Episodes over 99 [Absolute Ordering]
#1
prefacing this with things I have tried -
I have enabled absolute ordering in TVDB scraper settings
I have installed multiple regex including the one posted in the sticky on this section in the location C:\Users\(username)\AppData\Roaming\XBMC\userdata under the name "advancedsettings.xml"
renaming to S01EXXX works for anidb as its listed in 1 season for everything, but I'd rather not have to rename every tv show over 100 episodes

I am having trouble getting tv shows to scrape past episode 99 using absolute numbering
Lets take Dragon Ball Z for example. It has 291 episodes
all named as such
Dragon Ball Z - 001 - The new Threat
Dragon Ball Z - 002 - Reunions
etc
all the way to
Dragon Ball Z - 291 - Goku's Next Journey

It scrapes fine until it gets to episode 100
using TVDB I get episodes 1-99 fully scraped
using AniDB I get episodes 1-99 fully scraped along with another 99 episodes which are 100+ renamed to episodes 1-99 giving me 198 scraped episodes
neither are accurate nor account for 291 episodes

If I were to use the anidb scraper and rename all the episodes like follows it will work
Dragon Ball Z - S01E291 - Goku's Next Journey
However this looks weird to me and I'd rather not have to rename every TV show that has more than 99 episodes in absolute numbering


I've searched all day and I cannot find any way to fix this. I've tested countless regex for absolute ordering but none of them give me episodes higher than a count of 99.

One of the simpler regex I've tried is
<tvshowmatching action="append">
<regexp>[\._ \-]([0-9]+)()([\._\ \-][^\\/]*)</regexp>
<regexp>[\._ \- ]()([0-9]+)(-[0-9]+)?</regexp>
<regexp>[\._ \- ]([0-9]*)([0-9][0-9][0-9])([\._ \-][^\\/]*)</regexp>
<regexp>[/\._ \- ]()([0-9]+)([0-9]+)(-[0-9]+)?</regexp>
</tvshowmatching>
and many 1 line regexp as well as whole downloadable text files
found here http://forum.xbmc.org/showthread.php?t=51614
and one other large one that I cannot remember where I got it.


Debug Log: ANIDB scrape - http://pastebin.com/GzZs5mVn
Debug Log: TVDB scrape - (Had to cut into 2 since larger than 500 kb
Part 1 - http://pastebin.com/DgASwYA5
Part 2 - http://pastebin.com/KtuBk1rR

The best resource I found regarding what I'm looking for came from this
http://forum.xbmc.org/showthread.php?tid=74324
which said to look here
http://forum.xbmc.org/showthread.php?pid...3pid533763

Quote:Where you able to scan episode 100 and above for Naruto? For me they get scanned as

#101 - Episode 1
#102 - Episode 2


Hi

I used to get similar problems (#117 scanned as #17, etc). I made my own regexp for anime:

[\._ \-]([0-9]{2,3})()([\._\ \-][^\\/]*)

Or slightly simpler:

[\._ \-]([0-9]+)()([\._\ \-][^\\/]*)

The default regexp will not work for absolute numbering so you really need to use this one, or something along these lines.

I tried using his regexp, but it did not work for me.
edit: The above is incorrect. It does indeed work, but with errors where there were numbers in the episode name. The reason it did not work was user error. I didn't contain the tvshowmatching with <advancedsettings></advancedsettings>
Reply
#2
Not sure, but subscribing to this thread for the answer. I've been thinking of making an anime-specific guide for the wiki, and this is a situation that people will undoubtedly run into who have a large anime library.
Reply
#3
Void this post. It contained incorrect data due to user error.
Reply
#4
What I've found so far -
1. The Regex's most people use do not support the naming scheme 001, 002, etc and when you find one that does it can lead to duplicates via incorrect scraping. The below regex works.
Quote:<tvshowmatching>
<regexp>[\._ \-][Ss]([0-9]+)[\.\-]?[Ee]([0-9]+)([^\\/]*)</regexp>
<regexp>[\._ \-]\[?([0-9]+)x([0-9]+)\]?([^\\/]*)</regexp>
<regexp>[\._ \-]([0-9]{2,3})()([\._\ \-][^\\/]*)</regexp>
</tvshowmatching>

2. Locking your series after fixing the duplicates from scraping can fix issues of getting more duplicates from updating your library

3. If your anime have no duplicate errors you wont have duplicate problems changing your regex back to your normal one after scraping the 001 series.

4. The easiest way to get your anime collection working with 001 appears to just be renaming everything to S01EXXX and not bothering with this mess I am trying to figure out =P

Still trying to figure out a regex useable for all situations.

Reply
#5
Okay I've tested against this specific advancedsettings.xml that is specifically for anime, however by default it does not work with 001 EVEN THOUGH IT HAS THE CORRECT REGEX INSIDE OF IT. I'm guessing that since a regex above it finds the Season 0 shit out of 001 the correct regex never gets run. To fix this I moved the regex in front of the problematic regex.

The problematic regex appears to be
<regexp>[\._ \-]([0-9]+)([0-9][0-9])([\._ \-][^\\/]*)</regexp> <!-- foo.103 -->
This makes sense being that it deals with shows with 3 digits the first digit being a season normally.
So, now that this is cleared up we have to be sure never to use that format for non-anime tv shows.

As soon as I put the 001 regex above this one, I was able to scrape without issues, no duplicates, no weird season 0, etc

The excerpt
Quote:<!-- If nothing else is matched, then it's a normal anime episode -->
<regexp>[\._ \-]()([0-9]{2,3})([\&\-][0-9]{2,3})?[v\._ \-\[\(]</regexp>
was moved from the last line to where it is here
http://pastebin.com/nJWkmCLn

After much work, this appears to do everything I need it to do. If anyone has any comments or questions feel free. I don't remember where I got what is included in that paste bin nor do I claim ownership. It simply works for me thus far =D.

Currently, it doesn't appear that we can have one regex to rule them all if you have a show with absolute numbering > 99 episodes.

One thing that would help quite a bit is if we could run regex specific to whether absolute numbering is on. I remember reading this while doing my research as something someone else proposed.
There is prepend and append, why not something like isAbsoluteTrue. This would be a feature request I am guessing, but it would definitely help people in my situation as few as that may be.
Reply

Logout Mark Read Team Forum Stats Members Help
Episodes over 99 [Absolute Ordering]0