gh-86943: implement pathlib.WindowsPath.is_mount()

[barneygale]

Have pathlib.WindowsPath.is_mount() call ntpath.ismount(). Previously it raised NotImplementedError unconditionally.

https://bugs.python.org/issue42777

• Issue: WindowsPath does not implement is_mount but ntpath implements and offers a ismount method #86943

Automerge-Triggered-By: GH:brettcannon

[zmievsa]

Will review soon

[barneygale]

Will review soon

Don't worry about it - I'm trying to focus my efforts on two more important PRs: #31691 and #91882. The branch has conflicts anyway.

[barneygale]

Not sure what's going on with the test failure - looks like ntpath.ismount('C:/\x00') is true in main? It raises an exception in my Windows VM with Python 3.10.

[eryksun]

Not sure what's going on with the test failure - looks like ntpath.ismount('C:/\x00') is true in main? It raises an exception in my Windows VM with Python 3.10.

The new implementation of normpath() strips the trailing null character. The old implementation keeps it.

[eryksun]

The new implementation of normpath() strips the trailing null character. The old implementation keeps it.

Do you think this needs to be fixed, or is it an improvement?

[barneygale]

Do you think this needs to be fixed, or is it an improvement?

I think it's unlikely to come up in real world usage. IMO it's one of the quirks noted in 99fcf15:

This commit attempts to preserve every previously tested quirk, but these may be changed in the future to better align platforms.