Missing genres from releases dump

Hi,
It looks like the genres are no longer making it into the release dumps. I’m using the json dumps at http://ftp.musicbrainz.org/pub/musicbrainz/data/json-dumps/ and some automated tests recently started failing because of the missing data. When I grep the untarred release.tar.xz file, it shows this for every release:
“genres”:[]

So the key and array seem to be included, just not the genre values. Naturally, not all releases have genres defined, but the dump doesn’t currently include genres for any releases. I’m not sure exactly when it changed, but it was within the last week or two.

Many thanks for all your efforts,
Mike

Are you sure?

I have just downloaded and imported the newest dumps (21.05.2022) in my local docker and I get 4 genres for this (shortened and adjusted) example query:
https://musicbrainz.org/ws/2/release/0455b4c9-ba6e-4542-9351-15ff112bc458?inc=genres&fmt=json

image

1 Like

User-provided data such as tags and genres are available under a different license from the regular Musicbrainz data, and thus a separate file (mbdump-derived.tar.bz2). Did you download and import that file too?

1 Like

I’m sure. You appear to be using a different dump. The latest JSON dump available is from the 18th, not the 21st:
http://ftp.musicbrainz.org/pub/musicbrainz/data/json-dumps/20220518-001002/release.tar.xz

I’ve checked the release dump for the 18th and the 14th and neither have the genre data available. No idea why there’s no JSON dump for the 21st, but maybe they’re generated after the fullexport dumps?

Thanks

The genre data was available in the release json dump until recently. I can add support for processing another file if necessary, but the fact that the genres key is present but empty in the JSON release dump makes me think that the intent was to include it.

Thanks

Did you get any follow up for this? I am also having the same issue.

The recent schema change accidentally deleted tag/genre data from our mirror server which generates the JSON dumps (MBS-12508). I restored that data & am generating new dumps now.

P.S. Sorry for not seeing this earlier. The best way to report an issue is through https://tickets.metabrainz.org since tickets there go directly to developers’ mailboxes. :slight_smile:

2 Likes