> Date: Wed, 8 Feb 2006 11:54:09 -0500
> From: Will Sexton <[log in to unmask]>
> 
> SRW/U specifies an optional 'numberOfRecords' sub-parameter in the
> response for a scan operation.  I'm wondering how best to aggregate
> the 'numberOfRecords' data according to different fields.  For
> example, if a search index contains records from multiple
> collections, you might want to break down the number of records from
> each collection.  Or, by a particular metadata field, such as
> dc.type.
> 
> I can't find anything built into the specification that explicitly
> allows this kind of kind of aggregation.  Is it something one would
> do with an 'extraTermData' extension, similar to the 'requestedTerm'
> example?

Yes, precisely.

> Any advice is appreciated, including thoughts on the general
> question of the advisability of hacking on the SRW/U spec to meet
> local needs.

Depends what you mean by "hacking on" the spec.  You absolutely
shouldn't _change_ anything in the spec for your local implementation;
but it's totally reasonable to _add_ extensions within the extension
framework.  Positively encouraged, even.

 _/|_	 ___________________________________________________________________
/o ) \/  Mike Taylor  <[log in to unmask]>  http://www.miketaylor.org.uk
)_v__/\  "The cladistic defintion of Aves is: an unimportant offshoot of
	 the much cooler dinosaur family which somehow managed to survive
	 the K/T boundry intact" -- Eric Lurio.