While finding a solution for How to solve Perl’s `length ‘für’ == 4` for `LC_CTYPE=”en_US.UTF-8″`? I wrote another little test program:

It seems Perl (5.18.2) does not output UTF-8 encoded strings correctly in an UTF-8 Linux (SLES12 SP5) environment when the strings are interpreted as UTF-8.
The basic problem was that (e.g.) string “Gemäß” being read from a file had a length of 7 instead of 5, so I wrote this test program (“length.pl”, the first test is “commented out” in an odd way):

#!/usr/bin/perl
use warnings;
use strict;

=begin debug

use utf8;

print length('Gemäß'), "n";

=end debug

=cut

if (open(my $fh, "<:encoding(UTF-8)", 'length.txt')) {
    while (<$fh>) {
        chomp;
        print length($_), ':', $_, "n";
    }
    close($fh);
} else {
    warn "length.txt: $!n";
}

The input file “length.txt” just contains a single line, like this

> cat length.txt
Gemäß
> hexdump -C length.txt
00000000  47 65 6d c3 a4 c3 9f 0a                           |Gem.....|
00000008
> ./length.pl
5:Gem▒▒
>  locale
LANG=en_US.UTF-8
LC_CTYPE="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_PAPER="en_US.UTF-8"
LC_NAME="en_US.UTF-8"
LC_ADDRESS="en_US.UTF-8"
LC_TELEPHONE="en_US.UTF-8"
LC_MEASUREMENT="en_US.UTF-8"
LC_IDENTIFICATION="en_US.UTF-8"
LC_ALL=
> vi length.pl # remove the ":encoding(UTF-8)" from open
> ./length.pl
7:Gemäß

So the length is correct, but the output on the screen is wrong.
When dropping :encoding(UTF-8) from the open call, then the string length is wrong, but the output is correct.

I’m using an SSH session via PuTTY with setting “Remote character set:” set to “UTF-8” (just in case someone would ask for that).

Obviously I’d like to have both (for correct UTF-8 input), the string length and correct text output.