I have a SVN repository for my project; it keeps code, docs, graphs, etc. Everything related to the project is there and versioned. However, I am trying to get some kind of intelligent stats for my code. The stat program I am using is StatSVN - they seem to be best of breed as far as I can tell. However, I am getting results for old files, Word files, etc, and these results are skewing the graphs.
Ideas for battling through this and getting a reliable set of information about my cpp files? I don't just want to rebuild the repo, but I'm fearing that might be the best idea at the moment...