added better anti spot-sucking

[spider.git] / perl / Spot.pm

diff --git a/perl/Spot.pm b/perl/Spot.pm
index d49136b725284b57a4984195d8d41b8c330518b8..a7b2f76e6661943c2ba98736eb6af6736217e79c 100644 (file)
--- a/perl/Spot.pm
+++ b/perl/Spot.pm
@@ -27,7 +27,7 @@ $BRANCH = sprintf( "%d.%03d", q$Revision$ =~ /\d+\.\d+\.(\d+)\.(\d+)/  || (0,0))
 $main::build += $VERSION;
 $main::branch += $BRANCH;
 
-use vars qw($fp $statp $maxspots $defaultspots $maxdays $dirprefix $duplth $dupage $filterdef $totalspots $hfspots $vhfspots );
+use vars qw($fp $statp $maxspots $defaultspots $maxdays $dirprefix $duplth $dupage $filterdef $totalspots $hfspots $vhfspots $maxcalllth);
 
 $fp = undef;
 $statp = undef;
@@ -37,6 +37,7 @@ $maxdays = 100;                               # normal maximum no of days to go back
 $dirprefix = "spots";
 $duplth = 20;                                  # the length of text to use in the deduping
 $dupage = 3*3600;               # the length of time to hold spot dups
+$maxcalllth = 12;                               # the max length of call to take into account for dupes
 $filterdef = bless ([
                          # tag, sort, field, priv, special parser 
                          ['freq', 'r', 0, 0, \&decodefreq],
@@ -328,28 +329,41 @@ sub dup
        $d *= 60;
 
        $freq = sprintf "%.1f", $freq;       # normalise frequency
-       $call = substr($call, 0, 12) if length $call > 12;
+       $call = substr($call, 0, $maxcalllth) if length $call > $maxcalllth;
 
        chomp $text;
        $text =~ s/\%([0-9A-F][0-9A-F])/chr(hex($1))/eg;
+       $text = uc unpad($text);
        $text = substr($text, 0, $duplth) if length $text > $duplth; 
-       unpad($text);
        $text = pack("C*", map {$_ & 127} unpack("C*", $text));
-       $text =~ s/[^a-zA-Z0-9]//g;
-       my $ldupkey = "X$freq|$call|\L$text";
-       my $sdupkey = "X$freq|$call|$by";
+       my $ldupkey = "X$freq|$call|$by|";
        my $t = DXDupe::find($ldupkey);
-       if ($t) {
-               my $dt = $main::systime + $dupage - $t;
-               return 1 if $dt < 300;
-       }
-       $t = DXDupe::find($sdupkey);
-       if ($t) {
-               my $dt = $main::systime + $dupage - $t;
-               return 1 if $dt < 300;
+       if ($t && $t - $main::systime > 0) {
+               my ($prefix) = $text = /\b(\w{1,4})$/;
+               if ($prefix) {
+                       my @ans = Prefix::extract($prefix);
+                       if (@ans) {
+
+                               # if we find a prefix then chop it off
+                               # the end of the string and then look for
+                               # a spot with that text. If we find it then
+                               # it has be sucked from an AR-C node and is
+                               # a dupe.
+                               my $txt = $text;
+                               $txt =~ s/\b\w{1,4}$//;
+                               $txt =~ s/[^A-Z0-9]//g;
+                               $t = DXDupe::find($ldupkey . $txt);
+                               return 1 if $t && $t - $main::systime > 0;
+                       }
+               } 
        }
+       $text =~ s/[^A-Z0-9]//g;
+       $ldupkey .= $text;
        DXDupe::add($ldupkey, $main::systime+$dupage);
-       DXDupe::add($sdupkey, $main::systime+$dupage);
+#      my $sdupkey = "X$freq|$call|$by";
+#      $t = DXDupe::find($sdupkey);
+#      return 1 if $t && $t - $main::systime > 0;      
+#      DXDupe::add($sdupkey, $main::systime+$dupage);
        return 0;
 }