| 1 | ##########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # basebuildproc.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 | # This document processor outputs a document for indexing (should be
|
|---|
| 27 | # implemented by subclass) and storing in the database
|
|---|
| 28 |
|
|---|
| 29 | package basebuildproc;
|
|---|
| 30 |
|
|---|
| 31 | eval {require bytes};
|
|---|
| 32 |
|
|---|
| 33 | use classify;
|
|---|
| 34 | use dbutil;
|
|---|
| 35 | use doc;
|
|---|
| 36 | use docproc;
|
|---|
| 37 | use strict;
|
|---|
| 38 | no strict 'subs';
|
|---|
| 39 | no strict 'refs';
|
|---|
| 40 | use util;
|
|---|
| 41 | use FileUtils;
|
|---|
| 42 |
|
|---|
| 43 | BEGIN {
|
|---|
| 44 | @basebuildproc::ISA = ('docproc');
|
|---|
| 45 | }
|
|---|
| 46 |
|
|---|
| 47 | sub new()
|
|---|
| 48 | {
|
|---|
| 49 | my ($class, $collection, $source_dir, $build_dir, $keepold, $verbosity, $outhandle) = @_;
|
|---|
| 50 | my $self = new docproc ();
|
|---|
| 51 |
|
|---|
| 52 | # outhandle is where all the debugging info goes
|
|---|
| 53 | # output_handle is where the output of the plugins is piped
|
|---|
| 54 | # to (i.e. mg, database etc.)
|
|---|
| 55 | $outhandle = STDERR unless defined $outhandle;
|
|---|
| 56 |
|
|---|
| 57 | $self->{'collection'} = $collection;
|
|---|
| 58 | $self->{'source_dir'} = $source_dir;
|
|---|
| 59 | $self->{'build_dir'} = $build_dir;
|
|---|
| 60 | $self->{'keepold'} = $keepold;
|
|---|
| 61 | $self->{'verbosity'} = $verbosity;
|
|---|
| 62 | $self->{'outhandle'} = $outhandle;
|
|---|
| 63 |
|
|---|
| 64 | $self->{'classifiers'} = [];
|
|---|
| 65 | $self->{'mode'} = "text";
|
|---|
| 66 | $self->{'assocdir'} = $build_dir;
|
|---|
| 67 | $self->{'dontdb'} = {};
|
|---|
| 68 | $self->{'store_metadata_coverage'} = "false";
|
|---|
| 69 |
|
|---|
| 70 | $self->{'index'} = "section:text";
|
|---|
| 71 | $self->{'indexexparr'} = [];
|
|---|
| 72 |
|
|---|
| 73 | $self->{'separate_cjk'} = 0;
|
|---|
| 74 |
|
|---|
| 75 | my $found_num_data = 0;
|
|---|
| 76 | my $buildconfigfile = undef;
|
|---|
| 77 |
|
|---|
| 78 | if ($keepold) {
|
|---|
| 79 | # For incremental building need to seed num_docs etc from values
|
|---|
| 80 | # stored in build.cfg (if present)
|
|---|
| 81 | $buildconfigfile = &FileUtils::filenameConcatenate($build_dir, "build.cfg");
|
|---|
| 82 | if (-e $buildconfigfile) {
|
|---|
| 83 | $found_num_data = 1;
|
|---|
| 84 | }
|
|---|
| 85 | else {
|
|---|
| 86 | # try the index dir
|
|---|
| 87 | $buildconfigfile = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'},
|
|---|
| 88 | "index", "build.cfg");
|
|---|
| 89 | if (-e $buildconfigfile) {
|
|---|
| 90 | $found_num_data = 1;
|
|---|
| 91 | }
|
|---|
| 92 | }
|
|---|
| 93 |
|
|---|
| 94 | }
|
|---|
| 95 |
|
|---|
| 96 | if ($found_num_data)
|
|---|
| 97 | {
|
|---|
| 98 | #print STDERR "Found_Num_Data!\n";
|
|---|
| 99 | my $buildcfg = &colcfg::read_build_cfg($buildconfigfile);
|
|---|
| 100 | $self->{'starting_num_docs'} = $buildcfg->{'numdocs'};
|
|---|
| 101 | #print STDERR "- num_docs: $self->{'starting_num_docs'}\n";
|
|---|
| 102 | $self->{'starting_num_sections'} = $buildcfg->{'numsections'};
|
|---|
| 103 | #print STDERR "- num_sections: $self->{'starting_num_sections'}\n";
|
|---|
| 104 | $self->{'starting_num_bytes'} = $buildcfg->{'numbytes'};
|
|---|
| 105 | #print STDERR "- num_bytes: $self->{'starting_num_bytes'}\n";
|
|---|
| 106 | }
|
|---|
| 107 | else
|
|---|
| 108 | {
|
|---|
| 109 | #print STDERR "NOT Found_Num_Data!\n";
|
|---|
| 110 | $self->{'starting_num_docs'} = 0;
|
|---|
| 111 | $self->{'starting_num_sections'} = 0;
|
|---|
| 112 | $self->{'starting_num_bytes'} = 0;
|
|---|
| 113 | }
|
|---|
| 114 |
|
|---|
| 115 | $self->{'output_handle'} = "STDOUT";
|
|---|
| 116 | $self->{'num_docs'} = $self->{'starting_num_docs'};
|
|---|
| 117 | $self->{'num_sections'} = $self->{'starting_num_sections'};
|
|---|
| 118 | $self->{'num_bytes'} = $self->{'starting_num_bytes'};
|
|---|
| 119 |
|
|---|
| 120 | $self->{'num_processed_bytes'} = 0;
|
|---|
| 121 | $self->{'store_text'} = 1;
|
|---|
| 122 |
|
|---|
| 123 | # what level (section/document) the database - indexer intersection is
|
|---|
| 124 | $self->{'db_level'} = "section";
|
|---|
| 125 | #used by browse interface
|
|---|
| 126 | $self->{'doclist'} = [];
|
|---|
| 127 |
|
|---|
| 128 | $self->{'indexing_text'} = 0;
|
|---|
| 129 |
|
|---|
| 130 | return bless $self, $class;
|
|---|
| 131 |
|
|---|
| 132 | }
|
|---|
| 133 |
|
|---|
| 134 | sub reset {
|
|---|
| 135 | my $self = shift (@_);
|
|---|
| 136 |
|
|---|
| 137 | $self->{'num_docs'} = $self->{'starting_num_docs'};
|
|---|
| 138 | $self->{'num_sections'} = $self->{'starting_num_sections'};
|
|---|
| 139 | $self->{'num_bytes'} = $self->{'starting_num_bytes'};
|
|---|
| 140 |
|
|---|
| 141 | $self->{'num_processed_bytes'} = 0;
|
|---|
| 142 | }
|
|---|
| 143 |
|
|---|
| 144 | sub zero_reset {
|
|---|
| 145 | my $self = shift (@_);
|
|---|
| 146 |
|
|---|
| 147 | $self->{'num_docs'} = 0;
|
|---|
| 148 | $self->{'num_sections'} = 0;
|
|---|
| 149 | # reconstructed docs have no text, just metadata, so we need to
|
|---|
| 150 | # remember how many bytes we had initially
|
|---|
| 151 | #$self->{'num_bytes'} = $self->{'starting_num_bytes'};
|
|---|
| 152 | $self->{'num_bytes'} = 0; # we'll store num bytes in db for reconstructed docs.
|
|---|
| 153 | $self->{'num_processed_bytes'} = 0;
|
|---|
| 154 | }
|
|---|
| 155 |
|
|---|
| 156 | sub is_incremental_capable
|
|---|
| 157 | {
|
|---|
| 158 | # By default we return 'no' as the answer
|
|---|
| 159 | # Safer to assume non-incremental to start with, and then override in
|
|---|
| 160 | # inherited classes that are.
|
|---|
| 161 |
|
|---|
| 162 | return 0;
|
|---|
| 163 | }
|
|---|
| 164 |
|
|---|
| 165 | sub get_num_docs {
|
|---|
| 166 | my $self = shift (@_);
|
|---|
| 167 |
|
|---|
| 168 | return $self->{'num_docs'};
|
|---|
| 169 | }
|
|---|
| 170 |
|
|---|
| 171 | sub get_num_sections {
|
|---|
| 172 | my $self = shift (@_);
|
|---|
| 173 |
|
|---|
| 174 | return $self->{'num_sections'};
|
|---|
| 175 | }
|
|---|
| 176 |
|
|---|
| 177 | # num_bytes is the actual number of bytes in the collection
|
|---|
| 178 | # this is normally the same as what's processed during text compression
|
|---|
| 179 | sub get_num_bytes {
|
|---|
| 180 | my $self = shift (@_);
|
|---|
| 181 |
|
|---|
| 182 | return $self->{'num_bytes'};
|
|---|
| 183 | }
|
|---|
| 184 |
|
|---|
| 185 | # num_processed_bytes is the number of bytes actually passed
|
|---|
| 186 | # to mg for the current index
|
|---|
| 187 | sub get_num_processed_bytes {
|
|---|
| 188 | my $self = shift (@_);
|
|---|
| 189 |
|
|---|
| 190 | return $self->{'num_processed_bytes'};
|
|---|
| 191 | }
|
|---|
| 192 |
|
|---|
| 193 | sub set_output_handle {
|
|---|
| 194 | my $self = shift (@_);
|
|---|
| 195 | my ($handle) = @_;
|
|---|
| 196 |
|
|---|
| 197 | $self->{'output_handle'} = $handle;
|
|---|
| 198 | # The output handle isn't always an actual handle. In a couple of the
|
|---|
| 199 | # database drivers (MSSQL and GDBMServer) it's actually a reference
|
|---|
| 200 | # to an object. Thus we need to test the type before setting binmode.
|
|---|
| 201 | # [jmt12]
|
|---|
| 202 | if (ref $handle eq "GLOB")
|
|---|
| 203 | {
|
|---|
| 204 | binmode($handle,":utf8");
|
|---|
| 205 | }
|
|---|
| 206 | }
|
|---|
| 207 |
|
|---|
| 208 |
|
|---|
| 209 | sub set_mode {
|
|---|
| 210 | my $self = shift (@_);
|
|---|
| 211 | my ($mode) = @_;
|
|---|
| 212 |
|
|---|
| 213 | $self->{'mode'} = $mode;
|
|---|
| 214 | $doc::processor_mode = $mode; # doc.pm needs to know what buildcol pass we're at
|
|---|
| 215 | }
|
|---|
| 216 |
|
|---|
| 217 | sub get_mode {
|
|---|
| 218 | my $self = shift (@_);
|
|---|
| 219 |
|
|---|
| 220 | return $self->{'mode'};
|
|---|
| 221 | }
|
|---|
| 222 |
|
|---|
| 223 | sub set_assocdir {
|
|---|
| 224 | my $self = shift (@_);
|
|---|
| 225 | my ($assocdir) = @_;
|
|---|
| 226 |
|
|---|
| 227 | $self->{'assocdir'} = $assocdir;
|
|---|
| 228 | }
|
|---|
| 229 |
|
|---|
| 230 | sub set_dontdb {
|
|---|
| 231 | my $self = shift (@_);
|
|---|
| 232 | my ($dontdb) = @_;
|
|---|
| 233 |
|
|---|
| 234 | $self->{'dontdb'} = $dontdb;
|
|---|
| 235 | }
|
|---|
| 236 |
|
|---|
| 237 | sub set_infodbtype
|
|---|
| 238 | {
|
|---|
| 239 | my $self = shift(@_);
|
|---|
| 240 | my $infodbtype = shift(@_);
|
|---|
| 241 | $self->{'infodbtype'} = $infodbtype;
|
|---|
| 242 | }
|
|---|
| 243 |
|
|---|
| 244 | sub set_site
|
|---|
| 245 | {
|
|---|
| 246 | my $self = shift(@_);
|
|---|
| 247 | my $site = shift(@_);
|
|---|
| 248 | $self->{'site'} = $site;
|
|---|
| 249 | }
|
|---|
| 250 |
|
|---|
| 251 | sub set_index {
|
|---|
| 252 | my $self = shift (@_);
|
|---|
| 253 | my ($index, $indexexparr) = @_;
|
|---|
| 254 |
|
|---|
| 255 | $self->{'index'} = $index;
|
|---|
| 256 | $self->{'indexexparr'} = $indexexparr if defined $indexexparr;
|
|---|
| 257 | }
|
|---|
| 258 |
|
|---|
| 259 | sub set_index_languages {
|
|---|
| 260 | my $self = shift (@_);
|
|---|
| 261 | my ($lang_meta, $langarr) = @_;
|
|---|
| 262 | $lang_meta =~ s/^ex\.([^.]+)$/$1/; # strip any ex. namespace iff it's the only namespace prefix (will leave ex.dc.* intact)
|
|---|
| 263 |
|
|---|
| 264 | $self->{'lang_meta'} = $lang_meta;
|
|---|
| 265 | $self->{'langarr'} = $langarr;
|
|---|
| 266 | }
|
|---|
| 267 |
|
|---|
| 268 | sub get_index {
|
|---|
| 269 | my $self = shift (@_);
|
|---|
| 270 |
|
|---|
| 271 | return $self->{'index'};
|
|---|
| 272 | }
|
|---|
| 273 |
|
|---|
| 274 | sub set_classifiers {
|
|---|
| 275 | my $self = shift (@_);
|
|---|
| 276 | my ($classifiers) = @_;
|
|---|
| 277 |
|
|---|
| 278 | $self->{'classifiers'} = $classifiers;
|
|---|
| 279 | }
|
|---|
| 280 |
|
|---|
| 281 | sub set_indexing_text {
|
|---|
| 282 | my $self = shift (@_);
|
|---|
| 283 | my ($indexing_text) = @_;
|
|---|
| 284 |
|
|---|
| 285 | $self->{'indexing_text'} = $indexing_text;
|
|---|
| 286 | }
|
|---|
| 287 |
|
|---|
| 288 | sub get_indexing_text {
|
|---|
| 289 | my $self = shift (@_);
|
|---|
| 290 |
|
|---|
| 291 | return $self->{'indexing_text'};
|
|---|
| 292 | }
|
|---|
| 293 |
|
|---|
| 294 | sub set_store_text {
|
|---|
| 295 | my $self = shift (@_);
|
|---|
| 296 | my ($store_text) = @_;
|
|---|
| 297 |
|
|---|
| 298 | $self->{'store_text'} = $store_text;
|
|---|
| 299 | }
|
|---|
| 300 |
|
|---|
| 301 | sub set_store_metadata_coverage {
|
|---|
| 302 | my $self = shift (@_);
|
|---|
| 303 | my ($store_metadata_coverage) = @_;
|
|---|
| 304 |
|
|---|
| 305 | $self->{'store_metadata_coverage'} = $store_metadata_coverage || "";
|
|---|
| 306 | }
|
|---|
| 307 |
|
|---|
| 308 | sub get_doc_list {
|
|---|
| 309 | my $self = shift(@_);
|
|---|
| 310 |
|
|---|
| 311 | return @{$self->{'doclist'}};
|
|---|
| 312 | }
|
|---|
| 313 |
|
|---|
| 314 | # the standard database level is section, but you may want to change it to document
|
|---|
| 315 | sub set_db_level {
|
|---|
| 316 | my $self= shift (@_);
|
|---|
| 317 | my ($db_level) = @_;
|
|---|
| 318 |
|
|---|
| 319 | $self->{'db_level'} = $db_level;
|
|---|
| 320 | }
|
|---|
| 321 |
|
|---|
| 322 | sub set_sections_index_document_metadata {
|
|---|
| 323 | my $self= shift (@_);
|
|---|
| 324 | my ($index_type) = @_;
|
|---|
| 325 |
|
|---|
| 326 | $self->{'sections_index_document_metadata'} = $index_type;
|
|---|
| 327 | }
|
|---|
| 328 |
|
|---|
| 329 | sub set_separate_cjk {
|
|---|
| 330 | my $self = shift (@_);
|
|---|
| 331 | my ($sep_cjk) = @_;
|
|---|
| 332 |
|
|---|
| 333 | $self->{'separate_cjk'} = $sep_cjk;
|
|---|
| 334 | }
|
|---|
| 335 |
|
|---|
| 336 | sub process {
|
|---|
| 337 | my $self = shift (@_);
|
|---|
| 338 | my $method = $self->{'mode'};
|
|---|
| 339 |
|
|---|
| 340 | $self->$method(@_);
|
|---|
| 341 | }
|
|---|
| 342 |
|
|---|
| 343 | # post process text depending on field. Currently don't do anything here
|
|---|
| 344 | # except cjk separation, and only for indexing
|
|---|
| 345 | # should only do this for indexed text (if $self->{'indexing_text'}),
|
|---|
| 346 | # but currently search term highlighting doesn't work if you do that.
|
|---|
| 347 | # once thats fixed up, then fix this.
|
|---|
| 348 | sub filter_text {
|
|---|
| 349 | my $self = shift (@_);
|
|---|
| 350 | my ($field, $text) = @_;
|
|---|
| 351 |
|
|---|
| 352 | # lets do cjk seg here
|
|---|
| 353 | my $new_text =$text;
|
|---|
| 354 | if ($self->{'separate_cjk'}) {
|
|---|
| 355 | $new_text = &cnseg::segment($text);
|
|---|
| 356 | }
|
|---|
| 357 | return $new_text;
|
|---|
| 358 | }
|
|---|
| 359 |
|
|---|
| 360 |
|
|---|
| 361 | sub infodb_metadata_stats
|
|---|
| 362 | {
|
|---|
| 363 | my $self = shift (@_);
|
|---|
| 364 | my ($field,$edit_mode) = @_;
|
|---|
| 365 |
|
|---|
| 366 | # Keep some statistics relating to metadata sets used and
|
|---|
| 367 | # frequency of particular metadata fields within each set
|
|---|
| 368 |
|
|---|
| 369 | # Union of metadata prefixes and frequency of fields
|
|---|
| 370 | # (both scoped for this document alone, and across whole collection)
|
|---|
| 371 |
|
|---|
| 372 | if ($field =~ m/^(.+)\.(.*)$/) {
|
|---|
| 373 | my $prefix = $1;
|
|---|
| 374 | my $core_field = $2;
|
|---|
| 375 |
|
|---|
| 376 | if (($edit_mode eq "add") || ($edit_mode eq "update")) {
|
|---|
| 377 | $self->{'doc_mdprefix_fields'}->{$prefix}->{$core_field}++;
|
|---|
| 378 | $self->{'mdprefix_fields'}->{$prefix}->{$core_field}++;
|
|---|
| 379 | }
|
|---|
| 380 | else {
|
|---|
| 381 | # delete
|
|---|
| 382 | $self->{'doc_mdprefix_fields'}->{$prefix}->{$core_field}--;
|
|---|
| 383 | $self->{'mdprefix_fields'}->{$prefix}->{$core_field}--;
|
|---|
| 384 | }
|
|---|
| 385 |
|
|---|
| 386 | }
|
|---|
| 387 | elsif ($field =~ m/^[[:upper:]]/) {
|
|---|
| 388 | # implicit 'ex' metadata set
|
|---|
| 389 |
|
|---|
| 390 | if (($edit_mode eq "add") || ($edit_mode eq "update")) {
|
|---|
| 391 |
|
|---|
| 392 | $self->{'doc_mdprefix_fields'}->{'ex'}->{$field}++;
|
|---|
| 393 | $self->{'mdprefix_fields'}->{'ex'}->{$field}++;
|
|---|
| 394 | }
|
|---|
| 395 | else {
|
|---|
| 396 | # delete
|
|---|
| 397 | $self->{'doc_mdprefix_fields'}->{'ex'}->{$field}--;
|
|---|
| 398 | $self->{'mdprefix_fields'}->{'ex'}->{$field}--;
|
|---|
| 399 | }
|
|---|
| 400 | }
|
|---|
| 401 |
|
|---|
| 402 | }
|
|---|
| 403 |
|
|---|
| 404 |
|
|---|
| 405 | sub infodbedit {
|
|---|
| 406 | my $self = shift (@_);
|
|---|
| 407 | my ($doc_obj, $filename, $edit_mode) = @_;
|
|---|
| 408 |
|
|---|
| 409 | # print STDERR "**** basebuildproc::infodbedit(): $filename, $edit_mode\n";
|
|---|
| 410 |
|
|---|
| 411 | # only output this document if it is a "indexed_doc" or "info_doc" (database only) document
|
|---|
| 412 | my $doctype = $doc_obj->get_doc_type();
|
|---|
| 413 | return if ($doctype ne "indexed_doc" && $doctype ne "info_doc");
|
|---|
| 414 |
|
|---|
| 415 | my $archivedir = "";
|
|---|
| 416 | if (defined $filename)
|
|---|
| 417 | {
|
|---|
| 418 | # doc_obj derived directly from file
|
|---|
| 419 | my ($dir) = $filename =~ /^(.*?)(?:\/|\\)[^\/\\]*$/;
|
|---|
| 420 | $dir = "" unless defined $dir;
|
|---|
| 421 | $dir =~ s/\\/\//g;
|
|---|
| 422 | $dir =~ s/^\/+//;
|
|---|
| 423 | $dir =~ s/\/+$//;
|
|---|
| 424 |
|
|---|
| 425 | $archivedir = $dir;
|
|---|
| 426 |
|
|---|
| 427 | if ($edit_mode eq "delete") {
|
|---|
| 428 | # record this doc so we don't process the reconstructed doc later
|
|---|
| 429 | $self->{'dont_process_reconstructed'}->{$doc_obj->get_OID()} = 1;
|
|---|
| 430 | # we don't need to do anything else for the info database for a deleted document. The infodb starts from scratch each time, so no deletion is necessary
|
|---|
| 431 | $self->delete_assoc_files ($archivedir, "delete");
|
|---|
| 432 | return;
|
|---|
| 433 | }
|
|---|
| 434 | if ($edit_mode eq "update") {
|
|---|
| 435 | # we don't want to process the reconstructed doc later, but we will process this version now.
|
|---|
| 436 | $self->{'dont_process_reconstructed'}->{$doc_obj->get_OID()} = 1;
|
|---|
| 437 | # delete the old assoc files as they may have changed
|
|---|
| 438 | $self->delete_assoc_files ($archivedir, "update");
|
|---|
| 439 | }
|
|---|
| 440 |
|
|---|
| 441 | # resolve the final filenames of the files associated with this document
|
|---|
| 442 | # now save the new assoc files for an update/new doc.
|
|---|
| 443 | $self->assoc_files ($doc_obj, $archivedir);
|
|---|
| 444 | }
|
|---|
| 445 | else
|
|---|
| 446 | {
|
|---|
| 447 | # doc_obj reconstructed from database (has metadata, doc structure but no text)
|
|---|
| 448 | my $top_section = $doc_obj->get_top_section();
|
|---|
| 449 | $archivedir = $doc_obj->get_metadata_element($top_section,"archivedir");
|
|---|
| 450 | }
|
|---|
| 451 |
|
|---|
| 452 | # rest of code used for add and update. In both cases, we add to the classifiers and to the info database.
|
|---|
| 453 |
|
|---|
| 454 | #add this document to the browse structure
|
|---|
| 455 | push(@{$self->{'doclist'}},$doc_obj->get_OID())
|
|---|
| 456 | unless ($doctype eq "classification");
|
|---|
| 457 | $self->{'num_docs'} += 1 unless ($doctype eq "classification");
|
|---|
| 458 |
|
|---|
| 459 | if (!defined $filename) {
|
|---|
| 460 | # a reconstructed doc
|
|---|
| 461 | my $num_reconstructed_bytes = $doc_obj->get_metadata_element ($doc_obj->get_top_section (), "total_numbytes");
|
|---|
| 462 | if (defined $num_reconstructed_bytes) {
|
|---|
| 463 | $self->{'num_bytes'} += $num_reconstructed_bytes;
|
|---|
| 464 | }
|
|---|
| 465 | }
|
|---|
| 466 | # classify the document
|
|---|
| 467 | &classify::classify_doc ($self->{'classifiers'}, $doc_obj);
|
|---|
| 468 |
|
|---|
| 469 | # now add all the sections to the infodb.
|
|---|
| 470 |
|
|---|
| 471 | # is this a paged or a hierarchical document
|
|---|
| 472 | my ($thistype, $childtype) = $self->get_document_type ($doc_obj);
|
|---|
| 473 |
|
|---|
| 474 | my $section = $doc_obj->get_top_section ();
|
|---|
| 475 | my $doc_OID = $doc_obj->get_OID();
|
|---|
| 476 | my $first = 1;
|
|---|
| 477 | my $infodb_handle = $self->{'output_handle'};
|
|---|
| 478 |
|
|---|
| 479 | $self->{'doc_mdprefix_fields'} = {};
|
|---|
| 480 |
|
|---|
| 481 | while (defined $section)
|
|---|
| 482 | {
|
|---|
| 483 | my $section_OID = $doc_OID;
|
|---|
| 484 | if ($section ne "")
|
|---|
| 485 | {
|
|---|
| 486 | $section_OID = $doc_OID . "." . $section;
|
|---|
| 487 | }
|
|---|
| 488 | my %section_infodb = ();
|
|---|
| 489 |
|
|---|
| 490 | # update a few statistics
|
|---|
| 491 | $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
|
|---|
| 492 | $self->{'num_sections'} += 1 unless ($doctype eq "classification");
|
|---|
| 493 |
|
|---|
| 494 | # output the fact that this document is a document (unless doctype
|
|---|
| 495 | # has been set to something else from within a plugin
|
|---|
| 496 | my $dtype = $doc_obj->get_metadata_element ($section, "doctype");
|
|---|
| 497 | if (!defined $dtype || $dtype !~ /\w/) {
|
|---|
| 498 | $section_infodb{"doctype"} = [ "doc" ];
|
|---|
| 499 | }
|
|---|
| 500 |
|
|---|
| 501 | if ($first && defined $filename) {
|
|---|
| 502 | # if we are at the top level of the document, and we are not a reconstructed document, set the total_text_length - used to count bytes when we reconstruct later
|
|---|
| 503 | my $length = $doc_obj->get_total_text_length();
|
|---|
| 504 | $section_infodb{"total_numbytes"} = [ $length ];
|
|---|
| 505 | }
|
|---|
| 506 | # Output whether this node contains text
|
|---|
| 507 | #
|
|---|
| 508 | # If doc_obj reconstructed from database file then no need to
|
|---|
| 509 | # explicitly add <hastxt> as this is preserved as metadata when
|
|---|
| 510 | # the database file is loaded in
|
|---|
| 511 | if (defined $filename)
|
|---|
| 512 | {
|
|---|
| 513 | # doc_obj derived directly from file
|
|---|
| 514 | if ($doc_obj->get_text_length($section) > 0) {
|
|---|
| 515 | $section_infodb{"hastxt"} = [ "1" ];
|
|---|
| 516 | } else {
|
|---|
| 517 | $section_infodb{"hastxt"} = [ "0" ];
|
|---|
| 518 | }
|
|---|
| 519 | }
|
|---|
| 520 |
|
|---|
| 521 | # output all the section metadata
|
|---|
| 522 | my $metadata = $doc_obj->get_all_metadata ($section);
|
|---|
| 523 | foreach my $pair (@$metadata) {
|
|---|
| 524 | my ($field, $value) = (@$pair);
|
|---|
| 525 |
|
|---|
| 526 | if ($field ne "Identifier" && $field !~ /^gsdl/ &&
|
|---|
| 527 | defined $value && $value ne "") {
|
|---|
| 528 |
|
|---|
| 529 | # escape problematic stuff
|
|---|
| 530 | $value =~ s/([^\\])\\([^\\])/$1\\\\$2/g;
|
|---|
| 531 | $value =~ s/\n/\\n/g;
|
|---|
| 532 | $value =~ s/\r/\\r/g;
|
|---|
| 533 | # remove any ex. iff it's the only namespace prefix (will leave ex.dc.* intact)
|
|---|
| 534 | $field =~ s/^ex\.([^.]+)$/$1/; # $field =~ s/^ex\.//;
|
|---|
| 535 |
|
|---|
| 536 | # special case for UTF8URL metadata
|
|---|
| 537 | if ($field =~ m/^UTF8URL$/i) {
|
|---|
| 538 | &dbutil::write_infodb_entry($self->{'infodbtype'}, $infodb_handle,
|
|---|
| 539 | $value, { 'section' => [ $section_OID ] });
|
|---|
| 540 | }
|
|---|
| 541 |
|
|---|
| 542 | if (!defined $self->{'dontdb'}->{$field}) {
|
|---|
| 543 | push(@{$section_infodb{$field}}, $value);
|
|---|
| 544 |
|
|---|
| 545 | if ($section eq ""
|
|---|
| 546 | && (($self->{'store_metadata_coverage'} =~ /^true$/i)
|
|---|
| 547 | || $self->{'store_metadata_coverage'} eq "1"))
|
|---|
| 548 | {
|
|---|
| 549 | $self->infodb_metadata_stats($field,$edit_mode);
|
|---|
| 550 | }
|
|---|
| 551 | }
|
|---|
| 552 | }
|
|---|
| 553 | }
|
|---|
| 554 |
|
|---|
| 555 | if ($section eq "")
|
|---|
| 556 | {
|
|---|
| 557 | my $doc_mdprefix_fields = $self->{'doc_mdprefix_fields'};
|
|---|
| 558 |
|
|---|
| 559 | foreach my $prefix (keys %$doc_mdprefix_fields)
|
|---|
| 560 | {
|
|---|
| 561 | push(@{$section_infodb{"metadataset"}}, $prefix);
|
|---|
| 562 |
|
|---|
| 563 | foreach my $field (keys %{$doc_mdprefix_fields->{$prefix}})
|
|---|
| 564 | {
|
|---|
| 565 | push(@{$section_infodb{"metadatalist-$prefix"}}, $field);
|
|---|
| 566 |
|
|---|
| 567 | my $val = $doc_mdprefix_fields->{$prefix}->{$field};
|
|---|
| 568 | push(@{$section_infodb{"metadatafreq-$prefix-$field"}}, $val);
|
|---|
| 569 | }
|
|---|
| 570 | }
|
|---|
| 571 | }
|
|---|
| 572 |
|
|---|
| 573 | # If doc_obj reconstructed from database file then no need to
|
|---|
| 574 | # explicitly add <archivedir> as this is preserved as metadata when
|
|---|
| 575 | # the database file is loaded in
|
|---|
| 576 | if (defined $filename)
|
|---|
| 577 | {
|
|---|
| 578 | # output archivedir if at top level
|
|---|
| 579 | if ($section eq $doc_obj->get_top_section()) {
|
|---|
| 580 | $section_infodb{"archivedir"} = [ $archivedir ];
|
|---|
| 581 | }
|
|---|
| 582 | }
|
|---|
| 583 |
|
|---|
| 584 | # output document display type
|
|---|
| 585 | if ($first) {
|
|---|
| 586 | $section_infodb{"thistype"} = [ $thistype ];
|
|---|
| 587 | }
|
|---|
| 588 |
|
|---|
| 589 | if ($self->{'db_level'} eq "document") {
|
|---|
| 590 | # doc num is num_docs not num_sections
|
|---|
| 591 | # output the matching document number
|
|---|
| 592 | $section_infodb{"docnum"} = [ $self->{'num_docs'} ];
|
|---|
| 593 | }
|
|---|
| 594 | else {
|
|---|
| 595 | # output a list of children
|
|---|
| 596 | my $children = $doc_obj->get_children ($section);
|
|---|
| 597 | if (scalar(@$children) > 0) {
|
|---|
| 598 | $section_infodb{"childtype"} = [ $childtype ];
|
|---|
| 599 | my $contains = "";
|
|---|
| 600 | foreach my $child (@$children)
|
|---|
| 601 | {
|
|---|
| 602 | $contains .= ";" unless ($contains eq "");
|
|---|
| 603 | if ($child =~ /^.*?\.(\d+)$/)
|
|---|
| 604 | {
|
|---|
| 605 | $contains .= "\".$1";
|
|---|
| 606 | }
|
|---|
| 607 | else
|
|---|
| 608 | {
|
|---|
| 609 | $contains .= "\".$child";
|
|---|
| 610 | }
|
|---|
| 611 | }
|
|---|
| 612 | $section_infodb{"contains"} = [ $contains ];
|
|---|
| 613 | }
|
|---|
| 614 | # output the matching doc number
|
|---|
| 615 | $section_infodb{"docnum"} = [ $self->{'num_sections'} ];
|
|---|
| 616 | }
|
|---|
| 617 |
|
|---|
| 618 | if(defined $section_infodb{'assocfilepath'})
|
|---|
| 619 | {
|
|---|
| 620 | &dbutil::write_infodb_entry($self->{'infodbtype'}, $infodb_handle, $section_infodb{'assocfilepath'}[0], { 'contains' => [ $section_OID ]});
|
|---|
| 621 | }
|
|---|
| 622 | &dbutil::write_infodb_entry($self->{'infodbtype'}, $infodb_handle, $section_OID, \%section_infodb);
|
|---|
| 623 |
|
|---|
| 624 | # output a database entry for the document number, unless we are incremental
|
|---|
| 625 | unless ($self->is_incremental_capable())
|
|---|
| 626 | {
|
|---|
| 627 | if ($self->{'db_level'} eq "document") {
|
|---|
| 628 | &dbutil::write_infodb_entry($self->{'infodbtype'}, $infodb_handle, $self->{'num_docs'}, { 'section' => [ $doc_OID ] });
|
|---|
| 629 | }
|
|---|
| 630 | else {
|
|---|
| 631 | &dbutil::write_infodb_entry($self->{'infodbtype'}, $infodb_handle, $self->{'num_sections'}, { 'section' => [ $section_OID ] });
|
|---|
| 632 | }
|
|---|
| 633 | }
|
|---|
| 634 |
|
|---|
| 635 | $first = 0;
|
|---|
| 636 | $section = $doc_obj->get_next_section($section);
|
|---|
| 637 | last if ($self->{'db_level'} eq "document"); # if no sections wanted, only add the docs
|
|---|
| 638 | } # while defined section
|
|---|
| 639 |
|
|---|
| 640 | }
|
|---|
| 641 |
|
|---|
| 642 |
|
|---|
| 643 |
|
|---|
| 644 |
|
|---|
| 645 | sub infodb {
|
|---|
| 646 | my $self = shift (@_);
|
|---|
| 647 | my ($doc_obj, $filename) = @_;
|
|---|
| 648 |
|
|---|
| 649 | $self->infodbedit($doc_obj,$filename,"add");
|
|---|
| 650 | }
|
|---|
| 651 |
|
|---|
| 652 | sub infodbreindex {
|
|---|
| 653 | my $self = shift (@_);
|
|---|
| 654 | my ($doc_obj, $filename) = @_;
|
|---|
| 655 |
|
|---|
| 656 | $self->infodbedit($doc_obj,$filename,"update");
|
|---|
| 657 | }
|
|---|
| 658 |
|
|---|
| 659 | sub infodbdelete {
|
|---|
| 660 | my $self = shift (@_);
|
|---|
| 661 | my ($doc_obj, $filename) = @_;
|
|---|
| 662 |
|
|---|
| 663 | $self->infodbedit($doc_obj,$filename,"delete");
|
|---|
| 664 | }
|
|---|
| 665 |
|
|---|
| 666 |
|
|---|
| 667 | sub text {
|
|---|
| 668 | my $self = shift (@_);
|
|---|
| 669 | my ($doc_obj) = @_;
|
|---|
| 670 |
|
|---|
| 671 | my $handle = $self->{'outhandle'};
|
|---|
| 672 | print $handle "basebuildproc::text function must be implemented in sub classes\n";
|
|---|
| 673 | die "\n";
|
|---|
| 674 | }
|
|---|
| 675 |
|
|---|
| 676 | sub textreindex
|
|---|
| 677 | {
|
|---|
| 678 | my $self = shift @_;
|
|---|
| 679 |
|
|---|
| 680 | my $outhandle = $self->{'outhandle'};
|
|---|
| 681 | print $outhandle "basebuildproc::textreindex function must be implemented in sub classes\n";
|
|---|
| 682 | if (!$self->is_incremental_capable()) {
|
|---|
| 683 |
|
|---|
| 684 | print $outhandle " This operation is only possible with indexing tools with that support\n";
|
|---|
| 685 | print $outhandle " incremental building\n";
|
|---|
| 686 | }
|
|---|
| 687 | die "\n";
|
|---|
| 688 | }
|
|---|
| 689 |
|
|---|
| 690 | sub textdelete
|
|---|
| 691 | {
|
|---|
| 692 | my $self = shift @_;
|
|---|
| 693 |
|
|---|
| 694 | my $outhandle = $self->{'outhandle'};
|
|---|
| 695 | print $outhandle "basebuildproc::textdelete function must be implemented in sub classes\n";
|
|---|
| 696 | if (!$self->is_incremental_capable()) {
|
|---|
| 697 |
|
|---|
| 698 | print $outhandle " This operation is only possible with indexing tools with that support\n";
|
|---|
| 699 | print $outhandle " incremental building\n";
|
|---|
| 700 | }
|
|---|
| 701 | die "\n";
|
|---|
| 702 | }
|
|---|
| 703 |
|
|---|
| 704 |
|
|---|
| 705 | # should the document be indexed - according to the subcollection and language
|
|---|
| 706 | # specification.
|
|---|
| 707 | sub is_subcollection_doc {
|
|---|
| 708 | my $self = shift (@_);
|
|---|
| 709 | my ($doc_obj) = @_;
|
|---|
| 710 |
|
|---|
| 711 | my $indexed_doc = 1;
|
|---|
| 712 | foreach my $indexexp (@{$self->{'indexexparr'}}) {
|
|---|
| 713 | $indexed_doc = 0;
|
|---|
| 714 | my ($field, $exp, $options) = split /\//, $indexexp;
|
|---|
| 715 | if (defined ($field) && defined ($exp)) {
|
|---|
| 716 | my ($bool) = $field =~ /^(.)/;
|
|---|
| 717 | $field =~ s/^.// if $bool eq '!';
|
|---|
| 718 | my @metadata_values;
|
|---|
| 719 | if ($field =~ /^filename$/i) {
|
|---|
| 720 | push(@metadata_values, $doc_obj->get_source_filename());
|
|---|
| 721 | }
|
|---|
| 722 | else {
|
|---|
| 723 | $field =~ s/^ex\.([^.]+)$/$1/; # remove any ex. iff it's the only namespace prefix (will leave ex.dc.* intact)
|
|---|
| 724 | @metadata_values = @{$doc_obj->get_metadata($doc_obj->get_top_section(), $field)};
|
|---|
| 725 | }
|
|---|
| 726 | next unless @metadata_values;
|
|---|
| 727 | foreach my $metadata_value (@metadata_values) {
|
|---|
| 728 | if ($bool eq '!') {
|
|---|
| 729 | if (defined $options && $options =~ /^i$/i) {
|
|---|
| 730 | if ($metadata_value !~ /$exp/i) {$indexed_doc = 1; last;}
|
|---|
| 731 | } else {
|
|---|
| 732 | if ($metadata_value !~ /$exp/) {$indexed_doc = 1; last;}
|
|---|
| 733 | }
|
|---|
| 734 | } else {
|
|---|
| 735 | if (defined $options && $options =~ /^i$/i) {
|
|---|
| 736 | if ($metadata_value =~ /$exp/i) {$indexed_doc = 1; last;}
|
|---|
| 737 | } else {
|
|---|
| 738 | if ($metadata_value =~ /$exp/) {$indexed_doc = 1; last;}
|
|---|
| 739 | }
|
|---|
| 740 | }
|
|---|
| 741 | }
|
|---|
| 742 |
|
|---|
| 743 | last if ($indexed_doc == 1);
|
|---|
| 744 | }
|
|---|
| 745 | }
|
|---|
| 746 |
|
|---|
| 747 | # if this doc is so far in the sub collection, and we have lang info,
|
|---|
| 748 | # now we check the languages to see if it matches
|
|---|
| 749 | if($indexed_doc && defined $self->{'lang_meta'}) {
|
|---|
| 750 | $indexed_doc = 0;
|
|---|
| 751 | my $field = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $self->{'lang_meta'});
|
|---|
| 752 | if (defined $field) {
|
|---|
| 753 | foreach my $lang (@{$self->{'langarr'}}) {
|
|---|
| 754 | my ($bool) = $lang =~ /^(.)/;
|
|---|
| 755 | if ($bool eq '!') {
|
|---|
| 756 | $lang =~ s/^.//;
|
|---|
| 757 | if ($field !~ /$lang/) {
|
|---|
| 758 | $indexed_doc = 1; last;
|
|---|
| 759 | }
|
|---|
| 760 | } else {
|
|---|
| 761 | if ($field =~ /$lang/) {
|
|---|
| 762 | $indexed_doc = 1; last;
|
|---|
| 763 | }
|
|---|
| 764 | }
|
|---|
| 765 | }
|
|---|
| 766 | }
|
|---|
| 767 | }
|
|---|
| 768 | return $indexed_doc;
|
|---|
| 769 |
|
|---|
| 770 | }
|
|---|
| 771 |
|
|---|
| 772 | # use 'Paged' if document has no more than 2 levels
|
|---|
| 773 | # and each section at second level has a number for
|
|---|
| 774 | # Title metadata
|
|---|
| 775 | # also use Paged if gsdlthistype metadata is set to Paged
|
|---|
| 776 | sub get_document_type {
|
|---|
| 777 | my $self = shift (@_);
|
|---|
| 778 | my ($doc_obj) = @_;
|
|---|
| 779 |
|
|---|
| 780 | my $thistype = "VList";
|
|---|
| 781 | my $childtype = "VList";
|
|---|
| 782 | my $title;
|
|---|
| 783 | my @tmp = ();
|
|---|
| 784 |
|
|---|
| 785 | my $section = $doc_obj->get_top_section ();
|
|---|
| 786 |
|
|---|
| 787 | my $gsdlthistype = $doc_obj->get_metadata_element ($section, "gsdlthistype");
|
|---|
| 788 | if (defined $gsdlthistype) {
|
|---|
| 789 | if ($gsdlthistype =~ /^paged$/i) {
|
|---|
| 790 | $childtype = "Paged";
|
|---|
| 791 | if ($doc_obj->get_text_length ($doc_obj->get_top_section())) {
|
|---|
| 792 | $thistype = "Paged";
|
|---|
| 793 | } else {
|
|---|
| 794 | $thistype = "Invisible";
|
|---|
| 795 | }
|
|---|
| 796 |
|
|---|
| 797 | return ($thistype, $childtype);
|
|---|
| 798 | }
|
|---|
| 799 | # gs3 pagedhierarchy option
|
|---|
| 800 | elsif ($gsdlthistype =~ /^pagedhierarchy$/i) {
|
|---|
| 801 | $childtype = "PagedHierarchy";
|
|---|
| 802 | if ($doc_obj->get_text_length ($doc_obj->get_top_section())) {
|
|---|
| 803 | $thistype = "PagedHierarchy";
|
|---|
| 804 | } else {
|
|---|
| 805 | $thistype = "Invisible";
|
|---|
| 806 | }
|
|---|
| 807 |
|
|---|
| 808 | return ($thistype, $childtype);
|
|---|
| 809 | } elsif ($gsdlthistype =~ /^hierarchy$/i) {
|
|---|
| 810 | return ($thistype, $childtype); # use VList, VList
|
|---|
| 811 | }
|
|---|
| 812 | }
|
|---|
| 813 | my $first = 1;
|
|---|
| 814 | while (defined $section) {
|
|---|
| 815 | @tmp = split /\./, $section;
|
|---|
| 816 | if (scalar(@tmp) > 1) {
|
|---|
| 817 | return ($thistype, $childtype);
|
|---|
| 818 | }
|
|---|
| 819 | if (!$first) {
|
|---|
| 820 | $title = $doc_obj->get_metadata_element ($section, "Title");
|
|---|
| 821 | if (!defined $title || $title !~ /^\d+$/) {
|
|---|
| 822 | return ($thistype, $childtype);
|
|---|
| 823 | }
|
|---|
| 824 | }
|
|---|
| 825 | $first = 0;
|
|---|
| 826 | $section = $doc_obj->get_next_section($section);
|
|---|
| 827 | }
|
|---|
| 828 | if ($doc_obj->get_text_length ($doc_obj->get_top_section())) {
|
|---|
| 829 | $thistype = "Paged";
|
|---|
| 830 | } else {
|
|---|
| 831 | $thistype = "Invisible";
|
|---|
| 832 | }
|
|---|
| 833 | $childtype = "Paged";
|
|---|
| 834 | return ($thistype, $childtype);
|
|---|
| 835 | }
|
|---|
| 836 |
|
|---|
| 837 | sub assoc_files
|
|---|
| 838 | {
|
|---|
| 839 | my $self = shift (@_);
|
|---|
| 840 | my ($doc_obj, $archivedir) = @_;
|
|---|
| 841 | my ($afile);
|
|---|
| 842 |
|
|---|
| 843 | foreach my $assoc_file (@{$doc_obj->get_assoc_files()}) {
|
|---|
| 844 | #rint STDERR "Processing associated file - copy " . $assoc_file->[0] . " to " . $assoc_file->[1] . "\n";
|
|---|
| 845 | # if assoc file starts with a slash, we put it relative to the assoc
|
|---|
| 846 | # dir, otherwise it is relative to the HASH... directory
|
|---|
| 847 | if ($assoc_file->[1] =~ m@^[/\\]@) {
|
|---|
| 848 | $afile = &FileUtils::filenameConcatenate($self->{'assocdir'}, $assoc_file->[1]);
|
|---|
| 849 | } else {
|
|---|
| 850 | $afile = &FileUtils::filenameConcatenate($self->{'assocdir'}, $archivedir, $assoc_file->[1]);
|
|---|
| 851 | }
|
|---|
| 852 |
|
|---|
| 853 | &FileUtils::hardLink($assoc_file->[0], $afile, $self->{'verbosity'});
|
|---|
| 854 | }
|
|---|
| 855 | }
|
|---|
| 856 |
|
|---|
| 857 | sub delete_assoc_files
|
|---|
| 858 | {
|
|---|
| 859 | my $self = shift (@_);
|
|---|
| 860 | my ($archivedir, $edit_mode) = @_;
|
|---|
| 861 |
|
|---|
| 862 | my $assoc_dir = &FileUtils::filenameConcatenate($self->{'assocdir'}, $archivedir);
|
|---|
| 863 | if (-d $assoc_dir) {
|
|---|
| 864 | &FileUtils::removeFilesRecursive($assoc_dir);
|
|---|
| 865 | }
|
|---|
| 866 | }
|
|---|
