On Wed, Oct 30, 2013 at 8:07 PM, Johan Herland <johan@xxxxxxxxxxx> wrote: > On Tue, Oct 29, 2013 at 7:23 AM, Christian Couder > <christian.couder@xxxxxxxxx> wrote: >> >> I don't agree. Git doesn't need to dictate anything to be able to do >> these expansions. >> Git only needs some hints to do these expansions properly and it could >> just look at the commit template, or the config, to get those hints. >> >> For example, if there is a "Acked-by:" line in the commit template, >> then Git might decide that "ack" means "Acked-by", and then that "-by" >> means that "Peff" should be related to an author, and then that it is >> probably "Jeff King <peff@xxxxxxxx>". > > I don't like putting that much Magic into core Git... Especially not > into builtin/commit.c. However, if we - as you suggest further below - > put it into a separate helper, and we make that helper available (and > usable) from elsewhere (most importantly from hooks where > people/projects can add their own more specific functionality), then I > don't have a problem with it. Ok, great! I started working on "git interpret-trailers" and I will post an RFC patch soon. It will support both configuration as Junio suggested and reading a commit template file as you suggested. >> Ok, let's call the new plumbing command "git interpret-trailers". >> And let's suppose that "git commit" is passed "-f ack:Peff -f >> fix:security-bug" (or "--trailer ack=Peff --trailer >> fix=security-bug"). >> >> "git commit" would then call something like: >> >> git interpret-trailers --file commit_message_template.txt 'ack:Peff' >> 'fix:security-bug' >> >> And this command would output: >> >> ------------------ >> <<<upper part of commit_message_template.txt>>> >> >> Fixes: 1234beef56 (Commit message summmary) >> Reported-by: >> Suggested-by: >> Improved-by: >> Acked-by: Jeff King <peff@xxxxxxxx> >> Reviewed-by: >> Tested-by: >> Signed-off-by: Myself <myself@xxxxxxxxxxx> >> ------------------ >> >> Because it would have looked at the commit template it is passed and >> filled in the blanks it could fill using the arguments it is also >> passed. >> >> "git commit" would then put the above lines in the file that it passes >> to the prepare-commit-msg hook. >> >> Then the prepare-commit-msg could just do nothing. >> >> After the user has edited the commit message, the commit-msg hook >> could just call: >> >> git interpret-trailers --trim-empty --file commit_message.txt >> >> so that what the user changed is interpreted again. >> >> For example if the user changed the "Reviewed-by:" line to >> "Reviewed-by: Johan", then the output would be: >> >> ------------------ >> <<<upper part of commit_message.txt>>> >> >> Fixes: 1234beef56 (Commit message summmary) >> Acked-by: Jeff King <peff@xxxxxxxx> >> Reviewed-by: Johan Herland <johan@xxxxxxxxxxx> >> Signed-off-by: Myself <myself@xxxxxxxxxxx> >> ------------------ >> >> And that would be the final commit message in most cases. > > This approach looks OK to me, as long as we make sure that this > functionality is (a) optional, (b) flexible/reusable from hooks, and > (c) not bound tightly to core Git (and AFAICS, your proposal is just > that). As I said above, this stuff certainly does not belong in > builtin/commit.c... Ok, I think it will be very easy to do all with "git interpret-trailers". Best regards, Christian. -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html