CS212 Spring 2002
3/27/2002
Lecture 10: No More FBR! ;  Translation ; Abstract Syntax Trees
-------------------------------------------------------------------------------
[0] Announcements:
+ reading: Brookshear 5.1, 5.4, 7.5
+ extended due date 4/15, 1:00:00 pm
+ some useful online notes:
  http://www.slc.edu/~msiff/courses/compilers/notes/absyn.html
  http://www.cs.wisc.edu/~cs536-1/Notes/1.OVERVIEW
-------------------------------------------------------------------------------
[1] Summary from last time:
+ no need to continue with FBRs -- I think you understand it by now :-)
-------------------------------------------------------------------------------
[2] Overview for today:
+ more details on compilers and intro to syntax trees
+ CS211 introduces trees on Thurs 4/4, so we can do that now
-------------------------------------------------------------------------------
[3] Language Implementation

+ see section 5.4 in optional textbook
+ showing a bit more detail on what's going on
+ TRANSLATOR vs COMPILER? see pg 228
  translator: convert language into another language
  compiler: convert a high-level language into assembly/machine code
-------------------------------------------------------------------------------
[4] Translation

+ TRANSLATION: converting a program from one language to another
               SOURCE PROGRAM gets translated into OBJECT PROGRAM

+ process of translation (pg 255, Fig 5.12):

  source       lexical       parser     code         object
  program      analyzer                 generator    program







+ note: steps are usually intertwined
+ see also http://www.cs.wisc.edu/~cs536-1/Notes/1.OVERVIEW for another 
  description and terms
-------------------------------------------------------------------------------
[5] Lexical Analysis

+ LEXICAL ANALYSIS: recognizing which strings of symbols from the 
                    source program represent a single entity
+ TOKEN: lexical unit, or one of the entities recognized by the lex analyzer

-------------------------------------------------------------------------------
[6] Parsing

+ PARSING: process of identifying the grammatical structure of the program
           and recognizing the role of each component

+ formatting:
  FIXED-FORMAT LANGUAGE: appearance and position matters for parser
  FREE-FORMAT LANGUAGE: positioning of statements is irrelevant

|=================================| Page 1 |=================================|
[7] Parsing continued

+ SYNTAX DIAGRAMS:
  - see pp 257-258
  - graphical ways of representing the syntax rules
  - TERMINALS ("finished" nodes) and NONTERMINALS (need further expansion)




+ PARSE TREE:
  - see pp 259-260
  - graphical ways of representing how code conforms to syntax
  - represents parser's understanding of program's grammar

    ex) x + y





+ SYMBOL TABLE:
  - storing variable info
  - see book for details (pg 261)
-------------------------------------------------------------------------------
[8] Abstract Syntax Trees

+ Program Hierarchy

  public static void xor(boolean x, boolean y) {
     if (x != y)
       return true;
     return false;
  }

  hierarchy/flow?

  statement block:  __________

  selection statement: __________

  terminals: __________

+ AST:
  - high-level description of program hierarchy
  - relations of program syntax, not punctuation & other elements
  - effectively "condenses" a parse tree
  - helps to build your compiler!

+ Examples)

    ( x + 1 );


    parse tree                       AST







|=================================| Page 2 |=================================|
[11] Abstract Syntax Trees continued

    x := 1 ; y := (x + 1) ;

    See also Example 1 on
    http://www.slc.edu/~msiff/courses/compilers/notes/absyn.html

    parse tree                                     AST











    if ( ( x > 1 ) )
       return x ;
    else
       return 0 ;









    while ( ( x < 4 ) ) do {
      y := f( ( x + 1 ) ); 
      x := ( x + 1 );
    }





+ more examples? 
  - there are trees and lists for each "blah->something" in the grammar
  - see next lecture and Part 4 notes (forthcoming)

-------------------------------------------------------------------------------
[10] Code Generation

+ CODE GENERATION: process of constructing the machine-language instructions
                   to simulate the statements recognized by the parser

-------------------------------------------------------------------------------
[11] Linking and Loading

+ terms pop up quite a bit with C/C++ programming
+ see pp 262-264
+ LINKING: connections to needed operating system utilities
+ LOADING: put complete executable program into memory




|=================================| Page 3 |=================================|