Bug 126418 – Add toHTMLTableSectionElement() functions, and use it

• Component: DOM
• Status: RESOLVED FIXED

[Gyuyoung Kim]

When    9          generating toFoo(), we can use more plenty of helper functions. This patch generate toHTMLTableElement() using existing templates.

[Gyuyoung Kim]

Created attachment 220280 [details]
Patch

[Gyuyoung Kim]

CC'ing Kling.

[Andreas Kling]

Comment on attachment 220280 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=220280&action=review

> Source/WebCore/ChangeLog:3
> +        Generate toHTMLTableElement(), and use it

This patch seems to be about toHTMLTableSectionElement(), not toHTMLTableElement().

> Source/WebCore/html/HTMLTagNames.in:121
> -tbody interfaceName=HTMLTableSectionElement
> +tbody interfaceName=HTMLTableSectionElement, generateTypeHelpers

HTMLTableSectionElement can have one of three tags: <thead>, <tbody> or <tfoot>.
This would generate an isHTMLTableSectionElement() that only returns true for <tbody>.

[Gyuyoung Kim]

Comment on attachment 220280 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=220280&action=review

>> Source/WebCore/ChangeLog:3
>> +        Generate toHTMLTableElement(), and use it
> 
> This patch seems to be about toHTMLTableSectionElement(), not toHTMLTableElement().

Oops. my mistake.

>> Source/WebCore/html/HTMLTagNames.in:121
>> +tbody interfaceName=HTMLTableSectionElement, generateTypeHelpers
> 
> HTMLTableSectionElement can have one of three tags: <thead>, <tbody> or <tfoot>.
> This would generate an isHTMLTableSectionElement() that only returns true for <tbody>.

Ah, right. I forget it. I'm sorry. If so, how about adding manual toHTMLTableSectionElement() for HTMLTableSectionElement ?

[Gyuyoung Kim]

Created attachment 220293 [details]
Patch

[Gyuyoung Kim]

Created attachment 220294 [details]
Patch

[Gyuyoung Kim]

kling, how about this latest one ?

[WebKit Commit Bot]

Comment on attachment 220294 [details]
Patch

Clearing flags on attachment: 220294

Committed r161401: <http://trac.webkit.org/changeset/161401>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Alexey Proskuryakov]

This broke dozens of tests, which crash now.

http://build.webkit.org/results/Apple%20Mavericks%20Debug%20WK1%20(Tests)/r161401%20(1535)/results.html

[Gyuyoung Kim]

Reverted r161401 for reason:

REGRESSION(r161401): Break layout test on mac-wk1(Debug)

Committed r161408: <http://trac.webkit.org/changeset/161408>

[Andreas Kling]

Comment on attachment 220294 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=220294&action=review

Can't this just use NODE_TYPE_CASTS instead? SVGSMILElement does something similar.

> Source/WebCore/html/HTMLTableSectionElement.h:66
> +    return node.hasTagName(HTMLNames::trTag) || node.hasTagName(HTMLNames::tfootTag) || node.hasTagName(HTMLNames::tbodyTag);

trTag should be theadTag here.

> Source/WebCore/html/HTMLTableSectionElement.h:94
> +inline HTMLTableSectionElement* toHTMLTableSectionElement(Node* node)
> +{
> +    ASSERT_WITH_SECURITY_IMPLICATION(!node || isHTMLTableSectionElement(*node));
> +    return static_cast<HTMLTableSectionElement*>(node);
> +}
> +
> +inline const HTMLTableSectionElement* toHTMLTableSectionElement(const Node* node)
> +{
> +    ASSERT_WITH_SECURITY_IMPLICATION(!node || isHTMLTableSectionElement(*node));
> +    return static_cast<const HTMLTableSectionElement*>(node);
> +}
> +
> +inline HTMLTableSectionElement& toHTMLTableSectionElement(Node& node)
> +{
> +    ASSERT_WITH_SECURITY_IMPLICATION(isHTMLTableSectionElement(node));
> +    return static_cast<HTMLTableSectionElement&>(node);
> +}
> +
> +inline const HTMLTableSectionElement& toHTMLTableSectionElement(const Node& node)
> +{
> +    ASSERT_WITH_SECURITY_IMPLICATION(isHTMLTableSectionElement(node));
> +    return static_cast<const HTMLTableSectionElement&>(node);
> +}
> +
> +void toHTMLTableSectionElement(const HTMLTableSectionElement*);
> +void toHTMLTableSectionElement(const HTMLTableSectionElement&);

Can't

[Gyuyoung Kim]

Created attachment 220504 [details]
Patch

[Gyuyoung Kim]

(In reply to comment #12)
> (From update of attachment 220294 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=220294&action=review
> 
> Can't this just use NODE_TYPE_CASTS instead? SVGSMILElement does something similar.

There were many faults. I updated this patch according to your suggestion. Thanks.

[WebKit Commit Bot]

Comment on attachment 220504 [details]
Patch

Clearing flags on attachment: 220504

Committed r161420: <http://trac.webkit.org/changeset/161420>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.