• From: Manos Batsis <manos_lists@g...>
• To: Michael Kay <mike@s...>, XML Developers List <xml-dev@l...>
• Date: Mon, 06 Dec 2010 19:59:17 +0200

On 12/06/2010 06:56 PM, Michael Kay wrote:
> On 06/12/2010 16:34, David Lee wrote:
>> Nothing stopping your application from allowing that equivalence
>> [between attributes and child elements].
>
> Yes, but it's an interesting idea that the parser should treat them as
> equivalent, simplifying the API and allowing an effortless move from
> attributes to child elements when the data gets more complex. Anything
> we can do to simplify the API (and thus the mapping to data structures
> supported by programming languages) is a move in the right direction.

My feeling exactly. Just to clarify and in the context of an element 
node, I'm only suggesting the following two cases of equivalence:

child element node <> attribute node with the same expanded name
(normalized) child text node <> (normalized) attribute value

An interesting exercise might be default conversion rules from such an 
"XML 2.0" to XML 1.x (e.g. xml:foo to attributes). Schema-driven 
conversions should be proven relatively simple.

Cheers,

Manos

• References:
  • Fixing what's broke
    • From: Bill Lindsey <bill@b...>
  • Re: Fixing what's broke
    • From: rjelliffe <rjelliffe@a...>
  • Re: XML2.0: Truncated End Tags (Was: Fixing what's broke)
    • From: "Pete Cordell" <petexmldev@c...>
  • RE: XML2.0: Truncated End Tags (Was: Fixing what's broke)
    • From: "David Lee" <dlee@c...>
  • Just nodes please (was: Re: XML2.0: Truncated End Tags(Was: Fixing what's broke))
    • From: Manos Batsis <manos_lists@g...>
  • RE: Just nodes please (was: Re: XML2.0: Truncated End Tags (Was: Fixing what's broke))
    • From: "David Lee" <dlee@c...>
  • Re: RE: Just nodes please
    • From: Michael Kay <mike@s...>