source: gsdl/trunk/perllib/plugins/BasePlugin.pm@ 16643

Last change on this file since 16643 was 16643, checked in by kjdon, 16 years ago

removed a couple of 'use xxx' lines that are not needed
  • Property svn:keywords set to Author Date Id Revision
File size: 32.3 KB
Line 
1###########################################################################
2#
3# BasePlugin.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) 1999-2005 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
26package BasePlugin;
27
28use strict;
29no strict 'subs';
30no strict 'refs'; # allow filehandles to be variables and viceversa
31
32use File::Basename;
33
34use encodings;
35use unicode;
36use textcat;
37use doc;
38use ghtml;
39use gsprintf 'gsprintf';
40
41use PrintInfo;
42
43BEGIN {
44 @BasePlugin::ISA = ( 'PrintInfo' );
45}
46
47our $encoding_list =
48 [ { 'name' => "ascii",
49 'desc' => "{BasePlugin.encoding.ascii}" },
50 { 'name' => "utf8",
51 'desc' => "{BasePlugin.encoding.utf8}" },
52 { 'name' => "unicode",
53 'desc' => "{BasePlugin.encoding.unicode}" } ];
54
55
56my $e = $encodings::encodings;
57foreach my $enc (sort {$e->{$a}->{'name'} cmp $e->{$b}->{'name'}} keys (%$e))
58{
59 my $hashEncode =
60 {'name' => $enc,
61 'desc' => $e->{$enc}->{'name'}};
62
63 push(@{$encoding_list},$hashEncode);
64}
65
66our $encoding_plus_auto_list =
67 [ { 'name' => "auto",
68 'desc' => "{BasePlugin.filename_encoding.auto}" },
69 { 'name' => "auto-language-analysis",
70 'desc' => "{BasePlugin.filename_encoding.auto_language_analysis}" }, # textcat
71 { 'name' => "auto-filesystem-encoding",
72 'desc' => "{BasePlugin.filename_encoding.auto_filesystem_encoding}" }, # locale
73 { 'name' => "auto-fl",
74 'desc' => "{BasePlugin.filename_encoding.auto_fl}" }, # locale followed by textcat
75 { 'name' => "auto-lf",
76 'desc' => "{BasePlugin.filename_encoding.auto_lf}" } ]; # texcat followed by locale
77
78push(@{$encoding_plus_auto_list},@{$encoding_list});
79
80my $arguments =
81 [ { 'name' => "process_exp",
82 'desc' => "{BasePlugin.process_exp}",
83 'type' => "regexp",
84 'deft' => "",
85 'reqd' => "no" },
86 { 'name' => "no_blocking",
87 'desc' => "{BasePlugin.no_blocking}",
88 'type' => "flag",
89 'reqd' => "no"},
90 { 'name' => "block_exp",
91 'desc' => "{BasePlugin.block_exp}",
92 'type' => "regexp",
93 'deft' => "",
94 'reqd' => "no" },
95 { 'name' => "associate_ext",
96 'desc' => "{BasePlugin.associate_ext}",
97 'type' => "string",
98 'reqd' => "no" },
99 { 'name' => "associate_tail_re",
100 'desc' => "{BasePlugin.associate_tail_re}",
101 'type' => "string",
102 'reqd' => "no" },
103 { 'name' => "use_as_doc_identifier",
104 'desc' => "{BasePlugin.use_as_doc_identifier}",
105 'type' => "string",
106 'reqd' => "no" ,
107 'deft' => "" } ,
108 { 'name' => "no_cover_image",
109 'desc' => "{BasePlugin.no_cover_image}",
110 'type' => "flag",
111 'reqd' => "no" },
112 { 'name' => "filename_encoding",
113 'desc' => "{BasePlugin.filename_encoding}",
114 'type' => "enum",
115 'deft' => "auto",
116 'list' => $encoding_plus_auto_list,
117 'reqd' => "no" },
118 { 'name' => "smart_block",
119 'desc' => "{common.deprecated}. {BasePlugin.smart_block}",
120 'type' => "flag",
121 'reqd' => "no",
122 'hiddengli' => "yes" } # deprecated, but leave in for old collections
123
124
125 ];
126
127
128my $options = { 'name' => "BasePlugin",
129 'desc' => "{BasePlugin.desc}",
130 'abstract' => "yes",
131 'inherits' => "no",
132 'args' => $arguments };
133
134
135sub new {
136
137 my ($class) = shift (@_);
138 my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
139 push(@$pluginlist, $class);
140
141 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
142 push(@{$hashArgOptLists->{"OptList"}},$options);
143
144 my $self = new PrintInfo($pluginlist, $inputargs, $hashArgOptLists);
145
146 if ($self->{'info_only'}) {
147 # don't worry about any options etc
148 return bless $self, $class;
149 }
150
151 if ($self->{'smart_block'}) {
152 print STDERR "WARNING: -smart_block option has been deprecated and is no longer useful\n";
153 }
154 $self->{'smart_block'} = undef;
155
156 my $plugin_name = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
157 $self->{'plugin_type'} = $plugin_name;
158
159 $self->{'num_processed'} = 0;
160 $self->{'num_not_processed'} = 0;
161 $self->{'num_blocked'} = 0;
162 $self->{'num_archives'} = 0;
163 $self->{'cover_image'} = 1; # cover image is on by default
164 $self->{'cover_image'} = 0 if ($self->{'no_cover_image'});
165 #$self->{'option_list'} = $hashArgOptLists->{"OptList"};
166
167 my $associate_ext = $self->{'associate_ext'};
168 if ((defined $associate_ext) && ($associate_ext ne "")) {
169
170 my $associate_tail_re = $self->{'associate_tail_re'};
171 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
172 my $outhandle = $self->{'outhandle'};
173 print $outhandle "Warning: can only specify 'associate_ext' or 'associate_tail_re'\n";
174 print $outhandle " defaulting to 'associate_tail_re'\n";
175 }
176 else {
177 my @exts = split(/,/,$associate_ext);
178
179 my @exts_bracketed = map { $_ = "(?:\\.$_)" } @exts;
180 my $associate_tail_re = join("|",@exts_bracketed);
181 $self->{'associate_tail_re'} = $associate_tail_re;
182 }
183
184 delete $self->{'associate_ext'};
185 }
186
187 return bless $self, $class;
188
189}
190
191# initialize BasePlugin options
192# if init() is overridden in a sub-class, remember to call BasePlugin::init()
193sub init {
194 my $self = shift (@_);
195 my ($verbosity, $outhandle, $failhandle) = @_;
196
197 # verbosity is passed through from the processor
198 $self->{'verbosity'} = $verbosity;
199
200 # as are the outhandle and failhandle
201 $self->{'outhandle'} = $outhandle if defined $outhandle;
202 $self->{'failhandle'} = $failhandle;
203# $self->SUPER::init(@_);
204
205 # set process_exp and block_exp to defaults unless they were
206 # explicitly set
207
208 if ((!$self->is_recursive()) and
209 (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
210
211 $self->{'process_exp'} = $self->get_default_process_exp ();
212 if ($self->{'process_exp'} eq "") {
213 warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
214 }
215 }
216
217 if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
218 $self->{'block_exp'} = $self->get_default_block_exp ();
219 }
220
221}
222
223sub begin {
224 my $self = shift (@_);
225 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
226}
227
228sub end {
229 # potentially called at the end of each plugin pass
230 # import.pl only has one plugin pass, but buildcol.pl has multiple ones
231
232 my ($self) = shift (@_);
233}
234
235sub deinit {
236 # called only once, after all plugin passes have been done
237
238 my ($self) = @_;
239}
240
241sub set_incremental {
242 my $self = shift(@_);
243 my ($incremental) = @_;
244
245 $self->{'incremental'} = $incremental;
246}
247
248# this function should be overridden to return 1
249# in recursive plugins
250sub is_recursive {
251 my $self = shift (@_);
252
253 return 0;
254}
255
256sub get_default_block_exp {
257 my $self = shift (@_);
258
259 return "";
260}
261
262sub get_default_process_exp {
263 my $self = shift (@_);
264
265 return "";
266}
267
268# default implementation is to do nothing
269sub store_block_files {
270
271 my $self =shift (@_);
272 my ($filename_full_path, $block_hash) = @_;
273
274}
275
276# put files to block into hash
277sub use_block_expressions {
278
279 my $self =shift (@_);
280 my ($filename_full_path, $block_hash) = @_;
281
282 if ($self->{'block_exp'} ne "" && $filename_full_path =~ /$self->{'block_exp'}/) {
283 $block_hash->{'file_blocks'}->{$filename_full_path} = 1;
284 }
285
286}
287
288#default implementation is to block a file with same name as this, but extension jpg or JPG, if cover_images is on.
289sub block_cover_image
290{
291 my $self =shift;
292 my ($filename, $block_hash) = @_;
293
294 if ($self->{'cover_image'}) {
295 my $coverfile = $filename;
296 $coverfile =~ s/\.[^\\\/\.]+$/\.jpg/;
297 if (!-e $coverfile) {
298 $coverfile =~ s/jpg$/JPG/;
299 }
300 if (-e $coverfile) {
301 $block_hash->{'file_blocks'}->{$coverfile} = 1;
302 }
303 }
304
305 return;
306}
307
308
309# discover all the files that should be blocked by this plugin
310# check the args ...
311sub file_block_read {
312
313 my $self = shift (@_);
314 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $gli) = @_;
315 # Keep track of filenames with same root but different extensions
316 # Used to support -associate_ext and the more generalised
317 # -associate_tail_re
318 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
319
320 my $associate_tail_re = $self->{'associate_tail_re'};
321 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
322
323 my ($file_prefix,$file_ext)
324 = &util::get_prefix_and_tail_by_regex($filename_full_path,$associate_tail_re);
325
326 if ((defined $file_prefix) && (defined $file_ext)) {
327 my $shared_fileroot = $block_hash->{'shared_fileroot'};
328 if (!defined $shared_fileroot->{$file_prefix}) {
329 my $file_prefix_rec = { 'tie_to' => undef,
330 'exts' => {} };
331 $shared_fileroot->{$file_prefix} = $file_prefix_rec;
332 }
333
334 my $file_prefix_rec = $shared_fileroot->{$file_prefix};
335
336 if ($self->can_process_this_file($filename_full_path)) {
337 # This is the document the others should be tied to
338 $file_prefix_rec->{'tie_to'} = $file_ext;
339 }
340 else {
341 if ($file_ext =~ m/$associate_tail_re$/) {
342 # this file should be associated to the main one
343 $file_prefix_rec->{'exts'}->{$file_ext} = 1;
344 }
345 }
346
347 }
348 }
349
350 # check block expressions
351 $self->use_block_expressions($filename_full_path, $block_hash) unless $self->{'no_blocking'};
352
353 # now check whether we are actually processing this
354 if (!-f $filename_full_path || !$self->can_process_this_file($filename_full_path)) {
355 return undef; # can't recognise
356 }
357
358 $self->store_block_files($filename_full_path, $block_hash) unless $self->{'no_blocking'};
359
360 # block the cover image if there is one
361 if ($self->{'cover_image'}) {
362 $self->block_cover_image($filename_full_path, $block_hash) unless $self->{'no_blocking'};
363 }
364
365 return 1;
366}
367
368# plugins that rely on more than process_exp (eg XML plugins) can override this method
369sub can_process_this_file {
370 my $self = shift(@_);
371 my ($filename) = @_;
372
373 if ($self->{'process_exp'} ne "" && $filename =~ /$self->{'process_exp'}/) {
374 return 1;
375 }
376 return 0;
377
378}
379
380# just converts path as is to utf8.
381sub filepath_to_utf8 {
382 my $self = shift (@_);
383 my ($file, $file_encoding) = @_;
384 my $filemeta = $file;
385
386 my $filename_encoding = $self->{'filename_encoding'}; # filename encoding setting
387
388## print STDERR "**** User chose filename encoding setting: $filename_encoding\n";
389
390 # Whenever filename-encoding is set to any of the auto settings, we
391 # check if the filename is already in UTF8. If it is, then we're done.
392 if($filename_encoding =~ m/auto/) {
393 if(&unicode::check_is_utf8($filemeta))
394 {
395## print STDERR "**** It is already UTF8\n";
396 $filename_encoding = "utf8";
397 return $filemeta;
398 }
399 }
400
401 # Auto setting, but filename is not utf8
402 if ($filename_encoding eq "auto")
403 {
404 # try textcat
405 $filename_encoding = $self->textcat_encoding($filemeta);
406
407 # check the locale next
408 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
409
410
411 # now try the encoding of the document, if available
412 if ($filename_encoding eq "undefined" && defined $file_encoding) {
413 $filename_encoding = $file_encoding;
414 }
415
416 }
417
418 elsif ($filename_encoding eq "auto-language-analysis")
419 {
420 $filename_encoding = $self->textcat_encoding($filemeta);
421
422 # now try the encoding of the document, if available
423 if ($filename_encoding eq "undefined" && defined $file_encoding) {
424 $filename_encoding = $file_encoding;
425 }
426 }
427
428 elsif ($filename_encoding eq "auto-filesystem-encoding")
429 {
430 # try locale
431 $filename_encoding = $self->locale_encoding();
432 }
433
434 elsif ($filename_encoding eq "auto-fl")
435 {
436 # filesystem-encoding (locale) then language-analysis (textcat)
437 $filename_encoding = $self->locale_encoding();
438
439 # try textcat
440 $filename_encoding = $self->textcat_encoding($filemeta) if $filename_encoding eq "undefined";
441
442 # else assume filename encoding is encoding of file content, if that's available
443 if ($filename_encoding eq "undefined" && defined $file_encoding) {
444 $filename_encoding = $file_encoding;
445 }
446 }
447
448 elsif ($filename_encoding eq "auto-lf")
449 {
450 # language-analysis (textcat) then filesystem-encoding (locale)
451 $filename_encoding = $self->textcat_encoding($filemeta);
452
453 # guess filename encoding from encoding of file content, if available
454 if ($filename_encoding eq "undefined" && defined $file_encoding) {
455 $filename_encoding = $file_encoding;
456 }
457
458 # try locale
459 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
460 }
461
462## print STDERR "**** filename_encoding selected: $filename_encoding \n";
463
464 # if still undefined, use utf8 as fallback
465 if ($filename_encoding eq "undefined") {
466 $filename_encoding = "utf8";
467 }
468
469 # if the filename encoding is set to utf8 but it isn't utf8 already--such as when
470 # 1. the utf8 fallback is used, or 2. if the system locale is used and happens to
471 # be always utf8 (in which case the filename's encoding is also set as utf8 even
472 # though the filename need not be if it originates from another system)--in such
473 # cases attempt to make the filename utf8 to match.
474 if($filename_encoding eq "utf8" && !&unicode::check_is_utf8($filemeta)) {
475## print STDERR "**** BEFORE utf8 conversion: $filemeta\n";
476 &unicode::ensure_utf8(\$filemeta);
477## print STDERR "**** AFTER utf8 conversion: $filemeta\n";
478 }
479
480
481 # convert non-unicode encodings to utf8
482 if ($filename_encoding !~ m/(?:ascii|utf8|unicode)/) {
483 $filemeta = &unicode::unicode2utf8(
484 &unicode::convert2unicode($filename_encoding, \$filemeta)
485 );
486 }
487
488 print "*** filename encoding found: $filename_encoding\n";
489 print "*** utf8 encoded filename: $filemeta\n";
490
491 return $filemeta;
492}
493
494# gets the filename with no path, converts to utf8, and then dm safes it.
495#filename_encoding set by user
496sub filename_to_utf8_metadata
497{
498 my $self = shift (@_);
499 my ($file, $file_encoding) = @_;
500
501 my $outhandle = $self->{'outhandle'};
502
503 my ($filemeta) = $file =~ /([^\\\/]+)$/; # getting the tail of the filepath (skips all string parts containing slashes upto the end)
504 $filemeta = $self->filepath_to_utf8($filemeta, $file_encoding);
505
506 my $dmsafe_filemeta = &ghtml::dmsafe($filemeta);
507
508 return $dmsafe_filemeta;
509
510}
511
512sub locale_encoding {
513 my $self = shift(@_);
514
515 if (!defined $self->{'filesystem_encoding'}) {
516 $self->{'filesystem_encoding'} = $self->get_filesystem_encoding();
517 $self->{'filesystem_encoding'} = "undefined" if !defined $self->{'filesystem_encoding'};
518 }
519
520 print "filename encoding determined based on locale: " . $self->{'filesystem_encoding'} . "\n";
521 return $self->{'filesystem_encoding'}; # can be the string "undefined"
522}
523
524sub textcat_encoding {
525 my $self = shift(@_);
526 my ($filemeta) = @_;
527
528 # analyse filenames without extensions and digits (and trimmed of surrounding
529 # whitespace), so that irrelevant chars don't confuse textcat
530 my $strictfilemeta = $filemeta;
531 $strictfilemeta =~ s/\.[^\.]+$//g;
532 $strictfilemeta =~ s/\d//g;
533 $strictfilemeta =~ s/^\s*//g;
534 $strictfilemeta =~ s/\s*$//g;
535
536## print STDERR "**** strict filename is |$strictfilemeta|\n";
537 my $filename_encoding = $self->encoding_from_language_analysis($strictfilemeta);
538 if(!defined $filename_encoding) {
539 $filename_encoding = "undefined";
540 }
541
542## print STDERR "**** textcat found filename encoding: " . $file_textcat_encoding_map{$strictfilemeta} . "\n";
543 return $filename_encoding; # can be the string "undefined"
544}
545
546# performs textcat
547sub encoding_from_language_analysis {
548 my $self = shift(@_);
549 my ($text) = @_;
550
551 my $outhandle = $self->{'outhandle'};
552 my $best_encoding = undef;
553
554 # get the language/encoding of the file using textcat
555 $self->{'textcat'} = new textcat() unless defined($self->{'textcat'});
556 #my $results = $self->{'textcat'}->classify(\$text);
557 my $results = $self->{'textcat'}->classify_cached(\$text);
558
559
560 if (scalar @$results < 0) {
561 print STDERR "**** Textcat returned 0 results\n";
562 return undef;
563 }
564
565 print STDERR "**** TEXTCAT RESULTS for $text: ";
566 print STDERR join(",", @$results);
567 print STDERR "\n";
568
569 # We have some results, we choose the first
570 my ($language, $encoding) = $results->[0] =~ /^([^-]*)(?:-(.*))?$/;
571
572 $best_encoding = $encoding;
573 if (!defined $best_encoding) {
574## print STDERR "**** Textcat cannot determine encoding of filename: it's undefined.\n";
575 return undef;
576 }
577
578 if (defined $best_encoding && $best_encoding =~ m/^iso_8859/ && &unicode::check_is_utf8($text)) {
579 # the text is valid utf8, so assume that's the real encoding (since textcat is based on probabilities)
580## print STDERR "*** Filename turns out to be UTF8\n";
581 $best_encoding = 'utf8';
582 }
583
584
585 # check for equivalents where textcat doesn't have some encodings...
586 # eg MS versions of standard encodings
587 if (defined $best_encoding && $best_encoding =~ /^iso_8859_(\d+)/) {
588## print STDERR "**** best_encoding is ISO_8859: $best_encoding\n";
589
590 my $iso = $1; # which variant of the iso standard?
591 # iso-8859 sets don't use chars 0x80-0x9f, windows codepages do
592 if ($text =~ /[\x80-\x9f]/) {
593## print STDERR "**** best_encoding is some windows value: $best_encoding\n";
594 # Western Europe
595 if ($iso == 1 or $iso == 15) { $best_encoding = 'windows_1252' }
596 elsif ($iso == 2) {$best_encoding = 'windows_1250'} # Central Europe
597 elsif ($iso == 5) {$best_encoding = 'windows_1251'} # Cyrillic
598 elsif ($iso == 6) {$best_encoding = 'windows_1256'} # Arabic
599 elsif ($iso == 7) {$best_encoding = 'windows_1253'} # Greek
600 elsif ($iso == 8) {$best_encoding = 'windows_1255'} # Hebrew
601 elsif ($iso == 9) {$best_encoding = 'windows_1254'} # Turkish
602## print STDERR "**** best_encoding windows value: $best_encoding\n";
603 }
604 }
605
606 if (defined $best_encoding && $best_encoding !~ /^(ascii|utf8|unicode)$/ &&
607 !defined $encodings::encodings->{$best_encoding})
608 {
609 if ($self->{'verbosity'}) {
610 gsprintf($outhandle, "BasePlugin: {ReadTextFile.unsupported_encoding}\n", $text, $best_encoding, "undef");
611 }
612## print STDERR "***** unsupported encoding: $best_encoding. Setting it to undefined.\n";
613 $best_encoding = undef;
614 }
615## print STDERR "**** language: $language\n" if defined $language;
616## print STDERR "**** encoding: $best_encoding\n" if defined $encoding;
617
618 return $best_encoding;
619}
620
621# uses locale
622sub get_filesystem_encoding {
623
624 my $self = shift(@_);
625
626 my $outhandle = $self->{'outhandle'};
627 my $filesystem_encoding = undef;
628
629 eval {
630 use POSIX qw(locale_h);
631
632 # With only one parameter, setlocale retrieves the
633 # current value
634 my $current_locale = setlocale(LC_CTYPE);
635
636 if ($current_locale =~ m/^.*\.(.*?)$/) {
637 my $char_encoding = lc($1);
638 if ($char_encoding =~ m/^(iso)(8859)(\d{1,2})$/) {
639 $char_encoding = "$1\_$2\_$3";
640 }
641
642 $char_encoding =~ s/-/_/g;
643 $char_encoding =~ s/^utf_8$/utf8/;
644
645 if ($char_encoding =~ m/^\d+$/) {
646 if (defined $encodings::encodings->{"windows_$char_encoding"}) {
647 $char_encoding = "windows_$char_encoding";
648 }
649 elsif (defined $encodings::encodings->{"dos_$char_encoding"}) {
650 $char_encoding = "dos_$char_encoding";
651 }
652 }
653
654 if (($char_encoding =~ m/(?:ascii|utf8|unicode)/)
655 || (defined $encodings::encodings->{$char_encoding})) {
656 $filesystem_encoding = $char_encoding;
657 }
658 else {
659 print $outhandle "Warning: Unsupported character encoding '$char_encoding' from locale '$current_locale'\n";
660 }
661 }
662
663
664 };
665 if ($@) {
666 print $outhandle "$@\n";
667 print $outhandle "Warning: Unable to establish locale. Will assume filesytem is UTF-8\n";
668
669 }
670 return $filesystem_encoding;
671}
672
673# is there ever only one Source? Sometimes this will be called twice, for images etc that are converted.
674sub set_Source_metadata {
675 my $self = shift (@_);
676 my ($doc_obj, $filename_no_path, $file_encoding) = @_;
677
678 my $top_section = $doc_obj->get_top_section();
679
680 # UTF-8 version of filename
681 my $filemeta = $self->filename_to_utf8_metadata($filename_no_path, $file_encoding);
682 $doc_obj->set_utf8_metadata_element($top_section, "Source", $filemeta);
683
684}
685
686sub add_OID {
687
688 my $self = shift (@_);
689 my ($doc_obj) = @_;
690
691 # See if a metadata field is specified as the field
692 if ((defined $self->{'use_as_doc_identifier'}) && ($self->{'use_as_doc_identifier'} ne "")) {
693 my $metadata_doc_id = $self->{'use_as_doc_identifier'};
694
695 # Consider "tidying" up metadata_doc_id to be something
696 # suitable in a URL
697 # Could even support a user specified plugin RE for this.
698
699 my $top_section = $doc_obj->get_top_section();
700 my $oid = $doc_obj->get_metadata_element($top_section,$metadata_doc_id);
701## print STDERR "**** oid = $oid\n";
702 $doc_obj->set_OID($oid);
703 }
704 # See if there is a plugin-specific set_OID function...
705 elsif (defined ($self->can('set_OID'))) {
706 # it will need $doc_obj to set the Identifier metadata...
707 $self->set_OID(@_); # pass through any extra arguments supplied
708 } else {
709 # use the default set_OID() in doc.pm
710 $doc_obj->set_OID();
711 }
712}
713
714
715
716# The BasePlugin read_into_doc_obj() function. This function does all the
717# right things to make general options work for a given plugin. It doesn't do anything with the file other than setting reads in
718# a file and sets up a slew of metadata all saved in doc_obj, which
719# it then returns as part of a tuple (process_status,doc_obj)
720#
721# Much of this functionality used to reside in read, but it was broken
722# down into a supporting routine to make the code more flexible.
723#
724# recursive plugins (e.g. RecPlug) and specialized plugins like those
725# capable of processing many documents within a single file (e.g.
726# GMLPlug) will normally want to implement their own version of
727# read_into_doc_obj()
728#
729# Note that $base_dir might be "" and that $file might
730# include directories
731
732# currently blocking has been done before it gets here - does this affect secondary plugin stuff??
733sub read_into_doc_obj {
734 my $self = shift (@_);
735 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
736
737 my $outhandle = $self->{'outhandle'};
738
739 # should we move this to read? What about secondary plugins?
740 print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
741 print $outhandle "$self->{'plugin_type'} processing $file\n"
742 if $self->{'verbosity'} > 1;
743
744 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
745 # create a new document
746 my $doc_obj = new doc ($filename_full_path, "indexed_doc");
747 my $top_section = $doc_obj->get_top_section();
748
749 # this should look at the plugin option too...
750 $doc_obj->set_OIDtype ($processor->{'OIDtype'}, $processor->{'OIDmetadata'});
751 $doc_obj->add_utf8_metadata($top_section, "Plugin", "$self->{'plugin_type'}");
752 $doc_obj->add_utf8_metadata($top_section, "FileSize", (-s $filename_full_path));
753
754 $self->set_Source_metadata($doc_obj, $filename_no_path);
755
756 # plugin specific stuff - what args do we need here??
757 unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
758 print STDERR "<ProcessingError n='$file'>\n" if ($gli);
759 return -1;
760 }
761
762 # include any metadata passed in from previous plugins
763 # note that this metadata is associated with the top level section
764 my $section = $doc_obj->get_top_section();
765 # can we merge these two methods??
766 $self->add_associated_files($doc_obj, $filename_full_path);
767 $self->extra_metadata ($doc_obj, $section, $metadata);
768 $self->auto_extract_metadata($doc_obj);
769
770 # if we haven't found any Title so far, assign one
771 # this was shifted to here from inside read()
772 $self->title_fallback($doc_obj,$section,$filename_no_path);
773
774 $self->add_OID($doc_obj);
775
776 return (1,$doc_obj);
777}
778
779sub add_dummy_text {
780 my $self = shift(@_);
781 my ($doc_obj, $section) = @_;
782
783 # add NoText metadata so we can hide this dummy text in format statements
784 $doc_obj->add_metadata($section, "NoText", "1");
785 $doc_obj->add_text($section, &gsprintf::lookup_string("{BasePlugin.dummy_text}",1));
786
787}
788
789# does nothing. Can be overridden by subclass
790sub auto_extract_metadata {
791 my $self = shift(@_);
792 my ($doc_obj) = @_;
793}
794
795# adds cover image, associate_file options stuff. Should be called by sub class
796# read_into_doc_obj
797sub add_associated_files {
798 my $self = shift(@_);
799 # whatis filename??
800 my ($doc_obj, $filename) = @_;
801
802 # add in the cover image
803 if ($self->{'cover_image'}) {
804 $self->associate_cover_image($doc_obj, $filename);
805 }
806
807
808}
809
810# implement this if you are extracting metadata for other documents
811sub metadata_read {
812 my $self = shift (@_);
813 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $extrametakeys, $extrametadata, $processor, $maxdocs, $gli) = @_;
814
815 # can we process this file??
816 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
817 return undef unless $self->can_process_this_file($filename_full_path);
818
819 return 1; # we recognise the file, but don't actually do anything with it
820}
821
822
823# The BasePlugin read() function. This function calls read_into_doc_obj()
824# to ensure all the right things to make general options work for a
825# given plugin are done. It then calls the process() function which
826# does all the work specific to a plugin (like the old read functions
827# used to do). Most plugins should define their own process() function
828# and let this read() function keep control.
829#
830# recursive plugins (e.g. RecPlug) and specialized plugins like those
831# capable of processing many documents within a single file (e.g.
832# GMLPlug) might want to implement their own version of read(), but
833# more likely need to implement their own version of read_into_doc_obj()
834#
835# Return number of files processed, undef if can't recognise, -1 if can't
836# process
837
838sub read {
839 my $self = shift (@_);
840 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
841
842 # can we process this file??
843 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
844 return undef unless $self->can_process_this_file($filename_full_path);
845
846 my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
847
848 if ((defined $process_status) && ($process_status == 1)) {
849
850 # process the document
851 $processor->process($doc_obj);
852
853 $self->{'num_processed'} ++;
854 undef $doc_obj;
855 }
856 # delete any temp files that we may have created
857 $self->clean_up_after_doc_obj_processing();
858
859 # if process_status == 1, then the file has been processed.
860 return $process_status;
861
862}
863
864# returns undef if file is rejected by the plugin
865sub process {
866 my $self = shift (@_);
867 my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
868
869 gsprintf(STDERR, "BasePlugin::process {common.must_be_implemented}\n") && die "\n";
870
871 return undef; # never gets here
872}
873
874# overwrite this method to delete any temp files that we have created
875sub clean_up_after_doc_obj_processing {
876 my $self = shift(@_);
877
878}
879
880# write_file -- used by ConvertToPlug, for example in post processing
881#
882# where should this go, is here the best place??
883sub utf8_write_file {
884 my $self = shift (@_);
885 my ($textref, $filename) = @_;
886
887 if (!open (FILE, ">$filename")) {
888 gsprintf(STDERR, "ConvertToPlug::write_file {ConvertToPlug.could_not_open_for_writing} ($!)\n", $filename);
889 die "\n";
890 }
891 print FILE $$textref;
892
893 close FILE;
894}
895
896
897sub filename_based_title
898{
899 my $self = shift (@_);
900 my ($file) = @_;
901
902 my $file_derived_title = $file;
903 $file_derived_title =~ s/_/ /g;
904 $file_derived_title =~ s/\..*?$//;
905
906 return $file_derived_title;
907}
908
909
910sub title_fallback
911{
912 my $self = shift (@_);
913 my ($doc_obj,$section,$file) = @_;
914
915 if (!defined $doc_obj->get_metadata_element ($section, "Title") or $doc_obj->get_metadata_element($section, "Title") eq "") {
916
917 my $file_derived_title = $self->filename_to_utf8_metadata($self->filename_based_title($file));
918 if (!defined $doc_obj->get_metadata_element ($section, "Title")) {
919 $doc_obj->add_utf8_metadata ($section, "Title", $file_derived_title);
920 }
921 else {
922 $doc_obj->set_utf8_metadata ($section, "Title", $file_derived_title);
923 }
924 }
925
926}
927
928# add any extra metadata that's been passed around from one
929# plugin to another.
930# extra_metadata uses add_utf8_metadata so it expects metadata values
931# to already be in utf8
932sub extra_metadata {
933 my $self = shift (@_);
934 my ($doc_obj, $cursection, $metadata) = @_;
935
936 my $associate_tail_re = $self->{'associate_tail_re'};
937
938 foreach my $field (keys(%$metadata)) {
939 # $metadata->{$field} may be an array reference
940 if ($field eq "gsdlassocfile_tobe") {
941 # 'gsdlassocfile_tobe' is artificially introduced metadata
942 # that is used to signal that certain additional files should
943 # be tied to this document. Useful in situations where a
944 # metadata pass in the plugin pipeline works out some files
945 # need to be associated with a document, but the document hasn't
946 # been formed yet.
947 my $equiv_form = "";
948 foreach my $gaf (@{$metadata->{$field}}) {
949 my ($full_filename,$mimetype) = ($gaf =~ m/^(.*):(.*):$/);
950 my ($tail_filename) = ($full_filename =~ /^.*[\/\\](.+?)$/);
951 my $filename = $full_filename;
952 $doc_obj->associate_file($full_filename,$tail_filename,$mimetype);
953
954 # work out extended tail extension (i.e. matching tail re)
955
956 my ($file_prefix,$file_extended_ext)
957 = &util::get_prefix_and_tail_by_regex($tail_filename,$associate_tail_re);
958 my ($pre_doc_ext) = ($file_extended_ext =~ m/^(.*)\..*$/);
959
960 my ($doc_ext) = ($tail_filename =~ m/^.*\.(.*)$/);
961 my $start_doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/{Or}{[parent(Top):archivedir],[archivedir]}/$tail_filename\">";
962 my $srcicon = "_icon".$doc_ext."_";
963 my $end_doclink = "</a>";
964
965 my $assoc_form = "$start_doclink\{If\}{$srcicon,$srcicon,$doc_ext\}$end_doclink";
966
967 if (defined $pre_doc_ext) {
968 # for metadata such as [mp3._edited] [mp3._full] ...
969 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.$pre_doc_ext", $assoc_form);
970 }
971
972 # for multiple metadata such as [mp3.assoclink]
973 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.assoclink", $assoc_form);
974
975 $equiv_form .= " $assoc_form";
976 }
977 $doc_obj->add_utf8_metadata ($cursection, "equivlink", $equiv_form);
978 }
979 elsif (ref ($metadata->{$field}) eq "ARRAY") {
980 map {
981 $doc_obj->add_utf8_metadata ($cursection, $field, $_);
982 } @{$metadata->{$field}};
983 } else {
984 $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
985 }
986 }
987}
988
989
990sub compile_stats {
991 my $self = shift(@_);
992 my ($stats) = @_;
993
994 $stats->{'num_processed'} += $self->{'num_processed'};
995 $stats->{'num_not_processed'} += $self->{'num_not_processed'};
996 $stats->{'num_archives'} += $self->{'num_archives'};
997
998}
999
1000sub associate_cover_image {
1001 my $self = shift;
1002 my ($doc_obj, $filename) = @_;
1003
1004 $filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1005 if (exists $self->{'covers_missing_cache'}->{$filename}) {
1006 # don't stat() for existence eg for multiple document input files
1007 # (eg SplitPlug)
1008 return;
1009 }
1010
1011 my $top_section=$doc_obj->get_top_section();
1012
1013 if (-e $filename) {
1014 $doc_obj->associate_file($filename, "cover.jpg", "image/jpeg");
1015 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1016 } else {
1017 my $upper_filename = $filename;
1018 $upper_filename =~ s/jpg$/JPG/;
1019 if (-e $upper_filename) {
1020 $doc_obj->associate_file($upper_filename, "cover.jpg",
1021 "image/jpeg");
1022 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1023 } else {
1024 # file doesn't exist, so record the fact that it's missing so
1025 # we don't stat() again (stat is slow)
1026 $self->{'covers_missing_cache'}->{$filename} = 1;
1027 }
1028 }
1029
1030}
1031
1032
1033# Overridden by exploding plugins (eg. ISISPlug)
1034sub clean_up_after_exploding
1035{
1036 my $self = shift(@_);
1037}
1038
1039
1040
10411;
Note: See TracBrowser for help on using the repository browser.