root/trunk/gsdl/perllib/plugins/HTMLPlug.pm @ 900

Revision 900, 14.2 KB (checked in by sjboddie, 20 years ago)

tweaked the way associated files are handled at build time - some
individual plugins may still need catching up.

  • 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 "   -rename_assoc_files    Renames files associated with documents (e.g. images). Also\n";
74    print STDERR "                          creates much shallower directory structure (useful when creating\n";
75    print STDERR "                          collections to go on cd-rom).\n\n";
76}
77
78sub new {
79    my $class = shift (@_);
80    my $self = new BasPlug ();
81
82    if (!parsargv::parse(\@_,
83             q^process_exp/.*/(?i)\.html?$^, \$self->{'process_exp'},
84             q^nolinks^, \$self->{'nolinks'},
85             q^block_exp/.*/(?i)\.(gif|jpe?g|png|pdf|rtf|css)$^, \$self->{'block_exp'},
86             q^keep_head^, \$self->{'keep_head'},
87             q^no_metadata^, \$self->{'no_metadata'},
88             q^metadata_fields/.*/Title^, \$self->{'metadata_fields'},
89             q^w3mir^, \$self->{'w3mir'},
90             q^rename_assoc_files^, \$self->{'rename_assoc_files'})) {
91    &print_usage();
92    die "\n";
93    }
94
95    $self->{'aux_files'} = {};
96    $self->{'dir_num'} = 0;
97    $self->{'file_num'} = 0;
98
99    return bless $self, $class;
100}
101
102sub is_recursive {
103    my $self = shift (@_);
104
105    return 0; # this is not a recursive plugin
106}
107
108# return number of files processed, undef if can't process
109# Note that $base_dir might be "" and that $file might
110# include directories
111sub read {
112    my $self = shift (@_);
113    my ($pluginfo, $base_dir, $file, $metadata, $processor) = @_;
114
115    my $filename = &util::filename_cat($base_dir, $file);
116    return 0 if $filename =~ /$self->{'block_exp'}/;
117    if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
118    return undef;
119    }
120    $file =~ s/^[\/\\]+//;
121
122    $self->{'verbosity'} = $processor->{'verbosity'};
123    print STDERR "HTMLPlug: processing $file\n"
124    if $self->{'verbosity'} > 1;
125
126    # create a new document
127    my $doc_obj = new doc ($file, "indexed_doc");
128    my $cursection = $doc_obj->get_top_section();
129   
130    # read in HTML file
131    open (FILE, $filename) || die "HTMLPlug::read - can't open $filename\n";
132    undef $/;
133    my $text = <FILE>;
134    $/ = "\n";
135    close FILE;
136    if ($text !~ /\w/) {
137    print STDERR "HTMLPlug: ERROR: $file contains no text\n" if $self->{'verbosity'};
138    return 0;
139    }
140
141    $self->extra_metadata ($doc_obj, $cursection, $metadata);
142    $self->extract_metadata (\$text, $metadata, $doc_obj, $cursection)
143    unless $self->{'no_metadata'};
144
145    # Store URL for page as metadata - this can be used for an
146    # altavista style search interface. The URL won't be valid
147    # unless the file structure contains the domain name (i.e.
148    # like when w3mir is used to download a website).
149    my $web_url = "http://$file";
150    $web_url =~ s/\\/\//g; # for windows
151    $doc_obj->add_metadata($cursection, "URL", $web_url);
152
153    # remove header and footer
154    if (!$self->{'keep_head'}) {
155    $text =~ s/^.*?<body[^>]*>//is;
156    $text =~ s/(<\/body[^>]*>|<\/html[^>]*>)//isg;
157    }
158
159    # trap links
160    if (!$self->{'nolinks'}) {
161
162    # usemap="./#index" not handled correctly => change to "#index"
163    $text =~ s/(<img[^>]*?usemap\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
164        $self->replace_usemap_links($1, $2, $3)/isge;
165
166    $text =~ s/(<(?:a|area|frame|link)\s+[^>]*?(?:href|src)\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
167        $self->replace_href_links ($1, $2, $3, $base_dir, $file, $doc_obj, $cursection)/isge;
168    }
169
170    # trap images
171    $text =~ s/(<img[^>]*?src\s*=\s*\"?)([^\">\s]+)(\"?[^>]*>)/
172    $self->replace_images ($1, $2, $3, $base_dir, $file, $doc_obj, $cursection)/isge;
173
174    $doc_obj->add_text ($cursection, $text);
175
176    # add an OID
177    $doc_obj->set_OID();
178
179    # process the document
180    $processor->process($doc_obj);
181
182    return 1; # processed the file
183}
184
185sub replace_images {
186    my $self = shift (@_);
187    my ($front, $link, $back, $base_dir,
188    $file, $doc_obj, $section) = @_;
189   
190    $link =~ s/\n/ /g;
191   
192    my ($href, $hash_part, $rl) = $self->format_link ($link, $base_dir, $file);
193    return $front . $self->add_file ($href, $base_dir, $doc_obj, $section) . $back;
194}
195
196sub replace_href_links {
197    my $self = shift (@_);
198    my ($front, $link, $back, $base_dir, $file, $doc_obj, $section) = @_;
199
200    # attempt to sort out targets - frames are not handled
201    # well in this plugin and some cases will screw things
202    # up - e.g. the _parent target (so we'll just remove
203    # them all ;-)
204    $front =~ s/(target=\"?)_top(\"?)/$1_gsdltop_$2/is;
205    $back =~ s/(target=\"?)_top(\"?)/$1_gsdltop_$2/is;
206    $front =~ s/target=\"?_parent\"?//is;
207    $back =~ s/target=\"?_parent\"?//is;
208
209    return $front . $link . $back if $link =~ /^\#/s;
210    $link =~ s/\n/ /g;
211
212    my ($href, $hash_part, $rl) = $self->format_link ($link, $base_dir, $file);
213
214    my ($filename) = $href =~ /^(?:.*?):(?:\/\/)?(.*)/;
215
216    ##### leave all these links alone (they won't be picked up by intermediate
217    ##### pages). I think that's safest when dealing with frames, targets etc.
218    ##### (at least until I think of a better way to do it). Problems occur with
219    ##### mailto links from within small frames, the intermediate page is displayed
220    ##### within that frame and can't be seen. There is still potential for this to
221    ##### happen even with html pages - the solution seems to be to somehow tell
222    ##### the browser from the server side to display the page being sent (i.e.
223    ##### the intermediate page) in the top level window - I'm not sure if that's
224    ##### possible - the following line should probably be deleted if that can be done
225    return $front . $link . $back if $href =~ /^(mailto|news|gopher|nntp|telnet|javascript):/is;
226
227
228    if (($rl == 0) || ($filename =~ /$self->{'process_exp'}/) ||
229    ($href =~ /\/$/) || ($href =~ /^(mailto|news|gopher|nntp|telnet|javascript):/i)) {
230    &html::urlsafe ($href);
231    return $front . "_httpextlink_&rl=" . $rl . "&href=" . $href . $hash_part . $back;
232   
233    } else {
234    # link is to some other type of file (image, pdf etc.) so we'll
235    # need to associate that file
236    return $front . $self->add_file ($href, $base_dir, $doc_obj, $section) . $back;
237    }
238}
239
240sub add_file {
241    my $self = shift (@_);
242    my ($href, $base_dir, $doc_obj, $section) = @_;
243    my ($newname);
244
245    my $filename = $href;
246    $filename =~ s/^[^:]*:\/\///;
247    $filename = &util::filename_cat ($base_dir, $filename);
248    my ($ext) = $filename =~ /(\.[^\.]*)$/;
249    if ($self->{'rename_assoc_files'}) {
250    if (defined $self->{'aux_files'}->{$href}) {
251        $newname = $self->{'aux_files'}->{$href}->{'dir_num'} . "/" .
252        $self->{'aux_files'}->{$href}->{'file_num'} . $ext;
253    } else {
254        $newname = $self->{'dir_num'} . "/" . $self->{'file_num'} . $ext;
255        $self->inc_filecount ();
256    }
257    $doc_obj->associate_file($filename, $newname, undef, $section);
258    return "_httpcollimg_/$newname";
259
260    } else {
261    ($newname) = $filename =~ /([^\/\\]*)$/;
262    $doc_obj->associate_file($filename, $newname, undef, $section);
263    return "_httpcollimg_/_thisOID_/$newname";
264    }
265}
266
267
268sub format_link {
269    my $self = shift (@_);
270    my ($link, $base_dir, $file) = @_;
271
272    my ($before_hash, $hash_part) = $link =~ /^([^\#]*)(\#?.*)$/;
273    $hash_part = "" if !defined $hash_part;
274    if (!defined $before_hash || $before_hash !~ /[\w\.\/]/) {
275    print STDERR "HTMLPlug: ERROR - badly formatted tag ignored ($link)\n"
276        if $self->{'verbosity'};
277    return ($link, "", 0);
278    }
279   
280    if ($before_hash =~ s/^((?:http|ftp|file):\/\/)//i) {
281    my $type = $1;
282           
283    if ($link =~ /^(http|ftp):/i) {
284        # Turn url (using /) into file name (possibly using \ on windows)
285        my @http_dir_split = split('/', $before_hash);
286        $before_hash = &util::filename_cat(@http_dir_split);
287    }
288
289    $before_hash = $self->eval_dir_dots($before_hash);
290                     
291    my $linkfilename = &util::filename_cat ($base_dir, $before_hash);
292
293    my $rl = 0;
294    $rl = 1 if (-e $linkfilename);
295
296    # make sure there's a slash on the end if it's a directory
297    if ($before_hash !~ /\/$/) {
298        $before_hash .= "/" if (-d $linkfilename);
299    }
300
301    return ($type . $before_hash, $hash_part, $rl);
302   
303    } elsif ($link !~ /^(mailto|news|gopher|nntp|telnet|javascript):/i) {
304
305    if ($before_hash =~ s/^\///) {
306        # the first directory will be the domain name if w3mir was used
307        # to generate archives, otherwise we'll assume all files are
308        # from the same site and base_dir is the root
309        if ($self->{'w3mir'}) {
310        my @dirs = split /[\/\\]/, $file;
311        my $domname = shift (@dirs);
312        $before_hash = &util::filename_cat($domname, $before_hash);
313        $before_hash =~ s/\\/\//g; # for windows
314        }
315
316    } else {
317        # Turn relative file path into full path
318        my $dirname = &File::Basename::dirname($file);
319        $before_hash = &util::filename_cat($dirname, $before_hash);
320        $before_hash = $self->eval_dir_dots($before_hash);
321    }
322
323    # make sure there's a slash on the end if it's a directory
324    my $linkfilename = &util::filename_cat ($base_dir, $before_hash);
325    if ($before_hash !~ /\/$/) {
326        $before_hash .= "/" if (-d $linkfilename);
327    }
328
329    return ("http://" . $before_hash, $hash_part, 1);
330
331    } else {
332    # mailto, news, nntp, telnet, javascript or gopher link
333    return ($before_hash, "", 0);
334    }
335}
336
337sub extract_metadata {
338    my $self = shift (@_);
339    my ($textref, $metadata, $doc_obj, $section) = @_;
340
341    foreach my $field (split /,/, $self->{'metadata_fields'}) {
342   
343    # don't need to extract field if it was passed in from a previous
344    # (recursive) plugin
345    next if defined $metadata->{$field};
346
347    # see if there's a <meta> tag for this field
348    if ($$textref =~ /<meta(.*?)(?:name|http-equiv)\s*=\s*\"?$field\"?([^>]*)/is) {
349        my $content = $1 . $2;
350        if ($content =~ /content\s*=\s*\"?(.*?)\"?/is) {
351        if (defined $1) {
352            my $value = $1;
353            $value =~ s/\s+/ /gs;
354            $doc_obj->add_metadata($section, $field, $value);
355            next;
356        }
357        }
358    }
359   
360    # special case for Title metadata - try <title> tags
361    # then first 100 characters of text
362   
363    if ($field =~ /^title$/i) {
364
365        # see if there's a <title> tag
366        if ($$textref =~ /<title[^>]*>([^<]*)<\/title[^>]*>/is) {
367        if (defined $1) {
368            my $title = $1;
369            if ($title =~ /\w/) {
370            $title =~ s/\s+/ /gs;
371            $doc_obj->add_metadata ($section, $field, $title);
372            next;
373            }
374        }
375        }
376       
377        # if no title use first 100 characters
378        my $tmptext = $$textref;
379        $tmptext =~ s/<[^>]*>//g;
380        my $title = substr ($tmptext, 0, 100);
381        $title =~ s/\s+/ /gs;
382        $doc_obj->add_metadata ($section, $field, $title);
383    }
384    }
385}
386
387# evaluate any "../" to next directory up
388# evaluate any "./" as here
389sub eval_dir_dots {
390    my $self = shift (@_);
391    my ($filename) = @_;
392
393    my $dirsep_os = &util::get_os_dirsep();
394    my @dirsep = split(/$dirsep_os/,$filename);
395
396    my @eval_dirs = ();
397    foreach my $d (@dirsep) {
398    if ($d eq "..") {
399        pop(@eval_dirs);
400
401    } elsif ($d eq ".") {
402        # do nothing!
403
404    } else {
405        push(@eval_dirs,$d);
406    }
407    }
408
409    return &util::filename_cat(@eval_dirs);
410}
411
412sub replace_usemap_links {
413    my $self = shift (@_);
414    my ($front, $link, $back) = @_;
415
416    $link =~ s/^\.\///;
417    return $front . $link . $back;
418}
419
420sub inc_filecount {
421    my $self = shift (@_);
422
423    if ($self->{'file_num'} == 1000) {
424    $self->{'dir_num'} ++;
425    $self->{'file_num'} = 0;
426    } else {
427    $self->{'file_num'} ++;
428    }
429}
430
4311;
Note: See TracBrowser for help on using the browser.