RE: [xml] XPath empty default namespaces (bug?)

["Alex Dron" <avdron infomosaic com>]

Thanks, Daniel,
everything is fine now. I cannot reproduce this error, too :)
I guess it was my fault (probably using old version of the library, I have
several).
Now I have correct node-set, and default attribute appears as well.

About "different point", I'm not using XPath to implement canonicalization,
rather to check implementation - since Canonical XML operates with XPath
node-sets as well as with octet stream.

Thanks again,
Alex

-----Original Message-----
From: xml-admin gnome org [mailto:xml-admin gnome org]On Behalf Of Daniel
Veillard
Sent: Thursday, November 01, 2001 4:51 AM
To: Alex Dron
Cc: Xml Gnome  Org
Subject: Re: [xml] XPath empty default namespaces (bug?)


On Wed, Oct 31, 2001 at 11:46:25AM -0800, Alex Dron wrote:
> ----------------------------------------------
> <!DOCTYPE doc [
> <!ATTLIST e2 xml:space (default|preserve) 'preserve'>
> <!ATTLIST e3 id ID #IMPLIED>
> ]>
> <doc xmlns="http://www.ietf.org"; xmlns:w3c="http://www.w3.org";>
>    <e1>
>       <e2 xmlns="">
>          <e3 id="E3"/>
>       </e2>
>    </e1>
> </doc>
> ----------------------------------------------
> Should not return <e2> node, but it does! Xerces Xpath works fine with
this
> example... It adds default attribute, too - libxml doesn't do it by
itself,
> but I can live with it.

  Well, I can't reproduce this !
First I double checked that e2 is really built in memory with no namespace
associated to it, and this looks right.
Second to test the XPath engine with a namespace context I included it
into an XPointer expression with a single namespace defined:

orchis:~/XML -> cat tst.xml
<!DOCTYPE doc [
<!ATTLIST e2 xml:space (default | preserve) "preserve">
<!ATTLIST e3 id ID #IMPLIED>
]>
<doc xmlns="http://www.ietf.org"; xmlns:w3c="http://www.w3.org";>
  <e1>
    <e2 xmlns="">
      <e3 id="E3"/>
    </e2>
  </e1>
</doc>

orchis:~/XML -> cat tst.xpath
xmlns(ietf=http://www.ietf.org)xpointer((//. | //@* | //namespace::*) [
self::ietf:e1 or (parent::ietf:e1 and not(self::text() or self::e2)) or
count(id("E3")|ancestor-or-self::node()) = count(ancestor-or-self::node())])
orchis:~/XML ->

If you look at xpointer.c around line 1022, you will see that
xmlns(ietf=http://www.ietf.org) is directly mapped as a call to
xmlXPathRegisterNs() with the associated prefix and URI.
The xpointer() part will then be evaluated in that context

And the result looks correct:

orchis:~/XML -> ./testXPath --xptr -i tst.xml -f tst.xpath

========================
Expression: xmlns(ietf=http://www.ietf.org)xpointer((//. | //@* |
//namespace::*) [ self::ietf:e1 or (parent::ietf:e1 and not(self::text() or
self::e2)) or count(id("E3")|ancestor-or-self::node()) =
count(ancestor-or-self::node())])
Object is a Node Set :
Set contains 3 nodes:
1  ELEMENT e1
2  ELEMENT e3
    ATTRIBUTE id
      TEXT
        content=E3
3  ATTRIBUTE id
    TEXT
      content=E3
orchis:~/XML ->

  The element e2 is not included in the resulting nodeset, and I don't
understand where your problem comes from. Note that xmlXPathRegisterNs()
is tied to a given evaluation context too, the problem may come from there.

  On a different point, I don't understand why you want to use XPath to
implement canonalization, this looks far too heavy weight and not a good
approach at all from an implementation perspective.

Daniel

--
Daniel Veillard      | Red Hat Network https://rhn.redhat.com/
veillard redhat com  | libxml Gnome XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/
_______________________________________________
xml mailing list, project page  http://xmlsoft.org/
xml gnome org
http://mail.gnome.org/mailman/listinfo/xml