Add initial support for exceptions....

Add initial support for exceptions.

Move some code into the anonymous namespace since it doesn't need to be exposed
outside of pepper_var.cc.

R=brettw
BUG=none
TEST=To be written once the new framework is available.


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=47738

[darin (slow to review), 2010-05-19 20:38:59]

The HAVE_WEBBINDINGS_EXCEPTION_HANDLER #ifdef refers to this:
https://bugs.webkit.org/show_bug.cgi?id=39378

[brettw, 2010-05-19 21:22:07]

LGTM

http://codereview.chromium.org/2123009/diff/1/2
File webkit/glue/plugins/pepper_var.cc (right):

http://codereview.chromium.org/2123009/diff/1/2#newcode56
webkit/glue/plugins/pepper_var.cc:56: PP_Var* exception_;
Can you comment that this may be NULL if the caller isn't interested in catching
exceptions?

http://codereview.chromium.org/2123009/diff/1/2#newcode414
webkit/glue/plugins/pepper_var.cc:414: // An exception should have been raised.
Is it possible to validate this with
  DCHECK(!exception || try_catch.HasException());

[darin (slow to review), 2010-05-19 21:48:50]

On Wed, May 19, 2010 at 2:22 PM, <brettw@chromium.org> wrote:

> LGTM
>
>
> http://codereview.chromium.org/2123009/diff/1/2
> File webkit/glue/plugins/pepper_var.cc (right):
>
> http://codereview.chromium.org/2123009/diff/1/2#newcode56
> webkit/glue/plugins/pepper_var.cc:56: PP_Var* exception_;
> Can you comment that this may be NULL if the caller isn't interested in
> catching exceptions?
>

OK


>
> http://codereview.chromium.org/2123009/diff/1/2#newcode414
> webkit/glue/plugins/pepper_var.cc:414: // An exception should have been
> raised.
> Is it possible to validate this with
>  DCHECK(!exception || try_catch.HasException());
>
>
Actually, I should change the text to say "may" instead of "should" because
there
are some "failure" cases that do not result in an exception.

-Darin



>
> http://codereview.chromium.org/2123009/show
>

[darin (slow to review), 2010-05-19 22:11:25]

> Actually, I should change the text to say "may" instead of "should" because
> there
> are some "failure" cases that do not result in an exception.

the latest CL will throw an exception if one was not thrown, thereby ensuring
that we always report an exception when something failed.

[brettw, 2010-05-19 22:15:31]

LGTM