#!/usr/bin/env perl

##############################################################################
# RDF Grabber
#
# Downloads RDF files and copies them to a position in the local filesystem.
# Gets information on file URL and local destination from a tab seperated 
# input file, where each line is of the form: ^URL\tPath$
#
# NOTE - This script does not support path names that contain whitespace, even
#        escaped whitespace.
#
# $Header: /var/cvs/projects/websites/acherondevelopment/files/parse_rdf/rdfgrabber,v 1.1 2003/06/11 18:19:40 vrai Exp $
##############################################################################

# Get the RDF information file
my $infoFilename = $ARGV [ 0 ] or die ( "Usage: rdfgrabber rdf_info_filename\n" );

# Parse the file
my %rdfs = ParseRDFInfoFile ( $infoFilename );

# Process the RDFs in the hash
ProcessRDFs ( \%rdfs );

exit ( );


##############################################################################
# Function definitions


# Process the RDFs, downloads them and copies them to the given location
#
# Arg0: Reference to the RDF hash, URL=>Path
sub ProcessRDFs ( )
{
	my ( $rdfs  ) = @_;

	foreach my $url ( keys ( %$rdfs ) )
	{
		# Get the path
		my $path = $rdfs->{ $url };
		my $tmpPath = '/tmp/tmp_' . `/usr/bin/env /usr/bin/basename $path` or die ( "Cannot get basename of $path - $!\n" );
		$tmpPath = TrimString ( $tmpPath );

		# Remove any existing temporary file with this name
		unlink ( $tmpPath ) or die ( "Unable to remove temporary file $tmpPath - $!\n" ) if ( stat ( $tmpPath ) );

		# Download the RDF to a local file
		my $output = `/usr/bin/env /usr/bin/wget -O $tmpPath $url 2>&1`;

		# Check the output, if it's valid move the file to it's new location
		my $regTmpPath = quotemeta ( $tmpPath );
		my $mvOutput = `mv $tmpPath $path`;
        my $touchOutput = `touch $path`;
		
		# Remove the temporary file (if it exists) - not too bothered if this doesn't work
		unlink ( $tmpPath );
	}
}


# Parses the RDF info file in to a hash where each pair is: URL=>Path
#
# Ret:  Hash of the form URL=>Path
# Arg0:	File name of RDF Information File
sub ParseRDFInfoFile ( )
{
	my ( $filename ) = @_;
	my %rdfs;

	# Open the file and read it in a line at a time
	open ( INFOFILE, "<$filename" ) or die ( "Unable to open info file \"$filename\" - $!\n" );
	while ( my $line = <INFOFILE> )
	{
		
		# Check that the line is valid
		if ( $line =~ /^http:\/\/[^\t]+[\t]+[^\t]+$/ )
		{
			# Split the line in to its two components and add them to the hash
			my ( $url, $path ) = split ( /[\t]+/, $line, 2 );
			$url = TrimString ( $url );
			$path = TrimString ( $path );
			$rdfs { $url } = $path;
		}
	}
	close ( INFOFILE );

	return %rdfs;
}


# Removes ALL whitespace from a string, even escaped whitespace
#
# Ret:  String sans whitespace
# Arg0: String to trim
sub TrimString ( )
{
	my ( $string ) = @_;
	$string =~ s/[ \t\r\n]+//g;
	return $string;
}