/[webpac2]/trunk/lib/WebPAC/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!
ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 80 by dpavlin, Mon Nov 21 14:42:16 2005 UTC revision 211 by dpavlin, Mon Dec 5 17:47:10 2005 UTC
# Line 13  WebPAC::Search::Estraier - search Hyper Line 13  WebPAC::Search::Estraier - search Hyper
13    
14  =head1 VERSION  =head1 VERSION
15    
16  Version 0.01  Version 0.04
17    
18  =cut  =cut
19    
20  our $VERSION = '0.01';  our $VERSION = '0.04';
21    
22  =head1 SYNOPSIS  =head1 SYNOPSIS
23    
# Line 31  L<WebPAC::Output::Estraier>. Line 31  L<WebPAC::Output::Estraier>.
31  Connect to Hyper Estraier index using HTTP  Connect to Hyper Estraier index using HTTP
32    
33   my $est = new WebPAC::Search::Estraier(   my $est = new WebPAC::Search::Estraier(
34          url => 'http://localhost:1978/node/webpac2',          masterurl => 'http://localhost:1978/',
35            database => 'webpac2',
36          user => 'admin',          user => 'admin',
37          passwd => 'admin',          passwd => 'admin',
38          encoding => 'iso-8859-2',          encoding => 'iso-8859-2',
39          log => $Log::Log4Perl->log_object,          log => $Log::Log4perl->log_object,
40   );   );
41    
42  Options are:  Options are:
43    
44  =over 4  =over 4
45    
46  =item url  =item maseterurl
47    
48  URI to C<estmaster> node  URI to C<estmaster> node
49    
50    =item database
51    
52    name of C<estmaster> node
53    
54  =item user  =item user
55    
56  C<estmaster> user with read rights  C<estmaster> user with read rights
# Line 60  character encoding of C<data_structure> Line 65  character encoding of C<data_structure>
65  (and it probably is). This encoding will be converted to C<UTF-8> for  (and it probably is). This encoding will be converted to C<UTF-8> for
66  Hyper Estraier.  Hyper Estraier.
67    
68    =item log
69    
70    L<Log::Log4perl> object or equivalent (C<< $c->log >> can be used in
71    L<Catalyst> and there is support for it).
72    
73  =back  =back
74    
75  =cut  =cut
# Line 71  sub new { Line 81  sub new {
81    
82          my $log = $self->_get_logger;          my $log = $self->_get_logger;
83    
84          foreach my $p (qw/url user passwd/) {          foreach my $p (qw/masterurl user passwd/) {
85                  $log->logdie("need $p") unless ($self->{$p});                  $log->logdie("need $p") unless ($self->{$p});
86          }          }
87    
88            my $url = $self->{masterurl} . '/node/' . $self->{database};
89            $url =~ s#//#/#g;
90            $self->{url} = $url;
91    
92          $log->info("opening Hyper Estraier index $self->{'url'} as $self->{'user'}");          $log->info("opening Hyper Estraier index $self->{'url'} as $self->{'user'}");
93    
94          $self->{'db'} = HyperEstraier::Node->new($self->{'url'});          $self->{'db'} = HyperEstraier::Node->new($self->{'url'});
# Line 83  sub new { Line 97  sub new {
97          my $encoding = $self->{'encoding'} || 'ISO-8859-2';          my $encoding = $self->{'encoding'} || 'ISO-8859-2';
98          $log->info("using encoding $encoding");          $log->info("using encoding $encoding");
99    
100          $self->{'iconv'} = new Text::Iconv($encoding, 'UTF-8') or          $self->{'iconv'} = new Text::Iconv('UTF-8', $encoding) or
101                  $log->die("can't create conversion from $encoding to UTF-8");                  $log->die("can't create conversion from UTF-8 to $encoding");
102    
103          $self ? return $self : return undef;          $self ? return $self : return undef;
104  }  }
# Line 94  sub new { Line 108  sub new {
108    
109  Locate items in index  Locate items in index
110    
111    $est->search(    my @results = $est->search(
112          query => 'name of book or novel',          phrase => 'name of book or novel',
113          attr => [ qw/PersonalName Title/ ],          add_attr => [
114                    "filepath ISTRINC $q",
115                    "size NUMGT 100",
116            ],
117            get_attr => qw/PersonalName TitleProper/,
118            order => 'NUMD',
119            max => 100,
120            options => $HyperEstraier::Condition::SURE,
121            page => 42,
122    );    );
123    
124    Options are close match to Hyper Estraier API, except C<get_attr> which defines
125    attributes which will be returned in hash for each record.
126    
127    Results are returned as hash array with keys named by attributes
128    
129    Pages are numbered C< 1 ... hits/max >.
130    
131  =cut  =cut
132    
133  sub search {  sub search {
# Line 108  sub search { Line 137  sub search {
137    
138          my $log = $self->_get_logger;          my $log = $self->_get_logger;
139    
140          $log->logconfess('need db in object') unless ($self->{'db'});          #$log->debug( 'search args: ' . Dumper($args) );
141          $log->logconfess('need attr') unless ($args->{'attr'});  
142          $log->logconfess("need attr as array not " . ref($args->{'attr'}) ) unless (ref($args->{'attr'}) eq 'ARRAY');          $self->confess('need db in object') unless ($self->{db});
143            $self->confess('need get_attr') unless ($args->{get_attr});
144  #       my $database = $self->{'database'} || $log->logconfess('no database in $self');  
145  #       foreach my $p (qw/id ds type/) {          $self->confess("need get_attr as array not " . ref($args->{get_attr}) ) unless (ref($args->{get_attr}) eq 'ARRAY');
146  #               $log->logconfess("need $p") unless ($args->{$p});  
147  #       }          my $q = $args->{phrase};
148    
149            $log->debug("args: " . Dumper( $args ));
150    
151          my $cond = HyperEstraier::Condition->new();          my $cond = HyperEstraier::Condition->new();
152  #       $cond->add_attr("filepath ISTRINC $q");          if ( ref($args->{add_attr}) eq 'ARRAY' ) {
153                    $log->debug("adding search attributes: " . join(", ", @{ $args->{add_attr} }) );
154                    map {
155                            $cond->add_attr( $self->{iconv}->convert( $_ ) );
156                            $log->debug(" + $_");
157                    } @{ $args->{add_attr} };
158            };
159    
160            $cond->set_phrase( $self->{iconv}->convert($q) ) if ($q);
161            $cond->set_options( $args->{options} ) if ($args->{options});
162            $cond->set_order( $args->{order} ) if ($args->{order});
163    
164            my $max = $args->{max} || 7;
165            my $page = $args->{page} || 1;
166            if ($page < 1) {
167                    $log->warn("page number $page < 1");
168                    $page = 1;
169            }
170    
171          $cond->set_phrase( $args->{'query'} ) if ($args->{'query'});          $cond->set_max( $page * $max );
         $cond->set_max( $args->{'max'} ) if ($args->{'max'});  
 #       $cond->set_options( $HyperEstraier::Condition::SURE );  
 #       $cond->set_order( 'NUMD' );  
172    
173          my $result = $self->{'db'}->search($cond, 0) ||          my $result = $self->{db}->search($cond, 0) ||
174                  $log->die("can't search for ", sub { Dumper( $args ) });                  $log->die("can't search for ", sub { Dumper( $args ) });
175    
176          my $hits = $result->doc_num;          my $hits = $result->doc_num;
177          $log->debug("found $hits hits");          $log->debug("found $hits hits for '$q'");
   
         my @attrs = $args->{'attr'} || $self->confess("need attr");  
178    
179          my @results;          my @results;
180    
181          for my $i ( 0 .. $hits ) {          for my $i ( (($page - 1) * $max) .. ( $hits - 1 ) ) {
182    
183                  $log->debug("get_doc($i)");                  #$log->debug("get_doc($i)");
184                  my $doc = $result->get_doc( $i );                  my $doc = $result->get_doc( $i );
185                  if (! $doc) {                  if (! $doc) {
186                          $log->warn("can't find result $i");                          $log->warn("can't find result $i");
# Line 146  sub search { Line 189  sub search {
189    
190                  my $hash;                  my $hash;
191    
192                  foreach my $attr (@attrs) {                  foreach my $attr (@{ $args->{get_attr} }) {
193                          my $val = $doc->attr( $attr );                          my $val = $doc->attr( $attr );
194                          $log->debug("attr $attr = ", $val || 'undef');                          #$log->debug("attr $attr = ", $val || 'undef');
195                          $hash->{$attr} = $val if (defined($val));                          $hash->{$attr} = $self->{iconv}->convert( $val ) if (defined($val));
196                  }                  }
197    
198                  if ($hash) {                  if ($hash) {
# Line 158  sub search { Line 201  sub search {
201    
202          }          }
203    
204          $log->debug("results " . Dumper( \@results ));  #       $log->debug("results " . Dumper( \@results ));
205    
206          $log->logconfess("expected to return array") unless (wantarray);          $self->confess("expected to return array") unless (wantarray);
207    
208          return @results;          return @results;
209  }  }
# Line 176  C<die>. Line 219  C<die>.
219  sub confess {  sub confess {
220          my $self = shift;          my $self = shift;
221          if (my $log = $self->{'log'}) {          if (my $log = $self->{'log'}) {
222                  if ($log->can('confess')) {                  if ($log->can('logconfess')) {
223                          $log->confess(@_);                          $log->logconfess(@_);
224                  } elsif ($log->can('fatal')) {                  } elsif ($log->can('fatal')) {
225                          $log->fatal(@_);                          $log->fatal(@_);
226                            die @_;
227                  } elsif ($log->can('error')) {                  } elsif ($log->can('error')) {
228                          $log->error(@_);                          $log->error(@_);
229                  } else {                  } else {
Legend:
Removed from v.80  
changed lines
  Added in v.211
  ViewVC Help
Powered by ViewVC 1.1.26