xref: /webtrees/app/I18N.php (revision 6bef31b0e1f5edb64c54d1cb393c35824d985e44)
1<?php
2namespace Fisharebest\Webtrees;
3
4/**
5 * webtrees: online genealogy
6 * Copyright (C) 2015 webtrees development team
7 * This program is free software: you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation, either version 3 of the License, or
10 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License
16 * along with this program. If not, see <http://www.gnu.org/licenses/>.
17 */
18
19use Fisharebest\Localization\Locale;
20use Fisharebest\Localization\LocaleEnUs;
21use Fisharebest\Localization\Translation;
22use Fisharebest\Localization\Translator;
23use Patchwork\TurkishUtf8;
24
25/**
26 * Class I18N - Functions to support internationalization (i18n) functionality.
27 */
28class I18N {
29	/** @var Locale The current locale (e.g. LocaleEnGb) */
30	private static $locale;
31
32	/** @var Translator */
33	private static $translator;
34
35	// Digits are always rendered LTR, even in RTL text.
36	const DIGITS = '0123456789٠١٢٣٤٥٦٧٨٩۰۱۲۳۴۵۶۷۸۹';
37
38	// Reversable character conversions from the UNICODE 5.1 database.
39	// It excludes ambiguous (turkish dotless i) and mixed-case (Dz) characters.
40	// The characters should be arranged in default unicode-collation order.
41	const ALPHABET_LOWER = 'aàáâãäåāăąǎǟǡǻȁȃȧḁạảấầẩẫậắằẳẵặⓐaæǣǽbḃḅḇⓑbƀɓƃcçćĉċčḉⅽⓒcƈdďḋḍḏḑḓⅾⓓddždzđɖɗƌðeèéêëēĕėęěȅȇȩḕḗḙḛḝẹẻẽếềểễệⓔeǝəɛfḟⓕfƒgĝğġģǧǵḡⓖgǥɠɣƣhĥȟḣḥḧḩḫⓗhƕħiìíîïĩīĭįǐȉȋḭḯỉịⅰⓘiⅱⅲijⅳⅸɨɩjĵⓙjkķǩḱḳḵⓚkƙlĺļľḷḹḻḽⅼⓛlŀljłƚmḿṁṃⅿⓜmnñńņňǹṅṇṉṋⓝnnjɲƞŋoòóôõöōŏőơǒǫǭȍȏȫȭȯȱṍṏṑṓọỏốồổỗộớờởỡợⓞoœøǿɔɵȣpṕṗⓟpƥqⓠqrŕŗřȑȓṙṛṝṟⓡrʀsśŝşšșṡṣṥṧṩⓢsʃtţťțṫṭṯṱⓣtŧƭʈuùúûüũūŭůűųưǔǖǘǚǜȕȗṳṵṷṹṻụủứừửữựⓤuʉɯʊvṽṿⅴⓥvⅵⅶⅷʋʌwŵẁẃẅẇẉⓦwxẋẍⅹⓧxⅺⅻyýÿŷȳẏỳỵỷỹⓨyƴzźżžẑẓẕⓩzƶȥǯʒƹȝþƿƨƽƅάαἀἁἂἃἄἅἆἇὰάᾀᾁᾂᾃᾄᾅᾆᾇᾰᾱᾳβγδέεἐἑἒἓἔἕὲέϝϛζήηἠἡἢἣἤἥἦἧὴήᾐᾑᾒᾓᾔᾕᾖᾗῃθϊἰἱἲἳἴἵἶἷὶίῐῑκϗλμνξοόὀὁὂὃὄὅὸόπϟϙρῥσϲτυϋύὑὓὕὗὺύῠῡφχψωώὠὡὢὣὤὥὦὧὼώᾠᾡᾢᾣᾤᾥᾦᾧῳϡϸϻϣϥϧϩϫϭϯаӑӓәӛӕбвгґғҕдԁђԃѓҙеѐёӗєжӂӝҗзԅӟѕӡԇиѝӣҋӥіїйјкқӄҡҟҝлӆљԉмӎнӊңӈҥњԋоӧөӫпҧҁрҏсԍҫтԏҭћќуӯўӱӳүұѹфхҳһѡѿѽѻцҵчӵҷӌҹҽҿџшщъыӹьҍѣэӭюяѥѧѫѩѭѯѱѳѵѷҩաբգդեզէըթժիլխծկհձղճմյնշոչպջռսվտրցւփքօֆȼɂɇɉɋɍɏͱͳͷͻͼͽӏӷӻӽӿԑԓԕԗԙԛԝԟԡԣԥᵹᵽỻỽỿⅎↄⰰⰱⰲⰳⰴⰵⰶⰷⰸⰹⰺⰻⰼⰽⰾⰿⱀⱁⱂⱃⱄⱅⱆⱇⱈⱉⱊⱋⱌⱍⱎⱏⱐⱑⱒⱓⱔⱕⱖⱗⱘⱙⱚⱛⱜⱝⱞⱡⱨⱪⱬⱳⱶⲁⲃⲅⲇⲉⲋⲍⲏⲑⲓⲕⲗⲙⲛⲝⲟⲡⲣⲥⲧⲩⲫⲭⲯⲱⲳⲵⲷⲹⲻⲽⲿⳁⳃⳅⳇⳉⳋⳍⳏⳑⳓⳕⳗⳙⳛⳝⳟⳡⳣⳬⳮⴀⴁⴂⴃⴄⴅⴆⴇⴈⴉⴊⴋⴌⴍⴎⴏⴐⴑⴒⴓⴔⴕⴖⴗⴘⴙⴚⴛⴜⴝⴞⴟⴠⴡⴢⴣⴤⴥꙁꙃꙅꙇꙉꙋꙍꙏꙑꙓꙕꙗꙙꙛꙝꙟꙣꙥꙧꙩꙫꙭꚁꚃꚅꚇꚉꚋꚍꚏꚑꚓꚕꚗꜣꜥꜧꜩꜫꜭꜯꜳꜵꜷꜹꜻꜽꜿꝁꝃꝅꝇꝉꝋꝍꝏꝑꝓꝕꝗꝙꝛꝝꝟꝡꝣꝥꝧꝩꝫꝭꝯꝺꝼꝿꞁꞃꞅꞇꞌ';
42	const ALPHABET_UPPER = 'AÀÁÂÃÄÅĀĂĄǍǞǠǺȀȂȦḀẠẢẤẦẨẪẬẮẰẲẴẶⒶAÆǢǼBḂḄḆⒷBɃƁƂCÇĆĈĊČḈⅭⒸCƇDĎḊḌḎḐḒⅮⒹDDŽDZĐƉƊƋÐEÈÉÊËĒĔĖĘĚȄȆȨḔḖḘḚḜẸẺẼẾỀỂỄỆⒺEƎƏƐFḞⒻFƑGĜĞĠĢǦǴḠⒼGǤƓƔƢHĤȞḢḤḦḨḪⒽHǶĦIÌÍÎÏĨĪĬĮǏȈȊḬḮỈỊⅠⒾIⅡⅢIJⅣⅨƗƖJĴⒿJKĶǨḰḲḴⓀKƘLĹĻĽḶḸḺḼⅬⓁLĿLJŁȽMḾṀṂⅯⓂMNÑŃŅŇǸṄṆṈṊⓃNNJƝȠŊOÒÓÔÕÖŌŎŐƠǑǪǬȌȎȪȬȮȰṌṎṐṒỌỎỐỒỔỖỘỚỜỞỠỢⓄOŒØǾƆƟȢPṔṖⓅPƤQⓆQRŔŖŘȐȒṘṚṜṞⓇRƦSŚŜŞŠȘṠṢṤṦṨⓈSƩTŢŤȚṪṬṮṰⓉTŦƬƮUÙÚÛÜŨŪŬŮŰŲƯǓǕǗǙǛȔȖṲṴṶṸṺỤỦỨỪỬỮỰⓊUɄƜƱVṼṾⅤⓋVⅥⅦⅧƲɅWŴẀẂẄẆẈⓌWXẊẌⅩⓍXⅪⅫYÝŸŶȲẎỲỴỶỸⓎYƳZŹŻŽẐẒẔⓏZƵȤǮƷƸȜÞǷƧƼƄΆΑἈἉἊἋἌἍἎἏᾺΆᾈᾉᾊᾋᾌᾍᾎᾏᾸᾹᾼΒΓΔΈΕἘἙἚἛἜἝῈΈϜϚΖΉΗἨἩἪἫἬἭἮἯῊΉᾘᾙᾚᾛᾜᾝᾞᾟῌΘΪἸἹἺἻἼἽἾἿῚΊῘῙΚϏΛΜΝΞΟΌὈὉὊὋὌὍῸΌΠϞϘΡῬΣϹΤΥΫΎὙὛὝὟῪΎῨῩΦΧΨΩΏὨὩὪὫὬὭὮὯῺΏᾨᾩᾪᾫᾬᾭᾮᾯῼϠϷϺϢϤϦϨϪϬϮАӐӒӘӚӔБВГҐҒҔДԀЂԂЃҘЕЀЁӖЄЖӁӜҖЗԄӞЅӠԆИЍӢҊӤІЇЙЈКҚӃҠҞҜЛӅЉԈМӍНӉҢӇҤЊԊОӦӨӪПҦҀРҎСԌҪТԎҬЋЌУӮЎӰӲҮҰѸФХҲҺѠѾѼѺЦҴЧӴҶӋҸҼҾЏШЩЪЫӸЬҌѢЭӬЮЯѤѦѪѨѬѮѰѲѴѶҨԱԲԳԴԵԶԷԸԹԺԻԼԽԾԿՀՁՂՃՄՅՆՇՈՉՊՋՌՍՎՏՐՑՒՓՔՕՖȻɁɆɈɊɌɎͰͲͶϽϾϿӀӶӺӼӾԐԒԔԖԘԚԜԞԠԢԤꝽⱣỺỼỾℲↃⰀⰁⰂⰃⰄⰅⰆⰇⰈⰉⰊⰋⰌⰍⰎⰏⰐⰑⰒⰓⰔⰕⰖⰗⰘⰙⰚⰛⰜⰝⰞⰟⰠⰡⰢⰣⰤⰥⰦⰧⰨⰩⰪⰫⰬⰭⰮⱠⱧⱩⱫⱲⱵⲀⲂⲄⲆⲈⲊⲌⲎⲐⲒⲔⲖⲘⲚⲜⲞⲠⲢⲤⲦⲨⲪⲬⲮⲰⲲⲴⲶⲸⲺⲼⲾⳀⳂⳄⳆⳈⳊⳌⳎⳐⳒⳔⳖⳘⳚⳜⳞⳠⳢⳫⳭႠႡႢႣႤႥႦႧႨႩႪႫႬႭႮႯႰႱႲႳႴႵႶႷႸႹႺႻႼႽႾႿჀჁჂჃჄჅꙀꙂꙄꙆꙈꙊꙌꙎꙐꙒꙔꙖꙘꙚꙜꙞꙢꙤꙦꙨꙪꙬꚀꚂꚄꚆꚈꚊꚌꚎꚐꚒꚔꚖꜢꜤꜦꜨꜪꜬꜮꜲꜴꜶꜸꜺꜼꜾꝀꝂꝄꝆꝈꝊꝌꝎꝐꝒꝔꝖꝘꝚꝜꝞꝠꝢꝤꝦꝨꝪꝬꝮꝹꝻꝾꞀꞂꞄꞆꞋ';
43
44	// Alphabet for the currently selected locale
45	private static $alphabet_lower = 'abcdefghijklmnopqrstuvwxyz';
46	private static $alphabet_upper = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ';
47
48	// Lookup table to convert unicode code-points into scripts.
49	// See https://en.wikipedia.org/wiki/Unicode_block
50	// Note: we only need details for scripts of languages into which webtrees is translated.
51	private static $scripts = array(
52		array('Latn', 0x0041, 0x005A), // a-z
53		array('Latn', 0x0061, 0x007A), // A-Z
54		array('Latn', 0x0100, 0x02AF),
55		array('Grek', 0x0370, 0x03FF),
56		array('Cyrl', 0x0400, 0x052F),
57		array('Hebr', 0x0590, 0x05FF),
58		array('Arab', 0x0600, 0x06FF),
59		array('Arab', 0x0750, 0x077F),
60		array('Arab', 0x08A0, 0x08FF),
61		array('Deva', 0x0900, 0x097F),
62		array('Taml', 0x0B80, 0x0BFF),
63		array('Sinh', 0x0D80, 0x0DFF),
64		array('Thai', 0x0E00, 0x0E7F),
65		array('Geor', 0x10A0, 0x10FF),
66		array('Grek', 0x1F00, 0x1FFF),
67		array('Deva', 0xA8E0, 0xA8FF),
68		array('Hans', 0x3000, 0x303F), // Mixed CJK, not just Hans
69		array('Hans', 0x3400, 0xFAFF), // Mixed CJK, not just Hans
70		array('Hans', 0x20000, 0x2FA1F), // Mixed CJK, not just Hans
71	);
72
73	// Characters that are displayed in mirror form in RTL text.
74	private static $mirror_characters = array(
75		'(' => ')',
76		')' => '(',
77		'[' => ']',
78		']' => '[',
79		'{' => '}',
80		'}' => '{',
81		'<' => '>',
82		'>' => '<',
83		'‹' => '›',
84		'›' => '‹',
85		'«' => '»',
86		'»' => '«',
87		'﴾' => '﴿',
88		'﴿' => '﴾',
89		'“' => '”',
90		'”' => '“',
91		'‘' => '’',
92		'’' => '‘',
93	);
94
95	/** @var string Punctuation used to separate list items, typically a comma */
96	public static $list_separator;
97
98	/**
99	 * Initialise the translation adapter with a locale setting.
100	 *
101	 * @param string|null $code Use this locale/language code, or choose one automatically
102	 *
103	 * @return string $string
104	 */
105	public static function init($code = null) {
106		global $WT_SESSION, $WT_TREE;
107
108		if ($code !== null) {
109			// Create the specified locale
110			self::$locale = Locale::create($code);
111		} else {
112			// Negotiate a locale, but if we can't then use a failsafe
113			self::$locale = new LocaleEnUs;
114			if (Filter::get('lang')) {
115				// A request in the URL
116				try {
117					$locale = Locale::create(Filter::get('lang'));
118					if (file_exists(WT_ROOT . 'language/' . $locale->languageTag() . '.mo')) {
119						self::$locale = $locale;
120					}
121				} catch (\Exception $ex) {
122				}
123			} elseif ($WT_SESSION->locale) {
124				// Previously used
125				self::$locale = Locale::create($WT_SESSION->locale);
126			} elseif (isset($_SERVER['HTTP_ACCEPT_LANGUAGE'])) {
127				// Browser negotiation
128				$http_accept_language = strtolower(str_replace(' ', '', $_SERVER['HTTP_ACCEPT_LANGUAGE']));
129				if (preg_match_all('/(?:([a-z][a-z0-9_-]+)(?:;q=([0-9.]+))?)/', $http_accept_language, $match)) {
130					$preferences = array_combine($match[1], $match[2]);
131					array_walk($preferences, function(&$x) { $x = $x === '' ? 1.0 : (float) $x; });
132				} else {
133					$preferences = array();
134				}
135				// Add the tree’s default language as a low-priority
136				if ($WT_TREE && !isset($preferences[$WT_TREE->getPreference('LANGUAGE')])) {
137					$preferences[$WT_TREE->getPreference('LANGUAGE')] = 0.2;
138				}
139				arsort($preferences);
140				foreach (array_keys($preferences) as $code) {
141					try {
142						$locale = Locale::create($code);
143						if (file_exists(WT_ROOT . 'language/' . $locale->languageTag() . '.mo')) {
144							self::$locale = $locale;
145							break;
146						}
147					} catch (\Exception $ex) {
148						// The user's prefered locale does not exist
149					}
150				}
151			}
152		}
153
154		File::mkdir(WT_DATA_DIR . 'cache');
155		$cache_file = WT_DATA_DIR . 'cache/language-' . self::$locale->languageTag() . '-cache.php';
156		if (file_exists($cache_file)) {
157			$filemtime = filemtime($cache_file);
158		} else {
159			$filemtime = 0;
160		}
161
162		// Load the translation file(s)
163		// Note that glob() returns false instead of an empty array when open_basedir_restriction
164		// is in force and no files are found.  See PHP bug #47358.
165		$translation_files = array_merge(
166			array(WT_ROOT . 'language/' . self::$locale->languageTag() . '.mo'),
167			glob(WT_MODULES_DIR . '*/language/' . self::$locale->languageTag() . '.{csv,php,mo}', GLOB_BRACE) ?: array(),
168			glob(WT_DATA_DIR . 'language/' . self::$locale->languageTag() . '.{csv,php,mo}', GLOB_BRACE) ?: array()
169		);
170
171		$rebuild_cache = false;
172		foreach ($translation_files as $translation_file) {
173			if (filemtime($translation_file) > $filemtime) {
174				$rebuild_cache = true;
175				break;
176			}
177		}
178
179		if ($rebuild_cache) {
180			$translations = array();
181			foreach ($translation_files as $translation_file) {
182				$translation = new Translation($translation_file);
183				$translations = array_merge($translations, $translation->asArray());
184			}
185			file_put_contents($cache_file, '<' . '?php return ' . var_export($translations, true) . ';');
186		} else {
187			$translations = include $cache_file;
188		}
189
190		// Create a translator
191		self::$translator = new Translator($translations, self::$locale->pluralRule());
192
193		// Extract language settings from the translation file
194		global $DATE_FORMAT; // I18N: This is the format string for full dates.  See http://php.net/date for codes
195		$DATE_FORMAT = self::$translator->translate('%j %F %Y');
196
197		global $TIME_FORMAT; // I18N: This is the format string for the time-of-day.  See http://php.net/date for codes
198		$TIME_FORMAT = self::$translator->translate('%H:%i:%s');
199
200		// Alphabetic sorting sequence (upper-case letters), used by webtrees to sort strings
201		list(, self::$alphabet_upper) = explode('=', self::$translator->translate('ALPHABET_upper=ABCDEFGHIJKLMNOPQRSTUVWXYZ'));
202		// Alphabetic sorting sequence (lower-case letters), used by webtrees to sort strings
203		list(, self::$alphabet_lower) = explode('=', self::$translator->translate('ALPHABET_lower=abcdefghijklmnopqrstuvwxyz'));
204
205		global $WEEK_START;
206		$WEEK_START = self::$locale->territory()->firstDay();
207
208		self::$list_separator = /* I18N: This punctuation is used to separate lists of items */ self::translate(', ');
209
210		return self::$locale->languageTag();
211	}
212
213	/**
214	 * The prefered locales for this site, or a default list if no preference.
215	 *
216	 * @return Locale[]
217	 */
218	public static function activeLocales() {
219		$code_list = Site::getPreference('LANGUAGES');
220
221		if ($code_list) {
222			$codes = explode(',', $code_list);
223		} else {
224			$codes = array(
225				'ar', 'bg', 'bs', 'ca', 'cs', 'da', 'de', 'el', 'en-GB', 'en-US', 'es',
226				'et', 'fi', 'fr', 'he', 'hr', 'hu', 'is', 'it', 'ka', 'lt', 'mr', 'nb',
227				'nl', 'nn', 'pl', 'pt', 'ru', 'sk', 'sv', 'tr', 'uk', 'vi', 'zh-Hans',
228			);
229		}
230
231		$locales = array();
232		foreach ($codes as $code) {
233			if (file_exists(WT_ROOT . 'language/' . $code . '.mo')) {
234				try {
235					$locales[] = Locale::create($code);
236				} catch (\Exception $ex) {
237					// No such locale exists?
238				}
239			}
240		}
241		usort($locales, '\Fisharebest\Localization\Locale::compare');
242
243		return $locales;
244	}
245
246	/**
247	 * Which MySQL collation should be used for this locale?
248	 *
249	 * @return string
250	 */
251	public static function collation() {
252		$collation = self::$locale->collation();
253		switch ($collation) {
254		case 'german2_ci':
255		case 'vietnamese_ci':
256			// Only available in MySQL 5.6
257			return 'utf8_unicode_ci';
258		default:
259			return 'utf8_' . $collation;
260		}
261	}
262
263	/**
264	 * All locales for which a translation file exists.
265	 *
266	 * @return Locale[]
267	 */
268	public static function installedLocales() {
269		$locales = array();
270		foreach (glob(WT_ROOT . 'language/*.mo') as $file) {
271			try {
272				$locales[] = Locale::create(basename($file, '.mo'));
273			} catch (\Exception $ex) {
274				// Not a recognised locale
275			}
276		}
277		usort($locales, '\Fisharebest\Localization\Locale::compare');
278
279		return $locales;
280	}
281
282	/**
283	 * Generate i18n markup for the <html> tag, e.g. lang="ar" dir="rtl"
284	 *
285	 * @return string
286	 */
287	public static function htmlAttributes() {
288		return self::$locale->htmlAttributes();
289	}
290
291	/**
292	 * Translate a number into the local representation.
293	 *
294	 * e.g. 12345.67 becomes
295	 * en: 12,345.67
296	 * fr: 12 345,67
297	 * de: 12.345,67
298	 *
299	 * @param float   $n
300	 * @param integer $precision
301	 *
302	 * @return string
303	 */
304	public static function number($n, $precision = 0) {
305		return self::$locale->number(round($n, $precision));
306	}
307
308	/**
309	 * Convert the digits 0-9 into the local script
310	 *
311	 * Used for years, etc., where we do not want thousands-separators, decimals, etc.
312	 *
313	 * @param integer $n
314	 *
315	 * @return string
316	 */
317	public static function digits($n) {
318		return self::$locale->digits($n);
319	}
320
321	/**
322	 * Translate a fraction into a percentage.
323	 *
324	 * e.g. 0.123 becomes
325	 * en: 12.3%
326	 * fr: 12,3 %
327	 * de: 12,3%
328	 *
329	 * @param float   $n
330	 * @param integer $precision
331	 *
332	 * @return string
333	 */
334	public static function percentage($n, $precision = 0) {
335		return self::$locale->percent($n, $precision);
336	}
337
338	/**
339	 * What is the direction of the current locale
340	 *
341	 * @return string "ltr" or "rtl"
342	 */
343	public static function direction() {
344		return self::$locale->direction();
345	}
346
347	/**
348	 * Translate a string, and then substitute placeholders
349	 *
350	 * echo I18N::translate('Hello World!');
351	 * echo I18N::translate('The %s sat on the mat', 'cat');
352	 *
353	 * @return string
354	 */
355	public static function translate(/* var_args */) {
356		$args = func_get_args();
357		$args[0] = self::$translator->translate($args[0]);
358
359		return call_user_func_array('sprintf', $args);
360	}
361
362	/**
363	 * Context sensitive version of translate.
364	 *
365	 * echo I18N::translate_c('NOMINATIVE', 'January');
366	 * echo I18N::translate_c('GENITIVE',   'January');
367	 *
368	 * @return string
369	 */
370	public static function translateContext(/* var_args */) {
371		$args = func_get_args();
372		$args[0] = self::$translator->translateContext($args[0], $args[1]);
373		unset($args[1]);
374
375		return call_user_func_array('sprintf', $args);
376	}
377
378	/**
379	 * Similar to translate, but do perform "no operation" on it.
380	 *
381	 * This is necessary to fetch a format string (containing % characters) without
382	 * performing sustitution of arguments.
383	 *
384	 * @param string $string
385	 *
386	 * @return string
387	 */
388	public static function noop($string) {
389		return self::$translator->translate($string);
390	}
391
392	/**
393	 * Translate a plural string
394	 *
395	 * echo self::plural('There is an error', 'There are errors', $num_errors);
396	 * echo self::plural('There is one error', 'There are %s errors', $num_errors);
397	 * echo self::plural('There is %1$s %2$s cat', 'There are %1$s %2$s cats', $num, $num, $colour);
398	 *
399	 * @return string
400	 */
401	public static function plural(/* var_args */) {
402		$args = func_get_args();
403		$args[0] = self::$translator->plural($args[0], $args[1], $args[2]);
404		unset($args[1], $args[2]);
405
406		return call_user_func_array('sprintf', $args);
407	}
408
409	/**
410	 * Convert a GEDCOM age string into translated_text
411	 *
412	 * NB: The import function will have normalised this, so we don't need
413	 * to worry about badly formatted strings
414	 * NOTE: this function is not yet complete - eventually it will replace get_age_at_event()
415	 *
416	 * @param $string
417	 *
418	 * @return string
419	 */
420	public static function gedcomAge($string) {
421		switch ($string) {
422		case 'STILLBORN':
423			// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (stillborn)
424			return self::translate('(stillborn)');
425		case 'INFANT':
426			// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (in infancy)
427			return self::translate('(in infancy)');
428		case 'CHILD':
429			// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (in childhood)
430			return self::translate('(in childhood)');
431		}
432		$age = array();
433		if (preg_match('/(\d+)y/', $string, $match)) {
434			// I18N: Part of an age string. e.g. 5 years, 4 months and 3 days
435			$years = $match[1];
436			$age[] = self::plural('%s year', '%s years', $years, self::number($years));
437		} else {
438			$years = -1;
439		}
440		if (preg_match('/(\d+)m/', $string, $match)) {
441			// I18N: Part of an age string. e.g. 5 years, 4 months and 3 days
442			$age[] = self::plural('%s month', '%s months', $match[1], self::number($match[1]));
443		}
444		if (preg_match('/(\d+)w/', $string, $match)) {
445			// I18N: Part of an age string. e.g. 7 weeks and 3 days
446			$age[] = self::plural('%s week', '%s weeks', $match[1], self::number($match[1]));
447		}
448		if (preg_match('/(\d+)d/', $string, $match)) {
449			// I18N: Part of an age string. e.g. 5 years, 4 months and 3 days
450			$age[] = self::plural('%s day', '%s days', $match[1], self::number($match[1]));
451		}
452		// If an age is just a number of years, only show the number
453		if (count($age) === 1 && $years >= 0) {
454			$age = $years;
455		}
456		if ($age) {
457			if (!substr_compare($string, '<', 0, 1)) {
458				// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (aged less than 21 years)
459				return self::translate('(aged less than %s)', $age);
460			} elseif (!substr_compare($string, '>', 0, 1)) {
461				// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (aged more than 21 years)
462				return self::translate('(aged more than %s)', $age);
463			} else {
464				// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (aged 43 years)
465				return self::translate('(aged %s)', $age);
466			}
467		} else {
468			// Not a valid string?
469			return self::translate('(aged %s)', $string);
470		}
471	}
472
473	/**
474	 * Convert a number of seconds into a relative time.  For example, 630 => "10 hours, 30 minutes ago"
475	 *
476	 * @param integer $seconds
477	 *
478	 * @return string
479	 */
480	public static function timeAgo($seconds) {
481		$minute = 60;
482		$hour   = 60 * $minute;
483		$day    = 24 * $hour;
484		$month  = 30 * $day;
485		$year   = 365 * $day;
486
487		if ($seconds > $year) {
488			$years = (int) ($seconds / $year);
489			return self::plural('%s year ago', '%s years ago', $years, self::number($years));
490		} elseif ($seconds > $month) {
491			$months = (int) ($seconds / $month);
492			return self::plural('%s month ago', '%s months ago', $months, self::number($months));
493		} elseif ($seconds > $day) {
494			$days = (int) ($seconds / $day);
495			return self::plural('%s day ago', '%s days ago', $days, self::number($days));
496		} elseif ($seconds > $hour) {
497			$hours = (int) ($seconds / $hour);
498			return self::plural('%s hour ago', '%s hours ago', $hours, self::number($hours));
499		} elseif ($seconds > $minute) {
500			$minutes = (int) ($seconds / $minute);
501			return self::plural('%s minute ago', '%s minutes ago', $minutes, self::number($minutes));
502		} else {
503			return self::plural('%s second ago', '%s seconds ago', $seconds, self::number($seconds));
504		}
505	}
506
507	/**
508	 * Return the endonym for a given language - as per http://cldr.unicode.org/
509	 *
510	 * @param string $locale
511	 *
512	 * @return string
513	 */
514	public static function languageName($locale) {
515		return Locale::create($locale)->endonym();
516	}
517
518	/**
519	 * Return the script used by a given language
520	 *
521	 * @param string $locale
522	 *
523	 * @return string
524	 */
525	public static function languageScript($locale) {
526		return Locale::create($locale)->script()->code();
527	}
528
529	/**
530	 * Identify the script used for a piece of text
531	 *
532	 * @param $string
533	 *
534	 * @return string
535	 */
536	public static function textScript($string) {
537		$string = strip_tags($string); // otherwise HTML tags show up as latin
538		$string = html_entity_decode($string, ENT_QUOTES, 'UTF-8'); // otherwise HTML entities show up as latin
539		$string = str_replace(array('@N.N.', '@P.N.'), '', $string); // otherwise unknown names show up as latin
540		$pos = 0;
541		$strlen = strlen($string);
542		while ($pos < $strlen) {
543			// get the Unicode Code Point for the character at position $pos
544			$byte1 = ord($string[$pos]);
545			if ($byte1 < 0x80) {
546				$code_point = $byte1;
547				$chrlen = 1;
548			} elseif ($byte1 < 0xC0) {
549				// Invalid continuation character
550				return 'Latn';
551			} elseif ($byte1 < 0xE0) {
552				$code_point = (($byte1 & 0x1F) << 6) + (ord($string[$pos + 1]) & 0x3F);
553				$chrlen = 2;
554			} elseif ($byte1 < 0xF0) {
555				$code_point = (($byte1 & 0x0F) << 12) + ((ord($string[$pos + 1]) & 0x3F) << 6) + (ord($string[$pos + 2]) & 0x3F);
556				$chrlen = 3;
557			} elseif ($byte1 < 0xF8) {
558				$code_point = (($byte1 & 0x07) << 24) + ((ord($string[$pos + 1]) & 0x3F) << 12) + ((ord($string[$pos + 2]) & 0x3F) << 6) + (ord($string[$pos + 3]) & 0x3F);
559				$chrlen = 3;
560 			} else {
561				// Invalid UTF
562				return 'Latn';
563			}
564
565			foreach (self::$scripts as $range) {
566				if ($code_point >= $range[1] && $code_point <= $range[2]) {
567					return $range[0];
568				}
569			}
570			// Not a recognised script.  Maybe punctuation, spacing, etc.  Keep looking.
571			$pos += $chrlen;
572		}
573
574		return 'Latn';
575	}
576
577	/**
578	 * Return the direction (ltr or rtl) for a given script
579	 *
580	 * The PHP/intl library does not provde this information, so we need
581	 * our own lookup table.
582	 *
583	 * @param string $script
584	 *
585	 * @return string
586	 */
587	public static function scriptDirection($script) {
588		switch ($script) {
589		case 'Arab':
590		case 'Hebr':
591		case 'Mong':
592		case 'Thaa':
593			return 'rtl';
594		default:
595			return 'ltr';
596		}
597	}
598
599	/**
600	 * UTF8 version of PHP::strtoupper()
601	 *
602	 * Convert a string to upper case, using the rules from the current locale
603	 *
604	 * @param string $string
605	 *
606	 * @return string
607	 */
608	public static function strtoupper($string) {
609		if (self::$locale->language()->code() === 'tr' || self::$locale->language()->code() === 'az') {
610			return TurkishUtf8::strtoupper($string);
611		} else {
612			return mb_strtoupper($string);
613		}
614	}
615
616	/**
617	 * UTF8 version of PHP::strtolower()
618	 *
619	 * Convert a string to lower case, using the rules from the current locale
620	 *
621	 * @param string $string
622	 *
623	 * @return string
624	 */
625	public static function strtolower($string) {
626		if (self::$locale->language()->code() === 'tr' || self::$locale->language()->code() === 'az') {
627			return TurkishUtf8::strtolower($string);
628		} else {
629			return mb_strtolower($string);
630		}
631	}
632
633	/**
634	 * UTF8 version of PHP::strcasecmp()
635	 *
636	 * Perform a case-insensitive comparison of two strings, using rules from the current locale
637	 *
638	 * @param string $string1
639	 * @param string $string2
640	 *
641	 * @return integer
642	 */
643	public static function strcasecmp($string1, $string2) {
644		$strpos1 = 0;
645		$strpos2 = 0;
646		$strlen1 = strlen($string1);
647		$strlen2 = strlen($string2);
648		while ($strpos1 < $strlen1 && $strpos2 < $strlen2) {
649			$byte1 = ord($string1[$strpos1]);
650			$byte2 = ord($string2[$strpos2]);
651			if (($byte1 & 0xE0) === 0xC0) {
652				$chr1 = $string1[$strpos1++] . $string1[$strpos1++];
653			} elseif (($byte1 & 0xF0) === 0xE0) {
654				$chr1 = $string1[$strpos1++] . $string1[$strpos1++] . $string1[$strpos1++];
655			} else {
656				$chr1 = $string1[$strpos1++];
657			}
658			if (($byte2 & 0xE0) === 0xC0) {
659				$chr2 = $string2[$strpos2++] . $string2[$strpos2++];
660			} elseif (($byte2 & 0xF0) === 0xE0) {
661				$chr2 = $string2[$strpos2++] . $string2[$strpos2++] . $string2[$strpos2++];
662			} else {
663				$chr2 = $string2[$strpos2++];
664			}
665			if ($chr1 === $chr2) {
666				continue;
667			}
668			// Try the local alphabet first
669			$offset1 = strpos(self::$alphabet_lower, $chr1);
670			if ($offset1 === false) {
671				$offset1 = strpos(self::$alphabet_upper, $chr1);
672			}
673			$offset2 = strpos(self::$alphabet_lower, $chr2);
674			if ($offset2 === false) {
675				$offset2 = strpos(self::$alphabet_upper, $chr2);
676			}
677			if ($offset1 !== false && $offset2 !== false) {
678				if ($offset1 === $offset2) {
679					continue;
680				} else {
681					return $offset1 - $offset2;
682				}
683			}
684			// Try the global alphabet next
685			$offset1 = strpos(self::ALPHABET_LOWER, $chr1);
686			if ($offset1 === false) {
687				$offset1 = strpos(self::ALPHABET_UPPER, $chr1);
688			}
689			$offset2 = strpos(self::ALPHABET_LOWER, $chr2);
690			if ($offset2 === false) {
691				$offset2 = strpos(self::ALPHABET_UPPER, $chr2);
692			}
693			if ($offset1 !== false && $offset2 !== false) {
694				if ($offset1 === $offset2) {
695					continue;
696				} else {
697					return $offset1 - $offset2;
698				}
699			}
700			// Just compare by unicode order
701			return strcmp($chr1, $chr2);
702		}
703		// Shortest string comes first.
704		return ($strlen1 - $strpos1) - ($strlen2 - $strpos2);
705	}
706
707	/**
708	 * UTF8 version of PHP::strrev()
709	 *
710	 * Reverse RTL text for third-party libraries such as GD2 and googlechart.
711	 *
712	 * These do not support UTF8 text direction, so we must mimic it for them.
713	 *
714	 * Numbers are always rendered LTR, even in RTL text.
715	 * The visual direction of characters such as parentheses should be reversed.
716	 *
717	 * @param string $text Text to be reversed
718	 *
719	 * @return string
720	 */
721	public static function reverseText($text) {
722		// Remove HTML markup - we can't display it and it is LTR.
723		$text = Filter::unescapeHtml($text);
724
725		// LTR text doesn't need reversing
726		if (self::scriptDirection(self::textScript($text)) === 'ltr') {
727			return $text;
728		}
729
730		// Mirrored characters
731		$text = strtr($text, self::$mirror_characters);
732
733		$reversed = '';
734		$digits = '';
735		while ($text != '') {
736			$letter = mb_substr($text, 0, 1);
737			$text = mb_substr($text, 1);
738			if (strpos(self::DIGITS, $letter) !== false) {
739				$digits .= $letter;
740			} else {
741				$reversed = $letter . $digits . $reversed;
742				$digits = '';
743			}
744		}
745
746		return $digits . $reversed;
747	}
748
749	/**
750	 * Generate consistent I18N for datatables.js
751	 *
752	 * @param array|null $lengths An optional array of page lengths
753	 *
754	 * @return string
755	 */
756	public static function datatablesI18N(array $lengths = null) {
757		if ($lengths === null) {
758			$lengths = array(10, 20, 30, 50, 100, -1);
759		}
760
761		$length_menu = '';
762		foreach ($lengths as $length) {
763			$length_menu .=
764				'<option value="' . $length . '">' .
765				($length === -1 ? /* I18N: listbox option, e.g. “10,25,50,100,all” */ self::translate('All') : self::number($length)) .
766				'</option>';
767		}
768		$length_menu = '<select>' . $length_menu . '</select>';
769		$length_menu = /* I18N: Display %s [records per page], %s is a placeholder for listbox containing numeric options */ self::translate('Display %s', $length_menu);
770
771		$digits = self::$locale->digits('0123456789');
772		if ($digits === '0123456789') {
773			$callback = '';
774		} else {
775			$callback = ',
776				"infoCallback": function(oSettings, iStart, iEnd, iMax, iTotal, sPre) {
777					return sPre
778						.replace(/0/g, "' . mb_substr($digits, 0, 1) . '")
779						.replace(/1/g, "' . mb_substr($digits, 1, 1) . '")
780						.replace(/2/g, "' . mb_substr($digits, 2, 1) . '")
781						.replace(/3/g, "' . mb_substr($digits, 3, 1) . '")
782						.replace(/4/g, "' . mb_substr($digits, 4, 1) . '")
783						.replace(/5/g, "' . mb_substr($digits, 5, 1) . '")
784						.replace(/6/g, "' . mb_substr($digits, 6, 1) . '")
785						.replace(/7/g, "' . mb_substr($digits, 7, 1) . '")
786						.replace(/8/g, "' . mb_substr($digits, 8, 1) . '")
787						.replace(/9/g, "' . mb_substr($digits, 9, 1) . '");
788				},
789				"formatNumber": function(iIn) {
790					return String(iIn)
791						.replace(/0/g, "' . mb_substr($digits, 0, 1) . '")
792						.replace(/1/g, "' . mb_substr($digits, 1, 1) . '")
793						.replace(/2/g, "' . mb_substr($digits, 2, 1) . '")
794						.replace(/3/g, "' . mb_substr($digits, 3, 1) . '")
795						.replace(/4/g, "' . mb_substr($digits, 4, 1) . '")
796						.replace(/5/g, "' . mb_substr($digits, 5, 1) . '")
797						.replace(/6/g, "' . mb_substr($digits, 6, 1) . '")
798						.replace(/7/g, "' . mb_substr($digits, 7, 1) . '")
799						.replace(/8/g, "' . mb_substr($digits, 8, 1) . '")
800						.replace(/9/g, "' . mb_substr($digits, 9, 1) . '");
801				}
802			';
803		}
804
805		return
806			'"language": {' .
807			' "paginate": {' .
808			'  "first":    "' . /* I18N: button label, first page    */ self::translate('first') . '",' .
809			'  "last":     "' . /* I18N: button label, last page     */ self::translate('last') . '",' .
810			'  "next":     "' . /* I18N: button label, next page     */ self::translate('next') . '",' .
811			'  "previous": "' . /* I18N: button label, previous page */ self::translate('previous') . '"' .
812			' },' .
813			' "emptyTable":     "' . self::translate('No records to display') . '",' .
814			' "info":           "' . /* I18N: %s are placeholders for numbers */ self::translate('Showing %1$s to %2$s of %3$s', '_START_', '_END_', '_TOTAL_') . '",' .
815			' "infoEmpty":      "' . self::translate('Showing %1$s to %2$s of %3$s', 0, 0, 0) . '",' .
816			' "infoFiltered":   "' . /* I18N: %s is a placeholder for a number */ self::translate('(filtered from %s total entries)', '_MAX_') . '",' .
817			' "infoPostfix":    "",' .
818			' "lengthMenu":     "' . Filter::escapeJs($length_menu) . '",' .
819			' "loadingRecords": "' . self::translate('Loading…') . '",' .
820			' "processing":     "' . self::translate('Loading…') . '",' .
821			' "search":         "' . self::translate('Filter') . '",' .
822			' "url":            "",' .
823			' "zeroRecords":    "' . self::translate('No records to display') . '"' .
824			'}' .
825			$callback;
826	}
827}
828