An Architecture for Editing Complex Digital Documents

In several on-going projects we were faced with the dilemma of how to reconcile our goal of delivering standardly encoded digital editions, yet have the actual editing and annotation performed by researchers and students who had no knowledge of XML and the Text Encoding Initiative Guidelines (TEI), and, for the most part, no great interest in learning them. The developed solution consists of allowing the annotators use familiar and flexible editors, such as Microsoft Word (for structural annotation of documents) and Excel (for word-level linguistic annotation) and automatically converting these into XML/TEI. Given the unconstrained nature of such editors this sounds like a recipe for disaster. But the solution crucially depends on a dedicated Web service, to which the annotators can up-load their files; the service then, via XSLT scripts, converts them to the canonical encoding in XML/TEI, and from it back to a visual format, either HTML or Excel XML. These files are returned to the annotators, giving them immediate feedback about the quality of their encoding in the source, and can thus correct errors before they accumulate. The paper describes the web service and details its use in a project compiling a digital library and lemmatised corpus of XIX century Slovene books.