source: main/trunk/greenstone2/perllib/plugins/BasePlugin.pm@ 23363

Last change on this file since 23363 was 23363, checked in by davidb, 13 years ago

Plugin code upgrade to support Greenstone working with filenames under Windows when then go beyond Latin-1 and start turning up in their DOS abbreviated form (e.g. Test~1.txt)
  • Property svn:keywords set to Author Date Id Revision
File size: 44.0 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;
33use Encode;
34
35use encodings;
36use unicode;
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' => "store_original_file",
119 'desc' => "{BasePlugin.store_original_file}",
120 'type' => "flag",
121 'reqd' => "no" },
122 { 'name' => "associate_ext",
123 'desc' => "{BasePlugin.associate_ext}",
124 'type' => "string",
125 'reqd' => "no" },
126 { 'name' => "associate_tail_re",
127 'desc' => "{BasePlugin.associate_tail_re}",
128 'type' => "string",
129 'reqd' => "no" },
130 { 'name' => "OIDtype",
131 'desc' => "{import.OIDtype}",
132 'type' => "enum",
133 'list' => $oidtype_list,
134 # leave default empty so we can tell if its been set or not - if not set will use option from import.pl
135 'deft' => "auto",
136 'reqd' => "no" },
137 { 'name' => "OIDmetadata",
138 'desc' => "{import.OIDmetadata}",
139 'type' => "metadata",
140 'deft' => "dc.Identifier",
141 'reqd' => "no" },
142# { 'name' => "use_as_doc_identifier",
143# 'desc' => "{BasePlugin.use_as_doc_identifier}",
144# 'type' => "string",
145# 'reqd' => "no" ,
146# 'deft' => "" } ,
147 { 'name' => "no_cover_image",
148 'desc' => "{BasePlugin.no_cover_image}",
149 'type' => "flag",
150 'reqd' => "no" },
151 { 'name' => "filename_encoding",
152 'desc' => "{BasePlugin.filename_encoding}",
153 'type' => "enum",
154 'deft' => "auto",
155 'list' => $encoding_plus_auto_list,
156 'reqd' => "no" },
157 { 'name' => "smart_block",
158 'desc' => "{common.deprecated}. {BasePlugin.smart_block}",
159 'type' => "flag",
160 'reqd' => "no",
161 'hiddengli' => "yes" }, # deprecated, but leave in for old collections
162 { 'name' => "file_rename_method",
163 'desc' => "{BasePlugin.file_rename_method}",
164 'type' => "enum",
165 'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
166 'list' => $file_rename_method_list,
167 'reqd' => "no"
168 }
169
170 ];
171
172
173my $options = { 'name' => "BasePlugin",
174 'desc' => "{BasePlugin.desc}",
175 'abstract' => "yes",
176 'inherits' => "no",
177 'args' => $arguments };
178
179sub new {
180
181 my ($class) = shift (@_);
182 my ($pluginlist,$inputargs,$hashArgOptLists,$auxiliary) = @_;
183 push(@$pluginlist, $class);
184
185 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
186 push(@{$hashArgOptLists->{"OptList"}},$options);
187
188 my $self = new PrintInfo($pluginlist, $inputargs, $hashArgOptLists,$auxiliary);
189
190 if ($self->{'info_only'}) {
191 # don't worry about any options etc
192 return bless $self, $class;
193 }
194
195 if ($self->{'smart_block'}) {
196 print STDERR "WARNING: -smart_block option has been deprecated and is no longer useful\n";
197 }
198 $self->{'smart_block'} = undef;
199
200 my $plugin_name = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
201 $self->{'plugin_type'} = $plugin_name;
202
203 # remove ex. from OIDmetadata
204 $self->{'OIDmetadata'} =~ s/^ex\.// if defined $self->{'OIDmetadata'};
205 $self->{'num_processed'} = 0;
206 $self->{'num_not_processed'} = 0;
207 $self->{'num_blocked'} = 0;
208 $self->{'num_archives'} = 0;
209 $self->{'cover_image'} = 1; # cover image is on by default
210 $self->{'cover_image'} = 0 if ($self->{'no_cover_image'});
211 $self->{'can_process_directories'} = 0;
212 #$self->{'option_list'} = $hashArgOptLists->{"OptList"};
213
214 my $associate_ext = $self->{'associate_ext'};
215 if ((defined $associate_ext) && ($associate_ext ne "")) {
216
217 my $associate_tail_re = $self->{'associate_tail_re'};
218 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
219 my $outhandle = $self->{'outhandle'};
220 print $outhandle "Warning: can only specify 'associate_ext' or 'associate_tail_re'\n";
221 print $outhandle " defaulting to 'associate_tail_re'\n";
222 }
223 else {
224 my @exts = split(/,/,$associate_ext);
225
226 my @exts_bracketed = map { $_ = "(?:\\.$_)" } @exts;
227 my $associate_tail_re = join("|",@exts_bracketed);
228 $self->{'associate_tail_re'} = $associate_tail_re;
229 }
230
231 delete $self->{'associate_ext'};
232 }
233
234 return bless $self, $class;
235
236}
237
238sub merge_inheritance
239{
240 my $self = {};
241 my @child_selfs = @_;
242
243 foreach my $child_self (@child_selfs) {
244 foreach my $key (keys %$child_self) {
245 if (defined $self->{$key}) {
246 if ($self->{$key} ne $child_self->{$key}) {
247# print STDERR "Warning: Conflicting value in multiple inheritance for '$key'\n";
248# print STDERR "Existing stored value = $self->{$key}\n";
249# print STDERR "New (child) value = $child_self->{$key}\n";
250# print STDERR "Keeping existing value\n";
251 # Existing value seems to be option specified in collect.cfg
252
253 ### $self->{$key} = $child_self->{$key};
254
255 }
256 else {
257## print STDERR "****Info: Value $self->{$key} for $key already defined through multiple inheritance as the same value\n";
258 }
259
260 }
261 else {
262 $self->{$key} = $child_self->{$key};
263 }
264 }
265 }
266
267 return $self;
268}
269
270# initialize BasePlugin options
271# if init() is overridden in a sub-class, remember to call BasePlugin::init()
272sub init {
273 my $self = shift (@_);
274 my ($verbosity, $outhandle, $failhandle) = @_;
275
276 # verbosity is passed through from the processor
277 $self->{'verbosity'} = $verbosity;
278
279 # as are the outhandle and failhandle
280 $self->{'outhandle'} = $outhandle if defined $outhandle;
281 $self->{'failhandle'} = $failhandle;
282# $self->SUPER::init(@_);
283
284 # set process_exp and block_exp to defaults unless they were
285 # explicitly set
286
287 if ((!$self->is_recursive()) and
288 (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
289
290 $self->{'process_exp'} = $self->get_default_process_exp ();
291 if ($self->{'process_exp'} eq "") {
292 warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
293 }
294 }
295
296 if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
297 $self->{'block_exp'} = $self->get_default_block_exp ();
298 }
299
300}
301
302sub begin {
303 my $self = shift (@_);
304 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
305
306 if ($self->{'OIDtype'} eq "auto") {
307 # hasn't been set in the plugin, use the processor values
308 $self->{'OIDtype'} = $processor->{'OIDtype'};
309 $self->{'OIDmetadata'} = $processor->{'OIDmetadata'};
310 }
311 if ($self->{'OIDtype'} eq "hash") {
312 # should we hash on the file or on the doc xml??
313 $self->{'OIDtype'} = $self->get_oid_hash_type();
314 if ($self->{'OIDtype'} !~ /^(hash_on_file|hash_on_ga_xml)$/) {
315 $self->{'OIDtype'} = "hash_on_file";
316 }
317 }
318}
319
320# This is called once if removeold is set with import.pl. Most plugins will do
321# nothing but if a plugin does any stuff outside of creating doc obj, then
322# it may need to clear something.
323sub remove_all {
324 my $self = shift (@_);
325 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
326}
327
328# This is called per document for docs that have been deleted from the
329# collection. Most plugins will do nothing
330# but if a plugin does any stuff outside of creating doc obj, then it may need
331# to clear something.
332sub remove_one {
333 my $self = shift (@_);
334
335 my ($file, $oids, $archivedir) = @_;
336 return 0 if $self->can_process_this_file($file);
337 return undef;
338}
339
340sub end {
341 # potentially called at the end of each plugin pass
342 # import.pl only has one plugin pass, but buildcol.pl has multiple ones
343
344 my ($self) = shift (@_);
345}
346
347sub deinit {
348 # called only once, after all plugin passes have been done
349
350 my ($self) = @_;
351}
352
353# default hashing type is to hash on the original file (or converted file)
354# override this to return hash_on_ga_xml for filetypes where hashing on the
355# file is no good eg video
356sub get_oid_hash_type {
357
358 my $self = shift (@_);
359
360 return "hash_on_file";
361}
362
363
364# this function should be overridden to return 1
365# in recursive plugins
366sub is_recursive {
367 my $self = shift (@_);
368
369 return 0;
370}
371
372sub get_default_block_exp {
373 my $self = shift (@_);
374
375 return "";
376}
377
378sub get_default_process_exp {
379 my $self = shift (@_);
380
381 return "";
382}
383
384# rename imported files and assoc files using URL encoding by default
385# as this will work for most plugins and give more legible filenames
386sub get_default_file_rename_method() {
387 my $self = shift (@_);
388 return "url";
389}
390
391# returns this plugin's active (possibly user-selected) file_rename_method
392sub get_file_rename_method() {
393 my $self = shift (@_);
394 my $rename_method = $self->{'file_rename_method'};
395 if($rename_method) {
396 return $rename_method;
397 } else {
398 return $self->get_default_file_rename_method();
399 }
400}
401
402# default implementation is to do nothing
403sub store_block_files {
404
405 my $self =shift (@_);
406 my ($filename_full_path, $block_hash) = @_;
407
408}
409
410# put files to block into hash
411sub use_block_expressions {
412
413 my $self =shift (@_);
414 my ($filename_full_path, $block_hash) = @_;
415
416 $filename_full_path = &util::upgrade_if_dos_filename($filename_full_path);
417
418 if ($self->{'block_exp'} ne "" && $filename_full_path =~ /$self->{'block_exp'}/) {
419 $block_hash->{'file_blocks'}->{$filename_full_path} = 1;
420 }
421
422}
423
424#default implementation is to block a file with same name as this, but extension jpg or JPG, if cover_images is on.
425sub block_cover_image
426{
427 my $self =shift;
428 my ($filename, $block_hash) = @_;
429
430 $filename = &util::upgrade_if_dos_filename($filename);
431
432 if ($self->{'cover_image'}) {
433 my $coverfile = $filename;
434 $coverfile =~ s/\.[^\\\/\.]+$/\.jpg/;
435 if (!&util::fd_exists($coverfile)) {
436 $coverfile =~ s/jpg$/JPG/;
437 }
438 if (&util::fd_exists($coverfile)) {
439 $block_hash->{'file_blocks'}->{$coverfile} = 1;
440 }
441 }
442
443 return;
444}
445
446
447# discover all the files that should be blocked by this plugin
448# check the args ...
449sub file_block_read {
450
451 my $self = shift (@_);
452 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $gli) = @_;
453 # Keep track of filenames with same root but different extensions
454 # Used to support -associate_ext and the more generalised
455 # -associate_tail_re
456 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
457
458 if (!-d $filename_full_path) {
459 $block_hash->{'all_files'}->{$file} = 1;
460 }
461
462 my $associate_tail_re = $self->{'associate_tail_re'};
463 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
464 my ($file_prefix,$file_ext)
465 = &util::get_prefix_and_tail_by_regex($filename_full_path,$associate_tail_re);
466 if ((defined $file_prefix) && (defined $file_ext)) {
467 my $shared_fileroot = $block_hash->{'shared_fileroot'};
468 if (!defined $shared_fileroot->{$file_prefix}) {
469 my $file_prefix_rec = { 'tie_to' => undef,
470 'exts' => {} };
471 $shared_fileroot->{$file_prefix} = $file_prefix_rec;
472 }
473
474 my $file_prefix_rec = $shared_fileroot->{$file_prefix};
475
476 if ($self->can_process_this_file($filename_full_path)) {
477 # This is the document the others should be tied to
478 $file_prefix_rec->{'tie_to'} = $file_ext;
479 }
480 else {
481 if ($file_ext =~ m/$associate_tail_re$/) {
482 # this file should be associated to the main one
483 $file_prefix_rec->{'exts'}->{$file_ext} = 1;
484 }
485 }
486
487 }
488 }
489
490 # check block expressions
491 $self->use_block_expressions($filename_full_path, $block_hash) unless $self->{'no_blocking'};
492
493 # now check whether we are actually processing this
494 if (!-f $filename_full_path || !$self->can_process_this_file($filename_full_path)) {
495 return undef; # can't recognise
496 }
497
498 # if we have a block_exp, then this overrides the normal 'smart' blocking
499 $self->store_block_files($filename_full_path, $block_hash) unless ($self->{'no_blocking'} || $self->{'block_exp'} ne "");
500
501 # block the cover image if there is one
502 if ($self->{'cover_image'}) {
503 $self->block_cover_image($filename_full_path, $block_hash);
504 }
505
506 return 1;
507}
508
509# plugins that rely on more than process_exp (eg XML plugins) can override this method
510sub can_process_this_file {
511 my $self = shift(@_);
512 my ($filename) = @_;
513
514 if (-d $filename && !$self->{'can_process_directories'}) {
515 return 0;
516 }
517
518 if ($self->{'process_exp'} ne "" && $filename =~ /$self->{'process_exp'}/) {
519 return 1;
520 }
521 return 0;
522
523}
524
525# just converts path as is to utf8.
526sub filepath_to_utf8 {
527 my $self = shift (@_);
528 my ($file, $file_encoding) = @_;
529 my $filemeta = $file;
530
531 my $filename_encoding = $self->{'filename_encoding'}; # filename encoding setting
532
533 # Whenever filename-encoding is set to any of the auto settings, we
534 # check if the filename is already in UTF8. If it is, then we're done.
535 if($filename_encoding =~ m/auto/) {
536 if(&unicode::check_is_utf8($filemeta))
537 {
538 $filename_encoding = "utf8";
539 return $filemeta;
540 }
541 }
542
543 # Auto setting, but filename is not utf8
544 if ($filename_encoding eq "auto")
545 {
546 # try textcat
547 $filename_encoding = $self->textcat_encoding($filemeta);
548
549 # check the locale next
550 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
551
552
553 # now try the encoding of the document, if available
554 if ($filename_encoding eq "undefined" && defined $file_encoding) {
555 $filename_encoding = $file_encoding;
556 }
557
558 }
559
560 elsif ($filename_encoding eq "auto-language-analysis")
561 {
562 $filename_encoding = $self->textcat_encoding($filemeta);
563
564 # now try the encoding of the document, if available
565 if ($filename_encoding eq "undefined" && defined $file_encoding) {
566 $filename_encoding = $file_encoding;
567 }
568 }
569
570 elsif ($filename_encoding eq "auto-filesystem-encoding")
571 {
572 # try locale
573 $filename_encoding = $self->locale_encoding();
574 }
575
576 elsif ($filename_encoding eq "auto-fl")
577 {
578 # filesystem-encoding (locale) then language-analysis (textcat)
579 $filename_encoding = $self->locale_encoding();
580
581 # try textcat
582 $filename_encoding = $self->textcat_encoding($filemeta) if $filename_encoding eq "undefined";
583
584 # else assume filename encoding is encoding of file content, if that's available
585 if ($filename_encoding eq "undefined" && defined $file_encoding) {
586 $filename_encoding = $file_encoding;
587 }
588 }
589
590 elsif ($filename_encoding eq "auto-lf")
591 {
592 # language-analysis (textcat) then filesystem-encoding (locale)
593 $filename_encoding = $self->textcat_encoding($filemeta);
594
595 # guess filename encoding from encoding of file content, if available
596 if ($filename_encoding eq "undefined" && defined $file_encoding) {
597 $filename_encoding = $file_encoding;
598 }
599
600 # try locale
601 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
602 }
603
604 # if still undefined, use utf8 as fallback
605 if ($filename_encoding eq "undefined") {
606 $filename_encoding = "utf8";
607 }
608
609 #print STDERR "**** UTF8 encoding the filename $filemeta ";
610
611 # if the filename encoding is set to utf8 but it isn't utf8 already--such as when
612 # 1. the utf8 fallback is used, or 2. if the system locale is used and happens to
613 # be always utf8 (in which case the filename's encoding is also set as utf8 even
614 # though the filename need not be if it originates from another system)--in such
615 # cases attempt to make the filename utf8 to match.
616 if($filename_encoding eq "utf8" && !&unicode::check_is_utf8($filemeta)) {
617 &unicode::ensure_utf8(\$filemeta);
618 }
619
620 # convert non-unicode encodings to utf8
621 if ($filename_encoding !~ m/(?:ascii|utf8|unicode)/) {
622 $filemeta = &unicode::unicode2utf8(
623 &unicode::convert2unicode($filename_encoding, \$filemeta)
624 );
625 }
626
627 #print STDERR " from encoding $filename_encoding -> $filemeta\n";
628 return $filemeta;
629}
630
631# gets the filename with no path, converts to utf8, and then dm safes it.
632# filename_encoding set by user
633sub filename_to_utf8_metadata
634{
635 my $self = shift (@_);
636 my ($file, $file_encoding) = @_;
637
638 my $outhandle = $self->{'outhandle'};
639
640 print $outhandle "****!!!!**** BasePlugin::filename_to_utf8_metadata now deprecated\n";
641 my ($cpackage,$cfilename,$cline,$csubr,$chas_args,$cwantarray) = caller(0);
642 print $outhandle "Calling method: $cfilename:$cline $cpackage->$csubr\n";
643
644
645 my ($filemeta) = $file =~ /([^\\\/]+)$/; # getting the tail of the filepath (skips all string parts containing slashes upto the end)
646 $filemeta = $self->filepath_to_utf8($filemeta, $file_encoding);
647
648 return $filemeta;
649}
650
651sub locale_encoding {
652 my $self = shift(@_);
653
654 if (!defined $self->{'filesystem_encoding'}) {
655 $self->{'filesystem_encoding'} = $self->get_filesystem_encoding();
656 $self->{'filesystem_encoding'} = "undefined" if !defined $self->{'filesystem_encoding'};
657 }
658
659 #print STDERR "*** filename encoding determined based on locale: " . $self->{'filesystem_encoding'} . "\n";
660 return $self->{'filesystem_encoding'}; # can be the string "undefined"
661}
662
663sub textcat_encoding {
664 my $self = shift(@_);
665 my ($filemeta) = @_;
666
667 # analyse filenames without extensions and digits (and trimmed of
668 # surrounding whitespace), so that irrelevant chars don't confuse
669 # textcat
670 my $strictfilemeta = $filemeta;
671 $strictfilemeta =~ s/\.[^\.]+$//g;
672 $strictfilemeta =~ s/\d//g;
673 $strictfilemeta =~ s/^\s*//g;
674 $strictfilemeta =~ s/\s*$//g;
675
676 my $filename_encoding = $self->encoding_from_language_analysis($strictfilemeta);
677 if(!defined $filename_encoding) {
678 $filename_encoding = "undefined";
679 }
680
681 return $filename_encoding; # can be the string "undefined"
682}
683
684# performs textcat
685sub encoding_from_language_analysis {
686 my $self = shift(@_);
687 my ($text) = @_;
688
689 my $outhandle = $self->{'outhandle'};
690 my $best_encoding = undef;
691
692 # get the language/encoding of the textstring using textcat
693 require textcat; # Only load the textcat module if it is required
694 $self->{'textcat'} = new textcat() unless defined($self->{'textcat'});
695 my $results = $self->{'textcat'}->classify_cached_filename(\$text);
696
697
698 if (scalar @$results < 0) {
699 return undef;
700 }
701
702 # We have some results, we choose the first
703 my ($language, $encoding) = $results->[0] =~ /^([^-]*)(?:-(.*))?$/;
704
705 $best_encoding = $encoding;
706 if (!defined $best_encoding) {
707 return undef;
708 }
709
710 if (defined $best_encoding && $best_encoding =~ m/^iso_8859/ && &unicode::check_is_utf8($text)) {
711 # the text is valid utf8, so assume that's the real encoding (since textcat is based on probabilities)
712 $best_encoding = 'utf8';
713 }
714
715
716 # check for equivalents where textcat doesn't have some encodings...
717 # eg MS versions of standard encodings
718 if (defined $best_encoding && $best_encoding =~ /^iso_8859_(\d+)/) {
719 my $iso = $1; # which variant of the iso standard?
720 # iso-8859 sets don't use chars 0x80-0x9f, windows codepages do
721 if ($text =~ /[\x80-\x9f]/) {
722 # Western Europe
723 if ($iso == 1 or $iso == 15) { $best_encoding = 'windows_1252' }
724 elsif ($iso == 2) {$best_encoding = 'windows_1250'} # Central Europe
725 elsif ($iso == 5) {$best_encoding = 'windows_1251'} # Cyrillic
726 elsif ($iso == 6) {$best_encoding = 'windows_1256'} # Arabic
727 elsif ($iso == 7) {$best_encoding = 'windows_1253'} # Greek
728 elsif ($iso == 8) {$best_encoding = 'windows_1255'} # Hebrew
729 elsif ($iso == 9) {$best_encoding = 'windows_1254'} # Turkish
730 }
731 }
732
733 if (defined $best_encoding && $best_encoding !~ /^(ascii|utf8|unicode)$/ &&
734 !defined $encodings::encodings->{$best_encoding})
735 {
736 if ($self->{'verbosity'}) {
737 gsprintf($outhandle, "BasePlugin: {ReadTextFile.unsupported_encoding}\n", $text, $best_encoding, "undef");
738 }
739 $best_encoding = undef;
740 }
741
742 return $best_encoding;
743}
744
745# uses locale
746sub get_filesystem_encoding
747{
748
749 my $self = shift(@_);
750
751 my $outhandle = $self->{'outhandle'};
752 my $filesystem_encoding = undef;
753
754 eval {
755 # Works for Windows as well, returning the DOS code page in use
756 use POSIX qw(locale_h);
757
758 # With only one parameter, setlocale retrieves the
759 # current value
760 my $current_locale = setlocale(LC_CTYPE);
761
762 if ($current_locale =~ m/^.*\.(.*?)$/) {
763 my $char_encoding = lc($1);
764 if ($char_encoding =~ m/^(iso)(8859)(\d{1,2})$/) {
765 $char_encoding = "$1\_$2\_$3";
766 }
767
768 $char_encoding =~ s/-/_/g;
769 $char_encoding =~ s/^utf_8$/utf8/;
770
771 if ($char_encoding =~ m/^\d+$/) {
772 if (defined $encodings::encodings->{"windows_$char_encoding"}) {
773 $char_encoding = "windows_$char_encoding";
774 }
775 elsif (defined $encodings::encodings->{"dos_$char_encoding"}) {
776 $char_encoding = "dos_$char_encoding";
777 }
778 }
779
780 if (($char_encoding =~ m/(?:ascii|utf8|unicode)/)
781 || (defined $encodings::encodings->{$char_encoding})) {
782 $filesystem_encoding = $char_encoding;
783 }
784 else {
785 print $outhandle "Warning: Unsupported character encoding '$char_encoding' from locale '$current_locale'\n";
786 }
787 }
788
789
790 };
791 if ($@) {
792 print $outhandle "$@\n";
793 print $outhandle "Warning: Unable to establish locale. Will assume filesystem is UTF-8\n";
794
795 }
796
797 return $filesystem_encoding;
798}
799
800
801
802sub deduce_filename_encoding
803{
804 my $self = shift (@_);
805 my ($file,$metadata,$plugin_filename_encoding) = @_;
806
807 my $gs_filename_encoding = $metadata->{"gs.filename_encoding"};
808 my $deduced_filename_encoding = undef;
809
810 # Start by looking for manually assigned metadata
811 if (defined $gs_filename_encoding) {
812 if (ref ($gs_filename_encoding) eq "ARRAY") {
813 my $outhandle = $self->{'outhandle'};
814
815 $deduced_filename_encoding = $gs_filename_encoding->[0];
816
817 my $num_vals = scalar(@$gs_filename_encoding);
818 if ($num_vals>1) {
819 print $outhandle "Warning: gs.filename_encoding multiply defined for $file\n";
820 print $outhandle " Selecting first value: $deduced_filename_encoding\n";
821 }
822 }
823 else {
824 $deduced_filename_encoding = $gs_filename_encoding;
825 }
826 }
827
828 if (!defined $deduced_filename_encoding || ($deduced_filename_encoding =~ m/^\s*$/)) {
829 # Look to see if plugin specifies this value
830
831 if (defined $plugin_filename_encoding) {
832 # First look to see if we're using any of the "older" (i.e. deprecated auto-... plugin options)
833 if ($plugin_filename_encoding =~ m/^auto-.*$/) {
834 my $outhandle = $self->{'outhandle'};
835 print $outhandle "Warning: $plugin_filename_encoding is no longer supported\n";
836 print $outhandle " default to 'auto'\n";
837 $self->{'filename_encoding'} = $plugin_filename_encoding = "auto";
838 }
839
840 if ($plugin_filename_encoding ne "auto") {
841 # We've been given a specific filenamne encoding
842 # => so use it!
843 $deduced_filename_encoding = $plugin_filename_encoding;
844 }
845 }
846 }
847
848 if (!defined $deduced_filename_encoding || ($deduced_filename_encoding =~ m/^\s*$/)) {
849
850 # Look to file system to provide a character encoding
851
852 # If Windows NTFS, then -- assuming we work with long file names got through
853 # Win32::GetLongFilePath() -- then the underlying file system is UTF16
854
855 if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
856 # Can do better than working with the DOS character encoding returned by locale
857 $deduced_filename_encoding = "unicode";
858 }
859 else {
860 # Unix of some form or other
861
862 # See if we can determine the file system encoding through locale
863 $deduced_filename_encoding = $self->locale_encoding();
864
865 # if locale shows us filesystem is utf8, check to see filename is consistent
866 # => if not, then we have an "alien" filename on our hands
867
868 if ($deduced_filename_encoding =~ m/^utf-?8$/i) {
869 if (!&unicode::check_is_utf8($file)) {
870 # "alien" filename, so revert
871 $deduced_filename_encoding = undef;
872 }
873 }
874 }
875 }
876
877# if (!defined $deduced_filename_encoding || ($deduced_filename_encoding =~ m/^\s*$/)) {
878# # Last chance, apply textcat to deduce filename encoding
879# $deduced_filename_encoding = $self->textcat_encoding($file);
880# }
881
882 if ($self->{'verbosity'}>3) {
883 my $outhandle = $self->{'outhandle'};
884
885 if (defined $deduced_filename_encoding) {
886 print $outhandle " Deduced filename encoding as: $deduced_filename_encoding\n";
887 }
888 else {
889 print $outhandle " No filename encoding deduced\n";
890 }
891 }
892
893 return $deduced_filename_encoding;
894}
895
896
897
898
899# Notionally written to be called once for each document, it is however safe to
900# call multiple times (as in the case of ImagePlugin) which calls this later on
901# after the original image has potentially been converted to a *new* source image
902# format (e.g. TIFF to PNG)
903
904sub set_Source_metadata {
905 my $self = shift (@_);
906 my ($doc_obj, $raw_filename, $filename_encoding) = @_;
907
908 # 1. Sets the filename (Source) for display encoded as Unicode if possible,
909 # and (as a fallback) using %xx if not for non-ascii chars
910 # 2. Sets the url ref (SourceFile) to the URL encoded version
911 # of filename for generated files
912
913 my ($unused_full_rf, $raw_file) = &util::get_full_filenames("", $raw_filename);
914
915 my $top_section = $doc_obj->get_top_section();
916
917 my $octet_file = $raw_file;
918
919 # UTF-8 version of filename
920 if ((defined $ENV{"DEBUG_UNICODE"}) && ($ENV{"DEBUG_UNICODE"})) {
921 print STDERR "**** Setting Source Metadata given: $octet_file\n";
922 }
923
924 # Deal with (on Windows) raw filenames that are in their
925 # abbreviated DOS form
926
927 if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
928 if ((defined $filename_encoding) && ($filename_encoding eq "unicode")) {
929 if (-e $raw_filename) {
930 require Win32;
931
932## print STDERR "**** raw filename before LPN: $raw_filename\n";
933 my $unicode_filename = Win32::GetLongPathName($raw_filename);
934
935 my $unused_full_uf;
936 ($unused_full_uf, $octet_file) = &util::get_full_filenames("", $unicode_filename);
937
938## print STDERR "**** raw filename after LPN: $raw_filename\n";
939 }
940 }
941 }
942
943 my $url_encoded_filename;
944 if (defined $filename_encoding) {
945 # => Generate a pretty print version of filename that is mapped to Unicode
946
947 # Use filename_encoding to map raw filename to a Perl unicode-aware string
948 $url_encoded_filename = decode($filename_encoding,$octet_file);
949 }
950 else {
951 # otherwise generate %xx encoded version of filename for char > 127
952 $url_encoded_filename = &unicode::raw_filename_to_url_encoded($octet_file);
953 }
954
955 if ((defined $ENV{"DEBUG_UNICODE"}) && ($ENV{"DEBUG_UNICODE"})) {
956 print STDERR "****** saving Source as: $url_encoded_filename\n";
957 }
958
959
960 # Source is the UTF8 display name - not necessarily the name of the file on the system
961 $doc_obj->set_utf8_metadata_element($top_section, "Source", $url_encoded_filename);
962
963 my $renamed_raw_file = &util::rename_file($raw_file, $self->{'file_rename_method'});
964 # If using URL encoding, then SourceFile is the url-reference to url-encoded
965 # renamed_raw_url: it's a url that refers to the actual file on the system
966 my $renamed_raw_url = &unicode::filename_to_url($renamed_raw_file);
967
968 $doc_obj->set_utf8_metadata_element($top_section, "SourceFile",
969 $renamed_raw_url);
970
971 if ((defined $ENV{"DEBUG_UNICODE"}) && ($ENV{"DEBUG_UNICODE"})) {
972 print STDERR "****** saving SourceFile as: $renamed_raw_url\n";
973 }
974}
975
976# this should be called by all plugins to set the oid of the doc obj, rather
977# than calling doc_obj->set_OID directly
978sub add_OID {
979 my $self = shift (@_);
980 my ($doc_obj) = @_;
981
982 $doc_obj->set_OIDtype($self->{'OIDtype'}, $self->{'OIDmetadata'});
983
984 # see if there is a plugin specific set_OID function
985 if (defined ($self->can('set_OID'))) {
986 $self->set_OID(@_); # pass through doc_obj and any extra arguments
987 }
988 else {
989 # use the default set_OID() in doc.pm
990 $doc_obj->set_OID();
991 }
992
993}
994
995# The BasePlugin read_into_doc_obj() function. This function does all the
996# right things to make general options work for a given plugin. It doesn't do anything with the file other than setting reads in
997# a file and sets up a slew of metadata all saved in doc_obj, which
998# it then returns as part of a tuple (process_status,doc_obj)
999#
1000# Much of this functionality used to reside in read, but it was broken
1001# down into a supporting routine to make the code more flexible.
1002#
1003# recursive plugins (e.g. RecPlug) and specialized plugins like those
1004# capable of processing many documents within a single file (e.g.
1005# GMLPlug) will normally want to implement their own version of
1006# read_into_doc_obj()
1007#
1008# Note that $base_dir might be "" and that $file might
1009# include directories
1010
1011# currently blocking has been done before it gets here - does this affect secondary plugin stuff??
1012sub read_into_doc_obj {
1013 my $self = shift (@_);
1014 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
1015
1016 my $outhandle = $self->{'outhandle'};
1017
1018 # should we move this to read? What about secondary plugins?
1019 print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
1020 my $pp_file = &util::prettyprint_file($base_dir,$file);
1021 print $outhandle "$self->{'plugin_type'} processing $pp_file\n"
1022 if $self->{'verbosity'} > 1;
1023
1024 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
1025
1026 # create a new document
1027 my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
1028 my $top_section = $doc_obj->get_top_section();
1029
1030 $doc_obj->add_utf8_metadata($top_section, "Plugin", "$self->{'plugin_type'}");
1031 $doc_obj->add_utf8_metadata($top_section, "FileSize", (-s $filename_full_path));
1032
1033
1034 my $plugin_filename_encoding = $self->{'filename_encoding'};
1035 my $filename_encoding = $self->deduce_filename_encoding($file,$metadata,$plugin_filename_encoding);
1036 $self->set_Source_metadata($doc_obj,$filename_full_path,$filename_encoding,$filename_full_path);
1037
1038 # plugin specific stuff - what args do we need here??
1039 unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
1040 print STDERR "<ProcessingError n='$file'>\n" if ($gli);
1041 return -1;
1042 }
1043
1044 # include any metadata passed in from previous plugins
1045 # note that this metadata is associated with the top level section
1046 my $section = $doc_obj->get_top_section();
1047 # can we merge these two methods??
1048 $self->add_associated_files($doc_obj, $filename_full_path);
1049 $self->extra_metadata ($doc_obj, $section, $metadata);
1050 $self->auto_extract_metadata($doc_obj);
1051
1052 # if we haven't found any Title so far, assign one
1053 # this was shifted to here from inside read()
1054 $self->title_fallback($doc_obj,$section,$filename_no_path);
1055
1056 $self->add_OID($doc_obj);
1057
1058 $self->post_process_doc_obj($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli);
1059 return (1,$doc_obj);
1060}
1061
1062sub post_process_doc_obj {
1063 my $self = shift (@_);
1064 my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
1065
1066 return 1;
1067}
1068
1069sub add_dummy_text {
1070 my $self = shift(@_);
1071 my ($doc_obj, $section) = @_;
1072
1073 # add NoText metadata so we can hide this dummy text in format statements
1074 $doc_obj->add_metadata($section, "NoText", "1");
1075 $doc_obj->add_text($section, &gsprintf::lookup_string("{BasePlugin.dummy_text}",1));
1076
1077}
1078
1079# does nothing. Can be overridden by subclass
1080sub auto_extract_metadata {
1081 my $self = shift(@_);
1082 my ($doc_obj) = @_;
1083}
1084
1085# adds cover image, associate_file options stuff. Should be called by sub class
1086# read_into_doc_obj
1087sub add_associated_files {
1088 my $self = shift(@_);
1089 # whatis filename??
1090 my ($doc_obj, $filename) = @_;
1091
1092 # add in the cover image
1093 if ($self->{'cover_image'}) {
1094 $self->associate_cover_image($doc_obj, $filename);
1095 }
1096 # store the original (used for eg TextPlugin to store the original for OAI)
1097 if ($self->{'store_original_file'}) {
1098 $self->associate_source_file($doc_obj, $filename);
1099 }
1100
1101
1102}
1103
1104# implement this if you are extracting metadata for other documents
1105sub metadata_read {
1106 my $self = shift (@_);
1107 my ($pluginfo, $base_dir, $file, $block_hash,
1108 $extrametakeys, $extrametadata, $extrametafile,
1109 $processor, $gli, $aux) = @_;
1110
1111 # can we process this file??
1112 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
1113 return undef unless $self->can_process_this_file($filename_full_path);
1114
1115 return 1; # we recognise the file, but don't actually do anything with it
1116}
1117
1118
1119# The BasePlugin read() function. This function calls read_into_doc_obj()
1120# to ensure all the right things to make general options work for a
1121# given plugin are done. It then calls the process() function which
1122# does all the work specific to a plugin (like the old read functions
1123# used to do). Most plugins should define their own process() function
1124# and let this read() function keep control.
1125#
1126# recursive plugins (e.g. RecPlug) and specialized plugins like those
1127# capable of processing many documents within a single file (e.g.
1128# GMLPlug) might want to implement their own version of read(), but
1129# more likely need to implement their own version of read_into_doc_obj()
1130#
1131# Return number of files processed, undef if can't recognise, -1 if can't
1132# process
1133
1134sub read {
1135 my $self = shift (@_);
1136 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
1137
1138 # can we process this file??
1139 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
1140
1141 return undef unless $self->can_process_this_file($filename_full_path);
1142
1143 my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
1144
1145 if ((defined $process_status) && ($process_status == 1)) {
1146
1147 # process the document
1148 $processor->process($doc_obj);
1149
1150 $self->{'num_processed'} ++;
1151 undef $doc_obj;
1152 }
1153 # delete any temp files that we may have created
1154 $self->clean_up_after_doc_obj_processing();
1155
1156
1157 # if process_status == 1, then the file has been processed.
1158 return $process_status;
1159
1160}
1161
1162# returns undef if file is rejected by the plugin
1163sub process {
1164 my $self = shift (@_);
1165 my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
1166
1167 gsprintf(STDERR, "BasePlugin::process {common.must_be_implemented}\n") && die "\n";
1168
1169 return undef; # never gets here
1170}
1171
1172# overwrite this method to delete any temp files that we have created
1173sub clean_up_after_doc_obj_processing {
1174 my $self = shift(@_);
1175
1176}
1177
1178# write_file -- used by ConvertToPlug, for example in post processing
1179#
1180# where should this go, is here the best place??
1181sub utf8_write_file {
1182 my $self = shift (@_);
1183 my ($textref, $filename) = @_;
1184
1185 if (!open (FILE, ">:utf8", $filename)) {
1186 gsprintf(STDERR, "ConvertToPlug::write_file {ConvertToPlug.could_not_open_for_writing} ($!)\n", $filename);
1187 die "\n";
1188 }
1189 print FILE $$textref;
1190
1191 close FILE;
1192}
1193
1194
1195sub filename_based_title
1196{
1197 my $self = shift (@_);
1198 my ($file) = @_;
1199
1200 my $file_derived_title = $file;
1201 $file_derived_title =~ s/_/ /g;
1202 $file_derived_title =~ s/\.[^.]+$//;
1203
1204 return $file_derived_title;
1205}
1206
1207
1208sub title_fallback
1209{
1210 my $self = shift (@_);
1211 my ($doc_obj,$section,$file) = @_;
1212
1213 if (!defined $doc_obj->get_metadata_element ($section, "Title")
1214 || $doc_obj->get_metadata_element($section, "Title") eq "") {
1215
1216 my $source_file = $doc_obj->get_metadata_element($section, "Source");
1217 my $file_derived_title;
1218 if (defined $source_file) {
1219 $file_derived_title = $self->filename_based_title($source_file);
1220 }
1221 else {
1222 # pp = pretty print
1223 my $pp_file = (defined $source_file) ? $source_file : $file;
1224
1225 my $raw_title = $self->filename_based_title($file);
1226 my $file_derived_title = &unicode::raw_filename_to_url_encoded($raw_title);
1227 }
1228
1229
1230 if (!defined $doc_obj->get_metadata_element ($section, "Title")) {
1231 $doc_obj->add_utf8_metadata ($section, "Title", $file_derived_title);
1232 }
1233 else {
1234 $doc_obj->set_utf8_metadata_element ($section, "Title", $file_derived_title);
1235 }
1236 }
1237
1238}
1239
1240# add any extra metadata that's been passed around from one
1241# plugin to another.
1242# extra_metadata uses add_utf8_metadata so it expects metadata values
1243# to already be in utf8
1244sub extra_metadata {
1245 my $self = shift (@_);
1246 my ($doc_obj, $cursection, $metadata) = @_;
1247
1248 my $associate_tail_re = $self->{'associate_tail_re'};
1249
1250 foreach my $field (keys(%$metadata)) {
1251 # $metadata->{$field} may be an array reference
1252 if ($field eq "gsdlassocfile_tobe") {
1253 # 'gsdlassocfile_tobe' is artificially introduced metadata
1254 # that is used to signal that certain additional files should
1255 # be tied to this document. Useful in situations where a
1256 # metadata pass in the plugin pipeline works out some files
1257 # need to be associated with a document, but the document hasn't
1258 # been formed yet.
1259 my $equiv_form = "";
1260 foreach my $gaf (@{$metadata->{$field}}) {
1261 my ($full_filename,$mimetype) = ($gaf =~ m/^(.*):(.*):$/);
1262 my ($tail_filename) = ($full_filename =~ /^.*[\/\\](.+?)$/);
1263
1264 # we need to make sure the filename is valid utf-8 - we do
1265 # this by url or base64 encoding it
1266 # $tail_filename is the name that we store the file as
1267 $tail_filename = &util::rename_file($tail_filename, $self->{'file_rename_method'});
1268 $doc_obj->associate_file($full_filename,$tail_filename,$mimetype);
1269 $doc_obj->associate_source_file($full_filename);
1270 # If the filename is url_encoded, we need to encode the % signs
1271 # in the filename, so that it works in a url
1272 my $url_tail_filename = &unicode::filename_to_url($tail_filename);
1273 # work out extended tail extension (i.e. matching tail re)
1274
1275 my ($file_prefix,$file_extended_ext)
1276 = &util::get_prefix_and_tail_by_regex($tail_filename,$associate_tail_re);
1277 my ($pre_doc_ext) = ($file_extended_ext =~ m/^(.*)\..*$/);
1278 my ($doc_ext) = ($tail_filename =~ m/^.*\.(.*)$/);
1279 my $start_doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/{Or}{[parent(Top):assocfilepath],[assocfilepath]}/$url_tail_filename\">";
1280 my $srcicon = "_icon".$doc_ext."_";
1281 my $end_doclink = "</a>";
1282
1283 my $assoc_form = "$start_doclink\{If\}{$srcicon,$srcicon,$doc_ext\}$end_doclink";
1284
1285 if (defined $pre_doc_ext && $pre_doc_ext ne "") {
1286 # for metadata such as [mp3._edited] [mp3._full] ...
1287 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.$pre_doc_ext", $assoc_form);
1288 }
1289
1290 # for multiple metadata such as [mp3.assoclink]
1291 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.assoclink", $assoc_form);
1292
1293 $equiv_form .= " $assoc_form";
1294 }
1295 $doc_obj->add_utf8_metadata ($cursection, "equivlink", $equiv_form);
1296 }
1297 elsif ($field eq "gsdlzipfilename") {
1298 # special case for when files have come out of a zip. source_path
1299 # (used for archives dbs and keeping track for incremental import)
1300 # must be set to the zip file name
1301 my $zip_filename = $metadata->{$field};
1302 # overwrite the source_path
1303 $doc_obj->set_source_path($zip_filename);
1304 # and set the metadata
1305 $zip_filename = &util::filename_within_collection($zip_filename);
1306 $zip_filename = $doc_obj->encode_filename($zip_filename, $self->{'file_rename_method'});
1307 $doc_obj->add_utf8_metadata ($cursection, $field, $zip_filename);
1308 }
1309 elsif (ref ($metadata->{$field}) eq "ARRAY") {
1310 map {
1311 $doc_obj->add_utf8_metadata ($cursection, $field, $_);
1312 } @{$metadata->{$field}};
1313 } else {
1314 $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
1315 }
1316 }
1317}
1318
1319
1320sub compile_stats {
1321 my $self = shift(@_);
1322 my ($stats) = @_;
1323
1324 $stats->{'num_processed'} += $self->{'num_processed'};
1325 $stats->{'num_not_processed'} += $self->{'num_not_processed'};
1326 $stats->{'num_archives'} += $self->{'num_archives'};
1327
1328}
1329sub associate_source_file {
1330 my $self = shift(@_);
1331
1332 my ($doc_obj, $filename) = @_;
1333 my $cursection = $doc_obj->get_top_section();
1334 my $assocfilename = $doc_obj->get_assocfile_from_sourcefile();
1335
1336 $doc_obj->associate_file($filename, $assocfilename, undef, $cursection);
1337 $doc_obj->add_utf8_metadata ($cursection, "srclink_file", $doc_obj->get_sourcefile());
1338}
1339
1340sub associate_cover_image {
1341 my $self = shift(@_);
1342 my ($doc_obj, $filename) = @_;
1343
1344 my $upgraded_filename = &util::upgrade_if_dos_filename($filename);
1345
1346 $filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1347 $upgraded_filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1348
1349 if (exists $self->{'covers_missing_cache'}->{$upgraded_filename}) {
1350 # don't stat() for existence e.g. for multiple document input files
1351 # (eg SplitPlug)
1352 return;
1353 }
1354
1355 my $top_section=$doc_obj->get_top_section();
1356
1357 if (&util::fd_exists($upgraded_filename)) {
1358 $doc_obj->associate_source_file($filename);
1359 $doc_obj->associate_file($filename, "cover.jpg", "image/jpeg");
1360 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1361 } else {
1362 my $upper_filename = $filename;
1363 my $upgraded_upper_filename = $upgraded_filename;
1364
1365 $upper_filename =~ s/jpg$/JPG/;
1366 $upgraded_upper_filename =~ s/jpg$/JPG/;
1367
1368 if (&util::fd_exists($upgraded_upper_filename)) {
1369 $doc_obj->associate_source_file($upper_filename);
1370 $doc_obj->associate_file($upper_filename, "cover.jpg",
1371 "image/jpeg");
1372 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1373 } else {
1374 # file doesn't exist, so record the fact that it's missing so
1375 # we don't stat() again (stat is slow)
1376 $self->{'covers_missing_cache'}->{$upgraded_filename} = 1;
1377 }
1378 }
1379
1380}
1381
1382
1383# Overridden by exploding plugins (eg. ISISPlug)
1384sub clean_up_after_exploding
1385{
1386 my $self = shift(@_);
1387}
1388
1389
1390
13911;
Note: See TracBrowser for help on using the repository browser.