DataMaker

The official Python library for the Automators DataMaker API.

Installation

You can install the package using pip:

pip install datamaker-py

Quick start

Basic example:

from datamaker import DataMaker, Template

# Create an instance of DataMaker
datamaker = DataMaker()


def main():
    # Define the template
    template = Template(
        name="basic template",
        quantity=2,
        fields=[
            {"name": "first_name", "type": "First Name"},
            {"name": "last_name", "type": "Last Name"},
            {
                "name": "email",
                "type": "Derived",
                "options": {"value": "{{first_name}}.{{last_name}}@automators.com"},
            },
        ],
    )

    # Generate data using the template
    result = datamaker.generate(template)
    print(result)


if __name__ == "__main__":
    main()

Push a set into Tosca Cloud

push_tosca_dataset creates or replaces a native Tosca Cloud data set from a saved set, an input set version, a dataset or inline rows. A Tosca test step reads it with {TDM[<data set>][<subset>][<column>]}.

from datamaker import DataMaker, ToscaDatasetSchemaChangedError

dm = DataMaker()

try:
    result = dm.push_tosca_dataset(
        integration_id,
        name="Orders",
        input_set_id="is_123",        # or set_id=..., dataset_id=..., rows=[...]
        on_existing="overwrite",      # "fail" (default) | "overwrite" | "new"
    )
    print(result.dataset_id, result.rows_count, result.outcome, result.example)
except ToscaDatasetSchemaChangedError as e:
    # The rows dropped a column the data set has. Pass on_schema_change="recreate"
    # to delete and recreate it instead.
    print("removed:", e.removed_columns, "added:", e.added_columns)

An input set gets one Tosca subset per test case by default; subsets_by picks another column or turns it off with False. get_tosca_datasets(integration_id) lists what a space already holds.

Autocomplete Types

This SDK automatically generates TypeScript autocomplete definitions for use in the datamaker application.

The artifact is generated automatically on every push to the main branch and can be fetched from:

https://raw.githubusercontent.com/automators/datamaker-py/main/artifacts/autocomplete-types.ts

See artifacts/README.md for more details.

Development & Contibutions

See the contributing.md guide for details on how to contribute to this project.

Types

Response types are generated from the API's OpenAPI document, not hand-written:

spec/openapi.json  ->  src/datamaker/generated/schema.py  ->  src/datamaker/types.py

Import from datamaker.types, which resolves each spec schema name to the class that really matches it:

from datamaker.types import Project, Plan, Template

projects: list[Project] = dm.get_projects()

They are TypedDicts, so nothing changes at runtime — methods still return the plain dicts response.json() produces, and project["name"] works exactly as before. The types are checker-only.

To regenerate after the API changes:

python scripts/generate_schema.py

CI runs python scripts/generate_schema.py --check and fails if the committed types drift from the spec.

Five schemas cannot currently be generated (SetDetail, DatasetDetail, InputSetDetail, PackInstallListItem, SchemaGraphNavigation): they are composed with allOf, or only appear inside a union, so the generator does not emit them as a named class. Methods returning those keep Dict. The generator script lists them explicitly and fails if the set grows.