Expand and improve our e2e (headless) tests for Wasp
Martinsos opened this issue · comments
While we have e2e tests for wasp right now, they are not enough. They are not testing everything, and they are not super well organized, in the sense that adding new e2e tests doesn't have a clear process.
What we want to do is have a minimal set of test Wasp apps that all together, with their e2e tests, provide a big enough coverage of Wasp. On one hand we don't want to have too many of these apps, as little as possible, but on the other hand having one huge app is also hard to maintain, so we need to find the right spot. And we need multiple apps to be able to test different variations (unless we find a way to somehow generate those variations from one app or its skeleton?).
Then, we should introduce a rule of having to add a new e2e test when new API is added! Such PR should also include e2e tests.
Good start might be to review the state of our current e2e test apps (both headless and not headless, but headless are probably more important here: https://github.com/wasp-lang/wasp/tree/main/waspc/headless-test) and figure out what is being tested currently and what is not being tested. Then, we want to expand from there until we have good coverage.
What about waspc/examples? Should we combine those with apps that are headless-tested? So we have one set of apps, they are covered with tests, and that is it? Why would we have both waspc/examples and also separate apps that are headless tested?
Also covered in #2024
Ok there is a purpose though to having apps in waspc/examples, because they were easy to run and develop with. But hopefully we can preserve that while also having them covered with tests? If we can, awesome. If not, then let's figure out something better.