more on base concerns

In terms of XPath 2 we seem to be discussing whether:

http://www.w3.org/TR/xpath-datamodel/#dm-document-uri
[[
The dm:document-uri accessor returns the absolute URI of the resource 
from which the Document Node was constructed
]]

or

http://www.w3.org/TR/xpath-datamodel/#dm-base-uri
[[
The dm:base-uri accessor returns the base URI of a node [...], see [XML 
Base].
]]

is used to select the URI that is passed to the RDF parser reading the 
transform output.

There is also a URI passed to the XSLT processor, as in Java SAX terms, 
the 'systemID' of the SAXSource ...
I am pretty clear this should be dm:document-uri, the retrieval URI.

It gets muddier if we apply GRDDL to some node that is not a root node, 
but I guess that that lies outside our concern here.


i.e. document representation D retrieved from U.

GRDDL processor finds its baseURI B (via xml:base or html:base, or = U), 
and interprets any relative refs of transforms T using B.

GRDDL processor passes D to transform T along with D's URI, I believe 
that that is U in this case. In XSLT1.0 the transform cannot see U, so 
quite what is passed is moot. This produces output RX.

GRDDL processor passes RX to RDF/XML parser and a 'base-URI' which is 
either U or B.

I could live with either, but it's helpful if we decide.
It would also be nice if some spec told us.

Jeremy

-- 
Hewlett-Packard Limited
registered Office: Cain Road, Bracknell, Berks RG12 1HN
Registered No: 690597 England

Received on Thursday, 26 April 2007 20:30:30 UTC