xref: /webtrees/app/I18N.php (revision 3bdc890b36a371d2e439cb152023b7eb79257527)
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 The MySQL collation sequence used by this language, typically utf8_unicode_ci */
96	public static $collation;
97
98	/** @var string Punctuation used to separate list items, typically a comma */
99	public static $list_separator;
100
101	/**
102	 * Initialise the translation adapter with a locale setting.
103	 *
104	 * @param string|null $code Use this locale/language code, or choose one automatically
105	 *
106	 * @return string $string
107	 */
108	public static function init($code = null) {
109		global $WT_SESSION, $WT_TREE;
110
111		if ($code !== null) {
112			// Create the specified locale
113			self::$locale = Locale::create($code);
114		} else {
115			// Negotiate a locale, but if we can't then use a failsafe
116			self::$locale = new LocaleEnUs;
117			if (Filter::get('lang')) {
118				// A request in the URL
119				try {
120					$locale = Locale::create(Filter::get('lang'));
121					if (file_exists(WT_ROOT . 'language/' . $locale->languageTag() . '.mo')) {
122						self::$locale = $locale;
123					}
124				} catch (\Exception $ex) {
125				}
126			} elseif ($WT_SESSION->locale) {
127				// Previously used
128				self::$locale = Locale::create($WT_SESSION->locale);
129			} elseif (isset($_SERVER['HTTP_ACCEPT_LANGUAGE'])) {
130				// Browser negotiation
131				$http_accept_language = strtolower(str_replace(' ', '', $_SERVER['HTTP_ACCEPT_LANGUAGE']));
132				if (preg_match_all('/(?:([a-z][a-z0-9_-]+)(?:;q=([0-9.]+))?)/', $http_accept_language, $match)) {
133					$preferences = array_combine($match[1], $match[2]);
134					array_walk($preferences, function(&$x) { $x = $x === '' ? 1.0 : (float) $x; });
135				} else {
136					$preferences = array();
137				}
138				// Add the tree’s default language as a low-priority
139				if ($WT_TREE && !isset($preferences[$WT_TREE->getPreference('LANGUAGE')])) {
140					$preferences[$WT_TREE->getPreference('LANGUAGE')] = 0.2;
141				}
142				arsort($preferences);
143				foreach (array_keys($preferences) as $code) {
144					try {
145						$locale = Locale::create($code);
146						if (file_exists(WT_ROOT . 'language/' . $locale->languageTag() . '.mo')) {
147							self::$locale = $locale;
148							break;
149						}
150					} catch (\Exception $ex) {
151						// The user's prefered locale does not exist
152					}
153				}
154			}
155		}
156
157		File::mkdir( WT_DATA_DIR . 'cache');
158		$cache_file = WT_DATA_DIR . 'cache/language-' . self::$locale->languageTag() . '-cache.php';
159		if (file_exists($cache_file)) {
160			$filemtime = filemtime($cache_file);
161		} else {
162			$filemtime = 0;
163		}
164
165		// Load the translation file(s)
166		$translation_files = array_merge(
167			array(WT_ROOT . 'language/' . self::$locale->languageTag() . '.mo'),
168			glob(WT_MODULES_DIR . '*/language/*.{csv,php,mo}', GLOB_BRACE),
169			glob(WT_DATA_DIR . 'language/*.{csv,php,mo}', GLOB_BRACE)
170		);
171
172		$rebuild_cache = false;
173		foreach ($translation_files as $translation_file) {
174			if (filemtime($translation_file) > $filemtime) {
175				$rebuild_cache = true;
176				break;
177			}
178		}
179
180		if ($rebuild_cache) {
181			$translations = array();
182			foreach ($translation_files as $translation_file) {
183				$translation = new Translation($translation_file);
184				$translations = array_merge($translations, $translation->asArray());
185			}
186			file_put_contents($cache_file, '<' . '?php return ' . var_export($translations, true) . ';');
187		} else {
188			$translations = include $cache_file;
189		}
190
191		// Create a translator
192		self::$translator = new Translator($translations, self::$locale->pluralRule());
193
194		// Extract language settings from the translation file
195		global $DATE_FORMAT; // I18N: This is the format string for full dates.  See http://php.net/date for codes
196		$DATE_FORMAT = self::$translator->translate('%j %F %Y');
197
198		global $TIME_FORMAT; // I18N: This is the format string for the time-of-day.  See http://php.net/date for codes
199		$TIME_FORMAT = self::$translator->translate('%H:%i:%s');
200
201		// Alphabetic sorting sequence (upper-case letters), used by webtrees to sort strings
202		list(, self::$alphabet_upper) = explode('=', self::$translator->translate('ALPHABET_upper=ABCDEFGHIJKLMNOPQRSTUVWXYZ'));
203		// Alphabetic sorting sequence (lower-case letters), used by webtrees to sort strings
204		list(, self::$alphabet_lower) = explode('=', self::$translator->translate('ALPHABET_lower=abcdefghijklmnopqrstuvwxyz'));
205
206		global $WEEK_START;
207		$WEEK_START = self::$locale->territory()->firstDay();
208
209		self::$list_separator = /* I18N: This punctuation is used to separate lists of items */ self::translate(', ');
210		self::$collation = 'utf8_' . self::$locale->collation();
211
212		return self::$locale->languageTag();
213	}
214
215	/**
216	 * The prefered locales for this site, or a default list if no preference.
217	 *
218	 * @return Locale[]
219	 */
220	public static function activeLocales() {
221		$code_list = Site::getPreference('LANGUAGES');
222
223		if ($code_list) {
224			$codes = explode(',', $code_list);
225		} else {
226			$codes = array(
227				'ar', 'bg', 'bs', 'ca', 'cs', 'da', 'de', 'el', 'en-GB', 'en-US', 'es',
228				'et', 'fi', 'fr', 'he', 'hr', 'hu', 'is', 'it', 'ka', 'lt', 'mr', 'nb',
229				'nl', 'nn', 'pl', 'pt', 'ru', 'sk', 'sv', 'tr', 'uk', 'vi', 'zh-Hans',
230			);
231		}
232
233		$locales = array();
234		foreach ($codes as $code) {
235			if (file_exists(WT_ROOT . 'language/' . $code . '.mo')) {
236				try {
237					$locales[] = Locale::create($code);
238				} catch (\Exception $ex) {
239					// No such locale exists?
240				}
241			}
242		}
243		usort($locales, '\Fisharebest\Localization\Locale::compare');
244
245		return $locales;
246	}
247
248	/**
249	 * All locales for which a translation file exists.
250	 *
251	 * @return Locale[]
252	 */
253	public static function installedLocales() {
254		$locales = array();
255		foreach (glob(WT_ROOT . 'language/*.mo') as $file) {
256			try {
257				$locales[] = Locale::create(basename($file, '.mo'));
258			} catch (\Exception $ex) {
259				// Not a recognised locale
260			}
261		}
262		usort($locales, '\Fisharebest\Localization\Locale::compare');
263
264		return $locales;
265	}
266
267	/**
268	 * Generate i18n markup for the <html> tag, e.g. lang="ar" dir="rtl"
269	 *
270	 * @return string
271	 */
272	public static function htmlAttributes() {
273		return self::$locale->htmlAttributes();
274	}
275
276	/**
277	 * Translate a number into the local representation.
278	 *
279	 * e.g. 12345.67 becomes
280	 * en: 12,345.67
281	 * fr: 12 345,67
282	 * de: 12.345,67
283	 *
284	 * @param float   $n
285	 * @param integer $precision
286	 *
287	 * @return string
288	 */
289	public static function number($n, $precision = 0) {
290		return self::$locale->number(round($n, $precision));
291	}
292
293	/**
294	 * Convert the digits 0-9 into the local script
295	 *
296	 * Used for years, etc., where we do not want thousands-separators, decimals, etc.
297	 *
298	 * @param integer $n
299	 *
300	 * @return string
301	 */
302	public static function digits($n) {
303		return self::$locale->digits($n);
304	}
305
306	/**
307	 * Translate a fraction into a percentage.
308	 *
309	 * e.g. 0.123 becomes
310	 * en: 12.3%
311	 * fr: 12,3 %
312	 * de: 12,3%
313	 *
314	 * @param float   $n
315	 * @param integer $precision
316	 *
317	 * @return string
318	 */
319	public static function percentage($n, $precision = 0) {
320		return self::$locale->percent($n, $precision);
321	}
322
323	/**
324	 * What is the direction of the current locale
325	 *
326	 * @return string "ltr" or "rtl"
327	 */
328	public static function direction() {
329		return self::$locale->direction();
330	}
331
332	/**
333	 * Translate a string, and then substitute placeholders
334	 *
335	 * echo I18N::translate('Hello World!');
336	 * echo I18N::translate('The %s sat on the mat', 'cat');
337	 *
338	 * @return string
339	 */
340	public static function translate(/* var_args */) {
341		$args = func_get_args();
342		$args[0] = self::$translator->translate($args[0]);
343
344		return call_user_func_array('sprintf', $args);
345	}
346
347	/**
348	 * Context sensitive version of translate.
349	 *
350	 * echo I18N::translate_c('NOMINATIVE', 'January');
351	 * echo I18N::translate_c('GENITIVE',   'January');
352	 *
353	 * @return string
354	 */
355	public static function translateContext(/* var_args */) {
356		$args = func_get_args();
357		$args[0] = self::$translator->translateContext($args[0], $args[1]);
358		unset($args[1]);
359
360		return call_user_func_array('sprintf', $args);
361	}
362
363	/**
364	 * Similar to translate, but do perform "no operation" on it.
365	 *
366	 * This is necessary to fetch a format string (containing % characters) without
367	 * performing sustitution of arguments.
368	 *
369	 * @param string $string
370	 *
371	 * @return string
372	 */
373	public static function noop($string) {
374		return self::$translator->translate($string);
375	}
376
377	/**
378	 * Translate a plural string
379	 *
380	 * echo self::plural('There is an error', 'There are errors', $num_errors);
381	 * echo self::plural('There is one error', 'There are %s errors', $num_errors);
382	 * echo self::plural('There is %1$s %2$s cat', 'There are %1$s %2$s cats', $num, $num, $colour);
383	 *
384	 * @return string
385	 */
386	public static function plural(/* var_args */) {
387		$args = func_get_args();
388		$args[0] = self::$translator->plural($args[0], $args[1], $args[2]);
389		unset($args[1], $args[2]);
390
391		return call_user_func_array('sprintf', $args);
392	}
393
394	/**
395	 * Convert a GEDCOM age string into translated_text
396	 *
397	 * NB: The import function will have normalised this, so we don't need
398	 * to worry about badly formatted strings
399	 * NOTE: this function is not yet complete - eventually it will replace get_age_at_event()
400	 *
401	 * @param $string
402	 *
403	 * @return string
404	 */
405	public static function gedcomAge($string) {
406		switch ($string) {
407		case 'STILLBORN':
408			// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (stillborn)
409			return self::translate('(stillborn)');
410		case 'INFANT':
411			// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (in infancy)
412			return self::translate('(in infancy)');
413		case 'CHILD':
414			// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (in childhood)
415			return self::translate('(in childhood)');
416		}
417		$age = array();
418		if (preg_match('/(\d+)y/', $string, $match)) {
419			// I18N: Part of an age string. e.g. 5 years, 4 months and 3 days
420			$years = $match[1];
421			$age[] = self::plural('%s year', '%s years', $years, self::number($years));
422		} else {
423			$years = -1;
424		}
425		if (preg_match('/(\d+)m/', $string, $match)) {
426			// I18N: Part of an age string. e.g. 5 years, 4 months and 3 days
427			$age[] = self::plural('%s month', '%s months', $match[1], self::number($match[1]));
428		}
429		if (preg_match('/(\d+)w/', $string, $match)) {
430			// I18N: Part of an age string. e.g. 7 weeks and 3 days
431			$age[] = self::plural('%s week', '%s weeks', $match[1], self::number($match[1]));
432		}
433		if (preg_match('/(\d+)d/', $string, $match)) {
434			// I18N: Part of an age string. e.g. 5 years, 4 months and 3 days
435			$age[] = self::plural('%s day', '%s days', $match[1], self::number($match[1]));
436		}
437		// If an age is just a number of years, only show the number
438		if (count($age) === 1 && $years >= 0) {
439			$age = $years;
440		}
441		if ($age) {
442			if (!substr_compare($string, '<', 0, 1)) {
443				// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (aged less than 21 years)
444				return self::translate('(aged less than %s)', $age);
445			} elseif (!substr_compare($string, '>', 0, 1)) {
446				// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (aged more than 21 years)
447				return self::translate('(aged more than %s)', $age);
448			} else {
449				// I18N: Description of an individual’s age at an event.  For example, Died 14 Jan 1900 (aged 43 years)
450				return self::translate('(aged %s)', $age);
451			}
452		} else {
453			// Not a valid string?
454			return self::translate('(aged %s)', $string);
455		}
456	}
457
458	/**
459	 * Convert a number of seconds into a relative time.  For example, 630 => "10 hours, 30 minutes ago"
460	 *
461	 * @param integer $seconds
462	 *
463	 * @return string
464	 */
465	public static function timeAgo($seconds) {
466		$minute = 60;
467		$hour   = 60 * $minute;
468		$day    = 24 * $hour;
469		$month  = 30 * $day;
470		$year   = 365 * $day;
471
472		if ($seconds > $year) {
473			$years = (int) ($seconds / $year);
474			return self::plural('%s year ago', '%s years ago', $years, self::number($years));
475		} elseif ($seconds > $month) {
476			$months = (int) ($seconds / $month);
477			return self::plural('%s month ago', '%s months ago', $months, self::number($months));
478		} elseif ($seconds > $day) {
479			$days = (int) ($seconds / $day);
480			return self::plural('%s day ago', '%s days ago', $days, self::number($days));
481		} elseif ($seconds > $hour) {
482			$hours = (int) ($seconds / $hour);
483			return self::plural('%s hour ago', '%s hours ago', $hours, self::number($hours));
484		} elseif ($seconds > $minute) {
485			$minutes = (int) ($seconds / $minute);
486			return self::plural('%s minute ago', '%s minutes ago', $minutes, self::number($minutes));
487		} else {
488			return self::plural('%s second ago', '%s seconds ago', $seconds, self::number($seconds));
489		}
490	}
491
492	/**
493	 * Return the endonym for a given language - as per http://cldr.unicode.org/
494	 *
495	 * @param string $locale
496	 *
497	 * @return string
498	 */
499	public static function languageName($locale) {
500		return Locale::create($locale)->endonym();
501	}
502
503	/**
504	 * Return the script used by a given language
505	 *
506	 * @param string $locale
507	 *
508	 * @return string
509	 */
510	public static function languageScript($locale) {
511		return Locale::create($locale)->script()->code();
512	}
513
514	/**
515	 * Identify the script used for a piece of text
516	 *
517	 * @param $string
518	 *
519	 * @return string
520	 */
521	public static function textScript($string) {
522		$string = strip_tags($string); // otherwise HTML tags show up as latin
523		$string = html_entity_decode($string, ENT_QUOTES, 'UTF-8'); // otherwise HTML entities show up as latin
524		$string = str_replace(array('@N.N.', '@P.N.'), '', $string); // otherwise unknown names show up as latin
525		$pos = 0;
526		$strlen = strlen($string);
527		while ($pos < $strlen) {
528			// get the Unicode Code Point for the character at position $pos
529			$byte1 = ord($string[$pos]);
530			if ($byte1 < 0x80) {
531				$code_point = $byte1;
532				$chrlen = 1;
533			} elseif ($byte1 < 0xC0) {
534				// Invalid continuation character
535				return 'Latn';
536			} elseif ($byte1 < 0xE0) {
537				$code_point = (($byte1 & 0x1F) << 6) + (ord($string[$pos + 1]) & 0x3F);
538				$chrlen = 2;
539			} elseif ($byte1 < 0xF0) {
540				$code_point = (($byte1 & 0x0F) << 12) + ((ord($string[$pos + 1]) & 0x3F) << 6) + (ord($string[$pos + 2]) & 0x3F);
541				$chrlen = 3;
542			} elseif ($byte1 < 0xF8) {
543				$code_point = (($byte1 & 0x07) << 24) + ((ord($string[$pos + 1]) & 0x3F) << 12) + ((ord($string[$pos + 2]) & 0x3F) << 6) + (ord($string[$pos + 3]) & 0x3F);
544				$chrlen = 3;
545 			} else {
546				// Invalid UTF
547				return 'Latn';
548			}
549
550			foreach (self::$scripts as $range) {
551				if ($code_point >= $range[1] && $code_point <= $range[2]) {
552					return $range[0];
553				}
554			}
555			// Not a recognised script.  Maybe punctuation, spacing, etc.  Keep looking.
556			$pos += $chrlen;
557		}
558
559		return 'Latn';
560	}
561
562	/**
563	 * Return the direction (ltr or rtl) for a given script
564	 *
565	 * The PHP/intl library does not provde this information, so we need
566	 * our own lookup table.
567	 *
568	 * @param string $script
569	 *
570	 * @return string
571	 */
572	public static function scriptDirection($script) {
573		switch ($script) {
574		case 'Arab':
575		case 'Hebr':
576		case 'Mong':
577		case 'Thaa':
578			return 'rtl';
579		default:
580			return 'ltr';
581		}
582	}
583
584	/**
585	 * UTF8 version of PHP::strtoupper()
586	 *
587	 * Convert a string to upper case, using the rules from the current locale
588	 *
589	 * @param string $string
590	 *
591	 * @return string
592	 */
593	public static function strtoupper($string) {
594		if (self::$locale->language()->code() === 'tr' || self::$locale->language()->code() === 'az') {
595			return TurkishUtf8::strtoupper($string);
596		} else {
597			return mb_strtoupper($string);
598		}
599	}
600
601	/**
602	 * UTF8 version of PHP::strtolower()
603	 *
604	 * Convert a string to lower case, using the rules from the current locale
605	 *
606	 * @param string $string
607	 *
608	 * @return string
609	 */
610	public static function strtolower($string) {
611		if (self::$locale->language()->code() === 'tr' || self::$locale->language()->code() === 'az') {
612			return TurkishUtf8::strtolower($string);
613		} else {
614			return mb_strtolower($string);
615		}
616	}
617
618	/**
619	 * UTF8 version of PHP::strcasecmp()
620	 *
621	 * Perform a case-insensitive comparison of two strings, using rules from the current locale
622	 *
623	 * @param string $string1
624	 * @param string $string2
625	 *
626	 * @return integer
627	 */
628	public static function strcasecmp($string1, $string2) {
629		$strpos1 = 0;
630		$strpos2 = 0;
631		$strlen1 = strlen($string1);
632		$strlen2 = strlen($string2);
633		while ($strpos1 < $strlen1 && $strpos2 < $strlen2) {
634			$byte1 = ord($string1[$strpos1]);
635			$byte2 = ord($string2[$strpos2]);
636			if (($byte1 & 0xE0) === 0xC0) {
637				$chr1 = $string1[$strpos1++] . $string1[$strpos1++];
638			} elseif (($byte1 & 0xF0) === 0xE0) {
639				$chr1 = $string1[$strpos1++] . $string1[$strpos1++] . $string1[$strpos1++];
640			} else {
641				$chr1 = $string1[$strpos1++];
642			}
643			if (($byte2 & 0xE0) === 0xC0) {
644				$chr2 = $string2[$strpos2++] . $string2[$strpos2++];
645			} elseif (($byte2 & 0xF0) === 0xE0) {
646				$chr2 = $string2[$strpos2++] . $string2[$strpos2++] . $string2[$strpos2++];
647			} else {
648				$chr2 = $string2[$strpos2++];
649			}
650			if ($chr1 === $chr2) {
651				continue;
652			}
653			// Try the local alphabet first
654			$offset1 = strpos(self::$alphabet_lower, $chr1);
655			if ($offset1 === false) {
656				$offset1 = strpos(self::$alphabet_upper, $chr1);
657			}
658			$offset2 = strpos(self::$alphabet_lower, $chr2);
659			if ($offset2 === false) {
660				$offset2 = strpos(self::$alphabet_upper, $chr2);
661			}
662			if ($offset1 !== false && $offset2 !== false) {
663				if ($offset1 === $offset2) {
664					continue;
665				} else {
666					return $offset1 - $offset2;
667				}
668			}
669			// Try the global alphabet next
670			$offset1 = strpos(self::ALPHABET_LOWER, $chr1);
671			if ($offset1 === false) {
672				$offset1 = strpos(self::ALPHABET_UPPER, $chr1);
673			}
674			$offset2 = strpos(self::ALPHABET_LOWER, $chr2);
675			if ($offset2 === false) {
676				$offset2 = strpos(self::ALPHABET_UPPER, $chr2);
677			}
678			if ($offset1 !== false && $offset2 !== false) {
679				if ($offset1 === $offset2) {
680					continue;
681				} else {
682					return $offset1 - $offset2;
683				}
684			}
685			// Just compare by unicode order
686			return strcmp($chr1, $chr2);
687		}
688		// Shortest string comes first.
689		return ($strlen1 - $strpos1) - ($strlen2 - $strpos2);
690	}
691
692	/**
693	 * UTF8 version of PHP::strrev()
694	 *
695	 * Reverse RTL text for third-party libraries such as GD2 and googlechart.
696	 *
697	 * These do not support UTF8 text direction, so we must mimic it for them.
698	 *
699	 * Numbers are always rendered LTR, even in RTL text.
700	 * The visual direction of characters such as parentheses should be reversed.
701	 *
702	 * @param string $text Text to be reversed
703	 *
704	 * @return string
705	 */
706	public static function reverseText($text) {
707		// Remove HTML markup - we can't display it and it is LTR.
708		$text = Filter::unescapeHtml($text);
709
710		// LTR text doesn't need reversing
711		if (self::scriptDirection(self::textScript($text)) === 'ltr') {
712			return $text;
713		}
714
715		// Mirrored characters
716		$text = strtr($text, self::$mirror_characters);
717
718		$reversed = '';
719		$digits = '';
720		while ($text != '') {
721			$letter = mb_substr($text, 0, 1);
722			$text = mb_substr($text, 1);
723			if (strpos(self::DIGITS, $letter) !== false) {
724				$digits .= $letter;
725			} else {
726				$reversed = $letter . $digits . $reversed;
727				$digits = '';
728			}
729		}
730
731		return $digits . $reversed;
732	}
733
734	/**
735	 * Generate consistent I18N for datatables.js
736	 *
737	 * @param array|null $lengths An optional array of page lengths
738	 *
739	 * @return string
740	 */
741	public static function datatablesI18N(array $lengths = null) {
742		if ($lengths === null) {
743			$lengths = array(10, 20, 30, 50, 100, -1);
744		}
745
746		$length_menu = '';
747		foreach ($lengths as $length) {
748			$length_menu .=
749				'<option value="' . $length . '">' .
750				($length === -1 ? /* I18N: listbox option, e.g. “10,25,50,100,all” */ self::translate('All') : self::number($length)) .
751				'</option>';
752		}
753		$length_menu = '<select>' . $length_menu . '</select>';
754		$length_menu = /* I18N: Display %s [records per page], %s is a placeholder for listbox containing numeric options */ self::translate('Display %s', $length_menu);
755
756		$digits = self::$locale->digits('0123456789');
757		if ($digits === '0123456789') {
758			$callback = '';
759		} else {
760			$callback = ',
761				"infoCallback": function(oSettings, iStart, iEnd, iMax, iTotal, sPre) {
762					return sPre
763						.replace(/0/g, "' . mb_substr($digits, 0, 1) . '")
764						.replace(/1/g, "' . mb_substr($digits, 1, 1) . '")
765						.replace(/2/g, "' . mb_substr($digits, 2, 1) . '")
766						.replace(/3/g, "' . mb_substr($digits, 3, 1) . '")
767						.replace(/4/g, "' . mb_substr($digits, 4, 1) . '")
768						.replace(/5/g, "' . mb_substr($digits, 5, 1) . '")
769						.replace(/6/g, "' . mb_substr($digits, 6, 1) . '")
770						.replace(/7/g, "' . mb_substr($digits, 7, 1) . '")
771						.replace(/8/g, "' . mb_substr($digits, 8, 1) . '")
772						.replace(/9/g, "' . mb_substr($digits, 9, 1) . '");
773				},
774				"formatNumber": function(iIn) {
775					return String(iIn)
776						.replace(/0/g, "' . mb_substr($digits, 0, 1) . '")
777						.replace(/1/g, "' . mb_substr($digits, 1, 1) . '")
778						.replace(/2/g, "' . mb_substr($digits, 2, 1) . '")
779						.replace(/3/g, "' . mb_substr($digits, 3, 1) . '")
780						.replace(/4/g, "' . mb_substr($digits, 4, 1) . '")
781						.replace(/5/g, "' . mb_substr($digits, 5, 1) . '")
782						.replace(/6/g, "' . mb_substr($digits, 6, 1) . '")
783						.replace(/7/g, "' . mb_substr($digits, 7, 1) . '")
784						.replace(/8/g, "' . mb_substr($digits, 8, 1) . '")
785						.replace(/9/g, "' . mb_substr($digits, 9, 1) . '");
786				}
787			';
788		}
789
790		return
791			'"language": {' .
792			' "paginate": {' .
793			'  "first":    "' . /* I18N: button label, first page    */ self::translate('first') . '",' .
794			'  "last":     "' . /* I18N: button label, last page     */ self::translate('last') . '",' .
795			'  "next":     "' . /* I18N: button label, next page     */ self::translate('next') . '",' .
796			'  "previous": "' . /* I18N: button label, previous page */ self::translate('previous') . '"' .
797			' },' .
798			' "emptyTable":     "' . self::translate('No records to display') . '",' .
799			' "info":           "' . /* I18N: %s are placeholders for numbers */ self::translate('Showing %1$s to %2$s of %3$s', '_START_', '_END_', '_TOTAL_') . '",' .
800			' "infoEmpty":      "' . self::translate('Showing %1$s to %2$s of %3$s', 0, 0, 0) . '",' .
801			' "infoFiltered":   "' . /* I18N: %s is a placeholder for a number */ self::translate('(filtered from %s total entries)', '_MAX_') . '",' .
802			' "infoPostfix":    "",' .
803			' "lengthMenu":     "' . Filter::escapeJs($length_menu) . '",' .
804			' "loadingRecords": "' . self::translate('Loading…') . '",' .
805			' "processing":     "' . self::translate('Loading…') . '",' .
806			' "search":         "' . self::translate('Filter') . '",' .
807			' "url":            "",' .
808			' "zeroRecords":    "' . self::translate('No records to display') . '"' .
809			'}' .
810			$callback;
811	}
812}
813