David, Chris,
little thought experiment. Starting point: the info space (= sum total of all available XML documents) is a node forest. Now the experiment: (a) imagine a single invisible "space node" (analogous to the invisible document node) to be parent of all document nodes; (b) reinterpret doc("http://foo.xml") to mean: root()/../document-node()[document-uri(.) eq 'http://foo.xml'].
In other words: regard doc("http://foo.xml") to be a notation of the two-step navigation (i) up to the space node, (ii) down into the document node of interest. So that is my answer to Chris: the value of
doc("countries.xml")/countries/../..
is the space node. End of thought experiment.
Hans
Von: David Lee <dlee@calldei.com>
An: Chris Maloney <voldrani@gmail.com>; Michael Kay <mike@saxonica.com>
CC: "Costello, Roger L." <costello@mitre.org>; "xml-dev@lists.xml.org" <xml-dev@lists.xml.org>
Gesendet: 15:26 Mittwoch, 14.August 2013
Betreff: RE: [xml-dev] XPath and a continuous, uniform information space
For example, what does the following mean?
doc("countries.xml")/countries/../..
In the spirit of "The Information Space" that would be "The Information Space" which is not a node().
