Revert "handle unicode escapes the right way, now that i properly looked up what was going on" because it turns out that result languages like ja mix the escaped unicode and the natural unicode and that screws up every method i've found to parse so far, as they all appear to expect all escaped or all native. here's an example:

\u0026lt;ハイテク\u0026gt;

if you can figure that out, un-revert and apply your fix.

This reverts commit 9836ebf11f.

dr.botzo.py

@@ -181,10 +181,17 @@ class GoogleTranslate(Module):
             end_idx = translation.find('"}, "')
             translation = translation[:end_idx]
 
+            # do some text conversion
+            translation = translation.replace('\\u0026quot;', '"')
+            translation = translation.replace('\\u0026amp;', '&')
+            translation = translation.replace('\\u003c', '<')
+            translation = translation.replace('\\u003e', '>')
+            translation = translation.replace('\\u0026#39;', '\'')
+
             if replypath is None:
-                return translation.decode('unicode_escape')
+                return translation
             else:
-                connection.privmsg(replypath, translation.decode('unicode_escape'))
+                connection.privmsg(replypath, translation)
 
 class Countdown(Module):
     """Class that adds a countdown item to the bot