编译原理Review

1 Background knowledge

1.1 Programing language 5 generation

1.2 Programing language Processors 3 type

• Compiler
  

• Interpreter
  

• Hybrid Processor
  

1.3 Whole procedure of compilation

[Task] Role Function Input Output

1. Analysis Stage

   • Lexical Analysis
   • Syntax Analysis
   • Semantic Analysis

2. Synthesis Stage

   • Intermediate Code Generation TAC
   • Machine-Independent Code Optimizer
   • Code Generation 汇编代码
   • Machine-Dependent Code Optimizer
   • Reload & Link 加载链接形成可执行程序

   

1.4 Concepts

1. Language

   • Natural language
   • Artificial language (Formal language)
     • Vocabularies Table
     • Sentence: a sequence of lexemes from the alphabet table by some laws
     • Language: a set of all sentences

2. Grammar

   G = ( V T , V N , S , P ) L ( G ) = { α ∈ V T ∗ ∣ S ⇒ α } G = (V_{T},V_{N},S,P) \\ L(G) = \{{\alpha} \in V_{T}^{*} | S \Rightarrow \alpha \} G=(VT​,VN​,S,P)L(G)={α∈VT∗​∣S⇒α}
   

   4 types of grammars
   

1.5 Design grammar for a given language

how to design a grammar

$\epsilon$ - free grammar

S -> $\epsilon$ S is start symbol and S should not appear in right-side of any production

how to design a $\epsilon$ - free grammar

Ambiguous grammar

2 Inside of Compilation

2.1 Lexical Analyzer/Scanner

• Role

  

• Tasks

  • Scanning/Buffering Buffer pair

  • Lexical Analysis Tokens

    • Recognize Tokens

      • Lexical Patterns

        Regular expression

        Regular grammar

        Finite automata (FA) RE -> NFA -> DFA -> mDFA

    • Message Error or Warning

• Input/Output

  Input: Source Program

  Output: Sequence of tokens/Symbol table

2.2 Syntax Analyzer/Parser

• Role

  

• Tasks

  • Syntax Analysis

    • Top-down(Derivation) Pushdown Automata

      • Derivation & Backmatching

      • Basic 递归下降

        

        Left Recursion / Never halt

        Back Tracking / Time consuming / Inefficient

        Can’t Identify errors

      • LL(1) Algorithm

        Eliminate left recursion

        Lift left max common factors

        Construct prediction table

        

        FIRST & FOLLOW

    • Bottom-Up(Reduction)

      • Basic Algorithm

        

        Identify Handling

        Back Tracking / Inefficient

        Can’t identify errors exactly

      • LR Algorithm family

        SLR

        

        LR(1)

        

        

        LALR

        

        

  • messages

  • Identify error position

• Input/Output

  Input: Sequence of tokens/Symbol table

  Output: Parsing tree

2.3 Semantics Analysis

• Role

  Source Program

  Scanner

  Sequence of tokens/Symbol table

  Parser C、C++、Java

  Parsing tree

  semantics Analyzer

  Annotated parse tree

• Tasks

  • Evaluating the value of each Attributions of the grammar symbol in the Parse tree. (By semantics rules to do this)

    1. Type checking
    2. value
    3. code (TAC generator)

    Syntax directed Semantic Analysis 语法制导的语义分析

  • Handling errors.

• Input/Output

  Input: Parse Tree

  Output: Annotated Parse tree

设计语义规则

产生式

属性 Inherited / Synthesis 继承性 综合性

Synthesized, whose value is defined only in terms of attributes values at the children and itself. E.g., val

Inherited, whose value is defined only in terms of attributes values at the parents, siblings and itself. E.g., type

2.4 Intermediate Generation (TAC generation)

• Role

  

• Forms of three address code

• Storage Forms of three address code

• Short-curcuit(Jumping) code generation method

  

  

2.5 Machine Independent Optimization

• Role

  

• Tasks

  

  • Collecting Information

    • Control Flow Analysis

      • Blocks

        (1) The first statement is a leader.

        (2) Any statement that is the target of a conditional or unconditional goto is a leader.

        (3) Any statement that immediately follows a goto or conditional goto statement is a leader.

      • Loops

    • Data Flow Analysis
      • Dead Position
      • Live Assignment 有用的赋值

  • Transformation

    • Block/Local Optimization
    • Global Optimization
      • Loop 循环不变代码外提
      • 循环变量i优化

    Directed acyclic graph 有向无环图 (DAG)

    

• Input/Output

2.6 Run-time Environment Management

• Role

  Some Program

  Compilation

  .exe

  code area

  data area

  运行时环境管理包

• Task

  

  为什么用栈 栈好 可以用栈

  • 存储空集利用连续

  • 每次只用Top (一个size大小)

  Heap浪费空间和时间 但是灵活