Error Handling

NyanSQLite-specific errors inherit from NyanSQLiteError. Pydantic input failures raise pydantic.ValidationError; SQLite failures retain the exception type of the active backend.

python

from pydantic import BaseModel, ValidationError
from nyansqlite import NyanSQLite, QueryValidationError, SchemaMismatchError

class User(BaseModel):
    id: int
    age: int

with NyanSQLite("app.db") as db:
    try:
        db.register(User)
        db.update(User, where={"id": 1}, age="invalid")
    except SchemaMismatchError as exc:
        print(f"Existing table differs from the model: {exc}")
    except ValidationError as exc:
        print(f"Invalid update value: {exc}")
    except QueryValidationError as exc:
        print(f"Invalid filter: {exc}")

register() checks an existing table's columns, types, primary key, nullability, and FTS searchable columns against the model. A mismatch raises SchemaMismatchError; it does not migrate tables or FTS definitions. Back up and explicitly migrate existing data before registering the changed model.

For ordinary models, update() validates the types and constraints of changed fields. With custom field or model validators or model configuration, it validates affected rows inside the transaction before writing. A validation failure leaves the rows unchanged.

An exception inside atomic() rolls the transaction back. The async client also keeps the connection protected until an in-flight SQLite call completes after cancellation.

See the exception reference for the available classes.

Error Handling ​

Error Handling