CSE 374 Midterm Exam 4/30/10

Question 1. (18 points) When TAs download homework assignments from Catalyst, the
files from each user are stored in a separate directory whose name is the user’s login
name. For this problem, assume that the current directory contains subdirectories and
files like the following:

xyzzy/hw4.c frodo/homework4.c
xyzzy/test.exe frodo/foo.exe
xyzzy/core frodo/notes.txt
xyzzy/notes.txt gumby/project4.c
rabbit/stuff.c gumby/core
rabbit/stuff.exe gumby/notes.txt
rabbit/notes.txt

Give a bash shell command or set of commands to do the following tasks. You must use
wildcards (*), loops, or other shell expansions or control constructs to process multiple
directories. You may not write out individual commands, one for each directory, using
the example names shown above. You do not need to write complete shell scripts – just
give the necessary commands.

(a) Delete all of the core files and .exe files in all of the directories.

rm */core */*.exe

(b) Each of the .c files seems to have a different name. Change the name of each .c
file to hw4.c. You should assume there is exactly one .c file in each directory.

for d in *
do
mv $d/*.c $d/hw4.c
done

(An industrial-strength answer would need to deal with issues like spaces in
directory names, but we did not worry about that when grading the question.)

(c) The command mail user < file will email the contents of file to user. Give shell
command(s) to email the notes.txt files to the users whose names are given by the
directory names (so, for example, file rabbit/notes.txt should be mailed to user
rabbit.)

for d in *
do
mail $d < $d/notes.txt
done

Sample Solution Page 1 of 6

 CSE 374 Midterm Exam 4/30/10

Question 2. (18 points) Write regular expressions that could be used with grep or
egrep to locate words in a file containing a list of words, one word per line (like the one
demonstrated in class). Circle grep or egrep for each problem to indicate whether you
are using basic or extended regular expressions.

(a) All words that contain the three letters x, y, and z in that order, possibly with other
letters before, after, and/or in between the x, y, and z (possibly including more
occurrences of x, y, or z). (“oxygenize” is such a word, “dysoxidize” is not.)

Circle: grep egrep

x.*y.*z

(b) All words that contain at least two digit characters (0-9) where this is at least one
non-digit character between two of the digits. (For example: “30-30” and “2,4-d” are
words that should be selected, “11-point”, “10th”, and “RS232” are not.)

Circle: grep egrep

[0-9][^0-9].*[0-9]

(c) All words that start and end with the same letter that is not a vowel. Vowels are the
letters a, e, i, o, and u and their upper case versions (A, E, I, O, U). (Examples include
“bathtub”, “carcinogenic”, and “yummy”.)

Circle: grep egrep

^\([^aeiouAEIOU]\).*\1$

It’s possible to solve all of these problems using extended regular expressions of
course, but the answers here only use basic ones.

Sample Solution Page 2 of 6

 CSE 374 Midterm Exam 4/30/10

Question 3. (6 points) Suppose we have a file silly.txt that contains the following
lines:

eat school pizza

ride farm cow

What output is produced when we execute the following command?

sed -e 's/\(.*\) \(.*\) \(.*\)/I \1 a \3 @ \2/' silly.txt

I eat a pizza @ school

I ride a cow @ farm

Question 4. (6 points) One problem that users sometimes have with Linux is that there
is no undo command to, for example, recover a file that is accidentally deleted. As an
attempt to prevent accidental deletions, it is possible to use the –i option on rm, which
causes rm to ask the user for permission before deleting each file.

Give a shell command that creates an alias remove so that when the user enters a
command like remove a b c d, the command rm –i a b c d is executed
instead.

alias remove='rm –i'

Sample Solution Page 3 of 6

 CSE 374 Midterm Exam 4/30/10

Question 5. (16 points) People write telephone numbers in different ways, for example
(206) 555-1212, 206-555-1212, and 206.555.1212. For this problem, create a shell script
to read a file that includes such phone numbers and rewrite them in the format
aaa.bbb.cccc, like the third example given above.

The input file is formatted as follows: each line consists of a name, an address, a phone
number, and a city. The four fields are separated by colons (:), and there are no other
colons in the file (i.e., every occurrence of ‘:’ is used to separate fields). Telephone
numbers only appear in the above three formats. An example input file is:

Bill Gates:1 Infinite Loop:(425) 524-1234:Redmond

Bonnie Dunbar:Museum of Flight:206.764.5720:Seattle
Ichiro:Seattle Mariners:206-346-4000:Seattle

The shell script should have one argument giving the input file name and should use sed
to write a copy of the input file on stdout with all of the phone numbers rewritten in
the aaa.bbb.cccc format. The script should print “missing file name” and exit with a non-
zero exit status if there is not exactly one argument. Otherwise you can assume that there
are no errors – the data file exists, is properly formatted, the phone numbers are in one of
the three formats given above, etc. You may also assume that the name, address, and city
fields do not contain phone numbers.

Hint: You might find more than one editing operation and/or a pipeline to be useful.
Then again, you might not.

There are, of course, many ways to answer the problem. We gave credit to any
answer that properly edited the phone numbers as specified. An industrial-strength
version of this would need to check that the phone numbers were, in fact, in the 3rd
field of each line but we let that go.

Here’s one brute-force answer that would work. The \ at the end of each line
indicates a long line that doesn’t fit on the page, not a line break:

#!/bin/bash
if [$# -ne 1]
then
echo missing file name
exit 1
fi
sed –e 's/(\([0-9][0-9][0-9]\)) \([0-9][0-9][0-9]\)-\
\([0-9][0-9][0-9][0-9]\)/\1.\2.\3/' \
-e 's/\([0-9][0-9][0-9]\)-\([0-9][0-9][0-9]\)-\
\([0-9][0-9][0-9][0-9]\)/\1.\2.\3/' $1

Sample Solution Page 4 of 6

 CSE 374 Midterm Exam 4/30/10

Question 6. (16 points) Consider the following C program.

#include <stdio.h>

void foo(int *a, int *b) {

*a = *b;
*b = 17;
*a = *a + 1;
}

int main() {
int i, j;
int *x;
int *y;
x = &i;
y = &j;
*x = 7;
*y = 0;
printf("x = %d, y = %d\n", *x, *y);
foo(x,y);
printf("x = %d, y = %d\n", *x, *y);
return 0;
}

(a) What output does this program produce when it is executed? (It does execute
successfully.) Feel free to draw diagrams showing memory to help answer the question
and to help us award partial credit if needed.

x = 7, y = 0

x = 1, y = 17

(b) Now suppose we change function foo by rewriting its first statement as follows:

void foo(int *a, int *b) {

a = b;
*b = 17;
*a = *a + 1;
}

Now what output does the program produce when it is executed?

x = 7, y = 0

x = 7, y = 18

Sample Solution Page 5 of 6

 CSE 374 Midterm Exam 4/30/10

Question 7. (20 points) (The small C programming exercise.) A useful Unix utility is
the program head that prints the first few lines of its input file. For this problem,
complete the program below so it will read the file whose name is given on the command
line and print the first 5 lines of the file to stdout. The program can be written as a
single main function, although you may define additional functions if you wish. You
may assume that no input line has more than MAX_LINE_SIZE characters, including the
trailing ‘\0’. You do not need to check for errors – assume that the input file exists and
can be read successfully. Be sure your code works properly if the input file contains
fewer than 5 lines. Hints: FILE * fopen(char* filename, char* mode);
char * fgets(char *buffer, int max_len, FILE *stream);

#include <stdio.h>
#include <stdlib.h>

#define MAX_LINE_SIZE 1000 /* max # characters per line */

/* print the first 5 lines of the file named on the command */

int main(int argc, char** argv) {

int nlines_printed; /* # lines printed so far */
char line[MAX_LINE_SIZE]; /* current input line */
FILE * f; /* input file */
f = fopen(argv[1], "r");
nlines_printed = 0;
while (nlines_printed < 5 && fgets(line, MAX_LINE_SIZE, f)) {
fputs(line, stdout); /* printf("%s", line) also works */
nlines_printed++;
}
fclose(f); /* good practice, but no deduction if not done */

return 0;
}

Sample Solution Page 6 of 6