From: D. Ben Knoble <ben.knoble@gmail.com>
Date: Thu, 18 Jun 2026 21:29:33 GMT
Subject: Re: [PATCH v3 3/4] history: add squash subcommand to fold a range
Message-ID: <CALnO6CDK50QhiwfO3=gyrn0HNFC6=2FfjByAcLxtg0op1cpF+g@mail.gmail.com>
In-Reply-To: <xmqqzf0rpmn0.fsf@gitster.g>

On Thu, Jun 18, 2026 at 5:25 PM Junio C Hamano <gitster@pobox.com> wrote:
>
> Junio C Hamano <gitster@pobox.com> writes:
>
> > As t3454 is taken by another topic already in flight, I've queued a
> > trivial "rename it to t3455" patch on top before queuing the topic.
>
> Another tweak I had to make was to replace "grep" with "test_grep"
> to avoid triggering test lint added by another topic in flight.
>
> For the one in the second hunk, it may be much better to rewrite it
> to process "out" directly with the awk script without preprocessing
> it with "grep", as awk is a programming language capable enough to
> recognize a line that matches a pattern and process only those
> matching lines by itself.
>
> --- >8 ---
> Author: Junio C Hamano <gitster@pobox.com>
> Date:   Thu Jun 18 13:44:36 2026 -0700
>
>     SQUASH??? avoid test_grep lint triggering on uses of raw grep
>
> diff --git a/t/t3455-history-squash.sh b/t/t3455-history-squash.sh
> index 1edd148295..20370c0136 100755
> --- a/t/t3455-history-squash.sh
> +++ b/t/t3455-history-squash.sh
[snip]
> @@ -177,7 +177,7 @@ test_expect_success '--dry-run predicts the rewrite without performing it' '
>         head_before=$(git rev-parse HEAD) &&
>
>         git history squash --dry-run start.. >out &&
> -       grep "^update refs/heads/" out >update &&
> +       test_grep "^update refs/heads/" out >update &&
>         predicted=$(awk "{print \$3}" update) &&
>         test_cmp_rev "$head_before" HEAD &&

Odd: I thought the other topic acknowledged that bare grep as a filter
(here, with stdout redirected) was fine. My memory must not be right
:)

-- 
D. Ben Knoble