CS 47 - T. Howell - Spring, 2008

Homework Assignment #2

Note: This page will look best with Internet Explorer. Firefox does not display some of the fonts as intended.
Due Thursday, February 28, 2008 by 23:59 as an email attachment hw2.zip (details below).

You are to write two short C programs and solve two exam-style problems. Details on the programs are as follows.

Program #1: buildfloat.c

This program takes three inputs representing the sign, exponent, and significand of a float. It uses them to construct the float and print it both as a float (using %.8g format) and its 8-digit hexadecimal representation. Your program need not deal with infinity or NaN (not a number), but it should deal properly with both normalized and denormalized floats. Here are some sample outputs to show the behavior of this program:

> buildfloat
Usage: buildfloat sgn, exp, sig (parts of floating point number)
> buildfloat 1 127 1.fffffe
value = 3.4028235e+38 (7f7fffff)
> buildfloat -1 0 1.000100
value = -1.0000153 (bf800080)
> buildfloat 1 -2 1.e62414
value = 0.474747 (3ef3120a)
> buildfloat 1 -127 0.00dab0
value = 3.9225147e-41 (00006d58)
> buildfloat 1 127 2.000000
Usage: sgn(+1 or -1), exp(-127 to +127), sig(0 to 2-eps)
> buildfloat 1 -127 1.000000
Invalid parts for denormalized float
> buildfloat 1 -126 0.fffffe
Invalid parts for normalized float

As before, your code should always print an error message if a command line input is not of the proper form or a value is out of the proper range.
Hint: I suggest reading the integer part and fractional part (× 2²³) of the significand separately as two integers.

Program #2: floatparts.c

This program is roughly the inverse of Program #1. It takes a list of floats and prints out the parts: sign, exponent, and significand. Here are some sample outputs:

>floatparts
Usage: floatparts f1 f2 f3 ... (list of floating point numbers)

>floatparts .474747 3.9225147e-41 -3.4028235e+38
input 1 = .474747
sign = +, exponent = -2 significand = 1.e62414

input 2 = 3.9225147e-41
sign = +, exponent = -127 significand = 0.00dab0

input 3 = -3.4028235e+38
sign = -, exponent = 127 significand = 1.fffffe

Try to make your output look as close to these sample outputs as possible.

Here are two functions that you might find useful. The function l2f turns an int into a float with the same bit pattern. The function f2l turns a float into an int with the same bit pattern. This is a different result than you get with a cast: l2f(x) != (float)x . Try to figure out how they work, but that is not part of this homework:

float l2f(int l)
/* convert long integer to float with same bit pattern */
 {
  union {
    float f;
    int ll; 
  } a;
  a.ll = l;
  return a.f;
}

int f2l(float f)
/* convert float to long integer with same bit pattern */
 {
  union {
    float ff; 
    int l; 
  } a;
  a.ff = f;
  return a.l;
}

Finally, students looking for an extra challenge (and 50% extra credit for this assignment) may write analogous programs builddbl.c and dblparts.c. These build and dissect doubles instead of floats. You might want to use a 64-bit integer datatype called long long int which is supported by gcc. This is significantly more challenging, and extra credit will only be given for programs that are completely correct. Submit only one pair of programs (for float or double but not both).

The two written problems are contained in either file Homework 2 Problems (.doc) or Homework 2 Problems (.rtf). They should be submitted in the form of a MS Word file (.doc) or as a .txt or .rtf file along with your programs.

Submission requirements and grading system

As noted above, you are to email me your homework as a zip file hw2.zip, which should contain one problems file (.doc, .rtf. or .txt), and two source code files only, with the exact names as given above, and with no paths embedded in the zip file. Your email must be sent to me by 23:59 on Thursday, February 28. Your email must have the following subject line:

CS 47 Section x Homework #2 John Doe

but of course with your own name instead of "John Doe" and your own section number in place of "x" (1 = 1:30, 2 = 5:30). Please also preserve all spacing and capitalization in this subject line.

As with any code you submit, each code file must be appropriately documented with a header comment containing (as a minimum) your name, the class and section number, the homework problem number, and a brief description of the problem.

I will grade according to three criteria: execution tests (70%), source code (20%), and documentation (10%). I believe I have specified these three programs very precisely above, but if you have any questions about the behavior of the programs it is your responsibility to ask me before the submission date.

Failure to observe any of the submission requirements stated above may result in a grade of 0 on this homework!

Happy programming!