On Arch 'python' is 3, rely on shebang

git-svn-id: https://wikiteam.googlecode.com/svn/trunk@977 31edc4fc-5e31-b4c4-d58b-c8bc928bcb95

batchdownload/launcher.py

@@ -77,9 +77,9 @@ for wiki in wikis:
     # typically they don't provide any crawl-delay value in their robots.txt).
     if started and wikidir: #then resume
         print 'Resuming download, using directory', wikidir
-        subprocess.call('python dumpgenerator.py --api=%s --xml --images --resume --path=%s' % (wiki, wikidir), shell=True)
+        subprocess.call('./dumpgenerator.py --api=%s --xml --images --resume --path=%s' % (wiki, wikidir), shell=True)
     else: #download from scratch
-        subprocess.call('python dumpgenerator.py --api=%s --xml --images' % wiki, shell=True)
+        subprocess.call('./dumpgenerator.py --api=%s --xml --images' % wiki, shell=True)
         started = True
         #save wikidir now
         for dirname, dirnames, filenames in os.walk('.'):