G'day Skeeve,

If you have Unicode characters in your source code, you'll need to use the utf8 pragma. Also look at the raw option for Net::LDAP's new() constructor and search() method. If that's of no help, you'll need to show your code: I'm completely guessing here.

Thanks for guessing ;)

I didn't show the code as there is really nothing special about it - at least as far as I can see. But I will try to shrink it down to the bare minimum and post it later.

What I forgot to mention is: I already tried the "raw" option, but it did not make any difference.

---

s$$([},&%#}/&/]+}%&{})*;#$&&s&&$^X.($'^"%]=\&(|?*{%
+.+=%;.#_}\&"^"-+%*).}%:##%}={~=~:.")&e&&s""`$''`"e

Just for completeness. I think I found the reason for the problem. See below.

So here is a small example of my script having problems with utf-8.

use strict;
use warnings;
use Net::LDAP;
# binmode STDOUT, ':utf8'; # <- If this is active, output is corrupted

my $ldap_host= '...myhost...';
my $ldap_port= ...myport...;
my $ldap_user= '...myuser...';
my $ldap_pass= '...mypass...';

my $ldap = Net::LDAP->new( $ldap_host,
    port => $ldap_port,
    raw => qr/(?i:^jpegPhoto|;binary)/,
) or die "$@";

my $mesg = $ldap->bind( $ldap_user,
    password => $ldap_pass,
);
if ($mesg->is_error) {
    die $mesg->error_text;
}

$mesg= $ldap->search(
    base => '...mybase...',
    filter => 'uid=...myuser...',
    attrs => [ qw/ cn / ]
);
if ($mesg->is_error) {
    die $mesg->error_text;
}
my $result= $mesg->as_struct;
while (my($id, $data)= each %$result) {
    print $data->{'cn'}->[0], "\n";
}

$ldap->unbind;
[download]

Output is okay without the binmode STDOUT, ':utf8';:

$ ./ldap-test.pl         
Jörg
$ ./ldap-test.pl | od -xc
0000000    f64a    6772    000a
           J 366   r   g  \n
0000005
[download]

$ ./ldap-test.pl 
Jörg
$ ./ldap-test.pl | od -xc
0000000    c34a    72b6    0a67
           J 303 266   r   g  \n
0000006
[download]

Assumption: The UTF-8 Conversion is okay, but my Cygwin rxvt Terminal is too dumb. It's not expecting UTF-8 but some WinDOS encodeing.

When I switch to an xterm and enable UTF-8 encoding there, output seems okay.