#musicbrainz

no need to write a parser when the browser already has one

not when you need to convert that back to moin, or you want to be able to show post-edit HTML interpretation without submission

ie, offline-preview

hi, is there a way to get the release year for a release? I'm using the perl WebService::MusicBrainz

there's a date node for each label node, you can parse just the first 4 characters from there

for each release node, I mean

hmm

text_rep_script, release_event_list, track_list, score, artist, id, type, title, asin, text_rep_language, disc_list

these are all the fields I get

toss a release-rels on your request

sorry if I'm daft, but how do I do that?

np

and actually, I think you want labels, not release-rels, let me check

release_event_list

is something in the list?

I get a hash back, these are its keys

there should be dates

I could take a look at what's in the release_event_list

along with other info, like country, catalog#

luks, release_event_list?

brianfreud: he is not using the WS directly

ah

but yes, once you're getting that, they're under release - release event list - event, attribute date

ah, thanks :) that's some deep digging there :)

and submitting data? not via the website, but for example through Perl or in any other programmatic manner. Is it possible and is there (good) documentation about it?

would have been cool... we're processing hundreds, if not thousands of CDs -- there are a bunch of them only on CDDB, some nowhere

and we don't want MB to end up like CDDB or FreeDB :)

for example you have to match artists to the right artist entities

not just submit artist names

so you couldn't just automatically add albums anyway

Yeah, good point -- though here at least the moderation system helps keeping things tidy

Ralesk: with the current balance, i think we're more interested in ppl moderating than ppl _only_ adding new data into it.

gotcha

luks, quick question

in /<\/em>/ and /<em>/ is there some wierd syntax I'm accidentally tripping on without realizing?

I don't think so

hmmm

have a very wierd bug

if, and only if, I have <strong><em>WORDS</strong> WORDS</em>, the </strong> in the middle, instead of turning only into ''', turns into ''''''''

it trips replace(/<\/strong>/g, "'''a"); which it ought to, but then it also trips replace(/<em>/g, "''c"); and replace(/<\/em>/g, "''d"); for '''b''cBold''d'''a''c Italics''d '''b''c

too many 's for me to parse the result :)

since it's only ever that one tag, half tempted to just replace all '{8} with '{3}, but would rather figure out quite why it's happening

lol

French capitalisation standards make me sad.

the tag in any other situation is fine, but just in that one case the match is really wierd

s/<em>/<strong>/ before you do it. :P

that's just the text from the test suite I threw together many months ago - parsing '''s is turning out to be a pain :P

ah, I see...

(after scanning)

Text::wiki closes every strong and em, then immediately reopens it again if it shouldn't actually have been closed - keeps the tags well nested - FireFox was reinterpreting to ignore the em tags in foo<close em><close strong><open em>bar

http://www.uk.musicbrainz.org/ is back

:):)

eh?

Kind of sad... everything works so well now, it's better at HTML than text::wiki :P

my <li>s actually work :P

So, "saving" in Picard writes the tag information, yeah? Does Picard also keep its own database?

Yes, no.

So how's it remember "saved" metadata?

it doesn't?

It does

I'm in the process of tagging a pretty large collection, most of which already has metadata

Picard crashed, I started it back up and scanned the directory, and it pulled the albums that I'd already saved up in the right pane

and everything else went into "unclustered files" over on the left.

oh, yeah, it does that

koolatron: every track that has the musicbrainz track id saved in its tag gets to the right pane afaik

Ah, okay

so if you remove the musicbrainz ids from your tags, they should get to the left pane. If you really want that

I don't :)

Although, a good portion of my music collection does not as of now have a PUID

it kind of sucks to add the PUIDs with picard anyway

such a stupid task

What would you suggest I do it with?

but you need only do it once

Adding the PUIDs doesn't help *me*, just anyone else with that album :)

koolatron: do it with picard until there's something better doing it automatically perhaps

yeah, I did it too because I'm such a nice guy

I'm looking at Jaikoz right now, but i'm not sure it's better than picard

I even wrote me a little shell script to check for albums with no PUIDs

It's sluggish, but that's probably because I'm not doing this locally

never used Jaikoz

i tried Jaikoz ages ago, but after learning to use Picard, I couldnt be bother to learn a new tagging app, especially one i knew i wasnt going to use

It's not.. "bad"

if it's more stable than Picard, I'm down with it

I just want something that isn't going to require scads of babysitting

Ihi

if you're tagging files, you want to babysit, too much automation can result in thousands of incorrectly tagged files

and you're doing the tagging only once (updating later is easy)

That's true

I shouldn't have let my collection get to where it is

Actually, it's not BAD, I'm pretty fastidious about tagging

But the entire 30k+ track collection needs to be reconciled with MB

I guess it's stupid to think that such a thing would be fast or easy ;)

it takes time if you want it to be perfect, i.e. correct errors on MB

took me months for not even 10k tracks

Yeah, I'll put this in the "continuing projects" pile

anybody feel like assisting me with voting down a wrong edit ? :)

I have a release with John Lee Hooker called the "The Classic Savoy Sessions". its a 2CD. The first disc is previously released on another label and is in MB as

Should I enter a new release for the doubledisc anyway?

Tykling: Which edit is it?

and the two other edits (three total) are: http://musicbrainz.org/show/edit/?editid=8379483 and http://musicbrainz.org/show/edit/?editid=8379480

you should give the guy a chance to cancel the edits

Yes, thre additional tracks that don't belong right?

right

MClemo: Yes, that's what I thought also. 2 "No":s now is a pretty good indication something is wrong. Hopefully he/she will react

seems fair, thanks

Tykling: Maybe leave a note to the edit?

I did :)

on the first one I pasted he also replied

Aha... I'll check again...

Ah sorry, didn't scroll enough

Hmmm I'm inclined to give him a "No" since he could have been more polite in his reply.

thats what I thought

MClemo: I think he's got more than a chance reading the notes.