Python package settings : setup. py, a Fortran for custom processing wrapping
I
have a python package that I want to distribute. I have the installation package to download tarball, unzip it and install it :
python setup.py install
Works great.
I also want to upload the package to PyPi and install it using pip.
However, the package contains a f2py wrapper of fortran that needs to be compiled at build time and the generated .so file moved to the final installation folder. I’m confused about how to use :
python3 setup.py sdist
The second is:
pip3 install pkg_name_here.tar.gz
The reason is when I run
python3 setup.py sdist
A custom command is running, part of which is an attempt to move the compiled *so file to the installation folder, which has not yet been created. The example of the code outline I used in this case here
from setuptools.command.install import install
from setuptools.command.develop import develop
from setuptools.command.egg_info import egg_info
'''
BEGIN CUSTOM INSTALL COMMANDS
These classes are used to hook into setup.py's install process. Depending on the context:
$ pip install my-package
Can yield `setup.py install`, `setup.py egg_info`, or `setup.py develop`
'''
def custom_command():
import sys
if sys.platform in ['darwin', 'linux']:
os.system('./custom_command.sh')
class CustomInstallCommand(install):
def run(self):
install.run(self)
custom_command()
class CustomDevelopCommand(develop):
def run(self):
develop.run(self)
custom_command()
class CustomEggInfoCommand(egg_info):
def run(self):
egg_info.run(self)
custom_command()
'''
END CUSTOM INSTALL COMMANDS
'''
setup(
...
cmdclass={
'install': CustomInstallCommand,
'develop': CustomDevelopCommand,
'egg_info': CustomEggInfoCommand,
},
...
)
In my case, custom_command() compiles and wraps fortran and copies the lib file to the installation folder.
I wonder if there is a way to just run these custom commands during installation with pip? That is, avoid running custom_command() during packaging and only during installation.
Update
I made some progress at Pierre de Buyl’s advice, but it still doesn’t work.
setup.py file currently looks like this:
def setup_f90_ext(parent_package='',top_path=''):
from numpy.distutils.misc_util import Configuration
from os.path import join
config = Configuration('',parent_package,top_path)
tort_src = [join('PackageName/','tort.f90')]
config.add_library('tort', sources=tort_src,
extra_f90_compile_args=['-fopenmp -lgomp -O3'],
extra_link_args=['-lgomp'])
sources = [join('PackageName','f90wrap_tort.f90')]
config.add_extension(name='',
sources=sources,
extra_f90_compile_args=['-fopenmp -lgomp -O3'],
libraries=['tort'],
extra_link_args=['-lgomp'],
include_dirs=['build/temp*/'])
return config
if __name__ == '__main__':
from numpy.distutils.core import setup
import subprocess
import os
import sys
version_file = open(os.getcwd()+'/PackageName/'+ 'VERSION')
__version__ = version_file.read().strip()
subprocess.call(cmd, shell=True)
config = {'name':'PackageName',
'version':__version__,
'project_description':'Package description',
'description':'Description',
'long_description': open('README.txt').read(),#read('README.txt'),
}
config2 = dict(config,**setup_f90_ext(parent_package='PackageName',top_path='').todict())
setup(**config2)
where f90wrap_tort.f90 is
the fortran file for f90wrapped and tort.f90 is the original fortran.
If I run the command twice, this file works with python setup.py install
I get the following error the first time I run python setup.py install
:
gfortran:f90: ./PackageName/f90wrap_tort.f90
f951: Warning: Nonexistent include directory ‘build/temp*/’ [-Wmissing-include-dirs]
./PackageName/f90wrap_tort.f90:4:8:
use tort_mod, only: test_node
1
Fatal Error: Can't open module file ‘tort_mod.mod’ for reading at (1): No such file or directory
compilation terminated.
f951: Warning: Nonexistent include directory ‘build/temp*/’ [-Wmissing-include-dirs]
./PackageName/f90wrap_tort.f90:4:8:
use tort_mod, only: test_node
1
Fatal Error: Can't open module file ‘tort_mod.mod’ for reading at (1): No such file or directory
The reason I put the include_dirs=['build/temp*/']
parameter in the extension is because I noticed after running python setup.py install
that the first time tort_mod
was being built and stored there.
What I can’t figure out is how to get the right link so that it’s done in one step.
Who can see what I’m missing?
Solution
After some Google searching, I made the following suggestions:
- Use NumPy’s distutils
- Use the
add_library
keyword for normal Fortran files (see here )。 This builds Fortran files as libraries, but does not attempt to interact with them using f2py. - Use f90wrap to pre-build f90 wrappers, include them in your package archive, and specify these files as source files in your extension.
I didn’t test the whole solution because it was pip time-consuming, but that’s what SciPy did with some of their modules, see here .
There is an item in NumPy’s documentation that is exceeded add_ library
Edit 1: After configuring the build with include_dirs=['build/temp.linux-x86_64-2.7']),
I got this directory structure on my first build attempt.
build/lib.linux-x86_64-2.7
├── crystal_torture
│ ├── cluster.py
│ ├── dist.f90
│ ├── f90wrap_tort.f90
│ ├── graph.py
│ ├── __init__.py
│ ├── minimal_cluster.py
│ ├── node.py
│ ├── node.pyc
│ ├── pymatgen_doping.py
│ ├── pymatgen_interface.py
│ ├── tort.f90
│ ├── tort.py
│ └── tort.pyc
└── crystal_torture.so