--- trunk/all2xml.pl 2003/01/11 06:14:48 5
+++ trunk2/all2all.pl 2004/10/30 23:04:37 560
@@ -1,209 +1,377 @@
#!/usr/bin/perl -w
-use strict;
-use OpenIsis;
-use Getopt::Std;
-use Data::Dumper;
-use XML::Simple;
-use Text::Unaccent 1.02; # 1.01 won't compile on my platform,
-require Unicode::Map8;
-
-my $config=XMLin(undef, forcearray => [ 'isis' ], forcecontent => 1);
-
-my %opts;
-
-getopts('d:m:q', \%opts);
-
-my $db_dir = $opts{d} || "ps"; # FIX
-
-#die "usage: $0 -d [database_dir] -m [database1,database2] " if (! %opts);
-
-#print Dumper($config->{indexer});
-#print "-" x 70,"\n";
-
-# how to convert isis code page to UTF8?
-my $isis_map = Unicode::Map8->new($config->{isis_codepage}) || die;
-
-sub isis2xml {
+=head1 NAME
- my $row = shift @_;
+all2all.pl - basic script for all WebPAC needs
- my $xml;
+=cut
- sub isis_sf {
- my $row = shift @_;
- my $isis_id = shift @_;
- my $subfield = shift @_;
- if ($row->{$isis_id}->[0]) {
- my $sf = OpenIsis::subfields($row->{$isis_id}->[0]);
- if (! defined $subfield || length($subfield) == 0) {
- # subfield list undef, empty or no defined subfields for this record
- my $all_sf = $row->{$isis_id}->[0];
- $all_sf =~ s/\^./ /g; nuke definirions
- return $all_sf;
- } elsif ($sf->{$subfield}) {
- return $sf->{$subfield};
- }
- }
+use strict;
+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 $tmp = shift || return;
+ return undef unless ($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;
+ $tmp =~ s/\s*CROVOC.*$/ #C#/;
+ $tmp =~ s/\s*EUROVOC.*//;
+ return $tmp;
+ },
+ # TT filter
+ 'CROVOC_img' => sub {
+ my $tmp = shift;
+ $tmp =~ s/\s*#C#\s*/ /gis;
+ return $tmp;
}
+};
- foreach my $field (keys %{$config->{indexer}}) {
-
- my $display_data = "";
- my $index_data = "";
-
- foreach my $x (@{$config->{indexer}->{$field}->{isis}}) {
-
- my $display_tmp = "";
- my $index_tmp = "";
-
- my $format = $x->{content};
- my $i = 1; # index only
- my $d = 1; # display only
- $i = 0 if (lc($x->{type}) eq "display");
- $d = 0 if (lc($x->{type}) eq "index");
-#print "## i: $i d: $d ## $format ##";
- # parse format
- my $prefix = "";
- if ($format =~ s/^([^\d]+)//) {
- $prefix = $1;
- }
- while ($format) {
- if ($format =~ s/^(\d\d\d)(\w?)//) {
- my $isis_tmp = isis_sf($row,$1,$2);
- if ($isis_tmp) {
-# $display_tmp .= $prefix . "/$1/$2/".$isis_tmp if ($d);
- $display_tmp .= $prefix . $isis_tmp if ($d);
- $index_tmp .= $isis_tmp." " if ($i);
-#print " $isis_tmp <--\n";
- }
- $prefix = "";
- } elsif ($format =~ s/^([^\d]+)//) {
- $prefix = $1;
- } else {
- print STDERR "WARNING: unparsed format '$format'\n";
- last;
- };
- }
- # add suffix
- $display_tmp .= $prefix if ($display_tmp);
-
-# $display_data .= $display_tmp if ($display_tmp ne "");
-# $index_data .= $index_tmp if ($index_tmp ne "");
- $display_data .= $display_tmp;
- $index_data .= $index_tmp;
-
- }
-#print "--display:$display_data\n--index:$index_data\n";
- $xml->{$field}->{display} .= $isis_map->tou($display_data)->utf8 if ($display_data);
- $xml->{$field}->{index} .= unac_string($config->{isis_codepage},$index_data) if ($index_data);
-
- }
- if ($xml) {
- return XMLout($xml, rootname => 'xml', noattr => 1 );
- } else {
- return;
+## 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;
-my $last_tell=0;
+ $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;/$1/gi;
-my @isis_dirs = ( '.' ); # use dirname as database name
-
-if ($opts{m}) {
- @isis_dirs = split(/,/,$opts{m});
+ return $ent;
}
-my @isis_dbs;
-
-foreach (@isis_dirs) {
- if (-e $config->{isis_data}."/$db_dir/$_/LIBRI") {
- push @isis_dbs,$config->{isis_data}."/$db_dir/$_/LIBRI/LIBRI";
- }
- if (-e $config->{isis_data}."/$db_dir/$_/PERI") {
- push @isis_dbs,$config->{isis_data}."/$db_dir/$_/PERI/PERI";
+# create WebPAC object
+#
+my $webpac = new WebPAC(
+ code_page => $code_page,
+ limit_mfn => $limit_mfn,
+ start_mfn => $start_mfn,
+ debug => $debug,
+ low_mem => $low_mem,
+ filter => $filter,
+) || die;
+
+my $log = $webpac->_get_logger() || die "can't get logger";
+
+$log->debug("creating WebPAC::jsFind object");
+
+my $index = new WebPAC::jsFind(
+ index_path => $index_path,
+ keys => 62,
+) || die;
+
+my $thes;
+
+$|=1;
+
+my $maxmfn = $webpac->open_isis(
+ filename => shift @ARGV || '/data/hidra/THS/THS',
+ lookup => [
+ { 'key' => 'd:v900', 'val' => 'filter{CROVOC_tree}v250^a v800' },
+# { 'eval' => '"v901^a" eq "Podruèje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
+# { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+# { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+ { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+ { 'key' => '900_mfn:v900', 'val' => 'v000' },
+ # tree structure
+ { 'eval' => 'length("v251") == 2 && "v800" =~ m/EUROVOC/ || "v800" =~ m/CROVOC/ && "v251" =~ m/^(H|HD|L|Z|P)$/', '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' },
+ ],
+);
+
+$log->debug("isis file ",$webpac->{'isis_filename'}," opened");
+
+$log->info("rows: $maxmfn");
+
+$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) {
+ $log->debug("rec = ",Dump($rec));
+ $log->debug("ds = ",Dump(\@ds));
}
- if (-e $config->{isis_data}."/$db_dir/$_/AMS") {
- push @isis_dbs,$config->{isis_data}."/$db_dir/$_/AMS/AMS";
- }
- if (-e $config->{isis_data}."/$db_dir/$_/ARTI") {
-# push @isis_dbs,$config->{isis_data}."/$db_dir/$_/ARTI/ARTI";
- }
-}
-print STDERR "FATAL: Can't find isis database.\nPerhaps isis_data (".$config->{isis_data}.") has wrong value?\n" if (! @isis_dbs);
+ next if (! @ds);
-my $db;
+ my $filename = $webpac->{'current_filename'} || $log->logdie("no current_filename in webpac object");
-foreach my $isis_db (@isis_dbs) {
-
-
- my $db = OpenIsis::open( $isis_db );
- if (0) {
-# # FIX
-# if (! $db ) {
- print STDERR "WARNING: can't open '$isis_db'\n";
- next ;
+ if ($filename) {
+ $webpac->output_file(
+ file => $filename,
+ template => 'html.tt',
+ data => \@ds,
+ headline => $webpac->{'headline'},
+ );
+ } else {
+ print $webpac->output(
+ template => 'text.tt',
+ data => \@ds,
+ headline => $webpac->{'headline'},
+ );
}
- my $max_rowid = OpenIsis::maxRowid( $db );
-
- print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
-
- my $last_p = 0;
+ my $headline = $webpac->{'headline'};
-# { my $row_id = 1;
-# FIX
- for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) {
- my $row = OpenIsis::read( $db, $row_id );
- if ($row && $row->{mfn}) {
+ my $f = $filename;
+ $f =~ s!out/!!;
- # output current process indicator
- my $p = int($row->{mfn} * 100 / $max_rowid);
- if ($p != $last_p) {
- printf STDERR ("%5d / %5d [%-51s] %-2d %% \r",$row->{mfn},$max_rowid,"=" x ($p/2).">", $p ) if (! $opts{q});
- $last_p = $p;
- }
+ # save into index
+ 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,
+ path => $webpac->mfn,
+ headline => $h,
+ words => unac($words),
+ );
+ }
- if (my $xml = isis2xml($row)) {
- print "Path-Name: $isis_db#".$row->{mfn}."\n";
- print "Content-Length: ".(length($xml)+1)."\n";
- print "Document-Type: XML\n\n$xml\n";
- }
+ # save into sorted index (thesaurus)
+ foreach my $ds (@ds) {
+ next if (! $ds->{'index'});
+
+ $thes->{$ds->{'tag'}} ||= new WebPAC::Index( name => $ds->{'tag'} );
+
+ foreach my $h (@{$ds->{'index'}}) {
+ $thes->{$ds->{'tag'}}->insert(
+ sort_by => unac_2($h),
+ mfn => $webpac->mfn,
+ headline => $h,
+ );
}
}
- print STDERR "\n";
-}
+# print Dump(\@ds);
-1;
-__END__
-##########################################################################
+}
-=head1 NAME
+foreach my $t (keys %{$thes}) {
+
+ my @e = $thes->{$t}->elements;
+ if (! @e) {
+ $log->logwarn("no elements in sorted index $t?");
+ next;
+ }
-isis2xml.pl - read isis file and dump XML
+ my $file = "./out/bfilter/$t.txt";
+ $log->info("saving sorted index $t to '$file' [".scalar(@e)." elements]");
-=head1 DESCRIPTION
+ $webpac->output_file(
+ file => $file,
+ template => 'index.tt',
+ data => \@e,
+ index_name => $t,
+ );
+}
-This command will read ISIS data file using OpenIsis perl module and
-create XML file for usage with I
-indexer. Dispite it's name, this script B
-from isis files (isis allready has something like that). Output of this
-script is tailor-made for SWISH-E.
+if (0 && $log->is_debug) {
+ $log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
+ $log->debug("data hash: ",Dump($webpac->{'data'}));
+ foreach my $t (keys %{$thes}) {
+ $log->debug("thesaurus $t hash: ",Dump($thes->{$t}));
+ }
+}
-=head1 AUTHOR
+} # XXX if(0)
-Dobrica Pavlinusic
+#$log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
-=head1 COPYRIGHT
+$log->info("creating tree");
-GNU Public License (GPL) v2 or later
+#
+# define tree structure
+#
+
+my $l = $webpac->{'lookup'} || $log->logconfess("can't find lookup");
+
+my @tree = ({
+ # level 0
+ code_arr => sub { sort keys %{$l} },
+ filter_code => sub {
+ my $t = shift;
+ return $t if ($t =~ s/root://);
+ },
+ lookup_v900 => sub { shift @{$l->{"root:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[1]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[1]}} },
+ have_children => sub { return $l->{"a:".$_[0]."::"} },
+ have_children_at_level => sub {
+ return unless (defined($l->{"code:".$_[1]}));
+ my $code = shift @{$l->{"code:".$_[1]}};
+ print STDERR "## $_[1] -> $code\n";
+ return undef unless($code);
+ return(9, $l->{"a:$code:"} ) if (defined($l->{"a:$code:"}));
+ },
+ style => 'display: none',
+ },{
+ # 1
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift }, # nop
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":"} },
+ style => 'display: none',
+ },{
+ # 2
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ #style => 'display: none',
+ },{
+ # 3 u¾i pojam
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ },{
+ # 4
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ },{
+ # 5
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ },{
+ # 6
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ },{
+ # 7
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ },{
+ # 8
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+# have_children => sub { return $l->{"a:".$_[1].":".$_[0]} },
+ have_children => sub { 0 },
+ },{
+ # 9 - level which is never reached except explicitly
+ code_arr => sub { @{$_[0]} },
+ filter_code => sub { shift },
+ lookup_v900 => sub { shift @{$l->{"code:".$_[0]}} },
+ lookup_term => sub { shift @{$l->{"d:".$_[0]}} },
+ lookup_mfn => sub { shift @{$l->{"900_mfn:".$_[0]}} },
+ have_children => sub { 0 },
+ have_children_at_level => sub { defined($l->{"a:".$_[1].":".$_[0]}) && return (9,$l->{"a:".$_[1].":".$_[0]}) },
+ },{
+});
+
+my $tree = new WebPAC::Tree(
+ tree => \@tree,
+);
+
+$tree->output(
+ dir => './out',
+ html => 'browse.html',
+ template => './output_template/tree.tt',
+ js => 'tree-ids.js',
+);
+
+$tree->output(
+ dir => './eurovoc',
+ html => 'hijerarhija.html',
+ template => './output_template/hijerarhija.tt',
+ js => 'tree-ids.js',
+);
-=head1 SEE ALSO
-SWISH-E web site at http://www.swish-e.org
+$log->info("closing index");
+$index->close;
-=cut
+$log->info("elapsed time: ",$webpac->fmt_time(time() - $webpac->{'start_t'}));