Kodi Community Forum

(2017-02-23, 23:04)thezoggy Wrote: [ -> ]I'm not posting release names (like you are)

Touché. This is his mother. My son's been grounded for a week.

Will have a look once I have time, right now I'm working on KCleaner addon To submit strings use web site: http://cleansub.heliohost.org/

v5.1
- Bug fix

Sorry for bug. Seems no languages except "generic" were used if set to use "kodi subtitle language settings", my bad, should be fixed.

(2017-02-23, 00:18)thezoggy Wrote: [ -> ]Lang[i] = Lang[i].upper()
IndexError: list index out of range
-->End of Python script error report<--

Don't know why it did this. Strange. Hopefully it won't happen again.

(2017-02-23, 00:18)thezoggy Wrote: [ -> ]this went from:

Code:

00:10:30,230 --> 00:10:34,265
but [Groans] this --

to: (double space after but)

Code:

00:10:30,230 --> 00:10:34,265
but  this --
this is too much.

OK, will fix this in next release.

(2017-02-23, 00:18)thezoggy Wrote: [ -> ]then with the CC cleanup, its not cleaning up the 'music' stuff like:

Code:

00:14:38,544 --> 00:14:41,646
¶¶

Don't know if I should fix this.... it is not advertizing etc...

(2017-02-23, 00:18)thezoggy Wrote: [ -> ]then finally, looks like you need to add cleanup string for rarbg:

Code:

00:18:02,838 --> 00:18:04,838
Torrent downloaded by RARBG

Not a problem to add to definitions. Will be in next release.

(2017-02-23, 21:49)HeresJohnny Wrote: [ -> ]3. The cleaning itself seems to fail partially. Example:
American.Horror.Story.S02E12.en.srt vs. American.Horror.Story.S02E12.en.srt_ORIGINAL

Cleansubs manages to clean the last lines of the sub which are

Code:

728
00:42:42,598 --> 00:42:52,817
<font color="#ec14bd">Sync & corrections by honeybunny</font>
<font color="#ec14bd">www.addic7ed.com</font>

However, it fails to clean stuff from the top which still has

Code:

1
00:00:48,917 --> 00:00:51,152
Daddy?

2
00:00:51,220 --> 00:00:53,788
Daddy'll be there in a minute.

3
00:01:48,608 --> 00:01:58,632
<font color="#ec14bd">Sync & corrections by honeybunny</font>
<font color="#ec14bd">www.addic7ed.com</font>

I'll be back with some more tests about NFS.

The only way I can explain this is if you have only heuristics turned on and definitions off. Then by default heur will clean only first 2 lines and last 2 lines.

v5.2
- Clean double spaces after cleaning hearing impaired
- Added more codepages (hebrew, arabic, japanese)

The addon fails on K18 "Leia" with the following message (not from debug log):

Damn, and I just installed Krypton

Will have a look....

v5.3
- Bug fix

It was a bug not only in Lea... I was concentrated on making the "auto language selection" work and messed up the other part Should be fixed anyway.

Thumbs up, Dalai Lanik :-)
CleanSubs just had a great run, cleaning a third of my subs until it errorred out. I think you're very close to perfection. Here's what I can glean from the normal log, will run some debug tests later to identify the subs which caused this.

Here goes. This run, the script got much futher which leads me to believe that it might be a memory problem. There are no real indicators, though, except for that stats hint.

(2017-03-14, 11:11)HeresJohnny Wrote: [ -> ]Thumbs up, Dalai Lanik :-)
CleanSubs just had a great run, cleaning a third of my subs until it errorred out. I think you're very close to perfection. Here's what I can glean from the normal log, will run some debug tests later to identify the subs which caused this.

Code:

10:05:56.839 T:1652   ERROR: EXCEPTION Thrown (PythonToCppException) : -->Python callback/script returned the following error<--
                                             - NOTE: IGNORING THIS CAN LEAD TO MEMORY LEAKS!
                                            Error Type: <type 'exceptions.IndexError'>
                                            Error Contents: string index out of range
                                            Traceback (most recent call last):
                                              File "C:\Users\JoScha\AppData\Roaming\Kodi\addons\service.cleansubs\standalone.py", line 290, in <module>
                                                intCancel = scanPaths(manFolder, 1, 1, 3)
                                              File "C:\Users\JoScha\AppData\Roaming\Kodi\addons\service.cleansubs\standalone.py", line 176, in scanPaths
                                                process_subs(os.path.join(path, basePath, name), 1)
                                              File "C:\Users\JoScha\AppData\Roaming\Kodi\addons\service.cleansubs\default.py", line 414, in process_subs
                                                if not line.text[subsPos - 1].isalnum() or not line.text[subsPos + 1].isalnum():
                                            IndexError: string index out of range
                                            -->End of Python script error report<--

OK, I understand this error, shouldn't be a problem to fix... there simply isn't character - or + 1.... back to the drawing board

v5.4
- Bug fix
- Slightly optimized logic of reading definitions
- Progress bar during reading of definitions

So close... with the latest version, Cleansubs parsed half of my folders before it encountered an error

Will have a deeper look at a debug log.

UPDATE: The problem here is probably a combination of things. According to the debug log, Cleansub opened the problematic file with CP1256 (arabic) whereas it really is CP932 (Japanese Shift-JIS). Of, course, there would be only garbage to analyze.

Every time I start up Cleansubs it displays on top of the screen a countup in percent till 899. What is that? (cleansubs I see on top then)