root/main/trunk/greenstone2/perllib/plugins/ConvertBinaryFile.pm @ 32277

Revision 32277, 20.0 KB (checked in by ak19, 23 months ago)

First attempt at PDFv2Plugin.pm.

  • Property svn:keywords set to Author Date Id Revision
Line 
1###########################################################################
2#
3# ConvertBinaryFile.pm -- plugin that facilitates conversion of binary files
4# through gsConvert.pl
5#
6# A component of the Greenstone digital library software
7# from the New Zealand Digital Library Project at the
8# University of Waikato, New Zealand.
9#
10# Copyright (C) 1999 New Zealand Digital Library Project
11#
12# This program is free software; you can redistribute it and/or modify
13# it under the terms of the GNU General Public License as published by
14# the Free Software Foundation; either version 2 of the License, or
15# (at your option) any later version.
16#
17# This program is distributed in the hope that it will be useful,
18# but WITHOUT ANY WARRANTY; without even the implied warranty of
19# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
20# GNU General Public License for more details.
21#
22# You should have received a copy of the GNU General Public License
23# along with this program; if not, write to the Free Software
24# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
25#
26###########################################################################
27
28# This plugin is inherited by such plugins as WordPlugin, PowerPointPlugin,
29# PostScriptPlugin,
30# RTFPlugin and PDFPlugin. It facilitates the conversion of these document types
31# to either HTML, Text or a series of images. It works by dynamically loading
32# an appropriate secondary plugin (HTMLPlug, StructuredHTMLPlug,
33# PagedImagePlugin or TextPlugin) based on the plugin argument 'convert_to'.
34
35package ConvertBinaryFile;
36
37use AutoExtractMetadata;
38use ghtml;
39use HTMLPlugin;
40use TextPlugin;
41use PagedImagePlugin;
42
43use strict;
44no strict 'refs'; # allow filehandles to be variables and viceversa
45no strict 'subs';
46use util;
47use FileUtils;
48
49
50sub BEGIN {
51    @ConvertBinaryFile::ISA = ('AutoExtractMetadata');
52}
53
54my $convert_to_list =
55    [ { 'name' => "auto",
56    'desc' => "{ConvertBinaryFile.convert_to.auto}" },
57      { 'name' => "html",
58    'desc' => "{ConvertBinaryFile.convert_to.html}" },
59      { 'name' => "text",
60    'desc' => "{ConvertBinaryFile.convert_to.text}" }
61      ];
62
63my $arguments =
64    [ { 'name' => "convert_to",
65    'desc' => "{ConvertBinaryFile.convert_to}",
66    'type' => "enum",
67    'reqd' => "yes",
68    'list' => $convert_to_list,
69    'deft' => "auto" },
70      { 'name' => "keep_original_filename",
71    'desc' => "{ConvertBinaryFile.keep_original_filename}",
72    'type' => "flag" },
73      { 'name' => "title_sub",
74    'desc' => "{HTMLPlugin.title_sub}",
75    'type' => "string",
76    #'type' => "regexp",
77    'deft' => "" },
78      { 'name' => "apply_fribidi",
79    'desc' => "{ConvertBinaryFile.apply_fribidi}",
80    'type' => "flag",
81    'reqd' => "no" },
82      { 'name' => "use_strings",
83    'desc' => "{ConvertBinaryFile.use_strings}",
84    'type' => "flag",
85    'reqd' => "no" },
86      ];
87
88my $options = { 'name'     => "ConvertBinaryFile",
89        'desc'     => "{ConvertBinaryFile.desc}",
90        'abstract' => "yes",
91        'inherits' => "yes",
92        'args'     => $arguments };
93
94
95sub load_secondary_plugins
96{
97    my $self = shift (@_);
98    my ($class,$input_args,$hashArgOptLists) = @_;
99
100    my @convert_to_list = split(",",$self->{'convert_to_plugin'});
101    my $secondary_plugins = {};
102    # find the plugin
103
104    foreach my $convert_to (@convert_to_list) {
105    # load in "convert_to" plugin package
106    my $plugin_class = $convert_to;
107    my $plugin_package = $plugin_class.".pm";
108
109    my $colplugname = undef;
110    if (defined $ENV{'GSDLCOLLECTDIR'}) {
111        $colplugname = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'},
112                           "perllib","plugins",
113                           $plugin_package);
114    }
115
116    my $mainplugname = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'},
117                           "perllib","plugins",
118                           $plugin_package);
119
120    if ((defined $colplugname) && (-e $colplugname)) { require $colplugname;}
121    elsif (-e $mainplugname) { require $mainplugname; }
122    else {
123        &gsprintf(STDERR, "{plugin.could_not_find_plugin}\n",
124              $plugin_class);
125        die "\n";
126    }
127
128    # call its constructor with extra options that we've worked out!
129    my $arglist = $input_args->{$plugin_class};
130
131    my ($secondary_plugin);
132    eval("\$secondary_plugin = new $plugin_class([],\$arglist)");
133    die "$@" if $@;
134    $secondary_plugins->{$plugin_class} = $secondary_plugin;
135    }
136    $self->{'secondary_plugins'} = $secondary_plugins;
137}
138
139sub new {
140    my ($class) = shift (@_);
141    my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
142    push(@$pluginlist, $class);
143    my $classPluginName = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
144    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
145    push(@{$hashArgOptLists->{"OptList"}},$options);
146
147    my $self = new AutoExtractMetadata($pluginlist, $inputargs, $hashArgOptLists);
148   
149    return bless $self, $class;
150}
151
152# should be called by subclasses after checking and setting
153# $self->{'convert_to'}
154sub set_standard_convert_settings {
155    my $self =shift (@_);
156   
157    my $convert_to = $self->{'convert_to'};
158    if ($convert_to eq "auto") {
159    $convert_to = "html";
160    $self->{'convert_to'} = "html";
161    }
162
163    if ($convert_to =~ /^html/ || $convert_to =~ /pretty_html$/) { # may be html or html_multi, or )paged_)pretty_html with the new Xpdf's own pdftohtml
164    $self->{'convert_to_plugin'} = "HTMLPlugin";
165    $self->{'convert_to_ext'} = "html";
166    } elsif ($convert_to eq "text") {
167    $self->{'convert_to_plugin'} = "TextPlugin";
168    $self->{'convert_to_ext'} = "txt";
169    } elsif ($convert_to eq "structuredhtml") {
170    $self->{'convert_to_plugin'} = "StructuredHTMLPlugin";
171    $self->{'convert_to_ext'} = "html";
172    } elsif ($convert_to =~ /^pagedimg/) {
173    $self->{'convert_to_plugin'} = "PagedImagePlugin";
174    my ($convert_to_ext) = $convert_to =~ /pagedimg\_(jpg|gif|png)/i;
175    $convert_to_ext = 'jpg' unless defined $convert_to_ext;
176    $self->{'convert_to_ext'} = $convert_to_ext;
177    }
178}
179sub init {
180    my $self = shift (@_);
181    my ($verbosity, $outhandle, $failhandle) = @_;
182
183    $self->SUPER::init($verbosity,$outhandle,$failhandle);
184
185    my $secondary_plugins =  $self->{'secondary_plugins'};
186
187    foreach my $plug_name (keys %$secondary_plugins) {
188    my $plugin = $secondary_plugins->{$plug_name};
189    $plugin->init($verbosity,$outhandle,$failhandle);
190    }
191}
192
193sub deinit {
194    # called only once, after all plugin passes have been done
195
196    my ($self) = @_;
197
198    my $secondary_plugins =  $self->{'secondary_plugins'};
199
200    foreach my $plug_name (keys %$secondary_plugins) {
201    my $plugin = $secondary_plugins->{$plug_name};
202    $plugin->deinit();
203    }
204}
205
206sub convert_post_process
207{
208    # by default do no post processing
209    return;
210}
211
212
213# Run conversion utility on the input file. 
214#
215# The conversion takes place in a collection specific 'tmp' directory so
216# that we don't accidentally damage the input.
217#
218# The desired output type is indicated by $output_ext.  This is usually
219# something like "html" or "word", but can be "best" (or the empty string)
220# to indicate that the conversion utility should do the best it can.
221sub tmp_area_convert_file {
222    my $self = shift (@_);
223    my ($output_ext, $input_filename, $textref) = @_;
224   
225    my $outhandle = $self->{'outhandle'};
226    my $convert_to = $self->{'convert_to'};
227    my $failhandle = $self->{'failhandle'};
228    my $convert_to_ext = $self->{'convert_to_ext'};
229   
230
231    my $upgraded_input_filename = &util::upgrade_if_dos_filename($input_filename);
232
233    # derive tmp filename from input filename
234    my ($tailname, $dirname, $suffix)
235    = &File::Basename::fileparse($upgraded_input_filename, "\\.[^\\.]+\$");
236
237    # softlink to collection tmp dir
238    my $tmp_dirname = &util::get_timestamped_tmp_folder();
239    if (defined $tmp_dirname) {
240    $self->{'tmp_dir'} = $tmp_dirname;
241    } else {
242    $tmp_dirname = $dirname;
243    }
244   
245#    # convert to utf-8 otherwise we have problems with the doc.xml file later on
246#    my $utf8_tailname = (&unicode::check_is_utf8($tailname)) ? $tailname : $self->filepath_to_utf8($tailname);
247
248    # make sure filename to be used can be stored OK in a UTF-8 compliant doc.xml file
249     my $utf8_tailname = &unicode::raw_filename_to_utf8_url_encoded($tailname);
250
251
252    # URLEncode this since htmls with images where the html filename is utf8 don't seem
253    # to work on Windows (IE or Firefox), as browsers are looking for filesystem-encoded
254    # files on the filesystem.
255    $utf8_tailname = &util::rename_file($utf8_tailname, $self->{'file_rename_method'}, "without_suffix");
256
257    my $lc_suffix = lc($suffix);
258    my $tmp_filename = &FileUtils::filenameConcatenate($tmp_dirname, "$utf8_tailname$lc_suffix");
259   
260    # If gsdl is remote, we're given relative path to input file, of the form import/utf8_tailname.suffix
261    # But we can't softlink to relative paths. Therefore, we need to ensure that
262    # the input_filename is the absolute path, see http://perldoc.perl.org/File/Spec.html
263    my $ensure_path_absolute = 1; # true
264    &FileUtils::softLink($input_filename, $tmp_filename, $ensure_path_absolute);
265
266    my $output_filename = $self->run_conversion_command($tmp_dirname, $tmp_filename,
267                            $utf8_tailname, $lc_suffix, $tailname, $suffix);
268
269    return $output_filename;
270}
271
272# The latter half of tmp_area_convert_file: runs the conversion command and returns the output file name
273# Split from tmp_area_convert_file because UnknownConverterPlugin can then inherit all of
274# tmp_area_convert_file and only needs to override this part:
275sub run_conversion_command {
276    my $self = shift (@_);
277    my ($tmp_dirname, $tmp_filename, $utf8_tailname, $lc_suffix, $tailname, $suffix) = @_;   
278
279    my $outhandle = $self->{'outhandle'};
280    my $convert_to = $self->{'convert_to'};
281    my $failhandle = $self->{'failhandle'};
282
283    my $verbosity = $self->{'verbosity'};
284    if ($verbosity > 0) {
285    print $outhandle "Converting $tailname$suffix to $convert_to format\n";
286    }
287
288    my $errlog = &FileUtils::filenameConcatenate($tmp_dirname, "err.log");
289   
290    # Execute the conversion command and get the type of the result,
291    # making sure the converter gives us the appropriate output type
292    my $output_type=$self->{'convert_to'};
293#    if ($convert_to =~ m/PagedImage/i) {
294#   $output_type = lc($convert_to)."_".lc($convert_to_ext);
295#    } else {
296#   $output_type = lc($convert_to);
297#    }
298
299    my $cmd = "\"".&util::get_perl_exec()."\" -S gsConvert.pl -verbose $verbosity ";
300    if (defined $self->{'convert_options'}) {
301    $cmd .= $self->{'convert_options'} . " ";
302    }
303    if ($self->{'use_strings'}) {
304      $cmd .= "-use_strings ";
305    }
306    $cmd .= "-errlog \"$errlog\" -output $output_type \"$tmp_filename\"";
307    print STDERR "calling cmd $cmd\n";
308    $output_type = `$cmd`;
309   
310    # remove symbolic link to original file
311    &FileUtils::removeFiles($tmp_filename);
312   
313    # Check STDERR here
314    chomp $output_type;
315    if ($output_type eq "fail") {
316    print $outhandle "Could not convert $tailname$suffix to $convert_to format\n";
317    print $failhandle "$tailname$suffix: " . ref($self) . " failed to convert to $convert_to\n";
318    # The following  meant that if a conversion failed, the document would be counted twice - do we need it for anything?
319    #$self->{'num_not_processed'} ++;
320    if (-s "$errlog") {
321        open(ERRLOG, "$errlog");
322        while (<ERRLOG>) {
323        print $outhandle "$_";
324        }
325        print $outhandle "\n";
326        close ERRLOG;
327    }
328    &FileUtils::removeFiles("$errlog") if (-e "$errlog");
329    return "";
330    }
331
332    # store the *actual* output type and return the output filename
333    # it's possible we requested conversion to html, but only to text succeeded
334    #$self->{'convert_to_ext'} = $output_type;
335    if ($output_type =~ /html/i) {
336    $self->{'converted_to'} = "HTML";
337    } elsif ($output_type =~ /te?xt/i) {
338    $self->{'converted_to'} = "Text";
339    } elsif ($output_type =~ /item/i){
340    $self->{'converted_to'} = "PagedImage";
341    }
342   
343    my $output_filename = $tmp_filename;
344    if ($output_type =~ /item/i) {
345    # running under windows
346    if ($ENV{'GSDLOS'} =~ /^windows$/i) {
347        $output_filename = $tmp_dirname . "\\$utf8_tailname\\" . $utf8_tailname . ".$output_type";
348    } else {
349        $output_filename = $tmp_dirname . "\/$utf8_tailname\/" . $utf8_tailname . ".$output_type";
350    }
351    } else {
352    $output_filename =~ s/$lc_suffix$/.$output_type/;
353    }
354   
355    return $output_filename;
356}
357
358
359# Override BasPlug read_into_doc_obj - we need to call secondary plugin stuff
360sub read_into_doc_obj {
361    my $self = shift (@_);
362    my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
363
364    my $outhandle = $self->{'outhandle'};
365
366    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
367
368    my $output_ext = $self->{'convert_to_ext'};
369    my $conv_filename = "";
370    $conv_filename = $self->tmp_area_convert_file($output_ext, $filename_full_path);
371       
372    if ("$conv_filename" eq "") {return -1;} # had an error, will be passed down pipeline
373
374    my $output_type=$self->{'convert_to'};
375    if (!&FileUtils::fileExists($conv_filename)) {return -1;}
376    $self->{'conv_filename'} = $conv_filename;
377    $self->convert_post_process($conv_filename);
378
379    # Check if, after post-processing, the final expected output file has changed
380    # And if it has, check that the final output file now exists after post processing
381    if(defined $self->{'conv_filename_after_post_process'}) {   
382    $conv_filename = $self->{'conv_filename_after_post_process'};
383    if (!&FileUtils::fileExists($conv_filename)) {return -1;}
384    }
385
386    # Run the "fribidi" (http://fribidi.org) Unicode Bidirectional Algorithm program over the converted file
387    # Added for fixing up Persian PDFs after being processed by pdftohtml, but may be useful in other cases too
388    if ($self->{'apply_fribidi'} && $self->{'converted_to'} =~ /(HTML|Text)/) {
389    my $fribidi_command = "fribidi \"$conv_filename\" >\"${conv_filename}.tmp\"";
390    if (system($fribidi_command) != 0) {
391        print STDERR "ERROR: Cannot run fribidi on \"$conv_filename\".\n";
392    }
393    else {
394        &FileUtils::moveFiles("${conv_filename}.tmp", $conv_filename);
395    }   
396    }
397   
398    my $secondary_plugins =  $self->{'secondary_plugins'};
399    my $num_secondary_plugins = scalar(keys %$secondary_plugins);
400
401    if ($num_secondary_plugins == 0) {
402    print $outhandle "Warning: No secondary plugin to use in conversion.  Skipping $file\n";
403    return 0; # effectively block it
404    }
405
406    my @plugin_names = keys %$secondary_plugins;
407    my $plugin_name = shift @plugin_names;
408   
409    if ($num_secondary_plugins > 1) {
410    print $outhandle "Warning: Multiple secondary plugins not supported yet!  Choosing $plugin_name\n.";
411    }
412   
413    my $secondary_plugin = $secondary_plugins->{$plugin_name};
414
415    # note: metadata is not carried on to the next level
416## **** I just replaced $metadata with {} in following
417    my ($rv,$doc_obj)
418    = $secondary_plugin->read_into_doc_obj ($pluginfo,"", $conv_filename, $block_hash, {}, $processor, $maxdocs, $total_count, $gli);
419
420    if ((!defined $rv) || ($rv<1)) {
421    # wasn't processed
422    return $rv;
423    }
424   
425    # Override previous gsdlsourcefilename set by secondary plugin
426    my $collect_file = &util::filename_within_collection($filename_full_path);
427    my $collect_conv_file = &util::filename_within_collection($conv_filename);
428    $doc_obj->set_source_filename ($collect_file, $self->{'file_rename_method'});
429    ## set_source_filename does not set the doc_obj source_path which is used in archives dbs for incremental
430    # build. so set it manually.
431    $doc_obj->set_source_path($filename_full_path);
432    $doc_obj->set_converted_filename($collect_conv_file);
433
434    my $plugin_filename_encoding = $self->{'filename_encoding'};
435    my $filename_encoding = $self->deduce_filename_encoding($file,$metadata,$plugin_filename_encoding);
436    $self->set_Source_metadata($doc_obj, $filename_full_path, $filename_encoding);
437       
438    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
439    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FileSize", (-s $filename_full_path));
440
441    # ****
442    my ($tailname, $dirname, $suffix)
443    = &File::Basename::fileparse($filename_full_path, "\\.[^\\.]+\$");
444    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FilenameRoot", $tailname);
445
446    # do plugin specific processing of doc_obj
447    unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
448    print STDERR "<ProcessingError n='$file'>\n" if ($gli);
449    return -1;
450    }
451
452    my $topsection = $doc_obj->get_top_section();
453    $self->add_associated_files($doc_obj, $filename_full_path);
454
455    # extra_metadata is already called by sec plugin in process??
456    $self->extra_metadata($doc_obj, $topsection, $metadata); # do we need this here??
457    # do any automatic metadata extraction
458    $self->auto_extract_metadata ($doc_obj);
459
460    # have we found a Title??
461    $self->title_fallback($doc_obj,$topsection,$filename_no_path);
462
463    # force a new OID - this will use OIDtype option set for this plugin.
464    $self->add_OID($doc_obj, 1);
465
466    return (1, $doc_obj);
467
468}
469
470sub process {
471    my $self = shift (@_);
472    my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
473
474    return $self->process_type($base_dir, $file, $doc_obj);
475}
476
477# do plugin specific processing of doc_obj for doc_ext type
478sub process_type {
479    my $self = shift (@_);
480    my ($base_dir, $file, $doc_obj) = @_;
481   
482    # need to check that not empty
483    my ($doc_ext) = $file =~ /\.(\w+)$/;
484    $doc_ext = lc($doc_ext);
485    my $file_type = "unknown";
486    $file_type = $self->{'file_type'} if defined $self->{'file_type'};
487   
488    # associate original file with doc object
489    my $cursection = $doc_obj->get_top_section();
490    my $filename = &FileUtils::filenameConcatenate($base_dir, $file);
491    my $assocfilename = "doc.$doc_ext";
492    if ($self->{'keep_original_filename'} == 1) {
493    # this should be the same filename that was used for the Source and SourceFile metadata,
494    # as we will use SourceFile in the srclink (below)
495    $assocfilename = $doc_obj->get_assocfile_from_sourcefile();
496    }
497
498    $doc_obj->associate_file($filename, $assocfilename, undef, $cursection);
499
500    # We use set instead of add here because we only want one value
501    $doc_obj->set_utf8_metadata_element($cursection, "FileFormat", $file_type);
502    my $srclink_filename = "doc.$doc_ext";
503    if ($self->{'keep_original_filename'} == 1) {
504    $srclink_filename = $doc_obj->get_sourcefile();
505    }
506    # srclink_file is now deprecated because of the "_" in the metadataname. Use srclinkFile
507    $doc_obj->add_utf8_metadata ($cursection, "srcicon",  "_icon".$doc_ext."_");
508    $doc_obj->add_utf8_metadata ($cursection, "srclink_file", $srclink_filename);
509    $doc_obj->add_utf8_metadata ($cursection, "srclinkFile", $srclink_filename);
510    return 1;
511}
512
513sub clean_up_after_doc_obj_processing {
514     my $self = shift(@_);
515
516     my $tmp_dir = $self->{'tmp_dir'};
517     if (defined $tmp_dir && -d $tmp_dir) {
518     ##print STDERR "**** Suppressing clean up of tmp dir\n";
519     &FileUtils::removeFilesRecursive($tmp_dir);
520     $self->{'tmp_dir'} = undef;
521     }
522     
523
524}
525
526# This sub is shared across PowerPointPlugin and UnknownConverterPlugin,
527# so it's been copied into here from the former.
528sub generate_item_file {
529    my $self = shift(@_);
530    my ($input_filename) = @_;
531    my $outhandle = $self->{'outhandle'};
532    my ($tailname, $dirname, $suffix)
533    = &File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");
534
535    my $plugin_name = $self->{'plugin_type'}; # inherited from BaseImporter
536
537    # find all the files in the directory
538    if (!opendir (DIR, $dirname)) {
539    print $outhandle "$plugin_name: Couldn't read directory $dirname\n";
540    return $input_filename;
541    }
542
543    my @dir = readdir (DIR);
544    closedir (DIR);
545
546    # start the item file
547    my $itemfile_name = &util::filename_cat($dirname, "$tailname.item");
548
549    # encoding specification????
550    if (!open (ITEMFILE, ">$itemfile_name")) {
551    print $outhandle "$plugin_name: Couldn't open $itemfile_name for writing\n";
552    }
553    print ITEMFILE "<GeneratedBy>$plugin_name\n";
554    # print the first page
555    my @sorted_dir = sort alphanum_sort @dir;
556    for (my $i = 0; $i < scalar(@sorted_dir); $i++) {
557    my $file = $sorted_dir[$i];
558    if ($file =~ /^img(\d+)\.jpg$/) {
559        my $num = $1;
560        $self->tidy_up_html(&util::filename_cat($dirname, "text$num.html"));
561        print ITEMFILE "$num:img$num.jpg:text$num.html:\n";
562    }
563    }
564    close ITEMFILE;
565    return $itemfile_name;
566
567}
568
5691;
570
571
572
573
574
575
576
Note: See TracBrowser for help on using the browser.