Re: Antw: [EXT] Re: git add --interactive patch improvement for split hunks

[Jeff King <peff@peff.net>]

On Mon, Jun 28, 2021 at 01:20:46PM +0200, Ævar Arnfjörð Bjarmason wrote:

> > There always was a related bug (IMHO) that showed the context of the
> > previous function even though the actual change was within a new
> > function (that starts within the context lines). So if that bug were
> > fixed, my guess is that the other would be as well.
> > However I don't know how easy or hard the fix will be.
> > Maybe the "definition" of function context is just different; I don't really know.
> 
> Does that bug perhaps have anything to do with:
> https://lore.kernel.org/git/20210215155020.2804-2-avarab@gmail.com/

I think it's similar. The issue is that we search backwards for a
funcname match from the top of the hunk, _not_ from the first changed
line. IIRC, that has been discussed before and considered "not a bug",
but I could be mis-remembering (and it's a tricky thing to search in the
archive for[0]).

The problem with split hunks is different, though; we do not search for
a funcname line at all on the second half of the hunk.

-Peff

[0] I did come up with:

      https://lore.kernel.org/git/1399824596-4670-1-git-send-email-avarab@gmail.com/

    which has discussion between you and me on this very same splitting
    topic back in 2014! Double-curious, your patch there implements the
    same "keep the hunk header on split" we've been discussing here, and
    we were all positive on it. Yet it doesn't seem to have ever gotten
    applied.

    It looks like Junio carried it in "What's Cooking" for almost a
    year, marked as "waiting for re-roll" to handle the squash, but then
    eventually discarded it as stale. :(