SAX2-beta (2000-01-28)

[Jim Layer]

Miles Sabin <msabin at cromwellmedia.co.uk> wrote:
> Like it or not SAX2 is a completely new API, so why not leave 
> SAX1 untouched (ie. no @deprecated's) and have a completely 
> fresh org.xml.sax2.* tree? 

I view this (SAX2 in org.xml.sax) as a Good Thing - I can put
the new jar in play with no CLASSPATH change and no impact to
current apps - 'til they hit the next update cycle, rebuild,
and then spew forth deprecation warnings if they have been so
slack as to have ignored other forms of interface change
warnings (email, interface doc updates, etc.). Also serves to 
aid locating change points when everyone *does* pay attention.
If not deprecation on the old stuff, then maybe a greater chance
for fat, dumb, and happy 'til we pull the old jar in a subsequent
release cycle. 'Course deprecation implies (from SUN's doc
on the subject "...it will probably cease to exist in the future")
that David's on the hook for another SAX <grin/> (do I hear a
groan from up north?).

> And I still think there needs to be a SAX2 equivalent of 
> ParserFactory ... bizarrely it's easier to create a SAX1
> Parser in an implementation independent way than it is to do 
> the same for an XMLReader. A couple of people volunteered to 
> try out the XMLReaderImplementations stuff I put together, and 
> hopefully they'll be able to give some feedback on it to this 
> list.

I'm one - and will feedback here when I've had a bit more time
to thrash about with it offline.

> One issue that I hit in putting that together is that there's 
> currently no way to determine an XMLReaders feature support 
> until _after_ it's been created. That's a bit awkward, since if 
> we have multiple implementations available we might want to use 
> the feature set to decide which one to instantiate ...

Believe this to be a Good Idea in general, whether coupled or not
with the search component of Miles' stuff. Given a well known
interface, even if search doesn't play lightweight & general enough
for SAX2, then at least there's an official hook to use in creating
variations on Miles' theme. Creating a weighty parser like Xerces,
then inspecting feature, is expensive - a well known static method
*seems* like it would be much faster. Guess I'll try it and get back...

BTW, David, all of the hard work is appreciated... Won't have a chance
to take it out for a spin 'til Monday (but between you and
Miles I have plenty of reading material for the weekend ;-)

Regards,

Jim Layer  

____________________________________________________________________
Get your own FREE, personal Netscape WebMail account today at http://webmail.netscape.com.

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ or CD-ROM/ISBN 981-02-3594-1
Unsubscribe by posting to majordom at ic.ac.uk the message
unsubscribe xml-dev  (or)
unsubscribe xml-dev your-subscribed-email at your-subscribed-address

Please note: New list subscriptions now closed in preparation for transfer to OASIS.