[15699] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # dbutil.pm -- utility functions for writing to different databases
|
---|
| 4 | # Copyright (C) 2008 DL Consulting Ltd
|
---|
| 5 | #
|
---|
| 6 | # A component of the Greenstone digital library software
|
---|
| 7 | # from the New Zealand Digital Library Project at the
|
---|
| 8 | # University of Waikato, New Zealand.
|
---|
| 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 | package dbutil;
|
---|
| 27 |
|
---|
| 28 | use strict;
|
---|
| 29 |
|
---|
| 30 |
|
---|
[16176] | 31 | sub close_infodb_write_handle
|
---|
| 32 | {
|
---|
| 33 | my $infodb_type = shift(@_);
|
---|
| 34 | my $infodb_handle = shift(@_);
|
---|
| 35 |
|
---|
| 36 | if ($infodb_type eq "sqlite")
|
---|
| 37 | {
|
---|
| 38 | return &close_infodb_write_handle_sqlite($infodb_handle);
|
---|
| 39 | }
|
---|
| 40 |
|
---|
| 41 | # Use GDBM if the infodb type is empty or not one of the values above
|
---|
| 42 | return &close_infodb_write_handle_gdbm($infodb_handle);
|
---|
| 43 | }
|
---|
| 44 |
|
---|
| 45 |
|
---|
[15727] | 46 | sub get_default_infodb_type
|
---|
| 47 | {
|
---|
| 48 | return "gdbm";
|
---|
| 49 | }
|
---|
| 50 |
|
---|
| 51 |
|
---|
[15710] | 52 | sub get_infodb_file_path
|
---|
| 53 | {
|
---|
[15725] | 54 | my $infodb_type = shift(@_);
|
---|
[15710] | 55 | my $collection_name = shift(@_);
|
---|
| 56 | my $infodb_directory_path = shift(@_);
|
---|
| 57 |
|
---|
[15725] | 58 | if ($infodb_type eq "sqlite")
|
---|
| 59 | {
|
---|
| 60 | return &get_infodb_file_path_sqlite($collection_name, $infodb_directory_path);
|
---|
| 61 | }
|
---|
| 62 |
|
---|
| 63 | # Use GDBM if the infodb type is empty or not one of the values above
|
---|
[15710] | 64 | return &get_infodb_file_path_gdbm($collection_name, $infodb_directory_path);
|
---|
| 65 | }
|
---|
| 66 |
|
---|
| 67 |
|
---|
[15711] | 68 | sub open_infodb_write_handle
|
---|
| 69 | {
|
---|
[15725] | 70 | my $infodb_type = shift(@_);
|
---|
[15711] | 71 | my $infodb_file_path = shift(@_);
|
---|
| 72 |
|
---|
[15725] | 73 | if ($infodb_type eq "sqlite")
|
---|
| 74 | {
|
---|
| 75 | return &open_infodb_write_handle_sqlite($infodb_file_path);
|
---|
| 76 | }
|
---|
| 77 |
|
---|
| 78 | # Use GDBM if the infodb type is empty or not one of the values above
|
---|
[15711] | 79 | return &open_infodb_write_handle_gdbm($infodb_file_path);
|
---|
| 80 | }
|
---|
| 81 |
|
---|
| 82 |
|
---|
[15705] | 83 | sub read_infodb_file
|
---|
| 84 | {
|
---|
[15725] | 85 | my $infodb_type = shift(@_);
|
---|
[15705] | 86 | my $infodb_file_path = shift(@_);
|
---|
| 87 | my $infodb_map = shift(@_);
|
---|
| 88 |
|
---|
[15746] | 89 | if ($infodb_type eq "sqlite")
|
---|
| 90 | {
|
---|
| 91 | return &read_infodb_file_sqlite($infodb_file_path, $infodb_map);
|
---|
| 92 | }
|
---|
| 93 |
|
---|
| 94 | # Use GDBM if the infodb type is empty or not one of the values above
|
---|
| 95 | return &read_infodb_file_gdbm($infodb_file_path, $infodb_map);
|
---|
[15705] | 96 | }
|
---|
| 97 |
|
---|
| 98 |
|
---|
[15699] | 99 | sub write_infodb_entry
|
---|
| 100 | {
|
---|
[15725] | 101 | my $infodb_type = shift(@_);
|
---|
[15699] | 102 | my $infodb_handle = shift(@_);
|
---|
| 103 | my $infodb_key = shift(@_);
|
---|
| 104 | my $infodb_map = shift(@_);
|
---|
| 105 |
|
---|
[15725] | 106 | if ($infodb_type eq "sqlite")
|
---|
| 107 | {
|
---|
| 108 | return &write_infodb_entry_sqlite($infodb_handle, $infodb_key, $infodb_map);
|
---|
| 109 | }
|
---|
| 110 |
|
---|
| 111 | # Use GDBM if the infodb type is empty or not one of the values above
|
---|
| 112 | return &write_infodb_entry_gdbm($infodb_handle, $infodb_key, $infodb_map);
|
---|
[15699] | 113 | }
|
---|
| 114 |
|
---|
[15705] | 115 |
|
---|
| 116 |
|
---|
| 117 | # ----------------------------------------------------------------------------------------
|
---|
[15746] | 118 | # GDBM IMPLEMENTATION
|
---|
[15705] | 119 | # ----------------------------------------------------------------------------------------
|
---|
| 120 |
|
---|
[16176] | 121 | sub close_infodb_write_handle_gdbm
|
---|
| 122 | {
|
---|
| 123 | my $infodb_handle = shift(@_);
|
---|
| 124 |
|
---|
| 125 | close($infodb_handle);
|
---|
| 126 | }
|
---|
| 127 |
|
---|
| 128 |
|
---|
[15710] | 129 | sub get_infodb_file_path_gdbm
|
---|
| 130 | {
|
---|
| 131 | my $collection_name = shift(@_);
|
---|
| 132 | my $infodb_directory_path = shift(@_);
|
---|
| 133 |
|
---|
| 134 | my $infodb_file_extension = (&util::is_little_endian() ? ".ldb" : ".bdb");
|
---|
| 135 | my $infodb_file_name = &util::get_dirsep_tail($collection_name) . $infodb_file_extension;
|
---|
| 136 | return &util::filename_cat($infodb_directory_path, $infodb_file_name);
|
---|
| 137 | }
|
---|
| 138 |
|
---|
| 139 |
|
---|
[15711] | 140 | sub open_infodb_write_handle_gdbm
|
---|
| 141 | {
|
---|
| 142 | my $infodb_file_path = shift(@_);
|
---|
| 143 |
|
---|
| 144 | my $txt2db_exe = &util::filename_cat("$ENV{'GSDLHOME'}/bin/$ENV{'GSDLOS'}", "txt2db" . &util::get_os_exe());
|
---|
[15717] | 145 | my $infodb_file_handle = undef;
|
---|
[16339] | 146 | if (!-e "$txt2db_exe" || !open($infodb_file_handle, "| \"$txt2db_exe\" \"$infodb_file_path\""))
|
---|
[15711] | 147 | {
|
---|
| 148 | return undef;
|
---|
| 149 | }
|
---|
| 150 |
|
---|
[15717] | 151 | return $infodb_file_handle;
|
---|
[15711] | 152 | }
|
---|
| 153 |
|
---|
| 154 |
|
---|
[15705] | 155 | sub read_infodb_file_gdbm
|
---|
| 156 | {
|
---|
| 157 | my $infodb_file_path = shift(@_);
|
---|
| 158 | my $infodb_map = shift(@_);
|
---|
| 159 |
|
---|
| 160 | open (PIPEIN, "db2txt \"$infodb_file_path\" |") || die "couldn't open pipe from db2txt\n";
|
---|
| 161 | my $infodb_line = "";
|
---|
| 162 | my $infodb_key = "";
|
---|
| 163 | my $infodb_value = "";
|
---|
| 164 | while (defined ($infodb_line = <PIPEIN>))
|
---|
| 165 | {
|
---|
| 166 | if ($infodb_line =~ /^\[([^\]]+)\]$/)
|
---|
| 167 | {
|
---|
| 168 | $infodb_key = $1;
|
---|
| 169 | }
|
---|
| 170 | elsif ($infodb_line =~ /^-{70}$/)
|
---|
| 171 | {
|
---|
| 172 | $infodb_map->{$infodb_key} = $infodb_value;
|
---|
| 173 | $infodb_key = "";
|
---|
| 174 | $infodb_value = "";
|
---|
| 175 | }
|
---|
| 176 | else
|
---|
| 177 | {
|
---|
| 178 | $infodb_value .= $infodb_line;
|
---|
| 179 | }
|
---|
| 180 | }
|
---|
| 181 |
|
---|
| 182 | close (PIPEIN);
|
---|
| 183 | }
|
---|
| 184 |
|
---|
[15699] | 185 |
|
---|
| 186 | sub write_infodb_entry_gdbm
|
---|
| 187 | {
|
---|
| 188 | my $infodb_handle = shift(@_);
|
---|
| 189 | my $infodb_key = shift(@_);
|
---|
| 190 | my $infodb_map = shift(@_);
|
---|
| 191 |
|
---|
| 192 | print $infodb_handle "[$infodb_key]\n";
|
---|
| 193 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
| 194 | {
|
---|
| 195 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
| 196 | {
|
---|
| 197 | if ($infodb_value =~ /-{70,}/)
|
---|
| 198 | {
|
---|
| 199 | # if value contains 70 or more hyphens in a row we need to escape them
|
---|
| 200 | # to prevent txt2db from treating them as a separator
|
---|
| 201 | $infodb_value =~ s/-/&\#045;/gi;
|
---|
| 202 | }
|
---|
| 203 | print $infodb_handle "<$infodb_value_key>" . $infodb_value . "\n";
|
---|
| 204 | }
|
---|
| 205 | }
|
---|
| 206 | print $infodb_handle '-' x 70, "\n";
|
---|
| 207 | }
|
---|
| 208 |
|
---|
| 209 |
|
---|
[15722] | 210 |
|
---|
| 211 | # ----------------------------------------------------------------------------------------
|
---|
[15746] | 212 | # SQLITE IMPLEMENTATION
|
---|
[15722] | 213 | # ----------------------------------------------------------------------------------------
|
---|
| 214 |
|
---|
[16176] | 215 | sub close_infodb_write_handle_sqlite
|
---|
| 216 | {
|
---|
| 217 | my $infodb_handle = shift(@_);
|
---|
| 218 |
|
---|
[16223] | 219 | # Close the transaction we began after opening the file
|
---|
[16178] | 220 | print $infodb_handle "END TRANSACTION;\n";
|
---|
| 221 |
|
---|
[16223] | 222 | # This is crucial for efficient queries on the database!
|
---|
[16224] | 223 | print $infodb_handle "CREATE INDEX IF NOT EXISTS dme ON document_metadata(element);\n";
|
---|
[16223] | 224 |
|
---|
[16176] | 225 | close($infodb_handle);
|
---|
| 226 | }
|
---|
| 227 |
|
---|
| 228 |
|
---|
[15722] | 229 | sub get_infodb_file_path_sqlite
|
---|
| 230 | {
|
---|
| 231 | my $collection_name = shift(@_);
|
---|
| 232 | my $infodb_directory_path = shift(@_);
|
---|
| 233 |
|
---|
| 234 | my $infodb_file_extension = ".db";
|
---|
| 235 | my $infodb_file_name = &util::get_dirsep_tail($collection_name) . $infodb_file_extension;
|
---|
| 236 | return &util::filename_cat($infodb_directory_path, $infodb_file_name);
|
---|
| 237 | }
|
---|
| 238 |
|
---|
| 239 |
|
---|
| 240 | sub open_infodb_write_handle_sqlite
|
---|
| 241 | {
|
---|
| 242 | my $infodb_file_path = shift(@_);
|
---|
| 243 |
|
---|
| 244 | my $sqlite3_exe = &util::filename_cat("$ENV{'GSDLHOME'}/bin/$ENV{'GSDLOS'}", "sqlite3" . &util::get_os_exe());
|
---|
[16177] | 245 | my $infodb_handle = undef;
|
---|
| 246 | if (!-e "$sqlite3_exe" || !open($infodb_handle, "| $sqlite3_exe \"$infodb_file_path\""))
|
---|
[15722] | 247 | {
|
---|
| 248 | return undef;
|
---|
| 249 | }
|
---|
| 250 |
|
---|
[16225] | 251 | print $infodb_handle "CREATE TABLE IF NOT EXISTS data (key TEXT PRIMARY KEY, value TEXT);\n";
|
---|
| 252 | print $infodb_handle "CREATE TABLE IF NOT EXISTS document_metadata (id INTEGER PRIMARY KEY, docOID TEXT, element TEXT, value TEXT);\n";
|
---|
[15722] | 253 |
|
---|
[16240] | 254 | # This is crucial for efficiency when importing large amounts of data
|
---|
| 255 | print $infodb_handle "CREATE INDEX IF NOT EXISTS dmd ON document_metadata(docOID);\n";
|
---|
| 256 |
|
---|
[16178] | 257 | # This is very important for efficiency, otherwise each command will be actioned one at a time
|
---|
| 258 | print $infodb_handle "BEGIN TRANSACTION;\n";
|
---|
| 259 |
|
---|
[16177] | 260 | return $infodb_handle;
|
---|
[15722] | 261 | }
|
---|
| 262 |
|
---|
| 263 |
|
---|
[15746] | 264 | sub read_infodb_file_sqlite
|
---|
| 265 | {
|
---|
| 266 | my $infodb_file_path = shift(@_);
|
---|
| 267 | my $infodb_map = shift(@_);
|
---|
| 268 |
|
---|
| 269 | # !! TO IMPLEMENT
|
---|
| 270 | }
|
---|
| 271 |
|
---|
| 272 |
|
---|
[15722] | 273 | sub write_infodb_entry_sqlite
|
---|
| 274 | {
|
---|
| 275 | my $infodb_handle = shift(@_);
|
---|
| 276 | my $infodb_key = shift(@_);
|
---|
| 277 | my $infodb_map = shift(@_);
|
---|
| 278 |
|
---|
[15747] | 279 | # Add the key -> value mapping into the "data" table
|
---|
[15722] | 280 | my $infodb_entry_value = "";
|
---|
| 281 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
| 282 | {
|
---|
| 283 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
| 284 | {
|
---|
| 285 | $infodb_entry_value .= "<$infodb_value_key>" . $infodb_value . "\n";
|
---|
| 286 | }
|
---|
| 287 | }
|
---|
| 288 |
|
---|
[15748] | 289 | my $safe_infodb_key = &sqlite_safe($infodb_key);
|
---|
[16226] | 290 | print $infodb_handle "INSERT OR REPLACE INTO data (key, value) VALUES ('" . $safe_infodb_key . "', '" . &sqlite_safe($infodb_entry_value) . "');\n";
|
---|
[15748] | 291 |
|
---|
| 292 | # If this infodb entry is for a document, add all the interesting document metadata to the
|
---|
| 293 | # "document_metadata" table (for use by the dynamic classifiers)
|
---|
| 294 | if ($infodb_key !~ /\./ && $infodb_entry_value =~ /\<doctype\>doc\n/)
|
---|
| 295 | {
|
---|
| 296 | print $infodb_handle "DELETE FROM document_metadata WHERE docOID='" . $safe_infodb_key . "';\n";
|
---|
| 297 |
|
---|
| 298 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
| 299 | {
|
---|
| 300 | # We're not interested in most of the automatically added document metadata
|
---|
| 301 | next if ($infodb_value_key eq "archivedir" ||
|
---|
| 302 | $infodb_value_key eq "assocfilepath" ||
|
---|
| 303 | $infodb_value_key eq "childtype" ||
|
---|
| 304 | $infodb_value_key eq "contains" ||
|
---|
| 305 | $infodb_value_key eq "docnum" ||
|
---|
| 306 | $infodb_value_key eq "doctype" ||
|
---|
| 307 | $infodb_value_key eq "Encoding" ||
|
---|
| 308 | $infodb_value_key eq "FileSize" ||
|
---|
| 309 | $infodb_value_key eq "hascover" ||
|
---|
| 310 | $infodb_value_key eq "hastxt" ||
|
---|
| 311 | $infodb_value_key eq "lastmodified" ||
|
---|
| 312 | $infodb_value_key eq "metadataset" ||
|
---|
| 313 | $infodb_value_key eq "thistype" ||
|
---|
| 314 | $infodb_value_key =~ /^metadatafreq\-/ ||
|
---|
| 315 | $infodb_value_key =~ /^metadatalist\-/);
|
---|
| 316 |
|
---|
| 317 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
| 318 | {
|
---|
| 319 | print $infodb_handle "INSERT INTO document_metadata (docOID, element, value) VALUES ('" . $safe_infodb_key . "', '" . &sqlite_safe($infodb_value_key) . "', '" . &sqlite_safe($infodb_value) . "');\n";
|
---|
| 320 | }
|
---|
| 321 | }
|
---|
| 322 | }
|
---|
[15747] | 323 | }
|
---|
| 324 |
|
---|
| 325 |
|
---|
| 326 | sub sqlite_safe
|
---|
| 327 | {
|
---|
| 328 | my $value = shift(@_);
|
---|
| 329 |
|
---|
[15722] | 330 | # Escape any single quotes in the value
|
---|
[15747] | 331 | $value =~ s/\'/\'\'/g;
|
---|
[15722] | 332 |
|
---|
[15747] | 333 | return $value;
|
---|
[15722] | 334 | }
|
---|
| 335 |
|
---|
| 336 |
|
---|
[15699] | 337 | 1;
|
---|