Sickchill: Episode not found in file re-scan

[TheRealNoob]

So this is a pretty simple bug. Episode not found in file scan due to the name of the episode. I have followed recommended naming convention.
File name: Marvel's Agents of S.H.I.E.L.D. - S01E02 - 0-8-4 - 720p WEB-DL.mkv

I suspect it is the use of the dashes that causes the issue. When I remove the episode name (0-8-4) from the file it is identified correctly. Logs below.

Branch/Commit: Master / e1af3a8af8dacccfb9c7adf45f08090a68875af2
OS: Windows-2012ServerR2-6.3.9600
What you did: Re-Scan Files on TV Show
What happened: Episode not found
What you expected: Episode to be found
Logs:

2017-05-07 11:01:45 INFO SHOWQUEUE-REFRESH :: 263365: No episode number found in D:\TV Shows\Marvel's Agents of S.H.I.E.L.D\Season 1\Marvel's Agents of S.H.I.E.L.D. - S01E02 - 0-8-4 - 720p WEB-DL.mkv, ignoring it
2017-05-07 11:01:45 INFO SHOWQUEUE-REFRESH :: 263365: parse_result: S01 [ABD: False] [ANIME: False] [whichReg: [u'season_only']]
2017-05-07 11:01:41 INFO SHOWQUEUE-REFRESH :: Performing refresh on Marvel's Agents of S.H.I.E.L.D.

[h3llrais3r]

It's most likely certainly related to the dashes in the episode name like you mentioned.
Due to the dashes, the parser is confused and nothing could be found...
Can I advise you another naming pattern which should not give any problems?
I use it for all my series and didn't have an issue yet:

[h3llrais3r]

I did some debugging and I found out that there is something in the code that causes the 'no match' result.
@miigotu @sharkykh
Any idea why this is done?
https://github.com/SickRage/SickRage/blob/develop/sickbeard/name_parser/parser.py#L149

Because the 'no_season' regex parses more than 4 extra_ep_num values, the matches are cleared and nothing could be found. I just tested it myself by putting the value to > 10 and now it parses with success.

What's the purpose of this check? Perhaps this can even be removed?

PS: using this naming pattern:

[sharkykh]

@h3llrais3r Yeah I asked @miigotu about this some time ago: https://github.com/SickRage/SickRage/issues/2925#issuecomment-280966290

[sharkykh]

I just tested it myself by putting the value to > 10 and now it parses with success.

@h3llrais3r I think SR parses it as extra episodes... then it's gonna add them to the database and maybe cause some issues?

[h3llrais3r]

Ok, I saw the reasons, but shouldn't it be better to just skip that regex than to clear the matches and break the loop?

EDIT: like this?

if 'extra_ep_num' in named_groups and match.group('extra_ep_num'):
    tmp_episodes = range(ep_num, self._convert_number(match.group('extra_ep_num')) + 1)
    if len(tmp_episodes) > 4:
        continue
    else:
        tmp_episodes = [ep_num]

[TheRealNoob]

Is it not simpler and more fool-proof to do a query like: *X(S|s)##(E|e)##X*
example: XS01e01X
Where X is the same character on both sides of S01e01. This seems better to me.

[h3llrais3r]

@TheRealNoob
It's not that simple... all regexes are used when parsing... and the best one is chosen.
See https://github.com/SickRage/SickRage/blob/develop/sickbeard/name_parser/regexes.py

Due to the scenario explained above, it clears the results an nothing can be found.

[TheRealNoob]

I'm not expecting that changes be made to accomodate a single episode, but I think it must be possible. Especially considering I am following the recommended naming convention.

[h3llrais3r]

Can you show me your settings than and also where you got those recommended naming convention?
IMHO it's never a good idea to put spaces and dashes in names, certainly not for media files.
That's why I always use the pattern with the dots.

EDIT: and again like I explained above, it should normally work, the reason why it isn't working is due to the regex that thinks it's more than 4 episodes.
I'll have to check with the other collaborators if we can find a way to fix this. (I suggested a fix above, but we still need to wait for feedback of @miigotu)

[TheRealNoob]

It's Sickrage's default naming convention, and Plex's recommended.
Season %S/%SN - S%0SE%0E - %EN - %QN

EDIT: I should point out if not already obvious, the episode name in question is "0-8-4".

[miigotu]

@h3llrais3r add that show, add a file for that episode, and try your suggested changes. There is absolutely a better way to fix the >10 episodes problem, the code was just added to prevent mass wrong parses added to the db, so it definitely can be improved.

[xpedicion]

I just switched to SickRage and when I imported my library, I ran into the same problem with this episode, as well. But let me add to the frenzy.

2017-05-09 19:19:52 INFO SHOWQUEUE-REFRESH :: 78804: No episode number found in /share/Multimedia/Serien/Doctor Who (2005)/Season 00/Doctor Who (2005) - S00E61-E65 - Pond Life - SDTV.avi, ignoring it
2017-05-09 19:19:52 INFO SHOWQUEUE-REFRESH :: 78804: parse_result: S00 [ABD: False] [ANIME: False] [whichReg: [u'season_only']]
2017-05-09 19:19:47 INFO SHOWQUEUE-REFRESH :: Performing refresh on Doctor Who (2005)

I figured that there was a 4 episode limit. I tried a few different names, and it would catch it (incorrectly of course but correctly according to the name) when named for example

Doctor Who (2005) - S00E61E62E63E64 - Pond Life - SDTV.avi

but not when having a 5th episode added in any way.
Here's another:

72227: No episode number found in /share/Multimedia/Serien/Two and a Half Men (2003)/Season 11/Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv, ignoring it

[miigotu]

The seldom instance that it does not correctly match one file, is MUCh better than it incorrectly matching one like Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv and creating 900 episodes in the show.

[xpedicion]

Well, no doubt about that! Still, wouldn't it be feasible to distinguish between <whitespace>-<whitespace> and <alphanum>-<alphanum> or something? That is, for splitting up the string or something before regexing season/episode numbering.

[h3llrais3r]

@h3llrais3r add that show, add a file for that episode, and try your suggested changes. There is absolutely a better way to fix the >10 episodes problem, the code was just added to prevent mass wrong parses added to the db, so it definitely can be improved.

@miigotu I have tested the proposed change locally and it works for that show. So basicly we are skipping that specific regex match, not all matches...

So we only remove https://github.com/SickRage/SickRage/blob/develop/sickbeard/name_parser/parser.py#L150

Postprocessing log:

Processing in folder C:\Temp\postprocessing
Processing C:\Temp\postprocessing
Processing folder C:\Temp\postprocessing
Processing C:\Temp\postprocessing\Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv (None)
Retrieving episode object for 1x2
Snatch history had a quality in it, using that: 720p WEB-DL
Is ep a priority download: False
File C:\Temp\completed\Marvel's Agents of S.H.I.E.L.D\Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv doesn't exist so there's no worries about replacing it
No associated files for C:\Temp\completed\Marvel's Agents of S.H.I.E.L.D\Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv were found during this pass
Using original release name Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv
Destination folder for this episode: C:\Temp\completed\Marvel's Agents of S.H.I.E.L.D
Moving file from C:\Temp\postprocessing\Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv to C:\Temp\completed\Marvel's Agents of S.H.I.E.L.D\Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv
Processing succeeded for C:\Temp\postprocessing\Marvel's Agents of S.H.I.E.L.D - S01E02 - 0-8-4 - 720p WEB-DL.mkv
Processing C:\Temp\postprocessing\Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv (None)
Retrieving episode object for 11x21
Looking up quality for name postprocessing, got Unknown
Looking up quality for name Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv, got Unknown
Guessing quality for name Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv, got Unknown
Is ep a priority download: False
There is no existing file so there's no worries about replacing it
Using original release name Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv
Destination folder for this episode: C:\Temp\completed\Two and a Half Men
Moving file from C:\Temp\postprocessing\Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv to C:\Temp\completed\Two and a Half Men\Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - Unknown.mkv
Processing succeeded for C:\Temp\postprocessing\Two and a Half Men - S11E21 - Dial 1-900-Mix-a-Lot - 720p HDTV.mkv
Successfully processed

As you can see only 1 episode is added for both shows 😄

[sharkykh]

@h3llrais3r Don't you also need to change break to continue?

[miigotu]

if len(tmp_episodes) > 4:
    continue

[TheRealNoob]

I updated to the latest master. It did find my episode this time:
Marvel's Agents of S.H.I.E.L.D. - S01E02 - 0-8-4 - 720p WEB-DL.mkv

Thank you all for the great work you do :)
-TheRealNoob