--- trunk/all2xml.pl 2003/03/15 21:48:48 42
+++ trunk2/all2all.pl 2004/09/11 20:21:28 423
@@ -1,298 +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;
-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 ;-)
+=head1 NAME
-my $index = new index_DBI(); # open index
+all2all.pl - basic script for all WebPAC needs
-my %opts;
+=cut
-# usage:
-# -d directory name
-# -m multiple directories
-# -q quiet
-# -s run swish
+use strict;
+use Data::Dumper;
+use Carp;
-getopts('d:m:qs', \%opts);
+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;
-my $db_dir;
+my $thes;
-Text::Iconv->raise_error(1); # Conversion errors raise exceptions
+$|=1;
-# 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';
+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 $utf2cp = Text::Iconv->new('UTF-8',$codepage);
-# this function will convert data from XML files to local encoding
-sub x {
- return $utf2cp->convert($_[0]);
-}
+ my $headline = $webpac->{'headline'};
-# decode isis import codepage
-my $isis2cp;
+ my $f = $filename;
+ $f =~ s!out/!!;
-# outgoing xml must be in UTF-8
-my $cp2utf = Text::Iconv->new($codepage,'UTF-8');
+ # 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}}) {
-
- $field=x($field);
+$log->info("lookup hash: ",Dumper($webpac->{'lookup'}));
- $field_usage{$field}++;
+#
+# tree dump implementation which shouldn't be here :-)
+#
- my $swish_data = "";
- my $display_data = "";
- my $line_delimiter;
+my $file = 'out/browse.html';
+open(HTML, "> $file") || $log->logdie("can't open '$file'");
- my ($swish,$display);
+print HTML qq{
+
+
+ Browse
+
+
+
+
+
+};
- foreach my $x (@{$config->{indexer}->{$field}->{isis}}) {
+foreach my $code (sort keys %{$webpac->{'lookup'}}) {
- my $format = x($x->{content});
- my $delimiter = x($x->{delimiter}) || ' ';
+ my $l = $webpac->{'lookup'} || $log->logconfess("can't find lookup");
- my $isis_i = 0; # isis repeatable offset
+ if ($code =~ m/^root:/) {
+ my $v900 = shift @{ $l->{$code} } || $log->logconfess("can't lookup '$code'");
+ $code =~ s/^root:// || die;
- my ($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");
+ my $term = shift @{ $l->{"d:${v900}"} } || die;
+ my $mfn = shift @{ $l->{"900_mfn:${v900}"} } || die;
- # what will separate last line from this one?
- if ($display_data && $x->{append} && $x->{append} eq "1") {
- $line_delimiter = ' ';
- } elsif ($display_data) {
- $line_delimiter = '
';
- }
+ $log->debug("$code -> $v900 : $term [$mfn]");
- # init vars so that we go into while...
- ($swish,$display) = (1,1);
+ print HTML qq{ - $term »
\n} if (-e "out/thes/$mfn.html");
- 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));
-print STDERR "#### $swish_data\n";
- } 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, $db_dir);
- }
- } else {
- $index->insert($field, $index_data, $db_dir);
- }
- }
- }
+ unless ($l->{"a:${code}::"}) {
+ $log->warn("can't find 'a:${code}::'");
+ next;
}
+ print HTML qq{ \n};
- if ($display_data) {
-
- if ($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) {
- $html .= x($field_name);
- }
- $html .= $display_data."###\n";
+ foreach my $l2_v900 (@{ $l->{"a:${code}::"} }) {
+
+ my $l2_code = shift @{ $l->{"code:${l2_v900}"} } || die;
+ my $l2_term = shift @{ $l->{"d:$l2_v900"} } || die;
+ my $l2_mfn = shift @{ $l->{"900_mfn:${l2_v900}"} } || die;
+
+ $log->debug("$l2_code -> $l2_v900 : $l2_term [$l2_mfn]");
+
+ print HTML qq{ - $l2_term »
\n} if (-e "out/thes/$l2_mfn.html");
+
+ unless ($l->{"a:${l2_code}:"}) {
+ $log->warn("can't find 'a:${l2_code}:'");
+ next;
}
- }
- if ($swish_data) {
- # remove extra spaces
- $swish_data =~ s/ +/ /g;
- $swish_data =~ s/ +$//g;
-
- $xml .= xmlify($field."_swish", unac_string($codepage,$swish_data));
- }
-
- }
-
- # dump formatted output in
- if ($html) {
- $xml .= xmlify("html",$html);
- }
-
- if ($xml) {
- $xml .= $add_xml if ($add_xml);
- return "\n$xml\n";
- } else {
- return;
- }
-}
-
-##########################################################################
-
-my $cfg = new Config::IniFiles( -file => $config_file );
-
-foreach my $database ($cfg->Sections) {
-
- 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);
-
- my $db = OpenIsis::open( $isis_db );
- if (0) {
-# # FIX
-# if (! $db ) {
- print STDERR "WARNING: can't open '$isis_db'\n";
- next ;
- }
-
- my $max_rowid = OpenIsis::maxRowid( $db );
+ print HTML qq{ \n};
- print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
+ foreach my $l3_v900 (@{ $l->{"a:${l2_code}:"} }) {
- my $path = $database; # was $isis_db
+ my $l3_term = shift @{ $l->{"d:$l3_v900"} } || die;
+ my $l3_mfn = shift @{ $l->{"900_mfn:${l3_v900}"} } || die;
- my $last_p = 0;
+ $log->debug(" -> $l3_v900 : $l3_term [$l3_mfn]");
-# { 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;
+ print HTML qq{ - $l3_term »
\n} if (-e "out/thes/$l3_mfn.html");
}
- if (my $xml = $cp2utf->convert(isis2xml($row,$add_xml))) {
- use bytes; # as opposed to chars
- print "Path-Name: $path#".int($row->{mfn})."\n";
- print "Content-Length: ".(length($xml)+1)."\n";
- print "Document-Type: XML\n\n$xml\n";
- }
+ print HTML qq{
\n};
}
+ print HTML qq{
\n};
}
- print STDERR "\n";
}
-# call this to commit index
-$index->close;
-
-1;
-__END__
-##########################################################################
-
-=head1 NAME
-
-isis2xml.pl - read isis file and dump XML
-
-=head1 DESCRIPTION
-
-This command will read ISIS data file using OpenIsis perl module and
-create XML file for usage with I
-indexer. Dispite it's name, this script B
-from isis files (isis allready has something like that). Output of this
-script is tailor-made for SWISH-E.
+print HTML qq{
+
+
+
+};
-=head1 AUTHOR
-
-Dobrica Pavlinusic
-
-=head1 COPYRIGHT
-
-GNU Public License (GPL) v2 or later
-
-=head1 SEE ALSO
-
-SWISH-E web site at http://www.swish-e.org
+$index->close;
-=cut
+$log->info("elapsed time: ",$webpac->fmt_time(time() - $webpac->{'start_t'}));