J. Jrgensen and M. Leuschel. Efficiently generating efficient generating extensions in Prolog. In O. Danvy, R. Gluck, and P. Thiemann, editors, Dagstuhl Seminar on Partial Evaluation, LNCS 1110, pages 238--262, Schlo Dagstuhl, Germany, 1996. Springer-Verlag, Berlin.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... a practical but non self applicable partial evaluator for full Prolog [84, 85] Bondorf and Mogensen [76] constructed a self applicable partial evaluator for a Prolog subset, Gurr one for the logic language Godel [52] J rgensen and Leuschel created a generator of generating extensions for Prolog [62]. 9.3 Related topics McCarthy used program transformation rules in calculational proofs for recursive functional programs [72] Boyer and Moore automated some proofs of this kind [22] Burstall and Darlington viewed manual program transformation as the application of a few types of ....

J. Jørgensen and M. Leuschel. Efficiently generating efficient generating extensions in Prolog. Technical Report CW 221, Department of Computer Science, Katholieke Universiteit Leuven, Belgium, 1996.

No context found.

Jrgensen, J. and Leuschel, M. Efficiently generating efficient generating extensions in Prolog. In [18], pp. 238--262, 1996.

No context found.

J. Jrgensen and M. Leuschel. Efficiently generating efficient generating extensions in Prolog. In O. Danvy, R. Gluck, and P. Thiemann, editors, Dagstuhl Seminar on Partial Evaluation, (LNCS 1110), pages 238--262, Schlo Dagstuhl, Saarbrucken, 1996. Springer-Verlag, Berlin.

.... main reasons for using the off line approach are to make specialisation itself more efficient and, due to a simpler specialiser algorithm, enable effective self application (specialisation of the specialiser) 16] Few authors discuss off line specialisation in the context of logic programming [27, 17], mainly because so far no automated binding time analysers have been developed. This paper aims to remedy this problem. 3 Towards BTA for partial deduction 3.1 An on line specialiser The basic idea of BTA in functional programming is to model the flow of static input: the arguments of a ....

....in a domain fstatic; dynamicg whose least solution yields the best annotation. This approach does not immediately translate to logic programs. Problems are that the dataflow in unification is bidirectional and that the degree of instantiation of a variable can change over its lifetime (see also [17]) We follow a different approach and reconstruct binding time analysis from first principles. We start with a Prolog program which performs the unfolding decisions of an on line specialiser. However, whereas real on line specialisers base their unfolding decisions on the history of the ....

[Article contains additional citation context not shown here]

J. Jrgensen and M. Leuschel. Efficiently Generating Efficient Generating Extensions in Prolog. In O. Danvy, R. Gluck, and P. Thiemann, editors, Proceedings of the 1996 Dagstuhl Seminar on Partial Evaluation, number 1110 in LNCS, pages 238--262, Schlo Dagstuhl, February 1996. Springer-Verlag.

....a key property for tabled programs. Also, in a broader context, techniques for proving quasi termination can be of great value to ensure termination of off line specialisation of logic programs (whether tabled or not) Currently, in all off line partial evaluation methods for logic programs (e.g. [13, 10]) termination has to be ensured manually. In the context of off line partial evaluation, quasi termination is actually identical to termination of the partial evaluator. Thus, given a technique to establish quasitermination, one can also establish whether a given binding time annotation will ....

J. Jrgensen and M. Leuschel. Efficiently Generating Efficient Generating Extensions in Prolog. In O. Danvy, R. Gluck, and P. Thiemann, editors, Proceedings of the 1996 Dagstuhl Seminar on Partial Evaluation, number 1110 in LNCS, pages 238--262, Schlo Dagstuhl, 1996. Springer-Verlag.

....specialised clause per branch. 2. CONTROL OF AUTOMATIC PARTIAL DEDUCTION Most work on partial deduction is situated within the on line tradition: All the control decisions are performed fully automatically during the actual specialisation phase. The work presented here is no exception (see e.g. [Jrgensen and Leuschel 1996] for an off line approach) In partial deduction one usually distinguishes two levels of control: ffl the global control , determining the set A, i.e. deciding which atoms are partially deduced, and ffl the local control , guiding construction of the finite SLDNF trees for each individual atom ....

....overcoming these limitations. The essential aspect lies in the joint treatment of entire conjunctions of atoms, connected through shared variables, at the global level (complemented with some renaming to deliver program clauses) Leuschel et al. 1996; Gluck et al. 1996; Leuschel and Srensen 1996; Jrgensen et al. 1996]. Apart from this aspect, the conventional control notions described above also apply in a conjunctive setting. In essence, it was possible to consolidate partial deduction and unfold fold program transformation, incorporating most of the power of the latter while keeping the automatic control ....

Jrgensen, J. and Leuschel, M. 1996. Efficiently generating efficient generating extensions in Prolog. In Proceedings of the 1996 Dagstuhl Seminar on Partial Evaluation, O. Danvy, R. Gluck, and P. Thiemann, Eds. LNCS 1110. Springer-Verlag, 238--262.

....of gaining speedups by specialising the static part with respect to a number of different patterns (often just the relational symbols) for the dynamic part seemed extremely promising. Moreover, this would allow for a very flexible way of generating specialised up 1 Some notable exceptions are [71, 33, 34, 39]. 2 In the remainder of this paper we will only talk about deductive databases, but the discussions and results remain of course also valid for inductive or abductive logic programs with integrity constraints. 3 date procedures. Any kind of update pattern and any kind of partial knowledge could ....

J. Jrgensen and M. Leuschel. Efficiently generating efficient generating extensions in Prolog. In O. Danvy, R. Gluck, and P. Thiemann, editors, Proceedings of the 1996 Dagstuhl Seminar on Partial Evaluation, LNCS 1110, pages 238--262, Schlo Dagstuhl, 1996. Springer-Verlag. Extended version as Technical Report CW 221, K.U. Leuven.

.... issues of termination and of code and search explosion, and efficiency gains have been obtained [11, 58, 23, 24, 40, 50] Several fully automated systems (sp [22] sage [27] paddy[63] mixtus [67] ecce [40, 50, 51, 53] as well as semi automated ones (logimix [60] leupel [39, 46] cogen [30]) have been developed and successfully applied to at least medium size applications [46, 49, 18, 37] A similar development of automated techniques and systems has not been undertaken in the context of unfold fold transformations. The aim of this paper is to bring the advantages of these two ....

J. Jørgensen, M. Leuschel. Efficiently generating efficient generating extensions in Prolog. In [14], 238--262, 1996.

No context found.

J. Jrgensen and M. Leuschel. Efficiently generating efficient generating extensions in Prolog. In O. Danvy, R. Gluck, and P. Thiemann, editors, Dagstuhl Seminar on Partial Evaluation, LNCS 1110, pages 238--262, Schlo Dagstuhl, Germany, 1996. Springer-Verlag, Berlin.

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