3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (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.
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
26 * @defgroup Language Language
29 use CLDRPluralRuleParser\Evaluator
;
30 use MediaWiki\MediaWikiServices
;
31 use Wikimedia\Assert\Assert
;
34 * Internationalisation code
39 * Return autonyms in fetchLanguageName(s).
42 const AS_AUTONYMS
= null;
45 * Return all known languages in fetchLanguageName(s).
51 * Return in fetchLanguageName(s) only the languages for which we have at
52 * least some localisation.
55 const SUPPORTED
= 'mwfile';
58 * @var LanguageConverter|FakeConverter
62 public $mVariants, $mCode, $mLoaded = false;
63 public $mMagicExtensions = [];
64 private $mHtmlCode = null;
65 /** @var Language|false */
66 private $mParentLanguage = false;
68 public $dateFormatStrings = [];
69 public $mExtendedSpecialPageAliases;
71 /** @var array|null */
72 protected $namespaceNames;
73 protected $mNamespaceIds, $namespaceAliases;
76 * ReplacementArray object caches
78 public $transformData = [];
81 * @var LocalisationCache
83 public static $dataCache;
85 public static $mLangObjCache = [];
88 * Return a fallback chain for messages in getFallbacksFor
91 const MESSAGES_FALLBACKS
= 0;
94 * Return a strict fallback chain in getFallbacksFor
97 const STRICT_FALLBACKS
= 1;
99 public static $mWeekdayMsgs = [
100 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
104 public static $mWeekdayAbbrevMsgs = [
105 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
108 public static $mMonthMsgs = [
109 'january', 'february', 'march', 'april', 'may_long', 'june',
110 'july', 'august', 'september', 'october', 'november',
113 public static $mMonthGenMsgs = [
114 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
115 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
118 public static $mMonthAbbrevMsgs = [
119 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
120 'sep', 'oct', 'nov', 'dec'
123 public static $mIranianCalendarMonthMsgs = [
124 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
125 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
126 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
127 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
130 public static $mHebrewCalendarMonthMsgs = [
131 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
132 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
133 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
134 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
135 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
138 public static $mHebrewCalendarMonthGenMsgs = [
139 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
140 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
141 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
142 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
143 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
146 public static $mHijriCalendarMonthMsgs = [
147 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
148 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
149 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
150 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
157 public static $durationIntervals = [
158 'millennia' => 31556952000,
159 'centuries' => 3155695200,
160 'decades' => 315569520,
161 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
170 * Cache for language fallbacks.
171 * @see Language::getFallbacksIncludingSiteLanguage
175 private static $fallbackLanguageCache = [];
178 * Cache for grammar rules data
179 * @var MapCacheLRU|null
181 private static $grammarTransformations;
184 * Cache for language names
185 * @var HashBagOStuff|null
187 private static $languageNameCache;
190 * Unicode directional formatting characters, for embedBidi()
192 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
193 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
194 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
197 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
198 * - in group 1 if it is LTR
199 * - in group 2 if it is RTL
200 * Does not match if there is no strong directionality codepoint.
202 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
204 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
205 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
207 // @codeCoverageIgnoreStart
208 // phpcs:ignore Generic.Files.LineLength
209 private static $strongDirRegex = '/(?:([\x{41}-\x{5a}\x{61}-\x{7a}\x{aa}\x{b5}\x{ba}\x{c0}-\x{d6}\x{d8}-\x{f6}\x{f8}-\x{2b8}\x{2bb}-\x{2c1}\x{2d0}\x{2d1}\x{2e0}-\x{2e4}\x{2ee}\x{370}-\x{373}\x{376}\x{377}\x{37a}-\x{37d}\x{37f}\x{386}\x{388}-\x{38a}\x{38c}\x{38e}-\x{3a1}\x{3a3}-\x{3f5}\x{3f7}-\x{482}\x{48a}-\x{52f}\x{531}-\x{556}\x{559}-\x{55f}\x{561}-\x{587}\x{589}\x{903}-\x{939}\x{93b}\x{93d}-\x{940}\x{949}-\x{94c}\x{94e}-\x{950}\x{958}-\x{961}\x{964}-\x{980}\x{982}\x{983}\x{985}-\x{98c}\x{98f}\x{990}\x{993}-\x{9a8}\x{9aa}-\x{9b0}\x{9b2}\x{9b6}-\x{9b9}\x{9bd}-\x{9c0}\x{9c7}\x{9c8}\x{9cb}\x{9cc}\x{9ce}\x{9d7}\x{9dc}\x{9dd}\x{9df}-\x{9e1}\x{9e6}-\x{9f1}\x{9f4}-\x{9fa}\x{a03}\x{a05}-\x{a0a}\x{a0f}\x{a10}\x{a13}-\x{a28}\x{a2a}-\x{a30}\x{a32}\x{a33}\x{a35}\x{a36}\x{a38}\x{a39}\x{a3e}-\x{a40}\x{a59}-\x{a5c}\x{a5e}\x{a66}-\x{a6f}\x{a72}-\x{a74}\x{a83}\x{a85}-\x{a8d}\x{a8f}-\x{a91}\x{a93}-\x{aa8}\x{aaa}-\x{ab0}\x{ab2}\x{ab3}\x{ab5}-\x{ab9}\x{abd}-\x{ac0}\x{ac9}\x{acb}\x{acc}\x{ad0}\x{ae0}\x{ae1}\x{ae6}-\x{af0}\x{af9}\x{b02}\x{b03}\x{b05}-\x{b0c}\x{b0f}\x{b10}\x{b13}-\x{b28}\x{b2a}-\x{b30}\x{b32}\x{b33}\x{b35}-\x{b39}\x{b3d}\x{b3e}\x{b40}\x{b47}\x{b48}\x{b4b}\x{b4c}\x{b57}\x{b5c}\x{b5d}\x{b5f}-\x{b61}\x{b66}-\x{b77}\x{b83}\x{b85}-\x{b8a}\x{b8e}-\x{b90}\x{b92}-\x{b95}\x{b99}\x{b9a}\x{b9c}\x{b9e}\x{b9f}\x{ba3}\x{ba4}\x{ba8}-\x{baa}\x{bae}-\x{bb9}\x{bbe}\x{bbf}\x{bc1}\x{bc2}\x{bc6}-\x{bc8}\x{bca}-\x{bcc}\x{bd0}\x{bd7}\x{be6}-\x{bf2}\x{c01}-\x{c03}\x{c05}-\x{c0c}\x{c0e}-\x{c10}\x{c12}-\x{c28}\x{c2a}-\x{c39}\x{c3d}\x{c41}-\x{c44}\x{c58}-\x{c5a}\x{c60}\x{c61}\x{c66}-\x{c6f}\x{c7f}\x{c82}\x{c83}\x{c85}-\x{c8c}\x{c8e}-\x{c90}\x{c92}-\x{ca8}\x{caa}-\x{cb3}\x{cb5}-\x{cb9}\x{cbd}-\x{cc4}\x{cc6}-\x{cc8}\x{cca}\x{ccb}\x{cd5}\x{cd6}\x{cde}\x{ce0}\x{ce1}\x{ce6}-\x{cef}\x{cf1}\x{cf2}\x{d02}\x{d03}\x{d05}-\x{d0c}\x{d0e}-\x{d10}\x{d12}-\x{d3a}\x{d3d}-\x{d40}\x{d46}-\x{d48}\x{d4a}-\x{d4c}\x{d4e}\x{d57}\x{d5f}-\x{d61}\x{d66}-\x{d75}\x{d79}-\x{d7f}\x{d82}\x{d83}\x{d85}-\x{d96}\x{d9a}-\x{db1}\x{db3}-\x{dbb}\x{dbd}\x{dc0}-\x{dc6}\x{dcf}-\x{dd1}\x{dd8}-\x{ddf}\x{de6}-\x{def}\x{df2}-\x{df4}\x{e01}-\x{e30}\x{e32}\x{e33}\x{e40}-\x{e46}\x{e4f}-\x{e5b}\x{e81}\x{e82}\x{e84}\x{e87}\x{e88}\x{e8a}\x{e8d}\x{e94}-\x{e97}\x{e99}-\x{e9f}\x{ea1}-\x{ea3}\x{ea5}\x{ea7}\x{eaa}\x{eab}\x{ead}-\x{eb0}\x{eb2}\x{eb3}\x{ebd}\x{ec0}-\x{ec4}\x{ec6}\x{ed0}-\x{ed9}\x{edc}-\x{edf}\x{f00}-\x{f17}\x{f1a}-\x{f34}\x{f36}\x{f38}\x{f3e}-\x{f47}\x{f49}-\x{f6c}\x{f7f}\x{f85}\x{f88}-\x{f8c}\x{fbe}-\x{fc5}\x{fc7}-\x{fcc}\x{fce}-\x{fda}\x{1000}-\x{102c}\x{1031}\x{1038}\x{103b}\x{103c}\x{103f}-\x{1057}\x{105a}-\x{105d}\x{1061}-\x{1070}\x{1075}-\x{1081}\x{1083}\x{1084}\x{1087}-\x{108c}\x{108e}-\x{109c}\x{109e}-\x{10c5}\x{10c7}\x{10cd}\x{10d0}-\x{1248}\x{124a}-\x{124d}\x{1250}-\x{1256}\x{1258}\x{125a}-\x{125d}\x{1260}-\x{1288}\x{128a}-\x{128d}\x{1290}-\x{12b0}\x{12b2}-\x{12b5}\x{12b8}-\x{12be}\x{12c0}\x{12c2}-\x{12c5}\x{12c8}-\x{12d6}\x{12d8}-\x{1310}\x{1312}-\x{1315}\x{1318}-\x{135a}\x{1360}-\x{137c}\x{1380}-\x{138f}\x{13a0}-\x{13f5}\x{13f8}-\x{13fd}\x{1401}-\x{167f}\x{1681}-\x{169a}\x{16a0}-\x{16f8}\x{1700}-\x{170c}\x{170e}-\x{1711}\x{1720}-\x{1731}\x{1735}\x{1736}\x{1740}-\x{1751}\x{1760}-\x{176c}\x{176e}-\x{1770}\x{1780}-\x{17b3}\x{17b6}\x{17be}-\x{17c5}\x{17c7}\x{17c8}\x{17d4}-\x{17da}\x{17dc}\x{17e0}-\x{17e9}\x{1810}-\x{1819}\x{1820}-\x{1877}\x{1880}-\x{18a8}\x{18aa}\x{18b0}-\x{18f5}\x{1900}-\x{191e}\x{1923}-\x{1926}\x{1929}-\x{192b}\x{1930}\x{1931}\x{1933}-\x{1938}\x{1946}-\x{196d}\x{1970}-\x{1974}\x{1980}-\x{19ab}\x{19b0}-\x{19c9}\x{19d0}-\x{19da}\x{1a00}-\x{1a16}\x{1a19}\x{1a1a}\x{1a1e}-\x{1a55}\x{1a57}\x{1a61}\x{1a63}\x{1a64}\x{1a6d}-\x{1a72}\x{1a80}-\x{1a89}\x{1a90}-\x{1a99}\x{1aa0}-\x{1aad}\x{1b04}-\x{1b33}\x{1b35}\x{1b3b}\x{1b3d}-\x{1b41}\x{1b43}-\x{1b4b}\x{1b50}-\x{1b6a}\x{1b74}-\x{1b7c}\x{1b82}-\x{1ba1}\x{1ba6}\x{1ba7}\x{1baa}\x{1bae}-\x{1be5}\x{1be7}\x{1bea}-\x{1bec}\x{1bee}\x{1bf2}\x{1bf3}\x{1bfc}-\x{1c2b}\x{1c34}\x{1c35}\x{1c3b}-\x{1c49}\x{1c4d}-\x{1c7f}\x{1cc0}-\x{1cc7}\x{1cd3}\x{1ce1}\x{1ce9}-\x{1cec}\x{1cee}-\x{1cf3}\x{1cf5}\x{1cf6}\x{1d00}-\x{1dbf}\x{1e00}-\x{1f15}\x{1f18}-\x{1f1d}\x{1f20}-\x{1f45}\x{1f48}-\x{1f4d}\x{1f50}-\x{1f57}\x{1f59}\x{1f5b}\x{1f5d}\x{1f5f}-\x{1f7d}\x{1f80}-\x{1fb4}\x{1fb6}-\x{1fbc}\x{1fbe}\x{1fc2}-\x{1fc4}\x{1fc6}-\x{1fcc}\x{1fd0}-\x{1fd3}\x{1fd6}-\x{1fdb}\x{1fe0}-\x{1fec}\x{1ff2}-\x{1ff4}\x{1ff6}-\x{1ffc}\x{200e}\x{2071}\x{207f}\x{2090}-\x{209c}\x{2102}\x{2107}\x{210a}-\x{2113}\x{2115}\x{2119}-\x{211d}\x{2124}\x{2126}\x{2128}\x{212a}-\x{212d}\x{212f}-\x{2139}\x{213c}-\x{213f}\x{2145}-\x{2149}\x{214e}\x{214f}\x{2160}-\x{2188}\x{2336}-\x{237a}\x{2395}\x{249c}-\x{24e9}\x{26ac}\x{2800}-\x{28ff}\x{2c00}-\x{2c2e}\x{2c30}-\x{2c5e}\x{2c60}-\x{2ce4}\x{2ceb}-\x{2cee}\x{2cf2}\x{2cf3}\x{2d00}-\x{2d25}\x{2d27}\x{2d2d}\x{2d30}-\x{2d67}\x{2d6f}\x{2d70}\x{2d80}-\x{2d96}\x{2da0}-\x{2da6}\x{2da8}-\x{2dae}\x{2db0}-\x{2db6}\x{2db8}-\x{2dbe}\x{2dc0}-\x{2dc6}\x{2dc8}-\x{2dce}\x{2dd0}-\x{2dd6}\x{2dd8}-\x{2dde}\x{3005}-\x{3007}\x{3021}-\x{3029}\x{302e}\x{302f}\x{3031}-\x{3035}\x{3038}-\x{303c}\x{3041}-\x{3096}\x{309d}-\x{309f}\x{30a1}-\x{30fa}\x{30fc}-\x{30ff}\x{3105}-\x{312d}\x{3131}-\x{318e}\x{3190}-\x{31ba}\x{31f0}-\x{321c}\x{3220}-\x{324f}\x{3260}-\x{327b}\x{327f}-\x{32b0}\x{32c0}-\x{32cb}\x{32d0}-\x{32fe}\x{3300}-\x{3376}\x{337b}-\x{33dd}\x{33e0}-\x{33fe}\x{3400}-\x{4db5}\x{4e00}-\x{9fd5}\x{a000}-\x{a48c}\x{a4d0}-\x{a60c}\x{a610}-\x{a62b}\x{a640}-\x{a66e}\x{a680}-\x{a69d}\x{a6a0}-\x{a6ef}\x{a6f2}-\x{a6f7}\x{a722}-\x{a787}\x{a789}-\x{a7ad}\x{a7b0}-\x{a7b7}\x{a7f7}-\x{a801}\x{a803}-\x{a805}\x{a807}-\x{a80a}\x{a80c}-\x{a824}\x{a827}\x{a830}-\x{a837}\x{a840}-\x{a873}\x{a880}-\x{a8c3}\x{a8ce}-\x{a8d9}\x{a8f2}-\x{a8fd}\x{a900}-\x{a925}\x{a92e}-\x{a946}\x{a952}\x{a953}\x{a95f}-\x{a97c}\x{a983}-\x{a9b2}\x{a9b4}\x{a9b5}\x{a9ba}\x{a9bb}\x{a9bd}-\x{a9cd}\x{a9cf}-\x{a9d9}\x{a9de}-\x{a9e4}\x{a9e6}-\x{a9fe}\x{aa00}-\x{aa28}\x{aa2f}\x{aa30}\x{aa33}\x{aa34}\x{aa40}-\x{aa42}\x{aa44}-\x{aa4b}\x{aa4d}\x{aa50}-\x{aa59}\x{aa5c}-\x{aa7b}\x{aa7d}-\x{aaaf}\x{aab1}\x{aab5}\x{aab6}\x{aab9}-\x{aabd}\x{aac0}\x{aac2}\x{aadb}-\x{aaeb}\x{aaee}-\x{aaf5}\x{ab01}-\x{ab06}\x{ab09}-\x{ab0e}\x{ab11}-\x{ab16}\x{ab20}-\x{ab26}\x{ab28}-\x{ab2e}\x{ab30}-\x{ab65}\x{ab70}-\x{abe4}\x{abe6}\x{abe7}\x{abe9}-\x{abec}\x{abf0}-\x{abf9}\x{ac00}-\x{d7a3}\x{d7b0}-\x{d7c6}\x{d7cb}-\x{d7fb}\x{e000}-\x{fa6d}\x{fa70}-\x{fad9}\x{fb00}-\x{fb06}\x{fb13}-\x{fb17}\x{ff21}-\x{ff3a}\x{ff41}-\x{ff5a}\x{ff66}-\x{ffbe}\x{ffc2}-\x{ffc7}\x{ffca}-\x{ffcf}\x{ffd2}-\x{ffd7}\x{ffda}-\x{ffdc}\x{10000}-\x{1000b}\x{1000d}-\x{10026}\x{10028}-\x{1003a}\x{1003c}\x{1003d}\x{1003f}-\x{1004d}\x{10050}-\x{1005d}\x{10080}-\x{100fa}\x{10100}\x{10102}\x{10107}-\x{10133}\x{10137}-\x{1013f}\x{101d0}-\x{101fc}\x{10280}-\x{1029c}\x{102a0}-\x{102d0}\x{10300}-\x{10323}\x{10330}-\x{1034a}\x{10350}-\x{10375}\x{10380}-\x{1039d}\x{1039f}-\x{103c3}\x{103c8}-\x{103d5}\x{10400}-\x{1049d}\x{104a0}-\x{104a9}\x{10500}-\x{10527}\x{10530}-\x{10563}\x{1056f}\x{10600}-\x{10736}\x{10740}-\x{10755}\x{10760}-\x{10767}\x{11000}\x{11002}-\x{11037}\x{11047}-\x{1104d}\x{11066}-\x{1106f}\x{11082}-\x{110b2}\x{110b7}\x{110b8}\x{110bb}-\x{110c1}\x{110d0}-\x{110e8}\x{110f0}-\x{110f9}\x{11103}-\x{11126}\x{1112c}\x{11136}-\x{11143}\x{11150}-\x{11172}\x{11174}-\x{11176}\x{11182}-\x{111b5}\x{111bf}-\x{111c9}\x{111cd}\x{111d0}-\x{111df}\x{111e1}-\x{111f4}\x{11200}-\x{11211}\x{11213}-\x{1122e}\x{11232}\x{11233}\x{11235}\x{11238}-\x{1123d}\x{11280}-\x{11286}\x{11288}\x{1128a}-\x{1128d}\x{1128f}-\x{1129d}\x{1129f}-\x{112a9}\x{112b0}-\x{112de}\x{112e0}-\x{112e2}\x{112f0}-\x{112f9}\x{11302}\x{11303}\x{11305}-\x{1130c}\x{1130f}\x{11310}\x{11313}-\x{11328}\x{1132a}-\x{11330}\x{11332}\x{11333}\x{11335}-\x{11339}\x{1133d}-\x{1133f}\x{11341}-\x{11344}\x{11347}\x{11348}\x{1134b}-\x{1134d}\x{11350}\x{11357}\x{1135d}-\x{11363}\x{11480}-\x{114b2}\x{114b9}\x{114bb}-\x{114be}\x{114c1}\x{114c4}-\x{114c7}\x{114d0}-\x{114d9}\x{11580}-\x{115b1}\x{115b8}-\x{115bb}\x{115be}\x{115c1}-\x{115db}\x{11600}-\x{11632}\x{1163b}\x{1163c}\x{1163e}\x{11641}-\x{11644}\x{11650}-\x{11659}\x{11680}-\x{116aa}\x{116ac}\x{116ae}\x{116af}\x{116b6}\x{116c0}-\x{116c9}\x{11700}-\x{11719}\x{11720}\x{11721}\x{11726}\x{11730}-\x{1173f}\x{118a0}-\x{118f2}\x{118ff}\x{11ac0}-\x{11af8}\x{12000}-\x{12399}\x{12400}-\x{1246e}\x{12470}-\x{12474}\x{12480}-\x{12543}\x{13000}-\x{1342e}\x{14400}-\x{14646}\x{16800}-\x{16a38}\x{16a40}-\x{16a5e}\x{16a60}-\x{16a69}\x{16a6e}\x{16a6f}\x{16ad0}-\x{16aed}\x{16af5}\x{16b00}-\x{16b2f}\x{16b37}-\x{16b45}\x{16b50}-\x{16b59}\x{16b5b}-\x{16b61}\x{16b63}-\x{16b77}\x{16b7d}-\x{16b8f}\x{16f00}-\x{16f44}\x{16f50}-\x{16f7e}\x{16f93}-\x{16f9f}\x{1b000}\x{1b001}\x{1bc00}-\x{1bc6a}\x{1bc70}-\x{1bc7c}\x{1bc80}-\x{1bc88}\x{1bc90}-\x{1bc99}\x{1bc9c}\x{1bc9f}\x{1d000}-\x{1d0f5}\x{1d100}-\x{1d126}\x{1d129}-\x{1d166}\x{1d16a}-\x{1d172}\x{1d183}\x{1d184}\x{1d18c}-\x{1d1a9}\x{1d1ae}-\x{1d1e8}\x{1d360}-\x{1d371}\x{1d400}-\x{1d454}\x{1d456}-\x{1d49c}\x{1d49e}\x{1d49f}\x{1d4a2}\x{1d4a5}\x{1d4a6}\x{1d4a9}-\x{1d4ac}\x{1d4ae}-\x{1d4b9}\x{1d4bb}\x{1d4bd}-\x{1d4c3}\x{1d4c5}-\x{1d505}\x{1d507}-\x{1d50a}\x{1d50d}-\x{1d514}\x{1d516}-\x{1d51c}\x{1d51e}-\x{1d539}\x{1d53b}-\x{1d53e}\x{1d540}-\x{1d544}\x{1d546}\x{1d54a}-\x{1d550}\x{1d552}-\x{1d6a5}\x{1d6a8}-\x{1d6da}\x{1d6dc}-\x{1d714}\x{1d716}-\x{1d74e}\x{1d750}-\x{1d788}\x{1d78a}-\x{1d7c2}\x{1d7c4}-\x{1d7cb}\x{1d800}-\x{1d9ff}\x{1da37}-\x{1da3a}\x{1da6d}-\x{1da74}\x{1da76}-\x{1da83}\x{1da85}-\x{1da8b}\x{1f110}-\x{1f12e}\x{1f130}-\x{1f169}\x{1f170}-\x{1f19a}\x{1f1e6}-\x{1f202}\x{1f210}-\x{1f23a}\x{1f240}-\x{1f248}\x{1f250}\x{1f251}\x{20000}-\x{2a6d6}\x{2a700}-\x{2b734}\x{2b740}-\x{2b81d}\x{2b820}-\x{2cea1}\x{2f800}-\x{2fa1d}\x{f0000}-\x{ffffd}\x{100000}-\x{10fffd}])|([\x{590}\x{5be}\x{5c0}\x{5c3}\x{5c6}\x{5c8}-\x{5ff}\x{7c0}-\x{7ea}\x{7f4}\x{7f5}\x{7fa}-\x{815}\x{81a}\x{824}\x{828}\x{82e}-\x{858}\x{85c}-\x{89f}\x{200f}\x{fb1d}\x{fb1f}-\x{fb28}\x{fb2a}-\x{fb4f}\x{10800}-\x{1091e}\x{10920}-\x{10a00}\x{10a04}\x{10a07}-\x{10a0b}\x{10a10}-\x{10a37}\x{10a3b}-\x{10a3e}\x{10a40}-\x{10ae4}\x{10ae7}-\x{10b38}\x{10b40}-\x{10e5f}\x{10e7f}-\x{10fff}\x{1e800}-\x{1e8cf}\x{1e8d7}-\x{1edff}\x{1ef00}-\x{1efff}\x{608}\x{60b}\x{60d}\x{61b}-\x{64a}\x{66d}-\x{66f}\x{671}-\x{6d5}\x{6e5}\x{6e6}\x{6ee}\x{6ef}\x{6fa}-\x{710}\x{712}-\x{72f}\x{74b}-\x{7a5}\x{7b1}-\x{7bf}\x{8a0}-\x{8e2}\x{fb50}-\x{fd3d}\x{fd40}-\x{fdcf}\x{fdf0}-\x{fdfc}\x{fdfe}\x{fdff}\x{fe70}-\x{fefe}\x{1ee00}-\x{1eeef}\x{1eef2}-\x{1eeff}]))/u';
210 // @codeCoverageIgnoreEnd
213 * Get a cached or new language object for a given language code
214 * @param string $code
215 * @throws MWException
218 static function factory( $code ) {
219 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
221 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
222 $code = $wgDummyLanguageCodes[$code];
225 // get the language object to process
226 $langObj = self
::$mLangObjCache[$code] ?? self
::newFromCode( $code );
228 // merge the language object in to get it up front in the cache
229 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
230 // get rid of the oldest ones in case we have an overflow
231 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
237 * Create a language object for a given language code
238 * @param string $code
239 * @param bool $fallback Whether we're going through language fallback chain
240 * @throws MWException
243 protected static function newFromCode( $code, $fallback = false ) {
244 if ( !self
::isValidCode( $code ) ) {
245 throw new MWException( "Invalid language code \"$code\"" );
248 if ( !self
::isValidBuiltInCode( $code ) ) {
249 // It's not possible to customise this code with class files, so
250 // just return a Language object. This is to support uselang= hacks.
251 $lang = new Language
;
252 $lang->mCode
= $code;
256 // Check if there is a language class for the code
257 $class = self
::classFromCode( $code, $fallback );
258 // LanguageCode does not inherit Language
259 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
264 // Keep trying the fallback list until we find an existing class
265 $fallbacks = self
::getFallbacksFor( $code );
266 foreach ( $fallbacks as $fallbackCode ) {
267 if ( !self
::isValidBuiltInCode( $fallbackCode ) ) {
268 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
271 $class = self
::classFromCode( $fallbackCode );
272 if ( class_exists( $class ) ) {
274 $lang->mCode
= $code;
279 throw new MWException( "Invalid fallback sequence for language '$code'" );
283 * Intended for tests that may change configuration in a way that invalidates caches.
287 public static function clearCaches() {
288 if ( !defined( 'MW_PHPUNIT_TEST' ) ) {
289 throw new MWException( __METHOD__
. ' must not be used outside tests' );
291 self
::$dataCache = null;
292 // Reinitialize $dataCache, since it's expected to always be available
293 self
::getLocalisationCache();
294 self
::$mLangObjCache = [];
295 self
::$fallbackLanguageCache = [];
296 self
::$grammarTransformations = null;
297 self
::$languageNameCache = null;
301 * Checks whether any localisation is available for that language tag
302 * in MediaWiki (MessagesXx.php exists).
304 * @param string $code Language tag (in lower case)
305 * @return bool Whether language is supported
308 public static function isSupportedLanguage( $code ) {
309 if ( !self
::isValidBuiltInCode( $code ) ) {
313 if ( $code === 'qqq' ) {
317 return is_readable( self
::getMessagesFileName( $code ) ) ||
318 is_readable( self
::getJsonMessagesFileName( $code ) );
322 * Returns true if a language code string is a well-formed language tag
323 * according to RFC 5646.
324 * This function only checks well-formedness; it doesn't check that
325 * language, script or variant codes actually exist in the repositories.
327 * Based on regexes by Mark Davis of the Unicode Consortium:
328 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
330 * @param string $code
331 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
336 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
339 $alphanum = '[a-z0-9]';
340 $x = 'x'; # private use singleton
341 $singleton = '[a-wy-z]'; # other singleton
342 $s = $lenient ?
'[-_]' : '-';
344 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
345 $script = "$alpha{4}"; # ISO 15924
346 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
347 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
348 $extension = "$singleton(?:$s$alphanum{2,8})+";
349 $privateUse = "$x(?:$s$alphanum{1,8})+";
351 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
352 # Since these are limited, this is safe even later changes to the registry --
353 # the only oddity is that it might change the type of the tag, and thus
354 # the results from the capturing groups.
355 # https://www.iana.org/assignments/language-subtag-registry
357 $grandfathered = "en{$s}GB{$s}oed"
358 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
359 . "|no{$s}(?:bok|nyn)"
360 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
361 . "|zh{$s}min{$s}nan";
363 $variantList = "$variant(?:$s$variant)*";
364 $extensionList = "$extension(?:$s$extension)*";
366 $langtag = "(?:($language)"
369 . "(?:$s$variantList)?"
370 . "(?:$s$extensionList)?"
371 . "(?:$s$privateUse)?)";
373 # The final breakdown, with capturing groups for each of these components
374 # The variants, extensions, grandfathered, and private-use may have interior '-'
376 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
378 return (bool)preg_match( "/$root/", strtolower( $code ) );
382 * Returns true if a language code string is of a valid form, whether or
383 * not it exists. This includes codes which are used solely for
384 * customisation via the MediaWiki namespace.
386 * @param string $code
390 public static function isValidCode( $code ) {
392 Assert
::parameterType( 'string', $code, '$code' );
393 if ( !isset( $cache[$code] ) ) {
394 // People think language codes are html safe, so enforce it.
395 // Ideally we should only allow a-zA-Z0-9-
396 // but, .+ and other chars are often used for {{int:}} hacks
397 // see bugs T39564, T39587, T38938
399 // Protect against path traversal
400 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
401 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
403 return $cache[$code];
407 * Returns true if a language code is of a valid form for the purposes of
408 * internal customisation of MediaWiki, via Messages*.php or *.json.
410 * @param string $code
415 public static function isValidBuiltInCode( $code ) {
416 Assert
::parameterType( 'string', $code, '$code' );
418 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
422 * Returns true if a language code is an IETF tag known to MediaWiki.
429 public static function isKnownLanguageTag( $tag ) {
430 // Quick escape for invalid input to avoid exceptions down the line
431 // when code tries to process tags which are not valid at all.
432 if ( !self
::isValidBuiltInCode( $tag ) ) {
436 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
437 || self
::fetchLanguageName( $tag, $tag ) !== ''
446 * Get the LocalisationCache instance
448 * @return LocalisationCache
450 public static function getLocalisationCache() {
451 if ( is_null( self
::$dataCache ) ) {
452 global $wgLocalisationCacheConf;
453 $class = $wgLocalisationCacheConf['class'];
454 self
::$dataCache = new $class( $wgLocalisationCacheConf );
456 return self
::$dataCache;
459 function __construct() {
460 $this->mConverter
= new FakeConverter( $this );
461 // Set the code to the name of the descendant
462 if ( static::class === 'Language' ) {
465 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
467 self
::getLocalisationCache();
471 * Reduce memory usage
472 * @suppress PhanTypeSuspiciousNonTraversableForeach
474 function __destruct() {
475 foreach ( $this as $name => $value ) {
476 unset( $this->$name );
481 * Hook which will be called if this is the content language.
482 * Descendants can use this to register hook functions or modify globals
484 function initContLang() {
491 public function getFallbackLanguages() {
492 return self
::getFallbacksFor( $this->mCode
);
496 * Exports $wgBookstoreListEn
499 public function getBookstoreList() {
500 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
504 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
505 * available in localised form, it will be included in English.
507 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
509 public function getNamespaces() {
510 if ( is_null( $this->namespaceNames
) ) {
511 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
513 $validNamespaces = MediaWikiServices
::getInstance()->getNamespaceInfo()->
514 getCanonicalNamespaces();
516 $this->namespaceNames
= $wgExtraNamespaces +
517 self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
518 $this->namespaceNames +
= $validNamespaces;
520 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
521 if ( $wgMetaNamespaceTalk ) {
522 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
524 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
525 $this->namespaceNames
[NS_PROJECT_TALK
] =
526 $this->fixVariableInNamespace( $talk );
529 # Sometimes a language will be localised but not actually exist on this wiki.
530 foreach ( $this->namespaceNames
as $key => $text ) {
531 if ( !isset( $validNamespaces[$key] ) ) {
532 unset( $this->namespaceNames
[$key] );
536 # The above mixing may leave namespaces out of canonical order.
537 # Re-order by namespace ID number...
538 ksort( $this->namespaceNames
);
540 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
543 return $this->namespaceNames
;
547 * Arbitrarily set all of the namespace names at once. Mainly used for testing
548 * @param array $namespaces Array of namespaces (id => name)
550 public function setNamespaces( array $namespaces ) {
551 $this->namespaceNames
= $namespaces;
552 $this->mNamespaceIds
= null;
556 * Resets all of the namespace caches. Mainly used for testing
558 public function resetNamespaces() {
559 $this->namespaceNames
= null;
560 $this->mNamespaceIds
= null;
561 $this->namespaceAliases
= null;
565 * A convenience function that returns getNamespaces() with spaces instead of underscores
566 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
570 public function getFormattedNamespaces() {
571 $ns = $this->getNamespaces();
572 foreach ( $ns as $k => $v ) {
573 $ns[$k] = strtr( $v, '_', ' ' );
579 * Get a namespace value by key
582 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
583 * echo $mw_ns; // prints 'MediaWiki'
586 * @param int $index The array key of the namespace to return
587 * @return string|bool String if the namespace value exists, otherwise false
589 public function getNsText( $index ) {
590 $ns = $this->getNamespaces();
591 return $ns[$index] ??
false;
595 * A convenience function that returns the same thing as
596 * getNsText() except with '_' changed to ' ', useful for
600 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
601 * echo $mw_ns; // prints 'MediaWiki talk'
604 * @param int $index The array key of the namespace to return
605 * @return string Namespace name without underscores (empty string if namespace does not exist)
607 public function getFormattedNsText( $index ) {
608 $ns = $this->getNsText( $index );
609 return strtr( $ns, '_', ' ' );
613 * Returns gender-dependent namespace alias if available.
614 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
615 * @param int $index Namespace index
616 * @param string $gender Gender key (male, female... )
620 public function getGenderNsText( $index, $gender ) {
621 global $wgExtraGenderNamespaces;
623 $ns = $wgExtraGenderNamespaces +
624 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
626 return $ns[$index][$gender] ??
$this->getNsText( $index );
630 * Whether this language uses gender-dependent namespace aliases.
631 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
635 public function needsGenderDistinction() {
636 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
637 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
638 // $wgExtraGenderNamespaces overrides everything
640 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
641 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
642 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
645 // Check what is in i18n files
646 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
647 return count( $aliases ) > 0;
652 * Get a namespace key by value, case insensitive.
653 * Only matches namespace names for the current language, not the
654 * canonical ones defined in Namespace.php.
656 * @param string $text
657 * @return int|bool An integer if $text is a valid value otherwise false
659 function getLocalNsIndex( $text ) {
660 $lctext = $this->lc( $text );
661 $ids = $this->getNamespaceIds();
662 return $ids[$lctext] ??
false;
668 public function getNamespaceAliases() {
669 if ( is_null( $this->namespaceAliases
) ) {
670 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
674 foreach ( $aliases as $name => $index ) {
675 if ( $index === NS_PROJECT_TALK
) {
676 unset( $aliases[$name] );
677 $name = $this->fixVariableInNamespace( $name );
678 $aliases[$name] = $index;
683 global $wgExtraGenderNamespaces;
684 $genders = $wgExtraGenderNamespaces +
685 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
686 foreach ( $genders as $index => $forms ) {
687 foreach ( $forms as $alias ) {
688 $aliases[$alias] = $index;
692 # Also add converted namespace names as aliases, to avoid confusion.
693 $convertedNames = [];
694 foreach ( $this->getVariants() as $variant ) {
695 if ( $variant === $this->mCode
) {
698 foreach ( $this->getNamespaces() as $ns => $_ ) {
699 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
703 $this->namespaceAliases
= $aliases +
$convertedNames;
705 # Filter out aliases to namespaces that don't exist, e.g. from extensions
706 # that aren't loaded here but are included in the l10n cache.
707 # (array_intersect preserves keys from its first argument)
708 $this->namespaceAliases
= array_intersect(
709 $this->namespaceAliases
,
710 array_keys( $this->getNamespaces() )
714 return $this->namespaceAliases
;
720 public function getNamespaceIds() {
721 if ( is_null( $this->mNamespaceIds
) ) {
722 global $wgNamespaceAliases;
723 # Put namespace names and aliases into a hashtable.
724 # If this is too slow, then we should arrange it so that it is done
725 # before caching. The catch is that at pre-cache time, the above
726 # class-specific fixup hasn't been done.
727 $this->mNamespaceIds
= [];
728 foreach ( $this->getNamespaces() as $index => $name ) {
729 $this->mNamespaceIds
[$this->lc( $name )] = $index;
731 foreach ( $this->getNamespaceAliases() as $name => $index ) {
732 $this->mNamespaceIds
[$this->lc( $name )] = $index;
734 if ( $wgNamespaceAliases ) {
735 foreach ( $wgNamespaceAliases as $name => $index ) {
736 $this->mNamespaceIds
[$this->lc( $name )] = $index;
740 return $this->mNamespaceIds
;
744 * Get a namespace key by value, case insensitive. Canonical namespace
745 * names override custom ones defined for the current language.
747 * @param string $text
748 * @return int|bool An integer if $text is a valid value otherwise false
750 public function getNsIndex( $text ) {
751 $lctext = $this->lc( $text );
752 $ns = MediaWikiServices
::getInstance()->getNamespaceInfo()->
753 getCanonicalIndex( $lctext );
754 if ( $ns !== null ) {
757 $ids = $this->getNamespaceIds();
758 return $ids[$lctext] ??
false;
762 * short names for language variants used for language conversion links.
764 * @param string $code
765 * @param bool $usemsg Use the "variantname-xyz" message if it exists
768 public function getVariantname( $code, $usemsg = true ) {
769 $msg = "variantname-$code";
770 if ( $usemsg && wfMessage( $msg )->exists() ) {
771 return $this->getMessageFromDB( $msg );
773 $name = self
::fetchLanguageName( $code );
775 return $name; # if it's defined as a language name, show that
777 # otherwise, output the language code
783 * @return string[]|bool List of date format preference keys, or false if disabled.
785 public function getDatePreferences() {
786 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
792 function getDateFormats() {
793 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
797 * @return array|string
799 public function getDefaultDateFormat() {
800 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
801 if ( $df === 'dmy or mdy' ) {
802 global $wgAmericanDates;
803 return $wgAmericanDates ?
'mdy' : 'dmy';
812 public function getDatePreferenceMigrationMap() {
813 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
819 public function getExtraUserToggles() {
820 wfDeprecated( __METHOD__
, '1.34' );
828 function getUserToggle( $tog ) {
829 return $this->getMessageFromDB( "tog-$tog" );
833 * Get an array of language names, indexed by code.
834 * @param null|string $inLanguage Code of language in which to return the names
835 * Use self::AS_AUTONYMS for autonyms (native names)
836 * @param string $include One of:
837 * self::ALL all available languages
838 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
839 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
840 * @return array Language code => language name (sorted by key)
843 public static function fetchLanguageNames( $inLanguage = self
::AS_AUTONYMS
, $include = 'mw' ) {
844 $cacheKey = $inLanguage === self
::AS_AUTONYMS ?
'null' : $inLanguage;
845 $cacheKey .= ":$include";
846 if ( self
::$languageNameCache === null ) {
847 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
850 $ret = self
::$languageNameCache->get( $cacheKey );
852 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
853 self
::$languageNameCache->set( $cacheKey, $ret );
859 * Uncached helper for fetchLanguageNames
860 * @param null|string $inLanguage Code of language in which to return the names
861 * Use self::AS_AUTONYMS for autonyms (native names)
862 * @param string $include One of:
863 * self::ALL all available languages
864 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
865 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
866 * @return array Language code => language name (sorted by key)
868 private static function fetchLanguageNamesUncached(
869 $inLanguage = self
::AS_AUTONYMS
,
872 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
874 // If passed an invalid language code to use, fallback to en
875 if ( $inLanguage !== self
::AS_AUTONYMS
&& !self
::isValidCode( $inLanguage ) ) {
882 # TODO: also include when $inLanguage is null, when this code is more efficient
883 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
886 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
887 if ( $wgUsePigLatinVariant ) {
888 // Pig Latin (for variant development)
889 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
892 foreach ( $mwNames as $mwCode => $mwName ) {
893 # - Prefer own MediaWiki native name when not using the hook
894 # - For other names just add if not added through the hook
895 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
896 $names[$mwCode] = $mwName;
900 if ( $include === self
::ALL
) {
906 $coreCodes = array_keys( $mwNames );
907 foreach ( $coreCodes as $coreCode ) {
908 $returnMw[$coreCode] = $names[$coreCode];
911 if ( $include === self
::SUPPORTED
) {
913 # We do this using a foreach over the codes instead of a directory
914 # loop so that messages files in extensions will work correctly.
915 foreach ( $returnMw as $code => $value ) {
916 if ( is_readable( self
::getMessagesFileName( $code ) )
917 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
919 $namesMwFile[$code] = $names[$code];
923 ksort( $namesMwFile );
928 # 'mw' option; default if it's not one of the other two options (all/mwfile)
933 * @param string $code The code of the language for which to get the name
934 * @param null|string $inLanguage Code of language in which to return the name
935 * (SELF::AS_AUTONYMS for autonyms)
936 * @param string $include See fetchLanguageNames()
937 * @return string Language name or empty
940 public static function fetchLanguageName(
942 $inLanguage = self
::AS_AUTONYMS
,
945 $code = strtolower( $code );
946 $array = self
::fetchLanguageNames( $inLanguage, $include );
947 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
951 * Get a message from the MediaWiki namespace.
953 * @param string $msg Message name
956 public function getMessageFromDB( $msg ) {
957 return $this->msg( $msg )->text();
961 * Get message object in this language. Only for use inside this class.
963 * @param string $msg Message name
966 protected function msg( $msg ) {
967 return wfMessage( $msg )->inLanguage( $this );
974 public function getMonthName( $key ) {
975 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
981 public function getMonthNamesArray() {
982 $monthNames = [ '' ];
983 for ( $i = 1; $i < 13; $i++
) {
984 $monthNames[] = $this->getMonthName( $i );
993 public function getMonthNameGen( $key ) {
994 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1001 public function getMonthAbbreviation( $key ) {
1002 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1008 public function getMonthAbbreviationsArray() {
1009 $monthNames = [ '' ];
1010 for ( $i = 1; $i < 13; $i++
) {
1011 $monthNames[] = $this->getMonthAbbreviation( $i );
1017 * @param string $key
1020 public function getWeekdayName( $key ) {
1021 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1025 * @param string $key
1028 function getWeekdayAbbreviation( $key ) {
1029 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1033 * @param string $key
1036 function getIranianCalendarMonthName( $key ) {
1037 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1041 * @param string $key
1044 function getHebrewCalendarMonthName( $key ) {
1045 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1049 * @param string $key
1052 function getHebrewCalendarMonthNameGen( $key ) {
1053 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1057 * @param string $key
1060 function getHijriCalendarMonthName( $key ) {
1061 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1065 * Pass through result from $dateTimeObj->format()
1066 * @param DateTime|bool|null &$dateTimeObj
1068 * @param DateTimeZone|bool|null $zone
1069 * @param string $code
1072 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1073 if ( !$dateTimeObj ) {
1074 $dateTimeObj = DateTime
::createFromFormat(
1075 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1078 return $dateTimeObj->format( $code );
1082 * This is a workalike of PHP's date() function, but with better
1083 * internationalisation, a reduced set of format characters, and a better
1086 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1087 * the PHP manual for definitions. There are a number of extensions, which
1090 * xn Do not translate digits of the next numeric format character
1091 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1092 * xr Use roman numerals for the next numeric format character
1093 * xh Use hebrew numerals for the next numeric format character
1095 * xg Genitive month name
1097 * xij j (day number) in Iranian calendar
1098 * xiF F (month name) in Iranian calendar
1099 * xin n (month number) in Iranian calendar
1100 * xiy y (two digit year) in Iranian calendar
1101 * xiY Y (full year) in Iranian calendar
1102 * xit t (days in month) in Iranian calendar
1103 * xiz z (day of the year) in Iranian calendar
1105 * xjj j (day number) in Hebrew calendar
1106 * xjF F (month name) in Hebrew calendar
1107 * xjt t (days in month) in Hebrew calendar
1108 * xjx xg (genitive month name) in Hebrew calendar
1109 * xjn n (month number) in Hebrew calendar
1110 * xjY Y (full year) in Hebrew calendar
1112 * xmj j (day number) in Hijri calendar
1113 * xmF F (month name) in Hijri calendar
1114 * xmn n (month number) in Hijri calendar
1115 * xmY Y (full year) in Hijri calendar
1117 * xkY Y (full year) in Thai solar calendar. Months and days are
1118 * identical to the Gregorian calendar
1119 * xoY Y (full year) in Minguo calendar or Juche year.
1120 * Months and days are identical to the
1121 * Gregorian calendar
1122 * xtY Y (full year) in Japanese nengo. Months and days are
1123 * identical to the Gregorian calendar
1125 * Characters enclosed in double quotes will be considered literal (with
1126 * the quotes themselves removed). Unmatched quotes will be considered
1127 * literal quotes. Example:
1129 * "The month is" F => The month is January
1132 * Backslash escaping is also supported.
1134 * Input timestamp is assumed to be pre-normalized to the desired local
1135 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1136 * $ts is UTC if $zone is not given.
1138 * @param string $format
1139 * @param string $ts 14-character timestamp
1142 * @param DateTimeZone|null $zone Timezone of $ts
1143 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1144 * Only makes sense if $ts is the current time.
1145 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1147 * @throws MWException
1150 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1155 $dateTimeObj = false;
1164 $usedSecond = false;
1165 $usedMinute = false;
1172 $usedISOYear = false;
1173 $usedIsLeapYear = false;
1175 $usedHebrewMonth = false;
1176 $usedIranianMonth = false;
1177 $usedHijriMonth = false;
1178 $usedHebrewYear = false;
1179 $usedIranianYear = false;
1180 $usedHijriYear = false;
1181 $usedTennoYear = false;
1183 if ( strlen( $ts ) !== 14 ) {
1184 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1187 if ( !ctype_digit( $ts ) ) {
1188 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1191 $formatLength = strlen( $format );
1192 for ( $p = 0; $p < $formatLength; $p++
) {
1194 $code = $format[$p];
1195 if ( $code == 'x' && $p < $formatLength - 1 ) {
1196 $code .= $format[++
$p];
1199 if ( ( $code === 'xi'
1205 && $p < $formatLength - 1 ) {
1206 $code .= $format[++
$p];
1217 $rawToggle = !$rawToggle;
1227 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1230 $usedHebrewMonth = true;
1232 $hebrew = self
::tsToHebrew( $ts );
1234 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1238 $num = substr( $ts, 6, 2 );
1242 $s .= $this->getWeekdayAbbreviation(
1243 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1248 $num = intval( substr( $ts, 6, 2 ) );
1253 $iranian = self
::tsToIranian( $ts );
1260 $hijri = self
::tsToHijri( $ts );
1267 $hebrew = self
::tsToHebrew( $ts );
1273 $s .= $this->getWeekdayName(
1274 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1279 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1282 $usedIranianMonth = true;
1284 $iranian = self
::tsToIranian( $ts );
1286 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1289 $usedHijriMonth = true;
1291 $hijri = self
::tsToHijri( $ts );
1293 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1296 $usedHebrewMonth = true;
1298 $hebrew = self
::tsToHebrew( $ts );
1300 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1304 $num = substr( $ts, 4, 2 );
1308 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1312 $num = intval( substr( $ts, 4, 2 ) );
1315 $usedIranianMonth = true;
1317 $iranian = self
::tsToIranian( $ts );
1322 $usedHijriMonth = true;
1324 $hijri = self
::tsToHijri( $ts );
1329 $usedHebrewMonth = true;
1331 $hebrew = self
::tsToHebrew( $ts );
1336 $usedHebrewMonth = true;
1338 $hebrew = self
::tsToHebrew( $ts );
1344 $num = substr( $ts, 0, 4 );
1347 $usedIranianYear = true;
1349 $iranian = self
::tsToIranian( $ts );
1354 $usedHijriYear = true;
1356 $hijri = self
::tsToHijri( $ts );
1361 $usedHebrewYear = true;
1363 $hebrew = self
::tsToHebrew( $ts );
1370 $thai = self
::tsToYear( $ts, 'thai' );
1377 $minguo = self
::tsToYear( $ts, 'minguo' );
1382 $usedTennoYear = true;
1384 $tenno = self
::tsToYear( $ts, 'tenno' );
1390 $num = substr( $ts, 2, 2 );
1393 $usedIranianYear = true;
1395 $iranian = self
::tsToIranian( $ts );
1397 $num = substr( $iranian[0], -2 );
1400 $usedIranianYear = true;
1402 $iranian = self
::tsToIranian( $ts );
1404 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1407 $usedIranianYear = true;
1409 $iranian = self
::tsToIranian( $ts );
1415 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1419 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1423 $h = substr( $ts, 8, 2 );
1424 $num = $h %
12 ?
: 12;
1428 $num = intval( substr( $ts, 8, 2 ) );
1432 $h = substr( $ts, 8, 2 );
1433 $num = sprintf( '%02d', $h %
12 ?
: 12 );
1437 $num = substr( $ts, 8, 2 );
1441 $num = substr( $ts, 10, 2 );
1445 $num = substr( $ts, 12, 2 );
1455 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1461 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1465 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1469 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1472 $usedIsLeapYear = true;
1473 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1476 $usedISOYear = true;
1477 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1484 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1487 # Backslash escaping
1488 if ( $p < $formatLength - 1 ) {
1489 $s .= $format[++
$p];
1496 if ( $p < $formatLength - 1 ) {
1497 $endQuote = strpos( $format, '"', $p +
1 );
1498 if ( $endQuote === false ) {
1499 # No terminating quote, assume literal "
1502 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1506 # Quote at end of string, assume literal "
1513 if ( $num !== false ) {
1514 if ( $rawToggle ||
$raw ) {
1517 } elseif ( $roman ) {
1518 $s .= self
::romanNumeral( $num );
1520 } elseif ( $hebrewNum ) {
1521 $s .= self
::hebrewNumeral( $num );
1524 $s .= $this->formatNum( $num, true );
1529 if ( $ttl === 'unused' ) {
1530 // No need to calculate the TTL, the caller wont use it anyway.
1531 } elseif ( $usedSecond ) {
1533 } elseif ( $usedMinute ) {
1534 $ttl = 60 - substr( $ts, 12, 2 );
1535 } elseif ( $usedHour ) {
1536 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1537 } elseif ( $usedAMPM ) {
1538 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1539 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1543 $usedIranianMonth ||
1550 // @todo Someone who understands the non-Gregorian calendars
1551 // should write proper logic for them so that they don't need purged every day.
1552 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1553 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1556 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1557 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1560 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1561 $timeRemainingInDay;
1562 } elseif ( $usedISOYear ) {
1563 // December 28th falls on the last ISO week of the year, every year.
1564 // The last ISO week of a year can be 52 or 53.
1565 $lastWeekOfISOYear = DateTime
::createFromFormat(
1567 substr( $ts, 0, 4 ) . '1228',
1568 $zone ?
: new DateTimeZone( 'UTC' )
1570 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1571 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1572 $timeRemainingInWeek =
1573 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1574 +
$timeRemainingInDay;
1575 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1580 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1581 substr( $ts, 6, 2 ) ) * 86400
1582 +
$timeRemainingInDay;
1583 } elseif ( $usedYear ) {
1585 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1586 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1587 +
$timeRemainingInDay;
1588 } elseif ( $usedIsLeapYear ) {
1589 $year = substr( $ts, 0, 4 );
1590 $timeRemainingInYear =
1591 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1592 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1593 +
$timeRemainingInDay;
1595 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1596 // this isn't a leap year. see when the next one starts
1597 $nextCandidate = $year - $mod +
4;
1598 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1599 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1600 $timeRemainingInYear;
1602 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1603 $timeRemainingInYear;
1606 // this is a leap year, so the next year isn't
1607 $possibleTtls[] = $timeRemainingInYear;
1611 if ( $possibleTtls ) {
1612 $ttl = min( $possibleTtls );
1619 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1620 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1623 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1624 * Gregorian dates to Iranian dates. Originally written in C, it
1625 * is released under the terms of GNU Lesser General Public
1626 * License. Conversion to PHP was performed by Niklas Laxström.
1628 * Link: http://www.farsiweb.info/jalali/jalali.c
1634 private static function tsToIranian( $ts ) {
1635 $gy = substr( $ts, 0, 4 ) - 1600;
1636 $gm = substr( $ts, 4, 2 ) - 1;
1637 $gd = substr( $ts, 6, 2 ) - 1;
1639 # Days passed from the beginning (including leap years)
1641 +
floor( ( $gy +
3 ) / 4 )
1642 - floor( ( $gy +
99 ) / 100 )
1643 +
floor( ( $gy +
399 ) / 400 );
1645 // Add days of the past months of this year
1646 for ( $i = 0; $i < $gm; $i++
) {
1647 $gDayNo +
= self
::$GREG_DAYS[$i];
1651 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1655 // Days passed in current month
1656 $gDayNo +
= (int)$gd;
1658 $jDayNo = $gDayNo - 79;
1660 $jNp = floor( $jDayNo / 12053 );
1663 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1666 if ( $jDayNo >= 366 ) {
1667 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1668 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1673 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1674 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1680 return [ $jy, $jm, $jd, $jz ];
1684 * Converting Gregorian dates to Hijri dates.
1686 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1688 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1694 private static function tsToHijri( $ts ) {
1695 $year = substr( $ts, 0, 4 );
1696 $month = substr( $ts, 4, 2 );
1697 $day = substr( $ts, 6, 2 );
1705 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1706 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1708 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1709 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1710 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1713 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1714 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1717 $zl = $zjd - 1948440 +
10632;
1718 $zn = (int)( ( $zl - 1 ) / 10631 );
1719 $zl = $zl - 10631 * $zn +
354;
1720 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1721 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1722 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1723 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1724 $zm = (int)( ( 24 * $zl ) / 709 );
1725 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1726 $zy = 30 * $zn +
$zj - 30;
1728 return [ $zy, $zm, $zd ];
1732 * Converting Gregorian dates to Hebrew dates.
1734 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1735 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1736 * to translate the relevant functions into PHP and release them under
1739 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1740 * and Adar II is 14. In a non-leap year, Adar is 6.
1746 private static function tsToHebrew( $ts ) {
1748 $year = substr( $ts, 0, 4 );
1749 $month = substr( $ts, 4, 2 );
1750 $day = substr( $ts, 6, 2 );
1752 # Calculate Hebrew year
1753 $hebrewYear = $year +
3760;
1755 # Month number when September = 1, August = 12
1757 if ( $month > 12 ) {
1764 # Calculate day of year from 1 September
1766 for ( $i = 1; $i < $month; $i++
) {
1770 # Check if the year is leap
1771 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1774 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1781 # Calculate the start of the Hebrew year
1782 $start = self
::hebrewYearStart( $hebrewYear );
1784 # Calculate next year's start
1785 if ( $dayOfYear <= $start ) {
1786 # Day is before the start of the year - it is the previous year
1788 $nextStart = $start;
1792 # Add days since previous year's 1 September
1794 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1798 # Start of the new (previous) year
1799 $start = self
::hebrewYearStart( $hebrewYear );
1802 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1805 # Calculate Hebrew day of year
1806 $hebrewDayOfYear = $dayOfYear - $start;
1808 # Difference between year's days
1809 $diff = $nextStart - $start;
1810 # Add 12 (or 13 for leap years) days to ignore the difference between
1811 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1812 # difference is only about the year type
1813 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1819 # Check the year pattern, and is leap year
1820 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1821 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1822 # and non-leap years
1823 $yearPattern = $diff %
30;
1824 # Check if leap year
1825 $isLeap = $diff >= 30;
1827 # Calculate day in the month from number of day in the Hebrew year
1828 # Don't check Adar - if the day is not in Adar, we will stop before;
1829 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1830 $hebrewDay = $hebrewDayOfYear;
1833 while ( $hebrewMonth <= 12 ) {
1834 # Calculate days in this month
1835 if ( $isLeap && $hebrewMonth == 6 ) {
1836 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1838 if ( $hebrewDay <= $days ) {
1842 # Subtract the days of Adar I
1843 $hebrewDay -= $days;
1846 if ( $hebrewDay <= $days ) {
1851 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1852 # Cheshvan in a complete year (otherwise as the rule below)
1854 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1855 # Kislev in an incomplete year (otherwise as the rule below)
1858 # Odd months have 30 days, even have 29
1859 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1861 if ( $hebrewDay <= $days ) {
1862 # In the current month
1865 # Subtract the days of the current month
1866 $hebrewDay -= $days;
1867 # Try in the next month
1872 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1876 * This calculates the Hebrew year start, as days since 1 September.
1877 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1878 * Used for Hebrew date.
1884 private static function hebrewYearStart( $year ) {
1885 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1886 $b = intval( ( $year - 1 ) %
4 );
1887 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1891 $Mar = intval( $m );
1897 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1898 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1900 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1902 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1906 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1911 * Algorithm to convert Gregorian dates to Thai solar dates,
1912 * Minguo dates or Minguo dates.
1914 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1915 * https://en.wikipedia.org/wiki/Minguo_calendar
1916 * https://en.wikipedia.org/wiki/Japanese_era_name
1918 * @param string $ts 14-character timestamp
1919 * @param string $cName Calender name
1920 * @return array Converted year, month, day
1922 private static function tsToYear( $ts, $cName ) {
1923 $gy = substr( $ts, 0, 4 );
1924 $gm = substr( $ts, 4, 2 );
1925 $gd = substr( $ts, 6, 2 );
1927 if ( !strcmp( $cName, 'thai' ) ) {
1929 # Add 543 years to the Gregorian calendar
1930 # Months and days are identical
1931 $gy_offset = $gy +
543;
1932 # fix for dates between 1912 and 1941
1933 # https://en.wikipedia.org/?oldid=836596673#New_year
1934 if ( $gy >= 1912 && $gy <= 1940 ) {
1938 $gm = ( $gm - 3 ) %
12;
1940 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1942 # Deduct 1911 years from the Gregorian calendar
1943 # Months and days are identical
1944 $gy_offset = $gy - 1911;
1945 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1946 # Nengō dates up to Meiji period
1947 # Deduct years from the Gregorian calendar
1948 # depending on the nengo periods
1949 # Months and days are identical
1951 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1952 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1955 $gy_gannen = $gy - 1868 +
1;
1956 $gy_offset = $gy_gannen;
1957 if ( $gy_gannen == 1 ) {
1960 $gy_offset = '明治' . $gy_offset;
1962 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1963 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1964 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1965 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1966 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1969 $gy_gannen = $gy - 1912 +
1;
1970 $gy_offset = $gy_gannen;
1971 if ( $gy_gannen == 1 ) {
1974 $gy_offset = '大正' . $gy_offset;
1976 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1977 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1978 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1981 $gy_gannen = $gy - 1926 +
1;
1982 $gy_offset = $gy_gannen;
1983 if ( $gy_gannen == 1 ) {
1986 $gy_offset = '昭和' . $gy_offset;
1988 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1989 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1990 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1993 $gy_gannen = $gy - 1989 +
1;
1994 $gy_offset = $gy_gannen;
1995 if ( $gy_gannen == 1 ) {
1998 $gy_offset = '平成' . $gy_offset;
2001 $gy_gannen = $gy - 2019 +
1;
2002 $gy_offset = $gy_gannen;
2003 if ( $gy_gannen == 1 ) {
2006 $gy_offset = '令和' . $gy_offset;
2012 return [ $gy_offset, $gm, $gd ];
2016 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2018 * This is the rule the BIDI algorithm uses to determine the directionality of
2019 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
2020 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2022 * TODO: Does not handle BIDI control characters inside the text.
2023 * TODO: Does not handle unallocated characters.
2025 * @param string $text Text to test
2026 * @return null|string Directionality ('ltr' or 'rtl') or null
2028 private static function strongDirFromContent( $text = '' ) {
2029 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2032 if ( $matches[1] === '' ) {
2039 * Roman number formatting up to 10000
2045 static function romanNumeral( $num ) {
2047 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
2048 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
2049 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
2050 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2051 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2054 $num = intval( $num );
2055 if ( $num > 10000 ||
$num <= 0 ) {
2060 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2061 if ( $num >= $pow10 ) {
2062 $s .= $table[$i][(int)floor( $num / $pow10 )];
2064 $num = $num %
$pow10;
2070 * Hebrew Gematria number formatting up to 9999
2076 static function hebrewNumeral( $num ) {
2078 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2079 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2092 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2095 $num = intval( $num );
2096 if ( $num > 9999 ||
$num <= 0 ) {
2100 // Round thousands have special notations
2101 if ( $num === 1000 ) {
2103 } elseif ( $num %
1000 === 0 ) {
2104 return $table[0][$num / 1000] . "' אלפים";
2109 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2110 if ( $num >= $pow10 ) {
2111 if ( $num === 15 ||
$num === 16 ) {
2112 $letters[] = $table[0][9];
2113 $letters[] = $table[0][$num - 9];
2116 $letters = array_merge(
2118 (array)$table[$i][intval( $num / $pow10 )]
2121 if ( $pow10 === 1000 ) {
2127 $num = $num %
$pow10;
2130 $preTransformLength = count( $letters );
2131 if ( $preTransformLength === 1 ) {
2132 // Add geresh (single quote) to one-letter numbers
2135 $lastIndex = $preTransformLength - 1;
2136 $letters[$lastIndex] = str_replace(
2137 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2138 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2139 $letters[$lastIndex]
2142 // Add gershayim (double quote) to multiple-letter numbers,
2143 // but exclude numbers with only one letter after the thousands
2144 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2145 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2148 array_splice( $letters, -1, 0, '"' );
2152 return implode( $letters );
2156 * Used by date() and time() to adjust the time output.
2158 * @param string $ts The time in date('YmdHis') format
2159 * @param mixed $tz Adjust the time by this amount (default false, mean we
2160 * get user timecorrection setting)
2163 public function userAdjust( $ts, $tz = false ) {
2164 global $wgUser, $wgLocalTZoffset;
2166 if ( $tz === false ) {
2167 $tz = $wgUser->getOption( 'timecorrection' );
2170 $data = explode( '|', $tz, 3 );
2172 if ( $data[0] == 'ZoneInfo' ) {
2174 $userTZ = new DateTimeZone( $data[2] );
2175 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2176 $date->setTimezone( $userTZ );
2177 return $date->format( 'YmdHis' );
2178 } catch ( Exception
$e ) {
2179 // Unrecognized timezone, default to 'Offset' with the stored offset.
2180 $data[0] = 'Offset';
2184 if ( $data[0] == 'System' ||
$tz == '' ) {
2185 # Global offset in minutes.
2186 $minDiff = $wgLocalTZoffset;
2187 } elseif ( $data[0] == 'Offset' ) {
2188 $minDiff = intval( $data[1] );
2190 $data = explode( ':', $tz );
2191 if ( count( $data ) == 2 ) {
2192 $data[0] = intval( $data[0] );
2193 $data[1] = intval( $data[1] );
2194 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2195 if ( $data[0] < 0 ) {
2196 $minDiff = -$minDiff;
2199 $minDiff = intval( $data[0] ) * 60;
2203 # No difference ? Return time unchanged
2204 if ( $minDiff == 0 ) {
2208 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2209 # Generate an adjusted date; take advantage of the fact that mktime
2210 # will normalize out-of-range values so we don't have to split $minDiff
2211 # into hours and minutes.
2213 (int)substr( $ts, 8, 2 ) ), # Hours
2214 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2215 (int)substr( $ts, 12, 2 ), # Seconds
2216 (int)substr( $ts, 4, 2 ), # Month
2217 (int)substr( $ts, 6, 2 ), # Day
2218 (int)substr( $ts, 0, 4 ) ); # Year
2220 $date = date( 'YmdHis', $t );
2221 Wikimedia\restoreWarnings
();
2227 * This is meant to be used by time(), date(), and timeanddate() to get
2228 * the date preference they're supposed to use, it should be used in
2231 * function timeanddate([...], $format = true) {
2232 * $datePreference = $this->dateFormat($format);
2236 * @param int|string|bool $usePrefs If true, the user's preference is used
2237 * if false, the site/language default is used
2238 * if int/string, assumed to be a format.
2241 function dateFormat( $usePrefs = true ) {
2244 if ( is_bool( $usePrefs ) ) {
2246 $datePreference = $wgUser->getDatePreference();
2248 $datePreference = (string)User
::getDefaultOption( 'date' );
2251 $datePreference = (string)$usePrefs;
2255 if ( $datePreference == '' ) {
2259 return $datePreference;
2263 * Get a format string for a given type and preference
2264 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2265 * @param string $pref The format name as it appears in Messages*.php under
2268 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2272 function getDateFormatString( $type, $pref ) {
2273 $wasDefault = false;
2274 if ( $pref == 'default' ) {
2276 $pref = $this->getDefaultDateFormat();
2279 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2280 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2282 if ( $type === 'pretty' && $df === null ) {
2283 $df = $this->getDateFormatString( 'date', $pref );
2286 if ( !$wasDefault && $df === null ) {
2287 $pref = $this->getDefaultDateFormat();
2288 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2291 $this->dateFormatStrings
[$type][$pref] = $df;
2293 return $this->dateFormatStrings
[$type][$pref];
2297 * @param string $ts The time format which needs to be turned into a
2298 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2299 * @param bool $adj Whether to adjust the time output according to the
2300 * user configured offset ($timecorrection)
2301 * @param mixed $format True to use user's date format preference
2302 * @param string|bool $timecorrection The time offset as returned by
2303 * validateTimeZone() in Special:Preferences
2306 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2307 $ts = wfTimestamp( TS_MW
, $ts );
2309 $ts = $this->userAdjust( $ts, $timecorrection );
2311 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2312 return $this->sprintfDate( $df, $ts );
2316 * @param string $ts The time format which needs to be turned into a
2317 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2318 * @param bool $adj Whether to adjust the time output according to the
2319 * user configured offset ($timecorrection)
2320 * @param mixed $format True to use user's date format preference
2321 * @param string|bool $timecorrection The time offset as returned by
2322 * validateTimeZone() in Special:Preferences
2325 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2326 $ts = wfTimestamp( TS_MW
, $ts );
2328 $ts = $this->userAdjust( $ts, $timecorrection );
2330 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2331 return $this->sprintfDate( $df, $ts );
2335 * @param string $ts The time format which needs to be turned into a
2336 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2337 * @param bool $adj Whether to adjust the time output according to the
2338 * user configured offset ($timecorrection)
2339 * @param mixed $format What format to return, if it's false output the
2340 * default one (default true)
2341 * @param string|bool $timecorrection The time offset as returned by
2342 * validateTimeZone() in Special:Preferences
2345 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2346 $ts = wfTimestamp( TS_MW
, $ts );
2348 $ts = $this->userAdjust( $ts, $timecorrection );
2350 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2351 return $this->sprintfDate( $df, $ts );
2355 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2359 * @param int $seconds The amount of seconds.
2360 * @param array $chosenIntervals The intervals to enable.
2364 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2365 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2369 foreach ( $intervals as $intervalName => $intervalValue ) {
2370 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2371 // duration-years, duration-decades, duration-centuries, duration-millennia
2372 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2373 $segments[] = $message->inLanguage( $this )->escaped();
2376 return $this->listToText( $segments );
2380 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2381 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2385 * @param int $seconds The amount of seconds.
2386 * @param array $chosenIntervals The intervals to enable.
2390 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2391 if ( empty( $chosenIntervals ) ) {
2392 $chosenIntervals = [
2404 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2405 $sortedNames = array_keys( $intervals );
2406 $smallestInterval = array_pop( $sortedNames );
2410 foreach ( $intervals as $name => $length ) {
2411 $value = floor( $seconds / $length );
2413 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2414 $seconds -= $value * $length;
2415 $segments[$name] = $value;
2423 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2425 * @param string $type Can be 'date', 'time' or 'both'
2426 * @param string $ts The time format which needs to be turned into a
2427 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2428 * @param User $user User object used to get preferences for timezone and format
2429 * @param array $options Array, can contain the following keys:
2430 * - 'timecorrection': time correction, can have the following values:
2431 * - true: use user's preference
2432 * - false: don't use time correction
2433 * - int: value of time correction in minutes
2434 * - 'format': format to use, can have the following values:
2435 * - true: use user's preference
2436 * - false: use default preference
2437 * - string: format to use
2441 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2442 $ts = wfTimestamp( TS_MW
, $ts );
2443 $options +
= [ 'timecorrection' => true, 'format' => true ];
2444 if ( $options['timecorrection'] !== false ) {
2445 if ( $options['timecorrection'] === true ) {
2446 $offset = $user->getOption( 'timecorrection' );
2448 $offset = $options['timecorrection'];
2450 $ts = $this->userAdjust( $ts, $offset );
2452 if ( $options['format'] === true ) {
2453 $format = $user->getDatePreference();
2455 $format = $options['format'];
2457 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2458 return $this->sprintfDate( $df, $ts );
2462 * Get the formatted date for the given timestamp and formatted for
2465 * @param mixed $ts Mixed: the time format which needs to be turned into a
2466 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2467 * @param User $user User object used to get preferences for timezone and format
2468 * @param array $options Array, can contain the following keys:
2469 * - 'timecorrection': time correction, can have the following values:
2470 * - true: use user's preference
2471 * - false: don't use time correction
2472 * - int: value of time correction in minutes
2473 * - 'format': format to use, can have the following values:
2474 * - true: use user's preference
2475 * - false: use default preference
2476 * - string: format to use
2480 public function userDate( $ts, User
$user, array $options = [] ) {
2481 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2485 * Get the formatted time for the given timestamp and formatted for
2488 * @param mixed $ts The time format which needs to be turned into a
2489 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2490 * @param User $user User object used to get preferences for timezone and format
2491 * @param array $options Array, can contain the following keys:
2492 * - 'timecorrection': time correction, can have the following values:
2493 * - true: use user's preference
2494 * - false: don't use time correction
2495 * - int: value of time correction in minutes
2496 * - 'format': format to use, can have the following values:
2497 * - true: use user's preference
2498 * - false: use default preference
2499 * - string: format to use
2503 public function userTime( $ts, User
$user, array $options = [] ) {
2504 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2508 * Get the formatted date and time for the given timestamp and formatted for
2511 * @param mixed $ts The time format which needs to be turned into a
2512 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2513 * @param User $user User object used to get preferences for timezone and format
2514 * @param array $options Array, can contain the following keys:
2515 * - 'timecorrection': time correction, can have the following values:
2516 * - true: use user's preference
2517 * - false: don't use time correction
2518 * - int: value of time correction in minutes
2519 * - 'format': format to use, can have the following values:
2520 * - true: use user's preference
2521 * - false: use default preference
2522 * - string: format to use
2526 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2527 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2531 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2533 * Determine the difference between the timestamp and the current time, and
2534 * generate a readable timestamp by returning "<N> <units> ago", where the
2535 * largest possible unit is used.
2537 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2539 * @param MWTimestamp $time
2540 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2541 * @param User|null $user User the timestamp is being generated for
2542 * (or null to use main context's user)
2543 * @return string Formatted timestamp
2545 public function getHumanTimestamp(
2546 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2548 if ( $relativeTo === null ) {
2549 $relativeTo = new MWTimestamp();
2551 if ( $user === null ) {
2552 $user = RequestContext
::getMain()->getUser();
2555 // Adjust for the user's timezone.
2556 $offsetThis = $time->offsetForUser( $user );
2557 $offsetRel = $relativeTo->offsetForUser( $user );
2560 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2561 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2564 // Reset the timezone on the objects.
2565 $time->timestamp
->sub( $offsetThis );
2566 $relativeTo->timestamp
->sub( $offsetRel );
2572 * Convert an MWTimestamp into a pretty human-readable timestamp using
2573 * the given user preferences and relative base time.
2575 * @see Language::getHumanTimestamp
2576 * @param MWTimestamp $ts Timestamp to prettify
2577 * @param MWTimestamp $relativeTo Base timestamp
2578 * @param User $user User preferences to use
2579 * @return string Human timestamp
2582 private function getHumanTimestampInternal(
2583 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2585 $diff = $ts->diff( $relativeTo );
2586 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2587 (int)$relativeTo->timestamp
->format( 'w' ) );
2588 $days = $diff->days ?
: (int)$diffDay;
2589 if ( $diff->invert ||
$days > 5
2590 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2592 // Timestamps are in different years: use full timestamp
2593 // Also do full timestamp for future dates
2595 * @todo FIXME: Add better handling of future timestamps.
2597 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2598 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2599 } elseif ( $days > 5 ) {
2600 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2601 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2602 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2603 } elseif ( $days > 1 ) {
2604 // Timestamp within the past week: show the day of the week and time
2605 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2606 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2608 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2609 $ts = wfMessage( "$weekday-at" )
2610 ->inLanguage( $this )
2611 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2613 } elseif ( $days == 1 ) {
2614 // Timestamp was yesterday: say 'yesterday' and the time.
2615 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2616 $ts = wfMessage( 'yesterday-at' )
2617 ->inLanguage( $this )
2618 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2620 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2621 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2622 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2623 $ts = wfMessage( 'today-at' )
2624 ->inLanguage( $this )
2625 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2628 // From here on in, the timestamp was soon enough ago so that we can simply say
2629 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2630 } elseif ( $diff->h
== 1 ) {
2631 // Less than 90 minutes, but more than an hour ago.
2632 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2633 } elseif ( $diff->i
>= 1 ) {
2634 // A few minutes ago.
2635 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2636 } elseif ( $diff->s
>= 30 ) {
2637 // Less than a minute, but more than 30 sec ago.
2638 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2640 // Less than 30 seconds ago.
2641 $ts = wfMessage( 'just-now' )->text();
2648 * @param string $key
2649 * @return string|null
2651 public function getMessage( $key ) {
2652 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2658 function getAllMessages() {
2659 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2664 * @param string $out
2665 * @param string $string
2668 public function iconv( $in, $out, $string ) {
2669 # Even with //IGNORE iconv can whine about illegal characters in
2670 # *input* string. We just ignore those too.
2671 # REF: https://bugs.php.net/bug.php?id=37166
2672 # REF: https://phabricator.wikimedia.org/T18885
2673 Wikimedia\
suppressWarnings();
2674 $text = iconv( $in, $out . '//IGNORE', $string );
2675 Wikimedia\restoreWarnings
();
2679 // callback functions for ucwords(), ucwordbreaks()
2682 * @param array $matches
2683 * @return mixed|string
2685 function ucwordbreaksCallbackAscii( $matches ) {
2686 return $this->ucfirst( $matches[1] );
2690 * @param array $matches
2693 function ucwordbreaksCallbackMB( $matches ) {
2694 return mb_strtoupper( $matches[0] );
2698 * @param array $matches
2701 function ucwordsCallbackMB( $matches ) {
2702 return mb_strtoupper( $matches[0] );
2706 * Make a string's first character uppercase
2708 * @param string $str
2712 public function ucfirst( $str ) {
2714 if ( $o < 96 ) { // if already uppercase...
2716 } elseif ( $o < 128 ) {
2717 return ucfirst( $str ); // use PHP's ucfirst()
2719 // fall back to more complex logic in case of multibyte strings
2720 return $this->uc( $str, true );
2725 * Convert a string to uppercase
2727 * @param string $str
2728 * @param bool $first
2732 public function uc( $str, $first = false ) {
2734 if ( $this->isMultibyte( $str ) ) {
2735 return $this->mbUpperChar( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2737 return ucfirst( $str );
2740 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2745 * Convert character to uppercase, allowing overrides of the default mb_upper
2746 * behaviour, which is buggy in many ways. Having a conversion table can be
2747 * useful during transitions between PHP versions where unicode changes happen.
2748 * This can make some resources unreachable on-wiki, see discussion at T219279.
2749 * Providing such a conversion table can allow to manage the transition period.
2753 * @param string $char
2757 protected function mbUpperChar( $char ) {
2758 global $wgOverrideUcfirstCharacters;
2759 if ( array_key_exists( $char, $wgOverrideUcfirstCharacters ) ) {
2760 return $wgOverrideUcfirstCharacters[$char];
2762 return mb_strtoupper( $char );
2767 * @param string $str
2768 * @return mixed|string
2770 function lcfirst( $str ) {
2773 return strval( $str );
2774 } elseif ( $o >= 128 ) {
2775 return $this->lc( $str, true );
2776 } elseif ( $o > 96 ) {
2779 $str[0] = strtolower( $str[0] );
2785 * @param string $str
2786 * @param bool $first
2787 * @return mixed|string
2789 function lc( $str, $first = false ) {
2791 if ( $this->isMultibyte( $str ) ) {
2792 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2794 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2797 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2802 * @param string $str
2805 function isMultibyte( $str ) {
2806 return strlen( $str ) !== mb_strlen( $str );
2810 * @param string $str
2811 * @return mixed|string
2813 function ucwords( $str ) {
2814 if ( $this->isMultibyte( $str ) ) {
2815 $str = $this->lc( $str );
2817 // regexp to find first letter in each word (i.e. after each space)
2818 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2820 // function to use to capitalize a single char
2821 return preg_replace_callback(
2823 [ $this, 'ucwordsCallbackMB' ],
2827 return ucwords( strtolower( $str ) );
2832 * capitalize words at word breaks
2834 * @param string $str
2837 function ucwordbreaks( $str ) {
2838 if ( $this->isMultibyte( $str ) ) {
2839 $str = $this->lc( $str );
2841 // since \b doesn't work for UTF-8, we explicitely define word break chars
2842 $breaks = "[ \-\(\)\}\{\.,\?!]";
2844 // find first letter after word break
2845 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2846 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2848 return preg_replace_callback(
2850 [ $this, 'ucwordbreaksCallbackMB' ],
2854 return preg_replace_callback(
2855 '/\b([\w\x80-\xff]+)\b/',
2856 [ $this, 'ucwordbreaksCallbackAscii' ],
2863 * Return a case-folded representation of $s
2865 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2866 * and $s2 are the same except for the case of their characters. It is not
2867 * necessary for the value returned to make sense when displayed.
2869 * Do *not* perform any other normalisation in this function. If a caller
2870 * uses this function when it should be using a more general normalisation
2871 * function, then fix the caller.
2877 function caseFold( $s ) {
2878 return $this->uc( $s );
2882 * TODO: $s is not always a string per T218883
2886 function checkTitleEncoding( $s ) {
2887 if ( is_array( $s ) ) {
2888 throw new MWException( 'Given array to checkTitleEncoding.' );
2890 if ( StringUtils
::isUtf8( $s ) ) {
2894 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2900 function fallback8bitEncoding() {
2901 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2905 * Most writing systems use whitespace to break up words.
2906 * Some languages such as Chinese don't conventionally do this,
2907 * which requires special handling when breaking up words for
2912 function hasWordBreaks() {
2917 * Some languages such as Chinese require word segmentation,
2918 * Specify such segmentation when overridden in derived class.
2920 * @param string $string
2923 function segmentByWord( $string ) {
2928 * Some languages have special punctuation need to be normalized.
2929 * Make such changes here.
2931 * @param string $string
2934 function normalizeForSearch( $string ) {
2935 return self
::convertDoubleWidth( $string );
2939 * convert double-width roman characters to single-width.
2940 * range: ff00-ff5f ~= 0020-007f
2942 * @param string $string
2946 protected static function convertDoubleWidth( $string ) {
2947 static $full = null;
2948 static $half = null;
2950 if ( $full === null ) {
2951 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2952 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2953 $full = str_split( $fullWidth, 3 );
2954 $half = str_split( $halfWidth );
2957 $string = str_replace( $full, $half, $string );
2962 * @param string $string
2963 * @param string $pattern
2966 protected static function insertSpace( $string, $pattern ) {
2967 $string = preg_replace( $pattern, " $1 ", $string );
2968 $string = preg_replace( '/ +/', ' ', $string );
2973 * @param string[] $termsArray
2976 function convertForSearchResult( $termsArray ) {
2977 # some languages, e.g. Chinese, need to do a conversion
2978 # in order for search results to be displayed correctly
2983 * Get the first character of a string.
2988 function firstChar( $s ) {
2991 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2992 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2997 if ( isset( $matches[1] ) ) {
2998 if ( strlen( $matches[1] ) != 3 ) {
3002 // Break down Hangul syllables to grab the first jamo
3003 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3004 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3006 } elseif ( $code < 0xb098 ) {
3008 } elseif ( $code < 0xb2e4 ) {
3010 } elseif ( $code < 0xb77c ) {
3012 } elseif ( $code < 0xb9c8 ) {
3014 } elseif ( $code < 0xbc14 ) {
3016 } elseif ( $code < 0xc0ac ) {
3018 } elseif ( $code < 0xc544 ) {
3020 } elseif ( $code < 0xc790 ) {
3022 } elseif ( $code < 0xcc28 ) {
3024 } elseif ( $code < 0xce74 ) {
3026 } elseif ( $code < 0xd0c0 ) {
3028 } elseif ( $code < 0xd30c ) {
3030 } elseif ( $code < 0xd558 ) {
3041 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3042 * also cleans up certain backwards-compatible sequences, converting them
3043 * to the modern Unicode equivalent.
3045 * This is language-specific for performance reasons only.
3051 public function normalize( $s ) {
3052 global $wgAllUnicodeFixes;
3053 $s = UtfNormal\Validator
::cleanUp( $s );
3054 if ( $wgAllUnicodeFixes ) {
3055 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
3056 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3063 * Transform a string using serialized data stored in the given file (which
3064 * must be in the serialized subdirectory of $IP). The file contains pairs
3065 * mapping source characters to destination characters.
3067 * The data is cached in process memory. This will go faster if you have the
3068 * FastStringSearch extension.
3070 * @param string $file
3071 * @param string $string
3073 * @throws MWException
3076 protected function transformUsingPairFile( $file, $string ) {
3077 if ( !isset( $this->transformData
[$file] ) ) {
3079 $data = require "$IP/languages/data/{$file}";
3080 $this->transformData
[$file] = new ReplacementArray( $data );
3082 return $this->transformData
[$file]->replace( $string );
3086 * For right-to-left language support
3091 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3095 * Return the correct HTML 'dir' attribute value for this language.
3099 return $this->isRTL() ?
'rtl' : 'ltr';
3103 * Return 'left' or 'right' as appropriate alignment for line-start
3104 * for this language's text direction.
3106 * Should be equivalent to CSS3 'start' text-align value....
3110 function alignStart() {
3111 return $this->isRTL() ?
'right' : 'left';
3115 * Return 'right' or 'left' as appropriate alignment for line-end
3116 * for this language's text direction.
3118 * Should be equivalent to CSS3 'end' text-align value....
3122 function alignEnd() {
3123 return $this->isRTL() ?
'left' : 'right';
3127 * A hidden direction mark (LRM or RLM), depending on the language direction.
3128 * Unlike getDirMark(), this function returns the character as an HTML entity.
3129 * This function should be used when the output is guaranteed to be HTML,
3130 * because it makes the output HTML source code more readable. When
3131 * the output is plain text or can be escaped, getDirMark() should be used.
3133 * @param bool $opposite Get the direction mark opposite to your language
3137 function getDirMarkEntity( $opposite = false ) {
3139 return $this->isRTL() ?
'‎' : '‏';
3141 return $this->isRTL() ?
'‏' : '‎';
3145 * A hidden direction mark (LRM or RLM), depending on the language direction.
3146 * This function produces them as invisible Unicode characters and
3147 * the output may be hard to read and debug, so it should only be used
3148 * when the output is plain text or can be escaped. When the output is
3149 * HTML, use getDirMarkEntity() instead.
3151 * @param bool $opposite Get the direction mark opposite to your language
3154 function getDirMark( $opposite = false ) {
3155 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3156 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3158 return $this->isRTL() ?
$lrm : $rlm;
3160 return $this->isRTL() ?
$rlm : $lrm;
3166 function capitalizeAllNouns() {
3167 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3171 * An arrow, depending on the language direction.
3173 * @param string $direction The direction of the arrow: forwards (default),
3174 * backwards, left, right, up, down.
3177 function getArrow( $direction = 'forwards' ) {
3178 switch ( $direction ) {
3180 return $this->isRTL() ?
'←' : '→';
3182 return $this->isRTL() ?
'→' : '←';
3195 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3199 function linkPrefixExtension() {
3200 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3204 * Get all magic words from cache.
3207 function getMagicWords() {
3208 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3212 * Fill a MagicWord object with data from here
3214 * @param MagicWord $mw
3216 function getMagic( $mw ) {
3217 $rawEntry = $this->mMagicExtensions
[$mw->mId
] ??
3218 self
::$dataCache->getSubitem( $this->mCode
, 'magicWords', $mw->mId
);
3220 if ( !is_array( $rawEntry ) ) {
3221 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3223 $mw->mCaseSensitive
= $rawEntry[0];
3224 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3229 * Add magic words to the extension array
3231 * @param array $newWords
3233 function addMagicWordsByLang( $newWords ) {
3234 $fallbackChain = $this->getFallbackLanguages();
3235 $fallbackChain = array_reverse( $fallbackChain );
3236 foreach ( $fallbackChain as $code ) {
3237 if ( isset( $newWords[$code] ) ) {
3238 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3244 * Get special page names, as an associative array
3245 * canonical name => array of valid names, including aliases
3248 function getSpecialPageAliases() {
3249 // Cache aliases because it may be slow to load them
3250 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3252 $this->mExtendedSpecialPageAliases
=
3253 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3256 return $this->mExtendedSpecialPageAliases
;
3260 * Italic is unsuitable for some languages
3262 * @param string $text The text to be emphasized.
3265 function emphasize( $text ) {
3266 return "<em>$text</em>";
3270 * Normally we output all numbers in plain en_US style, that is
3271 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3272 * point twohundredthirtyfive. However this is not suitable for all
3273 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3274 * Icelandic just want to use commas instead of dots, and dots instead
3275 * of commas like "293.291,235".
3277 * An example of this function being called:
3279 * wfMessage( 'message' )->numParams( $num )->text()
3282 * See $separatorTransformTable on MessageIs.php for
3283 * the , => . and . => , implementation.
3285 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3286 * @param int|float $number The string to be formatted, should be an integer
3287 * or a floating point number.
3288 * @param bool $nocommafy Set to true for special numbers like dates
3291 public function formatNum( $number, $nocommafy = false ) {
3292 global $wgTranslateNumerals;
3293 if ( !$nocommafy ) {
3294 $number = $this->commafy( $number );
3295 $s = $this->separatorTransformTable();
3297 $number = strtr( $number, $s );
3301 if ( $wgTranslateNumerals ) {
3302 $s = $this->digitTransformTable();
3304 $number = strtr( $number, $s );
3308 return (string)$number;
3312 * Front-end for non-commafied formatNum
3314 * @param int|float $number The string to be formatted, should be an integer
3315 * or a floating point number.
3319 public function formatNumNoSeparators( $number ) {
3320 return $this->formatNum( $number, true );
3324 * @param string $number
3327 public function parseFormattedNumber( $number ) {
3328 $s = $this->digitTransformTable();
3330 // eliminate empty array values such as ''. (T66347)
3331 $s = array_filter( $s );
3332 $number = strtr( $number, array_flip( $s ) );
3335 $s = $this->separatorTransformTable();
3337 // eliminate empty array values such as ''. (T66347)
3338 $s = array_filter( $s );
3339 $number = strtr( $number, array_flip( $s ) );
3342 $number = strtr( $number, [ ',' => '' ] );
3347 * Adds commas to a given number
3349 * @param mixed $number
3352 function commafy( $number ) {
3353 $digitGroupingPattern = $this->digitGroupingPattern();
3354 $minimumGroupingDigits = $this->minimumGroupingDigits();
3355 if ( $number === null ) {
3359 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3360 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3361 // In some languages it's conventional not to insert a thousands separator
3362 // in numbers that are four digits long (1000-9999).
3363 if ( $minimumGroupingDigits ) {
3364 // Number of '#' characters after last comma in the grouping pattern.
3365 // The pattern is hardcoded here, but this would vary for different patterns.
3366 $primaryGroupingSize = 3;
3367 // Maximum length of a number to suppress digit grouping for.
3368 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3369 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3373 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3375 // Ref: http://cldr.unicode.org/translation/number-patterns
3377 if ( intval( $number ) < 0 ) {
3378 // For negative numbers apply the algorithm like positive number and add sign.
3380 $number = substr( $number, 1 );
3384 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3385 preg_match( "/\d+/", $number, $integerPart );
3386 preg_match( "/\.\d*/", $number, $decimalPart );
3387 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3388 if ( $groupedNumber === $number ) {
3389 // the string does not have any number part. Eg: .12345
3390 return $sign . $groupedNumber;
3392 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3393 while ( $start > 0 ) {
3394 $match = $matches[0][$numMatches - 1];
3395 $matchLen = strlen( $match );
3396 $start = $end - $matchLen;
3400 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3402 if ( $numMatches > 1 ) {
3403 // use the last pattern for the rest of the number
3407 $groupedNumber = "," . $groupedNumber;
3410 return $sign . $groupedNumber;
3417 function digitGroupingPattern() {
3418 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3424 function digitTransformTable() {
3425 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3431 function separatorTransformTable() {
3432 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3438 function minimumGroupingDigits() {
3439 return self
::$dataCache->getItem( $this->mCode
, 'minimumGroupingDigits' );
3443 * Take a list of strings and build a locale-friendly comma-separated
3444 * list, using the local comma-separator message.
3445 * The last two strings are chained with an "and".
3447 * @param string[] $list
3450 public function listToText( array $list ) {
3451 $itemCount = count( $list );
3452 if ( $itemCount < 1 ) {
3455 $text = array_pop( $list );
3456 if ( $itemCount > 1 ) {
3457 $and = $this->msg( 'and' )->escaped();
3458 $space = $this->msg( 'word-separator' )->escaped();
3460 if ( $itemCount > 2 ) {
3461 $comma = $this->msg( 'comma-separator' )->escaped();
3463 $text = implode( $comma, $list ) . $and . $space . $text;
3469 * Take a list of strings and build a locale-friendly comma-separated
3470 * list, using the local comma-separator message.
3471 * @param string[] $list Array of strings to put in a comma list
3474 function commaList( array $list ) {
3476 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3482 * Take a list of strings and build a locale-friendly semicolon-separated
3483 * list, using the local semicolon-separator message.
3484 * @param string[] $list Array of strings to put in a semicolon list
3487 function semicolonList( array $list ) {
3489 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3495 * Same as commaList, but separate it with the pipe instead.
3496 * @param string[] $list Array of strings to put in a pipe list
3499 function pipeList( array $list ) {
3501 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3507 * Truncate a string to a specified length in bytes, appending an optional
3508 * string (e.g. for ellipsis)
3510 * If $length is negative, the string will be truncated from the beginning
3514 * @param string $string String to truncate
3515 * @param int $length Maximum length in bytes
3516 * @param string $ellipsis String to append to the end of truncated text
3517 * @param bool $adjustLength Subtract length of ellipsis from $length
3521 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3522 return $this->truncateInternal(
3523 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3528 * Truncate a string to a specified number of characters, appending an optional
3529 * string (e.g. for ellipsis).
3531 * This provides multibyte version of truncateForDatabase() method of this class,
3532 * suitable for truncation based on number of characters, instead of number of bytes.
3534 * If $length is negative, the string will be truncated from the beginning.
3538 * @param string $string String to truncate
3539 * @param int $length Maximum number of characters
3540 * @param string $ellipsis String to append to the end of truncated text
3541 * @param bool $adjustLength Subtract length of ellipsis from $length
3545 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3546 // Passing encoding to mb_strlen and mb_substr is optional.
3547 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3548 // explicit specification of encoding is skipped.
3549 // Note: Both multibyte methods are callables invoked in truncateInternal.
3550 return $this->truncateInternal(
3551 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3556 * Internal method used for truncation. This method abstracts text truncation into
3557 * one common method, allowing users to provide length measurement function and
3558 * function for finding substring.
3560 * For usages, see truncateForDatabase and truncateForVisual.
3562 * @param string $string String to truncate
3563 * @param int $length Maximum length of final text
3564 * @param string $ellipsis String to append to the end of truncated text
3565 * @param bool $adjustLength Subtract length of ellipsis from $length
3566 * @param callable $measureLength Callable function used for determining the length of text
3567 * @param callable $getSubstring Callable function used for getting the substrings
3571 private function truncateInternal(
3572 $string, $length, $ellipsis, $adjustLength, callable
$measureLength, callable
$getSubstring
3574 # Check if there is no need to truncate
3575 if ( $measureLength( $string ) <= abs( $length ) ) {
3576 return $string; // no need to truncate
3579 # Use the localized ellipsis character
3580 if ( $ellipsis == '...' ) {
3581 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3583 if ( $length == 0 ) {
3584 return $ellipsis; // convention
3587 $stringOriginal = $string;
3588 # If ellipsis length is >= $length then we can't apply $adjustLength
3589 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3590 $string = $ellipsis; // this can be slightly unexpected
3591 # Otherwise, truncate and add ellipsis...
3593 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3594 if ( $length > 0 ) {
3595 $length -= $ellipsisLength;
3596 $string = $getSubstring( $string, 0, $length ); // xyz...
3597 $string = $this->removeBadCharLast( $string );
3598 $string = rtrim( $string ) . $ellipsis;
3600 $length +
= $ellipsisLength;
3601 $string = $getSubstring( $string, $length ); // ...xyz
3602 $string = $this->removeBadCharFirst( $string );
3603 $string = $ellipsis . ltrim( $string );
3607 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3608 # This check is *not* redundant if $adjustLength, due to the single case where
3609 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3610 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3613 return $stringOriginal;
3618 * Remove bytes that represent an incomplete Unicode character
3619 * at the end of string (e.g. bytes of the char are missing)
3621 * @param string $string
3624 protected function removeBadCharLast( $string ) {
3625 if ( $string != '' ) {
3626 $char = ord( $string[strlen( $string ) - 1] );
3628 if ( $char >= 0xc0 ) {
3629 # We got the first byte only of a multibyte char; remove it.
3630 $string = substr( $string, 0, -1 );
3631 } elseif ( $char >= 0x80 &&
3632 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3633 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3634 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3636 # We chopped in the middle of a character; remove it
3644 * Remove bytes that represent an incomplete Unicode character
3645 * at the start of string (e.g. bytes of the char are missing)
3647 * @param string $string
3650 protected function removeBadCharFirst( $string ) {
3651 if ( $string != '' ) {
3652 $char = ord( $string[0] );
3653 if ( $char >= 0x80 && $char < 0xc0 ) {
3654 # We chopped in the middle of a character; remove the whole thing
3655 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3662 * Truncate a string of valid HTML to a specified length in bytes,
3663 * appending an optional string (e.g. for ellipses), and return valid HTML
3665 * This is only intended for styled/linked text, such as HTML with
3666 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3667 * Also, this will not detect things like "display:none" CSS.
3669 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3671 * @param string $text HTML string to truncate
3672 * @param int $length (zero/positive) Maximum length (including ellipses)
3673 * @param string $ellipsis String to append to the truncated text
3676 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3677 # Use the localized ellipsis character
3678 if ( $ellipsis == '...' ) {
3679 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3681 # Check if there is clearly no need to truncate
3682 if ( $length <= 0 ) {
3683 return $ellipsis; // no text shown, nothing to format (convention)
3684 } elseif ( strlen( $text ) <= $length ) {
3685 return $text; // string short enough even *with* HTML (short-circuit)
3688 $dispLen = 0; // innerHTML legth so far
3689 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3690 $tagType = 0; // 0-open, 1-close
3691 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3692 $entityState = 0; // 0-not entity, 1-entity
3693 $tag = $ret = ''; // accumulated tag name, accumulated result string
3694 $openTags = []; // open tag stack
3695 $maybeState = null; // possible truncation state
3697 $textLen = strlen( $text );
3698 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3699 for ( $pos = 0; true; ++
$pos ) {
3700 # Consider truncation once the display length has reached the maximim.
3701 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3702 # Check that we're not in the middle of a bracket/entity...
3703 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3704 if ( !$testingEllipsis ) {
3705 $testingEllipsis = true;
3706 # Save where we are; we will truncate here unless there turn out to
3707 # be so few remaining characters that truncation is not necessary.
3708 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3709 $maybeState = [ $ret, $openTags ]; // save state
3711 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3712 # String in fact does need truncation, the truncation point was OK.
3713 // @phan-suppress-next-line PhanTypeInvalidExpressionArrayDestructuring
3714 list( $ret, $openTags ) = $maybeState; // reload state
3715 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3716 $ret .= $ellipsis; // add ellipsis
3720 if ( $pos >= $textLen ) {
3721 break; // extra iteration just for above checks
3724 # Read the next char...
3726 $lastCh = $pos ?
$text[$pos - 1] : '';
3727 $ret .= $ch; // add to result string
3729 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3730 $entityState = 0; // for bad HTML
3731 $bracketState = 1; // tag started (checking for backslash)
3732 } elseif ( $ch == '>' ) {
3733 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3734 $entityState = 0; // for bad HTML
3735 $bracketState = 0; // out of brackets
3736 } elseif ( $bracketState == 1 ) {
3738 $tagType = 1; // close tag (e.g. "</span>")
3740 $tagType = 0; // open tag (e.g. "<span>")
3743 $bracketState = 2; // building tag name
3744 } elseif ( $bracketState == 2 ) {
3748 // Name found (e.g. "<a href=..."), add on tag attributes...
3749 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3751 } elseif ( $bracketState == 0 ) {
3752 if ( $entityState ) {
3755 $dispLen++
; // entity is one displayed char
3758 if ( $neLength == 0 && !$maybeState ) {
3759 // Save state without $ch. We want to *hit* the first
3760 // display char (to get tags) but not *use* it if truncating.
3761 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3764 $entityState = 1; // entity found, (e.g. " ")
3766 $dispLen++
; // this char is displayed
3767 // Add the next $max display text chars after this in one swoop...
3768 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3769 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3770 $dispLen +
= $skipped;
3776 // Close the last tag if left unclosed by bad HTML
3777 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3778 while ( count( $openTags ) > 0 ) {
3779 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3785 * truncateHtml() helper function
3786 * like strcspn() but adds the skipped chars to $ret
3788 * @param string $ret
3789 * @param string $text
3790 * @param string $search
3792 * @param null|int $len
3795 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3796 if ( $len === null ) {
3797 $len = -1; // -1 means "no limit" for strcspn
3798 } elseif ( $len < 0 ) {
3802 if ( $start < strlen( $text ) ) {
3803 $skipCount = strcspn( $text, $search, $start, $len );
3804 $ret .= substr( $text, $start, $skipCount );
3810 * truncateHtml() helper function
3811 * (a) push or pop $tag from $openTags as needed
3812 * (b) clear $tag value
3813 * @param string &$tag Current HTML tag name we are looking at
3814 * @param int $tagType (0-open tag, 1-close tag)
3815 * @param string $lastCh Character before the '>' that ended this tag
3816 * @param array &$openTags Open tag stack (not accounting for $tag)
3818 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3819 $tag = ltrim( $tag );
3821 if ( $tagType == 0 && $lastCh != '/' ) {
3822 $openTags[] = $tag; // tag opened (didn't close itself)
3823 } elseif ( $tagType == 1 ) {
3824 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3825 array_pop( $openTags ); // tag closed
3833 * Grammatical transformations, needed for inflected languages
3834 * Invoked by putting {{grammar:case|word}} in a message
3836 * @param string $word
3837 * @param string $case
3840 function convertGrammar( $word, $case ) {
3841 global $wgGrammarForms;
3842 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3843 return $wgGrammarForms[$this->getCode()][$case][$word];
3846 $grammarTransformations = $this->getGrammarTransformations();
3848 if ( isset( $grammarTransformations[$case] ) ) {
3849 $forms = $grammarTransformations[$case];
3851 // Some names of grammar rules are aliases for other rules.
3852 // In such cases the value is a string rather than object,
3853 // so load the actual rules.
3854 if ( is_string( $forms ) ) {
3855 $forms = $grammarTransformations[$forms];
3858 foreach ( array_values( $forms ) as $rule ) {
3861 if ( $form === '@metadata' ) {
3865 $replacement = $rule[1];
3867 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3868 $patternMatches = preg_match( $regex, $word );
3870 if ( $patternMatches === false ) {
3872 'An error occurred while processing grammar. ' .
3873 "Word: '$word'. Regex: /$form/."
3875 } elseif ( $patternMatches === 1 ) {
3876 $word = preg_replace( $regex, $replacement, $word );
3887 * Get the grammar forms for the content language
3888 * @return array Array of grammar forms
3891 function getGrammarForms() {
3892 global $wgGrammarForms;
3893 if ( isset( $wgGrammarForms[$this->getCode()] )
3894 && is_array( $wgGrammarForms[$this->getCode()] )
3896 return $wgGrammarForms[$this->getCode()];
3903 * Get the grammar transformations data for the language.
3904 * Used like grammar forms, with {{GRAMMAR}} and cases,
3905 * but uses pairs of regexes and replacements instead of code.
3907 * @return array[] Array of grammar transformations.
3908 * @throws MWException
3911 public function getGrammarTransformations() {
3912 $languageCode = $this->getCode();
3914 if ( self
::$grammarTransformations === null ) {
3915 self
::$grammarTransformations = new MapCacheLRU( 10 );
3918 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3919 return self
::$grammarTransformations->get( $languageCode );
3924 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3925 if ( is_readable( $grammarDataFile ) ) {
3926 $data = FormatJson
::decode(
3927 file_get_contents( $grammarDataFile ),
3931 if ( $data === null ) {
3932 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3935 self
::$grammarTransformations->set( $languageCode, $data );
3942 * Provides an alternative text depending on specified gender.
3943 * Usage {{gender:username|masculine|feminine|unknown}}.
3944 * username is optional, in which case the gender of current user is used,
3945 * but only in (some) interface messages; otherwise default gender is used.
3947 * If no forms are given, an empty string is returned. If only one form is
3948 * given, it will be returned unconditionally. These details are implied by
3949 * the caller and cannot be overridden in subclasses.
3951 * If three forms are given, the default is to use the third (unknown) form.
3952 * If fewer than three forms are given, the default is to use the first (masculine) form.
3953 * These details can be overridden in subclasses.
3955 * @param string $gender
3956 * @param array $forms
3960 function gender( $gender, $forms ) {
3961 if ( !count( $forms ) ) {
3964 $forms = $this->preConvertPlural( $forms, 2 );
3965 if ( $gender === 'male' ) {
3968 if ( $gender === 'female' ) {
3971 return $forms[2] ??
$forms[0];
3975 * Plural form transformations, needed for some languages.
3976 * For example, there are 3 form of plural in Russian and Polish,
3977 * depending on "count mod 10". See [[w:Plural]]
3978 * For English it is pretty simple.
3980 * Invoked by putting {{plural:count|wordform1|wordform2}}
3981 * or {{plural:count|wordform1|wordform2|wordform3}}
3983 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3985 * @param int $count Non-localized number
3986 * @param array $forms Different plural forms
3987 * @return string Correct form of plural for $count in this language
3989 function convertPlural( $count, $forms ) {
3990 // Handle explicit n=pluralform cases
3991 $forms = $this->handleExplicitPluralForms( $count, $forms );
3992 if ( is_string( $forms ) ) {
3995 if ( !count( $forms ) ) {
3999 $pluralForm = $this->getPluralRuleIndexNumber( $count );
4000 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
4001 return $forms[$pluralForm];
4005 * Handles explicit plural forms for Language::convertPlural()
4007 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4008 * If an explicitly defined plural form matches the $count, then
4009 * string value returned, otherwise array returned for further consideration
4010 * by CLDR rules or overridden convertPlural().
4014 * @param int $count Non-localized number
4015 * @param array $forms Different plural forms
4017 * @return array|string
4019 protected function handleExplicitPluralForms( $count, array $forms ) {
4020 foreach ( $forms as $index => $form ) {
4021 if ( preg_match( '/\d+=/i', $form ) ) {
4022 $pos = strpos( $form, '=' );
4023 if ( substr( $form, 0, $pos ) === (string)$count ) {
4024 return substr( $form, $pos +
1 );
4026 unset( $forms[$index] );
4029 return array_values( $forms );
4033 * Checks that convertPlural was given an array and pads it to requested
4034 * amount of forms by copying the last one.
4036 * @param array $forms
4037 * @param int $count How many forms should there be at least
4038 * @return array Padded array of forms
4040 protected function preConvertPlural( /* Array */ $forms, $count ) {
4041 return array_pad( $forms, $count, end( $forms ) );
4045 * Wraps argument with unicode control characters for directionality safety
4047 * This solves the problem where directionality-neutral characters at the edge of
4048 * the argument string get interpreted with the wrong directionality from the
4049 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4051 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4052 * directionality of the argument string, using the BIDI algorithm's own "First
4053 * strong directional codepoint" rule. Essentially, this works round the fact that
4054 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4055 * direction inference). The latter is cleaner but still not widely supported.
4057 * @param string $text Text to wrap
4058 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4060 public function embedBidi( $text = '' ) {
4061 $dir = self
::strongDirFromContent( $text );
4062 if ( $dir === 'ltr' ) {
4063 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4064 return self
::$lre . $text . self
::$pdf;
4066 if ( $dir === 'rtl' ) {
4067 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4068 return self
::$rle . $text . self
::$pdf;
4070 // No strong directionality: do not wrap
4075 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4076 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4077 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4078 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4081 * @param string $str The validated block duration in English
4082 * @param User|null $user User object to use timezone from or null for $wgUser
4083 * @param int $now Current timestamp, for formatting relative block durations
4084 * @return string Somehow translated block duration
4085 * @see LanguageFi.php for example implementation
4087 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
4088 $duration = SpecialBlock
::getSuggestedDurations( $this );
4089 foreach ( $duration as $show => $value ) {
4090 if ( strcmp( $str, $value ) == 0 ) {
4091 return htmlspecialchars( trim( $show ) );
4095 if ( wfIsInfinity( $str ) ) {
4096 foreach ( $duration as $show => $value ) {
4097 if ( wfIsInfinity( $value ) ) {
4098 return htmlspecialchars( trim( $show ) );
4103 // If all else fails, return a standard duration or timestamp description.
4104 $time = strtotime( $str, $now );
4105 if ( $time === false ) { // Unknown format. Return it as-is in case.
4107 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
4108 // The result differs based on current time, so the difference
4109 // is a fixed duration length.
4110 return $this->formatDuration( $time - $now );
4111 } else { // It's an absolute timestamp.
4112 if ( $time === 0 ) {
4113 // wfTimestamp() handles 0 as current time instead of epoch.
4114 $time = '19700101000000';
4117 return $this->userTimeAndDate( $time, $user );
4119 return $this->timeanddate( $time );
4124 * languages like Chinese need to be segmented in order for the diff
4127 * @param string $text
4130 public function segmentForDiff( $text ) {
4135 * and unsegment to show the result
4137 * @param string $text
4140 public function unsegmentForDiff( $text ) {
4145 * Return the LanguageConverter used in the Language
4148 * @return LanguageConverter
4150 public function getConverter() {
4151 return $this->mConverter
;
4155 * convert text to a variant
4157 * @param string $text text to convert
4158 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4159 * variant (if logged in), or the project default variant
4160 * @return string the converted string
4162 public function autoConvert( $text, $variant = false ) {
4163 return $this->mConverter
->autoConvert( $text, $variant );
4167 * convert text to all supported variants
4169 * @param string $text
4172 public function autoConvertToAllVariants( $text ) {
4173 return $this->mConverter
->autoConvertToAllVariants( $text );
4177 * convert text to different variants of a language.
4179 * @warning Glossary state is maintained between calls. This means
4180 * if you pass unescaped text to this method it can cause an XSS
4181 * in later calls to this method, even if the later calls have properly
4182 * escaped the input. Never feed this method user controlled text that
4183 * is not properly escaped!
4184 * @param string $text Content that has been already escaped for use in HTML
4185 * @return string HTML
4187 public function convert( $text ) {
4188 return $this->mConverter
->convert( $text );
4192 * Convert a Title object to a string in the preferred variant
4194 * @param Title $title
4197 public function convertTitle( $title ) {
4198 return $this->mConverter
->convertTitle( $title );
4202 * Convert a namespace index to a string in the preferred variant
4204 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4205 * @param string|null $variant variant to convert to, or null to use the user's preferred
4206 * variant (if logged in), or the project default variant
4207 * @return string a string representation of the namespace
4209 public function convertNamespace( $ns, $variant = null ) {
4210 return $this->mConverter
->convertNamespace( $ns, $variant );
4214 * Check if this is a language with variants
4218 public function hasVariants() {
4219 return count( $this->getVariants() ) > 1;
4223 * Strict check if the language has the specific variant.
4225 * Compare to LanguageConverter::validateVariant() which does a more
4226 * lenient check and attempts to coerce the given code to a valid one.
4229 * @param string $variant
4232 public function hasVariant( $variant ) {
4233 return $variant && ( $variant === $this->mConverter
->validateVariant( $variant ) );
4237 * Perform output conversion on a string, and encode for safe HTML output.
4238 * @param string $text Text to be converted
4240 * @todo this should get integrated somewhere sane
4242 public function convertHtml( $text ) {
4243 return htmlspecialchars( $this->convert( $text ) );
4247 * @param string $key
4250 public function convertCategoryKey( $key ) {
4251 return $this->mConverter
->convertCategoryKey( $key );
4255 * Get the list of variants supported by this language
4256 * see sample implementation in LanguageZh.php
4258 * @return string[] An array of language codes
4260 public function getVariants() {
4261 return $this->mConverter
->getVariants();
4267 public function getPreferredVariant() {
4268 return $this->mConverter
->getPreferredVariant();
4274 public function getDefaultVariant() {
4275 return $this->mConverter
->getDefaultVariant();
4281 public function getURLVariant() {
4282 return $this->mConverter
->getURLVariant();
4286 * If a language supports multiple variants, it is
4287 * possible that non-existing link in one variant
4288 * actually exists in another variant. this function
4289 * tries to find it. See e.g. LanguageZh.php
4290 * The input parameters may be modified upon return
4292 * @param string &$link The name of the link
4293 * @param Title &$nt The title object of the link
4294 * @param bool $ignoreOtherCond To disable other conditions when
4295 * we need to transclude a template or update a category's link
4297 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4298 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4302 * returns language specific options used by User::getPageRenderHash()
4303 * for example, the preferred language variant
4307 function getExtraHashOptions() {
4308 return $this->mConverter
->getExtraHashOptions();
4312 * For languages that support multiple variants, the title of an
4313 * article may be displayed differently in different variants. this
4314 * function returns the apporiate title defined in the body of the article.
4318 public function getParsedTitle() {
4319 return $this->mConverter
->getParsedTitle();
4323 * Refresh the cache of conversion tables when
4324 * MediaWiki:Conversiontable* is updated.
4326 * @param Title $title The Title of the page being updated
4328 public function updateConversionTable( Title
$title ) {
4329 $this->mConverter
->updateConversionTable( $title );
4333 * A regular expression to match legal word-trailing characters
4334 * which should be merged onto a link of the form [[foo]]bar.
4338 public function linkTrail() {
4339 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4343 * A regular expression character set to match legal word-prefixing
4344 * characters which should be merged onto a link of the form foo[[bar]].
4348 public function linkPrefixCharset() {
4349 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4353 * Get the "parent" language which has a converter to convert a "compatible" language
4354 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4356 * @return Language|null
4359 public function getParentLanguage() {
4360 if ( $this->mParentLanguage
!== false ) {
4361 return $this->mParentLanguage
;
4364 $code = explode( '-', $this->getCode() )[0];
4365 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4366 $this->mParentLanguage
= null;
4369 $lang = self
::factory( $code );
4370 if ( !$lang->hasVariant( $this->getCode() ) ) {
4371 $this->mParentLanguage
= null;
4375 $this->mParentLanguage
= $lang;
4380 * Compare with an other language object
4383 * @param Language $lang
4386 public function equals( Language
$lang ) {
4387 return $lang === $this ||
$lang->getCode() === $this->mCode
;
4391 * Get the internal language code for this language object
4393 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4394 * htmlspecialchars() or similar
4398 public function getCode() {
4399 return $this->mCode
;
4403 * Get the code in BCP 47 format which we can use
4404 * inside of html lang="" tags.
4406 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4407 * htmlspecialchars() or similar.
4412 public function getHtmlCode() {
4413 if ( is_null( $this->mHtmlCode
) ) {
4414 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4416 return $this->mHtmlCode
;
4420 * Get the language code from a file name. Inverse of getFileName()
4421 * @param string $filename $prefix . $languageCode . $suffix
4422 * @param string $prefix Prefix before the language code
4423 * @param string $suffix Suffix after the language code
4424 * @return string Language code, or false if $prefix or $suffix isn't found
4426 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4428 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4429 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4430 if ( !count( $m ) ) {
4433 return str_replace( '_', '-', strtolower( $m[1] ) );
4437 * @param string $code
4438 * @param bool $fallback Whether we're going through language fallback chain
4439 * @return string Name of the language class
4441 public static function classFromCode( $code, $fallback = true ) {
4442 if ( $fallback && $code == 'en' ) {
4445 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4450 * Get the name of a file for a certain language code
4451 * @param string $prefix Prepend this to the filename
4452 * @param string $code Language code
4453 * @param string $suffix Append this to the filename
4454 * @throws MWException
4455 * @return string $prefix . $mangledCode . $suffix
4457 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4458 if ( !self
::isValidBuiltInCode( $code ) ) {
4459 throw new MWException( "Invalid language code \"$code\"" );
4462 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4466 * @param string $code
4469 public static function getMessagesFileName( $code ) {
4471 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4472 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4477 * @param string $code
4479 * @throws MWException
4482 public static function getJsonMessagesFileName( $code ) {
4485 if ( !self
::isValidBuiltInCode( $code ) ) {
4486 throw new MWException( "Invalid language code \"$code\"" );
4489 return "$IP/languages/i18n/$code.json";
4493 * Get the first fallback for a given language.
4495 * @param string $code
4497 * @return bool|string
4499 public static function getFallbackFor( $code ) {
4500 $fallbacks = self
::getFallbacksFor( $code );
4502 return $fallbacks[0];
4508 * Get the ordered list of fallback languages.
4511 * @param string $code Language code
4512 * @param int $mode Fallback mode, either MESSAGES_FALLBACKS (which always falls back to 'en'),
4513 * or STRICT_FALLBACKS (whic honly falls back to 'en' when explicitly defined)
4514 * @throws MWException
4515 * @return array List of language codes
4517 public static function getFallbacksFor( $code, $mode = self
::MESSAGES_FALLBACKS
) {
4518 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4522 case self
::MESSAGES_FALLBACKS
:
4523 // For unknown languages, fallbackSequence returns an empty array,
4524 // hardcode fallback to 'en' in that case as English messages are
4526 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4527 case self
::STRICT_FALLBACKS
:
4528 // Use this mode when you don't want to fallback to English unless
4529 // explicitly defined, for example when you have language-variant icons
4530 // and an international language-independent fallback.
4531 return self
::getLocalisationCache()->getItem( $code, 'originalFallbackSequence' );
4533 throw new MWException( "Invalid fallback mode \"$mode\"" );
4538 * Get the ordered list of fallback languages, ending with the fallback
4539 * language chain for the site language.
4542 * @param string $code Language code
4543 * @return array [ fallbacks, site fallbacks ]
4545 public static function getFallbacksIncludingSiteLanguage( $code ) {
4546 global $wgLanguageCode;
4548 // Usually, we will only store a tiny number of fallback chains, so we
4549 // keep them in static memory.
4550 $cacheKey = "{$code}-{$wgLanguageCode}";
4552 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4553 $fallbacks = self
::getFallbacksFor( $code );
4555 // Append the site's fallback chain, including the site language itself
4556 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4557 array_unshift( $siteFallbacks, $wgLanguageCode );
4559 // Eliminate any languages already included in the chain
4560 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4562 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4564 return self
::$fallbackLanguageCache[$cacheKey];
4568 * Get all messages for a given language
4569 * WARNING: this may take a long time. If you just need all message *keys*
4570 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4572 * @param string $code
4576 public static function getMessagesFor( $code ) {
4577 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4581 * Get a message for a given language
4583 * @param string $key
4584 * @param string $code
4588 public static function getMessageFor( $key, $code ) {
4589 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4593 * Get all message keys for a given language. This is a faster alternative to
4594 * array_keys( Language::getMessagesFor( $code ) )
4597 * @param string $code Language code
4598 * @return array Array of message keys (strings)
4600 public static function getMessageKeysFor( $code ) {
4601 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4605 * @param string $talk
4608 function fixVariableInNamespace( $talk ) {
4609 if ( strpos( $talk, '$1' ) === false ) {
4613 global $wgMetaNamespace;
4614 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4616 # Allow grammar transformations
4617 # Allowing full message-style parsing would make simple requests
4618 # such as action=raw much more expensive than they need to be.
4619 # This will hopefully cover most cases.
4620 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4621 [ $this, 'replaceGrammarInNamespace' ], $talk );
4622 return str_replace( ' ', '_', $talk );
4629 function replaceGrammarInNamespace( $m ) {
4630 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4634 * Decode an expiry (block, protection, etc) which has come from the DB
4636 * @param string $expiry Database expiry String
4637 * @param bool|int $format True to process using language functions, or TS_ constant
4638 * to return the expiry in a given timestamp
4639 * @param string $infinity If $format is not true, use this string for infinite expiry
4643 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4645 if ( $dbInfinity === null ) {
4646 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4649 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4650 return $format === true
4651 ?
$this->getMessageFromDB( 'infiniteblock' )
4654 return $format === true
4655 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4656 : wfTimestamp( $format, $expiry );
4661 * Formats a time given in seconds into a string representation of that time.
4663 * @param int|float $seconds
4664 * @param array $format An optional argument that formats the returned string in different ways:
4665 * If $format['avoid'] === 'avoidhours': don't show hours, just show days
4666 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4667 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4668 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4670 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4671 * or 'avoidminutes'.
4674 function formatTimePeriod( $seconds, $format = [] ) {
4675 if ( !is_array( $format ) ) {
4676 $format = [ 'avoid' => $format ]; // For backwards compatibility
4678 if ( !isset( $format['avoid'] ) ) {
4679 $format['avoid'] = false;
4681 if ( !isset( $format['noabbrevs'] ) ) {
4682 $format['noabbrevs'] = false;
4684 $secondsMsg = wfMessage(
4685 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4686 $minutesMsg = wfMessage(
4687 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4688 $hoursMsg = wfMessage(
4689 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4690 $daysMsg = wfMessage(
4691 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4693 if ( round( $seconds * 10 ) < 100 ) {
4694 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4695 $s = $secondsMsg->params( $s )->text();
4696 } elseif ( round( $seconds ) < 60 ) {
4697 $s = $this->formatNum( round( $seconds ) );
4698 $s = $secondsMsg->params( $s )->text();
4699 } elseif ( round( $seconds ) < 3600 ) {
4700 $minutes = floor( $seconds / 60 );
4701 $secondsPart = round( fmod( $seconds, 60 ) );
4702 if ( $secondsPart == 60 ) {
4706 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4708 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4709 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4710 $hours = floor( $seconds / 3600 );
4711 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4712 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4713 if ( $secondsPart == 60 ) {
4717 if ( $minutes == 60 ) {
4721 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4723 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4724 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes', 'avoidhours' ] ) ) {
4725 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4728 $days = floor( $seconds / 86400 );
4729 if ( $format['avoid'] === 'avoidhours' ) {
4730 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4731 if ( $hours == 24 ) {
4735 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4736 } elseif ( $format['avoid'] === 'avoidminutes' ) {
4737 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4738 if ( $hours == 24 ) {
4742 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4744 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4745 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4746 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4747 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4748 if ( $minutes == 60 ) {
4752 if ( $hours == 24 ) {
4756 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4758 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4760 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4762 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4764 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4771 * Format a bitrate for output, using an appropriate
4772 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4773 * the magnitude in question.
4775 * This use base 1000. For base 1024 use formatSize(), for another base
4776 * see formatComputingNumbers().
4781 function formatBitrate( $bps ) {
4782 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4786 * @param int $size Size of the unit
4787 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4788 * @param string $messageKey Message key to be uesd
4791 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4793 return str_replace( '$1', $this->formatNum( $size ),
4794 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4797 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4800 $maxIndex = count( $sizes ) - 1;
4801 while ( $size >= $boundary && $index < $maxIndex ) {
4806 // For small sizes no decimal places necessary
4809 // For MB and bigger two decimal places are smarter
4812 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4814 $size = round( $size, $round );
4815 $text = $this->getMessageFromDB( $msg );
4816 return str_replace( '$1', $this->formatNum( $size ), $text );
4820 * Format a size in bytes for output, using an appropriate
4821 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4823 * This method use base 1024. For base 1000 use formatBitrate(), for
4824 * another base see formatComputingNumbers()
4826 * @param int $size Size to format
4827 * @return string Plain text (not HTML)
4829 function formatSize( $size ) {
4830 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4834 * Make a list item, used by various special pages
4836 * @param string $page Page link
4837 * @param string $details HTML safe text between brackets
4838 * @param bool $oppositedm Add the direction mark opposite to your
4839 * language, to display text properly
4840 * @return string HTML escaped
4842 function specialList( $page, $details, $oppositedm = true ) {
4847 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4850 $this->msg( 'word-separator' )->escaped() .
4851 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4855 * Generate (prev x| next x) (20|50|100...) type links for paging
4857 * @param Title $title Title object to link
4858 * @param int $offset
4860 * @param array $query Optional URL query parameter string
4861 * @param bool $atend Optional param for specified if this is the last page
4863 * @deprecated since 1.34, use PrevNextNavigationRenderer::buildPrevNextNavigation()
4866 public function viewPrevNext( Title
$title, $offset, $limit,
4867 array $query = [], $atend = false
4869 wfDeprecated( __METHOD__
, '1.34' );
4870 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4872 # Make 'previous' link
4873 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4874 if ( $offset > 0 ) {
4875 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4876 $query, $prev, 'prevn-title', 'mw-prevlink' );
4878 $plink = htmlspecialchars( $prev );
4882 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4884 $nlink = htmlspecialchars( $next );
4886 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4887 $query, $next, 'nextn-title', 'mw-nextlink' );
4890 # Make links to set number of items per page
4892 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4893 $numLinks[] = $this->numLink( $title, $offset, $num,
4894 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4897 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4898 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4902 * Helper function for viewPrevNext() that generates links
4904 * @param Title $title Title object to link
4905 * @param int $offset
4907 * @param array $query Extra query parameters
4908 * @param string $link Text to use for the link; will be escaped
4909 * @param string $tooltipMsg Name of the message to use as tooltip
4910 * @param string $class Value of the "class" attribute of the link
4911 * @return string HTML fragment
4913 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4916 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4917 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4918 ->numParams( $limit )->text();
4920 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4921 'title' => $tooltip, 'class' => $class ], $link );
4925 * Get the conversion rule title, if any.
4929 public function getConvRuleTitle() {
4930 return $this->mConverter
->getConvRuleTitle();
4934 * Get the compiled plural rules for the language
4936 * @return array Associative array with plural form, and plural rule as key-value pairs
4938 public function getCompiledPluralRules() {
4939 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4940 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4941 if ( !$pluralRules ) {
4942 foreach ( $fallbacks as $fallbackCode ) {
4943 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4944 if ( $pluralRules ) {
4949 return $pluralRules;
4953 * Get the plural rules for the language
4955 * @return array Associative array with plural form number and plural rule as key-value pairs
4957 public function getPluralRules() {
4958 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4959 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4960 if ( !$pluralRules ) {
4961 foreach ( $fallbacks as $fallbackCode ) {
4962 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4963 if ( $pluralRules ) {
4968 return $pluralRules;
4972 * Get the plural rule types for the language
4974 * @return array Associative array with plural form number and plural rule type as key-value pairs
4976 public function getPluralRuleTypes() {
4977 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4978 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4979 if ( !$pluralRuleTypes ) {
4980 foreach ( $fallbacks as $fallbackCode ) {
4981 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4982 if ( $pluralRuleTypes ) {
4987 return $pluralRuleTypes;
4991 * Find the index number of the plural rule appropriate for the given number
4992 * @param int $number
4993 * @return int The index number of the plural rule
4995 public function getPluralRuleIndexNumber( $number ) {
4996 $pluralRules = $this->getCompiledPluralRules();
4997 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
5002 * Find the plural rule type appropriate for the given number
5003 * For example, if the language is set to Arabic, getPluralType(5) should
5006 * @param int $number
5007 * @return string The name of the plural rule type, e.g. one, two, few, many
5009 public function getPluralRuleType( $number ) {
5010 $index = $this->getPluralRuleIndexNumber( $number );
5011 $pluralRuleTypes = $this->getPluralRuleTypes();
5012 return $pluralRuleTypes[$index] ??
'other';