#!/usr/bin/perl -w

use LWP::Simple;
use DBI;
require("db_connect.inc");

# In development mode (use dev db, more output), 0 for normal/live
$dev="0";

if ($dev) { print "user $userid and password $passwd on host $host \n"; }




# details for open database for recording
$db="ircount";

$connectionInfo="dbi:mysql:$db;$host";

if ($dev) {
	$tbl_ircount = "ircount_dev";
}
else {
	$tbl_ircount = "ircount";
}

#
#
# get the data from eprints.org
# first set the url
$URL = "http://roar.eprints.org/rawlist.txt";
# now do the actually getting, easy thanks to lwp::simple
$content = get $URL;
die "Couldn't get it!" unless defined $content;


# split content into seperate lines
(@lines) = split /\n/, $content;

# set up vars
my $data;
my $field;
my $save;
my $id;
my $name;

my $records=0;
my $identifier;
my $currentdate;
my $query;
my $country;
my $type;



#get current time
#which is bloody annoying in perl
($day, $month, $year) = (localtime)[3,4,5];
# perl returns funny values
$year += 1900;
$month += 1;
# and single digit days and months should start with a 0
if ($month < 10) {
	# we're basically turning it into a string here
	# having just treated it like an int. 
	# who needs strongly typed languages!
	$month = "0" . "$month";
}
if ($day < 10) {
	$day = "0" . "$day";
}
$currentdate = "$year$month$day";



# make connection to database
$dbh = DBI->connect($connectionInfo,$userid,$passwd);

#make sure we use utf8 for, especially for the archive names
#(and make sure tbl and field are utf8 general as well!)
$dbh->{'mysql_enable_utf8'} = 1;
$dbh->do('SET NAMES utf8');


# open file for recording data
open (LOGFILE, ">>datalog.txt") or die "can not open datalog \n";
print LOGFILE "METAMETA\t$year $month $day\tstart\n"; 


########################################
# main loop
foreach $line(@lines)
	{
	
	# each record has one field per line, and each record
	# is seperated by an empty line
	# if this is an empty line we have reached the end of 
	# a record and so output to file/db
	# if line empty write stuff to file/db
	if ($line eq "") {
		# blank line means we have reached the end of this record
		# new record
		# a few things before we write the record.
		# skip this if not type:institutional
		next if ($type ne "institutional");
		#
		# write the record out to file
		$save  = "$identifier\t$name\t$records";
		$save .= "\t$identifier\t$currentdate\t$country\n";
		print LOGFILE "$save";	

		# write record to db
		$query = "INSERT INTO $tbl_ircount (archiveid, archivename,";
		$query .= " records,";
		$query .= " currentdate, identifier, ";
		$query .= "country, collected_date)";
		$query .= " VALUES ($identifier, \"$name\",";
		$query .= "$records, $currentdate, \"$identifier\",";
		$query .= " \"$country\", now())";
		#print "$query\n\n";

		$sth = $dbh->prepare($query);
		$sth->execute();

		$query = "";



		# clear values ready for next
		$oai = "";
		$id = "";
		$name = "";
		$records = "0";
		$identifier = "";
		$data = "";
		$country = "";
		$type = "";
		next;
	}

	# this splits and removes the space after the colon
	($field, $data) = split /: /, $line, 2;
	next if (!$data);	
	next if $data eq "";
	

	if ($field eq "oai_pmh") {
		$oai = $data;
	} elsif ($field eq "title") {
		$data =~ s/"/ /; # replace quote strings
		$name = $data;
	} elsif ($field eq "recordcount") {
		$records = $data;
		next if ($records eq "0"); 
	} elsif ($field eq "eprintid") {
		$identifier = $data;
	} elsif ($field eq "location_country") {
		$country = $data;
	} elsif ($field eq "type") {
		$type = $data;
	}
}

#byeeeeeeeeeeeeeeeeeeeeeeee
exit;