Add python validators for decimal constraints (max_digits and decimal_places)

[sydney-runkle]

Closes #10498

This specifically addresses issues where there is annotated metadata applied between the type and the decimal constraints.

For this sample:

import decimal
from pydantic import BaseModel, Field, BeforeValidator, ValidationError
from typing import Annotated

DecimalAnnotation = Annotated[decimal.Decimal, BeforeValidator(lambda v: v)]


class ExampleSchema(BaseModel):
    example_value: Annotated[DecimalAnnotation, Field(max_digits=10, decimal_places=4)] | None

e = ExampleSchema(example_value=decimal.Decimal('123.4567'))

try:
    ExampleSchema(example_value=decimal.Decimal('123.45678'))
except ValidationError as e:
    print(e)
    """
    example_value
    Decimal input should have no more than 4 decimal places [type=decimal_max_places, input_value=Decimal('123.45678'), input_type=Decimal]
        For further information visit https://errors.pydantic.dev/2.10/v/decimal_max_places
    """

try:
    ExampleSchema(example_value=decimal.Decimal('12345678.901'))
except ValidationError as e:
    print(e)
    """
    example_value
    Decimal input should have no more than 10 digits in total [type=decimal_max_digits, input_value=Decimal('12345678.901'), input_type=Decimal]
        For further information visit https://errors.pydantic.dev/2.10/v/decimal_max_digits
    """

The output used to be:

  File "/Users/programming/pydantic_work/pydantic/pydantic/_internal/_known_annotated_metadata.py", line 293, in apply_known_metadata
    raise RuntimeError(f"Unable to apply constraint '{constraint}' to schema of type '{schema_type}'")
RuntimeError: Unable to apply constraint 'max_digits' to schema of type 'function-before'

[cloudflare-workers-and-pages]

Deploying pydantic-docs with    Cloudflare Pages

Latest commit:	9adef0c
Status:	✅  Deploy successful!
Preview URL:	https://99e28bdc.pydantic-docs.pages.dev
Branch Preview URL:	https://add-decimal-validators.pydantic-docs.pages.dev

View logs

[codspeed-hq]

CodSpeed Performance Report

Merging #10506 will not alter performance

_{Comparing add-decimal-validators (9adef0c) with main (1f74db2)}

Summary

✅ 38 untouched benchmarks

[github-actions]

Coverage report

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
pydantic/_internal
_known_annotated_metadata.py					298-300
_validators.py					351, 385, 400
Project Total

This report was generated by python-coverage-comment-action

[sydney-runkle]

I think it's alright if I merge this (given the approval and iterations on feedback), but if there's anything left unaddressed, happy to open another PR on Monday!

[Viicos]

I'm wondering: why is the Python validation only called when a validator is already present? Can we avoid duplication between pydantic-core and here?

Other thing, seems like allow_inf_nan suffers from the same issue.

[sydney-runkle]

I'm wondering: why is the Python validation only called when a validator is already present? Can we avoid duplication between pydantic-core and here?

Good question. I think this issue helps to explain things: #10036

Specifically, we use these Python validators for cases where it'd violate our promise to maintain the order of annotations if we directly applied the constraint to the core schema. But, the validator is still compatible with the type after the custom validator is called, so we can use these python wrappers.