Kodi Community Forum
[RELEASE] Texture Cache Maintenance utility - Printable Version

+- Kodi Community Forum (https://forum.kodi.tv)
+-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33)
+--- Forum: Supplementary Tools for Kodi (https://forum.kodi.tv/forumdisplay.php?fid=116)
+--- Thread: [RELEASE] Texture Cache Maintenance utility (/showthread.php?tid=158373)



RE: [RELEASE] Texture Cache Maintenance utility - redglory - 2014-08-01

Can you post complete log and what you're trying to run?


RE: [RELEASE] Texture Cache Maintenance utility - un1versal - 2014-08-01

@Milhouse

The user nistsio is reporting issues with duplicates, I suggested he uses script form clenaing up and sorting his library http://forum.xbmc.org/showthread.php?tid=200919 so these errors are unrelated to his reported issue of dupes, since nistsio is also new user personally Ill give him some latitude which I think he may overcome when gets more familiar with XBMC and other tools, though not reading/undertsanding things seems to be in the way.


RE: [RELEASE] Texture Cache Maintenance utility - nistsio - 2014-08-01

@redglory
I am at work at the moment but I will try to upload the whole debug log asap.
As for what I am trying to run is update and clean library.

@uNiversal
One thing is surely true:
I cannot understand why for the errors indicating http mishaps (with tmdb.org) your suggestion is to clean my library. (Huh)


RE: [RELEASE] Texture Cache Maintenance utility - Milhouse - 2014-08-01

(2014-08-01, 15:44)uNiversal Wrote: @Milhouse

The user nistsio is reporting issues with duplicates, I suggested he uses script form clenaing up and sorting his library http://forum.xbmc.org/showthread.php?tid=200919 so these errors are unrelated to his reported issue of dupes, since nistsio is also new user personally Ill give him some latitude which I think he may overcome when gets more familiar with XBMC and other tools, though not reading/undertsanding things seems to be in the way.

Well, that background info would have been useful first time around. Smile

Sure, the script will help clean up his cache
Code:
./texturecache.py Xd
./texturecache.py R
./texturecache.py P

but it can't fix communication problems or scraper meta-data errors which he also seems to be getting.

If a scraper is creating bogus entries, it should be possible to view those entries with "./texturecache.py jd movies <moviename>". It might be useful to use this option on the movie library to ensure their is only one version of the movie, not two, although I'm not sure why Search would show two versions but the library only one, unless the database is corrupted somehow (duplicate links)?

Another way to identify duplicates is:
Code:
./texturecache.py duplicates

which will match duplicate movies based on imdbnumber.

At best, texturecache.py can only give an indication that the media library - from a JSON perspective - is correct, or not. If there is some internal MySQL query being executed by Search that returns duplicates then JSON - runing a different query - may not reveal this issue.


RE: [RELEASE] Texture Cache Maintenance utility - puntloos - 2014-08-04

TL;DR: texturecache.py c does not download artist thumbnails. What gives?

Newbie experiencing problems:

- XBMC Gotham from scratch (.xbmc dir removed when upgrading)
- Scanned my (very well tagged) media library into system
- texturecache.py c

It 'does a lot of stuff'.. but:

Code:
OpenELEC:~ #  ./texturecache.py s Cure
Matching row ids:        
OpenELEC:~ #  ./texturecache.py s Adele
008631|3/3cf61b3c.jpg|0720|0720|0004|2014-08-04 18:04:54|                   |http://assets.fanart.tv/fanart/music/cc2c9c3c-b7bc-4b8b-84d8-4fbd8779e493/artistthumb/adele-4fb28f0853f8d.jpg
008632|f/f454c288.jpg|1080|1920|0003|2014-08-04 18:04:54|                   |http://assets.fanart.tv/fanart/music/cc2c9c3c-b7bc-4b8b-84d8-4fbd8779e493/artistbackground/adele-4dc840394f5f4.jpg
Matching row ids: 8631 8632

As you can see Adele (which I manually played) has been cached, but the Cure - which I havent played yet, is not.

What gives? Am I misunderstanding? I assumed that at least an artist thumbnail would be fetched for every artist in my DB


RE: [RELEASE] Texture Cache Maintenance utility - Milhouse - 2014-08-04

Hard to say without knowing what artwork urls (if any) have been scraped for The Cure in your media library.

Try "./texturecache.py jd artists cure" to see what thumbnail and/or fanart urls have been scraped for this artist, then you should be able to search the cache using the correct substring(s).

Chances are this artist may not have any associated artwork, or there may be a problem accessing this artwork (upload a logfile: ./texturecache.py c artists cure @logfile=/tmp/tc.log).

However if the artwork is embedded (image urls start with music://), then texturecache.py can't pre-load the cache with this artwork - you need to browse the GUI for this artwork to be extracted and cached.


RE: [RELEASE] Texture Cache Maintenance utility - puntloos - 2014-08-04

(2014-08-04, 20:53)Milhouse Wrote: Hard to say without knowing what artwork urls (if any) have been scraped for The Cure in your media library.

Try "./texturecache.py jd artists cure" to see what thumbnail and/or fanart urls have been scraped for this artist, then you should be able to search the cache using the correct substring(s).

Code:
OpenELEC:~ # ./texturecache.py jd artists cure
[                          
  {
    "artist": "The Cure",
    "artistid": 668,
    "fanart": "",
    "label": "The Cure",
    "thumbnail": ""
  }
]
Quote:Chances are this artist may not have any associated artwork, or there may be a problem accessing this artwork (upload a logfile: ./texturecache.py c artists cure @logfile=/tmp/tc.log).
Code:
OpenELEC:~ # ./texturecache.py c artists cure @logfile=/tmp/tc.log
                            
Cache pre-load activity summary for "artists", filtered by "cure":

              |    fanart   |  thumbnail  |    TOTAL    
--------------+-------------+-------------+-------------
Cached        |      -      |      -      |      0      
Deleted       |      -      |      -      |      0      
Duplicate     |      -      |      -      |      0      
Error         |      -      |      -      |      0      
Ignored       |      -      |      -      |      0      
Skipped       |      -      |      -      |      0      
Undefined     |      1      |      1      |      2      
========================================================
TOTAL         |      1      |      1      |      2      

       Loading: 00:00:00.05
       Parsing: 00:00:00.00
     Comparing: 00:00:00.00
TOTAL RUNTIME: 00:00:00.05

With the logfile here:
http://sprunge.us/XjQg

Quote:However if the artwork is embedded (image urls start with music://), then texturecache.py can't pre-load the cache with this artwork - you need to browse the GUI for this artwork to be extracted and cached.
I don't think it's embedded.


RE: [RELEASE] Texture Cache Maintenance utility - Milhouse - 2014-08-04

You don't have any artwork for this artist, it's a scraper or tagging problem.


RE: [RELEASE] Texture Cache Maintenance utility - puntloos - 2014-08-04

Small extra note, album art seems to be present in some cases now, but also definitely not all, even with very famous artists that should have been filled by the fans/online DBs.. Surely John Coltrane's CD Art should be there?

(2014-08-04, 21:06)Milhouse Wrote: You don't have any artwork for this artist, it's a scraper or tagging problem.

Can you be a bit more specific? The tags are good as far as I know, very clean, many fields filled. How do I check if/when/how the scraper failed?

[edit]Or even more specifically: Im pretty much positive the files are tagged very properly, so why is The Cure data not fetched? It's a very very popular band so certainly the art will be available[/edit]


RE: [RELEASE] Texture Cache Maintenance utility - Milhouse - 2014-08-04

(2014-08-04, 21:07)puntloos Wrote: Can you be a bit more specific? The tags are good as far as I know, very clean, many fields filled. How do I check if/when/how the scraper failed?

[edit]Or even more specifically: Im pretty much positive the files are tagged very properly, so why is The Cure data not fetched? It's a very very popular band so certainly the art will be available[/edit]

I think you may be confused about the purpose of texturecache.py.

texturecache.py will pre-load your texture cache using the artwork urls that are present in your media library (it will also allow you to view the contents of your media library, which is handy for diagnosing problems).

It is the job of a scraper (eg. Universal Album Scraper, Universal Artist Scraper, maybe others) to create the relevant music metadata (ie. song, album and artists entries) in your media library and to also assign appropriate artwork urls.

In your case, although the artists have been created, the scraper has failed to find any artwork (local or remote) so the fanart and thumbnail urls are empty for some of your artists. As such, this script - texturecache.py - has no artwork with which to pre-load the texture cache for these artists. It can't pre-load what it doesn't have, and it's not the job of this script to find that missing information (texturecache.py is not a scraper).

What you need to do is ask in the relevant scraper support thread why the scraper is not finding artwork for your artists.


RE: [RELEASE] Texture Cache Maintenance utility - puntloos - 2014-08-04

Great, that makes sense. Thanks Milhouse. (pity your script can't call universal artist scraper to do the work for empty artists Smile )

(2014-08-04, 21:39)Milhouse Wrote:
(2014-08-04, 21:07)puntloos Wrote: Can you be a bit more specific? The tags are good as far as I know, very clean, many fields filled. How do I check if/when/how the scraper failed?

[edit]Or even more specifically: Im pretty much positive the files are tagged very properly, so why is The Cure data not fetched? It's a very very popular band so certainly the art will be available[/edit]

I think you may be confused about the purpose of texturecache.py.

texturecache.py will pre-load your texture cache using the artwork urls that are present in your media library (it will also allow you to view the contents of your media library, which is handy for diagnosing problems).

It is the job of a scraper (eg. Universal Album Scraper, Universal Artist Scraper, maybe others) to create the relevant music metadata (ie. song, album and artists entries) in your media library and to also assign appropriate artwork urls.

In your case, although the artists have been created, the scraper has failed to find any artwork (local or remote) so the fanart and thumbnail urls are empty for some of your artists. As such, this script - texturecache.py - has no artwork with which to pre-load the texture cache for these artists. It can't pre-load what it doesn't have, and it's not the job of this script to find that missing information (texturecache.py is not a scraper).

What you need to do is ask in the relevant scraper support thread why the scraper is not finding artwork for your artists.



RE: [RELEASE] Texture Cache Maintenance utility - jeinnor - 2014-08-15

Im looking forward to trying this, but just wanted to know since there are 70 pages if these commands are all I am looking for?


./texturecache.py c movies
./texturecache.py c sets
./texturecache.py c tvshows

My sets do not showup at all
If I highlight movies, it still shows the normal background and not the random movies like I set in the config of the Skin.
Same goes for TV Series?


RE: [RELEASE] Texture Cache Maintenance utility - Milhouse - 2014-08-15

Yes on the commands. Check you have sets with "./texturecache jd sets", if you get no results then your sets are not defined (in Confluence you'd also have a Sets option below the Movies menu).

I can't really help on your skin problem - ask in the relevant skin thread.


RE: [RELEASE] Texture Cache Maintenance utility - pgjensen - 2014-08-16

Is there a way to update the Top 250 for movies via the imdb flag?

i.e. texturecache.py imdb movies @imdb.fields=top250

i scanned the py code but didn't see any references to it


RE: [RELEASE] Texture Cache Maintenance utility - Milhouse - 2014-08-16

No, unfortunately the top250 field isn't yet supported by omdb.com.