Data contract validator

When to invoke

• You receive JSON lines / JSON arrays from another system and need to ensure they meet a contract.
• You want a machine-readable report of missing fields, type mismatches, and simple rule violations.

Inputs needed

• --contract: path to a contract JSON file describing fields, requiredness, and simple constraints.
• --data: path to a JSON file (either an array of objects or JSON Lines).

Contract format (example)

• fields: object keyed by field name with:
  • type: one of string|number|integer|boolean|object|array|null
  • required: true/false
  • min_length / max_length (strings)
  • min / max (numbers)
  • regex (strings)
• rules: list of cross-field rules:
  • name
  • if: list of conditions (field, op, value)
  • then: list of requirements (field, op, value)

Workflow

1. Load contract.
2. Stream records from the data file.
3. Validate each record:
   • Required fields present
   • Types match
   • Field constraints pass
   • Cross-field rules pass
4. Emit a JSON report with counts and per-record errors.

Output format

JSON:

• summary: totals + error counts
• errors: list of {index, record_id, field, code, message}

Guardrails

• Do not mutate the input data.
• Keep error messages deterministic.
• Validate types conservatively (e.g., do not coerce strings to numbers).

Reference code

validate_contract.py