Phrases are sometimes not censored

[GoogleCodeExporter]

It is possible for a phrase (two or more words separated by spaces such as "dog 
house") to not be censored. 

Currently, the webpage is parsed using XPath. The profanity filter is applied 
on each of the returned text nodes. If a phrase is split up into separate text 
nodes, then the phrase will not be censored.

This is very rare and difficult to reproduce.

Original issue reported on code.google.com by markyd13 on 8 Jul 2010 at 4:09

[GoogleCodeExporter]

Apols for going for a support question in an issue, but the issue is related 
and I couldn't find a more appropriate place-

Is it possible to filter multi-word profanities or repugnant phrases? Your 
example of not filtering "dog house" - is there a way, or going to be a way of 
specifying a phrase?

Original comment by ora...@gmail.com on 2 Feb 2011 at 4:16