Compiler Construction

|
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
 5 views
of 23

Please download to get full document.

View again

Description
Compiler Construction. Lecture 1 MAJID MUMTAZ. Department of Computer Science, CIIT Wah. Course Textbooks. Compilers – Principles, Techniques and Tools Authors: Alfred V. Aho Ravi Sethi Jeffrey D. Ullman. Course Objectives.
Share
Transcript
Compiler ConstructionLecture 1MAJID MUMTAZDepartment of Computer Science, CIIT WahCourse Textbooks
  • Compilers – Principles, Techniques and Tools
  • Authors:
  • Alfred V. Aho
  • Ravi Sethi
  • Jeffrey D. Ullman
  • Course Objectives
  • Introduce and Describe the Fundamentals of Compilation Techniques
  • Assist you in Writing your Own Compiler (or a part of it)
  • Overview & Basic Concepts
  • Lexical Analysis
  • Syntax Analysis / Syntax-Directed Translation
  • Semantic Analysis
  • Intermediate Code Generation
  • What is a Compiler?
  • Programming languages: Notations for describing computations to people and to machines
  • E.g. Java, C#, C++, Perl, Prolog etc.
  • A program written in any language must be translated into a form that is understood by the computer
  • This form is typically known as the Machine Language (ML), Machine Code, or Object Code
  • Consists of streams of 0’s and 1’s
  • The program that carries out the translation activity is called the compiler.
  • Basic Concept
  • The language to be translated: Source language
  • Input code is called the Source code
  • The language produced: Target language
  • Output code is called Target code
  • A compiler is a program that translates source code written in one language into the target code of another language.
  • Machine LanguageJavaWhy Compile?
  • Executables: Files that actually do something by carrying out a set of instructions (as compared to files that only contain data)
  • E.g., .exe files in Windows
  • If you look at their data, it won’t make sense
  • Hex Dump of a boot loader executableThe primary reason for compiling source code is to create an executable ML programWhy Compile?
  • Once the executable is there, it can be called by the user to process the given inputs to a program and produce the desired outputs
  • Initially, there is a compile phase which is followed by the run/execute phase
  • This approach is used in many languages, e.g., Java, C++, C# etc.
  • Interpretation is concept parallel to compilation
  • There is only one run/execute phase.
  • Program InputsTarget ExecutableDesired OutputsCompiler vs. Interpreter
  • Interpreter: A program that doesn’t produce the target executable. It can do three things:
  • In a line-by-line fashion, it directly executes the source code by using the given inputs, and producing the desired outputs
  • May translate source code into some intermediate language and then execute that immediately, e.g., Perl, Python, and Matlab
  • May also execute previously stored pre-compiled code, made by a compiler that is part of the interpreter system, e.g., Java and Pascal
  • Some systems, e.g., SmallTalk, may combine 2 and 3.
  • InterpreterSource CodeInterpreterDesired OutputsProgram InputsJava combines compilation and interpretationSource CodeCompilerVirtual Machine (Interpreter)Bytecodes: Intermediate ProgramDesired OutputsProgram InputsPortability: Bytecodes compiled on one machine can be interpreted on another oneFaster Execution through Just-in-Time (JIC) compilers: translate parts of bytecode into ML immediately before executionLanguage Processing SystemSource CodeCombines source code stored in different filesPreprocessorConverts source code into Assembly Language (which is easier to generate as an intermediate representation and also, easier to debug) Preprocessed Source CodeCompilerTarget Assembly ProgramAssembler produces MLAssemblerCombines different ML parts (of the same program) that have been compiled separatelyRe-locatable MLLibrary Files Re-locatable ML FilesLinker/LoaderLinker: resolves external referencesLoader: Puts all object files into memory for executionTarget MLStructure of a CompilerCharacter StreamMachine-Independent Code OptimizerLexical AnalyzerIntermediate RepresentationToken StreamMachine Code GeneratorSyntax AnalyzerSyntax TreeTarget Machine CodeSemantic AnalyzerMachine-Dependent Code OptimizerSemantically-correct Syntax TreeOptimized Target Machine CodeIntermediate Code GeneratorIntermediate RepresentationSYMBOL TABLECompilation – Analysis Part
  • Phases involved:
  • Lexical Analysis
  • Syntax Analysis
  • Semantic Analysis
  • Determines the operations implied by the source program which are recorded in a tree structure called the Syntax Tree
  • Breaks up the source code into constituent pieces, while storing information in the symbol table
  • Imposes a grammatical structure on these pieces.
  • Compilation – Synthesis Part
  • Phases involved:
  • Intermediate Code Generation
  • Intermediate Code Optimization
  • Machine Code Generator
  • Machine Dependent Code Optimization
  • Constructs the target code from the syntax tree, and from the information in the symbol table
  • Optimization is another important activity
  • Both the intermediate and machine codes are optimized in order to achieve an efficient translation, with the least possible use of computing resources.
  • Front End & Back End
  • Front End phases:
  • Lexical Analysis
  • Syntax Analysis
  • Semantic Analysis
  • Intermediate Code Generation
  • Intermediate Code Optimization
  • Front end is machine independent
  • Back End phases:
  • Machine Code generation
  • Machine Code optimization
  • Back end is machine dependent.
  • Lexical Analysis
  • Read the character stream of the source code and break it up into meaningful sequences called lexemes (a.k.aScanning)
  • Each lexeme is represented as a token
  • <token_name, attribute value>
  • Single, atomic unit of the language
  • E.g., force = mass * 60
  • Lexeme force: token <id,1>
  • Lexeme =: token <=>
  • Lexeme mass: token <id,2>
  • Lexeme *: token <*>
  • Lexeme 60:token <60>.
  • Output: <id,1>=<id,2>*60
  • SYMBOL TABLESyntax Analysis
  • Syntax analysis: Parsing the token stream to identify the grammatical structure of the stream (a.k.a parsing)
  • Typically builds a parse tree, which replaces the linear sequence of tokens with a tree structure
  • The tree is built according to the rules of a formal grammar which define the language's syntax
  • It is analyzed, augmented, and transformed by later phases of compilation.
  • Syntax tree for <id,1>=<id,2>*60=*<id,1><id,2>60More ExamplesSemantic Analysis
  • Semantic analysis: Adding semantic information to the parse tree
  • Performs semantic checks, e.g., type checking (checking for type errors),  object binding (associating variable and function references with their definitions), rejecting incorrect programs or issuing warnings
  • Suppose force and mass are floats and acceleration is integer: type casting required
  • =*<id,1>Semantically correct syntax tree for <id,1>=<id,2>*60<id,2>inttofloat60Example: English Language
  • Ahmad told us that Faisal was going to his place
  • Who does his refer to?
  • In programming language, such ambiguities are avoided
  • { int Ahmad = 10; { int Ahmad = 15; System.out.println(Ahmad); }}15 is printedIntermediate Code Generator
  • Intermediate Code Generation: After verifying the semantics of the source code, we convert it into a machine-like intermediate representation i.e., like a program for a machine
  • Typically, an assembly language-like form is used, because it is easy to generate and debug
  • Three address code: 3 operands/instruction
  • Suppose t1 and t2 are registers
  • t1=inttofloat(60)
  • t2=id2*t1
  • id1=t2.
  • =*<id,1><id,2>inttofloat60Machine-Independent Optimizer
  • Machine-Independent Optimizer: Optimize the intermediate code, so that a better target program can be generated
  • Faster, smaller code that consumes less computation power
  • E.g., X = Y * 0 is the same as X = 0
  • Example:
  • t1=inttofloat(60)
  • t2=id2*t1 t1=id2*60.0
  • id1=t2 id1=t1
  • Eliminating the one time use of register t2, and converting 60 to a floatMachine Code Generation
  • Machine Code Generator: Converts the optimized intermediate code into the target code (typically the Machine Code)
  • This is done through the Assembly Language
  • In case of Machine Code, registers (memory locations) are selected for each variable
  • Then, intermediate instructions are translated into sequences of machine code instructions that perform the same task
  • Example:
  • id1=id2*60.0 LDF R2,id2
  • MULF R1, R2, #60 Transfer id2 into R2, multiply and assign to R1Machine-Dependent Optimizer
  • Machine-Dependent Optimizer: In this phase, after the Machine Code has been generated, it is optimized further (if needed)
  • This completes the compilation process
  • Important: Optimization is an optional activity in compilation
  • One or both of the optimization phases might be missing.
  • Related Search
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks