Contribute

Overall guidance on contributing to a PyAnsys library appears in the Contributing topic in the PyAnsys Developer’s Guide. Ensure that you are thoroughly familiar with this guide before attempting to contribute to PySystemCoupling.

The following contribution information is specific to PySystemCoupling.

Clone the repository

To clone and install the latest PySystemCoupling release in development mode, run:

git clone https://github.com/pyansys/pysystem-coupling.git
cd pysystem-coupling
python -m pip install --upgrade pip
pip install -e .

Build the generated API code

In the packaged version of PySystemCoupling, Python classes are generated during the package build to provide an API to System Coupling facilities. The classes are largely generated from metadata queried from a running instance of System Coupling.

For local development, you must perform this additional generation step manually, after the preceding steps for cloning and installing the package.

pip install -e .[classesgen]
python scripts/generate_datamodel.py

The generated code is written to the directory src/ansys/systemcoupling/core/adaptor/api_<version>, where <version> is the version of the System Coupling instance that was run in the background by the generation script. For example, the version 23_1 corresponds to the System Coupling 2023 R1. The default is 23_1, which means that this release of System Coupling is expected to be at the installation location given by the AWP_ROOT231 environment variable.

You can override the default behavior and run a different version, and generate the API classes for this different version, by setting either the SYSC_ROOT environment variable to point to the root directory of your System Coupling installation or the AWP_ROOT environment variable to point to the root of an Ansys installation. If SYSC_ROOT and AWP_ROOT environment variables are both set, the former takes priority. Additionally, both of these environment variables take priority over the AWP_ROOT231 environment variable.

Build documentation

To build the PySystemCoupling documentation locally, you must have first generated the API classes as described in Build the generated API code. This is because some of the documentation is extracted from these classes.

Because multiple versions of the API classes can exist, you must set the PYSYC_DOC_BUILD_VERSION environment variable to tell the documentation build which version to use. Given that there is no default for this environment variable, you must set it. The value should be a string in the same form as the <version> component of the api_<version> directory. For example, “23_1”.

With this variable set, run this code to build the documentation:

pip install -e .[doc]
cd doc
make html

After the build completes, the HTML documentation is located in the _builds/html directory. You can load the index.html file in this directory into a web browser.

You can clear all HTML files from the _builds/html directory with:

make clean

Run Sphinx Gallery examples

The Sphinx Gallery examples are not run as part of a documentation build by default. This is because realistic runs of System Coupling, involving both System Coupling itself and the participant solvers, are not currently possible on GitHub. Therefore, examples are run manually from time to time and the resulting Sphinx files are committed to the repository.

To override the default behavior and rebuild the entire documentation, including regeneration of the Sphinx Gallery` examples, set the PYSYC_BUILD_SPHINX_GALLERY environment variable. Because only its existence is examined, you can give any value to this environment variable.

Post issues

Use the PySystemCoupling Issues page to submit questions, report bugs, and request new features.

Adhere to code style

PySystemCoupling is compliant with the PyAnsys code style. It uses the tool pre-commit to check the code style. You can install and activate this tool with:

python -m pip install pre-commit
pre-commit install

When pre-commit is active, it automatically runs style checks on every git commit to your branch.

Alternatively, you can directly run pre-commit at any time with this command:

pre-commit run --all-files --show-diff-on-failure