| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # ClassifyTreeModel.pm --
|
|---|
| 4 | #
|
|---|
| 5 | # A component of the Greenstone digital library software
|
|---|
| 6 | # from the New Zealand Digital Library Project at the
|
|---|
| 7 | # University of Waikato, New Zealand.
|
|---|
| 8 | #
|
|---|
| 9 | # Copyright (C) 2006-2010 DL Consulting Ltd
|
|---|
| 10 | #
|
|---|
| 11 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 12 | # it under the terms of the GNU General Public License as published by
|
|---|
| 13 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 14 | # (at your option) any later version.
|
|---|
| 15 | #
|
|---|
| 16 | # This program is distributed in the hope that it will be useful,
|
|---|
| 17 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 18 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 19 | # GNU General Public License for more details.
|
|---|
| 20 | #
|
|---|
| 21 | # You should have received a copy of the GNU General Public License
|
|---|
| 22 | # along with this program; if not, write to the Free Software
|
|---|
| 23 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 24 | #
|
|---|
| 25 | ###########################################################################
|
|---|
| 26 |
|
|---|
| 27 |
|
|---|
| 28 | package ClassifyTreeModel;
|
|---|
| 29 |
|
|---|
| 30 | use ClassifyTreeNode;
|
|---|
| 31 | use strict;
|
|---|
| 32 |
|
|---|
| 33 |
|
|---|
| 34 | # /** Constructor
|
|---|
| 35 | # *
|
|---|
| 36 | # * @param $class The name of the class to bless as a string
|
|---|
| 37 | # * @param $collection The name of the collection whose info database we
|
|---|
| 38 | # * will be accessing as a string
|
|---|
| 39 | # * @param $root The oid of the root node of the classifier as a
|
|---|
| 40 | # * string
|
|---|
| 41 | # * @return A reference to the ClassifyTreeModel object
|
|---|
| 42 | # *
|
|---|
| 43 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 44 | # */
|
|---|
| 45 | sub new()
|
|---|
| 46 | {
|
|---|
| 47 | my ($class, $collection, $infodbtype, $root) = @_;
|
|---|
| 48 | my $debug = 0;
|
|---|
| 49 | print STDERR "ClassifyTreeModel.new(\"$collection\", $infodbtype, \"$root\")\n" unless !$debug;
|
|---|
| 50 | # Store the variables
|
|---|
| 51 | my $self = {};
|
|---|
| 52 | $self->{'collection'} = $collection;
|
|---|
| 53 | $self->{'infodbtype'} = $infodbtype;
|
|---|
| 54 | $self->{'debug'} = $debug;
|
|---|
| 55 | $self->{'root'} = $root;
|
|---|
| 56 | # Bless me father for I have sinned
|
|---|
| 57 | bless $self, $class;
|
|---|
| 58 | return $self;
|
|---|
| 59 | }
|
|---|
| 60 | # /** new() **/
|
|---|
| 61 |
|
|---|
| 62 | # /** Given a path and a document id, add this document to the classifier tree
|
|---|
| 63 | # * creating any necessary tree nodes first.
|
|---|
| 64 | # *
|
|---|
| 65 | # * @param $value The path to store this document in
|
|---|
| 66 | # * @param $oid Unique identifier of a document
|
|---|
| 67 | # *
|
|---|
| 68 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 69 | # */
|
|---|
| 70 | sub addDocument()
|
|---|
| 71 | {
|
|---|
| 72 | my ($self, $value, $oid) = @_;
|
|---|
| 73 | print STDERR "ClassifyTreeModel.addDocument(\"$value\", \"$oid\")\n" unless !$self->{'debug'};
|
|---|
| 74 | # Generate a treepath object from the metadata value, remembering to prefix
|
|---|
| 75 | # with the root nodes path
|
|---|
| 76 | my $root_node_obj = $self->getRootNode();
|
|---|
| 77 | my $path_obj = $root_node_obj->getPath();
|
|---|
| 78 | $path_obj->addPathComponent($value);
|
|---|
| 79 | # Ensure that this classifier node, and if necessary its ancestor nodes,
|
|---|
| 80 | # exist in our tree.
|
|---|
| 81 | my $node_obj = $self->getNodeByPath($path_obj);
|
|---|
| 82 | if (!$node_obj)
|
|---|
| 83 | {
|
|---|
| 84 | # The node doesn't exist, so we need to add it
|
|---|
| 85 | $node_obj = $self->addNode($path_obj);
|
|---|
| 86 | }
|
|---|
| 87 | # Add the document to the node.
|
|---|
| 88 | $node_obj->addDocument($oid);
|
|---|
| 89 | # Done.
|
|---|
| 90 | }
|
|---|
| 91 | # /** addDocument() **/
|
|---|
| 92 |
|
|---|
| 93 | # /** Add a node into the tree first ensuring all its parent nodes are inserted
|
|---|
| 94 | # * to.
|
|---|
| 95 | # *
|
|---|
| 96 | # * @param $path_obj The path to insert the new node at
|
|---|
| 97 | # *
|
|---|
| 98 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 99 | # */
|
|---|
| 100 | sub addNode()
|
|---|
| 101 | {
|
|---|
| 102 | my ($self, $path_obj) = @_;
|
|---|
| 103 | print STDERR "ClassifyTreeModel.addNode(\"" . $path_obj->toString() . "\")\n" unless !$self->{'debug'};
|
|---|
| 104 | # Ensure the parent exists, assuming we aren't at the root
|
|---|
| 105 | my $parent_path_obj = $path_obj->getParentPath();
|
|---|
| 106 | #rint STDERR "* parent path: " . $parent_path_obj->toString() . "\n";
|
|---|
| 107 | my $parent_node_obj = $self->getNodeByPath($parent_path_obj);
|
|---|
| 108 | #rint STDERR "* does parent node already exist? " . $parent_node_obj . "\n";
|
|---|
| 109 | #rint STDERR "* are we at the root node yet? " . $parent_path_obj->isRootPath() . "\n";
|
|---|
| 110 | if (!$parent_node_obj && !$parent_path_obj->isRootPath())
|
|---|
| 111 | {
|
|---|
| 112 | #rint STDERR "* recursive call!\n";
|
|---|
| 113 | $parent_node_obj = $self->addNode($parent_path_obj);
|
|---|
| 114 | }
|
|---|
| 115 | # Insert this node into it's parent.
|
|---|
| 116 | return $parent_node_obj->insertNode($path_obj);
|
|---|
| 117 | }
|
|---|
| 118 | # /** addNode() **/
|
|---|
| 119 |
|
|---|
| 120 | # /** Retrieve the name of the collection this model is drawing from.
|
|---|
| 121 | # *
|
|---|
| 122 | # * @return The collection name as a string
|
|---|
| 123 | # *
|
|---|
| 124 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 125 | # */
|
|---|
| 126 | sub getCollection()
|
|---|
| 127 | {
|
|---|
| 128 | my ($self) = @_;
|
|---|
| 129 | print STDERR "ClassifyTreeModel.getCollection()\n" unless !$self->{'debug'};
|
|---|
| 130 | return $self->{'collection'};
|
|---|
| 131 | }
|
|---|
| 132 | # /** getCollection() **/
|
|---|
| 133 |
|
|---|
| 134 | sub getInfoDBType()
|
|---|
| 135 | {
|
|---|
| 136 | my $self = shift(@_);
|
|---|
| 137 | return $self->{'infodbtype'};
|
|---|
| 138 | }
|
|---|
| 139 |
|
|---|
| 140 | # /** Retrieve a node from this tree based upon its CLID (OID).
|
|---|
| 141 | # * @param $clid The CLID as a string
|
|---|
| 142 | # * @return The indicated ClassifyTreeNode or null
|
|---|
| 143 | sub getNodeByCLID()
|
|---|
| 144 | {
|
|---|
| 145 | my ($self, $clid) = @_;
|
|---|
| 146 | print STDERR "ClassifyTreeModel.getNodeByCLID(\"$clid\")\n" unless !$self->{'debug'};
|
|---|
| 147 | my $result = 0;
|
|---|
| 148 | # Test if this clid is even in our tree
|
|---|
| 149 | if($clid !~ /^$self->{'root'}/)
|
|---|
| 150 | {
|
|---|
| 151 | print STDERR "Requested node $clid, which isn't part of " . $self->{'root'} . "\n";
|
|---|
| 152 | return 0;
|
|---|
| 153 | }
|
|---|
| 154 | # Unfortunately I have to check that there is text to retrieve before I
|
|---|
| 155 | # create a new node.
|
|---|
| 156 |
|
|---|
| 157 | my $index_text_directory_path = &util::filename_cat($ENV{'GSDLHOME'}, "collect", $self->getCollection(), "index", "text");
|
|---|
| 158 | my $infodb_file_path = &dbutil::get_infodb_file_path($self->{'infodbtype'}, $self->getCollection(), $index_text_directory_path);
|
|---|
| 159 | if (&dbutil::read_infodb_rawentry($self->{'infodbtype'}, $infodb_file_path, $clid) =~ /\w+/)
|
|---|
| 160 | {
|
|---|
| 161 | # Since the CLID can directly reference the correct entry in the info database we
|
|---|
| 162 | # just create the node and return it
|
|---|
| 163 | $result = new ClassifyTreeNode($self, $clid);
|
|---|
| 164 | }
|
|---|
| 165 | return $result;
|
|---|
| 166 | }
|
|---|
| 167 |
|
|---|
| 168 | # /** Retrieve a node from this tree based upon a path
|
|---|
| 169 | # *
|
|---|
| 170 | # * @param $path The path to the node as a ClassifyTreePath
|
|---|
| 171 | # * @return The indicated ClassifyTreeNode or null
|
|---|
| 172 | # *
|
|---|
| 173 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 174 | # */
|
|---|
| 175 | sub getNodeByPath()
|
|---|
| 176 | {
|
|---|
| 177 | my ($self, $path_obj) = @_;
|
|---|
| 178 | print STDERR "ClassifyTreeModel.getNodeByPath(\"" . $path_obj->toString() . "\")\n" unless !$self->{'debug'};
|
|---|
| 179 | # Starting at the ROOT of the tree, and with the first path component,
|
|---|
| 180 | # recursively descend through the tree looking for the node - we can assume
|
|---|
| 181 | # that we've found the root node (otherwise we won't be in a tree)
|
|---|
| 182 | my $cur_node_obj = $self->getRootNode();
|
|---|
| 183 | my $cur_path_obj = $cur_node_obj->getPath();
|
|---|
| 184 | my $depth = 1;
|
|---|
| 185 | # Continue till we either find the node we want, or run out a nodes
|
|---|
| 186 | while(!$cur_node_obj->getPath()->equals($path_obj))
|
|---|
| 187 | {
|
|---|
| 188 | # Append the path component at this depth to the current path we
|
|---|
| 189 | # are searching for
|
|---|
| 190 | $cur_path_obj->addPathComponent($path_obj->getPathComponent($depth));
|
|---|
| 191 | $depth++;
|
|---|
| 192 | #rint STDERR "Searching " . $cur_node_obj->getPath()->toString() . "'s children for: " . $cur_path_obj->toString() . "\n";
|
|---|
| 193 | # Search through the current nodes children, looking for one that
|
|---|
| 194 | # matches the current path
|
|---|
| 195 | my $found = 0;
|
|---|
| 196 | foreach my $child_node_obj ($cur_node_obj->getChildren())
|
|---|
| 197 | {
|
|---|
| 198 | #rint STDERR "* testing " . $child_node_obj->getPath()->toString() . "\n";
|
|---|
| 199 | if($child_node_obj->getPath()->equals($cur_path_obj))
|
|---|
| 200 | {
|
|---|
| 201 | $cur_node_obj = $child_node_obj;
|
|---|
| 202 | $found = 1;
|
|---|
| 203 | last;
|
|---|
| 204 | }
|
|---|
| 205 | }
|
|---|
| 206 | # Couldn't find any node with this path
|
|---|
| 207 | if(!$found)
|
|---|
| 208 | {
|
|---|
| 209 | #rint STDERR "* no such node exists!\n";
|
|---|
| 210 | return 0;
|
|---|
| 211 | }
|
|---|
| 212 | }
|
|---|
| 213 | return $cur_node_obj;
|
|---|
| 214 | }
|
|---|
| 215 | # /** getChild() **/
|
|---|
| 216 |
|
|---|
| 217 | # /** Retrieve the parent node of the given node.
|
|---|
| 218 | # *
|
|---|
| 219 | # * @param $child_node The node whose parent we want to retrieve
|
|---|
| 220 | # * @return The parent node, or 0 if this is the root
|
|---|
| 221 | # *
|
|---|
| 222 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 223 | # */
|
|---|
| 224 | sub getParentNode()
|
|---|
| 225 | {
|
|---|
| 226 | my ($self, $child_node) = @_;
|
|---|
| 227 | print STDERR "ClassifyTreeModel.getParentNode()\n" unless !$self->{'debug'};
|
|---|
| 228 | return $child_node->getParentNode();
|
|---|
| 229 | }
|
|---|
| 230 | # /** getParentNode() **/
|
|---|
| 231 |
|
|---|
| 232 | sub getRootNode()
|
|---|
| 233 | {
|
|---|
| 234 | my ($self) = @_;
|
|---|
| 235 | print STDERR "ClassifyTreeModel.getRootNode()\n" unless !$self->{'debug'};
|
|---|
| 236 | return new ClassifyTreeNode($self, $self->{'root'});
|
|---|
| 237 | }
|
|---|
| 238 |
|
|---|
| 239 | # /** Remove the given document from the classifier tree, and then remove any
|
|---|
| 240 | # * empty nodes if required.
|
|---|
| 241 | # *
|
|---|
| 242 | # * @param $value The value which contains the path of the node to remove
|
|---|
| 243 | # * the document from
|
|---|
| 244 | # * @param $oid The unique identifier of the document to remove
|
|---|
| 245 | # * @param $remove_empty Sets whether empty nodes are removed
|
|---|
| 246 | # *
|
|---|
| 247 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 248 | # */
|
|---|
| 249 | sub removeDocument()
|
|---|
| 250 | {
|
|---|
| 251 | my ($self, $path, $oid, $remove_empty) = @_;
|
|---|
| 252 | print STDERR "ClassifyTreeModel.removeDocument(\"$path\",\"$oid\",$remove_empty)\n" unless !$self->{'debug'};
|
|---|
| 253 | # Append to root path
|
|---|
| 254 | my $root_node_obj = $self->getRootNode();
|
|---|
| 255 | my $path_obj = $root_node_obj->getPath();
|
|---|
| 256 | $path_obj->addPathComponent($path);
|
|---|
| 257 | # Retrieve the node in question
|
|---|
| 258 | my $node_obj = $self->getNodeByPath($path_obj);
|
|---|
| 259 | # Check we retrieved a node
|
|---|
| 260 | if ($node_obj)
|
|---|
| 261 | {
|
|---|
| 262 | # Remove the document
|
|---|
| 263 | $node_obj->removeDocument($oid);
|
|---|
| 264 | # If we have been asked to remove empty nodes, do so now.
|
|---|
| 265 | if ($remove_empty)
|
|---|
| 266 | {
|
|---|
| 267 | my $cur_node_obj = $node_obj;
|
|---|
| 268 | my $empty_node_obj = 0;
|
|---|
| 269 | while ($cur_node_obj->getNumLeafDocs() == 0)
|
|---|
| 270 | {
|
|---|
| 271 | $empty_node_obj = $cur_node_obj;
|
|---|
| 272 | $cur_node_obj = $cur_node_obj->getParentNode();
|
|---|
| 273 | }
|
|---|
| 274 | if ($empty_node_obj)
|
|---|
| 275 | {
|
|---|
| 276 | # Try to retrieve the parent of this node
|
|---|
| 277 | my $parent_node_obj = $empty_node_obj->getParentNode();
|
|---|
| 278 | # As long as we have a parent (i.e. we aren't the root node) go
|
|---|
| 279 | # ahead and delete this subtree starting at empty node
|
|---|
| 280 | if ($parent_node_obj)
|
|---|
| 281 | {
|
|---|
| 282 | $parent_node_obj->removeNode($empty_node_obj);
|
|---|
| 283 | }
|
|---|
| 284 | }
|
|---|
| 285 | }
|
|---|
| 286 | }
|
|---|
| 287 | # If the node doesn't exist in this tree, then we can't very well remove
|
|---|
| 288 | # anything from it!
|
|---|
| 289 | }
|
|---|
| 290 | # /** removeDocument() **/
|
|---|
| 291 |
|
|---|
| 292 | 1;
|
|---|
