Squire: a simple and efficient ingest tool for institutional repositories that utilises Fedora Commons

Squire: a simple and efficient ingest tool for institutional repositories that utilises Fedora Commons

One of the most important components of an Institutional Repository is the ingest mechanism. SQUIRE, developed in 2008, was created by programmers working within the repository space who had an understanding of this fundamental principle. It was funded by the ARROW project (Australian Research Repositories Online), and was created in a collaborative environment with the goal of meeting the needs of all involved. When using an ingest tool the user enters form data and uploads files via a web browser. Where do these files and form data go? Squire manages these files reliably and efficiently using Fedora Commons. The data and uploaded files go directly into Fedora as an object. This process is executed programatically in a way that strictly adheres to the Fedora API's. Squire is able to retrieve the submitted object for display in the users web browser. Once selected a submitted object can be modified or deleted. Each submitted object is able to accept multiple file uploads. These files are stored in Fedora as data streams belonging to the submitted object. Once fully approved, the submitted object is made visible to your Fedora Commons repository by once again communicating with Fedora, via its API's. One of the main requirements of repository managers was to have round trip functionality in their ingest tool. This functionality was previously not available in all repository solutions. This would enable them to comprehensively edit already submitted objects using the original submission form. Having this in mind during the design phase, helped create the current simple and effective architecture, reducing the complexity of the round trip task into an almost natural function. The round trip is made possible due to the fact that the form data is stored as a data stream in Fedora, the form can be easily re-populated at any time using this data, during the submission and review phases. Another significant challenge for repository managers when implementing an institutional repository is acquiring assistance from systems administrators with software installation and configuration. With this in mind the Java platform was chosen for developing Squire. Squire is able to be deployed by simply copying a single Java archive file into the servlet container that ships with Fedora Commons. This compatibility with Fedora Commons prevents the need to run separate servers or server processes for the repository software and ingest tool. By using SQUIRE, repository managers are able to make full use of Fedora's flexible functionality, at the same time as giving themselves the extra flexibility in editing objects within the repository. Squires simple architecture and common development platform means quick installation time.