[PATCH] memctx: fix manifest for removed files (issue4470)

[Sean Farley]

FUJIWARA Katsunori writes:

> At Tue, 16 Dec 2014 15:33:11 -0800,
> Pierre-Yves David wrote:
>> 
>> On 12/16/2014 11:39 AM, Jordi Gutiérrez Hermoso wrote:
>> > On Tue, 2014-12-16 at 10:43 -0500, Augie Fackler wrote:
>> >> # HG changeset patch
>> >> # User Augie Fackler <augie at google.com>
>> >> # Date 1418673654 18000
>> >> #      Mon Dec 15 15:00:54 2014 -0500
>> >> # Node ID 40bd85091fb37fc992e3af2128a8ed31f31a47ff
>> >> # Parent  65c854f92d6ba8861414ff3191182fba28777a82
>> >> memctx: fix manifest for removed files (issue4470)
>> >
>> > Nice! Thanks for the fix!
>> >
>> >> diff --git a/tests/test-commit.t b/tests/test-commit.t
>> >> --- a/tests/test-commit.t
>> >> +++ b/tests/test-commit.t
>> >> @@ -429,6 +429,68 @@ specific template keywords work well
>> > [snip]
>> >> +  $ cd ..
>> >> +  $ rm -r lol
>> >
>> > Maybe remove this last `rm -r` from the test? I had it in the original
>> > repro, but it seems out of place in the test. It's useful to keep the
>> > repos in the tests when doing run-tests.py --keep-tmpdir.
>> >
>> > Also, perhaps a better name than "lol" should be used here? I'm kind
>> > of cavalier about my metasyntactic variables.
>> 
>> I've renamed "lol" to issue4470, dropped the rm and pushted the result 
>> to the clowncopter.
>
> Oh! I also just finished patches to fix this issue !
>
> Even though it may be too late, I post my patches for reference :-)
>
> In addition to fixing issue4470, my series also fixes "missing newly
> added files in the manifest" issue of memctx (by #2 of series)

Wow, this is much more robust and helps with issues when trying to use
memctx in the merge machinery (I have a half-baked version of what
Katsunori wrote). Please queue this one if not too late.