CODEWITHSUNDEEP
perlsed
Igor Boguslavetz
Igor Boguslavetz

Reputation: 63

I want to replace multy line string with sed or perl

I want to delete:

         ,
.rdt(rdt)

from text:

.colldisn(1'b0),
.rdt(rdt)

while .colldisn could be various texts.

So, I have a lot of file with this substring and I want to delete it.

Upvotes: 0

Views: 165

Answers (4)

JJoao
JJoao

Reputation: 5347

myscript:

#!/usr/bin/perl -i.bak

undef $/;                   # get all the text
while(<>){ 
  s/,\n\.rdt\(rdt\)//g; 
  print;
}

the "-i.bak" changes de original "file" but creates a "file.bak" copy. This way you can use 

myscript *.txt

to change all txt files in current dir.
A similar switch exists for sed.

Upvotes: 0

Para&#237;so
Para&#237;so

Reputation: 384

sed -e ':a;N;$!ba;s/,\n.rdt(rdt)//g' tst.txt

Where the tst.txt is you file, to write to the same file don't forget to add the -i option.

The :a;N;$!ba; will let you to accomplish what you want, witch is to compare several lines.

Upvotes: 2

potong
potong

Reputation: 58430

This might work for you (GNU sed):

sed '$!N;s/,\s*\n\s*\.rdt(rdt)//;P;D' file

If you want to preserve the space and newline:

sed '$!N;s/,\(\s*\n\s*\)\.rdt(rdt)/\1/;P;D' file

Upvotes: 2

choroba
choroba

Reputation: 241918

Should be easy in Perl: just keep the previous line in a variable. If you encounter .rdt(rdt), check that the previous line ends in a comma. If so, print the previous line without the comma and forget the current line, otherwise print the whole previous line. Do not forget to output the last line at the end of the file.

Upvotes: 0

Related Questions