XOM API for Processing XML with Java

[Wolfgang Hoschek <wolfgang.hoschek AT mac.com>]

Peter,

What you are doing is computing the namespaces declared *locally* on  
the given element (excluding it's ancestors). If you want the  
namespaces in scope, that is, including it's ancestors, you need to  
repeat what you are doing for the element and all its ancestors, and  
in doing so stuff the info into a set or map.

Like this:

                private static Map getNamespacePrefixesInScope(Element 
element) {
                        HashMap namespaces = new HashMap();
                        
                        do {
                                int size = 
element.getNamespaceDeclarationCount();
                                for (int i = 0; i < size; i++) {
                                        String prefix = 
element.getNamespacePrefix(i);
                                        if (!namespaces.containsKey(prefix)) {
                                                String uri = 
element.getNamespaceURI(prefix);
                                                namespaces.put(prefix, uri);
                                        }
                                }
                                ParentNode parent = element.getParent();
                                element = (parent instanceof Element ? 
(Element) parent : null);
                        } while (element != null);
                        
                        return namespaces;
                }

[which is the same as the non-public method  
Element.getNamespacePrefixesInScope()].

Wolfgang.

On Feb 5, 2006, at 3:44 PM, peter murray-rust wrote:

> At 18:25 05/02/2006, Elliotte Harold wrote:
> > XOM's about as consistent on this as it's possible to be given the
> > general mess that is namespaces.
>
> I agree that namespaces are one of the worst specs to come out of the
> process. They were done in a great rush and generated much friction.
> The use of "http://..."; as a name rather than an address is probably
> one of the worst ever decisions.
>
> >  Given any element if you ask for the URI mapped to a given prefix,
> > then it will return the URI as specified by Namespaces in XML.
>
> Agreed. But if I want to ask for the namespaceURIs in scope, rather
> than the prefixes then I don't seem to be able to get them all:
>
>
> > If you think you've identified a case where this isn't true, then
> > please send in a sample that shows what XOM does report as the
> > namespace URI and what you think it should report instead.
>
> Running under Xom1.12/Junit/Eclipse/Java 5.0:
>
>          public void testNamespaceScope() {
>                  String s = "<a xmlns:ns='http://foo'><b  
> foo='ns:bar'/></a>";
>                  Document doc = null;
>                  try {doc = new Builder().build(new
> StringReader(s));} catch (Exception e) {}
>                  Element a = doc.getRootElement();
>                  System.out.println("NS: "+a.getNamespaceURI("ns"));
>                  Element b = a.getFirstChildElement("b");
>                  int nsCount = b.getNamespaceDeclarationCount();
>                  System.out.println("NSCOUNT: "+nsCount);
>                  for (int i = 0; i < nsCount; i++) {
>                          System.out.println("NSPREFIX"+i+"
> ["+b.getNamespacePrefix(i)+"]");
>                          System.out.println("NSURI"+i+"
> ["+b.getNamespaceURI(b.getNamespacePrefix(i))+"]");
>                  }
>                  System.out.println("NS1: "+b.getNamespaceURI("ns"));
>          }
> gives:
> NS: http://foo
> NSCOUNT: 1
> NSPREFIX0 []
> NSURI0 []
> NS1: http://foo
>
> It seems that the prefix "ns" is not in scope as evidenced by the
> namespace declaration count and there is no other means of iterating
> through the namespaces (which is what I sometimes wish to do as I
> want to see if a namespaceURI is in scope at that stage and I may not
> know its prefix).
>
> Thanks
>
> P.
>
>
> Peter Murray-Rust
> Unilever Centre for Molecular Sciences Informatics
> University of Cambridge,
> Lensfield Road,  Cambridge CB2 1EW, UK
> +44-1223-763069
>
> _______________________________________________
> XOM-interest mailing list
> XOM-interest AT lists.ibiblio.org
> http://lists.ibiblio.org/mailman/listinfo/xom-interest