NAME

Net::LibIDN - Perl bindings for GNU Libidn

SYNOPSIS

use Net::LibIDN ':all';

idn_to_ascii("Räksmörgås.Josefßon.ORG") eq
  idn_to_ascii(idn_to_unicode("xn--rksmrgs-5wao1o.josefsson.org"));

idn_prep_name("LibÜDN") eq "libüdn";

idn_punycode_encode("kistenmöhre") eq
  idn_punycode_encode(idn_punycode_decode("kistenmhre-kcb"));
#IF_TLD

my $errpos;
tld_check("mèrle.se", $errpos) eq undef;
  $errpos == 1;

tld_get("mainbase.mars") eq "mars";

my $hashref = Net::LibIDN::tld_get_table("de");

print "$hashref->{version}\n";
foreach (@{$hashref->{valid}})
{
  print "Unicode range from ".$_->{start}." to ".$_->{end}."\n";
}
#ENDIF_TLD

DESCRIPTION

Provides bindings for GNU Libidn, a C library for handling Internationalized Domain Names according to IDNA (RFC 3490), in a way very much inspired by Turbo Fredriksson's PHP-IDN.

Functions

Net::LibIDN::idn_to_ascii($clear_hostname, [$charset, [$flags]]);

Converts $clear_hostname which might contain characters outside the range allowed in DNS names, to IDNA ACE. If $charset is specified, treats string as being encoded in it, otherwise assumes it is ISO-8859-1 encoded. If flag IDNA_ALLOW_UNASSIGNED is set in $flags, accepts also unassigned Unicode characters, if IDNA_USE_STD3_ASCII_RULES is set, accepts only ASCII LDH characters (letter-digit-hyphen). Flags can be combined with ||. Returns result of conversion or undef on error.

Net::LibIDN::idn_to_unicode($idn_hostname, [$charset, [$flags]]);

Converts ASCII $idn_hostname, which might be IDNA ACE encoded, into the decoded form in $charset or ISO-8859-1. Flags are interpreted as above. Returns result of conversion or undef on error.

Net::LibIDN::idn_punycode_encode($string, [$charset]);

Encodes $string into "punycode" (RFC 3492). If $charset is present, treats $string as being in $charset, otherwise uses ISO-8859-1. Returns result of conversion or undef on error.

Net::LibIDN::idn_punycode_decode($string, [$charset]);

Decodes $string from "punycode" (RFC 3492). If $charset is present, result is converted to $charset, otherwise it is converted to ISO-8859-1. Returns result of conversion or undef on error.

Net::LibIDN::idn_prep_name($string, [$charset]);

Net::LibIDN::idn_prep_kerberos5($string, [$charset]);

Net::LibIDN::idn_prep_node($string, [$charset]);

Net::LibIDN::idn_prep_resource($string, [$charset]);

Net::LibIDN::idn_prep_plain($string, [$charset]);

Net::LibIDN::idn_prep_trace($string, [$charset]);

Net::LibIDN::idn_prep_sasl($string, [$charset]);

Net::LibIDN::idn_prep_iscsi($string, [$charset]);

Performs "stringprep" (RFC 3454) on $string according to the named profile (e.g. *_name -> "nameprep" (RFC 3491)). If $charset is present, converts from and to this charset before and after the operation respectively. Returns result string, or undef on error.

#IF_TLD

Net::LibIDN::tdl_check($string, $errpos, [$charset, [$tld]]);

Checks whether or not $string conforms to the restrictions on the sets of valid characters defined by TLD authorities around the World. Treats $string as a hostname if $tld is not present, determining the TLD from the hostname. If $tld is present, uses the restrictions defined by the parties responsible for TLD $tld. $charset may be used to specify the character set the $string is in. Should an invalid character be detected, returns 0 and the 0-based position of the offending character in $errpos. In case of other failure conditions, $errpos is not touched, and undef is returned. Should $string conform to the TLD restrictions, 1 is returned.

Net::LibIDN::tld_get($hostname);

Returns top level domain of $hostname, or undef if an error occurs or if no top level domain was found.

Net::LibIDN::tld_get_table($tld);

Retrieves a hash reference with the TLD restriction info of given TLD $tld, or undef if $tld is not found. The hash ref contains the following fields:

• $h->{name} ... name of TLD

• $h->{version} ... version string of this restriction table

• $h->{nvalid} ... number of Unicode intervals

• $h->{valid} ... [ {start => number, end => number}, ...] ... Unicode intervals

#ENDIF_TLD =back

Limitations

There is currently no support for Perl's unicode capabilities (man perlunicode). All input strings are assumed to be octet strings, all output strings are generated as octet strings. Thus, if you require Perl's unicode features, you will have to convert your strings manually. For example:

use Encode;

use Data::Dumper;

print Dumper(Net::LibIDN::idn_to_unicode('xn--uro-j50a.com', 'utf-8'));

print Dumper(decode('utf-8', Net::LibIDN::idn_to_unicode('xn--uro-j50a.com', 'utf-8')));

AUTHOR

Thomas Jacob, http://internet24.de

SEE ALSO

perl(1), RFC 3454, RFC 3490-3492, http://www.gnu.org/software/libidn.