Upstream service worker message event tests to WPT

Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116, 702352, 702361
R=falken@chromium.org

Review-Url: https://codereview.chromium.org/2751113005
Cr-Commit-Position: refs/heads/master@{#459099}
Committed: https://chromium.googlesource.com/chromium/src/+/7d998666c33c0eaaf9c6b2fa874644c1c03f31fc

[mike3, 2017-03-16 21:49:37]

Would you mind taking a look, Matt?

[falken, 2017-03-17 03:34:43]

I like what you've done to the tests, though I haven't fully examined the diff
yet.

The enable/disable test juggling should not be needed thanks to how the test
runner handles failed WPT tests: we just to put expected output from the test in
the -expected.txt file. Use run-webkit-tests --reset-results to create this
file. Example:

$ run-webkit-tests --target=Default --reset-results
external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/extendable-message-event-constructor*

Writing new expected result
"external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/extendable-message-event-constructor.https-expected.txt"

With the expected file, we don't need any entry in TestExpectations, and can
remove the test from http/serviceworker.

$ cat
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/extendable-message-event-constructor.https-expected.txt
This is a testharness.js-based test.
PASS ServiceWorkerGlobalScope: ExtendableMessageEvent Constructor 
FAIL no initializer specified Cannot read property 'length' of null
PASS `bubbles` is specified 
PASS `cancelable` is specified 
PASS `data` is specified 
PASS `origin` is specified 
PASS `lastEventId` is specified 
PASS `source` is specified 
FAIL `ports` is specified Cannot read property 'length' of null
PASS all initial values are specified 
Harness: the test ran to completion.

[mike3, 2017-03-17 14:58:03]

Description was changed from

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API. The corrected version
cannot be enabled immediately, but simply disabling it would diminish
test coverage. Therefor, this patch introduces a version of the original
test where the invalid assertions have simply been disabled.

This patch also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align
with the patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

to

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

[mike3, 2017-03-17 14:58:19]

Description was changed from

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

to

==========
This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

[mike3, 2017-03-17 14:59:18]

Description was changed from

==========
This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

to

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

[mike3, 2017-03-17 15:06:19]

On 2017/03/17 03:34:43, falken wrote:
> I like what you've done to the tests, though I haven't fully examined the diff
> yet.

Thanks!

> With the expected file, we don't need any entry in TestExpectations, and can
> remove the test from http/serviceworker.

That is much more straightforward! It doesn't seem like there is any way to
annotate the expectations with references to open bugs, but I suppose anyone
who fixes those issues will discover the file when the tests fail.

Anyway, I've simplified the patch's description to not include all that detail
about my previous attempts to preserve coverage.

[falken, 2017-03-17 15:37:47]

falken@chromium.org changed reviewers:
+ mek@chromium.org

[falken, 2017-03-17 15:37:48]

I'd like to give this a proper review but it's late now and I won't be around on
Monday due to a national holiday. I've added mek@ as an optional reviewer in
case he'd be able to review this.

I made some quick comments on a superficial pass through.

> It doesn't seem like there is any way to
annotate the expectations with references to open bugs, but I suppose anyone
who fixes those issues will discover the file when the tests fail.

I set the bugs to block our master bug 678905. You should also add those bugs to
the BUG= list:

BUG=688116, 702352, 702361

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event.js
(right):

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event.js:1:
var wptEME = {};
nit: This name is a bit inscrutable. Should we just expand it to
ExtendableMessageEventTest? I think it's more common for this kind of
namespace-like var to start with a capital letter too.

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event.js:6:
*/
nit: /* */ style comments are rare in Chromium and the service worker tests
upstreamed so far. I would use // unless there is an overriding WPT style guide.

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html
(left):

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html:1:
<!DOCTYPE html>
For some reason Rietveld annotates this file diff as M (modified) instead of D
(deleted). The patch is really deleting this file, and not just making it 0
bytes right?

[mike3, 2017-03-17 16:11:53]

Description was changed from

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116
R=falken@chromium.org
==========

to

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116, 702352, 702361
R=falken@chromium.org
==========

[mike3, 2017-03-17 16:31:41]

I've made the improvements you've requested and updated the patch description.
Thanks for the review, and enjoy the equinox!

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event.js
(right):

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event.js:1:
var wptEME = {};
On 2017/03/17 15:37:48, falken (ooo Mon Mar 20) wrote:
> nit: This name is a bit inscrutable. Should we just expand it to
> ExtendableMessageEventTest? I think it's more common for this kind of
> namespace-like var to start with a capital letter too.

Yeah, probably developer ergonomics need to take a back seat here. I've
used the name `ExtendableMessageEventUtils` so that the file name can
match without confusion about the "test versus utility" distinction.

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event.js:6:
*/
On 2017/03/17 15:37:48, falken (ooo Mon Mar 20) wrote:
> nit: /* */ style comments are rare in Chromium and the service worker tests
> upstreamed so far. I would use // unless there is an overriding WPT style
guide.

Acknowledged.

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html
(left):

https://codereview.chromium.org/2751113005/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html:1:
<!DOCTYPE html>
On 2017/03/17 15:37:48, falken (ooo Mon Mar 20) wrote:
> For some reason Rietveld annotates this file diff as M (modified) instead of D
> (deleted). The patch is really deleting this file, and not just making it 0
> bytes right?

As far as I can tell, yes, this file is deleted. (It may be rendered
inconsistently because this file was introduced in the first patch and then
removed in the second version.)

[falken, 2017-03-21 14:42:56]

I think this looks good.

For the CL description, it looks like the common format in other commits is
BUG=NUM,NUM,NUM without the space after commas. I know I recommended the spaces
before, sorry.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js
(right):

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:1:
importScripts('/resources/testharness.js');
A suggestion: For files like this that Rietveld can't provide a diff for, it'd
help for reviewing if you could possibly link to a diff of the existing vs new
content, and point out changes (or the lack of changes).

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:2:

I think we've lost a comment: "These test cases are mostly copied from
fast/events/constructors/message-event-constructor.html" which still seems
relevant.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:135:
assert_array_equals(createEvent({ ports: undefined }).ports, []);
The original file tested |null| also, is removing that intended?

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js
(right):

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js:1:
var ExtendableMessageEventUtils = {};
This file seems to be using a four space indent mostly, is there a reason for
that? In our style guide which most existing WPT SW tests are based on
<https://www.chromium.org/blink/serviceworker/testing> we use a two space
indent.

(BTW eventually I'd like the WPT test style to use a more common JS style, but I
don't want to do a big reformatting until we've finishing upstreaming the
existing tests.)

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js:8:
});
I think according to our style guide this is indented like so:

var ports = events.ports.map(function(port) {
      return {...};
  });

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js:25:
};
I think according to our style guide lines 10-24 should just be indented 2
spaces:

return {
  a: x,
  b: y
};

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html
(left):

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html:9:
// the case where postMessage() is called from a detached client.
I'm wondering why this file can be deleted given the comment. Is there a WPT now
testing this behavior?

[mike3, 2017-03-21 16:54:16]

Description was changed from

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116, 702352, 702361
R=falken@chromium.org
==========

to

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116,702352,702361
R=falken@chromium.org
==========

[mike3, 2017-03-21 17:15:29]

> For the CL description, it looks like the common format in other commits is
> BUG=NUM,NUM,NUM without the space after commas. I know I recommended the
> spaces before, sorry.

No problem! I've done this via the "Edit Issue" page available in the web UI,
but the list is still being rendered with whitespace between the commas. If I
re-open the "Edit Issue" page, the description is rendered as intended (e.g.
*without* the spaces), so I'm not sure what the state of the description
is/will be.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js
(right):

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:1:
importScripts('/resources/testharness.js');
On 2017/03/21 14:42:56, falken wrote:
> A suggestion: For files like this that Rietveld can't provide a diff for, it'd
> help for reviewing if you could possibly link to a diff of the existing vs new
> content, and point out changes (or the lack of changes).

Sure thing. It looks like you had to do this yourself for your prior
review--sorry about that. Here's a diff for this file (generated using `git
diff`'s `-w` option to limit noise):

https://gist.github.com/jugglinmike/05fd45204be5e0635cf06731802b4e39

For this file, I introduced `testharness.js` so we could segment the assertions
into more targeted tests. This included some trivial reformatting. Beyond that,
I updated three invalid assertions regarding the `ports` option--the invalidity
is the subject of the Chromium bugs I filed.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:2:

On 2017/03/21 14:42:56, falken wrote:
> I think we've lost a comment: "These test cases are mostly copied from
> fast/events/constructors/message-event-constructor.html" which still seems
> relevant.

I've re-added that comment, expanded with additional context to make it
meaningful from WPT.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:135:
assert_array_equals(createEvent({ ports: undefined }).ports, []);
On 2017/03/21 14:42:56, falken wrote:
> The original file tested |null| also, is removing that intended?

Good eye! This is intentional: that assertion was invalid. In this new
version of the file, we are still testing for `null`, but we're instead
asserting that it triggers an exception (included below, under "Invalid
message ports.")

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js
(right):

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js:1:
var ExtendableMessageEventUtils = {};
On 2017/03/21 14:42:56, falken wrote:
> This file seems to be using a four space indent mostly, is there a reason for
> that? In our style guide which most existing WPT SW tests are based on
> <https://www.chromium.org/blink/serviceworker/testing> we use a two space
> indent.
> 
> (BTW eventually I'd like the WPT test style to use a more common JS style, but
I
> don't want to do a big reformatting until we've finishing upstreaming the
> existing tests.)

There's some inconsistency in the WPT tests, but I may have chosen the
lesser-used
version. Happy to reformat, though.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js:8:
});
On 2017/03/21 14:42:56, falken wrote:
> I think according to our style guide this is indented like so:
> 
> var ports = events.ports.map(function(port) {
>       return {...};
>   });

Acknowledged.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-utils.js:25:
};
On 2017/03/21 14:42:56, falken wrote:
> I think according to our style guide lines 10-24 should just be indented 2
> spaces:
> 
> return {
>   a: x,
>   b: y
> };
> 

Acknowledged.

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html
(left):

https://codereview.chromium.org/2751113005/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/http/tests/serviceworker/chromium/extendable-message-event.html:9:
// the case where postMessage() is called from a detached client.
On 2017/03/21 14:42:56, falken wrote:
> I'm wondering why this file can be deleted given the comment. Is there a WPT
now
> testing this behavior?

It most certainly can *not* be deleted. This was my mistake; I was over-zelous
in removing the test I introduced in the previous version of this patch. This
test remains valid, but only for Chromium, so I've re-added it as originally
authored.

[falken, 2017-03-22 00:55:15]

lgtm, thanks!

Regarding the whitespace in the BUG= list, let's not worry about it. Probably
it'll work out fine (the thing that matters is just bugdroid being able to link
the commit to the bugs).

https://codereview.chromium.org/2751113005/diff/60001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js
(right):

https://codereview.chromium.org/2751113005/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:6:
var ports = [channel1.port1, channel1.port2, channel2.port1];
nit: Would it be worth naming these like const TEST_OBJECT, const CHANNEL_1,
const PORTS? It's hard to grep |ports| in this file and see where this global is
used.

https://codereview.chromium.org/2751113005/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:89:
// lastEventId is passed.
nit: You probably meant to remove this comment, like the others.

[mike3, 2017-03-22 22:45:28]

Thanks for the review!

https://codereview.chromium.org/2751113005/diff/60001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js
(right):

https://codereview.chromium.org/2751113005/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:6:
var ports = [channel1.port1, channel1.port2, channel2.port1];
On 2017/03/22 00:55:14, falken wrote:
> nit: Would it be worth naming these like const TEST_OBJECT, const CHANNEL_1,
> const PORTS? It's hard to grep |ports| in this file and see where this global
is
> used.

Sounds good to me.

https://codereview.chromium.org/2751113005/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/external/wpt/service-workers/service-worker/ServiceWorkerGlobalScope/resources/extendable-message-event-constructor-worker.js:89:
// lastEventId is passed.
On 2017/03/22 00:55:14, falken wrote:
> nit: You probably meant to remove this comment, like the others.

Indeed!

[mike3, 2017-03-23 14:53:08]

The CQ bit was checked by mike@mikepennisi.com

[mike3, 2017-03-23 14:53:08]

The patchset sent to the CQ was uploaded after l-g-t-m from falken@chromium.org
Link to the patchset: https://codereview.chromium.org/2751113005/#ps100001
(title: "Remove redundant comment")

[commit-bot: I haz the power, 2017-03-23 14:53:28]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-23 16:00:19]

CQ is committing da patch.

Bot data: {"patchset_id": 100001, "attempt_start_ts": 1490280788072360,
"parent_rev": "f57859facd3f9b3468693cb4869a4566c6624eb1", "commit_rev":
"7d998666c33c0eaaf9c6b2fa874644c1c03f31fc"}

[commit-bot: I haz the power, 2017-03-23 16:01:13]

Description was changed from

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116,702352,702361
R=falken@chromium.org
==========

to

==========
Upstream service worker message event tests to WPT

This patch corrects invalid assertions relating to the `ports` property
described by the ExtendableMessageEvent API (see crbug.com/702352 and
crbug.com/702361). It also re-formats the tests to improve test precision (by
reducing reliance on the `toString` operation) and more closely align with the
patterns used within the Web Platform Tests project.

BUG=688116,702352,702361
R=falken@chromium.org

Review-Url: https://codereview.chromium.org/2751113005
Cr-Commit-Position: refs/heads/master@{#459099}
Committed:
https://chromium.googlesource.com/chromium/src/+/7d998666c33c0eaaf9c6b2fa8746...
==========

[commit-bot: I haz the power, 2017-03-23 16:01:14]

Committed patchset #6 (id:100001) as
https://chromium.googlesource.com/chromium/src/+/7d998666c33c0eaaf9c6b2fa8746...