[21411] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # dbutil::jdbm -- utility functions for writing to jdbm databases
|
---|
| 4 | #
|
---|
| 5 | # A component of the Greenstone digital library software
|
---|
| 6 | # from the New Zealand Digital Library Project at the
|
---|
| 7 | # University of Waikato, New Zealand.
|
---|
| 8 | #
|
---|
| 9 | # Copyright (C) 2009
|
---|
| 10 | #
|
---|
| 11 | # This program is free software; you can redistribute it and/or modify
|
---|
| 12 | # it under the terms of the GNU General Public License as published by
|
---|
| 13 | # the Free Software Foundation; either version 2 of the License, or
|
---|
| 14 | # (at your option) any later version.
|
---|
| 15 | #
|
---|
| 16 | # This program is distributed in the hope that it will be useful,
|
---|
| 17 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 18 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 19 | # GNU General Public License for more details.
|
---|
| 20 | #
|
---|
| 21 | # You should have received a copy of the GNU General Public License
|
---|
| 22 | # along with this program; if not, write to the Free Software
|
---|
| 23 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
| 24 | #
|
---|
| 25 | ###########################################################################
|
---|
| 26 |
|
---|
| 27 | package dbutil::jdbm;
|
---|
| 28 |
|
---|
| 29 | use strict;
|
---|
| 30 |
|
---|
| 31 |
|
---|
| 32 | # -----------------------------------------------------------------------------
|
---|
| 33 | # JDBM IMPLEMENTATION
|
---|
| 34 | # -----------------------------------------------------------------------------
|
---|
| 35 |
|
---|
| 36 | # When DBUtil::* is properly structured with inheritence, then
|
---|
| 37 | # much of this code (along with GDBM and GDBM-TXT-GZ) can be grouped into
|
---|
| 38 | # a shared base class. Really it is only the the command that needs to
|
---|
| 39 | # be constructed that changes between much of the code that is used
|
---|
| 40 |
|
---|
| 41 |
|
---|
| 42 | sub open_infodb_write_handle
|
---|
| 43 | {
|
---|
| 44 | my $infodb_file_path = shift(@_);
|
---|
| 45 | my $opt_append = shift(@_);
|
---|
| 46 |
|
---|
| 47 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
| 48 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
| 49 |
|
---|
| 50 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
| 51 |
|
---|
| 52 | my $infodb_file_handle = undef;
|
---|
| 53 | my $txt2jdb_cmd = "java -cp \"$classpath\" Txt2Jdb";
|
---|
| 54 |
|
---|
| 55 | if ((defined $opt_append) && ($opt_append eq "append")) {
|
---|
| 56 | $txt2jdb_cmd .= " -append";
|
---|
[28126] | 57 | print STDERR "Append operation to $infodb_file_path\n";
|
---|
[21411] | 58 | }
|
---|
[28126] | 59 | else {
|
---|
| 60 | print STDERR "Create database $infodb_file_path\n";
|
---|
| 61 | }
|
---|
[21411] | 62 |
|
---|
| 63 | # Lop off file extension, as JDBM does not expect this to be present
|
---|
| 64 | $infodb_file_path =~ s/\.jdb$//;
|
---|
| 65 |
|
---|
| 66 | $txt2jdb_cmd .= " \"$infodb_file_path\"";
|
---|
| 67 |
|
---|
[28126] | 68 | ## print STDERR "**** txt2jdbm cmd = $txt2jdb_cmd\n\n";
|
---|
| 69 |
|
---|
[21411] | 70 | if (!open($infodb_file_handle, "| $txt2jdb_cmd"))
|
---|
| 71 | {
|
---|
[22076] | 72 | print STDERR "Error: Failed to open pipe to $txt2jdb_cmd";
|
---|
| 73 | print STDERR " $!\n";
|
---|
| 74 | return undef;
|
---|
[21411] | 75 | }
|
---|
[23166] | 76 |
|
---|
| 77 | binmode($infodb_file_handle,":utf8");
|
---|
[21411] | 78 | return $infodb_file_handle;
|
---|
| 79 | }
|
---|
| 80 |
|
---|
| 81 |
|
---|
| 82 |
|
---|
| 83 | sub close_infodb_write_handle
|
---|
| 84 | {
|
---|
| 85 | my $infodb_handle = shift(@_);
|
---|
| 86 |
|
---|
| 87 | close($infodb_handle);
|
---|
| 88 | }
|
---|
| 89 |
|
---|
| 90 |
|
---|
| 91 | sub get_infodb_file_path
|
---|
| 92 | {
|
---|
| 93 | my $collection_name = shift(@_);
|
---|
| 94 | my $infodb_directory_path = shift(@_);
|
---|
| 95 |
|
---|
| 96 | my $infodb_file_extension = ".jdb";
|
---|
| 97 | my $infodb_file_name = &util::get_dirsep_tail($collection_name) . $infodb_file_extension;
|
---|
| 98 | return &util::filename_cat($infodb_directory_path, $infodb_file_name);
|
---|
| 99 | }
|
---|
| 100 |
|
---|
| 101 |
|
---|
| 102 | sub read_infodb_file
|
---|
| 103 | {
|
---|
| 104 | my $infodb_file_path = shift(@_);
|
---|
| 105 | my $infodb_map = shift(@_);
|
---|
| 106 |
|
---|
| 107 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
| 108 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
| 109 |
|
---|
| 110 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
| 111 |
|
---|
| 112 | my $jdb2txt_cmd = "java -cp \"$classpath\" Jdb2Txt";
|
---|
| 113 |
|
---|
| 114 | open (PIPEIN, "$jdb2txt_cmd \"$infodb_file_path\" |") || die "couldn't open pipe from db2txt \$infodb_file_path\"\n";
|
---|
[23166] | 115 | binmode(PIPEIN,":utf8");
|
---|
[21411] | 116 | my $infodb_line = "";
|
---|
| 117 | my $infodb_key = "";
|
---|
| 118 | my $infodb_value = "";
|
---|
| 119 | while (defined ($infodb_line = <PIPEIN>))
|
---|
| 120 | {
|
---|
| 121 | if ($infodb_line =~ /^\[([^\]]+)\]$/)
|
---|
| 122 | {
|
---|
| 123 | $infodb_key = $1;
|
---|
| 124 | }
|
---|
| 125 | elsif ($infodb_line =~ /^-{70}$/)
|
---|
| 126 | {
|
---|
| 127 | $infodb_map->{$infodb_key} = $infodb_value;
|
---|
| 128 | $infodb_key = "";
|
---|
| 129 | $infodb_value = "";
|
---|
| 130 | }
|
---|
| 131 | else
|
---|
| 132 | {
|
---|
| 133 | $infodb_value .= $infodb_line;
|
---|
| 134 | }
|
---|
| 135 | }
|
---|
| 136 |
|
---|
| 137 | close (PIPEIN);
|
---|
| 138 | }
|
---|
| 139 |
|
---|
| 140 | sub read_infodb_keys
|
---|
| 141 | {
|
---|
| 142 | my $infodb_file_path = shift(@_);
|
---|
| 143 | my $infodb_map = shift(@_);
|
---|
| 144 |
|
---|
| 145 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
| 146 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
| 147 |
|
---|
| 148 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
| 149 |
|
---|
| 150 | my $jdbkeys_cmd = "java -cp \"$classpath\" JdbKeys";
|
---|
| 151 |
|
---|
| 152 | open (PIPEIN, "$jdbkeys_cmd \"$infodb_file_path\" |") || die "couldn't open pipe from jdbmkeys \$infodb_file_path\"\n";
|
---|
[23166] | 153 | binmode(PIPEIN,":utf8");
|
---|
[21411] | 154 | my $infodb_line = "";
|
---|
| 155 | my $infodb_key = "";
|
---|
| 156 | my $infodb_value = "";
|
---|
| 157 | while (defined ($infodb_line = <PIPEIN>))
|
---|
| 158 | {
|
---|
| 159 | chomp $infodb_line; # remove end of line
|
---|
| 160 |
|
---|
| 161 | $infodb_map->{$infodb_line} = 1;
|
---|
| 162 | }
|
---|
| 163 |
|
---|
| 164 | close (PIPEIN);
|
---|
| 165 | }
|
---|
| 166 |
|
---|
| 167 |
|
---|
| 168 |
|
---|
| 169 | sub write_infodb_entry
|
---|
| 170 | {
|
---|
| 171 |
|
---|
| 172 | my $infodb_handle = shift(@_);
|
---|
| 173 | my $infodb_key = shift(@_);
|
---|
| 174 | my $infodb_map = shift(@_);
|
---|
| 175 |
|
---|
| 176 | print $infodb_handle "[$infodb_key]\n";
|
---|
| 177 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
| 178 | {
|
---|
| 179 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
| 180 | {
|
---|
| 181 | if ($infodb_value =~ /-{70,}/)
|
---|
| 182 | {
|
---|
| 183 | # if value contains 70 or more hyphens in a row we need to escape them
|
---|
| 184 | # to prevent txt2db from treating them as a separator
|
---|
| 185 | $infodb_value =~ s/-/&\#045;/gi;
|
---|
| 186 | }
|
---|
| 187 | print $infodb_handle "<$infodb_value_key>" . $infodb_value . "\n";
|
---|
| 188 | }
|
---|
| 189 | }
|
---|
| 190 | print $infodb_handle '-' x 70, "\n";
|
---|
| 191 | }
|
---|
| 192 |
|
---|
| 193 |
|
---|
[21856] | 194 | sub write_infodb_rawentry
|
---|
| 195 | {
|
---|
| 196 |
|
---|
| 197 | my $infodb_handle = shift(@_);
|
---|
| 198 | my $infodb_key = shift(@_);
|
---|
| 199 | my $infodb_val = shift(@_);
|
---|
| 200 |
|
---|
| 201 | print $infodb_handle "[$infodb_key]\n";
|
---|
| 202 | print $infodb_handle "$infodb_val\n";
|
---|
| 203 | print $infodb_handle '-' x 70, "\n";
|
---|
| 204 | }
|
---|
| 205 |
|
---|
[23399] | 206 | sub set_infodb_entry
|
---|
| 207 | {
|
---|
[28108] | 208 | my $infodb_file_path = shift(@_);
|
---|
| 209 | my $infodb_key = shift(@_);
|
---|
| 210 | my $infodb_map = shift(@_);
|
---|
[23399] | 211 |
|
---|
[28108] | 212 | # HTML escape anything that is not part of the "contains" metadata value
|
---|
| 213 | foreach my $k (keys %$infodb_map) {
|
---|
| 214 | my @escaped_v = ();
|
---|
| 215 | foreach my $v (@{$infodb_map->{$k}}) {
|
---|
| 216 | if ($k eq "contains") {
|
---|
| 217 | push(@escaped_v, $v);
|
---|
| 218 | }
|
---|
| 219 | else {
|
---|
| 220 | my $ev = &ghtml::unescape_html($v);
|
---|
| 221 | push(@escaped_v, $ev);
|
---|
| 222 | }
|
---|
| 223 | }
|
---|
| 224 | $infodb_map->{$k} = \@escaped_v;
|
---|
| 225 | }
|
---|
| 226 |
|
---|
| 227 | # Generate the record string
|
---|
| 228 | my $serialized_infodb_map = &dbutil::convert_infodb_hash_to_string($infodb_map);
|
---|
| 229 | ### print STDERR "**** ser dr\n$serialized_infodb_map\n\n\n";
|
---|
| 230 |
|
---|
| 231 | # Store it into JDBM using 'Txt2Jdb .... -append' which despite its name
|
---|
| 232 | # actually replaces the record if it already exists
|
---|
| 233 |
|
---|
| 234 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
| 235 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
| 236 |
|
---|
| 237 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
| 238 |
|
---|
| 239 | # Lop off file extension, as JDBM does not expect this to be present
|
---|
| 240 | $infodb_file_path =~ s/\.jdb$//;
|
---|
| 241 |
|
---|
| 242 | my $cmd = "java -cp \"$classpath\" Txt2Jdb -append \"$infodb_file_path\"";
|
---|
| 243 |
|
---|
| 244 | my $status = undef;
|
---|
| 245 | if(!open(GOUT, "| $cmd"))
|
---|
| 246 | {
|
---|
| 247 | print STDERR "Error: jdbm::set_infodb_entry() failed to open pipe to: $cmd\n";
|
---|
| 248 | print STDERR " $!\n";
|
---|
| 249 | $status = -1;
|
---|
| 250 | }
|
---|
| 251 | else {
|
---|
| 252 | binmode(GOUT,":utf8");
|
---|
| 253 |
|
---|
| 254 | print GOUT "[$infodb_key]\n";
|
---|
| 255 | print GOUT "$serialized_infodb_map\n";
|
---|
| 256 |
|
---|
| 257 | close(GOUT);
|
---|
| 258 | $status = 0; # as in exit status of cmd OK
|
---|
| 259 | }
|
---|
| 260 |
|
---|
| 261 | return $status;
|
---|
[23399] | 262 | }
|
---|
| 263 |
|
---|
| 264 |
|
---|
[28108] | 265 |
|
---|
| 266 |
|
---|
[21411] | 267 | sub delete_infodb_entry
|
---|
| 268 | {
|
---|
| 269 | my $infodb_handle = shift(@_);
|
---|
| 270 | my $infodb_key = shift(@_);
|
---|
| 271 |
|
---|
| 272 | # A minus at the end of a key (after the ]) signifies 'delete'
|
---|
| 273 | print $infodb_handle "[$infodb_key]-\n";
|
---|
| 274 |
|
---|
| 275 | # The 70 minus signs are also needed, to help make the parsing by db2txt simple
|
---|
| 276 | print $infodb_handle '-' x 70, "\n";
|
---|
| 277 | }
|
---|
| 278 |
|
---|
| 279 |
|
---|
| 280 |
|
---|
| 281 |
|
---|
| 282 |
|
---|
| 283 | 1;
|
---|