********************************************************************* * Ecole Normale Supe'rieure * * * * Se'minaire * * SEMANTIQUE ET INTERPRETATION ABSTRAITE * * P. Cousot * * * * Vendredi, 14h--16h * * Salle W, 4e`me e'tage, toits du DMI * * DMI ENS 45 rue d'Ulm 75005 Paris * ********************************************************************* *** Vendredi 7 janvier 1993 ***************************************** A new scheme based on pruning for constructive negation in CLP and CC languages Francois FAGES (LIENS) Re'sume': Because of their importance in real-life applications, most constraint logic programming systems, such as CHIP, CLP(R) and Prolog III, include various higher-order predicates for op- timization. These constructs do not belong however to the formal framework of Jaffar and Lassez and they lack a declarative seman- tics. In this talk I shall present an approach based on the se- mantics of negation. Constructive negation and its compilative version named intensional negation, provide already correct and complete schemes w.r.t. the 3-valued interpretation of logic pro- grams with negation. But these rules are not easily amenable to practical implementations. A new scheme based on pruning in CSLD-derivation forests will be proposed and proved correct and complete w.r.t. the 3-valued completion of the program. We shall derive from that scheme a family of branch and bound like pro- cedures for optimization that can be practically implemented. Following the same line of development we shall also outline an extension of the CC family of languages with negation, and of its denotational semantics defined by non-linear closure operators. *** Vendredi 14 janvier 1993 **************************************** Improving Abstract Interpretations with Continuations Juarez Assumpcao MUYLAERT-FILHO (Imperial College, Londres) Re'sume' : In this talk we consider the benefits of continuations for abstract interpretations (a.i.) of programs written in simply-typed (lazy) functional languages. The a.i. framework we consider was first described in [BHA85] and uses finite domains. The idea is to do abstract interpretation of continuised terms. The basic slogan is continuations do improve abstract interpretations as they expose more properties that were otherwise hidden in the untranslated programs. [BHA85] Burn, Hankin and Abramsky, The Theory of Strictness Analysis for higher-order functions, LNCS 217, 1985. *** Vendredi 21 janvier 1993 **************************************** Rela^che. *** Vendredi 28 janvier 1993 **************************************** Proving Concurrent Constraint Programs Correct Maurizio GABBRIELLI (CWI, Amsterdam) Re'sume' : We develop a compositional proof-system for proving the partial correctness of concurrent constraint programs (ccp). We first introduce a specification logic to reason about the correctness of ccp programs in terms of properties of constraints. The rule of the calculus are then obtained by ``mirroring'' the equations of a simple denotational semantics which describes the strongest postcondition of a ccp process. Our results clarify the declarative nature of ccp since allow to consider a ccp program as a theory in the specification logic. The operators of the language are interpreted logically by considering action prefixing as implication, non-deterministic choice as disjunction and parallel composition as conjunction. Joint work with F.S. de Boer, E. Marchiori and C. Palamidessi. *********************************************************************