Re: How to watch files in a Git repository
- From
Jeff King <peff@peff.net>
- Date
- Jun 8, 2022, 23:55 UTC
- Message-ID
- <YqE27RU45kjNRwxf@coredump.intra.peff.net>
- In-Reply-To
- <68627d29-8ffd-2e22-46ca-c28c9e980177@gmail.com>
On Mon, Jun 06, 2022 at 06:04:11PM +0200, R. Diez wrote:
Show 7 quoted lines
> If there is nothing of the sort, I could write my own script in Bash > or Perl. I can handle cron and sending e-mails, but I do not know much > about Git's internals. Could someone provide a few pointers about how > to code this? I would expect there is some command to list commits, > and all files touched by a particular commit. And there would be some > way to interface with Bash or Perl, which does not need parsing > complicated text output from Git.
This sounds kind of like git-multimail:
https://github.com/git-multimail/git-multimail
That's usually triggered from a hook, I think, but it would not be hard to trigger it with arbitrary segments of history.
You'd probably want to keep a "seen" ref of processed commits, and move from that, like:
# assuming you just care about one branch on the remote, but this # concept can be extended to several branch=refs/remotes/origin/main seen=refs/heads/seen
git fetch
# I don't know what git-multimail expects, but this is similar to what # a server-side receive hook would show echo "$(git rev-parse $seen) $(git rev-parse $branch) $branch" | some-git-multimail-command
# now move your pointer forward for next time git update-ref $seen $branch
If multimail doesn't do what you want, then you can probably just script around:
git rev-list $seen..$branch -- $paths_you_care_about | git diff-tree --stdin -r --name-only --format="Commit %h touched: " -- $paths_you_care_about
depending how you want to format things.
-Peff