[SOLVED] match source pattern in destination and substitute in destination usin AWK

123raajesh · 09-16-2013, 11:44 AM

Please help with this. I have two files rspfile and myfile as below:

rspfile

#
wsdm_USERID=
#
WASGRP_GROUPID=
#
ND_HOSTNAME=
#
BASE_NODENAME=
#
BASE_HOSTNAME=
#
APP_NAME_1=
#APP_NAME_2=
#APP_NAME_3=
#
CLUSTER_NAME_1=
#CLUSTER_NAME_2=
#CLUSTER_NAME_3=

myfile

wsdm_USERID=500
WASGRP_GROUPID=500
ND_HOSTNAME=abcdefgh.was.tb.dsdc
BASE_NODENAME=abcdefgh
BASE_HOSTNAME=abcdefgh.was.tb.dsdc
APP_NAME_1=CCCCC01DD
APP_NAME_2=CCCCC02DD
CLUSTER_NAME_1=CCC01DD
CLUSTER_NAME_2=CCC02DD

Using awk like to have result as below:

rspfile

#
wsdm_USERID=500
#
WASGRP_GROUPID=500
#
ND_HOSTNAME=abcdefgh.was.tb.dsdc
#
BASE_NODENAME=abcdefgh
#
BASE_HOSTNAME=abcdefgh.was.tb.dsdc
#
APP_NAME_1=CCCCC01DD
APP_NAME_2=CCCCC02DD
#APP_NAME_3=

#
CLUSTER_NAME_1=CCC01DD
CLUSTER_NAME_2=CCC02DD
#CLUSTER_NAME_3=

I tried this using sed but its lengthy.
sed -e '/wsdm_USERID=/ c wsdm_USERID=500' -e '/WASGRP_GROUPID=/ c WASGRP_GROUPID=500' . . . /home/welcome/rspfile

Please suggest thanks in advance.

colucix · 09-16-2013, 02:09 PM

Code:

awk 'BEGIN { FS = OFS = "=" } FNR == NR { _[$1] = $2 } FNR < NR { sub(/^#/,"",$1); if ( $1 in _ ) print $1, _[$1]; else print "#" $0 }' myfile rspfile

The expression FNR == NR ensures that the rule is applied only to the first file (myfile) whereas the expression FNR < NR ensures the rule is applied to the second file. Here is a more readable version of the code:

Code:

awk 'BEGIN {

  FS = OFS = "="

}

FNR == NR {

  _[$1] = $2 

}

FNR < NR {

  sub(/^#/,"",$1)

  if ( $1 in _ )
    print $1, _[$1]
  else
    print "#" $0

}' myfile rspfile

123raajesh · 09-17-2013, 01:05 PM

Hi Colucix,

Thanks it worked for me. Could I know where I can learn AWK?

Like to add more my query.
1) How can I update the result in destination file itself(rspfile).
2) Could it be possible to restrict to update the rspfile, if the file is already updated with values = $2.

I would still be interested to learn awk.

Thank you

colucix · 09-17-2013, 02:34 PM

Quote:

Originally Posted by 123raajesh

1) How can I update the result in destination file itself(rspfile).

Inside the awk program you can redirect the output of the print statements to FILENAME, which is an internal variable containing the name of the file that awk is processing:

Code:

  ...
  if ( $1 in _ )
    print $1, _[$1] > FILENAME
  else
    print "#" $0 > FILENAME
  ...

However this doesn't work for long input files (I don't know what is the limit, anyway). In that case you have to redirect the output to a new file and then rename it as the original input file.

Quote:

2) Could it be possible to restrict to update the rspfile, if the file is already updated with values = $2.

Yes. You have to add a further condition, that is if the second field is not empty, then update the value, otherwise print the line without modification (see the part added in red)

Code:

  ...
  if ( $1 in _ )
    if ( ! $2 )
      print $1, _[$1]
    else
      print
  else
    print "#" $0
  ...

Finally to learn awk I suggest the official GNU Awk User's Guide, where the first 11 chapters are the basics, the rest is somewhat for advanced users.

123raajesh · 09-18-2013, 01:01 PM

Hi colucix,

Thank you very much.

colucix · 09-18-2013, 02:50 PM

You're welcome!

123raajesh · 11-13-2013, 12:13 AM

Hi colucix,

Have a query as u mentioned earlier the limit of input file matters. Now I have same issue, I have a file of 604L, 16481C and the its failing.

Could you please suggest any other way to get the result?

Thanks

Raajesh

zhjim · 11-13-2013, 04:22 AM

You could use split to get smaller files