Class 13: Predictive Parsing, Continued

Held Friday, September 27, 2002

Summary

Today we continue our exploration of predictive parsing by looking at steps for automating the construction of predictive parsers..

Notes

• Oh boy! Class 13 falls on a Friday. Luck is with us.
• Microsoft has just released a vision of the future. I particularly like the idea of voice mail over email. As one colleague suggested, Several contributors to the Slashdot commentary on this story are looking forward to the arrival of spam in the Cubicle of the Future, complete with Active X controls that crank up the volume before rendering those invitations to [spam topic of your choice].
• I brought poptarts and pudding today.

Due

• Project, Phase 1.

Assignments

• Homework 2: Grammars and Parsing

Overview

• Helpful tables: First, Follow, and Nullable
• Building the First table.
• Building the Nullable table.
• Building the Follow table.

Analyzing the Grammar

• As you can tell, we need to make a number of decisions when we convert a grammar into a predictive parser.
  • Given multiple right-hand sides that start with a nonterminal, which one do you choose?
  • If a nonterminal derives epsilon, how do you decide to apply that derivation?
  • What if a nonterminal derives a nonterminal that derives epsilon? Consider what to do when you see a b when matching an S in

    S :: = X b
    X ::= x X
       |  epsilon
    

• All of these analyses are aided by three tables (and corresponding functions) known as First (and first), Follow, and Nullable (and nullable).
• The tables and functions are recursively constructed. That is, the functions may be defined in terms of the tables and the tables may be defined in terms of the current versions of the functions.
• First maps nonterminals to sets of tokens that can begin strings derived from those nonterminals.
  • first maps sequences of nonterminals and terminals to the symbols that can begin strings derived from those sequences
• Follow maps nonterminals to sets of tokens that can follow those nonterminals in sequences derivable in the grammar
  • There is no nullable function as it is neither interesting nor useful.
• Nullable indicates which nonterminals can derive the empty string
  • nullable operates on sequences of symbols (nonterminals and terminals), returning true only when the sequence can derive epsilon

Building the First table and first function

• How do we define the first function?
  • We know that if a sequence begins with a token, then it can only derive strings that begin with that token.
• But what if the sequence begins with a nonterminal?
  • Then the sequence can derive strings that begin with any token that begins strings that the nonterminal can derive.
• What what if the nonterminal can derive the empty string?
  • Then we need to look further in the original sequence.
• Putting it all together

  TokenSet first(SymbolList string) {
    Symbol s1 = string.car();
    if (isToken(s1)) {
      return new TokenSet(s1);
    }
    else {
      tmp = First[s1];
      if (Nullable[s1]) {
        return new Union(tmp,
                         first(string.cdr()));
      } // if s1 is nullable
    } // if s1 is a nonterminal
  } // first
  

• Note that this relies on the First table and the Nullable table.
• How do we build the First table? Using the first function.

    for each nonterminal N, First(N) = emptySet();
    repeat
      for each production N ::= RHS
        First(N) = union(First(N), first(RHS))
      end for
    until no changes are made
  

• How do we know this terminates?

Building the Nullable table and the nullable function

• The nullable function is relatively easy to define

  boolean nullable(SymbolList string) {
    // The empty string can derive the empty string
    if (String.isEpsilon()) {
      return true;
    }
    Symbol s1 = string.car()
    // If we begin with a token, we're not nullable
    if (isToken(s1)) {
      return false;
    }
    // Otherwise, we begin with a nonterminal.  That nonterminal must
    // be nullable, as must the rest of the sequence.
    else {
      return Nullable[s1] && 
             nullable(string.cdr());
    }
  } // nullable
  

• And, once again, we can define the Nullable table in terms of the nullable function.

    foreach nonterminal N, Nullable(N) = false
    repeat
      for each production N := RHS
        if nullable(RHS) { 
          Nullable[N] = true; 
        }
      end for
    until no changes are made