CODEWITHSUNDEEP
regexawk
Sigularity
Sigularity

Reputation: 967

How can I find the match string in pattern(AWK)?

I have a text file with the pattern starting '-- Host' and ending ';', and this pattern occurs so many times in the file as follow:

-- Host (first) Sid (queen1)

-- prince princess#/king 1/1

;

-- Host (first) Sid (queen2)

-- prince princess#/king 2/2

;

-- Host (first) Sid (queen3)

-- dir princess#/king 3/3

;

My desired result is as follow:

queen1 1/1

queen2 2/2

queen3 3/3

With AWK, I think I can specify the block including this patten like below.However, I got error in putting the second pattern in pattern.

Can you help me accomplish this requirement? Thanks!

BEGIN {
}

/-- Host/,/;/ { /prince/ { print $3 } }

END

Upvotes: 2

Views: 67

Answers (1)

John1024
John1024

Reputation: 113984

$ awk -F'[ ()]+' '/-- Host/,/;/ { if (/-- Host/) printf "%s ",$5; else if (/prince/) print $4;}' file
queen1 1/1
queen2 2/2
queen3 3/3

How it works

  • -F'[ ()]+'

    This tells awk to use any combination of any number of the blanks and parens as the field separator.

  • /-- Host/,/;/

    This selects just groups of lines that start with -- Host and end with ;. The commands which follow in braces are only executed for lines in these groups.

  • { if (/-- Host/) printf "%s ",$5; else if (/prince/) print $4;}

    If the line contains -- Host, then we print its fifth field. Otherwise, if it contains prince, we print its fourth field.

Upvotes: 1

Related Questions