root/main/trunk/greenstone2/perllib/plugins/ConvertBinaryFile.pm @ 24362

Revision 24362, 17.2 KB (checked in by ak19, 8 years ago)

The method of locating perl has changed once more: util now defines the fuction get_perl_exec which is used by other scripts to obtain the path to the perl executable they should use.

  • 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;
47
48
49sub BEGIN {
50    @ConvertBinaryFile::ISA = ('AutoExtractMetadata');
51}
52
53my $convert_to_list =
54    [ { 'name' => "auto",
55    'desc' => "{ConvertBinaryFile.convert_to.auto}" },
56      { 'name' => "html",
57    'desc' => "{ConvertBinaryFile.convert_to.html}" },
58      { 'name' => "text",
59    'desc' => "{ConvertBinaryFile.convert_to.text}" }
60      ];
61
62my $arguments =
63    [ { 'name' => "convert_to",
64    'desc' => "{ConvertBinaryFile.convert_to}",
65    'type' => "enum",
66    'reqd' => "yes",
67    'list' => $convert_to_list,
68    'deft' => "auto" },
69      { 'name' => "keep_original_filename",
70    'desc' => "{ConvertBinaryFile.keep_original_filename}",
71    'type' => "flag" },
72      { 'name' => "title_sub",
73    'desc' => "{HTMLPlugin.title_sub}",
74    'type' => "string",
75    #'type' => "regexp",
76    'deft' => "" },
77      { 'name' => "apply_fribidi",
78    'desc' => "{ConvertBinaryFile.apply_fribidi}",
79    'type' => "flag",
80    'reqd' => "no" },
81      { 'name' => "use_strings",
82    'desc' => "{ConvertBinaryFile.use_strings}",
83    'type' => "flag",
84    'reqd' => "no" },
85      ];
86
87my $options = { 'name'     => "ConvertBinaryFile",
88        'desc'     => "{ConvertBinaryFile.desc}",
89        'abstract' => "yes",
90        'inherits' => "yes",
91        'args'     => $arguments };
92
93
94sub load_secondary_plugins
95{
96    my $self = shift (@_);
97    my ($class,$input_args,$hashArgOptLists) = @_;
98
99    my @convert_to_list = split(",",$self->{'convert_to_plugin'});
100    my $secondary_plugins = {};
101    # find the plugin
102
103    foreach my $convert_to (@convert_to_list) {
104    # load in "convert_to" plugin package
105    my $plugin_class = $convert_to;
106    my $plugin_package = $plugin_class.".pm";
107
108    my $colplugname = undef;
109    if (defined $ENV{'GSDLCOLLECTDIR'}) {
110        $colplugname = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},
111                           "perllib","plugins",
112                           $plugin_package);
113    }
114
115    my $mainplugname = &util::filename_cat($ENV{'GSDLHOME'},
116                           "perllib","plugins",
117                           $plugin_package);
118
119    if ((defined $colplugname) && (-e $colplugname)) { require $colplugname;}
120    elsif (-e $mainplugname) { require $mainplugname; }
121    else {
122        &gsprintf(STDERR, "{plugin.could_not_find_plugin}\n",
123              $plugin_class);
124        die "\n";
125    }
126
127    # call its constructor with extra options that we've worked out!
128    my $arglist = $input_args->{$plugin_class};
129
130    my ($secondary_plugin);
131    eval("\$secondary_plugin = new $plugin_class([],\$arglist)");
132    die "$@" if $@;
133    $secondary_plugins->{$plugin_class} = $secondary_plugin;
134    }
135    $self->{'secondary_plugins'} = $secondary_plugins;
136}
137
138sub new {
139    my ($class) = shift (@_);
140    my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
141    push(@$pluginlist, $class);
142    my $classPluginName = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
143    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
144    push(@{$hashArgOptLists->{"OptList"}},$options);
145
146    my $self = new AutoExtractMetadata($pluginlist, $inputargs, $hashArgOptLists);
147   
148    return bless $self, $class;
149}
150
151# should be called by subclasses after checking and setting
152# $self->{'convert_to'}
153sub set_standard_convert_settings {
154    my $self =shift (@_);
155   
156    my $convert_to = $self->{'convert_to'};
157    if ($convert_to eq "auto") {
158    $convert_to = "html";
159    $self->{'convert_to'} = "html";
160    }
161
162    if ($convert_to =~ /^html/) { # may be html or html_multi
163    $self->{'convert_to_plugin'} = "HTMLPlugin";
164    $self->{'convert_to_ext'} = "html";
165    } elsif ($convert_to eq "text") {
166    $self->{'convert_to_plugin'} = "TextPlugin";
167    $self->{'convert_to_ext'} = "txt";
168    } elsif ($convert_to eq "structuredhtml") {
169    $self->{'convert_to_plugin'} = "StructuredHTMLPlugin";
170    $self->{'convert_to_ext'} = "html";
171    } elsif ($convert_to =~ /^pagedimg/) {
172    $self->{'convert_to_plugin'} = "PagedImagePlugin";
173    my ($convert_to_ext) = $convert_to =~ /pagedimg\_(jpg|gif|png)/i;
174    $convert_to_ext = 'jpg' unless defined $convert_to_ext;
175    $self->{'convert_to_ext'} = $convert_to_ext;
176    }
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 $tmp_filename = &util::filename_cat($tmp_dirname, "$utf8_tailname$lc_suffix");
259   
260    # If gsdl is remote, we're given relative path to input file, of the form import/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 $ensure_path_absolute = 1; # true
264    &util::soft_link($input_filename, $tmp_filename, $ensure_path_absolute);
265    my $verbosity = $self->{'verbosity'};
266    if ($verbosity > 0) {
267    print $outhandle "Converting $tailname$suffix to $convert_to format\n";
268    }
269
270    my $errlog = &util::filename_cat($tmp_dirname, "err.log");
271   
272    # Execute the conversion command and get the type of the result,
273    # making sure the converter gives us the appropriate output type
274    my $output_type=$self->{'convert_to'};
275#    if ($convert_to =~ m/PagedImage/i) {
276#   $output_type = lc($convert_to)."_".lc($convert_to_ext);
277#    } else {
278#   $output_type = lc($convert_to);
279#    }
280
281    my $cmd = "\"".&util::get_perl_exec()."\" -S gsConvert.pl -verbose $verbosity ";
282    if (defined $self->{'convert_options'}) {
283    $cmd .= $self->{'convert_options'} . " ";
284    }
285    if ($self->{'use_strings'}) {
286      $cmd .= "-use_strings ";
287    }
288    $cmd .= "-errlog \"$errlog\" -output $output_type \"$tmp_filename\"";
289    print STDERR "calling cmd $cmd\n";
290    $output_type = `$cmd`;
291   
292    # remove symbolic link to original file
293    &util::rm($tmp_filename);
294   
295    # Check STDERR here
296    chomp $output_type;
297    if ($output_type eq "fail") {
298    print $outhandle "Could not convert $tailname$suffix to $convert_to format\n";
299    print $failhandle "$tailname$suffix: " . ref($self) . " failed to convert to $convert_to\n";
300    # The following  meant that if a conversion failed, the document would be counted twice - do we need it for anything?
301    #$self->{'num_not_processed'} ++;
302    if (-s "$errlog") {
303        open(ERRLOG, "$errlog");
304        while (<ERRLOG>) {
305        print $outhandle "$_";
306        }
307        print $outhandle "\n";
308        close ERRLOG;
309    }
310    &util::rm("$errlog") if (-e "$errlog");
311    return "";
312    }
313
314    # store the *actual* output type and return the output filename
315    # it's possible we requested conversion to html, but only to text succeeded
316    #$self->{'convert_to_ext'} = $output_type;
317    if ($output_type =~ /html/i) {
318    $self->{'converted_to'} = "HTML";
319    } elsif ($output_type =~ /te?xt/i) {
320    $self->{'converted_to'} = "Text";
321    } elsif ($output_type =~ /item/i){
322    $self->{'converted_to'} = "PagedImage";
323    }
324   
325    my $output_filename = $tmp_filename;
326    if ($output_type =~ /item/i) {
327    # running under windows
328    if ($ENV{'GSDLOS'} =~ /^windows$/i) {
329        $output_filename = $tmp_dirname . "\\$tailname\\" . $tailname . ".$output_type";
330    } else {
331        $output_filename = $tmp_dirname . "\/$tailname\/" . $tailname . ".$output_type";
332    }
333    } else {
334    $output_filename =~ s/$lc_suffix$/.$output_type/;
335    }
336   
337    return $output_filename;
338}
339
340
341# Override BasPlug read_into_doc_obj - we need to call secondary plugin stuff
342sub read_into_doc_obj {
343    my $self = shift (@_);
344    my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
345
346    my $outhandle = $self->{'outhandle'};
347
348    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
349
350    my $output_ext = $self->{'convert_to_ext'};
351    my $conv_filename = "";
352    $conv_filename = $self->tmp_area_convert_file($output_ext, $filename_full_path);
353       
354    if ("$conv_filename" eq "") {return -1;} # had an error, will be passed down pipeline
355    if (! -e "$conv_filename") {return -1;}
356    $self->{'conv_filename'} = $conv_filename;
357    $self->convert_post_process($conv_filename);
358
359    # Run the "fribidi" (http://fribidi.org) Unicode Bidirectional Algorithm program over the converted file
360    # Added for fixing up Persian PDFs after being processed by pdftohtml, but may be useful in other cases too
361    if ($self->{'apply_fribidi'} && $self->{'converted_to'} =~ /(HTML|Text)/) {
362    my $fribidi_command = "fribidi \"$conv_filename\" >\"${conv_filename}.tmp\"";
363    if (system($fribidi_command) != 0) {
364        print STDERR "ERROR: Cannot run fribidi on \"$conv_filename\".\n";
365    }
366    else {
367        &util::mv("${conv_filename}.tmp", $conv_filename);
368    }   
369    }
370   
371    my $secondary_plugins =  $self->{'secondary_plugins'};
372    my $num_secondary_plugins = scalar(keys %$secondary_plugins);
373
374    if ($num_secondary_plugins == 0) {
375    print $outhandle "Warning: No secondary plugin to use in conversion.  Skipping $file\n";
376    return 0; # effectively block it
377    }
378
379    my @plugin_names = keys %$secondary_plugins;
380    my $plugin_name = shift @plugin_names;
381   
382    if ($num_secondary_plugins > 1) {
383    print $outhandle "Warning: Multiple secondary plugins not supported yet!  Choosing $plugin_name\n.";
384    }
385   
386    my $secondary_plugin = $secondary_plugins->{$plugin_name};
387
388    # note: metadata is not carried on to the next level
389## **** I just replaced $metadata with {} in following
390    my ($rv,$doc_obj)
391    = $secondary_plugin->read_into_doc_obj ($pluginfo,"", $conv_filename, $block_hash, {}, $processor, $maxdocs, $total_count, $gli);
392
393    if ((!defined $rv) || ($rv<1)) {
394    # wasn't processed
395    return $rv;
396    }
397   
398    # Override previous gsdlsourcefilename set by secondary plugin
399    my $collect_file = &util::filename_within_collection($filename_full_path);
400    my $collect_conv_file = &util::filename_within_collection($conv_filename);
401    $doc_obj->set_source_filename ($collect_file, $self->{'file_rename_method'});
402    ## set_source_filename does not set the doc_obj source_path which is used in archives dbs for incremental
403    # build. so set it manually.
404    $doc_obj->set_source_path($filename_full_path);
405    $doc_obj->set_converted_filename($collect_conv_file);
406
407    my $plugin_filename_encoding = $self->{'filename_encoding'};
408    my $filename_encoding = $self->deduce_filename_encoding($file,$metadata,$plugin_filename_encoding);
409    $self->set_Source_metadata($doc_obj, $filename_full_path, $filename_encoding);
410       
411    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
412    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FileSize", (-s $filename_full_path));
413
414    # ****
415    my ($tailname, $dirname, $suffix)
416    = &File::Basename::fileparse($filename_full_path, "\\.[^\\.]+\$");
417    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FilenameRoot", $tailname);
418
419    # do plugin specific processing of doc_obj
420    unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
421    print STDERR "<ProcessingError n='$file'>\n" if ($gli);
422    return -1;
423    }
424
425    my $topsection = $doc_obj->get_top_section();
426    $self->add_associated_files($doc_obj, $filename_full_path);
427
428    # extra_metadata is already called by sec plugin in process??
429    $self->extra_metadata($doc_obj, $topsection, $metadata); # do we need this here??
430    # do any automatic metadata extraction
431    $self->auto_extract_metadata ($doc_obj);
432
433    # have we found a Title??
434    $self->title_fallback($doc_obj,$topsection,$filename_no_path);
435
436    $self->add_OID($doc_obj);
437
438    return (1, $doc_obj);
439
440}
441
442sub process {
443    my $self = shift (@_);
444    my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
445
446    return $self->process_type($base_dir, $file, $doc_obj);
447}
448
449# do plugin specific processing of doc_obj for doc_ext type
450sub process_type {
451    my $self = shift (@_);
452    my ($base_dir, $file, $doc_obj) = @_;
453   
454    # need to check that not empty
455    my ($doc_ext) = $file =~ /\.(\w+)$/;
456    $doc_ext = lc($doc_ext);
457    my $file_type = "unknown";
458    $file_type = $self->{'file_type'} if defined $self->{'file_type'};
459   
460    # associate original file with doc object
461    my $cursection = $doc_obj->get_top_section();
462    my $filename = &util::filename_cat($base_dir, $file);
463    my $assocfilename = "doc.$doc_ext";
464    if ($self->{'keep_original_filename'} == 1) {
465    # this should be the same filename that was used for the Source and SourceFile metadata,
466    # as we will use SourceFile in the srclink (below)
467    $assocfilename = $doc_obj->get_assocfile_from_sourcefile();
468    }
469
470    $doc_obj->associate_file($filename, $assocfilename, undef, $cursection);
471
472    # We use set instead of add here because we only want one value
473    $doc_obj->set_utf8_metadata_element($cursection, "FileFormat", $file_type);
474    my $srclink_filename = "doc.$doc_ext";
475    if ($self->{'keep_original_filename'} == 1) {
476    $srclink_filename = $doc_obj->get_sourcefile();
477    }
478    # srclink_file is now deprecated because of the "_" in the metadataname. Use srclinkFile
479    $doc_obj->add_utf8_metadata ($cursection, "srcicon",  "_icon".$doc_ext."_");
480    $doc_obj->add_utf8_metadata ($cursection, "srclink_file", $srclink_filename);
481    $doc_obj->add_utf8_metadata ($cursection, "srclinkFile", $srclink_filename);
482    return 1;
483}
484
485sub clean_up_after_doc_obj_processing {
486     my $self = shift(@_);
487
488     my $tmp_dir = $self->{'tmp_dir'};
489     if (defined $tmp_dir && -d $tmp_dir) {
490     ##print STDERR "**** Supressing clean up of tmp dir\n";
491     &util::rm_r($tmp_dir);
492     $self->{'tmp_dir'} = undef;
493     }
494     
495
496}
4971;
498
499
500
501
502
503
504
Note: See TracBrowser for help on using the browser.