| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # HTMLPlug.pm -- basic html plugin
|
|---|
| 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) 1999 New Zealand Digital Library Project
|
|---|
| 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 | #
|
|---|
| 28 | # Note that this plugin handles frames only in a very simple way
|
|---|
| 29 | # i.e. each frame is treated as a separate document. This means
|
|---|
| 30 | # search results will contain links to individual frames rather
|
|---|
| 31 | # than linking to the top level frameset.
|
|---|
| 32 | # There may also be some problems caused by the _parent target
|
|---|
| 33 | # (it's removed by this plugin)
|
|---|
| 34 | #
|
|---|
| 35 |
|
|---|
| 36 | package HTMLPlug;
|
|---|
| 37 |
|
|---|
| 38 | use BasPlug;
|
|---|
| 39 | use ghtml;
|
|---|
| 40 | use unicode;
|
|---|
| 41 | use util;
|
|---|
| 42 | use parsargv;
|
|---|
| 43 | use XMLParser;
|
|---|
| 44 |
|
|---|
| 45 | sub BEGIN {
|
|---|
| 46 | @HTMLPlug::ISA = ('BasPlug');
|
|---|
| 47 | }
|
|---|
| 48 |
|
|---|
| 49 | use strict; # every perl program should have this!
|
|---|
| 50 | no strict 'refs'; # make an exception so we can use variables as filehandles
|
|---|
| 51 |
|
|---|
| 52 | my $arguments =
|
|---|
| 53 | [ { 'name' => "process_exp",
|
|---|
| 54 | 'desc' => "{BasPlug.process_exp}",
|
|---|
| 55 | 'type' => "regexp",
|
|---|
| 56 | 'deft' => &get_default_process_exp() },
|
|---|
| 57 | { 'name' => "block_exp",
|
|---|
| 58 | 'desc' => "{BasPlug.block_exp}",
|
|---|
| 59 | 'type' => 'regexp',
|
|---|
| 60 | 'deft' => &get_default_block_exp() },
|
|---|
| 61 | { 'name' => "nolinks",
|
|---|
| 62 | 'desc' => "{HTMLPlug.nolinks}",
|
|---|
| 63 | 'type' => "flag" },
|
|---|
| 64 | { 'name' => "keep_head",
|
|---|
| 65 | 'desc' => "{HTMLPlug.keep_head}",
|
|---|
| 66 | 'type' => "flag" },
|
|---|
| 67 | { 'name' => "no_metadata",
|
|---|
| 68 | 'desc' => "{HTMLPlug.no_metadata}",
|
|---|
| 69 | 'type' => "flag" },
|
|---|
| 70 | { 'name' => "metadata_fields",
|
|---|
| 71 | 'desc' => "{HTMLPlug.metadata_fields}",
|
|---|
| 72 | 'type' => "string",
|
|---|
| 73 | 'deft' => "Title" },
|
|---|
| 74 | { 'name' => "hunt_creator_metadata",
|
|---|
| 75 | 'desc' => "{HTMLPlug.hunt_creator_metadata}",
|
|---|
| 76 | 'type' => "flag" },
|
|---|
| 77 | { 'name' => "file_is_url",
|
|---|
| 78 | 'desc' => "{HTMLPlug.file_is_url}",
|
|---|
| 79 | 'type' => "flag" },
|
|---|
| 80 | { 'name' => "assoc_files",
|
|---|
| 81 | 'desc' => "{HTMLPlug.assoc_files}",
|
|---|
| 82 | 'type' => "regexp",
|
|---|
| 83 | 'deft' => &get_default_block_exp() },
|
|---|
| 84 | { 'name' => "rename_assoc_files",
|
|---|
| 85 | 'desc' => "{HTMLPlug.rename_assoc_files}",
|
|---|
| 86 | 'type' => "flag" },
|
|---|
| 87 | { 'name' => "title_sub",
|
|---|
| 88 | 'desc' => "{HTMLPlug.title_sub}",
|
|---|
| 89 | 'type' => "string",
|
|---|
| 90 | 'deft' => "" },
|
|---|
| 91 | { 'name' => "description_tags",
|
|---|
| 92 | 'desc' => "{HTMLPlug.description_tags}",
|
|---|
| 93 | 'type' => "flag" },
|
|---|
| 94 | { 'name' => "no_strip_metadata_html",
|
|---|
| 95 | 'desc' => "{HTMLPlug.no_strip_metadata_html}",
|
|---|
| 96 | 'type' => "string",
|
|---|
| 97 | 'deft' => "",
|
|---|
| 98 | 'reqd' => "no"}
|
|---|
| 99 | ];
|
|---|
| 100 |
|
|---|
| 101 | my $options = { 'name' => "HTMLPlug",
|
|---|
| 102 | 'desc' => "{HTMLPlug.desc}",
|
|---|
| 103 | 'abstract' => "no",
|
|---|
| 104 | 'inherits' => "yes",
|
|---|
| 105 | 'args' => $arguments };
|
|---|
| 106 |
|
|---|
| 107 | sub new {
|
|---|
| 108 | my $class = shift (@_);
|
|---|
| 109 | my $self = new BasPlug ($class, @_);
|
|---|
| 110 | $self->{'plugin_type'} = "HTMLPlug";
|
|---|
| 111 | # 14-05-02 To allow for proper inheritance of arguments - John Thompson
|
|---|
| 112 | my $option_list = $self->{'option_list'};
|
|---|
| 113 | push( @{$option_list}, $options );
|
|---|
| 114 |
|
|---|
| 115 | if (!parsargv::parse(\@_,
|
|---|
| 116 | q^nolinks^, \$self->{'nolinks'},
|
|---|
| 117 | q^keep_head^, \$self->{'keep_head'},
|
|---|
| 118 | q^no_metadata^, \$self->{'no_metadata'},
|
|---|
| 119 | q^metadata_fields/.*/Title^, \$self->{'metadata_fields'},
|
|---|
| 120 | q^hunt_creator_metadata^, \$self->{'hunt_creator_metadata'},
|
|---|
| 121 | q^w3mir^, \$self->{'w3mir'},
|
|---|
| 122 | q^file_is_url^, \$self->{'file_is_url'},
|
|---|
| 123 | q^assoc_files/.*/(?i)\.(jpe?g|jpe|gif|png|css)$^, \$self->{'assoc_files'},
|
|---|
| 124 | q^rename_assoc_files^, \$self->{'rename_assoc_files'},
|
|---|
| 125 | q^title_sub/.*/^, \$self->{'title_sub'},
|
|---|
| 126 | q^description_tags^, \$self->{'description_tags'},
|
|---|
| 127 | q^no_strip_metadata_html/.*/^, \$self->{'no_strip_metadata_html'},
|
|---|
| 128 | "allow_extra_options")) {
|
|---|
| 129 |
|
|---|
| 130 | print STDERR "\nIncorrect options passed to HTMLPlug, check your collect.cfg configuration file\n";
|
|---|
| 131 | $self->print_txt_usage(""); # Use default resource bundle
|
|---|
| 132 | die "\n";
|
|---|
| 133 | }
|
|---|
| 134 |
|
|---|
| 135 | # retain this for backward compatibility (w3mir option was replaced by
|
|---|
| 136 | # file_is_url)
|
|---|
| 137 | if ($self->{'w3mir'}) {
|
|---|
| 138 | $self->{'file_is_url'} = 1;
|
|---|
| 139 | }
|
|---|
| 140 |
|
|---|
| 141 | $self->{'aux_files'} = {};
|
|---|
| 142 | $self->{'dir_num'} = 0;
|
|---|
| 143 | $self->{'file_num'} = 0;
|
|---|
| 144 |
|
|---|
| 145 | return bless $self, $class;
|
|---|
| 146 | }
|
|---|
| 147 |
|
|---|
| 148 | # may want to use (?i)\.(gif|jpe?g|jpe|png|css|js(?:@.*)?)$
|
|---|
| 149 | # if have eg <script language="javascript" src="img/lib.js@123">
|
|---|
| 150 | sub get_default_block_exp {
|
|---|
| 151 | my $self = shift (@_);
|
|---|
| 152 |
|
|---|
| 153 | return q^(?i)\.(gif|jpe?g|jpe|jpg|png|css)$^;
|
|---|
| 154 | }
|
|---|
| 155 |
|
|---|
| 156 | sub get_default_process_exp {
|
|---|
| 157 | my $self = shift (@_);
|
|---|
| 158 |
|
|---|
| 159 | # the last option is an attempt to encode the concept of an html query ...
|
|---|
| 160 | return q^(?i)(\.html?|\.shtml|\.shm|\.asp|\.php\d?|\.cgi|.+\?.+=.*)$^;
|
|---|
| 161 | }
|
|---|
| 162 |
|
|---|
| 163 | sub store_block_files
|
|---|
| 164 | {
|
|---|
| 165 | my $self =shift (@_);
|
|---|
| 166 | my ($filename) = @_;
|
|---|
| 167 | my $html_fname = $filename;
|
|---|
| 168 | my @file_blocks;
|
|---|
| 169 |
|
|---|
| 170 | my ($language, $encoding) = $self->textcat_get_language_encoding ($filename);
|
|---|
| 171 |
|
|---|
| 172 | # read in file ($text will be in utf8)
|
|---|
| 173 | my $text = "";
|
|---|
| 174 | $self->read_file ($filename, $encoding, $language, \$text);
|
|---|
| 175 | my $textref = \$text;
|
|---|
| 176 | my $opencom = '(?:<!--|<!(?:—|—|--))';
|
|---|
| 177 | my $closecom = '(?:-->|(?:—|—|--)>)';
|
|---|
| 178 | $$textref =~ s/$opencom(.*?)$closecom//gs;
|
|---|
| 179 |
|
|---|
| 180 | my $attval = "\\\"[^\\\"]+\\\"|[^\\s>]+";
|
|---|
| 181 | my @img_matches = ($$textref =~ m/<img[^>]*?src\s*=\s*($attval)[^>]*>/igs);
|
|---|
| 182 | my @usemap_matches = ($$textref =~ m/<img[^>]*?usemap\s*=\s*($attval)[^>]*>/igs);
|
|---|
| 183 | my @link_matches = ($$textref =~ m/<link[^>]*?href\s*=\s*($attval)[^>]*>/igs);
|
|---|
| 184 | my @embed_matches = ($$textref =~ m/<embed[^>]*?src\s*=\s*($attval)[^>]*>/igs);
|
|---|
| 185 |
|
|---|
| 186 | foreach my $link (@img_matches, @usemap_matches, @link_matches, @embed_matches) {
|
|---|
| 187 |
|
|---|
| 188 | # remove quotes from link at start and end if necessary
|
|---|
| 189 | if ($link=~/^\"/) {
|
|---|
| 190 | $link=~s/^\"//;
|
|---|
| 191 | $link=~s/\"$//;
|
|---|
| 192 | }
|
|---|
| 193 |
|
|---|
| 194 | $link =~ s/\#.*$//s; # remove any anchor names, e.g. foo.html#name becomes foo.html
|
|---|
| 195 |
|
|---|
| 196 | if ($link !~ m@^/@ && $link !~ m/^([A-Z]:?)\\/) {
|
|---|
| 197 | # Turn relative file path into full path
|
|---|
| 198 | my $dirname = &File::Basename::dirname($filename);
|
|---|
| 199 | $link = &util::filename_cat($dirname, $link);
|
|---|
| 200 | }
|
|---|
| 201 | $link = $self->eval_dir_dots($link);
|
|---|
| 202 |
|
|---|
| 203 | $self->{'file_blocks'}->{$link} = 1;
|
|---|
| 204 | }
|
|---|
| 205 | }
|
|---|
| 206 |
|
|---|
| 207 |
|
|---|
| 208 | # do plugin specific processing of doc_obj
|
|---|
| 209 | sub process {
|
|---|
| 210 | my $self = shift (@_);
|
|---|
| 211 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
|
|---|
| 212 | my $outhandle = $self->{'outhandle'};
|
|---|
| 213 |
|
|---|
| 214 | print STDERR "<Processing n='$file' p='HTMLPlug'>\n" if ($gli);
|
|---|
| 215 |
|
|---|
| 216 | print $outhandle "HTMLPlug: processing $file\n"
|
|---|
| 217 | if $self->{'verbosity'} > 1;
|
|---|
| 218 |
|
|---|
| 219 | if ($ENV{'GSDLOS'} =~ /^windows/i) {
|
|---|
| 220 | # this makes life so much easier... perl can cope with unix-style '/'s.
|
|---|
| 221 | $base_dir =~ s@(\\)+@/@g;
|
|---|
| 222 | $file =~ s@(\\)+@/@g;
|
|---|
| 223 | }
|
|---|
| 224 |
|
|---|
| 225 | # reset per-doc stuff...
|
|---|
| 226 | $self->{'aux_files'} = {};
|
|---|
| 227 | $self->{'dir_num'} = 0;
|
|---|
| 228 | $self->{'file_num'} = 0;
|
|---|
| 229 |
|
|---|
| 230 | my $cursection = $doc_obj->get_top_section();
|
|---|
| 231 |
|
|---|
| 232 | $self->extract_metadata ($textref, $metadata, $doc_obj, $cursection)
|
|---|
| 233 | unless $self->{'no_metadata'} || $self->{'description_tags'};
|
|---|
| 234 |
|
|---|
| 235 | # Store URL for page as metadata - this can be used for an
|
|---|
| 236 | # altavista style search interface. The URL won't be valid
|
|---|
| 237 | # unless the file structure contains the domain name (i.e.
|
|---|
| 238 | # like when w3mir is used to download a website).
|
|---|
| 239 |
|
|---|
| 240 | # URL metadata (even invalid ones) are used to support internal
|
|---|
| 241 | # links, so even if 'file_is_url' is off, still need to store info
|
|---|
| 242 |
|
|---|
| 243 | my $web_url = "http://$file";
|
|---|
| 244 | $doc_obj->add_metadata($cursection, "URL", $web_url);
|
|---|
| 245 |
|
|---|
| 246 | if ($self->{'file_is_url'}) {
|
|---|
| 247 | $doc_obj->add_metadata($cursection, "weblink", "<a href=\"$web_url\">");
|
|---|
| 248 | $doc_obj->add_metadata($cursection, "webicon", "_iconworld_");
|
|---|
| 249 | $doc_obj->add_metadata($cursection, "/weblink", "</a>");
|
|---|
| 250 | }
|
|---|
| 251 |
|
|---|
| 252 | if ($self->{'description_tags'}) {
|
|---|
| 253 | # remove the html header - note that doing this here means any
|
|---|
| 254 | # sections defined within the header will be lost (so all <Section>
|
|---|
| 255 | # tags must appear within the body of the HTML)
|
|---|
| 256 | my ($head_keep) = ($$textref =~ m/^(.*?)<body[^>]*>/is);
|
|---|
| 257 |
|
|---|
| 258 | $$textref =~ s/^.*?<body[^>]*>//is;
|
|---|
| 259 | $$textref =~ s/(<\/body[^>]*>|<\/html[^>]*>)//isg;
|
|---|
| 260 |
|
|---|
| 261 | my $opencom = '(?:<!--|<!(?:—|—|--))';
|
|---|
| 262 | my $closecom = '(?:-->|(?:—|—|--)>)';
|
|---|
| 263 |
|
|---|
| 264 | my $lt = '(?:<|<)';
|
|---|
| 265 | my $gt = '(?:>|>)';
|
|---|
| 266 | my $quot = '(?:"|"|”|“)';
|
|---|
| 267 |
|
|---|
| 268 | my $dont_strip = '';
|
|---|
| 269 | if ($self->{'no_strip_metadata_html'}) {
|
|---|
| 270 | ($dont_strip = $self->{'no_strip_metadata_html'}) =~ s{,}{|}g;
|
|---|
| 271 | }
|
|---|
| 272 |
|
|---|
| 273 | my $found_something = 0; my $top = 1;
|
|---|
| 274 | while ($$textref =~ s/^(.*?)$opencom(.*?)$closecom//s) {
|
|---|
| 275 | my $text = $1;
|
|---|
| 276 | my $comment = $2;
|
|---|
| 277 | if (defined $text) {
|
|---|
| 278 | # text before a comment - note that getting to here
|
|---|
| 279 | # doesn't necessarily mean there are Section tags in
|
|---|
| 280 | # the document
|
|---|
| 281 | $self->process_section(\$text, $base_dir, $file, $doc_obj, $cursection);
|
|---|
| 282 | }
|
|---|
| 283 | while ($comment =~ s/$lt(.*?)$gt//s) {
|
|---|
| 284 | my $tag = $1;
|
|---|
| 285 | if ($tag eq "Section") {
|
|---|
| 286 | $found_something = 1;
|
|---|
| 287 | $cursection = $doc_obj->insert_section($doc_obj->get_end_child($cursection)) unless $top;
|
|---|
| 288 | $top = 0;
|
|---|
| 289 | } elsif ($tag eq "/Section") {
|
|---|
| 290 | $found_something = 1;
|
|---|
| 291 | $cursection = $doc_obj->get_parent_section ($cursection);
|
|---|
| 292 | } elsif ($tag =~ /^Metadata name=$quot(.*?)$quot/s) {
|
|---|
| 293 | my $metaname = $1;
|
|---|
| 294 | my $accumulate = $tag =~ /mode=${quot}accumulate${quot}/ ? 1 : 0;
|
|---|
| 295 | $comment =~ s/^(.*?)$lt\/Metadata$gt//s;
|
|---|
| 296 | my $metavalue = $1;
|
|---|
| 297 | $metavalue =~ s/^\s+//;
|
|---|
| 298 | $metavalue =~ s/\s+$//;
|
|---|
| 299 | # assume that no metadata value intentionally includes
|
|---|
| 300 | # carriage returns or HTML tags (if they're there they
|
|---|
| 301 | # were probably introduced when converting to HTML from
|
|---|
| 302 | # some other format).
|
|---|
| 303 | # actually some people want to have html tags in their
|
|---|
| 304 | # metadata.
|
|---|
| 305 | $metavalue =~ s/[\cJ\cM]/ /sg;
|
|---|
| 306 | $metavalue =~ s/<[^>]+>//sg
|
|---|
| 307 | unless $dont_strip && ($dont_strip eq 'all' || $metaname =~ /^($dont_strip)$/);
|
|---|
| 308 | $metavalue =~ s/\s+/ /sg;
|
|---|
| 309 | if ($accumulate) {
|
|---|
| 310 | $doc_obj->add_utf8_metadata($cursection, $metaname, $metavalue);
|
|---|
| 311 | } else {
|
|---|
| 312 | $doc_obj->set_utf8_metadata_element($cursection, $metaname, $metavalue);
|
|---|
| 313 | }
|
|---|
| 314 | } elsif ($tag eq "Description" || $tag eq "/Description") {
|
|---|
| 315 | # do nothing with containing Description tags
|
|---|
| 316 | } else {
|
|---|
| 317 | # simple HTML tag (probably created by the conversion
|
|---|
| 318 | # to HTML from some other format) - we'll ignore it and
|
|---|
| 319 | # hope for the best ;-)
|
|---|
| 320 | }
|
|---|
| 321 | }
|
|---|
| 322 | }
|
|---|
| 323 | if ($cursection ne "") {
|
|---|
| 324 | print $outhandle "HTMLPlug: WARNING: $file contains unmatched <Section></Section> tags\n";
|
|---|
| 325 | }
|
|---|
| 326 |
|
|---|
| 327 | $$textref =~ s/^.*?<body[^>]*>//is;
|
|---|
| 328 | $$textref =~ s/(<\/body[^>]*>|<\/html[^>]*>)//isg;
|
|---|
| 329 | if ($$textref =~ /\S/) {
|
|---|
| 330 | if (!$found_something) {
|
|---|
| 331 | if ($self->{'verbosity'} > 2) {
|
|---|
| 332 | print $outhandle "HTMLPlug: WARNING: $file appears to contain no Section tags so\n";
|
|---|
| 333 | print $outhandle " will be processed as a single section document\n";
|
|---|
| 334 | }
|
|---|
| 335 |
|
|---|
| 336 | # go ahead and process single-section document
|
|---|
| 337 | $self->process_section($textref, $base_dir, $file, $doc_obj, $cursection);
|
|---|
| 338 |
|
|---|
| 339 | # if document contains no Section tags we'll go ahead
|
|---|
| 340 | # and extract metadata (this won't have been done
|
|---|
| 341 | # above as the -description_tags option prevents it)
|
|---|
| 342 | my $complete_text = $head_keep.$doc_obj->get_text($cursection);
|
|---|
| 343 | $self->extract_metadata (\$complete_text, $metadata, $doc_obj, $cursection)
|
|---|
| 344 | unless $self->{'no_metadata'};
|
|---|
| 345 |
|
|---|
| 346 | } else {
|
|---|
| 347 | print $outhandle "HTMLPlug: WARNING: $file contains the following text outside\n";
|
|---|
| 348 | print $outhandle " of the final closing </Section> tag. This text will\n";
|
|---|
| 349 | print $outhandle " be ignored.";
|
|---|
| 350 |
|
|---|
| 351 | my ($text);
|
|---|
| 352 | if (length($$textref) > 30) {
|
|---|
| 353 | $text = substr($$textref, 0, 30) . "...";
|
|---|
| 354 | } else {
|
|---|
| 355 | $text = $$textref;
|
|---|
| 356 | }
|
|---|
| 357 | $text =~ s/\n/ /isg;
|
|---|
| 358 | print $outhandle " ($text)\n";
|
|---|
| 359 | }
|
|---|
| 360 | } elsif (!$found_something) {
|
|---|
| 361 |
|
|---|
| 362 | if ($self->{'verbosity'} > 2) {
|
|---|
| 363 | # may get to here if document contained no valid Section
|
|---|
| 364 | # tags but did contain some comments. The text will have
|
|---|
| 365 | # been processed already but we should print the warning
|
|---|
| 366 | # as above and extract metadata
|
|---|
| 367 | print $outhandle "HTMLPlug: WARNING: $file appears to contain no Section tags and\n";
|
|---|
| 368 | print $outhandle " is blank or empty. Metadata will be assigned if present.\n";
|
|---|
| 369 | }
|
|---|
| 370 |
|
|---|
| 371 | my $complete_text = $head_keep.$doc_obj->get_text($cursection);
|
|---|
| 372 | $self->extract_metadata (\$complete_text, $metadata, $doc_obj, $cursection)
|
|---|
| 373 | unless $self->{'no_metadata'};
|
|---|
| 374 | }
|
|---|
| 375 |
|
|---|
| 376 | } else {
|
|---|
| 377 |
|
|---|
| 378 | # remove header and footer
|
|---|
| 379 | if (!$self->{'keep_head'} || $self->{'description_tags'}) {
|
|---|
| 380 | $$textref =~ s/^.*?<body[^>]*>//is;
|
|---|
| 381 | $$textref =~ s/(<\/body[^>]*>|<\/html[^>]*>)//isg;
|
|---|
| 382 | }
|
|---|
| 383 |
|
|---|
| 384 | # single section document
|
|---|
| 385 | $self->process_section($textref, $base_dir, $file, $doc_obj, $cursection);
|
|---|
| 386 | }
|
|---|
| 387 | return 1;
|
|---|
| 388 | }
|
|---|
| 389 |
|
|---|
| 390 | # note that process_section may be called multiple times for a single
|
|---|
| 391 | # section (relying on the fact that add_utf8_text appends the text to any
|
|---|
| 392 | # that may exist already).
|
|---|
| 393 | sub process_section {
|
|---|
| 394 | my $self = shift (@_);
|
|---|
| 395 | my ($textref, $base_dir, $file, $doc_obj, $cursection) = @_;
|
|---|
| 396 | # trap links
|
|---|
| 397 | if (!$self->{'nolinks'}) {
|
|---|
| 398 |
|
|---|
| 399 | # usemap="./#index" not handled correctly => change to "#index"
|
|---|
| 400 | $$textref =~ s/(<img[^>]*?usemap\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
|
|---|
| 401 | $self->replace_usemap_links($1, $2, $3)/isge;
|
|---|
| 402 |
|
|---|
| 403 | $$textref =~ s/(<(?:a|area|frame|link|script)\s+[^>]*?\s*(?:href|src)\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
|
|---|
| 404 | $self->replace_href_links ($1, $2, $3, $base_dir, $file, $doc_obj, $cursection)/isge;
|
|---|
| 405 | }
|
|---|
| 406 |
|
|---|
| 407 | # trap images
|
|---|
| 408 |
|
|---|
| 409 | # allow spaces if inside quotes - jrm21
|
|---|
| 410 | $$textref =~ s/(<(?:img|embed)[^>]*?src\s*=\s*)(\"[^\"]+\"|[^\s>]+)([^>]*>)/
|
|---|
| 411 | $self->replace_images ($1, $2, $3, $base_dir, $file, $doc_obj, $cursection)/isge;
|
|---|
| 412 |
|
|---|
| 413 | # add text to document object
|
|---|
| 414 | # turn \ into \\ so that the rest of greenstone doesn't think there
|
|---|
| 415 | # is an escape code following. (Macro parsing loses them...)
|
|---|
| 416 | $$textref =~ s/\\/\\\\/go;
|
|---|
| 417 | $doc_obj->add_utf8_text($cursection, $$textref);
|
|---|
| 418 | }
|
|---|
| 419 |
|
|---|
| 420 | sub replace_images {
|
|---|
| 421 | my $self = shift (@_);
|
|---|
| 422 | my ($front, $link, $back, $base_dir,
|
|---|
| 423 | $file, $doc_obj, $section) = @_;
|
|---|
| 424 |
|
|---|
| 425 | # remove quotes from link at start and end if necessary
|
|---|
| 426 | if ($link=~/^\"/) {
|
|---|
| 427 | $link=~s/^\"//;$link=~s/\"$//;
|
|---|
| 428 | $front.='"';
|
|---|
| 429 | $back="\"$back";
|
|---|
| 430 | }
|
|---|
| 431 |
|
|---|
| 432 | $link =~ s/\n/ /g;
|
|---|
| 433 |
|
|---|
| 434 | # Hack to overcome Windows wv 0.7.1 bug that causes embedded images to be broken
|
|---|
| 435 | # If the Word file path has spaces in it, wv messes up and you end up with
|
|---|
| 436 | # absolute paths for the images, and without the "file://" prefix
|
|---|
| 437 | # So check for this special case and massage the data to be correct
|
|---|
| 438 | if ($ENV{'GSDLOS'} =~ /^windows/i && $self->{'plugin_type'} eq "WordPlug" && $link =~ /^[A-Za-z]\:\\/) {
|
|---|
| 439 | $link =~ s/^.*\\([^\\]+)$/$1/;
|
|---|
| 440 | }
|
|---|
| 441 |
|
|---|
| 442 | my ($href, $hash_part, $rl) = $self->format_link ($link, $base_dir, $file);
|
|---|
| 443 |
|
|---|
| 444 | my $img_file = $self->add_file ($href, $rl, $hash_part, $base_dir, $doc_obj, $section);
|
|---|
| 445 |
|
|---|
| 446 | my $anchor_name = $img_file;
|
|---|
| 447 | $anchor_name =~ s/^.*\///;
|
|---|
| 448 | $anchor_name = "<a name=\"$anchor_name\">";
|
|---|
| 449 |
|
|---|
| 450 | return $front . $img_file . $back . $anchor_name;
|
|---|
| 451 | }
|
|---|
| 452 |
|
|---|
| 453 | sub replace_href_links {
|
|---|
| 454 | my $self = shift (@_);
|
|---|
| 455 | my ($front, $link, $back, $base_dir, $file, $doc_obj, $section) = @_;
|
|---|
| 456 |
|
|---|
| 457 | # attempt to sort out targets - frames are not handled
|
|---|
| 458 | # well in this plugin and some cases will screw things
|
|---|
| 459 | # up - e.g. the _parent target (so we'll just remove
|
|---|
| 460 | # them all ;-)
|
|---|
| 461 | $front =~ s/(target=\"?)_top(\"?)/$1_gsdltop_$2/is;
|
|---|
| 462 | $back =~ s/(target=\"?)_top(\"?)/$1_gsdltop_$2/is;
|
|---|
| 463 | $front =~ s/target=\"?_parent\"?//is;
|
|---|
| 464 | $back =~ s/target=\"?_parent\"?//is;
|
|---|
| 465 |
|
|---|
| 466 | return $front . $link . $back if $link =~ /^\#/s;
|
|---|
| 467 | $link =~ s/\n/ /g;
|
|---|
| 468 |
|
|---|
| 469 | my ($href, $hash_part, $rl) = $self->format_link ($link, $base_dir, $file);
|
|---|
| 470 | # href may use '\'s where '/'s should be on Windows
|
|---|
| 471 | $href =~ s/\\/\//g;
|
|---|
| 472 |
|
|---|
| 473 | my ($filename) = $href =~ /^(?:.*?):(?:\/\/)?(.*)/;
|
|---|
| 474 |
|
|---|
| 475 |
|
|---|
| 476 | ##### leave all these links alone (they won't be picked up by intermediate
|
|---|
| 477 | ##### pages). I think that's safest when dealing with frames, targets etc.
|
|---|
| 478 | ##### (at least until I think of a better way to do it). Problems occur with
|
|---|
| 479 | ##### mailto links from within small frames, the intermediate page is displayed
|
|---|
| 480 | ##### within that frame and can't be seen. There is still potential for this to
|
|---|
| 481 | ##### happen even with html pages - the solution seems to be to somehow tell
|
|---|
| 482 | ##### the browser from the server side to display the page being sent (i.e.
|
|---|
| 483 | ##### the intermediate page) in the top level window - I'm not sure if that's
|
|---|
| 484 | ##### possible - the following line should probably be deleted if that can be done
|
|---|
| 485 | return $front . $link . $back if $href =~ /^(mailto|news|gopher|nntp|telnet|javascript):/is;
|
|---|
| 486 |
|
|---|
| 487 |
|
|---|
| 488 | if (($rl == 0) || ($filename =~ /$self->{'process_exp'}/) ||
|
|---|
| 489 | ($href =~ /\/$/) || ($href =~ /^(mailto|news|gopher|nntp|telnet|javascript):/i)) {
|
|---|
| 490 | &ghtml::urlsafe ($href);
|
|---|
| 491 | return $front . "_httpextlink_&rl=" . $rl . "&href=" . $href . $hash_part . $back;
|
|---|
| 492 | } else {
|
|---|
| 493 | # link is to some other type of file (eg image) so we'll
|
|---|
| 494 | # need to associate that file
|
|---|
| 495 | return $front . $self->add_file ($href, $rl, $hash_part, $base_dir, $doc_obj, $section) . $back;
|
|---|
| 496 | }
|
|---|
| 497 | }
|
|---|
| 498 |
|
|---|
| 499 | sub add_file {
|
|---|
| 500 | my $self = shift (@_);
|
|---|
| 501 | my ($href, $rl, $hash_part, $base_dir, $doc_obj, $section) = @_;
|
|---|
| 502 | my ($newname);
|
|---|
| 503 |
|
|---|
| 504 | my $filename = $href;
|
|---|
| 505 | $filename =~ s/^[^:]*:\/\///;
|
|---|
| 506 | $filename = &util::filename_cat($base_dir, $filename);
|
|---|
| 507 |
|
|---|
| 508 | # Replace %20's in URL with a space if required. Note that the filename
|
|---|
| 509 | # may include the %20 in some situations
|
|---|
| 510 | if ($filename =~ /\%20/) {
|
|---|
| 511 | if (!-e $filename) {
|
|---|
| 512 | $filename =~ s/\%20/ /g;
|
|---|
| 513 | }
|
|---|
| 514 | }
|
|---|
| 515 |
|
|---|
| 516 | my ($ext) = $filename =~ /(\.[^\.]*)$/;
|
|---|
| 517 |
|
|---|
| 518 | if ($rl == 0) {
|
|---|
| 519 | if ((!defined $ext) || ($ext !~ /$self->{'assoc_files'}/)) {
|
|---|
| 520 | return "_httpextlink_&rl=0&el=prompt&href=" . $href . $hash_part;
|
|---|
| 521 | }
|
|---|
| 522 | else {
|
|---|
| 523 | return "_httpextlink_&rl=0&el=direct&href=" . $href . $hash_part;
|
|---|
| 524 | }
|
|---|
| 525 | }
|
|---|
| 526 |
|
|---|
| 527 | if ((!defined $ext) || ($ext !~ /$self->{'assoc_files'}/)) {
|
|---|
| 528 | return "_httpextlink_&rl=" . $rl . "&href=" . $href . $hash_part;
|
|---|
| 529 | }
|
|---|
| 530 | if ($self->{'rename_assoc_files'}) {
|
|---|
| 531 | if (defined $self->{'aux_files'}->{$href}) {
|
|---|
| 532 | $newname = $self->{'aux_files'}->{$href}->{'dir_num'} . "/" .
|
|---|
| 533 | $self->{'aux_files'}->{$href}->{'file_num'} . $ext;
|
|---|
| 534 | } else {
|
|---|
| 535 | $newname = $self->{'dir_num'} . "/" . $self->{'file_num'} . $ext;
|
|---|
| 536 | $self->{'aux_files'}->{$href} = {'dir_num' => $self->{'dir_num'}, 'file_num' => $self->{'file_num'}};
|
|---|
| 537 | $self->inc_filecount ();
|
|---|
| 538 | }
|
|---|
| 539 | $doc_obj->associate_file($filename, $newname, undef, $section);
|
|---|
| 540 | return "_httpdocimg_/$newname";
|
|---|
| 541 | } else {
|
|---|
| 542 | ($newname) = $filename =~ /([^\/\\]*)$/;
|
|---|
| 543 | $doc_obj->associate_file($filename, $newname, undef, $section);
|
|---|
| 544 | return "_httpdocimg_/$newname";
|
|---|
| 545 | }
|
|---|
| 546 | }
|
|---|
| 547 |
|
|---|
| 548 |
|
|---|
| 549 | sub format_link {
|
|---|
| 550 | my $self = shift (@_);
|
|---|
| 551 | my ($link, $base_dir, $file) = @_;
|
|---|
| 552 |
|
|---|
| 553 | my ($before_hash, $hash_part) = $link =~ /^([^\#]*)(\#?.*)$/;
|
|---|
| 554 |
|
|---|
| 555 | $hash_part = "" if !defined $hash_part;
|
|---|
| 556 | if (!defined $before_hash || $before_hash !~ /[\w\.\/]/) {
|
|---|
| 557 | my $outhandle = $self->{'outhandle'};
|
|---|
| 558 | print $outhandle "HTMLPlug: ERROR - badly formatted tag ignored ($link)\n"
|
|---|
| 559 | if $self->{'verbosity'};
|
|---|
| 560 | return ($link, "", 0);
|
|---|
| 561 | }
|
|---|
| 562 |
|
|---|
| 563 | if ($before_hash =~ s@^((?:http|ftp|file)://)@@i) {
|
|---|
| 564 | my $type = $1;
|
|---|
| 565 |
|
|---|
| 566 | if ($link =~ /^(http|ftp):/i) {
|
|---|
| 567 | # Turn url (using /) into file name (possibly using \ on windows)
|
|---|
| 568 | my @http_dir_split = split('/', $before_hash);
|
|---|
| 569 | $before_hash = &util::filename_cat(@http_dir_split);
|
|---|
| 570 | }
|
|---|
| 571 |
|
|---|
| 572 | $before_hash = $self->eval_dir_dots($before_hash);
|
|---|
| 573 |
|
|---|
| 574 | my $linkfilename = &util::filename_cat ($base_dir, $before_hash);
|
|---|
| 575 |
|
|---|
| 576 | my $rl = 0;
|
|---|
| 577 | $rl = 1 if (-e $linkfilename);
|
|---|
| 578 |
|
|---|
| 579 | # make sure there's a slash on the end if it's a directory
|
|---|
| 580 | if ($before_hash !~ /\/$/) {
|
|---|
| 581 | $before_hash .= "/" if (-d $linkfilename);
|
|---|
| 582 | }
|
|---|
| 583 |
|
|---|
| 584 | return ($type . $before_hash, $hash_part, $rl);
|
|---|
| 585 |
|
|---|
| 586 | } elsif ($link !~ /^(mailto|news|gopher|nntp|telnet|javascript):/i) {
|
|---|
| 587 | if ($before_hash =~ s@^/@@ || $before_hash =~ /\\/) {
|
|---|
| 588 |
|
|---|
| 589 | # the first directory will be the domain name if file_is_url
|
|---|
| 590 | # to generate archives, otherwise we'll assume all files are
|
|---|
| 591 | # from the same site and base_dir is the root
|
|---|
| 592 |
|
|---|
| 593 | if ($self->{'file_is_url'}) {
|
|---|
| 594 | my @dirs = split /[\/\\]/, $file;
|
|---|
| 595 | my $domname = shift (@dirs);
|
|---|
| 596 | $before_hash = &util::filename_cat($domname, $before_hash);
|
|---|
| 597 | $before_hash =~ s@\\@/@g; # for windows
|
|---|
| 598 | }
|
|---|
| 599 | else
|
|---|
| 600 | {
|
|---|
| 601 | # see if link shares directory with source document
|
|---|
| 602 | # => turn into relative link if this is so!
|
|---|
| 603 |
|
|---|
| 604 | if ($ENV{'GSDLOS'} =~ /^windows/i) {
|
|---|
| 605 | # too difficult doing a pattern match with embedded '\'s...
|
|---|
| 606 | my $win_before_hash=$before_hash;
|
|---|
| 607 | $win_before_hash =~ s@(\\)+@/@g;
|
|---|
| 608 | # $base_dir is already similarly "converted" on windows.
|
|---|
| 609 | if ($win_before_hash =~ s@^$base_dir/@@o) {
|
|---|
| 610 | # if this is true, we removed a prefix
|
|---|
| 611 | $before_hash=$win_before_hash;
|
|---|
| 612 | }
|
|---|
| 613 | }
|
|---|
| 614 | else {
|
|---|
| 615 | # before_hash has lost leading slash by this point,
|
|---|
| 616 | # -> add back in prior to substitution with $base_dir
|
|---|
| 617 | $before_hash = "/$before_hash";
|
|---|
| 618 |
|
|---|
| 619 | $before_hash = &util::filename_cat("",$before_hash);
|
|---|
| 620 | $before_hash =~ s@^$base_dir/@@;
|
|---|
| 621 | }
|
|---|
| 622 | }
|
|---|
| 623 | } else {
|
|---|
| 624 | # Turn relative file path into full path
|
|---|
| 625 | my $dirname = &File::Basename::dirname($file);
|
|---|
| 626 | $before_hash = &util::filename_cat($dirname, $before_hash);
|
|---|
| 627 | $before_hash = $self->eval_dir_dots($before_hash);
|
|---|
| 628 | }
|
|---|
| 629 |
|
|---|
| 630 | my $linkfilename = &util::filename_cat ($base_dir, $before_hash);
|
|---|
| 631 | # make sure there's a slash on the end if it's a directory
|
|---|
| 632 | if ($before_hash !~ /\/$/) {
|
|---|
| 633 | $before_hash .= "/" if (-d $linkfilename);
|
|---|
| 634 | }
|
|---|
| 635 | return ("http://" . $before_hash, $hash_part, 1);
|
|---|
| 636 | } else {
|
|---|
| 637 | # mailto, news, nntp, telnet, javascript or gopher link
|
|---|
| 638 | return ($before_hash, "", 0);
|
|---|
| 639 | }
|
|---|
| 640 | }
|
|---|
| 641 |
|
|---|
| 642 | sub extract_first_NNNN_characters {
|
|---|
| 643 | my $self = shift (@_);
|
|---|
| 644 | my ($textref, $doc_obj, $thissection) = @_;
|
|---|
| 645 |
|
|---|
| 646 | foreach my $size (split /,/, $self->{'first'}) {
|
|---|
| 647 | my $tmptext = $$textref;
|
|---|
| 648 | # skip to the body
|
|---|
| 649 | $tmptext =~ s/.*<body[^>]*>//i;
|
|---|
| 650 | # remove javascript
|
|---|
| 651 | $tmptext =~ s@<script.*?</script>@ @sig;
|
|---|
| 652 | $tmptext =~ s/<[^>]*>/ /g;
|
|---|
| 653 | $tmptext =~ s/ / /g;
|
|---|
| 654 | $tmptext =~ s/^\s+//;
|
|---|
| 655 | $tmptext =~ s/\s+$//;
|
|---|
| 656 | $tmptext =~ s/\s+/ /gs;
|
|---|
| 657 | $tmptext = &unicode::substr ($tmptext, 0, $size);
|
|---|
| 658 | $tmptext =~ s/\s\S*$/…/; # adds an ellipse (...)
|
|---|
| 659 | $doc_obj->add_utf8_metadata ($thissection, "First$size", $tmptext);
|
|---|
| 660 | }
|
|---|
| 661 | }
|
|---|
| 662 |
|
|---|
| 663 |
|
|---|
| 664 | sub extract_metadata {
|
|---|
| 665 | my $self = shift (@_);
|
|---|
| 666 | my ($textref, $metadata, $doc_obj, $section) = @_;
|
|---|
| 667 | my $outhandle = $self->{'outhandle'};
|
|---|
| 668 | # if we don't want metadata, we may as well not be here ...
|
|---|
| 669 | return if (!defined $self->{'metadata_fields'});
|
|---|
| 670 |
|
|---|
| 671 | # metadata fields to extract/save. 'key' is the (lowercase) name of the
|
|---|
| 672 | # html meta, 'value' is the metadata name for greenstone to use
|
|---|
| 673 | my %find_fields = ();
|
|---|
| 674 |
|
|---|
| 675 | my %creator_fields = (); # short-cut for lookups
|
|---|
| 676 |
|
|---|
| 677 |
|
|---|
| 678 | foreach my $field (split /,/, $self->{'metadata_fields'}) {
|
|---|
| 679 | # support tag<tagname>
|
|---|
| 680 | if ($field =~ /^(.*?)<(.*?)>$/) {
|
|---|
| 681 | # "$2" is the user's preferred gs metadata name
|
|---|
| 682 | $find_fields{lc($1)}=$2; # lc = lowercase
|
|---|
| 683 | } else { # no <tagname> for mapping
|
|---|
| 684 | # "$field" is the user's preferred gs metadata name
|
|---|
| 685 | $find_fields{lc($field)}=$field; # lc = lowercase
|
|---|
| 686 | }
|
|---|
| 687 | }
|
|---|
| 688 |
|
|---|
| 689 | if (defined $self->{'hunt_creator_metadata'} &&
|
|---|
| 690 | $self->{'hunt_creator_metadata'} == 1 ) {
|
|---|
| 691 | my @extra_fields =
|
|---|
| 692 | (
|
|---|
| 693 | 'author',
|
|---|
| 694 | 'author.email',
|
|---|
| 695 | 'creator',
|
|---|
| 696 | 'dc.creator',
|
|---|
| 697 | 'dc.creator.corporatename',
|
|---|
| 698 | );
|
|---|
| 699 |
|
|---|
| 700 | # add the creator_metadata fields to search for
|
|---|
| 701 | foreach my $field (@extra_fields) {
|
|---|
| 702 | $creator_fields{$field}=0; # add to lookup hash
|
|---|
| 703 | }
|
|---|
| 704 | }
|
|---|
| 705 |
|
|---|
| 706 |
|
|---|
| 707 | # find the header in the html file, which has the meta tags
|
|---|
| 708 | $$textref =~ m@<head>(.*?)</head>@si;
|
|---|
| 709 |
|
|---|
| 710 | my $html_header=$1;
|
|---|
| 711 | # go through every <meta... tag defined in the html and see if it is
|
|---|
| 712 | # one of the tags we want to match.
|
|---|
| 713 |
|
|---|
| 714 | # special case for title - we want to remember if its been found
|
|---|
| 715 | my $found_title = 0;
|
|---|
| 716 | # this assumes that ">" won't appear. (I don't think it's allowed to...)
|
|---|
| 717 | $html_header =~ /^/; # match the start of the string, for \G assertion
|
|---|
| 718 |
|
|---|
| 719 | while ($html_header =~ m/\G.*?<meta(.*?)>/sig) {
|
|---|
| 720 | my $metatag=$1;
|
|---|
| 721 | my ($tag, $value);
|
|---|
| 722 |
|
|---|
| 723 | # find the tag name
|
|---|
| 724 | $metatag =~ /(?:name|http-equiv)\s*=\s*([\"\'])?(.*?)\1/is;
|
|---|
| 725 | $tag=$2;
|
|---|
| 726 | # in case they're not using " or ', but they should...
|
|---|
| 727 | if (! $tag) {
|
|---|
| 728 | $metatag =~ /(?:name|http-equiv)\s*=\s*([^\s\>]+)/is;
|
|---|
| 729 | $tag=$1;
|
|---|
| 730 | }
|
|---|
| 731 |
|
|---|
| 732 | if (!defined $tag) {
|
|---|
| 733 | print $outhandle "HTMLPlug: can't find NAME in \"$metatag\"\n";
|
|---|
| 734 | next;
|
|---|
| 735 | }
|
|---|
| 736 |
|
|---|
| 737 | # don't need to assign this field if it was passed in from a previous
|
|---|
| 738 | # (recursive) plugin
|
|---|
| 739 | if (defined $metadata->{$tag}) {next}
|
|---|
| 740 |
|
|---|
| 741 | # find the tag content
|
|---|
| 742 | $metatag =~ /content\s*=\s*([\"\'])?(.*?)\1/is;
|
|---|
| 743 | $value=$2;
|
|---|
| 744 | if (! $value) {
|
|---|
| 745 | $metatag =~ /(?:name|http-equiv)\s*=\s*([^\s\>]+)/is;
|
|---|
| 746 | $value=$1;
|
|---|
| 747 | }
|
|---|
| 748 | if (!defined $value) {
|
|---|
| 749 | print $outhandle "HTMLPlug: can't find VALUE in \"$metatag\"\n";
|
|---|
| 750 | next;
|
|---|
| 751 | }
|
|---|
| 752 |
|
|---|
| 753 | # clean up and add
|
|---|
| 754 | $value =~ s/\s+/ /gs;
|
|---|
| 755 | chomp($value); # remove trailing \n, if any
|
|---|
| 756 | if (exists $creator_fields{lc($tag)}) {
|
|---|
| 757 | # map this value onto greenstone's "Creator" metadata
|
|---|
| 758 | $tag='Creator';
|
|---|
| 759 | } elsif (!exists $find_fields{lc($tag)}) {
|
|---|
| 760 | next; # don't want this tag
|
|---|
| 761 | } else {
|
|---|
| 762 | # get the user's preferred capitalisation
|
|---|
| 763 | $tag = $find_fields{lc($tag)};
|
|---|
| 764 | }
|
|---|
| 765 | if (lc($tag) eq "title") {
|
|---|
| 766 | $found_title = 1;
|
|---|
| 767 | }
|
|---|
| 768 | print $outhandle " extracted \"$tag\" metadata \"$value\"\n"
|
|---|
| 769 | if ($self->{'verbosity'} > 2);
|
|---|
| 770 | $doc_obj->add_utf8_metadata($section, $tag, $value);
|
|---|
| 771 |
|
|---|
| 772 | }
|
|---|
| 773 |
|
|---|
| 774 | # TITLE: extract the document title
|
|---|
| 775 | if (exists $find_fields{'title'} && !$found_title) {
|
|---|
| 776 | # we want a title, and didn't find one in the meta tags
|
|---|
| 777 | # see if there's a <title> tag
|
|---|
| 778 | my $title;
|
|---|
| 779 | my $from = ""; # for debugging output only
|
|---|
| 780 | if ($html_header =~ /<title[^>]*>([^<]+)<\/title[^>]*>/is) {
|
|---|
| 781 | $title = $1;
|
|---|
| 782 | $from = "<title> tags";
|
|---|
| 783 | }
|
|---|
| 784 |
|
|---|
| 785 | if (!defined $title) {
|
|---|
| 786 | $from = "first 100 chars";
|
|---|
| 787 | # if no title use first 100 or so characters
|
|---|
| 788 | $title = $$textref;
|
|---|
| 789 | $title =~ s/^\xFE\xFF//; # Remove unicode byte order mark
|
|---|
| 790 | $title =~ s/^.*?<body>//si;
|
|---|
| 791 | # ignore javascript!
|
|---|
| 792 | $title =~ s@<script.*?</script>@ @sig;
|
|---|
| 793 | $title =~ s/<\/([^>]+)><\1>//g; # (eg) </b><b> - no space
|
|---|
| 794 | $title =~ s/<[^>]*>/ /g; # remove all HTML tags
|
|---|
| 795 | $title = substr ($title, 0, 100);
|
|---|
| 796 | $title =~ s/\s\S*$/.../;
|
|---|
| 797 | }
|
|---|
| 798 | $title =~ s/<[^>]*>/ /g; # remove html tags
|
|---|
| 799 | $title =~ s/ / /g;
|
|---|
| 800 | $title =~ s/(?: |\xc2\xa0)/ /g; # utf-8 for nbsp...
|
|---|
| 801 | $title =~ s/\s+/ /gs; # collapse multiple spaces
|
|---|
| 802 | $title =~ s/^\s*//; # remove leading spaces
|
|---|
| 803 | $title =~ s/\s*$//; # remove trailing spaces
|
|---|
| 804 |
|
|---|
| 805 | $title =~ s/^$self->{'title_sub'}// if ($self->{'title_sub'});
|
|---|
| 806 | $title =~ s/^\s+//s; # in case title_sub introduced any...
|
|---|
| 807 | $doc_obj->add_utf8_metadata ($section, 'Title', $title);
|
|---|
| 808 | print $outhandle " extracted Title metadata \"$title\" from $from\n"
|
|---|
| 809 | if ($self->{'verbosity'} > 2);
|
|---|
| 810 | }
|
|---|
| 811 |
|
|---|
| 812 | # add FileFormat metadata
|
|---|
| 813 | $doc_obj->add_metadata($section,"FileFormat", "HTML");
|
|---|
| 814 |
|
|---|
| 815 | # Special, for metadata names such as tagH1 - extracts
|
|---|
| 816 | # the text between the first <H1> and </H1> tags into "H1" metadata.
|
|---|
| 817 |
|
|---|
| 818 | foreach my $field (keys %find_fields) {
|
|---|
| 819 | if ($field !~ /^tag([a-z0-9]+)$/i) {next}
|
|---|
| 820 | my $tag = $1;
|
|---|
| 821 | if ($$textref =~ m@<$tag[^>]*>(.*?)</$tag[^>]*>@g) {
|
|---|
| 822 | my $content = $1;
|
|---|
| 823 | $content =~ s/ / /g;
|
|---|
| 824 | $content =~ s/<[^>]*>/ /g;
|
|---|
| 825 | $content =~ s/^\s+//;
|
|---|
| 826 | $content =~ s/\s+$//;
|
|---|
| 827 | $content =~ s/\s+/ /gs;
|
|---|
| 828 | if ($content) {
|
|---|
| 829 | $tag=$find_fields{"tag$tag"}; # get the user's capitalisation
|
|---|
| 830 | $tag =~ s/^tag//i;
|
|---|
| 831 | $doc_obj->add_utf8_metadata ($section, $tag, $content);
|
|---|
| 832 | print $outhandle " extracted \"$tag\" metadata \"$content\"\n"
|
|---|
| 833 | if ($self->{'verbosity'} > 2);
|
|---|
| 834 | }
|
|---|
| 835 | }
|
|---|
| 836 | }
|
|---|
| 837 | }
|
|---|
| 838 |
|
|---|
| 839 |
|
|---|
| 840 | # evaluate any "../" to next directory up
|
|---|
| 841 | # evaluate any "./" as here
|
|---|
| 842 | sub eval_dir_dots {
|
|---|
| 843 | my $self = shift (@_);
|
|---|
| 844 | my ($filename) = @_;
|
|---|
| 845 | my $dirsep_os = &util::get_os_dirsep();
|
|---|
| 846 | my @dirsep = split(/$dirsep_os/,$filename);
|
|---|
| 847 |
|
|---|
| 848 | my @eval_dirs = ();
|
|---|
| 849 | foreach my $d (@dirsep) {
|
|---|
| 850 | if ($d eq "..") {
|
|---|
| 851 | pop(@eval_dirs);
|
|---|
| 852 |
|
|---|
| 853 | } elsif ($d eq ".") {
|
|---|
| 854 | # do nothing!
|
|---|
| 855 |
|
|---|
| 856 | } else {
|
|---|
| 857 | push(@eval_dirs,$d);
|
|---|
| 858 | }
|
|---|
| 859 | }
|
|---|
| 860 |
|
|---|
| 861 | # Need to fiddle with number of elements in @eval_dirs if the
|
|---|
| 862 | # first one is the empty string. This is because of a
|
|---|
| 863 | # modification to util::filename_cat that supresses the addition
|
|---|
| 864 | # of a leading '/' character (or \ if windows) (intended to help
|
|---|
| 865 | # filename cat with relative paths) if the first entry in the
|
|---|
| 866 | # array is the empty string. Making the array start with *two*
|
|---|
| 867 | # empty strings is a way to defeat this "smart" option.
|
|---|
| 868 | #
|
|---|
| 869 | if (scalar(@eval_dirs) > 0) {
|
|---|
| 870 | if ($eval_dirs[0] eq ""){
|
|---|
| 871 | unshift(@eval_dirs,"");
|
|---|
| 872 | }
|
|---|
| 873 | }
|
|---|
| 874 | return &util::filename_cat(@eval_dirs);
|
|---|
| 875 | }
|
|---|
| 876 |
|
|---|
| 877 | sub replace_usemap_links {
|
|---|
| 878 | my $self = shift (@_);
|
|---|
| 879 | my ($front, $link, $back) = @_;
|
|---|
| 880 |
|
|---|
| 881 | $link =~ s/^\.\///;
|
|---|
| 882 | return $front . $link . $back;
|
|---|
| 883 | }
|
|---|
| 884 |
|
|---|
| 885 | sub inc_filecount {
|
|---|
| 886 | my $self = shift (@_);
|
|---|
| 887 |
|
|---|
| 888 | if ($self->{'file_num'} == 1000) {
|
|---|
| 889 | $self->{'dir_num'} ++;
|
|---|
| 890 | $self->{'file_num'} = 0;
|
|---|
| 891 | } else {
|
|---|
| 892 | $self->{'file_num'} ++;
|
|---|
| 893 | }
|
|---|
| 894 | }
|
|---|
| 895 |
|
|---|
| 896 |
|
|---|
| 897 | # Extend the BasPlug read_file so that strings like é are
|
|---|
| 898 | # converted to UTF8 internally.
|
|---|
| 899 | #
|
|---|
| 900 | # We don't convert < or > or & or " in case
|
|---|
| 901 | # they interfere with the GML files
|
|---|
| 902 |
|
|---|
| 903 | sub read_file {
|
|---|
| 904 | my ($self, $filename, $encoding, $language, $textref) = @_;
|
|---|
| 905 |
|
|---|
| 906 | &BasPlug::read_file($self, $filename, $encoding, $language, $textref);
|
|---|
| 907 |
|
|---|
| 908 | # Convert entities to their UTF8 equivalents
|
|---|
| 909 | $$textref =~ s/&(lt|gt|amp|quot|nbsp);/&z$1;/go;
|
|---|
| 910 | $$textref =~ s/&([^;]+);/&ghtml::getcharequiv($1,1)/gseo;
|
|---|
| 911 | $$textref =~ s/&z(lt|gt|amp|quot|nbsp);/&$1;/go;
|
|---|
| 912 | }
|
|---|
| 913 |
|
|---|
| 914 | 1;
|
|---|
