[GNC-dev] Fwd: Testing python

[Mark]

Great. Glad that you managed to figure it out.
I tried a few times to see if I could find what is causing my error
messages and try to fix it, but never got anywhere. Just seemed to get
entangled in an ever-expanding maze of conflicting libs and versions...
Finally I just gave up and stopped worrying about it since my gnucash
always works without problem even showing those warnings.

cheers,

*Mark*


*epistemik at gmail.com <epistemik at gmail.com>*
*(613) 447-5385*



On Sun, Apr 26, 2020 at 10:57 AM Andy Goblins <andygoblins at gmx.com> wrote:

> Mark, I do not get any of the errors you get when running with '--debug
> --test' - the python test stuff runs fine for me: it says, "Hello from
> python!" and then prints a bunch of tests.
>
> So I did some more playing, and discovered:
> 1. If I run 'ninja install' the installed version loads the libraries
> correctly.
> 2. If I re-build gnucash with a build location of /gnucash-git/build
> instead of /some/other/place/build the libraries load correctly.
>
> I'm guessing that most people keep their 'build' directory in the root of
> the git repository rather than in a different folder, and there must be
> some hard-coded references to this that messes up builds in alternative
> directories.
>
>
> *Sent:* Saturday, April 25, 2020 at 5:17 PM
> *From:* "Mark" <epistemik at gmail.com>
> *To:* "andygoblins" <andygoblins at gmx.com>
> *Cc:* "Geert Janssens" <geert.gnucash at kobaltwit.be>, "gnucash-devel" <
> gnucash-devel at gnucash.org>
> *Subject:* Re: [GNC-dev] Fwd: Testing python
> Unfortunately, we are beyond where I can help you with the build details,
> but I can at least give my experience for comparison. I had never tried it
> before, but I went to my most recent build directory and ran the gnucash
> binary directly from there. It ran fine for me, in the sense that it didn't
> give any different errors than when I run the installed binary. And I find
> that, whatever these warnings may mean about the build, they DO NOT prevent
> gnucash from opening and DO NOT seem to cause any problems in the
> functioning at all that I ever notice. This is my output, for information:
>
> 2012 > bin/gnucash --version
> GnuCash 3.10
> Build ID: 3.10+(2020-04-11)
> 2013 > bin/gnucash --debug --extra
> Traceback (most recent call last):
>   File
> "/newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/lib/python3.9/site-packages/gnucash/gnucash_core_c.py",
> line 14, in swig_import_helper
>     return importlib.import_module(mname)
>   File "/usr/local/lib/python3.9/importlib/__init__.py", line 127, in
> import_module
>     return _bootstrap._gcd_import(name[level:], package, level)
>   File "<frozen importlib._bootstrap>", line 1030, in _gcd_import
>   File "<frozen importlib._bootstrap>", line 1007, in _find_and_load
>   File "<frozen importlib._bootstrap>", line 986, in
> _find_and_load_unlocked
>   File "<frozen importlib._bootstrap>", line 666, in _load_unlocked
>   File "<frozen importlib._bootstrap>", line 565, in module_from_spec
>   File "<frozen importlib._bootstrap_external>", line 1108, in
> create_module
>   File "<frozen importlib._bootstrap>", line 228, in
> _call_with_frames_removed
> ImportError:
> /newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/lib/python3.9/site-packages/gnucash/_gnucash_core_c.so:
> undefined symbol: forkpty
>
> During handling of the above exception, another exception occurred:
>
> Traceback (most recent call last):
>   File
> "/newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/share/gnucash/python/init.py",
> line 2, in <module>
>     import gnucash._sw_app_utils as _sw_app_utils
>   File
> "/newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/lib/python3.9/site-packages/gnucash/__init__.py",
> line 6, in <module>
>     from gnucash.gnucash_core import *
>   File
> "/newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/lib/python3.9/site-packages/gnucash/gnucash_core.py",
> line 31, in <module>
>     import gnucash.gnucash_core_c as gnucash_core_c
>   File
> "/newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/lib/python3.9/site-packages/gnucash/gnucash_core_c.py",
> line 17, in <module>
>     _gnucash_core_c = swig_import_helper()
>   File
> "/newdata/dev/Gnucash/RELEASES/gnucash-3.10/build_with-python_2020-04-12/lib/python3.9/site-packages/gnucash/gnucash_core_c.py",
> line 16, in swig_import_helper
>     return importlib.import_module('_gnucash_core_c')
>   File "/usr/local/lib/python3.9/importlib/__init__.py", line 127, in
> import_module
>     return _bootstrap._gcd_import(name[level:], package, level)
> ModuleNotFoundError: No module named '_gnucash_core_c'
> Found Finance::Quote version 1.49.
>
>
> *Mark*
>
>
> *epistemik at gmail.com <epistemik at gmail.com>*
> *(613) 447-5385*
>
>
>
> On Sat, Apr 25, 2020 at 2:54 PM andygoblins <andygoblins at gmx.com> wrote:
>
>> I get the same issue when running the gnucash binary out of the build
>> directory. Do I need to set a different environment variable so gnucash
>> loads the correct .so files?
>>
>> On April 24, 2020 12:16:22 PM UTC, Geert Janssens <
>> geert.gnucash at kobaltwit.be> wrote:
>>>
>>> Ok, so it looks like an issue with how gnucash finds its libraries from
>>> within the build directory, rather than from the installation directory.
>>>
>>>
>>>
>>> Can you run gnucash itself properly from the build directory ?
>>>
>>>
>>>
>>> Regards,
>>>
>>>
>>>
>>> Geert
>>>
>>>
>>>
>>> Op vrijdag 24 april 2020 00:37:37 CEST schreef andygoblins:
>>>
>>> > I'm building from git master. I just run "ninja build" without an
>>> install
>>>
>>> > and do my testing straight from the build directory.
>>>
>>> >
>>>
>>> > The last commit in my copy of master was on April 14. Does the master
>>> branch
>>>
>>> > include the 3.10 build fix?
>>>
>>> > On April 21, 2020 8:23:40 PM UTC, Mark <epistemik at gmail.com> wrote:
>>>
>>> > >Yea, I ran 'sudo ninja install'.
>>>
>>> > >Andy, from your initial email, can't tell for sure but possibly you
>>>
>>> > >built a
>>>
>>> > >venv right after your ninja build without doing an install?
>>>
>>> > >
>>>
>>> > >*Mark*
>>>
>>> > >
>>>
>>> > >
>>>
>>> > >*epistemik at gmail.com <epistemik at gmail.com>*
>>>
>>> > >*(613) 447-5385*
>>>
>>> > >
>>>
>>> > >
>>>
>>> > >
>>>
>>> > >On Tue, Apr 21, 2020 at 10:18 AM Geert Janssens
>>>
>>> > ><geert.gnucash at kobaltwit.be>
>>>
>>> > >
>>>
>>> > >wrote:
>>>
>>> > >> Op dinsdag 21 april 2020 15:01:33 CEST schreef Mark:
>>>
>>> > >> > My bindings are built from the git source code for the 3.10
>>> release
>>>
>>> > >> >
>>>
>>> > >> > (gnucash-3.10.tar.bz2):
>>>
>>> > >> >
>>>
>>> > >> >
>>>
>>> > >> >
>>>
>>> > >> > /opt/bin | Tue Apr 21 08:55:25 | marksa at Ares-A717-72G | bash
>>> 4.4.20
>>>
>>> > >> >
>>>
>>> > >> > 2013 > ./gnucash --version
>>>
>>> > >> >
>>>
>>> > >> > GnuCash 3.10
>>>
>>> > >> >
>>>
>>> > >> > Build ID: 3.10+(2020-04-11)
>>>
>>> > >> >
>>>
>>> > >> >
>>>
>>> > >> >
>>>
>>> > >> > So maybe I have a slightly later version of 3.10 where this bug
>>> was
>>>
>>> > >>
>>>
>>> > >> already
>>>
>>> > >>
>>>
>>> > >> > fixed?
>>>
>>> > >>
>>>
>>> > >> No, I don't think so. The fix was only included on 2020-04-12 and is
>>>
>>> > >
>>>
>>> > >not
>>>
>>> > >
>>>
>>> > >> part of the release tarball.
>>>
>>> > >>
>>>
>>> > >>
>>>
>>> > >>
>>>
>>> > >> But there are several factors at play. Do you run "make install" ?
>>> If
>>>
>>> > >
>>>
>>> > >so,
>>>
>>> > >
>>>
>>> > >> the python bindings would properly find the libraries in the
>>>
>>> > >
>>>
>>> > >installation
>>>
>>> > >
>>>
>>> > >> directory. If not, they should be found in the build directory, but
>>> I
>>>
>>> > >> expect that to fail without the aforementioned patch.
>>>
>>> > >>
>>>
>>> > >>
>>>
>>> > >>
>>>
>>> > >> Regards,
>>>
>>> > >>
>>>
>>> > >>
>>>
>>> > >>
>>>
>>> > >> Geert
>>>
>>>
>>>
>>>
>>>
>>
>> --
>> Sent from my Android device with K-9 Mail. Please excuse my brevity.
>>
>