On Tue, Jun 07, 2016 at 06:13:25PM -0400, Jeff King wrote:
> On Tue, Jun 07, 2016 at 03:11:47PM -0700, Junio C Hamano wrote:
>
> > On Tue, Jun 7, 2016 at 3:07 PM, Jeff King <peff@xxxxxxxx> wrote:
> > >>
> > >> Puzzled. I didn't even use --format=%(tagname) in the above.
> > >
> > > No, but you used --show-tagname, which does not exist today (and which
> > > IMHO should be implemented as --format). Would --show-tagname take
> > > either a tagname _or_ a sha1? I assume it would not be calling
> > > get_sha1(), as having it find refs/heads/$tag would be silly.
> >
> > And you do not even want to rely on where refs/tags/* it lives.
> > show-tagname, as I hinted in the first response, was meant to be
> > a short-hand for
> >
> > git cat-file tag $tag_object_name | sed -e '/^$/q' -e 's/^tag //p'
> >
> > so I am still puzzled.
>
> If you are suggesting that you can do the whole thing today by parsing
> the tag object yourself, then sure, I agree. I thought the point of the
> exercise was to make that less painful for the callers.
This is what I understand so far, it seems all of us are on the same
understanding here?
1.- we can do this right now by sed-ing out the tagname, but it might
not be optimal.
2.- We can, instead, provide a --format flag to git tag -v for the same
purpose. Which would only print the tag (if the appropriate format
string is provided)
I still agree with the rest of Peff's comments about this approach. I'm
not sure about which approach to take either.
-Santiago.
--
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