I've done a bit of updating against the Ockham Alerting Service.
This time of year, because things usually slow down, I always have time
to hack, and I have done a bit of hacking against the SRU-based Ockham
Alerting Service. Hacks include:
1. Re-write of the SRU server taking greater advantage of the SRU and
CQL Perl modules written by Ed Summers (good stuff) making the server
more compliant:
* http://alert.ockham.org/sru-server.cgi (link)
* http://alert.ockham.org/etc/sru-server.txt (present source code)
2. Re-write of the user interface acting as an SRU client on behalf of
the user-agent. This application takes input from a user-agent, creates
an SRU URL, sends it to the server, transforms the results, and lastly
provides additional functionality depending on the non-SRU input
garnered from the user-agent:
* http://alert.ockham.org/ (link)
* http://alert.ockham.org/etc/sru-client.txt (present source code)
* http://alert.ockham.org/?cmd=search&output=HTML&query=dodecahedron
(example query)
3. Search results sent to the SRU server through the user interface now
support linking to version 2.0 RSS feeds or emailing. These are
examples of "additional functionality" alluded to above in #2.
The to-do list is still very long, but I believe progress is being made:
* Allow users to create profiles.
* Demonstrate how to import MARC data.
* Demonstrate the use of MyLibrary as a place for the
centralized "pile."
* Enable implied Boolean intersection (and) searching
against the underlying SRU server as in the following
query: time frequency.
* Figure out how to deal with unicode display.
* Fill the implementation with more content.
* Implement a paging mechanism returning limted numbers
of search results at a time.
* Implement rudimentary Did You Mean functionality a la
Google.
* Improve documentation.
* Turn into more of a server and less of a
client/server combination.
* Support the SRU scan operation.
* Write a CSS file designed for print media.
--
Eric Lease Morgan
University Libraries of Notre Dame
|
