Test suite for MagicMirror²

roramirez

Hi,

Today I’ve working a problem of issue with Electron and other things, all details here
https://github.com/MichMich/MagicMirror/issues/577

After testing in two enviroment different for Lite and with Pixel on Raspbian, the process is long because you need the create SD and follow the instruction and check what could be the problem.

I think this can be automatic and running on a hosted system or my own machine. Someone know howto use a arm image for RaspberryPi and run like a Docker’s container? for run a test of enviroments. I think can be a help to include in the Test Suite.

morozgrafix

@roramirez I haven’t played with Docker in a while and never tried raspbian on a docker container. Quickly glanced at DockerHub and it seems that people are able to do that. https://hub.docker.com/search/?isAutomated=1&isOfficial=0&page=1&pullCount=0&q=raspbian&starCount=0
I may take a look into it, but it all depends on my availability.

roramirez

Hi, I recently pushed a Pull Request with two e2e testing.
https://github.com/MichMich/MagicMirror/pull/669

These use their own configurations like @qistoph mention and after added the MM_CONFIG_FILE variable for MagicMirror.

It’s really easy create new test for check a different part of MagicMirror and their modules.

strawberry 3.141

@roramirez what is the reason for a 10 sec timeout in each test? If there are a many tests this will sum up

roramirez

@strawberry-3.141 said in Test suite for MagicMirror²:

what is the reason for a 10 sec timeout in each test? If there are a many tests this will sum up

It’s does not accumulate. It’s a timeout, the maximun time for the test. You can set different timeout for every test.

morozgrafix

I haven’t looked thoroughly at the code and I wanted to suggest setting a default timeout that can be overwritten in individual tests as needed. It may make test code a little bit cleaner and less repetitive.

Good job getting it going!

roramirez

@morozgrafix Yes, I knowed is not more cleaner and repetitive but was the first proof of concept ;)

Good idea set default timeout. I added two new task in Trello board about you mentioned.

morozgrafix

@roramirez I’ve played a little with test suite and added basic test for clock module. Then I realized that if we want to test different config options for a given module we may need to have to create multiple configs, which may be challenging with current tests directory organization.
Also I believe that process.env.MM_CONFIG_FILE = "tests/confs/config_name.js" line needs to be moved into before step (seems to work there or it can go into beforeEach if needed). Otherwise when running npm run test:e2e first config that gets picked up seems to persist throughout all of the tests and always used by the app.js.

I was wondering if structure similar to this makes sense:

tests
├── configs
│   ├── env.js
│   └── modules
│   │   └── clock
│   │   │   └── clock_12hr.js
│   │   │   └── clock_24hr.js
│   │   │   └── clock_analog.js
│   │   └── helloworld
│   │       └── helloworld.js
├── e2e
│   ├── env_spec.js
│   └── modules
│       ├── clock_spec.js
│       └── helloworld_spec.js
├── functions
│   └── compare-version_spec.js
└── global_vars
    └── root_path.js

but this would involve changing app.use("/tests/confs", express.static(path.resolve(global.root_path + "/tests/confs"))); in the server.js to be somewhat dynamic and I’m not very familiar with express on how it can be done. I think that’s easily solved by just changing it to app.use("/tests/configs", express.static(path.resolve(global.root_path + "/tests/configs")));

roramirez

@morozgrafix Seem good sense going to a refactor of the structure.

I think we can take two way acord you mentioned.

1.- Include your test for clock module with new format for name and respective directories
2.- Do it the refactor to all tests that remain.

I really like see how you resolve the multiples instances configuration in a one tests file.

morozgrafix

@roramirez cool thanks. I have it all ready on my fork. Will submit PR for review.