PHP: Choosing the closest matching language from Accept-Language header

LanguageDetect.php

<?php
/**
 * So you have a list of languages your site supports
 * and want to automatically assign the user to the closest
 * one.
 * 
 * Usage:
 * 
 * 	 LanguageDetect::findBestMatch(['en', 'fr', 'gr']);
 * 	 LanguageDetect::findBestMatchOrFallback(['en', 'fr', 'gr']);
 * 
 * The above will automatically parse the browser's Accept-Language header.
 * You can alternatively pass a second parameter if you want to supply another
 * string or array instead of looking in the header.
 */
class LanguageDetect {
	
	/**
	 * Provide a list of supported languages and this function
	 * will scan the list supported by the user's browser and
	 * return the closest language the user supports.
	 * 
	 * False if no match is found. Use findBestMatchOrFallback()
	 * if you want to automatically default to the first language.
	 * 
	 * @param $supportedLanguages - array of supported languages (use lowercase)
	 * @param $userLanguages - leave null to use $_SERVER['HTTP_ACCEPT_LANGUAGE']
	 *                         Otherwise pass a string like 'en,fr,cn' or 'en,fr;q=0.5'
	 */
	public static function findBestMatch($supportedLanguages, $userLanguages = null)
	{
		if (empty($userLanguages))
			$userLanguages = @$_SERVER['HTTP_ACCEPT_LANGUAGE'];
		
		if (!is_array($userLanguages))
			$userLanguages = self::languageStringToArray($userLanguages);
		
		// Find exact matches
		foreach ($userLanguages as $l) {
			if (in_array($l, $supportedLanguages))
				return $l;
		}
		
		// No exact match? Let's look for partial matches ('en-us' == 'en')
		// Note: currently will NOT find matches when supportedLanguages is more specific ('en-us')
		foreach ($userLanguages as $l) {
			list($l) = explode('-', $l);
			if (in_array($l, $supportedLanguages))
				return $l;
		}
		
		// No match at all
		return false;		
	}

	public static function findBestMatchOrFallback($supportedLanguages, $userLanguages = null) 
	{
		$bestMatch = self::findBestMatch($supportedLanguages, $userLanguages);
		if (!empty($bestMatch))
			return $bestMatch;
		// No result? Return first supported language as a fallback
		return $supportedLanguages[0];
	}
	
	/**
	 * Converts a comma-separated string of accepted languages (with an
	 * optional priority) into an array of languages, ordered by priority
	 * 
	 * @param $languageString - eg. 'en-us,en;q=0.8,fr;q=0.5'
	 * @return an array of languages, lowercased, ordered by priority, e.g. ['en-us','en','fr']
	 */
	public static function languageStringToArray($languageString)
	{
		$languages = explode(',', trim($languageString));
		$result = [];
		foreach ($languages as $language) {
			if (preg_match('/(\*|[a-zA-Z0-9]{1,8}(?:-[a-zA-Z0-9]{1,8})*)(?:\s*;\s*q\s*=\s*(0(?:\.\d{0,3})|1(?:\.0{0,3})))?/', trim($language), $match)) {
				// $match[1] is the language, $match[2] is the priority (if present)
				if (!isset($match[2])) {
					$priority = '1.0';
				} else {
					$priority = (string) floatval($match[2]);
				}
				$result[strtolower($match[1])] = $priority;
			}
		}
		// Sort by priority, descending
		arsort($result);
		// Return only the keys, discarding the priorities
		return array_keys($result);
	}
	
}

// TESTS
// print_r(LanguageDetect::languageStringToArray('en-us,en;q=1,fr;q=0.5'));
// assert(LanguageDetect::findBestMatch(['en','cn'], 'en-us,en;q=0.8,fr;q=0.5') == 'en');
// assert(LanguageDetect::findBestMatch(['cn'], 'en-us,en;q=0.8,fr;q=0.5') == null);
// assert(LanguageDetect::findBestMatch(['en','en-us','cn'], 'en-us,en;q=0.8,fr;q=0.5') == 'en-us');
// assert(LanguageDetect::findBestMatch(['fr','en','en-us','cn'], 'en-us,en;q=0.8,fr;q=0.5') == 'en-us');
// assert(LanguageDetect::findBestMatch(['fr','cn'], 'en-us,en;q=0.8,fr;q=0.5') == 'fr');
// assert(LanguageDetect::findBestMatch(['en'], 'en-us;q=0.8') == 'en');
// assert(LanguageDetect::findBestMatch(['en'], 'fr,en-us;q=0.8') == 'en');
// assert(LanguageDetect::findBestMatch(['en-us','en'], 'fr,gr,en,en-us;q=0.8') == 'en');
// assert(LanguageDetect::findBestMatch(['en-us','en'], 'fr,gr,en-us,en;q=0.8') == 'en-us');
// assert(LanguageDetect::findBestMatch(['en-us','en'], 'fr,gr,en;q=0.8,en-us') == 'en-us');
// assert(LanguageDetect::findBestMatch(['en-us','en'], 'fr,gr,en-us;q=0.8,en') == 'en');
// assert(LanguageDetect::findBestMatchOrFallback(['en-us','en'], 'fr,gr,en-us;q=0.8,en') == 'en');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'fr,gr,en-us;q=0.8,en') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xa') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xa,xb') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xb,xa') == 'xb');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xb-xx,xa') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xb-xx,xa-xx') == 'xb');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xb;q=0.7,xa;q=1.0') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], 'xd;q=0.7,xc;q=1.0') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa'], 'fr,gr,en-us;q=0.8,en') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], '') == 'xa');
// assert(LanguageDetect::findBestMatchOrFallback(['xa','xb'], '') == 'xa');