Parsing Unicode CLDR Windows Timezones in PHP

parse_windowszones.php

<?php

// The location of the Unicode CLDR "windowsZones" mappings here
const XML_URL = 'http://unicode.org/cldr/data/common/supplemental/windowsZones.xml';

const XML_ZONE_MAP_XPATH = '/supplementalData/windowsZones/mapTimezones/mapZone';

const ZONE_TERRITORY_ATTRIBUTE = 'territory';
const ZONE_IANA_NAME_ATTRIBUTE = 'type';
const ZONE_WINDOWS_NAME_ATTRIBUTE = 'other';

const MAP_IANA_MULTIPLE_DELIMITER = ' ';


// Load the remote XML document
$xml = new SimpleXmlElement(XML_URL, 0, true);

// Build the Windows->IANA Timezone map
$windows_iana_timezone_map = array_reduce(
    $xml->xpath(XML_ZONE_MAP_XPATH),
    function (array $map, SimpleXmlElement $element) {
        $windows_name = (string) $element[ZONE_WINDOWS_NAME_ATTRIBUTE];
        $territory = (string) $element[ZONE_TERRITORY_ATTRIBUTE];
        $iana_names = explode(MAP_IANA_MULTIPLE_DELIMITER, (string) $element[ZONE_IANA_NAME_ATTRIBUTE]);

        if (!isset($map[$windows_name])) {
            $map[$windows_name] = [];
        }

        if (!isset($map[$windows_name][$territory])) {
            $map[$windows_name][$territory] = [];
        }

        $map[$windows_name][$territory] = array_merge(
            $map[$windows_name][$territory],
            $iana_names
        );

        return $map;
    },
    []
);

// Build the IANA->Windows Timezone map for faster inverted lookups
$iana_windows_timezone_map = [];
array_walk(
    $windows_iana_timezone_map,
    function (array $territories, $windows_name) use (&$iana_windows_timezone_map) {
        $iana_names = array_reduce(
            $territories,
            function ($all, $row) {
                return array_merge($all, $row);
            },
            []
        );

        foreach ($iana_names as $iana_name) {
            $iana_windows_timezone_map[$iana_name] = $windows_name;
        }
    }
);

// OPTIONAL: Sort the keys of the maps naturally for easier viewing and generated file "diffing"
ksort($iana_windows_timezone_map, SORT_NATURAL);
ksort($windows_iana_timezone_map, SORT_NATURAL);