Hi! This is what will make Expo never break ever. 🙏
If you have problems with the code in this repository, please file issues & bug reports at https://github.com/expo/expo. Thanks!
Just run this on XDE or exp and point Expo to the URL. You can navigate to <theurl>/+<regexp>
to only run tests whose names match the regexp. The tests are in tests/
and their name usually is the same as the filename.
There is also a published version of this at exp://exp.host/@nikki/test-suite. So you can go to, for example, exp://exp.host/@nikki/test-suite/+Assets.* to run the asset tests!
Add a file in tests/
, or, if your test falls into one of the categories already there, edit one of the files there. The name
export of the module defines the test name. The module must export a function test
that takes an argument t
(an object providing the Jasmine interface) and sets up the Jasmine suites and specs. Read http://jasmine.github.io/2.4/introduction.html to learn how to use Jasmine. Functions such as describe
, it
, expect
, etc. are available as t.describe
, t.it
, t.expect
, etc. where t
is the argument passed to test
. Check out one of the tests already in tests/
to get an idea of how stuff is done.
If you add a new file under tests/
you must add it to the testModules
list in index.js
to have it be registered.
Make sure to go over the Jasmine documentation at http://jasmine.github.io/2.4/introduction.html to get an idea of all the functionality supported. In addition to what Jasmine offers, this app patches the Jasmine functions so they support async
functions as well. Asynchronous exceptions are properly caught and displayed too. See tests/Contacts.js
for an example of using async/await
in a test context. This is important for us because basically most of the SDK's functionality is asynchronous.
One neat thing is that you can focus to certain subtrees of test using f.describe
or f.it
and the other tests are skipped. If you do this on a local instance of the app code and send the link to someone, they will run that test. Then you can change the test and have them refresh to see the changes. This way you can debug things on someone else's device, isolate the issue and maybe even fix it live (if it is a pure JS bug).