- The Recipe Project would have github actions which run the tests every time code is pushed to the repository. This is the best way for testing given the three options. Running tests locally after changes are made is essentially performing them with github actions except github actions automates the testing compared to manually running them. Running the tests at the end is not feasible as if a problem araises early in development, it will not be caught and will be deeply embedded in the code. Automated and even manual testing is a better alternative than to test only when development is finished. Automated testing is better than manual in the sense that it can be run as a github action so the developer does not need to consciously and constantly remember to test.
- No, that would be a unit test. An end to end test is for UI
- No since the writing and sending of the message would be considered multiple components.
- Yes since checking maximum length is a single component and will not have any effect on other components.
Here is the image of the tests passing:
