J. Launchbury. Projection Factorizations in Partial Evaluation. PhD thesis, University of Glasgow, Nov. 1989.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....of the Ackerman s function, and (n; n) is for all intents and purposes a small constant. His algorithm uses a constraint solving system to determine where annotations should be placed. ManyBTAs are based upon abstract analysis. BTAs for partially static data have been presented by Launchbury[11] and Mogenson[14] and polyvariant BTAs have been presented by Consel [6, 1] Dussart et al. 7] Ritz Gengler [16] amongst others. The techniques described here incorporate all these features in a simple framework based upon search. 16. CONTRIBUTIONS In this paper wehave described a ....

J. Launchbury. Projection Factorizations in Partial Evaluation. PhD thesis, University of Glasgow, Nov. 1989.

....w.r.t. some part of the input (designated as static) so that they will run faster, when the rest of the input (designated as dynamic) becomes available. Binding time analysis attempts to find those parts of the program that depend only on static input. It is equivalent to strictness analysis [13]. Automatic program debugging is the inference of program properties through automatic means with minimum user intervention. In this sense, they cover the area between type checkers like ML type inference algorithm and full scale program verifiers. Information about the values that program ....

....assertions are satisfied. One example would be array bounds checking. 6 Related Work In [1] dead code elimination is investigated for a first order imperative language. Inspite of this, there are parallels between our method for collecting flow information and data flow analysis. The work in [14, 19, 23, 13] makes use of projections for di#erent program analysis problems for first order languages. In comparison, our use of a type system o#ers several benefits. Type systems are a widely used and natural tool for expressing properties about higher order functions, which the above work does not address. ....

J. Launchbury. Projection Factorizations in Partial Evaluation. PhD thesis, Department of Computing, University of Glasgow, Glasgow, Scotland, June 1989.

....but Mogensen s was the rst analysis that could recover such information. Consel [2] used an analysis based upon cons points, i.e. labels labelling every constructor function in a program, to build similar graph like descriptions, where the nodes of the graph are the cons points. Launchbury[7] used projection analysis to assign each partially static object in a typed language, to a point in a lattice of partial domains. John Hughes[4] work on type based specialization may also be closely related. At some point, the term partially static took on the association of static data in the ....

J. Launchbury. Projection Factorizations in Partial Evaluation. PhD thesis, University of Glasgow, Nov. 1989.

....annotated functions allowing di#erent specializations at di#erent occurrences. Unlike our use of annotated types as binding time specifications, theirs is limited to first order functions. Many BTAs are based upon abstract analysis. BTAs for partially static data have been presented by Launchbury[13] and Mogensen[17] and polyvariant BTAs have been presented by Consel [6, 1] Dussart et al. 7] Rytz Gengler [19] amongst others. Glueck and Joregensen [8, 9] pioneered the use of BTA for multi level languages. Their work generalizes a standard abstract interpretation technique to multiple ....

J. Launchbury. Projection Factorizations in Partial Evaluation. PhD thesis, University of Glasgow, Nov. 1989.

....of the Ackerman s function, and (n; n) is for all intents and purposes a small constant. His algorithm uses a constraint solving system to determine where annotations should be placed. Many BTAs are based upon abstract analysis. BTAs for partially static data have been presented by Launchbury[11] and Mogenson[14] and polyvariant BTAs have been presented by Consel [6, 1] Dussart et al. 7] Ritz Gengler [16] amongst others. The techniques described here incorporate all these features in a simple framework based upon search. 16. CONTRIBUTIONS In this paper we have described a ....

J. Launchbury. Projection Factorizations in Partial Evaluation. PhD thesis, University of Glasgow, Nov. 1989.

....p and simplifying the function body in appropriate ways. There are a number of techniques developed for partial evaluation that are related to this goal. For example, projection functions have been used in binding time analysis for partial evaluation in the presence of partially static structures [19,22]. However, for slicing, we need to propagate projection functions backwards from function outputs to function arguments. Thus, the slicing problem has similarities with the algorithms that propagate projection functions backwards to perform strictness analysis of lazy functional languages ....

Launchbury, J., Projection Factorizations in Partial Evaluation, Cambridge University Press, Cambridge, UK (1991).

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