Hi Glenn and Arun, I'll make the fix for the namespace-prefix feature documentation. The other inconsistencies though are more serious, since they imply a clash between the XML 1.0 spec and SAX. Has anyone brought this up on the SAX list?
Since this change would impact on backward compatibility in Xerces1, I wouldn't think it would be appropriate to make it there. But in the case of Xerces2, if SAX really wants things that way then I can't see how we can avoid making this change. After all, the user still has the option whether to get the warnings or not--the parser is just biased towards giving them to him. Cheers, Neil Neil Graham XML Parser Development IBM Toronto Lab Phone: 905-413-3519, T/L 969-3519 E-mail: [EMAIL PROTECTED] Arun Yadav <[EMAIL PROTECTED]> on 11/13/2001 01:01:51 PM Please respond to [EMAIL PROTECTED] To: [EMAIL PROTECTED] cc: Subject: Re: patch for SAXParser and XMLValidator[xerces1.4.3] Glenn wrote:- > From "http://sax.sourceforge.net/?selected=namespaces": > > The http://xml.org/sax/features/namespace-prefixes feature controls > the reporting of qNames and Namespace declarations (xmlns* attributes): > when this feature is false (the default), qNames may optionally be > reported as empty strings for elements and attributes that have an > associated namespace URI, and xmlns* attributes must not be reported. > > If it is documented as being "true" somewhere, then that is a doc bug, > not a code bug. I agreed with u , that means it is doc. bug , so xerces feature list doc. should be updated i.e, http://xml.apache.org/xerces-j/features.html and also Xerces2 feature page i.e, http://xml.apache.org/xerces2-j/features.html. > > I do not think that the other two changes are consistent with the > XML spec and I believe that those changes should not me made. It has > always been the case that the phrase "an XML processor may at user > option issue a warning" is taken to mean "if the users asks you to". > This has been the behavior of Xerces at least as far back as 1.0.2, > which is the oldest copy I happen to have readily available. Were > the default behavior of the parser to change to start squawking at > users to warn them about constructs that the spec considers to be > legal in well-formed documents I for one would be most annoyed. If it is so , this should be update in feature list doc. It is documented in feature list that the default value warn-on-undeclared-elemdef and warn-on-duplicate-attdef is true for xerces1.x and also in xerces2. Regards, Arun Sun Microsystem, Inc. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
