Week 1 #

Translation #

The process of converting high-level source code into machine code.

A Programming Language (PL) is neither compiled or interpreted, it’s implementation can be.

Types of Translation #

• Compilation: translated before execution

  [Source Code] -> {Compiler} -> (Target Code)
  

  • Target code optimises for speed and security verification (potential memory leaks) for the specific machine

    • ‘Heavy’ information (i.e. variable types) only needed during compilation

  • Can expose potential errors/bugs before run-time

  • C, Rust are complied languages

• Interpretation: translated line by line

  [Source Code] -> (Interpreter)
  

  • Useful for prototyping, scripting

  • Easier to debug; stack trace uses the source code rather than binaries

  • More portable; only requires an interpreter

  • Python, JavaScript, Shell are interpreted languages

• Pseudo-Compilation: hybrid of compilation and interpretation

  [Source Code] -> {Compiler} -> [(Intermediate Code)] -> (Interpreter)
  

  • Compiler translate all of the source code before execution but into Intermediate Code

  • Interpreter executes the Intermediate Code line by line

  • Intermediate Code can be executed on any machine with the interpreter

  • Java is a hybrid language, it’s intermediate code is bytecode

The Process of Translation #

1. Lexical Analysis:

   • Creates a sequence of tokens; the smallest meaningful element (language specific)

2. Syntactic Analysis

   • Structures tokens into an Initial Parse Tree

     • internal nodes are syntax rules, leaves are tokens
     • Rules are determined by a Precedence Table

   • Where syntax errors are discovered

3. Symantic Analysis

   • Annotates parse tree with semantic actions

4. Code Generation

   • Produces machine code