Adding Element.closest() API

Adding Element.closest() API

Implementing Element.closest() API according to specs
https://dom.spec.whatwg.org/#dom-element-closest

This API will parse selectors and if fails then throw SyntaxError
else return closest ancestor that matches selectors
if there is no ancestor that matches selectors, it will
return nullptr.

Blink-Dev Intent to Implement and ship Link :
https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/yk_Fo6hyvb8

WebKit Change: http://trac.webkit.org/changeset/174324

BUG=422731
R=habib.virji@samsung.com

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=184946

[Paritosh Kumar, 2014-11-05 11:03:36]

Please have a look. 

Thanks and Regards
Paritosh

[Paritosh Kumar, 2014-11-06 09:02:42]

On 2014/11/05 11:03:36, Paritosh Kumar wrote:
> Please have a look. 
> 
> Thanks and Regards
> Paritosh

PTAL.

Thanks and Regards
Paritosh

[pdr., 2014-11-07 01:40:59]

pdr@chromium.org changed reviewers:
+ pdr@chromium.org, rune@opera.com

[pdr., 2014-11-07 01:40:59]

The CQ bit was checked by pdr@chromium.org

[pdr., 2014-11-07 01:41:00]

I think this looks great. LGTM.

+cc Rune, just FYI.

[commit-bot: I haz the power, 2014-11-07 01:41:14]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/704733002/1

[commit-bot: I haz the power, 2014-11-07 02:57:41]

Committed patchset #1 (id:1) as 184946

[pdr., 2014-11-07 22:24:02]

pdr@chromium.org changed reviewers:
+ l.gombos@samsung.com

[pdr., 2014-11-07 22:26:39]

Not LGTM. I'm going to roll this patch out.

A colleague at Apple brought to my attention that this patch is incorrect. The
following cases from the original patch do not pass:
FAIL theTarget.closest(":not(:scope)") should be [object HTMLBodyElement]. Was
null.
FAIL theTarget.closest("body:scope") should be null. Was [object
HTMLBodyElement].

What happened here? Did you just remove the failing cases and omit them in the
joined test file?

[pdr., 2014-11-07 22:28:32]

A revert of this CL (patchset #1 id:1) has been created in
https://codereview.chromium.org/693203004/ by pdr@chromium.org.

The reason for reverting is: Merged patch from WebKit contains errors..

[Paritosh Kumar, 2014-11-10 09:06:12]

Ohh!!!! Sorry... I'm really sorry. Big Mistake..

I uploaded the patch from my test branch not correct branch.

Surely, It won't work because, in this change I'm calling
selectorQuery->matches with argument currntElement, that changes each time.
and selectorQuery->matches calls SelectorDatalist::matches and it calls
SelectorDatalist::selectorMatches with same argument for ContainerNode and
Element at
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...

and this results in change in scope every time at
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...

Actually initially I was trying with the above method and those two missing
conditions were failing
so I done in different way in other branch but by mistake uploaded from the
incorrect branch. 
And by mistake got committed also.

I'm very sorry for this major and serious mistake and will always take care in
future.

Please have a look on the actual change.

Again Many Apologize for the caused inconvenience and thanks for reverting this
ptach and thanks to that Apple colleague also.

Please have a look again.

Thanks & Regards
Paritosh

[pdr., 2014-11-11 00:23:41]

On 2014/11/10 at 09:06:12, paritosh.in wrote:
> Ohh!!!! Sorry... I'm really sorry. Big Mistake..
> 
> I uploaded the patch from my test branch not correct branch.
> 
> Surely, It won't work because, in this change I'm calling
> selectorQuery->matches with argument currntElement, that changes each time.
> and selectorQuery->matches calls SelectorDatalist::matches and it calls
> SelectorDatalist::selectorMatches with same argument for ContainerNode and
Element at
>
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...
> 
> and this results in change in scope every time at
>
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...
> 
> Actually initially I was trying with the above method and those two missing
conditions were failing
> so I done in different way in other branch but by mistake uploaded from the
incorrect branch. 
> And by mistake got committed also.
> 
> I'm very sorry for this major and serious mistake and will always take care in
future.
> 
> Please have a look on the actual change.
> 
> Again Many Apologize for the caused inconvenience and thanks for reverting
this ptach and thanks to that Apple colleague also.
> 
> Please have a look again.
> 
> Thanks & Regards
> Paritosh

This was pretty bad but mistakes happen; please do be extra careful going
forward. Lets start a new review for re-landing this. Please reupload (just run
"git cl issue 0" and then "git cl upload").

Is there any reason to split up the tests from the webkit patch? Lets keep them
separate so this is easier to review.

I did not catch this in the original review but please include the original
author's name and email in the change description. I would add "This patch is a
merge from webkit http://trac.webkit.org/changeset/174324 by Dhi Aurrahman
<​diorahman@rockybars.com>".