[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: [ANN] smallx XML Infoset and Pipeline Released (Open Sourc

• To: Elliotte Harold <elharo@m...>
• Subject: Re: [ANN] smallx XML Infoset and Pipeline Released (Open Source)
• From: Alan Gutierrez <alan-xml-dev@e...>
• Date: Thu, 14 Apr 2005 06:53:33 -0400
• Cc: ElektonikaMail@m..., Xml-Dev <xml-dev@l...>
• In-reply-to: <425E453B.1000300@m...>
• Mail-followup-to: Elliotte Harold <elharo@m...>,ElektonikaMail@m..., Xml-Dev <xml-dev@l...>
• References: <c43751899c33f56c61d770be0bb561c9@m...> <42569A85.7040206@b...> <908e97ffe509e7fc4938581431c7251a@m...> <4257B244.1080908@b...> <3f91e4e3447db1be834e21c2ae446e66@m...> <425BE997.7040403@m...> <c9e6ac6ac81070dd9bc1aa083d9059fb@m...> <001801c540c1$c3e52f90$0115a8c0@E...> <425E453B.1000300@m...>
• User-agent: Mutt/1.4.1i

* Elliotte Harold <elharo@m...> [2005-04-14 06:26]:
> Alan Gutierrez wrote:
> 
> >    In my SAX library, XStrategy, I combined namespaces and events.
> >
> >    The XStrategy framwork maintains a stack of namespaces in scope
> >    by default. It seemed impossible to get anything done without it.
> 
> If you're int he unfortunate situation of needing to resolve namespace 
> prefixes in attribute values and element content, as is the case in an 
> XSLT processor or a schema validator, then you absolutely need to do 
> this. However, probably at least 90% of use cases are much simpler and 
> just need to know the namespaces of each element and attribute, and this 
> doesn't require any extra work.
> 
> It's not all that hard to keep track of namespaces in scope in SAX. You 
> just use a NamespaceSupport object, push the namespace on in 
> startPrefixMapping and pop it off in endPrefixMapping. Sadly this isn't 
> as well documented or obvious as it should be.

    The key to XStrategy is a second event system that sends
    messages up a stack event handlers. It's API is stabilizing, and
    I'm finding that the reusable strategies, and the bubbling event
    model solve a lot of problems easily.

    With SAX I find that I'm not trying to shave every instruction,
    simply avoiding the overhead of tree materialization.

    If I really needed to strip every last bit overhead, I'd use
    SAX. I like that I've got namespaces in scope and the stack of
    events, amoung other things, encapsulated. It is good to have
    them there when I need them.

    Cheers.

--
Alan Gutierrez - alan@e...
    - http://engrm.com/blogometer/index.html
    - http://engrm.com/blogometer/rss.2.0.xml

• References:
  • [ANN] smallx XML Infoset and Pipeline Released (Open Source)
    • From: Alex Milowski <alexml@m...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (OpenSource)
    • From: Erik Bruchez <erik@b...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (Open Source)
    • From: Alex Milowski <alexml@m...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (OpenSource)
    • From: Erik Bruchez <erik@b...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (Open Source)
    • From: Alex Milowski <alexml@m...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (OpenSource)
    • From: Elliotte Harold <elharo@m...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (Open Source)
    • From: Alex Milowski <alexml@m...>
  • Re: [ANN] smallx XML Infoset and Pipeline Released (OpenSource)
    • From: Elliotte Harold <elharo@m...>

• Prev by Date: Re: [ANN] XML Enhancements for Java 1.0
• Next by Date: Fw: [ANN] XML Enhancements for Java 1.0
• Previous by thread: Re: [ANN] smallx XML Infoset and Pipeline Released (OpenSource)
• Next by thread: Re: [ANN] smallx XML Infoset and Pipeline Released
• Index(es):
  • Date
  • Thread