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

Re: If SAX == lex, then what == yacc?

• From: "Anthony B. Coates" <abcoates@T...>
• To: "W3C XML Developers' List" <xml-dev@i...>
• Date: Wed, 24 Nov 1999 00:44:17 AET

** Reply to message from "James Tauber" <jtauber@j...> on Mon, 22 Nov
1999 09:20:43 -0500

Dear James,

	What you describe (included below) strikes me as using SAX events to build a
not-quite-DOM tree, which is exactly what I would expect FOP to do.  However, I
can envisage applications, operating on either large or streaming XML files,
where building a tree is not an appropriate solution.  Compilers (traditionally)
don't build a tree version of a source file before they start outputting the
object code to a file.  This is the kind of operation I am imagining.  I know I
myself would find it useful.  Mind you, I will be surprised if nobody has looked
into writing something like an XML version of 'yacc'.

	If someone published an XML Schema for C programs (don't laugh ...), what
would be the best way to write a compiler that reads this format but otherwise
runs like a typical C compiler?  Is there something that supercedes 'yacc' for
XML purposes?  I would be grateful for any suggestions from anyone.

	Cheers,
			Tony.

> FOP (and I suspect a lot of XML applications) takes SAX events and builds a
> tree structure out of objects that belong to a class depending on the
> element type or attribute. This tree building is handled through a largely
> generic builder and the mappings from element and attribute name to class
> name are provided to the builder at run-time. In the case of attributes (and
> perhaps eventually elements), the actual classes themselves are generated
> from XML schemas.
> 
> I don't know if it is the equivalent of "yacc" but it seems to be a fairly
> generic way of dealing with event streams that still allows
> application-specific tailoring.


** Anthony B. Coates    >> mailto:abcoates@T... <<
** Software Engineer (Java).  This is a 100% Pure Java e-mail.
** AJUG National Secretary (see http://www.ajug.org.au/)

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

• Follow-Ups:
  • Re: If SAX == lex, then what == yacc?
    • From: Sean Chen <schen@f...>

• References:
  • If SAX == lex, then what == yacc?
    • From: "Anthony B. Coates" <abcoates@T...>

• Prev by Date: Re: What do XML consultants actually do?
• Next by Date: Re: facts, conclusions, and exhortations re XML (long)
• Previous by thread: Re: If SAX == lex, then what == yacc?
• Next by thread: Re: If SAX == lex, then what == yacc?
• Index(es):
  • Date
  • Thread