--- trunk/all2xml.pl	2003/10/29 21:27:00	135
+++ trunk2/all2all.pl	2004/07/21 16:18:45	390
@@ -1,576 +1,104 @@
 #!/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;		# default DBI module for index
-use index_DBI_cache;	# faster DBI module using memory cache
-my $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(0);     # Conversion errors don't 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 Data::Dumper;
+use Carp;
 
-# decode isis/excel or other import codepage
-my $import2cp;
+use lib './lib';
+use WebPAC;
+use WebPAC::jsFind;
+
+my $webpac = new WebPAC(
+	code_page => 'ISO-8859-2',
+	limit_mfn => 100,
+#	debug => 1,
+) || die;
+
+my $log = $webpac->_get_logger();
+
+my $index = new WebPAC::jsFind(
+	index_path => './out/index',
+) || die;
 
-# outgoing xml must be in UTF-8
-my $cp2utf = Text::Iconv->new($codepage,'UTF-8');
+$|=1;
 
-# mapping between data type and tag which specify
-# format in XML file
-my %type2tag = (
-	'isis' => 'isis',
-	'excel' => 'column',
-	'marc' => 'marc',
-	'feed' => 'feed'
+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' },
+	],
 );
 
-sub data2xml {
-
-	use xmlify;
-
-	my $type = shift @_;
-	my $row = shift @_;
-	my $add_xml = shift @_;
-	# needed to read values from configuration file
-	my $cfg = shift @_;
-	my $database = shift @_;
+$log->info("rows: $maxmfn");
 
-	my $xml;
+$webpac->open_import_xml(type => 'isis_hidra_ths');
 
-	use parse_format;
+while (my $rec = $webpac->fetch_rec) {
 
-	my $html = "";		# html formatted display output
+	my @ds = $webpac->data_structure($rec);
 
-	my %field_usage;	# counter for usage of each field
-
-	# sort subrouting using order="" attribute
-	sub by_order {
-		my $va = $config->{indexer}->{$a}->{order} ||
-			$config->{indexer}->{$a};
-		my $vb = $config->{indexer}->{$b}->{order} ||
-			$config->{indexer}->{$b};
-
-		return $va <=> $vb;
+	if ($log->is_debug) {
+		$log->debug("rec = ",Dumper($rec));
+		$log->debug("ds = ",Dumper(\@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);
-
-		my $tag = $type2tag{$type} || die "can't find which tag to use for type $type";
-		foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) {
-
-			my $format = x($x->{content});
-			my $delimiter = x($x->{delimiter}) || ' ';
-
-			my $repeat_off = 0;		# repeatable offset
-
-			my ($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) {
-				$line_delimiter = '<br/>';
-			}
-
-			# init vars so that we go into while...
-			($swish,$display) = (1,1);
-
-			# placeholder for all repeatable entries for index
-			my @index_data;
-			my $index_filter;
-
-			# while because of repeatable fields
-			while ($swish || $display) {
-				($swish,$display) = parse_format($type, $format,$row,$repeat_off++,$import2cp);
-				if ($repeat_off > 1000) {
-					print STDERR "loop (more than 1000 repeatable fields) deteced in $row, $format\n";
-					last;
-				}
-				
-				# 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) {
-					push @index_data, $display;
-					$index_filter = $filter if ($filter);
-				}
-			}
-
-			# fill data in index
-			if (@index_data) {
-				if ($index_filter) {
-					no strict 'refs';
-					foreach my $d (@index_data) {
-						$index->insert($field, &$index_filter($d), $path);
-					}
-				} else {
-					foreach my $d (@index_data) {
-						$index->insert($field, $d, $path);
-					}
-				}
-			}
-		}
-
-		# now try to parse variables from configuration file
-		foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) {
+	next if (! @ds);
 
-			my $delimiter = x($x->{delimiter}) || ' ';
-			my $val = $cfg->val($database, x($x->{content}));
+	my $filename = $webpac->{'current_filename'};
 
-			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");
-
-			if ($val) {
-				$display_data .= $delimiter.$val if ($d);
-				$swish_data .= $val if ($s);
-				$index->insert($field, $val, $path) if ($i);
-			}
-
-		}
-
-
-		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 .= $display_data."###\n";
-			}
-		}
-		if ($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 <html>
-	if ($html) {
-		#$xml .= xmlify("html",$html);
-		$xml .= "<html><![CDATA[ $html ]]></html>";
-	}
-	
-	if ($xml) {
-		$xml .= $add_xml if ($add_xml);
-		return "<xml>\n$xml</xml>\n";
+	if ($filename) {
+		open(OUT,"> $filename") || $log->logdie("can't open output '$filename': $!");
+		print OUT $webpac->output(
+			template => 'html.tt',
+			data => \@ds,
+		);
+		close(OUT);
 	} else {
-		return;
-	}
-}
-
-##########################################################################
-
-# read configuration for this script
-my $cfg = new Config::IniFiles( -file => $config_file );
-
-# read global.conf configuration
-my $cfg_global = new Config::IniFiles( -file => 'global.conf' );
-
-# open index
-$index = new index_DBI(
-		$cfg_global->val('global', 'dbi_dbd'),
-		$cfg_global->val('global', 'dbi_dsn'),
-		$cfg_global->val('global', 'dbi_user'),
-		$cfg_global->val('global', 'dbi_passwd') || '',
-	);
-
-my $show_progress = $cfg_global->val('global', 'show_progress');
-
-foreach my $database ($cfg->Sections) {
-
-	my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined";
-	my $add_xml = $cfg -> val($database, 'xml');	# optional
-
-print STDERR "reading ./import_xml/$type.xml\n";
-
-	# extract just type basic
-	my $type_base = $type;
-	$type_base =~ s/_.+$//g;
-
-	$config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type_base}, 'config' ], forcecontent => 1);
-
-	# output current progress indicator
-	my $last_p = 0;
-	sub progress {
-		return if (! $show_progress);
-		my $current = shift;
-		my $total = shift || 1;
-		my $p = int($current * 100 / $total);
-		if ($p != $last_p) {
-			printf STDERR ("%5d / %5d [%-51s] %-2d %% \r",$current,$total,"=" x ($p/2).">", $p );
-			$last_p = $p;
-		}
+		print $webpac->output(
+			template => 'text.tt',
+			data => \@ds,
+		);
 	}
 
-	my $fake_dir = 1;
-	sub fakeprogress {
-		return if (! $show_progress);
-		my $current = shift @_;
-
-		my @ind = ('-','\\','|','/','-','\\','|','/', '-');
-
-		$last_p += $fake_dir;
-		$fake_dir = -$fake_dir if ($last_p > 1000 || $last_p < 0);
-		if ($last_p % 10 == 0) {
-			printf STDERR ("%5d / %5s [%-51s]\r",$current,"?"," " x ($last_p/20).$ind[($last_p/20) % $#ind]);
+	my $headline;
+	foreach my $ds (@ds) {
+		if ($ds->{'tag'} eq 'headline') {
+			$headline = join(" ",@{$ds->{'display'}});
+			last;
 		}
 	}
 
-	# now read database
-print STDERR "using: $type...\n";
-
-	if ($type_base eq "isis") {
-
-		my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!";
-
-		$import2cp = Text::Iconv->new($config->{isis_codepage},$codepage);
-		my $db = OpenIsis::open( $isis_db );
-
-		# check if .txt database for OpenIsis is zero length,
-		# if so, erase it and re-open database
-		sub check_txt_db {
-			my $isis_db = shift || die "need isis database name";
-			my $reopen = 0;
-
-			if (-e $isis_db.".TXT") {
-				print STDERR "WARNING: removing $isis_db.TXT OpenIsis database...\n";
-				unlink $isis_db.".TXT" || warn "FATAL: unlink error on '$isis_db.TXT': $!";
-				$reopen++;
-			}
-			if (-e $isis_db.".PTR") {
-				print STDERR "WARNING: removing $isis_db.PTR OpenIsis database...\n";
-				unlink $isis_db.".PTR" || warn "FATAL: unlink error on '$isis_db.PTR': $!";
-				$reopen++;
-			}
-			return OpenIsis::open( $isis_db ) if ($reopen);
-		}
-
-		# EOF error
-		if ($db == -1) {
-			$db = check_txt_db($isis_db);
-			if ($db == -1) {
-				print STDERR "FATAL: OpenIsis can't open zero size file $isis_db\n";
-				next;
-			}
-		}
-
-		# OpenIsis::ERR_BADF 
-		if ($db == -4) {
-			print STDERR "FATAL: OpenIsis can't find file $isis_db\n";
-			next;
-		# OpenIsis::ERR_IO
-		} elsif ($db == -5) {
-			print STDERR "FATAL: OpenIsis can't access file $isis_db\n";
-			next;
-		} elsif ($db < 0) {
-			print STDERR "FATAL: OpenIsis unknown error $db with file $isis_db\n";
-			next;
-		}
-
-		my $max_rowid = OpenIsis::maxRowid( $db );
-
-		# if 0 records, try to rease isis .txt database
-		if ($max_rowid == 0) {
-			# force removal of database
-			$db = check_txt_db($isis_db);
-			$max_rowid = OpenIsis::maxRowid( $db );
-		}
-
-		print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
-
-		my $path = $database;
-
-		for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) {
-			my $row = OpenIsis::read( $db, $row_id );
-			if ($row && $row->{mfn}) {
-	
-				progress($row->{mfn}, $max_rowid);
-
-				my $swishpath = $path."#".int($row->{mfn});
-
-				if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) {
-					$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$xml\n";
-				}
-			}
-		}
-		# for this to work with current version of OpenIsis (0.9.0)
-		# you might need my patch from
-		# http://www.rot13.org/~dpavlin/projects/openisis-0.9.0-perl_close.diff
-		OpenIsis::close($db);
-		print STDERR "\n";
-
-	} elsif ($type_base eq "excel") {
-		use Spreadsheet::ParseExcel;
-		use Spreadsheet::ParseExcel::Utility qw(int2col);
-		
-		$import2cp = Text::Iconv->new($config->{excel_codepage},$codepage);
-		my $excel_file = $cfg -> val($database, 'excel_file') || die "$database doesn't have 'excel_file' defined!";
-
-		my $sheet = x($config->{sheet}) || die "no sheet in $type.xml";
-		my $start_row = x($config->{start_row}) - 1 || die "no start_row in $type.xml";
-
-		my $oBook = Spreadsheet::ParseExcel::Workbook->Parse($excel_file) || die "can't open Excel file '$excel_file'";
-
-		my $sheet_nr = 0;
-		foreach my $oWks (@{$oBook->{Worksheet}}) {
-			#print STDERR "-- SHEET $sheet_nr:", $oWks->{Name}, "\n";
-			last if ($oWks->{Name} eq $sheet);
-			$sheet_nr++;
-		}
-
-		my $oWorksheet = $oBook->{Worksheet}[$sheet_nr];
-	
-		print STDERR "using sheet: ",$oWorksheet->{Name},"\n";
-		defined ($oWorksheet) || die "can't find sheet '$sheet' in $excel_file";
-		my $end_row = x($config->{end_row}) || $oWorksheet->{MaxRow};
-
-		for(my $iR = $start_row ; defined $end_row && $iR <= $end_row ; $iR++) {
-			my $row;
-			for(my $iC = $oWorksheet->{MinCol} ; defined $oWorksheet->{MaxCol} && $iC <= $oWorksheet->{MaxCol} ; $iC++) {
-				my $cell = $oWorksheet->{Cells}[$iR][$iC];
-				if ($cell) {
-					$row->{int2col($iC)} = $cell->Value;
-				}
-			}
-
-			progress($iR, $end_row);
-
-#			print "row[$iR/$end_row] ";
-#			foreach (keys %{$row}) {
-#				print "$_: ",$row->{$_},"\t";
-#			}
-#			print "\n";
-
-			my $swishpath = $database."#".$iR;
-
-			next if (! $row);
-
-			if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) {
-				$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$xml\n";
-			}
-		}
-	} elsif ($type_base eq "marc") {
-
-		use MARC;
-		
-		$import2cp = Text::Iconv->new($config->{marc_codepage},$codepage);
-		my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!";
-
-		# optional argument is format
-		my $format = x($config->{format}) || 'usmarc';
-
-		print STDERR "Reading MARC file '$marc_file'\n";
-
-		my $marc = new MARC;
-		my $nr = $marc->openmarc({
-				file=>$marc_file, format=>$format
-			}) || die "Can't open MARC file '$marc_file'";
-
-		my $i=0;	# record nr.
-
-		my $rec;
-
-		while ($marc->nextmarc(1)) {
-
-			# XXX
-			fakeprogress($i++);
-
-			my $swishpath = $database."#".$i;
-
-			if (my $xml = data2xml($type_base,$marc,$add_xml,$cfg,$database)) {
-				$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$xml\n";
-			}
-		}
-	} elsif ($type_base eq "feed") {
-
-		$import2cp = Text::Iconv->new($config->{feed_codepage},$codepage);
-		my $prog = x($config->{prog}) || die "$database doesn't have 'prog' defined!";
+	# save into index
+	foreach my $ds (@ds) {
+		next if (! $ds->{'swish'});
 
-		print STDERR "Reading feed from program '$prog'\n";
-
-		open(FEED,"feeds/$prog |") || die "can't start $prog: $!";
-
-		my $i=1;	# record nr.
-
-		my $data;
-		my $line=1;
-
-		while (<FEED>) {
-			chomp;
-
-			if (/^$/) {
-				my $swishpath = $database."#".$i++;
-
-				if (my $xml = data2xml($type_base,$data,$add_xml,$cfg,$database)) {
-					$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";
-				}
-				$line = 1;
-				$data = {};
-				next;
-			}
+		$index->insert(
+			index_name => $ds->{'tag'},
+			path => $filename,
+			headline => $headline,
+			words => join(" ",@{$ds->{'swish'}})
+		);
+	}
 
-			$line = $1 if (s/^(\d+):\s*//);
-			$data->{$line++} = $_;
+#	print Dumper(\@ds);
 
-			fakeprogress($i);
+}
 
-		}
-	}
+if ($log->is_debug) {
+	$log->debug("lookup hash: ",Dumper($webpac->{'lookup'}));
+	$log->debug("data hash: ",Dumper($webpac->{'data'}));
 }
 
-# call this to commit index
 $index->close;
 
-1;
-__END__
-##########################################################################
-
-=head1 NAME
-
-all2xml.pl - read various file formats and dump XML for SWISH-E
-
-=head1 DESCRIPTION
-
-This command will read ISIS data file using OpenIsis perl module, MARC
-records using MARC module and optionally Micro$oft Excel files to
-create one XML file for usage with I<SWISH-E> indexer. Dispite it's name,
-this script B<isn't general xml generator> from isis files (isis allready
-has something like that). Output of this script is tailor-made for SWISH-E.
-
-=head1 BUGS
-
-Documentation is really lacking. However, in true Open Source spirit, source
-is best documentation. I even made considerable effort to comment parts
-which are not intuitively clear, so...
-
-=head1 AUTHOR
-
-Dobrica Pavlinusic <dpavlin@rot13.org>
-
-=head1 COPYRIGHT
-
-GNU Public License (GPL) v2 or later
-
-=head1 SEE ALSO
-
-SWISH-E web site at http://www.swish-e.org
-
-=cut