[14912] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # mgppbuildproc.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
|
---|
| 27 | # for mgpp to process
|
---|
| 28 |
|
---|
| 29 |
|
---|
| 30 | package mgppbuildproc;
|
---|
| 31 |
|
---|
| 32 | use basebuildproc;
|
---|
| 33 | use strict;
|
---|
| 34 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
| 35 |
|
---|
| 36 |
|
---|
| 37 | BEGIN {
|
---|
| 38 | @mgppbuildproc::ISA = ('basebuildproc');
|
---|
| 39 | }
|
---|
| 40 |
|
---|
| 41 | #this must be the same as in mgppbuilder
|
---|
| 42 | our %level_map = ('document'=>'Doc',
|
---|
| 43 | 'section'=>'Sec',
|
---|
| 44 | 'paragraph'=>'Para');
|
---|
| 45 |
|
---|
| 46 | sub new {
|
---|
| 47 | my $class = shift @_;
|
---|
| 48 | my $self = new basebuildproc (@_);
|
---|
| 49 |
|
---|
| 50 | # use a different index specification to the default
|
---|
| 51 | $self->{'index'} = "text";
|
---|
| 52 |
|
---|
| 53 | $self->{'dontindex'} = {};
|
---|
| 54 | $self->{'indexfieldmap'} = {};
|
---|
| 55 | $self->{'indexfields'} = {}; # only put in the ones that are not specified directly in the index
|
---|
| 56 | $self->{'strip_html'}=1;
|
---|
| 57 |
|
---|
| 58 | return bless $self, $class;
|
---|
| 59 | }
|
---|
| 60 |
|
---|
| 61 |
|
---|
| 62 | sub set_indexfieldmap {
|
---|
| 63 | my $self = shift (@_);
|
---|
| 64 | my ($indexmap) = @_;
|
---|
| 65 |
|
---|
| 66 | $self->{'indexfieldmap'} = $indexmap;
|
---|
| 67 | }
|
---|
| 68 |
|
---|
| 69 | sub get_indexfieldmap {
|
---|
| 70 | my $self = shift (@_);
|
---|
| 71 |
|
---|
| 72 | return $self->{'indexfieldmap'};
|
---|
| 73 | }
|
---|
| 74 |
|
---|
| 75 | sub set_levels {
|
---|
| 76 | my $self = shift (@_);
|
---|
| 77 | my ($levels) = @_;
|
---|
| 78 |
|
---|
| 79 | $self->{'levels'} = $levels;
|
---|
| 80 | }
|
---|
| 81 |
|
---|
| 82 | sub set_strip_html {
|
---|
| 83 | my $self = shift (@_);
|
---|
| 84 | my ($strip) = @_;
|
---|
| 85 | $self->{'strip_html'}=$strip;
|
---|
| 86 | }
|
---|
| 87 |
|
---|
| 88 | #sub find_paragraphs {
|
---|
| 89 | # $_[1] =~ s/(<p\b)/<Paragraph>$1/gi;
|
---|
| 90 | #}
|
---|
| 91 |
|
---|
| 92 | sub remove_gtlt {
|
---|
| 93 | my $self =shift(@_);
|
---|
| 94 | my ($text, $para) = @_;
|
---|
| 95 | $text =~s/[<>]//g;
|
---|
| 96 | return "$para$text$para";
|
---|
| 97 | }
|
---|
| 98 |
|
---|
| 99 | sub process_tags {
|
---|
| 100 | my $self = shift(@_);
|
---|
| 101 | my ($text, $para) = @_;
|
---|
| 102 | if ($text =~ /^p\b/i) {
|
---|
| 103 | return $para;
|
---|
| 104 | }
|
---|
| 105 | return "";
|
---|
| 106 | }
|
---|
| 107 |
|
---|
| 108 | sub preprocess_text {
|
---|
| 109 | my $self = shift (@_);
|
---|
| 110 | my ($text, $strip_html, $para) = @_;
|
---|
| 111 | # at this stage, we do not do paragraph tags unless have strip_html -
|
---|
| 112 | # it will result in a huge mess of non-xml
|
---|
| 113 | return unless $strip_html;
|
---|
| 114 |
|
---|
| 115 | my $new_text = $text;
|
---|
| 116 |
|
---|
| 117 | # if we have <pre> tags, we can have < > inside them, need to delete
|
---|
| 118 | # the <> before stripping tags
|
---|
| 119 | $new_text =~ s/<pre>(.*?)<\/pre>/$self->remove_gtlt($1,$para)/gse;
|
---|
| 120 |
|
---|
| 121 | if ($para eq "") {
|
---|
| 122 | # just remove all tags
|
---|
| 123 | $new_text =~ s/<[^>]*>//gs;
|
---|
| 124 | } else {
|
---|
| 125 | # strip all tags except <p> tags which get turned into $para
|
---|
| 126 | $new_text =~ s/<([^>]*)>/$self->process_tags($1, $para)/gse;
|
---|
| 127 |
|
---|
| 128 | }
|
---|
| 129 | return $new_text;
|
---|
| 130 | }
|
---|
| 131 | #this function strips the html tags from the doc if ($strip_html) and
|
---|
| 132 | # if ($para) replaces <p> with <Paragraph> tags.
|
---|
| 133 | # if both are false, the original text is returned
|
---|
| 134 | #assumes that <pre> and </pre> have no spaces, and removes all < and > inside
|
---|
| 135 | #these tags
|
---|
| 136 | sub preprocess_text_old_and_slow {
|
---|
| 137 | my $self = shift (@_);
|
---|
| 138 | my ($text, $strip_html, $para) = @_;
|
---|
| 139 | my ($outtext) = "";
|
---|
| 140 | if ($strip_html) {
|
---|
| 141 | while ($text =~ /<([^>]*)>/ && $text ne "") {
|
---|
| 142 |
|
---|
| 143 | my $tag = $1;
|
---|
| 144 | $outtext .= $`." "; #add everything before the matched tag
|
---|
| 145 | $text = $'; #'everything after the matched tag
|
---|
| 146 | if ($para && $tag =~ /^\s*p\s/i) {
|
---|
| 147 | $outtext .= $para;
|
---|
| 148 | }
|
---|
| 149 | elsif ($tag =~ /^pre$/) { # a pre tag
|
---|
| 150 | $text =~ /<\/pre>/; # find the closing pre tag
|
---|
| 151 | my $tmp_text = $`; #everything before the closing pre tag
|
---|
| 152 | $text = $'; #'everything after the </pre>
|
---|
| 153 | $tmp_text =~ s/[<>]//g; # remove all < and >
|
---|
| 154 | $outtext.= $tmp_text . " ";
|
---|
| 155 | }
|
---|
| 156 | }
|
---|
| 157 |
|
---|
| 158 | $outtext .= $text; # add any remaining text
|
---|
| 159 | return $outtext;
|
---|
| 160 | } #if strip_html
|
---|
| 161 |
|
---|
| 162 | #if ($para) {
|
---|
| 163 | #$text =~ s/(<p\b)/$para$1/gi;
|
---|
| 164 | #return $text;
|
---|
| 165 | # }
|
---|
| 166 | return $text;
|
---|
| 167 | }
|
---|
| 168 |
|
---|
| 169 |
|
---|
| 170 |
|
---|
| 171 | sub filter_text {
|
---|
| 172 | # $self->filter_text ($field, $new_text);
|
---|
| 173 | # don't want to do anything for this version, however,
|
---|
| 174 | # in a particular collection you might want to override
|
---|
| 175 | # this method to post-process certain fields depending on
|
---|
| 176 | # the field, or whether we are outputting it for indexing
|
---|
| 177 | }
|
---|
| 178 |
|
---|
| 179 | sub text {
|
---|
| 180 | my $self = shift (@_);
|
---|
| 181 | my ($doc_obj) = @_;
|
---|
| 182 | my $handle = $self->{'output_handle'};
|
---|
| 183 | my $outhandle = $self->{'outhandle'};
|
---|
| 184 |
|
---|
| 185 | # only output this document if it is one to be indexed
|
---|
| 186 | return if ($doc_obj->get_doc_type() ne "indexed_doc");
|
---|
| 187 |
|
---|
| 188 | my $indexed_doc = $self->is_subcollection_doc($doc_obj);
|
---|
| 189 |
|
---|
| 190 | # this is another document
|
---|
| 191 | $self->{'num_docs'} += 1;
|
---|
| 192 |
|
---|
| 193 | # get the parameters for the output
|
---|
| 194 | # split on : just in case there is subcoll and lang stuff
|
---|
| 195 | my ($fields) = split (/:/, $self->{'index'});
|
---|
| 196 |
|
---|
| 197 | # we always do text and index on Doc and Sec levels
|
---|
| 198 | my ($documenttag) = "\n<". $level_map{'document'} . ">\n";
|
---|
| 199 | my ($documentendtag) = "\n</". $level_map{'document'} . ">\n";
|
---|
| 200 | my ($sectiontag) = "\n<". $level_map{'section'} . ">\n";
|
---|
| 201 | my ($sectionendtag) = "\n</". $level_map{'section'} . ">\n";
|
---|
| 202 |
|
---|
| 203 | my ($paratag) = "";
|
---|
| 204 |
|
---|
| 205 | # paragraph tags will only be used for indexing (can't retrieve
|
---|
| 206 | # paragraphs), and can ony be used if we are stripping HTML tags
|
---|
| 207 | if ($self->{'indexing_text'} && $self->{'levels'}->{'paragraph'}) {
|
---|
| 208 | if ($self->{'strip_html'}) {
|
---|
| 209 | $paratag = "<". $level_map{'paragraph'} . ">";
|
---|
| 210 | } else {
|
---|
| 211 | print $outhandle "Paragraph level can not be used with no_strip_html!. Not indexing Paragraphs.\n";
|
---|
| 212 | }
|
---|
| 213 | }
|
---|
| 214 |
|
---|
| 215 | my $doc_section = 0; # just for this document
|
---|
| 216 |
|
---|
| 217 | my $text = $documenttag;
|
---|
| 218 |
|
---|
| 219 | # get the text for this document
|
---|
| 220 | my $section = $doc_obj->get_top_section();
|
---|
| 221 |
|
---|
| 222 | while (defined $section) {
|
---|
| 223 | # update a few statistics
|
---|
| 224 | $doc_section++;
|
---|
| 225 | $self->{'num_sections'} += 1;
|
---|
| 226 | $text .= "$sectiontag";
|
---|
| 227 |
|
---|
| 228 | my $indexed_section = $doc_obj->get_metadata_element($section, "gsdldoctype") || "indexed_section";
|
---|
| 229 | if (($indexed_doc == 0) || ($indexed_section ne "indexed_section" && $indexed_section ne "indexed_doc")) {
|
---|
| 230 | # we are not actually indexing anything for this document,
|
---|
| 231 | # but we want to keep the section numbers the same, so we just
|
---|
| 232 | # output section tags for each section (which is done above)
|
---|
| 233 | $text .= "$sectionendtag";
|
---|
| 234 | $section = $doc_obj->get_next_section($section);
|
---|
| 235 | next;
|
---|
| 236 | }
|
---|
| 237 |
|
---|
| 238 | $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
|
---|
| 239 | foreach my $field (split (/;/, $fields)) {
|
---|
| 240 | # only deal with this field if it doesn't start with top or
|
---|
| 241 | # this is the first section
|
---|
| 242 | my $real_field = $field;
|
---|
| 243 | next if (($real_field =~ s/^top//) && ($doc_section != 1));
|
---|
| 244 |
|
---|
| 245 | my $new_text = "";
|
---|
| 246 |
|
---|
| 247 | # we get allfields by default - do nothing
|
---|
| 248 | if ($real_field eq "allfields") {
|
---|
| 249 |
|
---|
| 250 | }
|
---|
| 251 |
|
---|
| 252 | # metadata - output all metadata we know about except gsdl stuff
|
---|
| 253 | elsif ($real_field eq "metadata") {
|
---|
| 254 | my $shortname = "";
|
---|
| 255 | my $metadata = $doc_obj->get_all_metadata ($section);
|
---|
| 256 | foreach my $pair (@$metadata) {
|
---|
| 257 | my ($mfield, $mvalue) = (@$pair);
|
---|
| 258 | # check fields here, maybe others dont want - change to use dontindex!!
|
---|
| 259 | if ($mfield ne "Identifier"
|
---|
| 260 | && $mfield !~ /^gsdl/
|
---|
| 261 | && $mfield ne "classifytype"
|
---|
| 262 | && $mfield ne "assocfilepath"
|
---|
| 263 | && defined $mvalue && $mvalue ne "") {
|
---|
| 264 |
|
---|
| 265 | if (defined $self->{'indexfieldmap'}->{$mfield}) {
|
---|
| 266 | $shortname = $self->{'indexfieldmap'}->{$mfield};
|
---|
| 267 | }
|
---|
| 268 | else {
|
---|
| 269 | $shortname = $self->create_shortname($mfield);
|
---|
| 270 | $self->{'indexfieldmap'}->{$mfield} = $shortname;
|
---|
| 271 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
---|
| 272 | }
|
---|
| 273 | $new_text .= "$paratag<$shortname>$mvalue</$shortname>\n";
|
---|
| 274 | if (!defined $self->{'indexfields'}->{$mfield}) {
|
---|
| 275 | $self->{'indexfields'}->{$mfield} = 1;
|
---|
| 276 | }
|
---|
| 277 | }
|
---|
| 278 | }
|
---|
| 279 | }
|
---|
| 280 | else {
|
---|
| 281 | #individual metadata and or text specified - could be
|
---|
| 282 | # a comma separated list
|
---|
| 283 | my $shortname="";
|
---|
| 284 | if (defined $self->{'indexfieldmap'}->{$real_field}) {
|
---|
| 285 | $shortname = $self->{'indexfieldmap'}->{$real_field};
|
---|
| 286 | }
|
---|
| 287 | else {
|
---|
| 288 | $shortname = $self->create_shortname($real_field);
|
---|
| 289 | $self->{'indexfieldmap'}->{$real_field} = $shortname;
|
---|
| 290 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
---|
| 291 | }
|
---|
| 292 | my @metadata_list = ();
|
---|
| 293 | foreach my $submeta (split /,/, $real_field) {
|
---|
| 294 | if ($submeta eq "text") {
|
---|
| 295 | my $section_text = $doc_obj->get_text($section);
|
---|
| 296 | if ($self->{'indexing_text'}) {
|
---|
| 297 | # tag the text with <Text>...</Text>, add the <Paragraph> tags and strip out html if needed
|
---|
| 298 | $new_text .= "$paratag<$shortname>\n";
|
---|
| 299 | if ($paratag ne "") {
|
---|
| 300 | $section_text = $self->preprocess_text($section_text, $self->{'strip_html'}, "</$shortname>$paratag<$shortname>");
|
---|
| 301 | }
|
---|
| 302 | else {
|
---|
| 303 | $section_text = $self->preprocess_text($section_text, $self->{'strip_html'}, "");
|
---|
| 304 | }
|
---|
| 305 | $new_text .= "$section_text</$shortname>\n";
|
---|
| 306 | }
|
---|
| 307 | else {
|
---|
| 308 | # leave html stuff in, and don't add Paragraph tags - never retrieve paras at the moment
|
---|
| 309 | $new_text .= $section_text;
|
---|
| 310 | }
|
---|
| 311 | }
|
---|
| 312 | else {
|
---|
| 313 | my @section_metadata = @{$doc_obj->get_metadata ($section, $submeta)};
|
---|
| 314 | if ($section ne $doc_obj->get_top_section() && $self->{'indexing_text'} && defined ($self->{'sections_index_document_metadata'})) {
|
---|
| 315 | if ($self->{'sections_index_document_metadata'} eq "always" || ( scalar(@section_metadata) == 0 && $self->{'sections_index_document_metadata'} eq "unless_section_metadata_exists")) {
|
---|
| 316 | push (@section_metadata, @{$doc_obj->get_metadata ($doc_obj->get_top_section(), $submeta)});
|
---|
| 317 | }
|
---|
| 318 | }
|
---|
| 319 | push (@metadata_list, @section_metadata);
|
---|
| 320 | }
|
---|
| 321 | }
|
---|
| 322 | foreach my $item (@metadata_list) {
|
---|
| 323 | $new_text .= "$paratag<$shortname>$item</$shortname>\n";
|
---|
| 324 | }
|
---|
| 325 | }
|
---|
| 326 |
|
---|
| 327 | # filter the text
|
---|
| 328 | $self->filter_text ($field, $new_text);
|
---|
| 329 |
|
---|
| 330 | $self->{'num_processed_bytes'} += length ($new_text);
|
---|
| 331 | $text .= "$new_text";
|
---|
| 332 | } # foreach field
|
---|
| 333 |
|
---|
| 334 | $text .= "$sectionendtag";
|
---|
| 335 | $section = $doc_obj->get_next_section($section);
|
---|
| 336 | } # while defined section
|
---|
| 337 | print $handle "$text\n$documentendtag";
|
---|
| 338 |
|
---|
| 339 | }
|
---|
| 340 |
|
---|
| 341 | #chooses the first two letters or digits for the shortname
|
---|
| 342 | #now ignores non-letdig characters
|
---|
| 343 | sub create_shortname {
|
---|
| 344 | my $self = shift(@_);
|
---|
| 345 |
|
---|
| 346 | my ($realname) = @_;
|
---|
| 347 | #take the first two chars
|
---|
| 348 | my $shortname;
|
---|
| 349 | if ($realname =~ /^[^\w]*(\w)[^\w]*(\w)/) {
|
---|
| 350 | $shortname = "$1$2";
|
---|
| 351 | } else {
|
---|
| 352 | # there aren't two letdig's in the field - try arbitrary combinations
|
---|
| 353 | $realname = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
|
---|
| 354 | $shortname = "AB";
|
---|
| 355 | }
|
---|
| 356 | $shortname =~ tr/a-z/A-Z/;
|
---|
| 357 |
|
---|
| 358 | #if already used, take the first and third letdigs and so on
|
---|
| 359 | my $count = 1;
|
---|
| 360 | while (defined $self->{'indexfieldmap'}->{$shortname}) {
|
---|
| 361 | if ($realname =~ /^[^\w]*(\w)([^\w]*\w){$count}[^\w]*(\w)/) {
|
---|
| 362 | $shortname = "$1$3";
|
---|
| 363 | $count++;
|
---|
| 364 | $shortname =~ tr/a-z/A-Z/;
|
---|
| 365 |
|
---|
| 366 | }
|
---|
| 367 | else {
|
---|
| 368 | #remove up to and incl the first letdig
|
---|
| 369 | $realname =~ s/^[^\w]*\w//;
|
---|
| 370 | $count = 0;
|
---|
| 371 | }
|
---|
| 372 | }
|
---|
| 373 |
|
---|
| 374 | return $shortname;
|
---|
| 375 | }
|
---|
| 376 |
|
---|
| 377 | 1;
|
---|
| 378 |
|
---|