Re: [PATCH 2/2] Allow passing pipes for input pipes to diff --no-index

From: Junio C Hamano <gitster@pobox.com>
Date: Sep 18, 2020, 17:48 UTC
Message-ID: <xmqqbli3m0w6.fsf@gitster.c.googlers.com>
In-Reply-To: <20200918171950.GA183026@coredump.intra.peff.net>

Jeff King <peff@peff.net> writes:

Show 11 quoted lines

> Getting back to the overall feature, this is definitely something that
> has come up before. The last I know of is:
>
>   https://lore.kernel.org/git/20181220002610.43832-1-sandals@crustytoothpaste.net/
>
> which everybody seemed to like the direction of; I suspect the original
> author (cc'd) just never got around to it again. Compared to this
> approach, it uses a command-line option to avoid dereferencing symlinks.
> That puts an extra burden on the caller to pass the option, but it's way
> less magical; you could drop all of the "does this look like a symlink
> to a pipe" heuristics. It would also be much easier to test. ;)

Yes, I do remember liking the approach very much and wanted to take it once the "do not dereference symlinks everywhere---just limit it to what was given from the command line" was done.

To be quite honest, I think "git diff --no-index A B" should unconditionally dereference A and/or B if they are symlinks, whether they are symlinks to pipes, regular files or directories, and otherwise treat symlinks in A and B the same way as "git diff" if A and B are directories. But that is a design guideline that becomes needed only after we start resurrecting Brian's effort, not with these patches that started this thread.

Thanks.