--- trunk/Estraier.pm	2006/01/05 15:33:48	30
+++ trunk/Estraier.pm	2006/01/06 12:40:23	49
@@ -27,6 +27,10 @@
 
 =cut
 
+=head1 Inheritable common methods
+
+This methods should really move somewhere else.
+
 =head2 _s
 
 Remove multiple whitespaces from string, as well as whitespaces at beginning or end
@@ -56,6 +60,7 @@
 This class implements Document which is collection of attributes
 (key=value), vectors (also key value) display text and hidden text.
 
+
 =head2 new
 
 Create new document, empty or from draft.
@@ -175,6 +180,7 @@
 	push @{ $self->{htexts} }, $self->_s($text);
 }
 
+
 =head2 id
 
 Get the ID number of document. If the object has never been registred, C<-1> is returned.
@@ -188,6 +194,7 @@
 	return $self->{id};
 }
 
+
 =head2 attr_names
 
 Returns array with attribute names from document object.
@@ -233,6 +240,7 @@
 	return @{ $self->{dtexts} };
 }
 
+
 =head2 cat_texts
 
 Return whole text as single scalar.
@@ -246,6 +254,7 @@
 	return join(' ',@{ $self->{dtexts} });
 }
 
+
 =head2 dump_draft
 
 Dump draft data from document object.
@@ -272,12 +281,13 @@
 
 	$draft .= "\n";
 
-	$draft .= join("\n", @{ $self->{dtexts} }) . "\n";
-	$draft .= "\t" . join("\n\t", @{ $self->{htexts} }) . "\n";
+	$draft .= join("\n", @{ $self->{dtexts} }) . "\n" if ($self->{dtexts});
+	$draft .= "\t" . join("\n\t", @{ $self->{htexts} }) . "\n" if ($self->{htexts});
 
 	return $draft;
 }
 
+
 =head2 delete
 
 Empty document object
@@ -330,6 +340,7 @@
 	$self ? return $self : return undef;
 }
 
+
 =head2 set_phrase
 
   $cond->set_phrase('search phrase');
@@ -341,6 +352,7 @@
 	$self->{phrase} = $self->_s( shift );
 }
 
+
 =head2 add_attr
 
   $cond->add_attr('@URI STRINC /~dpavlin/');
@@ -353,6 +365,7 @@
 	push @{ $self->{attrs} }, $self->_s( $attr );
 }
 
+
 =head2 set_order
 
   $cond->set_order('@mdate NUMD');
@@ -364,6 +377,7 @@
 	$self->{order} = shift;
 }
 
+
 =head2 set_max
 
   $cond->set_max(42);
@@ -373,10 +387,11 @@
 sub set_max {
 	my $self = shift;
 	my $max = shift;
-	croak "set_max needs number" unless ($max =~ m/^\d+$/);
+	croak "set_max needs number, not '$max'" unless ($max =~ m/^\d+$/);
 	$self->{max} = $max;
 }
 
+
 =head2 set_options
 
   $cond->set_options( SURE => 1 );
@@ -405,6 +420,7 @@
 	$self->{options} ||= $options->{$option};
 }
 
+
 =head2 phrase
 
 Return search phrase.
@@ -418,6 +434,7 @@
 	return $self->{phrase};
 }
 
+
 =head2 order
 
 Return search result order.
@@ -431,6 +448,7 @@
 	return $self->{order};
 }
 
+
 =head2 attrs
 
 Return search result attrs.
@@ -445,6 +463,7 @@
 	return @{ $self->{attrs} };
 }
 
+
 =head2 max
 
 Return maximum number of results.
@@ -460,6 +479,7 @@
 	return $self->{max};
 }
 
+
 =head2 options
 
 Return options for this condition.
@@ -511,6 +531,7 @@
 	$self ? return $self : return undef;
 }
 
+
 =head2 uri
 
 Return URI of result document
@@ -539,6 +560,7 @@
 	return sort keys %{ $self->{attrs} };
 }
 
+
 =head2 attr
 
 Returns value of an attribute.
@@ -553,6 +575,7 @@
 	return $self->{attrs}->{ $name };
 }
 
+
 =head2 snippet
 
 Return snippet from result document
@@ -566,6 +589,7 @@
 	return $self->{snippet};
 }
 
+
 =head2 keywords
 
 Return keywords from result document
@@ -610,6 +634,7 @@
 	$self ? return $self : return undef;
 }
 
+
 =head2 doc_num
 
 Return number of documents
@@ -623,6 +648,7 @@
 	return $#{$self->{docs}};
 }
 
+
 =head2 get_doc
 
 Return single document
@@ -636,11 +662,12 @@
 sub get_doc {
 	my $self = shift;
 	my $num = shift;
-	croak "expect number as argument" unless ($num =~ m/^\d+$/);
+	croak "expect number as argument, not '$num'" unless ($num =~ m/^\d+$/);
 	return undef if ($num < 0 || $num > $self->{docs});
 	return $self->{docs}->[$num];
 }
 
+
 =head2 hint
 
 Return specific hint from results.
@@ -661,7 +688,11 @@
 
 package Search::Estraier::Node;
 
-use Carp qw/croak/;
+use Carp qw/carp croak confess/;
+use URI;
+use MIME::Base64;
+use IO::Socket::INET;
+use URI::Escape qw/uri_escape/;
 
 =head1 Search::Estraier::Node
 
@@ -675,7 +706,7 @@
 	my $class = shift;
 	my $self = {
 		pxport => -1,
-		timeout => -1,
+		timeout => 0,	# this used to be -1
 		dnum => -1,
 		wnum => -1,
 		size => -1.0,
@@ -686,9 +717,15 @@
 	};
 	bless($self, $class);
 
+	if (@_) {
+		$self->{debug} = shift;
+		warn "## Node debug on\n";
+	}
+
 	$self ? return $self : return undef;
 }
 
+
 =head2 set_url
 
 Specify URL to node server
@@ -702,6 +739,7 @@
 	$self->{url} = shift;
 }
 
+
 =head2 set_proxy
 
 Specify proxy server to connect to node server
@@ -713,11 +751,12 @@
 sub set_proxy {
 	my $self = shift;
 	my ($host,$port) = @_;
-	croak "proxy port must be number" unless ($port =~ m/^\d+$/);
+	croak "proxy port must be number, not '$port'" unless ($port =~ m/^\d+$/);
 	$self->{pxhost} = $host;
 	$self->{pxport} = $port;
 }
 
+
 =head2 set_timeout
 
 Specify timeout of connection in seconds
@@ -729,63 +768,532 @@
 sub set_timeout {
 	my $self = shift;
 	my $sec = shift;
-	croak "timeout must be number" unless ($sec =~ m/^\d+$/);
+	croak "timeout must be number, not '$sec'" unless ($sec =~ m/^\d+$/);
 	$self->{timeout} = $sec;
 }
 
-package Search::Estraier::Master;
 
-use Carp;
+=head2 set_auth
+
+Specify name and password for authentication to node server.
+
+  $node->set_auth('clint','eastwood');
+
+=cut
+
+sub set_auth {
+	my $self = shift;
+	my ($login,$passwd) = @_;
+	my $basic_auth = encode_base64( "$login:$passwd" );
+	chomp($basic_auth);
+	$self->{auth} = $basic_auth;
+}
+
+
+=head2 status
+
+Return status code of last request.
+
+  print $node->status;
+
+C<-1> means connection failure.
+
+=cut
+
+sub status {
+	my $self = shift;
+	return $self->{status};
+}
+
+
+=head2 put_doc
+
+Add a document
+
+  $node->put_doc( $document_draft ) or die "can't add document";
+
+Return true on success or false on failture.
+
+=cut
+
+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',
+		'text/x-estraier-draft',
+		$doc->dump_draft,
+		undef
+	) == 200;
+}
+
+
+=head2 out_doc
+
+Remove a document
+
+  $node->out_doc( document_id ) or "can't remove document";
+
+Return true on success or false on failture.
+
+=cut
+
+sub out_doc {
+	my $self = shift;
+	my $id = shift || return;
+	return unless ($self->{url});
+	croak "id must be number, not '$id'" unless ($id =~ m/^\d+$/);
+	$self->shuttle_url( $self->{url} . '/out_doc',
+		'application/x-www-form-urlencoded',
+		"id=$id",
+		undef
+	) == 200;
+}
+
+
+=head2 out_doc_by_uri
+
+Remove a registrated document using it's uri
+
+  $node->out_doc_by_uri( 'file:///document/uri/42' ) or "can't remove document";
+
+Return true on success or false on failture.
+
+=cut
+
+sub out_doc_by_uri {
+	my $self = shift;
+	my $uri = shift || return;
+	return unless ($self->{url});
+	$self->shuttle_url( $self->{url} . '/out_doc',
+		'application/x-www-form-urlencoded',
+		"uri=$uri",
+		undef
+	) == 200;
+}
+
+
+=head2 edit_doc
+
+Edit attributes of a document
+
+  $node->edit_doc( $document_draft ) or die "can't edit document";
+
+Return true on success or false on failture.
+
+=cut
+
+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',
+		'text/x-estraier-draft',
+		$doc->dump_draft,
+		undef
+	) == 200;
+}
+
+
+=head2 get_doc
+
+Retreive document
+
+  my $doc = $node->get_doc( document_id ) or die "can't get document";
+
+Return true on success or false on failture.
+
+=cut
+
+sub get_doc {
+	my $self = shift;
+	my $id = shift || return;
+	return $self->_fetch_doc( id => $id );
+}
+
+
+=head2 get_doc_by_uri
+
+Retreive document
+
+  my $doc = $node->get_doc_by_uri( 'file:///document/uri/42' ) or die "can't get document";
+
+Return true on success or false on failture.
+
+=cut
+
+sub get_doc_by_uri {
+	my $self = shift;
+	my $uri = shift || return;
+	return $self->_fetch_doc( uri => $uri );
+}
+
+
+=head2 get_doc_attr
+
+Retrieve the value of an atribute from object
+
+  my $val = $node->get_doc_attr( document_id, 'attribute_name' ) or
+  	die "can't get document attribute";
+
+=cut
+
+sub get_doc_attr {
+	my $self = shift;
+	my ($id,$name) = @_;
+	return unless ($id && $name);
+	return $self->_fetch_doc( id => $id, attr => $name );
+}
+
+
+=head2 get_doc_attr_by_uri
+
+Retrieve the value of an atribute from object
+
+  my $val = $node->get_doc_attr_by_uri( document_id, 'attribute_name' ) or
+  	die "can't get document attribute";
+
+=cut
+
+sub get_doc_attr_by_uri {
+	my $self = shift;
+	my ($uri,$name) = @_;
+	return unless ($uri && $name);
+	return $self->_fetch_doc( uri => $uri, attr => $name );
+}
+
+
+=head2 etch_doc
+
+Exctract document keywords
+
+  my $keywords = $node->etch_doc( document_id ) or die "can't etch document";
+
+=cut
+
+sub etch_doc {
+	my $self = shift;
+	my $id = shift || return;
+	return $self->_fetch_doc( id => $id, etch => 1 );
+}
+
+=head2 etch_doc_by_uri
+
+Retreive document
+
+  my $keywords = $node->etch_doc_by_uri( 'file:///document/uri/42' ) or die "can't etch document";
+
+Return true on success or false on failture.
+
+=cut
+
+sub etch_doc_by_uri {
+	my $self = shift;
+	my $uri = shift || return;
+	return $self->_fetch_doc( uri => $uri, etch => 1 );
+}
+
+
+=head2 uri_to_id
 
-=head1 Search::Estraier::Master
+Get ID of document specified by URI
 
-Controll node master. This requires user with administration priviledges.
+  my $id = $node->uri_to_id( 'file:///document/uri/42' );
 
 =cut
 
-{
-	package RequestAgent;
-	our @ISA = qw(LWP::UserAgent);
+sub uri_to_id {
+	my $self = shift;
+	my $uri = shift || return;
+	return $self->_fetch_doc( uri => $uri, path => '/uri_to_id', chomp_resbody => 1 );
+}
+
+
+=head2 _fetch_doc
+
+Private function used for implementing of C<get_doc>, C<get_doc_by_uri>,
+C<etch_doc>, C<etch_doc_by_uri>.
+
+ # this will decode received draft into Search::Estraier::Document object
+ my $doc = $node->_fetch_doc( id => 42 );
+ my $doc = $node->_fetch_doc( uri => 'file:///document/uri/42' );
 
-	sub new { 
-		my $self = LWP::UserAgent::new(@_);
-		$self->agent("Search-Estraier/$Search::Estraer::VERSION");
-		$self;
+ # to extract keywords, add etch
+ my $doc = $node->_fetch_doc( id => 42, etch => 1 );
+ my $doc = $node->_fetch_doc( uri => 'file:///document/uri/42', etch => 1 );
+
+ # to get document attrubute add attr
+ my $doc = $node->_fetch_doc( id => 42, attr => '@mdate' );
+ my $doc = $node->_fetch_doc( uri => 'file:///document/uri/42', attr => '@mdate' );
+
+ # more general form which allows implementation of
+ # uri_to_id
+ my $id = $node->_fetch_doc(
+ 	uri => 'file:///document/uri/42',
+	path => '/uri_to_id',
+	chomp_resbody => 1
+ );
+
+=cut
+
+sub _fetch_doc {
+	my $self = shift;
+	my $a = {@_};
+	return unless ( ($a->{id} || $a->{uri}) && $self->{url} );
+
+	my ($arg, $resbody);
+
+	my $path = $a->{path} || '/get_doc';
+	$path = '/etch_doc' if ($a->{etch});
+
+	if ($a->{id}) {
+		croak "id must be numberm not '$a->{id}'" unless ($a->{id} =~ m/^\d+$/);
+		$arg = 'id=' . $a->{id};
+	} elsif ($a->{uri}) {
+		$arg = 'uri=' . $a->{uri};
+	} else {
+		confess "unhandled argument. Need id or uri.";
+	}
+
+	if ($a->{attr}) {
+		$path = '/get_doc_attr';
+		$arg .= '&attr=' . uri_escape($a->{attr});
+		$a->{chomp_resbody} = 1;
 	}
 
-	sub get_basic_credentials {
-		my($self, $realm, $uri) = @_;
-#		return ($user, $password);
+	my $rv = $self->shuttle_url( $self->{url} . $path,
+		'application/x-www-form-urlencoded',
+		$arg,
+		\$resbody,
+	);
+
+	return if ($rv != 200);
+
+	if ($a->{etch}) {
+		$self->{kwords} = {};
+		return +{} unless ($resbody);
+		foreach my $l (split(/\n/, $resbody)) {
+			my ($k,$v) = split(/\t/, $l, 2);
+			$self->{kwords}->{$k} = $v if ($v);
+		}
+		return $self->{kwords};
+	} elsif ($a->{chomp_resbody}) {
+		return unless (defined($resbody));
+		chomp($resbody);
+		return $resbody;
+	} else {
+		return new Search::Estraier::Document($resbody);
 	}
 }
 
 
+=head2 name
 
-=head2 new
+  my $node_name = $node->name;
 
-Create new connection to node master.
+=cut
 
-  my $master = new Search::Estraier::Master(
-	url => 'http://localhost:1978',
-	user => 'admin',
-	passwd => 'admin',
-  );
+sub name {
+	my $self = shift;
+	$self->set_info unless ($self->{name});
+	return $self->{name};
+}
+
+
+=head2 label
+
+  my $node_label = $node->label;
 
 =cut
 
-sub new {
-	my $class = shift;
-	my $self = {@_};
-	bless($self, $class);
+sub label {
+	my $self = shift;
+	$self->set_info unless ($self->{label});
+	return $self->{label};
+}
+
+
+=head2 doc_num
+
+  my $documents_in_node = $node->doc_num;
+
+=cut
+
+sub doc_num {
+	my $self = shift;
+	$self->set_info if ($self->{dnum} < 0);
+	return $self->{dnum};
+}
+
+
+=head2 word_num
+
+  my $words_in_node = $node->word_num;
+
+=cut
+
+sub word_num {
+	my $self = shift;
+	$self->set_info if ($self->{wnum} < 0);
+	return $self->{wnum};
+}
+
+
+=head2 size
+
+  my $node_size = $node->size;
+
+=cut
+
+sub size {
+	my $self = shift;
+	$self->set_info if ($self->{size} < 0);
+	return $self->{size};
+}
+
+
+
+=head2 shuttle_url
+
+This is method which uses C<IO::Socket::INET> to communicate with Hyper Estraier node
+master.
+
+  my $rv = shuttle_url( $url, $content_type, \$req_body, \$resbody );
+
+C<$resheads> and C<$resbody> booleans controll if response headers and/or response
+body will be saved within object.
+
+=cut
+
+sub shuttle_url {
+	my $self = shift;
+
+	my ($url, $content_type, $reqbody, $resbody) = @_;
 
-	foreach my $p (qw/url user passwd/) {
-		croak "need $p" unless ($self->{$p});
+	$self->{status} = -1;
+
+	warn "## $url\n" if ($self->{debug});
+
+	$url = new URI($url);
+	if (
+			!$url || !$url->scheme || !$url->scheme eq 'http' ||
+			!$url->host || !$url->port || $url->port < 1
+		) {
+		carp "can't parse $url\n";
+		return -1;
 	}
 
-	$self ? return $self : return undef;
+	my ($host,$port,$query) = ($url->host, $url->port, $url->path);
+
+	if ($self->{pxhost}) {
+		($host,$port) = ($self->{pxhost}, $self->{pxport});
+		$query = "http://$host:$port/$query";
+	}
+
+	$query .= '?' . $url->query if ($url->query && ! $reqbody);
+
+	my $headers;
+
+	if ($reqbody) {
+		$headers .= "POST $query HTTP/1.0\r\n";
+	} else {
+		$headers .= "GET $query HTTP/1.0\r\n";
+	}
+
+	$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;
+	}
+
+	warn $headers if ($self->{debug});
+
+	print $sock $headers or
+		carp "can't send headers to network:\n$headers\n" and return -1;
+
+	if ($reqbody) {
+		warn "$reqbody\n" if ($self->{debug});
+		print $sock $reqbody or
+			carp "can't send request body to network:\n$$reqbody\n" and return -1;
+	}
+
+	my $line = <$sock>;
+	chomp($line);
+	my ($schema, $res_status, undef) = split(/  */, $line, 3);
+	return if ($schema !~ /^HTTP/ || ! $res_status);
+
+	$self->{status} = $res_status;
+	warn "## response status: $res_status\n" if ($self->{debug});
+
+	# skip rest of headers
+	$line = <$sock>;
+	while ($line) {
+		$line = <$sock>;
+		$line =~ s/[\r\n]+$//;
+		warn "## ", $line || 'NULL', " ##\n" if ($self->{debug});
+	};
+
+	# read body
+	$len = 0;
+	do {
+		$len = read($sock, my $buf, 8192);
+		$$resbody .= $buf if ($resbody);
+	} while ($len);
+
+	warn "## response body:\n$$resbody\n" if ($resbody && $self->{debug});
+
+	return $self->{status};
 }
 
 
+=head2 set_info
+
+Set information for node
+
+  $node->set_info;
+
+=cut
+
+sub set_info {
+	my $self = shift;
+
+	$self->{status} = -1;
+	return unless ($self->{url});
+
+	my $resbody;
+	my $rv = $self->shuttle_url( $self->{url} . '/inform',
+		'text/plain',
+		undef,
+		\$resbody,
+	);
+
+	return if ($rv != 200 || !$resbody);
+
+	chomp($resbody);
+
+	( $self->{name}, $self->{label}, $self->{dnum}, $self->{wnum}, $self->{size} ) =
+		split(/\t/, $resbody, 5);
+
+}
 
 ###