Our goal with this project is the traditional goal of Automatic Programming: we try to raise the level of programming in our domain and, thus, enable an increase in productivity, since the programmer won’t have to deal with low-level details.

We are working on data-structure abstractions and it would be very useful (both to us and to people using the library) to have a unifying model for our data structures. The cursor-container paradigm is this model. In other words, we view all data structures in our universe as pairs of containers and cursors (iterators). These two elements explicitly decouple the notions of element storage and element access. All the operations in DiSTiL are either performed on a cursor or on a container. For some data-structures this model is more intuitive than for others (for instance, it is easy to imagine a doubly linked list as a container with a cursor for traversal. A hash table, on the other hand may be more difficult to envision in this way). The model, however, is very general and can be used to represent a large variety of data structure constructs.