API Reference

API Reference

parse_ddl

Parses SQL CREATE TABLE DDL into a canonical ContractSchema.

from great_generator import parse_ddl

contract = parse_ddl(ddl_text, dialect="databricks", strict=True)
print(contract.fingerprint())

Supported dialect values are "ansi", "spark", and "databricks" for the documented subset. Strict mode raises structured diagnostics for unsupported contract-affecting syntax.

generate_from_schema

generate_from_schema(
    schema,
    rows=100,
    seed=None,
    engine="auto",
    spark=None,
    table_name="sample",
    realism="realistic",
    domain=None,
    custom_rules=None,
    plan=None,
    realistic=None,
    validate=False,
    return_report=False,
)

Parameter	Meaning
schema	Supported schema input, including a parsed ContractSchema from parse_ddl(...)
rows	Row count, or table-count mapping for DomainSchema and multi-table ContractSchema
seed	Optional reproducibility seed
engine	auto, pandas, or spark
spark	Optional SparkSession
table_name	Logical single-table name
realism	realistic or placeholder
domain	Optional semantic preset
custom_rules	Per-column generation rules
plan	Optional GenerationPlan from infer_generation_plan(...)
realistic	Backward-compatible boolean mode override
validate	Run supported validation
return_report	Return (data, report)

Advisor APIs

Advisors are optional and run before generation. They produce JSON artifacts that can be inspected, edited, saved, and reused.

from great_generator import generate_from_schema, infer_generation_plan

schema = "customer_id int, customer_name string, email string"
plan = infer_generation_plan(schema, advisor="none")
df = generate_from_schema(schema, rows=1000, plan=plan)

See Advisors and Plans and Tags.

Related APIs

API	Purpose
explain_generation_plan	Show inferred semantic behavior
infer_generation_plan	Create an editable generation plan
tag_schema	Create editable column tags
review_realism	Review a generated sample against a plan
validate_generated_data	Validate schema-generated Pandas data
parse_ddl	Parse SQL CREATE TABLE DDL into a canonical contract
generate_relational	Generate custom related tables
generate_domain	Generate a ready-made domain pack
generate_cdc	Generate CDC records
generate_history	Generate SCD2 history
generate_dimensional_model	Generate facts and dimensions
generate_data_vault_model	Generate hubs, links, and satellites
generate_from_recipe	Run JSON, TOML, or simple YAML recipes
export_data	Convenience domain dataset export