CODEWITHSUNDEEP
linuxawksedgrep
Hammad Ahmed
Hammad Ahmed

Reputation: 75

searching a multiple line pattern using grep regex

I am relatively new to linux I want to search a pattern in a file which starts with "Leonard is" and ends on "champion"

Also this pattern might be placed in multiple lines

the input file(input.txt) may look like:

1 rabbit eats carrot Leonard is a champion 
2 loin is the king of 
3 jungle Leonard is a 
4 Champion 
5 Leonard is An exemplary 
6 Champion

i would want to have all the occurrences of my pattern ignoring all the other characters other than the pattern in the output file:

1 Leonard is a champion
3 Leonard is a
4 Champion
5 Leonard is An exemplary
6 Champion

i have been very close with the following command:

cat input.txt | grep -ioE "Leonard.*Champion$"

as this command only returns 

1 Leonard is a champion

ignoring all the patterns occurring in multiple line

if any other approach of searching other than grep is useful kindly let me know Thanks!!

Upvotes: 1

Views: 1382

Answers (3)

choroba
choroba

Reputation: 242373

Perl to the rescue:

perl -l -0777 -e 'print for <> =~ /(.*Leonard(?s:.*?)[Cc]hampion.*)/g' -- input.txt
  • -l adds newlines to prints
  • -0777 reads the whole file instead of processing it line by line
  • the diamond operator <> reads the input
  • .*? is like .*, i.e. it matches anything, but the ? means the shortest possible match is enough. That prevents the regex from matching everything between the first Leonard and last Champion.
  • . in a regex doesn't match a newline normally, but it does with the s modifier. (?s:.*?) localizes the changed behaviour, so other dots still don't match newlines.

Upvotes: 1

Shani C
Shani C

Reputation: 166

The "." is referenced as "any character except new line", therefore, what you're trying to achieve with . is not possible, I suggest using \s with an addition of * or + as well (as suggested above), but need to find out how to implement it with the "grep" reg expression. There are also nice tools for regex testing - https://regexr.com/ for example.

Upvotes: 0

Tomas Votruba
Tomas Votruba

Reputation: 24298

You're looking for \s which stands for whitespace. + stands for one or more

Pattern: Leonard is a\s+Champion

See: https://regex101.com/r/qiNXhf/1

I use this tool with 0 knowledge of regex in my mind, and it helps me a lot. See the notes on the right bottom, where all these signs are explained.

enter image description here

Upvotes: 0

Related Questions