Semantic-Engine/EPrints/index.pl

#!/usr/bin/perl -w

use strict;
use Semantic::API;
use Data::Dump qw/dump/;

use EPrints qw/_x/;
use KinoSearch::Simple;

my $debug = shift @ARGV;
my $use = {
        score => 1,
        stem => 1,
        slogovi => 1,
};

my $dbh = EPrints->dbh;
my $sth = $dbh->prepare(qq{
SELECT
        archive_title.eprintid as id,
        title
FROM archive_title 
WHERE
        lang = 'hr'
}) || die $dbh->errstr();
$sth->execute() || die $sth->errstr();

my $indexer = Semantic::API::Index->new(
        storage => 'sqlite',
        database => 'eprints.db',
        collection => 'EPrints'
);


$indexer->add_word_filters( minimum_length => 3,
                            too_many_numbers => 10,
                            maximum_word_length => 15 );

# use this encoding for any incoming text
$indexer->set_default_encoding( "iso-8859-2" ); 

my $total = 0;

my $kino = KinoSearch::Simple->new(
        path => 'kinoindex/',
        language => 'ru',
);

while (my $row = $sth->fetchrow_hashref ) {
        my $id = $row->{id};
        EPrints->id( $id );

        my $parts = {
                title => [ _x( $row->{title} ), 3 ],
                keywords => [ EPrints->lookup( 'keywords' ), 2 ],
                abstract => [ EPrints->lookup( 'abstract' ), 1 ],
#               content => [ EPrints->fulltext_content, 1 ],
        };

        my $skip = 0;
        foreach my $part ( qw/title keywords abstract/ ) {
                if ( ! $parts->{$part}->[0] ) {
                        warn "skipped $id doesn't have required part $part\n";
                        $skip = 1;
                        last;
                }
        }
        next if $skip;

        my $body = '';

        foreach my $part ( qw/title keywords abstract content/ ) {
                my $content = $parts->{$part}->[0] || next;

                if ( $use->{slogovi} ) {
                        $body .= ' ' . EPrints->slogovi( $content );
                }

                if ( $use->{stem} ) {
                        my $stem = EPrints->stem( $content );
                        warn "stem of '$content' didn't return anything\n" unless $stem;
                        $content = $stem;
                }


                if ( $use->{score} ) {
                        map { $body .= "$content " } 1 .. $parts->{$part}->[1];
                } else {
                        $body .= "$content ";
                }

                warn ">>> $body <<<\n" if $debug;
        }

        $indexer->index( $row->{id}, $body );
        $kino->add_doc({
                id => $id,
                title => $parts->{title}->[0],
                keywords => $parts->{keywords}->[0],
                abstract => $parts->{abstract}->[0],
        });
        
        $total++;
        print STDERR "$total: ", $row->{id}, " ", _x( $row->{title} ), " - ", length($body), " bytes\n";
}


print STDERR "\nNow adding $total items to the database...";
$indexer->finish(); 
print STDERR "done!\n"; 

1	dpavlin	1	#!/usr/bin/perl -w
2
3			use strict;
4			use Semantic::API;
5			use Data::Dump qw/dump/;
6	dpavlin	13
7	dpavlin	1	use EPrints qw/_x/;
8	dpavlin	18	use KinoSearch::Simple;
9	dpavlin	13
10	dpavlin	1	my $debug = shift @ARGV;
11	dpavlin	17	my $use = {
12			score => 1,
13			stem => 1,
14			slogovi => 1,
15			};
16	dpavlin	1
17			my $dbh = EPrints->dbh;
18			my $sth = $dbh->prepare(qq{
19			SELECT
20			archive_title.eprintid as id,
21			title
22			FROM archive_title
23	dpavlin	17	WHERE
24			lang = 'hr'
25	dpavlin	1	}) \|\| die $dbh->errstr();
26			$sth->execute() \|\| die $sth->errstr();
27
28			my $indexer = Semantic::API::Index->new(
29			storage => 'sqlite',
30			database => 'eprints.db',
31			collection => 'EPrints'
32			);
33
34
35			$indexer->add_word_filters( minimum_length => 3,
36			too_many_numbers => 10,
37			maximum_word_length => 15 );
38
39			# use this encoding for any incoming text
40	dpavlin	8	$indexer->set_default_encoding( "iso-8859-2" );
41	dpavlin	1
42			my $total = 0;
43
44	dpavlin	18	my $kino = KinoSearch::Simple->new(
45			path => 'kinoindex/',
46			language => 'ru',
47			);
48
49	dpavlin	1	while (my $row = $sth->fetchrow_hashref ) {
50	dpavlin	17	my $id = $row->{id};
51			EPrints->id( $id );
52	dpavlin	18
53	dpavlin	14	my $parts = {
54	dpavlin	17	title => [ _x( $row->{title} ), 3 ],
55			keywords => [ EPrints->lookup( 'keywords' ), 2 ],
56			abstract => [ EPrints->lookup( 'abstract' ), 1 ],
57	dpavlin	14	# content => [ EPrints->fulltext_content, 1 ],
58			};
59
60	dpavlin	17	my $skip = 0;
61			foreach my $part ( qw/title keywords abstract/ ) {
62			if ( ! $parts->{$part}->[0] ) {
63			warn "skipped $id doesn't have required part $part\n";
64			$skip = 1;
65			last;
66			}
67			}
68			next if $skip;
69
70	dpavlin	8	my $body = '';
71	dpavlin	1
72	dpavlin	14	foreach my $part ( qw/title keywords abstract content/ ) {
73	dpavlin	17	my $content = $parts->{$part}->[0] \|\| next;
74	dpavlin	8
75	dpavlin	17	if ( $use->{slogovi} ) {
76			$body .= ' ' . EPrints->slogovi( $content );
77			}
78
79			if ( $use->{stem} ) {
80			my $stem = EPrints->stem( $content );
81			warn "stem of '$content' didn't return anything\n" unless $stem;
82			$content = $stem;
83			}
84
85
86			if ( $use->{score} ) {
87			map { $body .= "$content " } 1 .. $parts->{$part}->[1];
88	dpavlin	14	} else {
89	dpavlin	17	$body .= "$content ";
90	dpavlin	14	}
91	dpavlin	17
92			warn ">>> $body <<<\n" if $debug;
93	dpavlin	14	}
94	dpavlin	1
95	dpavlin	14	$indexer->index( $row->{id}, $body );
96	dpavlin	18	$kino->add_doc({
97			id => $id,
98	dpavlin	19	title => $parts->{title}->[0],
99			keywords => $parts->{keywords}->[0],
100			abstract => $parts->{abstract}->[0],
101	dpavlin	18	});
102
103	dpavlin	1	$total++;
104	dpavlin	14	print STDERR "$total: ", $row->{id}, " ", _x( $row->{title} ), " - ", length($body), " bytes\n";
105	dpavlin	1	}
106
107
108			print STDERR "\nNow adding $total items to the database...";
109			$indexer->finish();
110			print STDERR "done!\n";
111