Re: [PATCH 17/20] abbrev: unify the handling of empty values

On 9 June 2018 at 00:41, Ævar Arnfjörð Bjarmason [off-list ref] wrote:

For no good reason the --abbrev= command-line option was less strict
than the core.abbrev config option, which came down to the latter
using git_config_int() which rejects an empty string, but the rest of
the parsing using strtoul() which will convert it to 0.

It will still be less strict in that it accepts trailing garbage, e.g.,
`--abbrev=7a`. Probably ok to leave it at that in this series, but
possibly useful to mention here that this only makes these options "less
differently strict".

I also notice that the documentation of `--abbrev` starts with "Instead
of showing the full 40-byte hexadecimal object name" which doesn't seem
right. I get much shorter IDs and I can't see that I'd have any
configuration causing this. Anyway, that might not be anything this
series needs to do anything about.

+               if (!strcmp(arg, ""))
+                       die("--abbrev expects a value, got '%s'", arg);

+               if (!strcmp(arg, ""))
+                       return opterror(opt, "expects a value", 0);

+               if (!strcmp(optarg, ""))
+                       die("--abbrev expects a value, got '%s'", optarg);

+       test_must_fail git branch -v --abbrev= 2>stderr &&
+       test_i18ngrep "expects a value" stderr &&

+       test_must_fail git log --abbrev= -1 --pretty=format:%h 2>stderr &&
+       test_i18ngrep "expects a value" stderr &&

+       test_must_fail git diff --raw --abbrev= HEAD~ 2>stderr &&
+       test_i18ngrep "expects a value" stderr &&

Mismatch re i18n-ed or not between implementations and tests?

Martin