Minimize calls to Node::nodeType() as it is virtual

Minimize calls to Node::nodeType() as it is virtual

Minimize calls to Node::nodeType() as it is virtual and there are sometimes
better alternative (i.e. Node flags). This patch leverages Node flags when
possible (using methods such as isElementNode(), isDocumentFragment(), ...). It
also caches the result of Node::nodeType() when suitable to avoid calling it
several times in a row.

R=eseidel@chromium.org

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=175405

[aandrey, 2014-06-02 14:50:51]

FYI

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Document...
File Source/core/dom/Document.cpp (right):

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Document...
Source/core/dom/Document.cpp:3337: for (Node* c =
toDocumentFragment(newChild).firstChild(); c; c = c->nextSibling()) {
why change this?

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Node.cpp
File Source/core/dom/Node.cpp (right):

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Node.cpp...
Source/core/dom/Node.cpp:503: if (node->nodeType() == TEXT_NODE)
node->isTextNode() ?

[Inactive, 2014-06-02 15:48:02]

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Document...
File Source/core/dom/Document.cpp (right):

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Document...
Source/core/dom/Document.cpp:3337: for (Node* c =
toDocumentFragment(newChild).firstChild(); c; c = c->nextSibling()) {
On 2014/06/02 14:50:51, aandrey wrote:
> why change this?

Because this ends up calling ContainerNode::firstChild() instead of
Node::firstChild(). This is faster because it bypasses the useless
isContainerNode() check.

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Node.cpp
File Source/core/dom/Node.cpp (right):

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Node.cpp...
Source/core/dom/Node.cpp:503: if (node->nodeType() == TEXT_NODE)
On 2014/06/02 14:50:51, aandrey wrote:
> node->isTextNode() ?

This is not equivalent, isTextNode() returns true for both Text *and*
CDATASection. (node->nodeType() == TEXT_NODE) is true only for Text, not for
CDATASection.

[Inactive, 2014-06-02 15:49:02]

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Document...
File Source/core/dom/Document.cpp (right):

https://codereview.chromium.org/308963002/diff/20001/Source/core/dom/Document...
Source/core/dom/Document.cpp:3337: for (Node* c =
toDocumentFragment(newChild).firstChild(); c; c = c->nextSibling()) {
On 2014/06/02 15:48:02, Chris Dumez wrote:
> On 2014/06/02 14:50:51, aandrey wrote:
> > why change this?
> 
> Because this ends up calling ContainerNode::firstChild() instead of
> Node::firstChild(). This is faster because it bypasses the useless
> isContainerNode() check.

BTW, this is not the only place where we should do this. I have a follow-up
patch to do this in other places as well.

[Inactive, 2014-06-03 19:19:01]

Eric, what do you think about this one?

[eseidel, 2014-06-03 19:37:54]

lgtm

[commit-bot: I haz the power, 2014-06-03 19:38:45]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/ch.dumez@samsung.com/308963002/20001

[commit-bot: I haz the power, 2014-06-03 20:38:01]

Change committed as 175405