Pydantic-YAML

Pydantic-YAML adds YAML capabilities to Pydantic, which is an excellent Python library for data validation and settings management. If you aren't familiar with Pydantic, I would suggest you first check out their docs.

Documentation on ReadTheDocs.org

Basic Usage

from enum import Enum
from pydantic import BaseModel, validator
from pydantic_yaml import parse_yaml_raw_as, to_yaml_str

class MyEnum(str, Enum):
    """A custom enumeration that is YAML-safe."""

    a = "a"
    b = "b"

class InnerModel(BaseModel):
    """A normal pydantic model that can be used as an inner class."""

    fld: float = 1.0

class MyModel(BaseModel):
    """Our custom Pydantic model."""

    x: int = 1
    e: MyEnum = MyEnum.a
    m: InnerModel = InnerModel()

    @validator("x")
    def _chk_x(cls, v: int) -> int:  # noqa
        """You can add your normal pydantic validators, like this one."""
        assert v > 0
        return v

m1 = MyModel(x=2, e="b", m=InnerModel(fld=1.5))

# This dumps to YAML and JSON respectively
yml = to_yaml_str(m1)
jsn = m1.json()

# This parses YAML as the MyModel type
m2 = parse_yaml_raw_as(MyModel, yml)
assert m1 == m2

# JSON is also valid YAML, so this works too
m3 = parse_yaml_raw_as(MyModel, jsn)
assert m1 == m3

With Pydantic v2, you can also dump dataclasses:

from pydantic import RootModel
from pydantic.dataclasses import dataclass
from pydantic.version import VERSION as PYDANTIC_VERSION
from pydantic_yaml import to_yaml_str

assert PYDANTIC_VERSION >= "2"

@dataclass
class YourType:
    foo: str = "bar"

obj = YourType(foo="wuz")
assert to_yaml_str(RootModel[YourType](obj)) == 'foo: wuz\n'

Configuration

Currently we use the JSON dumping of Pydantic to perform most of the magic.

This uses the Config inner class, as in Pydantic:

class MyModel(BaseModel):
    # ...
    class Config:
        # You can override these fields, which affect JSON and YAML:
        json_dumps = my_custom_dumper
        json_loads = lambda x: MyModel()
        # As well as other Pydantic configuration:
        allow_mutation = False

You can control some YAML-specfic options via the keyword options:

to_yaml_str(model, indent=4)  # Makes it wider
to_yaml_str(model, map_indent=9, sequence_indent=7)  # ... you monster.

You can additionally pass your own YAML instance:

from ruamel.yaml import YAML
my_writer = YAML(typ="safe")
my_writer.default_flow_style = True
to_yaml_file("foo.yaml", model, custom_yaml_writer=my_writer)

A separate configuration for YAML specifically will be added later, likely in v2.

Breaking Changes for pydantic-yaml V1

The API for pydantic-yaml version 1.0.0 has been greatly simplified!

Mixin Class

This functionality has currently been removed! YamlModel and YamlModelMixin base classes are no longer needed. The plan is to re-add it before v1 fully releases, to allow the .yaml() or .parse_*() methods. However, this will be availble only for pydantic<2.

Versioned Models

This functionality has been removed, as it's questionably useful for most users. There is an example in the docs that's available.