The Hippo Repository is primarily a plain JCR repository.  It consists of the Apache Jackrabbit repository and is augmented with some additional features and improvements.  Most of these features are available though the normal JCR API or through small extensions of this API.  For instance, the facteted search requires only the usage of the plain JCR API.  Browsing though a faceted search uses the same retrieval JCR API calls as browsing though a normal tree.  The tree you are browsing is not persisted, but obtained using searches of the data stored elsewhere is the repository

 

The facted search is an example where no API apart from the JCR API is needed.  The content is exposed or processed using different views, also known as virtual laying.  It is possible to obtain more information by casting certain JCR objects to a more specific Hippo variant.  For instance, to directly invoke workflow you can obtain the workflow manager object by casting a javax.jcr.Workspace to a HippoWorkspace class object which provides you access to the right method.

 

Most of the Hippo repository is transparently available through the JCR API:

• The LDAP, Single sign on and repository based login mangement, and the accompanying faceted authorization are part of the default infrastructure of Jackrabbit and JCR;
• Updating content is automatically run when necessary when the repository starts up;
• Derived data (comparable to extractors) to automatically fill in properties based upon other characteristics of a document is automatically performed upon a save;
• Improvements to import and export functionality.

The workflow is probably the most important non-JCR API.  The important reason is that workflow is exposed and implemented using Plain-Old Java Objects (POJOs).  This enables relatively high level interaction with the repository content and functionality, whereas the JCR API calls are fine grained property and node handling calls.