Easy to make mistakes when defining nullable (Field() vs Field(Column())

[martin-greentrax]

First Check

• I added a very descriptive title to this issue.
• I used the GitHub search to find a similar issue and didn't find it.
• I searched the SQLModel documentation, with the integrated search.
• I already searched in Google "How to X in SQLModel" and didn't find any information.
• I already read and followed all the tutorial in the docs and didn't find an answer.
• I already checked if it is not related to SQLModel but to Pydantic.
• I already checked if it is not related to SQLModel but to SQLAlchemy.

Commit to Help

• I commit to help with one of those options 👆

Example Code

Here a sample model:


class FooDB(SQLModel, table=True):
    id: int = Field(primary_key=True, nullable=False)
    datetime_1: datetime = Field(sa_column=Column(DateTime), nullable=False)
    datetime_2: datetime = Field(sa_column=Column(DateTime, nullable=False))

    string_1: str = Field(sa_column=Column(String(50)))
    string_2: str = Field(max_length=50)

Alembic creates the table as following:

    op.create_table(
        "foodb",
        sa.Column("datetime_1", sa.DateTime(), nullable=True),
        sa.Column("datetime_2", sa.DateTime(), nullable=False),
        sa.Column("string_1", sa.String(length=50), nullable=True),
        sa.Column("string_2", sqlmodel.sql.sqltypes.AutoString(length=50), nullable=False),
        sa.PrimaryKeyConstraint("id"),
    )

### Description

I had some surprises when looking closer at what data my tables actually accept vs. what I saw in my models.py. I find that it is very easy that one thinks one is doing the right thing (while looking at the pydantic/sqlmodel docs), but the definition does actually never make it into the database.

In my sample model I would have expected that the nullable-Attribute would lead to the same in each pair.
  
Related to https://github.com/tiangolo/sqlmodel/issues/420

### Operating System

Linux

### Operating System Details

_No response_

### SQLModel Version

0.0.8

### Python Version

Python 3.10.7

### Additional Context

_No response_

[yixiongngvsys]

I overlooked your open issue, currently I am facing the same issue (#466) seems like if the nullable is set outside Column(), this field will be ignored

[tiangolo]

If you use the SQLAlchemy column that overrides everything. It's made as an escape hatch so that you can take the wheel and SQLModel won't interfere.

[martin-greentrax]

@tiangolo why is nullable=False in datetime_1 silently ignored though? To me that feels error prone :-|

[tiangolo]

Thank you! It was indeed error-prone, I added #681 to solve it.

Now it's not possible to combine sa_column with other arguments that would be invalid/not used. It will raise an error. And it also includes the typing tricks so that the editor will complain right there while editing if a developer uses sa_column with one of those incompatible arguments.

It will be available in the next version, 0.0.11. 🎉

[github-actions]

Assuming the original need was handled, this will be automatically closed now. But feel free to add more comments or create new issues or PRs.