Matching regex returned as UNDECIDED

[ghost]

Are there some default configs that make Flexget ignore results not featuring certain criteria? (ex: lack of proper quality keyword in title)
[NC-Raws] 转生成为魔剑 / Tensei shitara Ken Deshita - 01 (B-Global 3840x2160 HEVC AAC MKV)

Manual execute shows the entry is left undecided

pi@raspberrypi:~ $ flexget execute -v --tasks "seasonal-misc" | grep "Tensei shitara Ken Deshita"
2022-10-01 19:51:20 VERBOSE  verbose       seasonal-misc   UNDECIDED: `[NC-Raws] 转生成为魔剑 / Tensei shitara Ken Deshita - 01 (B-Global 3840x2160 HEVC AAC MKV)`

although the line in the config matches when trying regex interactively

-------------------------------------------------------------------------------
Hi there, welcome to try regexps in realtime!
Press ^D or type 'exit' to continue. Type 'continue' to continue non-interactive execution.
Task 'seasonal-misc' has 211 entries, enter regexp to see what matches it.
--> Tensei shitara Ken Deshita
Title: [NC-Raws] 转生成为魔剑 / Tensei shitara Ken Deshita - 01 (B-Global 3840x2160 HEVC AAC MKV) URL:
https://nyaa.si/download/1584037.torrent From: original_title
1 of 211 entries matched
--> exit
Bye!

My portion of the configs featuring the task

  seasonal-misc:
    template: seasonal
    inputs:
      - rss: https://nyaa.si/?page=rss&q=1080p+-batch&c=0_0&f=0&u=Erai-raws
      - rss: https://nyaa.si/?page=rss&q=1080p+-batch&c=0_0&f=0&u=gst
      - rss: https://nyaa.si/?page=rss&c=1_2&f=0&u=BraveSail
    regexp:
      reject:
        - \[v[2-9]\]
        - \d{1,2}v[2-9]
      from: title
    series:
      - Tensei shitara Ken Deshita
      - Isekai Ojisan:
          identified_by: sequence

[BrutuZ]

The entry is not being recognized because of the kanji between the group name and romaji, there's no built-in logic for that.
Either of those new lines fix it, pick your preference:

    series:
      - Tensei shitara Ken Deshita:
+         name_regexp: ['Tensei shitara Ken Deshita']
+         alternate_name: ['转生成为魔剑 / Tensei shitara Ken Deshita']

[BrutuZ]

Because series runs before if (see https://flexget.com/Plugins/plugin_priority).

You can either move the manipulate to the task root (it's redundant anyways since the expression itself already limits it) or reorder the plugins execution making series run after if (whether you raise the priority of if or lower series is up to you).

tasks:
  test:
    mock:
      - {title: '[NC-Raws] 死神 千年血战篇 / Bleach: Sennen Kessen-hen - 11 (B-Global 3840x2160 HEVC AAC MKV)'}
    if:
      - "'[NC-Raws]' in title":
          manipulate:
            - title:
                replace:
                  regexp: '^(\[NC-Raws\]) .*(?=[฀-﫿]).+? \/ '
                  format: '\1 '
            - title:
                replace:
                  regexp: ':'
                  format: ''
          set:
            content_filename: "{{ title }}"
    series:
      2160p:
        - Bleach Sennen Kessen-hen
    plugin_priority:    # removing these will go back to Undecided
      series: 75

[BrutuZ]

Probably worth noting: You don't have to remove semicolons from the title, just enclose the title in quotes

series:
  2160p:
    - 'Bleach: Sennen Kessen-hen'

[ghost]

Thank you for the explanation. I was trying to rearrange the order of plugins as well but I didn't know about plugin_priority.

(it's redundant anyways since the expression itself already limits it)

The reason why I added an IF statement is because I'm also using set to change the content_filename. I wish to leave that alone if regex doesn't match.

You don't have to remove semicolons from the title, just enclose the title in quotes

Since I'm using content_filename to also rename the output file, I want to make sure no special characters like \ / : * ? " < > | are left behind

[BrutuZ]

Illegal characters shouldn't be a problem because of pathscrub plugin, but if you want to be sure there's also the option of sanitizing it in the jinja itself with the filter {{ title|pathscrub('windows') }}

[ghost]

I just tested without the second manipulate and the output file is named like this. It appears a random unicode character is used in place of the illegal character

[NC-Raws] Bleach Sennen Kessen-hen - 11 (B-Global 3840x2160 HEVC AAC MKV).mkv

While looking in the documentation I found other jinja templates and I solved the issue with something like

set:
  content_filename: "{{ title|re_replace(':', '') }}"