Fix 6784

[artpelling]

Description

Fixes #6784 .

I (and others) encountered an issue where playlist queries returned empty results. At first, I thought it was because I had playlist in extended m3u format. I found that the line skipping mechanism did not work correctly because the playlist file is read as bytestrings and the comparator used regular strings.

Fixing that did not help (the previously unskipped lines were simply not found but did not break the search) and I then found #6784. I think the issue there is that different path normalizations are used in beets ls and playlist.py.

To Do

• Documentation. (If you've added a new command-line flag, for example, find the appropriate page under docs/ to describe it.)
• Changelog. (Add an entry to docs/changelog.rst to the bottom of one of the lists near the top of the document.)
• Tests. (Very much encouraged but not strictly required.)

[github-actions]

Thank you for the PR! The changelog has not been updated, so here is a friendly reminder to check if you need to add an entry.

[artpelling]

I just found #6792 ... @snejus what was wrong with that? I think the fix there is probably cleaner than what I did.

[semohr]

I just found #6792 ... @snejus what was wrong with that? I think the fix there is probably cleaner than what I did.

The issue with that PR wasn't the fix itself. It was that the author used AI to generate and submit a large number of pull requests with little apparent human involvement. We're generally fine with using LLMs as development tools, but we don't want communication or contributions to be entirely AI-generated. Our policy is outlined here: https://github.com/beetbox/beets?tab=contributing-ov-file#notes-on-ai-usage.

[artpelling]

Thanks for the quick reply @semohr

The issue with that PR wasn't the fix itself.

Should I then modify this PR to incorporate the fix in #6792 ?

I am not sure what fix is preferred by main devs. I saw that there are plans to migrate to pathlib. Please let me know how to proceed here