X-Git-Url: http://dxcluster.net/gitweb/gitweb.cgi?a=blobdiff_plain;f=perl%2FSpot.pm;h=056fd628557a1947706598f54f958358ef5807fb;hb=8e14149148baba63ce5ae2b95aacda8ab6dd0d87;hp=9b67cc5fb54f3a7d9e562e8247fc46c40b2ca29e;hpb=5b9ab81174d4a7734c92af52da6ddf89e980e321;p=spider.git diff --git a/perl/Spot.pm b/perl/Spot.pm index 9b67cc5f..056fd628 100644 --- a/perl/Spot.pm +++ b/perl/Spot.pm @@ -3,7 +3,7 @@ # # Copyright (c) - 1998 Dirk Koopman G1TLH # -# $Id$ +# # package Spot; @@ -17,16 +17,15 @@ use Julian; use Prefix; use DXDupe; use Data::Dumper; +use QSL; +use DXSql; +use Time::HiRes qw(gettimeofday tv_interval); -use strict; -use vars qw($VERSION $BRANCH); -$VERSION = sprintf( "%d.%03d", q$Revision$ =~ /(\d+)\.(\d+)/ ); -$BRANCH = sprintf( "%d.%03d", q$Revision$ =~ /\d+\.\d+\.(\d+)\.(\d+)/ || (0,0)); -$main::build += $VERSION; -$main::branch += $BRANCH; +use strict; -use vars qw($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 $can_encode $use_db_for_search); $fp = undef; $statp = undef; @@ -35,24 +34,51 @@ $defaultspots = 10; # normal number of spots to return $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 +$dupage = 1*3600; # the length of time to hold spot dups +$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], - ['on', 'r', 0, 0, \&decodefreq], - ['call', 'c', 1], - ['info', 't', 3], - ['by', 'c', 4], - ['call_dxcc', 'nc', 5], - ['by_dxcc', 'nc', 6], - ['origin', 'c', 7, 9], - ['call_itu', 'ni', 8], - ['call_zone', 'nz', 9], - ['by_itu', 'ni', 10], - ['by_zone', 'nz', 11], - ['channel', 'c', 12], - ], 'Filter::Cmd'); + # tag, sort, field, priv, special parser + ['freq', 'r', 0, 0, \&decodefreq], + ['on', 'r', 0, 0, \&decodefreq], + ['call', 'c', 1], + ['info', 't', 3], + ['spotter', 'c', 4], + ['by', 'c', 4], + ['dxcc', 'nc', 5], + ['call_dxcc', 'nc', 5], + ['by_dxcc', 'nc', 6], + ['origin', 'c', 7, 9], + ['call_itu', 'ni', 8], + ['itu', 'ni', 8], + ['call_zone', 'nz', 9], + ['cq', 'nz', 9], + ['zone', 'nz', 9], + ['by_itu', 'ni', 10], + ['byitu', 'ni', 10], + ['by_zone', 'nz', 11], + ['byzone', 'nz', 11], + ['bycq', 'nz', 11], + ['call_state', 'ns', 12], + ['state', 'ns', 12], + ['by_state', 'ns', 13], + ['bystate', 'ns', 13], + ['ip', 'c', 14], +# ['channel', 'c', 15], +# ['rbn', 'a', 4, 0, \&filterrbnspot], + ], 'Filter::Cmd'); $totalspots = $hfspots = $vhfspots = 0; +$use_db_for_search = 0; + +our %spotcache; # the cache of data within the last $spotcachedays 0 or 2+ days +our $spotcachedays = 2; # default 2 days worth +our $minselfspotqrg = 1240000; # minimum freq above which self spotting is allowed + +our $readback = $main::is_win ? 0 : 1; + +if ($readback) { + $readback = `which tac`; + chomp $readback; +} # create a Spot Object sub new @@ -72,7 +98,7 @@ sub decodefreq foreach $f (@f) { my ($a, $b); - if (m{^\d+/\d+$}) { + if ($f =~ m{^\d+/\d+$}) { push @out, $f; } elsif (($a, $b) = $f =~ m{^(\w+)(?:/(\w+))?$}) { $b = lc $b if $b; @@ -93,13 +119,104 @@ sub decodefreq return (0, join(',', @out)); } +# filter setup for rbn spot so return the regex to detect it +sub filterrbnspot +{ + my $dxchan = shift; + return ('-#$'); +} + sub init { mkdir "$dirprefix", 0777 if !-e "$dirprefix"; $fp = DXLog::new($dirprefix, "dat", 'd'); $statp = DXLog::new($dirprefix, "dys", 'd'); - system("rm -f $main::data/$dirprefix/200?/*.bys"); - system("rm -f $main::data/$dirprefix/200?/*.cys"); + my $today = Julian::Day->new(time); + + # load up any old spots + if ($main::dbh) { + unless (grep $_ eq 'spot', $main::dbh->show_tables) { + dbg('initialising spot tables'); + my $t = time; + my $total; + $main::dbh->spot_create_table; + + my $now = Julian::Day->alloc(1995, 0); + my $sth = $main::dbh->spot_insert_prepare; + while ($now->cmp($today) <= 0) { + my $fh = $fp->open($now); + if ($fh) { +# $main::dbh->{RaiseError} = 0; + $main::dbh->begin_work; + my $count = 0; + while (<$fh>) { + chomp; + my @s = split /\^/; + if (@s < 14) { + my @a = (Prefix::cty_data($s[1]))[1..3]; + my @b = (Prefix::cty_data($s[4]))[1..3]; + push @s, $b[1] if @s < 7; + push @s, '' if @s < 8; + push @s, @a[0,1], @b[0,1] if @s < 12; + push @s, $a[2], $b[2] if @s < 14; + } + $main::dbh->spot_insert(\@s, $sth); + $count++; + } + $main::dbh->commit; + dbg("inserted $count spots from $now->[0] $now->[1]"); + $fh->close; + $total += $count; + } + $now = $now->add(1); + } + $main::dbh->begin_work; + $main::dbh->spot_add_indexes; + $main::dbh->commit; +# $main::dbh->{RaiseError} = 1; + $t = time - $t; + my $min = int($t / 60); + my $sec = $t % 60; + dbg("$total spots converted in $min:$sec"); + } + unless ($main::dbh->has_ipaddr) { + $main::dbh->add_ipaddr; + dbg("added ipaddr field to spot table"); + } + } + + # initialise the cache if required + if ($spotcachedays > 0) { + my $t0 = [gettimeofday]; + $spotcachedays = 2 if $spotcachedays < 2; + dbg "Spot::init - reading in $spotcachedays days of spots into cache"; + for (my $i = 0; $i < $spotcachedays; ++$i) { + my $now = $today->sub($i); + my $fh = $fp->open($now); + if ($fh) { + my @in; + my $rec; + for ($rec = 0; <$fh>; ++$rec) { + chomp; + my @s = split /\^/; + if (@s < 14) { + my @a = (Prefix::cty_data($s[1]))[1..3]; + my @b = (Prefix::cty_data($s[4]))[1..3]; + push @s, $b[1] if @s < 7; + push @s, '' if @s < 8; + push @s, @a[0,1], @b[0,1] if @s < 12; + push @s, $a[2], $b[2] if @s < 14; + } + unshift @in, \@s; + } + $fh->close; + dbg("Spot::init read $rec spots from " . _cachek($now)); + $spotcache{_cachek($now)} = \@in; + } + $now->add(1); + } + dbg("Spot::init $spotcachedays files of spots read into cache in " . _diffms($t0) . "mS") + } } sub prefix @@ -108,45 +225,61 @@ sub prefix } # fix up the full spot data from the basic spot data +# input is +# freq, call, time, comment, spotter, origin[, ip_address] sub prepare { - # $freq, $call, $t, $comment, $spotter = @_ + # $freq, $call, $t, $comment, $spotter, node, ip address = @_ my @out = @_[0..4]; # just up to the spotter # normalise frequency - $_[0] = sprintf "%.1f", $_[0]; + $out[0] = sprintf "%.1f", $out[0]; - # remove ssids if present on spotter + # remove ssids and /xxx if present on spotter $out[4] =~ s/-\d+$//o; - # remove leading and trailing spaces - $_[3] = unpad($_[3]); + # remove leading and trailing spaces from comment field + $out[3] = unpad($out[3]); # add the 'dxcc' country on the end for both spotted and spotter, then the cluster call - my @dxcc = Prefix::extract($out[1]); - my $spotted_dxcc = (@dxcc > 0 ) ? $dxcc[1]->dxcc() : 666; - my $spotted_itu = (@dxcc > 0 ) ? $dxcc[1]->itu() : 0; - my $spotted_cq = (@dxcc > 0 ) ? $dxcc[1]->cq() : 0; - push @out, $spotted_dxcc; - @dxcc = Prefix::extract($out[4]); - my $spotter_dxcc = (@dxcc > 0 ) ? $dxcc[1]->dxcc() : 666; - my $spotter_itu = (@dxcc > 0 ) ? $dxcc[1]->itu() : 0; - my $spotter_cq = (@dxcc > 0 ) ? $dxcc[1]->cq() : 0; - push @out, $spotter_dxcc; + my @spd = Prefix::cty_data($out[1]); + push @out, $spd[0]; + my @spt = Prefix::cty_data($out[4]); + push @out, $spt[0]; push @out, $_[5]; - return (@out, $spotted_itu, $spotted_cq, $spotter_itu, $spotter_cq); + push @out, @spd[1,2], @spt[1,2], $spd[3], $spt[3]; + push @out, $_[6] if $_[6] && is_ipaddr($_[6]); + + # thus we now have: + # freq, call, time, comment, spotter, call country code, spotter country code, origin, call itu, call cqzone, spotter itu, spotter cqzone, call state, spotter state, spotter ip address + return @out; } sub add { - my $buf = join("\^", @_[0..7]); + my $buf = join('^', @_); $fp->writeunix($_[2], $buf); + if ($spotcachedays > 0) { + my $now = Julian::Day->new($_[2]); + my $day = _cachek($now); + my $r = (exists $spotcache{$day}) ? $spotcache{$day} : ($spotcache{$day} = []); + unshift @$r, \@_; + } + if ($main::dbh) { + $main::dbh->begin_work; + $main::dbh->spot_insert(\@_); + $main::dbh->commit; + } $totalspots++; if ($_[0] <= 30000) { $hfspots++; } else { $vhfspots++; } + if ($_[3] =~ /(?:QSL|VIA)/i) { + my $q = QSL::get($_[1]) || new QSL $_[1]; + $q->update($_[3], $_[2], $_[4]); + } } # search the spot database for records based on the field no and an expression @@ -163,7 +296,13 @@ sub add # $f5 = spotted dxcc country # $f6 = spotter dxcc country # $f7 = origin -# +# $f8 = spotted itu +# $f9 = spotted cq zone +# $f10 = spotter itu +# $f11 = spotter cq zone +# $f12 = spotted us state +# $f13 = spotter us state +# $f14 = ip address # # In addition you can specify a range of days, this means that it will start searching # from days less than today to days less than today @@ -176,8 +315,7 @@ sub add sub search { - my ($expr, $dayfrom, $dayto, $from, $to, $hint) = @_; - my $eval; + my ($expr, $dayfrom, $dayto, $from, $to, $hint, $dofilter, $dxchan) = @_; my @out; my $ref; my $i; @@ -198,47 +336,92 @@ sub search $to = $from + $maxspots if $to - $from > $maxspots || $to - $from <= 0; - $expr =~ s/\$f(\d)/\$ref->[$1]/g; # swap the letter n for the correct field name + if ($main::dbh && $use_db_for_search) { + return $main::dbh->spot_search($expr, $dayfrom, $dayto, $from, $to, $hint, $dofilter, $dxchan); + } + + # $expr =~ s/\$f(\d\d?)/\$ref->[$1]/g; # swap the letter n for the correct field name # $expr =~ s/\$f(\d)/\$spots[$1]/g; # swap the letter n for the correct field name - dbg("hint='$hint', expr='$expr', spotno=$from-$to, day=$dayfrom-$dayto\n") if isdbg('search'); + + dbg("Spot::search hint='$hint', expr='$expr', spotno=$from-$to, day=$dayfrom-$dayto\n") if isdbg('search'); # build up eval to execute - $eval = qq( - while (<\$fh>) { - $hint; - chomp; - push \@spots, [ split '\\^' ]; - } - my \$c; - my \$ref; - for (\$c = \$#spots; \$c >= 0; \$c--) { - \$ref = \$spots[\$c]; - if ($expr) { - \$count++; - next if \$count < \$from; # wait until from - push(\@out, \$ref); - last if \$count >= \$to; # stop after to - } - } - ); - $fp->close; # close any open files - - for ($i = $count = 0; $i < $maxdays; ++$i) { # look thru $maxdays worth of files only - my $now = $fromdate->sub($i); # but you can pick which $maxdays worth - last if $now->cmp($todate) <= 0; + dbg("Spot::search Spot eval: $expr") if isdbg('searcheval'); + $expr =~ s/\$r/\$_[0]/g; + my $eval = qq{ sub { return $expr; } }; + dbg("Spot::search Spot eval: $eval") if isdbg('searcheval'); + my $ecode = eval $eval; + return ("Spot search error", $@) if $@; + + my $fh; + my $now = $fromdate; + my $today = Julian::Day->new($main::systime); - my @spots = (); - my $fh = $fp->open($now); # get the next file - if ($fh) { - my $in; - eval $eval; # do the search on this file - last if $count >= $to; # stop after to - return ("Spot search error", $@) if $@; + for ($i = $count = 0; $count < $to && $i < $maxdays; ++$i) { # look thru $maxdays worth of files only + last if $now->cmp($todate) <= 0; + + + my $this = $now->sub($i); + my $fn = $fp->fn($this); + my $cachekey = _cachek($this); + my $rec = 0; + + if ($spotcachedays > 0 && $spotcache{$cachekey}) { + foreach my $r (@{$spotcache{$cachekey}}) { + ++$rec; + if ($dofilter && $dxchan && $dxchan->{spotsfilter}) { + my ($gotone, undef) = $dxchan->{spotsfilter}->it(@$r); + next unless $gotone; + } + if (&$ecode($r)) { + ++$count; + next if $count < $from; + push @out, $r; + last if $count >= $to; + } + } + dbg("Spot::search cache recs read: $rec") if isdbg('search'); + } else { + if ($readback) { + dbg("Spot::search search using tac fn: $fn $i") if isdbg('search'); + $fh = IO::File->new("$readback $fn |"); + } + else { + dbg("Spot::search search fn: $fp->{fn} $i") if isdbg('search'); + $fh = $fp->open($now->sub($i)); # get the next file + } + if ($fh) { + my $in; + while (<$fh>) { + chomp; + my @r = split /\^/; + ++$rec; + if ($dofilter && $dxchan && $dxchan->{spotsfilter}) { + my ($gotone, undef) = $dxchan->{spotsfilter}->it(@r); + next unless $gotone; + } + if (&$ecode(\@r)) { + ++$count; + next if $count < $from; + if ($readback) { + push @out, \@r; + last if $count >= $to; + } else { + push @out, \@r; + shift @out if $count >= $to; + } + } + } + dbg("Spot::search file recs read: $rec") if isdbg('search'); + last if $count >= $to; # stop after to + } } } + return ("Spot search error", $@) if $@; + @out = sort {$b->[2] <=> $a->[2]} @out if @out; return @out; } @@ -274,57 +457,70 @@ sub ftor # format a spot for user output in list mode sub formatl { - my $t = ztime($_[2]); - my $d = cldate($_[2]); - return sprintf "%8.1f %-11s %s %s %-28.28s%7s>", $_[0], $_[1], $d, $t, $_[3], "<$_[4]" ; -} - -# -# return all the spots from a day's file as an array of references -# the parameter passed is a julian day -sub readfile($) -{ - my @spots; + my $t = ztime($_[3]); + my $d = cldate($_[3]); + my $spotter = "<$_[5]>"; + my $comment = $_[4] || ''; + $comment =~ s/\t+/ /g; + my $cl = length $comment; + my $s = sprintf "%9.1f %-11s %s %s", $_[1], $_[2], $d, $t; + my $width = ($_[0] ? $_[0] : 80) - length($spotter) - length($s) - 4; - my $fh = $fp->open(shift); - if ($fh) { - my $in; - while (<$fh>) { - chomp; - push @spots, [ split '\^' ]; - } - } - return @spots; + $comment = substr $comment, 0, $width if $cl > $width; + $comment .= ' ' x ($width-$cl) if $cl < $width; + +# return sprintf "%8.1f %-11s %s %s %-28.28s%7s>", $_[0], $_[1], $d, $t, ($_[3]||''), "<$_[4]" ; + return "$s $comment$spotter"; } # enter the spot for dup checking and return true if it is already a dup sub dup { - my ($freq, $call, $d, $text) = @_; + my ($freq, $call, $d, $text, $by, $node, $just_find) = @_; # dump if too old return 2 if $d < $main::systime - $dupage; - - # turn the time into minutes (should be already but...) + + # turn the time into minutes (should be already but...) $d = int ($d / 60); $d *= 60; - $freq = sprintf "%.1f", $freq; # normalise frequency - $call = substr($call, 0, 12) if length $call > 12; + # remove SSID or area + $by =~ s|[-/]\d+$||; + +# $freq = sprintf "%.1f", $freq; # normalise frequency + $freq = int $freq; # normalise frequency + $call = substr($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); + my $otext = $text; +# $text = Encode::encode("iso-8859-1", $text) if $main::can_encode && Encode::is_utf8($text, 1); + $text =~ s/^\+\w+\s*//; # remove leading LoTW callsign + $text =~ s/\s{2,}[\dA-Z]?[A-Z]\d?$// if length $text > 24; + $text =~ s/[\W\x00-\x2F\x7B-\xFF]//g; # tautology, just to make quite sure! $text = substr($text, 0, $duplth) if length $text > $duplth; - unpad($text); - $text = pack("C*", map {$_ & 127} unpack("C*", $text)); - $text =~ s/[^a-zA-Z0-9]//g; - for (0,60,120,180,240,300) { - my $dt = $d - $_; - my $dupkey = "X$freq|$call|$dt|\L$text"; - return 1 if DXDupe::find($dupkey); + my $ldupkey = "X$|$call|$by|$node|$freq|$d|$text"; + + my $t = DXDupe::find($ldupkey); + return 1 if $t && $t - $main::systime > 0; + + DXDupe::add($ldupkey, $main::systime+$dupage) unless $just_find; + $otext = substr($otext, 0, $duplth) if length $otext > $duplth; + $otext =~ s/\s+$//; + if (length $otext && $otext ne $text) { + $ldupkey = "X$freq|$call|$by|$otext"; + $t = DXDupe::find($ldupkey); + return 1 if $t && $t - $main::systime > 0; + DXDupe::add($ldupkey, $main::systime+$dupage) unless $just_find; } - my $dupkey = "X$freq|$call|$d|\L$text"; - DXDupe::add($dupkey, $main::systime+$dupage); - return 0; + return undef; +} + +sub dup_find +{ + return dup(@_, 1); } sub listdups @@ -332,11 +528,11 @@ sub listdups return DXDupe::listdups('X', $dupage, @_); } -sub genstats($) +sub genstats { my $date = shift; - my $in = $fp->open($date); - my $out = $statp->open($date, 'w'); + my $in = $fp->open($date) or dbg("Spot::genstats: Cannot open " . $fp->fn($date) . " $!"); + my $out = $statp->open($date, 'w') or dbg("Spot::genstats: Cannot open " . $statp->fn($date) . " $!"); my @freq; my %list; my @tot; @@ -349,6 +545,7 @@ sub genstats($) my ($freq, $by, $dxcc) = (split /\^/)[0,4,6]; my $ref = $list{$by} || [0, $dxcc]; for (@freq) { + next unless defined $_; if ($freq >= $_->[1] && $freq <= $_->[2]) { $$ref[$_->[0]+2]++; $tot[$_->[0]+2]++; @@ -378,7 +575,7 @@ sub genstats($) } # return true if the stat file is newer than than the spot file -sub checkstats($) +sub checkstats { my $date = shift; my $in = $fp->mtime($date); @@ -391,6 +588,26 @@ sub daily { my $date = Julian::Day->new($main::systime)->sub(1); genstats($date) unless checkstats($date); + clean_cache(); +} + +sub _cachek +{ + return "$_[0]->[0]|$_[0]->[1]"; +} + +sub clean_cache +{ + if ($spotcachedays > 0) { + my $now = Julian::Day->new($main::systime); + for (my $i = $spotcachedays; $i < $spotcachedays + 5; ++$i ) { + my $k = _cachek($now->sub($i)); + if (exists $spotcache{$k}) { + dbg("Spot::spotcache deleting day $k, more than $spotcachedays days old"); + delete $spotcache{$k}; + } + } + } } 1;