编译器技术探索：工具、翻译器与语言接口实现

需积分: 9 121 浏览量更新于2024-07-22 收藏 4.43MB PDF 举报

编译器技术：工具、翻译器、语言接口编译器技术是一门重要的计算机科学领域，涉及到计算机语言的设计、实现和优化。编译器技术的主要目的是将源代码转换为目标机器代码，使得计算机能够正确地执行程序。编译器技术的应用非常广泛，涵盖了操作系统、编译器、解释器、数据库系统等领域。本书《编译器技术：工具、翻译器、语言接口》是一本关于编译器技术的经典著作，涵盖了编译器技术的基本概念、工具、翻译器、语言接口等方面的内容。该书由Derek Beng Kee Kiong博士编写，Derek Beng Kee Kiong博士是新加坡国立大学的教授，具有丰富的编译器技术教学和研究经验。本书的主要内容包括： 1. 编译器技术的基本概念：介绍编译器技术的定义、历史、应用领域等。 2. 工具：介绍编译器技术中常用的工具，如Lex、Yacc、ANTLR等。 3. 翻译器：介绍翻译器的概念、类型、实现方法等。 4. 语言接口：介绍语言接口的概念、类型、实现方法等。 5. 实现：介绍编译器技术的实现方法、技术和工具等。本书的特点是： 1. 系统性：本书系统地介绍了编译器技术的基本概念、工具、翻译器、语言接口等方面的内容。 2. 实践性：本书提供了许多实践例子和案例，帮助读者更好地理解编译器技术的应用。 3. 深入性：本书深入地介绍了编译器技术的实现方法、技术和工具等。《编译器技术：工具、翻译器、语言接口》是一本非常有价值的参考书籍，适合计算机科学和技术专业的学生、研究人员和从业人员阅读。知识点： 1. 编译器技术的定义和历史 2. 编译器技术的应用领域 3. Lex、Yacc、ANTLR等工具的应用 4. 翻译器的概念和类型 5. 语言接口的概念和类型 6. 编译器技术的实现方法和技术 7. 编译器技术在操作系统、编译器、解释器、数据库系统等领域的应用通过学习本书，读者将能够： 1. 了解编译器技术的基本概念和应用领域 2. 掌握编译器技术中常用的工具和翻译器 3.理解语言接口的概念和类型 4. 学习编译器技术的实现方法和技术 5. 应用编译器技术在实际项目中《编译器技术：工具、翻译器、语言接口》是一本非常有价值的参考书籍，对于计算机科学和技术专业的学生、研究人员和从业人员来说非常有帮助。

INTRODUCTION

LANGUAGE

IMPLEMENTATION

Many computer systems

the past were implemented using assembly code so

achieve efficient execution. However, most modem computer systems are

developed using high-level computer languages. These have been found to be more

amenable to human programmers who are involved with system design,

implementation, integration and maintenance. High-level languages allow

algorithms

be specified near the conceptual level

the problem rather than

terms

operational or machine specific implementation details. Proper use

high-

level languages promotes programmer productivity since it supports abstraction and

less error-proned when compared with systems developed using machine code or

assembly language.

However, while a program written

a high-level language

more programmer-

friendly, it is generally not directly executable by raw hardware. Instead, it must be

translated into an equivalent machine code before it may be executable. While such

translations may be specified and performed manually, automating the process

sensible since it is repeated for each new high-level language program. Such

automation is consistent with software engineering principles and the use

software

development tools.

A language compiler

one

the most crucial tools for system development.

Enhancements

the current framework might involve good run-time monitoring as

well

debugging facilities. The integration

an editing environment with the

source handler allows for program compilation to be performed during input and

D. B. K. Kiong, Compiler Technology

Introduction to Language Implementation

thus reveals program errors immediately. However, these issues are beyond the

scope

this book.

1.1

Translator Strategies

We define a compiler

the tool which translates a high-level language program into

an equivalent machine language program that

executable on the intended system.

The first compiler was developed for

Fortran - the high-level programming

language designed in the early 1960s. In the most general sense, this

a specific

instance

a language processor which analyses programs in a source language

representation to produce equivalent programs in some target representation.

Some language processors allow the execution

high-level language programs

without any explicit code being generated. Such an implementation

known

language interpreter. This strategy has been used in many implementations

the

Basic language.

allows for the direct execution

program statements: each line

program code

analyzed, converted to some internal data structure and

interpreted immediately.

Another common strategy

where the target program representation

not

directly executable by any hardware machine. Instead a machine simulator software

developed to interpret the generated code in a similar fashion that a hardware chip

does via silicon circuits. This hybrid language implementation strategy consists

part compilation and interpretation and has been adopted in early implementations

Pascal (such as UCSD Pascal), and more recently, the Java programming language.

1.2 Translator Components

Whatever strategy

used to implement a language system, the language processor

generally may be viewed as two logical components. The analyzer component

determines the structure and meaning

the source program.

illustrated in figure

1-1, it will typically tokenize the input program, identify program constructs, and

verifies that their associated semantics are meaningful and sensible, and does not

impeach on any language rules.

Introduction

Language

Implementation

source - analysis:

program

synthesis:

•

source handling

lexical analysis

syntactic analysis

intermediate

code

generation

machine independent optimisation

code

generation

with

storage allocation,

code

selection,

and

meaning

L-

______

-L~m~a~c~h~in~e~d~e~p<en~d~e~n~t~oLPtt=im~iz=a~ti~o~n

______

_J~

code

Figure 1-1: Compiler Components

Working from the semantics

the source program

determined by the analyzer

component, the synthesizer component generates suitable code for execution on the

target machine. The most important requirement

a compiler is that it works

correctly by generating the correct codes. The execution

generated code must

produce the same effects as that specified by the program accordingly to language

rules and semantics.

the case

an interpreter, the synthesizer component

executes the internal form immediately.

In generating appropriate target instructions, the synthesizer component can

principle select from a wide range

available instructions and data addressing

formats.

can optimize execution on the target machine by performing register

allocations so that frequently accessed variables may be effectively slaved. As such,

variables may thus be accessed from high-speed registers rather than slower memory

areas. Other optimization techniques include instruction resequencing

as to

reduce the number

redundant memory store and retrieve instructions, or to

eliminate the re-evaluation

common sub-expressions. Finally, the synthesizer

component could also incorporate machine dependent optimization to fully exploit

the characteristics

the target machine.

The analyzer and synthesizer components are also known as the front- and back-

ends

the language implementation system. With this framework, the front-end

merely dependent on source language, while the back-end is dependent on the target

host machine.

supports re-targeting

a language implementation in two ways:

• A language implementation for language X for machine P differs from that for

machine Q only

its back-end. Both systems require the same front-end

Introduction to Language Implementation

component to analyze the (same) source language, but must generate

instructions for different target machines.

• Language implementations for two similar languages

X and Y for the same

machine are likely to use similar target instruction sets and run-time machine

organization. It

thus possible to reuse the back-end generator when

considering the implementation

a different language but with similar program

structure.

The frontlback-end decomposition is consistent with software engineering principles

well

promoting software reuse. Conceptually, a compiler may be more easily

understood when viewed in smaller and more manageable units. The same

abstraction principle applies and aids understanding and ultimately, maintainability

overall language implementation. When modular components with clean

interfaces are well-defined, they can operate in other contexts and thus promote

reusability.

1.2.

1 Source Handling

The main function

the source handler

provide the lexical analyzer with an

input stream

characters from the source files.

an alternative to a separate

preprocessor

in the case for

or when required by a particular source language, a

source handler could also provide simple preprocessing functions such as file

inclusion or options handling. Some dialects

Pascal have adopted this approach,

where file inclusion and options may

specified within the input stream via

specially marked comments.

A source handler often also includes a suitable interface to the error reporting

mechanism, and in tum, to the output listing handler.

useful for a compiler to

detect as many program errors

possible in a single run.

this aspect, diagnostic

messages should

reported accurately, and as near

erroneous program fragments

possible. The close relationship between the input handler, output handler and

error reporting mechanisms is thus mutually beneficial.

1.2.2 Lexical Analysis

The role

the lexical scanner

to identify tokens or terminal symbols in the input

stream. A lexical scanner for a Pascal compiler will identify the string

procedure

the keyword which starts a procedure declaration. Similarly, the string

count

would

be identified as an user-declared identifier which would be used in the context

user-defined constant, type name, variable or procedure name.

Introduction

Language Implementation 5

Tokens may ultimately be represented by an integer sub-range. For example, the

if,

then,

else,

while,

and

keywords may be represented via arbitrary discrete

token values

12,

13, 14,

and

respectively. The representation

some tokens

however require additional values. The string

count

may be identified

an user-

declared identifier with symbol value

10,

but this does not differentiate it from the

string

endOfFile.

In this case, the lexer must also supply the actual spelling or an

equivalent unique hash value

the string. Such representation issues are further

discussed in Chapter

program

test

(output)

;

var

integer;

begin

for

j:=l

writeln(j)

end.

lexical analysis

•

program_sym

identifiecsym

"test"

lefc.parenthesis_sym

identifiecsym

"output"

semicolon_sym

vacsym

identifier_sym

"j"

colon

identifiecsym

"integer"

semicolon_sym

begin_sym

for_sym

identifiecsym

"j"

assignmenc.sym

integecsym

to_sym

etc

Figure 1-2: Lexical Tokens

As with identifiers, literal values must be encoded appropriately to differentiate the

token kind from its literal value. For example, integer and real literals may be

distinguished with token values

8 and

and the lexical scanner also returning the

literal value proper. Figure

1-2

illustrates the results

lexical analysis

a typical

source program.

Lexical scanners usually attempt to determine the longest symbol to match a

token. For example, the

string:

= could have been interpreted

two tokens

colon:

and equal

But in determining the longest string, it

matched with the assignment

symbol :

Similarly, the string

output

could have been interpreted

6 user-

declared identifiers

u and

but again on matching the longest token, it

interpreted

a 6-character identifier

output

instead.

The ability

determine the end

a token requires at least a one-character

lookahead. For example, in Pascal, a one-character lookahead

required

剩余220页未读，继续阅读

xiaofuss2

粉丝: 0
资源: 7

编译器技术探索：工具、翻译器与语言接口实现

嵌入式体系结构及接口技术：ARM ADS集成开发环境.ppt

计算机组成与设计：硬件、软件接口中文第4版pdf

HolyC编译器：TempleOS的HolyC语言的编译器和工具娱乐

编译器：

ARM编译器特性：数据包络分析与语言一致性

8051微处理器C编译器手册：KEIL C51中文翻译

编译器入门：词法分析与正则表达式解析

CodeVisionAVR C编译器详解：从注释到调试

PICC-16C编译器教程：从入门到精通

GCC 4.9.0 编译器详解： GNU 组织的编译原理指南

最新资源