CODEWITHSUNDEEP
awkmatch
CodeDevotion
CodeDevotion

Reputation: 307

How to print only the columns next to a match using AWK

Below is the sample file where I would like to print values in the columns next to the match..

$> cat sample.txt
DATABASE ORACLE, DB_TYPE Relational, OS OEL 5
VERSION 12.2.0.1

DATABASE MongoDB, DB_TYPE NOSQL, OS RedHat 7
VERSION 3.2
DATABASE MySQL, DB_TYPE Relational
VERSION 5.5

The lines starting with DATABASE and VERSION I joined them as following:

$> grep -i "^DATABASE\|^VERSION" sample.txt | sed 'N;s/,/\t/;s/\n/\t/'
DATABASE ORACLE  DB_TYPE Relational, OS OEL 5   VERSION 12.2.0.1
DATABASE MongoDB         DB_TYPE NOSQL, OS RedHat 7     VERSION 3.2
DATABASE MySQL   DB_TYPE Relational     VERSION 5.5

But now I would like to print only the columns next to DATABASE, DB_TYPE and VERSION and I did some thing like this.

$> grep -i "^DATABASE\|^VERSION" sample.txt | sed 'N;s/,/\t/;s/\n/\t/' |awk -v val1='DATABASE' -v val2='DB_TYPE' -v val3='VERSION' -F ' ' '{for (i=1; i<=NF; i++) if ($i==val1 || $i==val2 || $i==val3) {n=i+1; print $n} }' 
ORACLE
Relational,
12.2.0.1
MongoDB
NOSQL,
3.2
MySQL
Relational
5.5

But I am expecting some thing like this

ORACLE      Relational      12.2.0.1
MongoDB     NOSQL           3.2
MySQL       Relational      5.5

Upvotes: 0

Views: 151

Answers (2)

karakfa
karakfa

Reputation: 67497

awk to the rescue!

$ tr ',' '\n' <file | 
  awk '/DATABASE|DB_TYPE|VERSION/{printf "%s",$2 OFS; if(++c%3==0)print""}'

ORACLE Relational 12.2.0.1
MongoDB NOSQL 3.2
MySQL Relational 5.5

you can set OFS to tab or pipe to column -t for tabular formatting.

Upvotes: 2

LHWizard
LHWizard

Reputation: 2389

For your example, this will work: grep -i "^DATABASE\|^VERSION" sample.txt | sed 'N;s/,/\t/;s/\n/\t/' | awk 'BEGIN { OFS = " " } /DATABASE/ {print $2,$4,$NF}'

Searches for occurrences of DATABASE, output file separator (OFS) is tab - figure out how to insert a tab between the quotes for your particular OS

Upvotes: 0

Related Questions