source: trunk/gsdl/perllib/plugins/PDFPlug.pm@ 8795

###########################################################################
#
# PDFPlug.pm -- reasonably with-it pdf plugin
# A component of the Greenstone digital library software
# from the New Zealand Digital Library Project at the 
# University of Waikato, New Zealand.
#
# Copyright (C) 1999-2001 New Zealand Digital Library Project
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#
###########################################################################
use strict;
no strict 'refs'; # so we can use a var for filehandles (eg STDERR)

package PDFPlug;

use ConvertToPlug;
use unicode;

sub BEGIN {
@PDFPlug::ISA = ('ConvertToPlug');
}

my $arguments =
    [ { 'name' => "process_exp",
    'desc' => "{BasPlug.process_exp}",
    'type' => "regexp",
    'deft' => &get_default_process_exp(),
    'reqd' => "no" },
      { 'name' => "block_exp",
    'desc' => "{BasPlug.block_exp}",
    'type' => "regexp",
    'deft' => &get_default_block_exp() },
      { 'name' => "noimages",
    'desc' => "{PDFPlug.noimages}",
    'type' => "flag" },
      { 'name' => "complex",
    'desc' => "{PDFPlug.complex}",
    'type' => "flag" },
      { 'name' => "nohidden",
    'desc' => "{PDFPlug.nohidden}",
    'type' => "flag" },
      { 'name' => "zoom",
    'desc' => "{PDFPlug.zoom}",
    'deft' => "2",
    'range' => "1,3", # actually the range is 0.5-3 
    'type' => "int" },
      { 'name' => "use_sections",
    'desc' => "{PDFPlug.use_sections}",
    'type' => "flag" } ];

my $options = { 'name'     => "PDFPlug",
        'desc'     => "{PDFPlug.desc}",
        'abstract' => "no",
        'inherits' => "yes",
        'args'     => $arguments };

sub new {
    my $class = shift (@_);

    my ($noimages, $complex, $zoom, $use_sections, $nohidden);

    my @args=@_;

    if (!parsargv::parse(\@_, 
             q^noimages^, \$noimages,
             q^complex^, \$complex,
             q^zoom/\d+/2^, \$zoom,
             q^nohidden^, \$nohidden,
             q^use_sections^, \$use_sections,
             "allow_extra_options")) {

    my $self = new ConvertToPlug($class, @_);
    my $outhandle=$self->{'outhandle'};
    print $outhandle "\nIncorrect options passed to PDFPlug, check your collect.cfg configuration file\n";
    $self->print_txt_usage("");  # Use default resource bundle
    exit 1;
    }


    if ($use_sections) {
    push (@args, "-description_tags");
    }

    # following title_sub removes "Page 1" added by pdftohtml, and a leading
    # "1", which is often the page number at the top of the page. Bad Luck
    # if your document title actually starts with "1 " - is there a better way?

    my $self = new ConvertToPlug ($class, @args, "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
    $self->{'plugin_type'} = "PDFPlug";
    if ($use_sections) {
    $self->{'use_sections'}=1;
    }

    # 14-05-02 To allow for proper inheritance of arguments - John Thompson
    my $option_list = $self->{'option_list'};
    push( @{$option_list}, $options );

    # these are passed through to gsConvert.pl by ConvertToPlug.pm
    $self->{'convert_options'} = "-pdf_zoom $zoom";
    $self->{'convert_options'} .= " -pdf_complex" if $complex;
    $self->{'convert_options'} .= " -pdf_nohidden" if $nohidden;
    $self->{'convert_options'} .= " -pdf_ignore_images" if $noimages;

    # pdftohtml will always produce html files encoded as utf-8
    if ($self->{'input_encoding'} eq "auto") {
    $self->{'input_encoding'} = "utf8";
    $self->{'extract_language'} = 1;
    }

    return bless $self, $class;
}

sub get_default_process_exp {
    my $self = shift (@_);

    return q^(?i)\.pdf$^;
}

# so we don't inherit HTMLPlug's block exp...
sub get_default_block_exp {
    return "";
}
    

# do plugin specific processing of doc_obj for HTML type
sub process {
    my $self = shift (@_);
    my $outhandle=$self->{'outhandle'};

    my $textref=$_[0];

    if ($self->{'use_sections'}
    && $self->{'converted_to'} eq "HTML") {

    print $outhandle "PDFPlug: Calculating sections...\n";

    # we have "<a name=1></a>" etc for each page
    # it may be <A name=
    my @sections = split('<[Aa] name=', $$textref);

    if (scalar (@sections) == 1) { #only one section - no split!
        print $outhandle "PDFPlug: warning - no sections found\n";
    } else {
        shift @sections; # don't need HTML header, etc
    }

    # handle first section specially for title? Or all use first 100...
    
    my $title = $sections[0];
    $title =~ s/^\"?\w+\"?>//; # specific for pdftohtml...
    $title =~ s/<\/([^>]+)><\1>//g; # (eg) </b><b> - no space
    $title =~ s/<[^>]*>/ /g;
    $title =~ s/(?:&nbsp;|\xc2\xa0)/ /g; # utf-8 for nbsp...
    $title =~ s/^\s+//s;
    $title =~ s/\s+$//;
    $title =~ s/\s+/ /gs;
    $title =~ s/^$self->{'title_sub'}// if ($self->{'title_sub'});
    $title =~ s/^\s+//s; # in case title_sub introduced any...
    $title = substr ($title, 0, 100);
    $title =~ s/\s\S*$/.../;

    my $top_section;
    if (scalar (@sections) == 1) { # no sections found
        $top_section=$sections[0];
        @sections=();
    } else {
        $top_section = "<!--<Section>\n<Metadata name=\"Title\">$title</Metadata>\n-->\n <!--</Section>-->\n";
    }

    # add metadata per section...
    foreach my $section (@sections) {
        # section names are not always just digits, may be like "outline"
        $section =~ s@^\"?(\w+)\"?></a>@@; # leftover from split expression...

        $title = $1; # Greenstone does magic if sections are titled digits
        if (! defined($title) ) {
        print STDERR "no title: $section\n";
        $title = " "; # get rid of the undefined warning in next line
        }
        my $newsection = "<!-- from PDFPlug -->\n<!-- <Section>\n";
        $newsection .= "<Metadata name=\"Title\">" . $title
        . "</Metadata>\n--><p>\n";
        $newsection .= $section;
        $newsection .= "<!--</Section>-->\n";
        $section = $newsection;
    }

    $$textref=join('', ($top_section, @sections));
    }

    # turn any high bytes that aren't valid utf-8 into utf-8.
    unicode::ensure_utf8($textref);

    print $outhandle "PDFPlug: passing $_[3] on to $self->{'converted_to'}Plug\n"
    if $self->{'verbosity'} > 1;

    # tell htmlplug to extract these metadata fields from the HEAD META fields
    $self->{'metadata_fields'} .= ",date,author<Creator>";

    my $result = ConvertToPlug::process_type($self,"pdf",@_);

    my $doc_obj = pop(@_);

    # fix up the extracted date metadata to be in Greenstone date format,
    # and fix the capitalisation of 'date'
    my $cursection = $doc_obj->get_top_section();

    foreach my $datemeta (@{$doc_obj->get_metadata($cursection, "date")}) {
    $doc_obj->delete_metadata($cursection, "date", $datemeta);

    # We're just interested in the date bit, not the time
    # some pdf creators (eg "Acrobat 5.0 Scan Plug-in for Windows")
    # set a /CreationDate, and set /ModDate to 000000000. pdftohtml
    # extracts the ModDate, so it is 0...
    $datemeta =~ /(\d+)-(\d+)-(\d+)/;
    my ($year, $month, $day) = ($1,$2,$3);
    if (defined($year) && defined($month) && defined($day)) {
        if ($year == 0) {next}
        if ($year < 100) {$year += 1900} # just to be safe
        if ($month =~ /^\d$/) {$month="0$month"} # single digit
        if ($day =~ /^\d$/) {$day="0$day"} # single digit
        my $date="$year$month$day";
        $doc_obj->add_utf8_metadata($cursection, "Date", $date);
    }
    }

    # Add NumPages metadata (we have "<a name=1>" etc for each page)
    my @pages = ($$textref =~ /\<[Aa] name=\"?\w+\"?>/ig);
    $doc_obj->add_utf8_metadata($cursection, "NumPages", scalar(@pages));
    
    if ($self->{'use_sections'} && $self->{'converted_to'} eq "HTML") {
    # we explicitly make it a paged document, cos greenstone won't get it
    # right if any section has an empty title, or one with letters in it
    $doc_obj->set_utf8_metadata_element ($cursection, "gsdlthistype", "Paged");
    }
    return $result;
}

1;