Oh, one more idea. Could it be that your file contains extra "\r" characters? There shouldn't be any at all. You see, notepad, and anything else using the same textbox control, use CR+CR+LF as a flag for soft wrap. Change anything in your text and they all disappear. That's why it's possible to download a text file from the internet, a HTML file for example, try to edit one tiny thing, and end up with just one huge line for the whole file. Downloaded HTML files often contain two CR characters per LF, and notepad strips them all.

This snippet can delete them all for you, if present:

tr/\r//d;
[download]

It looks fine to me.

$regel = "Username; abcd; Username"; 
$regel =~ s/Username/\nGebruikersnaam/sgm; 
print "(BEGIN)$regel(END)\n";
__END__
output:
(BEGIN)                 # <-- \n here
Gebruikersnaam; abcd;   # <-- and \n here 
Gebruikersnaam(END)
[download]

Maybe the problem is in your output routine. Are you producing HTML? If so, a "\n" doesn't have any effect, unless inside <pre> </pre> tags.

$regel =~ s/Username/\nGebruikersnaam/sgm; </BR>

Do you have a filehandle open named /BR? (Is it even legal in Perl to start a filehandle name with a slash? I've *got* to remember that for future obfuscations...) Also, why are you reading from the filehandle in void context?

All the "Username" 's are changed in "Gebruikersnaam" 's but the "\n" has no effect and is not visible. So it remains one big line.

Where is the output being displayed? In a Windows command prompt box? Is that Win9x/Me's command.com, or the cmd.exe that is used in NT? Or is the output being sent to a web browser, and you forgot to mention that? (That would *sort-of* explain the </BR>, though not fully...) If the output is going to a web browser, then it is normal for newlines to have no effect. Or maybe the output is being sent to a file, which you're opening and looking at in notepad? (That would explain your (otherwise very odd) remark about Notepad not recognising the \r.) You need to tell us where it is that you're looking at this output and seeing it all on one line.

I have tried "\r", "\r\n" and "\n\r" but the "\r" is not recognized by notepad.

Usually, if you're not using binmode, Perl on Win32 will normally automagically insert a proper CRLF pair when you use \n. However, this isn't fully portable; if someone later runs the code on Unix, they'll get just linefeeds. Then again, that is portable, because on Unix linefeeds are what is normally wanted. Unless you're generating a file that will be transported across platforms without undergoing translation. Anyway, it isn't a very common thing, but if for some reason you do have the occasion to need to expressly specifiy a CRLF as such when you'd otherwise get just a linefeed, \r\n is the way to do it. But I very much doubt that's your problem.

If you're outputting to a web browser, try it this way:

Otherwise, you're going to have to tell us more about where your output is going.

$;=sub{$/};@;=map{my($a,$b)=($_,$;);$;=sub{$a.$b->()}}
split//,".rekcah lreP rehtona tsuJ";$\=$ ;->();print$/
[download]

Hallo,

I'm not writing HTML. Just plain old ascii :-)

Thanks for the effort :-)

Convex

If you are losing carriage returns, check that they have not been chomped from string contained in $regel.

If you are generating HTML then you will need to change the substitution to add the <br> tag e.g.
$regel =~ s/Username/<br>\nGebruikersnaam/sgm;

#! /usr/bin/perl -w

use strict;
use warnings;

undef $/;
my $regel = <DATA>;
$regel =~ s/Username/\nGebruikersnaam/sgm;
print "start\n";
print "$regel";
print "end\n";
__DATA__
Username=Joe Smith
Username=Jane Doe

start

Gebruikersnaam=Joe Smith

Gebruikersnaam=Jane Doe
end

Inman

I can't recreate your problem, I get newlines on both win and linux here. But perhaps you could work around it by using $/ - the input record separator instead of specifying \n. Like this:

$regel =~ s[Username][$/Gebruikersnaam]sgm;

Mats

Ok, guy's I found it. It was not the regex but an array slip-up.
I send the string to an array and there it was glue together again.
Silly me. ;-)

Let this be a lesson to me: The mistakes are not alway's
were you assume they are + long live the "perl -d".

Thanks for all the help (I learned something) and sorry for the trouble.
Nylon

undef $/;
$regel = <FH_input>
[download]

undef $/;
$regel = <FH_input>
[download]

All this does is read the entire contents of your input file into $regel in one fell swoop. It should still have the newlines in it that it had originally, and in any case how you do your input shouldn't have any impact on newlines that you explicitely insert in your output. I'd say that's not your problem. Maybe you should post the whole script here.

Maybe Notepad is part of the issue. What happens if you look at the file in another editor (UltraEdit or whatever, or even EDIT.COM) or pull up a command prompt and use the TYPE command to display it?

---

$;=sub{$/};@;=map{my($a,$b)=($_,$;);$;=sub{$a.$b->()}}
split//,".rekcah lreP rehtona tsuJ";$\=$ ;->();print$/
[download]