The Linux Command Line (2012)

Part IV. Writing Shell Scripts

Chapter 33. Flow Control: Looping with for

In this final chapter on flow control, we will look at another of the shell’s looping constructs. The for loop differs from the while and until loops in that it provides a means of processing sequences during a loop. This turns out to be very useful when programming. Accordingly, the for loop is a very popular construct in bash scripting.

A for loop is implemented, naturally enough, with the for command. In modern versions of bash, for is available in two forms.

for: Traditional Shell Form

The original for command’s syntax is as follows:

for variable [in words]; do
commands
done

where variable is the name of a variable that will increment during the execution of the loop, words is an optional list of items that will be sequentially assigned to variable, and commands are the commands that are to be executed on each iteration of the loop.

The for command is useful on the command line. We can easily demonstrate how it works:

[me@linuxbox ˜]$ for i in A B C D; do echo $i; done
A
B
C
D

In this example, for is given a list of four words: A, B, C, and D. With a list of four words, the loop is executed four times. Each time the loop is executed, a word is assigned to the variable i. Inside the loop, we have an echo command that displays the value of i to show the assignment. As with the while and until loops, the done keyword closes the loop.

The really powerful feature of for is the number of interesting ways we can create the list of words. For example, we can use brace expansion:

[me@linuxbox ˜]$ for i in {A..D}; do echo $i; done
A
B
C
D

or pathname expansion:

[me@linuxbox ˜]$ for i in distros*.txt; do echo $i; done
distros-by-date.txt
distros-dates.txt
distros-key-names.txt
distros-key-vernums.txt
distros-names.txt
distros.txt
distros-vernums.txt
distros-versions.txt

or command substitution:

#!/bin/bash

# longest-word : find longest string in a file

while [[ -n $1 ]]; do
if [[ -r $1 ]]; then
max_word=
max_len=0
for i in $(strings $1); do
len=$(echo $i | wc -c)
if (( len < max_len )); then
max_len=$len
max_word=$i
fi
done
echo "$1: '$max_word' ($max_len characters)"
fi
shift
done

In this example, we look for the longest string found within a file. When given one or more filenames on the command line, this program uses the strings program (which is included in the GNU binutils package) to generate a list of readable text “words” in each file. The for loop processes each word in turn and determines if the current word is the longest found so far. When the loop concludes, the longest word is displayed.

If the optional in words portion of the for command is omitted, for defaults to processing the positional parameters. We will modify our longest-word script to use this method:

#!/bin/bash

# longest-word2 : find longest string in a file

for i; do
if [[ -r $i ]]; then
max_word=
max_len=0
for j in $(strings $i); do
len=$(echo $j | wc -c)
if (( len > max_len )); then
max_len=$len
max_word=$j
fi
done
echo "$i: '$max_word' ($max_len characters)"
fi
done

As we can see, we have changed the outermost loop to use for in place of while. Because we omitted the list of words in the for command, the positional parameters are used instead. Inside the loop, previous instances of the variable i have been changed to the variable j. The use of shift has also been eliminated.