HW3: Integer/Hex Conversions Solution

Objectives



The objective of this assignment is to get you familiar with integer and character string manipulations. You are asked to covert an integer into its equivalent hex string, and then covert the hex string back to integer format. You are not allowed to use any integer to string manipulation library functions, such as itoa() etc. Make a subdirectory "hw3" in your cs240 directory for this assignment and copy the files from /courses/cs240/s15/cheungr/GROUP/hw3/. Use the gdb debugger to help you troubleshoot your program. Make a typescript that lists your programs, the execution results and intermediate debugging steps.




2. Integer <- Hex String Manipulation



(a) Write the C function itox() to covert an integer to a hex string. Its prototype is described in the header file xbits.h in this directory. Put your code in file xbits.c in the hw3 subdirectory of your cs240 directory. You can use the stub xbits.c from this directory to start with.




The idea behind the itox function is to convert an int variable (int is 32 bits on our machine) directly to an ascii string of hex numbers, '0','1'…’F’. In this exercise, we will only deal with positive integers.




The algorithm to covert decimal number to hex is to first divide the int number by 16, the remainder will form the first hex digit (the last character in the string). Use ‘0’ to represent reminder = 0; ‘1’ for remainder =1; ‘A’ for remainder =10; ‘B’ for remainder =11 and so on. Next, divide the quotient by 16 and the remainder will form the second hex digit (next to the last). This process is repeated until the quotient is equal to 0 or all the characters in the string have been filled up. Then the algorithm stops. If the algorithm stops with quotient equal to 0, it fills the rest of the hex string with ‘0’. Print out the input integer number and its corresponding hex string to see if it is correct. Remember to terminate the string with ‘\0’.




Note that the array hexstring declared by the caller can be declared with a size 2*sizeof(int) + 1. Now sizeof() is a function evaluated at precompile time which gives the number of BYTES in a variable of a given type -- 4 bytes to an int on our machine, but it might be 8 bytes on a different machine and sizeof() would reflect that. There are 8 bits to a byte and so we expect 2*sizeof(int) chars ‘0’, ‘1’,…’F’ in a representation of the hex value. The +1 is to leave space for the final '\0' in the string.







(b) Write the C function xtoi() to convert a hex string to an integer. Its prototype is also described in xbits.h. Put your code in file xbits.c in the hw3 subdirectory of your cs240 directory. Your algorithm should only accept the valid hex digit 0,1,2,3,4,5,6,7,8,9,A,B,C,D,E,F.




The algorithm to convert hex string to a decimal number is to start the string from the last character. Remember to skip over the ‘\0’ character. Convert the corresponding character to its decimal equivalent (e.g. a character A will become 10) and multiply the number by powers of 16. The decimal equivalent of the last character will be multiplied by 160; the next by 161 and so on. The result is obtained by summing all the products together. Print out the hex string and its equivalent decimal number to see if it is correct.







NOTE: The functions in xbits.c must be called by another program, and will not generate any output. So to test them you will have to write a driver (a main program that calls these functions and with appropriate arguments and outputs results -- this should be named showxbits.c). You must then compile the driver and link the compiled driver object code with the object code for the functions you write, combining them into a single executable. Note that there is a stub showxbits.c driver you can use to test the program and the code in xbits.c and xbits.h. To test bits initially, you can give the command to build and run the showxbits program:




gcc showxbits.c xbits.c -o showxbits

./showxbits




The xbits.h file is called a header file. The header file is the glue that makes sure xbits.c and the driver have the same prototypes for the functions in xbits.c. It guarantees that any correct driver can link with any correct implementation.




Interactive Integer <- Hex String Manipulation


Rewrite showxbits.c so that it reads integers one at a time from stdin using the library function scanf (the equivalent to printf that does input -- look this up in K&R pg.157).




For each integer it reads it should write the input decimal, converted hexadecimal, and the reconverted decimal representations. The hexadecimal output strings requested by showxbits are to be calculated by calling itox(). To generate the reconverted decimal, you can call xtoi(). Use printf with the %x format for integers input, %s format for hex strings you've created.




Use the value returned by scanf to end showxbits execution when it scans a non-integer (or EOF). When scanf is properly converting a single value under its conversion %d, it should return 1 (the number of tokens converted). See pg 245, fscanf, end of the first paragraph. The code you need to write in showxbits could look like this:




while (scanf("%d", &n) = = 1) {

itox( hexstring, n);

m = xtoi( hexstring);

printf("%12d %s %12d\n", n, hexstring, m);

}




Note: showxbits program should *not* prompt the user for input. "No prompt" is a UNIX convention for reading files from stdin. There's a good reason for the convention. It allows you run showxbits easily and cleanly either by typing in the input data or using input redirection. Create a typescript to show your results.




Debugger gdb



Your typescript file should show that you have demonstrated the use of the gdb debugger. First build your program with the gcc –g option for gdb. Show the following gdb commands (shown in bold letters) in your typescript file:




gdb progname

b(reak) main

r(un)

in some reasonable order:

s(tep) n(ext) (learn the difference!)

at any useful time, p(rint) some expression

l(ist) source code lines

bt see call stack

help for more information

q(uit) to conclude




*** NOTE: IF YOU SEE A RUNTIME ERROR THAT SAYS "core dumped", DELETE THE "core" FILE FROM YOUR SUBDIRECTORY RIGHT AWAY OR RISK RUNNING OUT OF DISK SPACE. *****