AnnoCultor - Technical overview

Technically speaking, AnnoCultor is the following:

a wrap-up on a SAX parser that extracts domain objects from their XML representation,
an object conversion rule that checks if a domain object should be converted and invokes property rules,
property conversion rules that rename specific properties, modify their values, and write RDF triples to named graphs,
named graphs that collect triples and store them in RDF files,
and more...
In AnnoCultor we build converters from the following classes (outdated):

Converter than are extended by users to create converters specific for a dataset.
Task than has a few documenting fields and a link to the Environment.
Environment specifies basic file locations, such as the temp directory, and loads vocabularies from their RDF files. You are likely to override it.
Graph, or RDF graphs that are created by converters as their output; these graphs are typically stored in separate RDF files. It is common to create a graph for works, and graph for their images, one for directory of people, etc.
During conversion we examine the XML tree of XML tags and their values and need to select the XML subtrees that correspond to domain objects that we represent with RDF resources in the output RDF.
ObjectRule, linked to a Task, specifies how RDF objects can be found in the XML files. In an Object Rule we specify the object separating path that separates domain objects in their XML representation. These objects are represented with the AnnoCultor class DataObject.
PropertyRules apply to an XML tag to create one or more RDF triples, typically retaining (or translating) the value of the XML tag.
AnnoCultor provides a number of standard rules to create RDF resources, literals, modify tag values, etc. In addition, it defines several lookup rules (LookupPersonRule, LookupPlaceRule, LookupTermRule) that allow looking the values up in external vocabularies and replacing them with the corresponding vocabulary codes.