CS 4850 - Programming Languages

HW1

SummerII 2007

Given: July 2, 2007

Due: PART1: July 11, 2007 - postponed to 7/13/07 (1:30pm); PART2: July 13, 2007 - postponed to 7/18/07 (1:30pm)

This assignment involves recognizing tokens, building a symbol table, and constructing token strings for a LISP-type language. You are to write all code in C++/Java. As an example of what your program should produce for its token strings, consider the following:

Input

( CAR '(B D C))
(+   +15 ( *  -20 2))
(CONS '(B D C) (CONS ``A (B)'C D" '(BB BB)))
^Z

Output: NOTE: Separate the output lines for each line of input, as shown)

( CAR '(B D C))
                     ( 157 161 ( 250  80  9 ) )

(+   +15 ( *  -20 2))
                     ( 35 100 ( 74 102  91 ) )

(CONS '(B D C) (CONS ``A (B)'C D" '(BB BB)))
                    ( 34 161 ( 250  80  9 ) ( 34  209 161 ( 65  65 ) ) )

:

Of course, the Symbol Table addresses shown are arbitrary; yours may be quite different. Print actual parentheses, but symbol addresses as shown for other symbols. For now, the only types you must recognize are "Token", "LeftParenTok", "RightParenTok", "QuoteTok", "Integer", and "String".

When testing your program you may use a terminal for I/O, but for the later runs use the input file hw1.dat.
Before printing the token string for each line of output, you are to print the input line (as in the example). Compress the output line to 6 characters per symbol except parentheses which should be two characters (ex: "( "). After the EOF (^Z) is encountered, print the non-empty symbol table elements in order.

Symbol Table Construction

Use 256 elements numbered 0 - 255. Each element should allow for a string of maximum length of 28 characters and a type code. Access the table through a very simple hashing function. Of course you must be able to handle collisions.

Recognizing Tokens and Constructing Token Strings

Implement this first through a finite state machine (fsm) and then using lex, both of which will be discussed in class. Use the following rules.

Parentheses, blanks, single quotes, and double quotes are delimiters. Each identifies the end of the current token. In addition, parentheses and single quotes themselves are tokens. Double quotes delineate strings and should be stored as a part of the string. Of course blanks, parentheses, and single quotes within strings are not delimiters.

C++/Java Design and Implementation

There are two parts to this assignment. Part1: Implement the FSM directly using a high level language such as C/C++/Java. Part2: Use lex to generate the FSM and the required output. Start by doing part1 and then move onto part2.

You should know that most of what you develop for this assignment in part2 will also be used in the semester project, so do a GOOD JOB!  Use top-down and object-oriented design. Use extensive modularization, abstraction, and information hiding. The main program can be a simple controller for the rest of the program. Document your program well. All packages/classes should tell where they are used and who uses them. You do not need to handle any exceptions for this assignment. Programming assignments that are just spaghetti / monolithic programs coded in Java/C++ will be heavily penalized or not accepted at all.

Implement token strings as linked lists of symbol table addresses.

Turn in: (with your name and WMU-ID number hand printed on the outside of your hardcopy submission)

• Hardcopy Submission:

1. Verified script showing input and output for assigned test runs.
2. Complete listing of program and documentation.
3. A complete table showing your fsm. This may be computer printed or hand printed, but should be carefully labeled for readability.
4. To get the script/log file use "script" command of Unix. (After script, cat the source program files and input files, run the program, cat the output files - for steps 1 and 2 [and 3 if you typed in your fsm].)
5. Your top-down and object-oriented design diagrams.
6. HIGHLIGHT APPROPRIATE CODE WITH YELLOW.
7. Label the items you turn in and turn them all together.

• Online Submission

1. A copy of your zipped file consisting of of your source codes, scripts (to run your program), a couple of sample executions of your solution and report.
2. Use <hw#cs4850_yourlastname_mmddyy.{zip,ppt,doc,tex}> as the naming convention for your zipped, ppt, MS-Word, or LaTeX files when emailing your submission to gupta@cs.wmich.edu. Replace '#' with the appropriate homework number.
   

Penalty for Late Submission

10% per day (including weekends). Turn in programs at the start of class when due.

Any student may be asked to show and discuss his solution in class, so be ready with your presentation.

For programming assignments, submit a zipped file of your source codes, scripts (to run your program) and report along with a hardcopy of your source codes, scripts, a couple of sample executions of your solution and report.

Use <hw#cs4850_yourlastname_mmddyy.{zip,ppt,doc,tex}> as the naming convention for your zipped, ppt, MS-Word, or LaTex files when emailing your submission to ajay.gupta@wmich.edu. . Replace '#' with the appropriate homework number.

REMINDER: You are responsible for making yourself aware of and understanding the policies and procedures in the undergraduate (pp. 268-270) [Graduate (pp. 24-26)] Catalog that pertain to Academic Integrity. Additionally, easy availability of information, material, source codes, lecture notes etc on the Internet may make it possible to find solutions to your assignments on the Internet or elsewhere. It is okay to refer to those, understand them and use them to enhance your solutions, generate your own ideas etc. However, you must give proper and full credit to original authors of the work, if you include their ideas. Failing to do so is part of academic and professional dishonesty. It will not be tolerated in this class. Do not give in to temptations....

Home