IIIF Experimentation at Georgetown - June 2017

We installed the Cantaloupe Image Server on one of our test boxes.

We configured the image server to use an HttpResolver to retrieve bitstream assets from DSpace.  This was configured through the delegates.rb file of the server.

 module HttpResolver

    ##
    # @param identifier [String] Image identifier: [Handle Prefix]_[Handle Suffix]_[Bitstream Sequence]
    # @return [String,nil] URL of the image corresponding to the given
    #                      identifier, or nil if not found.
    #
    def self.get_url(identifier)
		identifier.gsub(/^([0-9]+)_([0-9]+)_([0-9]+).*/, "https://repository-dev.library.georgetown.edu/bitstream/handle/\\1/\\2/?sequence=\\3")
    end

Using this resolver, we were able to retrieve DSpace bitstreams 

Viewer Access

Although the images were restricted to our network, we were able to use some cloud-based IIIF viewers (such as Stanford's Image Cropping Tool) to view the images.  These viewers must be using the browser to perform the asset retrieval.

We were unable to view these assets in Universal Viewer or Mirador until we (1) created a manifest file and (2) made the manifest file publicly accessible.

We created the manifest file using the following online manifest editor.

We posted the manifest file as a gist file and then referenced that gist file in Universal Viewer and Mirador.