Print

Print


A summary of the Scan discussion.

Issue:

We need to be able to browse/scan. This was always known but not defined
under v1.0.

Discussion:

*  We need a new operation with new request/response.
   - New Service:  How to duplicate Explain? How to link from service to
     service?  Rejected.
   - Adding to existing operaion:  Ugh!

*  Need to retrofit SRU with an operation field.
   - Default is 'searchRetrieve', can be set to 'scan'.

*  Request:

Rob's Version:
index           string          The index on which to scan.
term            string          The term to scan around/at.
numRequested    integer         Number of terms in the response
termPosition    integer         Position of the term in the response
stepSize        integer         Terms omitted between consecutive in resp.

Do we need stepSize? (I've never used it)

Matthew's version:
term            CQLtype         CQL searchClause
xTerm           XCQL:searchClauseType XCQL searchClause
numTerms        integer         aka numRequested
Position        integer         aka termPosition

Joe:
Do we need position?

Rob/Matthew:
Yes :)  Need it to cursor properly.  Useful locations are:
term is before first/after last in response
term is first/last in response
term is centered in response

Also some interfaces have a set number before the term (eg 5) and then
user requested afterwards.

Rob:
Splitting index and term is good, IMO. Otherwise what to do with the
Relation? Semantics of scan: dc.title > foo  vs dc.title < foo ?

Response:

List [Array? ;)] of terms with following subfields:

termValue
termFrequency
displayTerm
alternativeTerms ( list of alternativeTerm? )
totalTerms ?


Jannifer adds:

database / server address
CQL for direct access (index plus term)
Occurrence count of term in database, vs occs of term in Z39.50 scan)

----
In a library catalogue, scan indexes are typically of two types - those
that link directly to the bibliographic database and those that link
directly to the authority database and indirectly to the bibliographic
database. In the first case for each term there will be x (e.g. 12)
occurrences in the bibliographic database indicated by global occurrences.
One search, e.g. by internal identifier will retrieve the 12 occurrences
and nothing else.

In the second case there will be x occurrences (e.g. 2) in the authority
database indicated by global occurrences. Each authority record will also
have x occurrences in the bibliographic database.
----

--
      ,'/:.          Rob Sanderson ([log in to unmask])
    ,'-/::::.        http://www.o-r-g.org/~azaroth/
  ,'--/::(@)::.      Special Collections and Archives, extension 3142
,'---/::::::::::.    Twin Cathedrals:  telnet: liverpool.o-r-g.org 7777
____/:::::::::::::.              WWW:  http://liverpool.o-r-g.org:8000/
I L L U M I N A T I