faker-jsonschema

[ Docs ]

faker-jsonschema is a Faker provider that generates random data conforming to a JSON Schema.

Pass any JSON Schema dict to faker.from_jsonschema() and get back a valid, randomly generated value.

1800+ unit and PBT tests aiming to ensure correctness. Limitations of our strategies are acknowledged in the docs.

Install

pip install faker-jsonschema
# or
uv add faker-jsonschema

Usage

from faker import Faker
from faker_jsonschema.provider import JSONSchemaProvider

fake = Faker()
fake.add_provider(JSONSchemaProvider)

schema = {
    "type": "object",
    "properties": {
        "name": {"type": "string"},
        "score": {"type": "integer", "minimum": 0, "maximum": 100},
        "tags": {"type": "array", "items": {"type": "string"}, "maxItems": 3},
    },
    "required": ["name", "score"],
}

print(fake.from_jsonschema(schema))
# {'name': 'Doctor pass describe matter.', 'score': 8}

See the full documentation here.

Acknowledgements

This project exists because I have used joke2k's Faker lib many times at work and find it an invaluable tool in tests.

JSON Schema has some property specs defined as regex in JS syntax. So we need a way to translate those into Python/PCRE syntax that we can execute. We have vendored a fork of the now unmaintained Zac-HD/js-regex lib for that purpose, which unlocked that feature for us.

And then we needed a way to generate random values that match such a regex, for this we make use of the brilliant Hypothesis Property-based testing library, which has a handy from_regex() generation strategy. We also have some PBT tests using Hypothesis for its intended purpose.

Contributing

See MAINTAINER.md for development setup, common tasks, and release instructions.