After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 351667 - doesn't fetch album art correctly for non-ascii artists/albums
doesn't fetch album art correctly for non-ascii artists/albums
Status: RESOLVED OBSOLETE
Product: rhythmbox
Classification: Other
Component: Plugins (other)
0.9.5
Other All
: Normal normal
: ---
Assigned To: RhythmBox Maintainers
RhythmBox Maintainers
Depends on:
Blocks: 412853
 
 
Reported: 2006-08-16 20:15 UTC by Bill Nottingham
Modified: 2010-01-03 09:20 UTC
See Also:
GNOME target: ---
GNOME version: ---


Description Bill Nottingham 2006-08-16 20:15:24 UTC 
Please describe the problem:
At least, all of my tracks that have non-ascii chars in their name never show album art. This includes:

Queensrÿche
Béla Fleck and the Flecktones (the non-accented version works)
Shakira (album ¿Dónde Están los Ladrones?)

So, my only assumption is that there's an encoding issue somewhere in the stack.

Steps to reproduce:
1. 
2. 
3. 


Actual results:


Expected results:


Does this happen every time?


Other information:
Comment 1 Sven Arvidsson 2006-11-20 17:04:17 UTC 
Isn't this a duplicate of bug 347818?
Comment 2 Bill Nottingham 2006-11-20 19:07:09 UTC 
Possibly. The original bug there was about it getting the wrong cover... it appears to have expanded in scope since I filed this one.
Comment 3 Martin Szulecki 2007-01-15 17:18:22 UTC 
This is failing due to the Amazon search query behaviour. Surprisingly Amazon seems to optimize their search by using strippped down keywords which appears to also be locale independent (doesn't matter if you search on .com, .jp, .br, etc.). 

Example: "Salomé" vs "Salome", gives correct results for the non-accented version and similar.

By "converting" everything to non-accented characters the album search seems to return correct covers.

This was marked as a TODO in: 
http://svn.gnome.org/viewcvs/rhythmbox/trunk/plugins/artdisplay/artdisplay/AmazonCoverArtSearch.py?view=markup

Anyone knows an efficient way to shift the string characters automatically to non-accented characters? (Instead of just using .replace, some regex or lookup table)
Comment 4 Sven Arvidsson 2007-01-18 16:19:33 UTC 
Banshee seems to use a different method for fetching covers. They look up the ASIN (Amazon Standard Identification Number) by way of MusicBrainz and uses that to find the correct cover.

I'm not sure if this is feasible for Rhythmbox, but it does seem like a less error prone way.

Could be interesting to compare the methods of other applications which uses Amazon for cover art, like Amarok.
Comment 5 Alex Lancaster 2007-02-22 05:41:14 UTC 
(In reply to comment #4)
> Banshee seems to use a different method for fetching covers. They look up the
> ASIN (Amazon Standard Identification Number) by way of MusicBrainz and uses
> that to find the correct cover.

> I'm not sure if this is feasible for Rhythmbox, but it does seem like a less
> error prone way.

Support for this should be relatively straightforward to add:  bug #410684.
Comment 6 Jonathan Matthew 2010-01-03 09:20:54 UTC 
Amazon cover art search is no more, and the sites we use now handle non-ascii data correctly.