| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # lucenebuilder.pm -- perl wrapper for building index with Lucene
|
|---|
| 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 | ###########################################################################
|
|---|
| 27 | # /*
|
|---|
| 28 | # * @version 1.0 ?
|
|---|
| 29 | # * @version 2.0 Incremental building assistance added, including
|
|---|
| 30 | # * remove_document_from_database which implements the granddad's
|
|---|
| 31 | # * empty function to call the lucene_passes.pl and full_lucene_passes_exe
|
|---|
| 32 | # * so there is one place in the code that works out where the
|
|---|
| 33 | # * perl script is. John Rowe
|
|---|
| 34 | # *
|
|---|
| 35 | # * @author John Rowe, DL Consulting Ltd.
|
|---|
| 36 | # */
|
|---|
| 37 | ###########################################################################
|
|---|
| 38 |
|
|---|
| 39 | package lucenebuilder;
|
|---|
| 40 |
|
|---|
| 41 | # Use same basic XML structure setup by mgppbuilder/mgppbuildproc
|
|---|
| 42 |
|
|---|
| 43 | use mgppbuilder;
|
|---|
| 44 |
|
|---|
| 45 | # use IncrementalBuildUtils;
|
|---|
| 46 |
|
|---|
| 47 | sub BEGIN {
|
|---|
| 48 | @lucenebuilder::ISA = ('mgppbuilder');
|
|---|
| 49 | }
|
|---|
| 50 |
|
|---|
| 51 | # /**
|
|---|
| 52 | # * @author John Thompson, DL Consulting Ltd.
|
|---|
| 53 | # */
|
|---|
| 54 | sub new {
|
|---|
| 55 | my $class = shift(@_);
|
|---|
| 56 | my ($collection, $source_dir, $build_dir, $verbosity, $maxdocs, $debug, $keepold, $allclassifications, $outhandle, $no_text, $faillog, $gli, $incremental) = @_;
|
|---|
| 57 |
|
|---|
| 58 | my $self = new mgppbuilder (@_);
|
|---|
| 59 | $self = bless $self, $class;
|
|---|
| 60 |
|
|---|
| 61 | $self->{'buildtype'} = "lucene";
|
|---|
| 62 |
|
|---|
| 63 | # Do we need to put exe on the end?
|
|---|
| 64 | my $exe = &util::get_os_exe ();
|
|---|
| 65 | my $scriptdir = "$ENV{'GSDLHOME'}/bin/script";
|
|---|
| 66 |
|
|---|
| 67 | # So where is lucene_passes.pl anyway?
|
|---|
| 68 | my $lucene_passes_script = &util::filename_cat($scriptdir, "lucene_passes.pl");
|
|---|
| 69 |
|
|---|
| 70 | # So tack perl on the beginning to ensure execution
|
|---|
| 71 | $self->{'full_lucene_passes'} = "$lucene_passes_script";
|
|---|
| 72 | if ($exe eq ".exe")
|
|---|
| 73 | {
|
|---|
| 74 | $self->{'full_lucene_passes_exe'} = "perl$exe \"$lucene_passes_script\"";
|
|---|
| 75 | }
|
|---|
| 76 | else
|
|---|
| 77 | {
|
|---|
| 78 | $self->{'full_lucene_passes_exe'} = "perl -S \"$lucene_passes_script\"";
|
|---|
| 79 | }
|
|---|
| 80 |
|
|---|
| 81 | # We must also record whether we have been asked to do just an incremental
|
|---|
| 82 | # build (which makes no difference to the Lucene indexing bit, just the
|
|---|
| 83 | # building of the classifiers in the GDBM).
|
|---|
| 84 | $self->{'incremental'} = $incremental;
|
|---|
| 85 |
|
|---|
| 86 | return $self;
|
|---|
| 87 | }
|
|---|
| 88 | # /** new() **/
|
|---|
| 89 |
|
|---|
| 90 | # lucene has none of these options
|
|---|
| 91 | sub generate_index_options {
|
|---|
| 92 | my $self = shift (@_);
|
|---|
| 93 |
|
|---|
| 94 | $self->{'casefold'} = 0;
|
|---|
| 95 | $self->{'stem'} = 0;
|
|---|
| 96 | $self->{'accentfold'} = 0;
|
|---|
| 97 | $self->{'stemindexes'} = 0;
|
|---|
| 98 | }
|
|---|
| 99 |
|
|---|
| 100 | sub default_buildproc {
|
|---|
| 101 | my $self = shift (@_);
|
|---|
| 102 |
|
|---|
| 103 | return "lucenebuildproc";
|
|---|
| 104 | }
|
|---|
| 105 |
|
|---|
| 106 | # this writes a nice version of the text docs
|
|---|
| 107 | sub compress_text
|
|---|
| 108 | {
|
|---|
| 109 | my $self = shift (@_);
|
|---|
| 110 | # we don't do anything if we don't want compressed text
|
|---|
| 111 | return if $self->{'no_text'};
|
|---|
| 112 |
|
|---|
| 113 | my ($textindex) = @_;
|
|---|
| 114 | my $outhandle = $self->{'outhandle'};
|
|---|
| 115 | print STDERR "Saving the document text\n";
|
|---|
| 116 | # the text directory
|
|---|
| 117 | my $text_dir = &util::filename_cat($self->{'build_dir'}, "text");
|
|---|
| 118 | my $build_dir = &util::filename_cat($self->{'build_dir'},"");
|
|---|
| 119 | &util::mk_all_dir ($text_dir);
|
|---|
| 120 |
|
|---|
| 121 | my $osextra = "";
|
|---|
| 122 | if ($ENV{'GSDLOS'} =~ /^windows$/i)
|
|---|
| 123 | {
|
|---|
| 124 | $text_dir =~ s@/@\\@g;
|
|---|
| 125 | }
|
|---|
| 126 | else
|
|---|
| 127 | {
|
|---|
| 128 | if ($outhandle ne "STDERR")
|
|---|
| 129 | {
|
|---|
| 130 | # so lucene_passes doesn't print to stderr if we redirect output
|
|---|
| 131 | $osextra .= " 2>/dev/null";
|
|---|
| 132 | }
|
|---|
| 133 | }
|
|---|
| 134 |
|
|---|
| 135 | # get any os specific stuff
|
|---|
| 136 | my $scriptdir = "$ENV{'GSDLHOME'}/bin/script";
|
|---|
| 137 |
|
|---|
| 138 | # Find the perl script to call to run lucene
|
|---|
| 139 | my $full_lucene_passes = $self->{'full_lucene_passes'};
|
|---|
| 140 | my $full_lucene_passes_exe = $self->{'full_lucene_passes_exe'};
|
|---|
| 141 |
|
|---|
| 142 | my $lucene_passes_sections = "Doc";
|
|---|
| 143 |
|
|---|
| 144 | my ($handle);
|
|---|
| 145 |
|
|---|
| 146 | if ($self->{'debug'})
|
|---|
| 147 | {
|
|---|
| 148 | $handle = STDOUT;
|
|---|
| 149 | }
|
|---|
| 150 | else
|
|---|
| 151 | {
|
|---|
| 152 | print STDERR "Full Path: $full_lucene_passes\n";
|
|---|
| 153 | print STDERR "Executable: $full_lucene_passes_exe\n";
|
|---|
| 154 | print STDERR "Sections: $lucene_passes_sections\n";
|
|---|
| 155 | print STDERR "Build Dir: $build_dir\n";
|
|---|
| 156 | print STDERR "Cmd: $full_lucene_passes_exe text $lucene_passes_sections \"$build_dir\" \"dummy\" $osextra\n";
|
|---|
| 157 | if (!-e "$full_lucene_passes" ||
|
|---|
| 158 | !open (PIPEOUT, "| $full_lucene_passes_exe text $lucene_passes_sections \"$build_dir\" \"dummy\" $osextra"))
|
|---|
| 159 | {
|
|---|
| 160 | print STDERR "<FatalError name='NoRunLucenePasses'/>\n</Stage>\n" if $self->{'gli'};
|
|---|
| 161 | die "lucenebuilder::build_index - couldn't run $full_lucene_passes_exe\n";
|
|---|
| 162 | }
|
|---|
| 163 | $handle = lucenebuilder::PIPEOUT;
|
|---|
| 164 | }
|
|---|
| 165 | my $levels = $self->{'levels'};
|
|---|
| 166 | my $gdbm_level = "document";
|
|---|
| 167 | if ($levels->{'section'})
|
|---|
| 168 | {
|
|---|
| 169 | $gdbm_level = "section";
|
|---|
| 170 | }
|
|---|
| 171 |
|
|---|
| 172 | undef $levels->{'paragraph'}; # get rid of para if we had it.
|
|---|
| 173 | # set up the document processr
|
|---|
| 174 | $self->{'buildproc'}->set_output_handle ($handle);
|
|---|
| 175 | $self->{'buildproc'}->set_mode ('text');
|
|---|
| 176 | $self->{'buildproc'}->set_index ($textindex);
|
|---|
| 177 | $self->{'buildproc'}->set_indexing_text (0);
|
|---|
| 178 | $self->{'buildproc'}->set_indexfieldmap ($self->{'indexfieldmap'});
|
|---|
| 179 | $self->{'buildproc'}->set_levels ($levels);
|
|---|
| 180 | $self->{'buildproc'}->set_gdbm_level ($gdbm_level);
|
|---|
| 181 | $self->{'buildproc'}->reset();
|
|---|
| 182 | &plugin::begin($self->{'pluginfo'}, $self->{'source_dir'},
|
|---|
| 183 | $self->{'buildproc'}, $self->{'maxdocs'});
|
|---|
| 184 | &plugin::read ($self->{'pluginfo'}, $self->{'source_dir'},
|
|---|
| 185 | "", {}, $self->{'buildproc'}, $self->{'maxdocs'}, 0, $self->{'gli'});
|
|---|
| 186 | &plugin::end($self->{'pluginfo'});
|
|---|
| 187 | close ($handle) unless $self->{'debug'};
|
|---|
| 188 | close PIPEOUT;
|
|---|
| 189 | $self->print_stats();
|
|---|
| 190 |
|
|---|
| 191 | print STDERR "</Stage>\n" if $self->{'gli'};
|
|---|
| 192 | }
|
|---|
| 193 |
|
|---|
| 194 | sub build_indexes {
|
|---|
| 195 | my $self = shift (@_);
|
|---|
| 196 | my ($indexname) = @_;
|
|---|
| 197 | my $outhandle = $self->{'outhandle'};
|
|---|
| 198 |
|
|---|
| 199 | my $indexes = [];
|
|---|
| 200 | if (defined $indexname && $indexname =~ /\w/) {
|
|---|
| 201 | push @$indexes, $indexname;
|
|---|
| 202 | } else {
|
|---|
| 203 | $indexes = $self->{'collect_cfg'}->{'indexes'};
|
|---|
| 204 | }
|
|---|
| 205 |
|
|---|
| 206 | # create the mapping between the index descriptions
|
|---|
| 207 | # and their directory names (includes subcolls and langs)
|
|---|
| 208 | $self->{'index_mapping'} = $self->create_index_mapping ($indexes);
|
|---|
| 209 |
|
|---|
| 210 | # build each of the indexes
|
|---|
| 211 | foreach $index (@$indexes) {
|
|---|
| 212 | if ($self->want_built($index)) {
|
|---|
| 213 |
|
|---|
| 214 | my $idx = $self->{'index_mapping'}->{$index};
|
|---|
| 215 | foreach my $level (keys %{$self->{'levels'}}) {
|
|---|
| 216 | next if $level =~ /paragraph/; # we don't do para indexing
|
|---|
| 217 | my ($pindex) = $level =~ /^(.)/;
|
|---|
| 218 | # should probably check that new name with level
|
|---|
| 219 | # is unique ... but currently (with doc sec and para)
|
|---|
| 220 | # each has unique first letter.
|
|---|
| 221 | $self->{'index_mapping'}->{$index} = $pindex.$idx;
|
|---|
| 222 |
|
|---|
| 223 | my $llevel = $mgppbuilder::level_map{$level};
|
|---|
| 224 | print $outhandle "\n*** building index $index at level $llevel in subdirectory " .
|
|---|
| 225 | "$self->{'index_mapping'}->{$index}\n" if ($self->{'verbosity'} >= 1);
|
|---|
| 226 | print STDERR "<Stage name='Index' source='$index' level=$llevel>\n" if $self->{'gli'};
|
|---|
| 227 |
|
|---|
| 228 | $self->build_index($index,$llevel);
|
|---|
| 229 | }
|
|---|
| 230 | $self->{'index_mapping'}->{$index} = $idx;
|
|---|
| 231 |
|
|---|
| 232 | } else {
|
|---|
| 233 | print $outhandle "\n*** ignoring index $index\n" if ($self->{'verbosity'} >= 1);
|
|---|
| 234 | }
|
|---|
| 235 | }
|
|---|
| 236 |
|
|---|
| 237 | #define the final field lists
|
|---|
| 238 | $self->make_final_field_list();
|
|---|
| 239 | }
|
|---|
| 240 |
|
|---|
| 241 | # /** Lucene specific document removal function. This works by calling lucene_passes.pl with
|
|---|
| 242 | # * -remove and the document id on the command line.
|
|---|
| 243 | # *
|
|---|
| 244 | # * @param oid is the document identifier to be removed.
|
|---|
| 245 | # *
|
|---|
| 246 | # * @author John Rowe, DL Consulting Ltd.
|
|---|
| 247 | # */
|
|---|
| 248 | sub remove_document_from_database
|
|---|
| 249 | {
|
|---|
| 250 | my ($self, $oid) = @_;
|
|---|
| 251 | # Find the perl script to call to run lucene
|
|---|
| 252 | my $full_lucene_passes_exe = $self->{'full_lucene_passes_exe'};
|
|---|
| 253 | # Call lucene_passes.pl with -remove and the document ID on the command line
|
|---|
| 254 | `$full_lucene_passes_exe -remove "$oid"`;
|
|---|
| 255 | }
|
|---|
| 256 | # /** remove_document_from_database **/
|
|---|
| 257 |
|
|---|
| 258 | sub build_index {
|
|---|
| 259 | my $self = shift (@_);
|
|---|
| 260 | my ($index,$llevel) = @_;
|
|---|
| 261 | my $outhandle = $self->{'outhandle'};
|
|---|
| 262 | my $build_dir = $self->{'build_dir'};
|
|---|
| 263 |
|
|---|
| 264 | # get the full index directory path and make sure it exists
|
|---|
| 265 | my $indexdir = $self->{'index_mapping'}->{$index};
|
|---|
| 266 | &util::mk_all_dir (&util::filename_cat($build_dir, $indexdir));
|
|---|
| 267 |
|
|---|
| 268 | # get any os specific stuff
|
|---|
| 269 | my $exedir = "$ENV{'GSDLHOME'}/bin/$ENV{'GSDLOS'}";
|
|---|
| 270 | my $scriptdir = "$ENV{'GSDLHOME'}/bin/script";
|
|---|
| 271 |
|
|---|
| 272 | # Find the perl script to call to run lucene
|
|---|
| 273 | my $full_lucene_passes = $self->{'full_lucene_passes'};
|
|---|
| 274 | my $full_lucene_passes_exe = $self->{'full_lucene_passes_exe'};
|
|---|
| 275 |
|
|---|
| 276 | # define the section names for lucenepasses
|
|---|
| 277 | # define the section names and possibly the doc name for lucenepasses
|
|---|
| 278 | my $lucene_passes_sections = $llevel;
|
|---|
| 279 |
|
|---|
| 280 | my $opt_create_index = ($self->{'keepold'}) ? "" : "-create";
|
|---|
| 281 |
|
|---|
| 282 | my $osextra = "";
|
|---|
| 283 | if ($ENV{'GSDLOS'} =~ /^windows$/i) {
|
|---|
| 284 | $build_dir =~ s@/@\\@g;
|
|---|
| 285 | } else {
|
|---|
| 286 | if ($outhandle ne "STDERR") {
|
|---|
| 287 | # so lucene_passes doesn't print to stderr if we redirect output
|
|---|
| 288 | $osextra .= " 2>/dev/null";
|
|---|
| 289 | }
|
|---|
| 290 | }
|
|---|
| 291 |
|
|---|
| 292 | # get the index expression if this index belongs
|
|---|
| 293 | # to a subcollection
|
|---|
| 294 | my $indexexparr = [];
|
|---|
| 295 | my $langarr = [];
|
|---|
| 296 |
|
|---|
| 297 | # there may be subcollection info, and language info.
|
|---|
| 298 | my ($fields, $subcollection, $language) = split (":", $index);
|
|---|
| 299 | my @subcollections = ();
|
|---|
| 300 | @subcollections = split /,/, $subcollection if (defined $subcollection);
|
|---|
| 301 |
|
|---|
| 302 | foreach $subcollection (@subcollections) {
|
|---|
| 303 | if (defined ($self->{'collect_cfg'}->{'subcollection'}->{$subcollection})) {
|
|---|
| 304 | push (@$indexexparr, $self->{'collect_cfg'}->{'subcollection'}->{$subcollection});
|
|---|
| 305 | }
|
|---|
| 306 | }
|
|---|
| 307 |
|
|---|
| 308 | # add expressions for languages if this index belongs to
|
|---|
| 309 | # a language subcollection - only put languages expressions for the
|
|---|
| 310 | # ones we want in the index
|
|---|
| 311 | my @languages = ();
|
|---|
| 312 | my $language_metadata = "Language";
|
|---|
| 313 | if (defined ($self->{'collect_cfg'}->{'language_metadata'})) {
|
|---|
| 314 | $language_metadata = $self->{'collect_cfg'}->{'language_metadata'};
|
|---|
| 315 | }
|
|---|
| 316 | @languages = split /,/, $language if (defined $language);
|
|---|
| 317 | foreach my $language (@languages) {
|
|---|
| 318 | my $not=0;
|
|---|
| 319 | if ($language =~ s/^\!//) {
|
|---|
| 320 | $not = 1;
|
|---|
| 321 | }
|
|---|
| 322 | if($not) {
|
|---|
| 323 | push (@$langarr, "!$language");
|
|---|
| 324 | } else {
|
|---|
| 325 | push (@$langarr, "$language");
|
|---|
| 326 | }
|
|---|
| 327 | }
|
|---|
| 328 |
|
|---|
| 329 | # Build index dictionary. Uses verbatim stem method
|
|---|
| 330 | print $outhandle "\n creating index dictionary (lucene_passes -I1)\n" if ($self->{'verbosity'} >= 1);
|
|---|
| 331 | print STDERR "<Phase name='CreatingIndexDic'/>\n" if $self->{'gli'};
|
|---|
| 332 | my ($handle);
|
|---|
| 333 |
|
|---|
| 334 | if ($self->{'debug'}) {
|
|---|
| 335 | $handle = STDOUT;
|
|---|
| 336 | } else {
|
|---|
| 337 | print STDERR "Cmd: $full_lucene_passes_exe $opt_create_index index $lucene_passes_sections \"$build_dir\" \"$indexdir\" $osextra\n";
|
|---|
| 338 | if (!-e "$full_lucene_passes" ||
|
|---|
| 339 | !open (PIPEOUT, "| $full_lucene_passes_exe $opt_create_index index $lucene_passes_sections \"$build_dir\" \"$indexdir\" $osextra")) {
|
|---|
| 340 | print STDERR "<FatalError name='NoRunLucenePasses'/>\n</Stage>\n" if $self->{'gli'};
|
|---|
| 341 | die "lucenebuilder::build_index - couldn't run $lucene_passes_exe\n";
|
|---|
| 342 | }
|
|---|
| 343 | $handle = lucenebuilder::PIPEOUT;
|
|---|
| 344 | }
|
|---|
| 345 |
|
|---|
| 346 | my $store_levels = $self->{'levels'};
|
|---|
| 347 | my $gdbm_level = "document";
|
|---|
| 348 | if ($store_levels->{'section'}) {
|
|---|
| 349 | $gdbm_level = "section";
|
|---|
| 350 | }
|
|---|
| 351 |
|
|---|
| 352 | my $dom_level = "";
|
|---|
| 353 | foreach my $key (keys %$store_levels) {
|
|---|
| 354 | if ($mgppbuilder::level_map{$key} eq $llevel) {
|
|---|
| 355 | $dom_level = $key;
|
|---|
| 356 | }
|
|---|
| 357 | }
|
|---|
| 358 | if ($dom_level eq "") {
|
|---|
| 359 | print STDERR "Warning: unrecognized tag level $llevel\n";
|
|---|
| 360 | $dom_level = "document";
|
|---|
| 361 | }
|
|---|
| 362 |
|
|---|
| 363 | my $local_levels = { $dom_level => 1 }; # work on one level at a time
|
|---|
| 364 |
|
|---|
| 365 | # set up the document processr
|
|---|
| 366 | $self->{'buildproc'}->set_output_handle ($handle);
|
|---|
| 367 | $self->{'buildproc'}->set_mode ('text');
|
|---|
| 368 | $self->{'buildproc'}->set_index ($index, $indexexparr);
|
|---|
| 369 | $self->{'buildproc'}->set_index_languages ($language_metadata, $langarr) if (defined $language);
|
|---|
| 370 | $self->{'buildproc'}->set_indexing_text (1);
|
|---|
| 371 | $self->{'buildproc'}->set_indexfieldmap ($self->{'indexfieldmap'});
|
|---|
| 372 | $self->{'buildproc'}->set_levels ($local_levels);
|
|---|
| 373 | $self->{'buildproc'}->set_gdbm_level($gdbm_level);
|
|---|
| 374 | $self->{'buildproc'}->reset();
|
|---|
| 375 | &plugin::read ($self->{'pluginfo'}, $self->{'source_dir'},
|
|---|
| 376 | "", {}, $self->{'buildproc'}, $self->{'maxdocs'}, 0, $self->{'gli'});
|
|---|
| 377 | close ($handle) unless $self->{'debug'};
|
|---|
| 378 |
|
|---|
| 379 | $self->print_stats();
|
|---|
| 380 |
|
|---|
| 381 | $self->{'buildproc'}->set_levels ($store_levels);
|
|---|
| 382 | print STDERR "</Stage>\n" if $self->{'gli'};
|
|---|
| 383 | }
|
|---|
| 384 |
|
|---|
| 385 | # /** A modified version of the basebuilder.pm's function that generates the
|
|---|
| 386 | # * information database (GDBM) from the GA documents. We need to change this
|
|---|
| 387 | # * so that if we've been asked to do an incremental build we only add
|
|---|
| 388 | # * metadata to autohierarchy classifiers via the IncrementalBuildUtils
|
|---|
| 389 | # * module. All other classifiers and metadata will be ignored.
|
|---|
| 390 | # */
|
|---|
| 391 | sub make_infodatabase
|
|---|
| 392 | {
|
|---|
| 393 | my $self = shift (@_);
|
|---|
| 394 | my $outhandle = $self->{'outhandle'};
|
|---|
| 395 |
|
|---|
| 396 | my $dbext = ".bdb";
|
|---|
| 397 | $dbext = ".ldb" if &util::is_little_endian();
|
|---|
| 398 | my $infodb_file = &util::filename_cat($self->{'build_dir'}, "text", $self->{'collection'} . $dbext);
|
|---|
| 399 |
|
|---|
| 400 | # If we aren't doing an incremental addition, then we just call the super-
|
|---|
| 401 | # classes version
|
|---|
| 402 | # Note: Incremental addition can only occur if a text/<collection>.ldb
|
|---|
| 403 | # already exists. If it doesn't, let the super classes function be
|
|---|
| 404 | # called once to generate it.
|
|---|
| 405 | if (!$self->{'incremental'} || !(-e $infodb_file))
|
|---|
| 406 | {
|
|---|
| 407 | # basebuilder::make_infodatabase(@_);
|
|---|
| 408 | # Note: this doesn't work as the direct reference means all the $self
|
|---|
| 409 | # data is lost.
|
|---|
| 410 | $self->basebuilder::make_infodatabase(@_);
|
|---|
| 411 | return;
|
|---|
| 412 | }
|
|---|
| 413 |
|
|---|
| 414 | # Carry on with an incremental addition
|
|---|
| 415 | print $outhandle "\n*** performing an incremental addition to the info database\n" if ($self->{'verbosity'} >= 1);
|
|---|
| 416 | print STDERR "<Stage name='CreateInfoData'>\n" if $self->{'gli'};
|
|---|
| 417 |
|
|---|
| 418 | # 1. Init all the classifiers
|
|---|
| 419 | &classify::init_classifiers ($self->{'classifiers'});
|
|---|
| 420 | # 2. Init the buildproc settings.
|
|---|
| 421 | # Note: we still need this to process any associated files - but we
|
|---|
| 422 | # don't expect to pipe anything to txt2db so we can do away with the
|
|---|
| 423 | # complex output handle.
|
|---|
| 424 | my $assocdir = &util::filename_cat($self->{'build_dir'}, "assoc");
|
|---|
| 425 | &util::mk_all_dir ($assocdir);
|
|---|
| 426 | $self->{'buildproc'}->set_mode ('incinfodb'); # Very Important
|
|---|
| 427 | $self->{'buildproc'}->set_assocdir ($assocdir);
|
|---|
| 428 | # 3. Read in all the metadata from the files in the archives directory using
|
|---|
| 429 | # the GAPlug and using ourselves as the document processor!
|
|---|
| 430 | &plugin::read ($self->{'pluginfo'}, $self->{'source_dir'}, "", {}, $self->{'buildproc'}, $self->{'maxdocs'},0, $self->{'gli'});
|
|---|
| 431 |
|
|---|
| 432 | print STDERR "</Stage>\n" if $self->{'gli'};
|
|---|
| 433 | }
|
|---|
| 434 |
|
|---|
| 435 | 1;
|
|---|
| 436 |
|
|---|
| 437 |
|
|---|
