[PATCH] "git-tag -s" should create a signed annotated tag

[PATCH] "git-tag -s" should create a signed annotated tag

[Junio C Hamano]

The earlier patch 396865859918e9c7bf8ce74aae137c57da134610 (Make
builtin-tag.c use parse_options.) broke "git-tag -s".

Signed-off-by: Junio C Hamano <gitster@pobox.com>
---

 * It's _very_ disturbing that the testsuite did not catch this
   breakage, even though I though we had a very detailed coverage for
   "git tag" when it was rewritten in C.

 builtin-tag.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/builtin-tag.c b/builtin-tag.c
index 3a6467d..114c684 100644
--- a/builtin-tag.c
+++ b/builtin-tag.c
@@ -398,6 +398,9 @@ int cmd_tag(int argc, const char **argv, const char *prefix)
 
 	argc = parse_options(argc, argv, options, git_tag_usage, 0);
 
+	if (sign)
+		annotate = 1;
+
 	if (list)
 		return list_tags(list == no_pattern ? NULL : list, lines);
 	if (delete)

Re: [PATCH] "git-tag -s" should create a signed annotated tag

[Jeff King]

On Sun, Nov 25, 2007 at 03:21:42PM -0800, Junio C Hamano wrote:

>  * It's _very_ disturbing that the testsuite did not catch this
>    breakage, even though I though we had a very detailed coverage for
>    "git tag" when it was rewritten in C.

Then where is the new test to accompany your patch? :P

-- >8 --
git-tag: test that -s implies an annotated tag

This detects a regression introduced while moving git-tag to a C
builtin.

Signed-off-by: Jeff King <peff@peff.net>
---
 t/t7004-tag.sh |   16 ++++++++++++++++
 1 files changed, 16 insertions(+), 0 deletions(-)

diff --git a/t/t7004-tag.sh b/t/t7004-tag.sh
index 736f22e..c7130c4 100755
--- a/t/t7004-tag.sh
+++ b/t/t7004-tag.sh
@@ -667,6 +667,22 @@ test_expect_success 'creating a signed tag with -F - should succeed' '
 	git diff expect actual
 '
 
+cat >fakeeditor <<'EOF'
+#!/bin/sh
+test -n "$1" && exec >"$1"
+echo A signed tag message
+echo from a fake editor.
+EOF
+chmod +x fakeeditor
+get_tag_header implied-annotate $commit commit $time >expect
+./fakeeditor >>expect
+echo '-----BEGIN PGP SIGNATURE-----' >>expect
+test_expect_success '-s implies annotated tag' '
+	GIT_EDITOR=./fakeeditor git-tag -s implied-annotate &&
+	get_tag_msg implied-annotate >actual &&
+	git diff expect actual
+'
+
 test_expect_success \
 	'trying to create a signed tag with non-existing -F file should fail' '
 	! test -f nonexistingfile &&
-- 
1.5.3.6.2022.gd7807a-dirty

Re: [PATCH] "git-tag -s" should create a signed annotated tag

[Carlos Rica]

On Nov 26, 2007 12:21 AM, Junio C Hamano <gitster@pobox.com> wrote:
> The earlier patch 396865859918e9c7bf8ce74aae137c57da134610 (Make
> builtin-tag.c use parse_options.) broke "git-tag -s".
>
>  * It's _very_ disturbing that the testsuite did not catch this
>    breakage, even though I though we had a very detailed coverage for
>    "git tag" when it was rewritten in C.

Yes, you're right, "-s" without -m or -F is creating lightweight tags
in my proposal.

I think it is due to the absence of tests checking the "editor" option, that is,
calls without -m or -F options to provide the message, since the tests are using
only those and they are enabling automatically the creation of annotated tags.

I need to think in adding tests replacing the editor with a custom command
to check also the correct behaviour of git-tag when the EDITOR and/or
the other variables are set.

At the end, I've learned that it's better to test a little of every feature than
test completely only a few of them.

Thank you.