/eˈliːs/ → pronounced like “eh-LEES”
Grammar Rules, Documentation, Specification, Todos
A schema-driven data transformation language that compiles type-optimized bytecode from pipeline expressions over structured data. Write once, run against any conforming dataset.
NOTE: Only
.csvdata files are supported for now.
| Extension | Purpose |
|---|---|
.eli |
Source code |
.elt |
Schema / type definitions for input data |
.csv |
Input data file |
.elb |
Generated file with compiled bytecode |
elise --mode=run --source-code=sample.eli --data=data.csv --data-schema=data.elt-
Compiles in-memory (no
.elboutput) -
Performs full runtime validation of input data against schema
-
Executes immediately
Step 1 — Build an executable
elise --mode=build --source-code=sample.eli --data-schema=data.elt --output=program.elbStep 2 — Execute
elise --mode=exec --executable=program.elb --data=data.csv-
Requires precompiled .elb
-
Data must comply with schema that was used for building executable
-
Skips runtime validation
-
Executes fastest possible path
Use case: trusted, prevalidated data
elise --mode=validate --data=data.csv --data-schema=data.elt-
Full scan of data to ensure strict schema compliance
-
Can be used before unsafe execution