[537] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # DateList.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 |
|
---|
[408] | 26 | # classifier plugin for sorting by date
|
---|
| 27 |
|
---|
[2632] | 28 | # always sorts by 'Date' metadata
|
---|
[408] | 29 |
|
---|
| 30 | # date is assumed to be in the form yyyymmdd
|
---|
| 31 |
|
---|
| 32 | # at present dates are split by year - this should change
|
---|
[2632] | 33 | # jrm21 - added option "bymonth", which splits by year and month.
|
---|
[408] | 34 |
|
---|
| 35 | package DateList;
|
---|
| 36 |
|
---|
[1483] | 37 | use BasClas;
|
---|
[408] | 38 | use sorttools;
|
---|
| 39 |
|
---|
[1483] | 40 | sub BEGIN {
|
---|
| 41 | @ISA = ('BasClas');
|
---|
| 42 | }
|
---|
| 43 |
|
---|
[1839] | 44 | sub print_usage {
|
---|
| 45 | print STDERR "
|
---|
[2632] | 46 | usage: classify DateList [options]
|
---|
| 47 | options:
|
---|
| 48 | -bymonth [or bymonth=1] Classify by year and month
|
---|
[1839] | 49 |
|
---|
[2916] | 50 | Classifier plugin for sorting by date, and assumes that 'Date' metadata
|
---|
| 51 | exists. Date is assumed to be in the form yyyymmdd (all digits).
|
---|
| 52 | By default dates are classified by year.
|
---|
[1839] | 53 |
|
---|
| 54 | ";
|
---|
| 55 | }
|
---|
| 56 |
|
---|
[408] | 57 | sub new {
|
---|
[1839] | 58 | my $class = shift (@_);
|
---|
| 59 | my $self = new BasClas($class, @_);
|
---|
[408] | 60 |
|
---|
[1483] | 61 | $self->{'list'} = {};
|
---|
[2632] | 62 | if (!parsargv::parse(\@_,
|
---|
| 63 | q^bymonth^, \$self->{'bymonth'},
|
---|
| 64 | "allow_extra_options")) {
|
---|
| 65 | &print_usage();
|
---|
| 66 | die "\n";
|
---|
| 67 | }
|
---|
[1483] | 68 | return bless $self, $class;
|
---|
[408] | 69 | }
|
---|
| 70 |
|
---|
| 71 | sub init {
|
---|
| 72 | my $self = shift (@_);
|
---|
| 73 |
|
---|
| 74 | $self->{'list'} = {};
|
---|
| 75 | }
|
---|
| 76 |
|
---|
| 77 | sub classify {
|
---|
| 78 | my $self = shift (@_);
|
---|
| 79 | my ($doc_obj) = @_;
|
---|
| 80 |
|
---|
| 81 | my $doc_OID = $doc_obj->get_OID();
|
---|
| 82 | my $date = $doc_obj->get_metadata_element ($doc_obj->get_top_section(), 'Date');
|
---|
| 83 |
|
---|
| 84 | # if this document doesn't contain Date element we won't
|
---|
| 85 | # include it in this classification
|
---|
| 86 | if (defined $date && $date =~ /\d/) {
|
---|
| 87 | if (defined $self->{'list'}->{$doc_OID}) {
|
---|
[1483] | 88 | my $outhandle = $self->{'outhandle'};
|
---|
| 89 | print $outhandle "WARNING: DateList::classify called multiple times for $doc_OID\n";
|
---|
[408] | 90 | }
|
---|
| 91 | $self->{'list'}->{$doc_OID} = $date;
|
---|
| 92 | }
|
---|
| 93 | }
|
---|
| 94 |
|
---|
| 95 |
|
---|
| 96 | sub get_classify_info {
|
---|
| 97 | my $self = shift (@_);
|
---|
| 98 |
|
---|
| 99 | my @classlist = sort {$self->{'list'}->{$a} cmp $self->{'list'}->{$b};} keys %{$self->{'list'}};
|
---|
| 100 |
|
---|
| 101 | return $self->splitlist (\@classlist);
|
---|
| 102 | }
|
---|
| 103 |
|
---|
| 104 |
|
---|
| 105 | sub get_entry {
|
---|
| 106 | my $self = shift (@_);
|
---|
[677] | 107 | my ($title, $childtype, $thistype) = @_;
|
---|
[408] | 108 |
|
---|
| 109 | # organise into classification structure
|
---|
[677] | 110 | my %classifyinfo = ('childtype'=>$childtype,
|
---|
[408] | 111 | 'Title'=>$title,
|
---|
| 112 | 'contains'=>[]);
|
---|
[677] | 113 | $classifyinfo{'thistype'} = $thistype
|
---|
| 114 | if defined $thistype && $thistype =~ /\w/;
|
---|
[408] | 115 |
|
---|
| 116 | return \%classifyinfo;
|
---|
| 117 | }
|
---|
| 118 |
|
---|
[2632] | 119 | # splitlist takes an ordered list of classifications (@$classlistref) and
|
---|
| 120 | # splits it up into sub-sections by date
|
---|
[408] | 121 | sub splitlist {
|
---|
| 122 | my $self = shift (@_);
|
---|
| 123 | my ($classlistref) = @_;
|
---|
| 124 | my $classhash = {};
|
---|
| 125 |
|
---|
| 126 | # top level
|
---|
[677] | 127 | my $childtype = "HList";
|
---|
[2503] | 128 | if (scalar (@$classlistref) <= 39) {$childtype = "DateList";}
|
---|
[677] | 129 | my $classifyinfo = $self->get_entry ("Date", $childtype, "Invisible");
|
---|
[408] | 130 |
|
---|
[2503] | 131 | # don't need to do any splitting if there are less than 39 (max + min -1) classifications
|
---|
| 132 | if ((scalar @$classlistref) <= 39) {
|
---|
[408] | 133 | foreach $subOID (@$classlistref) {
|
---|
| 134 | push (@{$classifyinfo->{'contains'}}, {'OID'=>$subOID});
|
---|
| 135 | }
|
---|
| 136 | return $classifyinfo;
|
---|
| 137 | }
|
---|
| 138 |
|
---|
[2632] | 139 |
|
---|
| 140 | if ($self->{'bymonth'}) {
|
---|
| 141 | # first split up the list into separate year+month classifications
|
---|
| 142 | foreach $classification (@$classlistref) {
|
---|
| 143 | my $date = $self->{'list'}->{$classification};
|
---|
| 144 | $date =~ s/^(\d\d\d\d)(\d\d).*$/$1 _textmonth$2_/;
|
---|
| 145 | # sanity check if month is zero
|
---|
| 146 | if ($date =~ /00_$/) {
|
---|
| 147 | $date =~ s/^(\d\d\d\d).*$/$1/g;
|
---|
| 148 | }
|
---|
| 149 | $classhash->{$date} = [] unless defined $classhash->{$date};
|
---|
| 150 | push (@{$classhash->{$date}}, $classification);
|
---|
| 151 | }
|
---|
| 152 | } else {
|
---|
| 153 | # first split up the list into separate year classifications
|
---|
| 154 | foreach $classification (@$classlistref) {
|
---|
| 155 | my $date = $self->{'list'}->{$classification};
|
---|
| 156 | $date =~ s/^(\d\d\d\d).*$/$1/;
|
---|
| 157 | $classhash->{$date} = [] unless defined $classhash->{$date};
|
---|
| 158 | push (@{$classhash->{$date}}, $classification);
|
---|
| 159 | }
|
---|
[408] | 160 | }
|
---|
| 161 | $classhash = $self->compactlist ($classhash);
|
---|
| 162 |
|
---|
| 163 | foreach $subclass (sort keys %$classhash) {
|
---|
| 164 | my $tempclassify = $self->get_entry($subclass, "DateList");
|
---|
| 165 | foreach $subsubOID (@{$classhash->{$subclass}}) {
|
---|
| 166 | push (@{$tempclassify->{'contains'}}, {'OID'=>$subsubOID});
|
---|
| 167 | }
|
---|
| 168 | push (@{$classifyinfo->{'contains'}}, $tempclassify);
|
---|
| 169 | }
|
---|
| 170 |
|
---|
| 171 | return $classifyinfo;
|
---|
| 172 | }
|
---|
| 173 |
|
---|
| 174 | sub compactlist {
|
---|
| 175 | my $self = shift (@_);
|
---|
| 176 | my ($classhashref) = @_;
|
---|
| 177 | my $compactedhash = {};
|
---|
| 178 | my @currentOIDs = ();
|
---|
[2632] | 179 | my $currentfirstdate = "";
|
---|
| 180 | my $currentlastdate = "";
|
---|
[408] | 181 | my $lastkey = "";
|
---|
| 182 |
|
---|
| 183 | # minimum and maximum documents to be displayed per page.
|
---|
| 184 | # the actual maximum will be max + (min-1).
|
---|
| 185 | # the smallest sub-section is a single letter at present
|
---|
| 186 | # so in this case there may be many times max documents
|
---|
| 187 | # displayed on a page.
|
---|
| 188 | my $min = 10;
|
---|
| 189 | my $max = 30;
|
---|
[2632] | 190 | foreach my $subsection (sort keys %$classhashref) {
|
---|
| 191 | $currentfirstdate = $subsection if $currentfirstdate eq "";
|
---|
[408] | 192 | if ((scalar (@currentOIDs) < $min) ||
|
---|
| 193 | ((scalar (@currentOIDs) + scalar (@{$classhashref->{$subsection}})) <= $max)) {
|
---|
| 194 | push (@currentOIDs, @{$classhashref->{$subsection}});
|
---|
[2632] | 195 | $currentlastdate = $subsection;
|
---|
[408] | 196 | } else {
|
---|
| 197 |
|
---|
[2632] | 198 | if ($currentfirstdate eq $currentlastdate) {
|
---|
| 199 | @{$compactedhash->{$currentfirstdate}} = @currentOIDs;
|
---|
| 200 | $lastkey = $currentfirstdate;
|
---|
[408] | 201 | } else {
|
---|
[2632] | 202 | @{$compactedhash->{"$currentfirstdate-$currentlastdate"}} = @currentOIDs;
|
---|
| 203 | $lastkey = "$currentfirstdate-$currentlastdate";
|
---|
[408] | 204 | }
|
---|
| 205 | if (scalar (@{$classhashref->{$subsection}}) >= $max) {
|
---|
| 206 | $compactedhash->{$subsection} = $classhashref->{$subsection};
|
---|
| 207 | @currentOIDs = ();
|
---|
[2632] | 208 | $currentfirstdate = "";
|
---|
[677] | 209 | $lastkey = $subsection;
|
---|
[408] | 210 | } else {
|
---|
| 211 | @currentOIDs = @{$classhashref->{$subsection}};
|
---|
[2632] | 212 | $currentfirstdate = $subsection;
|
---|
| 213 | $currentlastdate = $subsection;
|
---|
[408] | 214 | }
|
---|
| 215 | }
|
---|
| 216 | }
|
---|
| 217 |
|
---|
| 218 | # add final OIDs to last sub-classification if there aren't many otherwise
|
---|
| 219 | # add final sub-classification
|
---|
[2503] | 220 | if ((scalar (@currentOIDs) < $min) && (scalar (@currentOIDs) > 0)) {
|
---|
[2685] | 221 | # want every thing in previous up to the dash
|
---|
| 222 | my ($newkey) = $lastkey =~ /^([^\-]+)/;
|
---|
[408] | 223 | @currentOIDs = (@{$compactedhash->{$lastkey}}, @currentOIDs);
|
---|
| 224 | delete $compactedhash->{$lastkey};
|
---|
[2632] | 225 | @{$compactedhash->{"$newkey-$currentlastdate"}} = @currentOIDs;
|
---|
[408] | 226 | } else {
|
---|
[2632] | 227 | if ($currentfirstdate eq $currentlastdate) {
|
---|
| 228 | @{$compactedhash->{$currentfirstdate}} = @currentOIDs;
|
---|
[408] | 229 | } else {
|
---|
[2632] | 230 | @{$compactedhash->{"$currentfirstdate-$currentlastdate"}} = @currentOIDs;
|
---|
[408] | 231 | }
|
---|
| 232 | }
|
---|
| 233 |
|
---|
| 234 | return $compactedhash;
|
---|
| 235 | }
|
---|
| 236 |
|
---|
| 237 | 1;
|
---|