M. Gabbrielli, R. Giacobazzi, and G. Levi. Goal independency and call patterns in the analysis of logic programs. In ACM Symposium on Applied Computing. ACM Press, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....properties which are different from success patterns, such as: failure, call patterns and partial answers cannot be modeled by applying directly the semantic construction in Section 4. As for call patterns, both the magic like transformation in [15] and the semantic based approximation in [32] can be easily extended to CLP languages in view of the present paper. The machinery of partial answers instead may require We thank an anonymous referee for this comment. an additional layer of abstraction, like the one applied in [16] for the compositional analysis of modular logic programs. We ....

M. Gabbrielli and R. Giacobazzi. Goal Independency and Call Patterns in the Analysis of Logic Programs. In E. Deaton, D. Oppenheim, J. Urban and H. Berghel editors, Proc. of the Ninth ACM Symposium on Applied Computing, pages 394--399, ACM Press, Phoenix AZ 1994.

....states that a SLD derivation of Q can be lifted to a SLD derivation of Q. We prove that some speci c arguments of Q, called derivation neutral , can be instantiated as well, while retaining the existence of the lifted derivation. A new application of binary unfoldings to left loop inference. [12] introduced the binary unfoldings of a logic program P as a goal independent technique to transform P into a possibly in nite set of binary clauses, which preserves the termination property [7] while abstracting the standard operational semantics associated to SLD resolution. We present an ....

....if there exists such that Q left loops w.r.t. P 0 , then Q left loops w.r.t. any P P 0 . Such properties are false for Prolog programs, due to the non logical features of the language. 3. 2 The binary unfoldings of a logic program Let us present the main ideas about the binary unfoldings [12] of a logic program, borrowed from [7] This technique transforms a logic program P (without any query of interest) into a possibly in nite set of binary clauses. Intuitively, each generated binary clause H B (where B is either an atom or the atom true which denotes the empty query) speci es ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proceedings of the ACM Symposium on applied computing, pages 394-399. ACM Press, 1994.

....: Rn and j j 1 ; j j n . Let i 2 f1; ng. If Q is de ned in R 1 [ R i then Q is strongly bounded wrt. R 1 [ R i and j j 1 ; j j i . In order to verify whether a query Q is strongly bounded wrt. a given program P one can perform a call pattern analysis [JB92, GG94, CD95] which allows us to infer information about the form of the call patterns, i.e. the atoms that will be possibly called during the execution of P [ fQg. However this is not the only way for guaranteeing strong boundedness. There are classes of programs and queries for which strong boundedness can ....

....be pursued for a fruitful integration with existing automatic tools. The rst one exploits the fact that in each single module it is sucient to synthesize a level mapping which does not need to measure atoms de ned in other modules. The second one concerns tools based on call patterns analysis [DVB92, GG94, CD95]. They can take advantage of the concept of strong boundedness which, as we show, can be implied by well behaviour of programs [DW88, Deb89] Acknowledgements. We wish to thank the anonymous referees for comments and suggestions which helped us to improve the paper. This work has been partially ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proceedings of the Ninth ACM Symposium on Applied Computing, pages 394-399. ACM Press, 1994.

.... Several semantic definitions for logic programs have been proposed to capture various notions of observables: the standard minimal model semantics models logical consequences, the c semantics and the s semantics [20] model correct and computed answers respectively, the semantic definition of [22] models the notion of call patterns, etc. In this paper we show that the definition of [22] is suitable to model also the termination This work was supported by a grant from the Israel Science Foundation Preprint submitted to Elsevier Preprint 8 March 1999 properties of programs. This provides ....

....notions of observables: the standard minimal model semantics models logical consequences, the c semantics and the s semantics [20] model correct and computed answers respectively, the semantic definition of [22] models the notion of call patterns, etc. In this paper we show that the definition of [22] is suitable to model also the termination This work was supported by a grant from the Israel Science Foundation Preprint submitted to Elsevier Preprint 8 March 1999 properties of programs. This provides a formal semantic basis for the analysis of termination of logic programs based on ....

[Article contains additional citation context not shown here]

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proceedings of the Ninth ACM Symposium on Applied Computing, pages 394--399. ACM Press, 1994.

....properties which are different from success patterns, such as: failure, call patterns and partial answers cannot be modeled by applying directly the semantic construction in Section 4. As for call patterns, both the magic like transformation in [15] and the semantic based approximation in [32] can be easily extended to CLP languages in view of the present paper. The machinery of partial answers instead may require an additional layer of abstraction, like the one applied in [16] for the compositional analysis of modular logic programs. We believe that our constraint system notion and ....

M. Gabbrielli and R. Giacobazzi. Goal Independency and Call Patterns in the Analysis of Logic Programs. In E. Deaton, D. Oppenheim, J. Urban and H. Berghel editors, Proc. of the Ninth ACM Symposium on Applied Computing, pages 394--399, ACM Press, Phoenix AZ 1994.

....initial goals are derived very quickly. Moreover, the complexity of analyses is reduced as we apply bottom up evaluation to smaller program modules. The effect is similar to that obtained using condensing as suggested by Jacobs and Langen in [25] It is also related to the work described in [20] which describes a goal independent analysis for call patterns based on the abstraction of the semantic characterization described in [21] We illustrate this type of analysis with an example. Consider again the naive reverse program P and its prop ositional counterpart P # from Figure 1. ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proc. SAC'94, 1994.

....provides a safe and finite approximation of some runtime behavior of the program. For instance, a number of abstract interpretations have been proposed to approximate success patterns of atoms or clause [2] 9] The ones in [6] 7] 9] are based on top down evaluation while the ones in [2] [5], 8] on bottom up evaluation. Recently, bottom up abstract interpretation of logic programs has gained much attention. The analysis is based on a concrete semantics which propagates the information in the opposite direction with respect to the Selective Linear resolution for Definite clauses ....

M. Gabbrielli and R. Giacobazzi, "Goal independency and call patterns in the analysis of logic programs, " In Proceedingsof ACM Symposiumon Applied Computing, ACM Press, 1994.

....where approximation is the rule of the game. Theorems valid for I observables are therefore weaker and denotations provide characterizations of semantic properties which are correct in the sense of abstract interpretation theory. We show how we can reconstruct the abstract semantics defined in [12], which allows us to derive groundness relations among the arguments of procedure calls. 3.1 An algebraic formalization of observables: the S observ ables Given an observable we want to be able to observe computations of conjunctive goals from the single conjuncts computations. Moreover we do ....

....2 B lm ff (p) 9 c 0 2 A 0 ; e 0 1 ; e 0 n 2 O ff (p) s.t. v 0 c 0 Delta (f e 0 1 g Theta : Thetaf e 0 n g) Example 4.6 Let AH be the Herbrand constraint system. We show how to reconstruct the ground dependency analysis for call patterns described in [12]. The abstract domain of computation is Prop [7] consisting of propositional formulas which provide a concise representation of abstract substitutions which describe ground dependency relations among arguments of a procedure call. A Prop is the algebra of possibly existentially quantified ....

[Article contains additional citation context not shown here]

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proc. SAC'94, 1994.

....present the ones most frequently used in optimizing compilers. However, we believe the techniques proposed are equally applicable to bottom up analyses. In the text, we consider in general goal dependent analyses, but point out solutions for goal independent analyses where appropriate (see, e.g. [16, 15, 8]) The pairs of call and success patterns computed by the analysis, be it topdown or bottom up, goal dependent or independent, will be denoted by AOT ff (P ) for a given program P . A most general goal pattern (or simply goal pattern, hereafter) of a predicate is a normalized goal for that ....

M. Gabbrielli, R. Giacobazzi, and G. Levi. Goal independency and call patterns in the analysis of logic programs. In ACM Symposium on Applied Computing. ACM, 1994.

....obtained from the answers computed for pure atomic goals. 6.2 Call Pattern Semantics The call patterns (with state) of a program P for a goal G are the atoms selected in any SLD derivation of G in P. A call pattern semantics was defined in [25] and used as a basis of call pattern analysis in [24]. We define the domain A cp [Goals ae (Atoms Theta Subst) For any X 2 A cp the interpretation of hC; #i 2 X(G) is the execution of the goal G generates a procedure call C with state (partial computed answer substitution) # . Note that C can be . The abstraction which allows us to obtain the ....

.... [20] and the call pattern semantics [27] The results on semi denotational observables apply to bottom up abstract interpretation frameworks, such as those whose collecting semantics is the computed answer semantics [4, 9] the correct answers semantics [30] and the call pattern semantics [24]. Finally, the results on semi perfect observables apply to top down abstract interpretation frameworks [7] There is something more we can learn from our theory of observables on the relation between denotational and operational definitions and the relation between goal independence and ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In J. Urban E. Deaton, D. Oppenheim and H. Berghel, editors, Proceedings of the Ninth ACM Symposium on Applied Computing, pages 394--399, Phoenix AZ, 1994. ACM Press.

....for logic programs. 3 In the sense of abstract interpretation ( 10] i.e. ff(lfp( P ) D lfp(ffffi P ffifl) The following is a list of some well known semantics for logic programs that can all be derived by abstract interpretation from h(T a P (SLD) P i. Some of them have been used in [4, 8, 18] for data flow analysis of logic programs. Each semantics is provided with the corresponding abstraction, hence with a G.c. indeed insertion) A more complete treatment for semantics derivable by abstract interpretation from a reference semantics is in [9, 21] In what follows, we denote fl x the ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proc. ACM SAC '94, pp. 394--399, 1994.

.... Several semantic definitions for logic programs have been proposed to capture various notions of observables: the standard minimal model semantics models logical consequences, the c semantics and the s semantics [20] model correct and computed answers respectively, the semantic definition of [22] models the notion of call patterns, etc. In this paper we show that the definition of [22] is suitable to model also the termination properties of programs. This provides a formal semantic basis for the analysis of termination of logic programs based on abstract interpretation [12] We study left ....

....notions of observables: the standard minimal model semantics models logical consequences, the c semantics and the s semantics [20] model correct and computed answers respectively, the semantic definition of [22] models the notion of call patterns, etc. In this paper we show that the definition of [22] is suitable to model also the termination properties of programs. This provides a formal semantic basis for the analysis of termination of logic programs based on abstract interpretation [12] We study left termination, i.e. the universal termination of logic programs executed by means of ....

[Article contains additional citation context not shown here]

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proceedings of the ACM Symposium on Applied Computing. ACM Press, 1994.

....bottom up semantics for call patterns with the semantics proposed in [13] They propose a goal independent, bottom up semantics for resultants. Since call patterns are an abstraction of resultants, their semantics can be used for computing call patterns. This abstraction is explicitly described in [12], where the authors provide a concrete semantics for call patterns and its abstraction for program analysis. The main difference with our approach is that they use sets of clauses rather than sequences of constraints as the semantical domain. Our choice was motivated by the fact that clauses are ....

M. Gabbrielli and R. Giacobazzi. Goal Independency and Call Patterns in the Analysis of Logic Programs. In J. Urban E. Deaton, D. Oppenheim and H. Berghel, editors, Proceedings of the Ninth ACM Symposium on Applied Computing, pages 394--399, Phoenix AZ, 1994. ACM Press.

....behavior for a specific goal can be determined from the goalindependent analysis by using an AND compositionality theorem. ffl The approach can be extended to observables which are more concrete than answer constraints, such as call patterns [16] thus allowing us to reconstruct the analyses in [11, 22]. Several interesting domains useful for type, groundness, aliasing and independence analysis can easily be formalized as constraint systems. The technique can be combined with an elegant model theoretic approach to the abstraction of the constraint system [8] Moreover, the approach can be ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proc. SAC'94, 1994.

....initial goals are derived very quickly. Moreover, the complexity of analyses is reduced as we apply bottom up evaluation to smaller program modules. The effect is similar to that obtained using condensing as suggested by Jacobs and Langen in [27] It is also related to the work described in [22] which describes a goal independent analysis for call patterns based on the abstraction of the semantic characterization described in [23] We illustrate this type of analysis with an example. Consider again the naive reverse program P and its prop ositional counterpart 13 qsort(X1,X2) ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proc. SAC'94, 1994.

....of the call patterns, we should use a concrete semantics which gives more information on the computation than just the computed answers. Namely, we have to model an observable consisting of all the procedure calls. The problem of analyzing properties of the call patterns has been considered in [53], where the concrete semantics is the call patterns semantics derived according to a local selection rule, as defined in [61, 59] The resulting abstract semantics are goal independent, parametric w.r.t. the (local) selection rule and allow us to characterize properties of the correct call ....

....T Pi P 0 of definition 3.13 can now be used to compute in a bottom up way information on the call patterns of G in P . Recently, the approach was made goal independent [33] by using the Omega Gammae 478 tics. The result is a denotation consisting of clauses very similar to the one in [53]. ffl It is worth noting that the top down operational or denotational frameworks [24, 94, 104, 74] do indeed contain a lot of information on the internal computation details. By choosing a semantics like the one of section 5.3, we can model the same observables and still get a goal ....

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proc. ACM Symposium on Applied Computing, ACM press, 1994.

No context found.

M. Gabbrielli, R. Giacobazzi, and G. Levi. Goal independency and call patterns in the analysis of logic programs. In ACM Symposium on Applied Computing. ACM Press, 1994.

No context found.

M. Gabbrielli and R. Giacobazzi. Goal independency and call patterns in the analysis of logic programs. In Proceedings of the ACM Symposium on applied computing, pages 394--399. ACM Press, 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC