Re: SAX2 NS processing (was Re: Namespace question)

james anderson (James.Anderson@mecomnet.de)
Fri, 07 Jan 2000 16:05:14 +0100

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
• Next message: Arnold, Curt: "RE: Requirement for attributes with schema scope"
• Previous message: David Megginson: "RE: SAX2: Parser interface"
• In reply to: Miles Sabin: "RE: SAX2: Parser interface"
• Next in thread: Stefan Haustein: "Attribute defaulting to element NS may cause problems (was: Re: SAX2 NS processing (was Re: Namespace question))"
• Next in thread: Simon St.Laurent: "Re: Namespace question: Do attributes inherit the element's namespace?"

The following is true only of application which don't live very long. For
example, for processes which correspond to a single selfcontained document
only. Where the processes live longer and continue to process additional
documents over their life span, that second little 
   || att.uri == ""
clause is going to make thier lives, and their coding, unnecessarily complicated.

In contrast with either of the illustrated clauses, i would expect a
standardized xml interface to support coding of the following sort.

 static QName symFoo = sax?.??.intern("http://www.foo.com/ns/", "foo");
 static QName symBar = sax?.??.intern("http://www.foo.com/ns/", "bar");
 >   for each att in atts do
>    
>       if att == symFoo
>         doFoo(att.value)
>       else if att == symBar
>         doBar(att.value)
>       end
>   end

Anything else is looking for problems.
Yes, one could code this over a SAX which follows "option #2".
Why should each application, or application library, need to? Where's the benefit?

David Megginson wrote:
> 
> "Simon St.Laurent" <simonstl@simonstl.com> writes:
> 
> > Will SAX2 namespace processing take this 'real-world' approach (which I
> > totally support), or are there going to be options for this?
> >
> > It sounded earlier like you expect namespace processing to be the domain of
> > the parser, not the application, so it seems like a live issue.
> 
> Basically, the difference in pseudo code is this:
> 
>   for each att in atts do
>     if att.uri == "http://www.foo.com/ns/" || att.uri == ""
>       if att.local == "foo"
>         doFoo(att.value)
>       else if att.local == "bar"
>         doBar(att.value)
>       end
>     end
>   end
> 
> vs. this:
> 
>   for each att in atts do
>     if att.uri == "http://www.foo.com/ns/"
>       if att.local == "foo"
>         doFoo(att.value)
>       else if att.local == "bar"
>         doBar(att.value)
>       end
>     end
>   end
> 
> In other words, there's not a very significant difference between the
> two.  The main problem is that the spec is just so much harder to read
> and understand that it had to be.
>


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo@ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)

• Next message: Arnold, Curt: "RE: Requirement for attributes with schema scope"
• Previous message: David Megginson: "RE: SAX2: Parser interface"
• In reply to: Miles Sabin: "RE: SAX2: Parser interface"
• Next in thread: Stefan Haustein: "Attribute defaulting to element NS may cause problems (was: Re: SAX2 NS processing (was Re: Namespace question))"
• Next in thread: Simon St.Laurent: "Re: Namespace question: Do attributes inherit the element's namespace?"

This archive was generated by hypermail 2.0b3 on Fri Jan 07 2000 - 15:03:54 GMT