You need to (find out and) tell Perl what encoding your i-with-double-dots letter is in. Then you need to Encode::decode it and then pass it to HTML::Entities for output, or as an alternative, tell the mail client in the headers what output encoding your mail uses.

Likely of help is perlunitut.

Thanks

perlunitut is of great help.

You aren't decoding your source string first.

use strict;
use warnings;

use HTML::Entities;
use Encode;

my $TestStr = 'ï';
print encode_entities($TestStr) . "\n";
print encode_entities(decode ('utf-8', $TestStr)) . "\n";
[download]

Have a read of perlunitut if you haven't already. It'll explain the basics.

decodeing from utf-8 only helps if the source code is actually encoded as UTF-8. This may or may not be the case.

At least according to Wikipedia, likely encodings are also ISO 8859-3, ISO 8859-9 or Windows-1254, if guessing that &iuml is supposed to depict a Turkish letter.

Indeed so - it is nigh on impossible to determine the encoding of a document from a single character, so the actual encoding of the source will only be known by gepebril69. UTF-8 seemed a reasonable first guess in this instance and it does produce the desired output for that one character.

I've checked the template file and it is

text/html; charset=utf-8

Now I understand why I had a similar problem in the past with parsing files. Perl don't seem to auto detect this formatting. It will have a logical reason I guess

Thanks hippo

That is very much explaining, so in my case when I want to define unsafe characters I have to use a similar methode.

my $UnsafeChar = 'ïé';
print encode_entities(decode ('utf-8', $TestStr), decode ('utf-8', $Un
+safeChar)) . "\n";
[download]

Also, note that your code needs to know its own encoding.

use strict;
use HTML::Entities;
{
    my $TestStr = 'ï';
    print encode_entities($TestStr), $/;
}
{
    use utf8;
    my $TestStr = 'ï';
    print encode_entities($TestStr), $/;
}

__END__
ï
ï
[download]

In addition, also be sure that when you send the HTML to the browser, you tell the browser what character-set (language ...) is the proper one to use for this content.   e.g. http://stackoverflow.com/questions/6157485/content-language-and-accept-language.   Browsers are reasonably-smart about such things these days, except when they are stupid.

Thanks!

I had to change this line of code to make my mail look OK on the webbrowsers I've tested

$Mail{'content-type'}    = 'text/html; charset="iso-8859-1"'; #Old, no
+t correct 
$Mail{'content-type'}    = 'text/html; charset="utf-8"';
[download]

Thanks all for aiming me to the right direction