First set of WebUSB layout tests with Mojo service mocks.

First set of WebUSB layout tests with Mojo service mocks.

This patch adds a minimal set of layout tests for the WebUSB API that
work by registering mock implementations of the Mojo services the Blink
implementation depends on. Since these mocks are implemented using the
Javascript Mojo bindings they can live right next to the tests that
depend on them.

BUG=492204
Committed: https://crrev.com/a6f677309a7741a1b9d6f6c510ee8282bfbff8c8
Cr-Commit-Position: refs/heads/master@{#377429}

[Reilly Grant (use Gerrit), 2016-02-23 21:32:34]

reillyg@chromium.org changed reviewers:
+ rockot@chromium.org

[Reilly Grant (use Gerrit), 2016-02-23 21:32:34]

Please take a look.

[Ken Rockot(use gerrit already), 2016-02-23 23:12:49]

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/usb/resources/fake-devices.js (right):

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/fake-devices.js:7: resolve([
Will we never want to refer to fake devices by name for testing? i.e. what if I
just want to add a fake printer to the mock service? Do I have to know that
"fake printer" is index 0?

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/fake-devices.js:27:
configuration_name: "Printer Mode",
nit: please use consistent quotes, so either '' or "" everywhere, except when
breaking convention is convenient for nesting.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js (right):

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:3: const usbMocks =
new Promise(resolve => {
How about making usbMocks a function which takes a |mojo| object and returns the
define here.


function usbMocks(mojo) {
  return define('USB Mocks', ... );
}

Then you don't need to manually import serviceRegistry, core, router.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:8:
'content/public/renderer/service_provider',
nit: I suppose we should keep these sorted

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:356:
usbMocks.then(usb => {
and then you can just call:

usbMocks(mojo).then(usb => {...


also, you want to return this promise from the mojo_test function... in fact, it
should probably just be:


mojo_test(mojo => usbMocks(mojo).then(usb => {
  // ...
}));


in case it's not obvious, () => { foo } returns undefined whereas () => foo
returns the value of foo.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:366: return result;
nit: You can return result.then(cleanUp, cleanUp)

[Reilly Grant (use Gerrit), 2016-02-24 20:18:47]

Please take a look at this as we discussed the tests hang for an unknown reason.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/usb/resources/fake-devices.js (right):

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/fake-devices.js:27:
configuration_name: "Printer Mode",
On 2016/02/23 23:12:48, Ken Rockot wrote:
> nit: please use consistent quotes, so either '' or "" everywhere, except when
> breaking convention is convenient for nesting.

Done.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js (right):

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:3: const usbMocks =
new Promise(resolve => {
On 2016/02/23 23:12:49, Ken Rockot wrote:
> How about making usbMocks a function which takes a |mojo| object and returns
the
> define here.
> 
> 
> function usbMocks(mojo) {
>   return define('USB Mocks', ... );
> }
> 
> Then you don't need to manually import serviceRegistry, core, router.

Done.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:8:
'content/public/renderer/service_provider',
On 2016/02/23 23:12:48, Ken Rockot wrote:
> nit: I suppose we should keep these sorted

No longer an issue now that it takes mojo as a parameter.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:356:
usbMocks.then(usb => {
On 2016/02/23 23:12:48, Ken Rockot wrote:
> and then you can just call:
> 
> usbMocks(mojo).then(usb => {...
> 
> 
> also, you want to return this promise from the mojo_test function... in fact,
it
> should probably just be:
> 
> 
> mojo_test(mojo => usbMocks(mojo).then(usb => {
>   // ...
> }));
> 
> 
> in case it's not obvious, () => { foo } returns undefined whereas () => foo
> returns the value of foo.

Done.

https://codereview.chromium.org/1726943002/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/usb/resources/usb-helpers.js:366: return result;
On 2016/02/23 23:12:49, Ken Rockot wrote:
> nit: You can return result.then(cleanUp, cleanUp)

Done.

[Reilly Grant (use Gerrit), 2016-02-24 22:23:16]

I think this fixes it.

[Ken Rockot(use gerrit already), 2016-02-24 22:26:25]

lgtm

[Reilly Grant (use Gerrit), 2016-02-24 22:55:28]

The CQ bit was checked by reillyg@chromium.org

[commit-bot: I haz the power, 2016-02-24 22:56:29]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1726943002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1726943002/40001

[commit-bot: I haz the power, 2016-02-25 00:10:56]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-02-25 00:12:28]

Description was changed from

==========
First set of WebUSB layout tests with Mojo service mocks.

This patch adds a minimal set of layout tests for the WebUSB API that
work by registering mock implementations of the Mojo services the Blink
implementation depends on. Since these mocks are implemented using the
Javascript Mojo bindings they can live right next to the tests that
depend on them.

BUG=492204
==========

to

==========
First set of WebUSB layout tests with Mojo service mocks.

This patch adds a minimal set of layout tests for the WebUSB API that
work by registering mock implementations of the Mojo services the Blink
implementation depends on. Since these mocks are implemented using the
Javascript Mojo bindings they can live right next to the tests that
depend on them.

BUG=492204

Committed: https://crrev.com/a6f677309a7741a1b9d6f6c510ee8282bfbff8c8
Cr-Commit-Position: refs/heads/master@{#377429}
==========

[commit-bot: I haz the power, 2016-02-25 00:12:29]

Patchset 3 (id:??) landed as
https://crrev.com/a6f677309a7741a1b9d6f6c510ee8282bfbff8c8
Cr-Commit-Position: refs/heads/master@{#377429}

[Dirk Pranke, 2016-02-25 05:47:59]

A revert of this CL (patchset #3 id:40001) has been created in
https://codereview.chromium.org/1734753003/ by dpranke@chromium.org.

The reason for reverting is: Odd, I'm not sure what happened, but these layout
tests appear to have passed the trybots and crashed on the waterfall bots:

https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7/builds/39956
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...
.

[Reilly Grant (use Gerrit), 2016-02-26 22:14:38]

On 2016/02/25 05:47:59, Dirk Pranke wrote:
> A revert of this CL (patchset #3 id:40001) has been created in
> https://codereview.chromium.org/1734753003/ by mailto:dpranke@chromium.org.
> 
> The reason for reverting is: Odd, I'm not sure what happened, but these layout
> tests appear to have passed the trybots and crashed on the waterfall bots:
> 
>
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7/builds/39956
>
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...

Do these waterfall bots have equivalent trybots I could run?