http://www.mediawiki.org/wiki/Special:Code/pywikipedia/10452

Revision: 10452 Author: xqt Date: 2012-07-09 13:18:08 +0000 (Mon, 09 Jul 2012) Log Message: ----------- enable fixArabicLetters() without fixing digits, fix for fa pipe andlinkstrails moved to cleanUpLinks()

Modified Paths: -------------- trunk/pywikipedia/cosmetic_changes.py

Modified: trunk/pywikipedia/cosmetic_changes.py =================================================================== --- trunk/pywikipedia/cosmetic_changes.py 2012-07-09 13:06:27 UTC (rev 10451) +++ trunk/pywikipedia/cosmetic_changes.py 2012-07-09 13:18:08 UTC (rev 10452) @@ -178,8 +178,8 @@ text = self.fixReferences(text) text = self.fixStyle(text) text = self.fixTypo(text) -## if self.site.lang in ['ckb', 'fa']: -## text = self.fixArabicLetters(text) + if self.site.lang in ['ckb', 'fa']: + text = self.fixArabicLetters(text) try: text = isbn.hyphenateIsbnNumbers(text) except isbn.InvalidIsbnException, error: @@ -515,6 +515,13 @@ text = pywikibot.replaceExcept(text, linkR, handleOneLink, ['comment', 'math', 'nowiki', 'pre', 'startspace']) + ##fixing pipe and trailing for fa. Thanks ZxxZxxZ + if self.site.lang == 'fa': + faChrs = u'ءاآأإئؤبپتثجچحخدذرزژسشصضطظعغفقکگلمنوهیةيك' + u'ًٌٍَُِّْٓٔ' + text = re.sub(u'[[([^]|]*)]]([‌%s]+)' % faChrs, + ur'[[\1|\1\2]]', text) + text = re.sub(u'[[([^]|]*)|(.+?)]]([‌%s]+)' % faChrs, + ur'[[\1|\2\3]]', text) return text

def resolveHtmlEntities(self, text): @@ -739,7 +746,9 @@ text = pywikibot.replaceExcept(text, u'ه', u'ھ', exceptions) text = pywikibot.replaceExcept(text, u'ك', u'ک', exceptions) text = pywikibot.replaceExcept(text, ur'[ىي]', u'ی', exceptions) + return text # replace persian/arabic digits + ## deactivated due to bug #3539407 for i in xrange(0, 10): text = pywikibot.replaceExcept(text, old[i], new[i], exceptions) # do not change digits in class, style and table params @@ -752,11 +761,6 @@ # replace digits for i in xrange(0, 10): text = pywikibot.replaceExcept(text, str(i), new[i], exceptions) - ##fixing pipe and trailing for fa. Thanks ZxxZxxZ - if self.site.lang=='fa': - faChrs = u'ءاآأإئؤبپتثجچحخدذرزژسشصضطظعغفقکگلمنوهیةيك' + u'ًٌٍَُِّْٓٔ' - text = re.sub(u'[[([^]|]*)]]([‌%s]+)' % faChrs, ur'[[\1|\1\2]]', text) - text = re.sub(u'[[([^]|]*)|(.+?)]]([‌%s]+)' % faChrs, ur'[[\1|\2\3]]', text) return text

# Retrieved from "http://commons.wikimedia.org/wiki/Commons:Tools/pywiki_file_description_clea..."