CS 301 Homework 1 - due 3:30pm, Wednesday, January 20, 1999

Interpreting the E Language

The E language is a small expression-based language described by the following grammar.

program $\rightarrow$ exp
     
exp $\rightarrow$ term
  $\rightarrow$ term ' +' term
  $\rightarrow$ term ' -' term
  $\rightarrow$ id ' =' exp
  $\rightarrow$ if0 exp then exp else exp
     
term $\rightarrow$ ID
  $\rightarrow$ NUM
  $\rightarrow$ ' (' exp ' )'
  $\rightarrow$ ' {' block ' }'
     
block $\rightarrow$ var vars ' ;' exps
  $\rightarrow$ exps
     
vars $\rightarrow$ var
  $\rightarrow$ id vars
     
exps $\rightarrow$ exp
  $\rightarrow$ exp ' ;' exps

The keywords are var, if0, then, and else. Identifiers (ID) are sequences of alphanumeric characters, beginning with a letter, excluding the keywords. Numbers (NUM) are sequences of digits.

A program is just an expression, which evaluates to a single integer result.

The semantics of expressions, terms, and blocks can be informally described as follows.

Your task in this assignment is to produce a static checker and interpreter for E programs. The checker makes sure - without actually evaluating the program -- that all variables are declared (in some enclosing block) before they are used. The interpreter evaluates the program and writes out the integer result.

Your interpreter should take a single command-line argument specifying the name of a file containing the program to interpret. If the program parses and checks correctly, the interpreted result should be written to standard output. Any parsing or checking errors should be written to standard error.

Don't worry about the possibility of integer overflow during evaluation.

Implementation and Assignment Submission

Much of the work for this assignment has already been done for you. The relevant files are available in /u/cs301acc/1 and via the course web page. Files ast.h and ast.c define a data type and constructor functions for representing the abstract syntax of E expressions. File parse0.c contains a lexical analyzer and parser for most of E, with the exception of the subtraction operator (-) and the if0 expression; check0.c contains a static checker for the same subset of E. File main0.c is a driver that reads, parses, and checks expressions conforming to the subset.

You are strongly encouraged, though not required, to use these files as the basis of your solution. If you do use them, all you need to do is:

(a) Extend the code in parse0.c and check0.c to support the subtraction operator and the if0 expression.

(b) Write a new function int interp(Exp e) that interprets a program in the E language (including subtraction and if0) and returns its value. You may assume that interp will only be called on programs that have passed the checker, as illustrated in the commented-out portion of the driver main0.c.

Hint: your interpreter can be structured as a recursive function quite similar to the checker.

A working solution to the assignment is in /u/cs301acc/1/interpreter. Your program should generate the same output as this one for correct programs; error messages may differ in format, though not in substance.

You will probably want to place your interpreter code in a new file interp.c. In any case, be sure to include in your submission a makefile that builds your interpreter given your source files and leaves the executable program in interpreter.


Andrew P. Tolmach
1999-01-06