#!/home/eric/bin/perl

# marc2oai - Convert MARC in communications format to simple OAI-XML files

# Eric Lease Morgan <emorgan@nd.edu>
# 2007-07-10 - added prefix
# 2007-06-17 - first cut

# define
use constant STYLE => './marc2oai.xsl';
use constant DATA  => '/home/eric/apache/htdocs/oai/data/books/';

# require
use MARC::Batch;
use MARC::File::XML ( BinaryEncoding => 'utf8', RecordFormat => 'USMARC' );
use strict;
use XML::LibXML;
use XML::LibXSLT;

# get the input
my $filename = shift @ARGV;
my $prefix   = shift @ARGV;
if ( ! $filename or ! $prefix ) {

	print "Usage: $0 filename prefix\n";
	exit;
	
}

# make i/o binary
binmode( STDOUT, ':utf8' );
binmode( $filename, ':bytes' );

# initialize XSLT
my $parser = XML::LibXML->new;
my $xslt   = XML::LibXSLT->new;
my $style  = $parser->parse_file( STYLE ) or croak $!;
my $stylesheet = $xslt->parse_stylesheet( $style ) or croak $!;

# loop through the marc records
my $index = 0;
my $batch  = MARC::Batch->new( 'USMARC', $filename );
while ( my $record = $batch->next ) {
		
	# sanity check
	eval { my $source = $parser->parse_string( MARC::File::XML::record( $record )) };
	if ( $@ ) { next }
	
	# parse and transform
	my $source  = $parser->parse_string( MARC::File::XML::record( $record )) or croak $!;
	my $results = $stylesheet->transform( $source )                          or croak $!;

	# increment the index and save
	$index++;
	open OUTPUT, " > " . DATA . "$prefix-$index.xml" or die "Can't open output ($!)\n";
	print $stylesheet->output_string( $results ) . "\n";
	print OUTPUT $stylesheet->output_string( $results );
	close OUTPUT;

} 

# done
exit;