Expose "webkitIndexedDB" to Workers
Created attachment 86394 [details]
This isn't ready for actual review.
I'm tearing my hair out trying to figure out why these WorkerContext changes don't make webkitIndexedDB available in a worker. Do you know what I'm missing? Do I need to add some stuff to the bindings generator?
Comment on attachment 86394 [details]
View in context: https://bugs.webkit.org/attachment.cgi?id=86394&action=review
> + shouldBeTrue("'webkitIndexedDB' in window");
I don't get it...you set window = self above (in some cases)...? What exactly is this testing?
> + No new tests. (OOPS!)
what about the above test?
Eventually, all indexedDB tests shoudl run against workers and the main window. It's OK to just do a few for now tho.
extra new line
I'm not sure. cc'ing others who can maybe help
Btw, you'll need to add some stuff to the idl about static tables or something like that IIRC.
Also, note that you'll need to land code to make the IndexedDBDispatcher (and maybe other things? not positive) thread-safe before landing this so that we don't break isolates.
Attachment 86394 [details] did not build on qt:
Build output: http://queues.webkit.org/results/8221096
Make sure to add " NoStaticTables" to the idl that you are exposing. (Search for where it is used in other places.)
Your syntax for EnabledAtRuntime appears to be incorrect. (Look for where it is used in other places.) When you get it correct, you'll actually need to do work to expose turn on the flag. (Search for what other places do for their flag).
Attachment 86394 [details] did not build on win:
Build output: http://queues.webkit.org/results/8221100
(In reply to comment #7)
> Your syntax for EnabledAtRuntime appears to be incorrect. (Look for where it is used in other places.)
Drew pointed out that this part of my comment was incorrect. :)
But the key is to understand this flag.
I would strongly recommend looking at the generated code for this attribute and then maybe step through the code in a debugger - apologies if you've already done this, but pretty much any time I've run into issues with creating new JS bindings (and I *always* run into issues) I have to actually step through the bindings code to figure it out.
I was mentioning to Dave that the bindings code is like 90% boilerplate, but small problems with the boilerplate result in really difficult to diagnose problems.
Attachment 86394 [details] did not build on gtk:
Build output: http://queues.webkit.org/results/8276377
*** This bug has been marked as a duplicate of bug 73609 ***