Re: [xml] building a doc from fragments

[Daniel Veillard <veillard redhat com>]

On Fri, Jun 30, 2006 at 10:54:09PM -0600, John Calcote wrote:
> Since I don't really build a "doc" per se, but rather a series of
> fragments, what is the most appropriate way to build these fragments,
> and then append them to an doc later? Also, my API allows the insertion
> at various points of non-well-defined fragments of XML text. How do I
> parse in such a fragment of XML text, and then insert the root node of
> that fragment into a parent node that I'm managing?

  In general XML doesn't define semantic for this. You parse document,
there is no notion of parsing a fragement. Many thing in parsing are contextual
like namespace, base, entities, etc...
I really suggest reading the specs before trying to do this kind of things.

> Pointers to existing (sample) code that does these things would be
> perfectly acceptable, but please don't point me to sample code that
> doesn't do what I asked for. :)

  The point is that what you ask for really doesn't match any standard
processing it it will be custom. The closest to what you seems to have in
mind are:
   http://xmlsoft.org/html/libxml-parser.html#xmlParseBalancedChunkMemory
   and 
   http://xmlsoft.org/html/libxml-parser.html#xmlParseInNodeContext

 but really it's outside the definition of normal document parsing as
defined in XML-1.0

> The doc is not very extensive on libxml,
> and so good sample code is kinda required.

  We take patches !

Daniel

-- 
Daniel Veillard      | Red Hat http://redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/