Describe your question When creating a Base class with @as_declarative the following errors shows up using mypy error: Name 'DeclarativeMeta' is not defined. Am I doing something wrong or is there a better way?

Example - please use the Minimal, Complete, and Verifiable guidelines if possible Using the example as provided in the documentation, it works without a problem:

Base = declarative_base()

class User(Base):
    __tablename__ = 'user'

    id = Column(Integer, primary_key=True)
    name = Column(String)

    addresses: Mapped[List["Address"]] = relationship("Address", back_populates="user")

class Address(Base):
    __tablename__ = 'address'

    id = Column(Integer, primary_key=True)
    user_id: int = Column(ForeignKey("user.id"))

    user: Mapped[User] = relationship(User, back_populates="addresses")

main  mypy --config mypy.ini --strict simple-test.py
Success: no issues found in 1 source file

When trying to use it as following it shows the error:

@as_declarative()
class Base(object):
    id = Column(Integer, primary_key=True)


class User(Base):
    __tablename__ = "user"

    name = Column(String)

    addresses: Mapped[List["Address"]] = relationship("Address", back_populates="user")


class Address(Base):
    __tablename__ = "address"

    user_id: int = Column(ForeignKey("user.id"))

    user: Mapped[User] = relationship(User, back_populates="addresses")

main  mypy --config mypy.ini --strict simple-test.py
simple-test.py: error: Name 'DeclarativeMeta' is not defined

mypy config

[mypy]
plugins = sqlalchemy.ext.mypy.plugin

Versions

• OS: MacOS 11.0
• Python: Python 3.9.2
• Mypy: mypy 0.812
• SQLAlchemy: 1.4.3
• sqlalchemy2-stub: 0.0.1a4
• Database: n/a
• DBAPI: n/a

Description

Updates code style and pre-commit checks to match sqlalchemy. A few things to note:

• # fmt: off + # fmt: on was needed to keep black and zimports from fighting
• I added flake8-pyi and did not include flake8-builtins, flake8-docstrings, and flake8-rst-docstrings because they are irrelevant for pyi files

Checklist

This pull request is:

• [ ] A documentation / typographical error fix
  • Good to go, no issue or tests are needed
• [x] A short code fix
  • please include the issue number, and create an issue if none exists, which must include a complete example of the issue. one line code fixes without an issue and demonstration will not be accepted.
  • Please include: Fixes: #<issue number> in the commit message
  • please include tests. one line code fixes without tests will not be accepted.
• [ ] A new feature implementation
  • please include the issue number, and create an issue if none exists, which must include a complete example of how the feature would look.
  • Please include: Fixes: #<issue number> in the commit message
  • please include tests.

Have a nice day!

Describe the bug mypy fails on asynchronous sessionmaker initialization.

Expected behavior No errors.

To Reproduce Please try to provide a Minimal, Complete, and Verifiable example. See also Reporting Bugs on the website, and some example issues.

Please do not use Flask-SQLAlchemy or any other third-party extensions or dependencies in test cases. The test case must illustrate the problem without using any third party SQLAlchemy extensions. Otherwise, please report the bug to those projects first.

import os

from sqlalchemy.ext.asyncio import AsyncSession
from sqlalchemy.ext.asyncio import create_async_engine
from sqlalchemy.orm import sessionmaker

engine = create_async_engine(os.getenv("DATABASE_URL"))
async_session = sessionmaker(engine, class_=AsyncSession)

Error

session.py:8: error: Value of type variable "_TSessionMakerType" of "sessionmaker" cannot be "AsyncSession"
Found 1 error in 1 file (checked 1 source file)

Versions.

• OS: Linux
• Python: 3.9.6
• SQLAlchemy: 1.4.23
• mypy: 0.910
• SQLAlchemy2-stubs: v0.0.2a9

Additional context The problem appeared on the latest release of sqlalchemy2-stubs and was not present on v0.0.2a8.

Have a nice day!

I was just trying out the new stubs on an existing project if mine, and it looks like you did some awesome work there - I can get rid of a lot of workarounds. Thanks! Nevertheless, there's one issue I wanted to share with you.

Describe the bug There are two ways to pass the first parameter of a relationship: by class or by str. When using the class directly everything works as expected, the type is detected when using uselist or relationsship. And I don't find any false positives afterwards.

When I use a str to configure my relationship things get pretty messy. First, I need to type the annotations manually, I wasn't able to find any way to configure it such that auto-detection works.

[SQLAlchemy Mypy plugin] Can't infer type from ORM mapped expression assigned to attribute 'addresses'; please specify a Python type or Mapped[<python type>] on the left hand side.

Second, even when annotating it manually, I'll have to cast it every single time I use it (when it's some kind of Iterable as shown in the example below).

error: "List?[Address?]" has no attribute "__iter__" (not iterable)
error: "List?[Address?]" has no attribute "append"

Expected behavior Type detection works for relationship independent of type of first argument.

To Reproduce

from __future__ import annotations

from typing import List

from sqlalchemy import Integer, Column, String, ForeignKey
from sqlalchemy.orm import declarative_base, relationship

Base = declarative_base()


class User(Base):

    __tablename__ = "users"

    id = Column(Integer, primary_key=True)
    name = Column(String)

    addresses: List[Address] = relationship("Address", back_populates="user", uselist=True)

    def clone(self) -> User:
        new_user = User(name=self.name)

        for existing_addr in self.addresses:
            new_user.addresses.append(
                Address(address=existing_addr.address)
            )
        return new_user


class Address(Base):

    __tablename__ = "addresses"

    id = Column(Integer, primary_key=True)
    address = Column(String)
    
    user_id = Column(Integer, ForeignKey("users.id"))
    
    user = relationship(User, back_populates="addresses", uselist=False)

Versions.

• OS: WIn10
• Python: 3.8.6
• SQLAlchemy: 1.4.9
• Database:
• DBAPI:

Have a nice day! You too ;)

The server default is any expression that is passed as DDL to the database and these usually don't have SQL types explicitly stated.

With the code as is, pylance is complaining about this:

Column(DateTime(), server_default=func.now())

and requiring I do this:

Column(DateTime(), server_default=func.now(type_=DateTime))

people don't need to do that, server_defaults SQL type always comes from the column type and doesn't normally need to be stated.

Also, column_server_default.py was importing "functions as func", which is wrong. "func" is not a module it's a namespace object, fixed that.

Describe the bug mypy is complaining about unpacking Row object, even though it works.

To Reproduce

test_table = Table(
    "test_table",
    metadata,
    Column("i", UUID(as_uuid=True), nullable=False, primary_key=True),
    Column("x", UUID(as_uuid=True), index=True),
    Column("y", UUID(as_uuid=True), index=True),
)
query = test_table.select()
async with engine.connect() as connection:
  result = await connection.execute(query)
row = result.fetchone()
if row:
  my_dict = {**row}

Error

Argument after ** must be a mapping, not "Row"

Versions.

• OS: Linux Mint 19.3 Cinnamon
• Python: 3.7.10
• SQLAlchemy: 1.4.15
• Database: PostgreSQL
• DBAPI: asyncpg

Description

Improve orm.session. Depends on #46. This is a large one because of _SessionProtocol and _AsyncSessionProtocol. These are needed to allow sessionmaker() to take classes that aren't required to inherit from Session or AsyncSession, but are duck-typed to be like Session or AsyncSession.

Fixes #17.

Checklist

This pull request is:

• [ ] A documentation / typographical error fix
  • Good to go, no issue or tests are needed
• [X] A short code fix
  • please include the issue number, and create an issue if none exists, which must include a complete example of the issue. one line code fixes without an issue and demonstration will not be accepted.
  • Please include: Fixes: #<issue number> in the commit message
  • please include tests. one line code fixes without tests will not be accepted.
• [ ] A new feature implementation
  • please include the issue number, and create an issue if none exists, which must include a complete example of how the feature would look.
  • Please include: Fixes: #<issue number> in the commit message
  • please include tests.

Have a nice day!

I installed the the type stubs and configured the mypy plugin as directed in the documentation

My database is defined like this:

@as_declarative()
class Base:
    def __repr__(self):
        name = self.__class__.__name__
        attrs = (
            "%s=%r" % (attr, getattr(self, attr))
            for attr in self._sa_class_manager.keys()
            if not (attr[-2:] == "id" or isinstance(getattr(self, attr), list))
        )
        return name + "(%s)" % ", ".join(attrs)


class Source(Base):
    __tablename__ = "sources"
    id = sa.Column(sa.Integer, primary_key=True)
    address = sa.Column(sa.String, index=True)
    match_id = sa.Column(sa.Integer, sa.ForeignKey("matches.id"))

    match = relationship("Match", back_populates="sources")


class Match(Base):
    __tablename__ = "matches"
    id = sa.Column(sa.Integer, primary_key=True)
    original_id = sa.Column(sa.Integer, sa.ForeignKey("original.id"))
    romanized_id = sa.Column(sa.Integer, sa.ForeignKey("romanized.id"))
    count = sa.Column(sa.Integer)

    original = relationship("Original", back_populates="matches")
    romanized = relationship("Romanized", back_populates="matches")
    sources = relationship("Source", back_populates="match")


sa.Index("matches_idx", Match.original_id, Match.romanized_id, unique=True)


class Standard(Base):
    __tablename__ = "standards"
    id = sa.Column(sa.Integer, primary_key=True)
    st = sa.Column(sa.String, index=True, unique=True)


class Original(Base):
    __tablename__ = "original"
    id = sa.Column(sa.Integer, primary_key=True)
    form = sa.Column(sa.String, index=True, unique=True)

    matches = relationship("Match", back_populates="original")


class Romanized(Base):
    __tablename__ = "romanized"
    id = sa.Column(sa.Integer, primary_key=True)
    form = sa.Column(sa.String, index=True)
    standard_id = sa.Column(sa.Integer, sa.ForeignKey("standards.id"))

    standard = relationship(Standard)
    matches = relationship("Match", back_populates="romanized")


sa.Index("standard_form", Romanized.form, Romanized.standard_id, unique=True)

When I run mypy, I get these errors:

$ mypy deromanize 
deromanize/cacheutils.py:201: error: Need type annotation for 'match'
deromanize/cacheutils.py:211: error: Need type annotation for 'original'
deromanize/cacheutils.py:212: error: Need type annotation for 'romanized'
deromanize/cacheutils.py:213: error: Need type annotation for 'sources'
deromanize/cacheutils.py:230: error: Need type annotation for 'matches'
deromanize/cacheutils.py:239: error: Need type annotation for 'standard'
deromanize/cacheutils.py:240: error: Need type annotation for 'matches'
Found 7 errors in 1 file (checked 6 source files)

This does not match the documented error message:

test3.py:22: error: [SQLAlchemy Mypy plugin] Can't infer scalar or
collection for ORM mapped expression assigned to attribute 'user'
if both 'uselist' and 'collection_class' arguments are absent from the
relationship(); please specify a type annotation on the left hand side.
Found 1 error in 1 file (checked 1 source file)

Furthermore, when I attempt to add annotations, that doesn't work either:

class Source(Base):
    __tablename__ = "sources"
    id = sa.Column(sa.Integer, primary_key=True)
    address = sa.Column(sa.String, index=True)
    match_id = sa.Column(sa.Integer, sa.ForeignKey("matches.id"))

    match: 'Match' = relationship("Match", back_populates="sources")

deromanize/cacheutils.py:201: error: Incompatible types in assignment (expression has type "RelationshipProperty[<nothing>]", variable has type "Match")

Am I doing something wrong or is this a bug?

Describe the bug Using on_conflict_do_update method with returning and PyCharm highlight return with "Cannot find reference 'returning' in 'None'"

Expected behavior Don't see any highlight.

To Reproduce Install packages, and try to use

insert(self.table).on_conflict_do_update(
      index_elements=(self.table.c.id,),
      set_={
          'count': self.table.c.count + insert_query.excluded.count,
      },
  ).returning(self.table.c.id)

Error

"Cannot find reference 'returning' in 'None'"

Proposition Change:

def on_conflict_do_update(
        self,
        constraint: Optional[Any] = ...,
        index_elements: Optional[Any] = ...,
        index_where: Optional[Any] = ...,
        set_: Optional[Any] = ...,
        where: Optional[Any] = ...,
    ) -> 'Insert': ...
    def on_conflict_do_nothing(
        self,
        constraint: Optional[Any] = ...,
        index_elements: Optional[Any] = ...,
        index_where: Optional[Any] = ...,
    ) -> 'Insert': ...

Versions.

• OS: Ubuntu 20.4
• IDE: PyCharm Professional 2022.1
• Python: 3.8
• SQLAlchemy: 1.4.32
• mypy: 0.782
• SQLAlchemy2-stubs: 0.0.2a21

Have a nice day!

Describe the bug If I use a DDL construct, mypy conplains that it can't be executed. However using .execute raises a deprecation warning:

testscript.py:9: RemovedIn20Warning: The DDLElement.execute() method is considered legacy as of the 1.x series of SQLAlchemy and will be removed in 2.0. All statement execution in SQLAlchemy 2.0 is performed by the Connection.execute() method of Connection, or in the ORM by the Session.execute() method of Session. (Background on SQLAlchemy 2.0 at: https://sqlalche.me/e/b8d9)

Expected behavior Mypy doesn't complain

To Reproduce

from sqlalchemy.schema import CreateSequence, Sequence
from sqlalchemy.orm import sessionmaker
from sqlalchemy import create_engine

engine = create_engine("sqlite:////:memory:")
session = sessionmaker(bind=engine)()

construct = CreateSequence(Sequence("foo"))
try:
    construct.execute(bind=session)  # warning
    session.execute(construct)  # mypy
except:
    pass

Error

testscript.py:10: error: Argument "bind" to "execute" of "DDLElement" has incompatible type "Session"; expected "Union[Engine, Connection, None]"

Versions.

• OS: Linux
• Python: 3.7.1
• SQLAlchemy: 1.4.27
• mypy: 0.910
• SQLAlchemy2-stubs: 0.0.2a19

Have a nice day!

In most cases a type only superclass can avoid repeating the definitions in one or more classes.

Examples: Session|ScopedSession|AsyncSession; Connection|AsyncConnection

Describe the bug SQLAlchemy release 1.4.33 added a new optional close: bool = True parameter to Engine#dispose(). This is not declared in the type definition, at https://github.com/sqlalchemy/sqlalchemy2-stubs/blob/dd0e6cf7e1c61cc1c25cd1d8d72c9f322fa73d90/sqlalchemy-stubs/engine/base.pyi#L200

Expected behavior The current type definition should declare the close parameter.

To Reproduce Call Engine.dispose(close) and then run mypy on it with sqlalchemy2-stubs.

engine.dispose(close=False)

Error

error: Too many arguments for "dispose" of "Engine"

Versions.

• OS: Darwin Kernel Version 21.6.0
• Python: 3.7.10
• SQLAlchemy: 1.4.39
• mypy: 0.931
• SQLAlchemy2-stubs: 0.0.2a29

Describe the bug mypy returns [Invald ](error: Invalid base class "Base" [misc]) when Base class is returned by a function.

To Reproduce

pip install sqlalchemy[mypy]==1.4.44
mypy --config-file=mypy.ini test.py

mypy.ini

[mypy]
plugins = sqlalchemy.ext.mypy.plugin

test.py

from sqlalchemy import String, select
from sqlalchemy.orm import declarative_base
from typing import Optional, Dict, Type
import sqlalchemy
from sqlalchemy.orm.ext import DeclarativeMeta

def get_base() -> Type[DeclarativeMeta]:
    return declarative_base()

Base: Type[DeclarativeMeta] = get_base()

class User(Base):
    __tablename__ = "user"

    id = sqlalchemy.Column(sqlalchemy.Integer, primary_key=True)
    name = sqlalchemy.Column(sqlalchemy.String)

Error

error: Invalid base class "Base"

Versions.

• OS: Ubuntu 22.04
• Python: 3.8
• SQLAlchemy: 1.4.44
• mypy: 0.991
• SQLAlchemy2-stubs: 0.0.2a29

Have a nice day!

Relevant block of code:

sa.and_(
    tb.c.VERSION == version,
    tb.c.ISIN.in_(isins) if isins is not None else True,
    tb.c.ISIN != None,  # pylint: disable=singleton-comparison
     b.c.DATE.in_(dates))
)

Error I get from pyright is:

error: Argument of type "ColumnElement[Boolean] | Literal[True]" cannot be assigned to parameter "clauses" of type "_CLE@BooleanClauseList" in function "and_"
    Type "ColumnElement[Boolean] | Literal[True]" cannot be assigned to type "ColumnElement[Boolean]"
      "Literal[True]" is incompatible with "ColumnElement[Boolean]" (reportGeneralTypeIssues)

Describe the bug The compile decorator is not fully typed. Therefore mypy complains when using it

Expected behavior Mypy doesn't complain

To Reproduce Use compile as a decorator

Error

error: Untyped decorator makes function "foo" untyped  [misc]

Versions.

• OS: Linux
• Python: 3.7.1
• SQLAlchemy: 1.4.41
• mypy: 0.982
• SQLAlchemy2-stubs: 0.0.2a29

I'm happy to give a PR a try

Is your feature request related to a problem? Please describe.

A typical typed SQLAlchemy model looks like this:

class Model(Base):
    name: Mapped[str] = Column(Unicode(100), nullable=True)

This works fine with mypy, but triggers the following error with Pylance/Pyright

Expression of type "Column[Unicode]" cannot be assigned to declared type "Mapped[str]"

Same for relationship

Describe the solution you'd like

Make Column and relationship inherit from Mapped

Additional context

I can make a PR if you're okay with this.

Have a nice day!