source: main/trunk/greenstone2/perllib/plugins/BasePlugin.pm@ 21286

Last change on this file since 21286 was 21286, checked in by kjdon, 14 years ago

added empty removeold method
  • Property svn:keywords set to Author Date Id Revision
File size: 35.9 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
47# the different methods that can be applied when renaming
48# imported documents and their associated files
49our $file_rename_method_list =
50 [ { 'name' => "url",
51 'desc' => "{BasePlugin.rename_method.url}" },
52 { 'name' => "base64",
53 'desc' => "{BasePlugin.rename_method.base64}" },
54 { 'name' => "none",
55 'desc' => "{BasePlugin.rename_method.none}",
56 'hiddengli' => "yes" } ];
57
58our $encoding_list =
59 [ { 'name' => "ascii",
60 'desc' => "{BasePlugin.encoding.ascii}" },
61 { 'name' => "utf8",
62 'desc' => "{BasePlugin.encoding.utf8}" },
63 { 'name' => "unicode",
64 'desc' => "{BasePlugin.encoding.unicode}" } ];
65
66
67my $e = $encodings::encodings;
68foreach my $enc (sort {$e->{$a}->{'name'} cmp $e->{$b}->{'name'}} keys (%$e))
69{
70 my $hashEncode =
71 {'name' => $enc,
72 'desc' => $e->{$enc}->{'name'}};
73
74 push(@{$encoding_list},$hashEncode);
75}
76
77our $encoding_plus_auto_list =
78 [ { 'name' => "auto",
79 'desc' => "{BasePlugin.filename_encoding.auto}" },
80 { 'name' => "auto-language-analysis",
81 'desc' => "{BasePlugin.filename_encoding.auto_language_analysis}" }, # textcat
82 { 'name' => "auto-filesystem-encoding",
83 'desc' => "{BasePlugin.filename_encoding.auto_filesystem_encoding}" }, # locale
84 { 'name' => "auto-fl",
85 'desc' => "{BasePlugin.filename_encoding.auto_fl}" }, # locale followed by textcat
86 { 'name' => "auto-lf",
87 'desc' => "{BasePlugin.filename_encoding.auto_lf}" } ]; # texcat followed by locale
88
89push(@{$encoding_plus_auto_list},@{$encoding_list});
90
91our $oidtype_list =
92 [ { 'name' => "auto",
93 'desc' => "{BasePlugin.OIDtype.auto}" },
94 { 'name' => "hash",
95 'desc' => "{import.OIDtype.hash}" },
96 { 'name' => "assigned",
97 'desc' => "{import.OIDtype.assigned}" },
98 { 'name' => "incremental",
99 'desc' => "{import.OIDtype.incremental}" },
100 { 'name' => "dirname",
101 'desc' => "{import.OIDtype.dirname}" } ];
102
103my $arguments =
104 [ { 'name' => "process_exp",
105 'desc' => "{BasePlugin.process_exp}",
106 'type' => "regexp",
107 'deft' => "",
108 'reqd' => "no" },
109 { 'name' => "no_blocking",
110 'desc' => "{BasePlugin.no_blocking}",
111 'type' => "flag",
112 'reqd' => "no"},
113 { 'name' => "block_exp",
114 'desc' => "{BasePlugin.block_exp}",
115 'type' => "regexp",
116 'deft' => "",
117 'reqd' => "no" },
118 { 'name' => "associate_ext",
119 'desc' => "{BasePlugin.associate_ext}",
120 'type' => "string",
121 'reqd' => "no" },
122 { 'name' => "associate_tail_re",
123 'desc' => "{BasePlugin.associate_tail_re}",
124 'type' => "string",
125 'reqd' => "no" },
126 { 'name' => "OIDtype",
127 'desc' => "{import.OIDtype}",
128 'type' => "enum",
129 'list' => $oidtype_list,
130 # leave default empty so we can tell if its been set or not - if not set will use option from import.pl
131 'deft' => "auto",
132 'reqd' => "no" },
133 { 'name' => "OIDmetadata",
134 'desc' => "{import.OIDmetadata}",
135 'type' => "metadata",
136 'deft' => "dc.Identifier",
137 'reqd' => "no" },
138# { 'name' => "use_as_doc_identifier",
139# 'desc' => "{BasePlugin.use_as_doc_identifier}",
140# 'type' => "string",
141# 'reqd' => "no" ,
142# 'deft' => "" } ,
143 { 'name' => "no_cover_image",
144 'desc' => "{BasePlugin.no_cover_image}",
145 'type' => "flag",
146 'reqd' => "no" },
147 { 'name' => "filename_encoding",
148 'desc' => "{BasePlugin.filename_encoding}",
149 'type' => "enum",
150 'deft' => "auto",
151 'list' => $encoding_plus_auto_list,
152 'reqd' => "no" },
153 { 'name' => "smart_block",
154 'desc' => "{common.deprecated}. {BasePlugin.smart_block}",
155 'type' => "flag",
156 'reqd' => "no",
157 'hiddengli' => "yes" }, # deprecated, but leave in for old collections
158 { 'name' => "file_rename_method",
159 'desc' => "{BasePlugin.file_rename_method}",
160 'type' => "enum",
161 'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
162 'list' => $file_rename_method_list,
163 'reqd' => "no"
164 }
165
166 ];
167
168
169my $options = { 'name' => "BasePlugin",
170 'desc' => "{BasePlugin.desc}",
171 'abstract' => "yes",
172 'inherits' => "no",
173 'args' => $arguments };
174
175sub new {
176
177 my ($class) = shift (@_);
178 my ($pluginlist,$inputargs,$hashArgOptLists,$auxiliary) = @_;
179 push(@$pluginlist, $class);
180
181 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
182 push(@{$hashArgOptLists->{"OptList"}},$options);
183
184 my $self = new PrintInfo($pluginlist, $inputargs, $hashArgOptLists,$auxiliary);
185
186 if ($self->{'info_only'}) {
187 # don't worry about any options etc
188 return bless $self, $class;
189 }
190
191 if ($self->{'smart_block'}) {
192 print STDERR "WARNING: -smart_block option has been deprecated and is no longer useful\n";
193 }
194 $self->{'smart_block'} = undef;
195
196 my $plugin_name = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
197 $self->{'plugin_type'} = $plugin_name;
198
199 # remove ex. from OIDmetadata
200 $self->{'OIDmetadata'} =~ s/^ex\.// if defined $self->{'OIDmetadata'};
201 $self->{'num_processed'} = 0;
202 $self->{'num_not_processed'} = 0;
203 $self->{'num_blocked'} = 0;
204 $self->{'num_archives'} = 0;
205 $self->{'cover_image'} = 1; # cover image is on by default
206 $self->{'cover_image'} = 0 if ($self->{'no_cover_image'});
207 $self->{'can_process_directories'} = 0;
208 #$self->{'option_list'} = $hashArgOptLists->{"OptList"};
209
210 my $associate_ext = $self->{'associate_ext'};
211 if ((defined $associate_ext) && ($associate_ext ne "")) {
212
213 my $associate_tail_re = $self->{'associate_tail_re'};
214 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
215 my $outhandle = $self->{'outhandle'};
216 print $outhandle "Warning: can only specify 'associate_ext' or 'associate_tail_re'\n";
217 print $outhandle " defaulting to 'associate_tail_re'\n";
218 }
219 else {
220 my @exts = split(/,/,$associate_ext);
221
222 my @exts_bracketed = map { $_ = "(?:\\.$_)" } @exts;
223 my $associate_tail_re = join("|",@exts_bracketed);
224 $self->{'associate_tail_re'} = $associate_tail_re;
225 }
226
227 delete $self->{'associate_ext'};
228 }
229
230 return bless $self, $class;
231
232}
233
234sub merge_inheritance
235{
236 my $self = {};
237 my @child_selfs = @_;
238
239 foreach my $child_self (@child_selfs) {
240 foreach my $key (keys %$child_self) {
241 if (defined $self->{$key}) {
242 if ($self->{$key} ne $child_self->{$key}) {
243# print STDERR "Warning: Conflicting value in multiple inheritance for '$key'\n";
244# print STDERR "Existing stored value = $self->{$key}\n";
245# print STDERR "New (child) value = $child_self->{$key}\n";
246# print STDERR "Keeping existing value\n";
247 # Existing value seems to be option specified in collect.cfg
248
249 ### $self->{$key} = $child_self->{$key};
250
251 }
252 else {
253## print STDERR "****Info: Value $self->{$key} for $key already defined through multiple inheritance as the same value\n";
254 }
255
256 }
257 else {
258 $self->{$key} = $child_self->{$key};
259 }
260 }
261 }
262
263 return $self;
264}
265
266# initialize BasePlugin options
267# if init() is overridden in a sub-class, remember to call BasePlugin::init()
268sub init {
269 my $self = shift (@_);
270 my ($verbosity, $outhandle, $failhandle) = @_;
271
272 # verbosity is passed through from the processor
273 $self->{'verbosity'} = $verbosity;
274
275 # as are the outhandle and failhandle
276 $self->{'outhandle'} = $outhandle if defined $outhandle;
277 $self->{'failhandle'} = $failhandle;
278# $self->SUPER::init(@_);
279
280 # set process_exp and block_exp to defaults unless they were
281 # explicitly set
282
283 if ((!$self->is_recursive()) and
284 (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
285
286 $self->{'process_exp'} = $self->get_default_process_exp ();
287 if ($self->{'process_exp'} eq "") {
288 warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
289 }
290 }
291
292 if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
293 $self->{'block_exp'} = $self->get_default_block_exp ();
294 }
295
296}
297
298sub begin {
299 my $self = shift (@_);
300 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
301
302 if ($self->{'OIDtype'} eq "auto") {
303 # hasn't been set in the plugin, use the processor values
304 $self->{'OIDtype'} = $processor->{'OIDtype'};
305 $self->{'OIDmetadata'} = $processor->{'OIDmetadata'};
306 }
307 if ($self->{'OIDtype'} eq "hash") {
308 # should we hash on the file or on the doc xml??
309 $self->{'OIDtype'} = $self->get_oid_hash_type();
310 if ($self->{'OIDtype'} !~ /^(hash_on_file|hash_on_ga_xml)$/) {
311 $self->{'OIDtype'} = "hash_on_file";
312 }
313 }
314}
315
316# In case a plugin (eg PharosImageIndexer) needs to do something extra
317sub removeold {
318 my $self = shift (@_);
319 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
320}
321
322sub end {
323 # potentially called at the end of each plugin pass
324 # import.pl only has one plugin pass, but buildcol.pl has multiple ones
325
326 my ($self) = shift (@_);
327}
328
329sub deinit {
330 # called only once, after all plugin passes have been done
331
332 my ($self) = @_;
333}
334
335# default hashing type is to hash on the original file (or converted file)
336# override this to return hash_on_ga_xml for filetypes where hashing on the
337# file is no good eg video
338sub get_oid_hash_type {
339
340 my $self = shift (@_);
341
342 return "hash_on_file";
343}
344
345
346# this function should be overridden to return 1
347# in recursive plugins
348sub is_recursive {
349 my $self = shift (@_);
350
351 return 0;
352}
353
354sub get_default_block_exp {
355 my $self = shift (@_);
356
357 return "";
358}
359
360sub get_default_process_exp {
361 my $self = shift (@_);
362
363 return "";
364}
365
366# rename imported files and assoc files using URL encoding by default
367# as this will work for most plugins and give more legible filenames
368sub get_default_file_rename_method() {
369 my $self = shift (@_);
370 return "url";
371}
372
373# returns this plugin's active (possibly user-selected) file_rename_method
374sub get_file_rename_method() {
375 my $self = shift (@_);
376 my $rename_method = $self->{'file_rename_method'};
377 if($rename_method) {
378 return $rename_method;
379 } else {
380 return $self->get_default_file_rename_method();
381 }
382}
383
384# default implementation is to do nothing
385sub store_block_files {
386
387 my $self =shift (@_);
388 my ($filename_full_path, $block_hash) = @_;
389
390}
391
392# put files to block into hash
393sub use_block_expressions {
394
395 my $self =shift (@_);
396 my ($filename_full_path, $block_hash) = @_;
397
398 if ($self->{'block_exp'} ne "" && $filename_full_path =~ /$self->{'block_exp'}/) {
399 $block_hash->{'file_blocks'}->{$filename_full_path} = 1;
400 }
401
402}
403
404#default implementation is to block a file with same name as this, but extension jpg or JPG, if cover_images is on.
405sub block_cover_image
406{
407 my $self =shift;
408 my ($filename, $block_hash) = @_;
409
410 if ($self->{'cover_image'}) {
411 my $coverfile = $filename;
412 $coverfile =~ s/\.[^\\\/\.]+$/\.jpg/;
413 if (!-e $coverfile) {
414 $coverfile =~ s/jpg$/JPG/;
415 }
416 if (-e $coverfile) {
417 $block_hash->{'file_blocks'}->{$coverfile} = 1;
418 }
419 }
420
421 return;
422}
423
424
425# discover all the files that should be blocked by this plugin
426# check the args ...
427sub file_block_read {
428
429 my $self = shift (@_);
430 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $gli) = @_;
431 # Keep track of filenames with same root but different extensions
432 # Used to support -associate_ext and the more generalised
433 # -associate_tail_re
434 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
435
436 if (!-d $filename_full_path) {
437 $block_hash->{'all_files'}->{$file} = 1;
438 }
439
440 my $associate_tail_re = $self->{'associate_tail_re'};
441 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
442 my ($file_prefix,$file_ext)
443 = &util::get_prefix_and_tail_by_regex($filename_full_path,$associate_tail_re);
444 if ((defined $file_prefix) && (defined $file_ext)) {
445 my $shared_fileroot = $block_hash->{'shared_fileroot'};
446 if (!defined $shared_fileroot->{$file_prefix}) {
447 my $file_prefix_rec = { 'tie_to' => undef,
448 'exts' => {} };
449 $shared_fileroot->{$file_prefix} = $file_prefix_rec;
450 }
451
452 my $file_prefix_rec = $shared_fileroot->{$file_prefix};
453
454 if ($self->can_process_this_file($filename_full_path)) {
455 # This is the document the others should be tied to
456 $file_prefix_rec->{'tie_to'} = $file_ext;
457 }
458 else {
459 if ($file_ext =~ m/$associate_tail_re$/) {
460 # this file should be associated to the main one
461 $file_prefix_rec->{'exts'}->{$file_ext} = 1;
462 }
463 }
464
465 }
466 }
467
468 # check block expressions
469 $self->use_block_expressions($filename_full_path, $block_hash) unless $self->{'no_blocking'};
470
471 # now check whether we are actually processing this
472 if (!-f $filename_full_path || !$self->can_process_this_file($filename_full_path)) {
473 return undef; # can't recognise
474 }
475
476 # if we have a block_exp, then this overrides the normal 'smart' blocking
477 $self->store_block_files($filename_full_path, $block_hash) unless ($self->{'no_blocking'} || $self->{'block_exp'} ne "");
478
479 # block the cover image if there is one
480 if ($self->{'cover_image'}) {
481 $self->block_cover_image($filename_full_path, $block_hash);
482 }
483
484 return 1;
485}
486
487# plugins that rely on more than process_exp (eg XML plugins) can override this method
488sub can_process_this_file {
489 my $self = shift(@_);
490 my ($filename) = @_;
491
492 if (-d $filename && !$self->{'can_process_directories'}) {
493 return 0;
494 }
495 if ($self->{'process_exp'} ne "" && $filename =~ /$self->{'process_exp'}/) {
496 return 1;
497 }
498 return 0;
499
500}
501
502# just converts path as is to utf8.
503sub filepath_to_utf8 {
504 my $self = shift (@_);
505 my ($file, $file_encoding) = @_;
506 my $filemeta = $file;
507
508 my $filename_encoding = $self->{'filename_encoding'}; # filename encoding setting
509
510 # Whenever filename-encoding is set to any of the auto settings, we
511 # check if the filename is already in UTF8. If it is, then we're done.
512 if($filename_encoding =~ m/auto/) {
513 if(&unicode::check_is_utf8($filemeta))
514 {
515 $filename_encoding = "utf8";
516 return $filemeta;
517 }
518 }
519
520 # Auto setting, but filename is not utf8
521 if ($filename_encoding eq "auto")
522 {
523 # try textcat
524 $filename_encoding = $self->textcat_encoding($filemeta);
525
526 # check the locale next
527 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
528
529
530 # now try the encoding of the document, if available
531 if ($filename_encoding eq "undefined" && defined $file_encoding) {
532 $filename_encoding = $file_encoding;
533 }
534
535 }
536
537 elsif ($filename_encoding eq "auto-language-analysis")
538 {
539 $filename_encoding = $self->textcat_encoding($filemeta);
540
541 # now try the encoding of the document, if available
542 if ($filename_encoding eq "undefined" && defined $file_encoding) {
543 $filename_encoding = $file_encoding;
544 }
545 }
546
547 elsif ($filename_encoding eq "auto-filesystem-encoding")
548 {
549 # try locale
550 $filename_encoding = $self->locale_encoding();
551 }
552
553 elsif ($filename_encoding eq "auto-fl")
554 {
555 # filesystem-encoding (locale) then language-analysis (textcat)
556 $filename_encoding = $self->locale_encoding();
557
558 # try textcat
559 $filename_encoding = $self->textcat_encoding($filemeta) if $filename_encoding eq "undefined";
560
561 # else assume filename encoding is encoding of file content, if that's available
562 if ($filename_encoding eq "undefined" && defined $file_encoding) {
563 $filename_encoding = $file_encoding;
564 }
565 }
566
567 elsif ($filename_encoding eq "auto-lf")
568 {
569 # language-analysis (textcat) then filesystem-encoding (locale)
570 $filename_encoding = $self->textcat_encoding($filemeta);
571
572 # guess filename encoding from encoding of file content, if available
573 if ($filename_encoding eq "undefined" && defined $file_encoding) {
574 $filename_encoding = $file_encoding;
575 }
576
577 # try locale
578 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
579 }
580
581 # if still undefined, use utf8 as fallback
582 if ($filename_encoding eq "undefined") {
583 $filename_encoding = "utf8";
584 }
585
586 #print STDERR "**** UTF8 encoding the filename $filemeta ";
587
588 # if the filename encoding is set to utf8 but it isn't utf8 already--such as when
589 # 1. the utf8 fallback is used, or 2. if the system locale is used and happens to
590 # be always utf8 (in which case the filename's encoding is also set as utf8 even
591 # though the filename need not be if it originates from another system)--in such
592 # cases attempt to make the filename utf8 to match.
593 if($filename_encoding eq "utf8" && !&unicode::check_is_utf8($filemeta)) {
594 &unicode::ensure_utf8(\$filemeta);
595 }
596
597 # convert non-unicode encodings to utf8
598 if ($filename_encoding !~ m/(?:ascii|utf8|unicode)/) {
599 $filemeta = &unicode::unicode2utf8(
600 &unicode::convert2unicode($filename_encoding, \$filemeta)
601 );
602 }
603
604 #print STDERR " from encoding $filename_encoding -> $filemeta\n";
605 return $filemeta;
606}
607
608# gets the filename with no path, converts to utf8, and then dm safes it.
609# filename_encoding set by user
610sub filename_to_utf8_metadata
611{
612 my $self = shift (@_);
613 my ($file, $file_encoding) = @_;
614
615 my $outhandle = $self->{'outhandle'};
616
617 my ($filemeta) = $file =~ /([^\\\/]+)$/; # getting the tail of the filepath (skips all string parts containing slashes upto the end)
618 $filemeta = $self->filepath_to_utf8($filemeta, $file_encoding);
619
620 my $dmsafe_filemeta = &ghtml::dmsafe($filemeta);
621
622 return $dmsafe_filemeta;
623
624}
625
626sub locale_encoding {
627 my $self = shift(@_);
628
629 if (!defined $self->{'filesystem_encoding'}) {
630 $self->{'filesystem_encoding'} = $self->get_filesystem_encoding();
631 $self->{'filesystem_encoding'} = "undefined" if !defined $self->{'filesystem_encoding'};
632 }
633
634 #print STDERR "*** filename encoding determined based on locale: " . $self->{'filesystem_encoding'} . "\n";
635 return $self->{'filesystem_encoding'}; # can be the string "undefined"
636}
637
638sub textcat_encoding {
639 my $self = shift(@_);
640 my ($filemeta) = @_;
641
642 # analyse filenames without extensions and digits (and trimmed of surrounding
643 # whitespace), so that irrelevant chars don't confuse textcat
644 my $strictfilemeta = $filemeta;
645 $strictfilemeta =~ s/\.[^\.]+$//g;
646 $strictfilemeta =~ s/\d//g;
647 $strictfilemeta =~ s/^\s*//g;
648 $strictfilemeta =~ s/\s*$//g;
649
650 my $filename_encoding = $self->encoding_from_language_analysis($strictfilemeta);
651 if(!defined $filename_encoding) {
652 $filename_encoding = "undefined";
653 }
654
655 return $filename_encoding; # can be the string "undefined"
656}
657
658# performs textcat
659sub encoding_from_language_analysis {
660 my $self = shift(@_);
661 my ($text) = @_;
662
663 my $outhandle = $self->{'outhandle'};
664 my $best_encoding = undef;
665
666 # get the language/encoding of the textstring using textcat
667 $self->{'textcat'} = new textcat() unless defined($self->{'textcat'});
668 my $results = $self->{'textcat'}->classify_cached_filename(\$text);
669
670
671 if (scalar @$results < 0) {
672 return undef;
673 }
674
675 # We have some results, we choose the first
676 my ($language, $encoding) = $results->[0] =~ /^([^-]*)(?:-(.*))?$/;
677
678 $best_encoding = $encoding;
679 if (!defined $best_encoding) {
680 return undef;
681 }
682
683 if (defined $best_encoding && $best_encoding =~ m/^iso_8859/ && &unicode::check_is_utf8($text)) {
684 # the text is valid utf8, so assume that's the real encoding (since textcat is based on probabilities)
685 $best_encoding = 'utf8';
686 }
687
688
689 # check for equivalents where textcat doesn't have some encodings...
690 # eg MS versions of standard encodings
691 if (defined $best_encoding && $best_encoding =~ /^iso_8859_(\d+)/) {
692 my $iso = $1; # which variant of the iso standard?
693 # iso-8859 sets don't use chars 0x80-0x9f, windows codepages do
694 if ($text =~ /[\x80-\x9f]/) {
695 # Western Europe
696 if ($iso == 1 or $iso == 15) { $best_encoding = 'windows_1252' }
697 elsif ($iso == 2) {$best_encoding = 'windows_1250'} # Central Europe
698 elsif ($iso == 5) {$best_encoding = 'windows_1251'} # Cyrillic
699 elsif ($iso == 6) {$best_encoding = 'windows_1256'} # Arabic
700 elsif ($iso == 7) {$best_encoding = 'windows_1253'} # Greek
701 elsif ($iso == 8) {$best_encoding = 'windows_1255'} # Hebrew
702 elsif ($iso == 9) {$best_encoding = 'windows_1254'} # Turkish
703 }
704 }
705
706 if (defined $best_encoding && $best_encoding !~ /^(ascii|utf8|unicode)$/ &&
707 !defined $encodings::encodings->{$best_encoding})
708 {
709 if ($self->{'verbosity'}) {
710 gsprintf($outhandle, "BasePlugin: {ReadTextFile.unsupported_encoding}\n", $text, $best_encoding, "undef");
711 }
712 $best_encoding = undef;
713 }
714
715 return $best_encoding;
716}
717
718# uses locale
719sub get_filesystem_encoding {
720
721 my $self = shift(@_);
722
723 my $outhandle = $self->{'outhandle'};
724 my $filesystem_encoding = undef;
725
726 eval {
727 use POSIX qw(locale_h);
728
729 # With only one parameter, setlocale retrieves the
730 # current value
731 my $current_locale = setlocale(LC_CTYPE);
732
733 if ($current_locale =~ m/^.*\.(.*?)$/) {
734 my $char_encoding = lc($1);
735 if ($char_encoding =~ m/^(iso)(8859)(\d{1,2})$/) {
736 $char_encoding = "$1\_$2\_$3";
737 }
738
739 $char_encoding =~ s/-/_/g;
740 $char_encoding =~ s/^utf_8$/utf8/;
741
742 if ($char_encoding =~ m/^\d+$/) {
743 if (defined $encodings::encodings->{"windows_$char_encoding"}) {
744 $char_encoding = "windows_$char_encoding";
745 }
746 elsif (defined $encodings::encodings->{"dos_$char_encoding"}) {
747 $char_encoding = "dos_$char_encoding";
748 }
749 }
750
751 if (($char_encoding =~ m/(?:ascii|utf8|unicode)/)
752 || (defined $encodings::encodings->{$char_encoding})) {
753 $filesystem_encoding = $char_encoding;
754 }
755 else {
756 print $outhandle "Warning: Unsupported character encoding '$char_encoding' from locale '$current_locale'\n";
757 }
758 }
759
760
761 };
762 if ($@) {
763 print $outhandle "$@\n";
764 print $outhandle "Warning: Unable to establish locale. Will assume filesystem is UTF-8\n";
765
766 }
767 return $filesystem_encoding;
768}
769
770# is there ever only one Source? Sometimes this will be called twice, for images etc that are converted.
771sub set_Source_metadata {
772 my $self = shift (@_);
773 my ($doc_obj, $filename_no_path, $file_encoding) = @_;
774
775 my $top_section = $doc_obj->get_top_section();
776
777 # UTF-8 version of filename
778 my $filemeta = $self->filename_to_utf8_metadata($filename_no_path, $file_encoding);
779
780 # Source is the UTF8 display name - not necessarily the name of the file on the system
781 $doc_obj->set_utf8_metadata_element($top_section, "Source", $filemeta);
782
783 $filemeta = &util::rename_file($filemeta, $self->{'file_rename_method'});
784 # If using URL encoding, then SourceFile is the url-reference to url-encoded
785 # filemeta: it's a url that refers to the actual file on the system
786 $filemeta = &unicode::filename_to_url($filemeta);
787
788 $doc_obj->set_utf8_metadata_element($top_section, "SourceFile", $filemeta);
789}
790
791# this should be called by all plugins to set the oid of the doc obj, rather
792# than calling doc_obj->set_OID directly
793sub add_OID {
794 my $self = shift (@_);
795 my ($doc_obj) = @_;
796
797 $doc_obj->set_OIDtype($self->{'OIDtype'}, $self->{'OIDmetadata'});
798
799 # see if there is a plugin specific set_OID function
800 if (defined ($self->can('set_OID'))) {
801 $self->set_OID(@_); # pass through doc_obj and any extra arguments
802 }
803 else {
804 # use the default set_OID() in doc.pm
805 $doc_obj->set_OID();
806 }
807
808}
809
810# The BasePlugin read_into_doc_obj() function. This function does all the
811# right things to make general options work for a given plugin. It doesn't do anything with the file other than setting reads in
812# a file and sets up a slew of metadata all saved in doc_obj, which
813# it then returns as part of a tuple (process_status,doc_obj)
814#
815# Much of this functionality used to reside in read, but it was broken
816# down into a supporting routine to make the code more flexible.
817#
818# recursive plugins (e.g. RecPlug) and specialized plugins like those
819# capable of processing many documents within a single file (e.g.
820# GMLPlug) will normally want to implement their own version of
821# read_into_doc_obj()
822#
823# Note that $base_dir might be "" and that $file might
824# include directories
825
826# currently blocking has been done before it gets here - does this affect secondary plugin stuff??
827sub read_into_doc_obj {
828 my $self = shift (@_);
829 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
830
831 my $outhandle = $self->{'outhandle'};
832
833 # should we move this to read? What about secondary plugins?
834 print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
835 print $outhandle "$self->{'plugin_type'} processing $file\n"
836 if $self->{'verbosity'} > 1;
837
838 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
839
840 # create a new document
841 my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
842 my $top_section = $doc_obj->get_top_section();
843
844 $doc_obj->add_utf8_metadata($top_section, "Plugin", "$self->{'plugin_type'}");
845 $doc_obj->add_utf8_metadata($top_section, "FileSize", (-s $filename_full_path));
846
847
848 # sets the UTF8 filename (Source) for display and sets the url ref to URL encoded version
849 # of the UTF8 filename (SourceFile) for generated files
850 $self->set_Source_metadata($doc_obj, $filename_no_path);
851
852
853 # plugin specific stuff - what args do we need here??
854 unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
855 print STDERR "<ProcessingError n='$file'>\n" if ($gli);
856 return -1;
857 }
858
859 # include any metadata passed in from previous plugins
860 # note that this metadata is associated with the top level section
861 my $section = $doc_obj->get_top_section();
862 # can we merge these two methods??
863 $self->add_associated_files($doc_obj, $filename_full_path);
864 $self->extra_metadata ($doc_obj, $section, $metadata);
865 $self->auto_extract_metadata($doc_obj);
866
867 # if we haven't found any Title so far, assign one
868 # this was shifted to here from inside read()
869 $self->title_fallback($doc_obj,$section,$filename_no_path);
870
871 $self->add_OID($doc_obj);
872
873 $self->post_process_doc_obj($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli);
874 return (1,$doc_obj);
875}
876
877sub post_process_doc_obj {
878 my $self = shift (@_);
879 my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
880
881 return 1;
882}
883
884sub add_dummy_text {
885 my $self = shift(@_);
886 my ($doc_obj, $section) = @_;
887
888 # add NoText metadata so we can hide this dummy text in format statements
889 $doc_obj->add_metadata($section, "NoText", "1");
890 $doc_obj->add_text($section, &gsprintf::lookup_string("{BasePlugin.dummy_text}",1));
891
892}
893
894# does nothing. Can be overridden by subclass
895sub auto_extract_metadata {
896 my $self = shift(@_);
897 my ($doc_obj) = @_;
898}
899
900# adds cover image, associate_file options stuff. Should be called by sub class
901# read_into_doc_obj
902sub add_associated_files {
903 my $self = shift(@_);
904 # whatis filename??
905 my ($doc_obj, $filename) = @_;
906
907 # add in the cover image
908 if ($self->{'cover_image'}) {
909 $self->associate_cover_image($doc_obj, $filename);
910 }
911
912
913}
914
915# implement this if you are extracting metadata for other documents
916sub metadata_read {
917 my $self = shift (@_);
918 my ($pluginfo, $base_dir, $file, $block_hash,
919 $extrametakeys, $extrametadata, $extrametafile,
920 $processor, $maxdocs, $gli) = @_;
921
922 # can we process this file??
923 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
924 return undef unless $self->can_process_this_file($filename_full_path);
925
926 return 1; # we recognise the file, but don't actually do anything with it
927}
928
929
930# The BasePlugin read() function. This function calls read_into_doc_obj()
931# to ensure all the right things to make general options work for a
932# given plugin are done. It then calls the process() function which
933# does all the work specific to a plugin (like the old read functions
934# used to do). Most plugins should define their own process() function
935# and let this read() function keep control.
936#
937# recursive plugins (e.g. RecPlug) and specialized plugins like those
938# capable of processing many documents within a single file (e.g.
939# GMLPlug) might want to implement their own version of read(), but
940# more likely need to implement their own version of read_into_doc_obj()
941#
942# Return number of files processed, undef if can't recognise, -1 if can't
943# process
944
945sub read {
946 my $self = shift (@_);
947 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
948
949 # can we process this file??
950 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
951
952 return undef unless $self->can_process_this_file($filename_full_path);
953
954 my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
955
956 if ((defined $process_status) && ($process_status == 1)) {
957
958 # process the document
959 $processor->process($doc_obj);
960
961 $self->{'num_processed'} ++;
962 undef $doc_obj;
963 }
964 # delete any temp files that we may have created
965 $self->clean_up_after_doc_obj_processing();
966
967
968 # if process_status == 1, then the file has been processed.
969 return $process_status;
970
971}
972
973# returns undef if file is rejected by the plugin
974sub process {
975 my $self = shift (@_);
976 my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
977
978 gsprintf(STDERR, "BasePlugin::process {common.must_be_implemented}\n") && die "\n";
979
980 return undef; # never gets here
981}
982
983# overwrite this method to delete any temp files that we have created
984sub clean_up_after_doc_obj_processing {
985 my $self = shift(@_);
986
987}
988
989# write_file -- used by ConvertToPlug, for example in post processing
990#
991# where should this go, is here the best place??
992sub utf8_write_file {
993 my $self = shift (@_);
994 my ($textref, $filename) = @_;
995
996 if (!open (FILE, ">$filename")) {
997 gsprintf(STDERR, "ConvertToPlug::write_file {ConvertToPlug.could_not_open_for_writing} ($!)\n", $filename);
998 die "\n";
999 }
1000 print FILE $$textref;
1001
1002 close FILE;
1003}
1004
1005
1006sub filename_based_title
1007{
1008 my $self = shift (@_);
1009 my ($file) = @_;
1010
1011 my $file_derived_title = $file;
1012 $file_derived_title =~ s/_/ /g;
1013 $file_derived_title =~ s/\..*?$//;
1014
1015 return $file_derived_title;
1016}
1017
1018
1019sub title_fallback
1020{
1021 my $self = shift (@_);
1022 my ($doc_obj,$section,$file) = @_;
1023
1024 if (!defined $doc_obj->get_metadata_element ($section, "Title") or $doc_obj->get_metadata_element($section, "Title") eq "") {
1025
1026 my $file_derived_title = $self->filename_to_utf8_metadata($self->filename_based_title($file));
1027 if (!defined $doc_obj->get_metadata_element ($section, "Title")) {
1028 $doc_obj->add_utf8_metadata ($section, "Title", $file_derived_title);
1029 }
1030 else {
1031 $doc_obj->set_utf8_metadata_element ($section, "Title", $file_derived_title);
1032 }
1033 }
1034
1035}
1036
1037# add any extra metadata that's been passed around from one
1038# plugin to another.
1039# extra_metadata uses add_utf8_metadata so it expects metadata values
1040# to already be in utf8
1041sub extra_metadata {
1042 my $self = shift (@_);
1043 my ($doc_obj, $cursection, $metadata) = @_;
1044
1045 my $associate_tail_re = $self->{'associate_tail_re'};
1046
1047 foreach my $field (keys(%$metadata)) {
1048 # $metadata->{$field} may be an array reference
1049 if ($field eq "gsdlassocfile_tobe") {
1050 # 'gsdlassocfile_tobe' is artificially introduced metadata
1051 # that is used to signal that certain additional files should
1052 # be tied to this document. Useful in situations where a
1053 # metadata pass in the plugin pipeline works out some files
1054 # need to be associated with a document, but the document hasn't
1055 # been formed yet.
1056 my $equiv_form = "";
1057 foreach my $gaf (@{$metadata->{$field}}) {
1058 my ($full_filename,$mimetype) = ($gaf =~ m/^(.*):(.*):$/);
1059 my ($tail_filename) = ($full_filename =~ /^.*[\/\\](.+?)$/);
1060
1061 # we need to make sure the filename is valid utf-8 - we do
1062 # this by url or base64 encoding it
1063 # $tail_filename is the name that we store the file as
1064 $tail_filename = &util::rename_file($tail_filename, $self->{'file_rename_method'});
1065 $doc_obj->associate_file($full_filename,$tail_filename,$mimetype);
1066 $doc_obj->associate_source_file($full_filename);
1067 # If the filename is url_encoded, we need to encode the % signs
1068 # in the filename, so that it works in a url
1069 my $url_tail_filename = &unicode::filename_to_url($tail_filename);
1070 # work out extended tail extension (i.e. matching tail re)
1071
1072 my ($file_prefix,$file_extended_ext)
1073 = &util::get_prefix_and_tail_by_regex($tail_filename,$associate_tail_re);
1074 my ($pre_doc_ext) = ($file_extended_ext =~ m/^(.*)\..*$/);
1075 my ($doc_ext) = ($tail_filename =~ m/^.*\.(.*)$/);
1076 my $start_doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/{Or}{[parent(Top):archivedir],[archivedir]}/$url_tail_filename\">";
1077 my $srcicon = "_icon".$doc_ext."_";
1078 my $end_doclink = "</a>";
1079
1080 my $assoc_form = "$start_doclink\{If\}{$srcicon,$srcicon,$doc_ext\}$end_doclink";
1081
1082 if (defined $pre_doc_ext && $pre_doc_ext ne "") {
1083 # for metadata such as [mp3._edited] [mp3._full] ...
1084 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.$pre_doc_ext", $assoc_form);
1085 }
1086
1087 # for multiple metadata such as [mp3.assoclink]
1088 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.assoclink", $assoc_form);
1089
1090 $equiv_form .= " $assoc_form";
1091 }
1092 $doc_obj->add_utf8_metadata ($cursection, "equivlink", $equiv_form);
1093 }
1094 elsif (ref ($metadata->{$field}) eq "ARRAY") {
1095 map {
1096 $doc_obj->add_utf8_metadata ($cursection, $field, $_);
1097 } @{$metadata->{$field}};
1098 } else {
1099 $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
1100 }
1101 }
1102}
1103
1104
1105sub compile_stats {
1106 my $self = shift(@_);
1107 my ($stats) = @_;
1108
1109 $stats->{'num_processed'} += $self->{'num_processed'};
1110 $stats->{'num_not_processed'} += $self->{'num_not_processed'};
1111 $stats->{'num_archives'} += $self->{'num_archives'};
1112
1113}
1114
1115sub associate_cover_image {
1116 my $self = shift;
1117 my ($doc_obj, $filename) = @_;
1118
1119 $filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1120 if (exists $self->{'covers_missing_cache'}->{$filename}) {
1121 # don't stat() for existence eg for multiple document input files
1122 # (eg SplitPlug)
1123 return;
1124 }
1125
1126 my $top_section=$doc_obj->get_top_section();
1127
1128 if (-e $filename) {
1129 $doc_obj->associate_source_file($filename);
1130 $doc_obj->associate_file($filename, "cover.jpg", "image/jpeg");
1131 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1132 } else {
1133 my $upper_filename = $filename;
1134 $upper_filename =~ s/jpg$/JPG/;
1135 if (-e $upper_filename) {
1136 $doc_obj->associate_source_file($upper_filename);
1137 $doc_obj->associate_file($upper_filename, "cover.jpg",
1138 "image/jpeg");
1139 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1140 } else {
1141 # file doesn't exist, so record the fact that it's missing so
1142 # we don't stat() again (stat is slow)
1143 $self->{'covers_missing_cache'}->{$filename} = 1;
1144 }
1145 }
1146
1147}
1148
1149
1150# Overridden by exploding plugins (eg. ISISPlug)
1151sub clean_up_after_exploding
1152{
1153 my $self = shift(@_);
1154}
1155
1156
1157
11581;
Note: See TracBrowser for help on using the repository browser.