The synthesis of a Java card tokenisation algorithm

Abstract

:We describe the development of a Java bytecode optimisation algorithm by the methodology of program extraction. We develop the algorithm as a collection of proofs and definitions in the Coq proof assistant, and then use Coq's extraction mechanism to automatically generate a program in OCaml. The extraction methodology guarantees that this program is correct. We discuss the feasibility of the methodology and suggest some improvements that could be made. The authors and the University of Edinburgh retain the right to reproduce and publish this paper for non-commercial purposes.Permission is granted for this report to be reproduced by others for non-commercial purposes as long as this copyright notice is reprinted in full in any reproduction. Applications to make other use of the material should be addressed in the first instance to Copyright Permissions, Division of Informatics, The University of Edinburgh, 80 South Bridge, Edinburgh EH1 1HN, Scotland. The Synthesis of a Java Card Tokenisation Algorithm

Extracting a Data Flow Analyser in Constructive Logic

“…Genet et al [GJKP03] use the generic proof assistant PVS for proving the correctness of algorithms for transforming Java Card byte code into the more compact CAP format. Similar works was done by Denney [Den01], using the program extraction mechanism of Coq. These optimisations do not involve any sophisticated static analysis.…”

Extracting a data flow analyser in constructive logic

“…Genet et al [11] use the generic proof assistant PVS for proving the correctness of algorithms for transforming Java Card byte code into the more compact CAP format. Similar work was done by Denney [10], using the program extraction mechanism of Coq. These optimizations do not involve any sophisticated static analysis.…”

“…We construct a function analyse of dependent type (P : Program) → State P which must verify ∀P : Program, P ٛanalyse(P ) (10) In addition, we want to obtain a non-trivial solution of the constraint system: e.g., an analyser returning the top element of the lattice for any input is a correct solution, but of poor interest. We thus add the requirement that our solution is the least solution of the constraint system:…”

xmonad in Coq (experience report)

“…The surprising lesson, for me at least, was the amount of effort it required to smooth out all the niggling issues with interfacing with other Haskell libraries, realizing axioms, custom module headers, type classes, and so forth. [Denney 2001;Filliâtre and Letouzey 2004;Leroy 2006]. Using extraction successfully requires a close tie between the theorem prover and target programming language.…”