--- trunk/all2xml.pl 2003/03/22 22:51:48 44
+++ trunk2/all2all.pl 2004/10/26 21:44:57 550
@@ -1,300 +1,368 @@
#!/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,
-use Text::Iconv;
-use Config::IniFiles;
-use Encode;
-
-$|=1;
-
-my $config_file = $0;
-$config_file =~ s/\.pl$/.conf/;
-die "FATAL: can't find configuration file '$config_file'" if (! -e $config_file);
-
-my $config;
-
-use index_DBI; # there is no other, right now ;-)
-
-my $index = new index_DBI(); # open index
-
-my %opts;
-
-# usage:
-# -d directory name
-# -m multiple directories
-# -q quiet
-# -s run swish
-
-getopts('d:m:qs', \%opts);
-
-my $path; # this is name of database
+=head1 NAME
-Text::Iconv->raise_error(1); # Conversion errors raise exceptions
+all2all.pl - basic script for all WebPAC needs
-# this is encoding of all files on disk, including import_xml/*.xml file and
-# filter/*.pm files! It will be used to store strings in perl internally!
-my $codepage = 'ISO-8859-2';
+=cut
-my $utf2cp = Text::Iconv->new('UTF-8',$codepage);
-# this function will convert data from XML files to local encoding
-sub x {
- return $utf2cp->convert($_[0]);
+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.*$/ */);
+ return $tmp;
+ },
+ 'CROVOC_tree' => sub {
+ my $tmp = shift || return;
+ $tmp =~ s/\s*CROVOC.*$/ */;
+ $tmp =~ s/\s*EUROVOC.*//;
+ 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);
}
-# decode isis import codepage
-my $isis2cp;
-
-# outgoing xml must be in UTF-8
-my $cp2utf = Text::Iconv->new($codepage,'UTF-8');
-
-sub isis2xml {
-
- use xmlify;
-
- my $row = shift @_;
- my $add_xml = shift @_;
+sub unac_entities {
+ my $ent = shift || return;
- my $xml;
+ $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;
- use parse_format;
+ return $ent;
+}
- my $html = ""; # html formatted display output
+# 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 => 10,
+) || die;
- my %field_usage; # counter for usage of each field
+my $thes;
- # sort subrouting using order="" attribute
- sub by_order {
- return 0 if (! $config->{indexer}->{$a}->{order});
- return 0 if (! $config->{indexer}->{$b}->{order});
+$|=1;
- return $config->{indexer}->{$a}->{order} <=>
- $config->{indexer}->{$b}->{order} ;
+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));
}
- foreach my $field (sort by_order keys %{$config->{indexer}}) {
-
- $field=x($field);
-
- $field_usage{$field}++;
-
- my $swish_data = "";
- my $display_data = "";
- my $line_delimiter;
-
- my ($swish,$display);
-
- foreach my $x (@{$config->{indexer}->{$field}->{isis}}) {
-
- my $format = x($x->{content});
- my $delimiter = x($x->{delimiter}) || ' ';
-
- my $isis_i = 0; # isis repeatable offset
-
- my ($s,$d,$i) = (1,1,0); # swish, display default
- $s = 0 if (lc($x->{type}) eq "display");
- $d = 0 if (lc($x->{type}) eq "swish");
- ($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index");
-
- # what will separate last line from this one?
- if ($display_data && $x->{append} && $x->{append} eq "1") {
- $line_delimiter = ' ';
- } elsif ($display_data) {
- $line_delimiter = '
';
- }
-
- # init vars so that we go into while...
- ($swish,$display) = (1,1);
-
- while ($swish || $display) {
- ($swish,$display) = parse_format($format,$row,$isis_i++,$isis2cp);
-
- # filter="name" ; filter this field through
- # filter/[name].pm
- my $filter = $x->{filter};
- if ($filter) {
- require "filter/".$filter.".pm";
- }
- # type="swish" ; field for swish
- if ($s && $swish) {
- if ($filter) {
- no strict 'refs';
- $swish_data .= join(" ",&$filter($swish));
- } else {
- $swish_data .= $swish;
- }
- }
-
- # type="display" ; field for display
- if ($d && $display) {
- if ($line_delimiter && $display_data) {
- $display_data .= $line_delimiter;
- undef $line_delimiter;
- }
- if ($filter) {
- no strict 'refs';
- $display_data .= join($delimiter,&$filter($display));
- } else {
- if ($display_data) {
- $display_data .= $delimiter.$display;
- } else {
- $display_data .= $display;
- }
- }
- }
-
- # type="index" ; insert into index
- if ($i && $display) {
- my $index_data = $display;
- if ($filter) {
- no strict 'refs';
- foreach my $d (&$filter($index_data)) {
- $index->insert($field, $d, $path);
- }
- } else {
- $index->insert($field, $index_data, $path);
- }
- }
- }
- }
-
+ next if (! @ds);
- if ($display_data) {
+ my $filename = $webpac->{'current_filename'} || $log->logdie("no current_filename in webpac object");
- if ($field eq "headline") {
- $xml .= xmlify("headline", $display_data);
- } else {
-
- # find field name (signular, plural)
- my $field_name = "";
- if ($config->{indexer}->{$field}->{name_singular} && $field_usage{$field} == 1) {
- $field_name = $config->{indexer}->{$field}->{name_singular}."#-#";
- } elsif ($config->{indexer}->{$field}->{name_plural}) {
- $field_name = $config->{indexer}->{$field}->{name_plural}."#-#";
- } elsif ($config->{indexer}->{$field}->{name}) {
- $field_name = $config->{indexer}->{$field}->{name}."#-#";
- } else {
- print STDERR "WARNING: field '$field' doesn't have 'name' attribute!";
- }
- if ($field_name) {
- $html .= x($field_name);
- }
- $html .= $display_data."###\n";
- }
- }
- if ($swish_data) {
- # remove extra spaces
- $swish_data =~ s/ +/ /g;
- $swish_data =~ s/ +$//g;
+ 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'},
+ );
+ }
- $xml .= xmlify($field."_swish", unac_string($codepage,$swish_data));
- }
+ my $headline = $webpac->{'headline'};
+ my $f = $filename;
+ $f =~ s!out/!!;
+ # 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),
+ );
}
- # dump formatted output in
- if ($html) {
- $xml .= xmlify("html",$html);
- }
-
- if ($xml) {
- $xml .= $add_xml if ($add_xml);
- return "\n$xml\n";
- } else {
- return;
+ # 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,
+ );
+ }
}
-}
-
-##########################################################################
-
-my $cfg = new Config::IniFiles( -file => $config_file );
-foreach my $database ($cfg->Sections) {
+# print Dump(\@ds);
- my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!";
- my $type = $cfg -> val($database, 'type') || die "$database doesn't have 'type' defined";
- my $add_xml = $cfg -> val($database, 'xml'); # optional
-
- $config=XMLin("./import_xml/$type.xml", forcearray => [ 'isis' ], forcecontent => 1);
+}
- $isis2cp = Text::Iconv->new($config->{isis_codepage},$codepage);
+foreach my $t (keys %{$thes}) {
- my $db = OpenIsis::open( $isis_db );
- if (0) {
-# # FIX
-# if (! $db ) {
- print STDERR "WARNING: can't open '$isis_db'\n";
- next ;
+ my @e = $thes->{$t}->elements;
+ if (! @e) {
+ $log->logwarn("no elements in sorted index $t?");
+ next;
}
- my $max_rowid = OpenIsis::maxRowid( $db );
-
- print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
+ my $file = "./out/bfilter/$t.txt";
+ $log->info("saving sorted index $t to '$file' [".scalar(@e)." elements]");
- my $path = $database; # was $isis_db
-
- my $last_p = 0;
+ $webpac->output_file(
+ file => $file,
+ template => 'index.tt',
+ data => \@e,
+ index_name => $t,
+ );
+}
-# { my $row_id = 4514;
-# FIX
- for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) {
- my $row = OpenIsis::read( $db, $row_id );
- if ($row && $row->{mfn}) {
- # 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;
- }
-
- my $swishpath = $path."#".int($row->{mfn});
-
- if (my $xml = isis2xml($row,$add_xml)) {
- $xml = $cp2utf->convert($xml);
- use bytes; # as opposed to chars
- print "Path-Name: $swishpath\n";
- print "Content-Length: ".(length($xml)+1)."\n";
- print "Document-Type: XML\n\n$xml\n";
- }
- }
+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->{$t}));
}
- print STDERR "\n";
}
-# call this to commit index
-$index->close;
-
-1;
-__END__
-##########################################################################
-
-=head1 NAME
+} # XXX if(0)
-isis2xml.pl - read isis file and dump XML
+#$log->debug("lookup hash: ",Dump($webpac->{'lookup'}));
-=head1 DESCRIPTION
+$log->info("creating tree");
-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.
+#
+# 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',
+);
+
+$tree->output(
+ dir => './eurovoc',
+ html => 'hijerarhija.html',
+ template => './output_template/hijerarhija.tt',
+ js => 'tree-ids.js',
+);
-=head1 AUTHOR
-Dobrica Pavlinusic
-
-=head1 COPYRIGHT
-
-GNU Public License (GPL) v2 or later
-
-=head1 SEE ALSO
-
-SWISH-E web site at http://www.swish-e.org
+$log->info("closing index");
+$index->close;
-=cut
+$log->info("elapsed time: ",$webpac->fmt_time(time() - $webpac->{'start_t'}));