OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: Announcement: SAX Java Implementation (pre-release)

[ Lists Home | Date Index | Thread Index ]
  • From: Tyler Baker <tyler@infinet.com>
  • To: Tim Bray <tbray@textuality.com>
  • Date: Sun, 12 Apr 1998 20:33:45 -0400

Tim Bray wrote:

> At 05:33 PM 12/04/98 -0400, David Megginson wrote:
> >Here is my question: do we need SAXByteStream and SAXCharacterStream
> >interfaces?  If so, should SAX also come with default implementations
> >for each language?
> At the bottom level, a characterStream interface is a good idea.  Having
> constructors for these that take an asciiByteStream, an iso8859ByteStream
> and (I guess) a UTF8InputStream and produce seems awfully convenient. -Tim

This seems like a pretty reasonable idea.  This would also allow parser writers
to provide their own byte to character converters which may in fact be more
efficient than the sun.io.ByteToCharConverter and sun.io.CharToByteConverter
classes which do all of this stuff now in the JDK.  These implementations are
undocumented which is another problem that a CharacterStream interface might be
nice to have as an alternative to the Reader classes.  Of course a character
stream interface could of course delegate to a Reader class if the parser writer
chose two which would be a relatively low cost in performance considering the
Reader classes are all screwed up anyways IMHO.  In this particular case, a
CharacterStream implementation may take a Reader object as an argument and the
rest is just delegation.

Most poeple already know that the I/O implementation of the JDK is not too great
as it is, for example every time you read a single character from the Reader
class, it instantiates a new Array object as an argument necessary for delegating
to another read method.  Why things work this way in the Reader class I don't
really know other than possibly the person who wrote it was either lazy or else
forgot to improve the implementation at a later time.

I think Tim's idea is a good one...


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/
To (un)subscribe, mailto:majordomo@ic.ac.uk the following message;
(un)subscribe 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)


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS