- Understand how to do integration testing in the context of Flask applications.
In sprint B we created two views: one to serve epithets and one to serve the vocabulary used to generate epithets. Application logic to generate these views was abstracted out into three helper classes that were then unit tested to ensure each helper class behaves as expected. However, at this point we have not ensured our application is actually functional. Assuming all unit tests pass, we know components work in isolation but we do not know if each component is working together. This difference, between testing if components work in isolation vs in cooperation with each other, is the difference between unit and integration testing.
In order to serve epithets, the file loader must read the .json file and parse it into a format that we can use. Then, the EpithetGenerator must select a random word from each column of the vocabulary and return a templated epithet. Finally, Flask must serve generated epithets for each incoming request. Where unit tests verify each step in this series is functional, we need to know if all steps of the series work together to serve epithets.
In this assignment we will continue where sprint B concluded by:
- adding a route to serve a random number of epithets.
- creating integration tests to verify each route serves its expected payload.
- Watch Web API Development with Flask: Using the Flask Test Client
- Read Flask-Testing's documentation.
- Add one route that serves a random number of epithets.
- Create integration tests using the Flask-Testing extension to verify each route serves its expected payload.