Re: [PATCH] describe: output tag's ref instead of embedded name

[Matheus Tavares Bernardino <matheus.bernardino@usp.br>]

On Thu, Feb 20, 2020 at 2:34 PM Junio C Hamano <gitster@pobox.com> wrote:
>
[...]
> Another thing that is not satisfying is what should happen in "all"
> mode.  We add "tags/" prefix so in the case we've been discussing,
> the output would be "tags/v1.0-0-g0123456", but the whole reason why
> we add the prefix is to say that the early part of the name, "v1.0",
> is a tag, and it would be natural to associate it with refs/tags/v1.0
> that is *not* Bob's tag.

Yeah, that is not very satisfying, but at least the emitted warning
would make the user think twice before wrongly associating
refs/tags/v1.0 with Bob's tag?

> Having said all that, here is what I have at this moment.
>
> -- >8 --
> Subject: describe: force long format for a name based on a mislocated tag
[...]
>
> The output from "git describe", at least in the modern Git, should
> be usable as an object name to name the exact commit given to the
> "git describe" command.  Using the tagname, when two names differ,
> breaks this property, when describing a commit that is directly
> pointed at by such a tag.  An annotated tag Bob made as "v1.0" may
> sit at "refs/tags/v1.0-bob" in the ref hierarchy, and output from
> "git describe v1.0-bob^0" would say "v1.0", but there may not be
> any tag at "refs/tags/v1.0" locally or there may be another tag that
> points at a different object.
>
> Note that this won't be a problem if a commit being described is not
> directly pointed at by such a mislocated tag.  In the example in the
> previous paragraph, "git describe v1.0-bob~1" would result in "v1.0"
> (i.e. the tagname taken from the tag object) followed by "-1-gXXXXX"

I now this is just an illustration, but shouldn't this example be "git
describe --contains v1.0-bob~1"? Otherwise, I think the tag wouldn't
be found as it comes after the given commit. Testing here without
--contains I get the error "fatal: No tags can describe <sha1 hash>"

However, when using --contains, the output is also not what I
expected. It doesn't fail, but I get "v1.0-bob~1". I.e., it not only
prints the refname instead of the tagname but also don't print any
warnings... This is not what we want, right?

> Show the name in the long format, i.e. with "-0-gXXXXX" suffix, when
> the name we give is based on a mislocated annotated tag to ensure
> that the output can be used as the object name for the object
> originally given to the command to fix the issue.

Another case that came to my mind is when the user runs `git describe
--abrev=0 HEAD` and v1.0-bob points to HEAD~. In this case, v1.0 will
be displayed without suffix, so it is not usable at get_oid(). Should
we, perhaps, also force the long output in this case?

And talking about the forced long output, using `--abbrev=0` in HEAD
(i.e. tag: v1.0-bob), after the patch, will make the complete hash be
displayed. This might be an unnecessary cosmetic nitpick, but we might
want to follow the default behavior when --long is given and use just
as many digits as necessary?