datafaker-types.example.yaml

# Custom Datafaker types for SeedStream.
#
# Datafaker exposes far more providers than SeedStream pre-registers. Map a SeedStream type key to
# a Datafaker method path (each dot-separated segment is a no-arg method, invoked left to right
# starting on the Faker instance). The chain is validated at load time, so a typo fails fast.
#
# Use with:  ./seedstream inspect schema.sql --faker-types config/datafaker-types.example.yaml
#       and: ./seedstream execute --job ... --faker-types config/datafaker-types.example.yaml
#
# Once registered, a structure can reference the key directly, e.g.  datatype: beer_style
# The inspector also auto-targets a custom type when a column name matches its key.

types:
  beer_style: beer.style                 # faker.beer().style()
  beer_name: beer.name                   # faker.beer().name()
  pokemon: pokemon.name                  # faker.pokemon().name()
  full_address: address.fullAddress      # faker.address().fullAddress()
  job_title: job.title                   # faker.job().title()
  company_industry: company.industry     # faker.company().industry()

  # A `regex:`-prefixed value generates strings matching the regex after `regex:` (via RgxGen, the
  # engine Datafaker uses). Use it for structured IDs/references. Invalid patterns fail fast at load;
  # generation is deterministic under the seed.
  # Quirks: `.` matches any printable ASCII (incl. punctuation) — prefer explicit classes like
  # [A-Za-z0-9]; unbounded quantifiers (+ * {n,}) are capped at 100 reps. See config/README.md.
  iso_msg_id: "regex:[A-Z0-9]{10,35}"    # e.g. R7ZK9444J9
  order_ref: "regex:ORD-\\d{8}"          # e.g. ORD-40021785

aliases:
  beerstyle: beer_style