###########################################################################
#
# ConvertToPlug.pm -- plugin that inherits from HTML or TEXT Plug, depending
#                     on plugin argument convert_to
#
# 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 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.
#
###########################################################################

# The plugin is inherited by such plugins as WordPlug and PDFPlug.
# It facilitates the conversion of these document types to either HTML
# or TEXT by setting up variable that instruct ConvertToBasPlug
# how to work.

# It works by dynamically inheriting HTMLPlug or TEXTPlug based on
# the plugin argument 'convert_to'.  If the argument is not present,
# the default is to inherit HTMLPlug.


package ConvertToPlug;

use BasPlug;
use HTMLPlug;
use TEXTPlug;
use ghtml;

sub BEGIN {
    @ISA = ('HTMLPlug');
#    @ISA = ('HTMLPlug', 'TEXTPlug');
#    @ISA = ('BasPlug'); #, 'HTMLPlug', 'TEXTPlug');
}

my $convert_to_list =
    [ {	'name' => "html",
	'desc' => "{ConvertToPlug.convert_to.html}" },
      {	'name' => "text",
	'desc' => "{ConvertToPlug.convert_to.text}" } ];

my $arguments =
    [ { 'name' => "convert_to",
	'desc' => "{ConvertToPlug.convert_to}",
	'type' => "enum",
	'reqd' => "yes",
	'list' => $convert_to_list, 
	'deft' => "html" },
      { 'name' => "use_strings",
	'desc' => "{ConvertToPlug.use_strings}",
	'type' => "flag",
	'reqd' => "no" } ];

my $options = { 'name'     => "ConvertToPlug",
		'desc'     => "{ConvertToPlug.desc}",
		'abstract' => "yes",
		'inherits' => "yes",
		'args'     => $arguments };

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

    my $plugin_name = $class;
    $plugin_name =~ s/\.pm$//;

    my $newargs = {};

    if (!parsargv::parse($args,  
			 q^extract_keyphrases^, \$newargs->{'kea'}, #with extra options (undocumented) 
			 q^extract_keyphrase_options/.*/^, \$newargs->{'kea_options'}, #no extra options (undocumented)
			 q^convert_to/(html|text)/html^, \$newargs->{'generate_format'},
			 q^use_strings^, \$newargs->{'use_strings'},
			 "allow_extra_options")) {

	print STDERR "\nIncorrect options passed to $plugin_name, ";
	print STDERR "check your collect.cfg configuration file\n";
	$self->print_txt_usage("");  # Use default resource bundle
	die "\n";
    }
   
    return ($plugin_name, $newargs);
}

sub new {
    my $class = shift (@_);
    if ($class eq "ConvertToPlug" && defined $_[0]) {$class = shift (@_);}
    my $self;
    # parsargv::parse might modify the list, so we do this by creating a copy
    # of the argument list.
    my @arglist = @_;
    my ($plugin_name, $args) = $class->parse_args(\@_);

    if ($class eq "PDFPlug" && $args->{'generate_format'} eq "text" && 
	$ENV{'GSDLOS'} =~ /^windows$/i) {
	print STDERR "Windows does not support pdf to text. PDFs will be converted to HTML instead\n";
	$args->{'generate_format'} = "html";
    }

    if ($args->{'generate_format'} eq "text")
    {
	$self = new TEXTPlug ($class, @arglist);
	$self->{'convert_to'} = "TEXT";
	$self->{'convert_to_ext'} = "txt";
    }
    else
    {
	$self = new HTMLPlug ($class, @arglist);
	$self->{'convert_to'} = "HTML";
	$self->{'convert_to_ext'} = "html";

	$self->{'rename_assoc_files'} = 1;
	$self->{'metadata_fields'} .= ",GENERATOR";
    }

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

    foreach my $key (keys %$args) {
      $self->{$key} = $args->{$key};
    }
 
    return bless $self, $class;
}



# Run conversion utility on the input file.  
#
# The conversion takes place in a collection specific 'tmp' directory so 
# that we don't accidentally damage the input.
#
# The desired output type is indicated by $output_ext.  This is usually
# something like "html" or "word", but can be "best" (or the empty string)
# to indicate that the conversion utility should do the best it can.

sub tmp_area_convert_file {
    my $self = shift (@_);
    my ($output_ext, $input_filename, $textref) = @_;

    my $outhandle = $self->{'outhandle'};
    my $convert_to = $self->{'convert_to'};
    my $failhandle = $self->{'failhandle'};

    # softlink to collection tmp dir
    my $tmp_dirname 
	= &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
    &util::mk_dir($tmp_dirname) if (!-e $tmp_dirname);

    # derive tmp filename from input filename
    my ($tailname, $dirname, $suffix)
	= &File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");

    # Remove any white space from filename -- no risk of name collision, and
    # makes later conversion by utils simpler. Leave spaces in path...
    $tailname =~ s/\s+//g; 

    my $tmp_filename = &util::filename_cat($tmp_dirname, "$tailname$suffix");

    &util::soft_link($input_filename, $tmp_filename);

    my $verbosity = $self->{'verbosity'};
    if ($verbosity > 0) {
	print $outhandle "Converting $tailname$suffix to $convert_to format\n";
    }

    my $errlog = &util::filename_cat($tmp_dirname, "err.log");
    
    # Execute the conversion command and get the type of the result,
    # making sure the converter gives us the appropriate output type
    my $output_type = lc($convert_to);
    my $cmd = "perl -S gsConvert.pl -verbose $verbosity ";
    if (defined $self->{'convert_options'}) {
	$cmd .= $self->{'convert_options'} . " ";
    }
    if ($self->{'use_strings'}) {
      $cmd .= "-use_strings ";
    }
    $cmd .= "-errlog \"$errlog\" -output $output_type \"$tmp_filename\"";

    $output_type = `$cmd`;

    # remove symbolic link to original file
    &util::rm($tmp_filename);

    # Check STDERR here
    chomp $output_type;
    if ($output_type eq "fail") {
	print $outhandle "Could not convert $tailname$suffix to $convert_to format\n";
	print $failhandle "$tailname$suffix: " . ref($self) . " failed to convert to $convert_to\n";
	$self->{'num_not_processed'} ++;
	if (-s "$errlog") {
	    open(ERRLOG, "$errlog");
	    while (<ERRLOG>) {
		print $outhandle "$_";
	    }
	    print $outhandle "\n";
	    close ERRLOG;
	}
	&util::rm("$errlog") if (-e "$errlog");
	return "";
    }

    # store the *actual* output type and return the output filename
    # it's possible we requested conversion to html, but only to text succeeded

    $self->{'convert_to_ext'} = $output_type;
    if ($output_type =~ /html/i) {
	$self->{'converted_to'} = "HTML";
    } elsif ($output_type =~ /te?xt/i) {
	$self->{'converted_to'} = "TEXT";
    }
    my $output_filename = $tmp_filename;

    $output_filename =~ s/$suffix$/.$output_type/;

    return $output_filename;
}


# Remove collection specific tmp directory and all its contents.

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

    my $tmp_dirname 
	= &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
    &util::rm_r($tmp_dirname);
    &util::mk_dir($tmp_dirname);
}




# Override BasPlug read
# We don't want to get language encoding stuff until after we've converted
# our file to either TEXT or HTML.
sub read {
    my $self = shift (@_);
    my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs) = @_;
#    if ($self->is_recursive()) {
#        die "BasPlug::read function must be implemented in sub-class for recursive plugins\n";
#    }

    my $outhandle = $self->{'outhandle'};

    my $filename = $file;
    $filename = &util::filename_cat ($base_dir, $file) if $base_dir =~ /\w/;

    if ($self->{'block_exp'} ne "" && $filename =~ /$self->{'block_exp'}/) {
	$self->{'num_blocked'} ++;
	return 0;
    }
    if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
        return undef;
    }
    $file =~ s/^[\/\\]+//; # $file often begins with / so we'll tidy it up

    # read in file ($text will be in utf8)
    my $text = "";

    my $output_ext = $self->{'convert_to_ext'};
    my $conv_filename = $self->tmp_area_convert_file($output_ext, $filename);

    if ("$conv_filename" eq "") {return 0;} # allows continue on errors
    if (! -e "$conv_filename") {return 0;} # allows continue on errors
    $self->{'conv_filename'} = $conv_filename;

    # Do encoding stuff
    my ($language, $encoding) = $self->textcat_get_language_encoding ($conv_filename);

    &BasPlug::read_file($self, $conv_filename, $encoding, $language, \$text);
    if (!length ($text)) {
	my $plugin_name = ref ($self);
        print $outhandle "$plugin_name: ERROR: $file contains no text\n" if $self->{'verbosity'};
        return 0;
    }

    # if we converted to HTML, convert &eacute; and etc to utf-8.
    # this should really happen before language_extraction, but that means
    # modifying a file on disk...
    $text =~ s/&([^;]+);/&ghtml::getcharequiv($1,0)/ge;

    # create a new document
    #my $doc_obj = new doc ($conv_filename, "indexed_doc");
    # now we use the original filename here
    my $doc_obj = new doc($filename, "indexed_doc");
    $doc_obj->set_converted_filename($conv_filename);
    $doc_obj->set_OIDtype ($processor->{'OIDtype'});
    $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Language", $language);
    $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Encoding", $encoding);
    my ($filemeta) = $file =~ /([^\\\/]+)$/;
    $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Source", &ghtml::dmsafe($filemeta));
    if ($self->{'cover_image'}) {
	$self->associate_cover_image($doc_obj, $filename);
    }
    $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
    # include any metadata passed in from previous plugins 
    # note that this metadata is associated with the top level section
    $self->extra_metadata ($doc_obj, $doc_obj->get_top_section(), $metadata);
    # do plugin specific processing of doc_obj
    return -1 unless defined ($self->process (\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj));
    # do any automatic metadata extraction
    $self->auto_extract_metadata ($doc_obj);
    # add an OID
    $doc_obj->set_OID();
    # process the document
    $processor->process($doc_obj);
    $self->cleanup_tmp_area();

    $self->{'num_processed'} ++;

    return 1;
}


# do plugin specific processing of doc_obj for HTML type
sub process_type {
    my $self = shift (@_);
    my ($doc_ext, $textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
    
    my $conv_filename = $self->{'conv_filename'};
    my $tmp_dirname = File::Basename::dirname($conv_filename);
    my $tmp_tailname = File::Basename::basename($conv_filename);
    
    my $converted_to = $self->{'converted_to'};
    my $ret_val;    

    if ($converted_to eq "TEXT")
    {

	$ret_val = &TEXTPlug::process($self, $textref, $pluginfo,
				      $tmp_dirname, $tmp_tailname,
				      $metadata, $doc_obj);
    }
    else
    {
	$ret_val = &HTMLPlug::process($self, $textref, $pluginfo,
				      $tmp_dirname, $tmp_tailname,
				      $metadata, $doc_obj);
    }

    # associate original file with doc object
    my $cursection = $doc_obj->get_top_section();
    my $filename = &util::filename_cat($base_dir, $file);
    $doc_obj->associate_file($filename, "doc.$doc_ext", undef, $cursection);

    my $file_type;

    if ($doc_ext eq "doc") {
    	$file_type = "Word";
    } elsif ($doc_ext eq "xls") {
	$file_type = "Excel";
    } elsif ($doc_ext eq "ppt") {
	$file_type = "PPT"; 
    } elsif ($doc_ext eq "pdf") {
	$file_type = "PDF"; 
    } elsif ($doc_ext eq "rtf") {
	$file_type = "RTF";
    }

    my $file_format = $file_type || "unknown";
	
    $doc_obj->add_metadata($cursection, "FileFormat", $file_format);

    my $doclink = "<a href=\"_httpcollection_/index/assoc/[archivedir]/doc.$doc_ext\">";
    $doc_obj->add_utf8_metadata ($cursection, "srclink",  $doclink); 
    $doc_obj->add_utf8_metadata ($cursection, "srcicon",  "_icon".$doc_ext."_"); 
    $doc_obj->add_utf8_metadata ($cursection, "/srclink", "</a>"); 

    return $ret_val;
}

1;