A Database Coprocessor for Haskell

Relational database management systems (RDBMSs) provide the best understood and most carefully engineered query processing infrastructure available today. However, RDBMSs are often operated as plain stores that do little more than reproduce stored data items for further processing outside the database host, in the general-purpose programming language heap. One reason for this is that the aforementioned query processing capabilities require mastering of advanced features of query languages (e.g., SQL) in addition to the general-purpose language the application is programmed in. Moreover, the query languages are often inadequately integrated into the host programming language. One way to solve these problems is to use RDBMSs as a coprocessor for general-purpose programming languages, for those program parts that carry out data-intensive and data-parallel computations. In this paper we present a library for database-supported program execution in Haskell. Data-intensive and data-parallel computations are expressed using familiar combinators from the standard list prelude and expressive list comprehension notation. The library, in addition to queries of basic types, supports computations over arbitrarily nested tuples and lists. The implementation minimises unnecessary data transfer and context switching between the database coprocessor and the programming language run-time by ensuring that the least possible number of queries is generated when executing the library functions. Although Haskell has inspired a number of language-integrated query facilities (most notably LINQ in Microsoft’s .NET framework), as far as we know this is the first proposal and implementation of database-supported program execution facility for Haskell.