Re: [JDEV] *Real* element.h usage doc

[dsmith@ai.uwf.edu]

On  2 May, Jeremie wrote:
> Everything else looks happy, here's a quickie note:
> 

> Can(and often does) get parsed by Expat into:
> 
> 	Element "tag"
> 		CDATA "this is som"
> 		CDATA "e "
> 		CDATA "text"
> 
> It will break a string into multiple CDATA sections, this is just a normal
> part of sream-based XML processing...
> 

Yeah, that's correct. The W3C DOM (and ours,soon) handles this by
merging adjacent CDATA sections, on CDATA insertion into the DOM tree.
Really, the element.h is just the basic storage and manipulation
routines for dealing with elements. I'll probably move the
element_get_child_cdata (and adjacent CDATA merging) to a higher level
of abstraction since it's not really a basic manipulation. 

Hopefully tonight I'll be able to start writing the next layer that
will integrate expat for packet level parsing... :)

I also want to try out the latest stuff you've committed. Sounds great.
:)


D.