Revision: 7967 Author: xqt Date: 2010-02-26 16:20:35 +0000 (Fri, 26 Feb 2010)

Log Message: ----------- enable logpagesPageGenerator for special:log entries

Modified Paths: -------------- trunk/pywikipedia/interwiki.py trunk/pywikipedia/pagegenerators.py trunk/pywikipedia/wikipedia.py

Modified: trunk/pywikipedia/interwiki.py =================================================================== --- trunk/pywikipedia/interwiki.py 2010-02-26 09:20:14 UTC (rev 7966) +++ trunk/pywikipedia/interwiki.py 2010-02-26 16:20:35 UTC (rev 7967) @@ -190,8 +190,9 @@ (note: without ending colon)

-initialredirect work on its target if a redirect or category redirect is - entered on the command line or by a generator. - (note: without ending colon) + entered on the command line or by a generator (note: without + ending colon). It is recommended to use this option with + -movelog pagegenerator.

-neverlink: used as -neverlink:xx where xx is a language code: Disregard any links found to language xx. You can also

Modified: trunk/pywikipedia/pagegenerators.py =================================================================== --- trunk/pywikipedia/pagegenerators.py 2010-02-26 09:20:14 UTC (rev 7966) +++ trunk/pywikipedia/pagegenerators.py 2010-02-26 16:20:35 UTC (rev 7967) @@ -115,6 +115,19 @@ delimited with ";" Example: -usercontribs:DumZiBoT;500 returns 500 distinct pages to work on. + +-<mode>log Work on articles that were on a specified special:log. + You have options for every type of logs given by the + <mode> parameter which could be one of the following: + block, protect, rights, delete, upload, move, import, + patrol, merge, suppress, review, stable, gblblock, + renameuser, globalauth, gblrights, abusefilter, newusers + Examples: + -movelog gives 500 pages from move log (should be redirects) + -deletelog:10 gives 10 pages from deletion log + -protect:Dummy gives 500 pages from protect by user Dummy + -patrol:Dummy;20 gives 20 pages patroled by user Dummy + In some cases this must be written as -patrol:"Dummy;20"

-weblink Work on all articles that contain an external link to a given URL; may be given as "-weblink:url" @@ -286,6 +299,12 @@ for page in site.prefixindex(prefix = title, namespace = namespace, includeredirects = includeredirects): yield page

+def LogpagesPageGenerator(number = 500, mode='', user=None, repeat = False, site = None, namespace=[]): + if site is None: + site = pywikibot.getSite() + for page in site.logpages(number=number, mode=mode, user=user, repeat=repeat, namespace=namespace): + yield page[0] + def NewpagesPageGenerator(number = 100, get_redirect = False, repeat = False, site = None, namespace = 0): if site is None: site = pywikibot.getSite() @@ -1145,7 +1164,7 @@ limit = arg[11:] or pywikibot.input( u'How many images do you want to load?') gen = NewimagesPageGenerator(number = int(limit)) - elif arg.startswith('-new'): + elif arg == ('-new') or arg.startswith('-new:'): if len(arg) >=5: gen = NewpagesPageGenerator(number = int(arg[5:])) else: @@ -1175,7 +1194,25 @@ elif arg.startswith('-yahoo'): gen = YahooSearchPageGenerator(arg[7:]) else: - pass + mode, log, user = arg.partition('log') + if log == 'log': + number = 500 + if not user: + user = None + else: + try: + number = int(user[1:]) + user = None + except ValueError: + user = user[1:] + if user: + result = user.split(';') + user = result[0] + try: + number = int(result[1]) + except: + pass + gen = LogpagesPageGenerator(number, mode[1:], user) if gen: self.gens.append(gen) return self.getCombinedGenerator()

Modified: trunk/pywikipedia/wikipedia.py =================================================================== --- trunk/pywikipedia/wikipedia.py 2010-02-26 09:20:14 UTC (rev 7966) +++ trunk/pywikipedia/wikipedia.py 2010-02-26 16:20:35 UTC (rev 7967) @@ -5928,6 +5928,59 @@ yield page, '', '', '', '', ''

# TODO: avoid code duplication for the following methods + + def logpages(self, number=50, mode='', user=None, repeat=False, namespace=[], offset=-1): + if config.use_api: + apiURL = self.api_address() + del apiURL + else: + raise NotImplementedError + if mode not in ('block', 'protect', 'rights', 'delete', 'upload', + 'move', 'import', 'patrol', 'merge', 'suppress', + 'review', 'stable', 'gblblock', 'renameuser', + 'globalauth', 'gblrights', 'abusefilter', 'newusers'): + raise NotImplementedError + params = { + 'action' : 'query', + 'list' : 'logevents', + 'letype' : mode, + 'lelimit' : int(number), + } + + if number > config.special_page_limit: + params['lelimit'] = config.special_page_limit + if number > 5000 and self.site().isAllowed('apihighlimits'): + params['lelimit'] = 5000 + if user: + params['leuser'] = user + nbresults = 0 + while True: + result = query.GetData(params, self) + if 'error' in result or 'warnings' in result: + output('%s' % result) + raise Error + for c in result['query']['logevents']: + if not namespace or c['ns'] in namespace: + yield (Page(self, c['title'], defaultNamespace=c['ns']), + c['user'], + parsetime2stamp(c['timestamp']), + c['comment'], + ) + + nbresults += 1 + if nbresults >= number: + break + if 'query-continue' in result and nbresults < number: + params['ucstart'] = result['query-continue']['logevents']['ucstart'] + elif repeat: + nbresults = 0 + try: + params.pop('ucstart') + except KeyError: + pass + else: + break + return def newpages(self, number = 10, get_redirect = False, repeat = False, namespace = 0): """Yield new articles (as Page objects) from Special:Newpages.