Since the WorldCat Grid API became available, I’ve spent odd periods of time playing with the services to see what kinds of things I could do and not do with the information being made available. This was partly due to a desire to create an open source version of WorldCat Local. It wouldn’t have been functionally complete (lacking facets, some data, etc), but it would be good enough, most likely, for folks wanting to switch to a catalog interface that utilized WorldCat as their source database, rather than their local ILS dbs. And while the project was pretty much completed, the present service terms the are attached to the API make this type of work out of bounds at present (though, I knew that when I started — it was more experimental with the hope that by the time I was finished, the API would be more open).
Anyway…as part of the process, I started working on some rudimentary components for ruby to make processing the OCLC provided data easier. Essentially, creating a handful of convenience functions that would make it much easier to call and utilize that data the grid services provided — my initial component was called wcapi.
Tomorrow (or this morning if you are in Providence), a few folks from OCLC will be putting on a preconference related to the OCLC API at Code4Lib. There, folks from OCLC will be giving some background and insight into the development of API, as well as looking at how the API works and some sample code snippets. I’d hoped to be there (I’m signed up), but an airline snafu has me stranded in transit until tomorrow — so by the time I get to Providence, all of, or most of the pre-conference will be over. Too bad, I know. Fortunately, there will be someone else from Oregon State in the audience, so I should be able to get some notes.
However, there’s no reason why I can’t participate, even while Iâ€?m jetting towards Providence — and so I submit my wcapi gem/code to the preconference folks to use however they see fit. It doesn’t have a rakefile or docs, but once you install the gem, there is a test.rb file in the wcapi folder that demonstrates how to use all the functions. The wrapper functions accept the same values as the API — and like the API, you need to have your OCLC developer’s key to make it work. The wcapi gem will utilize libxml for most xml processing (if it’s on your machine, otherwise it will downgrade to rexml), save for a handful of functions that done use named namespaces within the resultset (this confuses libxml for some reason — though their is a way around it, I just didn’t have the time to add it).
So, the gem/code can be found at: wcapi.tar.gz. Since I’ll likely make use of this in the future, I’ll likely move this to source forge in the near future.
*******Example OpenSearch Call*******
require ‘rubygems’
require ‘wcapi’
client = WCAPI::Client.new :wskey => ‘[insert_your_own_key_here’
response = client.OpenSearch(:q=>’civil war’, :format=>’atom’, :start => ‘1’, :count => ’25’, :cformat => ‘mla’)
puts "Total Results: " + response.header["totalResults"]
response.records.each {|rec|
puts "Title: " + rec[:title] + "\n"
puts "URL: " + rec[:link] + "\n"
puts "OCLC #: " + rec[:id] + "\n"
puts "Description: " + rec[:summary] + "\n"
puts "citation: " + rec[:citation] + "\n"
}
********End Sample Call****************
And that’s it. Pretty straightforward. So, I hope that everyone has a great pre-conference. I wish that I was there (and with some luck, I may catch the 2nd half) — but if not, I’ll see everyone in Providence Monday evening.
–TR
Comments
One response to “WorldCat API: draft ruby gem”
One note, I should preface the above where I note: “And while the project was pretty much completed, the present service terms the are attached to the API make this type of work out of bounds at present” — this is related to the restrictions on service levels and the data that is provided. My vision for an open source version of WorldCat Local was to provide an open replacement option for the ILS utilizing WorldCat data. And this is possible and folks have done it (David Walker I believe has integrated the WorldCat API into the Xerces project). The main difference between that work and my interest is I want to provide the interface without need for authentication. At present, to use the full service level, you must authenticate all users — essentially the same way that you would for article databases. The difference however, is that you would be talking about one’s ILS and that becomes a much stickier issue — since large swaths of users that have no relation to an organization likely use and rely on an organization’s ILS interface. So, it is these limits that I’m alluding to. And as I say, these were limits that I was aware of, though, I honestly believe will be short-lived as OCLC continues to evaluate how to make their data more available to the membership. That seems to be the trend with their other services, so I certainly can see this eventually moving in that direction.
–TR