root/main/trunk/greenstone2/perllib/plugins/ConvertBinaryFile.pm @ 22504

Revision 22504, 18.5 KB (checked in by kjdon, 9 years ago)

for WordPLugin, if openoffice_scripting is set, need to use structuredhtml plugin
  • 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';
46
47sub BEGIN {
48    @ConvertBinaryFile::ISA = ('AutoExtractMetadata');
49}
50
51my $convert_to_list =
52    [ { 'name' => "auto",
53    'desc' => "{ConvertBinaryFile.convert_to.auto}" },
54      { 'name' => "html",
55    'desc' => "{ConvertBinaryFile.convert_to.html}" },
56      { 'name' => "text",
57    'desc' => "{ConvertBinaryFile.convert_to.text}" }
58      ];
59
60my $arguments =
61    [ { 'name' => "convert_to",
62    'desc' => "{ConvertBinaryFile.convert_to}",
63    'type' => "enum",
64    'reqd' => "yes",
65    'list' => $convert_to_list,
66    'deft' => "auto" },
67      { 'name' => "keep_original_filename",
68    'desc' => "{ConvertBinaryFile.keep_original_filename}",
69    'type' => "flag" },
70      { 'name' => "title_sub",
71    'desc' => "{HTMLPlugin.title_sub}",
72    'type' => "string",
73    #'type' => "regexp",
74    'deft' => "" },
75      { 'name' => "apply_fribidi",
76    'desc' => "{ConvertBinaryFile.apply_fribidi}",
77    'type' => "flag",
78    'reqd' => "no" },
79      { 'name' => "use_strings",
80    'desc' => "{ConvertBinaryFile.use_strings}",
81    'type' => "flag",
82    'reqd' => "no" },
83      ];
84
85my $options = { 'name'     => "ConvertBinaryFile",
86        'desc'     => "{ConvertBinaryFile.desc}",
87        'abstract' => "yes",
88        'inherits' => "yes",
89        'args'     => $arguments };
90
91
92sub load_secondary_plugins
93{
94    my $self = shift (@_);
95    my ($class,$input_args,$hashArgOptLists) = @_;
96
97    my @convert_to_list = split(",",$self->{'convert_to'});
98    my $secondary_plugins = {};
99    # find the plugin
100
101    foreach my $convert_to (@convert_to_list) {
102    # load in "convert_to" plugin package
103    my $plugin_class = $convert_to."Plugin";
104    my $plugin_package = $plugin_class.".pm";
105
106    my $colplugname = undef;
107    if (defined $ENV{'GSDLCOLLECTDIR'}) {
108        $colplugname = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},
109                           "perllib","plugins",
110                           $plugin_package);
111    }
112
113    my $mainplugname = &util::filename_cat($ENV{'GSDLHOME'},
114                           "perllib","plugins",
115                           $plugin_package);
116
117    if ((defined $colplugname) && (-e $colplugname)) { require $colplugname;}
118    elsif (-e $mainplugname) { require $mainplugname; }
119    else {
120        &gsprintf(STDERR, "{plugin.could_not_find_plugin}\n",
121              $plugin_class);
122        die "\n";
123    }
124
125    # call its constructor with extra options that we've worked out!
126    my $arglist = $input_args->{$plugin_class};
127
128    my ($secondary_plugin);
129    eval("\$secondary_plugin = new $plugin_class([],\$arglist)");
130    die "$@" if $@;
131    $secondary_plugins->{$plugin_class} = $secondary_plugin;
132    }
133    $self->{'secondary_plugins'} = $secondary_plugins;
134}
135
136sub new {
137    my ($class) = shift (@_);
138    my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
139    push(@$pluginlist, $class);
140    my $classPluginName = (defined $pluginlist->[0]) ? $pluginlist->[0] : $class;
141    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
142    push(@{$hashArgOptLists->{"OptList"}},$options);
143
144    my $self = new AutoExtractMetadata($pluginlist, $inputargs, $hashArgOptLists);
145   
146    if ($self->{'info_only'}) {
147    # don't worry about any options etc
148    return bless $self, $class;
149    }
150
151    my $convert_to_type = $self->{'convert_to'};
152    if (!defined $convert_to_type || $convert_to_type eq "") {
153    $convert_to_type = "auto";
154    }
155    my $windows_scripting = $self->{'windows_scripting'};
156    $windows_scripting = 0 unless defined $windows_scripting;
157    if ($classPluginName eq "PDFPlugin") {
158    if ($convert_to_type eq "text" &&
159        $ENV{'GSDLOS'} =~ /^windows$/i) {
160        print STDERR "Windows does not support pdf to text. PDFs will be converted to HTML instead\n";
161        $convert_to_type = "html";
162    }
163    } elsif ($classPluginName eq "WordPlugin") {
164    if (($windows_scripting && $ENV{'GSDLOS'} =~ /^windows$/i && $convert_to_type =~ /^(html|auto)$/) || defined $self->{'openoffice_scripting'}) {
165        # we use structured HTML, not normal html
166        $convert_to_type = "structuredhtml";
167    }
168    } elsif ($classPluginName eq "PowerPointPlugin") {
169    if ($windows_scripting && $ENV{'GSDLOS'} =~ /^windows$/i && $convert_to_type eq "auto") {
170        # we use paged img
171        $convert_to_type = "pagedimg_jpg";
172    }
173    } elsif ($classPluginName eq "PostScriptPlugin") {
174    if ($convert_to_type eq "auto") {
175        # we use text
176        $convert_to_type = "text";
177    }
178    }
179   
180    if ($convert_to_type eq "auto") {
181    # choose html for now - should choose a format based on doc type
182    $convert_to_type = "html";
183    }
184   
185    if ($convert_to_type eq "html") {
186    $self->{'convert_to'} = "HTML";
187    $self->{'convert_to_ext'} = "html";
188    } elsif ($convert_to_type eq "text") {
189    $self->{'convert_to'} = "Text";
190    $self->{'convert_to_ext'} = "txt";
191    } elsif ($convert_to_type eq "structuredhtml") {
192    $self->{'convert_to'} = "StructuredHTML";
193    $self->{'convert_to_ext'} = "html";
194    } elsif ($convert_to_type =~ /^pagedimg/) {
195    $self->{'convert_to'} = "PagedImage";
196    my ($convert_to_ext) = $convert_to_type =~ /pagedimg\_(jpg|gif|png)/i;
197    $convert_to_ext = 'jpg' unless defined $convert_to_ext;
198    $self->{'convert_to_ext'} = $convert_to_ext;
199    }
200   
201    return bless $self, $class;
202}
203
204
205sub init {
206    my $self = shift (@_);
207    my ($verbosity, $outhandle, $failhandle) = @_;
208
209    $self->SUPER::init($verbosity,$outhandle,$failhandle);
210
211    my $secondary_plugins =  $self->{'secondary_plugins'};
212
213    foreach my $plug_name (keys %$secondary_plugins) {
214    my $plugin = $secondary_plugins->{$plug_name};
215    $plugin->init($verbosity,$outhandle,$failhandle);
216    }
217}
218
219sub deinit {
220    # called only once, after all plugin passes have been done
221
222    my ($self) = @_;
223
224    my $secondary_plugins =  $self->{'secondary_plugins'};
225
226    foreach my $plug_name (keys %$secondary_plugins) {
227    my $plugin = $secondary_plugins->{$plug_name};
228    $plugin->deinit();
229    }
230}
231
232sub convert_post_process
233{
234    # by default do no post processing
235    return;
236}
237
238
239# Run conversion utility on the input file. 
240#
241# The conversion takes place in a collection specific 'tmp' directory so
242# that we don't accidentally damage the input.
243#
244# The desired output type is indicated by $output_ext.  This is usually
245# something like "html" or "word", but can be "best" (or the empty string)
246# to indicate that the conversion utility should do the best it can.
247sub tmp_area_convert_file {
248    my $self = shift (@_);
249    my ($output_ext, $input_filename, $textref) = @_;
250   
251    my $outhandle = $self->{'outhandle'};
252    my $convert_to = $self->{'convert_to'};
253    my $failhandle = $self->{'failhandle'};
254    my $convert_to_ext = $self->{'convert_to_ext'};
255   
256    # derive tmp filename from input filename
257    my ($tailname, $dirname, $suffix)
258    = &File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");
259
260    # softlink to collection tmp dir
261    my $tmp_dirname = $dirname;
262    if(defined $ENV{'GSDLCOLLECTDIR'}) {
263    $tmp_dirname = $ENV{'GSDLCOLLECTDIR'};
264    } elsif(defined $ENV{'GSDLHOME'}) {
265    $tmp_dirname = $ENV{'GSDLHOME'};
266    }
267
268    $tmp_dirname = &util::filename_cat($tmp_dirname, "tmp");
269    &util::mk_dir($tmp_dirname) if (!-e $tmp_dirname);
270
271    # add the timestamp into the path otherwise we can run into problems
272    # if documents have the same name
273    my $timestamp = time;
274    my $time_tmp_dirname = &util::filename_cat($tmp_dirname, $timestamp);
275    $tmp_dirname = $time_tmp_dirname;
276    my $i = 1;
277    while (-e $tmp_dirname) {
278    $tmp_dirname = "$time_tmp_dirname$i";
279    $i++;
280    }
281    &util::mk_dir($tmp_dirname);
282
283    # The following is not necessary and will cause problems with
284    # replacing_srcdoc_with_html in the GSDLremote case:
285    # Remove any white space from filename -- no risk of name collision, and
286    # makes later conversion by utils simpler. Leave spaces in path...
287    # tidy up the filename with space, dot, hyphen between
288    #$tailname =~ s/\s+//g;
289    #$tailname =~ s/\.+//g;
290    #$tailname =~ s/\-+//g;
291
292    # convert to utf-8 otherwise we have problems with the doc.xml file later on
293#    print STDERR "**** filename $tailname$suffix is already UTF8\n" if &unicode::check_is_utf8($tailname);
294    $tailname = $self->SUPER::filepath_to_utf8($tailname) unless &unicode::check_is_utf8($tailname);
295
296    # URLEncode this since htmls with images where the html filename is utf8 don't seem
297    # to work on Windows (IE or Firefox), as browsers are looking for filesystem-encoded
298    # files on the filesystem.
299    $tailname = &util::rename_file($tailname, $self->{'file_rename_method'}, "without_suffix");
300
301    $suffix = lc($suffix);
302    my $tmp_filename = &util::filename_cat($tmp_dirname, "$tailname$suffix");
303   
304    # If gsdl is remote, we're given relative path to input file, of the form import/tailname.suffix
305    # But we can't softlink to relative paths. Therefore, we need to ensure that
306    # the input_filename is the absolute path, see http://perldoc.perl.org/File/Spec.html
307    my $ensure_path_absolute = 1; # true
308    &util::soft_link($input_filename, $tmp_filename, $ensure_path_absolute);
309    my $verbosity = $self->{'verbosity'};
310    if ($verbosity > 0) {
311    print $outhandle "Converting $tailname$suffix to $convert_to format\n";
312    }
313
314    my $errlog = &util::filename_cat($tmp_dirname, "err.log");
315   
316    # Execute the conversion command and get the type of the result,
317    # making sure the converter gives us the appropriate output type
318    my $output_type="";
319    if ($convert_to =~ m/PagedImage/i) {
320    $output_type = lc($convert_to)."_".lc($convert_to_ext);
321    } else {
322    $output_type = lc($convert_to);
323    }
324
325    my $cmd = "perl -S gsConvert.pl -verbose $verbosity ";
326    if (defined $self->{'convert_options'}) {
327    $cmd .= $self->{'convert_options'} . " ";
328    }
329    if ($self->{'use_strings'}) {
330      $cmd .= "-use_strings ";
331    }
332    $cmd .= "-errlog \"$errlog\" -output $output_type \"$tmp_filename\"";
333
334    $output_type = `$cmd`;
335
336    # remove symbolic link to original file
337    &util::rm($tmp_filename);
338   
339    # Check STDERR here
340    chomp $output_type;
341    if ($output_type eq "fail") {
342    print $outhandle "Could not convert $tailname$suffix to $convert_to format\n";
343    print $failhandle "$tailname$suffix: " . ref($self) . " failed to convert to $convert_to\n";
344    # The following  meant that if a conversion failed, the document would be counted twice - do we need it for anything?
345    #$self->{'num_not_processed'} ++;
346    if (-s "$errlog") {
347        open(ERRLOG, "$errlog");
348        while (<ERRLOG>) {
349        print $outhandle "$_";
350        }
351        print $outhandle "\n";
352        close ERRLOG;
353    }
354    &util::rm("$errlog") if (-e "$errlog");
355    return "";
356    }
357
358    # store the *actual* output type and return the output filename
359    # it's possible we requested conversion to html, but only to text succeeded
360    #$self->{'convert_to_ext'} = $output_type;
361    if ($output_type =~ /html/i) {
362    $self->{'converted_to'} = "HTML";
363    } elsif ($output_type =~ /te?xt/i) {
364    $self->{'converted_to'} = "Text";
365    } elsif ($output_type =~ /item/i){
366    $self->{'converted_to'} = "PagedImage";
367    }
368   
369    my $output_filename = $tmp_filename;
370    if ($output_type =~ /item/i) {
371    # running under windows
372    if ($ENV{'GSDLOS'} =~ /^windows$/i) {
373        $output_filename = $tmp_dirname . "\\$tailname\\" . $tailname . ".$output_type";
374    } else {
375        $output_filename = $tmp_dirname . "\/$tailname\/" . $tailname . ".$output_type";
376    }
377    } else {
378    $output_filename =~ s/$suffix$/.$output_type/;
379    }
380   
381    return $output_filename;
382}
383
384
385# Override BasPlug read_into_doc_obj - we need to call secondary plugin stuff
386sub read_into_doc_obj {
387    my $self = shift (@_);
388    my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
389
390    my $outhandle = $self->{'outhandle'};
391
392    my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
393
394    my $output_ext = $self->{'convert_to_ext'};
395    my $conv_filename = "";
396    $conv_filename = $self->tmp_area_convert_file($output_ext, $filename_full_path);
397       
398    if ("$conv_filename" eq "") {return -1;} # had an error, will be passed down pipeline
399    if (! -e "$conv_filename") {return -1;}
400    $self->{'conv_filename'} = $conv_filename;
401    $self->convert_post_process($conv_filename);
402
403    # Run the "fribidi" (http://fribidi.org) Unicode Bidirectional Algorithm program over the converted file
404    # Added for fixing up Persian PDFs after being processed by pdftohtml, but may be useful in other cases too
405    if ($self->{'apply_fribidi'} && $self->{'converted_to'} =~ /(HTML|Text)/) {
406    my $fribidi_command = "fribidi \"$conv_filename\" >\"${conv_filename}.tmp\"";
407    if (system($fribidi_command) != 0) {
408        print STDERR "ERROR: Cannot run fribidi on \"$conv_filename\".\n";
409    }
410    else {
411        &util::mv("${conv_filename}.tmp", $conv_filename);
412    }   
413    }
414
415    my $secondary_plugins =  $self->{'secondary_plugins'};
416    my $num_secondary_plugins = scalar(keys %$secondary_plugins);
417
418    if ($num_secondary_plugins == 0) {
419    print $outhandle "Warning: No secondary plugin to use in conversion.  Skipping $file\n";
420    return 0; # effectively block it
421    }
422
423    my @plugin_names = keys %$secondary_plugins;
424    my $plugin_name = shift @plugin_names;
425   
426    if ($num_secondary_plugins > 1) {
427    print $outhandle "Warning: Multiple secondary plugins not supported yet!  Choosing $plugin_name\n.";
428    }
429   
430    my $secondary_plugin = $secondary_plugins->{$plugin_name};
431
432    # note: metadata is not carried on to the next level
433## **** I just replaced $metadata with {} in following
434    my ($rv,$doc_obj)
435    = $secondary_plugin->read_into_doc_obj ($pluginfo,"", $conv_filename, $block_hash, {}, $processor, $maxdocs, $total_count, $gli);
436
437    if ((!defined $rv) || ($rv<1)) {
438    # wasn't processed
439    return $rv;
440    }
441   
442    # Override previous gsdlsourcefilename set by secondary plugin
443    my $collect_file = &util::filename_within_collection($filename_full_path);
444    my $collect_conv_file = &util::filename_within_collection($conv_filename);
445    $doc_obj->set_source_filename ($collect_file, $self->{'file_rename_method'});
446    ## set_source_filename does not set the doc_obj source_path which is used in archives dbs for incremental
447    # build. so set it manually.
448    $doc_obj->{'source_path'} = $filename_full_path;
449    $doc_obj->set_converted_filename($collect_conv_file);
450
451    $self->set_Source_metadata($doc_obj, $filename_no_path);
452       
453    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
454    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FileSize", (-s $filename_full_path));
455
456    # ****
457    my ($tailname, $dirname, $suffix)
458    = &File::Basename::fileparse($filename_full_path, "\\.[^\\.]+\$");
459$doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FilenameRoot", $tailname);
460
461    # do plugin specific processing of doc_obj
462    unless (defined ($self->process($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
463    print STDERR "<ProcessingError n='$file'>\n" if ($gli);
464    return -1;
465    }
466
467    my $topsection = $doc_obj->get_top_section();
468    $self->add_associated_files($doc_obj, $filename_full_path);
469
470    # extra_metadata is already called by sec plugin in process??
471    $self->extra_metadata($doc_obj, $topsection, $metadata); # do we need this here??
472    # do any automatic metadata extraction
473    $self->auto_extract_metadata ($doc_obj);
474
475    # have we found a Title??
476    $self->title_fallback($doc_obj,$topsection,$filename_no_path);
477
478    $self->add_OID($doc_obj);
479
480    return (1, $doc_obj);
481
482}
483
484sub process {
485    my $self = shift (@_);
486    my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
487
488    return $self->process_type($base_dir, $file, $doc_obj);
489}
490
491# do plugin specific processing of doc_obj for doc_ext type
492sub process_type {
493    my $self = shift (@_);
494    my ($base_dir, $file, $doc_obj) = @_;
495   
496    # need to check that not empty
497    my $doc_ext = $self->{'filename_extension'};
498    my $file_type = "unknown";
499    $file_type = $self->{'file_type'} if defined $self->{'file_type'};
500   
501    # associate original file with doc object
502    my $cursection = $doc_obj->get_top_section();
503    my $filename = &util::filename_cat($base_dir, $file);
504    my $assocfilename = "doc.$doc_ext";
505    if ($self->{'keep_original_filename'} == 1) {
506    # this should be the same filename that was used for the Source and SourceFile metadata,
507    # as we will use [SourceFile] in the srclink
508    $assocfilename = $doc_obj->get_assocfile_from_sourcefile();
509    }
510    $doc_obj->associate_file($filename, $assocfilename, undef, $cursection);
511
512    # We use set instead of add here because we only want one value
513    $doc_obj->set_utf8_metadata_element($cursection, "FileFormat", $file_type);
514    my $srclink_filename = "doc.$doc_ext";
515    #my $doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/[assocfilepath]/doc.$doc_ext\">";
516    if ($self->{'keep_original_filename'} == 1) {
517    $srclink_filename = "[SourceFile]";
518    #$doclink = "<a href=\"_httpprefix_/collect/[collection]/index/assoc/[assocfilepath]/[SourceFile]\">";
519    }
520    #$doc_obj->add_utf8_metadata ($cursection, "srclink",  $doclink);
521    $doc_obj->add_utf8_metadata ($cursection, "srcicon",  "_icon".$doc_ext."_");
522    #$doc_obj->add_utf8_metadata ($cursection, "/srclink", "</a>");
523    $doc_obj->add_utf8_metadata ($cursection, "srclink_file", $srclink_filename);
524    return 1;
525}
526
5271;
528
529
530
531
532
533
534
Note: See TracBrowser for help on using the browser.