nomad-parser-dmft

A compilation of NOMAD parsers for input/output files of DMFT codes.

This nomad plugin was generated with Cookiecutter along with @nomad's cookiecutter-nomad-plugin template.

Development

If you want to develop locally this plugin, clone the project and in the plugin folder, create a virtual environment (you can use Python 3.9, 3.10, or 3.11):

git clone https://github.com/JosePizarro3/nomad-parser-dmft.git
cd nomad-parser-dmft
python3.11 -m venv .pyenv
. .pyenv/bin/activate

Make sure to have pip upgraded:

pip install --upgrade pip

We recommend installing uv for fast pip installation of the packages:

pip install uv

Install the nomad-lab package (the flag -e ensures that you can develop the package by installing it in editable mode):

uv pip install -e '.[dev]'

Note! If you do not use uv and until we have an official pypi NOMAD release with the plugins functionality make sure to include NOMAD's internal package registry (via --index-url in the above command). The plugin is still under development.

pip install -e '.[dev]' --index-url https://gitlab.mpcdf.mpg.de/api/v4/projects/2187/packages/pypi/simple

Run the tests

You can run locally the tests:

python -m pytest -sv tests

where the -s and -v options toggle the output verbosity.

Our CI/CD pipeline produces a more comprehensive test report using the pytest-cov package. You can generate a local coverage report:

uv pip install pytest-cov
python -m pytest --cov=src tests

Run linting and auto-formatting

We use Ruff for linting and formatting the code. Ruff auto-formatting is also a part of the GitHub workflow actions. You can run locally:

ruff check .
ruff format . --check

Debugging

For interactive debugging of the tests, use pytest with the --pdb flag. We recommend using an IDE for debugging, e.g., VSCode. If that is the case, add the following snippet to your .vscode/launch.json:

{
  "configurations": [
      {
        "name": "<descriptive tag>",
        "type": "debugpy",
        "request": "launch",
        "cwd": "${workspaceFolder}",
        "program": "${workspaceFolder}/.pyenv/bin/pytest",
        "justMyCode": true,
        "env": {
            "_PYTEST_RAISE": "1"
        },
        "args": [
            "-sv",
            "--pdb",
            "<path-to-plugin-tests>",
        ]
    }
  ]
}

where <path-to-plugin-tests> must be changed to the local path to the test module to be debugged.

The settings configuration file .vscode/settings.json automatically applies the linting and formatting upon saving the modified file.

Documentation on Github pages

To view the documentation locally, install the related packages using:

uv pip install -r requirements_docs.txt

Run the documentation server:

mkdocs serve

Adding this plugin to NOMAD

Currently, NOMAD has two distinct flavors that are relevant depending on your role as an user:

1. A NOMAD Oasis: any user with a NOMAD Oasis instance.
2. Local NOMAD installation and the source code of NOMAD: internal developers.

Adding this plugin in your NOMAD Oasis

Read the NOMAD plugin documentation for all details on how to deploy the plugin on your NOMAD instance.

Adding this plugin in your local NOMAD installation and the source code of NOMAD

Modify the text file under /nomad/default_plugins.txt and add:

<other-content-in-default_plugins.txt>
nomad-parser-dmft==x.y.z

where x.y.z represents the released version of this plugin.

Then, go to your NOMAD folder, activate your NOMAD virtual environment and run:

deactivate
cd <route-to-NOMAD-folder>/nomad
source .pyenv/bin/activate
./scripts/setup_dev_env.sh

Alternatively and only valid for your local NOMAD installation, you can modify nomad.yaml to include this plugin, see NOMAD Oasis - Install plugins.

Build the python package

The pyproject.toml file contains everything that is necessary to turn the project into a pip installable python package. Run the python build tool to create a package distribution:

pip install build
python -m build --sdist

You can install the package with pip:

pip install dist/nomad-parser-dmft-0.1.0

Read more about python packages, pyproject.toml, and how to upload packages to PyPI on the PyPI documentation.

Template update

We use cruft to update the project based on template changes. A cruft-update.yml is included in Github workflows to automatically check for updates and create pull requests to apply updates. Follow the instructions on how to enable Github Actions to create pull requests.

To run the check for updates locally, follow the instructions on cruft website.

Main contributors

Name	E-mail
Jose Pizarro	[email protected]