Normalizers

Normalizers¶

Normalizers are functions that transform field values before validation. They run in order, and each normalizer receives the output of the previous one. This is useful for cleaning, formatting, or transforming data before it is validated and stored.

How It Works¶

When a value is set on a field, the processing order is:

Normalizers run first - transform the value
Validators run second - validate the transformed value

Basic Usage¶

Pass a list of normalizer functions to the normalizers parameter of Field:

from pydbm import DbmModel, Field


class UserModel(DbmModel):
    username: str = Field(normalizers=[lambda x: x.lower()])

user = UserModel(username="HakanCelik")
assert user.username == "hakancelik"

Chaining Multiple Normalizers¶

You can chain multiple normalizers. They run in order:

from pydbm import DbmModel, Field


class UserModel(DbmModel):
    username: str = Field(normalizers=[str.strip, str.lower])

user = UserModel(username="  HakanCelik  ")
assert user.username == "hakancelik"

Using Named Functions¶

For more complex transformations, define named functions:

from pydbm import DbmModel, Field


def remove_special_chars(value: str) -> str:
    return "".join(c for c in value if c.isalnum())


def truncate(value: str) -> str:
    return value[:20]


class UserModel(DbmModel):
    username: str = Field(
        normalizers=[str.strip, str.lower, remove_special_chars, truncate]
    )

Normalizer Signature¶

A normalizer is any callable that takes a value and returns the transformed value:

def normalizer(value: Any) -> Any:
    ...

The type is defined as NormalizationT = Callable[[Any], Any] and can be imported from pydbm:

from pydbm import NormalizationT

Normalizers with Validators¶

Normalizers run before validators, so validators see the already-transformed value:

from pydbm import DbmModel, Field


def min_length(value: str) -> None:
    if len(value) < 3:
        raise ValueError("Username must be at least 3 characters")


class UserModel(DbmModel):
    username: str = Field(
        normalizers=[str.strip, str.lower],
        validators=[min_length],
    )

In this example, the username is first stripped and lowered, then validated for minimum length.