CODEWITHSUNDEEP
linuxawk
Riley Willow
Riley Willow

Reputation: 594

trim the output of the command using AWK

I have been working to setup my own DNS server using FreeBSD and I have done almost everything. In order to write a custom script for logging and monitoring different domains in the zones, I am using host -a -l mydomain.com command, that displays all the sub domains in the give domain. I am writing the output of this command to a file using > output. The problem is, the output has some other data as well which is of no interest for me.

The output of this command have following sections,

;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 6149
;; flags: qr aa ra; QUERY: 1, ANSWER: 45, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;unix.                          IN      AXFR

;; ANSWER SECTION:
abc.unix
xyz.unix

Received 1108 bytes from 192.168.82.1#53 in 0 ms

Is there anyway that I can simply read write the section between ;; ANSWER SECTION: and Received 1108 bytes from 192.168.82.1#53 in 0 ms to a file using simple terminal command ? I am reading about awk but I don't understand how I can use it to print only certain part of the output of the command.

Expected output

abc.unix
xyz.unix

Thank you

Upvotes: 2

Views: 444

Answers (3)

sjsam
sjsam

Reputation: 21965

This is a simple requirement where sed could be used

hostname -a -l mydomain.com | sed -n '/;; ANSWER/{n;N;p;q}'

would do it.

Output 

abc.unix
xyz.unix

The q command is handy as it will stop processing once the required portion is printed.

Upvotes: 2

Jay jargot
Jay jargot

Reputation: 2868

Your are right the awk command is very handy for this situation.

Give a try to this:

awk '/^;; ANSWER SECTION:/ , /^Received [0-9]* bytes from [0-9.#]* in [0-9.] ms/ {if ($0 !~ /^;; ANSWER SECTION:/ && $0 !~ /^Received [0-9]* bytes from [0-9.#]* in [0-9.] ms/) print}'

The test:

$ host -a -l mydomain.com | awk '/^;; ANSWER SECTION:/ , /^Received [0-9]* bytes from [0-9.#]* in [0-9.] ms/ {if ($0 !~ /^;; ANSWER SECTION:/ && $0 !~ /^Received [0-9]* bytes from [0-9.#]* in [0-9.] ms/) print}'
abc.unix
xyz.unix

Upvotes: 3

F. Knorr
F. Knorr

Reputation: 3055

In awk you could try this:

awk '$0~/;; ANSWER SECTION:/{a=1; next}
     (a && $0 !~/^Received/)' input

The output for the given example

abc.unix
xyz.unix

The first line sets a=1 when the current line contains ";; ANSWER SECTION:". If a=1 and the current line does not start with "Received", then we print the entire line. (If you want to suppress empty lines, you could add && $0 != "" to (a && $0 !~/^Received/))

In combination with your command host -a -l mydomain.com: 

host -a -l mydomain.com | awk '$0~/;; ANSWER SECTION:/{a=1; next}
         (a && $0 !~/^Received/)'

Upvotes: 3

Related Questions