Database And Tables¶

The Ormdantic object is the registry for one database connection.

from ormdantic import Ormdantic

db = Ormdantic("sqlite:///app.sqlite3")

It stores:

the connection URL;
registered table metadata;
registered namespaces, sequences, and views;
event handlers;
the native Rust runtime once initialized;
the migration manager.

Register A Table¶

Use @db.table(...) on a Pydantic model:

from pydantic import BaseModel


@db.table(pk="id", indexed=["email"], unique=["email"])
class User(BaseModel):
    id: str
    email: str
    active: bool = True

The model remains a Pydantic model. Ormdantic stores the database metadata beside it.

Table Decorator Options¶

Common options:

Option	Purpose
`pk`	Primary key field name.
`tablename`	Override the generated table name.
`indexed`	Create simple indexes for fields.
`unique`	Create simple unique constraints for fields.
`columns`	Per-field `TableColumn` metadata.
`indexes`	Explicit `TableIndex` objects.
`check_constraints`	Table-level `TableCheck` objects.
`unique_constraints`	Table-level `TableUnique` objects.
`foreign_key_constraints`	Composite or named `TableForeignKey` objects.
`comment`	Table comment where the backend supports it.

Advanced backend options include PostgreSQL tablespaces and partitions, SQL Server filegroups and clustered primary keys, MySQL table options, SQLite conflict behavior, and Oracle table compression.

Initialize Schema¶

await db.init()

init() creates registered namespaces, sequences, tables, indexes, comments, backend-specific options, and views. For existing production databases, prefer migrations instead of calling create_all() blindly.

Table Handles¶

Access a model's table handle with db[Model]:

users = await db[User].find_many({"active": True})

The handle methods include:

find_one;
find_many;
insert;
update;
upsert;
delete;
count;
select;
update_where.

See Table API.