Wiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== AWK - AWK Patterns ======

Assuming a file exists with the following contents:

<file>
10   Peter     Terence   Roux        45
11   Virginia  Genevieve  Roux        45
12   Felix     Devon     Roux         5
13   David     Bruce     Stevenson   48
14   Bob       James     Smith       16
48   Adam      Winter    Ridley      23
</file>

----

===== Matching Patterns =====

To only print out lines which **match** a specific search criteria:

<code awk>
awk '/45/ {print $2,$4}' test.txt
</code>

returns:

<code>
Peter Roux
Virginia Roux
</code>

<WRAP info>
**NOTE:**  This only prints out the lines from the file which contained the number 45.
</WRAP>


----

===== Multi-line Output with Patterns =====

To output onto multiple lines:

<code awk>
awk '/45/ {print $2,$3,$4 ; print $1, $5}' test.txt
</code>

returns:

<code>
Peter Terence Roux
10 45
Virginia Genevieve Roux
11 45
</code>

<WRAP info>
**NOTE:**  This only prints out the lines from the file which contained the number 45.

  * But also prints out a 2nd line because of the use of the semicolon.

</WRAP>

<WRAP info>
**NOTE:**  A semicolon does not have to be used to allow using multiple commands.

For instance issue the following command instead:

<code awk>
awk '/45/ {print $2,$3,$4} {print $1, $5}' test.txt
</code>

returns:

<code>
Peter Terence Roux
10 45
Virginia Genevieve Roux
11 45
12 5
13 48
14 16
48 23
</code>

This time an altogether different result is produced, than when a semicolon was used.

The reason for this is that the pattern that is used, **/45/** is only applied to the command immediately following it.  In this example, this was only against fields $2, $3 and $4.

The 2nd print statement is not affected by the pattern and would apply to all fields in the input file.
</WRAP>

----

===== Match using Multiple Patterns =====

To search for more than one pattern match at a time, issue the following command:

<code awk>
awk '/45|48/ {print $2,$4}' test.txt
</code>

returns:

<code>
Peter Roux
Virginia Roux
David Stevenson
Adam Ridley
</code>

<WRAP info>
**NOTE:**  The pipe symbol **|** is used to provide multiple patterns.
</WRAP>


----

===== Match against specific field =====

To restrict the match to only a specific field:

<code awk>
awk '$5 ~ /45|48/ {print $2,$4}' test.txt
</code>

returns:

<code>
Peter Roux
Virginia Roux
David Stevenson
</code>

<WRAP info>
**NOTE:**  Adam Ridley is not displayed even though the source file has a 48 as part of his record.

  * The tilda ~ ties the match to a specific field.

</WRAP>

----

===== Match not against a specific field =====

To have the match be the opposite of matching a specific field:

<code awk>
awk '$5 !~ /45|48/ {print $2,$4}' test.txt
</code>

returns:

<code>
Felix Roux
Bob Smith
Adam Ridley
</code>

<WRAP info>
**NOTE:**  The **exclamation mark** and **tilda** (!~) together informs to not match to a specific field.
</WRAP>


----

===== Match in a Range =====

To have the match be the opposite of matching a specific field:

<code awk>
awk '/4[2-8]/ {print $2,$4}' test.txt
</code>

returns:

<code>
Peter Roux
Virginia Roux
David Stevenson
Adam Ridley
</code>

<WRAP info>
**NOTE:**  This finds all fields with values between 42 and 48.
</WRAP>


----