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