Don't install private script class without a context document

Don't install private script class without a context document

In HTMLMarqueeElement's constructor, don't try to install the private
script class if there is no context document. Doing so is doesn't work,
and will crash (safely).

BUG=456624
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=189958

[Jens Widell, 2015-02-11 09:10:47]

jl@opera.com changed reviewers:
+ haraken@chromium.org

[Jens Widell, 2015-02-11 09:10:48]

PTAL

Alternative/complementary fix. I don't know what negative side-effects not
installing the class has, but it might be better than crashing, at least.

Only potential problem I can think of would be that the marquee might not start
animating automatically, but without a context document, I can't imagine the
marquee being displayed, so that's probably irrelevant.

This also fixes the test added in https://codereview.chromium.org/919473004/.

[haraken, 2015-02-11 09:16:31]

On 2015/02/11 09:10:48, Jens Widell wrote:
> PTAL
> 
> Alternative/complementary fix. I don't know what negative side-effects not
> installing the class has, but it might be better than crashing, at least.
> 
> Only potential problem I can think of would be that the marquee might not
start
> animating automatically, but without a context document, I can't imagine the
> marquee being displayed, so that's probably irrelevant.
> 
> This also fixes the test added in https://codereview.chromium.org/919473004/.

I think https://codereview.chromium.org/919473004/ is a better fix, but what do
you think?

In this CL, <marquee> will just not processed.

[Jens Widell, 2015-02-11 09:19:06]

On 2015/02/11 09:16:31, haraken wrote:
> On 2015/02/11 09:10:48, Jens Widell wrote:
> > PTAL
> > 
> > Alternative/complementary fix. I don't know what negative side-effects not
> > installing the class has, but it might be better than crashing, at least.
> > 
> > Only potential problem I can think of would be that the marquee might not
> start
> > animating automatically, but without a context document, I can't imagine the
> > marquee being displayed, so that's probably irrelevant.
> > 
> > This also fixes the test added in
https://codereview.chromium.org/919473004/.
> 
> I think https://codereview.chromium.org/919473004/ is a better fix, but what
do
> you think?
> 
> In this CL, <marquee> will just not processed.

I think we might want both. 919473004 doesn't fully fix the crash, after all.
This does, and they don't conflict in any way.

[haraken, 2015-02-11 09:19:58]

On 2015/02/11 09:19:06, Jens Widell wrote:
> On 2015/02/11 09:16:31, haraken wrote:
> > On 2015/02/11 09:10:48, Jens Widell wrote:
> > > PTAL
> > > 
> > > Alternative/complementary fix. I don't know what negative side-effects not
> > > installing the class has, but it might be better than crashing, at least.
> > > 
> > > Only potential problem I can think of would be that the marquee might not
> > start
> > > animating automatically, but without a context document, I can't imagine
the
> > > marquee being displayed, so that's probably irrelevant.
> > > 
> > > This also fixes the test added in
> https://codereview.chromium.org/919473004/.
> > 
> > I think https://codereview.chromium.org/919473004/ is a better fix, but what
> do
> > you think?
> > 
> > In this CL, <marquee> will just not processed.
> 
> I think we might want both. 919473004 doesn't fully fix the crash, after all.
> This does, and they don't conflict in any way.

Makes sense; LGTM.

[Jens Widell, 2015-02-11 09:21:48]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2015-02-11 09:22:49]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/918433004/1

[commit-bot: I haz the power, 2015-02-11 10:34:51]

Committed patchset #1 (id:1) as
https://src.chromium.org/viewvc/blink?view=rev&revision=189958