Student Solutions for

UNIX and Shell Programming

A Textbook
Behrouz A. Forouzan and Richard F. Gilberg
Brooks/Cole Publishing
( ISBN 0 534-95159-7)

Contents
Chapter 1: Introduction 3
Chapter 2: Basic vi Editor 5
Chapter 3: File Systems 7
Chapter 4: Security and File Permission 11
Chapter 5: Introduction to Shells 15
Chapter 6: Filters 19
Chapter 7: Communications 23
Chapter 8: vi and ex 25
Chapter 9: Regular Expressions 27
Chapter 10: grep 31
Chapter 11: sed 33
Chapter 12: awk 37
Chapter 13: Interactive Korn Shell 41
Chapter 14: Korn Shell Programming 43
Chapter 15: Korn Shell Advanced Programming 47
Chapter 16: Interactive C Shell 49
Chapter 17: C Shell Programming 51
Chapter 18: C Shell Advanced Programming 53

COPYRIGHT © 2003 the Wadsworth Group. Brooks/Cole is an imprint of the Wadsworth

Group, a division of Thomson Learning, Inc. Thomson Learning™ is a trademark used
herein under license.
2 • Chapter •

This document is provided to students using UNIX and Shell Programming A Textbook.
All reprints in any form must cite the Wadsworth Group copyright on the previous page.

Student solutions contain the answers to the odd numbered questions. Before checking
answers, be sure to download the latest errata from the Brooks/Cole web site. It’s address
is http://www.brookscole.com/compsci_d/.

Behrouz Forouzan
Richard Gilberg
 CHAPTER 1

Chapter 1: Introduction

Review Questions
1. Portable refers to the ability of the operating system to be moved to new platforms
without the need for a major rewrite. Portable operating systems are typically imple-
mented in a high level language such as C, which is available on many platforms.
3. In a time-sharing environment many users are connected to one or more computers,
and all users share output and storage devices. All computing and shared resources are
controlled by a central computer. In a client-server environment computing functions
are split between a central computer (server) and a users’ computer (client).
5. Verb (command name), options, and arguments.
7. Bourne shell, Korn shell, and C shell.
9. Logon, interaction, and logout.
11. This chapter discussed one option for date — u. It is used do display the date in GMT
time. Using the man command you may discover other options depending on the
particular UNIX implementation.
13. whoami has no options. On some systems the whoami has been obsoleted by the id
utility.
15. This chapter did not discuss any options for passwd. Using the man command you
may discover other options depending on the particular UNIX implementation.
17. This chapter did not discuss any options for man. Using the man command itself, you
may discover other options depending on the particular UNIX implementation.
19. who:
a. user name
b. terminal
c. time of login
21. date:
a. one
b. 28
c. no [@@ fix question]
d. 10: day mmm dd
e. 12: hh:mm:ss zzz

Exercises
23. cal:
a. No error.
b. Invalid option.
4 • C h a p t e r 1 • Exercises

c. cat should be cal

d. Invalid syntax.
25. echo:
a. No error.
b. No error, unless "man echo" was intended.
c. The option –u is not recognized as an option to echo.
d. No error.
27. man:
a. Should be man passwd.
b. No error.
c. No error — man accepts multiple keywords.
d. Invalid option.
29. cal month:
a. Prints calendar for August 2000.
b. Prints calendar for August 2000.
c. Error: bad month. Range 1..12.
d. Prints calendar for year 0009, not September.
31. date format:
a. Prints two digit month, space, two digit minute.
b. Prints the date in hh:mm:ss [AM | PM] hh:mm.
c. Prints the abbreviated weekday name, space, full weekday name.
d. Prints the abbreviated mont name, space, full month name.
33. man:
a. Shows man page for cal command.
b. Shows man page for calendar, which is different than cal.
c. Shows man page for date.
d. Shows man page for whoami, which is different than who.
 CHAPTER 2

Chapter 2: Basic vi Editor

Review Questions
1. An editor is a UNIX utility that can be used to create and change text files. It differs
from a word processor in that it does not perform typographical formatting such as
bolding, centering, underlining, or varying fonts within a document.
3. sed and ex
5. It is a screen editor.
7. The current character is the character at the cursor.
9. Two vi modes were discussed: command mode and text mode.
11. vi starts in command mode.
13. When moving the cursor, you are in command mode.
15. This chapter gave six commands that transition from command mode to text mode (a,
A, i, I, o, and O).
17. Frequently saving avoids loss of data in case of a system failure.
19. Yes. Start vi with the new filename as the argument.

Exercises
21. b. To delete the current character, use the x key.
23. d. All of the listed keys move from the command mode to the text mode.
25. d. All of the listed keys move from the command mode to the text mode.
27. To move from the eighth to the ninth character on a line, use the l (lower case L) key.
29. To move to the beginning of the current line, use the 0 (zero) key.
31. After all four keystrokes, the cursor would be on line 12 character 3.
33. After the two keystrokes, the cursor is on the fifth character of line 11.
35. The number of the last line on the screen is 25.
37. To discard all changes, use c (:q!).
39. To save and continue editing, use a (:w).
6 • C h a p t e r 2 • Exercises
 CHAPTER 3

Chapter 3: File Systems

Review Questions
1. Some rules for naming files:
a. Start your names with an alphabetic character.
b. Use dividers to separate parts of the name. Good dividers are the underscore,
period, and the hyphen.
c. Use an extension at the end of the file name, even though UNIX doesn’t recognize
extensions. Some applications, such as compilers, require file extensions. In addi-
tion, file extensions help you classify files so that they can be easily identified.
d. Never start a file name with a period. File names that start with a period are hidden
files in UNIX. Generally, hidden files are created and used by the system. They will
not be seen when you list your files.
3. The * matches zero or more characters, the ? matches any single character, and […]
matches a single character in the given set.
5. The […] matches a single character from the set of characters within brackets
7. The seven UNIX file types are: regular, directory, character special, block special,
symbolic link, FIFO, and socket
9. A directory is a file that contains the names and locations of files stored in a file
system. A user’s home directory is one such example.
11. A block-special file represents a physical device, such as a disk, that reads or writes
data a block at a time.
13. A FIFO, or first-in, first-out file—also known as a named pipe, is a file that is used for
inter-process communication.
15. Regular files are divided into two categories — text files and binary files. UNIX views
both formats as a collection of bytes and leaves the interpretation of the file format to
the program that processes it.
17. A directory hierarchy is an organizational structure in which directories can contain
files and other subdirectories.
19. The working or current directory is the directory that the user is in at any point in a
session. It is denoted by a dot (.).
21. The parent directory is the directory immediately above the working directory. It is
denoted by two dots (..).
23. An absolute pathname starts at the root directory. A relative pathname is the path from
the working directory to the file. An absolute pathname starts with a "/", such as
/etc/profile.
25. A relative pathname starts from the working directory.
8 • C h a p t e r 3 • Exercises

27. No. For example, if we are in the profile subdirectory of /etc, the relative pathname
would be represented as ../etc/profile. So in this case, the relative pathname is
longer and the absolute pathname, /etc/profile.
29. A directory contains a mapping of names to inodes. The relationship of file names to
inodes is a many to one relationship, that is a physical file has only one inode, but it
can have many names. For example, two filenames can both be linked to the same file.
31. The file system areas are the boot block, the super block, the inode block, and the data
block.
33. The inode fields are owner, group, type, permission, time, address.
35. A symbolic link is a logical file that defines the location of another file somewhere
else in the system. Soft links can be used for both files and directories.
37. With a basic list command, five entries (3 subdirectories and 2 files) are displayed.
With the all option, seven entries are displayed: the three subdirectories, the two files,
the current directory (.) and the parent directory (..).
39. The four operations that can be used only with files are create (vi etc), edit (vi, etc),
display (more), and print (lpr).
41. Copying a file duplicates the contents. Moving a file, changes a link to the contents.
43.
a. ls lists the contents of a directory.
b. ln links one file to another file.
c. cp copies a file.
d. pwd displays the absolute pathname of the current directory.
45.
a. rm -r deletes a file or a directory.
b. cp -r copies a directory.
c. cd changes the current working directory.
d. mkdir creates a directory.

Exercises
47. Only ‘a’ is a valid file name.
49.
a. file[1-9]
b. file[1-9]*
c. file2[0-9]
d. *.c
e. f*
f. [a-z]*
g. [a-zA-Z]*
h. [0-9]*
 C hapter 3 • E xercise s • 9

i. *[0-9][0-9]
j. ./*
k. ../*
l. [!0-9]*
51. The answers a, b, f, and g are absolute pathnames.
53. The relative pathname to a user’s home directory is the tilde (~).
55. /usr/user/john as shown in the following figure.

usr

Note: john can be a directory or a file.

john

57. The absolute pathnames are:

a. /
b. /bin
c. /usr/jack
d. /usr/jack/letter
e. /usr/jack/report
59. The relative pathnames from the john directory are:

bin dev etc usr

john jack anne

report letter

a. ../..
b. ../../bin
c. .
d. ../jack/letter
e. ../jack/report
61. The absolute pathnames are:
10 • C h a p t e r 3 • Exercises

a. /
b. /bin
c. /usr/jack/report
d. /usr/john
63. The errors in the commands are:
a. There is an extra argument (hello).
b. There is a missing argument.
c. There is an extra argument (file3).
d. There is an illegal option (-s).
e. The ordering is wrong; options should come first.
f. There is a missing argument (pathname).
g. There is no error.
 CHAPTER 4

Chapter 4: Security and File Permission

Review Questions
1. In UNIX, everyone who logs into the system is called a user.
3. Yes. For example a user can be in a group called "staff" and a group called "admin".
5. The home directory of a superuser user is root.
7. The three security levels are: system level, directory level, and file level.
9. At the system level, who accesses the system is controlled by log ins and passwords. At
the directory and file levels, security is controlled by permission codes to determine
who can access and manipulate a directory or file.
11. There are seven fields in /etc/passwd: login name, password, user-id, group-id, user
info, home directory, login shell.
13. The field lengths in the /etc/passwd file cannot be predetermined since they contain
variable length information.
15. The directory structure is shown in the following figure:

etc

/etc/passwd

17. Each permission code contains three triplets.

19. The second permissions triplet controls the group read/write/execute permissions.
21. A directory has read (r), write (w), and execute (x) permissions. Although for a
directory, execute really means permission to navigate to and through a directory.
23. Write permission for a directory allows entries to be added to and deleted from a
directory.
25. A file has read (r), write (w), and execute (x) permissions.
27. Permission w for a file allows a file’s contents to be changed.
29. The command to change a directory or file permission is chmod.
31. Yes, symbolic codes can be used to totally reset permissions.
33. Yes. When using symbolic codes, unreferenced permissions are unchanged.
35. Yes. For example
12 • C h a p t e r 4 • Exercises

chmod 666 filename

will set the read and write permissions, and clear the execute permissions for user,
group, and others.
37. No. The octal code always uses three octal digits when specifying file permissions, so
all permissions are assigned by the three octal digits.
39. To copy the contents of a directory requires read and execute permissions (o+rx).
41. To execute a program, read permission (o+r) is required for the directory and execute
permission (o+x) is required for the programs in the directory.
43. The system default permission for a new directory is 777.
45. If there is no user mask, the permissions for a newly created directory is 777. If there
is a user mask, it must be xor’d by 777 to get the new user mask.
47. The command to set the user mask is umask.
49. The command to change the ownership of a file is chown.
51. There can be multiple permission sets for a file or a directory in the sense that there
are separate read/write/execute permissions for user, group, and others.
53. To restrict file deletions to the owner requires u=rwx,g=rx,o=rx permissions.
55. It is possible to destroy a file even if it cannot be deleted. To delete a file, you must
have write permission for the containing directory. But to "destroy" a file, you only
need to be able to change the file’s contents, which is determined by a separate write
permission for the file.

Exercises
57. Symbolic format:
a. --x --x --x is a=x
b. rwx --x --x is u=rwx,g=x,o=x
c. --x rwx --x is u=x,g=rwx,o=x
d. r-x r-x rwx is u=rx,g=rx,o=rwx
59. The symbolic format is:
a. u=r,g-rwx,o-rwx
b. u=rw,o-rwx
c. u=r,g-rwx,o=x
d. u=rwx,g=r,o=x
61. The octal format is:
a. 711
b. 6?? (can’t be done)
c. 7?? (can’t be done)
d. 7?? (can’t be done)
63. The symbolic code is:
a. ug=x,o-rwx
b. a=rwx
c. uo=rx,g-rwx
 C hapter 4 • E xercises • 13

d. a=x
65. For George to copy the file, the following permissions are required: (1)
/usr/john/report should have read permission for others; (2) the root directory
should have read and execute permission for others; and /usr/john should have read
and execute permission for others.
67. The permissions for George to copy the file /usr/john/report to his own directory
are:
a. report: others: o+r
b. root: others: o+rx
c. /usr: others: o+rx
d. /usr/john: others: o+rx
e. ~george: others: u+wx
69. Permissions for Anne to copy the file /usr/anne/sales to /usr/john are:
a. sales: users: u+r
b. /: others: o+rx
c. /usr: others: o+rx
d. /usr/john: others: o+wx
e. /usr/anne: users: u+rx
71. The file has the following permissions: user—execute, group—execute, others:—
execute.
73. The file has the following permissions: rw- rwx r--
75. The default permissions after umask 022 are:
a. rw- r-- r--
b. rwx r-x r-x
14 • C h a p t e r 4 • Exercises
 CHAPTER 5

Chapter 5: Introduction to Shells

Review Questions
1. The first is the interpreter. The interpreter reads user commands and works with the
kernel to execute them. The second part of the shell is programming capability that
allows you to write a shell (command) script.
3. The login shell can be verified by entering echo $SHELL at the command line.
5. No, the login shell is not always the same as the current shell. The user can change
shells.
7. Yes. Shells can be nested.
9. The file descriptor for standard input is 0, standard output is 1, and standard error is 2.
11. The standard output stream is normally associated with the monitor
13. Redirection is the process by which a user specifies that a file is to be used in place of
one of the standard streams.
15. To redirect standard output use 1> filename.
17. Pipe is an operator that temporarily saves the output of one command in a buffer that is
being used at the same time as the input of the next command. The pipe token is the
vertical bar ‘|’.
19. A sequence of commands can be entered on one line separated by semicolons (;).
Commands can be grouped together using parentheses. Commands can be chained
together using pipes.
21. Command line editing is a shell facility that permits the user to edit commands using
one of the screen editors (vi or emacs).
23. The command line editors are normally vi or emacs.
25. The backslash metacharacter (\) changes the interpretation of the character that
follows. Double quotes remove the special interpretation of most metacharacters.
Single quotes operate like double quotes, but their effect is stronger. They preserve
only the meaning of single quotes.
27. A job is a command or a set of command entered on one line. A foreground job is any
job running under the supervision of the user. A background job is any job running on
the background without the direct supervision of the user.
29. The six different job states are: (1) Running in foreground. (2) Stopped (suspended).
(3) Running in background. (4) Done (completed successfully). (5) Exit (completed
unsuccessfully). (6) Terminated (killed).
31. An alias provides a means of creating customized commands by assigning a name to a
command. Aliases are available in the Bash, Korn, and C Shells. To use an alias to
rename files in Bash and Korn, use alias rename=mv. In the C Shell’s, the syntax is
slightly different, alias rename mv, or to use the C Shell positional argument
substitution, the user could write alias rename 'mv \!:1 \!:2'.
16 • C h a p t e r 5 • Exercises

33. A user can store a value in a user variable, for example msg="Hello world" (or for C
Shell: set msg = "Hello world"), and recall the variable’s value, as in echo $msg.
A predefined variable is similar in concept, but is already set up for the user. For
example the primary prompt (PS1 in Korn and Bash Shells, or prompt for C Shell) is
a predefined variable.
35. Options are on/off switches that are used to control the shell. Three such options are
noclobber—do not allow redirection to clobber existing files, ignoreeof—do not
allow ctrl+d to exit the shell, and verbose—print commands before executing them.
37. The startup files for the Korn shell are the system wide /etc/profile file maintained
by the system administrator, and the local .profile file, an optional file maintain by the
user. In addition, the Korn shell allows for an environment file to run whenever a new
shell is started. The file name is determined by the exported shell variable ENV. The
Bash Shell is similar, except that the user .profile file is searched for under several
names in the following order: ~/bash_profile, ~/bash_login, and finally
~/.profile. The C shell is similar, but the file names are different. The system wide
login file is /etc/csh.login, and the system wide environment file (for all new
subshells) is /etc/csh.cshrc. The personal login file is ~/.login, and the personal
environment file is ~/.chsrc.

Exercises
39. b: echo $SHELL
41. b: bash
43. c: csh
45. b: 1
47. None of these commands read from standard input.
49. Yes. The lpr command accepts standard input.
51. lpr input redirection:
a. lpr prints from standard input (keyboard).
b. lpr prints file1.
c. lpr prints file1.
d. lpr prints file1.
53. input redirection errors:
a. cp requires filename arguments.
b. ls should be followed by zero or more directory or file names.
c. mkdir should be followed by a new directory name.
d. rm should be followed by a file name.
55. Commands that don’t take input redirection:
a. lpr can be used with redirection (see man lpr), but it is generally used with file-
name arguments.
b. ln is used to link files. Using it with redirection makes no sense.
 C hapter 5 • E xercises • 17

c. cd is used to change directories. It does not operate on standard input.

d. more browses through a text file. Since it uses terminal input to control pag-
ing/navigation, it cannot also take input from the terminal. However, it can still
use input redirection from a file, as in more < someFile.
57. Commands that can be used with output redirection:
a. a) cal can be used with output redirection.
b. b) date can be used with output redirection.
c. c) echo can be used with output redirection.
d. d) man can be used with output redirection.
e. e) passwd cannot be used with output redirection.
f. who can be used with output redirection.
g. whoami can be used with output redirection.
59. cal command usage:
a. cal results (calendar of current month) are sent to standard output.
b. Sends cal results (calendar of current month) to Outfile
c. Sends cal 1 results (calendar of year 1) to Outfile. Because there is a space
between the ‘1’ and the ‘>’ this is not the same as redirecting standard output.
d. Sends cal 2 results (calendar of year 2) to Outfile. Because there is a space
between the ‘2’ and the ‘>’ this is not the same as redirecting standard error.
61. date command usage:
a. Sends date output to file1 (unless noclobber option is enabled).
b. Sends date output to file1, overriding noclobber setting.
c. Appends date output to file1.
d. syntax error.
63. Command errors:
a. No error.
b. more and date are swapped.
c. more cannot take input from the terminal.
d. man requires an argument and it does not make sense to pipe to man.
65. Chained commands:
a. Command1 writes to standard output.
b. Command2 reads from standard input.
67. The command date | more works. The command more | date does not work
because more needs to have input and date does not read from standard input.
69. In the following command: ls -l | tee:
a. Input comes from the system.
b. No output file is created.
71. Yes, the tee command can be used to create a file. The statement tee file1 sends
keyboard input to a file1.
18 • C h a p t e r 5 • Exercises

73. In the command ls -l | tee > file1:

a. The output will go to both standard output (monitor) and to the file file1.
b. One file is created.
75. C shell errors:
a. C requires the set command: set a = 24.
b. OK.
c. OK.
d. C shell does not allow the ‘$’ in target variable of set command.
77. echo command output: The only differences are in the quoting of UNIX.
a. Hello to the user of UNIX operating system
b. Hello to the user of "UNIX" operating system
c. Hello to the user of 'UNIX' operating system
d. Hello to the user of "UNIX" operating system
79. Command errors:
a. There should not be a ‘$’ prefix on $a and cp does not generate output.
b. lpr does not write to standard out.
c. No ‘$’ prefix on $a.
d. The cal command’s formatting gets removed. To preserve the formatting, use echo
"`cal`".
81. Command errors:
a. No ‘$’ prefix for $a. The output is sent to a file, so it is not available for the vari-
able assignment.
b. There is nothing to echo because the cal output is sent to lpr.
c. No ‘$’ prefix for $a and file1 is not a command.
d. The cal command’s formatting gets removed. Use quotes to preserve it.
 CHAPTER 6

Chapter 6: Filters

Review Questions
1. In UNIX, a filter is any command that gets its input from the standard input stream,
manipulates the input, and then sends the result to the standard output stream.
3. Filters discussed in this chapter: cat, cmp, comm, cut, diff, head, paste, sort, tail, tr,
uniq, and wc. Other filters mentioned in this chapter: more, grep, sed, and awk.
5. head [-options] [file-list]—The head command outputs lines at the beginning
of a file. Input can come from the keyboard, a file, or several files. Output can go to the
monitor or be redirected to a file. The head command does not output to multiple files.
7. cut [-options] [file-list]—The cut command select portions of each line of a
file. Input can come from the keyboard, a file, or several files. Output can go to the
monitor or be redirected to a file. The cut command does not output to multiple files.
9. sort [-options] [field-specifiers] [file-list]—The sort command arranges
data in ascending or descending order. Input can come from the keyboard, a file, or
several files. Output can go to the monitor or be redirected to a file, or a specified
output file that can even be the same as one of the input files. The sort command does
not output to multiple files.
11. uniq [-options] [file-list]—The uniq command displays unique lines in a file
(adjacent repeated lines are treated as one line). Input can come from the keyboard, a
file, or input redirection. Output can go to the monitor or be redirected to a file, or can
go to an output file specified in a command line argument. The uniq command does
not output to multiple files.
13. The paste and cat commands both combine files. The cat command concatenates files
vertically, the paste command concatenates files horizontally.
15. The cmp command compares two files for equality and if they do not have identical
contents, reports where the two files start to differ. The comm command compares two
files and optionally shows the lines that are only in the first, lines that are only in the
second, and lines that are common to both.
17. The comm command can be used to show lines that are common to two files, whereas
the diff command can be used to show the differences between two files.

Exercises
19. a & c display the contents of a file; b & d create files.
21. The input sources are:
a. file1
b. keyboard
c. keyboard
d. file1
20 • C h a p t e r 6 • Exercises

23. Equivalent commands are:

a. equivalent
b. not equivalent. The first outputs to file1, the second outputs from file1.
c. If noclobber is not set, then the two are equivalent. Otherwise, the first command
will fail but the second overrides noclobber and will succeed.
d. Not equivalent. The first reads from the keyboard, the second reads from file1.
25. Command errors:
a. date does not read from standard input.
b. OK
c. Can’t pipe to a file.
d. Can’t pipe to a file.
27. The output of echo is piped to cat file1 > file2. Since cat is reading from file1, it
ignores the output from echo, and thus file1 is copied to file2.
29. In the first command, “Header” is written to standard output, then the contents of file1
are written to file2. In the second command, “Header” followed by file1 are written to
file2. In the third command the contents of file1 is written to file2, and nothing goes to
standard output.
31. d copies file 1 to file 2.
33. cat <file1 >file2
35. Not possible.
tail -r file1 | tail -50r > file2
37. Use tail to copy last 50 lines: tail +50 file1 > file2
39. Copy line one: head -1 file1 > file2.
41. Copy and append lines from two files:
head -50 file1 | tail -21 > file3; head -60 file2 | tail -41 >> file3
43. Differences in commands:
a. The first command counts characters, words, and lines from file1 and annotates
the output with the filename "file1." The second counts characters, words, and
lines from standard input, which was redirected from file1, so it provides the same
count results without the filename annotation.
b. The first counts characters, words, and lines from file1 then from file2. The second
processes the files in the reverse order.
c. The first counts characters, words, and lines from file2. The second processes the
files file1 and file2.
45. Count characters in first line of a file: head -1 file | wc -c.
47. Find errors:
a. Illegal option: (-l).
b. OK.
c. The c option supersedes the u option. In some implementations, this is an illegal
usage.
d. Missing argument to -s option.
 C hapter 6 • E xercises • 21

49. Change lowercase to uppercase: tr "[a-z]" "[A-Z]" < filename.

51. The command works. The tr command will map each input character found in the first
string "AB" to each corresponding character in the second string "BA".
53. Yes, we can change uppercase and lowercase in one command. The command is:
tr "[a-z][A-Z]" "[A-Z][a-z]"
55. The command will squeeze multiple sequential A and B characters into the character B.
It will also map original X characters to B. For example if the input were ABAABB-B-A-
X, the output would be B-B-B-B.
57. To create a file of 10 lines, enter the command tail +1 > filename and enter at least
10 lines from standard input.
59. Enter the command paste - > filename and enter as many lines as you wish from
standard input. The option - means read from standard input instead of a file.
61. Yes, the sort can be used to copy a file without actually sorting by using the -m sort
option: sort -m file1 > file2. The -m option tells sort that the input is assumed to
be sorted (whether or not it really is), so it can be used in a merge only mode to
prevent sorting.
63. The command head -5 file1 | head -3 | head -2 demonstrates that head is a
filter because it is used on the left of a pipe, between two pipes, and on the right of a
pipe.
65. The command cut -f1,3 file1 | cut -f1,2 | cut -f1 demonstrates that cut is
a filter because it is used on the left of a pipe, between two pipes, and on the right of a
pipe.
67. The command tr "l" "L" file1 | tr "i" "I" | tr "n" "N" demonstrates that
tr is a filter because it is used on the left of a pipe, between two pipes, and on the right
of a pipe.
69. The command wc -l file1 | wc -l | wc -l demonstrates that wc is a filter
because it is used on the left of a pipe, between two pipes, and on the right of a pipe.
The result is always one since the output of the wc –l is a single line, and the final wc
–l provides a count of the one line, which is 1.
71. None of the commands are correct.
73. None of the commands are correct.
75. None of the commands are correct.
22 • C h a p t e r 6 • Exercises
 CHAPTER 7

Chapter 7: Communications

Review Questions
1. The five communications utilities are: talk, write, mail, telnet, and ftp.
3. Yes, talk can be used by users on different systems.
5. To prevent talk sessions, enter the command mesg n.
7. The talk command is a two-way communication utility.
9. No. The write command can only be used to communicate with users on the same
system.
11. The write command is a one-way communication utility.
13. The mail utility is used to send and receive electronic mail. It is a store and forward
based system, which means that the receiver does not need to be logged in at the time
the message arrives.
15. The body contains the text of the message. The header consists of the subject, the
addressees (To:), sender (From:), a list of open copy recipients (Cc), and a list of blind
copy recipients (Bcc).
17. A mail address is made up of two parts: local address and domain name separated by a
n at-sign (@) character.
19. The domain address uniquely defines the destination mail server anywhere in the
world.
21. The original seven generic domain labels are: com, edu, gov, int, mil, net, and org.
The labels approved in November 2000 are: aero, biz, coop, info, museum, name, and
pro.
23. Yes. To send mail, the mail utility needs the username(s) of the message recipient(s).
25. In mail, when we are in the read mode, we can type a command without using a tilde
(~) character.
27. When composing a document in vi text mode, the escape character is used to toggle to
command mode. When composing a message mail, a line starting with tilde (~) is used
to indicate a command. In vi, there are a number of commands (such as insert and
append) that transition from command to text mode. In mail, a reply (r or R) command,
or a new message command (m) can be used to transition from read mode to send
mode. The send mode in mail is a very primitive editor; it is not a general purpose text
editor.
29. In a distribution list, an alias is formatted as: alias <alias-name> <alias-list>.
31. To forward mail, we need to use the m command to move from the read mode to the
send mode. While in the send mode, we can use ~m to retrieve received mail and send
it. Of course, when we send it we can add comments.
33. telnet is a general-purpose client/server utility because it lets a user access any
application program on a remote computer; in other words, it allows the user to log into
24 • C h a p t e r 7 • Exercises

a remote computer. After logging on, a user can use the services available on the
remote computer and transfer the results back to the local computer.
35. A terminal driver is the interface between a terminal and the operating system. In the
case of a remote terminal (e.g. telnet client), no physical terminal (or terminal driver)
is used. Since there is no terminal driver, a faux terminal is needed for applications to
communicate with. This is done using a pseudoterminal driver.
37. Synopsis: ftp [-option] [host]—ftp (File Transfer Protocol) allows the user to
transfer files between the local and remote computers.
39. The ftp utility uses two connections: control connection and data connection

Exercises
41. The mail utility does not use a prompt while composing a mail message, but you may
be prompted for header items such as subject.
43. To read message 7, key either 7 or t 7.
45. To list the headers for all messages in the read mode, key h.
47. To reply to only the sender, use R.
49. To forward a message, use m to send a new message or ~m to read in a message to
forward.
51. To include a file in a message key ~r <filename>.
53. To save a file before it is sent, key ~w <filename>.
55. No, you can use vi as your e-mail editor when connected remotely using ftp.
 CHAPTER 8

Chapter 8: vi and ex

Review Questions
1. There are three operating modes in vi: command, text, and ex.
3. Yes. Enter ex from the command line.
5. Yes. Enter the command vi from the colon (:) prompt.
7. To directly exit from ex, key wq, q, or q!.
9. In command mode, a powerful set of vi commands are available to move and make
changes.
11. A number of vi commands—such as insert (i and I), append (a and A), and new line (o
and O)—enter text mode.
13. Yes. The colon (:) key moves from the command mode to the ex mode..
15. No. To move from the vi mode to the ex mode, you must pass through the vi command
mode.
17. There is no text mode prompt.
19. Yes. The ex mode uses a colon (:) prompt.
21. Local vi commands are applied to the text at the current cursor location.
23. Global commands are applied to all of the text in the whole buffer, as contrasted to the
current window.
25. A range command is applied on a text object.
27. There are 17 local vi commands described: i, I, a, A, o, O, r, R, s, S, x, X, p, P, m, ~, and
J.
29. Four range commands were described: "nothing" (moves the cursor), d (delete), c
(change), and y (yank).
31. Fifteen global commands were described: ctrl+y, ctrl+e, ctrl+u, ctrl+d, ctrl+b,
ctrl+f, u, U, z<return>, z., z-, ctrl+L, . (dot), ctrl+G, and ZZ.
33. A single-line address in ex defines one line for processing.
35. A set-of-line address in ex defines one or more lines that are not necessarily
consecutive. The lines are chosen, not based on their relative positions, but based on
their contents.
37. A range address in ex defines one or more consecutive lines. A special range address is
the percent sign (%), which defines the entire file, or a range address consisting of two
addresses separated by a comma or a semicolon.
39. A nested address defines a set-of-line address inside a range.
41. There are 26 alphabetic named buffers that are controlled by the user. They are
identified by the lowercase letters a through z.
26 • C h a p t e r 8 • Exercises

Exercises
43. The number of lines deleted are:
a. dd: 1
b. 4dd: 4
c. 4dd2: 4
d. dd5: 1
45. Only (b), the r command, needs a text object.
47. The ex address types are:
a. 5: single-line address
b. 20,30: range address
c. $: single-line address
d. /pattern/: set-of-lines address
49. The ex address types are:
a. 20,60g/pattern/: nested address
b. ?pattern?;?pattern?: range address
c. /pattern/,?pattern?: range address
d. 45,80v?pattern?: nested address
51. The ex that defines a block containing lines 30 to 50: 30,50
53. The difference between the two ex commands is:
a. deletes all lines with UNIX
b. deletes the next line with UNIX
55. The difference between the two ex commands is:
a. copies lines 20-50 after line 1
b. copies lines 20-50 before line 1
57. The lines deleted are:
a. 1—line 20
b. 11—lines 20 through 30
c. 11—lines 20 through 30
d. 31—lines 20 through 51
 CHAPTER 9

Chapter 9: Regular Expressions

Review Questions
1. A regular expression is a character pattern that is matched against text. A regular
expression is like a mathematical expression. A mathematical expression is made of
operands (data) and operators. Similarly, a regular expression is made of atoms and
operators.
3. An atom specifies what we are looking for and where in the text the match is to be
made. Atoms corresponds to operands in a mathematical expression.
5. A single character atom matches itself. Its length is one.
7. A class atom defines a set of characters, any one of which may match the
corresponding character in text. Its length is one.
9. There are four types of anchors: ^, $, \<, and \>.
11. Operators are used to combine atoms into more complex regular expressions. The
regular expression operators play the same role as mathematical operators.
Mathematical expression operators combine mathematical operands; regular
expression operators combine regular expression atoms.
13. The sequence operator concatenates a series of atoms. It is conceptual. It has no
symbolic representation.
15. The repetition operator specifies that the atoms or expressions may be repeated. Its
symbolic formats are as follows: \{m,n\}—match previous atom m to n times;
\{m\}—match previous atom exactly m times; \{m,\}—match previous atom at least
m times; \{,n\}—matches previous atom at most n times. There are also several short
forms for the repetition operator.
17. The save operator saves one or more characters in a buffer to be matched later with its
corresponding back-reference atom. Its symbolic format is \(…\).
19. The ? operator’s format is: \{0,1\}
21. The + operator’s format is: \{1,\}

Exercises
23. The number of characters matched is:
a. 1
b. 1
c. zero or more
d. 3
25. The minimum length to be matched is:
28 • C h a p t e r 9 • Exercises

a. 3
b. 0
c. 3
d. 0
27. The number of characters matched is:
a. 4
b. 4-7
c. 0-7
d. 8 or more
29. The expressions can be rewritten as:
a. A\{5\}
b. A\{2,4\}
c. A\{2\}| A\{5\}| A\{8\}
d. A
31. The expressions can be rewritten as:
a. Can’t rewrite using +
b. A+
c. Can’t rewrite using +
d. Can’t rewrite using +
33. [ABC] matches all but the last one. The matched characters are underlined.
A
AB
ADBC
35. [A-G] matches all lines. The matched characters are underlined.
A
AB
MBC
D
37. [ABC][AB] matches only one line. The matched characters are underlined.
AB
39. ^[ABC][AB]$ matches only one line. The matched characters are underlined.
AB
41. bc.* matches three lines. The matched characters are underlined.
aaabbbcccddd
aaaabcsssss
aaaaabc
43. ^[a-z]... matches two lines. The matched characters are underlined.
abcdefg
a:237efg
45. s:?s* matches three lines. The matched characters are underlined.
efgs:sgfe
ssssssssss
 C hapter 9 • E xercises • 29

rsts
47. [^\$]$ matches two lines. The matched characters are underlined.
$$$$$$aaaaaa
abc
49. ^[^\$]$ does not match any lines.
51. ^\$[0-9][0-9]$ matches one line. The matched characters are underlined.
$10
53. Match blank line: ^$
55. Matches every line: .*
57. Matches a line with at least three characters .\{3,\}
59. Matches date (Month dd, yy): [A-Z][a-z]+ [0-9][0-9], [0-9][0-9]
61. Matches Social Security Number: [0-9]\{3\}-[0-9]\{2\}-[0-9]\{4\}
63. Matches any digit: [0-9]
65. Matches Zip Code (5 digits): [0-9]\{5\}
67. Matches HTML tags: <.+>
30 • C h a p t e r 9 • Exercises
 CHAPTER 10

Chapter 10: grep

Questions
1. The term grep is an acronym for Global Regular Expression Print; egrep is Extended
grep; and, fgrep is Fast grep.
3. There are three members of the grep family: grep, egrep, and fgrep.
5. fgrep cannot accept regular expressions.
7. The atoms supported by grep varies depending on the grep implementation you are
using. Some older implementations may not support the begin/end word anchors \<
and \>.
9. No. Although the degree of operator support could vary based on the particular
implementation of grep/egrep.
11. Not all egrep implementations support the save operator. Some implementations
support save by overloading the group syntax, so (…) provides the functionality of
\(…\).

Exercises
13. Select all lines with exactly three characters: grep "^...$" file1
15. Select lines that have three or less characters: grep "^.\{0,3\}$" file1
17. Count blank lines: grep -c "^$" file1
19. Select lines starting with any uppercase character: grep "^[A-Z]" file1
21. Select lines containing UNIX or variation: egrep "UNIX|Unix|unix" file1
23. Select lines ending with UNIX: grep "UNIX$" file1
25. Select lines with UNIX twice: grep "UNIX.*UNIX" file1
27. Write command copy a file: grep ".*" file1 > file2
29. Select lines that start with space: egrep "^ +" file1
31. Select lines that start and end with space: egrep "^ .* $" file1
33. Select lines that end with an uppercase character followed by zero or more spaces:
grep "[A-Z] *$" file1
35. Select lines with two consecutive digits: grep "[0-9]\{2,\}" file1
37. Select lines in which first nonblank character is uppercase:
grep "^[ ]*A" file1
39. Select lines that do not start with A: grep "^[^A]" file1
41. Select lines that start with A, B, or C: grep "^[ABC]" file1
43. Select lines that start with a character: grep "^[a-zA-Z]" file1
45. Select lines that contain only one floating point number:
egrep "^[0-9]+[.][0-9]*$" file1
32 • C h a p t e r 10 • Exercises

47. Select lines that contain only one hex number:

egrep "^0[xX][0-9A-Fa-f]+$" file1
49. Copy only lines with at least four digits:
grep "[0-9].*[0-9].*[0-9].*[0-9]" file1 > file2
51. Copy palindromes: grep "^\(.\)\(.\).\2\1" file1 > file2
53. Simulate commands:
a. cat: grep ".*" > file1
b. cut: Not possible.
c. paste: Not possible.
d. It is not possible because if we use grep ".*" file1 file2, each output line
will be preceded by the file name.
 CHAPTER 11

Chapter 11: sed

Questions
1. sed [options] script [file-list]—sed edits specified lines in the input files and process
them.
3. The sed command allows for an "in-line script," in which the script is specified on the
command line and also a file script, in which the script is stored in a file.
5. A script file is a text file that contains a sed script.
7. A script file can be created with a standard editor such as vi.
9. The hold space is a buffer used to temporarily store and retrieve lines as directed by
sed instructions.
11. The address selects the line(s) to be processed (or not processed) by the command.
13. A single-line address specifies one and only one line in the input file. It is either a line
number or the $ (last line) symbol.
15. The single-line addresses are: a & d.
17. The range addresses are: b & d.
19. The two scripts are equivalent.
21. The addresses and commands are:
a. address: 25; command: d
b. no address; command: d
c. address: 20,50; command: s/A/B/g
d. address: 25,/^A.*A$/; command: s/A//g
23. If no address is mentioned, the command applies to every line (a).
25. No, the sed command does not change the file.
27. The N command appends the next line to the pattern space while the D command
deletes the first line of the pattern space.
29. The p command prints the entire pattern space; The P command prints only the first
line of the pattern space.
31. The h command copies the pattern space to the hold space whereas the H command
appends the pattern space to the hold space.
33. The command comparisons are shown in the following table.
Command ex sed
Line number # =
Insert i i
Append a a
Change c c
Delete d d
Substitute s s
34 • C h a p t e r 11 • Exercises

Command ex sed
Translate n/a y
Input next line n/a (no pattern space) n, N
Print p p, P
List l l
Read r r
Write w w
Label n/a :label
Branch n/a b, T

Exercises
35. The result of the command: sed "s/bc*./Z/"
aaaZbcccddd
aaaaZssss
aaaaaZ
aaZss
37. The result of the command: sed "s/\..*\$/Z/"
rsZ
abcZ
abcZ
abc
39. The result of the command: sed "s/:?.?/Z/"
:a?????????
eeeeeefffff?hhhh
aaa:Z???
:?.\?
41. The result of the command: sed "s/^[^$]*$/Z/"
$$$$$$aaaaaa
bcdef$
$
Z
43. The result of the command: sed "s/^\$\d\d$/Z/"
$$$$$$$
$10
abc$
$
45. This command is effectively a no-op.
47. Deletes all blank lines.
49. Prints the first 80 lines.
51. Each line is repeated three times.
53. Removes every third line (printing lines 1, 2, 4, 5, 7, 8, 10, …).
55. Moves lines 10-15 after line 20 while removing lines 11-15.
57. Delete second character: sed "s/^\(.\)./\1/" file1
 C hapter 11 • E xercises • 35

59. Delete character before last character: sed "s/.\(.\)$/\1/" file1

61. Delete second word: sed "s/^\([^ ]* *\)[^ ]* */\1/" file1
63. Delete word before last word: sed "s/ *[^ ]*\( *[^ ]*\)$/\1/" file1
65. Swap first and last characters:
sed "s/^\(.\)\(.*\)\(.\)$/\3\2\1/" file1
67. Delete all digits: sed "s/[0-9]//g" file1
69. Replace single spaces at beginning of line with tabs:
sed "s/^ \([^ ]*\)/ \1/" file1
To only replace a lines starting with single space we find the start of line followed
by one space and zero or more non-space characters. Just finding start of line followed
by space could match lines that start with two or more spaces.)
71. Print every other line: p;p;p;d
This command works whether or not the -n option is used.
73. Copy two lines and delete the third: n;n;d
75. Move lines 22-33 after line 56: 22h;23,33H;22,33d;56G
77. Delete all trailing spaces: sed "s/ *$//" file1
79. Change date mm/dd/yy to yy/mm/dd:
sed "s/\([0-9][0-9]\)\/\([0-9][0-9]\)\/\([0-9][0-
9]\)/\3\/\1\/\2/g" file1
81. Extract month from mm/dd/yy:
sed -n "s/[^0-9]*\([0-9][0-9]\)\/\([0-9][0-9]\)\/\([0-9][0-
9]\).*/\1/p" file1
83. Extract year from mm/dd/yy:
sed -n "s/[^0-9]*\([0-9][0-9]\)\/\([0-9][0-9]\)\/\([0-9][0-
9]\).*/\3/p" file1
85. Copy lines containing at least five digits:
sed -n "/\([0-9].*\)\{5\}/p" file1 > file2
87. Copy file deleting second word:
sed "s/\([^ ]\{1,\} *\)\([^ ]\{1,\} *\)/\1/" file1 > file2
89. Simulate copy command: sed -n "p" file1 > file2
91. Simulate head command: sed "20q" file1
93. Simulate tail command:
sed -n '1!G;h;$p' file1 | sed -n "1,40p" | sed -n '1!G;h;$p'
This is equivalent to tail -40 file1! Analysis: Since we don’t know how many
lines are in the file, we also don’t know the line number that is 40 before the end.
Therefore sed is not suited to this task. But by reversing the file using sed’s hold buffer
commands, we can then extract the first 40 lines (which were the last 40 lines) and
then reverse the lines again. So there is a tricky solution to this problem, but it is not
very practical.
36 • C h a p t e r 11 • Exercises
 CHAPTER 12

Chapter 12: awk

Questions
1. The awk utility, takes its name from the initials of its authors (Alfred V. Aho, Peter J.
Weinberger, and Brian W. Kernighan).
3. The awk command supports the regular expressions supported by egrep. It does not
support word anchors or back reference atoms.
5. The awk command can use an inline script specified on the command line or a file
based script specified by the -f option.
7. A script file is a file that contains a script. This is useful for longer scripts that are too
long to reasonably put on a command line.
9. Script files can be created using a standard text editor such as vi.
11. Patterns identifies which records in the file are to receive an action.
13. A simple pattern matches one record.
15. There are four types of simple patterns: BEGIN, END, expression, and nothing (no
expression).
17. END is true after the last line is read and false everywhere else.
19. An empty expression (no expression) applies to every line.
21. Patterns a, b, and d are valid.
23. Answers b and c are true; answers a and d are false.
25. The pattern NR == 8 || NR==9 selects lines 8 and 9.
27. The pattern $0 ~/UNIX/ || $0 ~/DOS/ selects lines containing UNIX or DOS.
29. No, awk does not modify the input file..
31. next and getline both read the next input line, but getline continues execution,
whereas next terminates the processing for the current record and begins processing of
the next one.
33. Actions in awk are instructions or statements that act when the pattern is true.
35. print provides simple unformatted output. printf is a formatted print, and sprintf
is a formatted print that will print to a string instead of standard output.
37. Yes, the for loop is a pre-test loop.
39. No, the do…while loop is not a pre-test loop.
41. The built-in string functions are: length, index, substring, split, substitute, global
substitute, toupper, and tolower.

Exercises
43. Simulate sed addresses:
a. 5: NR==5
b. n/a
38 • C h a p t e r 12 • Exercises

c. 5!: !(NR==5)
d. n/a
45. Simulate sed addresses:
a. 5,9: NR==5,NR==9
b. 5,9!: NR<5 || NR>9
c. /UNIX/,/DOS/: /UNIX/,/DOS/
d. /UNIX/,10: /UNIX/,NR==10
47. Simulate sed script:
!/^A/;/^A/ { print $0; print " This is the test to"; print "be
added" }
49. Simulate sed script:
!/^A/
51. Simulate sed script:
# simulate sed script /^A/,/B$/y/aeio/AEIO/
/^A/,/B$/ {
gsub(/a/,"A");
gsub(/e/,"E");
gsub(/i/,"I");
gsub(/o/,"O");
}
{ print $0 }
53. Simulate sed script:
{print}; NR==20,NR==60 {print $0 > "file2"}
55. Simulate sed script:
{print}; /UNIX/ {exit}
57. Simulate cp command: awk '{print > "file2"}' file1
59. Simulate head command: awk '{ print }; NR==20 { exit }' file1
61. Simulate tail command:
awk '{ lines[NR]=$0 }; END { for (i=40; i>0; i--) print lines[NR-
i+1] }' file1
63. Print eighth line: NR==8 {print; exit}
65. Print total input lines: END {print NR}
67. Print last field in each line: {print $NF}
69. Print lines with more that four fields: NF>4 {print $0}
71. Print total number of fields in file:
{total_fields += NF}; END {print total_fields}
73. Print lines that contain UNIX: /UNIX/ {print $0}
75. Print all the lines with field order field 4, field 3, field 2, and field 1.
{print $4, $3, $2, $1}
77. Print sum of all values in the third field: {sum3 += $3}; END {print sum3}
79. Print lines in which third field value is greater than 5.00 and fourth field value is not 0:
$3>5.00 && $4!=0 {print $0}
81. Add one blank line after each line: {print $0, "\n"}
 C hapter 12 • E xercises • 39

83. Print any line that follows a line whose first field is greater than 9:
{ if (printNext) print $0; printNext = ($1 > 9) }
85. Print each line three times: { print $0; print $0; print $0 }
87. Copy two lines and delete the third:
{ print $0; getline; print $0; getline }
89. Copy lines 22 to 23 after line 56:
{
if (NR>=22 && NR<=23)
held[NR]=$0
print $0
if (NR==56)
{
print held[22]
print held[23]
}
}
91. Prints the first and the third word in each line: {print $1, $3}
93. Evaluate expressions to true or false:
a. !(3 + 3 >= 6): false
b. 1 + 6 == 7 || 3 + 2 == 1: true
c. 1 > 5 || 6 < 50 && 2 < 5: true
d. 14 != 55 && !(13 < 29) || 31 > 52: false
95. Write an if statement that assigns 1 to best if score is 90 or greater:
if (score >= 90) best = 1
97. Find the smallest of ten integers:
#Find the smallest in an array
function smallestN(array, size)
{
smallest = array[0]
for (i = 1; i < size; i++)
if (array[i] < smallest) smallest = array[i]
return smallest
}

#find the smallest of 10 integers

function smallest10(n0, n1, n2, n3, n4, n5, n6, n7, n8, n9)
{
val[0]=n0; val[1]=n1; val[2]=n2; val[3]=n3; val[4]=n4
val[5]=n5; val[6]=n6; val[7]=n7; val[8]=n8; val[9]=n9
return smallestN(val, 10)
}
40 • C h a p t e r 12 • Exercises
 CHAPTER 13

Chapter 13: Interactive Korn Shell

Questions
1. The Korn shell can be used either interactively to receive user commands and interpret
them, or as a programming language to create shell scripts.
3. No. You can change to a different subshell and therefore the current shell will not
necessarily be the same as the login shell.
5. To create a child shell, enter the command name for the shell (such as ksh or csh).
Then use exit to return to the parent shell.
7. The standard stream file descriptors are: standard input, 0; standard output, 1; and
standard error, 2.
9. The standard output stream is normally associated with the monitor.
11. To redirect the standard input stream to a file, use input redirection.
command 0< filename
13. To redirect the standard input stream to a file, use error redirection.
command 2> filename
Also redirection override (1>|) and redirection append (1>>) may be used.
15. A sequence of commands can be entered on one line separated by semicolons (;).
Commands can be grouped together using parentheses. Commands can be chained
together using pipes.
17. Quotes are used to change the predefined meanings of characters. Three types of
quotes are used for this purpose: backslash (\), a pair of double quotes ("…"), and a pair
of single quotes ('…').
19. Command substitution provides the capability to convert the result of a command to a
string. The command substitution operator that converts the output of a command to a
string is a dollar sign and a set of parentheses: $(command). Another syntax,
`command`, is also supported, but may be deprecated in the future.
21. To suspend a foreground job, use ctrl-z. The bg command moves a suspended job to
the background. The fg command moves a suspended or background job to the
foreground.
23. A variable is a location in memory where values can be stored.
25. The shell variables are used to configure the shell. The environmental variables control
the user environment and can be exported to subshells.
27. When a command is executed, it returns a value known as the exit status of the
command. The exit status of a command is stored in a shell variable called (?).
29. The environmental variable holds name of the login shell is: SHELL. It is displayed with
print $SHELL.
31. The environmental variable holds the primary prompt is PS1. It is displayed with print
$PS1.
42 • C h a p t e r 13 • Exercises

33. The environmental variable holds the path of your home directory is HOME. It is
displayed with print $HOME.
35. The environmental variable holds the path of your mailbox is MAIL. It is displayed
with print $MAIL.
37. The environmental variable holds your login name is LOGNAME. It is displayed with
print $LOGNAME.
39. Options are on/off switches that are used to control the shell. Three such options are
noclobber—do not allow redirection to clobber existing files, ignoreeof—disallow
ctrl+d to exit the shell, and verbose—print commands before executing them.

Exercises
41. The command that shows your login shell is print $SHELL (a).
43. The descriptor of the input standard stream is 0 (a). The descriptor of the output
stream is 1 (b). The descriptor of the error stream is 2 (c).
45. The valid Korn shell variables are cat (a), _first (b), and 2Var (d).
47. The output is:
a. a
44
b. <a blank line>
a44
c. 44a
49. Use a grep command and check its exit status.
grep root /etc/passwd > /dev/null; print $?
This command should always be successful, so the exit status is zero. If we substitute
"no_user" (or some other bogus username) for "root", the grep command will be
unsuccessful, so the exit status will be non-zero.
51. Use a awk command and check its exit status.
awk '$0 ~ /root/' /etc/passwd > /dev/null; print $?
This command should always be successful, so the exit status is zero. If we introduce
an error into the script, such as leave out the /root/ regular expression, the awk com-
mand will be unsuccessful, so the exit status will be non-zero.
53. To show the value of all environmental variable, use the set command.
55. The first print will print Bye, and the second print will print Hello. This is because the
variable x in the subshell is independent of the variable x in the parent shell.
 CHAPTER 14

Chapter 14: Korn Shell Programming

Questions
1. No, the Korn shell does not support goto.
3. Yes, an until loop can be coded to execute infinitely. Here is an example.
until false; do
date; sleep 2 # print the time every two seconds
done
5. The Korn shell expressions are:
a. 3 + 4: (( 3 + 4 ))
b. x * 14: (( x * 14 ))
c. 4 / x: (( 4 / x ))
d. x % 5: (( x % 5 ))
7. The Korn shell expressions are:
a. x > 9: (( x > 9 ))
b. x < = y: (( x <= y ))
c. x != y: (( x != y ))
d. x = = y: (( x == y ))
9. Write Korn shell expressions for:
a. file1 is empty: [[ ! -s file1 ]]
b. file2 is not readable: [[ ! -r file2 ]]
11.
a. while loop repeated 10 times
loop_count=10
while (( loop_count > 0 )) ; do
(( loop_count = loop_count - 1 ))
echo $loop_count
done
b. until loop repeated 10 times
loop_count=10
until (( loop_count == 0 )) ; do
(( loop_count = loop_count - 1 ))
echo $loop_count
done
c. for-in loop repeated 10 times
for i in 0 1 2 3 4 5 6 7 8 9; do
echo $i
done
d. Select is used for generating menus, not to control repetitions.
44 • C h a p t e r 14 • Exercises

Exercises
13. The value of x and y:
a. (( x = x%2 )): x = 1, y = 5
b. (( x = x + 4 )): x = 17, y = 5
c. (( x = x * 3 )): x = 39, y = 5
d. (( x = x / y )): x = 2, y = 5
15. The value of each expression:
a. (( x < 12 )): 1
b. (( y > 7 )): 1
c. (( x < y )): 1
d. (( y == x )): 1
17. The loop prints an infinite sequence of hello.
19. The loop prints an infinite sequence of hello.
21. This loop prints an infinite sequence of 12 because the expression (( x > 7 )) is
always true.
23.
a. This loop prints the date and time followed by Hello and infinitely repeats
because date is executed and always returns a true exit status.
b. This loop prints an error because the while evaluates the output of the date com-
mand, which is a string instead of an exit status.
25. This loop prints the values of i in the list: 12 11 10 8 7.
27. This prints Hello seven times because that is how many values are in the list.
28. This prints Hello four times because that is how many values (strings) are in the list.
29.
a. This loop prints Hello one time because the list contains the string date which is
not interpreted as a command in this context.
b. This loop prints Hello six times because the list contains the string output of the
date command.
31. This loop prints the first three space separated strings of the date command on three
separate lines. That is, the day of the week, the month, and the date, on separate lines.
33. The script prints five lines of 12345. The inner loop prints the variable j as it changes
values from 1 to 5 and the outer loop prints a newline. The outer loop repeats five
times.
35. This loop prints the numbers 2 to 11, which is the data in the for-in loop list plus 1.
The increment expression in the loop increments each value before being printed.
37. This loop prints the first nine odd numbers, from 1 to 17. As the outer loop increments
j from 1 to 9, the inner loop prints j*2-1.
39. Simulate while loop with a for-in loop.
IFS="
"
for line in $(cat file1)
 C hapter 14 • E xercises • 45

do
…
done
46 • C h a p t e r 14 • Exercises
 CHAPTER 15

Chapter 15: Korn Shell Advanced Programming

Questions
1. 24
3. 40
5. Hello is printed twice.
7. This is the world of UNIX
(The purpose of this question is to show that quotes preserve multiple spaces.)
9. print "Hello World" | sed "s/\(..\).*/\1/"
11. print "Hello World" | sed "s/.*\(..\)$/\1/"
13. Ignore a SIGHUP signal.
trap "" SIGHUP
15. Use getopts to capture two options. getopts a:b variable

Exercises
17. Write a sed script.
echo $1 | sed -n '/^\(.\).*\1$/p'
19. Create a new file containing only one word lines.
sed -n '/^[^ ]\{1,\}$/p' < $1 > ${1}.words
21. Write a script that deletes all even lines
sed -n 'N;P' $1
23. Write a script that combines odd and even lines together.
sed -n 'N;s/\n//;p' $1
25. Write a script that prints the name of the file that is newer.
#!/bin/ksh
#print newer filename
if [[ $1 -nt $2 ]]; then
print $1
else
print $2
fi
27. What is the exit status of the script: 0
29. What is the exit status of a sed command when it finds a pattern? … when it does not
find a pattern?
A legal sed script that executes without error will generate an exit status of 0
regardless of whether it finds a pattern. Here is a one line script that searches itself for
sed, which it finds and shows the exit status of 0. It also searches itself for SED, which
it does not find, and shows the exit status of 0.
sed -n "/sed/p" $0; echo $?; sed -n "/\S\E\D/p" $0; echo $?
31. Store the first argument in $1 using the getopts command.
48 • C h a p t e r 15 • Exercises

#!/bin/ksh
echo $OPTIND
while getopts stuv variable 2>/dev/null
do
case $variable in
s) echo "processing s" ;;
t) echo "processing t" ;;
u) echo "processing u" ;;
v) echo "processing v" ;;
\?) print "Invalid option: Quitting" ; exit 1
esac
done
# shift arguments to start at $1
(( argShift=OPTIND-1 ))
shift $argShift
# show arguments
echo $1 $2 $3
 CHAPTER 16

Chapter 16: Interactive C Shell

Questions
1. The C shell can be used either interactively to receive user commands and interpret
them or as a programming language to create shell scripts.
3. No, the login shell is not always the same as your current shell. You can change to a
different subshell, and therefore the current shell will not necessarily be the same as the
login shell.
5. To return to the parent shell, use the command exit.
7. The three standard streams are standard input, standard output, and standard error.
9. The standard output stream is normally associated with the monitor.
11. The standard input stream can be redirected to a file using input redirection:
command < filename.
13. The standard error stream can be redirected to the same file as the standard output by
using append redirection: command >& file1.
15. A sequence of commands can be entered on one line separated by semicolons (;).
Commands can be grouped together using parentheses. Commands can be chained
together using pipes.
17. Quotes are used to change the predefined meanings of characters. Three types of
quotes are used for this purpose: backslash (\), a pair of double quotes ("…"), and a pair
of single quotes ('…').
19. Command substitution provides the capability to convert the output of a command to a
string. The command substitution in the C shell consists of backquotes surrounding the
command, as in `date.`
21. To suspend a foreground job key ctrl-z. The bg command moves a suspended job to
the background. The fg command moves a suspended or background job to the
foreground.
23. A variable is a location in memory where values can be stored.
25. The shell variables are used to configure the shell. The environmental variables control
the user environment and can be exported to subshells.
27. When a command is executed, it returns a value known as the exit status of the
command. In the C shell, the exit status of a command is stored in a shell variable
called status.
29. The environmental variable that holds the name of the login shell is SHELL. It is
displayed with echo $SHELL.
31. The environmental variable that holds the primary prompt in the C shell is
PROMPT/prompt. It is displayed with echo $prompt.
33. The mail check interval is not stored in a shell variable in the C shell.
50 • C h a p t e r 16 • Exercises

35. The environmental variable that holds the type of your terminal is TERM/term. It is
displayed with echo $TERM.
37. The startup scripts are: /etc/csh.login, ~/.login, /etc/csh.cshrc, and
~/.cshrc.

Exercises
39. The command that shows the login shell is echo $SHELL (a).
41. Five commands that cannot be used with input redirection are: date, who, cal, set, and
cd.
43. The display output would be:
a. a
44
b. Undefined
c. 44a
4444
45. Use a grep command and check its exit status.
grep root /etc/passwd > /dev/null; echo $status
This command should always be successful, so the exit status is zero. If we substitute
"no_user" (or some other bogus username) for "root", the grep command will be
unsuccessful, so the exit status will be non-zero.
47. Use a awk command and check its exit status.
awk '$0 ~ /root/' /etc/passwd > /dev/null; echo $status
This command should always be successful, so the exit status is zero. If we introduce
an error into the script, such as leave out the /root/ regular expression, the awk com-
mand will be unsuccessful, so the exit status will be non-zero.
 CHAPTER 17

Chapter 17: C Shell Programming

Questions
1. Yes, although goto statements are discouraged.
3. C shell foreach loops are list-controlled and can only loop a finite number of times.
5. The C shell expressions are:
a. (3 + 4) * 8: @ result = (3 + 4) * 8
b. (x * 14) / y: @ result = ($x * 14) / $y
c. z = t * (4 / x) + 6: @ z = $t * (4 / $x) + 6
d. x = x + y + z: @ x = $x + $y + $z
7. The C shell expressions are:
a. (x == 5) && (y > 10): (($x == 5) && ($y > 10))
b. (x < = y) || (y > 10): (($x <= $y) || ($y > 10))
c. (x != y) || (x > 7) && (y <= x):
(($x != $y) || ($x > 7) && ($y <= $x))
d. (x = = y) || (x > 5) || (y > 6):
(($x == $y) || ($x > 5) || ($y > 6))
9.
a. while loop repeated 10 times:
set count = 10
while ($count > 0)
@ count -= 1
echo $count
end
b. foreach loop repeated 10 times
foreach i (0 1 2 3 4 5 6 7 8 9)
echo $i
end
c. repeat loop repeated 10 times
repeat 10 echo hello

Exercises
11. The value of x and y:
a. x = 1, y = 5
b. x = 17, y = 5
c. x = 39, y = 5
d. syntax error
13. The value of each expression:
a. ( $x < 12 && $y > 7 ): false
b. ( $y > 7 || $x < $x ): false
52 • C h a p t e r 17 • Exercises

c. ( ! $y < 7 ): true (! has high order of precedence.)

d. ( ! $x >= 8 && ! $x < 9 ): false
15. The loop prints nothing.
17. The loop prints nothing.
19. The loop prints 5, <, and 6 on three separate lines.
21. The prints 12, 11, 10, 9, and 8 on separate lines as x decrements inside the loop.
23. The loop prints Hello three times because x yields three values.
25. The loop prints Hello four times because x yields four values.
27.
a. The loop prints Hello one time because the foreach list consists of one item, the
word date (not the command date).
b. The loop prints Hello six times, once for each field output by the date command.
29. The script segment prints the values 1 through 10, the values in the foreach list.
31. Each line’s contents is controlled by the inner loop which counts from 1 to i, while
the outer loop increments i from 1 to 9, so this prints the following:
1
12
123
1234
12345
123456
1234567
12345678
123456789
33. The script segment prints the odd numbers 1 through 9 on separate lines. Each value
in the foreach list is multiplied by 2 and decremented and then printed if the result is
less than or equal to 9.
 CHAPTER 18

Chapter 18: C Shell Advanced Programming

Questions
1. The command displays This is the world of UNIX
(The purpose of this question is to show that quotes preserve multiple spaces.)
3. There are nine elements in the array.
5. This results in an error since the echo command will reference a non-existent element.
7. The code prints 44, 66, 77 on separate lines
9. This code prints the number of words in he file.

Exercises
11. Change the name of files passed as arguments to lower-case.
#!/bin/csh
#convert filename ($argv[1]) to lowercase
set lowername = `echo $argv[1] | tr "A-Z" "a-z"`
mv $argv[1] $lowername
13. Write a script that deletes all odd lines
#!/bin/csh
set filename = $argv[1]
sed -n 'n;p' $filename
15. Write a script that changes the first three characters of file names to lower-case.
#!/bin/csh
# change first three letters of filenames to lowercase
while ( $#argv > 0 )
set name = $argv[1]
set first3 = `echo $name | sed "s/^\(...\).*/\1/"`
set end = `echo $name | sed "s/^...\(.*\)/\1/"`
set lower3 = `echo $first3 | tr "[A-Z]" "[a-z]"`
set lower3Name = ${lower3}${end}
mv $name $lower3Name
shift
end
17. The exit value of the script is 0.
19. A legal sed script that executes without error will generate an exit status of 0 regardless
of whether it finds a pattern. Here is a one line script that finds a match and shows the
exit status of 0. It then does not find a match and shows the exit status of 0:
echo xx | sed -n "/xx/p"; echo $status; echo xx | sed -n "/yy/p";
echo $status
54 • C h a p t e r 18 • Exercises