use warnings;
use strict;
+use base qw/WebPAC::Common/;
+
+use HyperEstraier;
+use Text::Iconv;
+use Data::Dumper;
+
=head1 NAME
-WebPAC::Output::Estraier - The great new WebPAC::Output::Estraier!
+WebPAC::Output::Estraier - Create Hyper Estraier full text index
=head1 VERSION
=head1 SYNOPSIS
-Quick summary of what the module does.
+Create full text index using Hyper Estraier index from data with
+type C<search>.
+
+=head1 FUNCTIONS
-Perhaps a little code snippet.
+=head2 new
- use WebPAC::Output::Estraier;
+Connect to Hyper Estraier index using HTTP
- my $foo = WebPAC::Output::Estraier->new();
- ...
+ my $est = new WebPAC::Output::Estraier(
+ url => 'http://localhost:1978/node/webpac2',
+ user => 'admin',
+ passwd => 'admin',
+ database => 'demo',
+ encoding => 'iso-8859-2',
+ );
-=head1 EXPORT
+Options are:
-A list of functions that can be exported. You can delete this section
-if you don't export anything, such as for a purely object-oriented module.
+=over 4
-=head1 FUNCTIONS
+=item url
+
+URI to C<estmaster> node
+
+=item user
+
+C<estmaster> user with sufficient rights
+
+=item passwd
+
+password for user
+
+=item database
+
+name of database from which data comes
-=head2 function1
+=item encoding
+
+character encoding of C<data_structure> if it's differenet than C<ISO-8859-2>
+(and it probably is). This encoding will be converted to C<UTF-8> for
+Hyper Estraier.
+
+=back
+
+Name of database will be used to form URI of documents in index.
=cut
-sub function1 {
+sub new {
+ my $class = shift;
+ my $self = {@_};
+ bless($self, $class);
+
+ my $log = $self->_get_logger;
+
+ foreach my $p (qw/url user passwd/) {
+ $log->logdie("need $p") unless ($self->{$p});
+ }
+
+ $log->info("opening Hyper Estraier index $self->{'url'}");
+
+ $self->{'db'} = HyperEstraier::Node->new($self->{'url'});
+ $self->{'db'}->set_auth($self->{'user'}, $self->{'passwd'});
+
+ my $encoding = $self->{'encoding'} || 'ISO-8859-2';
+ $log->info("using encoding $encoding");
+
+ my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
+
+ $self ? return $self : return undef;
}
-=head2 function2
+=head2 add
+
+Adds one entry to database.
+
+ $est->add(
+ id => 42,
+ ds => $ds,
+ type => 'display',
+ url_prefix => 'database name',
+ text => 'optional text from which snippet is created',
+ );
+
+This function will create entries in index using following URI format:
+
+ C<file:///database%20name/000>
+
+Each tag in C<data_structure> with specified C<type> will create one
+attribute and corresponding hidden text (used for search).
=cut
-sub function2 {
-}
+sub add {
+ my $self = shift;
-=head1 AUTHOR
+ my $args = {@_};
-Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >>
+ my $log = $self->_get_logger;
+
+ my $database = $self->{'database'} || $log->logconfess('no database in $self');
+ $log->logconfess('need db in object') unless ($self->{'db'});
+
+ foreach my $p (qw/id ds type/) {
+ $log->logdie("need $p") unless ($args->{$p});
+ }
+
+ my $type = $args->{'type'};
+ my $mfn = $args->{'id'};
+
+ my $uri = "file:///$type/$database/$mfn";
+ $log->debug("creating $uri");
+
+ my $doc = HyperEstraier::Document->new;
+ $doc->add_attr('@uri', $uri);
+
+ $log->debug("ds = ", sub { Dumper($args->{'ds'}) } );
-=head1 BUGS
+ # filter all tags which have type defined
+ my @tags = grep {
+ defined( $args->{'ds'}->{$_}->{$type} )
+ } keys %{ $args->{'ds'} };
-Please report any bugs or feature requests to
-C<bug-webpac-output-estraier@rt.cpan.org>, or through the web interface at
-L<http://rt.cpan.org/NoAuth/ReportBug.html?Queue=WebPAC>.
-I will be notified, and then you'll automatically be notified of progress on
-your bug as I make changes.
+ $log->debug("tags = ", join(",", @tags));
-=head1 ACKNOWLEDGEMENTS
+ return unless (@tags);
+
+ foreach my $tag (@tags) {
+
+ my $vals = join(" ", @{ $args->{'ds'}->{$tag}->{$type} });
+
+ $log->logconfess("no values for $tag/$type") unless ($vals);
+
+ $doc->add_attr($tag, $vals);
+ $doc->add_hidden_text($vals);
+ }
+
+ my $text = $args->{'text'};
+ $doc->add_text( $text ) if ( $text );
+
+ $log->debug("adding ", sub { $doc->dump_draft } );
+ $self->{'db'}->put_doc($doc) || $log->die("can't add document $uri to index");
+
+ return 1;
+}
+
+=head1 AUTHOR
+
+Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >>
=head1 COPYRIGHT & LICENSE
--- /dev/null
+#!/usr/bin/perl -w
+
+use strict;
+
+use Cwd qw/abs_path/;
+use File::Temp qw/tempdir/;
+use Data::Dumper;
+use lib './lib';
+
+use WebPAC::Lookup;
+use WebPAC::Input::ISIS;
+use WebPAC::DB;
+use WebPAC::Normalize::XML;
+use WebPAC::Output::TT;
+use WebPAC::Output::Estraier;
+
+my $abs_path = abs_path($0);
+$abs_path =~ s#/[^/]*$#/#;
+
+my $isis_file = '/data/isis_data/ps/LIBRI/LIBRI';
+
+my $lookup = new WebPAC::Lookup(
+ lookup_file => "$abs_path/conf/lookup/isis.pm",
+);
+
+my $isis = new WebPAC::Input::ISIS(
+ code_page => 'ISO-8859-2', # application encoding
+ limit_mfn => 50,
+);
+
+my $maxmfn = $isis->open(
+ filename => $isis_file,
+ code_page => '852', # database encoding
+);
+
+my $path = './db/';
+
+my $db = new WebPAC::DB(
+ path => $path,
+);
+
+my $n = new WebPAC::Normalize::XML(
+# filter => { 'foo' => sub { shift } },
+ db => $db,
+ lookup_regex => $lookup->regex,
+ lookup => $lookup,
+);
+
+$n->open(
+ tag => 'isis',
+ xml_file => "$abs_path/conf/normalize/isis_ffzg.xml",
+);
+
+my $out = new WebPAC::Output::TT(
+ include_path => "$abs_path/conf/output/tt",
+ filters => { foo => sub { shift } },
+);
+
+my $est = new WebPAC::Output::Estraier(
+ url => 'http://localhost:1978/node/webpac2',
+ user => 'admin',
+ passwd => 'admin',
+ database => 'ps',
+);
+
+while (my $row = $isis->fetch) {
+
+ my $mfn = $row->{'000'}->[0] || die "can't find MFN";
+
+ my $ds = $n->data_structure($row);
+
+# print STDERR Dumper($row, $ds);
+
+ my $html = $out->apply(
+ template => 'html_ffzg.tt',
+ data => $ds,
+ );
+
+ # create test output
+
+ my $file = sprintf('out/%02d.html', $mfn );
+ open(my $fh, '>', $file) or die "can't open $file: $!";
+ print $fh $html;
+ close($fh);
+
+ $html =~ s#\s*[\n\r]+\s*##gs;
+
+# print STDERR $html;
+
+ $est->add(
+ id => $mfn,
+ ds => $ds,
+ type => 'search',
+ );
+
+};
projects / webpac2 / commitdiff