#!/usr/bin/env perl

use strict;
use warnings;
use 5.010;

use File::Slurp qw(read_file write_file);
use JSON;
use List::Util qw(any);
use XML::LibXML;

my $json_str = read_file('stations.json');
my $stations = JSON->new->utf8->decode($json_str);
@{$stations} = sort { $a->{name} cmp $b->{name} } @{$stations};

my %station_by_ds100;
for my $station ( @{$stations} ) {
	$station_by_ds100{ $station->{ds100} } = $station;
}

my %station_by_eva;
for my $station ( @{$stations} ) {
	$station_by_eva{ $station->{eva} } = $station;
}

my %station_by_name;
for my $station ( @{$stations} ) {
	$station_by_name{ $station->{name} } = $station;
}

my %xml_by_name;
my %xml_by_eva;

my $xml_str = read_file('stations.xml');
my $tree    = XML::LibXML->load_xml( string => $xml_str );

my @missing;

if ( -e "missing.txt" ) {
	for my $line ( read_file("missing.txt") ) {
		chomp $line;
		push( @missing, $line );
	}
}

for my $station ( $tree->findnodes('//station') ) {
	my $name  = $station->getAttribute('name');
	my $eva   = $station->getAttribute('eva');
	my $ds100 = $station->getAttribute('ds100');
	my $is_db = $station->getAttribute('db') eq 'true';

	my $xml_station = {
		name  => $name,
		eva   => $eva,
		ds100 => $ds100,
		is_db => $is_db,
	};
	$xml_by_name{$name} = $xml_station;
	$xml_by_eva{$eva}   = $xml_station;

	my $found = 0;

	if ( $station_by_name{$name} or $station_by_eva{$eva} ) {
		$found = 1;
	}

	if (    $station_by_ds100{$ds100}
		and $station_by_ds100{$ds100}{name} ne $name )
	{
		printf( "%8s has been renamed: %30s -> %30s\n",
			$ds100, $station_by_ds100{$ds100}{name}, $name );

		#$station_by_ds100{$ds100}{name} = $name;
	}
	elsif ( $station_by_eva{$eva} and $station_by_eva{$eva}{name} ne $name ) {
		printf(
			"%d mismatch: (%s -> %s), (%s -> %s)\n",
			$eva,  $station_by_eva{$eva}{name},
			$name, $station_by_eva{$eva}{ds100}, $ds100
		);
	}
	elsif ( $station_by_name{$name}
		and $station_by_name{$name}{ds100} ne $ds100
		and $is_db
		and $ds100 !~ m{ ^ PQ }x )
	{
		printf( "%30s has been recoded: %8s -> %8s\n",
			$name, $station_by_name{$name}{ds100}, $ds100 );
	}
	elsif ( $station_by_name{$name}
		and $station_by_name{$name}{eva} ne $eva
		and $is_db
		and $ds100 !~ m{ ^ PQ }x )
	{
		printf( "%30s has been recoded: %d -> %d\n",
			$name, $station_by_name{$name}{eva}, $eva );
	}

	if ( not $found
		and any { $_ eq $name } @missing )
	{
		say "missing $eva  $ds100  \"$name\"";
		push(
			@{$stations},
			{
				name  => $name,
				ds100 => $ds100,
				eva   => $eva,
			}
		);
	}
}

my @to_delete;

for my $i ( 0 .. $#{$stations} ) {
	$stations->[$i]{eva} = 0 + $stations->[$i]{eva};
	my $j_station = $stations->[$i];
	my $j_name    = $j_station->{name};
	my $j_ds100   = $j_station->{ds100};
	my $j_eva     = $j_station->{eva};

	if ( not( $xml_by_name{$j_name} or $xml_by_eva{$j_eva} ) ) {
		say "station no longer exists: $j_eva  $j_ds100  \"$j_name\"";
		unshift( @to_delete, $i );
	}
}

for my $i (@to_delete) {
	splice( @{$stations}, $i, 1 );
}

my $json_out = JSON->new->utf8->canonical->pretty->encode($stations);
write_file( 'stations.json', $json_out );