An API for validating survey schemas.
It is recommended that you use Pyenv to manage your Python installations.
curl -sSL https://install.python-poetry.org | python3 -
poetry install
To run the app:
make run
If you want to run the app locally using multiple server workers set reload to "False" in the Uvicorn settings in api.py:
uvicorn.run("api:app", workers=20, port=5001, reload=False)The validator can be called directly in the browser using the "/validate" endpoint and the "url" parameter for the address where the schema is located (eg. GitHub Gist raw json):
http://localhost:5001/validate?url=...
By default, all schemas in the tests/schemas/valid directory will be evaluated as part of the unit tests.
Any errors in these schemas will cause a failure.
To run the app's unit tests:
make test-unit
To run the app's unit tests and ajv validator tests:
make test
To test the apps functionality:
make run
Then, in another terminal window/tab, navigate to a checked out copy of eq-survey-runner:
make test
In the eq-schema-validator directory, install node version manager (nvm) and node using the following commands:
brew install nvm
nvm install
Install node dependencies:
npm install
To run the ajv validator tests:
make test-ajv
Run the following to format all json files in the schemas directory:
make format
Also included is a node based version of the json schema validation which may be used during development to assist with debugging errors. This returns more errors than we'd currently like due to the way polymorphism works for each of our blocks.
Run the ajv (server) based version of validator.
make start-ajv
To stop the ajv (server) based version of validator.
make stop-ajv
This returns either an empty json response when the questionnaire is valid, or a response containing an "errors" key. The errors are ordered by their path length and with first error message being the deepest path into the schema and should represent the best match for the questionnaire which has been posted.