The Delphin Project

horizontal rule

Home
Downloads
Publications
Contact Us

Objective

The objective of this research is to develop a novel programming method that embraces logical framework technology at its core. To achieve this goal we propose to design and implement a programming language called Delphin that allows users to program with proofs, theorems, algorithms, code, and in general all kind of knowledge and its properties as if they were numbers, lists, trees, or arrays. The representation is direct and elegant, because common concepts such as environments, variables, substitutions and so on are implicitly provided and need not be explicitly programmed. Delphin will have the following features:

bulletIt will be based on the logical framework LF. One goal is to design Delphin in such a way that it can be easily extended to other logical frameworks in the future.
bulletIt will have a sound type system and an appropriate operational semantics.
bulletIt will be implemented in Standard ML.

Delphin

Delphin has been an ongoing project for many years.  We have experimented with many different possibilities (available here), culminating in an elegant logically-motivated system with a ∇-construct similar to, albeit different from, Miller's.

What's  New

bullet

Programming with Higher-Order Encodings and Dependent Types. (ESOP 08).
bullet

Technical Report (TR-1375) (PDF)

bullet

Example File: examples08.d

bullet

More Info (including source code)

Related

bullet

Miller and Tiu, A Proof Theory for Generic Judgments: An extended abstract. (PDF)

 

Elphin

Towards this end, we first rigorously designed and implemented the ∇-calculus, a calculus for defining general recursive functions over higher-order encodings.  Elphin is limited to the simply-typed fragment and follows a more ad-hoc stack-based solution which did not extend well to dependent-types.

 
bullet

Runnable Implementation of Elphin is available, including many examples!

 

Acknowledgements

This material is based upon work supported by the National Science Foundation under Grant No. CCR-0133502. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation (NSF).

horizontal rule

Home | Downloads | Publications | Contact Us

For problems or questions regarding this web contact poswolsky@cs.yale.edu.
Last updated: 11/14/07.