Pycrop2ml

A Python library to generate components from Crop2ML declarative language.

For more information, refer to the documentation.

Installation

PyCropML relies on two Python packages to build the diagrams of a Crop2ML composition: pydot, a pure-Python library that describes the graph, and graphviz, which only talks to the separate Graphviz dot executable that actually renders it as an image. Both Python packages install automatically with PyCropML below, but the dot executable itself is a system-level dependency that pip cannot install for you. Without it, cyml -p ... fails as soon as a package links more than one ModelUnit, with:

graphviz.backend.execute.ExecutableNotFound: failed to execute PosixPath('dot')

Using conda

The conda-forge channel ships the Graphviz executables alongside the Python packages, so this route needs nothing extra:

conda install -c amei -c openalea3 -c conda-forge pycropml

Using pip

PyCropML is not yet published on PyPI, so install it from a source checkout. Install the Graphviz executables first, since pip cannot:

conda install -c conda-forge graphviz    # inside any conda environment
# or, on Debian/Ubuntu:
sudo apt install graphviz

Then install PyCropML itself:

git clone https://github.com/AgriculturalModelExchangeInitiative/PyCrop2ML.git
cd PyCrop2ML
pip install .

Developer installation

To work on PyCropML itself — running its test suite, building its documentation, or modifying its generators — install it in editable mode from the same source checkout, after installing Graphviz as above:

git clone https://github.com/AgriculturalModelExchangeInitiative/PyCrop2ML.git
cd PyCrop2ML
pip install -e ".[test,doc]"

-e (editable) makes changes under src/pycropml take effect immediately, with no reinstall step in between — the workflow used throughout PyCropML's own test suite. [test,doc] pulls in pytest and the Sphinx toolchain declared as optional dependencies in pyproject.toml; drop the extras (pip install -e .) for a plain editable install.

Usage

From Platforms to Crop2ML

cyml -c name_of_the_component_repository absolute_path_of_the_output source_language_or_platform

From Crop2ML to Platforms

cyml -p name_of_the_Crop2ML_package target_language_or_platform

Testing

Run the test suite using pytest or unittest:

Using pytest

# Run all tests
python -m pytest test/cyml/test_cyml_operations.py -v

# Run a specific test class
python -m pytest test/cyml/test_cyml_operations.py::TestCyMLOperations -v

# Run a single test
python -m pytest test/cyml/test_cyml_operations.py::TestCyMLOperations::test_modulo_operation -v

Using unittest

cd test/cyml
python -m unittest test_cyml_operations