Series: Penn State Logic Seminar

Date: Tuesday, January 23, 2001

Time: 2:30 - 3:20 PM

Place: 316 Willard Building

Speaker: Matthew S. Davis, Computer Science, Penn State

Title: Adding Subrecords to the ML Family of Languages

Abstract: 

Many programming languages are useful because they allow programmers
to extend the language by introducing new functions.  These new
functions are often monomorphic - they are defined to operate on a
specific "kind" of value.  For example, take a function A that
determines the length of a list of integers.  In order to determine
the length of a list of floating point values, a distinct function B
is necessary, even though the code will more than likely be similar if
not identical to the code in A.

ML-like programming languages utilize a very simple yet powerful
concept called polymorphic functions.  These functions are defined
once but are capable of operating on many different "kinds" of values.
So we could define a function that operates on lists of any kind of
values.  This has obvious advantages.

Most ML-like languages do not allow polymorphic functions to be
defined that operate on records.  This is unfortunate because there is
a natural way to describe "similar kinds of records."  This talk will
address this discrepancy and provide one possible solution to the
problem.