#!/usr/bin/perl -w

# Count the number of dc identifiers in each repository and 
# group into several characteristics.

use strict;

my $last_result = "http_responses/2005-02-27result.txt";
my $url_file = "urls_by_article_complete.txt";

my $purl_ids   = "all_purl_ids.txt";
my $doi_ids    = "all_doi_ids.txt";
my $handle_ids = "all_handle_ids.txt";


open(PURLS, ">$purl_ids") || die($!);
open(DOIS, ">$doi_ids") || die($!);
open(HANDLES, ">$handle_ids") || die($!);

print "PURLs\tHandles\tDOIs\n";

my $purl_count = 0;
my $handle_count = 0;
my $doi_count = 0;
		
open(FILE, $url_file) || die("Can't open $url_file: $!");
while (my $line = <FILE>) {
	chomp($line);
	
	# Only look at non-article urls
	if ($line =~ /^\t/) {		
		
		my ($id) = $line =~ /^\t(.+)/;

		if ($id =~ /^http:\/\/purl\..+\/.+/ ||
			$id =~ /^http:\/\/bibpurl\./) {
			$purl_count++;
			
			# See if url was found at last check
		
			print PURLS "$id\t" . found_url($id) . "\n";
		}

		if ($id =~ /^http:\/\/hdl.handle.net\//) {
			$handle_count++;
			print HANDLES "$id\t". found_url($id) . "\n";
		}
		
		# DOIs have a prefix and suffix. Example: 10.1000.10/123456 
		# Prefix must start with 10, prefix can have any number of ##.##, 
		# suffix can have any number and type of chars				
		if ($id =~ /\/10(\.[0-9]+)+\/.+/) {
			$doi_count++;
			print DOIS "$id\t" . found_url($id) . "\n";;
		}			
	}
}

close FILE;

print "$purl_count\t$handle_count\t$doi_count\n";


close HANDLES;
close PURLS;
close DOIS;



sub found_url {

	my $url_find = shift;

	my $file;

	open(F, $last_result) || die("Can't open $last_result: $!\n");
	while (my $line = <F>) {
		chomp($line);

		my ($url, $len, $code) = split(/ /, $line);

		if ($url eq $url_find) {
			$code = 200 if ($code == 304 || $code == 301 || $code == 302);
			$code = "200-0" if ($code == 200 && $len == 0);
			if ($code eq '200') {
				close F;
				return 1;
			}
			else {
				close F;
				return 0;
			}
		}
		
	}
	close F;

	# never found url
	return -1;   
}