/[Search-Estraier]/trunk/lib/Search/Estraier.pm

This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!

Diff of /trunk/lib/Search/Estraier.pm

Parent Directory | Revision Log | View Patch Patch

-revision 142 by dpavlin,
Wed May 10 14:57:50 2006 UTC
+revision 191 by dpavlin,
Sun Nov  5 16:26:57 2006 UTC
 Line 4 
 use 5.008;
  use strict;
  use warnings;
- our $VERSION = '0.06';
+ our $VERSION = '0.08';
  =head1 NAME
 Line 120 
 our @ISA = qw/Search::Estraier/;
  =head1 Search::Estraier::Document
- This class implements Document which is collection of attributes
+ This class implements Document which is single item in Hyper Estraier.
- (key=value), vectors (also key value) display text and hidden text.
+ It's is collection of:
+ =over 4
+ =item attributes
+ C<< 'key' => 'value' >> pairs which can later be used for filtering of results
+ You can add common filters to C<attrindex> in estmaster's C<_conf>
+ file for better performance. See C<attrindex> in
+ L<Hyper Estraier P2P Guide|http://hyperestraier.sourceforge.net/nguide-en.html>.
+ =item vectors
+ also C<< 'key' => 'value' >> pairs
+ =item display text
+ Text which will be used to create searchable corpus of your index and
+ included in snippet output.
+ =item hidden text
+ Text which will be searchable, but will not be included in snippet.
+ =back
  =head2 new
-Line 157 
 sub new {
+Line 182 
 sub new {
                          if ($line =~ m/^%VECTOR\t(.+)$/) {
                                  my @fields = split(/\t/, $1);
-                                 for my $i ( 0 .. ($#fields - 1) ) {
+                                 if ($#fields % 2 == 1) {
-                                         $self->{kwords}->{ $fields[ $i ] } = $fields[ $i + 1 ];
+                                         $self->{kwords} = { @fields };
-                                         $i++;
+                                 } else {
+                                         warn "can't decode $line\n";
                                  }
                                  next;
+                         } elsif ($line =~ m/^%SCORE\t(.+)$/) {
+                             $self->{score} = $1;
+                             next;
                          } elsif ($line =~ m/^%/) {
                                  # What is this? comment?
                                  #warn "$line\n";
-Line 243 
 sub add_hidden_text {
+Line 272 
 sub add_hidden_text {
          push @{ $self->{htexts} }, $self->_s($text);
  }
+ =head2 add_vectors
+ Add a vectors
+   $doc->add_vector(
+         'vector_name' => 42,
+         'another' => 12345,
+   );
+ =cut
+ sub add_vectors {
+         my $self = shift;
+         return unless (@_);
+         # this is ugly, but works
+         die "add_vector needs HASH as argument" unless ($#_ % 2 == 1);
+         $self->{kwords} = {@_};
+ }
+ =head2 set_score
+ Set the substitute score
+   $doc->set_score(12345);
+ =cut
+ sub set_score {
+     my $self = shift;
+     my $score = shift;
+     return unless (defined($score));
+     $self->{score} = $score;
+ }
+ =head2 score
+ Get the substitute score
+ =cut
+ sub score {
+     my $self = shift;
+     return -1 unless (defined($self->{score}));
+     return $self->{score};
+ }
  =head2 id
-Line 337 
 sub dump_draft {
+Line 413 
 sub dump_draft {
          }
          if ($self->{kwords}) {
-                 $draft .= '%%VECTOR';
+                 $draft .= '%VECTOR';
                  while (my ($key, $value) = each %{ $self->{kwords} }) {
                          $draft .= "\t$key\t$value";
                  }
                  $draft .= "\n";
          }
+         if (defined($self->{score}) && $self->{score} >= 0) {
+             $draft .= "%SCORE\t" . $self->{score} . "\n";
+         }
          $draft .= "\n";
          $draft .= join("\n", @{ $self->{dtexts} }) . "\n" if ($self->{dtexts});
-Line 631 
 sub skip {
+Line 711 
 sub skip {
  }
+ =head2 set_distinct
+   $cond->set_distinct('@author');
+ =cut
+ sub set_distinct {
+         my $self = shift;
+         $self->{distinct} = shift;
+ }
+ =head2 distinct
+ Return distinct attribute
+   print $cond->distinct;
+ =cut
+ sub distinct {
+         my $self = shift;
+         return $self->{distinct};
+ }
+ =head2 set_mask
+ Filter out some links when searching.
+ Argument array of link numbers, starting with 0 (current node).
+   $cond->set_mask(qw/0 1 4/);
+ =cut
+ sub set_mask {
+         my $self = shift;
+         return unless (@_);
+         $self->{mask} = \@_;
+ }
  package Search::Estraier::ResultDocument;
  use Carp qw/croak/;
-Line 1053 
 Add a document
+Line 1174 
 Add a document
    $node->put_doc( $document_draft ) or die "can't add document";
- Return true on success or false on failture.
+ Return true on success or false on failure.
  =cut
-Line 1061 
 sub put_doc {
+Line 1182 
 sub put_doc {
          my $self = shift;
          my $doc = shift || return;
          return unless ($self->{url} && $doc->isa('Search::Estraier::Document'));
-         $self->shuttle_url( $self->{url} . '/put_doc',
+         if ($self->shuttle_url( $self->{url} . '/put_doc',
                  'text/x-estraier-draft',
                  $doc->dump_draft,
                  undef
-         ) == 200;
+         ) == 200) {
+                 $self->_clear_info;
+                 return 1;
+         }
+         return undef;
  }
-Line 1084 
 sub out_doc {
+Line 1209 
 sub out_doc {
          my $id = shift || return;
          return unless ($self->{url});
          croak "id must be number, not '$id'" unless ($id =~ m/^\d+$/);
-         $self->shuttle_url( $self->{url} . '/out_doc',
+         if ($self->shuttle_url( $self->{url} . '/out_doc',
                  'application/x-www-form-urlencoded',
                  "id=$id",
                  undef
-         ) == 200;
+         ) == 200) {
+                 $self->_clear_info;
+                 return 1;
+         }
+         return undef;
  }
-Line 1106 
 sub out_doc_by_uri {
+Line 1235 
 sub out_doc_by_uri {
          my $self = shift;
          my $uri = shift || return;
          return unless ($self->{url});
-         $self->shuttle_url( $self->{url} . '/out_doc',
+         if ($self->shuttle_url( $self->{url} . '/out_doc',
                  'application/x-www-form-urlencoded',
                  "uri=" . uri_escape($uri),
                  undef
-         ) == 200;
+         ) == 200) {
+                 $self->_clear_info;
+                 return 1;
+         }
+         return undef;
  }
-Line 1128 
 sub edit_doc {
+Line 1261 
 sub edit_doc {
          my $self = shift;
          my $doc = shift || return;
          return unless ($self->{url} && $doc->isa('Search::Estraier::Document'));
-         $self->shuttle_url( $self->{url} . '/edit_doc',
+         if ($self->shuttle_url( $self->{url} . '/edit_doc',
                  'text/x-estraier-draft',
                  $doc->dump_draft,
                  undef
-         ) == 200;
+         ) == 200) {
+                 $self->_clear_info;
+                 return 1;
+         }
+         return undef;
  }
-Line 1290 
 sub _fetch_doc {
+Line 1427 
 sub _fetch_doc {
          $path = '/etch_doc' if ($a->{etch});
          if ($a->{id}) {
-                 croak "id must be numberm not '$a->{id}'" unless ($a->{id} =~ m/^\d+$/);
+                 croak "id must be number not '$a->{id}'" unless ($a->{id} =~ m/^\d+$/);
                  $arg = 'id=' . $a->{id};
          } elsif ($a->{uri}) {
                  $arg = 'uri=' . uri_escape($a->{uri});
-Line 1499 
 sub cond_to_query {
+Line 1636 
 sub cond_to_query {
          push @args, 'wwidth=' . $self->{wwidth};
          push @args, 'hwidth=' . $self->{hwidth};
          push @args, 'awidth=' . $self->{awidth};
-         push @args, 'skip=' . $self->{skip} if ($self->{skip});
+         push @args, 'skip=' . $cond->{skip} if ($cond->{skip});
+         if (my $distinct = $cond->distinct) {
+                 push @args, 'distinct=' . uri_escape($distinct);
+         }
+         if ($cond->{mask}) {
+                 my $mask = 0;
+                 map { $mask += ( 2 ** $_ ) } @{ $cond->{mask} };
+                 push @args, 'mask=' . $mask if ($mask);
+         }
          return join('&', @args);
  }
-Line 1646 
 sub set_user {
+Line 1794 
 sub set_user {
          croak "mode must be number, not '$mode'" unless ($mode =~ m/^\d+$/);
          $self->shuttle_url( $self->{url} . '/_set_user',
-                 'text/plain',
+                 'application/x-www-form-urlencoded',
                  'name=' . uri_escape($name) . '&mode=' . $mode,
                  undef
          ) == 200;
-Line 1679 
 sub set_link {
+Line 1827 
 sub set_link {
                  undef
          ) == 200) {
                  # refresh node info after adding link
-                 $self->_set_info;
+                 $self->_clear_info;
                  return 1;
          }
+         return undef;
  }
  =head2 admins
-Line 1726 
 sub links {
+Line 1875 
 sub links {
          return $self->{inform}->{links};
  }
+ =head2 cacheusage
+ Return cache usage for a node
+   my $cache = $node->cacheusage;
+ =cut
+ sub cacheusage {
+         my $self = shift;
+         return unless ($self->{url});
+         my $resbody;
+         my $rv = $self->shuttle_url( $self->{url} . '/cacheusage',
+                 'text/plain',
+                 undef,
+                 \$resbody,
+         );
+         return if ($rv != 200 || !$resbody);
+         return $resbody;
+ }
  =head2 master
  Set actions on Hyper Estraier node master (C<estmaster> process)
-Line 1828 
 sub master {
+Line 2002 
 sub master {
          ) or confess "shuttle_url failed";
          if ($status == $rest->{status}) {
+                 # refresh node info after sync
+                 $self->_clear_info if ($action eq 'sync' || $action =~ m/^node(?:add|del|clr)$/);
                  if ($rest->{returns} && wantarray) {
                          my @results;
-Line 1885 
 sub _set_info {
+Line 2063 
 sub _set_info {
          my @lines = split(/[\r\n]/,$resbody);
-         $self->{inform} = {};
+         $self->_clear_info;
          ( $self->{inform}->{name}, $self->{inform}->{label}, $self->{inform}->{dnum},
                  $self->{inform}->{wnum}, $self->{inform}->{size} ) = split(/\t/, shift @lines, 5);
-Line 1910 
 sub _set_info {
+Line 2088 
 sub _set_info {
  }
+ =head2 _clear_info
+ Clear information for node
+   $node->_clear_info;
+ On next call to C<name>, C<label>, C<doc_num>, C<word_num> or C<size> node
+ info will be fetch again from Hyper Estraier.
+ =cut
+ sub _clear_info {
+         my $self = shift;
+         $self->{inform} = {
+                 dnum => -1,
+                 wnum => -1,
+                 size => -1.0,
+         };
+ }
  ###
  =head1 EXPORT
-Line 1922 
 L<http://hyperestraier.sourceforge.net/>
+Line 2119 
 L<http://hyperestraier.sourceforge.net/>
  Hyper Estraier Ruby interface on which this module is based.
+ Hyper Estraier now also has pure-perl binding included in distribution. It's
+ a faster way to access databases directly if you are not running
+ C<estmaster> P2P server.
  =head1 AUTHOR
  Dobrica Pavlinusic, E<lt>dpavlin@rot13.orgE<gt>

 Legend:



Removed from v.142
 


changed lines


 
Added in v.191
 Legend:



Removed from v.142
 


changed lines


 
Added in v.191
-Removed from v.142
+Added in v.191

	ViewVC Help
Powered by ViewVC 1.1.26