root/trunk/gsdl/perllib/plugins/PDFPlug.pm @ 11122

Revision 11122, 11.9 KB (checked in by davidb, 14 years ago)

Introduction of -associate_tail_re option to BasPlug?. This is a generalisation
of the associate_ext (in fact associate_ext is now mapped into the equivalent
assoicate_tail_re expression). To work properly, plugins that use the
secondary plugin mechanism (such as PDFPlug) need to duplicate the
associate_tail_re value secondary plug ins too.

  • Property svn:keywords set to Author Date Id Revision
Line 
1###########################################################################
2#
3# PDFPlug.pm -- reasonably with-it pdf plugin
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) 1999-2001 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###########################################################################
25package PDFPlug;
26
27use ConvertToPlug;
28use unicode;
29use strict;
30no strict 'refs'; # so we can use a var for filehandles (eg STDERR)
31
32sub BEGIN {
33    @PDFPlug::ISA = ('ConvertToPlug');
34}
35
36my $convert_to_list =
37    [ { 'name' => "auto",
38    'desc' => "{ConvertToPlug.convert_to.auto}" },
39      { 'name' => "html",
40    'desc' => "{ConvertToPlug.convert_to.html}" },
41      { 'name' => "text",
42    'desc' => "{ConvertToPlug.convert_to.text}" },
43      { 'name' => "pagedimg_jpg",
44    'desc' => "{ConvertToPlug.convert_to.pagedimg_jpg}"},
45      { 'name' => "pagedimg_gif",
46    'desc' => "{ConvertToPlug.convert_to.pagedimg_gif}"},
47      { 'name' => "pagedimg_png",
48    'desc' => "{ConvertToPlug.convert_to.pagedimg_png}"},
49      ];
50
51
52my $arguments =
53    [
54     { 'name' => "convert_to",
55       'desc' => "{ConvertToPlug.convert_to}",
56       'type' => "enum",
57       'reqd' => "yes",
58       'list' => $convert_to_list,
59       'deft' => "html" },   
60     { 'name' => "process_exp",
61       'desc' => "{BasPlug.process_exp}",
62       'type' => "regexp",
63       'deft' => &get_default_process_exp(),
64       'reqd' => "no" },
65     { 'name' => "block_exp",
66       'desc' => "{BasPlug.block_exp}",
67       'type' => "regexp",
68       'deft' => &get_default_block_exp() },
69     { 'name' => "metadata_fields",
70       'desc' => "{HTMLPlug.metadata_fields}",
71       'type' => "string",
72       'deft' => "" },
73     { 'name' => "noimages",
74       'desc' => "{PDFPlug.noimages}",
75       'type' => "flag" },
76     { 'name' => "allowimagesonly",
77       'desc' => "{PDFPlug.allowimagesonly}",
78       'type' => "flag" },
79     { 'name' => "complex",
80       'desc' => "{PDFPlug.complex}",
81       'type' => "flag" },
82     { 'name' => "nohidden",
83       'desc' => "{PDFPlug.nohidden}",
84       'type' => "flag" },
85     { 'name' => "zoom",
86       'desc' => "{PDFPlug.zoom}",
87       'deft' => "2",
88       'range' => "1,3", # actually the range is 0.5-3
89       'type' => "int" },
90     { 'name' => "use_sections",
91       'desc' => "{PDFPlug.use_sections}",
92       'type' => "flag" },
93     { 'name' => "description_tags",
94       'desc' => "{HTMLPlug.description_tags}",
95       'type' => "flag" }
96     ];
97
98my $options = { 'name'     => "PDFPlug",
99        'desc'     => "{PDFPlug.desc}",
100        'abstract' => "no",
101        'inherits' => "yes",
102        'args'     => $arguments };
103
104sub new {
105    my ($class) = shift (@_);
106    my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
107    push(@$pluginlist, $class);
108
109    push(@$inputargs,"-title_sub");
110    push(@$inputargs,'^(Page\s+\d+)?(\s*1\s+)?');
111
112    if(defined $arguments){ push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});}
113    if(defined $options) { push(@{$hashArgOptLists->{"OptList"}},$options)};
114
115    my @arg_array = @$inputargs;
116    my $self = (defined $hashArgOptLists)? new ConvertToPlug($pluginlist,$inputargs,$hashArgOptLists): new ConvertToPlug($pluginlist,$inputargs);
117   
118    if ($self->{'info_only'}) {
119    # don't worry about any options etc
120    return bless $self, $class;
121    }
122
123    # these are passed through to gsConvert.pl by ConvertToPlug.pm
124    my $zoom = $self->{"zoom"};
125    $self->{'convert_options'} = "-pdf_zoom $zoom";
126    $self->{'convert_options'} .= " -pdf_complex" if $self->{"complex"};
127    $self->{'convert_options'} .= " -pdf_nohidden" if $self->{"nohidden"};
128    $self->{'convert_options'} .= " -pdf_ignore_images" if $self->{"noimages"};
129    $self->{'convert_options'} .= " -pdf_allow_images_only" if $self->{"allowimagesonly"};
130
131    my $secondary_plugin_options = $self->{'secondary_plugin_options'};
132
133    if (!defined $secondary_plugin_options->{'HTMLPlug'}) {
134    $secondary_plugin_options->{'HTMLPlug'} = [];
135    }
136    if (!defined $secondary_plugin_options->{'TEXTPlug'}) {
137    $secondary_plugin_options->{'TEXTPlug'} = [];
138    }
139    if (defined $self->{'convert_to'} && $self->{'convert_to'} =~ /pagedimg.*/i) {
140    if (!defined $secondary_plugin_options->{'PagedImgPlug'}){
141        $secondary_plugin_options->{'PagedImgPlug'} = [];
142        my $pagedimg_options = $secondary_plugin_options->{'PagedImgPlug'};
143        push(@$pagedimg_options, "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
144    }
145    }
146    my $html_options = $secondary_plugin_options->{'HTMLPlug'};
147    my $text_options = $secondary_plugin_options->{'TEXTPlug'};
148    my $pagedimg_options = $secondary_plugin_options->{'PagedImgPlug'};
149   
150    if ($self->{'input_encoding'} eq "auto") {
151    # pdftohtml will always produce html files encoded as utf-8
152    # => restrict primary PDFPlug and secondary HTML plugin to use
153    # utf8 and extract language.
154    $self->{'input_encoding'} = "utf8";
155    $self->{'extract_language'} = 1;
156
157    push(@$html_options,"-extract_language");
158    }
159    # if pdftohtml is always producing utf8, then htmlplug always needs this option
160    push(@$html_options,"-input_encoding", "utf8");
161
162    # Instruct HTMLPlug (when eventually accessed through read_into_doc_obj)
163    # to extract these metadata fields from the HEAD META fields
164    my $required_metadata;
165    if (defined $self->{'metadata_fields'} && $self->{'metadata_fields'} =~ /\S/) {
166    push(@$html_options,"-metadata_fields",$self->{'metadata_fields'});
167    } else {
168    push(@$html_options,"-metadata_fields","Title,GENERATOR,date,author<Creator>");
169    }
170    #push(@$html_options,"-metadata_fields","Title,GENERATOR,date,author<Creator>");
171       
172    if ($self->{'use_sections'} || $self->{'description_tags'}) {
173    $self->{'description_tags'} = 1;
174    push(@$html_options,"-description_tags");
175    }
176
177    # following title_sub removes "Page 1" added by pdftohtml, and a leading
178    # "1", which is often the page number at the top of the page. Bad Luck
179    # if your document title actually starts with "1 " - is there a better way?
180    push(@$html_options , "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
181    push(@$text_options , "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
182
183    my $associate_tail_re = $self->{'associate_tail_re'};
184    if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
185    push(@$html_options, "-associate_tail_re", $associate_tail_re);
186    push(@$text_options, "-associate_tail_re", $associate_tail_re);
187    push(@$pagedimg_options, "-associate_tail_re", $associate_tail_re);
188    }
189
190 
191    $self = bless $self, $class;
192    $self->load_secondary_plugins($class,$secondary_plugin_options,$hashArgOptLists);
193    return $self;
194}
195
196sub get_default_process_exp {
197    my $self = shift (@_);
198
199    return q^(?i)\.pdf$^;
200}
201
202# so we don't inherit HTMLPlug's block exp...
203sub get_default_block_exp {
204    return "";
205}
206   
207sub convert_post_process
208{
209    my $self = shift (@_);
210    my ($conv_filename) = @_;
211
212    my $outhandle=$self->{'outhandle'};
213
214    my ($language, $encoding) = $self->textcat_get_language_encoding ($conv_filename);
215
216    # read in file ($text will be in utf8)
217    my $text = "";
218    $self->read_file ($conv_filename, $encoding, $language, \$text);
219
220    # Calculate number of pages based on <a ...> tags (we have a <a name=1> etc
221    # for each page).  Metadata based on this calculation not set until process()
222    #
223    # Note: this is done even if we are not breaking to document into pages as it might
224    # be useful to give an indication of document length in browser through setting
225    # num_pages as metadata.
226    my @pages = ($text =~ /\<[Aa] name=\"?\w+\"?>/ig);
227    my $num_pages = scalar(@pages);
228    $self->{'num_pages'} = $num_pages;
229
230    if ($self->{'use_sections'}
231    && $self->{'converted_to'} eq "HTML") {
232
233    print $outhandle "PDFPlug: Calculating sections...\n";
234
235    # we have "<a name=1></a>" etc for each page
236    # it may be <A name=
237    my @sections = split('<[Aa] name=', $text);
238
239    my $top_section = "";
240
241    if (scalar (@sections) == 1) { #only one section - no split!
242        print $outhandle "PDFPlug: warning - no sections found\n";
243    } else {
244        $top_section .= shift @sections; # keep HTML header etc as top_section
245    }
246
247    # handle first section specially for title? Or all use first 100...
248   
249    my $title = $sections[0];
250    $title =~ s/^\"?\w+\"?>//; # specific for pdftohtml...
251    $title =~ s/<\/([^>]+)><\1>//g; # (eg) </b><b> - no space
252    $title =~ s/<[^>]*>/ /g;
253    $title =~ s/(?:&nbsp;|\xc2\xa0)/ /g; # utf-8 for nbsp...
254    $title =~ s/^\s+//s;
255    $title =~ s/\s+$//;
256    $title =~ s/\s+/ /gs;
257    $title =~ s/^$self->{'title_sub'}// if ($self->{'title_sub'});
258    $title =~ s/^\s+//s; # in case title_sub introduced any...
259    $title = substr ($title, 0, 100);
260    $title =~ s/\s\S*$/.../;
261
262
263    if (scalar (@sections) == 1) { # no sections found
264        $top_section .= $sections[0];
265        @sections=();
266    } else {
267        $top_section .= "<!--<Section>\n<Metadata name=\"Title\">$title</Metadata>\n-->\n <!--</Section>-->\n";
268    }
269
270    # add metadata per section...
271    foreach my $section (@sections) {
272        # section names are not always just digits, may be like "outline"
273        $section =~ s@^\"?(\w+)\"?></a>@@; # leftover from split expression...
274
275        $title = $1; # Greenstone does magic if sections are titled digits
276        if (! defined($title) ) {
277        print STDERR "no title: $section\n";
278        $title = " "; # get rid of the undefined warning in next line
279        }
280        my $newsection = "<!-- from PDFPlug -->\n<!-- <Section>\n";
281        $newsection .= "<Metadata name=\"Title\">" . $title
282        . "</Metadata>\n--><p>\n";
283        $newsection .= $section;
284        $newsection .= "<!--</Section>-->\n";
285        $section = $newsection;
286    }
287
288    $text=join('', ($top_section, @sections));
289    }
290
291    # turn any high bytes that aren't valid utf-8 into utf-8.
292    unicode::ensure_utf8(\$text);
293
294    # Write it out again!
295    $self->utf8_write_file (\$text, $conv_filename);
296}
297
298
299# do plugin specific processing of doc_obj for HTML type
300sub process {
301    my $self = shift (@_);
302    my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
303
304    my $result = $self->process_type("pdf",$base_dir,$file,$doc_obj);
305
306    # fix up the extracted date metadata to be in Greenstone date format,
307    # and fix the capitalisation of 'date'
308    my $cursection = $doc_obj->get_top_section();
309    foreach my $datemeta (@{$doc_obj->get_metadata($cursection, "date")}) {
310    $doc_obj->delete_metadata($cursection, "date", $datemeta);
311
312    # We're just interested in the date bit, not the time
313    # some pdf creators (eg "Acrobat 5.0 Scan Plug-in for Windows")
314    # set a /CreationDate, and set /ModDate to 000000000. pdftohtml
315    # extracts the ModDate, so it is 0...
316    $datemeta =~ /(\d+)-(\d+)-(\d+)/;
317    my ($year, $month, $day) = ($1,$2,$3);
318    if (defined($year) && defined($month) && defined($day)) {
319        if ($year == 0) {next}
320        if ($year < 100) {$year += 1900} # just to be safe
321        if ($month =~ /^\d$/) {$month="0$month"} # single digit
322        if ($day =~ /^\d$/) {$day="0$day"} # single digit
323        my $date="$year$month$day";
324        $doc_obj->add_utf8_metadata($cursection, "Date", $date);
325    }
326    }
327
328    $doc_obj->add_utf8_metadata($cursection, "NumPages", $self->{'num_pages'});
329   
330    if ($self->{'use_sections'} && $self->{'converted_to'} eq "HTML") {
331    # we explicitly make it a paged document, cos greenstone won't get it
332    # right if any section has an empty title, or one with letters in it
333    $doc_obj->set_utf8_metadata_element ($cursection, "gsdlthistype", "Paged");
334    }
335
336    return $result;
337}
338
3391;
Note: See TracBrowser for help on using the browser.