How could I manually run that file to troubleshoot the problem? my mailer.py file is located under hooks folder too and mailer.conf is located outside of hooks folder.
That's simple: Simply run the command manually.
$ cd $REPOS #Where ever that is...
$ REV=230 #Whatever the revision number is
$ hooks/mailer.py commit . $REV mailer.conf
You can copy the mailer.py into another file, and play around with it. Add a few print statements, run it in the debugger, etc. For example, you probably want to futz the actual mail part of your program not to really mail everything out to the developers until you're ready to debug the particular sending mechanism. Developers don't like their mailboxes bombarded with a bunch of post-commit hook tests.
$ cp hooks/mailer.py hooks/mailer2.py
$ cp mailer.conf mailer2.conf
$ hooks/mailer2.py commit . $REV mailer.conf
What I'd suggest is that you disable your post-commit hook (in Unix/Linux, take off the executable bit should be sufficient, or simply rename post-commit to post-commit.temp). then have the user commit their change.
Once the change is committed, you have the revision number of the commit that caused the problem. Now, you can use that revision number when you manually run your post-commit hook.
Helpful Hint
If you're making a pre-commit hook, allow the hook to take either a transaction number or a revision number:
# Transaction number in pre-commit script:
REPOS="$1"
TRX="$2"
"$REPOS/hooks/myscript" -t $TRX "$REPOS"
From the command line:
cd $REPOS
hooks/myscript -r $REVISION .
This way, you can debug your pre-commit hook with already committed revisions. This is much, much easier than trying to debug by committing the same change over and over again.
Best Answer
Looks like it can't find (and thus cannot execute) commit-email.pl and log-commit.py. You either need those in the path or specified with a full path when calling (as well as making sure you even have these).