An integrated architecture for scholarly digital editing

This proposal originated from the successful outcome of a Web technologies class that took place during a graduate course in Computer Science given at the University of Bologna. This course saw the cooperation between an expert in computer science, passionate about humanities, and a philologist interested in computational systems.

The project stems from the need to develop an integrated Web system of applications and services useful in order to manage literary texts annotated in XML/TEI (at different levels). Once given to students, organized in groups, the full TEI scheme, all the possible operations needed by a philologist, and derived from the use of TEI elements, attributes and classes, had to be contemplated.

As regard to the delivered XML/TEI text files, a number of solutions were proposed, different from both the linguistic and chronological point of view: Renaissance texts annotated for a multiple visualization (reading/scholarly); Latin but also contemporary texts with an apparatus in TEI “parallel texts” version, without accepted lemma, traditional “location-referenced” method, but also “double-end-point” method for managing overlapping structures; XV century autograph manuscripts annotated as primary sources (special letters, abbreviations, deletions, additions, etc..).

The task of each group was to prepare a different Web solution, designed to solve certain aspects of the editorial work of the philologist, and capable to interact with the applications created by other groups and integrated in other Web services. The overall aim was that each project wouldacquire all the possible applications, and unite them into a single system for managing digital editions.

Each single Web environment had to be able to allow different services: the construction of the stemma through the metadata; the collation on screen or the extraction of variant readings; the word index displayed as a concordance with frequency; the dynamic reconstruction of each witness through its own siglum; the visualization of different transcriptions (reading/scholarly); the management of the apparatus according to different modes (traditional, line spacing, in a separate window).

The final result was the creation of 12 Web applications, each one with original solutions, but each one with all services, some of them obtained from other groups : 12 web environments based on 12 different final-user interfaces.

The same service was developed by each groups in different ways, starting from a common technological basis: a server-side language like PHP or Perl or Java (recommended Java servlets on Tomcat) as a repository of documents, and as an environment for the creation of services and viewer (from TEI to HTML + CSS through XSLT); a client-side Javascript / Ajax, using a framework like GWT or ExtJS, for the selection of documents, the display and the activation of services.

The future purpose of the project is to test the functionality of these Web services and applications by submitting new XML/TEI files based on heterogeneous textual situations. By uploading new files it should be possible to understand the validity and extensibility of the provided services.