Page History

...

OpenSearch is a small set of conventions and documents for describing and using 'search engines', meaning any service that returns a set of results for a query. It is nearly ubiquitous‚Äîbut also nearly invisible‚Äîin invisible‚ in modern web sites with search capability. If you look at the page source of Wikipedia, Facebook, CNN, etc you will find buried a link element declaring OpenSearch support. It is very much a lowest-common-denominator abstraction (think Google box), but does provide a means to extend its expressive power. This first implementation for DSpace supports none of these extensions‚Äîmany of which are of potential value‚Äîso value‚ so it should be regarded as a foundation, not a finished solution. So the short answer is that DSpace appears as a 'search-engine' to OpenSearch-aware software.

...

• Browser IntegrationMany recent browsers (IE7+, FF2+) can detect, or 'autodiscover', links to the document describing the search engine. Thus you can easily add your or other DSpace instances to the drop-down list of search engines in your browser. This list typically appears in the upper right corner of the browser, with a search box. In Firefox, for example, when you visit a site supporting OpenSearch, the color of the drop-down list widget changes color, and if you open it to show the list of search engines, you are offered an opportunity to add the site to the list. IE works nearly the same way but instead labels the web sites 'search providers'. When you select a DSpace instance as the search engine and enter a search, you are simply sent to the regular search results page of the instance.
• Flexible, interesting RSS FeedsBecause one of the formats that OpenSearch specifies for its results is RSS (or Atom), you can turn any search query into an RSS feed. So if there are keywords highly discriminative of content in a collection or repository, these can be turned into a URL that a feed reader can subscribe to. Taken to the extreme, one could take any search a user makes, and dynamically compose an RSS feed URL for it in the page of returned results. To see an example, if you have a DSpace with OpenSearch enabled, try:

  Code Block
  http://dspace.mysite.edu/open-search/?query

  -

  =<your query>

  The <your query>The default format returned is Atom 1.0, so you should see an Atom document containing your search results.
• You can extend the syntax with a few other parameters, as follows:

  Parameter	Values
  format	atom, rss, html
  scope	<handle>—search is restricted to a collection or community with the indicated handle.
  rpp	number indicating the number of results per page (i.e. per request)
  start	number of page to start with (if paginating results)
  sort_by	number indicating sorting criteria (same as DSpace advanced search values

...