[uf-discuss] include pattern question

[Tantek Ç elik]

On 6/8/06 1:10 PM, "Ryan King" <ryan at technorati.com> wrote:

> On Jun 8, 2006, at 12:02 PM, Scott Reynen wrote:
> 
>> On Jun 8, 2006, at 1:44 PM, Tantek Çelik wrote:
>> 
>>>> At first thought, I don't see why not, but it might require some
>>>> more
>>>> research.
>>> 
>>> I just checked the spec (HTML4.01), and I agree with Scott's
>>> conclusion.
>> 
>> That was actually Ryan's conclusion.
>> 
>> I have a question about the include pattern: is it okay to use
>> inclusion on elements whose semantics are tied to parent elements?
>> In Michael's example, does <area> still carry the same meaning when
>> it has been copied via inclusion outside of its container <map>?
>> Can I include <li> outside of <ul>?  <td> outside of <table>?
> 
> I'm trying to think of a case that would cause a problem. One would
> be the inclusion of a table cell that makes use of headers/axis
> stuff. I have a feeling that those references should be followed.

Indeed. I think this should work fine from a tree perspective.  Since the
semantics of microformats are in the class names rather than the element
names, those should work fine.

Another question is, should object-includes resolve recursively (unless
there is a violation of the ancestor rule) ?

That is, consider an page of hReviews that uses object-include to include
the hCard for the reviewer which itself uses object-include to include its
fn from some other location on the page.

When parsing one of the hReviews, and handling its object-include of the
reviewer hCard, the nested object-include should be resolved and handled as
well.

Tantek