[Rhythmbox-devel] Merging very similar genera (in Jamendo)

[Peter <rhythmbox-devel maubp freeserve co uk>]

I was just having a look at the latest Jamendo catalogue using the 
plugin with rhythmbox 0.10.1 and noticed their genera fields need a 
little tidying up.

I should probably give some feedback directly to Jamendo - my guess is 
contributors are stuffing as many "tags" into the genre field as 
possible to try raise their own music's profile. Certainly a controlled 
vocabulary and enforced capitalisation scheme would be less chaotic!

In fact, with almost 2000 genres, I was almost wishing for a "genre" 
button on the search box so I could restrict a search.

For an example of some specific issues, RB shows 13 tracks under 
"acoustique" and 10 under "Acoustique" (different cases). Similarly for 
"Alternative", and there are three capitalisations for "Alternative 
Rock" etc.

Its not just case: There are five variants of "Rock", including "Rock " 
(with trailing space), "ROCK" and "rock".

I expect I could find some examples using accents too (given the high 
number of French speakers on Jamendo).

I can see at least two ways to deal with this:

(1) Do some clean up in the genre browser for all sources (see bug 
144283, which also covers the artist and album)

http://bugzilla.gnome.org/show_bug.cgi?id=144283

(2) Do some clean up in the Jamendo plugin when parsing the catalogue
i.e. Strip any leading/trailing spaces, and perhaps apply "Title Case" 
(lower case everything, except the first letter of each word).

Peter