W3C home > Mailing lists > Public > www-ql@w3.org > April to June 2004

RE: top-level location-path context dependencies

From: Howard Katz <howardk@fatdog.com>
Date: Sat, 17 Apr 2004 15:25:18 -0700
To: "Michael Kay" <mhk@mhk.me.uk>, <www-ql@w3.org>
Message-ID: <IKEOLCDFPBBPPAHGNKKOMEJHEKAA.howardk@fatdog.com>

responses inline below ...

> > I'm trying to understand the ways in
> > which implementations can effect the settings of the
> > top-level context item,
> > as outlined in "C.2 Dynamic Context Components" in the main
> > XQuery document> [1].
>
> The expectation is that implementations are likely to provide an API
> allowing the context item to be initialized. In Saxon, for example, the
> XQueryExpression#evaluate() method takes an argument of type
> DynamicQueryContext, and the DynamicQueryContext object has a method
> setContextItem() or setContextNode(). If invoked from the command
> line, the
> initial context item can be set to the document node of an XML document
> using the option "-s source.xml".

Any reason you couldn't set the context node to be *any* node within a
particular document (altho it would likely be a very odd thing to do)?

> >
> > To be more specific, assuming a two-document context ("bib.xml" and
> > "reviews.xml" from the use cases),
>
> You're using the term "context" loosely there, I think. The
> context item is
> always a single item. There is no such thing as a "two-document context".

You're right: very loosely.
>
> > what are the results of
> > the following
> > top-level queries? And which of these depends on the
> > particular environment
> > and which are never dependant on the environment?
>
> All of these queries return nodes from the document that contains the
> context node. In all cases the result nodes all come from the
> same document,
> and in all cases they depend on the way the context was
> initialized. If the
> context has not been initialized, they throw an error.

I think I see it -- maybe. Is the following reasoning correct then? If the
environment has set the context item to be the document node for "bib.xml"
[1], for example, you'd get

1) //title => 4 title elements in "bib.xml"
2) title => no elements (since no <title> element is child of the context
item)
3) /bib => the root bib element
4) bib => ditto

Similarly if the context item has been set to be the document node for
"reviews.xml" [2], you'd get

1) //title => 3 title elements in "reviews.xml"
2) title => no elements (no <title> element is child of the context item)
3) /bib => no elements (no <bib> element is at the root of the document
containing the context node
4) bib => no elements (no <bib> element is the child of the context node)

Howard

[1] http://www.w3.org/TR/xquery-use-cases/#xmp-data
[2] http://www.w3.org/TR/xquery-use-cases/#xmp-data-q5
> >
> > It would be very useful to me if somebody could show the
> > above results in a tabular fashion
>
> No table needed: there is no difference between these four cases.


>
> Michael Kay
>
Received on Saturday, 17 April 2004 18:24:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 22 July 2006 00:10:19 GMT