CONTENTdm and OpenSearch

By reeset / On / In CONTENTdm, Digital Libraries

I love OpenSearch.  It’s been one of those things that I’ve been wanting to spend more time looking at — maybe incorporate into Dspace or some of our other services like LibraryFind (which actually is now on the todo list).  Anyway, folks may not know it, but Kyle Banerjee and I are writing a book.  A how to of sorts, for folks doing digital repositories.  I’ve been lights out for most of December cranking out 4 finished and 1 1/2 nearly completed chapters.  So far so good.  Well, one of the parts that of the book deals with exposing resources to larger audiences and a discussion of OpenSearch falls into that section.  As  I was looking through the specification this afternoon, I thought, wow, this would be easy to implement just about everywhere.   So I took a 1/2 hour, and quickly whipped up some code that integrated OpenSearch into CONTENTdm.  I’ll post the code shortly.  However, what I thought cool was the number of resources that have embraced OpenSearch as a query method.  IE7 for example, utilizes OpenSearch as the method for querying search providers.  This means that by adding an OpenSearch server to my Contentdm instance, I instantly am able to add this resource as a search target in IE 7.



As I mentioned, writing the code took about 30 minutes and was much easier than I’d anticipated.  Given the speed at which OpenSearch has caught on outside the library community (I was surprised at how many applications and services support it) and how simple it is to implement — I’m thinking that its almost crazy not to spend the time and integrate the protocol into our organization’s services if only to give developers outside the library community an straighter line for service integration.



4 thoughts on “CONTENTdm and OpenSearch

  1. “I’m thinking that its almost crazy not to spend the time and integrate the protocol into our organization’s services”

    I’m thinking you’re right. 🙂

    I’m also thinking that libraries could help show the way when creating opensearch extensions and, for instance, how to properly leverage the Query* element from the 1.1 draft spec. There’s some really powerful stuff buried in there.


  2. Dorothea,

    Thanks for the offer. The publisher has asked for a few names of individuals that might be willing to review the text when finished. I’ll pass yours along and let them know that you are willing.


  3. Mike,

    You are right of course. I think in part, what makes OpenSearch so elegant is the simplicity of the protocol while still allowing users to say, utilize SRU as a query mechanism within the OpenSearch protocol. There are just so many things that it does right.