Re: [xml] XPath queries on document fragments without a document

[Daniel Veillard <veillard redhat com>]

On Thu, Sep 05, 2002 at 12:15:09PM -0700, Sean Chittenden wrote:
> > > [...]  my question is now: is there a reason to refuse xpath
> > > queries on document fragments?
> >
> >   yes, the XPath semantic is not defined. For example what does "/"
> > XPath evaluation should return ???  And an awful lot of XPath
> > queries use / , /foo or //foo to start a lookup, no really, having
> > element without a document breaks an awful lot of things in the XML
> > infrastructure, each time I suggest not to do it, people do it
> > anyway and have troubles, big surprise ...
>
> Return NULL on all queries that start with / and node->doc == NULL?  -sc

  foo[ id = /bar/value]

  and this breaks.

THE SOLUTION, is to keep a surrounding document element, as well as
a fake root element to encapsulate the well-balanced chunk and run the
query on this perfectly well defined document. 

A document fragment ain't an XML document. Make it a document and
everything becomes smooth,

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/