Make sure subprocess calls don`t ignore result codes indicating errors. Fix JS docs generation whil…

sitescripts/subscriptions/bin/updateSubscriptionDownloads.py

Index: sitescripts/subscriptions/bin/updateSubscriptionDownloads.py
===================================================================
--- a/sitescripts/subscriptions/bin/updateSubscriptionDownloads.py
+++ b/sitescripts/subscriptions/bin/updateSubscriptionDownloads.py
@@ -30,18 +30,18 @@ if __name__ == '__main__':
 
   sourceTemp = {}
   destTemp = None
   try:
     destTemp = tempfile.mkdtemp()
     for repoName, repoDir in sourceRepos.iteritems():
       tempDir = tempfile.mkdtemp()
       sourceTemp[repoName] = tempDir
-      subprocess.Popen(['hg', 'archive', '-R', repoDir, '-r', 'default', tempDir]).communicate()
-    subprocess.Popen(['rsync', '-a', '--delete', destDir + '/', destTemp]).communicate()
+      subprocess.check_call(['hg', 'archive', '-R', repoDir, '-r', 'default', tempDir])
+    subprocess.check_call(['rsync', '-a', '--delete', destDir + '/', destTemp])

[Sebastian Noack, 2013/07/04 13:57:51]

There is no need to hard-code path separators. Use os.path.sep instead (you
might also want to have a look at os.path.join).

[Wladimir Palant, 2013/07/05 11:24:33]

os.path.join() won't help here - rsync requires that the source path ends with a
slash, otherwise it won't work correctly. But - yes, using os.path.sep for that
is probably cleaner.

     combineSubscriptions(sourceTemp, destTemp)
-    subprocess.Popen(['rsync', '-au', '--delete', destTemp + '/', destDir]).communicate()
+    subprocess.check_call(['rsync', '-au', '--delete', destTemp + '/', destDir])

[Sebastian Noack, 2013/07/04 13:57:51]

See above.

   finally:
     for tempDir in sourceTemp.itervalues():
       if os.path.exists(tempDir):
         shutil.rmtree(tempDir, True)
     if destTemp and os.path.exists(destTemp):
       shutil.rmtree(destTemp, True)