The UI was updated so that the document labels were on the same tab as the images to make image annotation easier. The eve container was also updated to install MongoDB from a .tar.gz installer instead of a package. This is because the packages now require systemctl to be installed and it is not present in a Ubuntu container.
██████╗ ██╗███╗ ██╗███████╗
██╔══██╗██║████╗ ██║██╔════╝
██████╔╝██║██╔██╗ ██║█████╗
██╔═══╝ ██║██║╚██╗██║██╔══╝
██║ ██║██║ ╚████║███████╗
╚═╝ ╚═╝╚═╝ ╚═══╝╚══════╝
Pmap Interface for Nlp Experimentation
© 2019 The Johns Hopkins University Applied Physics Laboratory LLC.
Cypress
The GUI test cases are written using Cypress (https://www.cypress.io/).
The actual test cases can be seen in tests/cypress/integration/.
PyTest
The python unit test cases are writting using PyTest (https://docs.pytest.org/).
The actual test cases can be seen in tests/pytest/.
Running in developer environment
pipenv install --dev
pushd tests
npm install
Then, when the dev stack is running:
./open_with_dev_stack.sh
This will both run pytest and open the cypress dashboard.
Note that running with the dev stack is noticably slower than running with the docker-compose stack.
You should also be sure to freshly import the testing data before you run the tests
(./setup_test_data.sh), as the tests may add data to the database that will mess up future runs.
Running with docker
To run with docker, a standalone container is built with the cypress code in it.
To build the docker container(s), run
./build.sh
This will build the entire docker-compose stack and import the test data into a test database (not the same one as running normal docker-compose).
To run the docker-compose stack and then the standalone cypress container, you can then run
./run_docker_compose.sh <arguments>
For usage instructions, run the script with no arguments.
It is also possible to run the cypress container directly, as long as you set the following environment variables:
CYPRESS_BASE_URL: the URI to the frontendCYPRESS_API_URL: the URI to the backend