source: main/trunk/greenstone2/perllib/plugins/BasePlugin.pm@ 25743

Last change on this file since 25743 was 25743, checked in by kjdon, 12 years ago

if we happen to have a file and matching process expression that doesn't have a file extension, eg 'mbox' instead of 'xxx.mbox', then block_cover_image will block the file as replacing .xxx with .jpg will not do anything, and the file will exist and be blocked. So check that we have actually generated a new filename before seeing if it exists
  • Property svn:keywords set to Author Date Id Revision
File size: 46.2 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;
34use Unicode::Normalize 'normalize';
35
36use encodings;
37use unicode;
38use doc;
39use ghtml;
40use gsprintf 'gsprintf';
41
42use PrintInfo;
43
44BEGIN {
45 @BasePlugin::ISA = ( 'PrintInfo' );
46}
47
48# the different methods that can be applied when renaming
49# imported documents and their associated files
50our $file_rename_method_list =
51 [ { 'name' => "url",
52 'desc' => "{BasePlugin.rename_method.url}" },
53 { 'name' => "base64",
54 'desc' => "{BasePlugin.rename_method.base64}" },
55 { 'name' => "none",
56 'desc' => "{BasePlugin.rename_method.none}",
57 'hiddengli' => "yes" } ];
58
59our $encoding_list =
60 [ { 'name' => "ascii",
61 'desc' => "{BasePlugin.encoding.ascii}" },
62 { 'name' => "utf8",
63 'desc' => "{BasePlugin.encoding.utf8}" },
64 { 'name' => "unicode",
65 'desc' => "{BasePlugin.encoding.unicode}" } ];
66
67
68my $e = $encodings::encodings;
69foreach my $enc (sort {$e->{$a}->{'name'} cmp $e->{$b}->{'name'}} keys (%$e))
70{
71 my $hashEncode =
72 {'name' => $enc,
73 'desc' => $e->{$enc}->{'name'}};
74
75 push(@{$encoding_list},$hashEncode);
76}
77
78our $encoding_plus_auto_list =
79 [ { 'name' => "auto",
80 'desc' => "{BasePlugin.filename_encoding.auto}" },
81 { 'name' => "auto-language-analysis",
82 'desc' => "{BasePlugin.filename_encoding.auto_language_analysis}" }, # textcat
83 { 'name' => "auto-filesystem-encoding",
84 'desc' => "{BasePlugin.filename_encoding.auto_filesystem_encoding}" }, # locale
85 { 'name' => "auto-fl",
86 'desc' => "{BasePlugin.filename_encoding.auto_fl}" }, # locale followed by textcat
87 { 'name' => "auto-lf",
88 'desc' => "{BasePlugin.filename_encoding.auto_lf}" } ]; # texcat followed by locale
89
90push(@{$encoding_plus_auto_list},@{$encoding_list});
91
92our $oidtype_list =
93 [ { 'name' => "auto",
94 'desc' => "{BasePlugin.OIDtype.auto}" },
95 { 'name' => "hash",
96 'desc' => "{import.OIDtype.hash}" },
97 { 'name' => "hash_on_ga_xml",
98 'desc' => "{import.OIDtype.hash_on_ga_xml}" },
99 { 'name' => "assigned",
100 'desc' => "{import.OIDtype.assigned}" },
101 { 'name' => "incremental",
102 'desc' => "{import.OIDtype.incremental}" },
103 { 'name' => "dirname",
104 'desc' => "{import.OIDtype.dirname}" } ];
105
106my $arguments =
107 [ { 'name' => "process_exp",
108 'desc' => "{BasePlugin.process_exp}",
109 'type' => "regexp",
110 'deft' => "",
111 'reqd' => "no" },
112 { 'name' => "no_blocking",
113 'desc' => "{BasePlugin.no_blocking}",
114 'type' => "flag",
115 'reqd' => "no"},
116 { 'name' => "block_exp",
117 'desc' => "{BasePlugin.block_exp}",
118 'type' => "regexp",
119 'deft' => "",
120 'reqd' => "no" },
121 { 'name' => "store_original_file",
122 'desc' => "{BasePlugin.store_original_file}",
123 'type' => "flag",
124 'reqd' => "no" },
125 { 'name' => "associate_ext",
126 'desc' => "{BasePlugin.associate_ext}",
127 'type' => "string",
128 'reqd' => "no" },
129 { 'name' => "associate_tail_re",
130 'desc' => "{BasePlugin.associate_tail_re}",
131 'type' => "string",
132 'reqd' => "no" },
133 { 'name' => "OIDtype",
134 'desc' => "{import.OIDtype}",
135 'type' => "enum",
136 'list' => $oidtype_list,
137 # leave default empty so we can tell if its been set or not - if not set will use option from import.pl
138 'deft' => "auto",
139 'reqd' => "no" },
140 { 'name' => "OIDmetadata",
141 'desc' => "{import.OIDmetadata}",
142 'type' => "metadata",
143 'deft' => "dc.Identifier",
144 'reqd' => "no" },
145# { 'name' => "use_as_doc_identifier",
146# 'desc' => "{BasePlugin.use_as_doc_identifier}",
147# 'type' => "string",
148# 'reqd' => "no" ,
149# 'deft' => "" } ,
150 { 'name' => "no_cover_image",
151 'desc' => "{BasePlugin.no_cover_image}",
152 'type' => "flag",
153 'reqd' => "no" },
154 { 'name' => "filename_encoding",
155 'desc' => "{BasePlugin.filename_encoding}",
156 'type' => "enum",
157 'deft' => "auto",
158 'list' => $encoding_plus_auto_list,
159 'reqd' => "no" },
160 { 'name' => "smart_block",
161 'desc' => "{common.deprecated}. {BasePlugin.smart_block}",
162 'type' => "flag",
163 'reqd' => "no",
164 'hiddengli' => "yes" }, # deprecated, but leave in for old collections
165 { 'name' => "file_rename_method",
166 'desc' => "{BasePlugin.file_rename_method}",
167 'type' => "enum",
168 'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
169 'list' => $file_rename_method_list,
170 'reqd' => "no"
171 }
172
173 ];
174
175
176my $options = { 'name' => "BasePlugin",
177 'desc' => "{BasePlugin.desc}",
178 'abstract' => "yes",
179 'inherits' => "no",
180 'args' => $arguments };
181
182sub new {
183
184 my ($class) = shift (@_);
185 my ($pluginlist,$inputargs,$hashArgOptLists,$auxiliary) = @_;
186 push(@$pluginlist, $class);
187
188 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
189 push(@{$hashArgOptLists->{"OptList"}},$options);
190
191 my $self = new PrintInfo($pluginlist, $inputargs, $hashArgOptLists,$auxiliary);
192
193 if ($self->{'info_only'}) {
194 # don't worry about any options etc
195 return bless $self, $class;
196 }
197
198 if ($self->{'smart_block'}) {
199 print STDERR "WARNING: -smart_block option has been deprecated and is no longer useful\n";
200 }
201 $self->{'smart_block'} = undef;
202
203 my $plugin_name = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
204 $self->{'plugin_type'} = $plugin_name;
205
206 # remove ex. from OIDmetadata iff it's the only namespace prefix
207 $self->{'OIDmetadata'} =~ s/^ex\.([^.]+)$/$1/ if defined $self->{'OIDmetadata'};
208 $self->{'num_processed'} = 0;
209 $self->{'num_not_processed'} = 0;
210 $self->{'num_blocked'} = 0;
211 $self->{'num_archives'} = 0;
212 $self->{'cover_image'} = 1; # cover image is on by default
213 $self->{'cover_image'} = 0 if ($self->{'no_cover_image'});
214 $self->{'can_process_directories'} = 0;
215 #$self->{'option_list'} = $hashArgOptLists->{"OptList"};
216
217 my $associate_ext = $self->{'associate_ext'};
218 if ((defined $associate_ext) && ($associate_ext ne "")) {
219
220 my $associate_tail_re = $self->{'associate_tail_re'};
221 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
222 my $outhandle = $self->{'outhandle'};
223 print $outhandle "Warning: can only specify 'associate_ext' or 'associate_tail_re'\n";
224 print $outhandle " defaulting to 'associate_tail_re'\n";
225 }
226 else {
227 my @exts = split(/,/,$associate_ext);
228
229 my @exts_bracketed = map { $_ = "(?:\\.$_)" } @exts;
230 my $associate_tail_re = join("|",@exts_bracketed);
231 $self->{'associate_tail_re'} = $associate_tail_re;
232 }
233
234 delete $self->{'associate_ext'};
235 }
236
237 return bless $self, $class;
238
239}
240
241sub merge_inheritance
242{
243 my $self = {};
244 my @child_selfs = @_;
245
246 foreach my $child_self (@child_selfs) {
247 foreach my $key (keys %$child_self) {
248 if (defined $self->{$key}) {
249 if ($self->{$key} ne $child_self->{$key}) {
250# print STDERR "Warning: Conflicting value in multiple inheritance for '$key'\n";
251# print STDERR "Existing stored value = $self->{$key}\n";
252# print STDERR "New (child) value = $child_self->{$key}\n";
253# print STDERR "Keeping existing value\n";
254 # Existing value seems to be option specified in collect.cfg
255
256 ### $self->{$key} = $child_self->{$key};
257
258 }
259 else {
260## print STDERR "****Info: Value $self->{$key} for $key already defined through multiple inheritance as the same value\n";
261 }
262
263 }
264 else {
265 $self->{$key} = $child_self->{$key};
266 }
267 }
268 }
269
270 return $self;
271}
272
273# initialize BasePlugin options
274# if init() is overridden in a sub-class, remember to call BasePlugin::init()
275sub init {
276 my $self = shift (@_);
277 my ($verbosity, $outhandle, $failhandle) = @_;
278
279 # verbosity is passed through from the processor
280 $self->{'verbosity'} = $verbosity;
281
282 # as are the outhandle and failhandle
283 $self->{'outhandle'} = $outhandle if defined $outhandle;
284 $self->{'failhandle'} = $failhandle;
285# $self->SUPER::init(@_);
286
287 # set process_exp and block_exp to defaults unless they were
288 # explicitly set
289
290 if ((!$self->is_recursive()) and
291 (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
292
293 $self->{'process_exp'} = $self->get_default_process_exp ();
294 if ($self->{'process_exp'} eq "") {
295 warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
296 }
297 }
298
299 if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
300 $self->{'block_exp'} = $self->get_default_block_exp ();
301 }
302
303}
304
305sub begin {
306 my $self = shift (@_);
307 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
308
309 if ($self->{'OIDtype'} eq "auto") {
310 # hasn't been set in the plugin, use the processor values
311 $self->{'OIDtype'} = $processor->{'OIDtype'};
312 $self->{'OIDmetadata'} = $processor->{'OIDmetadata'};
313 }
314 if ($self->{'OIDtype'} eq "hash") {
315 # should we hash on the file or on the doc xml??
316 $self->{'OIDtype'} = $self->get_oid_hash_type();
317 if ($self->{'OIDtype'} !~ /^(hash_on_file|hash_on_ga_xml)$/) {
318 $self->{'OIDtype'} = "hash_on_file";
319 }
320 }
321}
322
323# This is called once if removeold is set with import.pl. Most plugins will do
324# nothing but if a plugin does any stuff outside of creating doc obj, then
325# it may need to clear something.
326sub remove_all {
327 my $self = shift (@_);
328 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
329}
330
331# This is called per document for docs that have been deleted from the
332# collection. Most plugins will do nothing
333# but if a plugin does any stuff outside of creating doc obj, then it may need
334# to clear something.
335sub remove_one {
336 my $self = shift (@_);
337
338 my ($file, $oids, $archivedir) = @_;
339 return 0 if $self->can_process_this_file($file);
340 return undef;
341}
342
343sub end {
344 # potentially called at the end of each plugin pass
345 # import.pl only has one plugin pass, but buildcol.pl has multiple ones
346
347 my ($self) = shift (@_);
348}
349
350sub deinit {
351 # called only once, after all plugin passes have been done
352
353 my ($self) = @_;
354}
355
356# default hashing type is to hash on the original file (or converted file)
357# override this to return hash_on_ga_xml for filetypes where hashing on the
358# file is no good eg video
359sub get_oid_hash_type {
360
361 my $self = shift (@_);
362
363 return "hash_on_file";
364}
365
366
367# this function should be overridden to return 1
368# in recursive plugins
369sub is_recursive {
370 my $self = shift (@_);
371
372 return 0;
373}
374
375sub get_default_block_exp {
376 my $self = shift (@_);
377
378 return "";
379}
380
381sub get_default_process_exp {
382 my $self = shift (@_);
383
384 return "";
385}
386
387
388# rename imported files and assoc files using URL encoding by default
389# as this will work for most plugins and give more legible filenames
390sub get_default_file_rename_method() {
391 my $self = shift (@_);
392 return "url";
393}
394
395# returns this plugin's active (possibly user-selected) file_rename_method
396sub get_file_rename_method() {
397 my $self = shift (@_);
398 my $rename_method = $self->{'file_rename_method'};
399 if($rename_method) {
400 return $rename_method;
401 } else {
402 return $self->get_default_file_rename_method();
403 }
404}
405
406# default implementation is to do nothing
407sub store_block_files {
408
409 my $self =shift (@_);
410 my ($filename_full_path, $block_hash) = @_;
411
412}
413
414# put files to block into hash
415sub use_block_expressions {
416
417 my $self =shift (@_);
418 my ($filename_full_path, $block_hash) = @_;
419
420 $filename_full_path = &util::upgrade_if_dos_filename($filename_full_path);
421
422 if ($self->{'block_exp'} ne "" && $filename_full_path =~ /$self->{'block_exp'}/) {
423 &util::block_filename($block_hash,$filename_full_path);
424 }
425
426}
427
428#default implementation is to block a file with same name as this, but extension jpg or JPG, if cover_images is on.
429sub block_cover_image
430{
431 my $self =shift;
432 my ($filename, $block_hash) = @_;
433
434 $filename = &util::upgrade_if_dos_filename($filename);
435
436 if ($self->{'cover_image'}) {
437 my $coverfile = $filename;
438 $coverfile =~ s/\.[^\\\/\.]+$/\.jpg/;
439
440 #if there is no file extension, coverfile will be the same as filename
441 return if $coverfile eq $filename;
442
443 if (!&util::fd_exists($coverfile)) {
444 $coverfile =~ s/jpg$/JPG/;
445 }
446 if (&util::fd_exists($coverfile)) {
447 &util::block_filename($block_hash,$coverfile);
448 }
449 }
450
451 return;
452}
453
454
455# discover all the files that should be blocked by this plugin
456# check the args ...
457sub file_block_read {
458
459 my $self = shift (@_);
460 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $gli) = @_;
461 # Keep track of filenames with same root but different extensions
462 # Used to support -associate_ext and the more generalised
463 # -associate_tail_re
464 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
465
466 if (!-d $filename_full_path) {
467 $block_hash->{'all_files'}->{$file} = 1;
468 }
469
470 my $associate_tail_re = $self->{'associate_tail_re'};
471 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
472 my ($file_prefix,$file_ext)
473 = &util::get_prefix_and_tail_by_regex($filename_full_path,$associate_tail_re);
474 if ((defined $file_prefix) && (defined $file_ext)) {
475 my $shared_fileroot = $block_hash->{'shared_fileroot'};
476 if (!defined $shared_fileroot->{$file_prefix}) {
477 my $file_prefix_rec = { 'tie_to' => undef,
478 'exts' => {} };
479 $shared_fileroot->{$file_prefix} = $file_prefix_rec;
480 }
481
482 my $file_prefix_rec = $shared_fileroot->{$file_prefix};
483
484 if ($self->can_process_this_file($filename_full_path)) {
485 # This is the document the others should be tied to
486 $file_prefix_rec->{'tie_to'} = $file_ext;
487 }
488 else {
489 if ($file_ext =~ m/$associate_tail_re$/) {
490 # this file should be associated to the main one
491 $file_prefix_rec->{'exts'}->{$file_ext} = 1;
492 }
493 }
494
495 }
496 }
497
498 # check block expressions
499 $self->use_block_expressions($filename_full_path, $block_hash) unless $self->{'no_blocking'};
500
501 # now check whether we are actually processing this
502 if (!-f $filename_full_path || !$self->can_process_this_file($filename_full_path)) {
503 return undef; # can't recognise
504 }
505
506 # if we have a block_exp, then this overrides the normal 'smart' blocking
507 $self->store_block_files($filename_full_path, $block_hash) unless ($self->{'no_blocking'} || $self->{'block_exp'} ne "");
508
509 # block the cover image if there is one
510 if ($self->{'cover_image'}) {
511 $self->block_cover_image($filename_full_path, $block_hash);
512 }
513
514 return 1;
515}
516
517# plugins that rely on more than process_exp (eg XML plugins) can override this method
518sub can_process_this_file {
519 my $self = shift(@_);
520 my ($filename) = @_;
521
522 if (-d $filename && !$self->{'can_process_directories'}) {
523 return 0;
524 }
525
526 if ($self->{'process_exp'} ne "" && $filename =~ /$self->{'process_exp'}/) {
527 return 1;
528 }
529 return 0;
530
531}
532
533# Even if a plugin can extract metadata in its metadata_read pass,
534# make the default return 'undef' so processing of the file continues
535# down the pipeline, so other plugins can also have the opportunity to
536# locate metadata and set it up in the extrametakeys variables that
537# are passed around.
538
539sub can_process_this_file_for_metadata {
540 my $self = shift(@_);
541
542 return undef;
543}
544
545
546# just converts path as is to utf8.
547sub filepath_to_utf8 {
548 my $self = shift (@_);
549 my ($file, $file_encoding) = @_;
550 my $filemeta = $file;
551
552 my $filename_encoding = $self->{'filename_encoding'}; # filename encoding setting
553
554 # Whenever filename-encoding is set to any of the auto settings, we
555 # check if the filename is already in UTF8. If it is, then we're done.
556 if($filename_encoding =~ m/auto/) {
557 if(&unicode::check_is_utf8($filemeta))
558 {
559 $filename_encoding = "utf8";
560 return $filemeta;
561 }
562 }
563
564 # Auto setting, but filename is not utf8
565 if ($filename_encoding eq "auto")
566 {
567 # try textcat
568 $filename_encoding = $self->textcat_encoding($filemeta);
569
570 # check the locale next
571 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
572
573
574 # now try the encoding of the document, if available
575 if ($filename_encoding eq "undefined" && defined $file_encoding) {
576 $filename_encoding = $file_encoding;
577 }
578
579 }
580
581 elsif ($filename_encoding eq "auto-language-analysis")
582 {
583 $filename_encoding = $self->textcat_encoding($filemeta);
584
585 # now try the encoding of the document, if available
586 if ($filename_encoding eq "undefined" && defined $file_encoding) {
587 $filename_encoding = $file_encoding;
588 }
589 }
590
591 elsif ($filename_encoding eq "auto-filesystem-encoding")
592 {
593 # try locale
594 $filename_encoding = $self->locale_encoding();
595 }
596
597 elsif ($filename_encoding eq "auto-fl")
598 {
599 # filesystem-encoding (locale) then language-analysis (textcat)
600 $filename_encoding = $self->locale_encoding();
601
602 # try textcat
603 $filename_encoding = $self->textcat_encoding($filemeta) if $filename_encoding eq "undefined";
604
605 # else assume filename encoding is encoding of file content, if that's available
606 if ($filename_encoding eq "undefined" && defined $file_encoding) {
607 $filename_encoding = $file_encoding;
608 }
609 }
610
611 elsif ($filename_encoding eq "auto-lf")
612 {
613 # language-analysis (textcat) then filesystem-encoding (locale)
614 $filename_encoding = $self->textcat_encoding($filemeta);
615
616 # guess filename encoding from encoding of file content, if available
617 if ($filename_encoding eq "undefined" && defined $file_encoding) {
618 $filename_encoding = $file_encoding;
619 }
620
621 # try locale
622 $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
623 }
624
625 # if still undefined, use utf8 as fallback
626 if ($filename_encoding eq "undefined") {
627 $filename_encoding = "utf8";
628 }
629
630 #print STDERR "**** UTF8 encoding the filename $filemeta ";
631
632 # if the filename encoding is set to utf8 but it isn't utf8 already--such as when
633 # 1. the utf8 fallback is used, or 2. if the system locale is used and happens to
634 # be always utf8 (in which case the filename's encoding is also set as utf8 even
635 # though the filename need not be if it originates from another system)--in such
636 # cases attempt to make the filename utf8 to match.
637 if($filename_encoding eq "utf8" && !&unicode::check_is_utf8($filemeta)) {
638 &unicode::ensure_utf8(\$filemeta);
639 }
640
641 # convert non-unicode encodings to utf8
642 if ($filename_encoding !~ m/(?:ascii|utf8|unicode)/) {
643 $filemeta = &unicode::unicode2utf8(
644 &unicode::convert2unicode($filename_encoding, \$filemeta)
645 );
646 }
647
648 #print STDERR " from encoding $filename_encoding -> $filemeta\n";
649 return $filemeta;
650}
651
652# gets the filename with no path, converts to utf8, and then dm safes it.
653# filename_encoding set by user
654sub filename_to_utf8_metadata
655{
656 my $self = shift (@_);
657 my ($file, $file_encoding) = @_;
658
659 my $outhandle = $self->{'outhandle'};
660
661 print $outhandle "****!!!!**** BasePlugin::filename_to_utf8_metadata now deprecated\n";
662 my ($cpackage,$cfilename,$cline,$csubr,$chas_args,$cwantarray) = caller(0);
663 print $outhandle "Calling method: $cfilename:$cline $cpackage->$csubr\n";
664
665
666 my ($filemeta) = $file =~ /([^\\\/]+)$/; # getting the tail of the filepath (skips all string parts containing slashes upto the end)
667 $filemeta = $self->filepath_to_utf8($filemeta, $file_encoding);
668
669 return $filemeta;
670}
671
672sub locale_encoding {
673 my $self = shift(@_);
674
675 if (!defined $self->{'filesystem_encoding'}) {
676 $self->{'filesystem_encoding'} = $self->get_filesystem_encoding();
677 }
678
679 #print STDERR "*** filename encoding determined based on locale: " . $self->{'filesystem_encoding'} . "\n";
680 return $self->{'filesystem_encoding'}; # can be the string "undefined"
681}
682
683sub textcat_encoding {
684 my $self = shift(@_);
685 my ($filemeta) = @_;
686
687 # analyse filenames without extensions and digits (and trimmed of
688 # surrounding whitespace), so that irrelevant chars don't confuse
689 # textcat
690 my $strictfilemeta = $filemeta;
691 $strictfilemeta =~ s/\.[^\.]+$//g;
692 $strictfilemeta =~ s/\d//g;
693 $strictfilemeta =~ s/^\s*//g;
694 $strictfilemeta =~ s/\s*$//g;
695
696 my $filename_encoding = $self->encoding_from_language_analysis($strictfilemeta);
697 if(!defined $filename_encoding) {
698 $filename_encoding = "undefined";
699 }
700
701 return $filename_encoding; # can be the string "undefined"
702}
703
704# performs textcat
705sub encoding_from_language_analysis {
706 my $self = shift(@_);
707 my ($text) = @_;
708
709 my $outhandle = $self->{'outhandle'};
710 my $best_encoding = undef;
711
712 # get the language/encoding of the textstring using textcat
713 require textcat; # Only load the textcat module if it is required
714 $self->{'textcat'} = new textcat() unless defined($self->{'textcat'});
715 my $results = $self->{'textcat'}->classify_cached_filename(\$text);
716
717
718 if (scalar @$results < 0) {
719 return undef;
720 }
721
722 # We have some results, we choose the first
723 my ($language, $encoding) = $results->[0] =~ /^([^-]*)(?:-(.*))?$/;
724
725 $best_encoding = $encoding;
726 if (!defined $best_encoding) {
727 return undef;
728 }
729
730 if (defined $best_encoding && $best_encoding =~ m/^iso_8859/ && &unicode::check_is_utf8($text)) {
731 # the text is valid utf8, so assume that's the real encoding (since textcat is based on probabilities)
732 $best_encoding = 'utf8';
733 }
734
735
736 # check for equivalents where textcat doesn't have some encodings...
737 # eg MS versions of standard encodings
738 if (defined $best_encoding && $best_encoding =~ /^iso_8859_(\d+)/) {
739 my $iso = $1; # which variant of the iso standard?
740 # iso-8859 sets don't use chars 0x80-0x9f, windows codepages do
741 if ($text =~ /[\x80-\x9f]/) {
742 # Western Europe
743 if ($iso == 1 or $iso == 15) { $best_encoding = 'windows_1252' }
744 elsif ($iso == 2) {$best_encoding = 'windows_1250'} # Central Europe
745 elsif ($iso == 5) {$best_encoding = 'windows_1251'} # Cyrillic
746 elsif ($iso == 6) {$best_encoding = 'windows_1256'} # Arabic
747 elsif ($iso == 7) {$best_encoding = 'windows_1253'} # Greek
748 elsif ($iso == 8) {$best_encoding = 'windows_1255'} # Hebrew
749 elsif ($iso == 9) {$best_encoding = 'windows_1254'} # Turkish
750 }
751 }
752
753 if (defined $best_encoding && $best_encoding !~ /^(ascii|utf8|unicode)$/ &&
754 !defined $encodings::encodings->{$best_encoding})
755 {
756 if ($self->{'verbosity'}) {
757 gsprintf($outhandle, "BasePlugin: {ReadTextFile.unsupported_encoding}\n", $text, $best_encoding, "undef");
758 }
759 $best_encoding = undef;
760 }
761
762 return $best_encoding;
763}
764
765# uses locale
766sub get_filesystem_encoding
767{
768
769 my $self = shift(@_);
770
771 my $outhandle = $self->{'outhandle'};
772 my $filesystem_encoding = undef;
773
774 eval {
775 # Works for Windows as well, returning the DOS code page in use
776 use POSIX qw(locale_h);
777
778 # With only one parameter, setlocale retrieves the
779 # current value
780 my $current_locale = setlocale(LC_CTYPE);
781
782 my $char_encoding = undef;
783 if ($current_locale =~ m/\./) {
784 ($char_encoding) = ($current_locale =~ m/^.*\.(.*?)$/);
785 $char_encoding = lc($char_encoding);
786 }
787 else {
788 if ($current_locale =~ m/^(posix|c)$/i) {
789 $char_encoding = "ascii";
790 }
791 }
792
793 if (defined $char_encoding) {
794 if ($char_encoding =~ m/^(iso)(8859)-?(\d{1,2})$/) {
795 $char_encoding = "$1\_$2\_$3";
796 }
797
798 $char_encoding =~ s/-/_/g;
799 $char_encoding =~ s/^utf_8$/utf8/;
800
801 if ($char_encoding =~ m/^\d+$/) {
802 if (defined $encodings::encodings->{"windows_$char_encoding"}) {
803 $char_encoding = "windows_$char_encoding";
804 }
805 elsif (defined $encodings::encodings->{"dos_$char_encoding"}) {
806 $char_encoding = "dos_$char_encoding";
807 }
808 }
809
810 if (($char_encoding =~ m/(?:ascii|utf8|unicode)/)
811 || (defined $encodings::encodings->{$char_encoding})) {
812 $filesystem_encoding = $char_encoding;
813 }
814 else {
815 print $outhandle "Warning: Unsupported character encoding '$char_encoding' from locale '$current_locale'\n";
816 }
817 }
818
819
820 };
821 if ($@) {
822 print $outhandle "$@\n";
823 print $outhandle "Warning: Unable to establish locale. Will assume filesystem is UTF-8\n";
824
825 }
826
827 return $filesystem_encoding;
828}
829
830
831
832sub deduce_filename_encoding
833{
834 my $self = shift (@_);
835 my ($file,$metadata,$plugin_filename_encoding) = @_;
836
837 my $gs_filename_encoding = $metadata->{"gs.filenameEncoding"};
838 my $deduced_filename_encoding = undef;
839
840 # Start by looking for manually assigned metadata
841 if (defined $gs_filename_encoding) {
842 if (ref ($gs_filename_encoding) eq "ARRAY") {
843 my $outhandle = $self->{'outhandle'};
844
845 $deduced_filename_encoding = $gs_filename_encoding->[0];
846
847 my $num_vals = scalar(@$gs_filename_encoding);
848 if ($num_vals>1) {
849 print $outhandle "Warning: gs.filenameEncoding multiply defined for $file\n";
850 print $outhandle " Selecting first value: $deduced_filename_encoding\n";
851 }
852 }
853 else {
854 $deduced_filename_encoding = $gs_filename_encoding;
855 }
856 }
857
858 if (!defined $deduced_filename_encoding || ($deduced_filename_encoding =~ m/^\s*$/)) {
859 # Look to see if plugin specifies this value
860
861 if (defined $plugin_filename_encoding) {
862 # First look to see if we're using any of the "older" (i.e. deprecated auto-... plugin options)
863 if ($plugin_filename_encoding =~ m/^auto-.*$/) {
864 my $outhandle = $self->{'outhandle'};
865 print $outhandle "Warning: $plugin_filename_encoding is no longer supported\n";
866 print $outhandle " default to 'auto'\n";
867 $self->{'filename_encoding'} = $plugin_filename_encoding = "auto";
868 }
869
870 if ($plugin_filename_encoding ne "auto") {
871 # We've been given a specific filenamne encoding
872 # => so use it!
873 $deduced_filename_encoding = $plugin_filename_encoding;
874 }
875 }
876 }
877
878 if (!defined $deduced_filename_encoding || ($deduced_filename_encoding =~ m/^\s*$/)) {
879
880 # Look to file system to provide a character encoding
881
882 # If Windows NTFS, then -- assuming we work with long file names got through
883 # Win32::GetLongFilePath() -- then the underlying file system is UTF16
884
885 if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
886 # Can do better than working with the DOS character encoding returned by locale
887 $deduced_filename_encoding = "unicode";
888 }
889 else {
890 # Unix of some form or other
891
892 # See if we can determine the file system encoding through locale
893 $deduced_filename_encoding = $self->locale_encoding();
894
895 # if locale shows us filesystem is utf8, check to see filename is consistent
896 # => if not, then we have an "alien" filename on our hands
897
898 if (defined $deduced_filename_encoding && $deduced_filename_encoding =~ m/^utf-?8$/i) {
899 if (!&unicode::check_is_utf8($file)) {
900 # "alien" filename, so revert
901 $deduced_filename_encoding = undef;
902 }
903 }
904 }
905 }
906
907# if (!defined $deduced_filename_encoding || ($deduced_filename_encoding =~ m/^\s*$/)) {
908# # Last chance, apply textcat to deduce filename encoding
909# $deduced_filename_encoding = $self->textcat_encoding($file);
910# }
911
912 if ($self->{'verbosity'}>3) {
913 my $outhandle = $self->{'outhandle'};
914
915 if (defined $deduced_filename_encoding) {
916 print $outhandle " Deduced filename encoding as: $deduced_filename_encoding\n";
917 }
918 else {
919 print $outhandle " No filename encoding deduced\n";
920 }
921 }
922
923 return $deduced_filename_encoding;
924}
925
926
927
928
929# Notionally written to be called once for each document, it is however safe to
930# call multiple times (as in the case of ImagePlugin) which calls this later on
931# after the original image has potentially been converted to a *new* source image
932# format (e.g. TIFF to PNG)
933
934sub set_Source_metadata {
935 my $self = shift (@_);
936 my ($doc_obj, $raw_filename, $filename_encoding, $section) = @_;
937
938 # 1. Sets the filename (Source) for display encoded as Unicode if possible,
939 # and (as a fallback) using %xx if not for non-ascii chars
940 # 2. Sets the url ref (SourceFile) to the URL encoded version
941 # of filename for generated files
942
943 my ($unused_full_rf, $raw_file) = &util::get_full_filenames("", $raw_filename);
944
945 my $this_section = (defined $section)? $section : $doc_obj->get_top_section();
946
947 my $octet_file = $raw_file;
948
949 # UTF-8 version of filename
950# if ((defined $ENV{"DEBUG_UNICODE"}) && ($ENV{"DEBUG_UNICODE"})) {
951# print STDERR "**** Setting Source Metadata given: $octet_file\n";
952# }
953
954 # Deal with (on Windows) raw filenames that are in their
955 # abbreviated DOS form
956
957 if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
958 if ((defined $filename_encoding) && ($filename_encoding eq "unicode")) {
959 if (-e $raw_filename) {
960 my $unicode_filename = Win32::GetLongPathName($raw_filename);
961
962 my $unused_full_uf;
963 ($unused_full_uf, $octet_file) = &util::get_full_filenames("", $unicode_filename);
964 }
965 }
966 }
967
968 my $url_encoded_filename;
969 if ((defined $filename_encoding) && ($filename_encoding ne "ascii")) {
970 # => Generate a pretty print version of filename that is mapped to Unicode
971
972 # Use filename_encoding to map raw filename to a Perl unicode-aware string
973 $url_encoded_filename = decode($filename_encoding,$octet_file);
974 }
975 else {
976 # otherwise generate %xx encoded version of filename for char > 127
977 $url_encoded_filename = &unicode::raw_filename_to_url_encoded($octet_file);
978 }
979
980# if ((defined $ENV{"DEBUG_UNICODE"}) && ($ENV{"DEBUG_UNICODE"})) {
981# print STDERR "****** saving Source as: $url_encoded_filename\n";
982# }
983
984 # In the case of converted files and (generalized) exploded documents, there
985 # will already be a source filename => store as OrigSource before overriding
986 my $orig_source = $doc_obj->get_metadata_element ($this_section, "Source");
987 if ((defined $orig_source) && ($orig_source !~ m/^\s*$/)) {
988 $doc_obj->set_utf8_metadata_element($this_section, "OrigSource", $orig_source);
989 }
990
991 # Source is the UTF8 display name - not necessarily the name of the file on the system
992 if ($ENV{'GSDLOS'} =~ m/^darwin$/i) {
993 # on Darwin want all display strings to be in composed form, then can search on that
994 $url_encoded_filename = normalize('C', $url_encoded_filename); # Normalisation Form 'C' (composition)
995 }
996 $doc_obj->set_utf8_metadata_element($this_section, "Source", $url_encoded_filename);
997
998
999 my $renamed_raw_file = &util::rename_file($raw_file, $self->{'file_rename_method'});
1000 # If using URL encoding, then SourceFile is the url-reference to url-encoded
1001 # renamed_raw_url: it's a url that refers to the actual file on the system
1002 my $renamed_raw_url = &unicode::filename_to_url($renamed_raw_file);
1003
1004 $doc_obj->set_utf8_metadata_element($this_section, "SourceFile",
1005 $renamed_raw_url);
1006
1007# if ((defined $ENV{"DEBUG_UNICODE"}) && ($ENV{"DEBUG_UNICODE"})) {
1008# print STDERR "****** saving SourceFile as: $renamed_raw_url\n";
1009# }
1010}
1011
1012# this should be called by all plugins to set the oid of the doc obj, rather
1013# than calling doc_obj->set_OID directly
1014sub add_OID {
1015 my $self = shift (@_);
1016 my ($doc_obj) = @_;
1017
1018 $doc_obj->set_OIDtype($self->{'OIDtype'}, $self->{'OIDmetadata'});
1019
1020 # see if there is a plugin specific set_OID function
1021 if (defined ($self->can('set_OID'))) {
1022 $self->set_OID(@_); # pass through doc_obj and any extra arguments
1023 }
1024 else {
1025 # use the default set_OID() in doc.pm
1026 $doc_obj->set_OID();
1027 }
1028
1029}
1030
1031# The BasePlugin read_into_doc_obj() function. This function does all the
1032# right things to make general options work for a given plugin. It doesn't do anything with the file other than setting reads in
1033# a file and sets up a slew of metadata all saved in doc_obj, which
1034# it then returns as part of a tuple (process_status,doc_obj)
1035#
1036# Much of this functionality used to reside in read, but it was broken
1037# down into a supporting routine to make the code more flexible.
1038#
1039# recursive plugins (e.g. RecPlug) and specialized plugins like those
1040# capable of processing many documents within a single file (e.g.
1041# GMLPlug) will normally want to implement their own version of
1042# read_into_doc_obj()
1043#
1044# Note that $base_dir might be "" and that $file might
1045# include directories
1046
1047# currently blocking has been done before it gets here - does this affect secondary plugin stuff??
1048sub read_into_doc_obj {
1049 my $self = shift (@_);
1050 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
1051
1052 my $outhandle = $self->{'outhandle'};
1053
1054 # should we move this to read? What about secondary plugins?
1055 my $pp_file = &util::prettyprint_file($base_dir,$file,$gli);
1056 print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
1057 print $outhandle "$self->{'plugin_type'} processing $pp_file\n"
1058 if $self->{'verbosity'} > 1;
1059
1060 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
1061
1062 # create a new document
1063 my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
1064 my $top_section = $doc_obj->get_top_section();
1065
1066 $doc_obj->add_utf8_metadata($top_section, "Plugin", "$self->{'plugin_type'}");
1067 $doc_obj->add_utf8_metadata($top_section, "FileSize", (-s $filename_full_path));
1068
1069
1070 my $plugin_filename_encoding = $self->{'filename_encoding'};
1071 my $filename_encoding = $self->deduce_filename_encoding($file,$metadata,$plugin_filename_encoding);
1072 $self->set_Source_metadata($doc_obj,$filename_full_path,$filename_encoding,$top_section);
1073
1074 # plugin specific stuff - what args do we need here??
1075 unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
1076 print STDERR "<ProcessingError n='$file'>\n" if ($gli);
1077 return -1;
1078 }
1079
1080 # include any metadata passed in from previous plugins
1081 # note that this metadata is associated with the top level section
1082 my $section = $doc_obj->get_top_section();
1083 # can we merge these two methods??
1084 $self->add_associated_files($doc_obj, $filename_full_path);
1085 $self->extra_metadata ($doc_obj, $section, $metadata);
1086 $self->auto_extract_metadata($doc_obj);
1087
1088 # if we haven't found any Title so far, assign one
1089 # this was shifted to here from inside read()
1090 $self->title_fallback($doc_obj,$section,$filename_no_path);
1091
1092 $self->add_OID($doc_obj);
1093
1094 $self->post_process_doc_obj($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli);
1095 return (1,$doc_obj);
1096}
1097
1098sub post_process_doc_obj {
1099 my $self = shift (@_);
1100 my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
1101
1102 return 1;
1103}
1104
1105sub add_dummy_text {
1106 my $self = shift(@_);
1107 my ($doc_obj, $section) = @_;
1108
1109 # add NoText metadata so we can hide this dummy text in format statements
1110 $doc_obj->add_metadata($section, "NoText", "1");
1111 $doc_obj->add_text($section, &gsprintf::lookup_string("{BasePlugin.dummy_text}",1));
1112
1113}
1114
1115# does nothing. Can be overridden by subclass
1116sub auto_extract_metadata {
1117 my $self = shift(@_);
1118 my ($doc_obj) = @_;
1119}
1120
1121# adds cover image, associate_file options stuff. Should be called by sub class
1122# read_into_doc_obj
1123sub add_associated_files {
1124 my $self = shift(@_);
1125 # whatis filename??
1126 my ($doc_obj, $filename) = @_;
1127
1128 # add in the cover image
1129 if ($self->{'cover_image'}) {
1130 $self->associate_cover_image($doc_obj, $filename);
1131 }
1132 # store the original (used for eg TextPlugin to store the original for OAI)
1133 if ($self->{'store_original_file'}) {
1134 $self->associate_source_file($doc_obj, $filename);
1135 }
1136
1137
1138}
1139
1140# implement this if you are extracting metadata for other documents
1141sub metadata_read {
1142 my $self = shift (@_);
1143 my ($pluginfo, $base_dir, $file, $block_hash,
1144 $extrametakeys, $extrametadata, $extrametafile,
1145 $processor, $gli, $aux) = @_;
1146
1147 # can we process this file??
1148 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
1149 return undef unless $self->can_process_this_file_for_metadata($filename_full_path);
1150
1151 return 1; # we recognise the file, but don't actually do anything with it
1152}
1153
1154
1155# The BasePlugin read() function. This function calls read_into_doc_obj()
1156# to ensure all the right things to make general options work for a
1157# given plugin are done. It then calls the process() function which
1158# does all the work specific to a plugin (like the old read functions
1159# used to do). Most plugins should define their own process() function
1160# and let this read() function keep control.
1161#
1162# recursive plugins (e.g. RecPlug) and specialized plugins like those
1163# capable of processing many documents within a single file (e.g.
1164# GMLPlug) might want to implement their own version of read(), but
1165# more likely need to implement their own version of read_into_doc_obj()
1166#
1167# Return number of files processed, undef if can't recognise, -1 if can't
1168# process
1169
1170sub read {
1171 my $self = shift (@_);
1172 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
1173
1174 # can we process this file??
1175 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
1176
1177 return undef unless $self->can_process_this_file($filename_full_path);
1178
1179 my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
1180
1181 if ((defined $process_status) && ($process_status == 1)) {
1182
1183 # process the document
1184 $processor->process($doc_obj);
1185
1186 $self->{'num_processed'} ++;
1187 undef $doc_obj;
1188 }
1189 # delete any temp files that we may have created
1190 $self->clean_up_after_doc_obj_processing();
1191
1192
1193 # if process_status == 1, then the file has been processed.
1194 return $process_status;
1195
1196}
1197
1198# returns undef if file is rejected by the plugin
1199sub process {
1200 my $self = shift (@_);
1201 my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
1202
1203 gsprintf(STDERR, "BasePlugin::process {common.must_be_implemented}\n");
1204
1205 my ($cpackage,$cfilename,$cline,$csubr,$chas_args,$cwantarray) = caller(1);
1206 print STDERR "Calling method: $cfilename:$cline $cpackage->$csubr\n";
1207
1208 die "\n";
1209
1210 return undef; # never gets here
1211}
1212
1213# overwrite this method to delete any temp files that we have created
1214sub clean_up_after_doc_obj_processing {
1215 my $self = shift(@_);
1216
1217}
1218
1219# write_file -- used by ConvertToPlug, for example in post processing
1220#
1221# where should this go, is here the best place??
1222sub utf8_write_file {
1223 my $self = shift (@_);
1224 my ($textref, $filename) = @_;
1225
1226 if (!open (FILE, ">:utf8", $filename)) {
1227 gsprintf(STDERR, "ConvertToPlug::write_file {ConvertToPlug.could_not_open_for_writing} ($!)\n", $filename);
1228 die "\n";
1229 }
1230 print FILE $$textref;
1231
1232 close FILE;
1233}
1234
1235
1236sub filename_based_title
1237{
1238 my $self = shift (@_);
1239 my ($file) = @_;
1240
1241 my $file_derived_title = $file;
1242 $file_derived_title =~ s/_/ /g;
1243 $file_derived_title =~ s/\.[^.]+$//;
1244
1245 return $file_derived_title;
1246}
1247
1248
1249sub title_fallback
1250{
1251 my $self = shift (@_);
1252 my ($doc_obj,$section,$file) = @_;
1253
1254 if (!defined $doc_obj->get_metadata_element ($section, "Title")
1255 || $doc_obj->get_metadata_element($section, "Title") eq "") {
1256
1257 my $source_file = $doc_obj->get_metadata_element($section, "Source");
1258 my $file_derived_title;
1259 if (defined $source_file) {
1260 $file_derived_title = $self->filename_based_title($source_file);
1261 }
1262 else {
1263 # pp = pretty print
1264 my $pp_file = (defined $source_file) ? $source_file : $file;
1265
1266 my $raw_title = $self->filename_based_title($file);
1267 my $file_derived_title = &unicode::raw_filename_to_url_encoded($raw_title);
1268 }
1269
1270
1271 if (!defined $doc_obj->get_metadata_element ($section, "Title")) {
1272 $doc_obj->add_utf8_metadata ($section, "Title", $file_derived_title);
1273 }
1274 else {
1275 $doc_obj->set_utf8_metadata_element ($section, "Title", $file_derived_title);
1276 }
1277 }
1278
1279}
1280
1281# add any extra metadata that's been passed around from one
1282# plugin to another.
1283# extra_metadata uses add_utf8_metadata so it expects metadata values
1284# to already be in utf8
1285sub extra_metadata {
1286 my $self = shift (@_);
1287 my ($doc_obj, $cursection, $metadata) = @_;
1288
1289 my $associate_tail_re = $self->{'associate_tail_re'};
1290
1291 foreach my $field (keys(%$metadata)) {
1292 # $metadata->{$field} may be an array reference
1293 if ($field eq "gsdlassocfile_tobe") {
1294 # 'gsdlassocfile_tobe' is artificially introduced metadata
1295 # that is used to signal that certain additional files should
1296 # be tied to this document. Useful in situations where a
1297 # metadata pass in the plugin pipeline works out some files
1298 # need to be associated with a document, but the document hasn't
1299 # been formed yet.
1300 my $equiv_form = "";
1301 foreach my $gaf (@{$metadata->{$field}}) {
1302 my ($full_filename,$mimetype) = ($gaf =~ m/^(.*):(.*):$/);
1303 my ($tail_filename) = ($full_filename =~ /^.*[\/\\](.+?)$/);
1304
1305 # we need to make sure the filename is valid utf-8 - we do
1306 # this by url or base64 encoding it
1307 # $tail_filename is the name that we store the file as
1308 $tail_filename = &util::rename_file($tail_filename, $self->{'file_rename_method'});
1309 $doc_obj->associate_file($full_filename,$tail_filename,$mimetype);
1310 $doc_obj->associate_source_file($full_filename);
1311 # If the filename is url_encoded, we need to encode the % signs
1312 # in the filename, so that it works in a url
1313 my $url_tail_filename = &unicode::filename_to_url($tail_filename);
1314 # work out extended tail extension (i.e. matching tail re)
1315
1316 my ($file_prefix,$file_extended_ext)
1317 = &util::get_prefix_and_tail_by_regex($tail_filename,$associate_tail_re);
1318 my ($pre_doc_ext) = ($file_extended_ext =~ m/^(.*)\..*$/);
1319 my ($doc_ext) = ($tail_filename =~ m/^.*\.(.*)$/);
1320
1321 # the greenstone 2 stuff
1322 my $start_doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/{Or}{[parent(Top):assocfilepath],[assocfilepath]}/$url_tail_filename\">";
1323 #my $start_doclink = "<a href=\'_httpprefix_/collect/[collection]/index/assoc/[assocfilepath]/$url_tail_filename\'>";
1324 my $start_doclink_gs3 = "<a href=\'_httpprefix_/collect/[collection]/index/assoc/[assocfilepath]/$url_tail_filename\'>";
1325
1326 my $srcicon = "_icon".$doc_ext."_";
1327 my $end_doclink = "</a>";
1328
1329 my $assoc_form = "$start_doclink\{If\}{$srcicon,$srcicon,$doc_ext\}$end_doclink";
1330
1331
1332 if (defined $pre_doc_ext && $pre_doc_ext ne "") {
1333 # for metadata such as [mp3._edited] [mp3._full] ...
1334 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.$pre_doc_ext", $assoc_form);
1335 }
1336
1337 # for multiple metadata such as [mp3.assoclink]
1338 $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.assoclink", $assoc_form);
1339
1340 $equiv_form .= " $assoc_form";
1341
1342 # following are used for greenstone 3,
1343 $doc_obj->add_utf8_metadata ($cursection, "equivDocLink", $start_doclink_gs3);
1344 $doc_obj->add_utf8_metadata ($cursection, "equivDocIcon", $srcicon);
1345 $doc_obj->add_utf8_metadata ($cursection, "/equivDocLink", $end_doclink);
1346
1347 }
1348 $doc_obj->add_utf8_metadata ($cursection, "equivlink", $equiv_form);
1349 }
1350 elsif ($field eq "gsdlzipfilename") {
1351 # special case for when files have come out of a zip. source_path
1352 # (used for archives dbs and keeping track for incremental import)
1353 # must be set to the zip file name
1354 my $zip_filename = $metadata->{$field};
1355 # overwrite the source_path
1356 $doc_obj->set_source_path($zip_filename);
1357 # and set the metadata
1358 $zip_filename = &util::filename_within_collection($zip_filename);
1359 $zip_filename = $doc_obj->encode_filename($zip_filename, $self->{'file_rename_method'});
1360 $doc_obj->add_utf8_metadata ($cursection, $field, $zip_filename);
1361 }
1362 elsif (ref ($metadata->{$field}) eq "ARRAY") {
1363 map {
1364 $doc_obj->add_utf8_metadata ($cursection, $field, $_);
1365 } @{$metadata->{$field}};
1366 } else {
1367 $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
1368 }
1369 }
1370}
1371
1372
1373sub compile_stats {
1374 my $self = shift(@_);
1375 my ($stats) = @_;
1376
1377 $stats->{'num_processed'} += $self->{'num_processed'};
1378 $stats->{'num_not_processed'} += $self->{'num_not_processed'};
1379 $stats->{'num_archives'} += $self->{'num_archives'};
1380
1381}
1382sub associate_source_file {
1383 my $self = shift(@_);
1384
1385 my ($doc_obj, $filename) = @_;
1386 my $cursection = $doc_obj->get_top_section();
1387 my $assocfilename = $doc_obj->get_assocfile_from_sourcefile();
1388
1389 $doc_obj->associate_file($filename, $assocfilename, undef, $cursection);
1390 # srclink_file is now deprecated because of the "_" in the metadataname. Use srclinkFile
1391 $doc_obj->add_utf8_metadata ($cursection, "srclink_file", $doc_obj->get_sourcefile());
1392 $doc_obj->add_utf8_metadata ($cursection, "srclinkFile", $doc_obj->get_sourcefile());
1393}
1394
1395sub associate_cover_image {
1396 my $self = shift(@_);
1397 my ($doc_obj, $filename) = @_;
1398
1399 my $upgraded_filename = &util::upgrade_if_dos_filename($filename);
1400
1401 $filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1402 $upgraded_filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1403
1404 if (exists $self->{'covers_missing_cache'}->{$upgraded_filename}) {
1405 # don't stat() for existence e.g. for multiple document input files
1406 # (eg SplitPlug)
1407 return;
1408 }
1409
1410 my $top_section=$doc_obj->get_top_section();
1411
1412 if (&util::fd_exists($upgraded_filename)) {
1413 $doc_obj->associate_source_file($filename);
1414 $doc_obj->associate_file($filename, "cover.jpg", "image/jpeg");
1415 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1416 } else {
1417 my $upper_filename = $filename;
1418 my $upgraded_upper_filename = $upgraded_filename;
1419
1420 $upper_filename =~ s/jpg$/JPG/;
1421 $upgraded_upper_filename =~ s/jpg$/JPG/;
1422
1423 if (&util::fd_exists($upgraded_upper_filename)) {
1424 $doc_obj->associate_source_file($upper_filename);
1425 $doc_obj->associate_file($upper_filename, "cover.jpg",
1426 "image/jpeg");
1427 $doc_obj->add_utf8_metadata($top_section, "hascover", 1);
1428 } else {
1429 # file doesn't exist, so record the fact that it's missing so
1430 # we don't stat() again (stat is slow)
1431 $self->{'covers_missing_cache'}->{$upgraded_filename} = 1;
1432 }
1433 }
1434
1435}
1436
1437
1438# Overridden by exploding plugins (eg. ISISPlug)
1439sub clean_up_after_exploding
1440{
1441 my $self = shift(@_);
1442}
1443
1444
1445
14461;
Note: See TracBrowser for help on using the repository browser.