Fix another source of utf8 characters being lost.

appengine/chromium_build/app.py

Index: appengine/chromium_build/app.py
diff --git a/appengine/chromium_build/app.py b/appengine/chromium_build/app.py
index a1cda78128ad812f33dec6322f626fe6cc41fe93..a81f5f74d23a9f991f403760f4865e22141c1028 100644
--- a/appengine/chromium_build/app.py
+++ b/appengine/chromium_build/app.py
@@ -298,8 +298,7 @@ class ConsoleData(object):
 
   @staticmethod
   def ContentsToHtml(contents):
-    return ''.join(unicode(content).encode('ascii', 'replace')
-                   for content in contents)
+    return ''.join(str(content).decode('utf-8') for content in contents)
 
   @property
   def last_row(self):
@@ -482,7 +481,7 @@ def console_merger(localpath, remoteurl, page_data,
 
   # Place merged data at |merged_tag|'s location in |merged_page|, and put the
   # result in |merged_content|.
-  merged_tag.replaceWith(str(merged_content))
+  merged_tag.replaceWith(merged_content)
   # .prettify() may damage the HTML but makes output nicer.  However, that
   # cost is a bunch of extra whitespace.  We reduce page size by not using
   # .prettify().
@@ -670,14 +669,14 @@ def get_position_number(commit_msg):
   return '0'
 
 
-def utf8_convert(bstring):
+def utf8_convert(bstag):

[jrobbins, 2015/02/12 17:55:31]

What does 'bstag' stand for?

[cmp, 2015/02/12 17:59:58]

BeautifulSoup tag.

   # cmp also investigated:
-  #   bstring.__str__(encoding='utf-8').decode('utf-8')
+  #   bstag.__str__(encoding='utf-8').decode('utf-8')
   # He found that the BeautifulSoup() __str__ method when used with a 'utf-8'
   # encoding returned effectively the same thing as str(), a Python built-in.
   # After a handful of tests, he switched to using str() to avoid the add'l
   # complexity of another BeautifulSoup method.
-  return str(bstring).decode('utf-8')
+  return str(bstag).decode('utf-8')
 
 
 # W0613:600,28:parse_master: Unused argument 'remoteurl'