| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # IncrementalDocument.pm -- An object to encapsulate the Greenstone
|
|---|
| 4 | # document retrieved from the info database.
|
|---|
| 5 | #
|
|---|
| 6 | # A component of the Greenstone digital library software
|
|---|
| 7 | # from the New Zealand Digital Library Project at the
|
|---|
| 8 | # University of Waikato, New Zealand.
|
|---|
| 9 | #
|
|---|
| 10 | # Copyright (C) 2006 DL Consulting Ltd and New Zealand Digital Library Project
|
|---|
| 11 | #
|
|---|
| 12 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 13 | # it under the terms of the GNU General Public License as published by
|
|---|
| 14 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 15 | # (at your option) any later version.
|
|---|
| 16 | #
|
|---|
| 17 | # This program is distributed in the hope that it will be useful,
|
|---|
| 18 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 20 | # GNU General Public License for more details.
|
|---|
| 21 | #
|
|---|
| 22 | # You should have received a copy of the GNU General Public License
|
|---|
| 23 | # along with this program; if not, write to the Free Software
|
|---|
| 24 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 25 | #
|
|---|
| 26 | ###########################################################################
|
|---|
| 27 | package IncrementalDocument;
|
|---|
| 28 |
|
|---|
| 29 | use strict;
|
|---|
| 30 |
|
|---|
| 31 |
|
|---|
| 32 | # /**
|
|---|
| 33 | # */
|
|---|
| 34 | sub new()
|
|---|
| 35 | {
|
|---|
| 36 | my ($class, $collection, $oid) = @_;
|
|---|
| 37 |
|
|---|
| 38 | #rint STDERR "IncrementalDocument::new($collection, $oid)\n";
|
|---|
| 39 |
|
|---|
| 40 | # Test the parameters
|
|---|
| 41 | die ("Error! Can't create a document that doesn't belong to a collection!") unless $collection;
|
|---|
| 42 | die ("Error! Can't create a document that doesn't have a unique id (OID)!") unless $oid;
|
|---|
| 43 |
|
|---|
| 44 | # Store the variables
|
|---|
| 45 | my $self = {};
|
|---|
| 46 |
|
|---|
| 47 | # The collection this document object has been loaded from.
|
|---|
| 48 | $self->{'collection'} = $collection;
|
|---|
| 49 |
|
|---|
| 50 | # An associative array of information retrieved from the info database
|
|---|
| 51 | # which maps a key string to a nested associative array listing values.
|
|---|
| 52 | $self->{'data'} = {};
|
|---|
| 53 |
|
|---|
| 54 | # The unique identifier of the document loaded
|
|---|
| 55 | $self->{'oid'} = $oid;
|
|---|
| 56 |
|
|---|
| 57 | # Stores the order in which metadata keys where discovered/added.
|
|---|
| 58 | $self->{'order'} = {};
|
|---|
| 59 |
|
|---|
| 60 | bless $self, $class;
|
|---|
| 61 | return $self;
|
|---|
| 62 | }
|
|---|
| 63 | # /** new() **/
|
|---|
| 64 |
|
|---|
| 65 | # /**
|
|---|
| 66 | # */
|
|---|
| 67 | sub addMetadata()
|
|---|
| 68 | {
|
|---|
| 69 | my ($self, $key, $value, $internal) = @_;
|
|---|
| 70 |
|
|---|
| 71 | # Validate the arguments
|
|---|
| 72 | die ("Error! Can't add a metadata value to a document without a valid key!") unless $key =~ /[\w]+/;
|
|---|
| 73 | die ("Error! Can't add a metadata key to a document without a valid value!") unless $value =~ /[\w\d]+/;
|
|---|
| 74 |
|
|---|
| 75 | # Is this a new key that we haven't encountered before? If so ensure an
|
|---|
| 76 | # array exists for its values, and record the order in which we encountered
|
|---|
| 77 | # this key.
|
|---|
| 78 | if (!defined($self->{'data'}->{$key}))
|
|---|
| 79 | {
|
|---|
| 80 | # Determine how many data keys we're already storing, so we can add the next
|
|---|
| 81 | # one at the appropriate index
|
|---|
| 82 | my $index = scalar(keys %{$self->{'order'}});
|
|---|
| 83 | $self->{'order'}->{$index} = $key;
|
|---|
| 84 | $self->{'data'}->{$key} = {};
|
|---|
| 85 | }
|
|---|
| 86 |
|
|---|
| 87 | # Set the value of the associative path to 1.
|
|---|
| 88 | $self->{'data'}->{$key}->{$value} = 1;
|
|---|
| 89 | }
|
|---|
| 90 | # /** addMetadata() **/
|
|---|
| 91 |
|
|---|
| 92 | # /** Retrieve all the metadata of this document as an array of pairs.
|
|---|
| 93 | # *
|
|---|
| 94 | # */
|
|---|
| 95 | sub getAllMetadata()
|
|---|
| 96 | {
|
|---|
| 97 | my ($self) = @_;
|
|---|
| 98 | my @all_metadata;
|
|---|
| 99 |
|
|---|
| 100 | print STDERR "IncrementalDocument.getAllMetadata()\n";
|
|---|
| 101 |
|
|---|
| 102 | my $key_count = scalar(keys %{$self->{'order'}});
|
|---|
| 103 | for (my $i = 0; $i < $key_count; $i++)
|
|---|
| 104 | {
|
|---|
| 105 | my $key = $self->{'order'}->{$i};
|
|---|
| 106 | # Check if this key has been set
|
|---|
| 107 | if ($self->{'data'}->{$key})
|
|---|
| 108 | {
|
|---|
| 109 | # Note: there may be zero values left
|
|---|
| 110 | foreach my $value (sort keys %{$self->{'data'}->{$key}})
|
|---|
| 111 | {
|
|---|
| 112 | if ($self->{'data'}->{$key}->{$value})
|
|---|
| 113 | {
|
|---|
| 114 | print STDERR "* Storing $key => $value\n";
|
|---|
| 115 | push(@all_metadata, [$key, $value]);
|
|---|
| 116 | }
|
|---|
| 117 | }
|
|---|
| 118 | }
|
|---|
| 119 | }
|
|---|
| 120 | print STDERR "Complete!\n";
|
|---|
| 121 | return \@all_metadata;
|
|---|
| 122 | }
|
|---|
| 123 | # /** getAllMetadata() **/
|
|---|
| 124 |
|
|---|
| 125 | # /**
|
|---|
| 126 | # */
|
|---|
| 127 | sub getDocNum()
|
|---|
| 128 | {
|
|---|
| 129 | my ($self) = @_;
|
|---|
| 130 | my $docnum = -1;
|
|---|
| 131 | # Check the docnum path exists in the associated data
|
|---|
| 132 | if(defined($self->{'data'}->{'docnum'}))
|
|---|
| 133 | {
|
|---|
| 134 | # Get the list of keys from that associative path
|
|---|
| 135 | my @values = keys (%{$self->{'data'}->{'docnum'}});
|
|---|
| 136 | # And since we know there will only ever be one value for docnum
|
|---|
| 137 | $docnum = $values[0];
|
|---|
| 138 | }
|
|---|
| 139 | return $docnum;
|
|---|
| 140 | }
|
|---|
| 141 | # /** getDocNum() **/
|
|---|
| 142 |
|
|---|
| 143 | # /**
|
|---|
| 144 | # */
|
|---|
| 145 | sub loadDocument()
|
|---|
| 146 | {
|
|---|
| 147 | my ($self) = @_;
|
|---|
| 148 | #rint STDERR "IncrementalDocument::loadDocument()\n";
|
|---|
| 149 | # Load the raw text for the document object from the info database
|
|---|
| 150 | my $collection = $self->{'collection'};
|
|---|
| 151 | my $index_text_directory_path = &util::filename_cat($ENV{'GSDLHOME'}, "collect", $collection, "index", "text");
|
|---|
| 152 | my $infodb_file_path = &dbutil::get_infodb_file_path("gdbm", $collection, $index_text_directory_path);
|
|---|
| 153 | my $text = &dbutil::read_infodb_entry("gdbm", $infodb_file_path, $self->{'oid'});
|
|---|
| 154 | # For each line in the raw text, extract the key (enclosed in angle
|
|---|
| 155 | # brackets) and the value
|
|---|
| 156 | $text =~ s/<([\w\d\.]+)>(.+?)\r?\n/&addMetadata($self, $1, $2, 1)/egs;
|
|---|
| 157 | # Done
|
|---|
| 158 | }
|
|---|
| 159 | # /** loadDocument() **/
|
|---|
| 160 |
|
|---|
| 161 | # /** Locates and removes the given key/value mappings from this document
|
|---|
| 162 | # * object.
|
|---|
| 163 | # *
|
|---|
| 164 | # * @param $self A reference to this IncrementalDocument object
|
|---|
| 165 | # * @param $key The metadata key as a string
|
|---|
| 166 | # * @param $value The obsolete metadata value as a string
|
|---|
| 167 | # *
|
|---|
| 168 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 169 | # */
|
|---|
| 170 | sub removeMetadata()
|
|---|
| 171 | {
|
|---|
| 172 | my ($self, $key, $value) = @_;
|
|---|
| 173 | # Ensure the value doesn't exist by simply setting to 0 the correct
|
|---|
| 174 | # associative path
|
|---|
| 175 | $self->{'data'}->{$key}->{$value} = 0;
|
|---|
| 176 | }
|
|---|
| 177 | # /*** removeMetadat() **/
|
|---|
| 178 |
|
|---|
| 179 | # /**
|
|---|
| 180 | # */
|
|---|
| 181 | sub saveDocument()
|
|---|
| 182 | {
|
|---|
| 183 | my ($self) = @_;
|
|---|
| 184 | # Get a textual version of this object
|
|---|
| 185 | my $text = $self->toString();
|
|---|
| 186 |
|
|---|
| 187 | # Now store the object in the info database
|
|---|
| 188 | my $collection = $self->{'collection'};
|
|---|
| 189 |
|
|---|
| 190 | my $index_text_directory_path = &util::filename_cat($ENV{'GSDLHOME'}, "collect", $collection, "index", "text");
|
|---|
| 191 | my $infodb_file_path = &dbutil::get_infodb_file_path("gdbm", $collection, $index_text_directory_path);
|
|---|
| 192 | my $infodb_file_handle = &dbutil::open_infodb_write_handle("gdbm", $infodb_file_path, "append");
|
|---|
| 193 | &dbutil::write_infodb_entry("gdbm", $infodb_file_handle, $self->{'oid'}, &dbutil::convert_infodb_string_to_hash($text));
|
|---|
| 194 | &dbutil::close_infodb_write_handle("gdbm", $infodb_file_handle);
|
|---|
| 195 |
|
|---|
| 196 | # There is a little bit of extra complexity when saving an incremental
|
|---|
| 197 | # document in that we should ensure that a reverse lookup-from DocNum or
|
|---|
| 198 | # nodeID to Greenstone document hash-exists in the database.
|
|---|
| 199 | my $doc_num = $self->getDocNum();
|
|---|
| 200 | if($doc_num >= 0)
|
|---|
| 201 | {
|
|---|
| 202 | my $text = &dbutil::read_infodb_entry("gdbm", $infodb_file_path, $doc_num);
|
|---|
| 203 |
|
|---|
| 204 | # If there is no reverse lookup, then add one now
|
|---|
| 205 | if($text !~ /<section>/)
|
|---|
| 206 | {
|
|---|
| 207 | $infodb_file_handle = &dbutil::open_infodb_write_handle("gdbm", $infodb_file_path, "append");
|
|---|
| 208 | &dbutil::write_infodb_entry("gdbm", $infodb_file_handle, $doc_num, &dbutil::convert_infodb_string_to_hash("<section>" . $self->{'oid'}));
|
|---|
| 209 | &dbutil::close_infodb_write_handle("gdbm", $infodb_file_handle);
|
|---|
| 210 | }
|
|---|
| 211 | }
|
|---|
| 212 | # Done
|
|---|
| 213 | #rint STDERR "Stored document:\n[" . $self->{'oid'} . "]\n$text\n";
|
|---|
| 214 | }
|
|---|
| 215 | # /** saveDocument() **/
|
|---|
| 216 |
|
|---|
| 217 | # /** Produces a textual representation of this object.
|
|---|
| 218 | # *
|
|---|
| 219 | # * @return A string which describes this incremental document object
|
|---|
| 220 | # *
|
|---|
| 221 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 222 | # */
|
|---|
| 223 | sub toString()
|
|---|
| 224 | {
|
|---|
| 225 | my ($self) = @_;
|
|---|
| 226 | my $text = "";
|
|---|
| 227 |
|
|---|
| 228 | my $key_count = scalar(keys %{$self->{'order'}});
|
|---|
| 229 | for (my $i = 0; $i < $key_count; $i++)
|
|---|
| 230 | {
|
|---|
| 231 | my $key = $self->{'order'}->{$i};
|
|---|
| 232 | # Check if this key has been set
|
|---|
| 233 | if ($self->{'data'}->{$key})
|
|---|
| 234 | {
|
|---|
| 235 | # Note: there may be zero values left
|
|---|
| 236 | foreach my $value (sort keys %{$self->{'data'}->{$key}})
|
|---|
| 237 | {
|
|---|
| 238 | if ($self->{'data'}->{$key}->{$value})
|
|---|
| 239 | {
|
|---|
| 240 | $text .= "<$key>$value\n";
|
|---|
| 241 | }
|
|---|
| 242 | }
|
|---|
| 243 | }
|
|---|
| 244 | }
|
|---|
| 245 | return $text;
|
|---|
| 246 | }
|
|---|
| 247 | # /** toString() **/
|
|---|
| 248 |
|
|---|
| 249 | 1;
|
|---|
