From: Markus Heidelberg <markus.heidelberg@web.de>
Date: Sat, 30 May 2009 16:26:11 GMT
Subject: Re: [PATCH v2] add --abbrev to 'git cherry'
Message-ID: <200905301826.11924.markus.heidelberg@web.de>
In-Reply-To: <20090530140349.GA25265@unpythonic.net>

Jeff Epler, 30.05.2009:
>  Documentation/git-cherry.txt |    5 ++++-
>  builtin-log.c                |   24 +++++++++++++++++++-----
>  2 files changed, 23 insertions(+), 6 deletions(-)

You could also add --abbrev= to the bash completion.

> diff --git a/Documentation/git-cherry.txt b/Documentation/git-cherry.txt
> index 7deefda..5c03da0 100644
> --- a/Documentation/git-cherry.txt
> +++ b/Documentation/git-cherry.txt
> @@ -49,6 +49,9 @@ OPTIONS
>  -v::
>  	Verbose.
>  
> +--abbrev[=<n>]::
> +	Abbreviate commit ids to the given number of characters

The full stop is missing :)
And you could add "The default value is 7." as in the git-branch docs.
Or even copy the whole description from there for consistency, it also
mentions that this sets the minimum length, the displayed SHA1 may be
longer, but more about this below.

> diff --git a/builtin-log.c b/builtin-log.c
> index f10cfeb..1f3093e 100644
> --- a/builtin-log.c
> +++ b/builtin-log.c
> @@ -1218,12 +1232,12 @@ int cmd_cherry(int argc, const char **argv, const char *prefix)
>  			struct strbuf buf = STRBUF_INIT;
>  			pretty_print_commit(CMIT_FMT_ONELINE, commit,
>  			                    &buf, 0, NULL, NULL, 0, 0);
> -			printf("%c %s %s\n", sign,
> +			printf("%c %.*s %s\n", sign, abbrev,
>  			       sha1_to_hex(commit->object.sha1), buf.buf);
>  			strbuf_release(&buf);
>  		}
>  		else {
> -			printf("%c %s\n", sign,
> +			printf("%c %.*s\n", sign, abbrev,
>  			       sha1_to_hex(commit->object.sha1));
>  		}

There is no test for unique ids. "git cherry --abbrev=4" always prints 4
chars per SHA1, so "git show" on these SHA1s mostly gives "error: short
SHA1 xxxx is ambiguous." in git.git.

find_unique_abbrev() will help.

Markus