Implement unregisterProtocolHandler() function on CUSTOM_SCHEME_HANDLER

Implement unregisterProtocolHandler() function
As a step to support CUSTOM_SCHEME_HANDLER, this cl implements unregisterProtocolHandler() first. The isProtocolRegistered() is not implemented yet. It will be supported after talking with W3C group. This patch introduces a customSchemeHandler runtime flag to support this feature as an experimental feature.
    
Additionally, this cl changes SYNTAX_ERR with SyntaxError in unregister-protocol-handler.html.

BUG= N/A

[gyuyoung-inactive, 2013-08-15 15:41:00]

Hi Abarth,

As you can see, it looks there was no objection to implement
*unregisterProtocolHandler()* and *isProtocolHandlerRegistered()*. However,
there was some suggestion to support the *isProtocolHandlerRegistered()*. So, I
would add *unregisterProtocolHandler()* at the moment. CUSTOM_SCHEME_HANDLER
will be enabled after finishing *isProtocolHandlerRegistered()*. I wonder if you
can agree with my direction.

https://groups.google.com/a/chromium.org/forum/#!searchin/blink-dev/unregiste...

[abarth-chromium, 2013-08-15 20:05:49]

https://codereview.chromium.org/22871006/diff/2001/Source/web/ChromeClientImpl.h
File Source/web/ChromeClientImpl.h (right):

https://codereview.chromium.org/22871006/diff/2001/Source/web/ChromeClientImp...
Source/web/ChromeClientImpl.h:211: #if ENABLE(CUSTOM_SCHEME_HANDLER)
We no longer use #ifdefs to scope in-progress feature development.  Instead, we
use RuntimeEnabledFeatures.in.

[gyuyoung-inactive, 2013-08-16 07:07:13]

> We no longer use #ifdefs to scope in-progress feature development.  Instead,
we
> use RuntimeEnabledFeatures.in.

Thank you for pointing it out. I try to use RuntimeEnabledFeatures.in to support
this feature. But, I'm not sure if I don't missing something to do. Please take
a look again.

https://codereview.chromium.org/22871006/diff/2001/Source/web/ChromeClientImpl.h
File Source/web/ChromeClientImpl.h (right):

https://codereview.chromium.org/22871006/diff/2001/Source/web/ChromeClientImp...
Source/web/ChromeClientImpl.h:211: #if ENABLE(CUSTOM_SCHEME_HANDLER)
On 2013/08/15 20:05:49, abarth wrote:
> We no longer use #ifdefs to scope in-progress feature development.  Instead,
we
> use RuntimeEnabledFeatures.in.

Thank you for pointing it out. I try to use RuntimeEnabledFeatures.in to support
this feature. But, I'm not sure if I don't missing something to do. Please take
a look again.

[gyuyoung-inactive, 2013-08-16 07:07:59]

>
https://codereview.chromium.org/22871006/diff/2001/Source/web/ChromeClientImpl.h
> File Source/web/ChromeClientImpl.h (right):
> 
>
https://codereview.chromium.org/22871006/diff/2001/Source/web/ChromeClientImp...
> Source/web/ChromeClientImpl.h:211: #if ENABLE(CUSTOM_SCHEME_HANDLER)
> On 2013/08/15 20:05:49, abarth wrote:
> > We no longer use #ifdefs to scope in-progress feature development.  Instead,
> we
> > use RuntimeEnabledFeatures.in.
> 
> Thank you for pointing it out. I try to use RuntimeEnabledFeatures.in to
support
> this feature. But, I'm not sure if I don't missing something to do. Please
take
> a look again.

Oops, my comment is duplicated. Please ignore it.

[gyuyoung-inactive, 2013-08-18 05:12:18]

Hi Abarth,

Could you take a look 22855008 first ? 

https://codereview.chromium.org/22855008/

In my humble opinion, the NAVIGATOR_CONTENT_UTILS feature can use a runtime
flag. So, I think this CL needs to be updated after the 22855008 is landed.

[gyuyoung-inactive, 2013-08-21 04:05:43]

Patch is updated regardless of issue 22855008.

[gyuyoung-inactive, 2013-08-21 06:41:18]

On 2013/08/21 04:05:43, gyuyoung wrote:
> Patch is updated regardless of issue 22855008.

Abarth, could you please take a look this patch ?

[abarth-chromium, 2013-08-21 17:10:19]

https://codereview.chromium.org/22871006/diff/16001/Source/web/ChromeClientIm...
File Source/web/ChromeClientImpl.h (right):

https://codereview.chromium.org/22871006/diff/16001/Source/web/ChromeClientIm...
Source/web/ChromeClientImpl.h:211: #if ENABLE(CUSTOM_SCHEME_HANDLER)
Please do not add any more ENABLE #if statements to the code base.  Using a
runtime enabled flag in the IDL is great and all you need here.

The way I would approach this change is to first remove
ENABLE(CUSTOM_SCHEME_HANDLER) entirely in favor of a runtime flag.  Then I would
add the new functionality.

https://codereview.chromium.org/22871006/diff/16001/public/web/WebRuntimeFeat...
File public/web/WebRuntimeFeatures.h (right):

https://codereview.chromium.org/22871006/diff/16001/public/web/WebRuntimeFeat...
public/web/WebRuntimeFeatures.h:144: WEBKIT_EXPORT static bool
isCustomSchemeHandlerEnabled();
There is no need to expose this API separately.  It's covered by the omnibus
"experimental" API.

[gyuyoung-inactive, 2013-08-22 01:12:38]

https://codereview.chromium.org/22871006/diff/16001/Source/web/ChromeClientIm...
File Source/web/ChromeClientImpl.h (right):

https://codereview.chromium.org/22871006/diff/16001/Source/web/ChromeClientIm...
Source/web/ChromeClientImpl.h:211: #if ENABLE(CUSTOM_SCHEME_HANDLER)
On 2013/08/21 17:10:19, abarth wrote:

> The way I would approach this change is to first remove
> ENABLE(CUSTOM_SCHEME_HANDLER) entirely in favor of a runtime flag.  Then I
would
> add the new functionality.

Abarth, I upload a patch to remove all ENABLE(CUSTOM_SCHEME_HANDLER) at
https://codereview.chromium.org/22859043/. I will update this patch after
landing it. Thanks.

https://codereview.chromium.org/22871006/diff/16001/public/web/WebRuntimeFeat...
File public/web/WebRuntimeFeatures.h (right):

https://codereview.chromium.org/22871006/diff/16001/public/web/WebRuntimeFeat...
public/web/WebRuntimeFeatures.h:144: WEBKIT_EXPORT static bool
isCustomSchemeHandlerEnabled();
On 2013/08/21 17:10:19, abarth wrote:
> There is no need to expose this API separately.  It's covered by the omnibus
> "experimental" API.

I wonder if the "experimental" API is generated automatically, right ? Or,
should I do something for it ?