Re: [sXBL] question on the <content> processing model

Ian Hickson wrote:
> On Mon, 28 Mar 2005, Boris Zbarsky wrote:
>> I'm not quite clear on exactly what <content> does, or more precisely on 
>> which nodes it operates.  Section 4.2 talks about "explicit children", 
>> but consider the situation represented by the diagram in that section -- 
>> node 3 has an sXBL binding bound to it which makes nodes 4 and 5 
>> children of the leftmost purple node (call it node A) in the flattened 
>> tree.
>>
>> Now say node A has an sXBL binding bound to it, and this binding has 
>> <content> elements.  Would those be able to reposition node 4 and node 5 
>> in the flattened tree under node A?  It seems that the intent is that 
>> they should, yet node 4 and node 5 are not at all explicit children of 
>> node A...  Some clarification here would be nice.
> 
> The <content> element is the one that is an explicit child of A. So that's 
> what gets "repositioned". And in the final flattened tree, it gets 
> replaced by what matches it, as normal.

So in other words, in the flattened tree nodes 4 and 5 can't be repositioned to 
different <content> insertion points in the binding attached to A (since it's 
the <content> node that's a child of A being repositioned, you can't filter 4 
and 5 into different insertion points).  Is that correct?

 > This all seems well-defined if you take the spec literally...

Perhaps the problem is that as I read I'm mapping onto the Mozilla XBL 
implementation, looking for the differences, and trying to understand them...

-Boris

Received on Monday, 18 July 2005 20:34:55 UTC