Re: [PATCH] {maint} dist: add support for lzip compression

[Stefano Lattarini]

On 12/30/2011 03:05 PM, Eric Blake wrote:
> On 12/30/2011 06:47 AM, Stefano Lattarini wrote:
>>
>> OK, I've applied the patch with a couple of very minor tweaks.  Attached is
>> what I'll push in your name once I've leant what is the kosher/blessed way
>> to mark a git commit with a "no paperwork required" attribute.
> 
> Alas, right now the "solution" was a gnulib-local patch to each project,
> as we did not yet feel comfortable including it upstream in gnulib:
> 
> https://lists.gnu.org/archive/html/bug-gnulib/2011-11/msg00234.html
>
Ouch.  This is quite annoying, especially because Automake only syncs from
gnulib, and doesn't use it as a submodule (doing so would be a ludicrous
overkill, since only a couple of gnulib-provided scripts are actually used).

>> * NEWS: Update.
>> * automake.in (handle_dist): Recognize dist-lzip.
>> (make_paragraphs): Map LZIP to dist-lzip.
>> * doc/automake.texi (Dist, Options): Describe dist-lzip.
>> * lib/Automake/Options.pm (_process_option_list): Recognize
>> `dist-lzip'.
>> * lib/am/distdir.am (dist-lzip): New target.
>> (DIST_ARCHIVES) [?LZIP?]: Add `dist-lzip'.
>> (dist, dist-all) [?LZIP?]: Add command to create an lzip-compressed
>> tarball.
>> (distcheck): Handle lzip-compressed tarballs just like the others.
>> * tests/lzip.test: New test.
>> * tests/Makefile.am (TESTS): Add it.
>> ---
>>  NEWS                    |    2 ++
>>  automake.in             |    3 ++-
>>  doc/automake.texi       |   16 ++++++++++++++--
>>  lib/Automake/Options.pm |    2 +-
>>  lib/am/distdir.am       |    9 +++++++++
>>  tests/Makefile.am       |    1 +
>>  tests/lzip.test         |   43 +++++++++++++++++++++++++++++++++++++++++++
>>  7 files changed, 72 insertions(+), 4 deletions(-)
> 
> That's an awfully big diffstat
>
The diffstat is big, but most of it is done by copying-and-pasting from similar
(or practically equal) existing code; in particular, the `lzip.test' test case
and the additions to the manual are a quasi-copy of the corresponding ones
dealing with `xz'.

> to be attributing to Antonio under the
> no-assignment rules.  Would it be better to split this into two commits
> - Antonio's actual (tiny) change, and separately the tests/ addition
> that you wrote to expose it?
>
But I didn't write any test myself, actually.  The change is truly 99%
Antonio's.

Anyway, nothing has been pushed yet, so there's plenty of time to sort
this situation out before 1.11.3.

Thanks,
  Stefano