Documentation

Pfg Gen Schema

Reference docs for pydantic-fixturegen.

main

pfg gen schema

Capabilities

pfg gen schema produces JSON Schema files for your models. Unlike pfg schema config, this command runs inside the generation pipeline so schemas share the same include/exclude filters, privacy profile, and watch mode as JSON/fixtures. It is ideal for bundling typed contracts alongside generated data.

Typical use cases

  • Ship per-model schemas with SDKs or documentation sites.
  • Validate how privacy profiles (for example pii-safe) transform schemas before regenerating artifacts.
  • Run in watch mode while iterating on models to immediately inspect schema diffs.

Inputs & outputs

  • Target: Python module containing models.
  • Output: templated --out path. Include {model} to emit one file per model, or point at a directory to leverage templating (for example schema/{model}.json).

Flag reference

  • --out/-o: required path or template.
  • --indent: override JSON indentation (default derived from config).
  • --include/-i, --exclude/-e: select which models to emit.
  • --profile: apply privacy bundles prior to schema generation, ensuring sensitive fields are masked/excluded the same way as JSON output.
  • --watch + --watch-debounce: keep the command running and regenerate schemas when any watched file changes.
  • --json-errors: (global) surface structured diagnostics when schema emission fails (for example due to invalid output templates).

Example workflows

Emit per-model schemas with privacy profile

pfg gen schema ./app/models.py \
  --out schema/{model}.json --include app.models.* \
  --profile pii-safe --indent 2
Writes each schema under `schema/`, ensuring PII-sensitive fields reflect the `pii-safe` bundle.

Sample output

[schema_emit] path=schema/app.models.User.json bytes=2417
[schema_emit] path=schema/app.models.Address.json bytes=1095
profile=pii-safe indent=2

Excerpt (schema/app.models.User.json)

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "title": "User",
  "type": "object",
  "properties": {
    "email": {"type": "string", "format": "email"},
    "profile": {
      "type": "object",
      "properties": {
        "marketing_opt_in": {"type": "boolean"},
        "timezone": {"type": "string"}
      }
    }
  }
}

Watch and regenerate a directory index

pfg gen schema ./app/models.py --out schema/index.json --watch
Keeps `schema/index.json` up to date as you tweak models.

Sample output

[watch] tracking=/repo/app/models.py,/repo/schema/index.json
Regenerated schema/index.json (models=17)

Operational notes

  • Watch mode tracks the module tree plus the resolved output directory so new files trigger reruns.
  • When templates create directories dynamically, fixturegen watches the parent folder to rebuild path previews safely.
  • Constraint summaries for schema emission are logged at debug level (helpful when diagnosing why a field is missing).
Edit this page