660619 - Silence a warning during tests

Status: RESOLVED FIXED

(DevTools :: General, defect)

Description

[Panos Astithas (he/him) [:past] (please ni?)]

Attachment: Simpl

Comment 1

[Panos Astithas (he/him) [:past] (please ni?)]

There is an annoying warning that keeps recurring when running tests:

JS Component Loader: WARNING chrome://mochitests/content/browser/toolkit/components/console/hudservice/tests/browser/head.js:46
                     anonymous function does not always return a value

This is a simple fix for getting rid of the warning, although I would prefer avoiding the try/catch altogether. I suppose there must have been a reason for putting it there in the first place, though, so this patch leaves it in place.

And yes, pressing ENTER in the bug submission form, occasionally posts unfinished comments. Thanks for asking.

Comment 2

[Panos Astithas (he/him) [:past] (please ni?)]

Comment on attachment 536062 [details] [diff] [review]
Simpl

This is ddahl's code AFAICT.

Comment 3

[:Gavin Sharp [email: [e-mail]]]

Comment on attachment 536062 [details] [diff] [review]
Simpl

I don't think this try/catch is useful, "return HUDService" isn't actually going to throw if the import succeeded. I'd say just get rid of it. Actually, the entire lazy getter isn't actually useful given that pretty much all tests use it eagerly (and the perf difference is negligible anyways), so unless you somehow care about "ConsoleUtils" polluting the scope (seems unlikely), just remove it entirely and import() HUDService.jsm unconditionally.

Comment 4

[Panos Astithas (he/him) [:past] (please ni?)]

Attachment: [in-devtools][checked-in] Unconditional import

Comment 5

[:Gavin Sharp [email: [e-mail]]]

Comment on attachment 536246 [details] [diff] [review]
[in-devtools][checked-in] Unconditional import

We really should get a bug on file about removing all these imports if there isn't one already. Test code shouldn't be re-importing all of these things - if they're really needed they're already imported by the chrome window the test is running in. I guess bug 610953 could cover that too.

Comment 6

[Panos Astithas (he/him) [:past] (please ni?)]

(In reply to comment #5)
> Comment on attachment 536246 [details] [diff] [review] [review]
> Unconditional import
> 
> We really should get a bug on file about removing all these imports if there
> isn't one already. Test code shouldn't be re-importing all of these things -
> if they're really needed they're already imported by the chrome window the
> test is running in. I guess bug 610953 could cover that too.

Thanks, added a note to that effect in bug 610953.

Comment 7

[Rob Campbell [:rc] (:robcee)]

Comment on attachment 536246 [details] [diff] [review]
[in-devtools][checked-in] Unconditional import

http://hg.mozilla.org/projects/devtools/rev/c4bd2219aafc

Comment 8

[Rob Campbell [:rc] (:robcee)]

whups, that's not the right link. I should refresh tbpl before grabbing the checkin url.

http://hg.mozilla.org/projects/devtools/rev/b9ceaebb878f

Comment 9

[Dave Camp (:dcamp)]

Comment on attachment 536246 [details] [diff] [review]
[in-devtools][checked-in] Unconditional import

http://hg.mozilla.org/mozilla-central/rev/b9ceaebb878f