fix: support StringConstraints metadata in SQLAlchemy type generation

[Sujit-1509]

Summary

When using Pydantic v2's StringConstraints inside Annotated, SQLModel correctly applies validation constraints but does not propagate max_length to the generated SQLAlchemy column type.

For example:

from typing import Annotated
from pydantic import StringConstraints

class Item(SQLModel, table=True):
    id: int | None = Field(default=None, primary_key=True)
    name: Annotated[str, StringConstraints(max_length=50)]

currently generates a column with no length constraint, even though max_length=50 is declared.

Root Cause

get_field_metadata() recognizes PydanticMetadata and MaxLen metadata objects, but does not recognize StringConstraints, causing the metadata extractor to return FakeMetadata() instead of the constraint object.

As a result, get_sqlalchemy_type() cannot read max_length and falls back to an unconstrained string type.

Changes

• Add support for StringConstraints in get_field_metadata().
• Add a regression test verifying that Annotated[str, StringConstraints(max_length=50)] produces a SQLAlchemy column with length == 50.

Testing

Added a regression test that fails on the current implementation and passes with this change.

[YuriiMotov]

@Sujit-1509, do you have specific use case there it's required to use StringConstraints instead of passing max_length via Field?

You marked it as fix, but this is not a fix because it's not documented in SQLMOdel's docs.
To accept this as a feature request we need stronger arguments.
Please, avoid opening PRs just to contribute to the project. We are overloaded with PRs..

[Sujit-1509]

@YuriiMotov Thanks for the feedback.

The use case I had in mind is projects migrating to Pydantic v2 patterns and reusing constrained types across multiple models.

For example:

from typing import Annotated
from pydantic import StringConstraints

Name = Annotated[str, StringConstraints(max_length=50)]

class User(SQLModel, table=True):
    id: int | None = Field(default=None, primary_key=True)
    name: Name

In this case validation correctly enforces max_length=50, but the generated SQLAlchemy column does not receive the corresponding length constraint.

My assumption was that this was an inconsistency because SQLModel already extracts length-related metadata to configure SQLAlchemy column types, but currently ignores StringConstraints metadata.

That said, I understand your point that SQLModel's documented approach is Field(max_length=...). If support for StringConstraints is considered a feature request rather than a bug fix, I'm happy to reclassify the PR accordingly or discuss whether this functionality would be useful for users adopting Pydantic v2 style constraints.