#!/usr/bin/perl

use strict;

my ($xml_source, $data_source, $return_type) = @ARGV;

#call data args into scope from xml document

use LWP::Simple;
my $content;

$content = get $xml_source;
die "Couldn't get $xml_source" unless defined $content;

use XML::Simple;
my $xs1 = XML::Simple->new();
my $doc = $xs1->XMLin($content, forcearray => 1);

=comment
#to view the contents of the $doc array
use Data::Dumper;
print Dumper($doc);
=cut

=comment
my $line;
my @line_array;
=cut

my ($filename, $institution_desc, $institution_url, $institution_id, $platform_id, $package_id, $latitude, $longitude, $data_url, $filename_search, $file_row_start, $file_row_comment, $file_field_separator, $file_field_missing_value, $column_time, $dods_url, $comment) ;

my $contact_info = $doc->{'global_attributes'}->[0]->{'contact_info'}->[0] ;

my $institution_desc = $doc->{'global_attributes'}->[0]->{'institution_desc'}->[0] ;
my $institution_url = $doc->{'global_attributes'}->[0]->{'institution_url'}->[0] ;

my $institution_id = $doc->{'global_attributes'}->[0]->{'institution_id'}->[0] ;
my $platform_id = $doc->{'global_attributes'}->[0]->{'platform_id'}->[0] ;
my $package_id = $doc->{'global_attributes'}->[0]->{'package_id'}->[0] ;

my $latitude = $doc->{'global_attributes'}->[0]->{'latitude'}->[0] ;
my $longitude = $doc->{'global_attributes'}->[0]->{'longitude'}->[0] ;

my $data_url = $doc->{'global_attributes'}->[0]->{'data_url'}->[0] ;
my $filename_search = $doc->{'global_attributes'}->[0]->{'filename_search'}->[0] ;

my $file_row_start = $doc->{'global_attributes'}->[0]->{'file_row_start'}->[0] ;
if ($file_row_start =~ /HASH/) { $file_row_start = 1; }

#assumes 'none' won't actually be used as a line comment
#not testing for escape characters
my $file_row_comment = $doc->{'global_attributes'}->[0]->{'file_row_comment'}->[0] ;
if ($file_row_comment =~ /HASH/) { $file_row_comment = 'none'; }

#not testing for escape characters
my $file_field_separator = $doc->{'global_attributes'}->[0]->{'file_field_separator'}->[0] ;
#forced filtering to below separator on files
my $file_field_separator = '_SEP_' ;

#assumes 'none' won't actually be used as a missing value
#not testing for escape characters
my $file_field_missing_value = $doc->{'global_attributes'}->[0]->{'file_field_missing_value'}->[0] ;
if ($file_field_missing_value =~ /HASH/) { $file_field_missing_value = 'none'; }

my $column_time = $doc->{'global_attributes'}->[0]->{'column_time'}->[0] ;
if ($column_time =~ /HASH/) { $column_time = 1; }

my $measurement_time_zone = $doc->{'global_attributes'}->[0]->{'measurement_time_zone'}->[0] ;
if ($measurement_time_zone =~ /HASH/) { $measurement_time_zone = '+00'; }

my $dods_url = $doc->{'global_attributes'}->[0]->{'dods_url'}->[0] ;
if ($dods_url =~ /HASH/) { $dods_url = ''; }

my $comment = $doc->{'global_attributes'}->[0]->{'comment'}->[0] ;
if ($comment =~ /HASH/) { $comment = ''; }

#populate local array from ascii column data using xml file processing args

$content = get $data_source;
die "Couldn't get $data_source" unless defined $content;

#split rows into array elements
my @data_content = split (/\n/, $content);

my @obs_values;
my $number_file_columns;

my $line;
my @line_array;
my $line_count_file = 0;
my $line_count_processed = 0;

foreach $line (@data_content) {
        $line_count_file++;

        #drop lines until first row start
        if ($line_count_file < $file_row_start) { next; }

	#skip lines which are commented
        if ($line =~ /^$file_row_comment/) { next; }

	#processing row
	#print "process\n";
        $line_count_processed++;

	#map missing values
	@line_array = split(/$file_field_separator/,$line);
	foreach my $element (@line_array) {
		if ($element eq $file_field_missing_value) { $element = '-99999'; }
	}	
	foreach my $element (@line_array) {
		#print "$element\n";
	}	

	push (@obs_values, @line_array);
}

my $number_file_columns = scalar(@line_array);
#print "line_count_processed: $line_count_processed\n";

#loop through observations building substitution strings

my $obs_declaration = '';
my $z_count = 0;
my $z_list = '';
my $obs_list = '';

foreach my $obs_returned ( @{$doc->{'dependent_variables'}->[0]->{'variable'}} )
{
        my $obs_column_number = $obs_returned->{'column_number'}->[0] ;
        my $obs_standard_name = $obs_returned->{'standard_name'}->[0] ;
        my $obs_units = $obs_returned->{'units'}->[0] ;
        my $obs_z = $obs_returned->{'z'}->[0] ;
	
	if (!($obs_column_number =~ /HASH/)) {

        $obs_declaration .= " \
	double $obs_standard_name(time) ; \
		$obs_standard_name:standard_name = \"$obs_standard_name\" ;
		$obs_standard_name:units = \"$obs_units\" ;
		$obs_standard_name:z = $obs_z ; ";

	$z_count++;
	$z_list .= "$obs_z,";
	$obs_list .= "$obs_standard_name = ".get_array_column(\@obs_values, $obs_column_number, $number_file_columns, 0)." ;\n\n"; 
	}
}

$z_list = substr($z_list,0,-1); #strip trailing comma

my ($time_list, $start_time, $end_time) = get_array_column(\@obs_values, $column_time, $number_file_columns, 1); 
#print $time_list."\n";
#print $start_time."\n";
#print $end_time."\n";

my $filename_prefix = $institution_id.'_'.$platform_id.'_'.$package_id; 

#my $timestamp_prefix = `date +"%Y%m%d%H%M%S"`;
#my $timestamp_prefix = `date --date='$end_time'  +"%Y%m%d%H%M%S"`;
#chomp($timestamp_prefix);

my $timestamp_prefix = substr($end_time,0,4).substr($end_time,5,2).substr($end_time,8,2).substr($end_time,11,2).substr($end_time,14,2).substr($end_time,17,2).substr($end_time,19,3);
my $filename_text = $filename_prefix."_".$timestamp_prefix."_latest.txt";

my $create_date = `date +"%Y-%m-%d %H:%M:%S"`;
chomp($create_date);

#substitute args into netcdf template
open (NETCDF_FILE, ">/tmp/ms_tmp/$filename_text");
$content = `cat seacoos_xml_netcdf_template.txt`;

$content =~ s/<CONTACT_INFO>/$contact_info/g;

#platform information

$content =~ s/<INSTITUTION_DESC>/$institution_desc/g;
$content =~ s/<INSTITUTION_URL>/$institution_url/g;

$content =~ s/<INSTITUTION_ID>/$institution_id/g;
$content =~ s/<PLATFORM_ID>/$platform_id/g;
$content =~ s/<PACKAGE_ID>/$package_id/g;

$content =~ s/<LATITUDE>/$latitude/g;
$content =~ s/<LONGITUDE>/$longitude/g;

#column information

$content =~ s/<OBS_DECLARATION>/$obs_declaration/g;

$content =~ s/<TIME_COUNT>/$line_count_processed/g;
$content =~ s/<TIME_LIST>/$time_list/g;

$content =~ s/<START_TIME>/$start_time/g;
$content =~ s/<END_TIME>/$end_time/g;

$content =~ s/<RELEASE_DATE>/$create_date/g;
$content =~ s/<CREATION_DATE>/$create_date/g;
$content =~ s/<MODIFICATION_DATE>/$create_date/g;

$content =~ s/<Z_COUNT>/$z_count/g;
$content =~ s/<Z_LIST>/$z_list/g;

$content =~ s/<OBS_LIST>/$obs_list/g;

#other  information

$content =~ s/<DODS_URL>/$dods_url/g;
$content =~ s/<COMMENT>/$comment/g;

print NETCDF_FILE $content;

close (NETCDF_FILE);

my $filename_netcdf = $filename_prefix."_".$timestamp_prefix."_latest.nc";
`/usr2/home/jcothran/storm_surge/ncgen -o /tmp/ms_tmp/$filename_netcdf /tmp/ms_tmp/$filename_text`;

#the filename handle is passed back to the calling php page for reference
if ($return_type eq 'txt') { print $filename_text; }
else { print $filename_netcdf; }

exit 0;

sub get_array_column {

#returns a comma separated listing of a vertical sectioning of a flat array based on known number of columns and converts time to netcdf representation if flag set

my ($array_line, $get_column, $array_num_columns, $convert_time) = @_;

my $column_count = 1;
my $array_list = '';
my ($start_time, $end_time);

foreach my $element (@{$array_line}) {
	if ($column_count == $get_column) {
		if ($convert_time) {
		        my $measurement_date = substr($element,0,19);

		        #my $measurement_time_zone = substr($element,19,3); #only considering hour(not minute) zone differences
			#if (!($measurement_time_zone)) { $measurement_time_zone = '+00'; }

			my $full_arg_measurement_date = $measurement_date.$measurement_time_zone.'00'; 
			#print "$full_arg_measurement_date\n";

			if (!($start_time)) { $start_time = $full_arg_measurement_date; }
			$end_time = $full_arg_measurement_date;

        		my $time_sec = `date --date='$full_arg_measurement_date' +%s`;
        		chomp($time_sec);
			$array_list .= $time_sec.',';
		}
		else { 
			$array_list .= $element.',';
		}
	}
	
	$column_count++;
	if ($column_count == $array_num_columns+1) { $column_count = 1; }
}

$array_list = substr($array_list,0,-1); #strip trailing comma

if ($convert_time) {

	my @array_list_more;
	push (@array_list_more, $array_list);	
	push (@array_list_more, $start_time);	
	push (@array_list_more, $end_time);
	
	return @array_list_more;
}
else {
	return $array_list;
}
} #get_array_column