query language

[Bob Kline]

On Mon, 31 Jan 2000, Simon St.Laurent wrote:

> Querying XML has been a pretty regular issue on this list, and I
> thought folks might want to see the Requirements draft that came out
> today:
> 
> http://www.w3.org/TR/xmlquery-req
> 

I posted a question on this topic a week or so back, but it seems to
have fallen through the cracks as I haven't seen any responses, so I'll
ask again.

There doesn't seem to be any recognition (neither in the XQL drafts nor
in the XML Query requirements you just cited) of the need to refer to
the results of one query in the formulation of a subsequent query.

This could take two forms.  In the more modest of the two, it would be
useful if XQL (and whatever becomes the official recommendation for
querying XML, if it is not XQL) had the moral equivalent of the "IN
(...)" clause of SQL.  This would allow the somewhat more elegant and
less verbose enumeration of the list of values obtained from the earlier
query when formulating the new query.  For example:

//CdrDoc[@Type="Physician and DocId in ("cdr02345","cdr45198",...)]/DocId

rather than
            
//CdrDoc[@Type="Physician and
        (DocId="cdr02345" or
         DocId="cdr45198" or
         ...)]/DocId

This addition to the existing XQL syntax seems to us to be a natural
extension of the set of operators currently available.

A significantly more powerful addition would be the ability to use this
IN clause with a URI referring to an external list of values.  Such an
extension would support combination of query results without requiring
the client to embed potentially very long lists of values within the
body of the query proper.  There are obviously decisions which have to
be made about some of the details (for example, the format in which the
values are represented; the flavor of the URI, etc.).  We could go off
and come up with our own solutions, but it seems silly for lots of
different solutions to spring up, all incompatible with each other,
rather than adopting a common interoperable approach.

So, I'll ask once again: has anyone else tackled this problem?  Has a
solution been hammered out which we just haven't succeeded in finding
yet (for example, is there a later version of the XQL draft than the
one from last summer)?

-- 
Bob Kline
mailto:bkline at rksystems.com
http://www.rksystems.com



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ or CD-ROM/ISBN 981-02-3594-1
Please note: New list subscriptions and unsubscriptions
are  now ***CLOSED*** in preparation for list transfer to OASIS.