Map localized namespace prefixes to namespace integer codes for all (public) wikimedia projects.

namespace.map.py

import requests
import json
import csv

headers = {
    'User-Agent': 'Bot Dan Andreescu parsing Special:PrefixIndex pages',
    'From': 'dandreescu@wikimedia.org'
}


def get_wikis():
    matrix = requests.get(
        'https://www.mediawiki.org/w/api.php?action=sitematrix'
        '&smsiteprop=url|dbname|code'
        '&smstate=all'
        '&format=json').json().get('sitematrix', {})

    wikis = [

        wiki

        for language in matrix.values()
        if type(language) is dict and 'site' in language

        for wiki in language['site']
    ] + [

        wiki

        for wiki in matrix.get('specials', [])
    ]

    return [
        wiki
        for wiki in wikis
        if 'private' not in wiki and 'closed' not in wiki
    ]


wikis = get_wikis()
namespaceDictionary = {}
query = '/w/api.php?action=query&format=json&meta=siteinfo&siprop=namespaces'

###
# Writes mapping as: (hostname, dbname, ns integer, ns canonical, ns localized)
# hostname      : ja.wikipedia.org
# dbname        : jawiki
# ns integer    : 2, 100, etc.
# ns canonical  : the english prefix if exists, otherwise the localized prefix
# ns localized  : the localized prefix
###
with open('namespace.dictionary.csv', 'wb') as w:
    spamwriter = csv.writer(w)

    for wiki in wikis:
        site = wiki.get('url', '')
        host = site.replace('https://', '')
        dbname = wiki.get('dbname', host)
        try:
            r = requests.get(site + query, headers=headers)
            ns = json.loads(r.text)['query']['namespaces']

            for k, v in ns.items():
                row = [host, dbname, k, v.get('canonical', ''), v.get('*', '')]
                spamwriter.writerow([unicode(s).encode("utf-8") for s in row])

        except Exception, e:
            print '*************************'
            print site + ' FAILED!!! with ' + str(e)
            print '*************************'
            continue