Re: Custom merge driver with no rename detection

From: Junio C Hamano <gitster@pobox.com>
Date: Feb 15, 2016, 11:03 UTC
Message-ID: <xmqqd1ryqlpr.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <CALMa68p9jdO63k2NLTUJXyms=Fc+woXx00UXCxzJ5_zV8jO8Rw@mail.gmail.com>

Felipe Gonçalves Assis <felipeg.assis@gmail.com> writes:

> However, if you do find this approach acceptable/desirable
> (rename-threshold > 100%), I can work on the issues pointed out and
> propose a proper patch.

The caller asks diffcore-rename to detect rename, and the algorithm compares things to come up with a similarity score and match things up. And you add an option to the rename detection logic to forbid finding any?

To be bluntly honest, the approach sounds like a crazy talk.

If your goal is not to allow rename detection at all, why not teach the caller of the diff machinery not to even ask for rename detection at all? merge-recursive.c has a helper function called get_renames(), and it calls into the diff machinery passing DIFF_DETECT_RENAME option. As a dirty hack, I think you would achieve your desired result if you stop passing that option there.

I called that a "dirty hack", because for the purpose of not allowing rename detection inside merge-recursive, I think an even better thing to do is to teach the get_renames() helper to report to its caller, under your new option, "I found no renames at all" without doing anything.

It might be just a simple matter of teaching its callers (there probably are two of them, one between the common ancestor and our branch and the other between the common ancestor and their branch) not call the get_renames() helper at all under your new option, but I didn't check these callers closely.