/[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 62 by dpavlin, Sat Jan 7 02:40:57 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 524  sub new { Line 525  sub new {
525          my $self = {@_};          my $self = {@_};
526          bless($self, $class);          bless($self, $class);
527    
528          foreach my $f (qw/uri attrs snippet keywords/) {          croak "missing uri for ResultDocument" unless defined($self->{uri});
                 croak "missing $f for ResultDocument" unless defined($self->{$f});  
         }  
529    
530          $self ? return $self : return undef;          $self ? return $self : return undef;
531  }  }
# Line 717  sub new { Line 716  sub new {
716          };          };
717          bless($self, $class);          bless($self, $class);
718    
719          if (@_) {          my $args = {@_};
720                  $self->{debug} = shift;  
721                  warn "## Node debug on\n";          $self->{debug} = $args->{debug};
722          }          warn "## Node debug on\n" if ($self->{debug});
723    
724          $self ? return $self : return undef;          $self ? return $self : return undef;
725  }  }
# Line 1095  sub _fetch_doc { Line 1094  sub _fetch_doc {
1094    
1095  sub name {  sub name {
1096          my $self = shift;          my $self = shift;
1097          $self->set_info unless ($self->{name});          $self->_set_info unless ($self->{name});
1098          return $self->{name};          return $self->{name};
1099  }  }
1100    
# Line 1108  sub name { Line 1107  sub name {
1107    
1108  sub label {  sub label {
1109          my $self = shift;          my $self = shift;
1110          $self->set_info unless ($self->{label});          $self->_set_info unless ($self->{label});
1111          return $self->{label};          return $self->{label};
1112  }  }
1113    
# Line 1121  sub label { Line 1120  sub label {
1120    
1121  sub doc_num {  sub doc_num {
1122          my $self = shift;          my $self = shift;
1123          $self->set_info if ($self->{dnum} < 0);          $self->_set_info if ($self->{dnum} < 0);
1124          return $self->{dnum};          return $self->{dnum};
1125  }  }
1126    
# Line 1134  sub doc_num { Line 1133  sub doc_num {
1133    
1134  sub word_num {  sub word_num {
1135          my $self = shift;          my $self = shift;
1136          $self->set_info if ($self->{wnum} < 0);          $self->_set_info if ($self->{wnum} < 0);
1137          return $self->{wnum};          return $self->{wnum};
1138  }  }
1139    
# Line 1147  sub word_num { Line 1146  sub word_num {
1146    
1147  sub size {  sub size {
1148          my $self = shift;          my $self = shift;
1149          $self->set_info if ($self->{size} < 0);          $self->_set_info if ($self->{size} < 0);
1150          return $self->{size};          return $self->{size};
1151  }  }
1152    
# Line 1176  sub search { Line 1175  sub search {
1175    
1176          my $rv = $self->shuttle_url( $self->{url} . '/search',          my $rv = $self->shuttle_url( $self->{url} . '/search',
1177                  'application/x-www-form-urlencoded',                  'application/x-www-form-urlencoded',
1178                  $self->cond_to_query( $cond ),                  $self->cond_to_query( $cond, $depth ),
1179                  \$resbody,                  \$resbody,
1180          );          );
1181          return if ($rv != 200);          return if ($rv != 200);
# Line 1268  sub search { Line 1267  sub search {
1267    
1268  =head2 cond_to_query  =head2 cond_to_query
1269    
1270    my $args = $node->cond_to_query( $cond );  Return URI encoded string generated from Search::Estraier::Condition
1271    
1272      my $args = $node->cond_to_query( $cond, $depth );
1273    
1274  =cut  =cut
1275    
# Line 1277  sub cond_to_query { Line 1278  sub cond_to_query {
1278    
1279          my $cond = shift || return;          my $cond = shift || return;
1280          croak "condition must be Search::Estraier::Condition, not '$cond->isa'" unless ($cond->isa('Search::Estraier::Condition'));          croak "condition must be Search::Estraier::Condition, not '$cond->isa'" unless ($cond->isa('Search::Estraier::Condition'));
1281            my $depth = shift;
1282    
1283          my @args;          my @args;
1284    
# Line 1304  sub cond_to_query { Line 1306  sub cond_to_query {
1306                  push @args, 'options=' . $options;                  push @args, 'options=' . $options;
1307          }          }
1308    
1309          push @args, 'depth=' . $self->{depth} if ($self->{depth});          push @args, 'depth=' . $depth if ($depth);
1310          push @args, 'wwidth=' . $self->{wwidth};          push @args, 'wwidth=' . $self->{wwidth};
1311          push @args, 'hwidth=' . $self->{hwidth};          push @args, 'hwidth=' . $self->{hwidth};
1312          push @args, 'awidth=' . $self->{awidth};          push @args, 'awidth=' . $self->{awidth};
# Line 1325  body will be saved within object. Line 1327  body will be saved within object.
1327    
1328  =cut  =cut
1329    
1330    use LWP::UserAgent;
1331    
1332  sub shuttle_url {  sub shuttle_url {
1333          my $self = shift;          my $self = shift;
1334    
# Line 1343  sub shuttle_url { Line 1347  sub shuttle_url {
1347                  return -1;                  return -1;
1348          }          }
1349    
1350          my ($host,$port,$query) = ($url->host, $url->port, $url->path);          my $ua = LWP::UserAgent->new;
1351            $ua->agent( "Search-Estraier/$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;  
1352    
1353            my $req;
1354          if ($reqbody) {          if ($reqbody) {
1355                  $headers .= "POST $query HTTP/1.0\r\n";                  $req = HTTP::Request->new(POST => $url);
1356          } else {          } else {
1357                  $headers .= "GET $query HTTP/1.0\r\n";                  $req = HTTP::Request->new(GET => $url);
1358          }          }
1359    
1360          $headers .= "Host: " . $url->host . ":" . $url->port . "\r\n";          $req->headers->header( 'Host' => $url->host . ":" . $url->port );
1361          $headers .= "Connection: close\r\n";          $req->headers->header( 'Connection', 'close' );
1362          $headers .= "User-Agent: Search-Estraier/$Search::Estraier::VERSION\r\n";          $req->headers->header( 'Authorization', 'Basic ' . $self->{auth} );
1363          $headers .= "Content-Type: $content_type\r\n";          $req->content_type( $content_type );
         $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";  
1364    
1365          my $sock = IO::Socket::INET->new(          warn $req->headers->as_string,"\n" if ($self->{debug});
                 PeerAddr        => $host,  
                 PeerPort        => $port,  
                 Proto           => 'tcp',  
                 Timeout         => $self->{timeout} || 90,  
         );  
1366    
1367          if (! $sock) {          if ($reqbody) {
1368                  carp "can't open socket to $host:$port";                  warn "$reqbody\n" if ($self->{debug});
1369                  return -1;                  $req->content( $reqbody );
1370          }          }
1371    
1372          warn $headers if ($self->{debug});          my $res = $ua->request($req) || croak "can't make request to $url: $!";
1373    
1374          print $sock $headers or          warn "## response status: ",$res->status_line,"\n" if ($self->{debug});
                 carp "can't send headers to network:\n$headers\n" and return -1;  
1375    
1376          if ($reqbody) {          return -1 if (! $res->is_success);
                 warn "$reqbody\n" if ($self->{debug});  
                 print $sock $reqbody or  
                         carp "can't send request body to network:\n$$reqbody\n" and return -1;  
         }  
1377    
1378          my $line = <$sock>;          ($self->{status}, $self->{status_message}) = split(/\s+/, $res->status_line, 2);
         chomp($line);  
         my ($schema, $res_status, undef) = split(/  */, $line, 3);  
         return if ($schema !~ /^HTTP/ || ! $res_status);  
   
         $self->{status} = $res_status;  
         warn "## response status: $res_status\n" if ($self->{debug});  
   
         # skip rest of headers  
         $line = <$sock>;  
         while ($line) {  
                 $line = <$sock>;  
                 $line =~ s/[\r\n]+$//;  
                 warn "## ", $line || 'NULL', " ##\n" if ($self->{debug});  
         };  
1379    
1380          # read body          $$resbody .= $res->content;
         $len = 0;  
         do {  
                 $len = read($sock, my $buf, 8192);  
                 $$resbody .= $buf if ($resbody);  
         } while ($len);  
1381    
1382          warn "## response body:\n$$resbody\n" if ($resbody && $self->{debug});          warn "## response body:\n$$resbody\n" if ($resbody && $self->{debug});
1383    
# Line 1425  sub shuttle_url { Line 1385  sub shuttle_url {
1385  }  }
1386    
1387    
1388  =head2 set_info  =head2 set_snippet_width
1389    
1390    Set width of snippets in results
1391    
1392      $node->set_snippet_width( $wwidth, $hwidth, $awidth );
1393    
1394    C<$wwidth> specifies whole width of snippet. It's C<480> by default. If it's C<0> snippet
1395    is not sent with results. If it is negative, whole document text is sent instead of snippet.
1396    
1397    C<$hwidth> specified width of strings from beginning of string. Default
1398    value is C<96>. Negative or zero value keep previous value.
1399    
1400    C<$awidth> specifies width of strings around each highlighted word. It's C<96> by default.
1401    If negative of zero value is provided previous value is kept unchanged.
1402    
1403    =cut
1404    
1405    sub set_snippet_width {
1406            my $self = shift;
1407    
1408            my ($wwidth, $hwidth, $awidth) = @_;
1409            $self->{wwidth} = $wwidth;
1410            $self->{hwidth} = $hwidth if ($hwidth >= 0);
1411            $self->{awidth} = $awidth if ($awidth >= 0);
1412    }
1413    
1414    
1415    =head2 set_user
1416    
1417    Manage users of node
1418    
1419      $node->set_user( 'name', $mode );
1420    
1421    C<$mode> can be one of:
1422    
1423    =over 4
1424    
1425    =item 0
1426    
1427    delete account
1428    
1429    =item 1
1430    
1431    set administrative right for user
1432    
1433    =item 2
1434    
1435    set user account as guest
1436    
1437    =back
1438    
1439    Return true on success, otherwise false.
1440    
1441    =cut
1442    
1443    sub set_user {
1444            my $self = shift;
1445            my ($name, $mode) = @_;
1446    
1447            return unless ($self->{url});
1448            croak "mode must be number, not '$mode'" unless ($mode =~ m/^\d+$/);
1449    
1450            $self->shuttle_url( $self->{url} . '/_set_user',
1451                    'text/plain',
1452                    'name=' . uri_escape($name) . '&mode=' . $mode,
1453                    undef
1454            ) == 200;
1455    }
1456    
1457    
1458    =head2 set_link
1459    
1460    Manage node links
1461    
1462      $node->set_link('http://localhost:1978/node/another', 'another node label', $credit);
1463    
1464    If C<$credit> is negative, link is removed.
1465    
1466    =cut
1467    
1468    sub set_link {
1469            my $self = shift;
1470            my ($url, $label, $credit) = @_;
1471    
1472            return unless ($self->{url});
1473            croak "mode credit be number, not '$credit'" unless ($credit =~ m/^\d+$/);
1474    
1475            my $reqbody = 'url=' . uri_escape($url) . '&label=' . uri_escape($label);
1476            $reqbody .= '&credit=' . $credit if ($credit > 0);
1477    
1478            $self->shuttle_url( $self->{url} . '/_set_link',
1479                    'text/plain',
1480                    $reqbody,
1481                    undef
1482            ) == 200;
1483    }
1484    
1485    
1486    =head1 PRIVATE METHODS
1487    
1488    You could call those directly, but you don't have to. I hope.
1489    
1490    =head2 _set_info
1491    
1492  Set information for node  Set information for node
1493    
1494    $node->set_info;    $node->_set_info;
1495    
1496  =cut  =cut
1497    
1498  sub set_info {  sub _set_info {
1499          my $self = shift;          my $self = shift;
1500    
1501          $self->{status} = -1;          $self->{status} = -1;
# Line 1448  sub set_info { Line 1510  sub set_info {
1510    
1511          return if ($rv != 200 || !$resbody);          return if ($rv != 200 || !$resbody);
1512    
1513          chomp($resbody);          # it seems that response can have multiple line endings
1514            $resbody =~ s/[\r\n]+$//;
1515    
1516          ( $self->{name}, $self->{label}, $self->{dnum}, $self->{wnum}, $self->{size} ) =          ( $self->{name}, $self->{label}, $self->{dnum}, $self->{wnum}, $self->{size} ) =
1517                  split(/\t/, $resbody, 5);                  split(/\t/, $resbody, 5);
Legend:
Removed from v.53  
changed lines
  Added in v.62
  ViewVC Help
Powered by ViewVC 1.1.26