Ah, the joy of Perl.
I've previously set up a system so that our trace files contain a dump of the version numbers for all of the code files. A few internal bug reports later, it's clear that this is going to be useful, and moreover, we're going to need an easy way to recreate the codebase that corresponds to a trace file.
120 lines of Perl later, I have such a beast. For a while I thought I would have to reinvent a chunk of Perl for parsing RCS logfiles that I'd previously written (but under the auspices of my old gig, so I don't have it any more). In the end I remembered to go to CPAN rather than re-re-inventing the wheel and found a nice module ready for reuse.
[A:37385 B:3218 C:346 D:9187 E:8073 Total:58209]