Sorting Input in Unix Shell Script

Supposing your inputs are in $a, $b, $c, $d you can do this

printf "%s\n%s\n%s\n%s\n" $a $b $c $d | sort

another way

read -p "Enter 1st number: " a
read -p "Enter 2nd number: " b
read -p "Enter 3rd number: " c
read -p "Enter 4th number: " d
printf "%s\n" $a $b $c $d | sort

You can use the "sort" utility, which reads from stdin and sorts each line lexicographycally.

As an example, you can check:

echo -e "1\n6\n2\n3" | sort -n

Which outputs:

1
2
3
6

Just output each number on a separate line, and pipe that into sort

You need to read the values and then echo them to the sort program.

echo -n "Enter 1st number: "
read a junk
echo -n "Enter 2nd number: "
read b junk
echo -n "Enter 3rd number: "
read c junk
echo -n "Enter 4th number: "
read d junk
{ echo $a; echo $b; echo $c; echo $d; } | sort -n

The only problem here is the notation used to suppress newlines after the echo command. There are two notations - and different systems use different notations. The older notation is the echo -n shown. The newer notation is:

echo "Enter Nth number: \c"

Which you need to use depends on your system, then.

The 'junk' variable collects any extra words after the first word on the input line. It's a form of paranoia, but means that if the user types '12 abc', then the single-letter variable gets just the '12' and not the rest. The code does not validate that the user entered a number. You could write a function to do that, and use it 4 times - it might as well take the prompt string and return a number.

Note that if you use bash, the read command is much more powerful and can do the prompting as well, side-stepping tricky issues with echo notation for non-portable (to other than bash) notations in read instead. Except for the echo notation, the code shown is portable to Bourne shell back to 7th Edition (or Version 7) UNIX (where echo was not a shell built-in and it used the -n option) and derivatives (Korn, Bash, POSIX, ...).