Intrallect intraLibrary 2.8: Integration Guide

This guide is for those responsible for managing software system which will need to integrate with intraLibrary.

The following integration points are described in this guide.

Metadata Harvesting: The metadata harvesting interface, based on the Open Archive Initiative - Protocol for Metadata Harvesting (OAI-PMH), allows the metadata from intraLibrary to be harvested in XML form for use in other systems such as catalogues, aggregators and archives.
Search and Retrieve: The Search and Retrieve URL (SRU) and Search and Retrieve Web-Service (SRW) are related specifications, which both support searching a remote catalogue and retrieving search results.
File Transfer: How organisations can automate the upload of packages and files.
Digital Repositories Interoperability: This interface offers a subset the repository services defined in the IMS Digital Repositories Interoperability Specification.
Repository Java-API: Deeper access to the underlying functionality of intraLibrary, used for writing custom integration tools and extensions.
Authentication: Various ways intraLibrary can share and exchange user authentication information with other systems.

An IntraLibrary repository can expose metadata for harvesting by other systems, using the OAI-PMH . This section describes how intraLibrary supports OAI-PMH. For full details of the protocol visit the OAI web site.

By default, the collections managed in an intraLibrary are not harvestable. An administrator-class user of intraLibrary can make a collection harvestable changing a single property of the collection.

The entry point for harvesting metadata under OAI-PMH is from intrlibrary is the root URL for the intraLibrary installation, with "/IntraLibrary-OAI" appended to it:

http://<intralibrary-root>/IntraLibrary-OAI

Metadata can be harvested in the following formats:

Name	Code
Dublin Core	dc
IMS Metadata	imsmd
IEEE LOM	lom
ODRL	o-ex

You can test the harvestability of a repository using the OAI Registry Explorer. This allows you to send commands to intraLibrary. Type in the URL including /IntraLibrary-OAI and then you can select commands such as "Identify", "List Metadata Formats", etc. Note that for some commands you need to use parameters.

IntraLibrary supports two related specifications for search and retrieve services, both open specifications published by the Library of Congress:

Search and Retrieve URL (SRU)
Search and Retrieve Web-Service (SRW), also known as "SRU Web Services"

Both these specifications support searching a remote catalogue and retrieving search results. They use the same query language, Common Query Language (CQL) and the similar XML formats for the results of an "explain" and "query" operation.

Search and Retrieve URL (SRU) is a REST-ful interface. This means it can be called with a simple HTTP (GET or POST) request, and returns results in plain XML. The entry point for the SRU service is:

http://<intraLibrary-root>/IntraLibrary-SRU

The entry point for the SRW service is:

http://<intralibrary-root>/services/SRW

The SOAP binding has been implemented as follows, based on guidance from the Library of Congress team and the standard WSDL for SRW:

Service Style: document/literal
Message Style: inline with no multi-refs
SOAP version: 1.1

Queries in SRU and SRW are defined in Common Query Language (CQL)". It is very straightforward to create simple queries in CQL, but it is also possible to create complex queries, with mulitple constraints.

IntraLibrary supports "Level 1" of CQL. To enable searches to be constrained by specific metadata fields, the Dublin Core context set is supported. There is currently no standard context set for IEEE LOM metadata.

Some simple queries:

world
dc.title=metadata
dc.creator="john smith"
dc.keyword=snow

IntraLibrary supports an additional context set, "collections", which enables searches to be constrained by collection. The index collection.name can be used as part of query to specify a particular collection. For example a query can:

constrain a results set to objects in one collection only
- dc.title = "cat" and collection.name = "e-Prints"
return the entire contents of a collection
- collection.name = "e-Prints"
obtain results from more than one collection
- dc.title = "cat" and (collection.name = "images" or collection.name = "learning objects")

It can return either DC (default) or LOM, depending on the value of "recordSchema" so you should have no problems. See the SRU explain for more details, http://bagel.intrallect.com:8080/intralibrary/IntraLibrary-SRU?version=1.1&operation=explain

IntraLibrary complies with the Base Profile Requirements for SRW. It conforms with "Level 1" of CQL.

IntraLibrary supports the DC context set, and returns metadata in Simple Dublin Core (default) and IEEE LOM formats.

Sometimes it is not practial to use the "upload" facility to put files and packages into intraLibrary. For example, a file/package may be too big to upload in a reasonable time over a restricted network collection, or there may be a large set of files that need to be transferred from an existing system. It is possible to provide other means for users to transfer files to the repository server, such as FTP, including Secure FTP (SFTP) and WebDav.

The "file import" action in an intraLibrary workflow allows users to import files from a specified folder on the server filesystem. The location of this folder is configurable by user group. Any FTP or WebDAV server can be used to give intraLibrary users controlled access to transfer files to the appropriate folder on the server. If you would like some more information about this, please contact Intrallect.

It is possible to automate the deposit of files/packages into the repository using the Repository Java API. A sample implementation is provided with the product.

IntraLibrary has a fully documented "repository" Java-API. This API enables Java classes running in the same application context (webapp) as intraLibrary to access aspects of the intraLibrary business layer.

The Repository Java-API can be used to enable external systems to:

gather information about the status of a repository
create and manage users
trigger workflow events
query the structure of a workflow and find learning objects within it
search for objects

The following support information is provided with the product:

complete set of JavaDocs
some sample implementations, including full source code and installation instructions:
- adding users to intraLibrary
- uploading multiple files into intraLibrary
- searching for learning objects in intraLibrary

Intrallect plan to implement the following interfaces in intraLibrary:

Harvesting from other repositories via OAI-PMH.
Searching other catalgoues via SRU/SRW.
Remote deposit into an intraLibrary repository via the JISC Deposit API, when the specification is finalised.
Enable an intraLibrary repository to be searched via the FIRE interface.

Intrallect are considering the following interfaces for implementation in intraLibrary:

Transfer content packages to/from intraLibrary using AICC's PENS specification
Manage user accounts in intraLibrary using the IMS Enterprise specification
Authorisation using Shibboleth
Open URL