Gnu Awk: One-line Examples

Print Nth column

By default, awk uses one or more spaces as the default column delimiter:

• print the second column:

  $ echo "foo   bar baz" | awk '{print $2}'
  bar
  

Print where line matches

• print Nth column when the line contains a string (anywhere):

  create a file input1.txt like this:

  foo 1
  bar 2
  baz 3
  quf 4
  

  and pipe it to the awk command:

  cat input1.txt | awk '/fo/ {print $2}'
  1
  

Print where line matches regex

See all regexp operators here: Regexp operators

• Print Nth column for lines starting with "f":

  # input1.txt
  foo 1
  bar 2
  baz 3
  quf 4
  

  $ cat input1.txt | awk '$0 ~ /^f/ {print $2}'
  1
  

Print where line does not match regex

Use !~ instead:

• Print Nth column for lines NOT starting with "f"

  # input1.txt
  foo 1
  bar 2
  baz 3
  quf 4
  

  $ cat input1.txt | awk '$0 !~ /^f/ {print $2}'
  2
  3
  4
  

Custom delimiter

Use -Fdelimiter

$ echo "foo,bar,baz" | awk -F, '{print $2}'
bar

Regexp delimiter

• split on two or more dashes (note that ba-z was not split)

  $ echo "foo---bar--ba-z" | awk -F '-{2,}' '{print $3}'
  ba-z
  

• split on two or more occurrences of X or Y:

  $ echo "fooXXbarYYbXaz" | awk -F '[X|Y]{2,}' '{print $3}'
  bXaz
  

Print all columns from Nth to last

• Example: Print all columns from 4th to last

  $ echo 'foo,bar,baz,quux,bax' | awk -F, '{ print substr($0, index($0,$4)) }'
  quux,bax
  

Use shell variables

Use -v modifier.

# input1.txt
foo 1
bar 2
baz 3
quf 4

In this example: print the contents of myvar plus the second column for every line:

$ myvar='foobar'
$ cat input1.txt | awk -v var="$myvar" '{print var " " $2}'
foobar 1
foobar 2
foobar 3
foobar 4

References

• The GNU Awk User's Guide