Bug 42194 - WebKitTestRunner needs a more-complete implementation of eventSender
Summary: WebKitTestRunner needs a more-complete implementation of eventSender
Status: ASSIGNED
Alias: None
Product: WebKit
Classification: Unclassified
Component: WebKit2 (show other bugs)
Version: 528+ (Nightly build)
Hardware: PC OS X 10.5
: P2 Normal
Assignee: Chang Shu
URL:
Keywords: InRadar
: 58059 61394 (view as bug list)
Depends on: 64285 68552 57515 57573 68108 68556 69946 86881 117070
Blocks: 55056 56484 76986
  Show dependency treegraph
 
Reported: 2010-07-13 13:52 PDT by Maciej Stachowiak
Modified: 2017-03-07 07:41 PST (History)
11 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Maciej Stachowiak 2010-07-13 13:52:28 PDT
WebKitTestRunner needs an implementation of eventSender. Many tests fail because of this (hundreds).
Comment 1 Maciej Stachowiak 2010-07-13 13:58:47 PDT
<rdar://problem/8186759>
Comment 2 Chang Shu 2011-03-07 14:53:09 PST
My plan is to:
1. implement a coding mechanism for EventSender under WebKit2/WebProcess/InjectedBundle.
I have tried to put this under Tools/WebKitTestRunner/InjectedBundle. However, it heavily depends on WebCore and CoreIPC code.
2. Use WKBundlePostMessage to send the EventSenderMessage over to UIProcess/WebContext.
3. When Tools/WebKitTestRunner/TestController receives message, call into WebKit2/WebProcess/InjectedBundle to decode the message and then trigger corresponding functions.

Does it make sense? Thanks
Comment 3 Darin Adler 2011-03-29 14:24:36 PDT
Have you made any progress on this?
Comment 4 Darin Adler 2011-03-29 14:25:24 PDT
I ask because I was about to work on this myself. There is already an eventSender in WebKitTestRunner, but we haven’t hooked any events up yet.
Comment 5 Chang Shu 2011-03-29 14:28:40 PDT
(In reply to comment #4)
> I ask because I was about to work on this myself. There is already an eventSender in WebKitTestRunner, but we haven’t hooked any events up yet.

Yes, I started with the KeyDown function on the Qt platform. https://bugs.webkit.org/show_bug.cgi?id=56485

The patch has been there for quite some time. It will be great if you can take a look. Btw, that patch needs some polish. I also want to get some feedback on the comment#7 of the above bug before I submit the 2nd-round patch.
Comment 6 Chang Shu 2011-03-29 14:29:43 PDT
I can also provide an equivalent patch based on Mac.
Comment 7 Darin Adler 2011-03-31 13:59:20 PDT
(In reply to comment #5)
> The patch has been there for quite some time. It will be great if you can take a look. Btw, that patch needs some polish. I also want to get some feedback on the comment#7 of the above bug before I submit the 2nd-round patch.

I’m sorry I never sent you feedback on that patch. I’m not sure what the best next step is.

Sam, Maciej, your thoughts?
Comment 8 Chang Shu 2011-03-31 14:05:15 PDT
(In reply to comment #7)
> (In reply to comment #5)
> > The patch has been there for quite some time. It will be great if you can take a look. Btw, that patch needs some polish. I also want to get some feedback on the comment#7 of the above bug before I submit the 2nd-round patch.
> 
> I’m sorry I never sent you feedback on that patch. I’m not sure what the best next step is.
> 
> Sam, Maciej, your thoughts?

That's fine. Actually, in my latest patch for bug 57515 (keydown), I created some generalized WK-api for decoding. Thus, I don't have to expose CoreIPC anymore. I think this approach is ok. Hope you guys agree.
Comment 9 Jessie Berlin 2011-04-07 11:10:24 PDT
*** Bug 58059 has been marked as a duplicate of this bug. ***
Comment 10 Jessie Berlin 2011-04-10 11:07:55 PDT
Added another test to the skipped list which uses eventSender.keyDown: http://trac.webkit.org/changeset/83402
Comment 11 Csaba Osztrogonác 2011-04-26 04:36:09 PDT
(In reply to comment #10)
> Added another test to the skipped list which uses eventSender.keyDown: http://trac.webkit.org/changeset/83402

One more test added: http://trac.webkit.org/changeset/84903
Comment 12 Ryosuke Niwa 2011-05-24 14:51:22 PDT
*** Bug 61394 has been marked as a duplicate of this bug. ***