/[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!
ViewVC logotype

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

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

revision 53 by dpavlin, Fri Jan 6 14:39:45 2006 UTC revision 61 by dpavlin, Sat Jan 7 01:21:28 2006 UTC
# Line 4  use 5.008; Line 4  use 5.008;
4  use strict;  use strict;
5  use warnings;  use warnings;
6    
7  our $VERSION = '0.00';  our $VERSION = '0.01';
8    
9  =head1 NAME  =head1 NAME
10    
# Line 205  Returns array with attribute names from Line 205  Returns array with attribute names from
205    
206  sub attr_names {  sub attr_names {
207          my $self = shift;          my $self = shift;
208          croak "attr_names return array, not scalar" if (! wantarray);          return unless ($self->{attrs});
209            #croak "attr_names return array, not scalar" if (! wantarray);
210          return sort keys %{ $self->{attrs} };          return sort keys %{ $self->{attrs} };
211  }  }
212    
# Line 221  Returns value of an attribute. Line 222  Returns value of an attribute.
222  sub attr {  sub attr {
223          my $self = shift;          my $self = shift;
224          my $name = shift;          my $name = shift;
225            return unless (defined($name) && $self->{attrs});
226          return $self->{'attrs'}->{ $name };          return $self->{attrs}->{ $name };
227  }  }
228    
229    
# Line 236  Returns array with text sentences. Line 237  Returns array with text sentences.
237    
238  sub texts {  sub texts {
239          my $self = shift;          my $self = shift;
240          confess "texts return array, not scalar" if (! wantarray);          #confess "texts return array, not scalar" if (! wantarray);
241          return @{ $self->{dtexts} };          return @{ $self->{dtexts} } if ($self->{dtexts});
242  }  }
243    
244    
# Line 251  Return whole text as single scalar. Line 252  Return whole text as single scalar.
252    
253  sub cat_texts {  sub cat_texts {
254          my $self = shift;          my $self = shift;
255          return join(' ',@{ $self->{dtexts} });          return join(' ',@{ $self->{dtexts} }) if ($self->{dtexts});
256  }  }
257    
258    
# Line 460  Return search result attrs. Line 461  Return search result attrs.
461  sub attrs {  sub attrs {
462          my $self = shift;          my $self = shift;
463          #croak "attrs return array, not scalar" if (! wantarray);          #croak "attrs return array, not scalar" if (! wantarray);
464          return @{ $self->{attrs} };          return @{ $self->{attrs} } if ($self->{attrs});
465  }  }
466    
467    
# Line 717  sub new { Line 718  sub new {
718          };          };
719          bless($self, $class);          bless($self, $class);
720    
721          if (@_) {          my $args = {@_};
722                  $self->{debug} = shift;  
723                  warn "## Node debug on\n";          $self->{debug} = $args->{debug};
724          }          warn "## Node debug on\n" if ($self->{debug});
725    
726          $self ? return $self : return undef;          $self ? return $self : return undef;
727  }  }
# Line 1095  sub _fetch_doc { Line 1096  sub _fetch_doc {
1096    
1097  sub name {  sub name {
1098          my $self = shift;          my $self = shift;
1099          $self->set_info unless ($self->{name});          $self->_set_info unless ($self->{name});
1100          return $self->{name};          return $self->{name};
1101  }  }
1102    
# Line 1108  sub name { Line 1109  sub name {
1109    
1110  sub label {  sub label {
1111          my $self = shift;          my $self = shift;
1112          $self->set_info unless ($self->{label});          $self->_set_info unless ($self->{label});
1113          return $self->{label};          return $self->{label};
1114  }  }
1115    
# Line 1121  sub label { Line 1122  sub label {
1122    
1123  sub doc_num {  sub doc_num {
1124          my $self = shift;          my $self = shift;
1125          $self->set_info if ($self->{dnum} < 0);          $self->_set_info if ($self->{dnum} < 0);
1126          return $self->{dnum};          return $self->{dnum};
1127  }  }
1128    
# Line 1134  sub doc_num { Line 1135  sub doc_num {
1135    
1136  sub word_num {  sub word_num {
1137          my $self = shift;          my $self = shift;
1138          $self->set_info if ($self->{wnum} < 0);          $self->_set_info if ($self->{wnum} < 0);
1139          return $self->{wnum};          return $self->{wnum};
1140  }  }
1141    
# Line 1147  sub word_num { Line 1148  sub word_num {
1148    
1149  sub size {  sub size {
1150          my $self = shift;          my $self = shift;
1151          $self->set_info if ($self->{size} < 0);          $self->_set_info if ($self->{size} < 0);
1152          return $self->{size};          return $self->{size};
1153  }  }
1154    
# Line 1176  sub search { Line 1177  sub search {
1177    
1178          my $rv = $self->shuttle_url( $self->{url} . '/search',          my $rv = $self->shuttle_url( $self->{url} . '/search',
1179                  'application/x-www-form-urlencoded',                  'application/x-www-form-urlencoded',
1180                  $self->cond_to_query( $cond ),                  $self->cond_to_query( $cond, $depth ),
1181                  \$resbody,                  \$resbody,
1182          );          );
1183          return if ($rv != 200);          return if ($rv != 200);
# Line 1268  sub search { Line 1269  sub search {
1269    
1270  =head2 cond_to_query  =head2 cond_to_query
1271    
1272    my $args = $node->cond_to_query( $cond );  Return URI encoded string generated from Search::Estraier::Condition
1273    
1274      my $args = $node->cond_to_query( $cond, $depth );
1275    
1276  =cut  =cut
1277    
# Line 1277  sub cond_to_query { Line 1280  sub cond_to_query {
1280    
1281          my $cond = shift || return;          my $cond = shift || return;
1282          croak "condition must be Search::Estraier::Condition, not '$cond->isa'" unless ($cond->isa('Search::Estraier::Condition'));          croak "condition must be Search::Estraier::Condition, not '$cond->isa'" unless ($cond->isa('Search::Estraier::Condition'));
1283            my $depth = shift;
1284    
1285          my @args;          my @args;
1286    
# Line 1304  sub cond_to_query { Line 1308  sub cond_to_query {
1308                  push @args, 'options=' . $options;                  push @args, 'options=' . $options;
1309          }          }
1310    
1311          push @args, 'depth=' . $self->{depth} if ($self->{depth});          push @args, 'depth=' . $depth if ($depth);
1312          push @args, 'wwidth=' . $self->{wwidth};          push @args, 'wwidth=' . $self->{wwidth};
1313          push @args, 'hwidth=' . $self->{hwidth};          push @args, 'hwidth=' . $self->{hwidth};
1314          push @args, 'awidth=' . $self->{awidth};          push @args, 'awidth=' . $self->{awidth};
# Line 1325  body will be saved within object. Line 1329  body will be saved within object.
1329    
1330  =cut  =cut
1331    
1332    use LWP::UserAgent;
1333    
1334  sub shuttle_url {  sub shuttle_url {
1335          my $self = shift;          my $self = shift;
1336    
# Line 1343  sub shuttle_url { Line 1349  sub shuttle_url {
1349                  return -1;                  return -1;
1350          }          }
1351    
1352          my ($host,$port,$query) = ($url->host, $url->port, $url->path);          my $ua = LWP::UserAgent->new;
1353            $ua->agent( "Search-Estraier/$Search::Estraier::VERSION" );
         if ($self->{pxhost}) {  
                 ($host,$port) = ($self->{pxhost}, $self->{pxport});  
                 $query = "http://$host:$port/$query";  
         }  
   
         $query .= '?' . $url->query if ($url->query && ! $reqbody);  
   
         my $headers;  
1354    
1355            my $req;
1356          if ($reqbody) {          if ($reqbody) {
1357                  $headers .= "POST $query HTTP/1.0\r\n";                  $req = HTTP::Request->new(POST => $url);
1358          } else {          } else {
1359                  $headers .= "GET $query HTTP/1.0\r\n";                  $req = HTTP::Request->new(GET => $url);
         }  
   
         $headers .= "Host: " . $url->host . ":" . $url->port . "\r\n";  
         $headers .= "Connection: close\r\n";  
         $headers .= "User-Agent: Search-Estraier/$Search::Estraier::VERSION\r\n";  
         $headers .= "Content-Type: $content_type\r\n";  
         $headers .= "Authorization: Basic $self->{auth}\r\n";  
         my $len = 0;  
         {  
                 use bytes;  
                 $len = length($reqbody) if ($reqbody);  
         }  
         $headers .= "Content-Length: $len\r\n";  
         $headers .= "\r\n";  
   
         my $sock = IO::Socket::INET->new(  
                 PeerAddr        => $host,  
                 PeerPort        => $port,  
                 Proto           => 'tcp',  
                 Timeout         => $self->{timeout} || 90,  
         );  
   
         if (! $sock) {  
                 carp "can't open socket to $host:$port";  
                 return -1;  
1360          }          }
1361    
1362          warn $headers if ($self->{debug});          $req->headers->header( 'Host' => $url->host . ":" . $url->port );
1363            $req->headers->header( 'Connection', 'close' );
1364            $req->headers->header( 'Authorization', 'Basic ' . $self->{auth} );
1365            $req->content_type( $content_type );
1366    
1367          print $sock $headers or          warn $req->headers->as_string,"\n" if ($self->{debug});
                 carp "can't send headers to network:\n$headers\n" and return -1;  
1368    
1369          if ($reqbody) {          if ($reqbody) {
1370                  warn "$reqbody\n" if ($self->{debug});                  warn "$reqbody\n" if ($self->{debug});
1371                  print $sock $reqbody or                  $req->content( $reqbody );
                         carp "can't send request body to network:\n$$reqbody\n" and return -1;  
1372          }          }
1373    
1374          my $line = <$sock>;          my $res = $ua->request($req) || croak "can't make request to $url: $!";
1375          chomp($line);  
1376          my ($schema, $res_status, undef) = split(/  */, $line, 3);          warn "## response status: ",$res->status_line,"\n" if ($self->{debug});
1377          return if ($schema !~ /^HTTP/ || ! $res_status);  
1378            return -1 if (! $res->is_success);
1379          $self->{status} = $res_status;  
1380          warn "## response status: $res_status\n" if ($self->{debug});          ($self->{status}, $self->{status_message}) = split(/\s+/, $res->status_line, 2);
   
         # skip rest of headers  
         $line = <$sock>;  
         while ($line) {  
                 $line = <$sock>;  
                 $line =~ s/[\r\n]+$//;  
                 warn "## ", $line || 'NULL', " ##\n" if ($self->{debug});  
         };  
1381    
1382          # read body          $$resbody .= $res->content;
         $len = 0;  
         do {  
                 $len = read($sock, my $buf, 8192);  
                 $$resbody .= $buf if ($resbody);  
         } while ($len);  
1383    
1384          warn "## response body:\n$$resbody\n" if ($resbody && $self->{debug});          warn "## response body:\n$$resbody\n" if ($resbody && $self->{debug});
1385    
# Line 1425  sub shuttle_url { Line 1387  sub shuttle_url {
1387  }  }
1388    
1389    
1390  =head2 set_info  =head2 set_snippet_width
1391    
1392    Set width of snippets in results
1393    
1394      $node->set_snippet_width( $wwidth, $hwidth, $awidth );
1395    
1396    C<$wwidth> specifies whole width of snippet. It's C<480> by default. If it's C<0> snippet
1397    is not sent with results. If it is negative, whole document text is sent instead of snippet.
1398    
1399    C<$hwidth> specified width of strings from beginning of string. Default
1400    value is C<96>. Negative or zero value keep previous value.
1401    
1402    C<$awidth> specifies width of strings around each highlighted word. It's C<96> by default.
1403    If negative of zero value is provided previous value is kept unchanged.
1404    
1405    =cut
1406    
1407    sub set_snippet_width {
1408            my $self = shift;
1409    
1410            my ($wwidth, $hwidth, $awidth) = @_;
1411            $self->{wwidth} = $wwidth;
1412            $self->{hwidth} = $hwidth if ($hwidth >= 0);
1413            $self->{awidth} = $awidth if ($awidth >= 0);
1414    }
1415    
1416    
1417    =head2 set_user
1418    
1419    Manage users of node
1420    
1421      $node->set_user( 'name', $mode );
1422    
1423    C<$mode> can be one of:
1424    
1425    =over 4
1426    
1427    =item 0
1428    
1429    delete account
1430    
1431    =item 1
1432    
1433    set administrative right for user
1434    
1435    =item 2
1436    
1437    set user account as guest
1438    
1439    =back
1440    
1441    Return true on success, otherwise false.
1442    
1443    =cut
1444    
1445    sub set_user {
1446            my $self = shift;
1447            my ($name, $mode) = @_;
1448    
1449            return unless ($self->{url});
1450            croak "mode must be number, not '$mode'" unless ($mode =~ m/^\d+$/);
1451    
1452            $self->shuttle_url( $self->{url} . '/_set_user',
1453                    'text/plain',
1454                    'name=' . uri_escape($name) . '&mode=' . $mode,
1455                    undef
1456            ) == 200;
1457    }
1458    
1459    
1460    =head2 set_link
1461    
1462    Manage node links
1463    
1464      $node->set_link('http://localhost:1978/node/another', 'another node label', $credit);
1465    
1466    If C<$credit> is negative, link is removed.
1467    
1468    =cut
1469    
1470    sub set_link {
1471            my $self = shift;
1472            my ($url, $label, $credit) = @_;
1473    
1474            return unless ($self->{url});
1475            croak "mode credit be number, not '$credit'" unless ($credit =~ m/^\d+$/);
1476    
1477            my $reqbody = 'url=' . uri_escape($url) . '&label=' . uri_escape($label);
1478            $reqbody .= '&credit=' . $credit if ($credit > 0);
1479    
1480            $self->shuttle_url( $self->{url} . '/_set_link',
1481                    'text/plain',
1482                    $reqbody,
1483                    undef
1484            ) == 200;
1485    }
1486    
1487    
1488    =head1 PRIVATE METHODS
1489    
1490    You could call those directly, but you don't have to. I hope.
1491    
1492    =head2 _set_info
1493    
1494  Set information for node  Set information for node
1495    
1496    $node->set_info;    $node->_set_info;
1497    
1498  =cut  =cut
1499    
1500  sub set_info {  sub _set_info {
1501          my $self = shift;          my $self = shift;
1502    
1503          $self->{status} = -1;          $self->{status} = -1;
# Line 1448  sub set_info { Line 1512  sub set_info {
1512    
1513          return if ($rv != 200 || !$resbody);          return if ($rv != 200 || !$resbody);
1514    
1515          chomp($resbody);          # it seems that response can have multiple line endings
1516            $resbody =~ s/[\r\n]+$//;
1517    
1518          ( $self->{name}, $self->{label}, $self->{dnum}, $self->{wnum}, $self->{size} ) =          ( $self->{name}, $self->{label}, $self->{dnum}, $self->{wnum}, $self->{size} ) =
1519                  split(/\t/, $resbody, 5);                  split(/\t/, $resbody, 5);
Legend:
Removed from v.53  
changed lines
  Added in v.61
  ViewVC Help
Powered by ViewVC 1.1.26