Type errors when using limit=None for reddit.inbox.unread

[foolo]

Describe the Bug

Steps:
make a call to reddit.inbox.unread(limit=None), as described in https://praw.readthedocs.io/en/stable/code_overview/reddit/inbox.html#praw.models.Inbox.unread

See code example below.

Observe the type checker errors (from pyright):

praw_hints_example.py:6:39 - error: Argument of type "None" cannot be assigned to parameter "limit" of type "str | int | dict[str, str]" in function "unread"
    Type "None" is not assignable to type "str | int | dict[str, str]"
      "None" is not assignable to "str"
      "None" is not assignable to "int"
      "None" is not assignable to "dict[str, str]" (reportArgumentType)

Or in the IDE:

Note: The program runs without crashing but I don't know if it behaves correctly because I don't have an available unread feed which exceeds the default limit of 100.

Desired Result

No type error should show

Code to reproduce the bug

import praw

def example(reddit: praw.Reddit):
    items = reddit.inbox.unread(limit=None)
    for item in items:
        print(item)

My code does not include sensitive credentials

• Yes, I have removed sensitive credentials from my code.

Relevant Logs

N/A

This code has previously worked as intended

I'm not sure, I haven't used this code before.

Operating System/Environment

Ubuntu 23.10

Python Version

3.11.6

PRAW Version

7.8.1

Links, references, and/or additional comments?

No response

[foolo]

Another note, in strict mode ("python.analysis.typeCheckingMode": "strict" in .vscode/settings.json) there are some warnings about missing stub files for praw.

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

Remove the Stale label or comment or this will be closed in 30 days.

[foolo]

This issue is stale because it has been open for 30 days with no activity.

Remove the Stale label or comment or this will be closed in 30 days.

Still a relevant issue

[bboe]

@foolo will you make a pull request that updates the code to have the following:

**generator_kwargs: str | int | dict[str, str] | None,

We don't guarantee pyright will pass, so this change may result in cascading changes.

[foolo]

@foolo will you make a pull request that updates the code to have the following:

**generator_kwargs: str | int | dict[str, str] | None,

We don't guarantee pyright will pass, so this change may result in cascading changes.

Thanks for replying! Though I don't really know how to test it all and fix any cascading changes it may cause (or have time to figure it out).
I mean, I could just blindly change that one line without testing anything, but I assume that's not what your'e asking for :)

[bboe]

I mean, I could just blindly change that one line without testing anything, but I assume that's not what your'e asking for :)

Yeah, you're right that I'm not quite asking for that. While we can fix this one error to make pyright happy in this particular case, it's likely to cause a cascading issue, which may or may not be visible to end-users. As a result, pyright presently remains in an unsupported state. I happen to be doing some work to address type checking related issues as reported by ruff, but I am not optimistic that I'll be able to address all the pyright issues in this package.

[foolo]

I mean, I could just blindly change that one line without testing anything, but I assume that's not what your'e asking for :)

Yeah, you're right that I'm not quite asking for that. While we can fix this one error to make pyright happy in this particular case, it's likely to cause a cascading issue, which may or may not be visible to end-users. As a result, pyright presently remains in an unsupported state. I happen to be doing some work to address type checking related issues as reported by ruff, but I am not optimistic that I'll be able to address all the pyright issues in this package.

Fully understand. This is not a blocker in any way. Just need to add # type: ignore here and there and the problem is "gone" :) I just felt that I should report it in case you want to fix it.

[arjunprakash027]

Is this issue still relevant?
If yes, I can pick up and fix any cascading changes

[alythobani]

Is this issue still relevant?

I believe it is yeah. Still seems to be no issues on the dev side with adding a # type: ignore to it though