--- trunk2/all2all.pl 2004/09/11 20:21:28 423
+++ trunk2/all2all.pl 2004/11/01 17:19:48 574
@@ -7,19 +7,99 @@
=cut
use strict;
-use Data::Dumper;
+use locale;
+use YAML;
use Carp;
+use Getopt::Long;
+use Text::Unaccent 1.02;
use lib './lib';
use WebPAC;
use WebPAC::jsFind;
use WebPAC::Index;
+use WebPAC::Tree;
+# options which can be changed via command line
+#
+my $code_page = 'ISO-8859-2';
+my ($limit_mfn, $start_mfn, $debug, $low_mem);
+my $index_path = './out/index';
+
+my $result = GetOptions(
+ "code_page=s" => \$code_page,
+ "limit_mfn=i" => \$limit_mfn,
+ "start_mfn=i" => \$start_mfn,
+ "debug!" => \$debug,
+ "low_mem!" => \$low_mem,
+);
+
+my $filter = {
+ 'CROVOC' => sub {
+ my $tmp = shift || return;
+ $tmp =~ s/\s*HIDRA.*//;
+ return undef unless ($tmp =~ s/\s*CROVOC.*$/ #C#/);
+ # remove repeating stars
+ # FIXME this should be fixed at right place, not here!
+ $tmp =~ s/(\s*#C#)+/ #C#/g;
+ return $tmp;
+ },
+ 'CROVOC_tree' => sub {
+ my $tmp = shift || return;
+ $tmp =~ s/\s*CROVOC.*$/ /;
+ $tmp =~ s/\s*EUROVOC.*//;
+ $tmp =~ s/\s*HIDRA.*//;
+ return $tmp;
+ },
+ # TT filter
+ 'CROVOC_img' => sub {
+ my $tmp = shift;
+ $tmp =~ s/\s*#C#\s*/
/gis;
+ $tmp =~ s/"img\/crovoc.png"/"..\/img\/crovoc.png"/gis;
+ return $tmp;
+ }
+};
+
+## remove accented characters
+#
+sub unac {
+ my $string = shift || return;
+ $string = unac_string($code_page,$string);
+ $string =~ tr/ðÐ/dD/;
+ $string = unac_entities($string);
+ return $string;
+}
+sub unac_2 {
+ my $string = shift || return;
+ if (length($string) > 2) {
+ my $pr = substr($string,0,2);
+ $string = unac_string($code_page,substr($string,2));
+ $string =~ tr/ðÐ/dD/;
+ $string = $pr . $string;
+ }
+ $string = unac_entities($string);
+ return lc($string);
+}
+
+sub unac_entities {
+ my $ent = shift || return;
+
+ $ent =~ s/&(\w)(acute|cedil|circ|grave|ring|slash|tilde|uml);/$1/gi;
+ $ent =~ s/ð/d/g;
+ $ent =~ s/&E[tT][hH];/D/g;
+ $ent =~ s/&(\w\w)lig;/$1/gi;
+
+ return $ent;
+}
+
+# create WebPAC object
+#
my $webpac = new WebPAC(
- code_page => 'ISO-8859-2',
- limit_mfn => 500,
-# debug => 1,
-# low_mem => 1,
+ code_page => $code_page,
+ limit_mfn => $limit_mfn,
+ start_mfn => $start_mfn,
+ debug => $debug,
+ low_mem => $low_mem,
+ filter => $filter,
) || die;
my $log = $webpac->_get_logger() || die "can't get logger";
@@ -27,8 +107,8 @@
$log->debug("creating WebPAC::jsFind object");
my $index = new WebPAC::jsFind(
- index_path => './out/index',
- keys => 10,
+ index_path => $index_path,
+ keys => 62,
) || die;
my $thes;
@@ -38,15 +118,16 @@
my $maxmfn = $webpac->open_isis(
filename => shift @ARGV || '/data/hidra/THS/THS',
lookup => [
- { 'key' => 'd:v900', 'val' => 'v250^a' },
+ { 'key' => 'd:v900', 'val' => 'filter{CROVOC_tree}v250^a v800' },
# { 'eval' => '"v901^a" eq "Podruèje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
# { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
# { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
{ 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
{ 'key' => '900_mfn:v900', 'val' => 'v000' },
# tree structure
- { 'eval' => 'length("v251") == 2', 'key' => 'root:v251', 'val' => 'v900' },
+ { 'eval' => 'length("v251") == 2 && "v800" =~ m/EUROVOC/ || "v800" =~ m/CROVOC/ && "v251" =~ m/^(H|HD|L|Z|P)$/', 'key' => 'root:v251', 'val' => 'v900' },
{ 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },
+ { 'eval' => '"v561^4" && "v562^4"', 'key' => 'code:v900', 'val' => 'v561^4:v562^4' },
],
);
@@ -56,13 +137,15 @@
$webpac->open_import_xml(type => 'isis_hidra_ths');
+if(1) { # XXX
+
while (my $rec = $webpac->fetch_rec) {
my @ds = $webpac->data_structure($rec);
if (0 && $log->is_debug) {
- $log->debug("rec = ",Dumper($rec));
- $log->debug("ds = ",Dumper(\@ds));
+ $log->debug("rec = ",Dump($rec));
+ $log->debug("ds = ",Dump(\@ds));
}
next if (! @ds);
@@ -93,11 +176,26 @@
foreach my $ds (@ds) {
next if (! $ds->{'swish'});
+ # strip all non word characters from beginning or end
+ # of word
+ my $words = join(" ",@{$ds->{'swish'}});
+ $words =~ s/^\W+//;
+ $words =~ s/\W*\s+\W*/ /g;
+ $words =~ s/\W+$//;
+
+ # first try to generate headline for this entry from index
+ my $h = $ds->{'index'}->[0];
+ # then, from display
+ $h ||= $ds->{'display'}->[0];
+ # and as last resport, fallback to headline
+ $h ||= $headline;
+
$index->insert(
index_name => $ds->{'tag'},
- path => $f,
- headline => $headline,
- words => join(" ",@{$ds->{'swish'}})
+ #path => $f,
+ path => $webpac->mfn,
+ headline => $h,
+ words => unac($words),
);
}
@@ -105,17 +203,18 @@
foreach my $ds (@ds) {
next if (! $ds->{'index'});
- $thes->{$ds->{'tag'}} ||= new WebPAC::Index;
+ $thes->{$ds->{'tag'}} ||= new WebPAC::Index( name => $ds->{'tag'} );
foreach my $h (@{$ds->{'index'}}) {
$thes->{$ds->{'tag'}}->insert(
- path => $f,
+ sort_by => unac_2($h),
+ mfn => $webpac->mfn,
headline => $h,
);
}
}
-# print Dumper(\@ds);
+# print Dump(\@ds);
}
@@ -139,94 +238,156 @@
}
if (0 && $log->is_debug) {
- $log->debug("lookup hash: ",Dumper($webpac->{'lookup'}));
- $log->debug("data hash: ",Dumper($webpac->{'data'}));
+ $log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
+ $log->debug("data hash: ",Dump($webpac->{'data'}));
foreach my $t (keys %{$thes}) {
- $log->debug("thesaurus $t hash: ",Dumper($thes->{$t}));
+ $log->debug("thesaurus $t hash: ",Dump($thes->{$t}));
}
}
-$log->info("lookup hash: ",Dumper($webpac->{'lookup'}));
-
-#
-# tree dump implementation which shouldn't be here :-)
-#
-
-my $file = 'out/browse.html';
-open(HTML, "> $file") || $log->logdie("can't open '$file'");
+} # XXX if(0)
-print HTML qq{
-
-