About

This is a repository for creating a python library pycrysfml based on the new automatically generated Python API for the crystallographic library CrysFML2008 (Fortran 2008) from https://code.ill.fr/scientific-software/CrysFML2008.

CI status

Workflow for testing the build process: debug-build-and-test

Python version: 3.11

Platform / Compiler	gfortran	ifx	ifort	nagfor
Windows 10	✅	❌ Issue #23	✅	⚙️ Testing
Ubuntu 22.04	✅	✅	✅	⚙️ Testing
macOS 12 (Intel)	✅	Unsupported	❌ Issue #22	⚙️ Testing
macOS 14 (Apple Silicon)	✅	Unsupported	Unsupported	⚙️ Testing

Workflow for building a python release package: release-build-and-test

Fortran compiler: gfortran

Platform / Python	3.10	3.11	3.12
Windows 10	✅	✅	✅
Ubuntu 22.04	✅	✅	✅
macOS 12 (Intel)	✅	✅	✅
macOS 14 (Apple Silicon)	✅	✅	✅

Installing

Standard installation

The pycrysfml python package is currently in beta testing, with pre-releases being published on our own PyPi server. To test it, please install it using the PIP package manager with additional option as shown below.

• Create and activate a python environment (optional)

  macOS and Linux

  python3.11 -m venv .venv
  source .venv/bin/activate
  

  Windows

  python3.11 -m venv .venv
  .venv\Scripts\activate
  

• Upgrade the package installer for Python (optional)

  pip install --upgrade pip
  

• Install pycrysfml using a link to our own online repository

  pip install pycrysfml --extra-index-url https://easyscience.github.io/pypi/
  

Installation from source

If the standard installation doesn't work for you, try building pycrysfml locally and installing it from the generated Python wheel.

• Create and activate a python environment (optional)

  macOS and Linux

  python3.11 -m venv .venv
  source .venv/bin/activate
  

  Windows

  python3.11 -m venv .venv
  .venv\Scripts\activate
  

• Upgrade the package installer for Python (optional)

  pip install --upgrade pip
  

• Install Python dependences, including extras (ci and test)

  pip install '.[ci,test]'
  

• Print possible options for creating job scripts (optional)

  python pybuild.py --help
  

• Create job scripts with default options

  python pybuild.py --create-scripts
  

• Print some build-specific variables (optional)

  scripts/print_build_variables.sh
  

• Create CFML and pyCFML directories

  scripts/create_cfml_repo_dir.sh
  scripts/create_cfml_build_dir.sh
  scripts/create_cfml_dist_dir.sh
  scripts/create_pycfml_src_dir.sh
  scripts/create_pycfml_build_dir.sh
  scripts/create_pycfml_dist_dir.sh
  

• Download CFML repository

  scripts/download_cfml_repo.sh
  

• Build CFML modules

  scripts/rename_global_deps_file.sh
  scripts/build_cfml_modules_obj.sh
  scripts/delete_renamed_global_deps_file.sh
  

• Build CFML static library

  scripts/build_cfml_static_lib.sh
  

• Make CFML distribution

  scripts/move_built_to_cfml_dist.sh
  

• Build und run CFML functional test programs

  scripts/build_cfml_test_programs.sh
  scripts/copy_cfml_test_programs_to_tests_dir.sh
  scripts/run_cfml_functional_tests_no_benchmarks.sh
  

• Create pyCFML source code

  scripts/create_pycfml_src.sh
  

• Build pyCFML modules

  scripts/build_pycfml_modules_obj.sh
  

• Build pyCFML shared obj / dynamic library

  scripts/build_pycfml_lib_obj.sh
  scripts/build_pycfml_shared_obj_or_dynamic_lib.sh
  

• Make pyCFML distribution

  scripts/copy_built_to_pycfml_dist.sh
  scripts/change_runpath_for_built_pycfml.sh
  scripts/copy_extra_libs_to_pycfml_dist.sh
  scripts/copy_py_api_files_to_pycfml_dist.sh
  scripts/copy_init_file_to_pycfml_dist.sh
  scripts/copy_cfml_databases_to_pycfml_dist.sh
  

• Create Python package wheel of pyCFML

  scripts/validate_pyproject_toml.sh
  scripts/create_pycfml_python_wheel.sh
  scripts/rename_pycfml_python_wheel.sh
  

• Install pyCFML from Python package wheel (with dev extras)

  scripts/install_pycfml_from_wheel.sh
  

• Run pyCFML unit tests

  scripts/run_pycfml_unit_tests.sh
  

• Run pyCFML functional tests

  scripts/run_pycfml_functional_tests_no_benchmarks.sh
  

How to use

Compute the powder diffraction pattern

Here is the simplest example of using pycrysfml to calculate a powder diffraction pattern from a Python dictionary, which contains a description of the crystallographic phase as well as parameters related to the instrument and experiment.

Install the matplotlib package used in the example below:

pip install matplotlib

Run this example srtio3-pattern-simulation.py to see the simulated powder diffraction pattern:

python examples/pyCFML/cfml_utilities/powder_pattern_from_json/srtio3-pattern-simulation.py

More features will be added as the project develops.