[Top level Bash Page] .... [References]

Example scripts

1. if [ -z "$FILENAME" ] ; then
2.   echo "Usage: ./bex1.sh [filename]"
3.   exit 0
4. fi
5. if [ -e $FILENAME ] ; then
6.   echo "Properties for $FILENAME:"
7.   echo "Size is $(ls -lh $FILENAME | awk '{ print $5 }')"
8.   echo "Type is $(file $FILENAME | cut -d":" -f2 -)"
9.   echo "Inode number is $(ls -i $FILENAME | cut -d" " -f1 -)"
10.   echo "$(df -h $FILENAME | grep -v Mounted | awk '{ print "On",$1", \
11. which is mounted as the",$6,"partition."}')"
12. else
13.   echo "File does not exist."
14. fi

• Line 1:  We could use the new test ([[ ]]) instead:  if [[ -z $FILENAME ]] ; then
• Line 7:   when we call awk '{print$5}'  ....what is this doing?
• ./bex1.sh   but modify line 7 to  echo "Size is $(ls -lh $FILENAME )"
• Properties for bex2.sh:
  Size is -rwx--x--x 1 jmarty cuuser 513 Feb 26 08:28 bex2.sh
• The line above is input to awk.  Awk by default will parse fields separated by spaces.  We are asking awk to return the value in the fifth field
• >echo 'Size is -rwx--x--x 1 jmarty cuuser 513 Feb 26 08:28 bex2.sh' |  awk '{print$5}'
  jmarty
• Line 8:  The file command returns the file type information:
• >file bex3.sh    returns :     Type is bex3.sh: Bourne-Again shell script, ASCII text executable
• The cut command is used to cut -d":" -f1 -   
  
• The  -d":" param tells it to delimit the line using the colon.  The -f1 says cut field 1.  The  last param, the minus, has the command read from standard input
  
• echo "Type is bex3.sh: Bourne-Again shell script, ASCII text executable" | cut -d":" -f2 -
   Bourne-Again shell script, ASCII text executable
• Note: it's possible to use awk to do the same.....would set the delimiter to ":" and then select the second field.
  

Example: bex2.sh. this is an example of a script that provides a user interface to call all scripts in the current directory

1. #!/bin/bash
2. while true; do
3. echo
4. echo "Select a CMD?"
5. echo
6. Files=("empty");
7. Count=1
8. for f in *.*;
9. do
   
10. if [ "${0:2}" != "$f" ] && [ "${f: -3}" == ".sh" ]; then
    
11. echo ${0:2};
    
12. echo $Count. $f;
    
13. let "Count +=1";
    
14. Files=("${Files[@]}" $f)
    
15. fi
    
16. done
17. echo -n "Enter your choice, or 0 for exit: "
    
18. read choice
    
19. echo
20. if [ "$choice" == "0" ]; then break; fi
    
21. if [ "$choice" -ge "${#Files[@]}" ]; then echo "Invalid choice."; continue; fi
22. echo Running ${Files[$choice]}...
    
23. echo
    
24. ./${Files[$choice]}
25. done
    

Example:  bex3.sh  : this is a simple example of looping through the set of files in the current directory.   It copies the *.sh files in the cd to the subdirectory ./tmp

1. #!/bin/bash
2. # specific conversion script to convert sh files to ./tmp/script
3. LIST="$(ls *.sh)"
4. for i in "$LIST"; do
5. echo $i
6. done
7. if [ ! -d "./tmp" ]; then
8. mkdir tmp;
9. fi
10. path=*.sh
11. for f in $path;
12. do
13. cp "./$f" "./tmp/$f"
14. done
15. for f in ./tmp/*.sh;
16. do
17. echo -e "\n$(cat $f)" > $f
18. echo -e "# ahayes@clemson.edu\n$(cat $f)" > $f
19. echo -e "# Created by: Austen Hayes\n$(cat $f)" > $f
20. done

Example script : filesUpdated.sh

This script can be used to identify all files that have been modified within a specified amount of time.

Syntax: filesUpdated.sh param1 param2 param3

• Param1: the starting directory. Your script should start here and step into all subdirectories
• Param2: amount of time that is of interest. We will define this to be the mtime parameter in find (n*24 hours). So the granularity of the time of interest is in units of days.
• Param3: script mode. The output of the script depends on the setting of this parameter:
  • Value ‘1’ : summary mode: indicate how many files per user (i.e., unique number of owners), sorted in descending order based on number of files modified.
  • Value ‘2’ : verbose mode: For each user (i.e., owner), list the files that were modified. The format for each file should be the results of ‘ls –lt filename’
    

Example output for script mode 1 :
Owner Number of files changes
jjm 3
root 1
---------------

Example output for mode 2

jjm
----------
w------- 1 jjm jjm 1998 Feb 8 20:30 plotPDF.m
-rw-rw-r-- 1 jjm jjm 45000 Feb 8 20:29 ex2.1
-rw-rw-r-- 1 jjm jjm 45000 Feb 8 20:29 ex2.2

Root
----------
w-r--r-- 1 root root 3798 Feb 5 20:37 boot.log
----------------------------

filesUpdated.sh (credit to Anna Kutch - CPSC4240 Spring 2015)

#!/bin/bash
# Anna Kutch (akutch@g.clemson.edu)
# CPSC 4240 HW2 Script

# init variables
startDir=$1
amtTime=$2
mode=$3
count=0

# make sure there's a correct number of arguments
if (( $# < $"3" )); then
echo "Usage: [starting dir] [amt time] [script mode]"
exit 0
fi

# OUTPUT will be a list in the form:
# <count of files> <user>
#
# find $startDir -mtime -$amtTime -ls : find the files modified within $amtTime
# -type f : it is a file (not a directory)
# awk '{print $5}' : get the user name field only
# sort -r : sort in descending order
# uniq -c : print the count for unique user names
# awk '{printf ("%s\t%s\n", $2, $1)}' : print the name followed by the number

OUTPUT="$(find $startDir -mtime -$amtTime -type f -ls | awk '{print $5}' | sort -r | uniq -c | awk '{printf ("%s\t%s\n", $2, $1)}')"

# for script mode 1 (summary mode):
if(($mode==1));
then
printf "%-20s\tNumber of files changed\n" "Owner"
for i in $OUTPUT; do

# print the number of modified files
if(($count%2));
then
# count%2 = 1 -- file count
printf "$i\n"

# print the user name
else
# count%2 = 0 -- user name
printf "%-20s\t" $i
fi
let "count++"
done
fi

# for script mode 2 (verbose mode):
if(($mode==2));
then

# VOUTPUT will be a list of filenames
#
# find $startDir -mtime -$amtTime -ls : find the files modified within $amtTime
# -type f : it is a file (not a directory)
# awk '{print $11}' : print the file name

VOUTPUT="$(find $startDir -mtime -$amtTime -type f -ls | awk '{print $11}')";

# for each unique person...
for i in $OUTPUT; do
# print the files associated with the user
if(($count%2)); then
# only print the files for the specified user
echo "$(ls -lt $VOUTPUT | grep $username)"

# print and set the user name
else
# count%2 = 0 -- user name
printf "\n%-20s\n" $i
username="$i"
fi
let "count++"
done
fi

Last update: 9/6/2017