Bash Cheat Sheet

A bash cheat sheet with code listings for common tasks. This document is a cookbook. It does not explain why you do things in depth it only shows you what to do. It is meant for people who already have some experience in bash and quickly need to hack something together.

The idea is that one can open this document on one monitor and open a text editor in the other one and copy paste sollutions with some small adjustments over to get a working script that does something.

Table of contents

• Strings & File content
  • Reading a file line by line
  • Splitting a string into parts via a delimiter
• Conditionals
  • Check if a file exists before using it
  • Check if a array element exists at a certain index
  • Check if an array is empty
• The Find command
  • Find/list all the subdirectories of a directory
  • Find/list only files of a certain type.
• The cut command
  • Get the n-th characters from every line.
  • Get the first n-th character from every line
  • Get the n-th remaining characters of every line
  • Get the n-th word of every line.

Strings & File content

This section covers the manipulation and processing of strings and file contents.

Reading a file line by line

FILE="filename"
while IFE= read line; do
	echo "$line"
done < "$FILE"

• You can possibly add a -r flag to the read command which will ignore escaping with \ characters.
• You can choose the delimiter with IFS="delim" in this case the default delimiter of a newline will be used.

Splitting a string into parts via a delimiter

IFS=":" read first second rest <<< "$STRING"

• You can chose the delimiter with IFS="delim"
• The first word before the delimiter will be put in the first variable, the second in the second variable and so on.
• If the amount of delimited strings is greater then the ammount of provided variables then the rest will of the string will be appended to the last provided variable. In this case the rest variable.

Conditionals

If you want to run a piece of code only when a logical condition is met do the following.

[[ condition ]] && echo "The condition was met"

These conditions can be chained. And will be checked in order.

[[ condition ]] && [[ condition2 ]] && echo "Some code"

NOTE: These are all syntactical suger for the plain old if statement. You can always use nested if statements but they are less readable. So the above statement could be written as follows.

if [[ condition ]]; then
	if [[ condition2 ]]; then
		echo "Some code"
	fi
fi

Check if a file exists before using it

Example, only read a file when it exists then print out all the lines.

[[ -a $FILE ]] && while read -r line; do echo $line; done < $FILE

Check if a array element exists at a certain index

You can check this by checking if the value at a certain index is equal to the the empty string.

array=()
index=1

[[ "${array[$index]}" == "" ]] && printf "The element at index %d was empty" "${index}"

Check if an array is empty

array=()
[[ ${#array[@]} -eq 0 ]] && echo "Array is empty"

You can also use arithmetic expansion.

array=()
(( ${#array[@]} == 0 )) && echo "Array is empty"

There is some nuances to be noted here: https://serverfault.com/questions/477503/check-if-array-is-empty-in-bash

The Find command

Find is not a shell builtin but a tool commonly used in scripts. You can find it by default on most UNIX environments.

find /dev -name 'sda*'

This will print out all the partitions of the disk linked to the device file /dev/sda and the device file itself. A find command is a bit different from normal GNU tools. The first argument is the directory where you want to search. Then a set of command called predicates follow. For each file in the directory find will evaluate all these predicates in order and only files that produce true for all of them will be printed out.

The above command only has one predicate -name which simply checks if the file name starts with sda (notice the use of the wildcard). But you could add multiple predicates.

find /dev -name 'sda*' -name '*1'

This will print out all the files that start with sda and end with 1. So only the first partition of /dev/sda. The -name predicate is the simpelest one but there are a lot of them. Check man find for a listing of them. Some examples:

• -regex checks if the file matches the regex.
• -empty evaluates to true if the file is empty.

You can also negate the predicate by providing a ! before the predicate. So to get only the partitions without the device file one can do something like this.

find /dev -name 'sda*' ! -name 'sda'

Which translates to: every file that starts with sda but not the file that matches sda exactly.

NOTE: The -print predicate always evaluates to true and prints out the files. This is default behavior on most machines but you can specify another print behavior with predicates like -print0 which does not print out a newline but adds a NULL character. Or the predicate -printf which allows formatted printing.

Find/list all the subdirectories of a directory

find /dev -type d -print

Using the -type predicate you can filter on the basis of file type. For a quick explanation on the syntax see introduction to this section. For a quick explanation on the syntax see the introduction to this section The Find command.

Find/list only files of a certain type.

For example find/list all the sockets in a directory.

find /run -type s -print

Using the -type predicate you can filter on the basis of file type.

-type c
              File is of type c:

              b      block (buffered) special

              c      character (unbuffered) special

              d      directory

              p      named pipe (FIFO)

              f      regular file

              l      symbolic  link;  this is never true if the -L option or the -follow option is in effect, unless
                     the symbolic link is broken.  If you want to search for symbolic links when -L  is  in  effect,
                     use -xtype.

              s      socket

              D      door (Solaris)

              To  search  for more than one type at once, you can supply the combined list of type letters separated
              by a comma `,' (GNU extension).

For a quick explanation on the syntax see the introduction to this section The Find command.

The cut command

Cut is not a shell builtin but can be found on most UNIX systems. The first line of the cut manpage says it all.

SYNOPSIS
       cut OPTION... [FILE]...

DESCRIPTION
       Print selected parts of lines from each FILE to standard output.

What parts it cuts from every line depends on the options you give it. See man cut for a listing.

NOTE: cut will cut from every line. So do not make the mistake of thinking that something like this

cut -c 1 file.txt

will only return the first character from the file. It returns the first character of every line in the file.

TIP: I found it handy to thing of the verb 'select' when thinking about the cut command. Because cut will print out the characters you 'select' with your options. One would be forgiven for thinking that a command like the one above prints out the lines with the first character cut out (because that is what it means to cut something out). But as the manpage tells you, its the other way around.

Get the n-th characters from every line.

For example get the second character.

cut -c 2 $FILE

NOTE: You can always pipe the output of a command into cut so the input does not have to be a file.

Get the first n-th character from every line

For example get the first 3 characters from every line.

cut -c -3 $FILE

NOTE: You can always pipe the output of a command into cut so the input does not have to be a file.

Get the n-th remaining characters of every line

For example give me the 5 remaining characters of every line.

cut -c 5- $FILE

NOTE: You can always pipe the output of a command into cut so the input does not have to be a file.

Get the n-th word of every line.

Words are delimited by spaces (the string " ") so you can do the following. Here we take the 4th word.

cut -d " " -f 4 $FILE

NOTE: You can always pipe the output of a command into cut so the input does not have to be a file.