schemap 0.3.0

Automatic Pydantic Schemas for SQLAlchemy models

Schemap

Automatic Pydantic v2 schemas from SQLAlchemy 2.0 ORM models.

pip install schemap

Quick Start

from schemap import AutoBase
from sqlalchemy.orm import Mapped, mapped_column

class User(AutoBase):
    __tablename__ = "users"

    id: Mapped[int] = mapped_column(primary_key=True)
    name: Mapped[str]
    email: Mapped[str]

User.Schema         # all columns
User.CreateSchema   # excludes PKs, server_defaults, defaults
User.UpdateSchema   # all fields Optional (partial updates)
User.PublicSchema   # excludes __-prefixed fields

# Round-trip ORM <-> Pydantic
data = User.CreateSchema(name="Alice", email="[e-mail]")
user = User.from_schema(data)
schema = user.to_schema()

SchemaConfig

Customize generated schemas per model using __schema_config__:

from schemap import AutoBase
from schemap.config import SchemaConfig
from sqlalchemy.orm import Mapped, mapped_column

class User(AutoBase):
    __tablename__ = "users"
    __schema_config__ = SchemaConfig(
        exclude_public=["email"],
        exclude_create=["internal_id"],
    )

    id: Mapped[int] = mapped_column(primary_key=True)
    name: Mapped[str]
    email: Mapped[str]
    internal_id: Mapped[int]

Option	Type	Description
exclude_always	list[str]	Exclude from all schema variants
exclude_create	list[str]	Exclude from CreateSchema only
exclude_update	list[str]	Exclude from UpdateSchema only
exclude_public	list[str]	Exclude from PublicSchema only
field_overrides	dict[str, Any]	Override a field's Python type
required_always	list[str]	Force fields to be required
optional_always	list[str]	Force fields to be optional

Standalone

from schemap import build_schema
from schemap.config import SchemaConfig

UserSchema = build_schema(User, "create", config=SchemaConfig(exclude_create=["internal_id"]))

Requirements

Python ≥ 3.12.7, SQLAlchemy ≥ 2.0.49, Pydantic ≥ 2.13.4.

License

MIT