1 |
On Wed, Aug 21, 2013 at 3:27 PM, Michał Górny <mgorny@g.o> wrote: |
2 |
> Dnia 2013-08-21, o godz. 14:31:01 |
3 |
> Mike Gilbert <floppym@g.o> napisał(a): |
4 |
> |
5 |
>> On Wed, Aug 21, 2013 at 6:43 AM, Michał Górny <mgorny@g.o> wrote: |
6 |
>> > --- |
7 |
>> > gx86/eclass/distutils-r1.eclass | 13 +++++++++++++ |
8 |
>> > 1 file changed, 13 insertions(+) |
9 |
>> > |
10 |
>> > diff --git a/gx86/eclass/distutils-r1.eclass b/gx86/eclass/distutils-r1.eclass |
11 |
>> > index e95df3e..39e4434 100644 |
12 |
>> > --- a/gx86/eclass/distutils-r1.eclass |
13 |
>> > +++ b/gx86/eclass/distutils-r1.eclass |
14 |
>> > @@ -322,6 +322,17 @@ _distutils-r1_disable_ez_setup() { |
15 |
>> > fi |
16 |
>> > } |
17 |
>> > |
18 |
>> > +# @FUNCTION: _distutils-r1_copy_egg_info |
19 |
>> > +# @DESCRIPTION: |
20 |
>> > +# Copy egg-info files to the ${BUILD_DIR} (that's going to become |
21 |
>> > +# egg-base in esetup.py). This way, we respect whatever's in upstream |
22 |
>> > +# egg-info. |
23 |
>> > +_distutils-r1_copy_egg_info() { |
24 |
>> > + mkdir -p "${BUILD_DIR}" || die |
25 |
>> > + # stupid freebsd can't do 'cp -t ${BUILD_DIR} {} +' |
26 |
>> > + find -name '*.egg-info' -exec cp -pr {} "${BUILD_DIR}"/ ';' || die |
27 |
>> > +} |
28 |
>> > + |
29 |
>> > # @FUNCTION: distutils-r1_python_prepare_all |
30 |
>> > # @DESCRIPTION: |
31 |
>> > # The default python_prepare_all(). It applies the patches from PATCHES |
32 |
>> > @@ -382,6 +393,8 @@ distutils-r1_python_configure() { |
33 |
>> > distutils-r1_python_compile() { |
34 |
>> > debug-print-function ${FUNCNAME} "${@}" |
35 |
>> > |
36 |
>> > + _distutils-r1_copy_egg_info |
37 |
>> > + |
38 |
>> > esetup.py "${@}" |
39 |
>> > } |
40 |
>> > |
41 |
>> > -- |
42 |
>> > 1.8.3.2 |
43 |
>> > |
44 |
>> |
45 |
>> I don't think there is any point in copying egg-info *files* (created |
46 |
>> via distutils); we only seem to run into issues with setuptools and |
47 |
>> the egg-info directories. |
48 |
> |
49 |
> You mean to use '-type d' here? |
50 |
> |
51 |
|
52 |
Yes. |