[Python-apps-team] Bug#522426: Info received ( Bug#522426: Bug#522426: Status on debian bug #522426 (Mercurial 1.2.1)?)

[Dmitrijs Ledkovs]

2009/4/28 Max Bowsher <maxb at f2s.com>:
> Dmitrijs Ledkovs wrote:
>> --- debian/rules      (revision 2804)
>> +++ debian/rules      (working copy)
>> @@ -10,17 +10,15 @@
>>  include /usr/share/cdbs/1/rules/debhelper.mk
>>  include /usr/share/cdbs/1/rules/patchsys-quilt.mk
>>  DEB_INSTALL_DOCS_ALL=
>> +DEB_PYTHON_DESTDIR = $(CURDIR)/debian/$(cdbs_curpkg)
>> +DEB_PYTHON_MODULE_PACKAGES = mercurial-common
>>
>>  # These part must be run between dh_install and dh_pysupport
>>  # So, they must be defined BEFORE including python-distutils.mk
>> -binary-install/mercurial::
>> -     # remove arch-independent python stuff
>> -     find debian/mercurial/usr/lib \
>> -             ! -name '*.so' ! -type d -delete , \
>> -             -type d -empty -delete
>> -
>>  binary-install/mercurial-common::
>>       # remove arch-dependent python stuff
>> +     (cd debian/mercurial-common && find usr/lib \
>> +             -name '*.so' | cpio -pmd ../mercurial/)
>>       find debian/mercurial-common/usr/lib \
>>               -name '*.so' ! -type d -delete , \
>>               -type d -empty -delete
>
> Surely this is going to break horribly when building arch-specific
> binary packages only?
>

Well I'm aware this is a dirty hack... But this is very similar to
what was done before. I was more concerned about byte compiling and
preinstall / postinstall scripts for python. But this hack is after
dh_install and before dh_pysupport. So imho it's an extended
dh_install so I think it's alright.

> What about simply adding:
>
> DEB_PYTHON_MODULE_PACKAGES = mercurial mercurial-common
>
> ?
>

I was thinking of that, but didn't actually tried it, because of the
"news" file.

> I agree the mention in the NEWS file that:
> +    As its name indicates, the new variable may support multiple Python
> +    packages in the future.
>
> specifically "in the future" is slightly troubling, and some
> clarification from the cdbs maintainers would probably be a good thing.
>

I know sounds very "early in the release cycle" so lets push this this
out. I hope there will be more clarity later on......

> However, it seems to work fine, confirmed by diffing the produced binary
> debs from a build with lenny-cdbs with those from a build with
> squeeze-cdbs plus that one added line.
>

The response on the bug report did say patches welcome to make it
build..... so yeah that's what I tried to do last night.

> Max.
>
>

I'm not running sid so can someone try to build it with this patch and
use it / run test suite? Or shall we upload to experimental and wait
for bugs?

Cause I don't personally use hg for anything yet....


-- 
With best regards


Dmitrijs Ledkovs (for short Dima),
Ледков Дмитрий Юрьевич