Index definition extraction from content packages is missing non-XML files

Index definitions can reference non-xml files, such as "stop.txt" for a list of stopwords.

Currently we do not include these in the cp2fm converter. 

We need to modify the code to include everything except blacklisted files such as .vlt .