From: Junio C Hamano <gitster@pobox.com>
Date: Sun, 18 Jan 2015 22:11:45 GMT
Subject: Re: [PATCH] test: add git apply whitespace expansion tests
Message-ID: <CAPc5daXVk_ROUy7rmzS0aosWvE2wqw8tHZgomHHkay9CZjhbiw@mail.gmail.com>
In-Reply-To: <102e322e68e78e39a7c227f3f3e102c@74d39fa044aa309eaea14b9f57fe79c>

On Sun, Jan 18, 2015 at 2:49 AM, Kyle J. McKay <mackyle@gmail.com> wrote:
> * Here's some tests.  With "apply: make update_pre_post_images() sanity
>   check the given postlen" but not "apply: count the size of postimage
>   correctly" test 1/4 and 4/4 trigger the 'die("BUG: postlen...' but
>   test 2/4 and 3/4 do not although they fail because git apply generates
>   garbage.

Do the failing cases that do not trigger the new postlen check fail
because the original (mis)counting thinks (incorrectly) that the
rewritten result _should_ fit within the original postlen (hence we
allow an in-place rewrite by passing postlen=0 to the helper), but
in fact after rewriting postimage->len ends up being longer due to
the miscounting?