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