scientific article; zbMATH DE number 236850

From MaRDI portal
Publication:5287554





This carefully written book introduces the reader to compiler construction. Each of the ten chapters is designed in a uniform and easy to follow way: first the goals of the chapters are stated clearly in a bullet list, then an introduction helps the reader to get started. The main text of each chapter is followed by a summary, a list of acronyms and keywords, resp. a set of (nongraded) exercises follows. Each chapter closes with a review quiz, some suggestions for a compiler project and a list of further references, which are briefly commented. No answers to questions or exercises are given (and I could not find a reference to the Dragon Book). The style of writing is clear and concise; occasionally the typesetting got a bit out of hand, and longer listings are sometimes unpleasant to read. So much for the appearance. The authors state in the beginning that compiler theory is an important part of the text. Well, theoretical issues are dealt with, but sometimes this reviewer would have preferred an more formal argumentation, e.g., when the authors discuss the finite automaton that controls on \(\text{LR}(k)\) parser. The overall strategy seems to informally tell the reader what the theoretical issues are, without going into formal discussions or constructions too deeply. This might help practically oriented readers appreciating the theory necessary to digest the formalisms underlying a compiler, but it is a bit too shallow for substantiating the claim established in the subtitle of incorporating theoretical issues of compiler construction. The first chapter gives an introduction to what the authors call ``The Compiler Theory Landscape. This is followed by a chapter sketching the Chomsky hierarchy. Chapter 3 deals with scanners and regular languages, and despite its title ``Parsers and Context-Free Languages, chapter 4 centers around \(\text{LL}(k)\) grammars. In chapter 5 we find a brief discussion of issues pertaining to semantic analysis and attribute grammars; this discussion is not very general, systematic approaches to evaluate attribute grammars are not discussed. Chapter 6 deals with syntax-directed code generation and introduces run time environments. The topic of bottom-up parsing is introduced in chapter 7, which is rather short, followed by the definition and discussion of transformational attribute grammars in chapter 8. This chapter is partially devoted to specifying code optimization by transformation using transformational attribute grammars. Code generation and optimization is introduced and discussed in chapter 9, focussing on loop optimization, register and memory allocation and briefly discussing specialized architectures. The final chapter deals with issues of compiling nonprocedural languages; here LISP and a compiler for a transformational attribute grammar are discussed, outlining for the latter an application of the techniques discussed in former chapters. The book can be recommended for a compiler construction course in one or two semesters on the junior or senior level in a typical computer science curriculum. It should be easy and pleasant to teach from it, in particular because the examples are done very carefully and are given much room. A course, however, which intends to discuss deeper theoretical or practical issues of compiler construction would not want to use this text: the book by \textit{R. Wilhelm} and \textit{D. Maurer} [Übersetzerbau (Springer, 1992; Zbl 0757.68032)] is superior in its theoretical approach, the book by \textit{W. M. Waite} and \textit{G. Goos} [Compiler Construction (Springer, 1984; Zbl 0527.68003)] is much more adequate as a basis for tool construction.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5287554)