wps-tools

Overview

The Web Processing Services (WPS) at PCIC, known as birds, are developed and tested using many common functions. The wps-tools repository was created to store these functions in a single place, mitigating redundant code across the birds, and to simplify the creation and testing of new WPS processes. Currently, this package is used in all the PCIC birds:

• thunderbird
• osprey
• sandpiper
• chickadee
• quail

Structure

io.py

This module contains a collection of commonly used PyWPS inputs and outputs and are also mainly used in wps_*.py files.

testing.py

These functions help run a bird's pytest suite and are thus mainly used in test_*.py files.

utils.py

These functions gather PyWPS inputs, build PyWPS outputs, and handle logging information within a process's class file and are thus mainly used in wps_*.py files.

Installation as GitHub Repository

Clone the repo onto the target machine. Python installation should be done using the poetry tool. The required dependencies can be installed using

poetry install

The additional R dependencies can be installed using

poetry install --extras "r"

This also creates a virtual environment that can be activated using

poetry shell

Installation and Usage as Package

While in a different repository, the wps-tools package can be installed by executing

pip install -i https://pypi.pacifcclimate.org/simple wps-tools[complete]

Afterwards, each function can be used in a .py file by importing it. For example, if one wishes to use the log_handler function from utils.py, they access it by writing

from wps_tools.utils import log_handler

Development

Once the repository is cloned and the required development packages are installed, one can add new functions or PyWPS i/o objects by writing them in one of the aforementioned three modules or a new one in the wps_tools directory. Each function should have a corresponding test in the tests directory.

Testing

More Python packages are required to run the tests and they can be installed by executing

poetry install --extras "complete"

The entire test suite can then be run by executing

poetry run pytest

and one can specify desired test functions as optional arguments. For example, one can run test_is_opendap_url from test_utils.py by executing

poetry run pytest tests/test_utils.py::test_is_opendap_url

Releasing

To create a versioned release:

1. Increment version in pyproject.toml
2. Summarize the changes from the last release in NEWS.md
3. Commit these changes, then tag the release:

git add pyproject.toml NEWS.md
git commit -m"Bump to version x.x.x"
git tag -a -m"x.x.x" x.x.x
git push --follow-tags

4. Github Actions will automatically build and publish the package to our pypi server