Revision: 8390 Author: multichill Date: 2010-08-08 15:59:04 +0000 (Sun, 08 Aug 2010)

Log Message: ----------- Now parsing {{Information}} with standard pywikipedia template parser functions. Should be much more flexible. Location field is now parsed and put at the start of the description field (same way as it's used at enwp).

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

Modified: trunk/pywikipedia/imagecopy_self.py =================================================================== --- trunk/pywikipedia/imagecopy_self.py 2010-08-08 14:58:29 UTC (rev 8389) +++ trunk/pywikipedia/imagecopy_self.py 2010-08-08 15:59:04 UTC (rev 8390) @@ -324,7 +324,7 @@ Build a new description based on the imagepage ''' if u'{{Information' in imagepage.get() or u'{{information' in imagepage.get(): - (description, date, source, author) = self.getNewFieldsFromInformation(imagepage) + (description, date, source, author, permission, other_versions) = self.getNewFieldsFromInformation(imagepage) else: (description, date, source, author) = self.getNewFieldsFromFreetext(imagepage)

@@ -336,44 +336,69 @@ ''' Try to extract fields from the current information template for the new information template. ''' + + fields = [u'location', u'description', u'source', u'date', u'author', u'permission', u'other versions'] + description = u'' + source = u'' date = u'' - source = u'' author = u'' permission = u'' other_versions = u'' - text = imagepage.get() - # Need to add the permission field - # Need to use pywikipedia template parser code - regexes =[u'{{Information[\s\r\n]*|[\s\r\n]*description[\s\r\n]*=(?P<description>.*)|[\s\r\n]*source[\s\r\n]*=(?P<source>.*)|[\s\r\n]*date[\s\r\n]*=(?P<date>.*)|[\s\r\n]*author[\s\r\n]*=(?P<author>.*)|[\s\r\n]*permission.*=(?P<permission>[^}]*)|[\s\r\n]*other_versions.*=(?P<other_versions>[^}]*)}}', - u'{{Information[\s\r\n]*|[\s\r\n]*description[\s\r\n]*=(?P<description>.*)|[\s\r\n]*source[\s\r\n]*=(?P<source>.*)|[\s\r\n]*date[\s\r\n]*=(?P<date>.*)|[\s\r\n]*author[\s\r\n]*=(?P<author>.*)|[\s\r\n]*other_versions.*=(?P<other_versions>[^}]*)}}', - ] - - for regex in regexes: - match =re.search(regex, text, re.IGNORECASE|re.DOTALL) - if match: - description = self.convertLinks(match.group(u'description').strip(), imagepage.site()) - - date = match.group(u'date').strip() - if date == u'': - date = self.getUploadDate(imagepage)

- source = self.getSource(imagepage, source=self.convertLinks(match.group(u'source').strip(), imagepage.site())) + contents = {}

- author = self.convertLinks(match.group(u'author').strip(), imagepage.site()) - if author == u'': - author = self.getAuthorText(imagepage) - - if u'permission' in match.groupdict(): - permission = self.convertLinks(match.group(u'permission').strip(), imagepage.site()) - if u'other_versions' in match.groupdict(): - other_versions = self.convertLinks(match.group(u'other_versions').strip(), imagepage.site()) - # Return the stuff we found - return (description, date, source, author) - - #We didn't find anything, return the empty strings - return (description, date, source, author) + for field in fields: + contents[field]=u''

+ templates = imagepage.templatesWithParams() + + for (template, params) in templates: + if template==u'Information': + for param in params: + #Split at = + (field, sep, value) = param.partition(u'=') + #To lowercase, remove underscores and strip of spaces + field = field.lower().replace(u'_', u' ').strip() + #See if first part is in fields list + if field in fields: + #Ok, field is good, store it. + contents[field] = value.strip() + + # We now got the contents from the old information template. Let's get the info for the new one + + # Description + if not contents[u'location']==u'': + description = self.convertLinks(contents[u'location']) + u'\n' + if not contents[u'description']==u'': + description = description + self.convertLinks(contents[u'description'], imagepage.site()) + + # Source + source = self.getSource(imagepage, source=self.convertLinks(contents[u'source'], imagepage.site())) + + # Date + if not contents[u'date']==u'': + date = contents[u'date'] + else: + date = self.getUploadDate(imagepage) + + # Author + if not (contents[u'author']==u'' or contents[u'author']==self.getAuthor(imagepage)): + author = self.convertLinks(contents[u'author'], imagepage.site()) + else: + author = self.getAuthorText(imagepage) + + # Permission + # Still have to filter out crap like "see below" or "yes" + if not contents[u'permission']==u'': + permission = self.convertLinks(contents[u'permission'], imagepage.site()) + + # Other_versions + if not contents[u'other versions']==u'': + other_versions = self.convertLinks(contents[u'other versions'], imagepage.site()) + + return (description, date, source, author, permission, other_versions) + def getNewFieldsFromFreetext(self, imagepage): ''' Try to extract fields from free text for the new information template.