合并多个代码块

Question

I'm programming in JS. I want to but multiple same "code blocks" into one code block using regex. Please note second block is another distinct block.

Example Data

BEFORE ENTRY
  Im on the top
  block
END

RANDOM NAME
  HAHA I MAKE A PROBLEM 
END

BEFORE ENTRY
  some randon
  strings with also 1 numbers 1
END

BEFORE ENTRY
  more strings
  very cool
END

BEFORE ENTRY
  stringi
  pizza
END

Desired Output

RANDOM NAME
  HAHA I MAKE A PROBLEM 
END

BEFORE ENTRY
  Im on the top
  block

  some randon
  strings with also 1 numbers 1

  more strings
  very cool

  stringi
  pizza
END

(i don't care whether RANDOM NAME is on top or on the bottom.)

My failed attempts

1. /(?<=BEFORE ENTRY[^]*?)END[^]*?BEFORE ENTRY/g replace with "" (works if BEFORE ENTRY blocks are consequent)

2. /(BEFORE ENTRY)([^]*?)(END)/g playing around with \\1 , \\2, \\3 . Would work if I could somehow merge the capturing groups.

Edit regarding @Archer comment: Data-type is a multiline string inside output: .

arr = [{id:"1", output:">>EXAMPLE DATA<<"},...,{id:"n", output:">>Some similar Data<<"}]

Answer 1

EDIT: sorry i didn't see the perl tag.. sould i remove the answer? the logic is still re-usable..

If you want to use regex, i would use the second one which is simpler, and capturing groups. In addition, i made the solution generic with an array of keys.

Note that using a function for the replace , in addition to be able to control that the result is inserted only for the first match, also disables the special replacements such as $$ that could be in the result.

You're not obliged to store results in an array, you could use a local var for result.

 var codeInput = `BEFORE ENTRY Im on the top block END RANDOM NAME HAHA I MAKE A PROBLEM END BEFORE ENTRY some randon strings with also 1 numbers 1 END BEFORE ENTRY more strings very cool END BEFORE ENTRY stringi pizza END`, ending = 'END', keys = ['BEFORE ENTRY', 'RANDOM NAME'], i, len = keys.length, reg, match, results = {}, inserted; for(i = 0; i < len; i++){ //getting all the contents reg = new RegExp(keys[i] + '([^]*?)' + ending, 'g'); while(match = reg.exec(codeInput)){ if(!results[keys[i]]){ results[keys[i]] = keys[i]; } results[keys[i]] += match[1]; } //inserting the result: replace the first one by result, remove others if(results[keys[i]]){ results[keys[i]] += ending + '\\n'; inserted = false; codeInput = codeInput.replace(reg , function(){ if(!inserted){ inserted = true; return results[keys[i]]; } return ''; }); } } document.getElementById('result').innerHTML = codeInput; 

 <pre id="result"><pre> 

Answer 2

Using perl

use warnings;
use strict;
my $f = 0;
my $before_entry;
while(<DATA>)
{
    $f = 1 ,  next if(/^BEFORE ENTRY/);
    print and  $f = 0 ,  if(/^END/ && $f == 2);
    $f = 0 if(/^END/);
    print and next if($f == 2);
    $f = 2  if(/^(?!BEFORE ENTRY|\s+)/ && $f != 1);
    $before_entry.=$_ if ($f == 1);
}

print "BEFORE ENTRY\n$before_entry\nEND\n"

__DATA__
BEFORE ENTRY
Im on the top
block
END

RANDOM NAME
HAHA I MAKE A PROBLEM 
END

BEFORE ENTRY
some randon
strings with also 1 numbers 1
END

BEFORE ENTRY
more strings
very cool
END

BEFORE ENTRY
stringi
pizza
END