Several generic programs for converting values from regular datatypes to some other format, together with their corresponding inverses, are constructed. Among the formats considered are shape plus contents, compact bit streams and pretty printed strings. The different data conversion programs are constructed using John Hughes' arrow combinators along with a proof that printing (from a regular datatype to another format) followed by parsing (from that format back to the regular datatype) is the identity. The printers and parsers are described in PolyP, a polytypic extension of the functional language Haskell.

Corecursive programs produce values of greatest fixpoint types, in contrast to recursive programs, which consume values of least fixpoint types. There are a number of widely used methods for proving properties of corecursive programs, including fixpoint induction, the take lemma, and coinduction. However, these methods are all rather low-level, in the sense that they do not exploit the common structure that is often present in corecursive definitions. We argue for a more structured approach to proving properties of corecursive programs. In particular, we show that by writing corecursive programs using an operator called unfold that encapsulates a common pattern of corecursive de nition, we can then use high-level algebraic properties of this operator to conduct proofs in a purely calculational style that avoids the use of either induction or coinduction.

This paper explores the idea of generic programming in which programs are parameterised by data types. Part of the constructive theory of lists, specically the part dealing with properties of segments, is generalised in two ways: from lists to arbitrary inductive data types, and from functions to relations. The new theory is used to solve a generic problem about segments. 1 Introduction To what extent is it possible to construct programs without knowing exactly what data types are involved? At rst sight this may seem a strange question, but consider the case of pattern matching. Over lists, this problem can be formulated in terms of two strings, a pattern and a text; the object is to determine if and where the pattern occurs as a segment of the text. Now, pattern matching can be generalised to other data types, including arrays and trees of various kinds; the essential step is to be able to dene the notion of `segment' in these types. So the intriguing question arises: can one...