cannot be combined with

any other values.

3. Choose one or more lookup values to use as record filtering criteria.

4. MDM appends [Filtered] to the name of the selected matching field or
field combination to indicate that record filtering is in effect.

MDM Import Manager Reference Guide 369

SETTING A DEFAULT IMPORT ACTION
You can set the Default Import Action for a group of source records as
described in this section.

 To set the Default Import Action for a group of source records:

1. In the Default Import Actions grid, move the cell to the Default Import
Action column of the Match Level / Match Type for which you want to
set the default import action.
2. Click on the down-triangle to open the drop-down control and choose
the appropriate action from the list, or choose Records > Set Default
Import Action from the main menu, choose the Match Level / Match
Type from the cascading menu, and choose the appropriate action
from the cascading menu:
For None/None (no matching destination records):
 Skip
 Create
For others (matching destination records):
 Skip
 Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
3. To set the import action for the source records and override the default
setting, choose any of the import actions, or alternatively, to restore
inheritance, choose Inherit Default.

NOTE ►► MDM places a check next to the cascading menu choice

that corresponds to the current import action unless you have selected
multiple records for which the import actions are not all the same.

370 MDM Import Manager Reference Guide

SETTING A SOURCE RECORD IMPORT ACTION
You can set the import action for one or more source records and
override inheritance from the Default Import Actions grid as described in
this section.
NOTE ►► Inherited import actions appear in the Source Records grid
in gray.

 To manually set the import action for one or more source records:
1. In the Source Records grid, select the records for which you want to
set the import action.
2. Right-click on one of the records and choose the appropriate action
from the context menu, or choose Records > Set Source Import Action
from the main menu, and choose the appropriate action from the
cascading menu:
 Skip
 Create
 Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
 Inherit Default

3. To set the import action for the source records and override the default
setting, choose any of the import actions, or alternatively, to restore
inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current import action unless you have selected
multiple records for which the import actions are not all the same.

MDM Import Manager Reference Guide 371

SETTING A DESTINATION RECORD IMPORT ACTION
You can set the import action for one or more destination records and
override inheritance from both the Default Import Actions grid and the
matching source record as described in this section.
NOTE ►► Inherited import actions appear in the Destination Records
grid in gray.

NOTE ►► A destination record can appear more than once for two
different source records.

 To manually set the import action for one or more destination

records:
1. In the Destination Records grid, select the records for which you want
to set the import action.
2. Right-click on one of the records and choose the appropriate action
from the context menu, or choose Records > Set Destination Import
Action from the main menu, and choose the appropriate action from
the cascading menu:
 None
 Create
 Update (NULL Fields Only)
 Update (All Mapped Fields)
 Replace
 Delete
 Inherit Default
3. To set the import action for the destination records and override the
default setting, choose any of the import actions, or alternatively, to
restore inheritance, choose Inherit Default.
NOTE ►► MDM places a check next to the cascading menu choice
that corresponds to the current import action unless you have selected
multiple records for which the import actions are not all the same.

372 MDM Import Manager Reference Guide

PART 12: IMPORTING DATA

This part of the reference guide explains how to actually perform the
import, along with the final “pre-flight” checks to make sure that all the
necessary steps have been performed prior to import.

373
Import Execution
Every step in the import workflow must be complete before you can
perform the import. Moreover, the MDM Import Manager will not permit
you to perform the import until two specific conditions have been met:
• Value mapping and conversion. For every mapped field, each of
its values must either be properly converted to the destination data
type or mapped to an existing or newly added destination data value.
• Record matching. You must specify one or more unique fields so
that the MDM Import Manager can determine whether to create new
records or update and replace existing destination records.

MONITORING IMPORT STATUS

You can always determine what needs to be done next by checking the
Action Items at the top of the Import Status tab, which identifies the next
required action, as shown in Figure 220.

Figure 220. Import Status tab

MDM Import Manager Reference Guide 375

The Import Status tab also lists pre-import statistics that indicate what
will happen when you perform the import, as summarized in Table 82.

Table 82. Pre-Import Statistics

Item Status Information
 Source table name
Table Mapping  Destination table name
 # of Mapped fields (m of n)
 From table/field name
Table Joins
 To table/field name
 From table/field name
Field Lookups
 To destination table field name
 Source field name
Field Mapping (without Value Mapping)  Destination field name
 # of values converted (m of n)
 Source field name
 Destination field name
Field Mapping (with Value Mapping)  # of values mapped (m of n)
 # of values added
 # of values overmapped
 Source field name
Matching Fields
 Destination field name
None/None  # of records matched
Single/Exact  # of records matched
Single/Partial  # of records matched
Single/Conflict  # of records matched
Record Matching Multiple/Exact  # of records matched
Multiple/Partial  # of records matched
Multiple/Conflict  # of records matched
Filtered/Source  # of invalid source records
Filtered/Destination  # of leftover destination records
Skip  # of source records
Create  # of source records
Update (NULL Fields Only)  # of source records
Import Actions
Update (All Mapped Fields)  # of source records
Replace  # of source records
Delete  # of destination records

376 MDM Import Manager Reference Guide

PERFORMING THE IMPORT
When the Action Items on the Import Status tab reads Ready to Import,
you can perform the import as described in this section.

 To execute the import:

♦ Click the Execute Import toolbar button (shown at left), or choose
Records > Execute Import from the main menu.
NOTE ►► The Execute Import toolbar button and Execute Import
command will not be enabled unless the Import Status tab is the active
tab, even if the Import Manager is ready to import.

NOTE ►► Import is “forgiving” around individual record failure. In

other words, when you import a file of source records, the import will
proceed even if some of the records fail to import. (See “Import,
Checkout, and Workflow” on page 378 for more information about
checkout- and workflow-related import failures.)

NOTE ►► Currently, the MDM Import Manager does not check any
security constraints set in the MDM Console. Specifically, with the
following exceptions: (1) the import will fail if any record constraints are
set for the import user; and (2) you must have write access to the
Family Hierarchy if the Synchronize Family Hierarchy After Import
configuration option is set to Yes.

NOTE ►► If two users are importing to the same repository, the

second import will wait for the first to finish, and then proceed; results
may be somewhat unexpected since new data imported by the first will
not be reflected in value mapping or record matching of the second.

REFRESHING THE RECORD MATCHING

To ensure that the current state of the record matching corresponds to
the current state of the records in the MDM repository, you can refresh
the record matching prior to performing the import, as described in this
section.

 To refresh the record matching prior to import:

♦ Click the Refresh Record Matching toolbar button (shown at left), or
choose Records > Refresh Record Matching from the main menu.
NOTE ►► The Execute Import toolbar button and Execute Import
command will not be enabled unless the Import Status tab is the active
tab, even if the Import Manager is ready to import.

MDM Import Manager Reference Guide 377

Import, Checkout, and Workflow
MDM offers a variety of mechanisms to deal with the challenges around
import, checkout and workflow, as described in the following sections.

THE AUTO-JOIN CHECKED OUT RECORDS OPTION

Recall that an MDM user within the MDM Data Manager can check out
a record, and that other users – subject to join permissions – can
become members of the checkout. Non-members of the checkout see
the original version of the record; members of the checkout see the
checked out version (unless they use the All Versions command).
A challenge arises when the MDM Import Manager attempts to import
source records into destination records that are checked out, but for
which the import user is not already a member of the checkout. In other
words, the import user’s view of these destination records would be the
original version of the record rather than the checked out version.
DATA INTEGRITY ►► The “safest” behavior would be to reject
source records that attempt to modify destination records that have
been checked out for which the import user is not already a member of
the checkout. However, the MDM Import Manager needs to be able to
support automation and high-volume imports without triggering multiple
exceptions, especially when the import user has join permissions and
joining the checkout group is not an issue.

MDM addresses this challenge in several ways. First, regardless of

whether or not the import user is already a member of the checkout, the
MDM Import Manager displays – and performs record matching against
– the checked out version of each checked out record (which makes
sense since the original version is protected anyway).
Second, MDM can join the checkout of checked out records
automatically within the MDM Import Manager based on the setting of a
configuration option (and subject to join permissions).
Specifically, you can use the Auto-Join Checked Out Records
configuration option to specify whether or not to automatically join the
checkout of checked out records whenever possible, as summarized in
Table 83.

Table 83. Auto-Join Checked Out Records Configuration Option

Setting Description
Do not automatically join the checkout of checked
No
out records.
Automatically join the checkout of checked out
Yes
records whenever possible.

378 MDM Import Manager Reference Guide

THE CHECKOUT/WORKFLOW OPTION
Recall that record checkout is a staging mechanism that creates a
private “checked out” copy of a record. It allows MDM users to make a
series of changes to the checked out version over an extended period of
time while the visible original version remains untouched, and then have
those changes become visible all at once only after the record has been
checked back in (or discarded if the checked out record is rolled back).
Recall also that workflow, in conjunction with checkout, takes staging
one step further to make it a guided staging mechanism. Workflow with
checkout encapsulates a set of one or more records as a job, checks
them out, guides the job through a sequence of steps that represent
tasks for the users associated with each step, and then checks them
back in at the end of the job (or rolls them back).
Within the MDM Data Manager, checking out records, adding records to
workflow jobs, and launching workflow jobs are operations that are all
performed manually.
By contrast, MDM can perform these operations automatically within the
MDM Import Manager based on the setting of a configuration option
(and subject to a variety of limitations that are designed to respect the
constraints of checkout and workflow while maintaining data integrity).
Specifically, you can use the Checkout/Workflow configuration option to
specify whether or not to check out the imported records or launch a
workflow job containing the imported records (which may itself check
out the records), as summarized in Table 84.

Table 84. Checkout/Workflow Configuration Option

Setting Description
Do not check out imported records or launch a
None
workflow job.
Check out the imported records non-exclusively
Checkout
prior to the import.
Launch the specified workflow job that contains
Workflowname
the imported records.

NOTE ►► Starting in MDM 7.1 SP3, the Checkout setting results in

either a Check Out Nonexclusive or a Check Out New Record
Nonexclusive depending upon whether the imported record already
exists or must be created.

MDM Import Manager Reference Guide 379

 NOTE ►► Import checkout and import workflow launch are mutually
exclusive. However, regardless of whether imported records are
checked out by the import itself or by the workflow job that is launched
by the import, they are checked out prior to updating existing
destination records or creating new records in the repository.

NOTE ►► The Import Manager automatically launches the specified

workflow immediately and without delay when the import is complete,
regardless of the autolaunch settings of the workflow itself.

CAUTION ►► Importing records into a workflow job has severe

performance implications.

CHECKOUT AND WORKFLOW FAILURE CONDITIONS

The previous sections describe MDM Import Manager behaviors – and
a pair of configuration options – that support import automation in the
face of the various constraints on checkout and workflow. Sometimes,
however, these constraints cannot be avoided for all records, and the
import of some of the source records may fail.
NOTE ►► Such constraints include: (1) records that are already
checked out cannot be added to a workflow; (2) records that are
already in a workflow cannot be checked out; and (3) records can be in
multiple workflows only if none of them checks out its records.

The conditions under which source records will fail to import for each of
the settings of the Checkout/Workflow configuration option (and
sometimes dependent upon the Auto-Join Checked Out Records
configuration option) are summarized in Table 85.

Table 85. Checkout and Workflow Failure Conditions

Setting Fails if Destination Record is…
 Checked out and Auto-Join=No – or –
None  Checked out and Auto-Join=Yes and no join permission

 Success independent of whether or not record is in workflow

 In workflow – or –
Checkout
 Checked out by another user

Workflow  Checked out

(w/o checkout)  Success independent of whether or not record is in workflow
Workflow  In workflow – or –
(w/ checkout)  Checked out

380 MDM Import Manager Reference Guide

 NOTE ►► The failure to import individual records helps to assure that
the integrity of the overall import is respected. For example, if the MDM
Import Manager were configured to launch a workflow following the
import, and a source record corresponding to an already checked out
destination record were allowed to be imported, the record import
would succeed but the subsequent workflow launch would fail.

A NOTE ABOUT IMPORT AND VALIDATIONS

MDM does not run validations upon import, which might initially seem
like a shortcoming of the system. In fact, this is by design; if validations
were run on import and invalid records simply rejected, there would be
no way for the user to determine which records failed to import and why.
By contrast, allowing the invalid records into the repository allows you to
use the full search and edit capabilities of MDM on the invalid records.
Moreover, rather than burden the import process with complex
validations, a better solution is to make the validations part of a
workflow in which import is the first step. Specifically, import launches a
workflow job with the records checked out and then performs the
validation as part of the workflow to validate the records before checking
them back in. In this way, the import can proceed quickly and efficiently
but invalid records will never become visible in the repository until and
unless they are corrected.

MDM Import Manager Reference Guide 381

PART 13: IMPORTING RELATIONSHIPS

This part of the reference guide explains how to import relationship links
for parent/child relationships.

383
Importing Parent/Child Relationship Links
You can use the MDM Import Manager to import relationship links into
parent/child product relationships, as described in this section.
Specifically, the import file that defines the parent/child relationship links
is relationship-centric, with one parent/child pair for each link placed on
each line or row of the import file, as shown in Figure 221.

SKU <Kits> Part No <Parts> Quantity Required

135-A 1234 3 0
135-A 5678 2 1
135-A 9012 1 1

Figure 221. File format of parent/child relationship pairs

When importing relationship links, the MDM Import Manager behaves

as follows:
• Relationship selection. The drop-down list of destination tables
includes the parent/child relationships as the last group of items.
Choose the relationship into which you want to import relationship
links (see “The Current Tables” on page 36 and “Selecting the
Current Tables” on page 50).
• Destination fields list. The Destination Fields grid of the Map
Fields/Values tab includes all the parent and child table field names,
each qualified with the corresponding relationship name in angular
brackets (e.g. “SKU <Kits>” and “Part Number <Parts>”), and each
appearing twice if the parent and child tables are the same. The list
also includes the relationship qualifiers, Quantity and Required.
• Field mapping. You must map the source fields to the
corresponding destination fields that uniquely identify each parent
relationship record and each child relationship record, as well as the
source fields that correspond to the relationship qualifiers.
• Value mapping. Since all of the mapped fields correspond to index
fields, no value mapping is required; however, you must map the
values in the source field mapped to the Required relationship
qualifier to the Boolean values No and Yes.
• Record matching. The Match Records tab is disabled when the
destination table is a relationship; all the matching is done by the
Master Data Server during the import itself based on the entire set of
mapped source and destination fields as a group.
NOTE ►► Any link records that refer to parent or child records that do
not already exist in the repository will be flagged as an error and the
corresponding link will not be imported.

MDM Import Manager Reference Guide 385

RELATIONSHIP CONFIGURATION OPTIONS
Several configuration options control how the set of relationship links for
each record are updated or replaced, as summarized in Table 86 and
described in the following subsections.

Table 86. Relationships Configuration Options

Option Value Description
Groups source records based on the
Parent
parent record.
Aggregate By
Groups source records based on the
Child
child record.
Appends to the set of existing links,
Append
leaving existing links untouched.
Deletes the set of existing links before
Links Update Replace
adding the imported links.
Updates the set of existing links and/or
Update
their qualifier values.

New Links Skip On Update, skips new links.

(Update Only) Create On Update, creates new links.
On Update, skips imported links that
Skip correspond to existing links, leaving
existing links untouched.
On Update, updates existing links by
overwriting just the specified
Update (Mapped Qualifiers)
relationship qualifier values with the
Existing Links specified import value.
(Update Only)
On Update, deletes the existing link
before adding the imported link,
Replace effectively overwriting all the
relationship qualifier values either with
the specified import value or the default.
Delete On Update, deletes existing links.

NOTE ►► Unlike Import Actions, which includes the Update (NULL

Fields Only) choice, Existing Links does not include an Update (NULL
Qualifiers Only) choice because relationship qualifiers cannot be NULL.

NOTE ►► See “Configuration Options” on page 62 for more

information about the Relationships options.

NOTE ►► See “Aggregation Option” on page 387 for more

information about the Aggregate By option.

NOTE ►► See “Update Options” on page 387 for more information

about the Update options.

386 MDM Import Manager Reference Guide

Aggregation Option
Recall that as part of the record matching process for the normal import
of source records into the main table or any of the subtables, the MDM
Import Manager can group duplicate source records based on the value
of the key field (see “Source Record Merging” on page 350), and then
either preserve or collapse these duplicate instances based on the
setting of the Merge Source Records configuration option.
NOTE ►► When you import relationship links, the Match Records tab
is disabled because the entire set of mapped fields collectively defines
the record matching field or field combination for both the parent table
records and the child table records.

With normal import, multiple source records correspond to the same

destination record only in certain circumstances. By contrast, when you
import relationship links, there are almost always likely to be multiple
instances of each parent table record (identifying multiple linked
children) and/or each child record (identifying multiple linked parents).
When the Links Update configuration option is set to Replace, MDM
wipes out existing links before importing the new links. In this case, the
end result of importing relationship links depends upon whether you
group the links based on the parent table records or the child table
records, as follows:
• Aggregate by parent. For each parent record for which links are
specified in the source file, MDM wipes out the entire set of existing
child links before importing the new links.
• Aggregate by child. For each child record for which links are
specified in the source file, MDM wipes out the entire set of existing
parent links before importing the new links.

You can use the Aggregate By configuration option to specify whether

to group by the parent table records or by the child table records (see
“Configuration Options” on page 62 for more information).

Update Options
Recall that the Qualified Update option has three settings: Append,
Replace, and Update. Moreover, for the Update setting, you can use the
Set Matching Qualifiers dialog not only to identify which qualifiers should
be used for matching, but also to specify a different Update action for
new links and for existing links.
Similarly, the Relationships > Links Update configuration option has
three settings: Append, Replace, and Update, and for the Update setting,
respects the setting of the New Links and Existing Links configuration
options. These three configuration options have the same set of
interrelated behaviors as the Qualified Update option.

MDM Import Manager Reference Guide 387

Specifically, the settings of the Links Update configuration option
behave as follows:
• Append. Appends to the set of existing related records (leaving the
set of existing links untouched).
• Replace. Completely replaces the set of existing related records (by
first eliminating the entire set of existing links).
• Update. Updates the set of existing related records based on the
settings of the New Links and Existing Links configuration options.
NOTE ►► Updating existing links is really about updating the qualifier
values; nothing else about an existing link is updated.

NOTE ►► The New Links option allows you to Skip or Create new
links. The Existing Links option allows you to Skip, Update, Replace, or
Delete existing links, which only affects the relationship qualifiers.

NOTE ►► MDM does not permit the relationship qualifiers to have

NULL values, so if: (1) no qualifier value is specified for an imported
link; and (2) an existing link does not already exist or is being replaced,
the default qualifier values are Quantity=1 and Required=No.

NOTE ►► Links Update=Append is equivalent to Links

Update=Update with New Links=Create and Existing Links=Skip.

NOTE ►► Existing Links= Delete allows the import file to specify the
links to delete, allowing you to conveniently remove existing links
rather than add new links or update existing links as part of the import.

RELATIONSHIP FILE FORMAT

The file format for importing parent/child relationship links using the
MDM Import Manager is the same as that used with the Relationship
commands within the MDM Client, and consists of multiple rows, each
identifying a single parent/child pair, and for each pair, a value for the
Quantity and Required relationship qualifiers, as described in Table 87.

Table 87. File Format for Importing Parent/Child Relationship Links

Relationship Import File Format

388 MDM Import Manager Reference Guide

  The fields include the display fields of the parent table and the display
fields of the child table (both sides of the relationship even if it is a
single-table relationship).
 The fields also include the Required and Quantity relationship
Parent/Child
qualifiers (if they are defined for the relationship).
 Each row defines a single pair of related parent/child records, listing
the field and qualifier values that identify a single parent record and a
single child record.
 The Required and Quantity fields are optional; if they appear, not all
parent/child pairs need to have a value specified in the data file.
 If the relationship link being imported does not exist in the repository
and no value is specified for a qualifier, the default qualifier values are
Required=No and Quantity=1 (MDM does not permit relationship
Qualifiers
qualifiers to have NULL values).
 If the relationship link being imported already exists in the repository,
the existing value for a qualifier is: (1) overwritten if a value is specified
in the data file; and (2) unchanged if no value is specified for the
qualifier in the data file.

MDM Import Manager Reference Guide 389

PART 14: MULTILINGUAL SUPPORT

This part of the reference guide contains a general overview of

multilingual support within the MDM system and a specific description of
the multilingual features within the MDM Import Manager. Multilingual
support allows you to import multilingual data into a single MDM
repository.

391
Multilingual Basics
Basic multilingual concepts are explained in the following sections.

LANGUAGE LAYERS
When you define an MDM repository as multilingual, MDM stores the
multilingual data and metadata in multiple language layers, one for each
language. A single language repository has a single layer; a multiple
language repository has multiple layers.
The best way to understand language layers is to start by considering a
typical unstructured approach to storing multiple languages for a field by
creating multiple instances of the same field, as shown in Figure 222.

Part Number Product Color (Eng) Color (Fre) Color (Ger)

113 T-Shirt Red Rouge Rot
114 T-Shirt Green Vert Grün
115 T-Shirt Blue Bleu Blau

Figure 222. A typical table with three Color fields for three languages

The table above contains three Color fields side-by-side, one for each
language (English, French, and German), and can be successfully used
to store the multilingual color data within the table.
Unfortunately, the system would know nothing of the relationship among
the fields, so it cannot offer the user a language-centric view of the data,
and the user has no way of knowing that the fields are related (except
that the field names above have been tagged with the corresponding
language). Finally, all the field names themselves exist only in English.
Now consider the MDM approach that uses multiple language layers to
represent the multiple languages, as shown in Figure 223.

German
French
English

Figure 223. An MDM table with three language layers

Multiple layers efficiently organize and structure both multilingual data

and metadata, with a single Color field above containing multiple data
buckets rather than multiple Color fields that are completely unrelated,
and multiple language-specific field names for all of the fields.

MDM Import Manager Reference Guide 393

LANGUAGE INHERITANCE
To support convenient user access to multilingual data, the MDM Import
Manager provide a language-centric view of data within a multilingual
repository, meaning that data is presented from the point of view of a
particular language layer at a time. This single language is called the
current language, and you select it when you first connect to a
multilingual repository.
NOTE ►► The language-centric view determines not only which
language of data is displayed, but also which language of metadata is
displayed, including table names, field names, and attribute names.

Now consider a multilingual field that is missing data in the current

language. In a single-language repository, the value is shown as empty
or NULL. However, MDM uses an innovative inheritance scheme to
display – and color-code – data from other language layers for data
values that are missing in the current language.
The actual value shown depends on the language inheritance defined
for the current language. The language inheritance identifies the priority
sequence of language layers from which to find a non-NULL value to
display when the current language layer is NULL.
NOTE ►► Language inheritance is a type of layer transparency that
allows individual data values to “show through” from other language
layers when the current layer is missing data.

Language inheritance is set for each language, and is defined by the

administrator as the ordering of all the other languages of the repository,
split into: (1) primary inheritance (for languages whose values are close
enough to the current language to be acceptable for publishing); and (2)
secondary inheritance (for languages whose values are too different
from the current language to be acceptable for publishing, but are
perhaps useful during data entry and/or translation).
Thus, for the MDM Import Manager, there is the current language and
two levels of inheritance, color-coded as follows:
• Black. The value is from the current language.
• Green. The value is from a primary inherited language.
• Red. The value is from a secondary inherited language.
NOTE ►► The MDM Import Manager displays: (1) actual values from
the current language; (2) primary inherited values; and (3) secondary
inherited values.

394 MDM Import Manager Reference Guide

Consider a repository with three language layers: (1) English [US]; (2)
English [UK]; and (3) German [DE]. Both English values are typically the
same, so you can set the value for one version of English and allow the
other to inherit it. However, you don’t want the English languages to
inherit German or vice versa. In this case, inheritance for each language
would be as shown in Table 88.

Table 88. Language Inheritance Example for Three Languages

Language Primary Inheritance Secondary Inheritance
English [US] English [UK] German [DE]
English [UK] English [US] German [DE]
German [DE] <none> English [US]; English [UK]

Given the above inheritance, a record with the Size field set to “Small”
for English [US] and to NULL for both English [UK] and German [DE]
would display and color-code the value from each language-centric view
as shown in Table 89.

Table 89. Language-Centric Display Example for Three Languages

Language Actual Value Inherits From Display Value
English [US] Small <actual value> Small
English [UK] NULL English [US] Small
German [DE] NULL English [US] Small

The terminology and behavior around inheritance in the MDM Import

Manager is summarized in Table 90.

Table 90. Inheritance in the MDM Import Manager

Inheritance Item MDM Client
Language ordering Language-specific
Actual values Current language
Levels of inheritance Two (primary and secondary)
Type of inherited values Metadata and data
Color coding Black / Green / Red

NOTE ►► The MDM Import Manager displays the current language in

the title bar after the MDM repository name. If the language has not
been renamed, the language displays as “language [co]” (where
“language” is the language name and “co” is the country name).

MDM Import Manager Reference Guide 395

Multilingual Operations
The following sections describe the MDM Import Manager operations for
connecting to and importing data into a multilingual MDM repository.
There are no explicit commands for any of these operations. Rather, the
multilingual data is processed using various extensions to the existing
tabs, dialogs, and controls, as summarized in Table 91.

Table 91. Multilingual MDM Import Manager Operations

Operation Description
Language field / Connects to a multilingual MDM
Connect to MDM repository dialog repository.
Destination Fields grid / Maps language-specific source fields to
Map Fields/Values tab the layers of a multilingual field.
Language Field / Destination Fields Identifies the applicable language layer
grid / Map Fields/Values tab of expanded multilingual fields.
Expand Multilingual Fields option / Whether or not – and how – to expand
Configuration Options dialog multilingual fields into multiple fields.
Value Mapping Map button / Populates missing values in the current
Map Fields/Values tab language layer when mapping values.

CONNECTING TO THE MDM IMPORT MANAGER

When you first connect to a multilingual MDM repository using the MDM
Import Manager, you must select the current language for that MDM
Import Manager session.
You can use the Language field in the Connect to MDM Repository
dialog to select the current language from the drop-down list of
repository languages, as shown in Figure 224.

Figure 224. Language field in Connect to MDM Repository dialog

396 MDM Import Manager Reference Guide

FIELD MAPPING TO MULTILINGUAL DESTINATION FIELDS
Recall that you can import source data into an MDM repository simply
by mapping from a source field to the corresponding MDM destination
field in the Field Mapping grids of the Map Field/Values tab.
When the MDM destination field is a multilingual field within a
multilingual MDM repository, the source data may be structured in
several different ways, including: (1) a single source field with data for
the current language layer; (2) multiple source fields, one for each
language; or (3) a pair of source fields, one containing the data for all
languages and the second identifying the corresponding language for
each data value.
MDM supports import into multilingual destination fields based on how
the source data is structured, as described in the following sections.

The Expand Multilingual Fields Option

You can use the Expand Multilingual Text Fields configuration option to
specify whether or not – and how – to expand multilingual text fields into
multiple fields in the Destination Fields grid of the Map Fields/Values
tab, as summarized in Table 92.

Table 92. Expand Multilingual Text Fields Configuration Option

Setting Description
Do not expand multilingual text fields; only display
No
a single field for the current language layer.
Expand multilingual text fields into multiple fields,
As Multiple Fields
one field per language layer.
Expand multilingual text fields into multiple fields,
As Field and Language
as the field itself and a language field.

NOTE ►► If you are importing data only for the current language
layer and do not wish to expand multilingual text fields, you can turn off
the expansion of multilingual text fields in the Destination Fields grid by
setting the Expand Multilingual Text Fields configuration option to No.

As Multiple Fields Setting

When you import source data into a multilingual field, there may be
multiple source fields corresponding to each multilingual MDM field, one
for each language.
For multilingual Text and Text Large fields, the MDM Import Manager
allows you to import these multiple source fields into the multiple
language layers of the MDM field at the same time using the As Multiple
Fields setting of the Expand Multilingual Text Fields configuration option

MDM Import Manager Reference Guide 397

It does this by expanding each multilingual Text and Text Large field
into an instance of the field in the Destination Fields grid for each
repository language, as shown for a Description field in Figure 225. You
can then field map from each language-specific source field to the
applicable language layer instance of the multilingual destination field.

Figure 225. Multilingual Text field expanded into multiple items

NOTE ►► The field name for each instance of the multilingual field is
language-specific (e.g. Descrizione for Italian).

NOTE ►► The Language column in the Destination Fields grid: (1)

displays the name of the applicable language layer for expanded
instances of multilingual fields; and (2) is empty for non-lingual fields.

TIP ►► You can import multilingual lookup data into multiple

language layers at the same time by making the lookup table itself the
destination table.

As Field and Language Setting

When you import source data into a multilingual field, there may be a
pair of source fields corresponding to each multilingual MDM field, one
containing the data for all languages and the second identifying the
corresponding language for each data value.
For multilingual Text and Text Large fields, the MDM Import Manager
allows you to import this pair of source fields into the multiple language
layers of the MDM field at the same time using the As Field and
Language setting of the Expand Multilingual Text Fields configuration
option
It does this by expanding each multilingual Text and Text Large field
into a pair of fields: (1) the field itself; and (2) a Language field, as
shown for a Description field in Figure 226. You can then field map from
the source field containing data values to the data field and the source
field containing language values to the Language field.

398 MDM Import Manager Reference Guide

 Figure 226. Multilingual Text field expanded into a pair of items

NOTE ►► The Language field includes a value for each repository

language in the Destination Values grid of the Value Mapping pane.

VALUE MAPPING TO MULTILINGUAL DESTINATION VALUES

Recall that the Value Mapping grids of the Map Fields/Values tab allow
you as part of the import process to map from source data values to
destination data values, to convert the source values before mapping,
and to explicitly add new values to the list of destination data values.
In a multilingual MDM repository, the Destination Values grid still
displays the list of destination values. However, the list now displays not
only actual values from the current language layer, but also for values
that are missing in the current language, inherited values from primary
language layers in green and from secondary language layers in red.
When you map from a source value to a destination value that is
inheriting from another language layer, the MDM Import Manager allows
you to populate the missing data value in the current language layer
with the converted source value.
Specifically, when the destination value is missing in the current
language layer, the Map button (between the values grids) pops up a
context menu with two command choices, as shown in Figure 227.

Figure 227. Pop-up Set menu

MDM Import Manager Reference Guide 399

You can set the missing destination value in the current language layer
as summarized in Table 93 and described in this section.

Table 93. Map to Missing Value Map Operations

Operation Description
Sets the missing destination value in the current language
Set Missing Value
layer.
Does not set the missing value in the current language
Do Not Set Missing Value
layer.

 To map to a destination value that is missing in the current language

layer:
1. In the Source Values grid, select the source value you want to map.
2. Click on the Map button (between the values grids) to map the selected
source value and the selected destination value and choose from the
cascading menu:
 Set Missing Value
 Do Not Set Missing Value
3. MDM maps the source value to the destination value. If you choose Set
Missing Value, sets the destination value in the current language layer
with the converted source value and highlights it in blue.

IMPORTING MULTILINGUAL DATA INTO LOOKUP TABLES

Importing multilingual data into a lookup table depends upon a number
of variables, including whether: (1) the table is a flat or hierarchy table;
(2) new records are being added or language layer(s) of existing
records are being updated; and (3) just a single language layer or
multiple language layers are being imported at the same time.
In particular, import into the hierarchy table itself is now roughly
equivalent and as flexible as importing into flat tables thanks to several
new features: (1) value mapping is now optional rather than required to
the primary display field of the hierarchy table; (2) both internal and leaf
nodes can be mapped; and (3) the [Remote Key] field is now available
not only for field mapping but also for record matching.
NOTE ►► Previously, adding new multilingual records and updating
the language layer of existing records was not possible with the Import
Manager.

Importing multilingual data into flat and hierarchy lookup tables is

summarized in Table 94 and described in detail in the following
subsections.

400 MDM Import Manager Reference Guide

Table 94. Importing Multilingual Data into Lookup Tables
Flat Hierarchy Notes
To add new multilingual records
n/a If necessary, create the Hierarchy: Only one language
hierarchy from the current layer Name field contains the
language layer Name field hierarchy information.
that contains the hierarchy
information (e.g. using the
Split Hierarchy command).
Map the current language Map the current language Hierarchy: Now supports
layer Name field. layer Name field. mapping of internal nodes.
No value mapping necessary. Map/Add the entire set of
hierarchy values.
Map the other language layer Map the other language layer Hierarchy: Map Value button
Name field(s). Name field(s). previously could not be
No value mapping necessary. No value mapping necessary. unchecked for hierarchy field
and so Name field had to be
value mapped.
Map the [Remote Key] field. Map the [Remote Key] field.
Record match using: (a) any Record match using: (a) the [Remote Key] previously not
Name field; (b) the [Remote Name field; (b) the [Remote available as a record
Key] field; or (c) the Key] field (if the Name field matching field, and so Name
combination of both. was not value mapped); or (c) field had to be used for record
the combination of both. matching.
To update one or more language layers of existing records
Map the other language layer Map the other language layer Hierarchy: Map Value button
Name field(s). Name field(s). previously could not be
No value mapping necessary. No value mapping necessary. unchecked for hierarchy field
and so Name field had to be
value mapped
Map the [Remote Key] field. Map the [Remote Key] field. [Remote Key] previously did
not exist, and so current
language layer Name field
had to be mapped.
Record match using the Record match using the [Remote Key] previously not
[Remote Key] field and set [Remote Key] field and set available as a record
the import action to Update the import action to Update matching field, and so Name
(All Mapped Fields). (All Mapped Fields). field had to be used for record
matching.
Hierarchy: Can use [Remote
Key] field for record matching
because Map Value button
can now be unchecked for
hierarchy field and so Name
field no longer needs to be
value mapped.

NOTE ►► With value mapping optional for the primary display field of
a hierarchy table, MDM automatically figures out the other language
layers. When value mapping was required, the primary display field for
each language layer had to be individually value mapped.

MDM Import Manager Reference Guide 401

Import into Flat Tables
Multilingual import into a flat table is described in this section.

 To add new multilingual records:

1. Map the current language layer Name field. No value mapping is
necessary.
2. Map the other language layer Name field(s). No value mapping is
necessary.
3. Map the [Remote Key] field.
4. Record match using: (a) any Name field; (b) the [Remote Key] field; or
(c) the combination of both.

 To update one or more language layers of existing records:

1. Map the other language layer Name field(s). No value mapping is
necessary.
2. Map the [Remote Key] field.
3. Record match using the [Remote Key] field and set the import action to
Update (All Mapped Fields).

Import into Hierarchy Tables

Multilingual import into a hierarchy table is described in this section.

 To add new multilingual records:

1. If necessary, create the hierarchy from the current language layer
Name field that contains the hierarchy information (e.g. using the Split
Hierarchy command).
2. Map the current language layer Name field and then Map/Add the
entire set of hierarchy values.
3. Map the other language layer Name field(s). No value mapping is
necessary.
4. Map the [Remote Key] field.
5. Record match using: (a) the Name field; (b) the [Remote Key] field (if
the Name field was not value mapped); or (c) the combination of both.
NOTE ►► The Replace import action is disabled for import into the
hierarchy table itself; the Import Manager and MDIS will substitute
Update (All Mapped Fields) in old maps that have the Replace import
action.

402 MDM Import Manager Reference Guide

 To update the language layer of existing records:
1. Map the other language layer Name field(s). No value mapping is
necessary.
2. Map the [Remote Key] field.
3. Record match using the [Remote Key] field and set the import action to
Update (All Mapped Fields).

RECORD MATCHING AND MULTILINGUAL DESTINATION FIELDS

In a multilingual repository, all of the language layer instances of a
multilingual Text field appear in the Mapped Destinations Fields list and
are available as Matching Fields (where each field appears as
“fieldname <language layer>”).
This is true even if you have expanded the multilingual Text field into an
instance of the field for each repository language in the Destination
Fields grid of the Map Fields/Values tab.

SOURCE RECORD MERGING AND MULTILINGUAL DATA

Consider three source records containing different layers of multilingual
data for the same record in MDM, as shown in Figure 228

SKU Color (Eng) Color (Fre) Color (Ger)

#1 113 Red
#2 113 Rouge
#3 113 Rot

Figure 228. Three records containing different layers of multilingual data

Prior to import, the source record merging logic of MDM creates the
single merged record shown in Figure 229.

SKU Color (Eng) Color (Fre) Color (Ger)

113 Red Rouge Rot

Figure 229. Single merged record of multilingual data

NOTE ►► MDM was fixed to behave as described above in SP2.

MDM Import Manager Reference Guide 403

PART 15: REMOTE SYSTEMS AND MDM

This part of the reference guide contains a general overview of remote

systems and MDM, and a specific description of the related features
within the MDM Import Manager, including remote systems, key
mapping, and ports, which allow MDM to synchronize data between
itself and other systems.

405
Introduction
MDM has special features that enable it to synchronize data between
itself and other systems, as described in the following sections.

WHAT IS A REMOTE SYSTEM?

Any logical system that can supply data to or receive data from MDM is
known as a remote system.
MDM can import data from remote systems and create/update master
data objects using that data. Master data objects include main table
records, subtable records / lookup values, and text attribute text values.
When data is changed in a remote system, the changes can be
imported into MDM. Using previously created structural and key
mappings, in conjunction with dynamically reconfigurable
transformations and mappings, the data is applied to the master data
objects. All changes to master data objects are tracked.
At any time, master data objects can then be distributed to all known
remote systems through a process known as syndication. This involves
determining which master data objects need to be distributed and
converting them into a form that the remote system can understand.
NOTE ►► The remote system concept supports a number of features
related to distribution by bundling all sorts of useful information related
to a particular outside system, including key mapping, import maps,
syndication maps, and various timestamps.

KEY MAPPING
A remote system’s objects are mapped to master data objects within
MDM using key mapping. A key mapping maintains the relationship
between the remote system’s identifier (or key) for an object and the
corresponding master data object in MDM.
A key in MDM is a remote system-specific and object-type-specific
unique identifier. Different remote systems can have their own separate
collection of keys. Within a remote system, each type or collection of
objects can have its own separate collection of keys as well. Key
mappings are subject to the requirement that two different objects of the
same type from the same remote system cannot have the same key.
Remote system objects of a particular type can map only to MDM
objects of a particular type. A key can map to only one MDM object.
However, an MDM object may map to multiple keys from the same
remote system. When an MDM object maps to multiple keys, one of the
keys is marked as the default key. The default key is the one that is
used when syndicating a reference to the mapped MDM object.

MDM Import Manager Reference Guide 407

For example, for a particular remote system, the two color objects Light
Red and Dark Red both map to the MDM object Red. An MDM product
object Shirt has a Color attribute set to Red. When this Shirt object is
syndicated back to the remote system, the default key is used to choose
the value to syndicate from the two objects Light Red and Dark Red.
MDM objects that can be mapped to remote system keys include user-
defined table records and text attribute text values. Key mapping must
be enabled on each collection of objects for MDM to maintain the keys.

[REMOTE SYSTEM] AND [REMOTE KEY] FIELDS

MDM uses the remote systems defined in the Remote Systems table
within the MDM Console to store and maintain key mapping information
for each record or text attribute. It does this using a virtual “key
mapping” field that you never see in the MDM Import Manager.
This virtual key mapping field is very much like a qualified lookup field
into a virtual key mapping qualified lookup table. Each record of the
virtual lookup table consists of just two fields:
• [Remote System]. A single-valued Text field that contains the
name of the remote system. This is a normal field.
• [Remote Key]. A single-valued Text field that contains a key value
for the corresponding remote system. This is a qualifier field.

In effect, each Remote Systems table record becomes a record of the

key mapping qualified lookup table, and each actual key mapping
becomes a link of the key mapping qualified lookup field, one per
[Remote System] / [Remote Key] value pair, as illustrated in Figure 230.

SKU Name Lookup [Key Mapping]

213 Widget MDM; 112
CRM; 103
CRM; 105
R/3; 55-77

Figure 230. Key Mapping information stored in virtual lookup field

The [Remote System] and [Remote Key] fields are normally not visible;
however they do appear in several places in the MDM Import Manager.
Specifically, you choose a [Remote System] when you enter the MDM
Import Manager (or you choose one indirectly by selecting a map), and
[Remote Key] appears in the Destination Hierarchy tree, the list of
Destination Fields in the Map Fields/Values tab, and the Value Matching
lists in the Match Records tab.

408 MDM Import Manager Reference Guide

KEY MAPPING BEHAVIOR
The key mapping behavior of the MDM Import Manager is summarized
in Table 95.

Table 95. Key Mapping Behavior of the MDM Import Manager

Level Description
Each MDM Import Manager session and each import map is associated
Session with a single remote system, which is treated as the issuing remote
system for all remote keys loaded during the session.
If the destination table has Key Mapping enabled in the MDM Console,
then the [Remote Key] virtual field added to every field list for the table.
Table
 Appears only for remote system Type=Import/Export
 Appears for table Type=Main, Flat, Hierarchy, Taxonomy, or Qualified
[Remote Key] appears in: (1) the Destination Hierarchy tree; (2) the list
of Destination Fields in the Map Fields/Values tab; and (3) the Value
Matching lists in the Match Records tab.
Field
 Field Type=Key Mapping for [Remote Key] in Destination Fields grid
 Available for field mapping like any other field
 When selected, no values are displayed
Each key in the source data is represented by a value in the list of
source values. When a Source Field is mapped to a Destination Field
that has key mapping enabled, the MDM Import Manager loads value
mapping from the table of remote keys rather than from the map,
automatically mapping each Source Value to a Destination Value if the
Value source/destination value pair exists in the key mapping (similar to the
way value mappings are restored when a map file is loaded).

 Destination fields that have key mapping include: (1) a text attribute
that has key mapping enabled; (2) a lookup field for a table that has
key mapping enabled; and (3) the display field of a lookup table that
has key mapping enabled (for imports to a subtable)
[Remote Key] can be used as a record matching field. Each source
record will match one destination record if a key mapping exists for the
source remote key, and each remote key is associated with a single
destination record during the matching step. If no matching record is
found and the Create option is selected, the remote key will be
associated with the newly created record.

 During source record aggregation, [Remote Key] is treated like a

Record multi-valued field (i.e. all values from a group of source records
combined into a single multi-value). This may result in multiple
remote keys for a single source record. In this case, all keys will be
mapped to the same single destination record.
 When [Remote Key] is used for record matching and a source record
matches multiple destination records, the two Update import action
options are disabled; the only available import action is to Replace all
destination records with the source record.

MDM Import Manager Reference Guide 409

 Level Description
Key mappings are saved only for: (1) tables that have key mapping
enabled in the MDM Console; and (2) text attributes that have key
mapping enabled in Taxonomy mode of the MDM Client.
The key in the source data is designated – and key mapping value pairs
are established – as follows: (1) for destination table records, through
Record Matching in the Match Records tab, by mapping the key field in
the source to the [Remote Key] field in the Destination Fields list; and
(2) for subtables of the destination table and for text attributes, through
Value Mapping, by mapping source values to destination values.
The key mapping value pairs are stored in the repository as follows: (1)
for destination table records (but not for source records that are
skipped), subtables of the destination table, and text attributes, when
the import is executed; and (2) for mappings to existing records of
Key Mapping subtables of the destination table and to text values of text attributes,
when the map is saved
Source keys are stored literally as they appear in: (1) the source field
for each record (for the destination table); and (2) the source value list
(for subtables of the destination table and text attributes).

 Key mappings are saved only for remote system Type=Inbound or

Inbound/Outbound
 When multiple source records map to the same destination record,
one key is designated as the default. Upon subsequent export of the
destination record, the single default key will be used.
 When an inbound key mapping is already assigned to an existing
destination record, whether or not it is reassigned depends upon the
setting of the Reassign Existing Key Mappings configuration option.

KEY MAPPING REASSIGNMENT

Recall that each key mapping value pair consists of a remote system
and a remote key value, that each key mapping value pair must be
uniquely associated with a single record in an MDM repository, and
therefore that each remote key value must be unique for a particular
remote system.
A challenge arises when a source record contains a key mapping value
pair that already exists in the MDM repository but is associated with a
destination record other than the one that matches the source record.
Because of the uniqueness of key mapping value pairs, the key
mapping could not be added to the matching destination record without
first removing it from the existing destination record.
DATA INTEGRITY ►► The “safest” behavior would be to reject
source records that attempt to reassign a key mapping to another
record. However, the MDM Import Manager needs to be able to
support automation and high-volume imports without triggering multiple
exceptions, especially when the reassignment of key mappings is not
an issue.

MDM can automatically reassign key mapping value pairs within the
MDM Import Manager based on the setting of a configuration option.

410 MDM Import Manager Reference Guide

Specifically, you can use the Reassign Existing Key Mappings
configuration option to specify whether or not – and how – to reassign
existing key mappings during import, as summarized in Table 96.
Table 96. Reassign Existing Key Mappings Configuration Option
Setting Description
Do not reassign existing key mappings; fail to
No
import source records that contain them.
Into Original Reassign existing key mappings, but only into
Records Only original records.
Into Original and Reassign existing key mappings into both original
Checked Out Records and checked out records.

DATA INTEGRITY ►► When a key mapping is reassigned to a new

record: (1) it is immediately removed from the old record to which it
was previously assigned; (2) if the old record is a checked out record,
the key mapping is removed from both the original and the checked out
versions of the record; and (3) if the new record is a checked out
record, you will lose the key mapping if the new record is subsequently
rolled back rather than checked in, resulting in data loss.

WHAT IS A PORT?
With remote systems as a foundation, an MDM port encapsulates all of
the configuration and logistical information associated with inbound and
outbound processing of data from these remote systems.
In so doing, it simplifies the process of: (1) delivery and consolidation of
raw data from remote systems into MDM using the MDM Import
Manager; and (2) extraction and distribution of data from MDM to
remote systems using the MDM Syndicator.
In each case, the port represents the logical point of contact between
MDM and the outside world (e.g. XI or a user of the MDM Import
Manager). Within the MDM Import Manager and MDM Syndicator, it
represents the physical staging location of data and a logical handle by
which to identify all of the encapsulated information.
NOTE ►► Along with simplifying the user interaction, ports also lay
the foundation for future automation of the consolidation and
distribution processes within MDM.

MDM Import Manager Reference Guide 411

Remote System Operations
The following sections describe the MDM Import Manager operations for
managing the remote systems, key mapping, and ports of an MDM
repository, as summarized in Table 97.

Table 97. Remote System MDM Import Manager Operations

Operation Description
Remote System field / Enters the MDM Import Manager on
Connect to Source dialog behalf of the selected remote system.
Port field / Enters the MDM Import Manager on
Connect to Source dialog behalf of the selected port.
Get Next Source File command / Selects the next source data file waiting
main menu on the current port.
Field Mapping grids / Maps to MDM records of the current
Map Fields/Values tab destination table.
Value Mapping grids / Maps to MDM lookup table records and
Map Fields/Values tab text attribute text values.
Value Matching lists / Matches source records to MDM
Match Records tab records using the [Remote Key] field.

CONNECTING TO A REMOTE SYSTEM

Whenever you connect Import Manager to a data source, you must
select a remote system as the source of imported data.
You can use the Remote System field in the Connect to Source dialog
to select a remote system from the drop-down list of remote systems, as
shown in Figure 231.

Figure 231. Connect to Source dialog

NOTE ►► The drop-down list includes: (1) the built-in remote system
named MDM; and (2) user-defined remote systems of type Inbound.

412 MDM Import Manager Reference Guide

CONNECTING TO A PORT
An inbound port identifies a file system location that contains source
data files waiting for import into MDM. All information required by MDM
or an MDM Import Manager user to identify and process the inbound
data – including not only the location, but also the remote system, the
object type, and the MDM Import Manager map to automatically load –
is contained in the port’s definition.
When the source Type=Port, you can use the Port field in the Connect
to Source dialog to select a port from the drop-down list of ports, as
shown in Figure 231.

Figure 232. Connecting to a port using the Connect to Source dialog

NOTE ►► The Port drop-down list includes an entry for each inbound
port you define in the MDM Console for the selected Remote System;
those that do not have files waiting or are locked by another user are
disabled in the drop-down list.

NOTE ►► When you select a port, the MDM Import Manager

automatically selects the first source data file waiting to be processed;
you can then use the Get Next Source File command to select each
succeeding file.

FIELD MAPPING
To map between source records and MDM records of the current
destination table, use the Field Mapping grids in the Map Fields/Values
tab to create a field mapping between the remote system key field in the
Source Fields list and the [Remote Key] field in the Destination Fields
list.
NOTE ►► The [Remote Key] virtual field appears only when Key
Mapping is enabled for the destination table and the remote system
Type is Inbound or Inbound/Outbound.

MDM Import Manager Reference Guide 413

During the import process, source records will be matched against MDM
records using the matching rules you define. The result of this matching
will be used to create key mappings between the remote system key
values and the MDM objects.

VALUE MAPPING
To map between source records and MDM lookup table records or text
attribute text values, use the Value Mapping grids in the Map
Fields/Values tab to create value mappings.
NOTE ►► The source key used for existing and new MDM
destination values is the Original Value in the Source Values list.

RECORD MATCHING
When importing to a table that has Key Mapping enabled, the [Remote
Key] field is available for record matching purposes when it has been
mapped. Selecting this field will use the existing key mappings to match
the source records.
When mapping to MDM subtable records, existing key mappings are
also used to automatically map values when Key Mapping is enabled for
the underlying destination records.
NOTE ►► When the [Remote Key] field is used for record matching,
the two Update import action options are disabled for multiple matches.

NOTE ►► If Key Mapping is not enabled, the standard value mapping

functionality will try to remember the mapped values. However the
import map only remembers the values actually mapped from the last
import and discards unused value mappings. The Key Mapping retains
mappings whether they are used or not.

Lookup Table Imports Using the [Remote Key] Field

When the current table is a lookup table, the [Remote Key] field can
also be used as a record matching field.
However, to enforce the key mapping semantic that a key value can be
associated with at most a single record in the MDM repository, MDM
imposes a restriction on the use of the [Remote Key] field as a matching
field.
Specifically, if you map values for the display field(s) of a flat table or the
primary display field of a hierarchy table, the [Remote Key] field can be
used as a matching field only if it is combined with the display field(s) or
the primary display field.

414 MDM Import Manager Reference Guide

This constraint exists because value mapping allows you to add new
values, which correspond to records that do not yet exist in the table but
will be created as part of the import. If the [Remote Key] field could be
used as the matching field by itself, MDM might find an existing record
based on the key value even though a new value was just created
based on the value mapping.
Requiring that the [Remote Key] field be used in conjunction with the
value mapped display field(s) guarantees that the key field combination
will not match an existing or newly created record, allowing MDM to
create a new record as part of the record matching process.
NOTE ►► In addition, to maintain the constraint that a key value can
be associated with at most a single record, if the [Remote Key] value is
already associated with an existing record, MDM automatically
removes it from that record and associates it with the new record.

NOTE ►► Prior to key mapping support within MDM, when the

current table was a lookup table, only the display field(s) of a flat
lookup table or the primary display field of a hierarchy table could be
used as the matching field(s).

Conversely, if you do not map values for the display field(s), there is no
restriction on the use of the [Remote Key] field as a matching field. In
this case, the [Remote Key] field can be used as the matching field by
itself because without value mapping, there is no opportunity to create
new values, and therefore no possibility that the [Remote Key] value
already exists for a node whose corresponding record was just created.

[REMOTE KEY] VALUES AND LOOKUP FIELD AUTOMAPPING

When you map a source field containing remote key values to a
destination lookup field for a table that has key mapping enabled, the
MDM Import Manager automatically maps the source values to
destination values based on the remote keys of the lookup table for that
remote system.
NOTE ►► Unmapping a source value does not automatically remove
the corresponding key mapping from the lookup table (you must
instead use the MDM Data Manager’s Edit Key Mapping dialog). The
only time MDM does automatically delete a key mapping is if mapping
a source value causes the same key for a remote system to be
assigned to two destination values. If this happens, MDM resolves the
conflict by automatically deleting the key for the “old” destination value.

NOTE ►► Key mapping information for lookup field values is saved

whenever you save a map, unlike key mapping information for record
matching, which is only saved during import.

MDM Import Manager Reference Guide 415

 NOTE ►► For destination hierarchy lookup fields that have key
mapping enabled, you can disable the automatic value mapping
feature by setting the Disable automatic hierarchy value mapping
according to remote key configuration option to No (see "Configuration
Options" on page 62 for more information).

[REMOTE KEY] VALUES AND RECORD REPLACEMENT

Existing [Remote Key] values are always preserved, even for
destination records set to be replaced with the Replace action, as
follows: (1) if a source field is mapped to the [Remote Key] field, MDM
will reject the source record rather than perform the replace; and (2) if a
source field is not mapped to the [Remote Key] field, MDM will perform
the replace after first transferring the [Remote Key] values of the target
replaced records to the new record. In each case, the Replace action will
not result in the loss of [Remote Key] values.

416 MDM Import Manager Reference Guide

PART 16: REFERENCE

This part provides reference information in one convenient location,

including a menu command reference, glossary, and index.

417
Command Reference
FILE MENU
Main Menu Context Menu Description

Open… n/a Opens the selected map

Close n/a Closes the current map

Save n/a Saves the current map

Save As… n/a Saves the current map to a new name

Save Update n/a Saves an update of the current map

Selects the next source data file waiting on the

Get Next Source File n/a
current port

Connect to Source… n/a Connects to a new data source

Import… n/a Imports a map file into the repository

Export… n/a Exports a map file from the repository

Exit n/a Quits the application

VIEW MENU
Main Menu Context Menu Description

Select Source
[Drop-Down List] Selects the current source table
Table > table

Select Destination
[Drop-Down List] Selects the current destination table
Table > table

Switch Layout n/a Switches the main window layout

MDM Import Manager Reference Guide 419

SOURCE MENU
Main Menu Context Menu Description

Source
Add Field Adds an empty field to the hierarchy
Hierarchy Tree

Source
Clone Field Clones the selected field and its values
Hierarchy Tree

Source
Rename Field Renames the selected field
Hierarchy Tree

Source
Delete Field Removes the selected field from the hierarchy
Hierarchy Tree

Source
Join > table > field Joins the selected field to a field in another table
Hierarchy Tree

Source
Unjoin Unjoins the selected field
Hierarchy Tree

Source
Lookup > field Looks up a field from another joined table
Hierarchy Tree

Source Removes the selected lookup field from the

Unlookup
Hierarchy Tree hierarchy

Source Splits the data values within the selected field into
Split into Hierarchy
Hierarchy Tree a hierarchy value

Source Splits the data values within the selected field into
Split into Multiple Fields
Hierarchy Tree multiple fields

Source Creates a new field based on the selected fields

Create Hierarchy Field
Hierarchy Tree and then splits it into a hierarchy

Combines fields that have been mapped to the

Create Compound Source
multiple display fields of the selected destination
Field > compound field Hierarchy Tree
lookup field

Create Source Creates a new table by pivoting on the selected

Pivot Table… Hierarchy Tree fields

Create Source Creates a new table by reverse pivoting on the

Reverse Pivot Table… Hierarchy Tree selected fields

FIELDS MENU
Main Menu Context Menu Description

Find Next Matching Source Finds the first or next similarly destination field
Destination Field Fields Grid with a name similar to the selected source field

Set Flat Source Sets the Flat NULL Handling option for the
NULL Handling > setting Fields Grid selected source fields

420 MDM Import Manager Reference Guide

 Main Menu Context Menu Description

Set Hierarchy Source Sets the Hierarchy NULL Handling option for the
NULL Handling > setting Fields Grid selected source fields

Source
Set Case Sensitivity > Sets the Case Sensitivity option for the selected
Hierarchy Tree &
setting Source Fields Grid
source field

Set Value Delimiter > Source Sets the Split Hierarchy delimiter for the selected
Split Hierarchy… Hierarchy Tree source field

Set Value Delimiter > Source Sets the Split Multiple Fields delimiter for the
Split Multiple Fields… Hierarchy Tree selected source field

Set Value Delimiter > Sets the Partition Combine delimiter for the
Partition By List
Partition Combine… selected source field

Set Value Delimiter > Source Sets the Multi-Valued delimiter for the selected
Multi-Valued… Fields Grid source fields

Set Value Delimiter > Source Sets the Coupled Numeric delimiter for the
Coupled Numeric… Fields Grid selected source fields

Set Value Delimiter > Restores inheritance for all the value delimiters
n/a
Inherit Defaults for the selected source fields

Set Value Conversion Source Sets the value conversion filter for the selected
Filter > filter Fields Grid source field

Set Value Conversion Source Sets multiple value conversion filters for the
Filter > Multiple… Fields Grid selected source field

Destination Filters the attributes that are not linked to the

Filter Unlinked Attributes
Fields Grid selected categories (single taxonomy field)

Filter Unlinked Attributes Destination Filters the attributes that are not linked to the
> taxonomy field Fields Grid selected categories (multiple taxonomy fields)

Set NULL Destination Sets the NULL Interpretation option for the
Interpretation > setting Fields Grid selected destination fields

Set Multi-Valued Destination Sets the Multi-Valued Update option for the
Update > setting Fields Grid selected multi-valued destination fields

Set Qualified Destination Sets the Qualified Update option for the selected
Update > setting Fields Grid qualified lookup destination fields

Set Data Group Display Destination Sets the Data Group Display option for the
> setting Fields Grid selected object lookup destination fields

Set MDIS Unmapped Destination Sets the MDIS Unmapped Value Handling option
Value Handling > setting Fields Grid for the selected destination field

MDM Import Manager Reference Guide 421

VALUES MENU
Main Menu Context Menu Description

Source
Display Distinct Values Displays the distinct values for the selected field
Hierarchy Tree

Source
Remove Distinct Values Removes the distinct values for the selected field
Hierarchy Tree

Apply Value Conversion Source Converts the selected source values using the
Filter > filter Values Grid specified value conversion filter

Find Next Unmapped Source Finds the first or next unconverted or unmapped
Source Value Values Grid source value

Reorder Mapped Source Reorders the mapped destination values for the
Destination Values Values Grid selected source value

Find Next Mapped Source Finds the first or next destination value to which
Destination Value Values Grid the selected source value is mapped

Select Entire Branch > Source Selects the selected internal node and all of its
All Nodes Values Grid internal and leaf node children

Select Entire Branch > Source Selects all of the leaf node children of the
All Leaf Nodes Values Grid selected internal node

Select Entire Branch > Source Selects all of the unmapped leaf node children the
Unmapped Leaf Nodes Values Grid selected internal node

Add Destination Destination

Adds a new destination value
Value Values Grid

Add Destination
Destination Adds a new destination hierarchy value as a
Hierarchy Value >
Values Grid sibling
As Sibling
Add Destination
Destination
Hierarchy Value > Adds a new destination hierarchy value as a child
Values Grid
As Child

Remove Destination Destination

Removes the selected new destination values
Value Values Grid

RECORDS MENU
Main Menu Context Menu Description

Matching Sets the selected matching field to be the key

Set Key Field
Fields List field used to identify duplicate source records

Set Active Matching

Sets the value matches that should be active
Records > value match Fields List

Matching Sets the record filtering criteria for the selected

Set Filter
Fields List matching field

422 MDM Import Manager Reference Guide

 Main Menu Context Menu Description

Set Default Default Import Sets the default import action for a Match Level /
Import Action > action Actions Grid Match Type group of source records

Set Source Source Sets the import action for the selected source
Import Action > action Records Grid records

Set Destination Destination Sets the import action for the selected destination
Import Action > action Records Grid records

Refresh
n/a Refreshes the record matching results
Record Matching

Execute Import n/a Executes the import

CONFIGURATION MENU
Main Menu Context Menu Description

Options n/a Changes program options

Changes the connected MDM user’s repository

Change Password n/a
password

HELP MENU
Main Menu Context Menu Description

About SAP Displays program information, version, and

n/a
MDM Import Manager… copyright

PARTITION FIELD/VALUE TAB BUTTONS

Button Context Menu Description

Adds the selected item to the list of partitions for

Add n/a
the selected tree node (single item selected)

Add > Adds the selected items as individual partitions

n/a
Add as Individual (multiple items selected)

Add > Adds the selected items as a combined partition

n/a
Add as Combination (multiple items selected)

Removes the selected partition items from the list

Remove n/a
of partitions for the selected tree node

Combines the selected partition items for the

Combine Partition By List
selected tree node

Splits the previously combined partition items for

Split Partition By List
the selected tree node

MDM Import Manager Reference Guide 423

MAP FIELDS/VALUES TAB BUTTONS
Button Context Menu Description

Field Mapping Grids

Automatically maps each source field to a

Automap n/a
destination field with the same name

Maps the selected source field to the selected

Map n/a
destination field

Unmap n/a Unmaps the selected fields

Adds a new attribute to the destination taxonomy

Add n/a
(single taxonomy field)

Add > Adds a new attribute to the selected destination

n/a
taxonomy field taxonomy (multiple taxonomy fields)

Value Conversion and Mapping Grids

Automatically maps each source value to

Automap n/a
destination values that are equivalent

Maps the selected source values to the selected

Map n/a
destination values

Map > Sets the missing destination value in the current

n/a
Set Missing Value language layer (missing value)

Map >
Does not set the missing destination value in the
Do Not Set Missing n/a
current language layer (missing value)
Value

Unmap n/a Unmaps the selected values

Destination Adds the selected source values as new

Add
Values Grid destination values (flat destination)

Add > Adds the selected source values as siblings of the

n/a
Add as Sibling selected destination value (destination hierarchy)

Adds the selected source values as children of

Add >
n/a the selected destination value (destination
Add as Child hierarchy)
Adds the selected internal source values and all
Add >
n/a of their children as siblings of the selected
Add Branch as Sibling destination value (destination hierarchy)
Adds the selected internal source values and all
Add >
n/a of their children as children of the selected
Add Branch as Child destination value (destination hierarchy)

Destination
Remove Removes the selected new destination values
Values Grid

424 MDM Import Manager Reference Guide

MATCH RECORDS TAB BUTTONS
Button Context Menu Description

Adds the selected field to the list of matching

Add n/a
fields (single field selected)

Add > Adds the selected fields as individual matching

n/a
Add as Individual fields (multiple fields selected)

Add > Adds the selected fields as a matching field

n/a
Add as Combination combination (multiple fields selected)

Removes the selected fields from the list of

Remove n/a
matching fields

Matching Combines the selected matching fields into a

Combine
Fields List matching field combination

Matching Splits the selected matching field combination into

Split
Fields List individual matching fields

MISCELLANEOUS CONTEXT MENUS

Command Context Menu Description

Set Value
Modify… Modifies the specified value conversion
Conversions Dialog

MDM Import Manager Reference Guide 425

Display and Format Conventions
The MDM Import Manager is like the heads-up display of a cockpit. It
uses color and a variety of syntactical format conventions to efficiently
represent and convey a tremendous amount of information about the
source and destination data and the current state of the import, as listed
in Table 98.
Many of these are standard Windows conventions; others correspond to
the functional capabilities of the MDM system.

Table 98. Display and Format Conventions of the MDM Import Manager
Where What Description
Source Hierarchy Pane
field [Join] Field is joined
Source
Hierarchy [To: table.field] Field is joined to table.field.
Tree
[From: table.field] Field is joined from table.field
Destination Hierarchy Pane
Node is a text attribute
Node is a text attribute with no links
Destination Node is a numeric attribute
Hierarchy
Tree Node is a numeric attribute with no links
Node is a coupled attribute
Node is a coupled attribute with no links
Source Preview Tab
table [Lookup] Table has lookup fields
Tables List table <Pivot> Table is a pivot table
table <Reverse Pivot> Table is a reverse pivot table
Records Grid field <table> Field is a lookup from table
Partition Field/Value Tab
field Partition item is the current field
field [value] Partition item is the current field value
Partition field Partition item
Field/Value
Tab field1; field2 Combined partition of field1 by field2
 Inherit Partitions are inherited
 Inherit Partitions are not inherited
Map Fields/Values Tab
Fields Grids value Field is inheriting the global default value
Field is mapped to a destination field
Source Fields
field [Partition] Field has been partitioned
Grid
field <table> Field is a lookup from table

426 MDM Import Manager Reference Guide

 Where What Description
A source field is mapped to the destination
field, attribute, or qualifier
F Item is a field
D Item is a promoted lookup display field
Q Item is a qualifier
A Item is an attribute
Destination
Fields Grid field [DF] Display field of the current table
lookup field [field1; field2] Lookup consists of multiple display fields
field [lookup field] Promoted display field of lookup field
qualifier <qualified field> Qualifier is associated with qualified field
attribute <taxonomy field> Attribute is associated with taxonomy field
attribute Attribute not linked to selected categories
<NULL> Value is NULL
 Map Value mapping is required
Values Grids
 Map Value mapping not required nor permitted
 Map Value mapping is optional
Value does not require mapping and is
properly converted
Value is mapped to an existing destination
value
Value is mapped to multiple existing
destination values
Source value is mapped to a new
Source destination value
Values Grid Source value is mapped to multiple new
destination values
Source value is mapped to multiple
destination values, some of which are new
Converted value is inheriting value
converted value
conversion filters from the field level
Converted value has been overridden with
converted value
manual edits and conversions
A single source value is mapped to the
destination value
Multiple source values are mapped to the
destination value
A single source value is mapped to the
Destination new destination value
Values Grid
Multiple source values are mapped to the
new destination value
[n] Order of mapped destination value
value Existing destination value
value New destination value

MDM Import Manager Reference Guide 427

 Where What Description
 Inherit default Field is inheriting default
Delimiter  Inherit default Field is not inheriting default
Dialog
Multiple fields are selected, some
 Inherit default
inheriting the default and some not
Match Records Tab
field Matching field
field1; field2 Matching field combination
field [DF] Matching field is a display field

Matching field [Key] Matching field is the key field

Fields List field [All] All value matches active for matching field
field [Eql] Equal value matches active
field [Und] Undefined value matches active
field [Neq] Not Equal value matches active
None Source records with no matches
Match Level
Single Source records with single match
Column
Multiple Source records with multiple matches
None Source records with no matches

Match Type Exact Source records with Exact match type

Column Partial Source records with Partial match type
Conflict Source records with Conflict match type
Active
Record Exact match record
Matching
Grids Partial match record
Conflict match record

428 MDM Import Manager Reference Guide

Glossary

Field mapping A mapping at the field level between a source field and a
destination field (or attribute).
Join Combines information from two tables by performing a
lookup on every record of the primary table.
Lookup Uses one or more pairs of matching fields from two tables,
taking the value of the field for a single record in the first
(primary) table to “look up” additional information in a
single corresponding record in the second (lookup) table.
Mapping Establishes a one-to-one correspondence between a
single source item and a single destination item.
Partition A partition creates a set of value combinations based on
the value of one field and all of the corresponding values of
a second field (or attribute). A single value for a field can
be partitioned, resulting in a single set of value
combinations, or an entire field can be partitioned, resulting
in a set of value combinations for each value of the
partitioned field. Partitions can be inherited from parent
nodes in a hierarchy, and for a particular node in the
hierarchy, can be nested (which creates hierarchy among
value combinations) or combined (which merges value
combinations).
Table Rectangular row/column arrangement of related data
values. Each horizontal row in the table represents a single
record and consists of the same set of fields. Each vertical
column of the table represents one field that is stored for
each row in the table.
Unique field/ Key field (or compound key field combination) for a record
unique field that uniquely identifies the record. A table may have more
combination than one unique field (or field combination).
A field mapping to a unique field requires a value match to
detect existing records that match each source record.
Value domain The enumerated set of valid values for a destination text
field (or attribute). Value domains exist primarily for lookup
fields and text-valued attributes, but can also exist for a
non-lookup field in a lookup table or when a non-lookup
field has a well-defined set of valid values (such as the
two-letter abbreviations for state). A field mapping to a
destination with a value domain requires a value mapping.
Value mapping A mapping at the field value level between a particular
source value and a particular destination value.

MDM Import Manager Reference Guide 429

Index
C P
Configuration Options, 62 Partitions
Create Clone, 90 Multiple, 186
Nested and Combined, 187
F
Field properties, 161, 164, 165, S
166 Source Hierarchy operations,
85
I
Inheritance, 185 U
User interface, display
M conventions, 426
Main window, 49
MDM Client main window, 44

430 MDM Import Manager Reference Guide