reduce the (default) length of announces to dupe

[spider.git] / perl / AnnTalk.pm

diff --git a/perl/AnnTalk.pm b/perl/AnnTalk.pm
index 341857450be90e8e45dd5003fed742875d645682..c075765f6eaf1523fc4a1b022370d527f7a23c5c 100644 (file)
--- a/perl/AnnTalk.pm
+++ b/perl/AnnTalk.pm
@@ -3,7 +3,7 @@
 #
 # Copyright (c) 2000 Dirk Koopman
 #
-# $Id$
+#
 #
 
 package AnnTalk;
@@ -15,23 +15,43 @@ use DXDebug;
 use DXDupe;
 use DXVars;
 
-use vars qw(%dup $duplth $dupage);
-
-$duplth = 60;                                  # the length of text to use in the deduping
-$dupage = 5*24*3600;                   # the length of time to hold spot dups
+use vars qw(%dup $duplth $dupage $filterdef);
 
+$duplth = 30;                                  # the length of text to use in the deduping
+$dupage = 5*3600;                              # the length of time to hold ann dups
+$filterdef = bless ([
+                         # tag, sort, field, priv, special parser 
+                         ['by', 'c', 0],
+                         ['dest', 'c', 1],
+                         ['info', 't', 2],
+                         ['group', 't', 3],
+                         ['origin', 'c', 4],
+                         ['wx', 't', 5],
+                         ['channel', 'c', 6],
+                         ['by_dxcc', 'nc', 7],
+                         ['by_itu', 'ni', 8],
+                         ['by_zone', 'nz', 9],
+                         ['origin_dxcc', 'nc', 10],
+                         ['origin_itu', 'ni', 11],
+                         ['origin_zone', 'nz', 12],
+                         ['by_state', 'nz', 13],
+                         ['origin_state', 'nz', 14],
+                        ], 'Filter::Cmd');
 
 # enter the spot for dup checking and return true if it is already a dup
 sub dup
 {
-       my ($call, $to, $text) = @_; 
+       my ($call, $to, $text, $t) = @_; 
 
+       $t ||= $main::systime + $dupage;
        chomp $text;
        unpad($text);
-       $text =~ s/[^a-zA-Z0-9]//g;
+       $text =~ s/\%([0-9A-F][0-9A-F])/chr(hex($1))/eg;
+       $text = Encode::encode("iso-8859-1", $text) if $main::can_encode && Encode::is_utf8($text, 1);
+       $text =~ s/[^\#a-zA-Z0-9]//g;
        $text = substr($text, 0, $duplth) if length $text > $duplth; 
-       my $dupkey = "A$to|\L$text";
-       return DXDupe::check($dupkey, $main::systime + $dupage);
+       my $dupkey = "A$call|$to|\L$text";
+       return DXDupe::check($dupkey, $t);
 }
 
 sub listdups
@@ -39,6 +59,21 @@ sub listdups
        return DXDupe::listdups('A', $dupage, @_);
 }
 
-
+# is this text field a likely announce to talk substitution?
+# this may involve all sorts of language dependant heuristics, but 
+# then again, it might not
+sub is_talk_candidate
+{
+       my ($from, $text) = @_;
+       my $call;
+       ($call) = $text =~ /^\s*(?:[Xx]|[Tt][Oo]?:?)\s+([\w-]+)/;
+       ($call) = $text =~ /^\s*>\s*([\w-]+)\b/ unless $call;
+       ($call) = $text =~ /^\s*([\w-]+):?\b/ unless $call;
+       if ($call) {
+               $call = uc $call;
+               return is_callsign($call);
+       }
+    return undef;
+}
 1;