Matcher Unhandled error due to folder structure about kyoo HOT 6 CLOSED

It seems Medusa does this too: https://github.com/pymedusa/Medusa/blob/master/medusa/name_parser/rules/rules.py#L256
but they only search for the "title - alternative_title" combo. I'll push a change to use that instead of the first element only. Please report back if it worked or if we need fallbacks to single "title" search.

from kyoo.

Seems to be good for all the examples I gave, thank you! I still have many import issues I need to investigate 😭

from kyoo.

I did not know guessit could return lists for titles. Thanks for the detailed report, I'll fix it right away.

from kyoo.

Thanks for your fast reply @zoriya

After applying your patch I can give you another example of an episode that was not correctly identified.

/media/series/Star Trek - Discovery/Season 4/Star Trek - Discovery - S04E03 - Choose to Live SDTV.mkv

2024-04-28 10:41:42 INFO:matcher.matcher:Identied /video/Star Trek - Discovery/Season 4/Star Trek - Discovery - S04E03 - Choose to Live SDTV.mkv: MatchesDict({'title': ['Star Trek', 'Discovery'], 'season': 4, 'episode_title': 'Star Trek', 'alternative_title': 'Discovery', 'episode': 3, 'source': 'TV', 'container': 'mkv', 'mimetype': 'video/x-matroska', 'type': 'episode'})
2024-04-28 10:41:42 ERROR:matcher.matcher:Could not find episode 3 of season 4 of serie 253 (absolute: 82)

Here keeping the first element results in searching Star Trek in tmdb which is show 253. But the actual show is Star Trek : Discovery id 67198

I'm not sure if it could break others setup but why not giving guessit only the filename instead of the entire path?

Jellyfin tests for reference: https://github.com/jellyfin/jellyfin/blob/master/tests/Jellyfin.Naming.Tests/TV/EpisodePathParserTest.cs

from kyoo.

Using the filename gives the same result. Guessit thinks the - means an alternative title, and kyoo does not search based on alternative titles.
We use thexem to identify some alternative titles of value. For example, here is a custom rule:

Example: "JoJo's Bizarre Adventure - Diamond is Unbreakable - 12.mkv"
Default:
```json
{
        "title": "JoJo's Bizarre Adventure",
        "alternative_title": "Diamond is Unbreakable",
        "episode": 12,
}
```
Expected:
```json
{
        "title": "JoJo's Bizarre Adventure - Diamond is Unbreakable",
        "episode": 12,
}
```

I think the correct fix would be to first search multiples times and taking the first hit. Using something like this list: ["title - alternative_title", "title", "alternative_title"]. This might have an impact on scan time tho.

from kyoo.

Please get back with them, for now some anime are badly passed (most part2 shows don't work). I know some others issues but the goal is to slowly fix them all.

For anime, I want to use anilist or something like that for better parsing but that's way harder than It should (look at issue #352 for more info).

from kyoo.