Implement ScriptPromise and ScriptFunction

Implement ScriptPromise and ScriptFunction

These two classes allow C++ code to hook into a Promise's resolution chain.
Documentation provided in ScriptPromise.h

[abarth-chromium, 2013-08-19 22:03:18]

https://codereview.chromium.org/23254004/diff/3001/Source/core/testing/Intern...
File Source/core/testing/Internals.cpp (right):

https://codereview.chromium.org/23254004/diff/3001/Source/core/testing/Intern...
Source/core/testing/Internals.cpp:2165: return function->releaseToV8Function();
I'm not 100% happy with this line.  I'd rather if it read more like:

return adoptByGarbageCollector(new AddOneFunction);

or something like that.

[alecflett, 2013-08-20 17:40:42]

this looks fine to me - agreed on the adoptByGarbageCollector() or something
adopt-like. It doesn't seem like ScriptFunction should have a GC-specific method
visible by subclasses.

[alecflett, 2013-08-26 23:14:50]

On 2013/08/20 17:40:42, alecflett wrote:
> this looks fine to me - agreed on the adoptByGarbageCollector() or something
> adopt-like. It doesn't seem like ScriptFunction should have a GC-specific
method
> visible by subclasses.

Is this on hold for the Promise refactoring, or can this go in, perhaps with an
adopt* method?

[abarth-chromium, 2013-08-27 23:07:40]

We're just missing a review.  Do you feel comfortable reviewing it, or should I
find another reviewer?

[alecflett, 2013-08-27 23:45:10]

On 2013/08/27 23:07:40, abarth wrote:
> We're just missing a review.  Do you feel comfortable reviewing it, or should
I
> find another reviewer?

I do; lgtm!

[alecflett, 2013-09-09 18:31:44]

On 2013/08/27 23:45:10, alecflett wrote:
> On 2013/08/27 23:07:40, abarth wrote:
> > We're just missing a review.  Do you feel comfortable reviewing it, or
should
> I
> > find another reviewer?
> 
> I do; lgtm!

So it looks like this patch has now broken -I was using it in a branch, but now
there is a real ScriptPromise class in bindings/v8, but without the then()
method.

[alecflett, 2013-09-16 19:03:53]

On 2013/09/09 18:31:44, alecflett wrote:
> On 2013/08/27 23:45:10, alecflett wrote:
> > On 2013/08/27 23:07:40, abarth wrote:
> > > We're just missing a review.  Do you feel comfortable reviewing it, or
> should
> > I
> > > find another reviewer?
> > 
> > I do; lgtm!
> 
> So it looks like this patch has now broken -I was using it in a branch, but
now
> there is a real ScriptPromise class in bindings/v8, but without the then()
> method.

I just revisited this again, attempting to fix. I see the new isolate stuff is
now conflicting - in particular V8ScriptRunner::callFunction() now needs an
isolate. I'm not sure where the best place to pull that in from is. (i.e. should
it be passed in manually to ScriptPromise::then(), or should it be pulled back
out of result?)

Though I now see that's kind of in-the-works:
https://codereview.chromium.org/23450039/diff/6001/Source/bindings/v8/ScriptP...

[abarth-chromium, 2013-09-16 21:08:07]

These are good questions for Marja.  I see that she's pushing isolate parameters
further up the call chain.  I'm not sure where the boundary is going to end up.

You can always get the isolate from thread-local storage via
v8::Isolate::GetCurrent().

[haraken, 2013-09-16 23:55:19]

On 2013/09/16 21:08:07, abarth wrote:
> These are good questions for Marja.  I see that she's pushing isolate
parameters
> further up the call chain.  I'm not sure where the boundary is going to end
up.
> 
> You can always get the isolate from thread-local storage via
> v8::Isolate::GetCurrent().

We will implement ScriptPromise::m_isolate in
https://codereview.chromium.org/23450039/ , so you can use it.