Introduction

$ pipenv run ./mig.py recommend examples/recommend/ex001.toml 
Fields
    `aaa` SIZE(4)
    `bbb` SIZE(4)
    `ccc` SIZE(255)
    `xxx` SIZE(255)

Query #0
    SELECT * FROM `whatever`
        WHERE `aaa` = 123
Query #1
    SELECT * FROM `whatever`
        WHERE (`aaa` >= 22) AND (`aaa` <= 44)
Query #2
    SELECT * FROM `whatever`
        WHERE (`aaa` >= 22) AND (`bbb` = 1)
Query #3
    SELECT * FROM `whatever`
        WHERE (`aaa` >= 22) AND (`ccc` > "xyz")
Query #4
    SELECT * FROM `whatever`
        WHERE `aaa` >= 22
        ORDER BY `aaa` ASC
        LIMIT 1
Query #5
    SELECT * FROM `whatever`
        ORDER BY `aaa` DESC
Query #6
    SELECT * FROM `whatever`
        WHERE (`aaa` = 123) AND (`bbb` = 1)
        GROUP BY `ccc`
Query #7
    SELECT * FROM `whatever`
        WHERE `aaa` = 123
        GROUP BY `ccc`
        ORDER BY `ddd` ASC
Query #8
    SELECT * FROM `whatever`
        WHERE `aaa` = 123
        GROUP BY `ccc`
        ORDER BY `ccc` ASC
Query #9
    SELECT * FROM `whatever`
        WHERE `ccc` >= 432
        ORDER BY `ccc` ASC
        LIMIT 10
Query #10
    SELECT * FROM `whatever`
        WHERE `ccc` in (1,2,3)
        ORDER BY `aaa` ASC
        LIMIT 10
Query #11
    SELECT * FROM `whatever`
        WHERE ((`aaa` = 123) AND (`bbb` = 1)) OR ((`ccc` = 22) AND (`aaa` = 123))

PK recommendation #0 with columns `auto_id` and max size per row of 1311 bytes
--------------------------------------------------------------------------------
Index #0
    Recommendation: PRIMARY KEY(`auto_id`)
    To use by query:
    Max size with PK: 4 bytes
    Log messages:
      - RECOMMENDATION: The table don't have an autoincremental column, so a fake column with an autoincremental has been added as a recommendation.
    Pattern:
        Ordered subindex with the columns
            0. `auto_id`
Index #1
    Recommendation: KEY(`ccc`)
    To use by query: #9, #10
    Max size with PK: 259 bytes
    Log messages:
      - WARNING: Isn't possible optimize the `ORDER BY` in the query #10 because some of the `WHERE` filter a column with multiple values, probably with an `IN` or an `OR`. Remove the `ORDER BY` or execute multiple queries concatenating the output with an `UNION`
    Pattern:
        Ordered subindex with the columns
            0. `ccc`
            1. `auto_id`
Index #2
    Recommendation: KEY(`aaa`, `ccc`)
    To use by query: #0, #1, #3, #4, #5, #7, #8, #11
    Max size with PK: 263 bytes
    Log messages:
      - RECOMMENDATION: If you are going to get all the values in the query #5 it is probably better to remove the `ORDER BY` and sort the values in the application
      - WARNING: Isn't possible optimize both the `ORDER BY` and the `GROUP BY` in the query #7, so only the `GROUP BY` will be optimized
      - WARNING: Isn't possible optimize all the `OR` operator of the query #11, try to refactor it to execute multiple queries concatenating the output with an `UNION`
    Pattern:
        Ordered subindex with the columns
            0. `aaa`
            1. `ccc`
            2. `auto_id`
Index #3
    Recommendation: KEY(`bbb`, `aaa`, `ccc`)
    To use by query: #2, #6
    Max size with PK: 267 bytes
    Pattern:
        Ordered subindex with the columns
            0. `bbb`
            1. `aaa`
            2. `ccc`
            3. `auto_id`