Reference no: EM133140294

KIT501 ICT Systems Administration Fundamentals - University of Tasmania

Goal

The main purpose of this practical is to give you more familiarity with iteration (looping) within UNIX shell scripts, and it also introduces another loop syntax, the for loop.

1. UNIX Shell Scripting - Looping - First Sample Script

The following sample script (and the sample script in section 2) reinforce the looping concept in UNIX scripts.

Hands-on exercises

A Under your kit501 directory, change into the directory named Looping which you created last week. If you have removed it, create it now (and change into it).
B Create a script which must be named as cpback.sh, with the following content. For time considerations, you do not need to type the comment lines into your script (except for the first line), however in any scripts you write (for example the scripting assignment), comments are needed. Comments are there to help you understand the script. The first special comment the beginning is needed as it determines which shell is used to run the commands that follow.

#! /bin/sh

# This program copies a file into a safe directory without
# overwriting. For example, it copies a file named foo to foo.1 # if foo exists in the destination directory, or to foo.2
# if foo.1 exists in the destination directory.

# To run the script, two arguments are required in the command line, # a source file, followed by a destination directory. Without two
# arguments, the scrip displays the usage information, and exits

if [ $# -ne 2 ]; then
echo "Usage: $0 source destination"; exit

# The script also requires that the second argument must be an # existing directory file, if not, displays an error message # and exits

elif [ ! -d $2 ]; then
echo "Directory does not exist"; exit else
# assign the first argument (the source filename) as value # to variable file
file=$1

# If the source file does not exist in the destination directory, # then simply copy it into the directory
if [ ! -f $2/$file ]; then cp $1 $2

# But if the source file does exist in the destination directory, # then we have to do something to avoid overwriting
else
copies=1 # The variable copies takes 1 as initial value while true
do
if [ ! -f $2/$file.$copies ]; then cp $1 $2/$1.$copies
echo "File $1 copied to $1.$copies" break
else
copies=`expr $copies + 1`
fi done
fi
fi
Save the content of the script. Assign execute permission to the script.

To run the script, you need a support source file and a support destination directory.

Create a support text file named foo. One quick way to do this is to redirect the output of a command line into the file: (this example copies the output of the man page for the man command into the file called foo)

$ man man > foo
Create a support directory named safedir. Run the script as follows:
$ ./cpback foo safedir

Run the command line two more times:
$ ./cpback foo safedir
$ ./cpback foo safedir

Use ls to list all the file names stored under safedir. Has any file overwriting occurred?

2. UNIX Shell Scripting - Looping - Second Sample Script

The following example shows some pattern-matching using case statement based on user's input.

Hands-on exercises

Create a script which must be named as dentry.sh, with the following content. You do not need to type the comment lines into your script. They are there to help you understand the script.

Please note that this is a script with nested loop structure, that is, an inner loop is contained within an outer loop.

# This script repeatedly accepts a two-digit code (such as 01, 02, etc) and
# its corresponding description (such as engineer) from keyboard, performs
# some validation checks, and then adds the entry into a text file named # newlist

# First create a new blank text file named newlist, if it does not currently
# exist
if [ ! -f newlist ]; then touch newlist
fi

# Start the outer loop by prompting the user to enter a code while echo -e "Designation code: \c"
do
read code
case "$code" in
# Ensure the two-digit code is in correct format, # and does not currently exist
[0-9][0-9]) if grep "$code" newlist > /dev/null; then echo "Code exists"
continue fi ;;
*) echo "Invalid code" continue ;;
esac

# Start the inner loop by prompting the user to enter a description while echo -e "Description: \c";
do
read desc

case "$desc" in
# This is how to ensure that the user has entered a valid # description,n which contains letters and spaces only
*[!\ a-zA-Z]*) echo "Can contain only letters and spaces" continue ;;

# If the user simply presses the Enter key, without typing anything "") echo "Description not entered"
continue ;;

# The user has entered a valid description, add it together with # the previously entered valid code into the file
*) echo "$code|$desc" >> newlist break ;;
esac done

echo -e "\nEnter another entry? (y/n): \c" read answer
case "$answer" in [yY]*) continue ;;
*) break ;;
esac done

B Save the content of the script. Assign execute permission to the script.

Run the script by entering the following codes and descriptions, respectively:

01 student
02 engineer
03 chef
C Check the content of the newlist file to verify the above entries.

3. UNIX Shell Scripting - The for Loop - Looping with a list

The general format of for loop is as follows:

for variable in list
do
commands
done

• list can be any sequence of items, for example, strings, numbers, or even filenames.
• variable is the control variable. Each iteration its value is the next item in the list. Use
$variable to evaluate (retrieve) the value.
• commands are one or more commands to execute each iteration.

Iterating for all items in an explicit list
The following example shows how to do something with every item in an explicit list:

for item in item1 item2 item3 do
echo "The item is: $item"
done

3.1 Hands-on exercises

A Ensure that your present working directory is Looping, which is under your kit501 directory.

Make a new directory named forloop (under Looping) and change into the new directory.
B Create 3 text files which must be named as chap20, chap21, and chap22, respectively, by simply redirecting outputs of some echo commands:

$ echo "content of chap20" > chap20
$ echo "content of chap21" > chap21
$ echo "content of chap22" > chap22

Verify that the 3 files have been created with the correct content.
C Create a script which must be named as ftest.sh, with the following content:

for file in chap20 chap21 chap22 do
cp $file "$file.bak"
echo $file copied to "$file.bak"
done

In this script, file is a variable, which takes chap20 as its value first. Then the loop body is entered, which copies chap20 to chap20.bak, and displays a message to confirm the operation. Then the loop is repeated, with the variable file taking chap21 as its value this time, making a backup copy for chap21 and displaying a message confirming the operation. Then the loop is repeated again, with the variable file taking chap22 as its value, making a backup copy for chap22 and displaying a message confirming the operation. Now that all the files in the list have been manipulated, the loop ends.

Save the content. Assign execute permission to the script.

Run the script. Observe the outputs. Verify that all the backup files have been created.
D What if you want to make a backup copy of each file stored under the current directory, and there are currently 300 files available (under the current directory). Do you have to specify each filename in the list explicitly? You do not want to do this!
An approach is the following:

for file in `ls -p | grep -v /` do
cp $file "$file.bak"
echo $file copied to "$file.bak"
done

Note the command substitution used in the first line with the ls command. No matter how many files are stored under the current directory, the output of ls will retrieve all of them! The -p option tells ls to put a slash character after directory names. The grep -v / part means only show lines that do not contain the / character - this is one way to only find files (and exclude directory names). Of course, if the current directory only contains files, then just
`ls` by itself will work as well.

Modify the first statement in your ftest.sh so that it uses `ls` rather than an explicit list. Remove all the existing .bak files under the current directory. Run the script again. You will see that it also makes a backup copy of the script itself.

Remove all the .bak files again.
E Think about what the following script does:

for file in *.html *.htm do
gzip $file
done

Attachment:- ICT Systems Administration Fundamentals.rar