Re: A simple guy with a simple problem

["Simon St.Laurent" <simonstl@simonstl.com>]

At 04:00 PM 3/15/01 -0500, Jonathan Borden wrote:
>Right, and so we define 'equivalence' between two documents as
>operationally:
>
>1) parsing two documents produces the same SAX events
>2) parsing two documents produces an equivalent DOM object
>
>(that's the simple way of saying that these two documents have the same
>infoset)

Except that isn't always good enough, for regulatory reasons if nothing 
else.  (Just as it isn't wise to throw away original faxes and order forms 
when the data's in the computer.)  Remember, there are still people who 
don't find post-parse infosets to be a complete enough answer.

>ok, so the point about validating vs. nonvalidating parsers is a good one
>and so just tell Bob that if he accepts XML documents from external sources
>with DOCTYPE definitions, then be sure to use a validating XML parser (or at
>least one that handles external entities).

A validating parser won't work when he has to deal with docs with no 
DTD.  The non-validating that handles entities may be useful, so long as he 
wasn't planning to slap on a DOCTYPE (per my DOCTYPEChanger) and validate 
that way - the possibilities of internal subsets get creative and weird.

I guess this keeps my book-writing gig in business - XML Elements of Style 
goes into detail on these issues - but I can't say these are the reasons I 
want people buying books.

>no fancy wrench needed.

Too much understanding needed, though, for too little gain.



Simon St.Laurent - Associate Editor, O'Reilly and Associates
XML Elements of Style / XML: A Primer, 2nd Ed.
XHTML: Migrating Toward XML
http://www.simonstl.com - XML essays and books