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
62 public $mVariants, $mCode, $mLoaded = false;
63 public $mMagicExtensions = [];
64 private $mHtmlCode = null, $mParentLanguage = false;
66 public $dateFormatStrings = [];
67 public $mExtendedSpecialPageAliases;
69 /** @var array|null */
70 protected $namespaceNames;
71 protected $mNamespaceIds, $namespaceAliases;
74 * ReplacementArray object caches
76 public $transformData = [];
79 * @var LocalisationCache
81 public static $dataCache;
83 public static $mLangObjCache = [];
86 * Return a fallback chain for messages in getFallbacksFor
89 const MESSAGES_FALLBACKS
= 0;
92 * Return a strict fallback chain in getFallbacksFor
95 const STRICT_FALLBACKS
= 1;
97 public static $mWeekdayMsgs = [
98 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
102 public static $mWeekdayAbbrevMsgs = [
103 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
106 public static $mMonthMsgs = [
107 'january', 'february', 'march', 'april', 'may_long', 'june',
108 'july', 'august', 'september', 'october', 'november',
111 public static $mMonthGenMsgs = [
112 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
113 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
116 public static $mMonthAbbrevMsgs = [
117 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
118 'sep', 'oct', 'nov', 'dec'
121 public static $mIranianCalendarMonthMsgs = [
122 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
123 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
124 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
125 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
128 public static $mHebrewCalendarMonthMsgs = [
129 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
130 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
131 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
132 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
133 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
136 public static $mHebrewCalendarMonthGenMsgs = [
137 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
138 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
139 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
140 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
141 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
144 public static $mHijriCalendarMonthMsgs = [
145 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
146 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
147 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
148 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
155 public static $durationIntervals = [
156 'millennia' => 31556952000,
157 'centuries' => 3155695200,
158 'decades' => 315569520,
159 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
168 * Cache for language fallbacks.
169 * @see Language::getFallbacksIncludingSiteLanguage
173 private static $fallbackLanguageCache = [];
176 * Cache for grammar rules data
177 * @var MapCacheLRU|null
179 private static $grammarTransformations;
182 * Cache for language names
183 * @var HashBagOStuff|null
185 private static $languageNameCache;
188 * Unicode directional formatting characters, for embedBidi()
190 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
191 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
192 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
195 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
196 * - in group 1 if it is LTR
197 * - in group 2 if it is RTL
198 * Does not match if there is no strong directionality codepoint.
200 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
202 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
203 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
205 // @codeCoverageIgnoreStart
206 // phpcs:ignore Generic.Files.LineLength
207 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';
208 // @codeCoverageIgnoreEnd
211 * Get a cached or new language object for a given language code
212 * @param string $code
213 * @throws MWException
216 static function factory( $code ) {
217 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
219 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
220 $code = $wgDummyLanguageCodes[$code];
223 // get the language object to process
224 $langObj = self
::$mLangObjCache[$code] ?? self
::newFromCode( $code );
226 // merge the language object in to get it up front in the cache
227 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
228 // get rid of the oldest ones in case we have an overflow
229 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
235 * Create a language object for a given language code
236 * @param string $code
237 * @param bool $fallback Whether we're going through language fallback chain
238 * @throws MWException
241 protected static function newFromCode( $code, $fallback = false ) {
242 if ( !self
::isValidCode( $code ) ) {
243 throw new MWException( "Invalid language code \"$code\"" );
246 if ( !self
::isValidBuiltInCode( $code ) ) {
247 // It's not possible to customise this code with class files, so
248 // just return a Language object. This is to support uselang= hacks.
249 $lang = new Language
;
250 $lang->mCode
= $code;
254 // Check if there is a language class for the code
255 $class = self
::classFromCode( $code, $fallback );
256 // LanguageCode does not inherit Language
257 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
262 // Keep trying the fallback list until we find an existing class
263 $fallbacks = self
::getFallbacksFor( $code );
264 foreach ( $fallbacks as $fallbackCode ) {
265 if ( !self
::isValidBuiltInCode( $fallbackCode ) ) {
266 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
269 $class = self
::classFromCode( $fallbackCode );
270 if ( class_exists( $class ) ) {
272 $lang->mCode
= $code;
277 throw new MWException( "Invalid fallback sequence for language '$code'" );
281 * Intended for tests that may change configuration in a way that invalidates caches.
285 public static function clearCaches() {
286 if ( !defined( 'MW_PHPUNIT_TEST' ) ) {
287 throw new MWException( __METHOD__
. ' must not be used outside tests' );
289 self
::$dataCache = null;
290 // Reinitialize $dataCache, since it's expected to always be available
291 self
::getLocalisationCache();
292 self
::$mLangObjCache = [];
293 self
::$fallbackLanguageCache = [];
294 self
::$grammarTransformations = null;
295 self
::$languageNameCache = null;
299 * Checks whether any localisation is available for that language tag
300 * in MediaWiki (MessagesXx.php exists).
302 * @param string $code Language tag (in lower case)
303 * @return bool Whether language is supported
306 public static function isSupportedLanguage( $code ) {
307 if ( !self
::isValidBuiltInCode( $code ) ) {
311 if ( $code === 'qqq' ) {
315 return is_readable( self
::getMessagesFileName( $code ) ) ||
316 is_readable( self
::getJsonMessagesFileName( $code ) );
320 * Returns true if a language code string is a well-formed language tag
321 * according to RFC 5646.
322 * This function only checks well-formedness; it doesn't check that
323 * language, script or variant codes actually exist in the repositories.
325 * Based on regexes by Mark Davis of the Unicode Consortium:
326 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
328 * @param string $code
329 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
334 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
337 $alphanum = '[a-z0-9]';
338 $x = 'x'; # private use singleton
339 $singleton = '[a-wy-z]'; # other singleton
340 $s = $lenient ?
'[-_]' : '-';
342 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
343 $script = "$alpha{4}"; # ISO 15924
344 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
345 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
346 $extension = "$singleton(?:$s$alphanum{2,8})+";
347 $privateUse = "$x(?:$s$alphanum{1,8})+";
349 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
350 # Since these are limited, this is safe even later changes to the registry --
351 # the only oddity is that it might change the type of the tag, and thus
352 # the results from the capturing groups.
353 # https://www.iana.org/assignments/language-subtag-registry
355 $grandfathered = "en{$s}GB{$s}oed"
356 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
357 . "|no{$s}(?:bok|nyn)"
358 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
359 . "|zh{$s}min{$s}nan";
361 $variantList = "$variant(?:$s$variant)*";
362 $extensionList = "$extension(?:$s$extension)*";
364 $langtag = "(?:($language)"
367 . "(?:$s$variantList)?"
368 . "(?:$s$extensionList)?"
369 . "(?:$s$privateUse)?)";
371 # The final breakdown, with capturing groups for each of these components
372 # The variants, extensions, grandfathered, and private-use may have interior '-'
374 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
376 return (bool)preg_match( "/$root/", strtolower( $code ) );
380 * Returns true if a language code string is of a valid form, whether or
381 * not it exists. This includes codes which are used solely for
382 * customisation via the MediaWiki namespace.
384 * @param string $code
388 public static function isValidCode( $code ) {
390 Assert
::parameterType( 'string', $code, '$code' );
391 if ( !isset( $cache[$code] ) ) {
392 // People think language codes are html safe, so enforce it.
393 // Ideally we should only allow a-zA-Z0-9-
394 // but, .+ and other chars are often used for {{int:}} hacks
395 // see bugs T39564, T39587, T38938
397 // Protect against path traversal
398 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
399 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
401 return $cache[$code];
405 * Returns true if a language code is of a valid form for the purposes of
406 * internal customisation of MediaWiki, via Messages*.php or *.json.
408 * @param string $code
413 public static function isValidBuiltInCode( $code ) {
414 Assert
::parameterType( 'string', $code, '$code' );
416 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
420 * Returns true if a language code is an IETF tag known to MediaWiki.
427 public static function isKnownLanguageTag( $tag ) {
428 // Quick escape for invalid input to avoid exceptions down the line
429 // when code tries to process tags which are not valid at all.
430 if ( !self
::isValidBuiltInCode( $tag ) ) {
434 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
435 || self
::fetchLanguageName( $tag, $tag ) !== ''
444 * Get the LocalisationCache instance
446 * @return LocalisationCache
448 public static function getLocalisationCache() {
449 if ( is_null( self
::$dataCache ) ) {
450 global $wgLocalisationCacheConf;
451 $class = $wgLocalisationCacheConf['class'];
452 self
::$dataCache = new $class( $wgLocalisationCacheConf );
454 return self
::$dataCache;
457 function __construct() {
458 $this->mConverter
= new FakeConverter( $this );
459 // Set the code to the name of the descendant
460 if ( static::class === 'Language' ) {
463 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
465 self
::getLocalisationCache();
469 * Reduce memory usage
470 * @suppress PhanTypeSuspiciousNonTraversableForeach
472 function __destruct() {
473 foreach ( $this as $name => $value ) {
474 unset( $this->$name );
479 * Hook which will be called if this is the content language.
480 * Descendants can use this to register hook functions or modify globals
482 function initContLang() {
489 public function getFallbackLanguages() {
490 return self
::getFallbacksFor( $this->mCode
);
494 * Exports $wgBookstoreListEn
497 public function getBookstoreList() {
498 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
502 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
503 * available in localised form, it will be included in English.
505 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
507 public function getNamespaces() {
508 if ( is_null( $this->namespaceNames
) ) {
509 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
511 $validNamespaces = MediaWikiServices
::getInstance()->getNamespaceInfo()->
512 getCanonicalNamespaces();
514 $this->namespaceNames
= $wgExtraNamespaces +
515 self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
516 $this->namespaceNames +
= $validNamespaces;
518 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
519 if ( $wgMetaNamespaceTalk ) {
520 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
522 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
523 $this->namespaceNames
[NS_PROJECT_TALK
] =
524 $this->fixVariableInNamespace( $talk );
527 # Sometimes a language will be localised but not actually exist on this wiki.
528 // @phan-suppress-next-line PhanTypeMismatchForeach
529 foreach ( $this->namespaceNames
as $key => $text ) {
530 if ( !isset( $validNamespaces[$key] ) ) {
531 unset( $this->namespaceNames
[$key] );
535 # The above mixing may leave namespaces out of canonical order.
536 # Re-order by namespace ID number...
537 ksort( $this->namespaceNames
);
539 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
542 return $this->namespaceNames
;
546 * Arbitrarily set all of the namespace names at once. Mainly used for testing
547 * @param array $namespaces Array of namespaces (id => name)
549 public function setNamespaces( array $namespaces ) {
550 $this->namespaceNames
= $namespaces;
551 $this->mNamespaceIds
= null;
555 * Resets all of the namespace caches. Mainly used for testing
557 public function resetNamespaces() {
558 $this->namespaceNames
= null;
559 $this->mNamespaceIds
= null;
560 $this->namespaceAliases
= null;
564 * A convenience function that returns getNamespaces() with spaces instead of underscores
565 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
569 public function getFormattedNamespaces() {
570 $ns = $this->getNamespaces();
571 foreach ( $ns as $k => $v ) {
572 $ns[$k] = strtr( $v, '_', ' ' );
578 * Get a namespace value by key
581 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
582 * echo $mw_ns; // prints 'MediaWiki'
585 * @param int $index The array key of the namespace to return
586 * @return string|bool String if the namespace value exists, otherwise false
588 public function getNsText( $index ) {
589 $ns = $this->getNamespaces();
590 return $ns[$index] ??
false;
594 * A convenience function that returns the same thing as
595 * getNsText() except with '_' changed to ' ', useful for
599 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
600 * echo $mw_ns; // prints 'MediaWiki talk'
603 * @param int $index The array key of the namespace to return
604 * @return string Namespace name without underscores (empty string if namespace does not exist)
606 public function getFormattedNsText( $index ) {
607 $ns = $this->getNsText( $index );
608 return strtr( $ns, '_', ' ' );
612 * Returns gender-dependent namespace alias if available.
613 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
614 * @param int $index Namespace index
615 * @param string $gender Gender key (male, female... )
619 public function getGenderNsText( $index, $gender ) {
620 global $wgExtraGenderNamespaces;
622 $ns = $wgExtraGenderNamespaces +
623 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
625 return $ns[$index][$gender] ??
$this->getNsText( $index );
629 * Whether this language uses gender-dependent namespace aliases.
630 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
634 public function needsGenderDistinction() {
635 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
636 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
637 // $wgExtraGenderNamespaces overrides everything
639 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
640 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
641 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
644 // Check what is in i18n files
645 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
646 return count( $aliases ) > 0;
651 * Get a namespace key by value, case insensitive.
652 * Only matches namespace names for the current language, not the
653 * canonical ones defined in Namespace.php.
655 * @param string $text
656 * @return int|bool An integer if $text is a valid value otherwise false
658 function getLocalNsIndex( $text ) {
659 $lctext = $this->lc( $text );
660 $ids = $this->getNamespaceIds();
661 return $ids[$lctext] ??
false;
667 public function getNamespaceAliases() {
668 if ( is_null( $this->namespaceAliases
) ) {
669 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
673 foreach ( $aliases as $name => $index ) {
674 if ( $index === NS_PROJECT_TALK
) {
675 unset( $aliases[$name] );
676 $name = $this->fixVariableInNamespace( $name );
677 $aliases[$name] = $index;
682 global $wgExtraGenderNamespaces;
683 $genders = $wgExtraGenderNamespaces +
684 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
685 foreach ( $genders as $index => $forms ) {
686 foreach ( $forms as $alias ) {
687 $aliases[$alias] = $index;
691 # Also add converted namespace names as aliases, to avoid confusion.
692 $convertedNames = [];
693 foreach ( $this->getVariants() as $variant ) {
694 if ( $variant === $this->mCode
) {
697 foreach ( $this->getNamespaces() as $ns => $_ ) {
698 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
702 $this->namespaceAliases
= $aliases +
$convertedNames;
704 # Filter out aliases to namespaces that don't exist, e.g. from extensions
705 # that aren't loaded here but are included in the l10n cache.
706 # (array_intersect preserves keys from its first argument)
707 $this->namespaceAliases
= array_intersect(
708 $this->namespaceAliases
,
709 array_keys( $this->getNamespaces() )
713 return $this->namespaceAliases
;
719 public function getNamespaceIds() {
720 if ( is_null( $this->mNamespaceIds
) ) {
721 global $wgNamespaceAliases;
722 # Put namespace names and aliases into a hashtable.
723 # If this is too slow, then we should arrange it so that it is done
724 # before caching. The catch is that at pre-cache time, the above
725 # class-specific fixup hasn't been done.
726 $this->mNamespaceIds
= [];
727 foreach ( $this->getNamespaces() as $index => $name ) {
728 $this->mNamespaceIds
[$this->lc( $name )] = $index;
730 foreach ( $this->getNamespaceAliases() as $name => $index ) {
731 $this->mNamespaceIds
[$this->lc( $name )] = $index;
733 if ( $wgNamespaceAliases ) {
734 foreach ( $wgNamespaceAliases as $name => $index ) {
735 $this->mNamespaceIds
[$this->lc( $name )] = $index;
739 return $this->mNamespaceIds
;
743 * Get a namespace key by value, case insensitive. Canonical namespace
744 * names override custom ones defined for the current language.
746 * @param string $text
747 * @return int|bool An integer if $text is a valid value otherwise false
749 public function getNsIndex( $text ) {
750 $lctext = $this->lc( $text );
751 $ns = MediaWikiServices
::getInstance()->getNamespaceInfo()->
752 getCanonicalIndex( $lctext );
753 if ( $ns !== null ) {
756 $ids = $this->getNamespaceIds();
757 return $ids[$lctext] ??
false;
761 * short names for language variants used for language conversion links.
763 * @param string $code
764 * @param bool $usemsg Use the "variantname-xyz" message if it exists
767 public function getVariantname( $code, $usemsg = true ) {
768 $msg = "variantname-$code";
769 if ( $usemsg && wfMessage( $msg )->exists() ) {
770 return $this->getMessageFromDB( $msg );
772 $name = self
::fetchLanguageName( $code );
774 return $name; # if it's defined as a language name, show that
776 # otherwise, output the language code
782 * @return string[]|bool List of date format preference keys, or false if disabled.
784 public function getDatePreferences() {
785 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
791 function getDateFormats() {
792 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
796 * @return array|string
798 public function getDefaultDateFormat() {
799 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
800 if ( $df === 'dmy or mdy' ) {
801 global $wgAmericanDates;
802 return $wgAmericanDates ?
'mdy' : 'dmy';
811 public function getDatePreferenceMigrationMap() {
812 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
818 public function getExtraUserToggles() {
819 wfDeprecated( __METHOD__
, '1.34' );
827 function getUserToggle( $tog ) {
828 return $this->getMessageFromDB( "tog-$tog" );
832 * Get an array of language names, indexed by code.
833 * @param null|string $inLanguage Code of language in which to return the names
834 * Use self::AS_AUTONYMS for autonyms (native names)
835 * @param string $include One of:
836 * self::ALL all available languages
837 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
838 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
839 * @return array Language code => language name (sorted by key)
842 public static function fetchLanguageNames( $inLanguage = self
::AS_AUTONYMS
, $include = 'mw' ) {
843 $cacheKey = $inLanguage === self
::AS_AUTONYMS ?
'null' : $inLanguage;
844 $cacheKey .= ":$include";
845 if ( self
::$languageNameCache === null ) {
846 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
849 $ret = self
::$languageNameCache->get( $cacheKey );
851 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
852 self
::$languageNameCache->set( $cacheKey, $ret );
858 * Uncached helper for fetchLanguageNames
859 * @param null|string $inLanguage Code of language in which to return the names
860 * Use self::AS_AUTONYMS for autonyms (native names)
861 * @param string $include One of:
862 * self::ALL all available languages
863 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
864 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
865 * @return array Language code => language name (sorted by key)
867 private static function fetchLanguageNamesUncached(
868 $inLanguage = self
::AS_AUTONYMS
,
871 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
873 // If passed an invalid language code to use, fallback to en
874 if ( $inLanguage !== self
::AS_AUTONYMS
&& !self
::isValidCode( $inLanguage ) ) {
881 # TODO: also include when $inLanguage is null, when this code is more efficient
882 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
885 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
886 if ( $wgUsePigLatinVariant ) {
887 // Pig Latin (for variant development)
888 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
891 foreach ( $mwNames as $mwCode => $mwName ) {
892 # - Prefer own MediaWiki native name when not using the hook
893 # - For other names just add if not added through the hook
894 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
895 $names[$mwCode] = $mwName;
899 if ( $include === self
::ALL
) {
905 $coreCodes = array_keys( $mwNames );
906 foreach ( $coreCodes as $coreCode ) {
907 $returnMw[$coreCode] = $names[$coreCode];
910 if ( $include === self
::SUPPORTED
) {
912 # We do this using a foreach over the codes instead of a directory
913 # loop so that messages files in extensions will work correctly.
914 foreach ( $returnMw as $code => $value ) {
915 if ( is_readable( self
::getMessagesFileName( $code ) )
916 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
918 $namesMwFile[$code] = $names[$code];
922 ksort( $namesMwFile );
927 # 'mw' option; default if it's not one of the other two options (all/mwfile)
932 * @param string $code The code of the language for which to get the name
933 * @param null|string $inLanguage Code of language in which to return the name
934 * (SELF::AS_AUTONYMS for autonyms)
935 * @param string $include See fetchLanguageNames()
936 * @return string Language name or empty
939 public static function fetchLanguageName(
941 $inLanguage = self
::AS_AUTONYMS
,
944 $code = strtolower( $code );
945 $array = self
::fetchLanguageNames( $inLanguage, $include );
946 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
950 * Get a message from the MediaWiki namespace.
952 * @param string $msg Message name
955 public function getMessageFromDB( $msg ) {
956 return $this->msg( $msg )->text();
960 * Get message object in this language. Only for use inside this class.
962 * @param string $msg Message name
965 protected function msg( $msg ) {
966 return wfMessage( $msg )->inLanguage( $this );
973 public function getMonthName( $key ) {
974 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
980 public function getMonthNamesArray() {
981 $monthNames = [ '' ];
982 for ( $i = 1; $i < 13; $i++
) {
983 $monthNames[] = $this->getMonthName( $i );
992 public function getMonthNameGen( $key ) {
993 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1000 public function getMonthAbbreviation( $key ) {
1001 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1007 public function getMonthAbbreviationsArray() {
1008 $monthNames = [ '' ];
1009 for ( $i = 1; $i < 13; $i++
) {
1010 $monthNames[] = $this->getMonthAbbreviation( $i );
1016 * @param string $key
1019 public function getWeekdayName( $key ) {
1020 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1024 * @param string $key
1027 function getWeekdayAbbreviation( $key ) {
1028 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1032 * @param string $key
1035 function getIranianCalendarMonthName( $key ) {
1036 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1040 * @param string $key
1043 function getHebrewCalendarMonthName( $key ) {
1044 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1048 * @param string $key
1051 function getHebrewCalendarMonthNameGen( $key ) {
1052 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1056 * @param string $key
1059 function getHijriCalendarMonthName( $key ) {
1060 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1064 * Pass through result from $dateTimeObj->format()
1065 * @param DateTime|bool|null &$dateTimeObj
1067 * @param DateTimeZone|bool|null $zone
1068 * @param string $code
1071 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1072 if ( !$dateTimeObj ) {
1073 $dateTimeObj = DateTime
::createFromFormat(
1074 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1077 return $dateTimeObj->format( $code );
1081 * This is a workalike of PHP's date() function, but with better
1082 * internationalisation, a reduced set of format characters, and a better
1085 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1086 * the PHP manual for definitions. There are a number of extensions, which
1089 * xn Do not translate digits of the next numeric format character
1090 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1091 * xr Use roman numerals for the next numeric format character
1092 * xh Use hebrew numerals for the next numeric format character
1094 * xg Genitive month name
1096 * xij j (day number) in Iranian calendar
1097 * xiF F (month name) in Iranian calendar
1098 * xin n (month number) in Iranian calendar
1099 * xiy y (two digit year) in Iranian calendar
1100 * xiY Y (full year) in Iranian calendar
1101 * xit t (days in month) in Iranian calendar
1102 * xiz z (day of the year) in Iranian calendar
1104 * xjj j (day number) in Hebrew calendar
1105 * xjF F (month name) in Hebrew calendar
1106 * xjt t (days in month) in Hebrew calendar
1107 * xjx xg (genitive month name) in Hebrew calendar
1108 * xjn n (month number) in Hebrew calendar
1109 * xjY Y (full year) in Hebrew calendar
1111 * xmj j (day number) in Hijri calendar
1112 * xmF F (month name) in Hijri calendar
1113 * xmn n (month number) in Hijri calendar
1114 * xmY Y (full year) in Hijri calendar
1116 * xkY Y (full year) in Thai solar calendar. Months and days are
1117 * identical to the Gregorian calendar
1118 * xoY Y (full year) in Minguo calendar or Juche year.
1119 * Months and days are identical to the
1120 * Gregorian calendar
1121 * xtY Y (full year) in Japanese nengo. Months and days are
1122 * identical to the Gregorian calendar
1124 * Characters enclosed in double quotes will be considered literal (with
1125 * the quotes themselves removed). Unmatched quotes will be considered
1126 * literal quotes. Example:
1128 * "The month is" F => The month is January
1131 * Backslash escaping is also supported.
1133 * Input timestamp is assumed to be pre-normalized to the desired local
1134 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1135 * $ts is UTC if $zone is not given.
1137 * @param string $format
1138 * @param string $ts 14-character timestamp
1141 * @param DateTimeZone|null $zone Timezone of $ts
1142 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1143 * Only makes sense if $ts is the current time.
1144 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1146 * @throws MWException
1149 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1154 $dateTimeObj = false;
1163 $usedSecond = false;
1164 $usedMinute = false;
1171 $usedISOYear = false;
1172 $usedIsLeapYear = false;
1174 $usedHebrewMonth = false;
1175 $usedIranianMonth = false;
1176 $usedHijriMonth = false;
1177 $usedHebrewYear = false;
1178 $usedIranianYear = false;
1179 $usedHijriYear = false;
1180 $usedTennoYear = false;
1182 if ( strlen( $ts ) !== 14 ) {
1183 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1186 if ( !ctype_digit( $ts ) ) {
1187 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1190 $formatLength = strlen( $format );
1191 for ( $p = 0; $p < $formatLength; $p++
) {
1193 $code = $format[$p];
1194 if ( $code == 'x' && $p < $formatLength - 1 ) {
1195 $code .= $format[++
$p];
1198 if ( ( $code === 'xi'
1204 && $p < $formatLength - 1 ) {
1205 $code .= $format[++
$p];
1216 $rawToggle = !$rawToggle;
1226 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1229 $usedHebrewMonth = true;
1231 $hebrew = self
::tsToHebrew( $ts );
1233 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1237 $num = substr( $ts, 6, 2 );
1241 $s .= $this->getWeekdayAbbreviation(
1242 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1247 $num = intval( substr( $ts, 6, 2 ) );
1252 $iranian = self
::tsToIranian( $ts );
1259 $hijri = self
::tsToHijri( $ts );
1266 $hebrew = self
::tsToHebrew( $ts );
1272 $s .= $this->getWeekdayName(
1273 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1278 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1281 $usedIranianMonth = true;
1283 $iranian = self
::tsToIranian( $ts );
1285 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1288 $usedHijriMonth = true;
1290 $hijri = self
::tsToHijri( $ts );
1292 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1295 $usedHebrewMonth = true;
1297 $hebrew = self
::tsToHebrew( $ts );
1299 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1303 $num = substr( $ts, 4, 2 );
1307 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1311 $num = intval( substr( $ts, 4, 2 ) );
1314 $usedIranianMonth = true;
1316 $iranian = self
::tsToIranian( $ts );
1321 $usedHijriMonth = true;
1323 $hijri = self
::tsToHijri( $ts );
1328 $usedHebrewMonth = true;
1330 $hebrew = self
::tsToHebrew( $ts );
1335 $usedHebrewMonth = true;
1337 $hebrew = self
::tsToHebrew( $ts );
1343 $num = substr( $ts, 0, 4 );
1346 $usedIranianYear = true;
1348 $iranian = self
::tsToIranian( $ts );
1353 $usedHijriYear = true;
1355 $hijri = self
::tsToHijri( $ts );
1360 $usedHebrewYear = true;
1362 $hebrew = self
::tsToHebrew( $ts );
1369 $thai = self
::tsToYear( $ts, 'thai' );
1376 $minguo = self
::tsToYear( $ts, 'minguo' );
1381 $usedTennoYear = true;
1383 $tenno = self
::tsToYear( $ts, 'tenno' );
1389 $num = substr( $ts, 2, 2 );
1392 $usedIranianYear = true;
1394 $iranian = self
::tsToIranian( $ts );
1396 $num = substr( $iranian[0], -2 );
1399 $usedIranianYear = true;
1401 $iranian = self
::tsToIranian( $ts );
1403 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1406 $usedIranianYear = true;
1408 $iranian = self
::tsToIranian( $ts );
1414 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1418 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1422 $h = substr( $ts, 8, 2 );
1423 $num = $h %
12 ?
: 12;
1427 $num = intval( substr( $ts, 8, 2 ) );
1431 $h = substr( $ts, 8, 2 );
1432 $num = sprintf( '%02d', $h %
12 ?
: 12 );
1436 $num = substr( $ts, 8, 2 );
1440 $num = substr( $ts, 10, 2 );
1444 $num = substr( $ts, 12, 2 );
1454 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1460 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1464 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1468 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1471 $usedIsLeapYear = true;
1472 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1475 $usedISOYear = true;
1476 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1483 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1486 # Backslash escaping
1487 if ( $p < $formatLength - 1 ) {
1488 $s .= $format[++
$p];
1495 if ( $p < $formatLength - 1 ) {
1496 $endQuote = strpos( $format, '"', $p +
1 );
1497 if ( $endQuote === false ) {
1498 # No terminating quote, assume literal "
1501 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1505 # Quote at end of string, assume literal "
1512 if ( $num !== false ) {
1513 if ( $rawToggle ||
$raw ) {
1516 } elseif ( $roman ) {
1517 $s .= self
::romanNumeral( $num );
1519 } elseif ( $hebrewNum ) {
1520 $s .= self
::hebrewNumeral( $num );
1523 $s .= $this->formatNum( $num, true );
1528 if ( $ttl === 'unused' ) {
1529 // No need to calculate the TTL, the caller wont use it anyway.
1530 } elseif ( $usedSecond ) {
1532 } elseif ( $usedMinute ) {
1533 $ttl = 60 - substr( $ts, 12, 2 );
1534 } elseif ( $usedHour ) {
1535 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1536 } elseif ( $usedAMPM ) {
1537 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1538 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1542 $usedIranianMonth ||
1549 // @todo Someone who understands the non-Gregorian calendars
1550 // should write proper logic for them so that they don't need purged every day.
1551 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1552 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1555 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1556 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1559 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1560 $timeRemainingInDay;
1561 } elseif ( $usedISOYear ) {
1562 // December 28th falls on the last ISO week of the year, every year.
1563 // The last ISO week of a year can be 52 or 53.
1564 $lastWeekOfISOYear = DateTime
::createFromFormat(
1566 substr( $ts, 0, 4 ) . '1228',
1567 $zone ?
: new DateTimeZone( 'UTC' )
1569 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1570 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1571 $timeRemainingInWeek =
1572 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1573 +
$timeRemainingInDay;
1574 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1579 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1580 substr( $ts, 6, 2 ) ) * 86400
1581 +
$timeRemainingInDay;
1582 } elseif ( $usedYear ) {
1584 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1585 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1586 +
$timeRemainingInDay;
1587 } elseif ( $usedIsLeapYear ) {
1588 $year = substr( $ts, 0, 4 );
1589 $timeRemainingInYear =
1590 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1591 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1592 +
$timeRemainingInDay;
1594 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1595 // this isn't a leap year. see when the next one starts
1596 $nextCandidate = $year - $mod +
4;
1597 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1598 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1599 $timeRemainingInYear;
1601 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1602 $timeRemainingInYear;
1605 // this is a leap year, so the next year isn't
1606 $possibleTtls[] = $timeRemainingInYear;
1610 if ( $possibleTtls ) {
1611 $ttl = min( $possibleTtls );
1618 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1619 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1622 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1623 * Gregorian dates to Iranian dates. Originally written in C, it
1624 * is released under the terms of GNU Lesser General Public
1625 * License. Conversion to PHP was performed by Niklas Laxström.
1627 * Link: http://www.farsiweb.info/jalali/jalali.c
1633 private static function tsToIranian( $ts ) {
1634 $gy = substr( $ts, 0, 4 ) - 1600;
1635 $gm = substr( $ts, 4, 2 ) - 1;
1636 $gd = substr( $ts, 6, 2 ) - 1;
1638 # Days passed from the beginning (including leap years)
1640 +
floor( ( $gy +
3 ) / 4 )
1641 - floor( ( $gy +
99 ) / 100 )
1642 +
floor( ( $gy +
399 ) / 400 );
1644 // Add days of the past months of this year
1645 for ( $i = 0; $i < $gm; $i++
) {
1646 $gDayNo +
= self
::$GREG_DAYS[$i];
1650 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1654 // Days passed in current month
1655 $gDayNo +
= (int)$gd;
1657 $jDayNo = $gDayNo - 79;
1659 $jNp = floor( $jDayNo / 12053 );
1662 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1665 if ( $jDayNo >= 366 ) {
1666 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1667 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1672 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1673 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1679 return [ $jy, $jm, $jd, $jz ];
1683 * Converting Gregorian dates to Hijri dates.
1685 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1687 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1693 private static function tsToHijri( $ts ) {
1694 $year = substr( $ts, 0, 4 );
1695 $month = substr( $ts, 4, 2 );
1696 $day = substr( $ts, 6, 2 );
1704 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1705 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1707 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1708 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1709 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1712 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1713 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1716 $zl = $zjd - 1948440 +
10632;
1717 $zn = (int)( ( $zl - 1 ) / 10631 );
1718 $zl = $zl - 10631 * $zn +
354;
1719 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1720 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1721 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1722 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1723 $zm = (int)( ( 24 * $zl ) / 709 );
1724 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1725 $zy = 30 * $zn +
$zj - 30;
1727 return [ $zy, $zm, $zd ];
1731 * Converting Gregorian dates to Hebrew dates.
1733 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1734 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1735 * to translate the relevant functions into PHP and release them under
1738 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1739 * and Adar II is 14. In a non-leap year, Adar is 6.
1745 private static function tsToHebrew( $ts ) {
1747 $year = substr( $ts, 0, 4 );
1748 $month = substr( $ts, 4, 2 );
1749 $day = substr( $ts, 6, 2 );
1751 # Calculate Hebrew year
1752 $hebrewYear = $year +
3760;
1754 # Month number when September = 1, August = 12
1756 if ( $month > 12 ) {
1763 # Calculate day of year from 1 September
1765 for ( $i = 1; $i < $month; $i++
) {
1769 # Check if the year is leap
1770 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1773 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1780 # Calculate the start of the Hebrew year
1781 $start = self
::hebrewYearStart( $hebrewYear );
1783 # Calculate next year's start
1784 if ( $dayOfYear <= $start ) {
1785 # Day is before the start of the year - it is the previous year
1787 $nextStart = $start;
1791 # Add days since previous year's 1 September
1793 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1797 # Start of the new (previous) year
1798 $start = self
::hebrewYearStart( $hebrewYear );
1801 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1804 # Calculate Hebrew day of year
1805 $hebrewDayOfYear = $dayOfYear - $start;
1807 # Difference between year's days
1808 $diff = $nextStart - $start;
1809 # Add 12 (or 13 for leap years) days to ignore the difference between
1810 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1811 # difference is only about the year type
1812 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1818 # Check the year pattern, and is leap year
1819 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1820 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1821 # and non-leap years
1822 $yearPattern = $diff %
30;
1823 # Check if leap year
1824 $isLeap = $diff >= 30;
1826 # Calculate day in the month from number of day in the Hebrew year
1827 # Don't check Adar - if the day is not in Adar, we will stop before;
1828 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1829 $hebrewDay = $hebrewDayOfYear;
1832 while ( $hebrewMonth <= 12 ) {
1833 # Calculate days in this month
1834 if ( $isLeap && $hebrewMonth == 6 ) {
1835 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1837 if ( $hebrewDay <= $days ) {
1841 # Subtract the days of Adar I
1842 $hebrewDay -= $days;
1845 if ( $hebrewDay <= $days ) {
1850 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1851 # Cheshvan in a complete year (otherwise as the rule below)
1853 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1854 # Kislev in an incomplete year (otherwise as the rule below)
1857 # Odd months have 30 days, even have 29
1858 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1860 if ( $hebrewDay <= $days ) {
1861 # In the current month
1864 # Subtract the days of the current month
1865 $hebrewDay -= $days;
1866 # Try in the next month
1871 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1875 * This calculates the Hebrew year start, as days since 1 September.
1876 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1877 * Used for Hebrew date.
1883 private static function hebrewYearStart( $year ) {
1884 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1885 $b = intval( ( $year - 1 ) %
4 );
1886 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1890 $Mar = intval( $m );
1896 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1897 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1899 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1901 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1905 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1910 * Algorithm to convert Gregorian dates to Thai solar dates,
1911 * Minguo dates or Minguo dates.
1913 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1914 * https://en.wikipedia.org/wiki/Minguo_calendar
1915 * https://en.wikipedia.org/wiki/Japanese_era_name
1917 * @param string $ts 14-character timestamp
1918 * @param string $cName Calender name
1919 * @return array Converted year, month, day
1921 private static function tsToYear( $ts, $cName ) {
1922 $gy = substr( $ts, 0, 4 );
1923 $gm = substr( $ts, 4, 2 );
1924 $gd = substr( $ts, 6, 2 );
1926 if ( !strcmp( $cName, 'thai' ) ) {
1928 # Add 543 years to the Gregorian calendar
1929 # Months and days are identical
1930 $gy_offset = $gy +
543;
1931 # fix for dates between 1912 and 1941
1932 # https://en.wikipedia.org/?oldid=836596673#New_year
1933 if ( $gy >= 1912 && $gy <= 1940 ) {
1937 $gm = ( $gm - 3 ) %
12;
1939 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1941 # Deduct 1911 years from the Gregorian calendar
1942 # Months and days are identical
1943 $gy_offset = $gy - 1911;
1944 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1945 # Nengō dates up to Meiji period
1946 # Deduct years from the Gregorian calendar
1947 # depending on the nengo periods
1948 # Months and days are identical
1950 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1951 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1954 $gy_gannen = $gy - 1868 +
1;
1955 $gy_offset = $gy_gannen;
1956 if ( $gy_gannen == 1 ) {
1959 $gy_offset = '明治' . $gy_offset;
1961 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1962 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1963 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1964 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1965 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1968 $gy_gannen = $gy - 1912 +
1;
1969 $gy_offset = $gy_gannen;
1970 if ( $gy_gannen == 1 ) {
1973 $gy_offset = '大正' . $gy_offset;
1975 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1976 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1977 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1980 $gy_gannen = $gy - 1926 +
1;
1981 $gy_offset = $gy_gannen;
1982 if ( $gy_gannen == 1 ) {
1985 $gy_offset = '昭和' . $gy_offset;
1987 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1988 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1989 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1992 $gy_gannen = $gy - 1989 +
1;
1993 $gy_offset = $gy_gannen;
1994 if ( $gy_gannen == 1 ) {
1997 $gy_offset = '平成' . $gy_offset;
2000 $gy_gannen = $gy - 2019 +
1;
2001 $gy_offset = $gy_gannen;
2002 if ( $gy_gannen == 1 ) {
2005 $gy_offset = '令和' . $gy_offset;
2011 return [ $gy_offset, $gm, $gd ];
2015 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2017 * This is the rule the BIDI algorithm uses to determine the directionality of
2018 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
2019 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2021 * TODO: Does not handle BIDI control characters inside the text.
2022 * TODO: Does not handle unallocated characters.
2024 * @param string $text Text to test
2025 * @return null|string Directionality ('ltr' or 'rtl') or null
2027 private static function strongDirFromContent( $text = '' ) {
2028 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2031 if ( $matches[1] === '' ) {
2038 * Roman number formatting up to 10000
2044 static function romanNumeral( $num ) {
2046 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
2047 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
2048 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
2049 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2050 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2053 $num = intval( $num );
2054 if ( $num > 10000 ||
$num <= 0 ) {
2059 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2060 if ( $num >= $pow10 ) {
2061 $s .= $table[$i][(int)floor( $num / $pow10 )];
2063 $num = $num %
$pow10;
2069 * Hebrew Gematria number formatting up to 9999
2075 static function hebrewNumeral( $num ) {
2077 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2078 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2091 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2094 $num = intval( $num );
2095 if ( $num > 9999 ||
$num <= 0 ) {
2099 // Round thousands have special notations
2100 if ( $num === 1000 ) {
2102 } elseif ( $num %
1000 === 0 ) {
2103 return $table[0][$num / 1000] . "' אלפים";
2108 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2109 if ( $num >= $pow10 ) {
2110 if ( $num === 15 ||
$num === 16 ) {
2111 $letters[] = $table[0][9];
2112 $letters[] = $table[0][$num - 9];
2115 $letters = array_merge(
2117 (array)$table[$i][intval( $num / $pow10 )]
2120 if ( $pow10 === 1000 ) {
2126 $num = $num %
$pow10;
2129 $preTransformLength = count( $letters );
2130 if ( $preTransformLength === 1 ) {
2131 // Add geresh (single quote) to one-letter numbers
2134 $lastIndex = $preTransformLength - 1;
2135 $letters[$lastIndex] = str_replace(
2136 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2137 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2138 $letters[$lastIndex]
2141 // Add gershayim (double quote) to multiple-letter numbers,
2142 // but exclude numbers with only one letter after the thousands
2143 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2144 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2147 array_splice( $letters, -1, 0, '"' );
2151 return implode( $letters );
2155 * Used by date() and time() to adjust the time output.
2157 * @param string $ts The time in date('YmdHis') format
2158 * @param mixed $tz Adjust the time by this amount (default false, mean we
2159 * get user timecorrection setting)
2162 public function userAdjust( $ts, $tz = false ) {
2163 global $wgUser, $wgLocalTZoffset;
2165 if ( $tz === false ) {
2166 $tz = $wgUser->getOption( 'timecorrection' );
2169 $data = explode( '|', $tz, 3 );
2171 if ( $data[0] == 'ZoneInfo' ) {
2173 $userTZ = new DateTimeZone( $data[2] );
2174 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2175 $date->setTimezone( $userTZ );
2176 return $date->format( 'YmdHis' );
2177 } catch ( Exception
$e ) {
2178 // Unrecognized timezone, default to 'Offset' with the stored offset.
2179 $data[0] = 'Offset';
2183 if ( $data[0] == 'System' ||
$tz == '' ) {
2184 # Global offset in minutes.
2185 $minDiff = $wgLocalTZoffset;
2186 } elseif ( $data[0] == 'Offset' ) {
2187 $minDiff = intval( $data[1] );
2189 $data = explode( ':', $tz );
2190 if ( count( $data ) == 2 ) {
2191 $data[0] = intval( $data[0] );
2192 $data[1] = intval( $data[1] );
2193 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2194 if ( $data[0] < 0 ) {
2195 $minDiff = -$minDiff;
2198 $minDiff = intval( $data[0] ) * 60;
2202 # No difference ? Return time unchanged
2203 if ( $minDiff == 0 ) {
2207 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2208 # Generate an adjusted date; take advantage of the fact that mktime
2209 # will normalize out-of-range values so we don't have to split $minDiff
2210 # into hours and minutes.
2212 (int)substr( $ts, 8, 2 ) ), # Hours
2213 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2214 (int)substr( $ts, 12, 2 ), # Seconds
2215 (int)substr( $ts, 4, 2 ), # Month
2216 (int)substr( $ts, 6, 2 ), # Day
2217 (int)substr( $ts, 0, 4 ) ); # Year
2219 $date = date( 'YmdHis', $t );
2220 Wikimedia\restoreWarnings
();
2226 * This is meant to be used by time(), date(), and timeanddate() to get
2227 * the date preference they're supposed to use, it should be used in
2230 * function timeanddate([...], $format = true) {
2231 * $datePreference = $this->dateFormat($format);
2235 * @param int|string|bool $usePrefs If true, the user's preference is used
2236 * if false, the site/language default is used
2237 * if int/string, assumed to be a format.
2240 function dateFormat( $usePrefs = true ) {
2243 if ( is_bool( $usePrefs ) ) {
2245 $datePreference = $wgUser->getDatePreference();
2247 $datePreference = (string)User
::getDefaultOption( 'date' );
2250 $datePreference = (string)$usePrefs;
2254 if ( $datePreference == '' ) {
2258 return $datePreference;
2262 * Get a format string for a given type and preference
2263 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2264 * @param string $pref The format name as it appears in Messages*.php under
2267 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2271 function getDateFormatString( $type, $pref ) {
2272 $wasDefault = false;
2273 if ( $pref == 'default' ) {
2275 $pref = $this->getDefaultDateFormat();
2278 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2279 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2281 if ( $type === 'pretty' && $df === null ) {
2282 $df = $this->getDateFormatString( 'date', $pref );
2285 if ( !$wasDefault && $df === null ) {
2286 $pref = $this->getDefaultDateFormat();
2287 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2290 $this->dateFormatStrings
[$type][$pref] = $df;
2292 return $this->dateFormatStrings
[$type][$pref];
2296 * @param string $ts The time format which needs to be turned into a
2297 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2298 * @param bool $adj Whether to adjust the time output according to the
2299 * user configured offset ($timecorrection)
2300 * @param mixed $format True to use user's date format preference
2301 * @param string|bool $timecorrection The time offset as returned by
2302 * validateTimeZone() in Special:Preferences
2305 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2306 $ts = wfTimestamp( TS_MW
, $ts );
2308 $ts = $this->userAdjust( $ts, $timecorrection );
2310 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2311 return $this->sprintfDate( $df, $ts );
2315 * @param string $ts The time format which needs to be turned into a
2316 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2317 * @param bool $adj Whether to adjust the time output according to the
2318 * user configured offset ($timecorrection)
2319 * @param mixed $format True to use user's date format preference
2320 * @param string|bool $timecorrection The time offset as returned by
2321 * validateTimeZone() in Special:Preferences
2324 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2325 $ts = wfTimestamp( TS_MW
, $ts );
2327 $ts = $this->userAdjust( $ts, $timecorrection );
2329 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2330 return $this->sprintfDate( $df, $ts );
2334 * @param string $ts The time format which needs to be turned into a
2335 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2336 * @param bool $adj Whether to adjust the time output according to the
2337 * user configured offset ($timecorrection)
2338 * @param mixed $format What format to return, if it's false output the
2339 * default one (default true)
2340 * @param string|bool $timecorrection The time offset as returned by
2341 * validateTimeZone() in Special:Preferences
2344 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2345 $ts = wfTimestamp( TS_MW
, $ts );
2347 $ts = $this->userAdjust( $ts, $timecorrection );
2349 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2350 return $this->sprintfDate( $df, $ts );
2354 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2358 * @param int $seconds The amount of seconds.
2359 * @param array $chosenIntervals The intervals to enable.
2363 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2364 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2368 foreach ( $intervals as $intervalName => $intervalValue ) {
2369 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2370 // duration-years, duration-decades, duration-centuries, duration-millennia
2371 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2372 $segments[] = $message->inLanguage( $this )->escaped();
2375 return $this->listToText( $segments );
2379 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2380 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2384 * @param int $seconds The amount of seconds.
2385 * @param array $chosenIntervals The intervals to enable.
2389 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2390 if ( empty( $chosenIntervals ) ) {
2391 $chosenIntervals = [
2403 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2404 $sortedNames = array_keys( $intervals );
2405 $smallestInterval = array_pop( $sortedNames );
2409 foreach ( $intervals as $name => $length ) {
2410 $value = floor( $seconds / $length );
2412 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2413 $seconds -= $value * $length;
2414 $segments[$name] = $value;
2422 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2424 * @param string $type Can be 'date', 'time' or 'both'
2425 * @param string $ts The time format which needs to be turned into a
2426 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2427 * @param User $user User object used to get preferences for timezone and format
2428 * @param array $options Array, can contain the following keys:
2429 * - 'timecorrection': time correction, can have the following values:
2430 * - true: use user's preference
2431 * - false: don't use time correction
2432 * - int: value of time correction in minutes
2433 * - 'format': format to use, can have the following values:
2434 * - true: use user's preference
2435 * - false: use default preference
2436 * - string: format to use
2440 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2441 $ts = wfTimestamp( TS_MW
, $ts );
2442 $options +
= [ 'timecorrection' => true, 'format' => true ];
2443 if ( $options['timecorrection'] !== false ) {
2444 if ( $options['timecorrection'] === true ) {
2445 $offset = $user->getOption( 'timecorrection' );
2447 $offset = $options['timecorrection'];
2449 $ts = $this->userAdjust( $ts, $offset );
2451 if ( $options['format'] === true ) {
2452 $format = $user->getDatePreference();
2454 $format = $options['format'];
2456 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2457 return $this->sprintfDate( $df, $ts );
2461 * Get the formatted date for the given timestamp and formatted for
2464 * @param mixed $ts Mixed: the time format which needs to be turned into a
2465 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2466 * @param User $user User object used to get preferences for timezone and format
2467 * @param array $options Array, can contain the following keys:
2468 * - 'timecorrection': time correction, can have the following values:
2469 * - true: use user's preference
2470 * - false: don't use time correction
2471 * - int: value of time correction in minutes
2472 * - 'format': format to use, can have the following values:
2473 * - true: use user's preference
2474 * - false: use default preference
2475 * - string: format to use
2479 public function userDate( $ts, User
$user, array $options = [] ) {
2480 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2484 * Get the formatted time for the given timestamp and formatted for
2487 * @param mixed $ts The time format which needs to be turned into a
2488 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2489 * @param User $user User object used to get preferences for timezone and format
2490 * @param array $options Array, can contain the following keys:
2491 * - 'timecorrection': time correction, can have the following values:
2492 * - true: use user's preference
2493 * - false: don't use time correction
2494 * - int: value of time correction in minutes
2495 * - 'format': format to use, can have the following values:
2496 * - true: use user's preference
2497 * - false: use default preference
2498 * - string: format to use
2502 public function userTime( $ts, User
$user, array $options = [] ) {
2503 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2507 * Get the formatted date and time for the given timestamp and formatted for
2510 * @param mixed $ts The time format which needs to be turned into a
2511 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2512 * @param User $user User object used to get preferences for timezone and format
2513 * @param array $options Array, can contain the following keys:
2514 * - 'timecorrection': time correction, can have the following values:
2515 * - true: use user's preference
2516 * - false: don't use time correction
2517 * - int: value of time correction in minutes
2518 * - 'format': format to use, can have the following values:
2519 * - true: use user's preference
2520 * - false: use default preference
2521 * - string: format to use
2525 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2526 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2530 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2532 * Determine the difference between the timestamp and the current time, and
2533 * generate a readable timestamp by returning "<N> <units> ago", where the
2534 * largest possible unit is used.
2536 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2538 * @param MWTimestamp $time
2539 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2540 * @param User|null $user User the timestamp is being generated for
2541 * (or null to use main context's user)
2542 * @return string Formatted timestamp
2544 public function getHumanTimestamp(
2545 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2547 if ( $relativeTo === null ) {
2548 $relativeTo = new MWTimestamp();
2550 if ( $user === null ) {
2551 $user = RequestContext
::getMain()->getUser();
2554 // Adjust for the user's timezone.
2555 $offsetThis = $time->offsetForUser( $user );
2556 $offsetRel = $relativeTo->offsetForUser( $user );
2559 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2560 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2563 // Reset the timezone on the objects.
2564 $time->timestamp
->sub( $offsetThis );
2565 $relativeTo->timestamp
->sub( $offsetRel );
2571 * Convert an MWTimestamp into a pretty human-readable timestamp using
2572 * the given user preferences and relative base time.
2574 * @see Language::getHumanTimestamp
2575 * @param MWTimestamp $ts Timestamp to prettify
2576 * @param MWTimestamp $relativeTo Base timestamp
2577 * @param User $user User preferences to use
2578 * @return string Human timestamp
2581 private function getHumanTimestampInternal(
2582 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2584 $diff = $ts->diff( $relativeTo );
2585 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2586 (int)$relativeTo->timestamp
->format( 'w' ) );
2587 $days = $diff->days ?
: (int)$diffDay;
2588 if ( $diff->invert ||
$days > 5
2589 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2591 // Timestamps are in different years: use full timestamp
2592 // Also do full timestamp for future dates
2594 * @todo FIXME: Add better handling of future timestamps.
2596 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2597 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2598 } elseif ( $days > 5 ) {
2599 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2600 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2601 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2602 } elseif ( $days > 1 ) {
2603 // Timestamp within the past week: show the day of the week and time
2604 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2605 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2607 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2608 $ts = wfMessage( "$weekday-at" )
2609 ->inLanguage( $this )
2610 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2612 } elseif ( $days == 1 ) {
2613 // Timestamp was yesterday: say 'yesterday' and the time.
2614 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2615 $ts = wfMessage( 'yesterday-at' )
2616 ->inLanguage( $this )
2617 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2619 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2620 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2621 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2622 $ts = wfMessage( 'today-at' )
2623 ->inLanguage( $this )
2624 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2627 // From here on in, the timestamp was soon enough ago so that we can simply say
2628 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2629 } elseif ( $diff->h
== 1 ) {
2630 // Less than 90 minutes, but more than an hour ago.
2631 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2632 } elseif ( $diff->i
>= 1 ) {
2633 // A few minutes ago.
2634 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2635 } elseif ( $diff->s
>= 30 ) {
2636 // Less than a minute, but more than 30 sec ago.
2637 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2639 // Less than 30 seconds ago.
2640 $ts = wfMessage( 'just-now' )->text();
2647 * @param string $key
2648 * @return string|null
2650 public function getMessage( $key ) {
2651 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2657 function getAllMessages() {
2658 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2663 * @param string $out
2664 * @param string $string
2667 public function iconv( $in, $out, $string ) {
2668 # Even with //IGNORE iconv can whine about illegal characters in
2669 # *input* string. We just ignore those too.
2670 # REF: https://bugs.php.net/bug.php?id=37166
2671 # REF: https://phabricator.wikimedia.org/T18885
2672 Wikimedia\
suppressWarnings();
2673 $text = iconv( $in, $out . '//IGNORE', $string );
2674 Wikimedia\restoreWarnings
();
2678 // callback functions for ucwords(), ucwordbreaks()
2681 * @param array $matches
2682 * @return mixed|string
2684 function ucwordbreaksCallbackAscii( $matches ) {
2685 return $this->ucfirst( $matches[1] );
2689 * @param array $matches
2692 function ucwordbreaksCallbackMB( $matches ) {
2693 return mb_strtoupper( $matches[0] );
2697 * @param array $matches
2700 function ucwordsCallbackMB( $matches ) {
2701 return mb_strtoupper( $matches[0] );
2705 * Make a string's first character uppercase
2707 * @param string $str
2711 public function ucfirst( $str ) {
2713 if ( $o < 96 ) { // if already uppercase...
2715 } elseif ( $o < 128 ) {
2716 return ucfirst( $str ); // use PHP's ucfirst()
2718 // fall back to more complex logic in case of multibyte strings
2719 return $this->uc( $str, true );
2724 * Convert a string to uppercase
2726 * @param string $str
2727 * @param bool $first
2731 public function uc( $str, $first = false ) {
2733 if ( $this->isMultibyte( $str ) ) {
2734 return $this->mbUpperChar( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2736 return ucfirst( $str );
2739 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2744 * Convert character to uppercase, allowing overrides of the default mb_upper
2745 * behaviour, which is buggy in many ways. Having a conversion table can be
2746 * useful during transitions between PHP versions where unicode changes happen.
2747 * This can make some resources unreachable on-wiki, see discussion at T219279.
2748 * Providing such a conversion table can allow to manage the transition period.
2752 * @param string $char
2756 protected function mbUpperChar( $char ) {
2757 global $wgOverrideUcfirstCharacters;
2758 if ( array_key_exists( $char, $wgOverrideUcfirstCharacters ) ) {
2759 return $wgOverrideUcfirstCharacters[$char];
2761 return mb_strtoupper( $char );
2766 * @param string $str
2767 * @return mixed|string
2769 function lcfirst( $str ) {
2772 return strval( $str );
2773 } elseif ( $o >= 128 ) {
2774 return $this->lc( $str, true );
2775 } elseif ( $o > 96 ) {
2778 $str[0] = strtolower( $str[0] );
2784 * @param string $str
2785 * @param bool $first
2786 * @return mixed|string
2788 function lc( $str, $first = false ) {
2790 if ( $this->isMultibyte( $str ) ) {
2791 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2793 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2796 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2801 * @param string $str
2804 function isMultibyte( $str ) {
2805 return strlen( $str ) !== mb_strlen( $str );
2809 * @param string $str
2810 * @return mixed|string
2812 function ucwords( $str ) {
2813 if ( $this->isMultibyte( $str ) ) {
2814 $str = $this->lc( $str );
2816 // regexp to find first letter in each word (i.e. after each space)
2817 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2819 // function to use to capitalize a single char
2820 return preg_replace_callback(
2822 [ $this, 'ucwordsCallbackMB' ],
2826 return ucwords( strtolower( $str ) );
2831 * capitalize words at word breaks
2833 * @param string $str
2836 function ucwordbreaks( $str ) {
2837 if ( $this->isMultibyte( $str ) ) {
2838 $str = $this->lc( $str );
2840 // since \b doesn't work for UTF-8, we explicitely define word break chars
2841 $breaks = "[ \-\(\)\}\{\.,\?!]";
2843 // find first letter after word break
2844 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2845 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2847 return preg_replace_callback(
2849 [ $this, 'ucwordbreaksCallbackMB' ],
2853 return preg_replace_callback(
2854 '/\b([\w\x80-\xff]+)\b/',
2855 [ $this, 'ucwordbreaksCallbackAscii' ],
2862 * Return a case-folded representation of $s
2864 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2865 * and $s2 are the same except for the case of their characters. It is not
2866 * necessary for the value returned to make sense when displayed.
2868 * Do *not* perform any other normalisation in this function. If a caller
2869 * uses this function when it should be using a more general normalisation
2870 * function, then fix the caller.
2876 function caseFold( $s ) {
2877 return $this->uc( $s );
2881 * TODO: $s is not always a string per T218883
2885 function checkTitleEncoding( $s ) {
2886 if ( is_array( $s ) ) {
2887 throw new MWException( 'Given array to checkTitleEncoding.' );
2889 if ( StringUtils
::isUtf8( $s ) ) {
2893 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2899 function fallback8bitEncoding() {
2900 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2904 * Most writing systems use whitespace to break up words.
2905 * Some languages such as Chinese don't conventionally do this,
2906 * which requires special handling when breaking up words for
2911 function hasWordBreaks() {
2916 * Some languages such as Chinese require word segmentation,
2917 * Specify such segmentation when overridden in derived class.
2919 * @param string $string
2922 function segmentByWord( $string ) {
2927 * Some languages have special punctuation need to be normalized.
2928 * Make such changes here.
2930 * @param string $string
2933 function normalizeForSearch( $string ) {
2934 return self
::convertDoubleWidth( $string );
2938 * convert double-width roman characters to single-width.
2939 * range: ff00-ff5f ~= 0020-007f
2941 * @param string $string
2945 protected static function convertDoubleWidth( $string ) {
2946 static $full = null;
2947 static $half = null;
2949 if ( $full === null ) {
2950 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2951 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2952 $full = str_split( $fullWidth, 3 );
2953 $half = str_split( $halfWidth );
2956 $string = str_replace( $full, $half, $string );
2961 * @param string $string
2962 * @param string $pattern
2965 protected static function insertSpace( $string, $pattern ) {
2966 $string = preg_replace( $pattern, " $1 ", $string );
2967 $string = preg_replace( '/ +/', ' ', $string );
2972 * @param string[] $termsArray
2975 function convertForSearchResult( $termsArray ) {
2976 # some languages, e.g. Chinese, need to do a conversion
2977 # in order for search results to be displayed correctly
2982 * Get the first character of a string.
2987 function firstChar( $s ) {
2990 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2991 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2996 if ( isset( $matches[1] ) ) {
2997 if ( strlen( $matches[1] ) != 3 ) {
3001 // Break down Hangul syllables to grab the first jamo
3002 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3003 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3005 } elseif ( $code < 0xb098 ) {
3007 } elseif ( $code < 0xb2e4 ) {
3009 } elseif ( $code < 0xb77c ) {
3011 } elseif ( $code < 0xb9c8 ) {
3013 } elseif ( $code < 0xbc14 ) {
3015 } elseif ( $code < 0xc0ac ) {
3017 } elseif ( $code < 0xc544 ) {
3019 } elseif ( $code < 0xc790 ) {
3021 } elseif ( $code < 0xcc28 ) {
3023 } elseif ( $code < 0xce74 ) {
3025 } elseif ( $code < 0xd0c0 ) {
3027 } elseif ( $code < 0xd30c ) {
3029 } elseif ( $code < 0xd558 ) {
3040 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3041 * also cleans up certain backwards-compatible sequences, converting them
3042 * to the modern Unicode equivalent.
3044 * This is language-specific for performance reasons only.
3050 public function normalize( $s ) {
3051 global $wgAllUnicodeFixes;
3052 $s = UtfNormal\Validator
::cleanUp( $s );
3053 if ( $wgAllUnicodeFixes ) {
3054 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
3055 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3062 * Transform a string using serialized data stored in the given file (which
3063 * must be in the serialized subdirectory of $IP). The file contains pairs
3064 * mapping source characters to destination characters.
3066 * The data is cached in process memory. This will go faster if you have the
3067 * FastStringSearch extension.
3069 * @param string $file
3070 * @param string $string
3072 * @throws MWException
3075 protected function transformUsingPairFile( $file, $string ) {
3076 if ( !isset( $this->transformData
[$file] ) ) {
3078 $data = require "$IP/languages/data/{$file}";
3079 $this->transformData
[$file] = new ReplacementArray( $data );
3081 return $this->transformData
[$file]->replace( $string );
3085 * For right-to-left language support
3090 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3094 * Return the correct HTML 'dir' attribute value for this language.
3098 return $this->isRTL() ?
'rtl' : 'ltr';
3102 * Return 'left' or 'right' as appropriate alignment for line-start
3103 * for this language's text direction.
3105 * Should be equivalent to CSS3 'start' text-align value....
3109 function alignStart() {
3110 return $this->isRTL() ?
'right' : 'left';
3114 * Return 'right' or 'left' as appropriate alignment for line-end
3115 * for this language's text direction.
3117 * Should be equivalent to CSS3 'end' text-align value....
3121 function alignEnd() {
3122 return $this->isRTL() ?
'left' : 'right';
3126 * A hidden direction mark (LRM or RLM), depending on the language direction.
3127 * Unlike getDirMark(), this function returns the character as an HTML entity.
3128 * This function should be used when the output is guaranteed to be HTML,
3129 * because it makes the output HTML source code more readable. When
3130 * the output is plain text or can be escaped, getDirMark() should be used.
3132 * @param bool $opposite Get the direction mark opposite to your language
3136 function getDirMarkEntity( $opposite = false ) {
3138 return $this->isRTL() ?
'‎' : '‏';
3140 return $this->isRTL() ?
'‏' : '‎';
3144 * A hidden direction mark (LRM or RLM), depending on the language direction.
3145 * This function produces them as invisible Unicode characters and
3146 * the output may be hard to read and debug, so it should only be used
3147 * when the output is plain text or can be escaped. When the output is
3148 * HTML, use getDirMarkEntity() instead.
3150 * @param bool $opposite Get the direction mark opposite to your language
3153 function getDirMark( $opposite = false ) {
3154 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3155 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3157 return $this->isRTL() ?
$lrm : $rlm;
3159 return $this->isRTL() ?
$rlm : $lrm;
3165 function capitalizeAllNouns() {
3166 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3170 * An arrow, depending on the language direction.
3172 * @param string $direction The direction of the arrow: forwards (default),
3173 * backwards, left, right, up, down.
3176 function getArrow( $direction = 'forwards' ) {
3177 switch ( $direction ) {
3179 return $this->isRTL() ?
'←' : '→';
3181 return $this->isRTL() ?
'→' : '←';
3194 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3198 function linkPrefixExtension() {
3199 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3203 * Get all magic words from cache.
3206 function getMagicWords() {
3207 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3211 * Fill a MagicWord object with data from here
3213 * @param MagicWord $mw
3215 function getMagic( $mw ) {
3216 $rawEntry = $this->mMagicExtensions
[$mw->mId
] ??
3217 self
::$dataCache->getSubitem( $this->mCode
, 'magicWords', $mw->mId
);
3219 if ( !is_array( $rawEntry ) ) {
3220 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3222 $mw->mCaseSensitive
= $rawEntry[0];
3223 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3228 * Add magic words to the extension array
3230 * @param array $newWords
3232 function addMagicWordsByLang( $newWords ) {
3233 $fallbackChain = $this->getFallbackLanguages();
3234 $fallbackChain = array_reverse( $fallbackChain );
3235 foreach ( $fallbackChain as $code ) {
3236 if ( isset( $newWords[$code] ) ) {
3237 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3243 * Get special page names, as an associative array
3244 * canonical name => array of valid names, including aliases
3247 function getSpecialPageAliases() {
3248 // Cache aliases because it may be slow to load them
3249 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3251 $this->mExtendedSpecialPageAliases
=
3252 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3255 return $this->mExtendedSpecialPageAliases
;
3259 * Italic is unsuitable for some languages
3261 * @param string $text The text to be emphasized.
3264 function emphasize( $text ) {
3265 return "<em>$text</em>";
3269 * Normally we output all numbers in plain en_US style, that is
3270 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3271 * point twohundredthirtyfive. However this is not suitable for all
3272 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3273 * Icelandic just want to use commas instead of dots, and dots instead
3274 * of commas like "293.291,235".
3276 * An example of this function being called:
3278 * wfMessage( 'message' )->numParams( $num )->text()
3281 * See $separatorTransformTable on MessageIs.php for
3282 * the , => . and . => , implementation.
3284 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3285 * @param int|float $number The string to be formatted, should be an integer
3286 * or a floating point number.
3287 * @param bool $nocommafy Set to true for special numbers like dates
3290 public function formatNum( $number, $nocommafy = false ) {
3291 global $wgTranslateNumerals;
3292 if ( !$nocommafy ) {
3293 $number = $this->commafy( $number );
3294 $s = $this->separatorTransformTable();
3296 $number = strtr( $number, $s );
3300 if ( $wgTranslateNumerals ) {
3301 $s = $this->digitTransformTable();
3303 $number = strtr( $number, $s );
3307 return (string)$number;
3311 * Front-end for non-commafied formatNum
3313 * @param int|float $number The string to be formatted, should be an integer
3314 * or a floating point number.
3318 public function formatNumNoSeparators( $number ) {
3319 return $this->formatNum( $number, true );
3323 * @param string $number
3326 public function parseFormattedNumber( $number ) {
3327 $s = $this->digitTransformTable();
3329 // eliminate empty array values such as ''. (T66347)
3330 $s = array_filter( $s );
3331 $number = strtr( $number, array_flip( $s ) );
3334 $s = $this->separatorTransformTable();
3336 // eliminate empty array values such as ''. (T66347)
3337 $s = array_filter( $s );
3338 $number = strtr( $number, array_flip( $s ) );
3341 $number = strtr( $number, [ ',' => '' ] );
3346 * Adds commas to a given number
3348 * @param mixed $number
3351 function commafy( $number ) {
3352 $digitGroupingPattern = $this->digitGroupingPattern();
3353 $minimumGroupingDigits = $this->minimumGroupingDigits();
3354 if ( $number === null ) {
3358 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3359 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3360 // In some languages it's conventional not to insert a thousands separator
3361 // in numbers that are four digits long (1000-9999).
3362 if ( $minimumGroupingDigits ) {
3363 // Number of '#' characters after last comma in the grouping pattern.
3364 // The pattern is hardcoded here, but this would vary for different patterns.
3365 $primaryGroupingSize = 3;
3366 // Maximum length of a number to suppress digit grouping for.
3367 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3368 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3372 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3374 // Ref: http://cldr.unicode.org/translation/number-patterns
3376 if ( intval( $number ) < 0 ) {
3377 // For negative numbers apply the algorithm like positive number and add sign.
3379 $number = substr( $number, 1 );
3383 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3384 preg_match( "/\d+/", $number, $integerPart );
3385 preg_match( "/\.\d*/", $number, $decimalPart );
3386 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3387 if ( $groupedNumber === $number ) {
3388 // the string does not have any number part. Eg: .12345
3389 return $sign . $groupedNumber;
3391 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3392 while ( $start > 0 ) {
3393 $match = $matches[0][$numMatches - 1];
3394 $matchLen = strlen( $match );
3395 $start = $end - $matchLen;
3399 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3401 if ( $numMatches > 1 ) {
3402 // use the last pattern for the rest of the number
3406 $groupedNumber = "," . $groupedNumber;
3409 return $sign . $groupedNumber;
3416 function digitGroupingPattern() {
3417 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3423 function digitTransformTable() {
3424 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3430 function separatorTransformTable() {
3431 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3437 function minimumGroupingDigits() {
3438 return self
::$dataCache->getItem( $this->mCode
, 'minimumGroupingDigits' );
3442 * Take a list of strings and build a locale-friendly comma-separated
3443 * list, using the local comma-separator message.
3444 * The last two strings are chained with an "and".
3446 * @param string[] $list
3449 public function listToText( array $list ) {
3450 $itemCount = count( $list );
3451 if ( $itemCount < 1 ) {
3454 $text = array_pop( $list );
3455 if ( $itemCount > 1 ) {
3456 $and = $this->msg( 'and' )->escaped();
3457 $space = $this->msg( 'word-separator' )->escaped();
3459 if ( $itemCount > 2 ) {
3460 $comma = $this->msg( 'comma-separator' )->escaped();
3462 $text = implode( $comma, $list ) . $and . $space . $text;
3468 * Take a list of strings and build a locale-friendly comma-separated
3469 * list, using the local comma-separator message.
3470 * @param string[] $list Array of strings to put in a comma list
3473 function commaList( array $list ) {
3475 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3481 * Take a list of strings and build a locale-friendly semicolon-separated
3482 * list, using the local semicolon-separator message.
3483 * @param string[] $list Array of strings to put in a semicolon list
3486 function semicolonList( array $list ) {
3488 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3494 * Same as commaList, but separate it with the pipe instead.
3495 * @param string[] $list Array of strings to put in a pipe list
3498 function pipeList( array $list ) {
3500 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3506 * Truncate a string to a specified length in bytes, appending an optional
3507 * string (e.g. for ellipsis)
3509 * If $length is negative, the string will be truncated from the beginning
3513 * @param string $string String to truncate
3514 * @param int $length Maximum length in bytes
3515 * @param string $ellipsis String to append to the end of truncated text
3516 * @param bool $adjustLength Subtract length of ellipsis from $length
3520 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3521 return $this->truncateInternal(
3522 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3527 * Truncate a string to a specified number of characters, appending an optional
3528 * string (e.g. for ellipsis).
3530 * This provides multibyte version of truncateForDatabase() method of this class,
3531 * suitable for truncation based on number of characters, instead of number of bytes.
3533 * If $length is negative, the string will be truncated from the beginning.
3537 * @param string $string String to truncate
3538 * @param int $length Maximum number of characters
3539 * @param string $ellipsis String to append to the end of truncated text
3540 * @param bool $adjustLength Subtract length of ellipsis from $length
3544 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3545 // Passing encoding to mb_strlen and mb_substr is optional.
3546 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3547 // explicit specification of encoding is skipped.
3548 // Note: Both multibyte methods are callables invoked in truncateInternal.
3549 return $this->truncateInternal(
3550 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3555 * Internal method used for truncation. This method abstracts text truncation into
3556 * one common method, allowing users to provide length measurement function and
3557 * function for finding substring.
3559 * For usages, see truncateForDatabase and truncateForVisual.
3561 * @param string $string String to truncate
3562 * @param int $length Maximum length of final text
3563 * @param string $ellipsis String to append to the end of truncated text
3564 * @param bool $adjustLength Subtract length of ellipsis from $length
3565 * @param callable $measureLength Callable function used for determining the length of text
3566 * @param callable $getSubstring Callable function used for getting the substrings
3570 private function truncateInternal(
3571 $string, $length, $ellipsis, $adjustLength, callable
$measureLength, callable
$getSubstring
3573 # Check if there is no need to truncate
3574 if ( $measureLength( $string ) <= abs( $length ) ) {
3575 return $string; // no need to truncate
3578 # Use the localized ellipsis character
3579 if ( $ellipsis == '...' ) {
3580 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3582 if ( $length == 0 ) {
3583 return $ellipsis; // convention
3586 $stringOriginal = $string;
3587 # If ellipsis length is >= $length then we can't apply $adjustLength
3588 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3589 $string = $ellipsis; // this can be slightly unexpected
3590 # Otherwise, truncate and add ellipsis...
3592 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3593 if ( $length > 0 ) {
3594 $length -= $ellipsisLength;
3595 $string = $getSubstring( $string, 0, $length ); // xyz...
3596 $string = $this->removeBadCharLast( $string );
3597 $string = rtrim( $string ) . $ellipsis;
3599 $length +
= $ellipsisLength;
3600 $string = $getSubstring( $string, $length ); // ...xyz
3601 $string = $this->removeBadCharFirst( $string );
3602 $string = $ellipsis . ltrim( $string );
3606 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3607 # This check is *not* redundant if $adjustLength, due to the single case where
3608 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3609 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3612 return $stringOriginal;
3617 * Remove bytes that represent an incomplete Unicode character
3618 * at the end of string (e.g. bytes of the char are missing)
3620 * @param string $string
3623 protected function removeBadCharLast( $string ) {
3624 if ( $string != '' ) {
3625 $char = ord( $string[strlen( $string ) - 1] );
3627 if ( $char >= 0xc0 ) {
3628 # We got the first byte only of a multibyte char; remove it.
3629 $string = substr( $string, 0, -1 );
3630 } elseif ( $char >= 0x80 &&
3631 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3632 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3633 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3635 # We chopped in the middle of a character; remove it
3643 * Remove bytes that represent an incomplete Unicode character
3644 * at the start of string (e.g. bytes of the char are missing)
3646 * @param string $string
3649 protected function removeBadCharFirst( $string ) {
3650 if ( $string != '' ) {
3651 $char = ord( $string[0] );
3652 if ( $char >= 0x80 && $char < 0xc0 ) {
3653 # We chopped in the middle of a character; remove the whole thing
3654 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3661 * Truncate a string of valid HTML to a specified length in bytes,
3662 * appending an optional string (e.g. for ellipses), and return valid HTML
3664 * This is only intended for styled/linked text, such as HTML with
3665 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3666 * Also, this will not detect things like "display:none" CSS.
3668 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3670 * @param string $text HTML string to truncate
3671 * @param int $length (zero/positive) Maximum length (including ellipses)
3672 * @param string $ellipsis String to append to the truncated text
3675 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3676 # Use the localized ellipsis character
3677 if ( $ellipsis == '...' ) {
3678 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3680 # Check if there is clearly no need to truncate
3681 if ( $length <= 0 ) {
3682 return $ellipsis; // no text shown, nothing to format (convention)
3683 } elseif ( strlen( $text ) <= $length ) {
3684 return $text; // string short enough even *with* HTML (short-circuit)
3687 $dispLen = 0; // innerHTML legth so far
3688 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3689 $tagType = 0; // 0-open, 1-close
3690 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3691 $entityState = 0; // 0-not entity, 1-entity
3692 $tag = $ret = ''; // accumulated tag name, accumulated result string
3693 $openTags = []; // open tag stack
3694 $maybeState = null; // possible truncation state
3696 $textLen = strlen( $text );
3697 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3698 for ( $pos = 0; true; ++
$pos ) {
3699 # Consider truncation once the display length has reached the maximim.
3700 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3701 # Check that we're not in the middle of a bracket/entity...
3702 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3703 if ( !$testingEllipsis ) {
3704 $testingEllipsis = true;
3705 # Save where we are; we will truncate here unless there turn out to
3706 # be so few remaining characters that truncation is not necessary.
3707 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3708 $maybeState = [ $ret, $openTags ]; // save state
3710 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3711 # String in fact does need truncation, the truncation point was OK.
3712 // @phan-suppress-next-line PhanTypeInvalidExpressionArrayDestructuring
3713 list( $ret, $openTags ) = $maybeState; // reload state
3714 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3715 $ret .= $ellipsis; // add ellipsis
3719 if ( $pos >= $textLen ) {
3720 break; // extra iteration just for above checks
3723 # Read the next char...
3725 $lastCh = $pos ?
$text[$pos - 1] : '';
3726 $ret .= $ch; // add to result string
3728 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3729 $entityState = 0; // for bad HTML
3730 $bracketState = 1; // tag started (checking for backslash)
3731 } elseif ( $ch == '>' ) {
3732 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3733 $entityState = 0; // for bad HTML
3734 $bracketState = 0; // out of brackets
3735 } elseif ( $bracketState == 1 ) {
3737 $tagType = 1; // close tag (e.g. "</span>")
3739 $tagType = 0; // open tag (e.g. "<span>")
3742 $bracketState = 2; // building tag name
3743 } elseif ( $bracketState == 2 ) {
3747 // Name found (e.g. "<a href=..."), add on tag attributes...
3748 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3750 } elseif ( $bracketState == 0 ) {
3751 if ( $entityState ) {
3754 $dispLen++
; // entity is one displayed char
3757 if ( $neLength == 0 && !$maybeState ) {
3758 // Save state without $ch. We want to *hit* the first
3759 // display char (to get tags) but not *use* it if truncating.
3760 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3763 $entityState = 1; // entity found, (e.g. " ")
3765 $dispLen++
; // this char is displayed
3766 // Add the next $max display text chars after this in one swoop...
3767 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3768 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3769 $dispLen +
= $skipped;
3775 // Close the last tag if left unclosed by bad HTML
3776 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3777 while ( count( $openTags ) > 0 ) {
3778 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3784 * truncateHtml() helper function
3785 * like strcspn() but adds the skipped chars to $ret
3787 * @param string $ret
3788 * @param string $text
3789 * @param string $search
3791 * @param null|int $len
3794 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3795 if ( $len === null ) {
3796 $len = -1; // -1 means "no limit" for strcspn
3797 } elseif ( $len < 0 ) {
3801 if ( $start < strlen( $text ) ) {
3802 $skipCount = strcspn( $text, $search, $start, $len );
3803 $ret .= substr( $text, $start, $skipCount );
3809 * truncateHtml() helper function
3810 * (a) push or pop $tag from $openTags as needed
3811 * (b) clear $tag value
3812 * @param string &$tag Current HTML tag name we are looking at
3813 * @param int $tagType (0-open tag, 1-close tag)
3814 * @param string $lastCh Character before the '>' that ended this tag
3815 * @param array &$openTags Open tag stack (not accounting for $tag)
3817 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3818 $tag = ltrim( $tag );
3820 if ( $tagType == 0 && $lastCh != '/' ) {
3821 $openTags[] = $tag; // tag opened (didn't close itself)
3822 } elseif ( $tagType == 1 ) {
3823 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3824 array_pop( $openTags ); // tag closed
3832 * Grammatical transformations, needed for inflected languages
3833 * Invoked by putting {{grammar:case|word}} in a message
3835 * @param string $word
3836 * @param string $case
3839 function convertGrammar( $word, $case ) {
3840 global $wgGrammarForms;
3841 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3842 return $wgGrammarForms[$this->getCode()][$case][$word];
3845 $grammarTransformations = $this->getGrammarTransformations();
3847 if ( isset( $grammarTransformations[$case] ) ) {
3848 $forms = $grammarTransformations[$case];
3850 // Some names of grammar rules are aliases for other rules.
3851 // In such cases the value is a string rather than object,
3852 // so load the actual rules.
3853 if ( is_string( $forms ) ) {
3854 $forms = $grammarTransformations[$forms];
3857 foreach ( array_values( $forms ) as $rule ) {
3860 if ( $form === '@metadata' ) {
3864 $replacement = $rule[1];
3866 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3867 $patternMatches = preg_match( $regex, $word );
3869 if ( $patternMatches === false ) {
3871 'An error occurred while processing grammar. ' .
3872 "Word: '$word'. Regex: /$form/."
3874 } elseif ( $patternMatches === 1 ) {
3875 $word = preg_replace( $regex, $replacement, $word );
3886 * Get the grammar forms for the content language
3887 * @return array Array of grammar forms
3890 function getGrammarForms() {
3891 global $wgGrammarForms;
3892 if ( isset( $wgGrammarForms[$this->getCode()] )
3893 && is_array( $wgGrammarForms[$this->getCode()] )
3895 return $wgGrammarForms[$this->getCode()];
3902 * Get the grammar transformations data for the language.
3903 * Used like grammar forms, with {{GRAMMAR}} and cases,
3904 * but uses pairs of regexes and replacements instead of code.
3906 * @return array[] Array of grammar transformations.
3907 * @throws MWException
3910 public function getGrammarTransformations() {
3911 $languageCode = $this->getCode();
3913 if ( self
::$grammarTransformations === null ) {
3914 self
::$grammarTransformations = new MapCacheLRU( 10 );
3917 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3918 return self
::$grammarTransformations->get( $languageCode );
3923 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3924 if ( is_readable( $grammarDataFile ) ) {
3925 $data = FormatJson
::decode(
3926 file_get_contents( $grammarDataFile ),
3930 if ( $data === null ) {
3931 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3934 self
::$grammarTransformations->set( $languageCode, $data );
3941 * Provides an alternative text depending on specified gender.
3942 * Usage {{gender:username|masculine|feminine|unknown}}.
3943 * username is optional, in which case the gender of current user is used,
3944 * but only in (some) interface messages; otherwise default gender is used.
3946 * If no forms are given, an empty string is returned. If only one form is
3947 * given, it will be returned unconditionally. These details are implied by
3948 * the caller and cannot be overridden in subclasses.
3950 * If three forms are given, the default is to use the third (unknown) form.
3951 * If fewer than three forms are given, the default is to use the first (masculine) form.
3952 * These details can be overridden in subclasses.
3954 * @param string $gender
3955 * @param array $forms
3959 function gender( $gender, $forms ) {
3960 if ( !count( $forms ) ) {
3963 $forms = $this->preConvertPlural( $forms, 2 );
3964 if ( $gender === 'male' ) {
3967 if ( $gender === 'female' ) {
3970 return $forms[2] ??
$forms[0];
3974 * Plural form transformations, needed for some languages.
3975 * For example, there are 3 form of plural in Russian and Polish,
3976 * depending on "count mod 10". See [[w:Plural]]
3977 * For English it is pretty simple.
3979 * Invoked by putting {{plural:count|wordform1|wordform2}}
3980 * or {{plural:count|wordform1|wordform2|wordform3}}
3982 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3984 * @param int $count Non-localized number
3985 * @param array $forms Different plural forms
3986 * @return string Correct form of plural for $count in this language
3988 function convertPlural( $count, $forms ) {
3989 // Handle explicit n=pluralform cases
3990 $forms = $this->handleExplicitPluralForms( $count, $forms );
3991 if ( is_string( $forms ) ) {
3994 if ( !count( $forms ) ) {
3998 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3999 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
4000 return $forms[$pluralForm];
4004 * Handles explicit plural forms for Language::convertPlural()
4006 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4007 * If an explicitly defined plural form matches the $count, then
4008 * string value returned, otherwise array returned for further consideration
4009 * by CLDR rules or overridden convertPlural().
4013 * @param int $count Non-localized number
4014 * @param array $forms Different plural forms
4016 * @return array|string
4018 protected function handleExplicitPluralForms( $count, array $forms ) {
4019 foreach ( $forms as $index => $form ) {
4020 if ( preg_match( '/\d+=/i', $form ) ) {
4021 $pos = strpos( $form, '=' );
4022 if ( substr( $form, 0, $pos ) === (string)$count ) {
4023 return substr( $form, $pos +
1 );
4025 unset( $forms[$index] );
4028 return array_values( $forms );
4032 * Checks that convertPlural was given an array and pads it to requested
4033 * amount of forms by copying the last one.
4035 * @param array $forms
4036 * @param int $count How many forms should there be at least
4037 * @return array Padded array of forms
4039 protected function preConvertPlural( /* Array */ $forms, $count ) {
4040 return array_pad( $forms, $count, end( $forms ) );
4044 * Wraps argument with unicode control characters for directionality safety
4046 * This solves the problem where directionality-neutral characters at the edge of
4047 * the argument string get interpreted with the wrong directionality from the
4048 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4050 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4051 * directionality of the argument string, using the BIDI algorithm's own "First
4052 * strong directional codepoint" rule. Essentially, this works round the fact that
4053 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4054 * direction inference). The latter is cleaner but still not widely supported.
4056 * @param string $text Text to wrap
4057 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4059 public function embedBidi( $text = '' ) {
4060 $dir = self
::strongDirFromContent( $text );
4061 if ( $dir === 'ltr' ) {
4062 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4063 return self
::$lre . $text . self
::$pdf;
4065 if ( $dir === 'rtl' ) {
4066 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4067 return self
::$rle . $text . self
::$pdf;
4069 // No strong directionality: do not wrap
4074 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4075 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4076 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4077 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4080 * @param string $str The validated block duration in English
4081 * @param User|null $user User object to use timezone from or null for $wgUser
4082 * @param int $now Current timestamp, for formatting relative block durations
4083 * @return string Somehow translated block duration
4084 * @see LanguageFi.php for example implementation
4086 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
4087 $duration = SpecialBlock
::getSuggestedDurations( $this );
4088 foreach ( $duration as $show => $value ) {
4089 if ( strcmp( $str, $value ) == 0 ) {
4090 return htmlspecialchars( trim( $show ) );
4094 if ( wfIsInfinity( $str ) ) {
4095 foreach ( $duration as $show => $value ) {
4096 if ( wfIsInfinity( $value ) ) {
4097 return htmlspecialchars( trim( $show ) );
4102 // If all else fails, return a standard duration or timestamp description.
4103 $time = strtotime( $str, $now );
4104 if ( $time === false ) { // Unknown format. Return it as-is in case.
4106 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
4107 // The result differs based on current time, so the difference
4108 // is a fixed duration length.
4109 return $this->formatDuration( $time - $now );
4110 } else { // It's an absolute timestamp.
4111 if ( $time === 0 ) {
4112 // wfTimestamp() handles 0 as current time instead of epoch.
4113 $time = '19700101000000';
4116 return $this->userTimeAndDate( $time, $user );
4118 return $this->timeanddate( $time );
4123 * languages like Chinese need to be segmented in order for the diff
4126 * @param string $text
4129 public function segmentForDiff( $text ) {
4134 * and unsegment to show the result
4136 * @param string $text
4139 public function unsegmentForDiff( $text ) {
4144 * Return the LanguageConverter used in the Language
4147 * @return LanguageConverter
4149 public function getConverter() {
4150 return $this->mConverter
;
4154 * convert text to a variant
4156 * @param string $text text to convert
4157 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4158 * variant (if logged in), or the project default variant
4159 * @return string the converted string
4161 public function autoConvert( $text, $variant = false ) {
4162 return $this->mConverter
->autoConvert( $text, $variant );
4166 * convert text to all supported variants
4168 * @param string $text
4171 public function autoConvertToAllVariants( $text ) {
4172 return $this->mConverter
->autoConvertToAllVariants( $text );
4176 * convert text to different variants of a language.
4178 * @warning Glossary state is maintained between calls. This means
4179 * if you pass unescaped text to this method it can cause an XSS
4180 * in later calls to this method, even if the later calls have properly
4181 * escaped the input. Never feed this method user controlled text that
4182 * is not properly escaped!
4183 * @param string $text Content that has been already escaped for use in HTML
4184 * @return string HTML
4186 public function convert( $text ) {
4187 return $this->mConverter
->convert( $text );
4191 * Convert a Title object to a string in the preferred variant
4193 * @param Title $title
4196 public function convertTitle( $title ) {
4197 return $this->mConverter
->convertTitle( $title );
4201 * Convert a namespace index to a string in the preferred variant
4203 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4204 * @param string|null $variant variant to convert to, or null to use the user's preferred
4205 * variant (if logged in), or the project default variant
4206 * @return string a string representation of the namespace
4208 public function convertNamespace( $ns, $variant = null ) {
4209 return $this->mConverter
->convertNamespace( $ns, $variant );
4213 * Check if this is a language with variants
4217 public function hasVariants() {
4218 return count( $this->getVariants() ) > 1;
4222 * Strict check if the language has the specific variant.
4224 * Compare to LanguageConverter::validateVariant() which does a more
4225 * lenient check and attempts to coerce the given code to a valid one.
4228 * @param string $variant
4231 public function hasVariant( $variant ) {
4232 return $variant && ( $variant === $this->mConverter
->validateVariant( $variant ) );
4236 * Perform output conversion on a string, and encode for safe HTML output.
4237 * @param string $text Text to be converted
4239 * @todo this should get integrated somewhere sane
4241 public function convertHtml( $text ) {
4242 return htmlspecialchars( $this->convert( $text ) );
4246 * @param string $key
4249 public function convertCategoryKey( $key ) {
4250 return $this->mConverter
->convertCategoryKey( $key );
4254 * Get the list of variants supported by this language
4255 * see sample implementation in LanguageZh.php
4257 * @return string[] An array of language codes
4259 public function getVariants() {
4260 return $this->mConverter
->getVariants();
4266 public function getPreferredVariant() {
4267 return $this->mConverter
->getPreferredVariant();
4273 public function getDefaultVariant() {
4274 return $this->mConverter
->getDefaultVariant();
4280 public function getURLVariant() {
4281 return $this->mConverter
->getURLVariant();
4285 * If a language supports multiple variants, it is
4286 * possible that non-existing link in one variant
4287 * actually exists in another variant. this function
4288 * tries to find it. See e.g. LanguageZh.php
4289 * The input parameters may be modified upon return
4291 * @param string &$link The name of the link
4292 * @param Title &$nt The title object of the link
4293 * @param bool $ignoreOtherCond To disable other conditions when
4294 * we need to transclude a template or update a category's link
4296 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4297 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4301 * returns language specific options used by User::getPageRenderHash()
4302 * for example, the preferred language variant
4306 function getExtraHashOptions() {
4307 return $this->mConverter
->getExtraHashOptions();
4311 * For languages that support multiple variants, the title of an
4312 * article may be displayed differently in different variants. this
4313 * function returns the apporiate title defined in the body of the article.
4317 public function getParsedTitle() {
4318 return $this->mConverter
->getParsedTitle();
4322 * Refresh the cache of conversion tables when
4323 * MediaWiki:Conversiontable* is updated.
4325 * @param Title $title The Title of the page being updated
4327 public function updateConversionTable( Title
$title ) {
4328 $this->mConverter
->updateConversionTable( $title );
4332 * A regular expression to match legal word-trailing characters
4333 * which should be merged onto a link of the form [[foo]]bar.
4337 public function linkTrail() {
4338 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4342 * A regular expression character set to match legal word-prefixing
4343 * characters which should be merged onto a link of the form foo[[bar]].
4347 public function linkPrefixCharset() {
4348 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4352 * Get the "parent" language which has a converter to convert a "compatible" language
4353 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4355 * @return Language|null
4358 public function getParentLanguage() {
4359 if ( $this->mParentLanguage
!== false ) {
4360 return $this->mParentLanguage
;
4363 $code = explode( '-', $this->getCode() )[0];
4364 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4365 $this->mParentLanguage
= null;
4368 $lang = self
::factory( $code );
4369 if ( !$lang->hasVariant( $this->getCode() ) ) {
4370 $this->mParentLanguage
= null;
4374 $this->mParentLanguage
= $lang;
4379 * Compare with an other language object
4382 * @param Language $lang
4385 public function equals( Language
$lang ) {
4386 return $lang === $this ||
$lang->getCode() === $this->mCode
;
4390 * Get the internal language code for this language object
4392 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4393 * htmlspecialchars() or similar
4397 public function getCode() {
4398 return $this->mCode
;
4402 * Get the code in BCP 47 format which we can use
4403 * inside of html lang="" tags.
4405 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4406 * htmlspecialchars() or similar.
4411 public function getHtmlCode() {
4412 if ( is_null( $this->mHtmlCode
) ) {
4413 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4415 return $this->mHtmlCode
;
4419 * Get the language code from a file name. Inverse of getFileName()
4420 * @param string $filename $prefix . $languageCode . $suffix
4421 * @param string $prefix Prefix before the language code
4422 * @param string $suffix Suffix after the language code
4423 * @return string Language code, or false if $prefix or $suffix isn't found
4425 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4427 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4428 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4429 if ( !count( $m ) ) {
4432 return str_replace( '_', '-', strtolower( $m[1] ) );
4436 * @param string $code
4437 * @param bool $fallback Whether we're going through language fallback chain
4438 * @return string Name of the language class
4440 public static function classFromCode( $code, $fallback = true ) {
4441 if ( $fallback && $code == 'en' ) {
4444 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4449 * Get the name of a file for a certain language code
4450 * @param string $prefix Prepend this to the filename
4451 * @param string $code Language code
4452 * @param string $suffix Append this to the filename
4453 * @throws MWException
4454 * @return string $prefix . $mangledCode . $suffix
4456 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4457 if ( !self
::isValidBuiltInCode( $code ) ) {
4458 throw new MWException( "Invalid language code \"$code\"" );
4461 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4465 * @param string $code
4468 public static function getMessagesFileName( $code ) {
4470 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4471 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4476 * @param string $code
4478 * @throws MWException
4481 public static function getJsonMessagesFileName( $code ) {
4484 if ( !self
::isValidBuiltInCode( $code ) ) {
4485 throw new MWException( "Invalid language code \"$code\"" );
4488 return "$IP/languages/i18n/$code.json";
4492 * Get the first fallback for a given language.
4494 * @param string $code
4496 * @return bool|string
4498 public static function getFallbackFor( $code ) {
4499 $fallbacks = self
::getFallbacksFor( $code );
4501 return $fallbacks[0];
4507 * Get the ordered list of fallback languages.
4510 * @param string $code Language code
4511 * @param int $mode Fallback mode, either MESSAGES_FALLBACKS (which always falls back to 'en'),
4512 * or STRICT_FALLBACKS (whic honly falls back to 'en' when explicitly defined)
4513 * @throws MWException
4514 * @return array List of language codes
4516 public static function getFallbacksFor( $code, $mode = self
::MESSAGES_FALLBACKS
) {
4517 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4521 case self
::MESSAGES_FALLBACKS
:
4522 // For unknown languages, fallbackSequence returns an empty array,
4523 // hardcode fallback to 'en' in that case as English messages are
4525 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4526 case self
::STRICT_FALLBACKS
:
4527 // Use this mode when you don't want to fallback to English unless
4528 // explicitly defined, for example when you have language-variant icons
4529 // and an international language-independent fallback.
4530 return self
::getLocalisationCache()->getItem( $code, 'originalFallbackSequence' );
4532 throw new MWException( "Invalid fallback mode \"$mode\"" );
4537 * Get the ordered list of fallback languages, ending with the fallback
4538 * language chain for the site language.
4541 * @param string $code Language code
4542 * @return array [ fallbacks, site fallbacks ]
4544 public static function getFallbacksIncludingSiteLanguage( $code ) {
4545 global $wgLanguageCode;
4547 // Usually, we will only store a tiny number of fallback chains, so we
4548 // keep them in static memory.
4549 $cacheKey = "{$code}-{$wgLanguageCode}";
4551 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4552 $fallbacks = self
::getFallbacksFor( $code );
4554 // Append the site's fallback chain, including the site language itself
4555 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4556 array_unshift( $siteFallbacks, $wgLanguageCode );
4558 // Eliminate any languages already included in the chain
4559 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4561 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4563 return self
::$fallbackLanguageCache[$cacheKey];
4567 * Get all messages for a given language
4568 * WARNING: this may take a long time. If you just need all message *keys*
4569 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4571 * @param string $code
4575 public static function getMessagesFor( $code ) {
4576 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4580 * Get a message for a given language
4582 * @param string $key
4583 * @param string $code
4587 public static function getMessageFor( $key, $code ) {
4588 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4592 * Get all message keys for a given language. This is a faster alternative to
4593 * array_keys( Language::getMessagesFor( $code ) )
4596 * @param string $code Language code
4597 * @return array Array of message keys (strings)
4599 public static function getMessageKeysFor( $code ) {
4600 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4604 * @param string $talk
4607 function fixVariableInNamespace( $talk ) {
4608 if ( strpos( $talk, '$1' ) === false ) {
4612 global $wgMetaNamespace;
4613 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4615 # Allow grammar transformations
4616 # Allowing full message-style parsing would make simple requests
4617 # such as action=raw much more expensive than they need to be.
4618 # This will hopefully cover most cases.
4619 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4620 [ $this, 'replaceGrammarInNamespace' ], $talk );
4621 return str_replace( ' ', '_', $talk );
4628 function replaceGrammarInNamespace( $m ) {
4629 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4633 * Decode an expiry (block, protection, etc) which has come from the DB
4635 * @param string $expiry Database expiry String
4636 * @param bool|int $format True to process using language functions, or TS_ constant
4637 * to return the expiry in a given timestamp
4638 * @param string $infinity If $format is not true, use this string for infinite expiry
4642 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4644 if ( $dbInfinity === null ) {
4645 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4648 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4649 return $format === true
4650 ?
$this->getMessageFromDB( 'infiniteblock' )
4653 return $format === true
4654 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4655 : wfTimestamp( $format, $expiry );
4660 * Formats a time given in seconds into a string representation of that time.
4662 * @param int|float $seconds
4663 * @param array $format An optional argument that formats the returned string in different ways:
4664 * If $format['avoid'] === 'avoidhours': don't show hours, just show days
4665 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4666 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4667 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4669 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4670 * or 'avoidminutes'.
4673 function formatTimePeriod( $seconds, $format = [] ) {
4674 if ( !is_array( $format ) ) {
4675 $format = [ 'avoid' => $format ]; // For backwards compatibility
4677 if ( !isset( $format['avoid'] ) ) {
4678 $format['avoid'] = false;
4680 if ( !isset( $format['noabbrevs'] ) ) {
4681 $format['noabbrevs'] = false;
4683 $secondsMsg = wfMessage(
4684 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4685 $minutesMsg = wfMessage(
4686 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4687 $hoursMsg = wfMessage(
4688 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4689 $daysMsg = wfMessage(
4690 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4692 if ( round( $seconds * 10 ) < 100 ) {
4693 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4694 $s = $secondsMsg->params( $s )->text();
4695 } elseif ( round( $seconds ) < 60 ) {
4696 $s = $this->formatNum( round( $seconds ) );
4697 $s = $secondsMsg->params( $s )->text();
4698 } elseif ( round( $seconds ) < 3600 ) {
4699 $minutes = floor( $seconds / 60 );
4700 $secondsPart = round( fmod( $seconds, 60 ) );
4701 if ( $secondsPart == 60 ) {
4705 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4707 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4708 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4709 $hours = floor( $seconds / 3600 );
4710 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4711 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4712 if ( $secondsPart == 60 ) {
4716 if ( $minutes == 60 ) {
4720 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4722 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4723 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes', 'avoidhours' ] ) ) {
4724 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4727 $days = floor( $seconds / 86400 );
4728 if ( $format['avoid'] === 'avoidhours' ) {
4729 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4730 if ( $hours == 24 ) {
4734 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4735 } elseif ( $format['avoid'] === 'avoidminutes' ) {
4736 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4737 if ( $hours == 24 ) {
4741 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4743 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4744 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4745 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4746 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4747 if ( $minutes == 60 ) {
4751 if ( $hours == 24 ) {
4755 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4757 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4759 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4761 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4763 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4770 * Format a bitrate for output, using an appropriate
4771 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4772 * the magnitude in question.
4774 * This use base 1000. For base 1024 use formatSize(), for another base
4775 * see formatComputingNumbers().
4780 function formatBitrate( $bps ) {
4781 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4785 * @param int $size Size of the unit
4786 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4787 * @param string $messageKey Message key to be uesd
4790 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4792 return str_replace( '$1', $this->formatNum( $size ),
4793 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4796 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4799 $maxIndex = count( $sizes ) - 1;
4800 while ( $size >= $boundary && $index < $maxIndex ) {
4805 // For small sizes no decimal places necessary
4808 // For MB and bigger two decimal places are smarter
4811 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4813 $size = round( $size, $round );
4814 $text = $this->getMessageFromDB( $msg );
4815 return str_replace( '$1', $this->formatNum( $size ), $text );
4819 * Format a size in bytes for output, using an appropriate
4820 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4822 * This method use base 1024. For base 1000 use formatBitrate(), for
4823 * another base see formatComputingNumbers()
4825 * @param int $size Size to format
4826 * @return string Plain text (not HTML)
4828 function formatSize( $size ) {
4829 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4833 * Make a list item, used by various special pages
4835 * @param string $page Page link
4836 * @param string $details HTML safe text between brackets
4837 * @param bool $oppositedm Add the direction mark opposite to your
4838 * language, to display text properly
4839 * @return string HTML escaped
4841 function specialList( $page, $details, $oppositedm = true ) {
4846 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4849 $this->msg( 'word-separator' )->escaped() .
4850 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4854 * Generate (prev x| next x) (20|50|100...) type links for paging
4856 * @param Title $title Title object to link
4857 * @param int $offset
4859 * @param array $query Optional URL query parameter string
4860 * @param bool $atend Optional param for specified if this is the last page
4862 * @deprecated since 1.34, use PrevNextNavigationRenderer::buildPrevNextNavigation()
4865 public function viewPrevNext( Title
$title, $offset, $limit,
4866 array $query = [], $atend = false
4868 wfDeprecated( __METHOD__
, '1.34' );
4869 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4871 # Make 'previous' link
4872 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4873 if ( $offset > 0 ) {
4874 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4875 $query, $prev, 'prevn-title', 'mw-prevlink' );
4877 $plink = htmlspecialchars( $prev );
4881 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4883 $nlink = htmlspecialchars( $next );
4885 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4886 $query, $next, 'nextn-title', 'mw-nextlink' );
4889 # Make links to set number of items per page
4891 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4892 $numLinks[] = $this->numLink( $title, $offset, $num,
4893 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4896 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4897 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4901 * Helper function for viewPrevNext() that generates links
4903 * @param Title $title Title object to link
4904 * @param int $offset
4906 * @param array $query Extra query parameters
4907 * @param string $link Text to use for the link; will be escaped
4908 * @param string $tooltipMsg Name of the message to use as tooltip
4909 * @param string $class Value of the "class" attribute of the link
4910 * @return string HTML fragment
4912 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4915 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4916 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4917 ->numParams( $limit )->text();
4919 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4920 'title' => $tooltip, 'class' => $class ], $link );
4924 * Get the conversion rule title, if any.
4928 public function getConvRuleTitle() {
4929 return $this->mConverter
->getConvRuleTitle();
4933 * Get the compiled plural rules for the language
4935 * @return array Associative array with plural form, and plural rule as key-value pairs
4937 public function getCompiledPluralRules() {
4938 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4939 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4940 if ( !$pluralRules ) {
4941 foreach ( $fallbacks as $fallbackCode ) {
4942 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4943 if ( $pluralRules ) {
4948 return $pluralRules;
4952 * Get the plural rules for the language
4954 * @return array Associative array with plural form number and plural rule as key-value pairs
4956 public function getPluralRules() {
4957 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4958 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4959 if ( !$pluralRules ) {
4960 foreach ( $fallbacks as $fallbackCode ) {
4961 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4962 if ( $pluralRules ) {
4967 return $pluralRules;
4971 * Get the plural rule types for the language
4973 * @return array Associative array with plural form number and plural rule type as key-value pairs
4975 public function getPluralRuleTypes() {
4976 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4977 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4978 if ( !$pluralRuleTypes ) {
4979 foreach ( $fallbacks as $fallbackCode ) {
4980 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4981 if ( $pluralRuleTypes ) {
4986 return $pluralRuleTypes;
4990 * Find the index number of the plural rule appropriate for the given number
4991 * @param int $number
4992 * @return int The index number of the plural rule
4994 public function getPluralRuleIndexNumber( $number ) {
4995 $pluralRules = $this->getCompiledPluralRules();
4996 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
5001 * Find the plural rule type appropriate for the given number
5002 * For example, if the language is set to Arabic, getPluralType(5) should
5005 * @param int $number
5006 * @return string The name of the plural rule type, e.g. one, two, few, many
5008 public function getPluralRuleType( $number ) {
5009 $index = $this->getPluralRuleIndexNumber( $number );
5010 $pluralRuleTypes = $this->getPluralRuleTypes();
5011 return $pluralRuleTypes[$index] ??
'other';