- From: Simon Pieters <simonp@opera.com>
- Date: Wed, 03 Jun 2009 20:03:52 +0200
- To: "John Cowan" <cowan@ccil.org>, public-xml-core-wg@w3.org
On Wed, 03 Jun 2009 19:32:45 +0200, John Cowan <cowan@ccil.org> wrote: > Simon Pieters scripsit: > >> I think the issue here is that the XML spec doesn't define how to >> convert >> a stream of bytes into a parsed tree (in terms of some tree model -- >> HTML5 >> uses the DOM as the model but this does not restrict implementations to >> use DOM). The XML spec just states what is the allowed syntax, and the >> mapping to a tree model is implied. > > The Infoset spec does that. No, it just defines a model of a parsed tree, just like the DOM Core spec defines a separate model of a parsed tree. The mapping from bytes to the tree model is not explicitly defined like it is in HTML5. >> I also think it's an issue here that the XML spec doesn't say what >> an XML processor should do if it does not abort parsing upon a syntax >> error. > > No known XML processor does anything but abort parsing on any fatal > error. That's because if they don't, they are by definition not an XML processor. :-) There are user agents and libraries that process content that is labeled as XML and do not abort upon errors, though. > In particular, making any further changes to the DOM or issuing > any further SAX events other than endDocument is forbidden. See the > definition of 'fatal error' in 1.2. I understand that this is the case as currently specified. I was just trying to state my understanding of "HTML folks think the XML spec is broken because it doesn't define error recovery". -- Simon Pieters Opera Software
Received on Wednesday, 3 June 2009 18:04:47 UTC