#!/usr/bin/perl
# massage B.E. ciders XML files
# MUST RUN ON SOLARIS TO DEAL WITH UNICODE ISSUES
print "\ncdl-ciders v1.1.0\n\n";
print "C F Baum  Apr 2002\n\n";
use LWP::Simple;
use Text::Wrap;
$str = get("http://repositories.cdlib.org/cgi/oai.cgi?verb=ListRecords&metadataPrefix=repec&set=publication:6369646572");
$out[0]=">ciders.rdf";
$jnl[0]="Center for International and Development Economics Research, Working Paper Series";
$han[0]="ciders";
open(OUTA,$out[0]) || die "Cannot open $outa for output";

$nrtem=0;

$str =~ s/\n/ /ig;
$str =~ s/\r/ /ig;
$str =~ s/"/"/ig;
$str =~ s/&/&/ig;
$str =~ s/“/"/ig;
$str =~ s/”/"/ig;
$str =~ s/–/:/ig;
$str =~ s/—/--/ig;

$str =~ s/^.+?<record>//;
(@str) = split("<record>",$str);
$nstr=@str;
print "\n $nstr articles...\n\n";
$debug=1;
foreach $str (@str) {
	$str=&utf8_to_iso8859($str);
	($id) = ($str =~ /<identifier>oai:cdlib1:(.+?)<\/identifier/);
	($credt) = ($str =~ /<datestamp>(.+?)<\/datestamp/);
	($pubyr) = ($credt =~ /^(\d\d\d\d)/);
	($ti) = ($str =~ /<ti>(.+?)<\/ti/);
	(@au) = ($str =~ /<au>(.+?)<\/au/ig);
	$nau = @au;
	($ab) = ($str =~ /<ab>(.+?)<\/ab/);
	($jnl) = ($str =~ /<jrnti>(.+?)<\/jrnti/);
	($vol) = ($str =~ /<vol>(.+?)<\/vol/);
#	($iss) = ($str =~ /<iss>(.+?)<\/iss/);
	($ppf) = ($str =~ /<ppf>(.+?)<\/ppf/);
	($ppl) = ($str =~ /<ppl>(.+?)<\/ppl/);
	($url) = ($str =~ /<url>(.+?)<\/url/);
	$wh = 0;
	$jj="A";
#	print "$id $jnl \n";
	(@kw) = ($str =~ /<kwd>(.+?)<\/kwd/ig);
	$nkw = @kw;
	($jel) = ($str =~ /<categ>(.+?)<\/categ/);
	$oo = "OUT".$jj;
	$jt = $jnl[$wh];
	$hl = $han[$wh];
	print $oo "\nTemplate-Type: ReDIF-Paper 1.0\n";
	print $oo "Title: $ti\n";
	foreach $i (@au) {
		($f) = ($i =~ m/fnm>(.+?)<\/fnm/ig);
		($l) = ($i =~ m/snm>(.+?)<\/snm/ig);
		($afl) = ($i =~ m/affl>(.+?)<\/affl/ig);
	print $oo "Author-Name: $f $l \n";
	print $oo "Author-Workplace-Name: $afl \n";
	}
	if (length $ab > 1) {
		print $oo  "Abstract:\n";
		print $oo  wrap(" "," ",$ab);
		print $oo "\n";
	}
	if ($nkw > 0) {
		print $oo "Keywords: ";
		foreach $k (@kw) {
			print $oo "$k, ";
			}
		print $oo "\n";
	}
	print $oo "Classification-JEL: $jel \n";
	print $oo "Series: $jt \n";
	print $oo "Number: $vol$ppf \n";
    print $oo "Creation-Date: $credt \n";
#	print $oo "Volume: $vol \n";
#	print $oo "Pages: $ppf-$ppl \n";
	print $oo "Note: oai:cdlib1:$id\n";
	print $oo "File-URL: $url \n";
	print $oo "File-Format: application/pdf \n";
	print $oo "Handle: RePEc:cdl:$hl:$vol$ppf \n";
	$nrtem++;
}
print " \n$nrtem templates processed... \n\n";
exit;
sub utf8_to_iso8859
{
# Converts UTF-8 as long as it just encodes ISO-8859-1
	my ($str) = @_;
	my $retval;
	my @chars = split(//,$str);
	while (@chars) {
		my $ch = shift(@chars);
		my $val1 = ord($ch);
		if($val1 >= 128) {
			# XXXX could add error checking to make sure there's a char available.
			my $val2 = ord(shift(@chars));
			$val1 = ($val1 & 31) << 6;
			$val2 &= 63;
			my $realch = chr($val1 + $val2);
			print STDERR "Hello: converted upper ascii char $realch(", ord($realch), ")\n" if ($debug);
			$retval .= $realch;
		} else {
			$retval .= $ch;
		}
	}
	return $retval;
}
__END__