--- trunk/all2xml.pl 2003/02/23 07:53:01 32
+++ trunk2/all2all.pl 2004/09/11 20:21:28 423
@@ -1,314 +1,232 @@
#!/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;
-
-$|=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
+=head1 NAME
-my %opts;
+all2all.pl - basic script for all WebPAC needs
-# usage:
-# -d directory name
-# -m multiple directories
-# -q quiet
-# -s run swish
+=cut
-getopts('d:m:qs', \%opts);
+use strict;
+use Data::Dumper;
+use Carp;
-my $db_dir;
+use lib './lib';
+use WebPAC;
+use WebPAC::jsFind;
+use WebPAC::Index;
+
+my $webpac = new WebPAC(
+ code_page => 'ISO-8859-2',
+ limit_mfn => 500,
+# debug => 1,
+# low_mem => 1,
+) || die;
+
+my $log = $webpac->_get_logger() || die "can't get logger";
+
+$log->debug("creating WebPAC::jsFind object");
+
+my $index = new WebPAC::jsFind(
+ index_path => './out/index',
+ keys => 10,
+) || die;
-#die "usage: $0 -d [database_dir] -m [database1,database2] " if (! %opts);
+my $thes;
-#print Dumper($config->{indexer});
-#print "-" x 70,"\n";
+$|=1;
-Text::Iconv->raise_error(1); # Conversion errors raise exceptions
+my $maxmfn = $webpac->open_isis(
+ filename => shift @ARGV || '/data/hidra/THS/THS',
+ lookup => [
+ { 'key' => 'd:v900', 'val' => 'v250^a' },
+# { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
+# { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+# { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+ { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+ { 'key' => '900_mfn:v900', 'val' => 'v000' },
+ # tree structure
+ { 'eval' => 'length("v251") == 2', 'key' => 'root:v251', 'val' => 'v900' },
+ { 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' },
+ ],
+);
+
+$log->debug("isis file ",$webpac->{'isis_filename'}," opened");
+
+$log->info("rows: $maxmfn");
+
+$webpac->open_import_xml(type => 'isis_hidra_ths');
+
+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));
+ }
+
+ next if (! @ds);
+
+ my $filename = $webpac->{'current_filename'} || $log->logdie("no current_filename in webpac object");
+
+ if ($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 $isis_codepage;
-my $index_codepage;
-my $cludge_codepage = Text::Iconv->new('UTF8','ISO8859-1');
-my $xml_codepage;
+ my $headline = $webpac->{'headline'};
-my $XML_CHARSET = 'UTF8';
+ my $f = $filename;
+ $f =~ s!out/!!;
+ # save into index
+ foreach my $ds (@ds) {
+ next if (! $ds->{'swish'});
-sub isis2xml {
+ $index->insert(
+ index_name => $ds->{'tag'},
+ path => $f,
+ headline => $headline,
+ words => join(" ",@{$ds->{'swish'}})
+ );
+ }
- use xmlify;
+ # save into sorted index (thesaurus)
+ foreach my $ds (@ds) {
+ next if (! $ds->{'index'});
- my $row = shift @_;
- my $add_xml = shift @_;
+ $thes->{$ds->{'tag'}} ||= new WebPAC::Index;
- my $xml;
+ foreach my $h (@{$ds->{'index'}}) {
+ $thes->{$ds->{'tag'}}->insert(
+ path => $f,
+ headline => $h,
+ );
+ }
+ }
- use parse_format;
+# print Dumper(\@ds);
- my $html = ""; # html formatted display output
+}
- my %field_usage; # counter for usage of each field
+foreach my $t (keys %{$thes}) {
- # sort subrouting using order="" attribute
- sub by_order {
- return 0 if (! $config->{indexer}->{$a}->{order});
- return 0 if (! $config->{indexer}->{$b}->{order});
+ my @e = $thes->{$t}->elements;
+ if (! @e) {
+ $log->logwarn("no elements in sorted index $t?");
+ next;
+ }
+
+ my $file = "./out/bfilter/$t.txt";
+ $log->info("saving sorted index $t to '$file' [".scalar(@e)." elements]");
+
+ $webpac->output_file(
+ file => $file,
+ template => 'index.tt',
+ data => \@e,
+ index_name => $t,
+ );
+}
- return $config->{indexer}->{$a}->{order} <=>
- $config->{indexer}->{$b}->{order} ;
+if (0 && $log->is_debug) {
+ $log->debug("lookup hash: ",Dumper($webpac->{'lookup'}));
+ $log->debug("data hash: ",Dumper($webpac->{'data'}));
+ foreach my $t (keys %{$thes}) {
+ $log->debug("thesaurus $t hash: ",Dumper($thes->{$t}));
}
+}
- foreach my $field (sort by_order keys %{$config->{indexer}}) {
+$log->info("lookup hash: ",Dumper($webpac->{'lookup'}));
- $field_usage{$field}++;
+#
+# tree dump implementation which shouldn't be here :-)
+#
- my $swish_data = "";
- my $display_data = "";
- my $line_delimiter = "";
-
- foreach my $x (@{$config->{indexer}->{$field}->{isis}}) {
-
- my $format = $x->{content};
- my $delimiter = $x->{delimiter} || ' ';
-
- # FIX: this is ugly, UGLY, cludge string is returned
- # in UTF8 encoding , but as if source charset
- # is ISO8859-1 and not some other. This break other
- # character encodings, so we convert it first
- # back to ISO8859-1
- $format = $xml_codepage->convert($format);
- $delimiter = $xml_codepage->convert($delimiter) if ($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");
-#print STDERR "## s: $s d: $d i: $i ## $format ##\n";
-
- $display_data .= $line_delimiter if ($display_data && $display_data !~ /$line_delimiter$/);
-
- my ($swish,$display) = (1,1);
-
- while ($swish || $display) {
- ($swish,$display) = parse_format($format,$row,$isis_i++);
-#print STDERR "s: $swish\nd: $display\n" if ($swish);
-
-#print STDERR "swish: $swish<-- display: $display<--\n";
-
- # 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) {
-#print STDERR "using filter '$filter'\n";
- no strict 'refs';
- $swish_data .= join(" ",&$filter($swish));
- } else {
- $swish_data .= $swish;
- }
- }
-
- # type="display" ; field for display
- if ($d && $display) {
- if ($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 = $index_codepage->convert($display) || $display;
- if ($filter) {
- no strict 'refs';
- foreach my $d (&$filter($index_data)) {
- $index->insert($field, $d, $db_dir);
- }
- } else {
- $index->insert($field, $index_data, $db_dir);
- }
- }
- }
- if ($x->{append}) {
- $line_delimiter = ' ';
- } else {
- $line_delimiter = '
';
- }
- }
+my $file = 'out/browse.html';
+open(HTML, "> $file") || $log->logdie("can't open '$file'");
+print HTML qq{
+
+