[Pywikipedia-l] SVN: [4692] trunk/pywikipedia/wikipedia.py

[btongminh at svn.wikimedia.org]

Revision: 4692
Author:   btongminh
Date:     2007-12-10 21:22:16 +0000 (Mon, 10 Dec 2007)

Log Message:
-----------
Fix incomplete gets. (bug 1831698)

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

Modified: trunk/pywikipedia/wikipedia.py
===================================================================
--- trunk/pywikipedia/wikipedia.py	2007-12-10 17:29:45 UTC (rev 4691)
+++ trunk/pywikipedia/wikipedia.py	2007-12-10 21:22:16 UTC (rev 4692)
@@ -3806,16 +3806,7 @@
                 return self.getUrl(path, retry, sysop, data, compress)
 
             text = response.read()
-            contentType = response.getheader('Content-Type')
-            contentEncoding = response.getheader('Content-Encoding')
-            
-            # Ensure that all sent data is received
-            if int(response.getheader('Content-Length', '0')) != len(text):
-                output(u'Warning! len(text) does not match content-length: %s != %s' % \
-                    (len(text), response.getheader('Content-Length', '0')))
-                self.conn.close()
-                self.conn.connect()
-                return self.getUrl(path, retry, sysop, data, compress)
+            headers = dict(response.getheaders())
                 
         else:
             if self.hostname() in config.authenticate.keys():
@@ -3867,11 +3858,20 @@
                     else:
                         raise
             text = f.read()
+            
+            headers = f.info()
+            
+        contentType = headers.get('content-type', '')
+        contentEncoding = headers.get('content-encoding', '')
+            
+        # Ensure that all sent data is received
+        if int(headers.get('content-length', '0')) != len(text) and 'content-length' in headers:
+                output(u'Warning! len(text) does not match content-length: %s != %s' % \
+                    (len(text), headers.get('content-length')))
+                self.conn.close()
+                self.conn.connect()
+                return self.getUrl(path, retry, sysop, data, compress)
 
-            # Find charset in the content-type meta tag
-            contentType = f.info()['Content-Type']
-            contentEncoding = f.headers.get('Content-Encoding')
-
         if compress and contentEncoding == 'gzip':
             # Use cStringIO if available
             # TODO: rewrite gzip.py such that it supports unseekable fileobjects.