Results 1 
2 of
2
A Perfect Hash Function Generator
"... gperf is a "softwaretool generatingtool" designed to automate the generation of perfect hash functions. This paper describes the features, algorithms, and objectoriented design and implementation strategies incorporated in gperf.Italso presents the results from an empirical comparison between gp ..."
Abstract

Cited by 51 (35 self)
 Add to MetaCart
gperf is a "softwaretool generatingtool" designed to automate the generation of perfect hash functions. This paper describes the features, algorithms, and objectoriented design and implementation strategies incorporated in gperf.Italso presents the results from an empirical comparison between gperfgenerated recognizers and other popular techniques for reserved word lookup. gperf is distributed with the GNU libg++ library and is used to generate the keyword recognizers for the GNU C and GNU C++ compilers. 1 Introduction Perfect hash functions are a time and space efficient implementation of static search sets, which are ADTs with operations like initialize, insert,andretrieve. Static search sets are common in system software applications. Typical static search sets include compiler and interpreter reserved words, assembler instruction mnemonics, and shell interpreter builtin commands. Search set elements are called keywords.Key words are inserted into the set once, usually at c...
Exact Analyses of a Simple Heuristic Employed In Array Compression
, 2002
"... ... this paper is to precisely analyse the behaviour of one such extremely simple heuristic which is known to give modest compression in practice. For the heuristic we prove that the expected asymptotic space requirement is, at worst, a(k)n+ b(k)x and that although its dependency on n is inherent ..."
Abstract
 Add to MetaCart
... this paper is to precisely analyse the behaviour of one such extremely simple heuristic which is known to give modest compression in practice. For the heuristic we prove that the expected asymptotic space requirement is, at worst, a(k)n+ b(k)x and that although its dependency on n is inherent, it can be made arbitrarily small. Here k is a parameter and a(k) and b(k) are, respectively, monotonically decreasing and increasing functions. Thus k allows a tradeoff between dependency on n and x;for example, pairs (a(k), b(k)) can be (0.1, 3.26), (0.03, 5.57) and (6 10 4 , 33). We also show that for some applications the dependency of the space requirement on n canbemadesublinear. The heuristic allows constant time access to any element. Our analyses are over two different models for the uniform probability distribution and we derive exact formulae for the expected space used. We prove that the heuristic gives the same asymptotic performance in both models