[537] | 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 |
|
---|
[4] | 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 |
|
---|
| 35 | package arcinfo;
|
---|
| 36 |
|
---|
[10157] | 37 | use constant ORDER_OID_INDEX => 0;
|
---|
| 38 | use constant ORDER_SORT_INDEX => 1;
|
---|
| 39 |
|
---|
| 40 | use constant INFO_FILE_INDEX => 0;
|
---|
| 41 | use constant INFO_STATUS_INDEX => 1;
|
---|
| 42 |
|
---|
| 43 | # File format read in: OID <tab> Filename <tab> Optional-Index-Status
|
---|
| 44 |
|
---|
| 45 | # Index status can be:
|
---|
| 46 | # I = Index for the first time
|
---|
| 47 | # R = Reindex
|
---|
| 48 | # D = Delete
|
---|
| 49 | # B = Been indexed
|
---|
| 50 |
|
---|
[4] | 51 | sub new {
|
---|
| 52 | my ($class) = @_;
|
---|
[14] | 53 | my $self = {'info'=>{},
|
---|
[15073] | 54 | 'order'=>[],
|
---|
| 55 | 'reverse_sort'=>0};
|
---|
[4] | 56 |
|
---|
| 57 | return bless $self, $class;
|
---|
| 58 | }
|
---|
| 59 |
|
---|
| 60 | sub load_info {
|
---|
| 61 | my $self = shift (@_);
|
---|
| 62 | my ($filename) = @_;
|
---|
| 63 |
|
---|
| 64 | $self->{'info'} = {};
|
---|
[7904] | 65 |
|
---|
[12328] | 66 | if (defined $filename && -e $filename) {
|
---|
[4] | 67 | open (INFILE, $filename) ||
|
---|
| 68 | die "arcinfo::load_info couldn't read $filename\n";
|
---|
| 69 |
|
---|
| 70 | my ($line, @line);
|
---|
| 71 | while (defined ($line = <INFILE>)) {
|
---|
| 72 | $line =~ s/\cM|\cJ//g; # remove end-of-line characters
|
---|
| 73 | @line = split ("\t", $line); # filename,
|
---|
| 74 | if (scalar(@line) >= 2) {
|
---|
[14] | 75 | $self->add_info (@line);
|
---|
[4] | 76 | }
|
---|
| 77 | }
|
---|
| 78 | close (INFILE);
|
---|
| 79 | }
|
---|
| 80 | }
|
---|
| 81 |
|
---|
| 82 | sub save_info {
|
---|
| 83 | my $self = shift (@_);
|
---|
| 84 | my ($filename) = @_;
|
---|
| 85 |
|
---|
[14] | 86 | my ($OID, $info);
|
---|
[4] | 87 |
|
---|
| 88 | open (OUTFILE, ">$filename") ||
|
---|
| 89 | die "arcinfo::save_info couldn't write $filename\n";
|
---|
[7904] | 90 |
|
---|
[14] | 91 | foreach $info (@{$self->get_OID_list()}) {
|
---|
[4] | 92 | if (defined $info) {
|
---|
[14] | 93 | print OUTFILE join("\t", @$info), "\n";
|
---|
[4] | 94 | }
|
---|
| 95 | }
|
---|
| 96 | close (OUTFILE);
|
---|
| 97 | }
|
---|
| 98 |
|
---|
[14] | 99 | sub delete_info {
|
---|
| 100 | my $self = shift (@_);
|
---|
| 101 | my ($OID) = @_;
|
---|
| 102 |
|
---|
| 103 | if (defined $self->{'info'}->{$OID}) {
|
---|
| 104 | delete $self->{'info'}->{$OID};
|
---|
| 105 |
|
---|
| 106 | my $i = 0;
|
---|
| 107 | while ($i < scalar (@{$self->{'order'}})) {
|
---|
[10157] | 108 | if ($self->{'order'}->[$i]->[ORDER_OID_INDEX] eq $OID) {
|
---|
[14] | 109 | splice (@{$self->{'order'}}, $i, 1);
|
---|
| 110 | last;
|
---|
| 111 | }
|
---|
| 112 |
|
---|
| 113 | $i ++;
|
---|
| 114 | }
|
---|
| 115 | }
|
---|
| 116 | }
|
---|
| 117 |
|
---|
[4] | 118 | sub add_info {
|
---|
| 119 | my $self = shift (@_);
|
---|
[10157] | 120 | my ($OID, $doc_file, $index_status, $sortmeta) = @_;
|
---|
[1287] | 121 | $sortmeta = "" unless defined $sortmeta;
|
---|
[10157] | 122 | $index_status = "I" unless defined $index_status; # I = needs indexing
|
---|
[4] | 123 |
|
---|
[3416] | 124 | if (! defined($OID)) {
|
---|
| 125 | # only happens when no files can be processed?
|
---|
| 126 | return undef;
|
---|
| 127 | }
|
---|
[10157] | 128 |
|
---|
[14] | 129 | $self->delete_info ($OID);
|
---|
[10157] | 130 | $self->{'info'}->{$OID} = [$doc_file,$index_status];
|
---|
[1287] | 131 | push (@{$self->{'order'}}, [$OID, $sortmeta]);
|
---|
[4] | 132 | }
|
---|
| 133 |
|
---|
[10157] | 134 | sub set_status_info {
|
---|
[4] | 135 | my $self = shift (@_);
|
---|
[10157] | 136 | my ($OID, $index_status) = @_;
|
---|
[4] | 137 |
|
---|
[10157] | 138 | my $OID_info = $self->{'info'}->{$OID};
|
---|
| 139 | $OID_info->[INFO_STATUS_INDEX] = $index_status;
|
---|
| 140 | }
|
---|
| 141 |
|
---|
| 142 |
|
---|
| 143 | sub get_status_info {
|
---|
| 144 | my $self = shift (@_);
|
---|
| 145 | my ($OID) = @_;
|
---|
| 146 |
|
---|
| 147 | my $index_status = undef;
|
---|
| 148 |
|
---|
| 149 | my $OID_info = $self->{'info'}->{$OID};
|
---|
| 150 | if (defined $OID_info) {
|
---|
| 151 | $index_status = $OID_info->[INFO_STATUS_INDEX];
|
---|
| 152 | }
|
---|
| 153 | else {
|
---|
| 154 | die "Unable to find document id $OID\n";
|
---|
| 155 | }
|
---|
| 156 |
|
---|
| 157 | return $index_status;
|
---|
| 158 |
|
---|
| 159 | }
|
---|
| 160 |
|
---|
[15073] | 161 | sub reverse_sort
|
---|
| 162 | {
|
---|
| 163 | my $self = shift(@_);
|
---|
| 164 | $self->{'reverse_sort'} = 1;
|
---|
| 165 | }
|
---|
[10157] | 166 |
|
---|
| 167 | # returns a list of the form [[OID, doc_file, index_status], ...]
|
---|
| 168 | sub get_OID_list
|
---|
| 169 | {
|
---|
| 170 | my $self = shift (@_);
|
---|
| 171 |
|
---|
| 172 | my $order = $self->{'order'};
|
---|
| 173 |
|
---|
[15073] | 174 | my @sorted_order;
|
---|
| 175 | if ($self->{'reverse_sort'}) {
|
---|
| 176 | @sorted_order = sort {$b->[ORDER_SORT_INDEX] cmp $a->[ORDER_SORT_INDEX]} @$order;
|
---|
| 177 | } else {
|
---|
| 178 | @sorted_order = sort {$a->[ORDER_SORT_INDEX] cmp $b->[ORDER_SORT_INDEX]} @$order;
|
---|
| 179 | }
|
---|
[10157] | 180 |
|
---|
[4] | 181 | my @list = ();
|
---|
| 182 |
|
---|
[10157] | 183 | foreach my $OID_order (@sorted_order) {
|
---|
| 184 | my $OID = $OID_order->[ORDER_OID_INDEX];
|
---|
| 185 | my $OID_info = $self->{'info'}->{$OID};
|
---|
| 186 |
|
---|
| 187 | push (@list, [$OID, $OID_info->[INFO_FILE_INDEX],
|
---|
| 188 | $OID_info->[INFO_STATUS_INDEX]]);
|
---|
[4] | 189 | }
|
---|
| 190 |
|
---|
| 191 | return \@list;
|
---|
| 192 | }
|
---|
| 193 |
|
---|
| 194 | # returns a list of the form [[doc_file, OID], ...]
|
---|
| 195 | sub get_file_list {
|
---|
| 196 | my $self = shift (@_);
|
---|
| 197 |
|
---|
[10157] | 198 | my $order = $self->{'order'};
|
---|
| 199 |
|
---|
[15073] | 200 | my @sorted_order;
|
---|
| 201 | if ($self->{'reverse_sort'}) {
|
---|
| 202 | @sorted_order = sort {$b->[ORDER_SORT_INDEX] cmp $a->[ORDER_SORT_INDEX]} @$order;
|
---|
| 203 | } else {
|
---|
| 204 | @sorted_order = sort {$a->[ORDER_SORT_INDEX] cmp $b->[ORDER_SORT_INDEX]} @$order;
|
---|
| 205 | }
|
---|
[10157] | 206 |
|
---|
[4] | 207 | my @list = ();
|
---|
| 208 |
|
---|
[10157] | 209 | foreach $OID_order (@sorted_order) {
|
---|
| 210 | my $OID = $OID_order->[ORDER_OID_INDEX];
|
---|
| 211 | my $OID_info = $self->{'info'}->{$OID};
|
---|
| 212 |
|
---|
| 213 | push (@list, [$OID_info->[INFO_FILE_INDEX], $OID]);
|
---|
[4] | 214 | }
|
---|
| 215 |
|
---|
| 216 | return \@list;
|
---|
| 217 | }
|
---|
| 218 |
|
---|
| 219 |
|
---|
| 220 | # returns a list of the form [doc_file]
|
---|
| 221 | sub get_info {
|
---|
| 222 | my $self = shift (@_);
|
---|
| 223 | my ($OID) = @_;
|
---|
| 224 |
|
---|
| 225 | if (defined $self->{'info'}->{$OID}) {
|
---|
| 226 | return $self->{'info'}->{$OID};
|
---|
| 227 | }
|
---|
| 228 |
|
---|
| 229 | return undef;
|
---|
| 230 | }
|
---|
| 231 |
|
---|
| 232 |
|
---|
[98] | 233 | # returns the number of documents so far
|
---|
| 234 | sub size {
|
---|
| 235 | my $self = shift (@_);
|
---|
| 236 | return (scalar(@{$self->{'order'}}));
|
---|
| 237 | }
|
---|
| 238 |
|
---|
[4] | 239 | 1;
|
---|
| 240 |
|
---|