1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # arcinfo.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 |
|
---|
27 | # This module stores information about the archives. At the moment
|
---|
28 | # this information just consists of the file name (relative to the
|
---|
29 | # directory the archives information file is in) and its OID.
|
---|
30 |
|
---|
31 | # This module assumes there is a one to one correspondance between
|
---|
32 | # a file in the archives directory and an OID.
|
---|
33 |
|
---|
34 | package arcinfo;
|
---|
35 |
|
---|
36 | use constant ORDER_OID_INDEX => 0;
|
---|
37 | use constant ORDER_SORT_INDEX => 1;
|
---|
38 |
|
---|
39 | use constant INFO_FILE_INDEX => 0;
|
---|
40 | use constant INFO_STATUS_INDEX => 1;
|
---|
41 |
|
---|
42 | use strict;
|
---|
43 |
|
---|
44 |
|
---|
45 | # File format read in: OID <tab> Filename <tab> Optional-Index-Status
|
---|
46 |
|
---|
47 | # Index status can be:
|
---|
48 | # I = Index for the first time
|
---|
49 | # R = Reindex
|
---|
50 | # D = Delete
|
---|
51 | # B = Been indexed
|
---|
52 |
|
---|
53 | sub new {
|
---|
54 | my ($class) = @_;
|
---|
55 | my $self = {'info'=>{},
|
---|
56 | 'order'=>[],
|
---|
57 | 'reverse_sort'=>0};
|
---|
58 |
|
---|
59 | return bless $self, $class;
|
---|
60 | }
|
---|
61 |
|
---|
62 | sub load_info {
|
---|
63 | my $self = shift (@_);
|
---|
64 | my ($filename) = @_;
|
---|
65 |
|
---|
66 | $self->{'info'} = {};
|
---|
67 |
|
---|
68 | if (defined $filename && -e $filename) {
|
---|
69 | open (INFILE, $filename) ||
|
---|
70 | die "arcinfo::load_info couldn't read $filename\n";
|
---|
71 |
|
---|
72 | my ($line, @line);
|
---|
73 | while (defined ($line = <INFILE>)) {
|
---|
74 | $line =~ s/\cM|\cJ//g; # remove end-of-line characters
|
---|
75 | @line = split ("\t", $line); # filename,
|
---|
76 | if (scalar(@line) >= 2) {
|
---|
77 | $self->add_info (@line);
|
---|
78 | }
|
---|
79 | }
|
---|
80 | close (INFILE);
|
---|
81 | }
|
---|
82 | }
|
---|
83 |
|
---|
84 | sub save_info {
|
---|
85 | my $self = shift (@_);
|
---|
86 | my ($filename) = @_;
|
---|
87 |
|
---|
88 | my ($OID, $info);
|
---|
89 |
|
---|
90 | open (OUTFILE, ">$filename") ||
|
---|
91 | die "arcinfo::save_info couldn't write $filename\n";
|
---|
92 |
|
---|
93 | foreach $info (@{$self->get_OID_list()}) {
|
---|
94 | if (defined $info) {
|
---|
95 | print OUTFILE join("\t", @$info), "\n";
|
---|
96 | }
|
---|
97 | }
|
---|
98 | close (OUTFILE);
|
---|
99 | }
|
---|
100 |
|
---|
101 | sub delete_info {
|
---|
102 | my $self = shift (@_);
|
---|
103 | my ($OID) = @_;
|
---|
104 |
|
---|
105 | if (defined $self->{'info'}->{$OID}) {
|
---|
106 | delete $self->{'info'}->{$OID};
|
---|
107 |
|
---|
108 | my $i = 0;
|
---|
109 | while ($i < scalar (@{$self->{'order'}})) {
|
---|
110 | if ($self->{'order'}->[$i]->[ORDER_OID_INDEX] eq $OID) {
|
---|
111 | splice (@{$self->{'order'}}, $i, 1);
|
---|
112 | last;
|
---|
113 | }
|
---|
114 |
|
---|
115 | $i ++;
|
---|
116 | }
|
---|
117 | }
|
---|
118 | }
|
---|
119 |
|
---|
120 | sub add_info {
|
---|
121 | my $self = shift (@_);
|
---|
122 | my ($OID, $doc_file, $index_status, $sortmeta) = @_;
|
---|
123 | $sortmeta = "" unless defined $sortmeta;
|
---|
124 | $index_status = "I" unless defined $index_status; # I = needs indexing
|
---|
125 |
|
---|
126 | if (! defined($OID)) {
|
---|
127 | # only happens when no files can be processed?
|
---|
128 | return undef;
|
---|
129 | }
|
---|
130 |
|
---|
131 | $self->delete_info ($OID);
|
---|
132 | $self->{'info'}->{$OID} = [$doc_file,$index_status];
|
---|
133 | push (@{$self->{'order'}}, [$OID, $sortmeta]);
|
---|
134 | }
|
---|
135 |
|
---|
136 | sub set_status_info {
|
---|
137 | my $self = shift (@_);
|
---|
138 | my ($OID, $index_status) = @_;
|
---|
139 |
|
---|
140 | my $OID_info = $self->{'info'}->{$OID};
|
---|
141 | $OID_info->[INFO_STATUS_INDEX] = $index_status;
|
---|
142 | }
|
---|
143 |
|
---|
144 |
|
---|
145 | sub get_status_info {
|
---|
146 | my $self = shift (@_);
|
---|
147 | my ($OID) = @_;
|
---|
148 |
|
---|
149 | my $index_status = undef;
|
---|
150 |
|
---|
151 | my $OID_info = $self->{'info'}->{$OID};
|
---|
152 | if (defined $OID_info) {
|
---|
153 | $index_status = $OID_info->[INFO_STATUS_INDEX];
|
---|
154 | }
|
---|
155 | else {
|
---|
156 | die "Unable to find document id $OID\n";
|
---|
157 | }
|
---|
158 |
|
---|
159 | return $index_status;
|
---|
160 |
|
---|
161 | }
|
---|
162 |
|
---|
163 | sub reverse_sort
|
---|
164 | {
|
---|
165 | my $self = shift(@_);
|
---|
166 | $self->{'reverse_sort'} = 1;
|
---|
167 | }
|
---|
168 |
|
---|
169 | # returns a list of the form [[OID, doc_file, index_status], ...]
|
---|
170 | sub get_OID_list
|
---|
171 | {
|
---|
172 | my $self = shift (@_);
|
---|
173 |
|
---|
174 | my $order = $self->{'order'};
|
---|
175 |
|
---|
176 | my @sorted_order;
|
---|
177 | if ($self->{'reverse_sort'}) {
|
---|
178 | @sorted_order = sort {$b->[ORDER_SORT_INDEX] cmp $a->[ORDER_SORT_INDEX]} @$order;
|
---|
179 | } else {
|
---|
180 | @sorted_order = sort {$a->[ORDER_SORT_INDEX] cmp $b->[ORDER_SORT_INDEX]} @$order;
|
---|
181 | }
|
---|
182 |
|
---|
183 | my @list = ();
|
---|
184 |
|
---|
185 | foreach my $OID_order (@sorted_order) {
|
---|
186 | my $OID = $OID_order->[ORDER_OID_INDEX];
|
---|
187 | my $OID_info = $self->{'info'}->{$OID};
|
---|
188 |
|
---|
189 | push (@list, [$OID, $OID_info->[INFO_FILE_INDEX],
|
---|
190 | $OID_info->[INFO_STATUS_INDEX]]);
|
---|
191 | }
|
---|
192 |
|
---|
193 | return \@list;
|
---|
194 | }
|
---|
195 |
|
---|
196 | # returns a list of the form [[doc_file, OID], ...]
|
---|
197 | sub get_file_list {
|
---|
198 | my $self = shift (@_);
|
---|
199 |
|
---|
200 | my $order = $self->{'order'};
|
---|
201 |
|
---|
202 | my @sorted_order;
|
---|
203 | if ($self->{'reverse_sort'}) {
|
---|
204 | @sorted_order = sort {$b->[ORDER_SORT_INDEX] cmp $a->[ORDER_SORT_INDEX]} @$order;
|
---|
205 | } else {
|
---|
206 | @sorted_order = sort {$a->[ORDER_SORT_INDEX] cmp $b->[ORDER_SORT_INDEX]} @$order;
|
---|
207 | }
|
---|
208 |
|
---|
209 | my @list = ();
|
---|
210 |
|
---|
211 | foreach my $OID_order (@sorted_order) {
|
---|
212 | my $OID = $OID_order->[ORDER_OID_INDEX];
|
---|
213 | my $OID_info = $self->{'info'}->{$OID};
|
---|
214 |
|
---|
215 | push (@list, [$OID_info->[INFO_FILE_INDEX], $OID]);
|
---|
216 | }
|
---|
217 |
|
---|
218 | return \@list;
|
---|
219 | }
|
---|
220 |
|
---|
221 |
|
---|
222 | # returns a list of the form [doc_file]
|
---|
223 | sub get_info {
|
---|
224 | my $self = shift (@_);
|
---|
225 | my ($OID) = @_;
|
---|
226 |
|
---|
227 | if (defined $self->{'info'}->{$OID}) {
|
---|
228 | return $self->{'info'}->{$OID};
|
---|
229 | }
|
---|
230 |
|
---|
231 | return undef;
|
---|
232 | }
|
---|
233 |
|
---|
234 |
|
---|
235 | # returns the number of documents so far
|
---|
236 | sub size {
|
---|
237 | my $self = shift (@_);
|
---|
238 | return (scalar(@{$self->{'order'}}));
|
---|
239 | }
|
---|
240 |
|
---|
241 | 1;
|
---|
242 |
|
---|