Introduce toHTMLSelectElement and remove isHTMLSelectElement

Source/core/html/HTMLSelectElement.h

Index: Source/core/html/HTMLSelectElement.h
diff --git a/Source/core/html/HTMLSelectElement.h b/Source/core/html/HTMLSelectElement.h
index 65ab1d5a5dac86f4f931ae260e7a40283b60270b..ff3fb8c01ac886aeed4c97a9d5896756009542e7 100644
--- a/Source/core/html/HTMLSelectElement.h
+++ b/Source/core/html/HTMLSelectElement.h
@@ -206,25 +206,12 @@ private:
     bool m_isParsingInProgress;
 };
 
-inline bool isHTMLSelectElement(const Node* node)
-{
-    return node->hasTagName(HTMLNames::selectTag);
-}
-
 inline HTMLSelectElement* toHTMLSelectElement(Node* node)
 {
-    ASSERT_WITH_SECURITY_IMPLICATION(!node || isHTMLSelectElement(node));
+    ASSERT_WITH_SECURITY_IMPLICATION(!node || node->hasTagName(HTMLNames::selectTag));
     return static_cast<HTMLSelectElement*>(node);
 }
 
-inline const HTMLSelectElement* toHTMLSelectElement(const Node* node)

[tkent, 2013/07/17 01:47:56]

Why do you remove this function?

[kangil_, 2013/07/17 02:16:22]

Just because we don't use this at the moment.
But, I don't have strong opinion on this if you want to keep this as it is. :)

[tkent, 2013/07/17 03:47:05]

Hmm, it seems this function is actually unused.
I think it's ok to remove it because 'const' is rarely used for DOM nodes.

-{
-    ASSERT_WITH_SECURITY_IMPLICATION(!node || isHTMLSelectElement(node));
-    return static_cast<const HTMLSelectElement*>(node);
-}
-
-void toHTMLSelectElement(const HTMLSelectElement*); // This overload will catch anyone doing an unnecessary cast.

[tkent, 2013/07/17 01:47:56]

Ditto.

[kangil_, 2013/07/17 02:16:22]

I don't have strong opinion on this either, but should ask for clarity :)
  1. Why there isn't void toHTMLSelectElement(HTMLSelectElement*);
  2. Some HTMLFooElement have this function but others haven't. Do we need to
introduce same to others?

[kangil_, 2013/07/17 02:21:46]

Ah, this is wrong question. Please ignore this. :(

[tkent, 2013/07/17 03:47:05]

Removing this function is not a goal of this CL, right? I don't think you should
remove it in this patch.
IMO, this function is not so helpful, but it sometimes helps to improve code
quality.  I have no opinion about 2.

-
 } // namespace
 
 #endif