Bug 32936 – HTML5 <section> element support

• Component: DOM
• Status: RESOLVED FIXED

[Kent Tamura]

Implement <section> element. It should be very similar to <nav> we already have.

[Kent Tamura]

Created attachment 45494 [details]
Proposed patch

The patch is very similar to http://trac.webkit.org/changeset/47489 for <nav>.

[WebKit Review Bot]

style-queue ran check-webkit-style on attachment 45494 [details] without any errors.

[Darin Adler]

Comment on attachment 45494 [details]
Proposed patch

> +debug('<section> can be nest nested inside <section>:');

Looks like "nest" here is an extra word.

[Kent Tamura]

(In reply to comment #3)
> (From update of attachment 45494 [details])
> > +debug('<section> can be nest nested inside <section>:');
> 
> Looks like "nest" here is an extra word.

Right! Removed and reset the expected result.

Landed as r52564: <http://trac.webkit.org/changeset/52564>

[Lars Gunther]

As far as I can tell, this bug did not include the actual sectioning behaviour? I.e.

<h1>foo</h1>
<p>Lorem</p>
<section>
  <h1>bar</h1>
  <p>Lorem</p>
</section>

Should be functionally equivalent to:

<h1>Foo</h1>
<p>Lorem</p>
<h2>Bar</h2>
<p>Lorem</p>

I see no open bug about sectioning. Should one be opened?

[Kent Tamura]

(In reply to comment #5)
> As far as I can tell, this bug did not include the actual sectioning behaviour?

What part of the HTML5 spec are you talking about?

[Lars Gunther]

Sectioning is described here:

http://dev.w3.org/html5/spec/dom.html#sectioning-content-0

http://dev.w3.org/html5/spec/semantics.html#headings-and-sections

We also have this unofficial resource: http://gsnedders.html5.org/outliner/

[Kent Tamura]

(In reply to comment #7)

Two fragments in Comment #5 are equivalent in the outline algorithm, and they should not be equivalent as DOM trees.
I don't know we have reasons to implement the outline algorithm in WebKit.

[Lars Gunther]

(In reply to comment #8)

> Two fragments in Comment #5 are equivalent in the outline algorithm, and they
> should not be equivalent as DOM trees.
> I don't know we have reasons to implement the outline algorithm in WebKit.

It's not about the DOM tree, since they should not be equivalent. But it is about semantics and thus about accessibility and default styling.

When users jump between headings in programs like JAWS, an H1 within a section is not equal to an H1 outside of it.

Also, H1 within a section should probably not have the same default styling as an H1 outside of it.