From: Junio C Hamano <gitster@pobox.com>
Date: Wed, 30 Sep 2015 20:34:02 GMT
Subject: Re: [PATCH] rebase: accept indented comments (fixes regression)
Message-ID: <xmqq8u7n7ikl.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <vpqh9mbhek2.fsf@grenoble-inp.fr>

Matthieu Moy <Matthieu.Moy@grenoble-inp.fr> writes:

> Junio C Hamano <gitster@pobox.com> writes:
>
>> +		pick|p|drop|d|reword|r|edit|e|squash|s|fixup|f)
>> +			if ! check_commit_sha "${rest%% *}" "$lineno" "$1"
>
> This does not pass my "tabs" test, as it parses the sha1 out of the line
> assuming it's separated with a space.

Yes, it was very much deliberate to match what transform_todo_ids does.

I do not think we mind loosening this SP to SP/HT at all, but if we
are going to go in that direction, I'd very much more prefer to do

	"${rest%%[     ]*}"

here _and_ in transform_todo_ids so that we can keep the identical
parsing, which is the primary point of the "regression fix", rather
than doing this:

> I changed it to
>
> 	while read -r command sha1 rest
>
> which is a bit more lazy.

which is to invite the two parsers drift apart over time.

Thanks.