typeshed icon indicating copy to clipboard operation
typeshed copied to clipboard

Collection of library stubs for Python, with static types

Results 615 typeshed issues
Sort by recently updated
recently updated
newest added

When we submitted PR that adds the `stdlib/_ctypes.pyi`(#8582), imports some stuffs from `ctypes`. But [the implementation imports some stuffs from `_ctypes` to `ctypes`](https://github.com/python/cpython/blob/0ace820bec8892d621a4aadc1feb6c56e25560bf/Lib/ctypes/__init__.py#L8-L14). - also related to https://github.com/python/typeshed/issues/8571#issuecomment-1221316973 However, trying...

```python from datetime import date from dateutil.rrule import MONTHLY, rrule datetimes = [dt for dt in rrule(MONTHLY, dtstart=date(2001, 1, 1), until=date(2005, 1, 1))] reveal_type(datetimes) # Revealed type is "builtins.list[Any]" ```

type-incomplete
help wanted

I've encountered cases where the changes from `int` to a literal type in stubs causes regressions for mypy users. Here's one (simplified) example: ```py import logging class C: def __init__(self)...

project-discussion

_Originally posted [here](https://github.com/python/mypy/issues/4128#issuecomment-1196762683)_ Let me know if I need to scope this better. Even if this gets closed, it's also fine - if you have a better scope / new...

now: `_ECT = Callable[[Optional[Type[_CData]], _FuncPointer, Tuple[_CData, ...]], _CData]` should be: `_ECT = Callable[[Optional[_CData], _FuncPointer, Tuple[_CData, ...]], _CData]` Because, at least, the return type of this function must be the same...

size-small

An extension of #2566, seems like the typings for `parse_known_args`, `parse_intermixed_args`, `parse_known_intermixed_args` (and the undocumented `_parse_known_args`) of `argparse.ArgumentParser` were overlooked, as they all share the same pattern as `parse_args` for...

#6858 added stubs for urllib3. It might have been a mistake, because urllib3 seems to have its own type annotations and even a `py.typed` marker, making the stubs unnecessary.

removal

Consider the following: ```python3 names: set[str] = {"a", "b"} name: str | None = None names.discard(name) ``` It would be nice if this wasn't a type checker error. Unlike `.remove()`,...

The type of the first parameter of `http.server.BaseHTTPRequestHandler.__init__` shall be `socketserver._RequestType`, not `bytes`. To fix this issue, simply remove the `__init__` method and make `BaseHTTPRequestHandler` inherit from `StreamRequestHandler`.