CODEWITHSUNDEEP
bashshellsedawk
Gil
Gil

Reputation: 1538

Append to the previous line for a match

can I use sed or awk to append to the previous line if a match is found ?

I have a file which has the format :

          INT32
          FSHL (const TP  Buffer)
             {
          INT32
          FSHL_lm (const TP  Buffer) 
          { WORD32 ugo = 0; ...

What I am trying to do is scan for independant open braces {and append it to the previous non-blank line .The match should not occur for an open brace appended by anything in the same line .

The expected output :

           INT32
          FSHL (const TP  Buffer){
          INT32
          FSHL_lm (const TP  Buffer) 
          { WORD32 ugo = 0; ...

Thanks for the replies .

Upvotes: 3

Views: 4535

Answers (4)

Shyam Gupta
Shyam Gupta

Reputation: 513

[shyam@localhost ~]$ perl -lne 's/^/\n/ if $.>1 && /^\d+/; printf "%s",$_' appendDateText.txt

that will work

i/p:

06/12/2016 20:30 Test Test Test

TestTest

06/12/2019 20:30 abbs abcbcb abcbc

06/11/2016 20:30 test test

i123312331233123312331233123312331233123312331233Test

06/12/2016 20:30 abc

o/p:

06/12/2016 20:30 Test Test TestTestTest

06/12/2019 20:30 abbs abcbcb abcbc

06/11/2016 20:30 test ##testi123312331233123312331233123312331233123312331233Test

Upvotes: 0

potong
potong

Reputation: 58351

This might work for you (GNU sed):

sed '$!N;s/\n\s*{\s*$/{/;P;D' file

Explanation:

  • $!N unless the last line append the next line to the pattern space.
  • s/\n\s*{\s*$/{/ replace a linefeed followed by no or any amount of white space followed by an opening curly brace followed by no or any amount of white space to the end of the string, by an opening curly brace.
  • P print upto and including the first newline.
  • D delete upto and including the first newline (if so do not start a new cycle).

Upvotes: 7

Steve
Steve

Reputation: 54392

One way using awk:

awk '!(NF == 1 && $1 == "{") { if (line) print line; line = $0; next; } { sub(/^[ \t]+/, "", $0); line = line $0; } END { print line }' file.txt

Or broken out on multiple lines:

!(NF == 1 && $1 == "{") { 
    if (line) print line
    line = $0
    next
}

{
    sub(/^[ \t]+/, "", $0)
    line = line $0
}

END {
    print line
}

Results:

INT32
          FSHL (const TP  Buffer){
          INT32
          FSHL_lm (const TP  Buffer) 
          { WORD32 ugo = 0; ...

HTH

Upvotes: 2

Birei
Birei

Reputation: 36252

One way using perl. I read all file in slurp mode and use a regular expression to search lines with only a curly brace and remove its leading spaces.

perl -ne '
    do {
        local $/ = undef;
        $data = <>;
    };
    $data =~ s/\n^\s*(\{\s*)$/\1/mg;
    print $data
' infile

Assuming infile with the content of the question, output will be:

FSHL (const TP  Buffer){
INT32
FSHL_lm (const TP  Buffer)
{ WORD32 ugo = 0; ...

Upvotes: 2

Related Questions