Ignorant meta-document (from schema?) question
 

I am writing a post-processor for an existing XML document (that I have
no control over), this entails a lot of inserting and deleting of
nodes.

I don't want to have to hard code the (schema-defined) 'parent' XPath
of every node (so that I know what to clone and where to insert). It
seems that there should be someway of dynamically determining (via the
schema?) the 'parent' paths such that I am able to create valid
documents.

Does anything like this exist or is it just wishful thinking?

(I use 'parent' here not to mean the physical parent of a node, rather
the logical parent from the schema's point of view (as in if I clone
this node then I really have to clone this entire path and insert it
here)).
I'm hoping I've missed something painfully obvious.

Thanks-- Mike

Re: Ignorant meta-document (from schema?) question
 

Mike wrote:
> I am writing a post-processor for an existing XML document (that I have
> no control over), this entails a lot of inserting and deleting of
> nodes.
>
> I don't want to have to hard code the (schema-defined) 'parent' XPath
> of every node (so that I know what to clone and where to insert). It
> seems that there should be someway of dynamically determining (via the
> schema?) the 'parent' paths such that I am able to create valid
> documents.
>
> Does anything like this exist or is it just wishful thinking?
>
> (I use 'parent' here not to mean the physical parent of a node, rather
> the logical parent from the schema's point of view (as in if I clone
> this node then I really have to clone this entire path and insert it
> here)).

I think MSXML 4 when validating XML instance files against XML schemas
tries to display the XPath to the relevant element e.g.
/root/el1[3]/el2[4]/el3[5]
and it is certainly possible to construct such paths using XSLT. I can't
point you to an existing XSLT stylesheet or other XML software tools
that solves that.

--

Martin Honnen
http://JavaScript.FAQTs.com/

Re: Ignorant meta-document (from schema?) question
 

Martin,

Thanks for the thoughts, I already have access to the unique path
though.

What I'm looking for is something that would do an XPath-like
ancestor-or-self:: search but use what it knows about the schema to
find the "real" root node.

Thanks again-- Mike

Re: Ignorant meta-document (from schema?) question
 

Mike wrote:

> What I'm looking for is something that would do an XPath-like
> ancestor-or-self:: search but use what it knows about the schema to
> find the "real" root node.

Some schema aware parsers implement a schema object model that allows
you to find out schema related information.
For MSXML 4 (and 5) look at
http://msdn.microsoft.com/library/de...e_Overview.asp
For Xerces-J have a look at the documentation at
http://xml.apache.org/


--

Martin Honnen
http://JavaScript.FAQTs.com/

Re: Ignorant meta-document (from schema?) question
 

The Xerces-J schema stuff looks interesting, I might be able to build
something out of that.
http://xmlbeans.apache.org/ also looks promising.

Thanks for the help.