I’m new here and have just started working with MusicBrainz in the last couple of weeks. I’m very interested in using MusicBrainz as a source to make sense of output of a speech recogniser in the case where the input is something like ‘play muse starlight’, without saying anything that would give us a clue as to which part of the input is artist and which is song title. On the website, doing an indexed-search on Recordings:
does not give any helpful output, because it is only searching against the recording name and not against artist. My problem is that the WS/2 query that I’d like to use here appears to default to the advanced type, so aligns with the 2nd query above and not the first. Can anyone advise on whether the web query syntax supports the first, ‘Indexed’, version which is rather more useful to me? I don’t want to have to web-scrape if I can avoid it! Many thanks, David James.
Thank you Ulrich for the speedy reply - yep, that is working for me! I’ll investigate some more tomorrow and will certainly take care in due course given what you say about this not being officially supported. I’ve been running my own copy of the server (via the preconfigured Linux image) so I figure I have more control than if I were using the main site. But for the moment, that is brilliant, many thanks!
Thank you Jesus. In fact, I need a query format that is agnostic as to which element(s) are part of the band name and which the song name - I want the results of the lookup to tell me that! So I don’t think that the queries you suggest are quite what I am looking for. Using dismax, I can search for ‘starlight+muse’ and ‘muse+starlight’ and the results are sufficiently identical that it effectively doesn’t matter which order they’re in. Which means that, from the point of view of my application (think Amazon Echo-like control of music playback) I can be very flexible in allowing people to say pretty much what they like.
Is your application (going to be) open source licensed? It definitely sounds interesting. Please keep us updated on the development! (And if you end up publishing something (FLOSS and/or commercially), it would also be nice if you’d join our list of supporters.)