AWK split input by…

awk is designed to process text for reports. Let’s think about different methods of splitting text.

Default behaviour

Split by white characters and take second part

echo 'what is first part what IS second part' | awk '{print $2}'
is

Awk prints second word is

Split by string - it is regex but lets keep it step by step

To set splitting string use -F parameter. Lets split by ‘is’ (small letters)

echo 'what is first part what IS second part' | awk -F 'is' '{print $2}'
 first part what IS second part

When field separator was set to is (case sensitive) result was: ` first part what IS second part`

Now let’s split by IS (big letters).

echo 'what is first part what IS second part' | awk -F 'IS' '{print $2}'
 second part

Right now we only get ` second part` as second field.

Split by regex

Now let’s split by is but, case insensitive.

Command parameter

echo 'what is first part what IS second part' | awk -F '[Ii][Ss]' '{print $2}'
 first part what

Due to used regex, whole line was splitted into 3 fields. Where second field was: ` first part what . Both IS and is` was interpreted as field separator

Syntax

The same but lets use syntax instead of parameter

echo 'what is first part what IS second part' | awk 'BEGIN  {FS="[Ii][Ss]"}; {print $2}'
 first part what

Split by position of string and format

awk allows also ussage of substr method and can find possition of string.

echo 'Lay me down Let the only sound Be the overflow Pockets full of stones' | awk '{p=index($0,"Be");print "First part of line: \"", substr($0,0,p-1) "\",\nSecond part of line: \"", substr($0,p), "\""}'
First part of line: " Lay me down Let the only sound ",
Second part of line: " Be the overflow Pockets full of stones "