Skip to content

Merge pull request #133 from imcs-compsim/revert-132-fix_test_name #428

Merge pull request #133 from imcs-compsim/revert-132-fix_test_name

Merge pull request #133 from imcs-compsim/revert-132-fix_test_name #428

Workflow file for this run

# Workflow for testing meshpy
name: Test meshpy
on:
push:
branches:
- main
pull_request:
types:
- opened
- reopened
- synchronize
workflow_dispatch:
type: choice
env:
# Indicated the testing script, that it is performed on GitHub. In this
# case all test have to pass.
TESTING_GITHUB: 1
# Some tests require the 4C executable to check if the created input file
# works with 4C.
MESHPY_FOUR_C_EXE: /home_local/github-runner/testing_lib/four_c_master/release/4C
# Python executable and virtual environment name
PYTHON_EXE: python
PYTHON_VENV: python-testing-environment
# Meshpy can interact with cubitpy, to perform the corresponding tests,
# the following path has to be set
CUBIT_ROOT: /imcs/public/compsim/opt/cubit-15.2
# Meshpy allows to perform geometric search functions with ArborX,
# for this functionality we need Kokkos. On IMCS Ares this is provided
# with the spack installation under the following path.
SPACK_ACTIVATION_SCRIPT: /home_local/github-runner/testing_lib/spack/share/spack/setup-env.sh
jobs:
meshpy-testing:
name: meshpy-testing
runs-on: self-hosted
steps:
- uses: actions/checkout@v3
with:
submodules: true
- name: Test meshpy
run: |
whoami
pwd
MESHPY_PATH=$(pwd)
echo "Value of MESHPY_FOUR_C_EXE: \"${MESHPY_FOUR_C_EXE}\""
# Load the python interpreter
. ${SPACK_ACTIVATION_SCRIPT}
spack load [email protected]
# Create the virtual environment
${PYTHON_EXE} -m venv ${PYTHON_VENV}
source ${PYTHON_VENV}/bin/activate
# We have to update pip, otherwise there is a bug in the compilation of the cython code
pip install --upgrade pip
# Install meshpy
pip install .[CI-CD]
# Check git configuration
git config --list
# Print information on the python environment
python --version
pip list
# Run tests
cd tests
coverage run --rcfile=coverage.config testing_main.py
coverage html
coverage report
coverage-badge -o htmlcov/coverage.svg
# Check codestyle
cd $MESHPY_PATH
black . --check --exclude="${PYTHON_VENV}" && exit 0
# If we did not exit earlier, raise an error here
exit 1
ls
- name: Upload Test Results
if: failure()
uses: actions/upload-artifact@v3
with:
name: ${{github.job}}-${{github.run_number}}
path: ${{github.workspace}}/tests/testing-tmp/
geometric-search-testing:
name: geometric-search-testing
runs-on: self-hosted
steps:
- uses: actions/checkout@v3
with:
submodules: true
- name: Additionally test the geometric search module with ArborX
run: |
whoami
pwd
MESHPY_PATH=$(pwd)
# Load the external modules
git submodule update --init
# Load the required packages via spack
. ${SPACK_ACTIVATION_SCRIPT}
spack load [email protected]
spack load [email protected]
spack load [email protected]
spack load [email protected]
# Build the binaries
mkdir -p build/arborx
cd build/arborx
cmake ../../meshpy/geometric_search/src -G Ninja
ninja
cd $MESHPY_PATH
# Create the virtual environment
${PYTHON_EXE} -m venv ${PYTHON_VENV}
source ${PYTHON_VENV}/bin/activate
# We have to update pip, otherwise there is a bug in the compilation of the cython code
pip install --upgrade pip
# Install meshpy, it is not yet possible to globally install the ArborX binaries, so we
# use the developement mode here.
pip install -e .[CI-CD]
# Print information on the python environment
python --version
pip list
# Run tests
cd tests
python testing_geometric_search.py
performance-test:
name: meshpy-performance-testing
runs-on: self-hosted
continue-on-error: true
steps:
- uses: actions/checkout@v3
with:
submodules: true
- name: Performace test
run: |
whoami
pwd
MESHPY_PATH=$(pwd)
# Load the python interpreter
. ${SPACK_ACTIVATION_SCRIPT}
spack load [email protected]
# Create the virtual environment
${PYTHON_EXE} -m venv ${PYTHON_VENV}
source ${PYTHON_VENV}/bin/activate
# We have to update pip, otherwise there is a bug in the compilation of the cython code
pip install --upgrade pip
# Install meshpy
pip install .[CI-CD]
# Print information on the python environment
python --version
pip list
# Run performance tests
cd tests
python performance_testing.py