[Pywikipedia-l] SVN: [4835] trunk/pywikipedia/replace.py

wikipedian at svn.wikimedia.org wikipedian at svn.wikimedia.org
Wed Jan 9 17:04:10 UTC 2008


Revision: 4835
Author:   wikipedian
Date:     2008-01-09 17:04:10 +0000 (Wed, 09 Jan 2008)

Log Message:
-----------
re-enabled preloading for XML parsing, with a lower page number though

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

Modified: trunk/pywikipedia/replace.py
===================================================================
--- trunk/pywikipedia/replace.py	2008-01-09 17:00:47 UTC (rev 4834)
+++ trunk/pywikipedia/replace.py	2008-01-09 17:04:10 UTC (rev 4835)
@@ -589,10 +589,10 @@
     if namespaces != []:
         gen = pagegenerators.NamespaceFilterPageGenerator(gen, namespaces)
     if xmlFilename:
-        # XML parsing is slow enough that preloading would make bot even slower
-        preloadingGen = gen
+        # XML parsing can be quite slow, so we preload less pages each time.
+        preloadingGen = pagegenerators.PreloadingGenerator(gen, pageNumber = 20)
     else:
-        preloadingGen = pagegenerators.PreloadingGenerator(gen, pageNumber = 50)
+        preloadingGen = pagegenerators.PreloadingGenerator(gen, pageNumber = 60)
     bot = ReplaceRobot(preloadingGen, replacements, exceptions, acceptall, allowoverlap, recursive)
     bot.run()
 





More information about the Pywikipedia-l mailing list