root/trunk/gsdl/perllib/plugins/HTMLPlug.pm @ 965

Revision 965, 14.8 KB (checked in by sjboddie, 20 years ago)

fixed bug - added assoc_files option

  • Property svn:executable set to *
  • Property svn:keywords set to Author Date Id Revision
Line 
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
38package HTMLPlug;
39
40use BasPlug;
41use html;
42use util;
43use parsargv;
44
45sub BEGIN {
46    @ISA = ('BasPlug');
47}
48
49use strict;
50
51sub print_usage {
52    print STDERR "\nIncorrect options passed to HTMLPlug, check your collect.cfg configuration file\n";
53
54    print STDERR "\n  usage: plugin HTMLPlug [options]\n\n";
55    print STDERR "  options:\n";
56    print STDERR "   -process_exp           A perl regular expression to match against filenames.\n";
57    print STDERR "                          Matching filenames will be processed by this plugin.\n";
58    print STDERR "                          Defaults to '(?i)\.html?\$' i.e. all documents ending in\n";
59    print STDERR "                          .htm or .html (case-insensitive).\n";
60    print STDERR "   -nolinks               Don't make any attempt to trap links (setting this flag may\n";
61    print STDERR "                          improve speed of building/importing but any relative links within\n";
62    print STDERR "                          documents will be broken).\n";
63    print STDERR "   -block_exp             Files matching this regular expression will be blocked from\n";
64    print STDERR "                          being passed to any further plugins in the list. By default\n";
65    print STDERR "                          HTMLPlug blocks any files with .gif, .jpg, .jpeg, .png, .pdf,\n";
66    print STDERR "                          .rtf or .css file extensions.\n";
67    print STDERR "   -keep_head             Don't remove headers from html files.\n";
68    print STDERR "   -no_metadata           Don't attempt to extract any metadata from files.\n";
69    print STDERR "   -metadata_fields       Comma separated list of metadata fields to attempt to extract.\n";
70    print STDERR "                          Defaults to 'Title'\n";
71    print STDERR "   -w3mir                 Set if w3mir was used to generate input file structure.\n";
72    print STDERR "                          w3mir \n";
73    print STDERR "   -assoc_files           Perl regular expression of file extensions to associate with\n";
74    print STDERR "                          html documents. Defaults to '(?i)\.(jpe?g|gif|png|css|pdf)$'\n";
75    print STDERR "   -rename_assoc_files    Renames files associated with documents (e.g. images). Also\n";
76    print STDERR "                          creates much shallower directory structure (useful when creating\n";
77    print STDERR "                          collections to go on cd-rom).\n\n";
78}
79
80sub new {
81    my $class = shift (@_);
82    my $self = new BasPlug ();
83
84    if (!parsargv::parse(\@_,
85             q^process_exp/.*/(?i)\.html?$^, \$self->{'process_exp'},
86             q^nolinks^, \$self->{'nolinks'},
87             q^block_exp/.*/(?i)\.(gif|jpe?g|png|pdf|rtf|css)$^, \$self->{'block_exp'},
88             q^keep_head^, \$self->{'keep_head'},
89             q^no_metadata^, \$self->{'no_metadata'},
90             q^metadata_fields/.*/Title^, \$self->{'metadata_fields'},
91             q^w3mir^, \$self->{'w3mir'},
92             q^assoc_files/.*/(?i)\.(jpe?g|gif|png|css|pdf)$^, \$self->{'assoc_files'},
93             q^rename_assoc_files^, \$self->{'rename_assoc_files'})) {
94    &print_usage();
95    die "\n";
96    }
97
98    $self->{'aux_files'} = {};
99    $self->{'dir_num'} = 0;
100    $self->{'file_num'} = 0;
101
102    return bless $self, $class;
103}
104
105sub is_recursive {
106    my $self = shift (@_);
107
108    return 0; # this is not a recursive plugin
109}
110
111# return number of files processed, undef if can't process
112# Note that $base_dir might be "" and that $file might
113# include directories
114sub read {
115    my $self = shift (@_);
116    my ($pluginfo, $base_dir, $file, $metadata, $processor) = @_;
117
118    my $filename = &util::filename_cat($base_dir, $file);
119    return 0 if $filename =~ /$self->{'block_exp'}/;
120    if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
121    return undef;
122    }
123    $file =~ s/^[\/\\]+//;
124
125    $self->{'verbosity'} = $processor->{'verbosity'};
126    print STDERR "HTMLPlug: processing $file\n"
127    if $self->{'verbosity'} > 1;
128
129    # create a new document
130    my $doc_obj = new doc ($file, "indexed_doc");
131    my $cursection = $doc_obj->get_top_section();
132   
133    # read in HTML file
134    open (FILE, $filename) || die "HTMLPlug::read - can't open $filename\n";
135    undef $/;
136    my $text = <FILE>;
137    $/ = "\n";
138    close FILE;
139    if (!defined $text || $text !~ /\w/) {
140    print STDERR "HTMLPlug: ERROR: $file contains no text\n" if $self->{'verbosity'};
141    return 0;
142    }
143
144    $self->extra_metadata ($doc_obj, $cursection, $metadata);
145    $self->extract_metadata (\$text, $metadata, $doc_obj, $cursection)
146    unless $self->{'no_metadata'};
147
148    # Store URL for page as metadata - this can be used for an
149    # altavista style search interface. The URL won't be valid
150    # unless the file structure contains the domain name (i.e.
151    # like when w3mir is used to download a website).
152    my $web_url = "http://$file";
153    $web_url =~ s/\\/\//g; # for windows
154    $doc_obj->add_metadata($cursection, "URL", $web_url);
155
156    # remove header and footer
157    if (!$self->{'keep_head'}) {
158    $text =~ s/^.*?<body[^>]*>//is;
159    $text =~ s/(<\/body[^>]*>|<\/html[^>]*>)//isg;
160    }
161
162    # trap links
163    if (!$self->{'nolinks'}) {
164
165    # usemap="./#index" not handled correctly => change to "#index"
166    $text =~ s/(<img[^>]*?usemap\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
167        $self->replace_usemap_links($1, $2, $3)/isge;
168
169    $text =~ s/(<(?:a|area|frame|link)\s+[^>]*?(?:href|src)\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
170        $self->replace_href_links ($1, $2, $3, $base_dir, $file, $doc_obj, $cursection)/isge;
171    }
172
173    # trap images
174    $text =~ s/(<img[^>]*?src\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
175    $self->replace_images ($1, $2, $3, $base_dir, $file, $doc_obj, $cursection)/isge;
176
177    $doc_obj->add_text ($cursection, $text);
178
179    # add an OID
180    $doc_obj->set_OID();
181
182    # process the document
183    $processor->process($doc_obj);
184
185    return 1; # processed the file
186}
187
188sub replace_images {
189    my $self = shift (@_);
190    my ($front, $link, $back, $base_dir,
191    $file, $doc_obj, $section) = @_;
192   
193    $link =~ s/\n/ /g;
194   
195    my ($href, $hash_part, $rl) = $self->format_link ($link, $base_dir, $file);
196    return $front . $self->add_file ($href, $rl, $hash_part, $base_dir, $doc_obj, $section) . $back;
197}
198
199sub replace_href_links {
200    my $self = shift (@_);
201    my ($front, $link, $back, $base_dir, $file, $doc_obj, $section) = @_;
202
203    # attempt to sort out targets - frames are not handled
204    # well in this plugin and some cases will screw things
205    # up - e.g. the _parent target (so we'll just remove
206    # them all ;-)
207    $front =~ s/(target=\"?)_top(\"?)/$1_gsdltop_$2/is;
208    $back =~ s/(target=\"?)_top(\"?)/$1_gsdltop_$2/is;
209    $front =~ s/target=\"?_parent\"?//is;
210    $back =~ s/target=\"?_parent\"?//is;
211
212    return $front . $link . $back if $link =~ /^\#/s;
213    $link =~ s/\n/ /g;
214
215    my ($href, $hash_part, $rl) = $self->format_link ($link, $base_dir, $file);
216
217    my ($filename) = $href =~ /^(?:.*?):(?:\/\/)?(.*)/;
218
219    ##### leave all these links alone (they won't be picked up by intermediate
220    ##### pages). I think that's safest when dealing with frames, targets etc.
221    ##### (at least until I think of a better way to do it). Problems occur with
222    ##### mailto links from within small frames, the intermediate page is displayed
223    ##### within that frame and can't be seen. There is still potential for this to
224    ##### happen even with html pages - the solution seems to be to somehow tell
225    ##### the browser from the server side to display the page being sent (i.e.
226    ##### the intermediate page) in the top level window - I'm not sure if that's
227    ##### possible - the following line should probably be deleted if that can be done
228    return $front . $link . $back if $href =~ /^(mailto|news|gopher|nntp|telnet|javascript):/is;
229
230
231    if (($rl == 0) || ($filename =~ /$self->{'process_exp'}/) ||
232    ($href =~ /\/$/) || ($href =~ /^(mailto|news|gopher|nntp|telnet|javascript):/i)) {
233    &html::urlsafe ($href);
234    return $front . "_httpextlink_&rl=" . $rl . "&href=" . $href . $hash_part . $back;
235   
236    } else {
237    # link is to some other type of file (image, pdf etc.) so we'll
238    # need to associate that file
239    return $front . $self->add_file ($href, $rl, $hash_part, $base_dir, $doc_obj, $section) . $back;
240    }
241}
242
243sub add_file {
244    my $self = shift (@_);
245    my ($href, $rl, $hash_part, $base_dir, $doc_obj, $section) = @_;
246    my ($newname);
247
248    my $filename = $href;
249    $filename =~ s/^[^:]*:\/\///;
250    $filename = &util::filename_cat ($base_dir, $filename);
251    my ($ext) = $filename =~ /(\.[^\.]*)$/;
252
253    if ((!defined $ext) || ($ext !~ /$self->{'assoc_files'}/)) {
254    return "_httpextlink_&rl=" . $rl . "&href=" . $href . $hash_part;
255    }
256
257    if ($self->{'rename_assoc_files'}) {
258    if (defined $self->{'aux_files'}->{$href}) {
259        $newname = $self->{'aux_files'}->{$href}->{'dir_num'} . "/" .
260        $self->{'aux_files'}->{$href}->{'file_num'} . $ext;
261    } else {
262        $newname = $self->{'dir_num'} . "/" . $self->{'file_num'} . $ext;
263        $self->{'aux_files'}->{$href} = {'dir_num' => $self->{'dir_num'}, 'file_num' => $self->{'file_num'}};
264        $self->inc_filecount ();
265    }
266    $doc_obj->associate_file($filename, $newname, undef, $section);
267    return "_httpcollimg_/$newname";
268
269    } else {
270    ($newname) = $filename =~ /([^\/\\]*)$/;
271    $doc_obj->associate_file($filename, $newname, undef, $section);
272    return "_httpcollimg_/_thisOID_/$newname";
273    }
274}
275
276
277sub format_link {
278    my $self = shift (@_);
279    my ($link, $base_dir, $file) = @_;
280
281    my ($before_hash, $hash_part) = $link =~ /^([^\#]*)(\#?.*)$/;
282    $hash_part = "" if !defined $hash_part;
283    if (!defined $before_hash || $before_hash !~ /[\w\.\/]/) {
284    print STDERR "HTMLPlug: ERROR - badly formatted tag ignored ($link)\n"
285        if $self->{'verbosity'};
286    return ($link, "", 0);
287    }
288   
289    if ($before_hash =~ s/^((?:http|ftp|file):\/\/)//i) {
290    my $type = $1;
291           
292    if ($link =~ /^(http|ftp):/i) {
293        # Turn url (using /) into file name (possibly using \ on windows)
294        my @http_dir_split = split('/', $before_hash);
295        $before_hash = &util::filename_cat(@http_dir_split);
296    }
297
298    $before_hash = $self->eval_dir_dots($before_hash);
299                     
300    my $linkfilename = &util::filename_cat ($base_dir, $before_hash);
301
302    my $rl = 0;
303    $rl = 1 if (-e $linkfilename);
304
305    # make sure there's a slash on the end if it's a directory
306    if ($before_hash !~ /\/$/) {
307        $before_hash .= "/" if (-d $linkfilename);
308    }
309
310    return ($type . $before_hash, $hash_part, $rl);
311   
312    } elsif ($link !~ /^(mailto|news|gopher|nntp|telnet|javascript):/i) {
313
314    if ($before_hash =~ s/^\///) {
315        # the first directory will be the domain name if w3mir was used
316        # to generate archives, otherwise we'll assume all files are
317        # from the same site and base_dir is the root
318        if ($self->{'w3mir'}) {
319        my @dirs = split /[\/\\]/, $file;
320        my $domname = shift (@dirs);
321        $before_hash = &util::filename_cat($domname, $before_hash);
322        $before_hash =~ s/\\/\//g; # for windows
323        }
324
325    } else {
326        # Turn relative file path into full path
327        my $dirname = &File::Basename::dirname($file);
328        $before_hash = &util::filename_cat($dirname, $before_hash);
329        $before_hash = $self->eval_dir_dots($before_hash);
330    }
331
332    # make sure there's a slash on the end if it's a directory
333    my $linkfilename = &util::filename_cat ($base_dir, $before_hash);
334    if ($before_hash !~ /\/$/) {
335        $before_hash .= "/" if (-d $linkfilename);
336    }
337
338    return ("http://" . $before_hash, $hash_part, 1);
339
340    } else {
341    # mailto, news, nntp, telnet, javascript or gopher link
342    return ($before_hash, "", 0);
343    }
344}
345
346sub extract_metadata {
347    my $self = shift (@_);
348    my ($textref, $metadata, $doc_obj, $section) = @_;
349
350    foreach my $field (split /,/, $self->{'metadata_fields'}) {
351   
352    # don't need to extract field if it was passed in from a previous
353    # (recursive) plugin
354    next if defined $metadata->{$field};
355
356    # see if there's a <meta> tag for this field
357    if ($$textref =~ /<meta(.*?)(?:name|http-equiv)\s*=\s*\"?$field\"?([^>]*)/is) {
358        my $content = $1 . $2;
359        if ($content =~ /content\s*=\s*\"?(.*?)\"?/is) {
360        if (defined $1) {
361            my $value = $1;
362            $value =~ s/\s+/ /gs;
363            $doc_obj->add_metadata($section, $field, $value);
364            next;
365        }
366        }
367    }
368   
369    # special case for Title metadata - try <title> tags
370    # then first 100 characters of text
371   
372    if ($field =~ /^title$/i) {
373
374        # see if there's a <title> tag
375        if ($$textref =~ /<title[^>]*>([^<]*)<\/title[^>]*>/is) {
376        if (defined $1) {
377            my $title = $1;
378            if ($title =~ /\w/) {
379            $title =~ s/\s+/ /gs;
380            $doc_obj->add_metadata ($section, $field, $title);
381            next;
382            }
383        }
384        }
385       
386        # if no title use first 100 characters
387        my $tmptext = $$textref;
388        $tmptext =~ s/<[^>]*>//g;
389        my $title = substr ($tmptext, 0, 100);
390        $title =~ s/\s+/ /gs;
391        $doc_obj->add_metadata ($section, $field, $title);
392    }
393    }
394}
395
396# evaluate any "../" to next directory up
397# evaluate any "./" as here
398sub eval_dir_dots {
399    my $self = shift (@_);
400    my ($filename) = @_;
401
402    my $dirsep_os = &util::get_os_dirsep();
403    my @dirsep = split(/$dirsep_os/,$filename);
404
405    my @eval_dirs = ();
406    foreach my $d (@dirsep) {
407    if ($d eq "..") {
408        pop(@eval_dirs);
409
410    } elsif ($d eq ".") {
411        # do nothing!
412
413    } else {
414        push(@eval_dirs,$d);
415    }
416    }
417
418    return &util::filename_cat(@eval_dirs);
419}
420
421sub replace_usemap_links {
422    my $self = shift (@_);
423    my ($front, $link, $back) = @_;
424
425    $link =~ s/^\.\///;
426    return $front . $link . $back;
427}
428
429sub inc_filecount {
430    my $self = shift (@_);
431
432    if ($self->{'file_num'} == 1000) {
433    $self->{'dir_num'} ++;
434    $self->{'file_num'} = 0;
435    } else {
436    $self->{'file_num'} ++;
437    }
438}
439
4401;
Note: See TracBrowser for help on using the browser.