--- trunk/all2xml.pl 2003/02/23 08:06:07 34
+++ trunk2/all2all.pl 2004/09/05 22:22:37 411
@@ -1,317 +1,139 @@
#!/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
-
-my %opts;
-
-# usage:
-# -d directory name
-# -m multiple directories
-# -q quiet
-# -s run swish
-
-getopts('d:m:qs', \%opts);
-
-my $db_dir;
-
-#die "usage: $0 -d [database_dir] -m [database1,database2] " if (! %opts);
-
-#print Dumper($config->{indexer});
-#print "-" x 70,"\n";
-
-Text::Iconv->raise_error(1); # Conversion errors raise exceptions
-
-my $isis_codepage;
-my $index_codepage;
-my $cludge_codepage = Text::Iconv->new('UTF8','ISO8859-1');
-my $xml_codepage;
-
-my $XML_CHARSET = 'UTF8';
-
-
-sub isis2xml {
-
- use xmlify;
+=head1 NAME
- my $row = shift @_;
- my $add_xml = shift @_;
+all2all.pl - basic script for all WebPAC needs
- my $xml;
+=cut
- use parse_format;
+use strict;
+use Data::Dumper;
+use Carp;
- my $html = ""; # html formatted display output
+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,
+) || die;
+
+my $log = $webpac->_get_logger();
+
+my $index = new WebPAC::jsFind(
+ index_path => './out/index',
+ keys => 80,
+) || 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 = $webpac->open_isis(
+ filename => shift @ARGV || '/data/hidra/THS/THS',
+ lookup => [
+ { 'key' => 'd:v900', 'val' => 'v250^a' },
+# { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' },
+# { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+# { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+ { 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' },
+ { 'key' => '900_mfn:v900', 'val' => 'v000' },
+ ],
+);
+
+$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 = $webpac->{'current_filename'};
+
+ if ($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'},
+ );
}
- foreach my $field (sort by_order keys %{$config->{indexer}}) {
-
- $field_usage{$field}++;
+ my $headline = $webpac->{'headline'};
- my $swish_data = "";
- my $display_data = "";
- my $line_delimiter = "";
-
- my ($swish,$display);
-
- 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 ## $format ##\n";
-
- if ($x->{append}) {
- $line_delimiter = ' ';
- } else {
- $line_delimiter = '
';
- }
-
- $display_data .= $line_delimiter if ($display_data && $display);
- # init vars so that we go into while...
- ($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);
- }
- }
- }
- }
-
-
-#print STDERR "s_d: $swish_data\nd_d: $display_data\n" if ($swish_data);
- if ($display_data) {
-
- # remove last
- $display_data =~ s/$line_delimiter$//;
-
- $display_data = $isis_codepage->convert($display_data) || die "Can't convert '$display_data' !";
- # FIX: this is removed and replaced by html tag.
- #$xml .= xmlify($field."_display", $display_data);
-
- if ($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 .= $xml_codepage->convert($field_name);
-# $html .= "-->".$field_name."<--";
- }
- $html .= $display_data."###\n";
- }
- }
- if ($swish_data) {
- my $i = Text::Iconv->new($config->{isis_codepage},'ISO8859-2');
- # remove extra spaces
- $swish_data =~ s/ +/ /g;
- $swish_data =~ s/ +$//g;
-
- $swish_data = $i->convert($swish_data);
- $xml .= xmlify($field."_swish",unac_string('ISO8859-2',$swish_data));
- #$swish_data = $isis_codepage->convert($swish_data)."##" || $swish_data;
- #$xml .= xmlify($field."_swish",unac_string($config->{isis_codepage},$swish_data));
- }
+ my $f = $filename;
+ $f =~ s!out/!!;
+ # save into index
+ foreach my $ds (@ds) {
+ next if (! $ds->{'swish'});
+ $index->insert(
+ index_name => $ds->{'tag'},
+ path => $f,
+ headline => $headline,
+ words => join(" ",@{$ds->{'swish'}})
+ );
}
- # dump formatted output in
- if ($html) {
- $xml .= xmlify("html",$html);
- }
-
- if ($xml) {
-#print STDERR "x: $xml\n";
- $xml .= $add_xml if ($add_xml);
- return "\n$xml\n";
- } else {
- return;
- }
-}
-
-##########################################################################
-
-my $cfg = new Config::IniFiles( -file => $config_file );
+ # save into sorted index (thesaurus)
+ foreach my $ds (@ds) {
+ next if (! $ds->{'index'});
-foreach my $database ($cfg->Sections) {
+ $thes->{$ds->{'tag'}} ||= new WebPAC::Index;
- my $isis_db = $cfg -> val($database, 'isis_db');
- my $type = $cfg -> val($database, 'type');
- my $add_xml = $cfg -> val($database, 'xml');
-
- # read configuration for this type
- $config=XMLin("./import_xml/$type.xml", forcearray => [ 'isis' ], forcecontent => 1);
- $isis_codepage = Text::Iconv->new($config->{isis_codepage},$XML_CHARSET);
- $index_codepage = Text::Iconv->new($config->{isis_codepage},$config->{index_codepage});
- $xml_codepage = Text::Iconv->new($cfg->val($database,'xml_codepage'),$XML_CHARSET);
-
- my $db = OpenIsis::open( $isis_db );
- if (0) {
-# # FIX
-# if (! $db ) {
- print STDERR "WARNING: can't open '$isis_db'\n";
- next ;
+ $thes->{$ds->{'tag'}}->insert(
+ path => $f,
+ headline => $headline,
+ );
}
- my $max_rowid = OpenIsis::maxRowid( $db );
+# print Dumper(\@ds);
- print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
+}
- my $path = $database; # was $isis_db
+foreach my $t (keys %{$thes}) {
- my $last_p = 0;
+ my @e = $thes->{$t}->elements;
+ if (! @e) {
+ $log->logwarn("no elements in sorted index $t?");
+ next;
+ }
+
+ $log->debug("saving sorted index $t [".scalar(@e)." elements]");
+
+ $webpac->output_file(
+ file => "./out/thes_$t.html",
+ 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 && $row->{mfn}) {
-#print STDERR "mfn: ",$row->{mfn},"\n";
- # 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;
- }
-
- if (my $xml = isis2xml($row,$add_xml)) {
-#print STDERR "--ret-->$xml\n";
- print "Path-Name: $path#".int($row->{mfn})."\n";
- print "Content-Length: ".(length($xml)+1)."\n";
- print "Document-Type: XML\n\n$xml\n";
- }
- }
+if (0 && $log->is_debug) {
+ $log->debug("lookup hash: ",Dumper($webpac->{'lookup'}));
+ $log->debug("data hash: ",Dumper($webpac->{'data'}));
+ foreach my $t (keys %{$thes}) {
+ $log->debug("thesaurus $t hash: ",Dumper($thes->{$t}));
}
- print STDERR "\n";
}
-# call this to commit index
$index->close;
-1;
-__END__
-##########################################################################
-
-=head1 NAME
-
-isis2xml.pl - read isis file and dump XML
-
-=head1 DESCRIPTION
-
-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.
-
-=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
-
-=cut