this one would be more difficult, but could be
successfully processed by processing the cvs log output
on a file, for eample, once you know that a file is
different from that in cvs, yet the file exists in the
- capture the revision from the diff output for the tag
or the head.
- run cvs log on the file
- extract the log entry for that revision
log entries after the revision line are generally of
"date: <date> <time>; author: <cvsuser>; state:
[lines: <+/- lines>]"
where lines can be optional.
yank out the author, and add it to the report. Very
useful when tracking down who broke the build...