[Home] [By Thread] [By Date] [Recent Entries]
Tim Bray wrote: > > At 01:37 PM 04/01/98 -0500, David Megginson wrote: > > > Agreed. Could we change the first method to > > > startDocument(String root, String DTDSysID, String DTDPubID) > > > these obviously each being null in the event the document doesn't > > > provide them? -Tim > > > >This is a great idea, but it will require an implementation to queue > >events. For example, if I have > > > > <?foo bar?> > > <!DOCTYPE doc SYSTEM "doc.dtd"> > > > >I have to queue the PI event(s) until I have found the DOCTYPE > > Why? The PI is 'before' the doc in any meaningful sense... so they > get some PI events before they get the startDocument event. Is this > a problem? -Tim I think this is inconsistent with the XML spec: document ::= prolog element Misc* prolog ::= XMLDecl? Misc* (doctypedecl Misc*)? PIs before the doctypedecl are part of the document. I also feel that it's wrong to give special treatment to the external identifier in the doctypedecl. <!DOCTYPE doc SYSTEM "doc.dtd"> is just a convenient shorthand for <!DOCTYPE doc [ <!ENTITY % whatever SYSTEM "doc.dtd"> %doc; ]> James 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/ To (un)subscribe, mailto:majordomo@i... the following message; (un)subscribe xml-dev To subscribe to the digests, mailto:majordomo@i... the following message; subscribe xml-dev-digest List coordinator, Henry Rzepa (mailto:rzepa@i...)
|

Cart



