- From: David A. Lee <dlee@calldei.com>
- Date: Sat, 27 Dec 2008 12:46:53 -0500
- To: "Norman Walsh" <ndw@nwalsh.com>, "XProc Dev" <xproc-dev@w3.org>
Thanks for the discussion and pointers. Especially about serialization ideas. I'm particularly encouraged by your comment about "roll your own" because I too am not too happy with the existing streaming API's either but thought perhaps it was just my ignorance and maybe I was missing something everyone else thought was obvious :) ( although I think Binary XML could be relevent exactly because it is an alternate serialization of the infoset ... maybe its an option if it preserves the entire infoset ... but my first pass over that spec was that its prety complicated and likely inefficient ... but maybe a minimalist implementation would be a viable option) One thing. My read of "Except where the semantics of a step explicitly require changes, processors are required to preserve the information in the documents and fragments they manipulate. In particular, the information corresponding to the [Infoset] properties [attributes], [base URI], [children], [local name], [namespace name], [normalized value], [owner], and [parent] must be preserved." --- In particular the fragment "[attributes] ... must be preserved" implies that adding xml:base attributes without explicitly requested is in direct violation of the spec. the logic (in my head) is that the [attributes] properties is not preserved if you add to them. Thus an implementation that adds any attributes (even magic ones like xml:base or xml:id or xmlns) without explicitly requested is in violation .... -David ----------------------------------------------------------- David A. Lee dlee@calldei.com http://www.calldei.com http://www.xmlsh.org
Received on Saturday, 27 December 2008 17:47:41 UTC