How to get a next line of a matched word?

Shell Programming and Scripting
1

Hi ,

If I match a word in a file, I want to pick the next line of that matched word.

My file is a below format-

The ntrag trace has auditError
Line5005 contains transaction Ntrag data

------------

Here I wanted if I match a word auditError, I need to get the next line
"Line5005 contains transaction Ntrag data"

2

I don't know what utility you are using, but with awk, you can do something like this:

awk '/auditError/{getline; print}' inputfile
3

if you have GNU grep, you can use the -A switch

4

Thanks blowtorch.

awk '/auditError/{getline; print}' inputfile this has helped me a lot.

But how if I wanted to print the next two lines of matched word.

5

Muktesh,
Try to always specify the entire request at once.
Any change in the original specification could change the entire solution.

In any event, here is a flexible and parametized way to do what you want:

#!/bin/ksh
typeset -i mCntLines=0
mFound="N"
mTotLines=2
mPattern='YourPattern'
while read mLine
do
  mCntGrep=`echo ${mLine} | grep -c "${mPattern}"`
  if [ ${mCntGrep} = "1" ]; then
    mCntLines=0
    mFound="Y"
    continue
  fi
  if [ "$mFound" = "Y" -a ${mCntLines} -lt ${mTotLines} ]; then
    echo ${mLine}
    mCntLines=${mCntLines}+1
  fi
done < input_file

Where:
mPattern --> Your search pattern.
mTotLines --> Number of lines to print after pattern line.

6

A sed solution:

sed -n '/auditError/ {n;p;}' file
7

line_no=`cat <file-name> | grep -c <pattern-name>`
req_line=`cat <file-name> | awk '{NE=$line_no-1}'`

PS: Not tested the above one. Pls correct me if I m wrong

8

Oops.. its NR..not NE

9

Please ignore my reply.

grep -c returns the no of times, the string appears in file
grep -n returns the line no of the pattern.
awk '{NR=$line_no+1}' returns the line next to the line of pattern

With the above combination, v can get the desired output

10

Inside the braces, use another 'getline; print' pair. IMO, the simplest way you could do it.

11

nskworld,
testing a solution on a sample file - might be a good idea! :wink: