This post is collection of 10 useful unix \/ linux shell script code snippets I’ve often used.<\/p>\n\n\n\n
Following these code snippet you will be able to answer following questions:
1. How to calculate script run-time in shell.
2. How to parse parameters \/ arguments to shell script.
3. How to safely change a directory in shell script.
4. How to check if a command was successful in shell script.
5. How to add timestamps in script log output.
6. How to check if a process is running.
7. How to colour your script output.
8. How to read variables from config file.
9. How to loop over files in a folder.
10. How to perform actions based on variable value using SWITCH…CASE<\/p>\n\n\n\n
And a bonus snippet
11: How to send emails from shell script.<\/p>\n\n\n\n
Let’s begin.
<\/p>\n\n\n\n\n\n\n\n
It’s always good to note the script duration and having knowledge of the run-time. Following is the code snippet for noting the script execution time.<\/p>\n\n\n\n
Add this to the beginning of the script:<\/p>\n\n\n\n
\nSTART=$(date +%s)\n<\/pre>\n<\/blockquote>\n\n\n\nAdd following at the end or exit location:<\/p>\n\n\n\n
\nEND=$(date +%s)\nDIFF=$(( $END - $START ))\nDIFF=$(( $DIFF \/ 60 ))\n<\/pre>\n<\/blockquote>\n\n\n\n$DIFF will give you the runtime of the script in minutes.<\/p>\n\n\n\n
2. Parsing & processing script parameters \/ arguments:<\/h1>\n\n\n\n
Scripts are often required to have user inputs in terms of arguments passed after script. Below code snippet is a simple method to capture the arguments passed.<\/p>\n\n\n\n
\nwhile [ \"$1\" != \"\" ]; do\n case $1 in\n -s ) shift\t\n\t\tSERVER=$1 ;; \n -d ) shift\n\t\tDATE=$1 ;;\n\t--paramter|p ) shift\n\t\tPARAMETER=$1;;\n -h|help ) usage # function call\n exit ;;\n * ) usage # All other parameters\n exit 1\n esac\n shift\ndone\n<\/pre>\n<\/blockquote>\n\n\n\nAbove snippet in shell script paramter_test.sh will behave as follows:<\/p>\n\n\n\n
\nsh parameter_test.sh -s myserver -d 20151225 --parameter SomeValue\n<\/pre>\n<\/blockquote>\n\n\n\nThough we may need to check for correctness of input or mandatory parameters; using say a parser function.
# Confirm the mandatory parameters<\/p>\n\n\n\n\nif [ -z $SERVER ] || [ -z $DATE ]; then\n\techo \"Please specify both server and date\";\n\texit 1;\nfi;\n<\/pre>\n<\/blockquote>\n\n\n\n3. Change directory before processing your script:<\/h1>\n\n\n\n
Sanitizing & verifying the directory status is very important before progressing the script. You do not want to mistakenly execute the commands in a wrong directory. We should verify if directory is present and have sufficient permission for the user.
Following code snippet takes care of the same.<\/p>\n\n\n\n\nchangedir()\n{\n\tDIR_NAME=$1\n\t# Check if the directory exist?\n\t[ -d \"$DIR_NAME\" ] || {\n\t\techo Dir: $DIR_NAME does not exist \n\t\texit 1\n\t}\n\n\t# Check if the directory is readable\n\t[ -r \"$DIR_NAME\" ] || {\n\t\techo Dir: $DIR_NAME not readable\n\t\texit 2\n\t}\n\n\t# Check if we have execute perms on directory\n\t[ -x \"$DIR_NAME\" ] || {\n\t\techo Dir: cannot cd to $DIR_NAME\n\t\texit 3\n\t}\n\n\t# Check if the directory is writable\n\t[ -w \"$DIR_NAME\" ] || {\n\t\techo Dir: $DIR_NAME not writeable\n\t\texit 4\n\t}\n\n\tcd $DIR_NAME\n\techo \"Present directory $DIR_NAME\"\n}\n<\/pre>\n<\/blockquote>\n\n\n\nTo safely change the directory call the function as follows:<\/p>\n\n\n\n
\nchangedir \/data\/app\/\n<\/pre>\n<\/blockquote>\n\n\n\n4. Verifying if previous command was successful<\/h1>\n\n\n\n
When a processing of command depends on success or failure of previous command we can use the exit status. It can be checked using $?.<\/p>\n\n\n\n
\nif [ $? -ne 0 ]; then\n\techo \"The command was not successful.\";\n\t#do the needful \/ exit\nfi;\n<\/pre>\n<\/blockquote>\n\n\n\n5. Generating script logs with timestamp<\/h1>\n\n\n\n
Just as the duration of script it’s useful to have timestampped log. Use following function to log time for every output.<\/p>\n\n\n\n
\nlog() {\n echo [`date +%Y-%m-%d\\ %H:%M:%S`] $*\n}\n<\/pre>\n<\/blockquote>\n\n\n\nCall the function as follows instead of simply “echo”ing.<\/p>\n\n\n\n
\nlog \"my string to be logged\"\n<\/pre>\n<\/blockquote>\n\n\n\n6. Checking if process is running:<\/h1>\n\n\n\n
This code snippet is used often when a script progress requires the known status of any particular system process.
A usual scenario say we need MySQL down before progressing further in script, we will use following shell function.<\/p>\n\n\n\n\n# Define shell function\ncheck_process() {\n\techo \"Checking if process $1 exists...\"\n\t[ \"$1\" = \"\" ] && return 0\n\tPROCESS_NUM=$(ps -ef | grep \"$1\" | grep -v \"grep\" | wc -l)\n\tif [ $PROCESS_NUM -ge 1 ];\n\tthen\n\t return 1\n\telse\n\t return 0\n\tfi\n}\n<\/pre>\n<\/blockquote>\n\n\n\n# Check for MySQL process and make the decision<\/p>\n\n\n\n
\ncheck_process mysql;\nCHECK_RET=$?;\nif [ $CHECK_RET -ne 0 ]; \n\t# code block for process exists \nelse\n\t# code block for process not present\nfi;\n<\/pre>\n<\/blockquote>\n\n\n\n7. Colouring your script output<\/h1>\n\n\n\n
A readable and formatted output is good to have. Following code snippet helps you beautify your script output, colorizing or highlighting them.<\/p>\n\n\n\n
\nDefine variables:\ntxtund=$(tput sgr 0 1) # Underline\ntxtbld=$(tput bold) # Bold\ntxtred=$(tput setaf 1) # Red\ntxtgrn=$(tput setaf 2) # Green\ntxtylw=$(tput setaf 3) # Yellow\ntxtblu=$(tput setaf 4) # Blue\ntxtpur=$(tput setaf 5) # Purple\ntxtcyn=$(tput setaf 6) # Cyan\ntxtwht=$(tput setaf 7) # White\ntxtrst=$(tput sgr0) # Text reset\n\nUse them as:\necho \"${txtbld}This is bold text output from shell script${txtrst}\"\necho \"${txtred}This is coloured red except ${txtblu}this is blue${txtrst}\"\n\n${txtrst} will reset the terminal.\n<\/pre>\n<\/blockquote>\n\n\n\n