On constructing efficient evaluators for attribute grammars

Saarinen, Mikko

doi:10.1007/3-540-08860-1_29

Cited by 25 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the absolutely noncircular grammars, Kennedy and Warren proposed a tree walk algorithm for attribute evaluation [12], and Saarinen improved it in the output-oriented form [21]. Their tree walk evaluator is a recursive procedure that visits nodes of derivation trees and evaluates their attributes.…”

Section: Introductionmentioning

confidence: 99%

Translation of attribute grammars into procedures

Katayama

1984

ACM Trans. Program. Lang. Syst.

View full text Add to dashboard Cite

An efficient method for evaluating attribute grammars by translating them into sets of procedures is presented. The basic idea behind the method is to consider nonterminal symbols of the grammar as functions that map their inherited attributes to their synthesized attributes. Associated with the nonterminal symbols are procedures that realize the functions. The attribute grammar is translated into a program consisting of these procedures. The essential point about this method is that attribute grammars are completely compiled into procedures, in contrast with evaluation algorithms that work interpretively in a table-driven manner. No information is stored in the nodes of derivation trees.Although this evaluation method is applicable principally to absolutely noncircular attribute grammars in which the dependency relation among attribute occurrences of every production rule does not contain cycles, it is shown how the method is extended to the general noncircular attribute grammars. Problems of evaluating a set of attributes simultaneously and of recursive descent evaluation are also discussed.

show abstract

Section: Introductionmentioning

confidence: 99%

Translation of attribute grammars into procedures

Katayama

1984

ACM Trans. Program. Lang. Syst.

View full text Add to dashboard Cite

show abstract

“…For example, Rgiha [27] cites experiments in which a recycling algorithm led to a multi-pass evaluator using only 20% to 28% of the storage areas required by the previous "unoptimized" algorithm. Tree-walk evaluators [15,28] determine a traversal by analyzing attribute dependencies in the grammar rules, but storage properties are not considered in the analysis.…”

mentioning

confidence: 99%

The global storage needs of a subcomputation

Raoult

Sethi²

1984

Proceedings of the 11th ACM SIGACT-SIGPLAN Symposium on Principles of Programming Languages - POPL '84

View full text Add to dashboard Cite

A defining characteristic of "functional" specifications is the absence of assignments: updates of tables and data structures are expressed by giving the relationship between the new and old values. An obvious implementation allocates separate space for new and old values and may consume a lot of storage. However, even when updates of attributes like symbol tables are expressed functionally, we would like to avoid making copies of the symbol table during attribute evaluation. In other words, if possible, the implementation should have a single global copy of the table that is updated using assignments. Since the value of the global copy changes during computation, the order of evaluation has to be chosen carefully. In this paper, we partition attributes into classes, the problem being to determine if there exists an evaluation order that allows each class to share a global storage area. The solution extends to handle symbol tables for block structured languages.More precisely, consider a directed acyclic graph D in which vertices represent attribute values to be computed. Associated with each vertex is a label indicating the storage area to be shared by the vertex. Storage used during the evaluation of D is studied by playing a pebble game on D with the following steps: (1) pebble (i.e. place a pebble) on a source; (2) pebble a vertex if all its successors have pebbles (a pebble may be moved to the vertex from one of its successors); (3) pick up a pebble. Each vertex must be pebbled exactly once. The use of global storage is formalized by defining single pebblings, in which at most one of the vertices with a given label has a pebble at any time. The results. also apply to a form of pebbling, called chain pebbling, that allows symbol tables for block structured languages to be studied.

show abstract

“…(Note that a precise characterization of W could provide strong evidence concerning the question whether or not it is decidable in polynomial time that an attribute grammar is simple m-waddle for a fixed m. This idea is based on the observation that, for any X ∈ { pass, sweep, visit }, the complexity of deciding that an attribute grammar is simple m-X for a fixed m is polynomial if and only if for simple m-X evaluators a finite set of data structures is needed for the (a, a + 1)-allocation of the instances of an applied occurrence. For further information, compare Figure 11 In the introduction of this dissertation it was stated that improving the results of Saarinen [32] required a more thorough analysis at evaluator construction time. This was considered to be rather difficult because the order in which absolutely non-circular attribute evaluators compute attribute instances, determined at evaluator construction time, depended on the derivation trees to be evaluated.…”

Section: Implementation Aspectsmentioning

confidence: 99%

“…Global storage allocation techniques found in the literature are divisible into three groups: G1: Global storage allocation techniques in which allocation is decided at evaluator construction time by analyzing the evaluation strategy of a given time-efficient evaluator. These techniques, found in [10,12,13,15,18,20,32], are independent of specific derivation trees and leave the evaluation strategy of the given time-efficient evaluator unchanged.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Linear data structures for storage allocation in attribute evaluators

Sluiman¹

View full text Add to dashboard Cite

Practical and theoretical results have been found concerning the use of global storage allocation for the instances of applied occurrences of an attribute.The practical results focus on the necessary and sufficient conditions to decide at evaluator construction time whether an evaluator can allocate the instances of an applied occurrence to a number of global variables, stacks and queues. Checking these conditions takes polynomial time for a simple multi-visit evaluator and exponential time for an absolutely non-circular evaluator.The theoretical results are concerned with the data structures that are required for the global storage allocation of the instances of applied occurrences in simple multi-X evaluators, where X ∈ { pass, sweep, visit }. For this purpose, the general class of basic linear data structures is introduced. This class of data structures can also be used to explore the theoretical possibilities and limitations of storage allocation techniques in domains other than attribute grammars.

show abstract

On constructing efficient evaluators for attribute grammars

Cited by 25 publications

References 12 publications

Translation of attribute grammars into procedures

Translation of attribute grammars into procedures

The global storage needs of a subcomputation

Linear data structures for storage allocation in attribute evaluators

Contact Info

Product

Resources

About