1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # RecentDocumentsList.pm --
|
---|
4 | # A component of the Greenstone digital library software
|
---|
5 | # from the New Zealand Digital Library Project at the
|
---|
6 | # University of Waikato, New Zealand.
|
---|
7 | #
|
---|
8 | # Copyright (C) 1999 New Zealand Digital Library Project
|
---|
9 | #
|
---|
10 | # This program is free software; you can redistribute it and/or modify
|
---|
11 | # it under the terms of the GNU General Public License as published by
|
---|
12 | # the Free Software Foundation; either version 2 of the License, or
|
---|
13 | # (at your option) any later version.
|
---|
14 | #
|
---|
15 | # This program is distributed in the hope that it will be useful,
|
---|
16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
18 | # GNU General Public License for more details.
|
---|
19 | #
|
---|
20 | # You should have received a copy of the GNU General Public License
|
---|
21 | # along with this program; if not, write to the Free Software
|
---|
22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
23 | #
|
---|
24 | ###########################################################################
|
---|
25 |
|
---|
26 | # simple list classifier plugin
|
---|
27 | # to see the options, run "perl -S classinfo.pl RecentDocumentsList"
|
---|
28 |
|
---|
29 | use BasClas;
|
---|
30 | package RecentDocumentsList;
|
---|
31 |
|
---|
32 | use strict;
|
---|
33 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
34 | use sorttools;
|
---|
35 | use Time::Local;
|
---|
36 |
|
---|
37 | sub BEGIN {
|
---|
38 | @RecentDocumentsList::ISA = ('BasClas');
|
---|
39 | }
|
---|
40 |
|
---|
41 | my $arguments =
|
---|
42 | [ { 'name' => "include_docs_added_since",
|
---|
43 | 'desc' => "{RecentDocumentsList.include_docs_added_since}",
|
---|
44 | 'type' => "string",
|
---|
45 | 'reqd' => "no" },
|
---|
46 | { 'name' => "include_most_recently_added",
|
---|
47 | 'desc' => "{RecentDocumentsList.include_most_recently_added}",
|
---|
48 | 'type' => "int",
|
---|
49 | 'deft' => "20",
|
---|
50 | 'reqd' => "no"},
|
---|
51 | { 'name' => "sort",
|
---|
52 | 'desc' => "{RecentDocumentsList.sort}",
|
---|
53 | 'type' => "metadata",
|
---|
54 | 'reqd' => "no"}
|
---|
55 | ];
|
---|
56 |
|
---|
57 | my $options = { 'name' => "RecentDocumentsList",
|
---|
58 | 'desc' => "{RecentDocumentsList.desc}",
|
---|
59 | 'abstract' => "no",
|
---|
60 | 'inherits' => "yes",
|
---|
61 | 'args' => $arguments };
|
---|
62 |
|
---|
63 |
|
---|
64 | sub new {
|
---|
65 | my ($class) = shift (@_);
|
---|
66 |
|
---|
67 | my ($classifierslist,$inputargs,$hashArgOptLists) = @_;
|
---|
68 | push(@$classifierslist, $class);
|
---|
69 |
|
---|
70 | if(defined $arguments){ push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});}
|
---|
71 | if(defined $options) { push(@{$hashArgOptLists->{"OptList"}},$options)};
|
---|
72 |
|
---|
73 | my $self = new BasClas($classifierslist, $inputargs, $hashArgOptLists);
|
---|
74 |
|
---|
75 | if ($self->{'info_only'}) {
|
---|
76 | # don't worry about any options etc
|
---|
77 | return bless $self, $class;
|
---|
78 | }
|
---|
79 | # check the arguments
|
---|
80 |
|
---|
81 | if (!$self->{"buttonname"}) {
|
---|
82 | $self->{"buttonname"} = 'RecentDocuments';
|
---|
83 | }
|
---|
84 |
|
---|
85 | # we want either include_docs_added_since, or include_most_recently_added, but not both.
|
---|
86 | if (defined $self->{'include_docs_added_since'} && !($self->{'include_docs_added_since'} eq "")){
|
---|
87 | $self->{'classify_by_date'} = 1;
|
---|
88 | my ($year, $month, $day) = $self->{'include_docs_added_since'} =~
|
---|
89 | /^(\d\d\d\d)-?(\d\d)?-?(\d\d)?$/;
|
---|
90 | if (!defined $year) {
|
---|
91 | &gsprintf($self->{'outhandle'}, "RecentDocumentsList::init {RecentDocumentsList.date_wrong_format}\n");
|
---|
92 | die "\n";
|
---|
93 | }
|
---|
94 | if (!defined $month || $month < 1 || $month > 12) {
|
---|
95 | $month = "01";
|
---|
96 | $day = "01";
|
---|
97 | } elsif (!defined $day || $day < 1 || $day > 31) {
|
---|
98 | $day = "01";
|
---|
99 | }
|
---|
100 |
|
---|
101 | $self->{'classification_date'} = timelocal(0,0,0,$day,$month-1, $year);
|
---|
102 |
|
---|
103 | } else {
|
---|
104 | $self->{'classify_by_date'} = 0;
|
---|
105 | }
|
---|
106 | if ($self->{'sort'} eq "") {
|
---|
107 | undef $self->{'sort'};
|
---|
108 | }
|
---|
109 | # Further setup
|
---|
110 | $self->{'list'} = {};
|
---|
111 | # if we are getting top X docs, and sorting by meta, we need to store the
|
---|
112 | # date and the metadata
|
---|
113 | if (!$self->{'classify_by_date'} && $self->{'sort'}) {
|
---|
114 | $self->{'meta_list'} = {};
|
---|
115 | }
|
---|
116 | return bless $self, $class;
|
---|
117 | }
|
---|
118 |
|
---|
119 | sub init {
|
---|
120 | my $self = shift (@_);
|
---|
121 |
|
---|
122 | }
|
---|
123 |
|
---|
124 | sub classify {
|
---|
125 | my $self = shift (@_);
|
---|
126 | my ($doc_obj) = @_;
|
---|
127 |
|
---|
128 | my $doc_OID = $doc_obj->get_OID();
|
---|
129 | my $lastmodified = $doc_obj->get_metadata_element($doc_obj->get_top_section(), "lastmodified");
|
---|
130 | if (!defined $lastmodified || $lastmodified eq "") {
|
---|
131 | print $self->{'outhandle'}, "RecentDocumentsList: $doc_OID has no lastmodified metadata, not classifying\n";
|
---|
132 | return;
|
---|
133 | }
|
---|
134 |
|
---|
135 | # doc goes into classification if we are not classifying by date, or the date is after the cutoff date.
|
---|
136 | if ($self->{'classify_by_date'}) {
|
---|
137 | if ($lastmodified > $self->{'classification_date'}) {
|
---|
138 | my $sort_meta = $lastmodified;
|
---|
139 | if (defined $self->{'sort'}) {
|
---|
140 | $sort_meta = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $self->{'sort'});
|
---|
141 | }
|
---|
142 | $self->{'list'}->{$doc_OID} = $sort_meta;
|
---|
143 | $doc_obj->add_metadata($doc_obj->get_top_section(), "memberof", "CL".$self->get_number());
|
---|
144 | }
|
---|
145 | } else {
|
---|
146 |
|
---|
147 | # need to store metadata as well...
|
---|
148 | $self->{'list'}->{$doc_OID} = $lastmodified;
|
---|
149 | if (defined $self->{'sort'}) {
|
---|
150 | my $sort_meta = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $self->{'sort'});
|
---|
151 | $self->{'meta_list'}->{$doc_OID} = $sort_meta;
|
---|
152 | }
|
---|
153 | }
|
---|
154 |
|
---|
155 | }
|
---|
156 |
|
---|
157 |
|
---|
158 | sub get_classify_info {
|
---|
159 | my $self = shift (@_);
|
---|
160 | my $return_doc_size=0;
|
---|
161 |
|
---|
162 | my $list = $self->{'list'};
|
---|
163 |
|
---|
164 |
|
---|
165 | # organise into classification structure
|
---|
166 | my %classifyinfo = ('thistype'=>'Invisible',
|
---|
167 | 'childtype'=>'VList',
|
---|
168 | 'Title'=>$self->{'buttonname'},
|
---|
169 | 'contains'=>[]);
|
---|
170 |
|
---|
171 |
|
---|
172 | # may or may not support memberof, depending on options set
|
---|
173 | $classifyinfo{'supportsmemberof'} = &supports_memberof();
|
---|
174 |
|
---|
175 | # get either all documents (sorted by date), or the top X docs
|
---|
176 | my @sorted_docs = sort {$self->date_or_metadata_sort($a,$b)} keys %{$self->{'list'}};
|
---|
177 | my $numdocs = $self->{'include_most_recently_added'};
|
---|
178 | if ($self->{'classify_by_date'}) {
|
---|
179 | # just include all docs in the list
|
---|
180 | $numdocs = scalar (@sorted_docs);
|
---|
181 | } else {
|
---|
182 | if ($numdocs > scalar (@sorted_docs)) {
|
---|
183 | $numdocs = scalar (@sorted_docs);
|
---|
184 | }
|
---|
185 | if ($self->{'sort'}) {
|
---|
186 | # we need to sort further by metadata
|
---|
187 | # cut off the list
|
---|
188 | @sorted_docs = @sorted_docs[0..$numdocs-1];
|
---|
189 | # sort again
|
---|
190 | @sorted_docs = sort {$self->external_meta_sort($a,$b)}@sorted_docs;
|
---|
191 | }
|
---|
192 | }
|
---|
193 | for (my $i=0; $i<$numdocs; $i++) {
|
---|
194 | push (@{$classifyinfo{'contains'}}, {'OID'=> $sorted_docs[$i]});
|
---|
195 | }
|
---|
196 |
|
---|
197 |
|
---|
198 | return \%classifyinfo;
|
---|
199 | }
|
---|
200 |
|
---|
201 | # we can only support memberof if we have the include_docs_added_since option, otherwise we don't know at the time of classification of a document if it will be in the classifier or not.
|
---|
202 | sub supports_memberof {
|
---|
203 | my $self = shift(@_);
|
---|
204 |
|
---|
205 | if ($self->{'classify_by_date'}) {
|
---|
206 | return "true";
|
---|
207 | }
|
---|
208 | return "false";
|
---|
209 | }
|
---|
210 |
|
---|
211 | sub date_or_metadata_sort {
|
---|
212 | my ($self,$a,$b) = @_;
|
---|
213 | # make it do metadata too
|
---|
214 | my $date_a = $self->{'list'}->{$a};
|
---|
215 | my $date_b = $self->{'list'}->{$b};
|
---|
216 | if (!$self->{'sort'} || !$self->{'classify_by_date'}) {
|
---|
217 | # want reverse order (latest to earliest)
|
---|
218 | return ($date_b <=> $date_a);
|
---|
219 | }
|
---|
220 | # meta sorting, use string cmp
|
---|
221 | return ($date_a cmp $date_b);
|
---|
222 | }
|
---|
223 |
|
---|
224 | sub external_meta_sort {
|
---|
225 | my ($self,$a,$b) = @_;
|
---|
226 |
|
---|
227 | my $meta_a = $self->{'meta_list'}->{$a};
|
---|
228 | my $meta_b = $self->{'meta_list'}->{$b};
|
---|
229 |
|
---|
230 | return ($meta_a cmp $meta_b);
|
---|
231 | }
|
---|
232 |
|
---|
233 |
|
---|
234 | 1;
|
---|
235 |
|
---|
236 |
|
---|
237 |
|
---|
238 |
|
---|