For now it seems that I’ve stuck on apbs package tests.
One of them should be (these lines should appear in debian/tests/control):
Test-Command: python -c "from abpslib import *"
Depends: @
It won’t work, since apbs source in upstream source archive doesn’t contain something which is compiled to _apbslib.so and apbslib.py.
For now I found that APBS github repository really contains code, which produces these files, but they are not included to SourceForge tar.gz archive.
I could produce these files when I run this command from apbs/ repository’s subdirectory:
cmake -DCMAKE_BUILD_TYPE=Release -DBUILD_DOC=OFF -DBUILD_SHARED_LIBS=On -DCMAKE_SKIP_RPATH=On -DBUILD_TOOLS=Off \
-DENABLE_OPENMP=On -DENABLE_MPI=On -DENABLE_ZLIB=On -DENABLE_PYTHON=On -DENABLE_READLINE=Off -DFETK_PATH=/usr \
-DENABLE_FETK=Off -DENABLE_VERBOSE_DEBUG=On -DCMAKE_INSTALL_PREFIX=/data/src/apbs-pdb2pqr/apbs-install .
I also looked at issues page and found issue 381. It seems that for now this doesn’t work and test I wrote to pdb2pqr, which also gets loaded from SourceForge as source archive, won’t work for now and for some time. It makes me depressed. Other reason is that I can’t decide if I write them will it help to make it work NOW or not )
My friend would call this ‘to be a perfectionist’. And it’s hard not to be.
When I start working on pymol, I found that there are several big directories with examples, and to test it the best possible way would be to run special script to run all examples from given directory. I wrote it:
It is not perfect (I should have used pipes for console output processing, but I didn’t), but still works.
The main thing left to solve is to check all these files by hand and to fix paths, to make them not to raise various exception inside PyMol shell.
I also want to split pymol to 2 binary packages, and move to one of them architecture-independent examples part.
apbs requires the same fix, since examples folder is too big to be processed as a part of the main binary architecture-dependent package, but still useful.
Good thing is that I’ll probably distract for this weekend. Today ICFPC has started, yay! This will blow my depression away, since I like origami, which is this year’s problem :) If I’ll decide not to solve this, I’ll go outdoors, since the best thing which helps not to stuck is to look aside for a moment (or a couple of hours).