Some hyphen from MusicBrainz are the wrong hyphen

[jones1008]

Problem

In some releases/release groups/artists/recordings (etc....) there is the following hyphen -> Unicode: (U+2010)

‐

but it should be the following hyphen -> Unicode: (U+002D)

-

example: Third Eye Blind - Semi-Charmed Life

This can lead to some problems with plugins when searching in other databases than MusicBrainz.

possible solution

I know that this is a MusicBrainz database mistake but it could be replaced within the beets code before writing a tag to an item.
maybe like this:
tag = tag.replace("‐", "-")

But since I don't know where to add this code I am asking here...

[sampsyo]

Hello! If you think data is wrong in MusicBrainz, that's the right place to fix it—you can read up on MB's style guidelines and make edits there. (But I think you may find that they actually recommend dashes, not hyphens, in cases where that's typographically correct. Reading the Chicago Manual of Style might also help. 😃)

If you prefer ASCII hyphens over proper dashes, you can change that in your down database with beet modify. Or see #488 for a proposal that would help with automatically replacing these characters in tags.

[lbesnard]

@sampsyo I would like to comment on this, but happy to lodge a new issue but as it seems to be the same, I decided to add it here as it could also help someone else like I.

I'm importing this artist named "-M-" . However, to cd to the artist folder, I have to do the following
cd `echo -e '\u2010'm'\u2010'

Definitely not practical.

I tried to go around the issue by adding the following lines in my beet config (see http://unicode.scarfboy.com/?s=%E2%80%90):

replace:                                                                                                                                                                   
    b'\xe2\x80\x90': '-'                                                                      
    u'\u2010': '-' 

However after doing a beet move, i get

beet move albumartist:"-M-"
Moving 0 items (11 already in place).

I'm pretty sure my unicode is probably not written the proper way

[wisp3rwind]

replace:                                                                                                                                                                   
    b'\xe2\x80\x90': '-'                                                                      
    u'\u2010': '-' 

I'm pretty sure my unicode is probably not written the proper way

I don't think you should put those b and u prefixes there: The configuration file is YAML after all, not a Python source file, cf. the documentation.

[lbesnard]

thanks, for the record the following fixed it (at least one of the three...):

replace:
    '[\xe2\x80\x90]': '-'
    '[\u2010]': '-'
    '‐': '-'