1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # BasPlug.pm -- base class for all the import plugins
|
---|
4 | # A component of the Greenstone digital library software
|
---|
5 | # from the New Zealand Digital Library Project at the
|
---|
6 | # University of Waikato, New Zealand.
|
---|
7 | #
|
---|
8 | # Copyright (C) 1999 New Zealand Digital Library Project
|
---|
9 | #
|
---|
10 | # This program is free software; you can redistribute it and/or modify
|
---|
11 | # it under the terms of the GNU General Public License as published by
|
---|
12 | # the Free Software Foundation; either version 2 of the License, or
|
---|
13 | # (at your option) any later version.
|
---|
14 | #
|
---|
15 | # This program is distributed in the hope that it will be useful,
|
---|
16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
18 | # GNU General Public License for more details.
|
---|
19 | #
|
---|
20 | # You should have received a copy of the GNU General Public License
|
---|
21 | # along with this program; if not, write to the Free Software
|
---|
22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
23 | #
|
---|
24 | ###########################################################################
|
---|
25 |
|
---|
26 | package BasPlug;
|
---|
27 |
|
---|
28 | use parsargv;
|
---|
29 | use multiread;
|
---|
30 | use cnseg;
|
---|
31 | use acronym;
|
---|
32 | use textcat;
|
---|
33 | use strict;
|
---|
34 | use doc;
|
---|
35 | use diagnostics;
|
---|
36 |
|
---|
37 | sub print_general_usage {
|
---|
38 | my ($plugin_name) = @_;
|
---|
39 |
|
---|
40 | print STDERR "\n usage: plugin $plugin_name [options]\n\n";
|
---|
41 | print STDERR " -input_encoding The encoding of the source documents. Documents will be\n";
|
---|
42 | print STDERR " converted from these encodings and stored internally as\n";
|
---|
43 | print STDERR " utf8. The default input_encoding is ascii. Accepted values\n";
|
---|
44 | print STDERR " are:\n";
|
---|
45 | print STDERR " iso_8859_1 (extended ascii)\n";
|
---|
46 | print STDERR " Latin1 (the same as iso-8859-1)\n";
|
---|
47 | print STDERR " ascii (7 bit ascii -- may be faster than Latin1 as no\n";
|
---|
48 | print STDERR " conversion is neccessary)\n";
|
---|
49 | print STDERR " gb (GB or GBK simplified Chinese)\n";
|
---|
50 | print STDERR " iso_8859_6 (8 bit Arabic)\n";
|
---|
51 | print STDERR " windows_1256 (Windows codepage 1256 (Arabic))\n";
|
---|
52 | print STDERR " Arabic (the same as windows_1256)\n";
|
---|
53 | print STDERR " utf8 (either utf8 or unicode -- automatically detected)\n";
|
---|
54 | print STDERR " unicode (just unicode -- doesn't currently do endian\n";
|
---|
55 | print STDERR " detection)\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 " Each plugin has its own default process_exp. e.g HTMLPlug\n";
|
---|
59 | print STDERR " defaults to '(?i)\.html?\$' i.e. all documents ending in\n";
|
---|
60 | print STDERR " .htm or .html (case-insensitive).\n";
|
---|
61 | print STDERR " -block_exp Files matching this regular expression will be blocked from\n";
|
---|
62 | print STDERR " being passed to any further plugins in the list. This has no\n";
|
---|
63 | print STDERR " real effect other than to prevent lots of warning messages\n";
|
---|
64 | print STDERR " about input files you don't care about. Each plugin may or may\n";
|
---|
65 | print STDERR " not have a default block_exp. e.g. by default HTMLPlug blocks\n";
|
---|
66 | print STDERR " any files with .gif, .jpg, .jpeg, .png, .pdf, .rtf or .css\n";
|
---|
67 | print STDERR " file extensions.\n";
|
---|
68 | print STDERR " -extract_acronyms Extract acronyms from within text and set as metadata\n\n";
|
---|
69 | print STDERR " -markup_acronyms Added acronym metadata into document text\n\n";
|
---|
70 | print STDERR " -extract_langauge Identify the language of the text and set as metadata\n\n";
|
---|
71 | }
|
---|
72 |
|
---|
73 | # print_usage should be overridden for any sub-classes having
|
---|
74 | # their own plugin specific options
|
---|
75 | sub print_usage {
|
---|
76 | print STDERR "\nThis plugin has no plugin specific options\n\n";
|
---|
77 |
|
---|
78 | }
|
---|
79 |
|
---|
80 | sub new {
|
---|
81 | my $class = shift (@_);
|
---|
82 | my $plugin_name = shift (@_);
|
---|
83 |
|
---|
84 | my $self = {};
|
---|
85 | my $encodings = "^(iso_8859_1|Latin1|ascii|gb|iso_8859_6|windows_1256|Arabic|utf8|unicode)\$";
|
---|
86 |
|
---|
87 | # general options available to all plugins
|
---|
88 | if (!parsargv::parse(\@_,
|
---|
89 | qq^input_encoding/$encodings/ascii^, \$self->{'input_encoding'},
|
---|
90 | q^process_exp/.*/^, \$self->{'process_exp'},
|
---|
91 | q^block_exp/.*/^, \$self->{'block_exp'},
|
---|
92 | q^extract_acronyms^, \$self->{'extract_acronyms'},
|
---|
93 | q^markup_acronyms^, \$self->{'markup_acronyms'},
|
---|
94 | q^extract_language^, \$self->{'extract_language'},
|
---|
95 | "allow_extra_options")) {
|
---|
96 |
|
---|
97 | print STDERR "\nThe $plugin_name plugin uses an incorrect general option (general options are those\n";
|
---|
98 | print STDERR "available to all plugins). Check your collect.cfg configuration file.\n";
|
---|
99 | &print_general_usage($plugin_name);
|
---|
100 | die "\n";
|
---|
101 | }
|
---|
102 |
|
---|
103 | return bless $self, $class;
|
---|
104 | }
|
---|
105 |
|
---|
106 | # initialize BasPlug options
|
---|
107 | # if init() is overridden in a sub-class, remember to call BasPlug::init()
|
---|
108 | sub init {
|
---|
109 | my $self = shift (@_);
|
---|
110 | my ($verbosity) = @_;
|
---|
111 |
|
---|
112 | # verbosity is passed through from the processor
|
---|
113 | $self->{'verbosity'} = $verbosity;
|
---|
114 |
|
---|
115 | # set process_exp and block_exp to defaults unless they were
|
---|
116 | # explicitly set
|
---|
117 |
|
---|
118 | if ((!$self->is_recursive()) and
|
---|
119 | (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
|
---|
120 |
|
---|
121 | $self->{'process_exp'} = $self->get_default_process_exp ();
|
---|
122 | if ($self->{'process_exp'} eq "") {
|
---|
123 | warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
|
---|
124 | }
|
---|
125 | }
|
---|
126 |
|
---|
127 | if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
|
---|
128 | $self->{'block_exp'} = $self->get_default_block_exp ();
|
---|
129 | }
|
---|
130 |
|
---|
131 | # handle input_encoding aliases
|
---|
132 | $self->{'input_encoding'} = "iso_8859_1" if $self->{'input_encoding'} eq "Latin1";
|
---|
133 | $self->{'input_encoding'} = "windows_1256" if $self->{'input_encoding'} eq "Arabic";
|
---|
134 | }
|
---|
135 |
|
---|
136 | sub begin {
|
---|
137 | my $self = shift (@_);
|
---|
138 | my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
|
---|
139 | }
|
---|
140 |
|
---|
141 | sub end {
|
---|
142 | my ($self) = @_;
|
---|
143 | }
|
---|
144 |
|
---|
145 | # this function should be overridden to return 1
|
---|
146 | # in recursive plugins
|
---|
147 | sub is_recursive {
|
---|
148 | my $self = shift (@_);
|
---|
149 |
|
---|
150 | return 0;
|
---|
151 | }
|
---|
152 |
|
---|
153 | sub get_default_block_exp {
|
---|
154 | my $self = shift (@_);
|
---|
155 |
|
---|
156 | return "";
|
---|
157 | }
|
---|
158 |
|
---|
159 | sub get_default_process_exp {
|
---|
160 | my $self = shift (@_);
|
---|
161 |
|
---|
162 | return "";
|
---|
163 | }
|
---|
164 |
|
---|
165 | # The BasPlug read() function. This function does all the right things
|
---|
166 | # to make general options work for a given plugin. It calls the process()
|
---|
167 | # function which does all the work specific to a plugin (like the old
|
---|
168 | # read functions used to do). Most plugins should define their own
|
---|
169 | # process() function and let this read() function keep control.
|
---|
170 | #
|
---|
171 | # recursive plugins (e.g. RecPlug) and specialized plugins like those
|
---|
172 | # capable of processing many documents within a single file (e.g.
|
---|
173 | # GMLPlug) should normally implement their own version of read()
|
---|
174 | #
|
---|
175 | # Return number of files processed, undef if can't process
|
---|
176 | # Note that $base_dir might be "" and that $file might
|
---|
177 | # include directories
|
---|
178 |
|
---|
179 | sub read {
|
---|
180 | my $self = shift (@_);
|
---|
181 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs) = @_;
|
---|
182 |
|
---|
183 | if ($self->is_recursive()) {
|
---|
184 | die "BasPlug::read function must be implemented in sub-class for recursive plugins\n";
|
---|
185 | }
|
---|
186 |
|
---|
187 | my $filename = &util::filename_cat($base_dir, $file);
|
---|
188 | return 0 if $self->{'block_exp'} ne "" && $filename =~ /$self->{'block_exp'}/;
|
---|
189 | if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
|
---|
190 | return undef;
|
---|
191 | }
|
---|
192 | my $plugin_name = ref ($self);
|
---|
193 | $file =~ s/^[\/\\]+//; # $file often begins with / so we'll tidy it up
|
---|
194 |
|
---|
195 | # create a new document
|
---|
196 | my $doc_obj = new doc ($filename, "indexed_doc");
|
---|
197 |
|
---|
198 | # read in file ($text will be in utf8)
|
---|
199 | my $text = "";
|
---|
200 | $self->read_file ($filename, \$text);
|
---|
201 |
|
---|
202 | if ($text !~ /\w/) {
|
---|
203 | print STDERR "$plugin_name: ERROR: $file contains no text\n" if $self->{'verbosity'};
|
---|
204 | return 0;
|
---|
205 | }
|
---|
206 |
|
---|
207 | # include any metadata passed in from previous plugins
|
---|
208 | # note that this metadata is associated with the top level section
|
---|
209 | $self->extra_metadata ($doc_obj, $doc_obj->get_top_section(), $metadata);
|
---|
210 |
|
---|
211 | # do plugin specific processing of doc_obj
|
---|
212 | return undef unless defined ($self->process (\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj));
|
---|
213 |
|
---|
214 | # do any automatic metadata extraction
|
---|
215 | $self->auto_extract_metadata ($doc_obj);
|
---|
216 |
|
---|
217 | # add an OID
|
---|
218 | $doc_obj->set_OID();
|
---|
219 |
|
---|
220 | # process the document
|
---|
221 | $processor->process($doc_obj);
|
---|
222 |
|
---|
223 | return 1; # processed the file
|
---|
224 | }
|
---|
225 |
|
---|
226 | # returns undef if file is rejected by the plugin
|
---|
227 | sub process {
|
---|
228 | my $self = shift (@_);
|
---|
229 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
---|
230 |
|
---|
231 | die "Basplug::process function must be implemented in sub-class\n";
|
---|
232 |
|
---|
233 | return undef; # never gets here
|
---|
234 | }
|
---|
235 |
|
---|
236 | # uses the multiread package to read in the entire file pointed to
|
---|
237 | # by filename and loads the resulting text into $$textref. Input text
|
---|
238 | # may be in any of the encodings handled by multiread, output text
|
---|
239 | # will be in utf8
|
---|
240 | sub read_file {
|
---|
241 | my $self = shift (@_);
|
---|
242 | my ($filename, $textref) = @_;
|
---|
243 |
|
---|
244 | $$textref = "";
|
---|
245 |
|
---|
246 | open (FILE, $filename) || die "BasPlug::read_file could not open $filename for reading ($!)\n";
|
---|
247 |
|
---|
248 | if ($self->{'input_encoding'} eq "ascii") {
|
---|
249 | undef $/;
|
---|
250 | $$textref = <FILE>;
|
---|
251 | $/ = "\n";
|
---|
252 | } else {
|
---|
253 | my $reader = new multiread();
|
---|
254 | $reader->set_handle ('BasPlug::FILE');
|
---|
255 | $reader->set_encoding ($self->{'input_encoding'});
|
---|
256 | $reader->read_file ($textref);
|
---|
257 |
|
---|
258 | if ($self->{'input_encoding'} eq "gb") {
|
---|
259 | # segment the Chinese words
|
---|
260 | $$textref = &cnseg::segment($$textref);
|
---|
261 | }
|
---|
262 | }
|
---|
263 |
|
---|
264 | close FILE;
|
---|
265 | }
|
---|
266 |
|
---|
267 | # add any extra metadata that's been passed around from one
|
---|
268 | # plugin to another.
|
---|
269 | # extra_metadata uses add_utf8_metadata so it expects metadata values
|
---|
270 | # to already be in utf8
|
---|
271 | sub extra_metadata {
|
---|
272 | my $self = shift (@_);
|
---|
273 | my ($doc_obj, $cursection, $metadata) = @_;
|
---|
274 |
|
---|
275 | foreach my $field (keys(%$metadata)) {
|
---|
276 | # $metadata->{$field} may be an array reference
|
---|
277 | if (ref ($metadata->{$field}) eq "ARRAY") {
|
---|
278 | map {
|
---|
279 | $doc_obj->add_utf8_metadata ($cursection, $field, $_);
|
---|
280 | } @{$metadata->{$field}};
|
---|
281 | } else {
|
---|
282 | $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
|
---|
283 | }
|
---|
284 | }
|
---|
285 | }
|
---|
286 |
|
---|
287 | # extract acronyms (and hopefully other stuff soon too).
|
---|
288 | sub auto_extract_metadata {
|
---|
289 | my $self = shift (@_);
|
---|
290 | my ($doc_obj) = @_;
|
---|
291 |
|
---|
292 | if ($self->{'extract_acronyms'}) {
|
---|
293 | my $thissection = $doc_obj->get_top_section();
|
---|
294 | while (defined $thissection) {
|
---|
295 | my $text = $doc_obj->get_text($thissection);
|
---|
296 | $self->extract_acronyms (\$text, $doc_obj, $thissection) if $text =~ /./;
|
---|
297 | $thissection = $doc_obj->get_next_section ($thissection);
|
---|
298 | }
|
---|
299 | }
|
---|
300 |
|
---|
301 | if ($self->{'markup_acronyms'}) {
|
---|
302 | my $thissection = $doc_obj->get_top_section();
|
---|
303 | while (defined $thissection) {
|
---|
304 | my $text = $doc_obj->get_text($thissection);
|
---|
305 | $text = $self->markup_acronyms ($text, $doc_obj, $thissection);
|
---|
306 | $doc_obj->delete_text($thissection);
|
---|
307 | $doc_obj->add_text($thissection, $text);
|
---|
308 | $thissection = $doc_obj->get_next_section ($thissection);
|
---|
309 | }
|
---|
310 | }
|
---|
311 |
|
---|
312 | if ($self->{'extract_language'}) {
|
---|
313 | my $thissection = $doc_obj->get_top_section();
|
---|
314 | while (defined $thissection) {
|
---|
315 | my $text = $doc_obj->get_text($thissection);
|
---|
316 | $self->extract_language (\$text, $doc_obj, $thissection) if $text =~ /./;
|
---|
317 | $thissection = $doc_obj->get_next_section ($thissection);
|
---|
318 | }
|
---|
319 | }
|
---|
320 |
|
---|
321 | }
|
---|
322 |
|
---|
323 |
|
---|
324 | # Identify the language of a section and add it to the metadata
|
---|
325 | sub extract_language {
|
---|
326 | my $self = shift (@_);
|
---|
327 | my ($textref, $doc_obj, $thissection) = @_;
|
---|
328 |
|
---|
329 | # remove all HTML tags
|
---|
330 | my $text = $$textref;
|
---|
331 | $text =~ s/<P[^>]*>/\n/sgi;
|
---|
332 | $text =~ s/<H[^>]*>/\n/sgi;
|
---|
333 | $text =~ s/<[^>]*>//sgi;
|
---|
334 | $text =~ tr/\n/\n/s;
|
---|
335 |
|
---|
336 | # get the language
|
---|
337 | my @results = textcat::classify($text);
|
---|
338 | @results = ("unknown") if ($#results > 2);
|
---|
339 |
|
---|
340 | # create language string and remove encoding information
|
---|
341 | my $language = join(" or ", @results);
|
---|
342 | $language =~ s/\-\w+//g;
|
---|
343 | $doc_obj->add_utf8_metadata($thissection, "Language", $language);
|
---|
344 | # print "Language: ", time, "-> $language\n";
|
---|
345 |
|
---|
346 | }
|
---|
347 |
|
---|
348 | # extract acronyms from a section in a document. progress is
|
---|
349 | # reported to STDERR based on the verbosity. both the Acronym
|
---|
350 | # and the AcronymKWIC metadata items are created.
|
---|
351 |
|
---|
352 | sub extract_acronyms {
|
---|
353 | my $self = shift (@_);
|
---|
354 | my ($textref, $doc_obj, $thissection) = @_;
|
---|
355 |
|
---|
356 | print STDERR " extracting acronyms ...\n"
|
---|
357 | if ($self->{'verbosity'} >= 2);
|
---|
358 |
|
---|
359 | my $acro_array = &acronym::acronyms($textref);
|
---|
360 |
|
---|
361 | foreach my $acro (@$acro_array) {
|
---|
362 |
|
---|
363 | #check that this is the first time ...
|
---|
364 | my $seen_before = "false";
|
---|
365 | my $previous_data = $doc_obj->get_metadata($thissection, "Acronym");
|
---|
366 | foreach my $thisAcro (@$previous_data) {
|
---|
367 | if ($thisAcro eq $acro->to_string())
|
---|
368 | {
|
---|
369 | $seen_before = "true";
|
---|
370 | print STDERR " already seen ". $acro->to_string() . "\n"
|
---|
371 | if ($self->{'verbosity'} >= 2);
|
---|
372 | }
|
---|
373 | }
|
---|
374 |
|
---|
375 | if ($seen_before eq "false")
|
---|
376 | {
|
---|
377 | #write it to the file ...
|
---|
378 | $acro->write_to_file();
|
---|
379 |
|
---|
380 | #do the normal acronym
|
---|
381 | $doc_obj->add_utf8_metadata($thissection, "Acronym", $acro->to_string());
|
---|
382 | print STDERR " adding ". $acro->to_string() . "\n"
|
---|
383 | if ($self->{'verbosity'} >= 1);
|
---|
384 |
|
---|
385 | # # do the KWIC (Key Word In Context) acronym
|
---|
386 | # my @kwic = $acro->to_string_kwic();
|
---|
387 | # foreach my $kwic (@kwic) {
|
---|
388 | # $doc_obj->add_utf8_metadata($thissection, "AcronymKWIC", $kwic);
|
---|
389 | # print STDERR " adding ". $kwic . "\n"
|
---|
390 | # if ($self->{'verbosity'} >= 2);
|
---|
391 | # }
|
---|
392 | }
|
---|
393 | }
|
---|
394 | print STDERR " done extracting acronyms. \n"
|
---|
395 | if ($self->{'verbosity'} >= 2);
|
---|
396 | }
|
---|
397 |
|
---|
398 | sub markup_acronyms {
|
---|
399 | my $self = shift (@_);
|
---|
400 | my ($text, $doc_obj, $thissection) = @_;
|
---|
401 |
|
---|
402 | print STDERR " marking up acronyms ...\n"
|
---|
403 | if ($self->{'verbosity'} >= 2);
|
---|
404 |
|
---|
405 | #self is passed in to check for verbosity ...
|
---|
406 | $text = &acronym::markup_acronyms($text, $self);
|
---|
407 |
|
---|
408 | print STDERR " done marking up acronyms. \n"
|
---|
409 | if ($self->{'verbosity'} >= 2);
|
---|
410 |
|
---|
411 | return $text;
|
---|
412 | }
|
---|
413 |
|
---|
414 | 1;
|
---|