Print

Print


My main reason for a distinction between requesting records via SRW and
searching via CQL is that at least in our case  both are requests for
different servers and I expect that there will be many more situations
in which the index server that gets the CQL query is quite different
from the database server(s) that gets the record request. In the current
situation the CQL may have to be parsed twice.

Theo


>>> [log in to unmask] 06/03 4:48 nm >>>
> Date: Tue, 3 Jun 2003 15:32:57 +0200
> From: Theo van Veen <[log in to unmask]>
>
> I have the impression that there still is a lot of confusion. URI's
> to address metadata records are different from URI's to adress the
> digital objects that are described by these metadata records.

Yes.

> The first one is part of the metadata profile (e.g. <dc: identifier
> xsi:type=dcterms:URI>) .

No, that's the second one.  Dublin core _defines_ metadata elements
which _describe_ (and in this case point to) data.  It doesn't define
record-data elements for describing metadata.

SRW/U does, however.  It provides for locating a particular metadata
record by means of two mechanisms:

* A record's unique ID _within its SRW/U database_, to be returned as
  a part of a record using the "id" element from the Record Metadata
  Schema described at
        http://srw.o-r-g.org/schemas/rec/1.0/

* That record-ID can be forged into a URI for the record using the CQL
  query rec.id=<value>, where the "id" index is in the Record Metadata
  Index Set described at
        http://srw.o-r-g.org/indexSets/rec/1.0/

So if I find a record that I like which searching an SRU database at
        http://foo.bar.org/sru/db1
and that record includes an element "id" (from the Record Metadata
Schema) with value "12368", then I can reliably retrieve that record
later with URI
        http://foo.bar.org/sru/db1?query=rec.id=12368

> Having a special SRW parameter for recordid (as well in the response
> as in the URL request) allows a clear distinction between requesting
> records via SRW and searching via CQL, but maybe wanting that is a
> matter of taste.

The so-called "distinction" is fictitious.

 _/|_
_______________________________________________________________
/o ) \/  Mike Taylor  <[log in to unmask]>
http://www.miketaylor.org.uk
)_v__/\  "To bleed the lyric for this song, to write the rites to
         right my wrongs" -- Marillion, "Script for a Jester's Tear"

--
Listen to my wife's new CD of kids' music, _Child's Play_, at
        http://www.pipedreaming.org.uk/childsplay/