Emergency call manually sets document's lang/dir and uses 'localized' event to initialize. Instead it should use mozL10n.once.

Comment on attachment 8409880 [details] pull request Sorry, I'm only reviewer of the build system in Gaia. Tim, it looks like you already reviewed some code in emergency.

Comment on attachment 8409880 [details] pull request I am setting r? flag to the original author. Do we have a document on use cases of API methods for mozL10n?

No, not yet. We first need to clean the API (bug 993188), and that requires us to clean up current use cases.

Comment on attachment 8409880 [details] pull request Tried on real device and have no errors and problems, and the tests are all passed, so there is the r+.

Not directly related to the bug, but... (In reply to Tim Guan-tin Chien [:timdream] (MoCo-TPE) (please ni?) from comment #3) > Do we have a document on use cases of API methods for mozL10n? I spent some time today putting together the following docs: https://developer.mozilla.org/en-US/docs/Web/API/Navigator.mozL10n https://developer.mozilla.org/en-US/docs/Web/API/L10n It's a work in progress, but should be a good start!

Master: https://github.com/mozilla-b2g/gaia/commit/b64555b9ed961923b923239e96f87d6ba83cf951

I had to back out this patch for three reasons, sorry. https://github.com/mozilla-b2g/gaia/commit/49a096509f5ea53dbb18b7e0a51d1018900da499 1) It should be reviewed by a peer of the Dialer component. 2) It's missing new tests. 3) It's using function arrows for no reason.

(In reply to Anthony Ricaud (:rik) from comment #8) > 2) It's missing new tests. We have tests for .once in l10n testsuite. Can you help me identify how you'd like the test for .once in the Emergency Call app to look like? > 3) It's using function arrows for no reason. Are arrow functions discouraged? I thought they will be encouraged over anonymous functions.

Tests: We need to test that we wait for a mock of navigator.mozL10n to fire once before removing the hidden class. Arrow function: I'm not sure what the general policy is for Gaia in general but in this case, there is no need for one since we don't use this.

(In reply to Anthony Ricaud (:rik) (out until May 13th) from comment #10) > Tests: We need to test that we wait for a mock of navigator.mozL10n to fire > once before removing the hidden class. Ok, I added this test. Let me know if that looks like the test you want to have here. (I find testing bootstrapping code non-trivial)

Comment on attachment 8416370 [details] pull request with a test The logic of the test looks ok (and I agree with you that keypad.js is currently annoying to test). I've left comments on the pull request to use mocksHelper and move the mocks to shared. I also think we should use sinon but that's up to you.

Comment on attachment 8416370 [details] pull request with a test Two very very small nits on github. Otherwise, good to go!

Thanks for help! Patch: https://github.com/mozilla-b2g/gaia/commit/a04cecbabe0af14c50fd275e3a3aca223e07a08c Merge: https://github.com/mozilla-b2g/gaia/commit/356a468972cb2e29961aeb960c9adfa080651f9e