[Pywikipedia-l] SVN: [5398] trunk/pywikipedia/wikipedia.py

[nicdumz at svn.wikimedia.org]

Revision: 5398
Author:   nicdumz
Date:     2008-05-18 14:51:39 +0000 (Sun, 18 May 2008)

Log Message:
-----------
Fixing #1966402 "wrong mediawiki_message when not using the default site lang" :
Not fetching anymore every single message through Mediawiki:message_name, fetching instead *once* Special:Allmessages. It is slower indeed, but I think that this is the only way to get the properly localized messages.

Any comments are welcome...

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

Modified: trunk/pywikipedia/wikipedia.py
===================================================================
--- trunk/pywikipedia/wikipedia.py	2008-05-18 12:12:24 UTC (rev 5397)
+++ trunk/pywikipedia/wikipedia.py	2008-05-18 14:51:39 UTC (rev 5398)
@@ -146,6 +146,7 @@
 except ValueError:
     WIDEBUILD = False
 
+
 # Local exceptions
 
 class Error(Exception):
@@ -4420,54 +4421,43 @@
 
     def mediawiki_message(self, key):
         """Return the MediaWiki message text for key "key" """
-        global mwpage, tree
-        if key.lower() not in self._mediawiki_messages.keys() \
-                and not hasattr(self, "_phploaded"):
-            get_throttle()
-            mwpage = self.getUrl("%s?title=%s:%s&action=edit"
-                     % (self.path(), urllib.quote(
-                            self.namespace(8).replace(' ', '_').encode(
-                                self.encoding())),
-                        key))
-            tree = BeautifulSoup(mwpage,
-                                 convertEntities=BeautifulSoup.HTML_ENTITIES,
-                                 parseOnlyThese=SoupStrainer("textarea"))
-            if tree.textarea is not None and tree.textarea.string is not None:
-                value = tree.textarea.string.strip()
-            else:
-                value = None
-            if value:
-                self._mediawiki_messages[key.lower()] = value
-            else:
-                self._mediawiki_messages[key.lower()] = None
-                # Fallback in case MediaWiki: page method doesn't work
-                if verbose:
-                    output(
-                      u"Retrieving mediawiki messages from Special:Allmessages")
-                retry_idle_time = 1
-                while True:
-                    get_throttle()
-                    phppage = self.getUrl(self.get_address("Special:Allmessages")
-                                      + "&ot=php")
-                    Rphpvals = re.compile(r"(?ms)'([^']*)' => '(.*?[^\\])',")
-                    count = 0
-                    for (phpkey, phpval) in Rphpvals.findall(phppage):
-                        count += 1
-                        self._mediawiki_messages[str(phpkey).lower()] = phpval
-                    if count == 0:
-                        # No messages could be added.
-                        # We assume that the server is down.
-                        # Wait some time, then try again.
-                        output('WARNING: No messages found in Special:Allmessages. Maybe the server is down. Retrying in %i minutes...' % retry_idle_time)
-                        time.sleep(retry_idle_time * 60)
-                        # Next time wait longer, but not longer than half an hour
-                        retry_idle_time *= 2
-                        if retry_idle_time > 30:
-                            retry_idle_time = 30
-                        continue
-                    break
-                self._phploaded = True
+        try:
+            from xml.etree.cElementTree import XML
+        except ImportError:
+            from cElementTree import XML
 
+        # Allmessages is retrieved once for all in a session
+        if not self._mediawiki_messages:
+            if verbose:
+                output(
+                  u"Retrieving mediawiki messages from Special:Allmessages")
+            retry_idle_time = 1
+            while True:
+                get_throttle()
+                xml = self.getUrl(self.get_address("Special:Allmessages") 
+                                    + "&ot=xml")
+                decode = xml.encode(self.encoding())
+                tree = XML(decode)
+                # xml structure is :
+                # <messages lang="fr">
+                #    <message name="about">À propos</message>
+                #    ...
+                # </messages>
+                for msg in tree.getiterator('message'):
+                    self._mediawiki_messages[msg.get('name').lower()] = msg.text
+                if not self._mediawiki_messages:
+                    # No messages could be added.
+                    # We assume that the server is down.
+                    # Wait some time, then try again.
+                    output('WARNING: No messages found in Special:Allmessages. Maybe the server is down. Retrying in %i minutes...' % retry_idle_time)
+                    time.sleep(retry_idle_time * 60)
+                    # Next time wait longer, but not longer than half an hour
+                    retry_idle_time *= 2
+                    if retry_idle_time > 30:
+                        retry_idle_time = 30
+                    continue
+                break
+
         key = key.lower()
         if self._mediawiki_messages[key] is None:
             raise KeyError("MediaWiki key '%s' does not exist on %s"