Re: [PATCH] add-patch: roll over to next undecided hunk

From: Junio C Hamano <gitster@pobox.com>
Date: Oct 3, 2025, 20:39 UTC
Message-ID: <xmqqcy73u3de.fsf@gitster.g>
In-Reply-To: <737e78f5-6337-4964-8385-9c35897f5dff@web.de>

René Scharfe <l.s.r@web.de> writes:

> Weird that one can switch between use and skip, but there's no
> way to revert back to undecided.

Yes, but Phillip's "if you split the resulting hunks will revert to undecided" topic, together with "you can split one hunk into one" bug that is caused by the "permitted is never reset" bug, if you can navigate back to what you already decided to use or skip, you can say "split" to revert it undecided ;-).

Show 21 quoted lines

> This should be easy to fix by resetting permitted at the start of the
> loop, no?  Patch below.
>
>> With this bug, however, we have gained a bit of useful feature, I
>> think.  Even though j/J should not be offered when we are at the
>> last hunk for a file, we do wrap-around to the first hunk.  I just
>> checked the original code before the C rewrite, and even though it
>> were written defensively so that incrementing the current hunk
>> number to 5 when you have only 4 hunks would take you back to the
>> initial hunk (instead of barfing), because we did not have this
>> "permitted is never reset" bug, it actually did not allow you to go
>> beyond the end with j/J.  Today's code seems to have inherited this
>> defensive adjustment to stay within the available hunks, and with
>> the "permitted is never reset" bug, we are taken back to the first
>> hunk.
> y/n/e on the last hunk roll over, which makes sense to me.  Their
> movement part is not mentioned in the documentation, by the way.
>
> With the patch below j/J are stopped by the floor, as seemingly
> intended.  Not sure if the (now accidental) roll-over behavior is
> better for them.

Yes. Even if it is accidental, people are too used the roll-over behaviour. So at least we should always allow J/K and probably allow j/k as long as there at least is a single undecided hunk, if we were to do this fix, and make the prompt string to match.

I only am aware of this bugginess in "j,k,J,K,s" but that is only because I did not look at others. I wouldn't be surprised if they were even buggier.

Thanks.