File Coverage

blib/lib/AnyEvent/IRC/Util.pm

Criterion	Covered	Total	%
statement	52	123	42.2
branch	14	48	29.1
condition	5	12	41.6
subroutine	11	21	52.3
pod	13	18	72.2
total	95	222	42.7

package AnyEvent::IRC::Util;
use common::sense;
use Exporter;
use Encode;
our @ISA = qw/Exporter/;
our @EXPORT_OK =
   qw(mk_msg parse_irc_msg split_prefix prefix_nick
      decode_ctcp encode_ctcp filter_ctcp_text_attr prefix_user prefix_host
      rfc_code_to_name filter_colors is_nick_prefix join_prefix
      split_unicode_string);

=head1 NAME

AnyEvent::IRC::Util - Common utilities that help with IRC protocol handling

=head1 SYNOPSIS

   use AnyEvent::IRC::Util qw/parse_irc_msg mk_msg/;

   my $msgdata = mk_msg (undef, PRIVMSG => "mcmanus", "my hands glow!");

=head1 FUNCTIONS

These are some utility functions that might come in handy when
handling the IRC protocol.

You can export these with eg.:

   use AnyEvent::IRC::Util qw/parse_irc_msg/;

=over 4

=item B<parse_irc_msg ($ircline)>

This method parses the C<$ircline>, which is one line of the IRC protocol
without the trailing "\015\012".

It returns a hash which has the following entrys:

=over 4

=item prefix

The message prefix.

=item command

The IRC command.

=item params

The parameters to the IRC command in a array reference,
this includes the trailing parameter (the one after the ':' or
the 14th parameter).

=back

=cut

sub parse_irc_msg {
   my ($msg) = @_;

   $msg =~ s/^(?::([^ ]+)[ ])?([A-Za-z]+|\d{3})//
      or return undef;
   my %msg;
   ($msg{prefix}, $msg{command}, $msg{params}) = ($1, $2, []);

   my $cnt = 0;
   while ($msg =~ s/^[ ]([^ :\015\012\0][^ \015\012\0]*)//) {
      push @{$msg{params}}, $1 if defined $1;
      last if ++$cnt > 13;
   }

   if ($cnt == 14) {
      if ($msg =~ s/^[ ]:?([^\015\012\0]*)//) {
         push @{$msg{params}}, $1 if defined $1;
      }

   } else {
      if ($msg =~ s/^[ ]:([^\015\012\0]*)//) {
         push @{$msg{params}}, $1 if defined $1;
      }
   }

   \%msg
}

=item B<mk_msg ($prefix, $command, @params)>

This function assembles a IRC message. The generated
message will look like (pseudo code!)

   :<prefix> <command> <params> :<trail>

Please refer to RFC 1459 how IRC messages normally look like.

The prefix will be omitted if they are C<undef>.

Please note that only the last parameter may contain spaces, and if it
contains spaces it will be quoted as the trailing part of the
IRC message.

NOTE: The trailing "\015\012" is NOT added by this function!

EXAMPLES:

   mk_msg (undef, "PRIVMSG", "magnus", "you suck!");
   # will return: "PRIVMSG magnus :you suck!"

   mk_msg (undef, "PRIVMSG", "magnus", "Hi!");
   # will return: "PRIVMSG magnus :Hi!"

   mk_msg (undef, "JOIN", "#test");
   # will return: "JOIN #test"

=cut

sub mk_msg {
  my ($prefix, $command, @params) = @_;
  my $msg = "";

  $msg .= defined $prefix ? ":$prefix " : "";
  $msg .= "$command";

  my $trail;
  if (@params && ($params[-1] =~ /\x20/ || $params[-1] =~ /^:/)) {
     $trail = pop @params;
  }

  # FIXME: params must be counted, and if > 13 they have to be
  # concationated with $trail
  map { $msg .= " $_" } @params;

  $msg .= defined $trail ? " :$trail" : "";

  return $msg;
}

my @_ctcp_lowlevel_escape = ("\000", "0", "\012", "n", "\015", "r", "\020", "\020");

sub unescape_lowlevel {
   my ($data) = @_;
   my %map = reverse @_ctcp_lowlevel_escape;
   $data =~ s/\020(.)/defined $map{$1} ? $map{$1} : $1/ge;
   $data
}

sub escape_lowlevel {
   my ($data) = @_;
   my %map = @_ctcp_lowlevel_escape;
   $data =~ s/([\000\012\015\020])/"\020$map{$1}"/ge;
   $data
}

sub unescape_ctcp {
   my ($data) = @_;
   $data =~ s/\\(.)/$1 eq 'a' ? "\001" : ($1 eq "\\" ? "\\" : $1)/eg;
   $data
}

sub escape_ctcp {
   my ($data) = @_;
   $data =~ s/([\\\001])/$1 eq "\001" ? "\\a" : "\\\\"/eg;
   $data
}

=item B<decode_ctcp ($data)>

This function decodes CTCP messages contained in an IRC message.
C<$data> should be the last parameter of a IRC PRIVMSG or NOTICE.

It will first unescape the lower layer, extract CTCP messages
and then return a list with two elements: the line without the CTCP messages
and an array reference which contains array references of CTCP messages.
Those CTCP message array references will have the CTCP message tag as
first element (eg. "VERSION") and the rest of the CTCP message as the second
element.

=cut

sub decode_ctcp {
   my ($line) = @_;

   $line = unescape_lowlevel ($line);
   my @ctcp;
   while ($line =~ /\G\001([^\001]*)\001/g) {
      my $msg = unescape_ctcp ($1);
      my ($tag, $data) = split / /, $msg, 2;
      push @ctcp, [$tag, $data];
   }

   $line =~ s/\001[^\001]*\001//g;

   # try to parse broken ctcp messages anyway
   if ($line =~ s/\001([^\001]*)$//) {
      my $msg = unescape_ctcp ($1);
      my ($tag, $data) = split / /, $msg, 2;
      push @ctcp, [$tag, $data];
   }

   return ($line, \@ctcp)
}

=item B<encode_ctcp (@msg)>

This function encodes a CTCP message for the transmission via the NOTICE
or PRIVMSG command. C<@msg> is an array of strings or array references.
If an array reference occurs in the C<@msg> array it's first
element will be interpreted as CTCP TAG (eg. one of PING, VERSION, .. whatever)
the rest of the array ref will be appended to the tag and separated by
spaces.

All parts of the message will be concatenated and lowlevel quoted.
That means you can embed _any_ character from 0 to 255 in this message (thats
what the lowlevel quoting allows).

=cut

sub encode_ctcp {
   my (@args) = @_;
   escape_lowlevel (
      join "", map {
         ref $_
            ? "\001" . escape_ctcp (join " ", @$_) . "\001"
            : $_
      } @args
   )
}

=item B<filter_colors ($line)>

This function will filter out any mIRC colors and (most) ansi escape sequences.
Unfortunately the mIRC color coding will destroy improper colored numbers. So this
function may destroy the message in some occasions a bit.

=cut

sub filter_colors($) {
   my ($line) = @_;
   $line =~ s/\x1B\[.*?[\x00-\x1F\x40-\x7E]//g; # see ECMA-48 + advice by urxvt author
   $line =~ s/\x03\d\d?(?:,\d\d?)?//g;          # see http://www.mirc.co.uk/help/color.txt
   $line =~ s/[\x03\x16\x02\x1f\x0f]//g;        # see some undefined place :-)
   $line
}


# implemented after the below CTCP spec, but
# doesnt seem to be used by anyone... so it's untested.
sub filter_ctcp_text_attr_bogus {
   my ($line, $cb) = @_;
   return unless $cb;
   $line =~ s/\006([BVUSI])/{warn "FIL\n"; my $c = $cb->($1); defined $c ? $c : "\006$1"}/ieg;
   $line =~ s/\006CA((?:I[0-9A-F]|#[0-9A-F]{3}){2})/{my $c = $cb->($1); defined $c ? $c : "\006CA$1"}/ieg;
   $line =~ s/\006C([FB])(I[0-9A-F]|#[0-9A-F]{3})/{my $c = $cb->($1, $2); defined $c ? $c : "\006C$1$2"}/ieg;
   $line =~ s/\006CX([AFB])/{my $c = $cb->($1); defined $c ? $c : "\006CX$1"}/ieg;
   return $line;
}

=item B<split_prefix ($prefix)>

This function splits an IRC user prefix as described by RFC 2817
into the three parts: nickname, user and host. Which will be
returned as a list with that order.

C<$prefix> can also be a hash like it is returned by C<parse_irc_msg>.

=cut

sub split_prefix {
   my ($prfx) = @_;

   if (ref ($prfx) eq 'HASH') {
      $prfx = $prfx->{prefix};
   }

   # this splitting does indeed use the servername as nickname, but there
   # is no way for a client to distinguish.
   $prfx =~ m/^\s*([^!]*)(?:!([^@]*))?(?:@(.*?))?\s*$/;
   return ($1, $2, $3);
}

=item B<is_nick_prefix ($prefix)>

Returns true if the prefix is a nick prefix, containing user and host.

=cut

sub is_nick_prefix {
   my ($prfx) = @_;
   $prfx =~ m/^\s*([^!]+)!([^@]+)@(.+)?\s*$/;
}

=item B<join_prefix ($nick, $user, $host)>

Joins C<$nick>, C<$user> and C<$host> together to form a prefix.

=cut

sub join_prefix {
   my ($n, $u, $h) = @_;
   "$n!$u\@$h"
}

=item B<prefix_nick ($prefix)>

A shortcut to extract the nickname from the C<$prefix>.

C<$prefix> can also be a hash like it is returned by C<parse_irc_msg>.

=cut

sub prefix_nick {
   my ($prfx) = @_;
   return (split_prefix ($prfx))[0];
}

=item B<prefix_user ($prefix)>

A shortcut to extract the username from the C<$prefix>.

C<$prefix> can also be a hash like it is returned by C<parse_irc_msg>.

=cut

sub prefix_user {
   my ($prfx) = @_;
   return (split_prefix ($prfx))[1];
}

=item B<prefix_host ($prefix)>

A shortcut to extract the hostname from the C<$prefix>.

C<$prefix> can also be a hash like it is returned by C<parse_irc_msg>.

=cut

sub prefix_host {
   my ($prfx) = @_;
   return (split_prefix ($prfx))[2];
}


=item B<rfc_code_to_name ($code)>

This function is a interface to the internal mapping or numeric
replies to the reply name in RFC 2812 (which you may also consult).

C<$code> is returned if no name for C<$code> exists
(as some server may extended the protocol).

=cut

our %RFC_NUMCODE_MAP = (
   '001' => 'RPL_WELCOME',
   '002' => 'RPL_YOURHOST',
   '003' => 'RPL_CREATED',
   '004' => 'RPL_MYINFO',
   '005' => 'RPL_BOUNCE',
   '200' => 'RPL_TRACELINK',
   '201' => 'RPL_TRACECONNECTING',
   '202' => 'RPL_TRACEHANDSHAKE',
   '203' => 'RPL_TRACEUNKNOWN',
   '204' => 'RPL_TRACEOPERATOR',
   '205' => 'RPL_TRACEUSER',
   '206' => 'RPL_TRACESERVER',
   '207' => 'RPL_TRACESERVICE',
   '208' => 'RPL_TRACENEWTYPE',
   '209' => 'RPL_TRACECLASS',
   '210' => 'RPL_TRACERECONNECT',
   '211' => 'RPL_STATSLINKINFO',
   '212' => 'RPL_STATSCOMMANDS',
   '219' => 'RPL_ENDOFSTATS',
   '221' => 'RPL_UMODEIS',
   '233' => 'RPL_SERVICE',
   '234' => 'RPL_SERVLIST',
   '235' => 'RPL_SERVLISTEND',
   '242' => 'RPL_STATSUPTIME',
   '243' => 'RPL_STATSOLINE',
   '250' => 'RPL_STATSDLINE',
   '251' => 'RPL_LUSERCLIENT',
   '252' => 'RPL_LUSEROP',
   '253' => 'RPL_LUSERUNKNOWN',
   '254' => 'RPL_LUSERCHANNELS',
   '255' => 'RPL_LUSERME',
   '256' => 'RPL_ADMINME',
   '257' => 'RPL_ADMINLOC1',
   '258' => 'RPL_ADMINLOC2',
   '259' => 'RPL_ADMINEMAIL',
   '261' => 'RPL_TRACELOG',
   '262' => 'RPL_TRACEEND',
   '263' => 'RPL_TRYAGAIN',
   '301' => 'RPL_AWAY',
   '302' => 'RPL_USERHOST',
   '303' => 'RPL_ISON',
   '305' => 'RPL_UNAWAY',
   '306' => 'RPL_NOWAWAY',
   '311' => 'RPL_WHOISUSER',
   '312' => 'RPL_WHOISSERVER',
   '313' => 'RPL_WHOISOPERATOR',
   '314' => 'RPL_WHOWASUSER',
   '315' => 'RPL_ENDOFWHO',
   '317' => 'RPL_WHOISIDLE',
   '318' => 'RPL_ENDOFWHOIS',
   '319' => 'RPL_WHOISCHANNELS',
   '321' => 'RPL_LISTSTART',
   '322' => 'RPL_LIST',
   '323' => 'RPL_LISTEND',
   '324' => 'RPL_CHANNELMODEIS',
   '325' => 'RPL_UNIQOPIS',
   '331' => 'RPL_NOTOPIC',
   '332' => 'RPL_TOPIC',
   '341' => 'RPL_INVITING',
   '342' => 'RPL_SUMMONING',
   '346' => 'RPL_INVITELIST',
   '347' => 'RPL_ENDOFINVITELIST',
   '348' => 'RPL_EXCEPTLIST',
   '349' => 'RPL_ENDOFEXCEPTLIST',
   '351' => 'RPL_VERSION',
   '352' => 'RPL_WHOREPLY',
   '353' => 'RPL_NAMREPLY',
   '364' => 'RPL_LINKS',
   '365' => 'RPL_ENDOFLINKS',
   '366' => 'RPL_ENDOFNAMES',
   '367' => 'RPL_BANLIST',
   '368' => 'RPL_ENDOFBANLIST',
   '369' => 'RPL_ENDOFWHOWAS',
   '371' => 'RPL_INFO',
   '372' => 'RPL_MOTD',
   '374' => 'RPL_ENDOFINFO',
   '375' => 'RPL_MOTDSTART',
   '376' => 'RPL_ENDOFMOTD',
   '381' => 'RPL_YOUREOPER',
   '382' => 'RPL_REHASHING',
   '383' => 'RPL_YOURESERVICE',
   '384' => 'RPL_MYPORTIS',
   '391' => 'RPL_TIME',
   '392' => 'RPL_USERSSTART',
   '393' => 'RPL_USERS',
   '394' => 'RPL_ENDOFUSERS',
   '395' => 'RPL_NOUSERS',
   '401' => 'ERR_NOSUCHNICK',
   '402' => 'ERR_NOSUCHSERVER',
   '403' => 'ERR_NOSUCHCHANNEL',
   '404' => 'ERR_CANNOTSENDTOCHAN',
   '405' => 'ERR_TOOMANYCHANNELS',
   '406' => 'ERR_WASNOSUCHNICK',
   '407' => 'ERR_TOOMANYTARGETS',
   '408' => 'ERR_NOSUCHSERVICE',
   '409' => 'ERR_NOORIGIN',
   '411' => 'ERR_NORECIPIENT',
   '412' => 'ERR_NOTEXTTOSEND',
   '413' => 'ERR_NOTOPLEVEL',
   '414' => 'ERR_WILDTOPLEVEL',
   '415' => 'ERR_BADMASK',
   '421' => 'ERR_UNKNOWNCOMMAND',
   '422' => 'ERR_NOMOTD',
   '423' => 'ERR_NOADMININFO',
   '424' => 'ERR_FILEERROR',
   '431' => 'ERR_NONICKNAMEGIVEN',
   '432' => 'ERR_ERRONEUSNICKNAME',
   '433' => 'ERR_NICKNAMEINUSE',
   '436' => 'ERR_NICKCOLLISION',
   '437' => 'ERR_UNAVAILRESOURCE',
   '441' => 'ERR_USERNOTINCHANNEL',
   '442' => 'ERR_NOTONCHANNEL',
   '443' => 'ERR_USERONCHANNEL',
   '444' => 'ERR_NOLOGIN',
   '445' => 'ERR_SUMMONDISABLED',
   '446' => 'ERR_USERSDISABLED',
   '451' => 'ERR_NOTREGISTERED',
   '461' => 'ERR_NEEDMOREPARAMS',
   '462' => 'ERR_ALREADYREGISTRED',
   '463' => 'ERR_NOPERMFORHOST',
   '464' => 'ERR_PASSWDMISMATCH',
   '465' => 'ERR_YOUREBANNEDCREEP',
   '466' => 'ERR_YOUWILLBEBANNED',
   '467' => 'ERR_KEYSET',
   '471' => 'ERR_CHANNELISFULL',
   '472' => 'ERR_UNKNOWNMODE',
   '473' => 'ERR_INVITEONLYCHAN',
   '474' => 'ERR_BANNEDFROMCHAN',
   '475' => 'ERR_BADCHANNELKEY',
   '476' => 'ERR_BADCHANMASK',
   '477' => 'ERR_NOCHANMODES',
   '478' => 'ERR_BANLISTFULL',
   '481' => 'ERR_NOPRIVILEGES',
   '482' => 'ERR_CHANOPRIVSNEEDED',
   '483' => 'ERR_CANTKILLSERVER',
   '484' => 'ERR_RESTRICTED',
   '485' => 'ERR_UNIQOPPRIVSNEEDED',
   '491' => 'ERR_NOOPERHOST',
   '492' => 'ERR_NOSERVICEHOST',
   '501' => 'ERR_UMODEUNKNOWNFLAG',
   '502' => 'ERR_USERSDONTMATCH',
);

sub rfc_code_to_name {
   my ($code) = @_;
   return $RFC_NUMCODE_MAP{$code} || $code;
}

=item my (@lines) = split_unicode_string ($encoding, $string, $maxlinebytes)

This function splits up C<$string> into multiple C<@lines> which are
not longer than C<$maxlinebytes> bytes. Encoding can be given in C<$encoding>.
(eg. 'utf-8'). But the output will not be encoded.

This function takes care that your characters are not garbled.

=cut

sub split_unicode_string {
   my ($enc, $str, $maxlen) = @_;

   return $str unless length (encode ($enc, $str)) > $maxlen;

   my $cur_out = '';
   my @lines;

   while (length ($str) > 0) {

      while (length (encode ($enc, $cur_out)) <= $maxlen
             && length ($str) > 0) {

         $cur_out .= substr $str, 0, 1, '';
      }

      push @lines, $cur_out;
      $cur_out = '';
   }

   @lines
}

=back

=head1 AUTHOR

Robin Redeker, C<< <[e-mail]> >>

=head1 SEE ALSO

Internet Relay Chat Client To Client Protocol from February 2, 1997
http://www.invlogic.com/irc/ctcp.html

RFC 1459 - Internet Relay Chat: Client Protocol

=head1 COPYRIGHT & LICENSE

Copyright 2006-2009 Robin Redeker, all rights reserved.

This program is free software; you can redistribute it and/or modify it
under the same terms as Perl itself.

=cut

1;