Re: [PATCH v5 0/2] CRLF: ce_compare_data() checks for a sha1 of a path

From: Junio C Hamano <gitster@pobox.com>
Date: May 19, 2016, 23:10 UTC
Message-ID: <xmqqfutdhcej.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <1463667675-25930-1-git-send-email-tboegi@web.de>

tboegi@web.de writes:

Show 5 quoted lines

> From: Torsten Bögershausen <tboegi@web.de>
>
> Break up the old 10/10 series about CLRF handling into smaller
> series. This is a small bugfix, when merge.renormalize is used
> with core.autocrlf (and no attributes are set).

Is it worth protecting the fix with a new test? Or does this flip an existing "expect-failure" to "expect-success"?

> Prepare the refactoring to use the streaming interface.

Show 16 quoted lines

> Changes since v4:
>  - Rename #define in cache.h into HASH_USE_SHA_NOT_PATH
>  - convert.c: Rename has_cr_in_index into blob_has_cr()
>    Better logic when sha1 != NULL,
>    Adjusted the commit message
>
> Torsten Bögershausen (2):
>   read-cache: factor out get_sha1_from_index() helper
>   convert: ce_compare_data() checks for a sha1 of a path
>
>  cache.h      |  4 ++++
>  convert.c    | 34 ++++++++++++++++++++++------------
>  convert.h    | 23 +++++++++++++++++++----
>  read-cache.c | 33 +++++++++++++++++++++------------
>  sha1_file.c  | 17 +++++++++++++----
>  5 files changed, 79 insertions(+), 32 deletions(-)