UMBC CMSC202, Computer Science II, Fall 1998, Sections 0101, 0102, 0103, 0104

24. Expression Trees

Tuesday December 01, 1998

Assigned Reading:  none

Handouts (available on-line):

Programs from this lecture:

• Expression Trees

Topics Covered:

• Discussed the extra credit portion of Project 5. Some highlights:

  • Suppose that we have a Binary Search Tree (BST) that satisfies the condition that at every subtree with 5 or more nodes, neither subtree is more than twice the size of the other subtree. We argued that the height of such a BST is O(log n). In fact, it is roughly log_(3/2) n.

  • Since the height of such a balanced tree is O(log n), BST operations such as find and max take O(log n) time.

  • Insert and delete might take O(n) time in the worst case, since we may have to rebalance the tree during an insertion or a deletion.

  • However, not every insert and delete would require us to rebalance the tree. We argued (not very rigorously) that on average the running time of insert and delete is also O(log n). This average is taken over the operations performed on the BST (we are not taking the average over the data stored in the tree). In such a situation, we say that the amortized running times of insert and delete are O(log n).
   

• Another kind of binary tree is an expression tree --- a representation of arithmetic expressions. For example, the expression 2 + 3 can be represented as a binary tree with a root labeled with +. The root would have two children labeled with 2 and 3. For more complicated arithmetic expressions, the root of the tree is the operation that is performed last. The left subtree and right subtree of the root are the left and right operands of that last operation. For example, in the expression (2 + 3) * (4 + 5), the * is performed last, so the left and right subtrees represent (2 + 3) and (4 + 5), respectively.
  

  Note that in an expression tree, we do not need to store parentheses, since the order of evaluation is implicit in the structure of the tree.

• We can use the template Tree class from Lecture 22 to instantiate an expression tree class. Each node of the expression tree will be an object from a Token class which we describe below. We use the expression tree to help us build a calculator program. This program considers operator precedence of * over + and works with nested parentheses. Another feature of this program is its ability to report errors (and point to the location of the syntax error). See sample run.

• The calculator program is divided into 2 parts. The parser (header file and implementation) works with higher level concepts (expressions, terms and factors). The tokenizer (header file and implementation) does the string and character manipulations for the parser. The TokenStream class is initialized with a string which contains an arithmetic expression. The member functions look() and eat() allows us to obtain the next syntactic unit of the string (the next number, operator, etc.). This syntactic unit is returned as an object of type Token.

• The parser essentially follows the derivation rules for a legal arithmetic expression. The grammar we use to describe the set of allowed expressions is:

      Expression -> Term + Expression
      Expression -> Term - Expression
      Expression -> Term
  
      Term -> Factor * Term
      Term -> Factor / Term
      Term -> Factor
  
      Factor -> ( Expression )
      Factor -> Number
  

• We showed how a simple expression like (3 + 4) * (72 - 70) can be derived using the grammar above. This is a mechanical process that is tedious to do by hand, but perfect for a program to follow.

• Our parser is simply a set of mutually recursive functions (ParseExpression(), ParseTerm() and ParseFactor()) that follow the rules for deriving these arithmetic expressions. If any of these functions find an error, it throws an exception. Each of these functions also catches exceptions. When an exception is caught, memory for any partially built expression trees is deallocated. Then the exception is rethrown to be caught again and rethrown again ... until the function Parse() finally catches the exceptions for the last time.

• The function evaluate() recursively computes the value of an expression tree. Thus, we have all the parts of a calculator program. Build an expression tree from an arithmetic expression and use evaluate() to determine the value of the expression.

  • First main program and sample run.

  • Second main program and sample run.