4 |
use strict; |
use strict; |
5 |
use warnings; |
use warnings; |
6 |
|
|
7 |
our $VERSION = '0.04_1'; |
our $VERSION = '0.05'; |
8 |
|
|
9 |
=head1 NAME |
=head1 NAME |
10 |
|
|
17 |
use Search::Estraier; |
use Search::Estraier; |
18 |
|
|
19 |
# create and configure node |
# create and configure node |
20 |
my $node = new Search::Estraier::Node; |
my $node = new Search::Estraier::Node( |
21 |
$node->set_url("http://localhost:1978/node/test"); |
url => 'http://localhost:1978/node/test', |
22 |
$node->set_auth("admin","admin"); |
user => 'admin', |
23 |
|
passwd => 'admin' |
24 |
|
); |
25 |
|
|
26 |
# create document |
# create document |
27 |
my $doc = new Search::Estraier::Document; |
my $doc = new Search::Estraier::Document; |
34 |
$doc->add_text("Somewhere over the rainbow. Way up high."); |
$doc->add_text("Somewhere over the rainbow. Way up high."); |
35 |
$doc->add_text("There's a land that I heard of once in a lullaby."); |
$doc->add_text("There's a land that I heard of once in a lullaby."); |
36 |
|
|
37 |
die "error: ", $node->status,"\n" unless ($node->put_doc($doc)); |
die "error: ", $node->status,"\n" unless (eval { $node->put_doc($doc) }); |
38 |
|
|
39 |
=head2 Simple searcher |
=head2 Simple searcher |
40 |
|
|
41 |
use Search::Estraier; |
use Search::Estraier; |
42 |
|
|
43 |
# create and configure node |
# create and configure node |
44 |
my $node = new Search::Estraier::Node; |
my $node = new Search::Estraier::Node( |
45 |
$node->set_url("http://localhost:1978/node/test"); |
url => 'http://localhost:1978/node/test', |
46 |
$node->set_auth("admin","admin"); |
user => 'admin', |
47 |
|
passwd => 'admin', |
48 |
|
croak_on_error => 1, |
49 |
|
); |
50 |
|
|
51 |
# create condition |
# create condition |
52 |
my $cond = new Search::Estraier::Condition; |
my $cond = new Search::Estraier::Condition; |
55 |
$cond->set_phrase("rainbow AND lullaby"); |
$cond->set_phrase("rainbow AND lullaby"); |
56 |
|
|
57 |
my $nres = $node->search($cond, 0); |
my $nres = $node->search($cond, 0); |
58 |
|
|
59 |
if (defined($nres)) { |
if (defined($nres)) { |
60 |
|
print "Got ", $nres->hits, " results\n"; |
61 |
|
|
62 |
# for each document in results |
# for each document in results |
63 |
for my $i ( 0 ... $nres->doc_num - 1 ) { |
for my $i ( 0 ... $nres->doc_num - 1 ) { |
64 |
# get result document |
# get result document |
100 |
=cut |
=cut |
101 |
|
|
102 |
sub _s { |
sub _s { |
103 |
my $text = $_[1] || return; |
my $text = $_[1]; |
104 |
|
return unless defined($text); |
105 |
$text =~ s/\s\s+/ /gs; |
$text =~ s/\s\s+/ /gs; |
106 |
$text =~ s/^\s+//; |
$text =~ s/^\s+//; |
107 |
$text =~ s/\s+$//; |
$text =~ s/\s+$//; |
166 |
} elsif ($line =~ m/^$/) { |
} elsif ($line =~ m/^$/) { |
167 |
$in_text = 1; |
$in_text = 1; |
168 |
next; |
next; |
169 |
} elsif ($line =~ m/^(.+)=(.+)$/) { |
} elsif ($line =~ m/^(.+)=(.*)$/) { |
170 |
$self->{attrs}->{ $1 } = $2; |
$self->{attrs}->{ $1 } = $2; |
171 |
next; |
next; |
172 |
} |
} |
173 |
|
|
174 |
warn "draft ignored: $line\n"; |
warn "draft ignored: '$line'\n"; |
175 |
} |
} |
176 |
} |
} |
177 |
|
|
329 |
my $draft; |
my $draft; |
330 |
|
|
331 |
foreach my $attr_name (sort keys %{ $self->{attrs} }) { |
foreach my $attr_name (sort keys %{ $self->{attrs} }) { |
332 |
$draft .= $attr_name . '=' . $self->{attrs}->{$attr_name} . "\n"; |
next unless defined(my $v = $self->{attrs}->{$attr_name}); |
333 |
|
$draft .= $attr_name . '=' . $v . "\n"; |
334 |
} |
} |
335 |
|
|
336 |
if ($self->{kwords}) { |
if ($self->{kwords}) { |
378 |
|
|
379 |
package Search::Estraier::Condition; |
package Search::Estraier::Condition; |
380 |
|
|
381 |
use Carp qw/confess croak/; |
use Carp qw/carp confess croak/; |
382 |
|
|
383 |
use Search::Estraier; |
use Search::Estraier; |
384 |
our @ISA = qw/Search::Estraier/; |
our @ISA = qw/Search::Estraier/; |
456 |
|
|
457 |
=head2 set_options |
=head2 set_options |
458 |
|
|
459 |
$cond->set_options( SURE => 1 ); |
$cond->set_options( 'SURE' ); |
460 |
|
|
461 |
|
$cond->set_options( qw/AGITO NOIDF SIMPLE/ ); |
462 |
|
|
463 |
|
Possible options are: |
464 |
|
|
465 |
|
=over 8 |
466 |
|
|
467 |
|
=item SURE |
468 |
|
|
469 |
|
check every N-gram |
470 |
|
|
471 |
|
=item USUAL |
472 |
|
|
473 |
|
check every second N-gram |
474 |
|
|
475 |
|
=item FAST |
476 |
|
|
477 |
|
check every third N-gram |
478 |
|
|
479 |
|
=item AGITO |
480 |
|
|
481 |
|
check every fourth N-gram |
482 |
|
|
483 |
|
=item NOIDF |
484 |
|
|
485 |
|
don't perform TF-IDF tuning |
486 |
|
|
487 |
|
=item SIMPLE |
488 |
|
|
489 |
|
use simplified query phrase |
490 |
|
|
491 |
|
=back |
492 |
|
|
493 |
|
Skipping N-grams will speed up search, but reduce accuracy. Every call to C<set_options> will reset previous |
494 |
|
options; |
495 |
|
|
496 |
|
This option changed in version C<0.04> of this module. It's backwards compatibile. |
497 |
|
|
498 |
=cut |
=cut |
499 |
|
|
500 |
my $options = { |
my $options = { |
|
# check N-gram keys skipping by three |
|
501 |
SURE => 1 << 0, |
SURE => 1 << 0, |
|
# check N-gram keys skipping by two |
|
502 |
USUAL => 1 << 1, |
USUAL => 1 << 1, |
|
# without TF-IDF tuning |
|
503 |
FAST => 1 << 2, |
FAST => 1 << 2, |
|
# with the simplified phrase |
|
504 |
AGITO => 1 << 3, |
AGITO => 1 << 3, |
|
# check every N-gram key |
|
505 |
NOIDF => 1 << 4, |
NOIDF => 1 << 4, |
|
# check N-gram keys skipping by one |
|
506 |
SIMPLE => 1 << 10, |
SIMPLE => 1 << 10, |
507 |
}; |
}; |
508 |
|
|
509 |
sub set_options { |
sub set_options { |
510 |
my $self = shift; |
my $self = shift; |
511 |
my $option = shift; |
my $opt = 0; |
512 |
confess "unknown option" unless ($options->{$option}); |
foreach my $option (@_) { |
513 |
$self->{options} ||= $options->{$option}; |
my $mask; |
514 |
|
unless ($mask = $options->{$option}) { |
515 |
|
if ($option eq '1') { |
516 |
|
next; |
517 |
|
} else { |
518 |
|
croak "unknown option $option"; |
519 |
|
} |
520 |
|
} |
521 |
|
$opt += $mask; |
522 |
|
} |
523 |
|
$self->{options} = $opt; |
524 |
} |
} |
525 |
|
|
526 |
|
|
599 |
} |
} |
600 |
|
|
601 |
|
|
602 |
|
=head2 set_skip |
603 |
|
|
604 |
|
Set number of skipped documents from beginning of results |
605 |
|
|
606 |
|
$cond->set_skip(42); |
607 |
|
|
608 |
|
Similar to C<offset> in RDBMS. |
609 |
|
|
610 |
|
=cut |
611 |
|
|
612 |
|
sub set_skip { |
613 |
|
my $self = shift; |
614 |
|
$self->{skip} = shift; |
615 |
|
} |
616 |
|
|
617 |
|
=head2 skip |
618 |
|
|
619 |
|
Return skip for this condition. |
620 |
|
|
621 |
|
print $cond->skip; |
622 |
|
|
623 |
|
=cut |
624 |
|
|
625 |
|
sub skip { |
626 |
|
my $self = shift; |
627 |
|
return $self->{skip}; |
628 |
|
} |
629 |
|
|
630 |
|
|
631 |
package Search::Estraier::ResultDocument; |
package Search::Estraier::ResultDocument; |
632 |
|
|
633 |
use Carp qw/croak/; |
use Carp qw/croak/; |
771 |
|
|
772 |
print $res->doc_num; |
print $res->doc_num; |
773 |
|
|
774 |
|
This will return real number of documents (limited by C<max>). |
775 |
|
If you want to get total number of hits, see C<hits>. |
776 |
|
|
777 |
=cut |
=cut |
778 |
|
|
779 |
sub doc_num { |
sub doc_num { |
805 |
|
|
806 |
Return specific hint from results. |
Return specific hint from results. |
807 |
|
|
808 |
print $rec->hint( 'VERSION' ); |
print $res->hint( 'VERSION' ); |
809 |
|
|
810 |
Possible hints are: C<VERSION>, C<NODE>, C<HIT>, C<HINT#n>, C<DOCNUM>, C<WORDNUM>, |
Possible hints are: C<VERSION>, C<NODE>, C<HIT>, C<HINT#n>, C<DOCNUM>, C<WORDNUM>, |
811 |
C<TIME>, C<LINK#n>, C<VIEW>. |
C<TIME>, C<LINK#n>, C<VIEW>. |
818 |
return $self->{hints}->{$key}; |
return $self->{hints}->{$key}; |
819 |
} |
} |
820 |
|
|
821 |
|
=head2 hints |
822 |
|
|
823 |
|
More perlish version of C<hint>. This one returns hash. |
824 |
|
|
825 |
|
my %hints = $res->hints; |
826 |
|
|
827 |
|
=cut |
828 |
|
|
829 |
|
sub hints { |
830 |
|
my $self = shift; |
831 |
|
return $self->{hints}; |
832 |
|
} |
833 |
|
|
834 |
|
=head2 hits |
835 |
|
|
836 |
|
Syntaxtic sugar for total number of hits for this query |
837 |
|
|
838 |
|
print $res->hits; |
839 |
|
|
840 |
|
It's same as |
841 |
|
|
842 |
|
print $res->hint('HIT'); |
843 |
|
|
844 |
|
but shorter. |
845 |
|
|
846 |
|
=cut |
847 |
|
|
848 |
|
sub hits { |
849 |
|
my $self = shift; |
850 |
|
return $self->{hints}->{'HIT'} || 0; |
851 |
|
} |
852 |
|
|
853 |
package Search::Estraier::Node; |
package Search::Estraier::Node; |
854 |
|
|
868 |
|
|
869 |
my $node = new Search::HyperEstraier::Node( 'http://localhost:1978/node/test' ); |
my $node = new Search::HyperEstraier::Node( 'http://localhost:1978/node/test' ); |
870 |
|
|
871 |
|
or in more verbose form |
872 |
|
|
873 |
|
my $node = new Search::HyperEstraier::Node( |
874 |
|
url => 'http://localhost:1978/node/test', |
875 |
|
debug => 1, |
876 |
|
croak_on_error => 1 |
877 |
|
); |
878 |
|
|
879 |
|
with following arguments: |
880 |
|
|
881 |
|
=over 4 |
882 |
|
|
883 |
|
=item url |
884 |
|
|
885 |
|
URL to node |
886 |
|
|
887 |
|
=item debug |
888 |
|
|
889 |
|
dumps a B<lot> of debugging output |
890 |
|
|
891 |
|
=item croak_on_error |
892 |
|
|
893 |
|
very helpful during development. It will croak on all errors instead of |
894 |
|
silently returning C<-1> (which is convention of Hyper Estraier API in other |
895 |
|
languages). |
896 |
|
|
897 |
|
=back |
898 |
|
|
899 |
=cut |
=cut |
900 |
|
|
901 |
sub new { |
sub new { |
903 |
my $self = { |
my $self = { |
904 |
pxport => -1, |
pxport => -1, |
905 |
timeout => 0, # this used to be -1 |
timeout => 0, # this used to be -1 |
|
dnum => -1, |
|
|
wnum => -1, |
|
|
size => -1.0, |
|
906 |
wwidth => 480, |
wwidth => 480, |
907 |
hwidth => 96, |
hwidth => 96, |
908 |
awidth => 96, |
awidth => 96, |
909 |
status => -1, |
status => -1, |
910 |
}; |
}; |
911 |
|
|
912 |
bless($self, $class); |
bless($self, $class); |
913 |
|
|
914 |
if ($#_ == 0) { |
if ($#_ == 0) { |
916 |
} else { |
} else { |
917 |
my $args = {@_}; |
my $args = {@_}; |
918 |
|
|
919 |
$self->{debug} = $args->{debug}; |
%$self = ( %$self, @_ ); |
920 |
|
|
921 |
warn "## Node debug on\n" if ($self->{debug}); |
warn "## Node debug on\n" if ($self->{debug}); |
922 |
} |
} |
923 |
|
|
924 |
|
$self->{inform} = { |
925 |
|
dnum => -1, |
926 |
|
wnum => -1, |
927 |
|
size => -1.0, |
928 |
|
}; |
929 |
|
|
930 |
$self ? return $self : return undef; |
$self ? return $self : return undef; |
931 |
} |
} |
932 |
|
|
1205 |
|
|
1206 |
my $id = $node->uri_to_id( 'file:///document/uri/42' ); |
my $id = $node->uri_to_id( 'file:///document/uri/42' ); |
1207 |
|
|
1208 |
|
This method won't croak, even if using C<croak_on_error>. |
1209 |
|
|
1210 |
=cut |
=cut |
1211 |
|
|
1212 |
sub uri_to_id { |
sub uri_to_id { |
1213 |
my $self = shift; |
my $self = shift; |
1214 |
my $uri = shift || return; |
my $uri = shift || return; |
1215 |
return $self->_fetch_doc( uri => $uri, path => '/uri_to_id', chomp_resbody => 1 ); |
return $self->_fetch_doc( uri => $uri, path => '/uri_to_id', chomp_resbody => 1, croak_on_error => 0 ); |
1216 |
} |
} |
1217 |
|
|
1218 |
|
|
1272 |
'application/x-www-form-urlencoded', |
'application/x-www-form-urlencoded', |
1273 |
$arg, |
$arg, |
1274 |
\$resbody, |
\$resbody, |
1275 |
|
$a->{croak_on_error}, |
1276 |
); |
); |
1277 |
|
|
1278 |
return if ($rv != 200); |
return if ($rv != 200); |
1303 |
|
|
1304 |
sub name { |
sub name { |
1305 |
my $self = shift; |
my $self = shift; |
1306 |
$self->_set_info unless ($self->{name}); |
$self->_set_info unless ($self->{inform}->{name}); |
1307 |
return $self->{name}; |
return $self->{inform}->{name}; |
1308 |
} |
} |
1309 |
|
|
1310 |
|
|
1316 |
|
|
1317 |
sub label { |
sub label { |
1318 |
my $self = shift; |
my $self = shift; |
1319 |
$self->_set_info unless ($self->{label}); |
$self->_set_info unless ($self->{inform}->{label}); |
1320 |
return $self->{label}; |
return $self->{inform}->{label}; |
1321 |
} |
} |
1322 |
|
|
1323 |
|
|
1329 |
|
|
1330 |
sub doc_num { |
sub doc_num { |
1331 |
my $self = shift; |
my $self = shift; |
1332 |
$self->_set_info if ($self->{dnum} < 0); |
$self->_set_info if ($self->{inform}->{dnum} < 0); |
1333 |
return $self->{dnum}; |
return $self->{inform}->{dnum}; |
1334 |
} |
} |
1335 |
|
|
1336 |
|
|
1342 |
|
|
1343 |
sub word_num { |
sub word_num { |
1344 |
my $self = shift; |
my $self = shift; |
1345 |
$self->_set_info if ($self->{wnum} < 0); |
$self->_set_info if ($self->{inform}->{wnum} < 0); |
1346 |
return $self->{wnum}; |
return $self->{inform}->{wnum}; |
1347 |
} |
} |
1348 |
|
|
1349 |
|
|
1355 |
|
|
1356 |
sub size { |
sub size { |
1357 |
my $self = shift; |
my $self = shift; |
1358 |
$self->_set_info if ($self->{size} < 0); |
$self->_set_info if ($self->{inform}->{size} < 0); |
1359 |
return $self->{size}; |
return $self->{inform}->{size}; |
1360 |
} |
} |
1361 |
|
|
1362 |
|
|
1519 |
push @args, 'wwidth=' . $self->{wwidth}; |
push @args, 'wwidth=' . $self->{wwidth}; |
1520 |
push @args, 'hwidth=' . $self->{hwidth}; |
push @args, 'hwidth=' . $self->{hwidth}; |
1521 |
push @args, 'awidth=' . $self->{awidth}; |
push @args, 'awidth=' . $self->{awidth}; |
1522 |
|
push @args, 'skip=' . $self->{skip} if ($self->{skip}); |
1523 |
|
|
1524 |
return join('&', @args); |
return join('&', @args); |
1525 |
} |
} |
1542 |
sub shuttle_url { |
sub shuttle_url { |
1543 |
my $self = shift; |
my $self = shift; |
1544 |
|
|
1545 |
my ($url, $content_type, $reqbody, $resbody) = @_; |
my ($url, $content_type, $reqbody, $resbody, $croak_on_error) = @_; |
1546 |
|
|
1547 |
|
$croak_on_error = $self->{croak_on_error} unless defined($croak_on_error); |
1548 |
|
|
1549 |
$self->{status} = -1; |
$self->{status} = -1; |
1550 |
|
|
1571 |
|
|
1572 |
$req->headers->header( 'Host' => $url->host . ":" . $url->port ); |
$req->headers->header( 'Host' => $url->host . ":" . $url->port ); |
1573 |
$req->headers->header( 'Connection', 'close' ); |
$req->headers->header( 'Connection', 'close' ); |
1574 |
$req->headers->header( 'Authorization', 'Basic ' . $self->{auth} ); |
$req->headers->header( 'Authorization', 'Basic ' . $self->{auth} ) if ($self->{auth}); |
1575 |
$req->content_type( $content_type ); |
$req->content_type( $content_type ); |
1576 |
|
|
1577 |
warn $req->headers->as_string,"\n" if ($self->{debug}); |
warn $req->headers->as_string,"\n" if ($self->{debug}); |
1587 |
|
|
1588 |
($self->{status}, $self->{status_message}) = split(/\s+/, $res->status_line, 2); |
($self->{status}, $self->{status_message}) = split(/\s+/, $res->status_line, 2); |
1589 |
|
|
1590 |
return -1 if (! $res->is_success); |
if (! $res->is_success) { |
1591 |
|
if ($croak_on_error) { |
1592 |
|
croak("can't get $url: ",$res->status_line); |
1593 |
|
} else { |
1594 |
|
return -1; |
1595 |
|
} |
1596 |
|
} |
1597 |
|
|
1598 |
$$resbody .= $res->content; |
$$resbody .= $res->content; |
1599 |
|
|
1693 |
my $reqbody = 'url=' . uri_escape($url) . '&label=' . uri_escape($label); |
my $reqbody = 'url=' . uri_escape($url) . '&label=' . uri_escape($label); |
1694 |
$reqbody .= '&credit=' . $credit if ($credit > 0); |
$reqbody .= '&credit=' . $credit if ($credit > 0); |
1695 |
|
|
1696 |
$self->shuttle_url( $self->{url} . '/_set_link', |
if ($self->shuttle_url( $self->{url} . '/_set_link', |
1697 |
'application/x-www-form-urlencoded', |
'application/x-www-form-urlencoded', |
1698 |
$reqbody, |
$reqbody, |
1699 |
undef |
undef |
1700 |
) == 200; |
) == 200) { |
1701 |
|
# refresh node info after adding link |
1702 |
|
$self->_set_info; |
1703 |
|
return 1; |
1704 |
|
} |
1705 |
|
} |
1706 |
|
|
1707 |
|
=head2 admins |
1708 |
|
|
1709 |
|
my @admins = @{ $node->admins }; |
1710 |
|
|
1711 |
|
Return array of users with admin rights on node |
1712 |
|
|
1713 |
|
=cut |
1714 |
|
|
1715 |
|
sub admins { |
1716 |
|
my $self = shift; |
1717 |
|
$self->_set_info unless ($self->{inform}->{name}); |
1718 |
|
return $self->{inform}->{admins}; |
1719 |
|
} |
1720 |
|
|
1721 |
|
=head2 guests |
1722 |
|
|
1723 |
|
my @guests = @{ $node->guests }; |
1724 |
|
|
1725 |
|
Return array of users with guest rights on node |
1726 |
|
|
1727 |
|
=cut |
1728 |
|
|
1729 |
|
sub guests { |
1730 |
|
my $self = shift; |
1731 |
|
$self->_set_info unless ($self->{inform}->{name}); |
1732 |
|
return $self->{inform}->{guests}; |
1733 |
|
} |
1734 |
|
|
1735 |
|
=head2 links |
1736 |
|
|
1737 |
|
my $links = @{ $node->links }; |
1738 |
|
|
1739 |
|
Return array of links for this node |
1740 |
|
|
1741 |
|
=cut |
1742 |
|
|
1743 |
|
sub links { |
1744 |
|
my $self = shift; |
1745 |
|
$self->_set_info unless ($self->{inform}->{name}); |
1746 |
|
return $self->{inform}->{links}; |
1747 |
} |
} |
1748 |
|
|
1749 |
|
|
1774 |
|
|
1775 |
return if ($rv != 200 || !$resbody); |
return if ($rv != 200 || !$resbody); |
1776 |
|
|
1777 |
# it seems that response can have multiple line endings |
my @lines = split(/[\r\n]/,$resbody); |
1778 |
$resbody =~ s/[\r\n]+$//; |
|
1779 |
|
$self->{inform} = {}; |
1780 |
|
|
1781 |
|
( $self->{inform}->{name}, $self->{inform}->{label}, $self->{inform}->{dnum}, |
1782 |
|
$self->{inform}->{wnum}, $self->{inform}->{size} ) = split(/\t/, shift @lines, 5); |
1783 |
|
|
1784 |
|
return $resbody unless (@lines); |
1785 |
|
|
1786 |
|
shift @lines; |
1787 |
|
|
1788 |
|
while(my $admin = shift @lines) { |
1789 |
|
push @{$self->{inform}->{admins}}, $admin; |
1790 |
|
} |
1791 |
|
|
1792 |
|
while(my $guest = shift @lines) { |
1793 |
|
push @{$self->{inform}->{guests}}, $guest; |
1794 |
|
} |
1795 |
|
|
1796 |
|
while(my $link = shift @lines) { |
1797 |
|
push @{$self->{inform}->{links}}, $link; |
1798 |
|
} |
1799 |
|
|
1800 |
( $self->{name}, $self->{label}, $self->{dnum}, $self->{wnum}, $self->{size} ) = |
return $resbody; |
|
split(/\t/, $resbody, 5); |
|
1801 |
|
|
1802 |
} |
} |
1803 |
|
|