Active Development. This product is in active development; breaking changes as well as major refactors should be expected. As with all software, make sure the project works for you before using it.
Browser based Gherkin Testing Made Easy
VeriPy is a testing framework that can test a web-based system programmatically and output a report for inclusion in validation documentation. The purpose of VeriPy is to help automate the validation process encountered in clinical systems, especially regressive validations. VeriPy can reduce the time it takes for groups with regulatory requirements to introduce new software versions into their pipelines/workflows.
Installing VeriPy is really simple but has a few more steps than most Python projects.
- Clone this repo.
- If you're using a Virtual Environment (which you should) then create it with a Python version of 3.6 or higher.
- Install the requirements
# Development
pip install -e veripy[develop]
#Production
pip install veripy- Set up your preferred browser driver (default is chrome). Use these instructions
# For macOS
brew cask install chromedriver
# For Linux
cd $HOME/Downloads
wget https://chromedriver.storage.googleapis.com/2.37/chromedriver_linux64.zip
unzip chromedriver_linux64.zip
mkdir -p $HOME/bin
mv chromedriver $HOME/bin
echo "export PATH=$PATH:$HOME/bin" >> $HOME/.bash_profileFor more information, refer to the Splinter Installation Instructions
Behave is configured using environment variables. Please refer to the settings.py file for full documentation but the required variables are listed below. It is recommended that you create a file called environment.sh with the following contents and source that file before running VeriPy.
cat > environment.sh << EOF
export ENVIRONMENT=dev
EOF
source environment.shVeriPy is built on Behave. As such any files ending in .feature inside of the features/ directory will be run when the application starts. VeriPy comes with a sample set of tests demonstrating how to use the statements.
Veripy will also look for setup fixtures in the setup/ directory (or the directory configured in settings.py) for features that are tagged as @configure.NAME, where name is the setup's name. When features are run, veripy will inspect the tags of each feature for either the tags @fixture.setup.NAME or @fixture.setup.teardown.NAME, and run the matching setup/ feature with the same configured name. Using @fixture.setup.NAME will NOT run any of the setup's scenarios that are tagged as @teardown, while @fixture.setup.teardown.NAME will add a cleanup callback with the scenarios tagged as @teardown. This allows veripy to set up a through the web data set for testing, and clean it up afterwards. If multiple features rely on the same setup, it will only re-run the setup if the teardown has occurred. This allows multiple features to use the system under test with specific data that may be cumbersome to add throughout the testing lifecycle. @fixture.setup tags are run in the order they appear.
Also, veripy has a default browser configured so that any feature/scenario tagged with @fixture.browser.chrome will open a browser for testing, and close it at the end of the feature/scenario, respectively.
To run these or any custom tests, use the following command:
behave example/ --tags ~@xfailTo identify where sentences are used (and which are undefined), use the dry-run mode to output that data to step_usage.txt:
behave example/ --dry-runVeriPy uses Sphinx for documentation. Once you've installed the dependencies, simply run the following command to generate the docs.
make -C docs htmlBehave can also add step documentation. To generate, run the following command
behave -f sphinx.steps -o docs/step_definitions --dry-run -qYou should now see a _build directory. Either open index.html in a browser or serve the directory directly using the following command.
python -m http.server --directory docs/_build/htmlYou should now be able to navigate to http://localhost:8000/ and see the documentation.