#musicbrainz

Argh

I'm so annoyed with the Discogs importer breaking classical releases *and* with users not checking half the time

Maybe I should take some time to try and fix the importer... has anyone tried?

hi, new to musicbrainz but is there any way to query the database and then download the result instead of doing it locally?

you can download a snapshot of the database from https://beta.musicbrainz.org/doc/MusicBrainz_Da... which you can import into a local db engine to run queries on, if you like

er, i guess that's what you mean by "doing it locally"

no, the musicbrainz folks will not run arbitrary sql queries on your behalf :)

(anything you can do via the apis you're welcome to do, of course)

yeah lol sorry just was wondering if it was hosted online anywhere, thank you for the help!!

em7272, what exactly are you trying to achieve? If you’re trying to import a subset of MBz data into a local database, there are a few different things you could do… but if there’s a particular problem you’re trying to solve or question to answer, there might be a variety of solutions.

there's a lot you can do with just the api https://beta.musicbrainz.org/doc/MusicBrainz_API and search api https://beta.musicbrainz.org/doc/MusicBrainz_AP...

not super well-versed with this so bear with me but i'd like to get just the artists and area columns from the basic database without the rest

that api search link looks promising thank you kepstin!

I'm not sure what you mean by "artist columns"

artists aren't a single column pretty much anywhere in musicbrainz - and artists of what? tracks, recordings, releases, release groups, etc.?

Be warned that “the basic database” isn’t… it’s a brilliantly complicated mess of join tables, aggressively normalized.

sorry literally just artist names

but artists have multiple names

and recordings, releases, etc can have multiple artists

you have to be more specific about what you want :)

a list of artist names with their associated area .. is that possible

i don't see any way to do that without building a custom query against the full db.

(note that our artist-area information isn't particularly complete)

Also bear in mind that there are just over 1.9 million artist records in the database.

technically that information is all in the search index together

but paging through all the results of the search api for all 1.9 million artists.. would take a while :)

Yup. Sounds like running a local copy of the database is the way to go.

I would just download the full thing, do some subsetting queries, export the results, and run with that. If you’re hurting for space, spin up a free-tier AWS instance to do the work.

👍

thank you so much for your help crism kepstin rdswift this is all very good to know

(if you go at max permitted rate and don't get blocked it would take about 5½ hours to pull this all from the api, heh)

yeah, but if you decide you want some extra information later you can amortize the database setup time

em7272, download information: https://musicbrainz.org/doc/MusicBrainz_Databas...

for your use case since you'd just want to do queries in the database, consider using the 'mbdata' tool (linked from the docs) rather than doing a full server setup.

alternatively, there's a docker setup nowadays that's simpler than the old days of manual setup :)

"but if you decide you want some extra information later you can amortize the database setup time"... Exactly the reason I would download the whole thing and query it locally.

just doublechecking then, i should follow the instructions here https://github.com/metabrainz/musicbrainz-docker (not setting up replication or live indexing)

I think that will set up the whole server. If all you want is the database, you might want to look at https://github.com/lalinsky/mbdata