Invoke JS messaging for UIWebView via window.requestAnimationFrame

Invoke JS messaging for UIWebView via window.requestAnimationFrame

JS messaging to native code for UIWebView is accomplished by adding
and removing an iframe to the window. Some pages interact badly with
this mechanism, including iforgot.apple.com. To circumvent this for
such pages, the functionality in message_dynamic.sendQueue is wrapped
in a function and invoked from window.requestAnimationFrame to delay
message sending.

BUG=470595
Committed: https://crrev.com/b593a61fafbdfb29a4382263f138586913b8a3bd
Cr-Commit-Position: refs/heads/master@{#330430}

[Jackie Quinn, 2015-05-14 19:16:51]

jyquinn@chromium.org changed reviewers:
+ eugenebut@chromium.org

[Eugene But (OOO till 7-30), 2015-05-14 21:51:38]

eugenebut@chromium.org changed reviewers:
+ stuartmorgan@chromium.org

[Eugene But (OOO till 7-30), 2015-05-14 21:51:38]

This is something, that needs to be checked with Stuart

[Jackie Quinn, 2015-05-14 23:23:29]

On 2015/05/14 21:51:38, eugenebut wrote:
> This is something, that needs to be checked with Stuart

That's what I figured.

[Jackie Quinn, 2015-05-18 19:40:36]

As discussed online, wrapped the entire dynamic.sendQueue method and call after
delay only for iforgot.apple.com.

[stuartmorgan, 2015-05-18 19:57:08]

CL description/title need updating for the current implementation.

Since the animation timeout should be ~10x less than this, and worked in
testing, maybe we should go ahead with that after all given that we're
hard-coding this anyway? It occurred to me that we're creating a window of
message loss at the point where we navigate away from the page, which is not
ideal, so cutting that down by an order of magnitude seems worth the risk of it
being slower in rare cases.

https://codereview.chromium.org/1132163007/diff/20001/ios/web/web_state/js/re...
File ios/web/web_state/js/resources/message_dynamic_ui.js (right):

https://codereview.chromium.org/1132163007/diff/20001/ios/web/web_state/js/re...
ios/web/web_state/js/resources/message_dynamic_ui.js:47: if
(window.location.origin === "https://iforgot.apple.com") {
Let's do a bit of combination self-documentation and future-proofing, and make a
pageRequiresDelayedSend; that way if the list grows it'll be self-contained
already.

[Jackie Quinn, 2015-05-18 20:36:35]

On 2015/05/18 19:57:08, stuartmorgan wrote:
> CL description/title need updating for the current implementation.
>
 
Done.

> Since the animation timeout should be ~10x less than this, and worked in
> testing, maybe we should go ahead with that after all given that we're
> hard-coding this anyway? It occurred to me that we're creating a window of
> message loss at the point where we navigate away from the page, which is not
> ideal, so cutting that down by an order of magnitude seems worth the risk of
it
> being slower in rare cases.

Sgtm, changed it to window.requestAnimationFrame.

> 
>
https://codereview.chromium.org/1132163007/diff/20001/ios/web/web_state/js/re...
> File ios/web/web_state/js/resources/message_dynamic_ui.js (right):
> 
>
https://codereview.chromium.org/1132163007/diff/20001/ios/web/web_state/js/re...
> ios/web/web_state/js/resources/message_dynamic_ui.js:47: if
> (window.location.origin === "https://iforgot.apple.com") {
> Let's do a bit of combination self-documentation and future-proofing, and make
a
> pageRequiresDelayedSend; that way if the list grows it'll be self-contained
> already.

[Jackie Quinn, 2015-05-18 20:36:42]

https://codereview.chromium.org/1132163007/diff/20001/ios/web/web_state/js/re...
File ios/web/web_state/js/resources/message_dynamic_ui.js (right):

https://codereview.chromium.org/1132163007/diff/20001/ios/web/web_state/js/re...
ios/web/web_state/js/resources/message_dynamic_ui.js:47: if
(window.location.origin === "https://iforgot.apple.com") {
On 2015/05/18 19:57:08, stuartmorgan wrote:
> Let's do a bit of combination self-documentation and future-proofing, and make
a
> pageRequiresDelayedSend; that way if the list grows it'll be self-contained
> already.

Done.

[stuartmorgan, 2015-05-18 20:53:28]

LGTM. I really wish we knew more about the underlying bug so that we could do
something less horrible than this, but getting this page working in the short
term is worth the hack.

[Jackie Quinn, 2015-05-18 20:59:37]

The CQ bit was checked by jyquinn@chromium.org

[commit-bot: I haz the power, 2015-05-18 21:00:00]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1132163007/60001

[commit-bot: I haz the power, 2015-05-18 21:35:54]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2015-05-18 22:44:03]

Patchset 4 (id:??) landed as
https://crrev.com/b593a61fafbdfb29a4382263f138586913b8a3bd
Cr-Commit-Position: refs/heads/master@{#330430}