root/main/trunk/greenstone2/perllib/plugins/BasePlugin.pm @ 21742

Revision 21742, 36.4 KB (checked in by kjdon, 10 years ago)

should be using assocfilepath instead of archivedir
  • Property svn:keywords set to Author Date Id Revision
Line 
1###########################################################################
2#
3# BasePlugin.pm -- base class for all the import plugins
4# A component of the Greenstone digital library software
5# from the New Zealand Digital Library Project at the
6# University of Waikato, New Zealand.
7#
8# Copyright (C) 1999-2005 New Zealand Digital Library Project
9#
10# This program is free software; you can redistribute it and/or modify
11# it under the terms of the GNU General Public License as published by
12# the Free Software Foundation; either version 2 of the License, or
13# (at your option) any later version.
14#
15# This program is distributed in the hope that it will be useful,
16# but WITHOUT ANY WARRANTY; without even the implied warranty of
17# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
18# GNU General Public License for more details.
19#
20# You should have received a copy of the GNU General Public License
21# along with this program; if not, write to the Free Software
22# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
23#
24###########################################################################
25
26package BasePlugin;
27
28use strict;
29no strict 'subs';
30no strict 'refs'; # allow filehandles to be variables and viceversa
31
32use File::Basename;
33
34use encodings;
35use unicode;
36use textcat;
37use doc;
38use ghtml;
39use gsprintf 'gsprintf';
40
41use PrintInfo;
42
43BEGIN {
44    @BasePlugin::ISA = ( 'PrintInfo' );
45}
46
47# the different methods that can be applied when renaming
48# imported documents and their associated files
49our $file_rename_method_list =
50    [ { 'name' => "url",
51    'desc' => "{BasePlugin.rename_method.url}" },
52      { 'name' => "base64",
53    'desc' => "{BasePlugin.rename_method.base64}" },
54      { 'name' => "none",
55    'desc' => "{BasePlugin.rename_method.none}",
56    'hiddengli' => "yes" } ];
57
58our $encoding_list =
59    [ { 'name' => "ascii",
60    'desc' => "{BasePlugin.encoding.ascii}" },
61      { 'name' => "utf8",
62    'desc' => "{BasePlugin.encoding.utf8}" },
63      { 'name' => "unicode",
64    'desc' => "{BasePlugin.encoding.unicode}" } ];
65     
66
67my $e = $encodings::encodings;
68foreach my $enc (sort {$e->{$a}->{'name'} cmp $e->{$b}->{'name'}} keys (%$e))
69{
70    my $hashEncode =
71    {'name' => $enc,
72     'desc' => $e->{$enc}->{'name'}};
73   
74    push(@{$encoding_list},$hashEncode);
75}
76
77our $encoding_plus_auto_list =
78    [ { 'name' => "auto",
79    'desc' => "{BasePlugin.filename_encoding.auto}" },
80       { 'name' => "auto-language-analysis",
81    'desc' => "{BasePlugin.filename_encoding.auto_language_analysis}" }, # textcat
82      { 'name' => "auto-filesystem-encoding",
83    'desc' => "{BasePlugin.filename_encoding.auto_filesystem_encoding}" }, # locale
84      { 'name' => "auto-fl",
85    'desc' => "{BasePlugin.filename_encoding.auto_fl}" }, # locale followed by textcat
86      { 'name' => "auto-lf",
87    'desc' => "{BasePlugin.filename_encoding.auto_lf}" } ]; # texcat followed by locale
88
89push(@{$encoding_plus_auto_list},@{$encoding_list});
90
91our $oidtype_list =
92    [ { 'name' => "auto",
93    'desc' => "{BasePlugin.OIDtype.auto}" },
94      { 'name' => "hash",
95        'desc' => "{import.OIDtype.hash}" },
96      { 'name' => "assigned",
97        'desc' => "{import.OIDtype.assigned}" },
98      { 'name' => "incremental",
99        'desc' => "{import.OIDtype.incremental}" },
100      { 'name' => "dirname",
101        'desc' => "{import.OIDtype.dirname}" } ];
102
103my $arguments =
104    [ { 'name' => "process_exp",
105    'desc' => "{BasePlugin.process_exp}",
106    'type' => "regexp",
107    'deft' => "",
108    'reqd' => "no" },
109      { 'name' => "no_blocking",
110    'desc' => "{BasePlugin.no_blocking}",
111    'type' => "flag",
112    'reqd' => "no"},
113      { 'name' => "block_exp",
114    'desc' => "{BasePlugin.block_exp}",
115    'type' => "regexp",
116    'deft' => "",
117    'reqd' => "no" },
118      { 'name' => "associate_ext",
119    'desc' => "{BasePlugin.associate_ext}",
120    'type' => "string",
121    'reqd' => "no" },
122      { 'name' => "associate_tail_re",
123    'desc' => "{BasePlugin.associate_tail_re}",
124    'type' => "string",
125    'reqd' => "no" },
126      { 'name' => "OIDtype",
127    'desc' => "{import.OIDtype}",
128    'type' => "enum",
129    'list' => $oidtype_list,
130    # leave default empty so we can tell if its been set or not - if not set will use option from import.pl
131    'deft' => "auto",
132    'reqd' => "no" },
133      { 'name' => "OIDmetadata",
134    'desc' => "{import.OIDmetadata}",
135    'type' => "metadata",
136    'deft' => "dc.Identifier",
137    'reqd' => "no" },     
138#      { 'name' => "use_as_doc_identifier",
139#   'desc' => "{BasePlugin.use_as_doc_identifier}",
140#   'type' => "string",
141#   'reqd' => "no" ,
142#   'deft' => "" } ,
143      { 'name' => "no_cover_image",
144    'desc' => "{BasePlugin.no_cover_image}",
145    'type' => "flag",
146    'reqd' => "no" },
147      { 'name' => "filename_encoding",
148    'desc' => "{BasePlugin.filename_encoding}",
149    'type' => "enum",
150    'deft' => "auto",
151    'list' => $encoding_plus_auto_list,
152    'reqd' => "no" },
153      { 'name' => "smart_block",
154        'desc' => "{common.deprecated}. {BasePlugin.smart_block}",
155        'type' => "flag",
156        'reqd' => "no",
157    'hiddengli' => "yes" }, # deprecated, but leave in for old collections
158      { 'name' => "file_rename_method",
159    'desc' => "{BasePlugin.file_rename_method}",
160    'type' => "enum",
161    'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
162    'list' => $file_rename_method_list,
163    'reqd' => "no"
164      }
165     
166      ];
167
168
169my $options = { 'name'     => "BasePlugin",
170        'desc'     => "{BasePlugin.desc}",
171        'abstract' => "yes",
172        'inherits' => "no",
173        'args'     => $arguments };
174
175sub new {
176
177    my ($class) = shift (@_);
178    my ($pluginlist,$inputargs,$hashArgOptLists,$auxiliary) = @_;
179    push(@$pluginlist, $class);
180
181    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
182    push(@{$hashArgOptLists->{"OptList"}},$options);
183
184    my $self = new PrintInfo($pluginlist, $inputargs, $hashArgOptLists,$auxiliary);
185   
186    if ($self->{'info_only'}) {
187        # don't worry about any options etc
188        return bless $self, $class;
189    }
190
191    if ($self->{'smart_block'}) {
192    print STDERR "WARNING: -smart_block option has been deprecated and is no longer useful\n";
193    }
194    $self->{'smart_block'} = undef;
195
196    my $plugin_name = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
197    $self->{'plugin_type'} = $plugin_name;
198
199    # remove ex. from OIDmetadata
200    $self->{'OIDmetadata'} =~ s/^ex\.// if defined $self->{'OIDmetadata'};
201    $self->{'num_processed'} = 0;
202    $self->{'num_not_processed'} = 0;
203    $self->{'num_blocked'} = 0;
204    $self->{'num_archives'} = 0;
205    $self->{'cover_image'} = 1; # cover image is on by default
206    $self->{'cover_image'} = 0 if ($self->{'no_cover_image'});
207    $self->{'can_process_directories'} = 0;
208    #$self->{'option_list'} = $hashArgOptLists->{"OptList"};
209   
210    my $associate_ext = $self->{'associate_ext'};
211    if ((defined $associate_ext) && ($associate_ext ne "")) {
212
213    my $associate_tail_re = $self->{'associate_tail_re'};
214    if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
215        my $outhandle = $self->{'outhandle'};
216        print $outhandle "Warning: can only specify 'associate_ext' or 'associate_tail_re'\n";
217        print $outhandle "         defaulting to 'associate_tail_re'\n";
218    }
219    else {
220        my @exts = split(/,/,$associate_ext);
221
222        my @exts_bracketed = map { $_ = "(?:\\.$_)" } @exts;
223        my $associate_tail_re = join("|",@exts_bracketed);
224        $self->{'associate_tail_re'} = $associate_tail_re;
225    }
226
227    delete $self->{'associate_ext'};
228    }
229
230    return bless $self, $class;
231
232}
233
234sub merge_inheritance
235{
236    my $self = {};
237    my @child_selfs = @_;
238
239    foreach my $child_self (@child_selfs) {
240    foreach my $key (keys %$child_self) {
241        if (defined $self->{$key}) {
242        if ($self->{$key} ne $child_self->{$key}) {
243#           print STDERR "Warning: Conflicting value in multiple inheritance for '$key'\n";
244#           print STDERR "Existing stored value = $self->{$key}\n";
245#           print STDERR "New (child) value     = $child_self->{$key}\n";
246#           print STDERR "Keeping existing value\n";
247            # Existing value seems to be option specified in collect.cfg
248
249            ### $self->{$key} = $child_self->{$key};
250           
251        }
252        else {
253##          print STDERR "****Info: Value $self->{$key} for $key already defined through multiple inheritance as the same value\n";
254        }
255
256        }
257        else {
258        $self->{$key} = $child_self->{$key};
259        }
260    }
261    }
262
263    return $self;   
264}
265
266# initialize BasePlugin options
267# if init() is overridden in a sub-class, remember to call BasePlugin::init()
268sub init {
269    my $self = shift (@_);
270    my ($verbosity, $outhandle, $failhandle) = @_;
271
272    # verbosity is passed through from the processor
273    $self->{'verbosity'} = $verbosity;
274
275    # as are the outhandle and failhandle
276    $self->{'outhandle'} = $outhandle if defined $outhandle;
277    $self->{'failhandle'} = $failhandle;
278#    $self->SUPER::init(@_);
279   
280    # set process_exp and block_exp to defaults unless they were
281    # explicitly set
282
283    if ((!$self->is_recursive()) and
284    (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
285
286    $self->{'process_exp'} = $self->get_default_process_exp ();
287    if ($self->{'process_exp'} eq "") {
288        warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
289    }
290    }
291
292    if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
293    $self->{'block_exp'} = $self->get_default_block_exp ();
294    }
295
296}
297
298sub begin {
299    my $self = shift (@_);
300    my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
301
302    if ($self->{'OIDtype'} eq "auto") {
303    # hasn't been set in the plugin, use the processor values
304    $self->{'OIDtype'} = $processor->{'OIDtype'};
305    $self->{'OIDmetadata'} = $processor->{'OIDmetadata'};
306    }
307    if ($self->{'OIDtype'} eq "hash") {
308    # should we hash on the file or on the doc xml??
309    $self->{'OIDtype'} = $self->get_oid_hash_type();
310    if ($self->{'OIDtype'} !~ /^(hash_on_file|hash_on_ga_xml)$/) {
311        $self->{'OIDtype'} = "hash_on_file";
312    }
313    }
314}
315
316# This is called once if removeold is set with import.pl. Most plugins will do
317# nothing but if a plugin does any stuff outside of creating doc obj, then
318# it may need to clear something.
319sub remove_all {
320    my $self = shift (@_);
321    my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
322}
323
324# This is called per document for docs that have been deleted from the
325# collection. Most plugins will do nothing
326# but if a plugin does any stuff outside of creating doc obj, then it may need
327# to clear something.
328sub remove_one {
329    my $self = shift (@_);
330   
331    my ($file, $oids, $archivedir) = @_;
332    return 0 if $self->can_process_this_file($file);
333    return undef;
334}
335
336sub end {
337    # potentially called at the end of each plugin pass
338    # import.pl only has one plugin pass, but buildcol.pl has multiple ones
339
340    my ($self) = shift (@_);
341}
342
343sub deinit {
344    # called only once, after all plugin passes have been done
345
346    my ($self) = @_;
347}
348
349# default hashing type is to hash on the original file (or converted file)
350# override this to return hash_on_ga_xml for filetypes where hashing on the
351# file is no good eg video
352sub get_oid_hash_type {
353
354    my $self = shift (@_);
355
356    return "hash_on_file";
357}
358
359
360# this function should be overridden to return 1
361# in recursive plugins
362sub is_recursive {
363    my $self = shift (@_);
364
365    return 0;
366}
367
368sub get_default_block_exp {
369    my $self = shift (@_);
370
371    return "";
372}
373
374sub get_default_process_exp {
375    my $self = shift (@_);
376
377    return "";
378}
379
380# rename imported files and assoc files using URL encoding by default
381# as this will work for most plugins and give more legible filenames
382sub get_default_file_rename_method() {
383    my $self = shift (@_);
384    return "url";
385}
386
387# returns this plugin's active (possibly user-selected) file_rename_method
388sub get_file_rename_method() {
389    my $self = shift (@_);
390    my $rename_method = $self->{'file_rename_method'};
391    if($rename_method) {
392    return $rename_method;
393    } else {   
394    return $self->get_default_file_rename_method();
395    }
396}
397
398# default implementation is to do nothing
399sub store_block_files {
400   
401    my $self =shift (@_);
402    my ($filename_full_path, $block_hash) = @_;
403
404}
405
406# put files to block into hash
407sub use_block_expressions {
408
409    my $self =shift (@_);
410    my ($filename_full_path, $block_hash) = @_;
411
412    if ($self->{'block_exp'} ne "" && $filename_full_path =~ /$self->{'block_exp'}/) {
413    $block_hash->{'file_blocks'}->{$filename_full_path} = 1;
414    }
415
416}
417
418#default implementation is to block a file with same name as this, but extension jpg or JPG, if cover_images is on.
419sub block_cover_image
420{
421    my $self =shift;
422    my ($filename, $block_hash) = @_;
423
424    if ($self->{'cover_image'}) {
425    my $coverfile = $filename;
426    $coverfile =~ s/\.[^\\\/\.]+$/\.jpg/;
427    if (!-e $coverfile) {
428        $coverfile =~ s/jpg$/JPG/;
429    }   
430    if (-e $coverfile) {
431        $block_hash->{'file_blocks'}->{$coverfile} = 1;
432    }
433    }
434
435    return;
436}
437
438
439# discover all the files that should be blocked by this plugin
440# check the args ...
441sub file_block_read {
442
443    my $self = shift (@_); 
444    my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $gli) = @_;
445    # Keep track of filenames with same root but different extensions
446    # Used to support -associate_ext and the more generalised
447    # -associate_tail_re
448    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
449
450    if (!-d $filename_full_path) {
451    $block_hash->{'all_files'}->{$file} = 1;
452    }
453
454    my $associate_tail_re = $self->{'associate_tail_re'};
455    if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
456    my ($file_prefix,$file_ext)
457        = &util::get_prefix_and_tail_by_regex($filename_full_path,$associate_tail_re);
458    if ((defined $file_prefix) && (defined $file_ext)) {
459        my $shared_fileroot = $block_hash->{'shared_fileroot'};
460        if (!defined $shared_fileroot->{$file_prefix}) {
461        my $file_prefix_rec = { 'tie_to'  => undef,
462                        'exts'    => {} };
463        $shared_fileroot->{$file_prefix} = $file_prefix_rec;
464        }
465       
466        my $file_prefix_rec = $shared_fileroot->{$file_prefix};
467
468        if ($self->can_process_this_file($filename_full_path)) {
469        # This is the document the others should be tied to
470        $file_prefix_rec->{'tie_to'} = $file_ext;
471        }
472        else {
473        if ($file_ext =~ m/$associate_tail_re$/) {
474            # this file should be associated to the main one
475            $file_prefix_rec->{'exts'}->{$file_ext} = 1;
476        }
477        }
478
479    }
480    }
481
482    # check block expressions
483    $self->use_block_expressions($filename_full_path, $block_hash) unless $self->{'no_blocking'};
484
485    # now check whether we are actually processing this
486    if (!-f $filename_full_path || !$self->can_process_this_file($filename_full_path)) {
487    return undef; # can't recognise
488    }
489   
490    # if we have a block_exp, then this overrides the normal 'smart' blocking
491    $self->store_block_files($filename_full_path, $block_hash) unless ($self->{'no_blocking'} || $self->{'block_exp'} ne "");
492
493    # block the cover image if there is one
494    if ($self->{'cover_image'}) {
495    $self->block_cover_image($filename_full_path, $block_hash);
496    }
497       
498    return 1;
499}
500
501# plugins that rely on more than process_exp (eg XML plugins) can override this method
502sub can_process_this_file {
503    my $self = shift(@_);
504    my ($filename) = @_;
505
506    if (-d $filename && !$self->{'can_process_directories'}) {
507    return 0;
508    }
509    if ($self->{'process_exp'} ne "" && $filename =~ /$self->{'process_exp'}/) {
510    return 1;
511    }
512    return 0;
513   
514}
515
516# just converts path as is to utf8.
517sub filepath_to_utf8 {
518    my $self = shift (@_); 
519    my ($file, $file_encoding) = @_;
520    my $filemeta = $file;
521
522    my $filename_encoding = $self->{'filename_encoding'}; # filename encoding setting
523
524    # Whenever filename-encoding is set to any of the auto settings, we
525    # check if the filename is already in UTF8. If it is, then we're done.
526    if($filename_encoding =~ m/auto/) {
527    if(&unicode::check_is_utf8($filemeta))
528    {
529        $filename_encoding = "utf8";
530        return $filemeta;
531    }
532    }
533   
534    # Auto setting, but filename is not utf8
535    if ($filename_encoding eq "auto")
536    {
537    # try textcat
538    $filename_encoding = $self->textcat_encoding($filemeta);
539   
540    # check the locale next
541    $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
542   
543
544    # now try the encoding of the document, if available
545    if ($filename_encoding eq "undefined" && defined $file_encoding) {
546        $filename_encoding = $file_encoding;
547    }
548
549    }
550
551    elsif ($filename_encoding eq "auto-language-analysis")
552    {   
553    $filename_encoding = $self->textcat_encoding($filemeta);
554
555    # now try the encoding of the document, if available
556    if ($filename_encoding eq "undefined" && defined $file_encoding) {
557        $filename_encoding = $file_encoding;
558    }
559    }
560
561    elsif ($filename_encoding eq "auto-filesystem-encoding")
562    {
563    # try locale
564    $filename_encoding = $self->locale_encoding();
565    }
566
567    elsif ($filename_encoding eq "auto-fl")
568    {
569    # filesystem-encoding (locale) then language-analysis (textcat)
570    $filename_encoding = $self->locale_encoding();
571   
572    # try textcat
573    $filename_encoding = $self->textcat_encoding($filemeta) if $filename_encoding eq "undefined";
574       
575    # else assume filename encoding is encoding of file content, if that's available
576    if ($filename_encoding eq "undefined" && defined $file_encoding) {
577        $filename_encoding = $file_encoding;
578    }
579    }
580   
581    elsif ($filename_encoding eq "auto-lf")
582    {
583    # language-analysis (textcat) then filesystem-encoding (locale)
584    $filename_encoding = $self->textcat_encoding($filemeta);
585   
586    # guess filename encoding from encoding of file content, if available
587    if ($filename_encoding eq "undefined" && defined $file_encoding) {
588        $filename_encoding = $file_encoding;
589    }
590
591    # try locale
592    $filename_encoding = $self->locale_encoding() if $filename_encoding eq "undefined";
593    }
594       
595    # if still undefined, use utf8 as fallback
596    if ($filename_encoding eq "undefined") {
597    $filename_encoding = "utf8";
598    }
599
600    #print STDERR "**** UTF8 encoding the filename $filemeta ";
601   
602    # if the filename encoding is set to utf8 but it isn't utf8 already--such as when
603    # 1. the utf8 fallback is used, or 2. if the system locale is used and happens to
604    # be always utf8 (in which case the filename's encoding is also set as utf8 even
605    # though the filename need not be if it originates from another system)--in such
606    # cases attempt to make the filename utf8 to match.
607    if($filename_encoding eq "utf8" && !&unicode::check_is_utf8($filemeta)) {
608    &unicode::ensure_utf8(\$filemeta);
609    }
610
611    # convert non-unicode encodings to utf8
612    if ($filename_encoding !~ m/(?:ascii|utf8|unicode)/) {
613    $filemeta = &unicode::unicode2utf8(
614      &unicode::convert2unicode($filename_encoding, \$filemeta)
615    );
616    }
617
618    #print STDERR " from encoding $filename_encoding -> $filemeta\n";
619    return $filemeta;
620}
621
622# gets the filename with no path, converts to utf8, and then dm safes it.
623# filename_encoding set by user
624sub filename_to_utf8_metadata
625{
626    my $self = shift (@_); 
627    my ($file, $file_encoding) = @_;
628
629    my $outhandle = $self->{'outhandle'};
630
631    my ($filemeta) = $file =~ /([^\\\/]+)$/; # getting the tail of the filepath (skips all string parts containing slashes upto the end)
632    $filemeta = $self->filepath_to_utf8($filemeta, $file_encoding);
633   
634    my $dmsafe_filemeta = &ghtml::dmsafe($filemeta);
635
636    return $dmsafe_filemeta;
637
638}
639
640sub locale_encoding {
641    my $self = shift(@_);
642   
643    if (!defined $self->{'filesystem_encoding'}) {
644    $self->{'filesystem_encoding'} = $self->get_filesystem_encoding();
645    $self->{'filesystem_encoding'} = "undefined" if !defined $self->{'filesystem_encoding'};
646    }
647
648    #print STDERR "*** filename encoding determined based on locale: " . $self->{'filesystem_encoding'} . "\n";
649    return $self->{'filesystem_encoding'}; # can be the string "undefined"
650}
651
652sub textcat_encoding {
653    my $self = shift(@_);
654    my ($filemeta) = @_;
655
656    # analyse filenames without extensions and digits (and trimmed of surrounding
657    # whitespace), so that irrelevant chars don't confuse textcat
658    my $strictfilemeta = $filemeta;
659    $strictfilemeta =~ s/\.[^\.]+$//g;
660    $strictfilemeta =~ s/\d//g;
661    $strictfilemeta =~ s/^\s*//g;
662    $strictfilemeta =~ s/\s*$//g;
663   
664    my $filename_encoding = $self->encoding_from_language_analysis($strictfilemeta);
665    if(!defined $filename_encoding) {
666    $filename_encoding = "undefined";
667    }
668
669    return $filename_encoding; # can be the string "undefined"
670}
671
672# performs textcat
673sub encoding_from_language_analysis {
674    my $self = shift(@_);
675    my ($text) = @_;
676
677    my $outhandle = $self->{'outhandle'};
678    my $best_encoding = undef;
679   
680    # get the language/encoding of the textstring using textcat
681    $self->{'textcat'} = new textcat() unless defined($self->{'textcat'});
682    my $results = $self->{'textcat'}->classify_cached_filename(\$text);
683
684
685    if (scalar @$results < 0) {
686    return undef;
687    }
688   
689    # We have some results, we choose the first
690    my ($language, $encoding) = $results->[0] =~ /^([^-]*)(?:-(.*))?$/;
691   
692    $best_encoding = $encoding;
693    if (!defined $best_encoding) {
694    return undef;
695    }
696       
697    if (defined $best_encoding && $best_encoding =~ m/^iso_8859/ && &unicode::check_is_utf8($text)) {
698    # the text is valid utf8, so assume that's the real encoding (since textcat is based on probabilities)
699    $best_encoding = 'utf8';
700    }
701   
702   
703    # check for equivalents where textcat doesn't have some encodings...
704    # eg MS versions of standard encodings
705    if (defined $best_encoding && $best_encoding =~ /^iso_8859_(\d+)/) {
706    my $iso = $1; # which variant of the iso standard?
707    # iso-8859 sets don't use chars 0x80-0x9f, windows codepages do
708    if ($text =~ /[\x80-\x9f]/) {
709        # Western Europe
710        if ($iso == 1 or $iso == 15) { $best_encoding = 'windows_1252' }
711        elsif ($iso == 2) {$best_encoding = 'windows_1250'} # Central Europe
712        elsif ($iso == 5) {$best_encoding = 'windows_1251'} # Cyrillic
713        elsif ($iso == 6) {$best_encoding = 'windows_1256'} # Arabic
714        elsif ($iso == 7) {$best_encoding = 'windows_1253'} # Greek
715        elsif ($iso == 8) {$best_encoding = 'windows_1255'} # Hebrew
716        elsif ($iso == 9) {$best_encoding = 'windows_1254'} # Turkish
717    }
718    }
719   
720    if (defined $best_encoding && $best_encoding !~ /^(ascii|utf8|unicode)$/ &&
721    !defined $encodings::encodings->{$best_encoding})
722    {
723    if ($self->{'verbosity'}) {
724        gsprintf($outhandle, "BasePlugin: {ReadTextFile.unsupported_encoding}\n", $text, $best_encoding, "undef");
725    }
726    $best_encoding = undef;
727    }
728   
729    return $best_encoding;
730}
731
732# uses locale
733sub get_filesystem_encoding {
734
735    my $self = shift(@_);
736
737    my $outhandle = $self->{'outhandle'};
738    my $filesystem_encoding = undef;
739
740    eval {
741    use POSIX qw(locale_h);
742   
743    # With only one parameter, setlocale retrieves the
744    # current value
745    my $current_locale = setlocale(LC_CTYPE);
746   
747    if ($current_locale =~ m/^.*\.(.*?)$/) {
748        my $char_encoding = lc($1);
749        if ($char_encoding =~ m/^(iso)(8859)(\d{1,2})$/) {
750        $char_encoding = "$1\_$2\_$3";
751        }
752
753        $char_encoding =~ s/-/_/g;
754        $char_encoding =~ s/^utf_8$/utf8/;
755       
756        if ($char_encoding =~ m/^\d+$/) {
757        if (defined $encodings::encodings->{"windows_$char_encoding"}) {
758            $char_encoding = "windows_$char_encoding";
759        }
760        elsif (defined $encodings::encodings->{"dos_$char_encoding"}) {
761            $char_encoding = "dos_$char_encoding";
762        }
763        }
764       
765        if (($char_encoding =~ m/(?:ascii|utf8|unicode)/)
766        || (defined $encodings::encodings->{$char_encoding})) {
767        $filesystem_encoding = $char_encoding;
768        }
769        else {
770        print $outhandle "Warning: Unsupported character encoding '$char_encoding' from locale '$current_locale'\n";
771        }
772    }
773   
774
775    };
776    if ($@) {
777    print $outhandle "$@\n";
778    print $outhandle "Warning: Unable to establish locale.  Will assume filesystem is UTF-8\n";
779   
780    }
781    return $filesystem_encoding;
782}
783
784# is there ever only one Source? Sometimes this will be called twice, for images etc that are converted.
785sub set_Source_metadata {
786    my $self = shift (@_); 
787    my ($doc_obj, $filename_no_path, $file_encoding) = @_;
788
789    my $top_section = $doc_obj->get_top_section();
790   
791    # UTF-8 version of filename
792    my $filemeta = $self->filename_to_utf8_metadata($filename_no_path, $file_encoding);
793   
794    # Source is the UTF8 display name - not necessarily the name of the file on the system
795    $doc_obj->set_utf8_metadata_element($top_section, "Source", $filemeta);
796
797    $filemeta = &util::rename_file($filemeta, $self->{'file_rename_method'});
798    # If using URL encoding, then SourceFile is the url-reference to url-encoded
799    # filemeta: it's a url that refers to the actual file on the system
800    $filemeta = &unicode::filename_to_url($filemeta);
801
802    $doc_obj->set_utf8_metadata_element($top_section, "SourceFile", $filemeta);
803}
804   
805# this should be called by all plugins to set the oid of the doc obj, rather
806# than calling doc_obj->set_OID directly
807sub add_OID {
808    my $self = shift (@_); 
809    my ($doc_obj) = @_;
810
811    $doc_obj->set_OIDtype($self->{'OIDtype'}, $self->{'OIDmetadata'});
812
813    # see if there is a plugin specific set_OID function
814    if (defined ($self->can('set_OID'))) {
815    $self->set_OID(@_); # pass through doc_obj and any extra arguments
816    }
817    else {
818    # use the default set_OID() in doc.pm
819    $doc_obj->set_OID();
820    }
821
822}
823 
824# The BasePlugin read_into_doc_obj() function. This function does all the
825# right things to make general options work for a given plugin.  It doesn't do anything with the file other than setting reads in
826# a file and sets up a slew of metadata all saved in doc_obj, which
827# it then returns as part of a tuple (process_status,doc_obj)
828#
829# Much of this functionality used to reside in read, but it was broken
830# down into a supporting routine to make the code more flexible. 
831#
832# recursive plugins (e.g. RecPlug) and specialized plugins like those
833# capable of processing many documents within a single file (e.g.
834# GMLPlug) will normally want to implement their own version of
835# read_into_doc_obj()
836#
837# Note that $base_dir might be "" and that $file might
838# include directories
839
840# currently blocking has been done before it gets here - does this affect secondary plugin stuff??
841sub read_into_doc_obj {
842    my $self = shift (@_); 
843    my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
844
845    my $outhandle = $self->{'outhandle'};
846
847    # should we move this to read? What about secondary plugins?
848    print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
849    print $outhandle "$self->{'plugin_type'} processing $file\n"
850        if $self->{'verbosity'} > 1;
851
852    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
853
854    # create a new document
855    my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
856    my $top_section = $doc_obj->get_top_section();
857
858    $doc_obj->add_utf8_metadata($top_section, "Plugin", "$self->{'plugin_type'}");
859    $doc_obj->add_utf8_metadata($top_section, "FileSize", (-s $filename_full_path));
860 
861
862    # sets the UTF8 filename (Source) for display and sets the url ref to URL encoded version
863    # of the UTF8 filename (SourceFile) for generated files
864    $self->set_Source_metadata($doc_obj, $filename_no_path);
865
866
867    # plugin specific stuff - what args do we need here??
868    unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
869    print STDERR "<ProcessingError n='$file'>\n" if ($gli);
870    return -1;
871    }
872   
873    # include any metadata passed in from previous plugins
874    # note that this metadata is associated with the top level section
875    my $section = $doc_obj->get_top_section();
876    # can we merge these two methods??
877    $self->add_associated_files($doc_obj, $filename_full_path);
878    $self->extra_metadata ($doc_obj, $section, $metadata);
879    $self->auto_extract_metadata($doc_obj);
880
881    # if we haven't found any Title so far, assign one
882    # this was shifted to here from inside read()
883    $self->title_fallback($doc_obj,$section,$filename_no_path);
884   
885    $self->add_OID($doc_obj);
886   
887    $self->post_process_doc_obj($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli);
888    return (1,$doc_obj);
889}
890
891sub post_process_doc_obj {
892    my $self = shift (@_); 
893    my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
894
895    return 1;
896}
897
898sub add_dummy_text {
899    my $self = shift(@_);
900    my ($doc_obj, $section) = @_;
901
902    # add NoText metadata so we can hide this dummy text in format statements
903    $doc_obj->add_metadata($section, "NoText", "1");
904    $doc_obj->add_text($section, &gsprintf::lookup_string("{BasePlugin.dummy_text}",1));
905   
906}
907
908# does nothing. Can be overridden by subclass
909sub auto_extract_metadata {
910    my $self = shift(@_);
911    my ($doc_obj) = @_;
912}
913
914# adds cover image, associate_file options stuff. Should be called by sub class
915# read_into_doc_obj
916sub add_associated_files {
917    my $self = shift(@_);
918    # whatis filename??
919    my ($doc_obj, $filename) = @_;
920   
921    # add in the cover image
922    if ($self->{'cover_image'}) {
923    $self->associate_cover_image($doc_obj, $filename);
924    }
925   
926
927}
928
929# implement this if you are extracting metadata for other documents
930sub metadata_read {
931    my $self = shift (@_);
932    my ($pluginfo, $base_dir, $file, $block_hash,
933    $extrametakeys, $extrametadata, $extrametafile,
934    $processor, $maxdocs, $gli) = @_;
935   
936    # can we process this file??
937    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
938    return undef unless $self->can_process_this_file($filename_full_path);
939
940    return 1; # we recognise the file, but don't actually do anything with it
941}
942
943
944# The BasePlugin read() function. This function calls read_into_doc_obj()
945# to ensure all the right things to make general options work for a
946# given plugin are done. It then calls the process() function which
947# does all the work specific to a plugin (like the old read functions
948# used to do). Most plugins should define their own process() function
949# and let this read() function keep control. 
950#
951# recursive plugins (e.g. RecPlug) and specialized plugins like those
952# capable of processing many documents within a single file (e.g.
953# GMLPlug) might want to implement their own version of read(), but
954# more likely need to implement their own version of read_into_doc_obj()
955#
956# Return number of files processed, undef if can't recognise, -1 if can't
957# process
958
959sub read {
960    my $self = shift (@_); 
961    my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
962
963    # can we process this file??
964    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
965
966    return undef unless $self->can_process_this_file($filename_full_path);
967   
968    my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
969   
970    if ((defined $process_status) && ($process_status == 1)) {
971   
972    # process the document
973    $processor->process($doc_obj);
974
975    $self->{'num_processed'} ++;
976    undef $doc_obj;
977    }
978    # delete any temp files that we may have created
979    $self->clean_up_after_doc_obj_processing();
980
981
982    # if process_status == 1, then the file has been processed.
983    return $process_status;
984
985}
986
987# returns undef if file is rejected by the plugin
988sub process {
989    my $self = shift (@_);
990    my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
991
992    gsprintf(STDERR, "BasePlugin::process {common.must_be_implemented}\n") && die "\n";
993
994    return undef; # never gets here
995}
996
997# overwrite this method to delete any temp files that we have created
998sub clean_up_after_doc_obj_processing {
999    my $self = shift(@_);
1000
1001}
1002
1003# write_file -- used by ConvertToPlug, for example in post processing
1004#
1005# where should this go, is here the best place??
1006sub utf8_write_file {
1007    my $self = shift (@_);
1008    my ($textref, $filename) = @_;
1009   
1010    if (!open (FILE, ">$filename")) {
1011    gsprintf(STDERR, "ConvertToPlug::write_file {ConvertToPlug.could_not_open_for_writing} ($!)\n", $filename);
1012     die "\n";
1013     }
1014    print FILE $$textref;
1015   
1016    close FILE;
1017}
1018
1019
1020sub filename_based_title
1021{
1022    my $self = shift (@_);
1023    my ($file) = @_;
1024
1025    my $file_derived_title = $file;
1026    $file_derived_title =~ s/_/ /g;
1027    $file_derived_title =~ s/\..*?$//;
1028
1029    return $file_derived_title;
1030}
1031
1032
1033sub title_fallback
1034{
1035    my $self = shift (@_);
1036    my ($doc_obj,$section,$file) = @_;
1037
1038    if (!defined $doc_obj->get_metadata_element ($section, "Title") or $doc_obj->get_metadata_element($section, "Title") eq "") {
1039
1040    my $file_derived_title = $self->filename_to_utf8_metadata($self->filename_based_title($file));
1041    if (!defined $doc_obj->get_metadata_element ($section, "Title")) {
1042        $doc_obj->add_utf8_metadata ($section, "Title", $file_derived_title);
1043    }
1044    else {
1045        $doc_obj->set_utf8_metadata_element ($section, "Title", $file_derived_title);
1046    }
1047    }
1048       
1049}
1050 
1051# add any extra metadata that's been passed around from one
1052# plugin to another.
1053# extra_metadata uses add_utf8_metadata so it expects metadata values
1054# to already be in utf8
1055sub extra_metadata {
1056    my $self = shift (@_);
1057    my ($doc_obj, $cursection, $metadata) = @_;
1058
1059    my $associate_tail_re = $self->{'associate_tail_re'};
1060
1061    foreach my $field (keys(%$metadata)) {
1062    # $metadata->{$field} may be an array reference
1063    if ($field eq "gsdlassocfile_tobe") {
1064        # 'gsdlassocfile_tobe' is artificially introduced metadata
1065        # that is used to signal that certain additional files should
1066        # be tied to this document.  Useful in situations where a
1067        # metadata pass in the plugin pipeline works out some files
1068        # need to be associated with a document, but the document hasn't
1069        # been formed yet.
1070        my $equiv_form = "";
1071        foreach my $gaf (@{$metadata->{$field}}) {
1072        my ($full_filename,$mimetype) = ($gaf =~ m/^(.*):(.*):$/);
1073        my ($tail_filename) = ($full_filename =~ /^.*[\/\\](.+?)$/);
1074       
1075        # we need to make sure the filename is valid utf-8 - we do
1076        # this by url or base64 encoding it
1077        # $tail_filename is the name that we store the file as
1078        $tail_filename = &util::rename_file($tail_filename, $self->{'file_rename_method'});
1079        $doc_obj->associate_file($full_filename,$tail_filename,$mimetype);
1080        $doc_obj->associate_source_file($full_filename);
1081        # If the filename is url_encoded, we need to encode the % signs
1082        # in the filename, so that it works in a url
1083        my $url_tail_filename = &unicode::filename_to_url($tail_filename);
1084        # work out extended tail extension (i.e. matching tail re)
1085
1086        my ($file_prefix,$file_extended_ext)
1087            = &util::get_prefix_and_tail_by_regex($tail_filename,$associate_tail_re);
1088        my ($pre_doc_ext) = ($file_extended_ext =~ m/^(.*)\..*$/);
1089        my ($doc_ext) = ($tail_filename =~ m/^.*\.(.*)$/);
1090        my $start_doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/{Or}{[parent(Top):assocfilepath],[assocfilepath]}/$url_tail_filename\">";
1091        my $srcicon = "_icon".$doc_ext."_";
1092        my $end_doclink = "</a>";
1093
1094        my $assoc_form = "$start_doclink\{If\}{$srcicon,$srcicon,$doc_ext\}$end_doclink";
1095
1096        if (defined $pre_doc_ext && $pre_doc_ext ne "") {
1097            # for metadata such as [mp3._edited] [mp3._full] ...
1098            $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.$pre_doc_ext", $assoc_form);
1099        }
1100
1101        # for multiple metadata such as [mp3.assoclink]
1102        $doc_obj->add_utf8_metadata ($cursection, "$doc_ext.assoclink", $assoc_form);
1103       
1104        $equiv_form .= " $assoc_form"; 
1105        }
1106        $doc_obj->add_utf8_metadata ($cursection, "equivlink", $equiv_form);
1107    }
1108    elsif (ref ($metadata->{$field}) eq "ARRAY") {
1109        map {
1110        $doc_obj->add_utf8_metadata ($cursection, $field, $_);
1111        } @{$metadata->{$field}};
1112    } else {
1113        $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
1114    }
1115    }
1116}
1117
1118
1119sub compile_stats {
1120    my $self = shift(@_);
1121    my ($stats) = @_;
1122
1123    $stats->{'num_processed'} += $self->{'num_processed'};
1124    $stats->{'num_not_processed'} += $self->{'num_not_processed'};
1125    $stats->{'num_archives'} += $self->{'num_archives'};
1126
1127}
1128
1129sub associate_cover_image {
1130    my $self = shift;
1131    my ($doc_obj, $filename) = @_;
1132
1133    $filename =~ s/\.[^\\\/\.]+$/\.jpg/;
1134    if (exists $self->{'covers_missing_cache'}->{$filename}) {
1135    # don't stat() for existence eg for multiple document input files
1136    # (eg SplitPlug)
1137    return;
1138    }
1139
1140    my $top_section=$doc_obj->get_top_section();
1141
1142    if (-e $filename) {
1143    $doc_obj->associate_source_file($filename);
1144        $doc_obj->associate_file($filename, "cover.jpg", "image/jpeg");
1145    $doc_obj->add_utf8_metadata($top_section, "hascover",  1);
1146    } else {
1147    my $upper_filename = $filename;
1148    $upper_filename =~ s/jpg$/JPG/;
1149    if (-e $upper_filename) {
1150        $doc_obj->associate_source_file($upper_filename);
1151        $doc_obj->associate_file($upper_filename, "cover.jpg",
1152                     "image/jpeg");
1153        $doc_obj->add_utf8_metadata($top_section, "hascover",  1);
1154    } else {
1155        # file doesn't exist, so record the fact that it's missing so
1156        # we don't stat() again (stat is slow)
1157        $self->{'covers_missing_cache'}->{$filename} = 1;
1158    }
1159    }
1160
1161}
1162
1163
1164# Overridden by exploding plugins (eg. ISISPlug)
1165sub clean_up_after_exploding
1166{
1167    my $self = shift(@_);
1168}
1169
1170
1171
11721;
Note: See TracBrowser for help on using the browser.