I think an example will help to illustrate why such a method makes no
sense in the API. Consider the following document:
<a:root xmlns:a="http://a"; xmlns:b="http://b";
xmlns:y="http://y"; xmlns:z="http://z";>
<a:a xmlns:a="http://www.abc.com"; xmlns:x="http://www.xyz.com";>
<a:a xmlns:a="http://www.xyz.com"; xmlns:x="http://www.abc.com"/>
</a:a>
<y:a xmlns:b="http://www.abc.com"; xmlns:y="http://www.xyz.com";>
<y:a xmlns:b="http://www.xyz.com"; xmlns:y="http://www.abc.com"/>
</y:a>
</a:root>
What is the prefix for "http://www.abc.com";? Without also considering
the context you can't give a sensible answer. An XSModel may be composed
of multiple schema documents which could also have different prefix
mappings in each of them for the same namespaces.
Thanks.
On Fri, 22 Oct 2010, Mark Brucks wrote:
Thanks for the responses.
Some clarification about my problem. I have the schema, but no instance
documents, and hence no worries about the fluid nature of prefixes between
schema and instance documents. My current task is to generate a "starter"
document from the schema. What I would like to do is use the prefix values
from the schema in my instance document, under the assumption that the
person that wrote the schema found the prefix to be a good mnemonic, and
hence it would serve the same purpose in the instance document. If I had
the prefix, it appears that I could then use
Element elem = CoreDocumentImpl.createElementNS(namespace, name);
elem.setPrefix(prefix);
to create the element in the given namespace and set the prefix to be
associated with that namespace (for that element).
Doing this, I hope to generate an instance document that has
<s1:elementOne xmlns:s2="http://www.foo.com/SchemaTwo"; ...>
<s2:elementTwo ... />
instead of having elementTwo defined with
<elementTwo xmlns="http://www.foo.com/SchemaTwo"; ... />
From what Michael says, it appears I can't get the prefix information from
XSModel. From his phrase "not sure what it would mean to return such a
list", it seems I am missing something fundamental about why this wouldn't
be considered useful information. The list would simply be a default
association of prefix with namespace. It would be useful for my purpose,
but useless for processing any existing document. Is there some reason that
what I want to do is a bad idea?
Thanks - Mark
On Oct 21, 2010, at 12:30 PM, Michael Glavassevich wrote:
Hi Mark,
Mark Brucks <bru...@charter.net> wrote on 10/21/2010 03:17:47
PM:
> Here is the beginning of my schema file:
>
> <xs:schema targetNamespace="http://www.foo.com/SchemaOne";
> xmlns:s1="http://www.foo.com/SchemaOne";
> xmlns:s2="http://www.foo.com/SchemaTwo";
> xmlns:s3="http://www.foo.com/SchemaThree";
> ...>
>
> Using the XML Schema API, I parse the schema and have access
to the
> resulting XSModel instance. Using the method
getNamespaceItems() I
> can get a list of XSNamespaceItem and using the
getSchemaNamespace()
> method I can get all the namespaces.
You're getting the list of all the target namespaces for the
schema components in the XSModel. This has nothing to do with
the namespace declarations which may have been in the schema
documents.
> How do I get the associated prefix for each prefix?
You can't and not sure what it would mean to return such a list.
The XSModel is a schema component view. There is no way to
access the schema documents which composed it.
> Thanks - Mark
Thanks.
Michael Glavassevich
XML Parser Development
IBM Toronto Lab
E-mail: mrgla...@ca.ibm.com
E-mail: mrgla...@apache.org
---------------------------
Michael Glavassevich
XML Parser Development
IBM Toronto Lab
E-mail: mrgla...@ca.ibm.com
E-mail: mrgla...@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: j-users-unsubscr...@xerces.apache.org
For additional commands, e-mail: j-users-h...@xerces.apache.org
