Re: automatically adding artists' subscriptions via python-musicbrainz2 or musicbrainzngs [solved]

Am 19.03.2015 um 15:45 schrieb Maximilian Bräutigam:
> Hi Alastair,
> 
> thanks a lot for your answer. I was really afraid that the solution will
> be so bloated since it means a lot of traffic and requests on the mb
> database. Anyway, I do not see a better solution and just wanted to get
> a second opinion on this.
> 
> Thank you very much again!
> Best wishes,
> Max
> 
> Am 19.03.2015 um 15:35 schrieb Alastair Porter:
>> Hi Max,
>>
>> You're right, the API doesn't support subscribing to entities, so we
>> don't expose that in musicbrainzngs.
>>
>> Subscribing is just a matter of visiting a specific url: 
>> https://musicbrainz.org/account/subscriptions/artist/add?id=859749
>> while you are logged in. If you're comfortable with python-requests or
>> scrapy you could do this yourself.
>>
>> Unfortunately the biggest problem is that these urls take row ids
>> instead of musicbrainz ids. We spoke about it a bit
>> (http://chatlogs.musicbrainz.org/musicbrainz-devel/2015/2015-03/2015-03-19.html#T14-20-26-909477)
>> and the easiest way is probably to load the artist page, find the
>> subscribe link, and click on it automatically.
>> Meanwhile, we've filed some issues
>> (http://tickets.musicbrainz.org/browse/MBS-8305) to make this easier in
>> the future.
>>
>> Alastair
>>
>> On Thu, Mar 19, 2015 at 3:00 PM, Maximilian Bräutigam <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>>     Dear all,
>>
>>     I generated a huge list (python pickled) of artists' IDs to which I want
>>     to subscribe. Unfortunately, I cannot find the entry to do this.
>>     Obviously, with python-musicbrainzngs I can add them to my collection,
>>     but I want to subscribe to them.
>>
>>     Any comments, tips, and remarks are highly appreciated.
>>
>>     Best wishes,
>>     Max

Just for completeness, here is what I did:

$ ./artistid2rownumber.py > rows
$ wget --save-cookies cookies.txt --keep-session-cookies \
  --post-data 'username=eew4Weap&password=W4fkaVfj' \
  https://musicbrainz.org/login
$ ./subscribe.sh

It took about 1 hour or so for 1800 subscriptions.

Best wishes,
Max

_______________________________________________
MusicBrainz-devel mailing list
[email protected]
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-devel

#!/bin/bash

# get in the row numbers somehow:
list=`sed 's/.*;//' rows`

for i in $list
do
    # the cookie with the login info is crucial:
    wget --load-cookies cookies.txt "http://musicbrainz.org/account/subscriptions/artist/add?id=${i}"
    # musicbrainz accepts 45 requests per 20 s, just to be on the safe side:
    sleep 0.5s
done

#!/usr/bin/env python3

import csv
import urllib.request
import re

# get in the musicbrainz artist ids somehow:
d = csv.reader(open("artists.csv", "r"), delimiter=";")

# artist_id in the site's code is the row in the database that is needed:
reg = re.compile("artist_id=(\d*)")

for i in d:
    try:
        # get the artist's page; i[1] is the musicbrainz artist id
        f = urllib.request.urlopen("http://musicbrainz.org/artist/"+i[1])
        s = f.read().decode('utf-8')
        m = reg.search(s)
        # simply print out artist's name and row in the mb db
        print(i[0]+";"+m.group(1))
    except:
        raise("Error on: "+i)