The awk command is used to only display certain fields of output. In this context, a field is a string of data, delimited by whitespace. For example, the df -h command has fields of data delimited by whitespace.

Piping the output to awk '{print}' produces the same exact output.

[root@server1 ~]# df -h | awk '{print}'
Filesystem     Size    Used    Avail    Use%    Mounted on
/dev/md127     1.8T    1.1T     604G     65%    /
/dev/sda1      488M    100M     354M     22%    /boot

$1 will only print field 1.

[root@server1 ~]# df -h | awk '{print $1}'
Filesystem
/dev/md127
/dev/sda1

$5 will only print print 5.

[root@server1 ~]# df -h | awk '{print $5}'
Size
65%
22%

Let's add some custom text.

[root@server1 ~]# df -h | awk '{print $5, "of", $1, "used"}'
Filesystem is mounted on Mounted on
%65 of /dev/md127 used
22% of /dev/sda1 used

Specify delimiter

By default, whitespace separates fields. The /etc/passwd file is delimited by a colon.

user1:x:1001:1001::/home/user1:
user2:x:1002:1002::/home/user2:
user3:x:1003:1003::/home/user3:

The -F or --field-separator option can be used to specify a delimiter. For example, colon can be specified as the delimiter.

[root@server1 ~]# cat /etc/passwd | awk -F':' '{print $1}'
user1
user2
user3

Do not print certain lines

NR!=n can be used to not print certain lines. In this context, n is an integer, such as 1 or 2 or 3. As an example, NR!=1 will not print line 1.

~]# df -h | awk 'NR!=1{print $5, "of", $1, "used"}'
%65 of /dev/md127 used
22% of /dev/sda1 used

Only print certain lines

NR==n can be used to only print certain lines. In this example, only line 2 is printed.

~]# df -h | awk 'NR==2{print}'
/dev/md127     1.8T    1.1T     604G     65%    /

A range of lines can be specified. In this example, lines 1 through 3 are printed.

~]# df -h | awk 'NR==1,NR==3{print}'
Filesystem     Size    Used    Avail    Use%    Mounted on
/dev/md127     1.8T    1.1T     604G     65%    /
/dev/sda1      488M    100M     354M     22%    /boot

Do not print certain fields

The following can be used to print every field except for field 1.

df -h | awk '{for (i=2; i<NF; i++) printf $i " "; print $NF}'

Which should produce the following.

Size Used Avail Use% Mounted on
1.8T 1.1T 604G  65%  /
488M 100M 354M  22%  /boot

gsub can be used to do a regular expression replacement. In this example, "dev" will be replaced with "bar" and $0 is used to only return the results that were modified.

]$ df -h | awk 'gsub("dev", "bar", $0)'
bartmpfs                          3.7G     0  3.7G   0% /bar
tmpfs                             3.7G     0  3.7G   0% /bar/shm
/bar/mapper/centos_ansible1-root   50G  9.2G   41G  19% /
/bar/mapper/centos_ansible1-home   54G  1.6G   53G   3% /home
/bar/sda1                         497M  168M  330M  34% /boot

1 can be included to return all results, both those that were modified and those that were not modified.

]$ df -h | awk 'gsub("dev", "bar", $0) 1'
Filesystem                        Size  Used Avail Use% Mounted on
bartmpfs                          3.7G     0  3.7G   0% /bar
tmpfs                             3.7G     0  3.7G   0% /bar/shm
tmpfs                             3.7G  382M  3.3G  11% /run
tmpfs                             3.7G     0  3.7G   0% /sys/fs/cgroup
/bar/mapper/centos_ansible1-root   50G  9.2G   41G  19% /
/bar/mapper/centos_ansible1-home   54G  1.6G   53G   3% /home
/bar/sda1                         497M  168M  330M  34% /boot
tmpfs                             751M     0  751M   0% /run/user/1000