First Match the string pattern from a row then replace spacific value from just after row in unix

Question

I have below content in my parameter file.

[s_m_imgpstna]
$DBConnection1=Insight_QAT331_DB

[s_m_drcgctln]
$DBConnection1=Insight_QAT291_DB

[s_m_ioxgciln]
$DBConnection1=Insight_QAT291_DB

[s_m_reddecna]
$DBConnection1=Insight_QAT291_DB

[s_m_cd2gcpna]
$DBConnection1=Insight_QAT291_DB

[s_m_cd2gcpna]
$DBConnection1=Insight_QAT291_DB

[s_m_cd2gctna]
$DBConnection1=Insight_QAT291_DB

[s_m_g92gctsp]
$DBConnection1=Insight_QAT291_DB

[s_m_os2gcceu]
$DBConnection1=Insight_QAT291_DB

[s_m_ccpgctdf]
$DBConnection1=Insight_QAT291_DB

[s_m_ew4gciln]
$DBConnection1=Insight_QAT307_DB

[s_m_dr6gctln]
$DBConnection1=Insight_QAT291_DB

[s_m_dr2gctln] 
$DBConnection1=Insight_QAT291_DB

[s_m_dr4gctln]
$DBConnection1=Insight_QAT291_DB

Now I want to search for the pattern "cd2gcpna" which has two occurrence in file

[s_m_cd2gcpna]
$DBConnection1=Insight_QAT291_DB

[s_m_cd2gctna]
$DBConnection1=Insight_QAT291_DB

Now I want to replace string which is just below of that([s_m_cd2gctna]) row for all occurrences.

$DBConnection1=Insight_**QAT291**_DB

replace by

$DBConnection1=Insight_**QAT308**_DB

format your question, see https://stackoverflow.com/editing-help ... and add the code you tried to solve this — Sundeep, Oct 27 '17 at 12:59
@Amol, please spend your some of the time for re-phrasing your question and use code tags for sample(s) inputs and outputs too. — RavinderSingh13, Oct 27 '17 at 13:09
@amol, I reformatted your question. Fix it if I broke anything. — glenn jackman, Oct 27 '17 at 13:59

Marc Lambrichs · Answer 1 · 2017-10-27T14:54:51.447

0

awk solution:

$ awk '/cd2gcpna/{p=1;print;next} \
   p{sub(/291/,"308");p=0}1' file

If you want to change your file inplace, with GNU awk you can do this:

awk -i inplace '/cd2gcpna/{p=1;print;next} \
   p{sub(/291/,"308");p=0}1' file

And with params:

$ awk -v pat="cd2gcpna" -v val1=291 -v val2=308 '$0~pat{p=1;print;next} \
    p{sub(val1,val2);p=0}1' file

From a shell script:

$ cat test.sh
#!/bin/bash

pat=$1
val1=$2
val2=$3
file=$4

awk -v pat="$1" -v val1=$2 -v val2=$3 '$0~pat{p=1;print;next} \
    p{sub(val1,val2);p=0}1' $file

And call it with:

$ ./test.sh cd2gcpna 291 308 file

edited Oct 27 '17 at 14:54

answered Oct 27 '17 at 14:14

Marc Lambrichs

2,864
2
13
14

It is working for me but I want to make it parameterized. I want to pass cd2gcpna and 308 as a argument to script. – Amol Murkute Oct 27 '17 at 14:29
@AmolMurkute see https://stackoverflow.com/questions/19075671/how-do-i-use-shell-variables-in-an-awk-script – Sundeep Oct 27 '17 at 14:33
1

@AmolMurkute example with params added. – Marc Lambrichs Oct 27 '17 at 14:38
Thanks Marc...How I can wrap up AWK in shell script. – Amol Murkute Oct 27 '17 at 14:47
@AmolMurkute Added it to answer. – Marc Lambrichs Oct 27 '17 at 14:55
Thanks a lot for your effort @Marc. Really pleased with your help. Again thanks for Stackoverflow team for this wonderful platform. – Amol Murkute Oct 27 '17 at 15:32

score 0 · Answer 2 · answered Oct 29 '17 at 16:42

sed method:

sed -i '/cd2gcpna/{n;s/QAT291/QAT308/}' file

To wrap that in a function, (easier to test than a script, the code's virtually the same either way):

# usage: rep_after flagstring fromstring tostring file
rep_after() { sed -i '/'"$1"'/{n;s/'"$2"'/'"$3"'/}' "$4" ; }

And call it like so:

rep_after  cd2gcpna  QAT291 QAT308  file

First Match the string pattern from a row then replace spacific value from just after row in unix

2 Answers2