[13900] | 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 | package RecentDocumentsList;
|
---|
| 30 |
|
---|
[17209] | 31 | use BaseClassifier;
|
---|
[13900] | 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 {
|
---|
[17209] | 38 | @RecentDocumentsList::ISA = ('BaseClassifier');
|
---|
[13900] | 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, $class);
|
---|
| 69 |
|
---|
[17209] | 70 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
---|
| 71 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
---|
[13900] | 72 |
|
---|
[17209] | 73 | my $self = new BaseClassifier($classifierslist, $inputargs, $hashArgOptLists);
|
---|
[13900] | 74 |
|
---|
| 75 | if ($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 | }
|
---|
[20454] | 109 | $self->{'sort'} = $self->strip_ex_from_metadata($self->{'sort'});
|
---|
| 110 |
|
---|
[13900] | 111 | # Further setup
|
---|
| 112 | $self->{'list'} = {};
|
---|
| 113 | # if we are getting top X docs, and sorting by meta, we need to store the
|
---|
| 114 | # date and the metadata
|
---|
| 115 | if (!$self->{'classify_by_date'} && $self->{'sort'}) {
|
---|
| 116 | $self->{'meta_list'} = {};
|
---|
| 117 | }
|
---|
| 118 | return bless $self, $class;
|
---|
| 119 | }
|
---|
| 120 |
|
---|
| 121 | sub init {
|
---|
| 122 | my $self = shift (@_);
|
---|
| 123 |
|
---|
| 124 | }
|
---|
| 125 |
|
---|
| 126 | sub classify {
|
---|
| 127 | my $self = shift (@_);
|
---|
[23116] | 128 | my ($doc_obj) = @_;
|
---|
[13900] | 129 |
|
---|
| 130 | my $doc_OID = $doc_obj->get_OID();
|
---|
| 131 | my $lastmodified = $doc_obj->get_metadata_element($doc_obj->get_top_section(), "lastmodified");
|
---|
| 132 | if (!defined $lastmodified || $lastmodified eq "") {
|
---|
| 133 | print $self->{'outhandle'}, "RecentDocumentsList: $doc_OID has no lastmodified metadata, not classifying\n";
|
---|
| 134 | return;
|
---|
| 135 | }
|
---|
| 136 |
|
---|
| 137 | # doc goes into classification if we are not classifying by date, or the date is after the cutoff date.
|
---|
| 138 | if ($self->{'classify_by_date'}) {
|
---|
| 139 | if ($lastmodified > $self->{'classification_date'}) {
|
---|
| 140 | my $sort_meta = $lastmodified;
|
---|
| 141 | if (defined $self->{'sort'}) {
|
---|
| 142 | $sort_meta = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $self->{'sort'});
|
---|
| 143 | }
|
---|
| 144 | $self->{'list'}->{$doc_OID} = $sort_meta;
|
---|
| 145 | $doc_obj->add_metadata($doc_obj->get_top_section(), "memberof", "CL".$self->get_number());
|
---|
| 146 | }
|
---|
| 147 | } else {
|
---|
| 148 |
|
---|
| 149 | # need to store metadata as well...
|
---|
| 150 | $self->{'list'}->{$doc_OID} = $lastmodified;
|
---|
| 151 | if (defined $self->{'sort'}) {
|
---|
| 152 | my $sort_meta = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $self->{'sort'});
|
---|
| 153 | $self->{'meta_list'}->{$doc_OID} = $sort_meta;
|
---|
| 154 | }
|
---|
| 155 | }
|
---|
| 156 |
|
---|
| 157 | }
|
---|
| 158 |
|
---|
| 159 |
|
---|
| 160 | sub get_classify_info {
|
---|
| 161 | my $self = shift (@_);
|
---|
| 162 | my $return_doc_size=0;
|
---|
| 163 |
|
---|
| 164 | my $list = $self->{'list'};
|
---|
| 165 |
|
---|
| 166 |
|
---|
| 167 | # organise into classification structure
|
---|
| 168 | my %classifyinfo = ('thistype'=>'Invisible',
|
---|
| 169 | 'childtype'=>'VList',
|
---|
| 170 | 'Title'=>$self->{'buttonname'},
|
---|
| 171 | 'contains'=>[]);
|
---|
| 172 |
|
---|
| 173 |
|
---|
| 174 | # may or may not support memberof, depending on options set
|
---|
| 175 | $classifyinfo{'supportsmemberof'} = &supports_memberof();
|
---|
| 176 |
|
---|
| 177 | # get either all documents (sorted by date), or the top X docs
|
---|
| 178 | my @sorted_docs = sort {$self->date_or_metadata_sort($a,$b)} keys %{$self->{'list'}};
|
---|
| 179 | my $numdocs = $self->{'include_most_recently_added'};
|
---|
| 180 | if ($self->{'classify_by_date'}) {
|
---|
| 181 | # just include all docs in the list
|
---|
| 182 | $numdocs = scalar (@sorted_docs);
|
---|
| 183 | } else {
|
---|
| 184 | if ($numdocs > scalar (@sorted_docs)) {
|
---|
| 185 | $numdocs = scalar (@sorted_docs);
|
---|
| 186 | }
|
---|
| 187 | if ($self->{'sort'}) {
|
---|
| 188 | # we need to sort further by metadata
|
---|
| 189 | # cut off the list
|
---|
| 190 | @sorted_docs = @sorted_docs[0..$numdocs-1];
|
---|
| 191 | # sort again
|
---|
| 192 | @sorted_docs = sort {$self->external_meta_sort($a,$b)}@sorted_docs;
|
---|
| 193 | }
|
---|
| 194 | }
|
---|
| 195 | for (my $i=0; $i<$numdocs; $i++) {
|
---|
| 196 | push (@{$classifyinfo{'contains'}}, {'OID'=> $sorted_docs[$i]});
|
---|
| 197 | }
|
---|
| 198 |
|
---|
| 199 |
|
---|
| 200 | return \%classifyinfo;
|
---|
| 201 | }
|
---|
| 202 |
|
---|
| 203 | # 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.
|
---|
| 204 | sub supports_memberof {
|
---|
| 205 | my $self = shift(@_);
|
---|
| 206 |
|
---|
| 207 | if ($self->{'classify_by_date'}) {
|
---|
| 208 | return "true";
|
---|
| 209 | }
|
---|
| 210 | return "false";
|
---|
| 211 | }
|
---|
| 212 |
|
---|
| 213 | sub date_or_metadata_sort {
|
---|
| 214 | my ($self,$a,$b) = @_;
|
---|
| 215 | # make it do metadata too
|
---|
| 216 | my $date_a = $self->{'list'}->{$a};
|
---|
| 217 | my $date_b = $self->{'list'}->{$b};
|
---|
| 218 | if (!$self->{'sort'} || !$self->{'classify_by_date'}) {
|
---|
| 219 | # want reverse order (latest to earliest)
|
---|
| 220 | return ($date_b <=> $date_a);
|
---|
| 221 | }
|
---|
| 222 | # meta sorting, use string cmp
|
---|
| 223 | return ($date_a cmp $date_b);
|
---|
| 224 | }
|
---|
| 225 |
|
---|
| 226 | sub external_meta_sort {
|
---|
| 227 | my ($self,$a,$b) = @_;
|
---|
| 228 |
|
---|
| 229 | my $meta_a = $self->{'meta_list'}->{$a};
|
---|
| 230 | my $meta_b = $self->{'meta_list'}->{$b};
|
---|
| 231 |
|
---|
| 232 | return ($meta_a cmp $meta_b);
|
---|
| 233 | }
|
---|
| 234 |
|
---|
| 235 |
|
---|
| 236 | 1;
|
---|
| 237 |
|
---|
| 238 |
|
---|
| 239 |
|
---|
| 240 |
|
---|