Jorge Juan Garcia Garcia  <jorge-juan.garcia-gar...@ensimag.imag.fr>
writes:

> Some people always run 'git status -s'.
> The configuration variable status.short allows to set it by default.
>
> Signed-off-by: Jorge Juan Garcia Garcia 
> <jorge-juan.garcia-gar...@ensimag.imag.fr>
> Signed-off-by: Mathieu Lienard--Mayor <mathieu.lienard--ma...@ensimag.imag.fr>
> Signed-off-by: Matthieu Moy <matthieu....@grenoble-inp.fr>
> ---
>
> Changes since v2:
>  -removal of double quotes in test
>  -use of git config --unset to clean up test environment
>
>  Documentation/config.txt |    4 ++++
>  builtin/commit.c         |    5 +++++
>  t/t7508-status.sh        |   35 +++++++++++++++++++++++++++++++++++
>  3 files changed, 44 insertions(+), 0 deletions(-)
>
> diff --git a/Documentation/config.txt b/Documentation/config.txt
> index 6e53fc5..1983bf7 100644
> --- a/Documentation/config.txt
> +++ b/Documentation/config.txt
> @@ -2066,6 +2066,10 @@ status.relativePaths::
>       relative to the repository root (this was the default for Git
>       prior to v1.5.4).
>  
> +status.short::
> +     Set to true to enable --short by default in linkgit:git-status[1].
> +     The option --no-short takes precedence over this variable.
> +
>  status.showUntrackedFiles::
>       By default, linkgit:git-status[1] and linkgit:git-commit[1] show
>       files which are not currently tracked by Git. Directories which
> diff --git a/builtin/commit.c b/builtin/commit.c
> index 1621dfc..287f1cb 100644
> --- a/builtin/commit.c
> +++ b/builtin/commit.c
> @@ -1112,6 +1112,11 @@ static int git_status_config(const char *k, const char 
> *v, void *cb)
>                       s->submodule_summary = -1;
>               return 0;
>       }
> +     if (!strcmp(k, "status.short")) {
> +             if (git_config_bool(k, v))
> +                     status_format = STATUS_FORMAT_SHORT;

And if the user has

        [status]
                short = no

in $GIT_DIR/config for this particular project, perhaps in order to
override a blanket setting

        [status]
                short

that is in $HOME/.gitconfig, what should happen?

Perhaps you need

        if (git_config_bool(...))
                status_format = STATUS_FORMAT_SHORT;
        else            
                status_format = STATUS_FORMAT_NONE; /* default */

or something here?

> +             return 0;
> +     }
>       if (!strcmp(k, "status.color") || !strcmp(k, "color.status")) {
>               s->use_color = git_config_colorbool(k, v);
>               return 0;
> diff --git a/t/t7508-status.sh b/t/t7508-status.sh
> index e2ffdac..d99ca9f 100755
> --- a/t/t7508-status.sh
> +++ b/t/t7508-status.sh
> @@ -1335,4 +1335,39 @@ test_expect_failure '.git/config ignore=all suppresses 
> submodule summary' '
>       git config -f .gitmodules  --remove-section submodule.subname
>  '
>  
> +test_expect_success 'Setup of test environment' '
> +     git config status.showUntrackedFiles no
> +'
> +
> +test_expect_success '"status.short=true" same as "-s"' '
> +     git -c status.short=true status >actual &&
> +     git status -s >expected_short &&
> +     test_cmp actual expected_short
> +'
> +
> +test_expect_success '"status.short=true" different from "--no-short"' '
> +     git status --no-short >expected_noshort &&
> +     test_must_fail test_cmp actual expected_noshort
> +'
> +
> +test_expect_success '"status.short=true" weaker than "--no-short"' '
> +     git -c status.short=true status --no-short >actual &&
> +     test_cmp actual expected_noshort
> +'
> +
> +test_expect_success '"status.short=false" same as "--no-short"' '
> +     git -c status.short=false status >actual &&
> +     git status -s >expected_short &&
> +     test_cmp actual expected_noshort
> +'
> +
> +test_expect_success '"status.short=false" weaker than "-s"' '
> +     git -c status.short=false status -s >actual &&
> +     test_cmp actual expected_short
> +'
> +
> +test_expect_success 'Restore default test environment' '
> +     git config --unset status.showUntrackedFiles
> +'

A few observations.

 * It is very good that you check not just positive cases that show
   off how well this new feature works (i.e. status.short set
   without command line override gives a short output) but also
   negative cases that make sure the new feature does not kick in
   when it should not.  You test all four combinations, which is
   good.

 * If any of the first three fails, you may not have the correct
   string in expected_short or expected_noshort when running later
   tests that depend on them.

 * Similarly, if the first one to set showUntrackedFiles fails, the
   last one to --unset would also fail.

Perhaps limiting the number of tests that must pass (otherwise the
remainder becomes useless) by doing something like this is a better
alternative:


        test_expect_success 'setup for status.short' '
                git status --short >expected_short &&
                git status --no-short >expected_noshort
        '

        test_expect_success '-c status.short=true == status -s' '
                test_config status.showUntrackedFile no &&
                test_config status.short yes &&
                git status >actual &&
                test_cmp expected_short actual
        '

        test_expect_success 'status --no-short defeats status.short=true' '
                test_config status.showUntrackedFile no &&
                test_config status.short yes &&
                git status --no-short >actual &&
                test_cmp expected_noshort actual
        '

        ... other two combinations here ...

Points to note:

 * test_config takes care of setting configuration variables and
   then unsetting them when the test is done.

 * test_cmp should compare expected with actual, not the other way,
   so that "./t7508-status.sh -v" shows the diff between the two 
   shows how the actual output differs from what is expected.
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to