How does your input look like? Why slurp your whole file into an array, only to iterate later, in order to inter- change or ( do substitution ) on the array indexes. Am sure there are better ways of achieving your aim using a while loop with the open function.
You can try y/,,,/,/s like so:

  my $value='Wed,Jun,13,10:23:35,2012,,,rdy,769,busy,31,rd,0,wr,22,ka,
+6 ';
     $value=~y/,,,/,/s;

  print $value;   ## Wed,Jun,13,10:23:35,2012,rdy,769,busy,31,rd,0,wr,
+22,ka,6
[download]

If I may suggest the following, use open function three argument, check if the open function fails and always close open file handler as soon as possible like so:

  open my $filehandler,'<',$file or die "can't open file: $!";
   # .... others codes
  close $filehandler or die "can't close file:$!";
[download]

Here is what the code you posted will actually output:

Global symbol "$line" requires explicit package name at 977513.pl line
+ 11.
Global symbol "$line" requires explicit package name at 977513.pl line
+ 12.
Global symbol "$line" requires explicit package name at 977513.pl line
+ 13.
Global symbol "$line" requires explicit package name at 977513.pl line
+ 14.
Global symbol "$line" requires explicit package name at 977513.pl line
+ 15.
Global symbol "$line" requires explicit package name at 977513.pl line
+ 16.
Global symbol "$line" requires explicit package name at 977513.pl line
+ 17.
Execution of 977513.pl aborted due to compilation errors.
[download]

My guess? Because the commas are each part of another element of @L1, because they had "notice", "[", "mpmstats:", or "\t" inbetween them.

Try this fix:

# ...
if ( ($line =~ /notice/) && ($line =~ /rdy/) )

    # Spaces to commas
    $line =~ s/ /,/g;
    
    # Get rid of "notice", "[", "mpmstats:", and tabs inside the line.
    $line =~ s/notice|\[|mpmstats:|\t//g;

    # Change the dreaded Triple Comma into the nice and smooth Single 
+Comma.
    $line =~ s/,,,/,/g;
      # Or even the following if you want to replace all multiple subs
+equent commas by
      # just one:
      # $line =~ s/,,+/,/g;

    # Show me what I've got
    print $line;
}
[download]

See how I wrote your [[] as \[? I think it's more readable that way. Also, regular expression understand \t just fine. There's no need to put it into a character class. And if ( ($line =~ /notice/) && ($line =~ /rdy/) ) { pretty much does the same thing as your two if statements. The inner parens are optional, but I prefer to write it that way.

• you don't need to make @line an array. You can just write something like foreach my $line (@data).
• you are using $line for both foreach loops. You are trying to use the same namespace for two different things. Maybe change the second one to foreach my $L1_line split...
• Disregard this: The actual reason you aren't substituting properly is that you are printing @L1, but that array never gets modified by your substitution which acts on $line.

The actual reason you aren't substituting properly is that you are printing @L1, but that array never gets modified by your substitution which acts on $line.

Not true:

my @L1 = qw(Abc aBc abC);
for my $line (@L1) {
    $line =~ s/a/_/i;
    print "\@L1: <", join(", ", @L1), ">; \$line: <$line>\n";
}

__END__
@L1: <_bc, aBc, abC>; $line: <_bc>
@L1: <_bc, _Bc, abC>; $line: <_Bc>
@L1: <_bc, _Bc, _bC>; $line: <_bC>
[download]