3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
26 * @defgroup Language Language
29 use CLDRPluralRuleParser\Evaluator
;
30 use MediaWiki\MediaWikiServices
;
31 use Wikimedia\Assert\Assert
;
34 * Internationalisation code
39 * Return autonyms in fetchLanguageName(s).
42 const AS_AUTONYMS
= null;
45 * Return all known languages in fetchLanguageName(s).
51 * Return in fetchLanguageName(s) only the languages for which we have at
52 * least some localisation.
55 const SUPPORTED
= 'mwfile';
58 * @var LanguageConverter|FakeConverter
62 public $mVariants, $mCode, $mLoaded = false;
63 public $mMagicExtensions = [];
64 private $mHtmlCode = null;
65 /** @var Language|false */
66 private $mParentLanguage = false;
68 public $dateFormatStrings = [];
69 public $mExtendedSpecialPageAliases;
71 /** @var array|null */
72 protected $namespaceNames;
73 protected $mNamespaceIds, $namespaceAliases;
76 * ReplacementArray object caches
78 public $transformData = [];
81 * @var LocalisationCache
83 public static $dataCache;
85 public static $mLangObjCache = [];
88 * Return a fallback chain for messages in getFallbacksFor
91 const MESSAGES_FALLBACKS
= 0;
94 * Return a strict fallback chain in getFallbacksFor
97 const STRICT_FALLBACKS
= 1;
99 public static $mWeekdayMsgs = [
100 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
104 public static $mWeekdayAbbrevMsgs = [
105 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
108 public static $mMonthMsgs = [
109 'january', 'february', 'march', 'april', 'may_long', 'june',
110 'july', 'august', 'september', 'october', 'november',
113 public static $mMonthGenMsgs = [
114 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
115 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
118 public static $mMonthAbbrevMsgs = [
119 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
120 'sep', 'oct', 'nov', 'dec'
123 public static $mIranianCalendarMonthMsgs = [
124 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
125 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
126 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
127 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
130 public static $mHebrewCalendarMonthMsgs = [
131 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
132 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
133 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
134 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
135 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
138 public static $mHebrewCalendarMonthGenMsgs = [
139 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
140 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
141 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
142 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
143 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
146 public static $mHijriCalendarMonthMsgs = [
147 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
148 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
149 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
150 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
157 public static $durationIntervals = [
158 'millennia' => 31556952000,
159 'centuries' => 3155695200,
160 'decades' => 315569520,
161 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
170 * Cache for language fallbacks.
171 * @see Language::getFallbacksIncludingSiteLanguage
175 private static $fallbackLanguageCache = [];
178 * Cache for grammar rules data
179 * @var MapCacheLRU|null
181 private static $grammarTransformations;
184 * Cache for language names
185 * @var HashBagOStuff|null
187 private static $languageNameCache;
190 * Unicode directional formatting characters, for embedBidi()
192 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
193 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
194 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
197 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
198 * - in group 1 if it is LTR
199 * - in group 2 if it is RTL
200 * Does not match if there is no strong directionality codepoint.
202 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
204 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
205 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
207 // @codeCoverageIgnoreStart
208 // phpcs:ignore Generic.Files.LineLength
209 private static $strongDirRegex = '/(?:([\x{41}-\x{5a}\x{61}-\x{7a}\x{aa}\x{b5}\x{ba}\x{c0}-\x{d6}\x{d8}-\x{f6}\x{f8}-\x{2b8}\x{2bb}-\x{2c1}\x{2d0}\x{2d1}\x{2e0}-\x{2e4}\x{2ee}\x{370}-\x{373}\x{376}\x{377}\x{37a}-\x{37d}\x{37f}\x{386}\x{388}-\x{38a}\x{38c}\x{38e}-\x{3a1}\x{3a3}-\x{3f5}\x{3f7}-\x{482}\x{48a}-\x{52f}\x{531}-\x{556}\x{559}-\x{55f}\x{561}-\x{587}\x{589}\x{903}-\x{939}\x{93b}\x{93d}-\x{940}\x{949}-\x{94c}\x{94e}-\x{950}\x{958}-\x{961}\x{964}-\x{980}\x{982}\x{983}\x{985}-\x{98c}\x{98f}\x{990}\x{993}-\x{9a8}\x{9aa}-\x{9b0}\x{9b2}\x{9b6}-\x{9b9}\x{9bd}-\x{9c0}\x{9c7}\x{9c8}\x{9cb}\x{9cc}\x{9ce}\x{9d7}\x{9dc}\x{9dd}\x{9df}-\x{9e1}\x{9e6}-\x{9f1}\x{9f4}-\x{9fa}\x{a03}\x{a05}-\x{a0a}\x{a0f}\x{a10}\x{a13}-\x{a28}\x{a2a}-\x{a30}\x{a32}\x{a33}\x{a35}\x{a36}\x{a38}\x{a39}\x{a3e}-\x{a40}\x{a59}-\x{a5c}\x{a5e}\x{a66}-\x{a6f}\x{a72}-\x{a74}\x{a83}\x{a85}-\x{a8d}\x{a8f}-\x{a91}\x{a93}-\x{aa8}\x{aaa}-\x{ab0}\x{ab2}\x{ab3}\x{ab5}-\x{ab9}\x{abd}-\x{ac0}\x{ac9}\x{acb}\x{acc}\x{ad0}\x{ae0}\x{ae1}\x{ae6}-\x{af0}\x{af9}\x{b02}\x{b03}\x{b05}-\x{b0c}\x{b0f}\x{b10}\x{b13}-\x{b28}\x{b2a}-\x{b30}\x{b32}\x{b33}\x{b35}-\x{b39}\x{b3d}\x{b3e}\x{b40}\x{b47}\x{b48}\x{b4b}\x{b4c}\x{b57}\x{b5c}\x{b5d}\x{b5f}-\x{b61}\x{b66}-\x{b77}\x{b83}\x{b85}-\x{b8a}\x{b8e}-\x{b90}\x{b92}-\x{b95}\x{b99}\x{b9a}\x{b9c}\x{b9e}\x{b9f}\x{ba3}\x{ba4}\x{ba8}-\x{baa}\x{bae}-\x{bb9}\x{bbe}\x{bbf}\x{bc1}\x{bc2}\x{bc6}-\x{bc8}\x{bca}-\x{bcc}\x{bd0}\x{bd7}\x{be6}-\x{bf2}\x{c01}-\x{c03}\x{c05}-\x{c0c}\x{c0e}-\x{c10}\x{c12}-\x{c28}\x{c2a}-\x{c39}\x{c3d}\x{c41}-\x{c44}\x{c58}-\x{c5a}\x{c60}\x{c61}\x{c66}-\x{c6f}\x{c7f}\x{c82}\x{c83}\x{c85}-\x{c8c}\x{c8e}-\x{c90}\x{c92}-\x{ca8}\x{caa}-\x{cb3}\x{cb5}-\x{cb9}\x{cbd}-\x{cc4}\x{cc6}-\x{cc8}\x{cca}\x{ccb}\x{cd5}\x{cd6}\x{cde}\x{ce0}\x{ce1}\x{ce6}-\x{cef}\x{cf1}\x{cf2}\x{d02}\x{d03}\x{d05}-\x{d0c}\x{d0e}-\x{d10}\x{d12}-\x{d3a}\x{d3d}-\x{d40}\x{d46}-\x{d48}\x{d4a}-\x{d4c}\x{d4e}\x{d57}\x{d5f}-\x{d61}\x{d66}-\x{d75}\x{d79}-\x{d7f}\x{d82}\x{d83}\x{d85}-\x{d96}\x{d9a}-\x{db1}\x{db3}-\x{dbb}\x{dbd}\x{dc0}-\x{dc6}\x{dcf}-\x{dd1}\x{dd8}-\x{ddf}\x{de6}-\x{def}\x{df2}-\x{df4}\x{e01}-\x{e30}\x{e32}\x{e33}\x{e40}-\x{e46}\x{e4f}-\x{e5b}\x{e81}\x{e82}\x{e84}\x{e87}\x{e88}\x{e8a}\x{e8d}\x{e94}-\x{e97}\x{e99}-\x{e9f}\x{ea1}-\x{ea3}\x{ea5}\x{ea7}\x{eaa}\x{eab}\x{ead}-\x{eb0}\x{eb2}\x{eb3}\x{ebd}\x{ec0}-\x{ec4}\x{ec6}\x{ed0}-\x{ed9}\x{edc}-\x{edf}\x{f00}-\x{f17}\x{f1a}-\x{f34}\x{f36}\x{f38}\x{f3e}-\x{f47}\x{f49}-\x{f6c}\x{f7f}\x{f85}\x{f88}-\x{f8c}\x{fbe}-\x{fc5}\x{fc7}-\x{fcc}\x{fce}-\x{fda}\x{1000}-\x{102c}\x{1031}\x{1038}\x{103b}\x{103c}\x{103f}-\x{1057}\x{105a}-\x{105d}\x{1061}-\x{1070}\x{1075}-\x{1081}\x{1083}\x{1084}\x{1087}-\x{108c}\x{108e}-\x{109c}\x{109e}-\x{10c5}\x{10c7}\x{10cd}\x{10d0}-\x{1248}\x{124a}-\x{124d}\x{1250}-\x{1256}\x{1258}\x{125a}-\x{125d}\x{1260}-\x{1288}\x{128a}-\x{128d}\x{1290}-\x{12b0}\x{12b2}-\x{12b5}\x{12b8}-\x{12be}\x{12c0}\x{12c2}-\x{12c5}\x{12c8}-\x{12d6}\x{12d8}-\x{1310}\x{1312}-\x{1315}\x{1318}-\x{135a}\x{1360}-\x{137c}\x{1380}-\x{138f}\x{13a0}-\x{13f5}\x{13f8}-\x{13fd}\x{1401}-\x{167f}\x{1681}-\x{169a}\x{16a0}-\x{16f8}\x{1700}-\x{170c}\x{170e}-\x{1711}\x{1720}-\x{1731}\x{1735}\x{1736}\x{1740}-\x{1751}\x{1760}-\x{176c}\x{176e}-\x{1770}\x{1780}-\x{17b3}\x{17b6}\x{17be}-\x{17c5}\x{17c7}\x{17c8}\x{17d4}-\x{17da}\x{17dc}\x{17e0}-\x{17e9}\x{1810}-\x{1819}\x{1820}-\x{1877}\x{1880}-\x{18a8}\x{18aa}\x{18b0}-\x{18f5}\x{1900}-\x{191e}\x{1923}-\x{1926}\x{1929}-\x{192b}\x{1930}\x{1931}\x{1933}-\x{1938}\x{1946}-\x{196d}\x{1970}-\x{1974}\x{1980}-\x{19ab}\x{19b0}-\x{19c9}\x{19d0}-\x{19da}\x{1a00}-\x{1a16}\x{1a19}\x{1a1a}\x{1a1e}-\x{1a55}\x{1a57}\x{1a61}\x{1a63}\x{1a64}\x{1a6d}-\x{1a72}\x{1a80}-\x{1a89}\x{1a90}-\x{1a99}\x{1aa0}-\x{1aad}\x{1b04}-\x{1b33}\x{1b35}\x{1b3b}\x{1b3d}-\x{1b41}\x{1b43}-\x{1b4b}\x{1b50}-\x{1b6a}\x{1b74}-\x{1b7c}\x{1b82}-\x{1ba1}\x{1ba6}\x{1ba7}\x{1baa}\x{1bae}-\x{1be5}\x{1be7}\x{1bea}-\x{1bec}\x{1bee}\x{1bf2}\x{1bf3}\x{1bfc}-\x{1c2b}\x{1c34}\x{1c35}\x{1c3b}-\x{1c49}\x{1c4d}-\x{1c7f}\x{1cc0}-\x{1cc7}\x{1cd3}\x{1ce1}\x{1ce9}-\x{1cec}\x{1cee}-\x{1cf3}\x{1cf5}\x{1cf6}\x{1d00}-\x{1dbf}\x{1e00}-\x{1f15}\x{1f18}-\x{1f1d}\x{1f20}-\x{1f45}\x{1f48}-\x{1f4d}\x{1f50}-\x{1f57}\x{1f59}\x{1f5b}\x{1f5d}\x{1f5f}-\x{1f7d}\x{1f80}-\x{1fb4}\x{1fb6}-\x{1fbc}\x{1fbe}\x{1fc2}-\x{1fc4}\x{1fc6}-\x{1fcc}\x{1fd0}-\x{1fd3}\x{1fd6}-\x{1fdb}\x{1fe0}-\x{1fec}\x{1ff2}-\x{1ff4}\x{1ff6}-\x{1ffc}\x{200e}\x{2071}\x{207f}\x{2090}-\x{209c}\x{2102}\x{2107}\x{210a}-\x{2113}\x{2115}\x{2119}-\x{211d}\x{2124}\x{2126}\x{2128}\x{212a}-\x{212d}\x{212f}-\x{2139}\x{213c}-\x{213f}\x{2145}-\x{2149}\x{214e}\x{214f}\x{2160}-\x{2188}\x{2336}-\x{237a}\x{2395}\x{249c}-\x{24e9}\x{26ac}\x{2800}-\x{28ff}\x{2c00}-\x{2c2e}\x{2c30}-\x{2c5e}\x{2c60}-\x{2ce4}\x{2ceb}-\x{2cee}\x{2cf2}\x{2cf3}\x{2d00}-\x{2d25}\x{2d27}\x{2d2d}\x{2d30}-\x{2d67}\x{2d6f}\x{2d70}\x{2d80}-\x{2d96}\x{2da0}-\x{2da6}\x{2da8}-\x{2dae}\x{2db0}-\x{2db6}\x{2db8}-\x{2dbe}\x{2dc0}-\x{2dc6}\x{2dc8}-\x{2dce}\x{2dd0}-\x{2dd6}\x{2dd8}-\x{2dde}\x{3005}-\x{3007}\x{3021}-\x{3029}\x{302e}\x{302f}\x{3031}-\x{3035}\x{3038}-\x{303c}\x{3041}-\x{3096}\x{309d}-\x{309f}\x{30a1}-\x{30fa}\x{30fc}-\x{30ff}\x{3105}-\x{312d}\x{3131}-\x{318e}\x{3190}-\x{31ba}\x{31f0}-\x{321c}\x{3220}-\x{324f}\x{3260}-\x{327b}\x{327f}-\x{32b0}\x{32c0}-\x{32cb}\x{32d0}-\x{32fe}\x{3300}-\x{3376}\x{337b}-\x{33dd}\x{33e0}-\x{33fe}\x{3400}-\x{4db5}\x{4e00}-\x{9fd5}\x{a000}-\x{a48c}\x{a4d0}-\x{a60c}\x{a610}-\x{a62b}\x{a640}-\x{a66e}\x{a680}-\x{a69d}\x{a6a0}-\x{a6ef}\x{a6f2}-\x{a6f7}\x{a722}-\x{a787}\x{a789}-\x{a7ad}\x{a7b0}-\x{a7b7}\x{a7f7}-\x{a801}\x{a803}-\x{a805}\x{a807}-\x{a80a}\x{a80c}-\x{a824}\x{a827}\x{a830}-\x{a837}\x{a840}-\x{a873}\x{a880}-\x{a8c3}\x{a8ce}-\x{a8d9}\x{a8f2}-\x{a8fd}\x{a900}-\x{a925}\x{a92e}-\x{a946}\x{a952}\x{a953}\x{a95f}-\x{a97c}\x{a983}-\x{a9b2}\x{a9b4}\x{a9b5}\x{a9ba}\x{a9bb}\x{a9bd}-\x{a9cd}\x{a9cf}-\x{a9d9}\x{a9de}-\x{a9e4}\x{a9e6}-\x{a9fe}\x{aa00}-\x{aa28}\x{aa2f}\x{aa30}\x{aa33}\x{aa34}\x{aa40}-\x{aa42}\x{aa44}-\x{aa4b}\x{aa4d}\x{aa50}-\x{aa59}\x{aa5c}-\x{aa7b}\x{aa7d}-\x{aaaf}\x{aab1}\x{aab5}\x{aab6}\x{aab9}-\x{aabd}\x{aac0}\x{aac2}\x{aadb}-\x{aaeb}\x{aaee}-\x{aaf5}\x{ab01}-\x{ab06}\x{ab09}-\x{ab0e}\x{ab11}-\x{ab16}\x{ab20}-\x{ab26}\x{ab28}-\x{ab2e}\x{ab30}-\x{ab65}\x{ab70}-\x{abe4}\x{abe6}\x{abe7}\x{abe9}-\x{abec}\x{abf0}-\x{abf9}\x{ac00}-\x{d7a3}\x{d7b0}-\x{d7c6}\x{d7cb}-\x{d7fb}\x{e000}-\x{fa6d}\x{fa70}-\x{fad9}\x{fb00}-\x{fb06}\x{fb13}-\x{fb17}\x{ff21}-\x{ff3a}\x{ff41}-\x{ff5a}\x{ff66}-\x{ffbe}\x{ffc2}-\x{ffc7}\x{ffca}-\x{ffcf}\x{ffd2}-\x{ffd7}\x{ffda}-\x{ffdc}\x{10000}-\x{1000b}\x{1000d}-\x{10026}\x{10028}-\x{1003a}\x{1003c}\x{1003d}\x{1003f}-\x{1004d}\x{10050}-\x{1005d}\x{10080}-\x{100fa}\x{10100}\x{10102}\x{10107}-\x{10133}\x{10137}-\x{1013f}\x{101d0}-\x{101fc}\x{10280}-\x{1029c}\x{102a0}-\x{102d0}\x{10300}-\x{10323}\x{10330}-\x{1034a}\x{10350}-\x{10375}\x{10380}-\x{1039d}\x{1039f}-\x{103c3}\x{103c8}-\x{103d5}\x{10400}-\x{1049d}\x{104a0}-\x{104a9}\x{10500}-\x{10527}\x{10530}-\x{10563}\x{1056f}\x{10600}-\x{10736}\x{10740}-\x{10755}\x{10760}-\x{10767}\x{11000}\x{11002}-\x{11037}\x{11047}-\x{1104d}\x{11066}-\x{1106f}\x{11082}-\x{110b2}\x{110b7}\x{110b8}\x{110bb}-\x{110c1}\x{110d0}-\x{110e8}\x{110f0}-\x{110f9}\x{11103}-\x{11126}\x{1112c}\x{11136}-\x{11143}\x{11150}-\x{11172}\x{11174}-\x{11176}\x{11182}-\x{111b5}\x{111bf}-\x{111c9}\x{111cd}\x{111d0}-\x{111df}\x{111e1}-\x{111f4}\x{11200}-\x{11211}\x{11213}-\x{1122e}\x{11232}\x{11233}\x{11235}\x{11238}-\x{1123d}\x{11280}-\x{11286}\x{11288}\x{1128a}-\x{1128d}\x{1128f}-\x{1129d}\x{1129f}-\x{112a9}\x{112b0}-\x{112de}\x{112e0}-\x{112e2}\x{112f0}-\x{112f9}\x{11302}\x{11303}\x{11305}-\x{1130c}\x{1130f}\x{11310}\x{11313}-\x{11328}\x{1132a}-\x{11330}\x{11332}\x{11333}\x{11335}-\x{11339}\x{1133d}-\x{1133f}\x{11341}-\x{11344}\x{11347}\x{11348}\x{1134b}-\x{1134d}\x{11350}\x{11357}\x{1135d}-\x{11363}\x{11480}-\x{114b2}\x{114b9}\x{114bb}-\x{114be}\x{114c1}\x{114c4}-\x{114c7}\x{114d0}-\x{114d9}\x{11580}-\x{115b1}\x{115b8}-\x{115bb}\x{115be}\x{115c1}-\x{115db}\x{11600}-\x{11632}\x{1163b}\x{1163c}\x{1163e}\x{11641}-\x{11644}\x{11650}-\x{11659}\x{11680}-\x{116aa}\x{116ac}\x{116ae}\x{116af}\x{116b6}\x{116c0}-\x{116c9}\x{11700}-\x{11719}\x{11720}\x{11721}\x{11726}\x{11730}-\x{1173f}\x{118a0}-\x{118f2}\x{118ff}\x{11ac0}-\x{11af8}\x{12000}-\x{12399}\x{12400}-\x{1246e}\x{12470}-\x{12474}\x{12480}-\x{12543}\x{13000}-\x{1342e}\x{14400}-\x{14646}\x{16800}-\x{16a38}\x{16a40}-\x{16a5e}\x{16a60}-\x{16a69}\x{16a6e}\x{16a6f}\x{16ad0}-\x{16aed}\x{16af5}\x{16b00}-\x{16b2f}\x{16b37}-\x{16b45}\x{16b50}-\x{16b59}\x{16b5b}-\x{16b61}\x{16b63}-\x{16b77}\x{16b7d}-\x{16b8f}\x{16f00}-\x{16f44}\x{16f50}-\x{16f7e}\x{16f93}-\x{16f9f}\x{1b000}\x{1b001}\x{1bc00}-\x{1bc6a}\x{1bc70}-\x{1bc7c}\x{1bc80}-\x{1bc88}\x{1bc90}-\x{1bc99}\x{1bc9c}\x{1bc9f}\x{1d000}-\x{1d0f5}\x{1d100}-\x{1d126}\x{1d129}-\x{1d166}\x{1d16a}-\x{1d172}\x{1d183}\x{1d184}\x{1d18c}-\x{1d1a9}\x{1d1ae}-\x{1d1e8}\x{1d360}-\x{1d371}\x{1d400}-\x{1d454}\x{1d456}-\x{1d49c}\x{1d49e}\x{1d49f}\x{1d4a2}\x{1d4a5}\x{1d4a6}\x{1d4a9}-\x{1d4ac}\x{1d4ae}-\x{1d4b9}\x{1d4bb}\x{1d4bd}-\x{1d4c3}\x{1d4c5}-\x{1d505}\x{1d507}-\x{1d50a}\x{1d50d}-\x{1d514}\x{1d516}-\x{1d51c}\x{1d51e}-\x{1d539}\x{1d53b}-\x{1d53e}\x{1d540}-\x{1d544}\x{1d546}\x{1d54a}-\x{1d550}\x{1d552}-\x{1d6a5}\x{1d6a8}-\x{1d6da}\x{1d6dc}-\x{1d714}\x{1d716}-\x{1d74e}\x{1d750}-\x{1d788}\x{1d78a}-\x{1d7c2}\x{1d7c4}-\x{1d7cb}\x{1d800}-\x{1d9ff}\x{1da37}-\x{1da3a}\x{1da6d}-\x{1da74}\x{1da76}-\x{1da83}\x{1da85}-\x{1da8b}\x{1f110}-\x{1f12e}\x{1f130}-\x{1f169}\x{1f170}-\x{1f19a}\x{1f1e6}-\x{1f202}\x{1f210}-\x{1f23a}\x{1f240}-\x{1f248}\x{1f250}\x{1f251}\x{20000}-\x{2a6d6}\x{2a700}-\x{2b734}\x{2b740}-\x{2b81d}\x{2b820}-\x{2cea1}\x{2f800}-\x{2fa1d}\x{f0000}-\x{ffffd}\x{100000}-\x{10fffd}])|([\x{590}\x{5be}\x{5c0}\x{5c3}\x{5c6}\x{5c8}-\x{5ff}\x{7c0}-\x{7ea}\x{7f4}\x{7f5}\x{7fa}-\x{815}\x{81a}\x{824}\x{828}\x{82e}-\x{858}\x{85c}-\x{89f}\x{200f}\x{fb1d}\x{fb1f}-\x{fb28}\x{fb2a}-\x{fb4f}\x{10800}-\x{1091e}\x{10920}-\x{10a00}\x{10a04}\x{10a07}-\x{10a0b}\x{10a10}-\x{10a37}\x{10a3b}-\x{10a3e}\x{10a40}-\x{10ae4}\x{10ae7}-\x{10b38}\x{10b40}-\x{10e5f}\x{10e7f}-\x{10fff}\x{1e800}-\x{1e8cf}\x{1e8d7}-\x{1edff}\x{1ef00}-\x{1efff}\x{608}\x{60b}\x{60d}\x{61b}-\x{64a}\x{66d}-\x{66f}\x{671}-\x{6d5}\x{6e5}\x{6e6}\x{6ee}\x{6ef}\x{6fa}-\x{710}\x{712}-\x{72f}\x{74b}-\x{7a5}\x{7b1}-\x{7bf}\x{8a0}-\x{8e2}\x{fb50}-\x{fd3d}\x{fd40}-\x{fdcf}\x{fdf0}-\x{fdfc}\x{fdfe}\x{fdff}\x{fe70}-\x{fefe}\x{1ee00}-\x{1eeef}\x{1eef2}-\x{1eeff}]))/u';
210 // @codeCoverageIgnoreEnd
213 * Get a cached or new language object for a given language code
214 * @param string $code
215 * @throws MWException
218 static function factory( $code ) {
219 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
221 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
222 $code = $wgDummyLanguageCodes[$code];
225 // get the language object to process
226 $langObj = self
::$mLangObjCache[$code] ?? self
::newFromCode( $code );
228 // merge the language object in to get it up front in the cache
229 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
230 // get rid of the oldest ones in case we have an overflow
231 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
237 * Create a language object for a given language code
238 * @param string $code
239 * @param bool $fallback Whether we're going through language fallback chain
240 * @throws MWException
243 protected static function newFromCode( $code, $fallback = false ) {
244 if ( !self
::isValidCode( $code ) ) {
245 throw new MWException( "Invalid language code \"$code\"" );
248 if ( !self
::isValidBuiltInCode( $code ) ) {
249 // It's not possible to customise this code with class files, so
250 // just return a Language object. This is to support uselang= hacks.
251 $lang = new Language
;
252 $lang->mCode
= $code;
256 // Check if there is a language class for the code
257 $class = self
::classFromCode( $code, $fallback );
258 // LanguageCode does not inherit Language
259 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
264 // Keep trying the fallback list until we find an existing class
265 $fallbacks = self
::getFallbacksFor( $code );
266 foreach ( $fallbacks as $fallbackCode ) {
267 if ( !self
::isValidBuiltInCode( $fallbackCode ) ) {
268 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
271 $class = self
::classFromCode( $fallbackCode );
272 if ( class_exists( $class ) ) {
274 $lang->mCode
= $code;
279 throw new MWException( "Invalid fallback sequence for language '$code'" );
283 * Intended for tests that may change configuration in a way that invalidates caches.
287 public static function clearCaches() {
288 if ( !defined( 'MW_PHPUNIT_TEST' ) ) {
289 throw new MWException( __METHOD__
. ' must not be used outside tests' );
291 self
::$dataCache = null;
292 // Reinitialize $dataCache, since it's expected to always be available
293 self
::getLocalisationCache();
294 self
::$mLangObjCache = [];
295 self
::$fallbackLanguageCache = [];
296 self
::$grammarTransformations = null;
297 self
::$languageNameCache = null;
301 * Checks whether any localisation is available for that language tag
302 * in MediaWiki (MessagesXx.php exists).
304 * @param string $code Language tag (in lower case)
305 * @return bool Whether language is supported
308 public static function isSupportedLanguage( $code ) {
309 if ( !self
::isValidBuiltInCode( $code ) ) {
313 if ( $code === 'qqq' ) {
317 return is_readable( self
::getMessagesFileName( $code ) ) ||
318 is_readable( self
::getJsonMessagesFileName( $code ) );
322 * Returns true if a language code string is a well-formed language tag
323 * according to RFC 5646.
324 * This function only checks well-formedness; it doesn't check that
325 * language, script or variant codes actually exist in the repositories.
327 * Based on regexes by Mark Davis of the Unicode Consortium:
328 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
330 * @param string $code
331 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
336 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
339 $alphanum = '[a-z0-9]';
340 $x = 'x'; # private use singleton
341 $singleton = '[a-wy-z]'; # other singleton
342 $s = $lenient ?
'[-_]' : '-';
344 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
345 $script = "$alpha{4}"; # ISO 15924
346 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
347 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
348 $extension = "$singleton(?:$s$alphanum{2,8})+";
349 $privateUse = "$x(?:$s$alphanum{1,8})+";
351 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
352 # Since these are limited, this is safe even later changes to the registry --
353 # the only oddity is that it might change the type of the tag, and thus
354 # the results from the capturing groups.
355 # https://www.iana.org/assignments/language-subtag-registry
357 $grandfathered = "en{$s}GB{$s}oed"
358 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
359 . "|no{$s}(?:bok|nyn)"
360 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
361 . "|zh{$s}min{$s}nan";
363 $variantList = "$variant(?:$s$variant)*";
364 $extensionList = "$extension(?:$s$extension)*";
366 $langtag = "(?:($language)"
369 . "(?:$s$variantList)?"
370 . "(?:$s$extensionList)?"
371 . "(?:$s$privateUse)?)";
373 # The final breakdown, with capturing groups for each of these components
374 # The variants, extensions, grandfathered, and private-use may have interior '-'
376 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
378 return (bool)preg_match( "/$root/", strtolower( $code ) );
382 * Returns true if a language code string is of a valid form, whether or
383 * not it exists. This includes codes which are used solely for
384 * customisation via the MediaWiki namespace.
386 * @param string $code
390 public static function isValidCode( $code ) {
392 Assert
::parameterType( 'string', $code, '$code' );
393 if ( !isset( $cache[$code] ) ) {
394 // People think language codes are html safe, so enforce it.
395 // Ideally we should only allow a-zA-Z0-9-
396 // but, .+ and other chars are often used for {{int:}} hacks
397 // see bugs T39564, T39587, T38938
399 // Protect against path traversal
400 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
401 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
403 return $cache[$code];
407 * Returns true if a language code is of a valid form for the purposes of
408 * internal customisation of MediaWiki, via Messages*.php or *.json.
410 * @param string $code
415 public static function isValidBuiltInCode( $code ) {
416 Assert
::parameterType( 'string', $code, '$code' );
418 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
422 * Returns true if a language code is an IETF tag known to MediaWiki.
429 public static function isKnownLanguageTag( $tag ) {
430 // Quick escape for invalid input to avoid exceptions down the line
431 // when code tries to process tags which are not valid at all.
432 if ( !self
::isValidBuiltInCode( $tag ) ) {
436 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
437 || self
::fetchLanguageName( $tag, $tag ) !== ''
446 * Get the LocalisationCache instance
448 * @return LocalisationCache
450 public static function getLocalisationCache() {
451 if ( is_null( self
::$dataCache ) ) {
452 global $wgLocalisationCacheConf;
453 $class = $wgLocalisationCacheConf['class'];
454 self
::$dataCache = new $class( $wgLocalisationCacheConf );
456 return self
::$dataCache;
459 function __construct() {
460 $this->mConverter
= new FakeConverter( $this );
461 // Set the code to the name of the descendant
462 if ( static::class === 'Language' ) {
465 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
467 self
::getLocalisationCache();
471 * Reduce memory usage
472 * @suppress PhanTypeSuspiciousNonTraversableForeach
474 function __destruct() {
475 foreach ( $this as $name => $value ) {
476 unset( $this->$name );
481 * Hook which will be called if this is the content language.
482 * Descendants can use this to register hook functions or modify globals
484 function initContLang() {
491 public function getFallbackLanguages() {
492 return self
::getFallbacksFor( $this->mCode
);
496 * Exports $wgBookstoreListEn
499 public function getBookstoreList() {
500 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
504 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
505 * available in localised form, it will be included in English.
507 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
509 public function getNamespaces() {
510 if ( is_null( $this->namespaceNames
) ) {
511 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
513 $validNamespaces = MediaWikiServices
::getInstance()->getNamespaceInfo()->
514 getCanonicalNamespaces();
516 $this->namespaceNames
= $wgExtraNamespaces +
517 self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
518 $this->namespaceNames +
= $validNamespaces;
520 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
521 if ( $wgMetaNamespaceTalk ) {
522 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
524 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
525 $this->namespaceNames
[NS_PROJECT_TALK
] =
526 $this->fixVariableInNamespace( $talk );
529 # Sometimes a language will be localised but not actually exist on this wiki.
530 // @phan-suppress-next-line PhanTypeMismatchForeach
531 foreach ( $this->namespaceNames
as $key => $text ) {
532 if ( !isset( $validNamespaces[$key] ) ) {
533 unset( $this->namespaceNames
[$key] );
537 # The above mixing may leave namespaces out of canonical order.
538 # Re-order by namespace ID number...
539 // @phan-suppress-next-line PhanTypeMismatchArgumentInternal
540 ksort( $this->namespaceNames
);
542 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
545 return $this->namespaceNames
;
549 * Arbitrarily set all of the namespace names at once. Mainly used for testing
550 * @param array $namespaces Array of namespaces (id => name)
552 public function setNamespaces( array $namespaces ) {
553 $this->namespaceNames
= $namespaces;
554 $this->mNamespaceIds
= null;
558 * Resets all of the namespace caches. Mainly used for testing
560 public function resetNamespaces() {
561 $this->namespaceNames
= null;
562 $this->mNamespaceIds
= null;
563 $this->namespaceAliases
= null;
567 * A convenience function that returns getNamespaces() with spaces instead of underscores
568 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
572 public function getFormattedNamespaces() {
573 $ns = $this->getNamespaces();
574 foreach ( $ns as $k => $v ) {
575 $ns[$k] = strtr( $v, '_', ' ' );
581 * Get a namespace value by key
584 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
585 * echo $mw_ns; // prints 'MediaWiki'
588 * @param int $index The array key of the namespace to return
589 * @return string|bool String if the namespace value exists, otherwise false
591 public function getNsText( $index ) {
592 $ns = $this->getNamespaces();
593 return $ns[$index] ??
false;
597 * A convenience function that returns the same thing as
598 * getNsText() except with '_' changed to ' ', useful for
602 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
603 * echo $mw_ns; // prints 'MediaWiki talk'
606 * @param int $index The array key of the namespace to return
607 * @return string Namespace name without underscores (empty string if namespace does not exist)
609 public function getFormattedNsText( $index ) {
610 $ns = $this->getNsText( $index );
611 return strtr( $ns, '_', ' ' );
615 * Returns gender-dependent namespace alias if available.
616 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
617 * @param int $index Namespace index
618 * @param string $gender Gender key (male, female... )
622 public function getGenderNsText( $index, $gender ) {
623 global $wgExtraGenderNamespaces;
625 $ns = $wgExtraGenderNamespaces +
626 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
628 return $ns[$index][$gender] ??
$this->getNsText( $index );
632 * Whether this language uses gender-dependent namespace aliases.
633 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
637 public function needsGenderDistinction() {
638 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
639 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
640 // $wgExtraGenderNamespaces overrides everything
642 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
643 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
644 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
647 // Check what is in i18n files
648 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
649 return count( $aliases ) > 0;
654 * Get a namespace key by value, case insensitive.
655 * Only matches namespace names for the current language, not the
656 * canonical ones defined in Namespace.php.
658 * @param string $text
659 * @return int|bool An integer if $text is a valid value otherwise false
661 function getLocalNsIndex( $text ) {
662 $lctext = $this->lc( $text );
663 $ids = $this->getNamespaceIds();
664 return $ids[$lctext] ??
false;
670 public function getNamespaceAliases() {
671 if ( is_null( $this->namespaceAliases
) ) {
672 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
676 foreach ( $aliases as $name => $index ) {
677 if ( $index === NS_PROJECT_TALK
) {
678 unset( $aliases[$name] );
679 $name = $this->fixVariableInNamespace( $name );
680 $aliases[$name] = $index;
685 global $wgExtraGenderNamespaces;
686 $genders = $wgExtraGenderNamespaces +
687 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
688 foreach ( $genders as $index => $forms ) {
689 foreach ( $forms as $alias ) {
690 $aliases[$alias] = $index;
694 # Also add converted namespace names as aliases, to avoid confusion.
695 $convertedNames = [];
696 foreach ( $this->getVariants() as $variant ) {
697 if ( $variant === $this->mCode
) {
700 foreach ( $this->getNamespaces() as $ns => $_ ) {
701 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
705 $this->namespaceAliases
= $aliases +
$convertedNames;
707 # Filter out aliases to namespaces that don't exist, e.g. from extensions
708 # that aren't loaded here but are included in the l10n cache.
709 # (array_intersect preserves keys from its first argument)
710 $this->namespaceAliases
= array_intersect(
711 $this->namespaceAliases
,
712 array_keys( $this->getNamespaces() )
716 return $this->namespaceAliases
;
722 public function getNamespaceIds() {
723 if ( is_null( $this->mNamespaceIds
) ) {
724 global $wgNamespaceAliases;
725 # Put namespace names and aliases into a hashtable.
726 # If this is too slow, then we should arrange it so that it is done
727 # before caching. The catch is that at pre-cache time, the above
728 # class-specific fixup hasn't been done.
729 $this->mNamespaceIds
= [];
730 foreach ( $this->getNamespaces() as $index => $name ) {
731 $this->mNamespaceIds
[$this->lc( $name )] = $index;
733 foreach ( $this->getNamespaceAliases() as $name => $index ) {
734 $this->mNamespaceIds
[$this->lc( $name )] = $index;
736 if ( $wgNamespaceAliases ) {
737 foreach ( $wgNamespaceAliases as $name => $index ) {
738 $this->mNamespaceIds
[$this->lc( $name )] = $index;
742 return $this->mNamespaceIds
;
746 * Get a namespace key by value, case insensitive. Canonical namespace
747 * names override custom ones defined for the current language.
749 * @param string $text
750 * @return int|bool An integer if $text is a valid value otherwise false
752 public function getNsIndex( $text ) {
753 $lctext = $this->lc( $text );
754 $ns = MediaWikiServices
::getInstance()->getNamespaceInfo()->
755 getCanonicalIndex( $lctext );
756 if ( $ns !== null ) {
759 $ids = $this->getNamespaceIds();
760 return $ids[$lctext] ??
false;
764 * short names for language variants used for language conversion links.
766 * @param string $code
767 * @param bool $usemsg Use the "variantname-xyz" message if it exists
770 public function getVariantname( $code, $usemsg = true ) {
771 $msg = "variantname-$code";
772 if ( $usemsg && wfMessage( $msg )->exists() ) {
773 return $this->getMessageFromDB( $msg );
775 $name = self
::fetchLanguageName( $code );
777 return $name; # if it's defined as a language name, show that
779 # otherwise, output the language code
785 * @return string[]|bool List of date format preference keys, or false if disabled.
787 public function getDatePreferences() {
788 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
794 function getDateFormats() {
795 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
799 * @return array|string
801 public function getDefaultDateFormat() {
802 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
803 if ( $df === 'dmy or mdy' ) {
804 global $wgAmericanDates;
805 return $wgAmericanDates ?
'mdy' : 'dmy';
814 public function getDatePreferenceMigrationMap() {
815 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
821 public function getExtraUserToggles() {
822 wfDeprecated( __METHOD__
, '1.34' );
830 function getUserToggle( $tog ) {
831 return $this->getMessageFromDB( "tog-$tog" );
835 * Get an array of language names, indexed by code.
836 * @param null|string $inLanguage Code of language in which to return the names
837 * Use self::AS_AUTONYMS for autonyms (native names)
838 * @param string $include One of:
839 * self::ALL all available languages
840 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
841 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
842 * @return array Language code => language name (sorted by key)
845 public static function fetchLanguageNames( $inLanguage = self
::AS_AUTONYMS
, $include = 'mw' ) {
846 $cacheKey = $inLanguage === self
::AS_AUTONYMS ?
'null' : $inLanguage;
847 $cacheKey .= ":$include";
848 if ( self
::$languageNameCache === null ) {
849 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
852 $ret = self
::$languageNameCache->get( $cacheKey );
854 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
855 self
::$languageNameCache->set( $cacheKey, $ret );
861 * Uncached helper for fetchLanguageNames
862 * @param null|string $inLanguage Code of language in which to return the names
863 * Use self::AS_AUTONYMS for autonyms (native names)
864 * @param string $include One of:
865 * self::ALL all available languages
866 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
867 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
868 * @return array Language code => language name (sorted by key)
870 private static function fetchLanguageNamesUncached(
871 $inLanguage = self
::AS_AUTONYMS
,
874 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
876 // If passed an invalid language code to use, fallback to en
877 if ( $inLanguage !== self
::AS_AUTONYMS
&& !self
::isValidCode( $inLanguage ) ) {
884 # TODO: also include when $inLanguage is null, when this code is more efficient
885 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
888 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
889 if ( $wgUsePigLatinVariant ) {
890 // Pig Latin (for variant development)
891 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
894 foreach ( $mwNames as $mwCode => $mwName ) {
895 # - Prefer own MediaWiki native name when not using the hook
896 # - For other names just add if not added through the hook
897 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
898 $names[$mwCode] = $mwName;
902 if ( $include === self
::ALL
) {
908 $coreCodes = array_keys( $mwNames );
909 foreach ( $coreCodes as $coreCode ) {
910 $returnMw[$coreCode] = $names[$coreCode];
913 if ( $include === self
::SUPPORTED
) {
915 # We do this using a foreach over the codes instead of a directory
916 # loop so that messages files in extensions will work correctly.
917 foreach ( $returnMw as $code => $value ) {
918 if ( is_readable( self
::getMessagesFileName( $code ) )
919 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
921 $namesMwFile[$code] = $names[$code];
925 ksort( $namesMwFile );
930 # 'mw' option; default if it's not one of the other two options (all/mwfile)
935 * @param string $code The code of the language for which to get the name
936 * @param null|string $inLanguage Code of language in which to return the name
937 * (SELF::AS_AUTONYMS for autonyms)
938 * @param string $include See fetchLanguageNames()
939 * @return string Language name or empty
942 public static function fetchLanguageName(
944 $inLanguage = self
::AS_AUTONYMS
,
947 $code = strtolower( $code );
948 $array = self
::fetchLanguageNames( $inLanguage, $include );
949 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
953 * Get a message from the MediaWiki namespace.
955 * @param string $msg Message name
958 public function getMessageFromDB( $msg ) {
959 return $this->msg( $msg )->text();
963 * Get message object in this language. Only for use inside this class.
965 * @param string $msg Message name
968 protected function msg( $msg ) {
969 return wfMessage( $msg )->inLanguage( $this );
976 public function getMonthName( $key ) {
977 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
983 public function getMonthNamesArray() {
984 $monthNames = [ '' ];
985 for ( $i = 1; $i < 13; $i++
) {
986 $monthNames[] = $this->getMonthName( $i );
995 public function getMonthNameGen( $key ) {
996 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1000 * @param string $key
1003 public function getMonthAbbreviation( $key ) {
1004 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1010 public function getMonthAbbreviationsArray() {
1011 $monthNames = [ '' ];
1012 for ( $i = 1; $i < 13; $i++
) {
1013 $monthNames[] = $this->getMonthAbbreviation( $i );
1019 * @param string $key
1022 public function getWeekdayName( $key ) {
1023 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1027 * @param string $key
1030 function getWeekdayAbbreviation( $key ) {
1031 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1035 * @param string $key
1038 function getIranianCalendarMonthName( $key ) {
1039 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1043 * @param string $key
1046 function getHebrewCalendarMonthName( $key ) {
1047 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1051 * @param string $key
1054 function getHebrewCalendarMonthNameGen( $key ) {
1055 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1059 * @param string $key
1062 function getHijriCalendarMonthName( $key ) {
1063 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1067 * Pass through result from $dateTimeObj->format()
1068 * @param DateTime|bool|null &$dateTimeObj
1070 * @param DateTimeZone|bool|null $zone
1071 * @param string $code
1074 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1075 if ( !$dateTimeObj ) {
1076 $dateTimeObj = DateTime
::createFromFormat(
1077 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1080 return $dateTimeObj->format( $code );
1084 * This is a workalike of PHP's date() function, but with better
1085 * internationalisation, a reduced set of format characters, and a better
1088 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1089 * the PHP manual for definitions. There are a number of extensions, which
1092 * xn Do not translate digits of the next numeric format character
1093 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1094 * xr Use roman numerals for the next numeric format character
1095 * xh Use hebrew numerals for the next numeric format character
1097 * xg Genitive month name
1099 * xij j (day number) in Iranian calendar
1100 * xiF F (month name) in Iranian calendar
1101 * xin n (month number) in Iranian calendar
1102 * xiy y (two digit year) in Iranian calendar
1103 * xiY Y (full year) in Iranian calendar
1104 * xit t (days in month) in Iranian calendar
1105 * xiz z (day of the year) in Iranian calendar
1107 * xjj j (day number) in Hebrew calendar
1108 * xjF F (month name) in Hebrew calendar
1109 * xjt t (days in month) in Hebrew calendar
1110 * xjx xg (genitive month name) in Hebrew calendar
1111 * xjn n (month number) in Hebrew calendar
1112 * xjY Y (full year) in Hebrew calendar
1114 * xmj j (day number) in Hijri calendar
1115 * xmF F (month name) in Hijri calendar
1116 * xmn n (month number) in Hijri calendar
1117 * xmY Y (full year) in Hijri calendar
1119 * xkY Y (full year) in Thai solar calendar. Months and days are
1120 * identical to the Gregorian calendar
1121 * xoY Y (full year) in Minguo calendar or Juche year.
1122 * Months and days are identical to the
1123 * Gregorian calendar
1124 * xtY Y (full year) in Japanese nengo. Months and days are
1125 * identical to the Gregorian calendar
1127 * Characters enclosed in double quotes will be considered literal (with
1128 * the quotes themselves removed). Unmatched quotes will be considered
1129 * literal quotes. Example:
1131 * "The month is" F => The month is January
1134 * Backslash escaping is also supported.
1136 * Input timestamp is assumed to be pre-normalized to the desired local
1137 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1138 * $ts is UTC if $zone is not given.
1140 * @param string $format
1141 * @param string $ts 14-character timestamp
1144 * @param DateTimeZone|null $zone Timezone of $ts
1145 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1146 * Only makes sense if $ts is the current time.
1147 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1149 * @throws MWException
1152 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1157 $dateTimeObj = false;
1166 $usedSecond = false;
1167 $usedMinute = false;
1174 $usedISOYear = false;
1175 $usedIsLeapYear = false;
1177 $usedHebrewMonth = false;
1178 $usedIranianMonth = false;
1179 $usedHijriMonth = false;
1180 $usedHebrewYear = false;
1181 $usedIranianYear = false;
1182 $usedHijriYear = false;
1183 $usedTennoYear = false;
1185 if ( strlen( $ts ) !== 14 ) {
1186 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1189 if ( !ctype_digit( $ts ) ) {
1190 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1193 $formatLength = strlen( $format );
1194 for ( $p = 0; $p < $formatLength; $p++
) {
1196 $code = $format[$p];
1197 if ( $code == 'x' && $p < $formatLength - 1 ) {
1198 $code .= $format[++
$p];
1201 if ( ( $code === 'xi'
1207 && $p < $formatLength - 1 ) {
1208 $code .= $format[++
$p];
1219 $rawToggle = !$rawToggle;
1229 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1232 $usedHebrewMonth = true;
1234 $hebrew = self
::tsToHebrew( $ts );
1236 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1240 $num = substr( $ts, 6, 2 );
1244 $s .= $this->getWeekdayAbbreviation(
1245 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1250 $num = intval( substr( $ts, 6, 2 ) );
1255 $iranian = self
::tsToIranian( $ts );
1262 $hijri = self
::tsToHijri( $ts );
1269 $hebrew = self
::tsToHebrew( $ts );
1275 $s .= $this->getWeekdayName(
1276 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1281 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1284 $usedIranianMonth = true;
1286 $iranian = self
::tsToIranian( $ts );
1288 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1291 $usedHijriMonth = true;
1293 $hijri = self
::tsToHijri( $ts );
1295 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1298 $usedHebrewMonth = true;
1300 $hebrew = self
::tsToHebrew( $ts );
1302 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1306 $num = substr( $ts, 4, 2 );
1310 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1314 $num = intval( substr( $ts, 4, 2 ) );
1317 $usedIranianMonth = true;
1319 $iranian = self
::tsToIranian( $ts );
1324 $usedHijriMonth = true;
1326 $hijri = self
::tsToHijri( $ts );
1331 $usedHebrewMonth = true;
1333 $hebrew = self
::tsToHebrew( $ts );
1338 $usedHebrewMonth = true;
1340 $hebrew = self
::tsToHebrew( $ts );
1346 $num = substr( $ts, 0, 4 );
1349 $usedIranianYear = true;
1351 $iranian = self
::tsToIranian( $ts );
1356 $usedHijriYear = true;
1358 $hijri = self
::tsToHijri( $ts );
1363 $usedHebrewYear = true;
1365 $hebrew = self
::tsToHebrew( $ts );
1372 $thai = self
::tsToYear( $ts, 'thai' );
1379 $minguo = self
::tsToYear( $ts, 'minguo' );
1384 $usedTennoYear = true;
1386 $tenno = self
::tsToYear( $ts, 'tenno' );
1392 $num = substr( $ts, 2, 2 );
1395 $usedIranianYear = true;
1397 $iranian = self
::tsToIranian( $ts );
1399 $num = substr( $iranian[0], -2 );
1402 $usedIranianYear = true;
1404 $iranian = self
::tsToIranian( $ts );
1406 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1409 $usedIranianYear = true;
1411 $iranian = self
::tsToIranian( $ts );
1417 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1421 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1425 $h = substr( $ts, 8, 2 );
1426 $num = $h %
12 ?
: 12;
1430 $num = intval( substr( $ts, 8, 2 ) );
1434 $h = substr( $ts, 8, 2 );
1435 $num = sprintf( '%02d', $h %
12 ?
: 12 );
1439 $num = substr( $ts, 8, 2 );
1443 $num = substr( $ts, 10, 2 );
1447 $num = substr( $ts, 12, 2 );
1457 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1463 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1467 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1471 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1474 $usedIsLeapYear = true;
1475 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1478 $usedISOYear = true;
1479 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1486 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1489 # Backslash escaping
1490 if ( $p < $formatLength - 1 ) {
1491 $s .= $format[++
$p];
1498 if ( $p < $formatLength - 1 ) {
1499 $endQuote = strpos( $format, '"', $p +
1 );
1500 if ( $endQuote === false ) {
1501 # No terminating quote, assume literal "
1504 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1508 # Quote at end of string, assume literal "
1515 if ( $num !== false ) {
1516 if ( $rawToggle ||
$raw ) {
1519 } elseif ( $roman ) {
1520 $s .= self
::romanNumeral( $num );
1522 } elseif ( $hebrewNum ) {
1523 $s .= self
::hebrewNumeral( $num );
1526 $s .= $this->formatNum( $num, true );
1531 if ( $ttl === 'unused' ) {
1532 // No need to calculate the TTL, the caller wont use it anyway.
1533 } elseif ( $usedSecond ) {
1535 } elseif ( $usedMinute ) {
1536 $ttl = 60 - substr( $ts, 12, 2 );
1537 } elseif ( $usedHour ) {
1538 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1539 } elseif ( $usedAMPM ) {
1540 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1541 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1545 $usedIranianMonth ||
1552 // @todo Someone who understands the non-Gregorian calendars
1553 // should write proper logic for them so that they don't need purged every day.
1554 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1555 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1558 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1559 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1562 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1563 $timeRemainingInDay;
1564 } elseif ( $usedISOYear ) {
1565 // December 28th falls on the last ISO week of the year, every year.
1566 // The last ISO week of a year can be 52 or 53.
1567 $lastWeekOfISOYear = DateTime
::createFromFormat(
1569 substr( $ts, 0, 4 ) . '1228',
1570 $zone ?
: new DateTimeZone( 'UTC' )
1572 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1573 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1574 $timeRemainingInWeek =
1575 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1576 +
$timeRemainingInDay;
1577 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1582 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1583 substr( $ts, 6, 2 ) ) * 86400
1584 +
$timeRemainingInDay;
1585 } elseif ( $usedYear ) {
1587 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1588 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1589 +
$timeRemainingInDay;
1590 } elseif ( $usedIsLeapYear ) {
1591 $year = substr( $ts, 0, 4 );
1592 $timeRemainingInYear =
1593 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1594 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1595 +
$timeRemainingInDay;
1597 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1598 // this isn't a leap year. see when the next one starts
1599 $nextCandidate = $year - $mod +
4;
1600 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1601 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1602 $timeRemainingInYear;
1604 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1605 $timeRemainingInYear;
1608 // this is a leap year, so the next year isn't
1609 $possibleTtls[] = $timeRemainingInYear;
1613 if ( $possibleTtls ) {
1614 $ttl = min( $possibleTtls );
1621 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1622 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1625 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1626 * Gregorian dates to Iranian dates. Originally written in C, it
1627 * is released under the terms of GNU Lesser General Public
1628 * License. Conversion to PHP was performed by Niklas Laxström.
1630 * Link: http://www.farsiweb.info/jalali/jalali.c
1636 private static function tsToIranian( $ts ) {
1637 $gy = substr( $ts, 0, 4 ) - 1600;
1638 $gm = substr( $ts, 4, 2 ) - 1;
1639 $gd = substr( $ts, 6, 2 ) - 1;
1641 # Days passed from the beginning (including leap years)
1643 +
floor( ( $gy +
3 ) / 4 )
1644 - floor( ( $gy +
99 ) / 100 )
1645 +
floor( ( $gy +
399 ) / 400 );
1647 // Add days of the past months of this year
1648 for ( $i = 0; $i < $gm; $i++
) {
1649 $gDayNo +
= self
::$GREG_DAYS[$i];
1653 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1657 // Days passed in current month
1658 $gDayNo +
= (int)$gd;
1660 $jDayNo = $gDayNo - 79;
1662 $jNp = floor( $jDayNo / 12053 );
1665 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1668 if ( $jDayNo >= 366 ) {
1669 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1670 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1675 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1676 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1682 return [ $jy, $jm, $jd, $jz ];
1686 * Converting Gregorian dates to Hijri dates.
1688 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1690 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1696 private static function tsToHijri( $ts ) {
1697 $year = substr( $ts, 0, 4 );
1698 $month = substr( $ts, 4, 2 );
1699 $day = substr( $ts, 6, 2 );
1707 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1708 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1710 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1711 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1712 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1715 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1716 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1719 $zl = $zjd - 1948440 +
10632;
1720 $zn = (int)( ( $zl - 1 ) / 10631 );
1721 $zl = $zl - 10631 * $zn +
354;
1722 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1723 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1724 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1725 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1726 $zm = (int)( ( 24 * $zl ) / 709 );
1727 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1728 $zy = 30 * $zn +
$zj - 30;
1730 return [ $zy, $zm, $zd ];
1734 * Converting Gregorian dates to Hebrew dates.
1736 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1737 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1738 * to translate the relevant functions into PHP and release them under
1741 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1742 * and Adar II is 14. In a non-leap year, Adar is 6.
1748 private static function tsToHebrew( $ts ) {
1750 $year = substr( $ts, 0, 4 );
1751 $month = substr( $ts, 4, 2 );
1752 $day = substr( $ts, 6, 2 );
1754 # Calculate Hebrew year
1755 $hebrewYear = $year +
3760;
1757 # Month number when September = 1, August = 12
1759 if ( $month > 12 ) {
1766 # Calculate day of year from 1 September
1768 for ( $i = 1; $i < $month; $i++
) {
1772 # Check if the year is leap
1773 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1776 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1783 # Calculate the start of the Hebrew year
1784 $start = self
::hebrewYearStart( $hebrewYear );
1786 # Calculate next year's start
1787 if ( $dayOfYear <= $start ) {
1788 # Day is before the start of the year - it is the previous year
1790 $nextStart = $start;
1794 # Add days since previous year's 1 September
1796 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1800 # Start of the new (previous) year
1801 $start = self
::hebrewYearStart( $hebrewYear );
1804 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1807 # Calculate Hebrew day of year
1808 $hebrewDayOfYear = $dayOfYear - $start;
1810 # Difference between year's days
1811 $diff = $nextStart - $start;
1812 # Add 12 (or 13 for leap years) days to ignore the difference between
1813 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1814 # difference is only about the year type
1815 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1821 # Check the year pattern, and is leap year
1822 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1823 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1824 # and non-leap years
1825 $yearPattern = $diff %
30;
1826 # Check if leap year
1827 $isLeap = $diff >= 30;
1829 # Calculate day in the month from number of day in the Hebrew year
1830 # Don't check Adar - if the day is not in Adar, we will stop before;
1831 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1832 $hebrewDay = $hebrewDayOfYear;
1835 while ( $hebrewMonth <= 12 ) {
1836 # Calculate days in this month
1837 if ( $isLeap && $hebrewMonth == 6 ) {
1838 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1840 if ( $hebrewDay <= $days ) {
1844 # Subtract the days of Adar I
1845 $hebrewDay -= $days;
1848 if ( $hebrewDay <= $days ) {
1853 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1854 # Cheshvan in a complete year (otherwise as the rule below)
1856 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1857 # Kislev in an incomplete year (otherwise as the rule below)
1860 # Odd months have 30 days, even have 29
1861 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1863 if ( $hebrewDay <= $days ) {
1864 # In the current month
1867 # Subtract the days of the current month
1868 $hebrewDay -= $days;
1869 # Try in the next month
1874 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1878 * This calculates the Hebrew year start, as days since 1 September.
1879 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1880 * Used for Hebrew date.
1886 private static function hebrewYearStart( $year ) {
1887 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1888 $b = intval( ( $year - 1 ) %
4 );
1889 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1893 $Mar = intval( $m );
1899 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1900 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1902 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1904 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1908 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1913 * Algorithm to convert Gregorian dates to Thai solar dates,
1914 * Minguo dates or Minguo dates.
1916 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1917 * https://en.wikipedia.org/wiki/Minguo_calendar
1918 * https://en.wikipedia.org/wiki/Japanese_era_name
1920 * @param string $ts 14-character timestamp
1921 * @param string $cName Calender name
1922 * @return array Converted year, month, day
1924 private static function tsToYear( $ts, $cName ) {
1925 $gy = substr( $ts, 0, 4 );
1926 $gm = substr( $ts, 4, 2 );
1927 $gd = substr( $ts, 6, 2 );
1929 if ( !strcmp( $cName, 'thai' ) ) {
1931 # Add 543 years to the Gregorian calendar
1932 # Months and days are identical
1933 $gy_offset = $gy +
543;
1934 # fix for dates between 1912 and 1941
1935 # https://en.wikipedia.org/?oldid=836596673#New_year
1936 if ( $gy >= 1912 && $gy <= 1940 ) {
1940 $gm = ( $gm - 3 ) %
12;
1942 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1944 # Deduct 1911 years from the Gregorian calendar
1945 # Months and days are identical
1946 $gy_offset = $gy - 1911;
1947 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1948 # Nengō dates up to Meiji period
1949 # Deduct years from the Gregorian calendar
1950 # depending on the nengo periods
1951 # Months and days are identical
1953 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1954 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1957 $gy_gannen = $gy - 1868 +
1;
1958 $gy_offset = $gy_gannen;
1959 if ( $gy_gannen == 1 ) {
1962 $gy_offset = '明治' . $gy_offset;
1964 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1965 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1966 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1967 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1968 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1971 $gy_gannen = $gy - 1912 +
1;
1972 $gy_offset = $gy_gannen;
1973 if ( $gy_gannen == 1 ) {
1976 $gy_offset = '大正' . $gy_offset;
1978 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1979 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1980 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1983 $gy_gannen = $gy - 1926 +
1;
1984 $gy_offset = $gy_gannen;
1985 if ( $gy_gannen == 1 ) {
1988 $gy_offset = '昭和' . $gy_offset;
1990 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1991 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1992 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1995 $gy_gannen = $gy - 1989 +
1;
1996 $gy_offset = $gy_gannen;
1997 if ( $gy_gannen == 1 ) {
2000 $gy_offset = '平成' . $gy_offset;
2003 $gy_gannen = $gy - 2019 +
1;
2004 $gy_offset = $gy_gannen;
2005 if ( $gy_gannen == 1 ) {
2008 $gy_offset = '令和' . $gy_offset;
2014 return [ $gy_offset, $gm, $gd ];
2018 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2020 * This is the rule the BIDI algorithm uses to determine the directionality of
2021 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
2022 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2024 * TODO: Does not handle BIDI control characters inside the text.
2025 * TODO: Does not handle unallocated characters.
2027 * @param string $text Text to test
2028 * @return null|string Directionality ('ltr' or 'rtl') or null
2030 private static function strongDirFromContent( $text = '' ) {
2031 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2034 if ( $matches[1] === '' ) {
2041 * Roman number formatting up to 10000
2047 static function romanNumeral( $num ) {
2049 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
2050 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
2051 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
2052 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2053 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2056 $num = intval( $num );
2057 if ( $num > 10000 ||
$num <= 0 ) {
2062 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2063 if ( $num >= $pow10 ) {
2064 $s .= $table[$i][(int)floor( $num / $pow10 )];
2066 $num = $num %
$pow10;
2072 * Hebrew Gematria number formatting up to 9999
2078 static function hebrewNumeral( $num ) {
2080 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2081 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2094 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2097 $num = intval( $num );
2098 if ( $num > 9999 ||
$num <= 0 ) {
2102 // Round thousands have special notations
2103 if ( $num === 1000 ) {
2105 } elseif ( $num %
1000 === 0 ) {
2106 return $table[0][$num / 1000] . "' אלפים";
2111 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2112 if ( $num >= $pow10 ) {
2113 if ( $num === 15 ||
$num === 16 ) {
2114 $letters[] = $table[0][9];
2115 $letters[] = $table[0][$num - 9];
2118 $letters = array_merge(
2120 (array)$table[$i][intval( $num / $pow10 )]
2123 if ( $pow10 === 1000 ) {
2129 $num = $num %
$pow10;
2132 $preTransformLength = count( $letters );
2133 if ( $preTransformLength === 1 ) {
2134 // Add geresh (single quote) to one-letter numbers
2137 $lastIndex = $preTransformLength - 1;
2138 $letters[$lastIndex] = str_replace(
2139 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2140 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2141 $letters[$lastIndex]
2144 // Add gershayim (double quote) to multiple-letter numbers,
2145 // but exclude numbers with only one letter after the thousands
2146 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2147 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2150 array_splice( $letters, -1, 0, '"' );
2154 return implode( $letters );
2158 * Used by date() and time() to adjust the time output.
2160 * @param string $ts The time in date('YmdHis') format
2161 * @param mixed $tz Adjust the time by this amount (default false, mean we
2162 * get user timecorrection setting)
2165 public function userAdjust( $ts, $tz = false ) {
2166 global $wgUser, $wgLocalTZoffset;
2168 if ( $tz === false ) {
2169 $tz = $wgUser->getOption( 'timecorrection' );
2172 $data = explode( '|', $tz, 3 );
2174 if ( $data[0] == 'ZoneInfo' ) {
2176 $userTZ = new DateTimeZone( $data[2] );
2177 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2178 $date->setTimezone( $userTZ );
2179 return $date->format( 'YmdHis' );
2180 } catch ( Exception
$e ) {
2181 // Unrecognized timezone, default to 'Offset' with the stored offset.
2182 $data[0] = 'Offset';
2186 if ( $data[0] == 'System' ||
$tz == '' ) {
2187 # Global offset in minutes.
2188 $minDiff = $wgLocalTZoffset;
2189 } elseif ( $data[0] == 'Offset' ) {
2190 $minDiff = intval( $data[1] );
2192 $data = explode( ':', $tz );
2193 if ( count( $data ) == 2 ) {
2194 $data[0] = intval( $data[0] );
2195 $data[1] = intval( $data[1] );
2196 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2197 if ( $data[0] < 0 ) {
2198 $minDiff = -$minDiff;
2201 $minDiff = intval( $data[0] ) * 60;
2205 # No difference ? Return time unchanged
2206 if ( $minDiff == 0 ) {
2210 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2211 # Generate an adjusted date; take advantage of the fact that mktime
2212 # will normalize out-of-range values so we don't have to split $minDiff
2213 # into hours and minutes.
2215 (int)substr( $ts, 8, 2 ) ), # Hours
2216 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2217 (int)substr( $ts, 12, 2 ), # Seconds
2218 (int)substr( $ts, 4, 2 ), # Month
2219 (int)substr( $ts, 6, 2 ), # Day
2220 (int)substr( $ts, 0, 4 ) ); # Year
2222 $date = date( 'YmdHis', $t );
2223 Wikimedia\restoreWarnings
();
2229 * This is meant to be used by time(), date(), and timeanddate() to get
2230 * the date preference they're supposed to use, it should be used in
2233 * function timeanddate([...], $format = true) {
2234 * $datePreference = $this->dateFormat($format);
2238 * @param int|string|bool $usePrefs If true, the user's preference is used
2239 * if false, the site/language default is used
2240 * if int/string, assumed to be a format.
2243 function dateFormat( $usePrefs = true ) {
2246 if ( is_bool( $usePrefs ) ) {
2248 $datePreference = $wgUser->getDatePreference();
2250 $datePreference = (string)User
::getDefaultOption( 'date' );
2253 $datePreference = (string)$usePrefs;
2257 if ( $datePreference == '' ) {
2261 return $datePreference;
2265 * Get a format string for a given type and preference
2266 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2267 * @param string $pref The format name as it appears in Messages*.php under
2270 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2274 function getDateFormatString( $type, $pref ) {
2275 $wasDefault = false;
2276 if ( $pref == 'default' ) {
2278 $pref = $this->getDefaultDateFormat();
2281 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2282 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2284 if ( $type === 'pretty' && $df === null ) {
2285 $df = $this->getDateFormatString( 'date', $pref );
2288 if ( !$wasDefault && $df === null ) {
2289 $pref = $this->getDefaultDateFormat();
2290 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2293 $this->dateFormatStrings
[$type][$pref] = $df;
2295 return $this->dateFormatStrings
[$type][$pref];
2299 * @param string $ts The time format which needs to be turned into a
2300 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2301 * @param bool $adj Whether to adjust the time output according to the
2302 * user configured offset ($timecorrection)
2303 * @param mixed $format True to use user's date format preference
2304 * @param string|bool $timecorrection The time offset as returned by
2305 * validateTimeZone() in Special:Preferences
2308 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2309 $ts = wfTimestamp( TS_MW
, $ts );
2311 $ts = $this->userAdjust( $ts, $timecorrection );
2313 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2314 return $this->sprintfDate( $df, $ts );
2318 * @param string $ts The time format which needs to be turned into a
2319 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2320 * @param bool $adj Whether to adjust the time output according to the
2321 * user configured offset ($timecorrection)
2322 * @param mixed $format True to use user's date format preference
2323 * @param string|bool $timecorrection The time offset as returned by
2324 * validateTimeZone() in Special:Preferences
2327 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2328 $ts = wfTimestamp( TS_MW
, $ts );
2330 $ts = $this->userAdjust( $ts, $timecorrection );
2332 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2333 return $this->sprintfDate( $df, $ts );
2337 * @param string $ts The time format which needs to be turned into a
2338 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2339 * @param bool $adj Whether to adjust the time output according to the
2340 * user configured offset ($timecorrection)
2341 * @param mixed $format What format to return, if it's false output the
2342 * default one (default true)
2343 * @param string|bool $timecorrection The time offset as returned by
2344 * validateTimeZone() in Special:Preferences
2347 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2348 $ts = wfTimestamp( TS_MW
, $ts );
2350 $ts = $this->userAdjust( $ts, $timecorrection );
2352 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2353 return $this->sprintfDate( $df, $ts );
2357 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2361 * @param int $seconds The amount of seconds.
2362 * @param array $chosenIntervals The intervals to enable.
2366 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2367 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2371 foreach ( $intervals as $intervalName => $intervalValue ) {
2372 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2373 // duration-years, duration-decades, duration-centuries, duration-millennia
2374 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2375 $segments[] = $message->inLanguage( $this )->escaped();
2378 return $this->listToText( $segments );
2382 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2383 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2387 * @param int $seconds The amount of seconds.
2388 * @param array $chosenIntervals The intervals to enable.
2392 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2393 if ( empty( $chosenIntervals ) ) {
2394 $chosenIntervals = [
2406 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2407 $sortedNames = array_keys( $intervals );
2408 $smallestInterval = array_pop( $sortedNames );
2412 foreach ( $intervals as $name => $length ) {
2413 $value = floor( $seconds / $length );
2415 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2416 $seconds -= $value * $length;
2417 $segments[$name] = $value;
2425 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2427 * @param string $type Can be 'date', 'time' or 'both'
2428 * @param string $ts The time format which needs to be turned into a
2429 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2430 * @param User $user User object used to get preferences for timezone and format
2431 * @param array $options Array, can contain the following keys:
2432 * - 'timecorrection': time correction, can have the following values:
2433 * - true: use user's preference
2434 * - false: don't use time correction
2435 * - int: value of time correction in minutes
2436 * - 'format': format to use, can have the following values:
2437 * - true: use user's preference
2438 * - false: use default preference
2439 * - string: format to use
2443 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2444 $ts = wfTimestamp( TS_MW
, $ts );
2445 $options +
= [ 'timecorrection' => true, 'format' => true ];
2446 if ( $options['timecorrection'] !== false ) {
2447 if ( $options['timecorrection'] === true ) {
2448 $offset = $user->getOption( 'timecorrection' );
2450 $offset = $options['timecorrection'];
2452 $ts = $this->userAdjust( $ts, $offset );
2454 if ( $options['format'] === true ) {
2455 $format = $user->getDatePreference();
2457 $format = $options['format'];
2459 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2460 return $this->sprintfDate( $df, $ts );
2464 * Get the formatted date for the given timestamp and formatted for
2467 * @param mixed $ts Mixed: the time format which needs to be turned into a
2468 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2469 * @param User $user User object used to get preferences for timezone and format
2470 * @param array $options Array, can contain the following keys:
2471 * - 'timecorrection': time correction, can have the following values:
2472 * - true: use user's preference
2473 * - false: don't use time correction
2474 * - int: value of time correction in minutes
2475 * - 'format': format to use, can have the following values:
2476 * - true: use user's preference
2477 * - false: use default preference
2478 * - string: format to use
2482 public function userDate( $ts, User
$user, array $options = [] ) {
2483 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2487 * Get the formatted time for the given timestamp and formatted for
2490 * @param mixed $ts The time format which needs to be turned into a
2491 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2492 * @param User $user User object used to get preferences for timezone and format
2493 * @param array $options Array, can contain the following keys:
2494 * - 'timecorrection': time correction, can have the following values:
2495 * - true: use user's preference
2496 * - false: don't use time correction
2497 * - int: value of time correction in minutes
2498 * - 'format': format to use, can have the following values:
2499 * - true: use user's preference
2500 * - false: use default preference
2501 * - string: format to use
2505 public function userTime( $ts, User
$user, array $options = [] ) {
2506 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2510 * Get the formatted date and time for the given timestamp and formatted for
2513 * @param mixed $ts The time format which needs to be turned into a
2514 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2515 * @param User $user User object used to get preferences for timezone and format
2516 * @param array $options Array, can contain the following keys:
2517 * - 'timecorrection': time correction, can have the following values:
2518 * - true: use user's preference
2519 * - false: don't use time correction
2520 * - int: value of time correction in minutes
2521 * - 'format': format to use, can have the following values:
2522 * - true: use user's preference
2523 * - false: use default preference
2524 * - string: format to use
2528 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2529 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2533 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2535 * Determine the difference between the timestamp and the current time, and
2536 * generate a readable timestamp by returning "<N> <units> ago", where the
2537 * largest possible unit is used.
2539 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2541 * @param MWTimestamp $time
2542 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2543 * @param User|null $user User the timestamp is being generated for
2544 * (or null to use main context's user)
2545 * @return string Formatted timestamp
2547 public function getHumanTimestamp(
2548 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2550 if ( $relativeTo === null ) {
2551 $relativeTo = new MWTimestamp();
2553 if ( $user === null ) {
2554 $user = RequestContext
::getMain()->getUser();
2557 // Adjust for the user's timezone.
2558 $offsetThis = $time->offsetForUser( $user );
2559 $offsetRel = $relativeTo->offsetForUser( $user );
2562 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2563 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2566 // Reset the timezone on the objects.
2567 $time->timestamp
->sub( $offsetThis );
2568 $relativeTo->timestamp
->sub( $offsetRel );
2574 * Convert an MWTimestamp into a pretty human-readable timestamp using
2575 * the given user preferences and relative base time.
2577 * @see Language::getHumanTimestamp
2578 * @param MWTimestamp $ts Timestamp to prettify
2579 * @param MWTimestamp $relativeTo Base timestamp
2580 * @param User $user User preferences to use
2581 * @return string Human timestamp
2584 private function getHumanTimestampInternal(
2585 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2587 $diff = $ts->diff( $relativeTo );
2588 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2589 (int)$relativeTo->timestamp
->format( 'w' ) );
2590 $days = $diff->days ?
: (int)$diffDay;
2591 if ( $diff->invert ||
$days > 5
2592 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2594 // Timestamps are in different years: use full timestamp
2595 // Also do full timestamp for future dates
2597 * @todo FIXME: Add better handling of future timestamps.
2599 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2600 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2601 } elseif ( $days > 5 ) {
2602 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2603 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2604 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2605 } elseif ( $days > 1 ) {
2606 // Timestamp within the past week: show the day of the week and time
2607 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2608 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2610 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2611 $ts = wfMessage( "$weekday-at" )
2612 ->inLanguage( $this )
2613 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2615 } elseif ( $days == 1 ) {
2616 // Timestamp was yesterday: say 'yesterday' and the time.
2617 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2618 $ts = wfMessage( 'yesterday-at' )
2619 ->inLanguage( $this )
2620 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2622 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2623 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2624 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2625 $ts = wfMessage( 'today-at' )
2626 ->inLanguage( $this )
2627 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2630 // From here on in, the timestamp was soon enough ago so that we can simply say
2631 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2632 } elseif ( $diff->h
== 1 ) {
2633 // Less than 90 minutes, but more than an hour ago.
2634 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2635 } elseif ( $diff->i
>= 1 ) {
2636 // A few minutes ago.
2637 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2638 } elseif ( $diff->s
>= 30 ) {
2639 // Less than a minute, but more than 30 sec ago.
2640 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2642 // Less than 30 seconds ago.
2643 $ts = wfMessage( 'just-now' )->text();
2650 * @param string $key
2651 * @return string|null
2653 public function getMessage( $key ) {
2654 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2660 function getAllMessages() {
2661 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2666 * @param string $out
2667 * @param string $string
2670 public function iconv( $in, $out, $string ) {
2671 # Even with //IGNORE iconv can whine about illegal characters in
2672 # *input* string. We just ignore those too.
2673 # REF: https://bugs.php.net/bug.php?id=37166
2674 # REF: https://phabricator.wikimedia.org/T18885
2675 Wikimedia\
suppressWarnings();
2676 $text = iconv( $in, $out . '//IGNORE', $string );
2677 Wikimedia\restoreWarnings
();
2681 // callback functions for ucwords(), ucwordbreaks()
2684 * @param array $matches
2685 * @return mixed|string
2687 function ucwordbreaksCallbackAscii( $matches ) {
2688 return $this->ucfirst( $matches[1] );
2692 * @param array $matches
2695 function ucwordbreaksCallbackMB( $matches ) {
2696 return mb_strtoupper( $matches[0] );
2700 * @param array $matches
2703 function ucwordsCallbackMB( $matches ) {
2704 return mb_strtoupper( $matches[0] );
2708 * Make a string's first character uppercase
2710 * @param string $str
2714 public function ucfirst( $str ) {
2716 if ( $o < 96 ) { // if already uppercase...
2718 } elseif ( $o < 128 ) {
2719 return ucfirst( $str ); // use PHP's ucfirst()
2721 // fall back to more complex logic in case of multibyte strings
2722 return $this->uc( $str, true );
2727 * Convert a string to uppercase
2729 * @param string $str
2730 * @param bool $first
2734 public function uc( $str, $first = false ) {
2736 if ( $this->isMultibyte( $str ) ) {
2737 return $this->mbUpperChar( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2739 return ucfirst( $str );
2742 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2747 * Convert character to uppercase, allowing overrides of the default mb_upper
2748 * behaviour, which is buggy in many ways. Having a conversion table can be
2749 * useful during transitions between PHP versions where unicode changes happen.
2750 * This can make some resources unreachable on-wiki, see discussion at T219279.
2751 * Providing such a conversion table can allow to manage the transition period.
2755 * @param string $char
2759 protected function mbUpperChar( $char ) {
2760 global $wgOverrideUcfirstCharacters;
2761 if ( array_key_exists( $char, $wgOverrideUcfirstCharacters ) ) {
2762 return $wgOverrideUcfirstCharacters[$char];
2764 return mb_strtoupper( $char );
2769 * @param string $str
2770 * @return mixed|string
2772 function lcfirst( $str ) {
2775 return strval( $str );
2776 } elseif ( $o >= 128 ) {
2777 return $this->lc( $str, true );
2778 } elseif ( $o > 96 ) {
2781 $str[0] = strtolower( $str[0] );
2787 * @param string $str
2788 * @param bool $first
2789 * @return mixed|string
2791 function lc( $str, $first = false ) {
2793 if ( $this->isMultibyte( $str ) ) {
2794 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2796 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2799 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2804 * @param string $str
2807 function isMultibyte( $str ) {
2808 return strlen( $str ) !== mb_strlen( $str );
2812 * @param string $str
2813 * @return mixed|string
2815 function ucwords( $str ) {
2816 if ( $this->isMultibyte( $str ) ) {
2817 $str = $this->lc( $str );
2819 // regexp to find first letter in each word (i.e. after each space)
2820 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2822 // function to use to capitalize a single char
2823 return preg_replace_callback(
2825 [ $this, 'ucwordsCallbackMB' ],
2829 return ucwords( strtolower( $str ) );
2834 * capitalize words at word breaks
2836 * @param string $str
2839 function ucwordbreaks( $str ) {
2840 if ( $this->isMultibyte( $str ) ) {
2841 $str = $this->lc( $str );
2843 // since \b doesn't work for UTF-8, we explicitely define word break chars
2844 $breaks = "[ \-\(\)\}\{\.,\?!]";
2846 // find first letter after word break
2847 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2848 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2850 return preg_replace_callback(
2852 [ $this, 'ucwordbreaksCallbackMB' ],
2856 return preg_replace_callback(
2857 '/\b([\w\x80-\xff]+)\b/',
2858 [ $this, 'ucwordbreaksCallbackAscii' ],
2865 * Return a case-folded representation of $s
2867 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2868 * and $s2 are the same except for the case of their characters. It is not
2869 * necessary for the value returned to make sense when displayed.
2871 * Do *not* perform any other normalisation in this function. If a caller
2872 * uses this function when it should be using a more general normalisation
2873 * function, then fix the caller.
2879 function caseFold( $s ) {
2880 return $this->uc( $s );
2884 * TODO: $s is not always a string per T218883
2888 function checkTitleEncoding( $s ) {
2889 if ( is_array( $s ) ) {
2890 throw new MWException( 'Given array to checkTitleEncoding.' );
2892 if ( StringUtils
::isUtf8( $s ) ) {
2896 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2902 function fallback8bitEncoding() {
2903 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2907 * Most writing systems use whitespace to break up words.
2908 * Some languages such as Chinese don't conventionally do this,
2909 * which requires special handling when breaking up words for
2914 function hasWordBreaks() {
2919 * Some languages such as Chinese require word segmentation,
2920 * Specify such segmentation when overridden in derived class.
2922 * @param string $string
2925 function segmentByWord( $string ) {
2930 * Some languages have special punctuation need to be normalized.
2931 * Make such changes here.
2933 * @param string $string
2936 function normalizeForSearch( $string ) {
2937 return self
::convertDoubleWidth( $string );
2941 * convert double-width roman characters to single-width.
2942 * range: ff00-ff5f ~= 0020-007f
2944 * @param string $string
2948 protected static function convertDoubleWidth( $string ) {
2949 static $full = null;
2950 static $half = null;
2952 if ( $full === null ) {
2953 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2954 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2955 $full = str_split( $fullWidth, 3 );
2956 $half = str_split( $halfWidth );
2959 $string = str_replace( $full, $half, $string );
2964 * @param string $string
2965 * @param string $pattern
2968 protected static function insertSpace( $string, $pattern ) {
2969 $string = preg_replace( $pattern, " $1 ", $string );
2970 $string = preg_replace( '/ +/', ' ', $string );
2975 * @param string[] $termsArray
2978 function convertForSearchResult( $termsArray ) {
2979 # some languages, e.g. Chinese, need to do a conversion
2980 # in order for search results to be displayed correctly
2985 * Get the first character of a string.
2990 function firstChar( $s ) {
2993 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2994 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2999 if ( isset( $matches[1] ) ) {
3000 if ( strlen( $matches[1] ) != 3 ) {
3004 // Break down Hangul syllables to grab the first jamo
3005 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3006 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3008 } elseif ( $code < 0xb098 ) {
3010 } elseif ( $code < 0xb2e4 ) {
3012 } elseif ( $code < 0xb77c ) {
3014 } elseif ( $code < 0xb9c8 ) {
3016 } elseif ( $code < 0xbc14 ) {
3018 } elseif ( $code < 0xc0ac ) {
3020 } elseif ( $code < 0xc544 ) {
3022 } elseif ( $code < 0xc790 ) {
3024 } elseif ( $code < 0xcc28 ) {
3026 } elseif ( $code < 0xce74 ) {
3028 } elseif ( $code < 0xd0c0 ) {
3030 } elseif ( $code < 0xd30c ) {
3032 } elseif ( $code < 0xd558 ) {
3043 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3044 * also cleans up certain backwards-compatible sequences, converting them
3045 * to the modern Unicode equivalent.
3047 * This is language-specific for performance reasons only.
3053 public function normalize( $s ) {
3054 global $wgAllUnicodeFixes;
3055 $s = UtfNormal\Validator
::cleanUp( $s );
3056 if ( $wgAllUnicodeFixes ) {
3057 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
3058 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3065 * Transform a string using serialized data stored in the given file (which
3066 * must be in the serialized subdirectory of $IP). The file contains pairs
3067 * mapping source characters to destination characters.
3069 * The data is cached in process memory. This will go faster if you have the
3070 * FastStringSearch extension.
3072 * @param string $file
3073 * @param string $string
3075 * @throws MWException
3078 protected function transformUsingPairFile( $file, $string ) {
3079 if ( !isset( $this->transformData
[$file] ) ) {
3081 $data = require "$IP/languages/data/{$file}";
3082 $this->transformData
[$file] = new ReplacementArray( $data );
3084 return $this->transformData
[$file]->replace( $string );
3088 * For right-to-left language support
3093 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3097 * Return the correct HTML 'dir' attribute value for this language.
3101 return $this->isRTL() ?
'rtl' : 'ltr';
3105 * Return 'left' or 'right' as appropriate alignment for line-start
3106 * for this language's text direction.
3108 * Should be equivalent to CSS3 'start' text-align value....
3112 function alignStart() {
3113 return $this->isRTL() ?
'right' : 'left';
3117 * Return 'right' or 'left' as appropriate alignment for line-end
3118 * for this language's text direction.
3120 * Should be equivalent to CSS3 'end' text-align value....
3124 function alignEnd() {
3125 return $this->isRTL() ?
'left' : 'right';
3129 * A hidden direction mark (LRM or RLM), depending on the language direction.
3130 * Unlike getDirMark(), this function returns the character as an HTML entity.
3131 * This function should be used when the output is guaranteed to be HTML,
3132 * because it makes the output HTML source code more readable. When
3133 * the output is plain text or can be escaped, getDirMark() should be used.
3135 * @param bool $opposite Get the direction mark opposite to your language
3139 function getDirMarkEntity( $opposite = false ) {
3141 return $this->isRTL() ?
'‎' : '‏';
3143 return $this->isRTL() ?
'‏' : '‎';
3147 * A hidden direction mark (LRM or RLM), depending on the language direction.
3148 * This function produces them as invisible Unicode characters and
3149 * the output may be hard to read and debug, so it should only be used
3150 * when the output is plain text or can be escaped. When the output is
3151 * HTML, use getDirMarkEntity() instead.
3153 * @param bool $opposite Get the direction mark opposite to your language
3156 function getDirMark( $opposite = false ) {
3157 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3158 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3160 return $this->isRTL() ?
$lrm : $rlm;
3162 return $this->isRTL() ?
$rlm : $lrm;
3168 function capitalizeAllNouns() {
3169 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3173 * An arrow, depending on the language direction.
3175 * @param string $direction The direction of the arrow: forwards (default),
3176 * backwards, left, right, up, down.
3179 function getArrow( $direction = 'forwards' ) {
3180 switch ( $direction ) {
3182 return $this->isRTL() ?
'←' : '→';
3184 return $this->isRTL() ?
'→' : '←';
3197 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3201 function linkPrefixExtension() {
3202 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3206 * Get all magic words from cache.
3209 function getMagicWords() {
3210 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3214 * Fill a MagicWord object with data from here
3216 * @param MagicWord $mw
3218 function getMagic( $mw ) {
3219 $rawEntry = $this->mMagicExtensions
[$mw->mId
] ??
3220 self
::$dataCache->getSubitem( $this->mCode
, 'magicWords', $mw->mId
);
3222 if ( !is_array( $rawEntry ) ) {
3223 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3225 $mw->mCaseSensitive
= $rawEntry[0];
3226 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3231 * Add magic words to the extension array
3233 * @param array $newWords
3235 function addMagicWordsByLang( $newWords ) {
3236 $fallbackChain = $this->getFallbackLanguages();
3237 $fallbackChain = array_reverse( $fallbackChain );
3238 foreach ( $fallbackChain as $code ) {
3239 if ( isset( $newWords[$code] ) ) {
3240 // @phan-suppress-next-line PhanTypeMismatchProperty
3241 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3247 * Get special page names, as an associative array
3248 * canonical name => array of valid names, including aliases
3251 function getSpecialPageAliases() {
3252 // Cache aliases because it may be slow to load them
3253 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3255 $this->mExtendedSpecialPageAliases
=
3256 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3259 return $this->mExtendedSpecialPageAliases
;
3263 * Italic is unsuitable for some languages
3265 * @param string $text The text to be emphasized.
3268 function emphasize( $text ) {
3269 return "<em>$text</em>";
3273 * Normally we output all numbers in plain en_US style, that is
3274 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3275 * point twohundredthirtyfive. However this is not suitable for all
3276 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3277 * Icelandic just want to use commas instead of dots, and dots instead
3278 * of commas like "293.291,235".
3280 * An example of this function being called:
3282 * wfMessage( 'message' )->numParams( $num )->text()
3285 * See $separatorTransformTable on MessageIs.php for
3286 * the , => . and . => , implementation.
3288 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3289 * @param int|float $number The string to be formatted, should be an integer
3290 * or a floating point number.
3291 * @param bool $nocommafy Set to true for special numbers like dates
3294 public function formatNum( $number, $nocommafy = false ) {
3295 global $wgTranslateNumerals;
3296 if ( !$nocommafy ) {
3297 $number = $this->commafy( $number );
3298 $s = $this->separatorTransformTable();
3300 $number = strtr( $number, $s );
3304 if ( $wgTranslateNumerals ) {
3305 $s = $this->digitTransformTable();
3307 $number = strtr( $number, $s );
3311 return (string)$number;
3315 * Front-end for non-commafied formatNum
3317 * @param int|float $number The string to be formatted, should be an integer
3318 * or a floating point number.
3322 public function formatNumNoSeparators( $number ) {
3323 return $this->formatNum( $number, true );
3327 * @param string $number
3330 public function parseFormattedNumber( $number ) {
3331 $s = $this->digitTransformTable();
3333 // eliminate empty array values such as ''. (T66347)
3334 $s = array_filter( $s );
3335 $number = strtr( $number, array_flip( $s ) );
3338 $s = $this->separatorTransformTable();
3340 // eliminate empty array values such as ''. (T66347)
3341 $s = array_filter( $s );
3342 $number = strtr( $number, array_flip( $s ) );
3345 $number = strtr( $number, [ ',' => '' ] );
3350 * Adds commas to a given number
3352 * @param mixed $number
3355 function commafy( $number ) {
3356 $digitGroupingPattern = $this->digitGroupingPattern();
3357 $minimumGroupingDigits = $this->minimumGroupingDigits();
3358 if ( $number === null ) {
3362 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3363 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3364 // In some languages it's conventional not to insert a thousands separator
3365 // in numbers that are four digits long (1000-9999).
3366 if ( $minimumGroupingDigits ) {
3367 // Number of '#' characters after last comma in the grouping pattern.
3368 // The pattern is hardcoded here, but this would vary for different patterns.
3369 $primaryGroupingSize = 3;
3370 // Maximum length of a number to suppress digit grouping for.
3371 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3372 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3376 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3378 // Ref: http://cldr.unicode.org/translation/number-patterns
3380 if ( intval( $number ) < 0 ) {
3381 // For negative numbers apply the algorithm like positive number and add sign.
3383 $number = substr( $number, 1 );
3387 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3388 preg_match( "/\d+/", $number, $integerPart );
3389 preg_match( "/\.\d*/", $number, $decimalPart );
3390 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3391 if ( $groupedNumber === $number ) {
3392 // the string does not have any number part. Eg: .12345
3393 return $sign . $groupedNumber;
3395 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3396 while ( $start > 0 ) {
3397 $match = $matches[0][$numMatches - 1];
3398 $matchLen = strlen( $match );
3399 $start = $end - $matchLen;
3403 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3405 if ( $numMatches > 1 ) {
3406 // use the last pattern for the rest of the number
3410 $groupedNumber = "," . $groupedNumber;
3413 return $sign . $groupedNumber;
3420 function digitGroupingPattern() {
3421 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3427 function digitTransformTable() {
3428 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3434 function separatorTransformTable() {
3435 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3441 function minimumGroupingDigits() {
3442 return self
::$dataCache->getItem( $this->mCode
, 'minimumGroupingDigits' );
3446 * Take a list of strings and build a locale-friendly comma-separated
3447 * list, using the local comma-separator message.
3448 * The last two strings are chained with an "and".
3450 * @param string[] $list
3453 public function listToText( array $list ) {
3454 $itemCount = count( $list );
3455 if ( $itemCount < 1 ) {
3458 $text = array_pop( $list );
3459 if ( $itemCount > 1 ) {
3460 $and = $this->msg( 'and' )->escaped();
3461 $space = $this->msg( 'word-separator' )->escaped();
3463 if ( $itemCount > 2 ) {
3464 $comma = $this->msg( 'comma-separator' )->escaped();
3466 $text = implode( $comma, $list ) . $and . $space . $text;
3472 * Take a list of strings and build a locale-friendly comma-separated
3473 * list, using the local comma-separator message.
3474 * @param string[] $list Array of strings to put in a comma list
3477 function commaList( array $list ) {
3479 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3485 * Take a list of strings and build a locale-friendly semicolon-separated
3486 * list, using the local semicolon-separator message.
3487 * @param string[] $list Array of strings to put in a semicolon list
3490 function semicolonList( array $list ) {
3492 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3498 * Same as commaList, but separate it with the pipe instead.
3499 * @param string[] $list Array of strings to put in a pipe list
3502 function pipeList( array $list ) {
3504 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3510 * Truncate a string to a specified length in bytes, appending an optional
3511 * string (e.g. for ellipsis)
3513 * If $length is negative, the string will be truncated from the beginning
3517 * @param string $string String to truncate
3518 * @param int $length Maximum length in bytes
3519 * @param string $ellipsis String to append to the end of truncated text
3520 * @param bool $adjustLength Subtract length of ellipsis from $length
3524 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3525 return $this->truncateInternal(
3526 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3531 * Truncate a string to a specified number of characters, appending an optional
3532 * string (e.g. for ellipsis).
3534 * This provides multibyte version of truncateForDatabase() method of this class,
3535 * suitable for truncation based on number of characters, instead of number of bytes.
3537 * If $length is negative, the string will be truncated from the beginning.
3541 * @param string $string String to truncate
3542 * @param int $length Maximum number of characters
3543 * @param string $ellipsis String to append to the end of truncated text
3544 * @param bool $adjustLength Subtract length of ellipsis from $length
3548 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3549 // Passing encoding to mb_strlen and mb_substr is optional.
3550 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3551 // explicit specification of encoding is skipped.
3552 // Note: Both multibyte methods are callables invoked in truncateInternal.
3553 return $this->truncateInternal(
3554 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3559 * Internal method used for truncation. This method abstracts text truncation into
3560 * one common method, allowing users to provide length measurement function and
3561 * function for finding substring.
3563 * For usages, see truncateForDatabase and truncateForVisual.
3565 * @param string $string String to truncate
3566 * @param int $length Maximum length of final text
3567 * @param string $ellipsis String to append to the end of truncated text
3568 * @param bool $adjustLength Subtract length of ellipsis from $length
3569 * @param callable $measureLength Callable function used for determining the length of text
3570 * @param callable $getSubstring Callable function used for getting the substrings
3574 private function truncateInternal(
3575 $string, $length, $ellipsis, $adjustLength, callable
$measureLength, callable
$getSubstring
3577 # Check if there is no need to truncate
3578 if ( $measureLength( $string ) <= abs( $length ) ) {
3579 return $string; // no need to truncate
3582 # Use the localized ellipsis character
3583 if ( $ellipsis == '...' ) {
3584 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3586 if ( $length == 0 ) {
3587 return $ellipsis; // convention
3590 $stringOriginal = $string;
3591 # If ellipsis length is >= $length then we can't apply $adjustLength
3592 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3593 $string = $ellipsis; // this can be slightly unexpected
3594 # Otherwise, truncate and add ellipsis...
3596 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3597 if ( $length > 0 ) {
3598 $length -= $ellipsisLength;
3599 $string = $getSubstring( $string, 0, $length ); // xyz...
3600 $string = $this->removeBadCharLast( $string );
3601 $string = rtrim( $string ) . $ellipsis;
3603 $length +
= $ellipsisLength;
3604 $string = $getSubstring( $string, $length ); // ...xyz
3605 $string = $this->removeBadCharFirst( $string );
3606 $string = $ellipsis . ltrim( $string );
3610 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3611 # This check is *not* redundant if $adjustLength, due to the single case where
3612 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3613 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3616 return $stringOriginal;
3621 * Remove bytes that represent an incomplete Unicode character
3622 * at the end of string (e.g. bytes of the char are missing)
3624 * @param string $string
3627 protected function removeBadCharLast( $string ) {
3628 if ( $string != '' ) {
3629 $char = ord( $string[strlen( $string ) - 1] );
3631 if ( $char >= 0xc0 ) {
3632 # We got the first byte only of a multibyte char; remove it.
3633 $string = substr( $string, 0, -1 );
3634 } elseif ( $char >= 0x80 &&
3635 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3636 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3637 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3639 # We chopped in the middle of a character; remove it
3647 * Remove bytes that represent an incomplete Unicode character
3648 * at the start of string (e.g. bytes of the char are missing)
3650 * @param string $string
3653 protected function removeBadCharFirst( $string ) {
3654 if ( $string != '' ) {
3655 $char = ord( $string[0] );
3656 if ( $char >= 0x80 && $char < 0xc0 ) {
3657 # We chopped in the middle of a character; remove the whole thing
3658 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3665 * Truncate a string of valid HTML to a specified length in bytes,
3666 * appending an optional string (e.g. for ellipses), and return valid HTML
3668 * This is only intended for styled/linked text, such as HTML with
3669 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3670 * Also, this will not detect things like "display:none" CSS.
3672 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3674 * @param string $text HTML string to truncate
3675 * @param int $length (zero/positive) Maximum length (including ellipses)
3676 * @param string $ellipsis String to append to the truncated text
3679 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3680 # Use the localized ellipsis character
3681 if ( $ellipsis == '...' ) {
3682 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3684 # Check if there is clearly no need to truncate
3685 if ( $length <= 0 ) {
3686 return $ellipsis; // no text shown, nothing to format (convention)
3687 } elseif ( strlen( $text ) <= $length ) {
3688 return $text; // string short enough even *with* HTML (short-circuit)
3691 $dispLen = 0; // innerHTML legth so far
3692 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3693 $tagType = 0; // 0-open, 1-close
3694 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3695 $entityState = 0; // 0-not entity, 1-entity
3696 $tag = $ret = ''; // accumulated tag name, accumulated result string
3697 $openTags = []; // open tag stack
3698 $maybeState = null; // possible truncation state
3700 $textLen = strlen( $text );
3701 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3702 for ( $pos = 0; true; ++
$pos ) {
3703 # Consider truncation once the display length has reached the maximim.
3704 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3705 # Check that we're not in the middle of a bracket/entity...
3706 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3707 if ( !$testingEllipsis ) {
3708 $testingEllipsis = true;
3709 # Save where we are; we will truncate here unless there turn out to
3710 # be so few remaining characters that truncation is not necessary.
3711 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3712 $maybeState = [ $ret, $openTags ]; // save state
3714 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3715 # String in fact does need truncation, the truncation point was OK.
3716 // @phan-suppress-next-line PhanTypeInvalidExpressionArrayDestructuring
3717 list( $ret, $openTags ) = $maybeState; // reload state
3718 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3719 $ret .= $ellipsis; // add ellipsis
3723 if ( $pos >= $textLen ) {
3724 break; // extra iteration just for above checks
3727 # Read the next char...
3729 $lastCh = $pos ?
$text[$pos - 1] : '';
3730 $ret .= $ch; // add to result string
3732 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3733 $entityState = 0; // for bad HTML
3734 $bracketState = 1; // tag started (checking for backslash)
3735 } elseif ( $ch == '>' ) {
3736 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3737 $entityState = 0; // for bad HTML
3738 $bracketState = 0; // out of brackets
3739 } elseif ( $bracketState == 1 ) {
3741 $tagType = 1; // close tag (e.g. "</span>")
3743 $tagType = 0; // open tag (e.g. "<span>")
3746 $bracketState = 2; // building tag name
3747 } elseif ( $bracketState == 2 ) {
3751 // Name found (e.g. "<a href=..."), add on tag attributes...
3752 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3754 } elseif ( $bracketState == 0 ) {
3755 if ( $entityState ) {
3758 $dispLen++
; // entity is one displayed char
3761 if ( $neLength == 0 && !$maybeState ) {
3762 // Save state without $ch. We want to *hit* the first
3763 // display char (to get tags) but not *use* it if truncating.
3764 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3767 $entityState = 1; // entity found, (e.g. " ")
3769 $dispLen++
; // this char is displayed
3770 // Add the next $max display text chars after this in one swoop...
3771 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3772 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3773 $dispLen +
= $skipped;
3779 // Close the last tag if left unclosed by bad HTML
3780 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3781 while ( count( $openTags ) > 0 ) {
3782 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3788 * truncateHtml() helper function
3789 * like strcspn() but adds the skipped chars to $ret
3791 * @param string $ret
3792 * @param string $text
3793 * @param string $search
3795 * @param null|int $len
3798 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3799 if ( $len === null ) {
3800 $len = -1; // -1 means "no limit" for strcspn
3801 } elseif ( $len < 0 ) {
3805 if ( $start < strlen( $text ) ) {
3806 $skipCount = strcspn( $text, $search, $start, $len );
3807 $ret .= substr( $text, $start, $skipCount );
3813 * truncateHtml() helper function
3814 * (a) push or pop $tag from $openTags as needed
3815 * (b) clear $tag value
3816 * @param string &$tag Current HTML tag name we are looking at
3817 * @param int $tagType (0-open tag, 1-close tag)
3818 * @param string $lastCh Character before the '>' that ended this tag
3819 * @param array &$openTags Open tag stack (not accounting for $tag)
3821 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3822 $tag = ltrim( $tag );
3824 if ( $tagType == 0 && $lastCh != '/' ) {
3825 $openTags[] = $tag; // tag opened (didn't close itself)
3826 } elseif ( $tagType == 1 ) {
3827 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3828 array_pop( $openTags ); // tag closed
3836 * Grammatical transformations, needed for inflected languages
3837 * Invoked by putting {{grammar:case|word}} in a message
3839 * @param string $word
3840 * @param string $case
3843 function convertGrammar( $word, $case ) {
3844 global $wgGrammarForms;
3845 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3846 return $wgGrammarForms[$this->getCode()][$case][$word];
3849 $grammarTransformations = $this->getGrammarTransformations();
3851 if ( isset( $grammarTransformations[$case] ) ) {
3852 $forms = $grammarTransformations[$case];
3854 // Some names of grammar rules are aliases for other rules.
3855 // In such cases the value is a string rather than object,
3856 // so load the actual rules.
3857 if ( is_string( $forms ) ) {
3858 $forms = $grammarTransformations[$forms];
3861 foreach ( array_values( $forms ) as $rule ) {
3864 if ( $form === '@metadata' ) {
3868 $replacement = $rule[1];
3870 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3871 $patternMatches = preg_match( $regex, $word );
3873 if ( $patternMatches === false ) {
3875 'An error occurred while processing grammar. ' .
3876 "Word: '$word'. Regex: /$form/."
3878 } elseif ( $patternMatches === 1 ) {
3879 $word = preg_replace( $regex, $replacement, $word );
3890 * Get the grammar forms for the content language
3891 * @return array Array of grammar forms
3894 function getGrammarForms() {
3895 global $wgGrammarForms;
3896 if ( isset( $wgGrammarForms[$this->getCode()] )
3897 && is_array( $wgGrammarForms[$this->getCode()] )
3899 return $wgGrammarForms[$this->getCode()];
3906 * Get the grammar transformations data for the language.
3907 * Used like grammar forms, with {{GRAMMAR}} and cases,
3908 * but uses pairs of regexes and replacements instead of code.
3910 * @return array[] Array of grammar transformations.
3911 * @throws MWException
3914 public function getGrammarTransformations() {
3915 $languageCode = $this->getCode();
3917 if ( self
::$grammarTransformations === null ) {
3918 self
::$grammarTransformations = new MapCacheLRU( 10 );
3921 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3922 return self
::$grammarTransformations->get( $languageCode );
3927 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3928 if ( is_readable( $grammarDataFile ) ) {
3929 $data = FormatJson
::decode(
3930 file_get_contents( $grammarDataFile ),
3934 if ( $data === null ) {
3935 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3938 self
::$grammarTransformations->set( $languageCode, $data );
3945 * Provides an alternative text depending on specified gender.
3946 * Usage {{gender:username|masculine|feminine|unknown}}.
3947 * username is optional, in which case the gender of current user is used,
3948 * but only in (some) interface messages; otherwise default gender is used.
3950 * If no forms are given, an empty string is returned. If only one form is
3951 * given, it will be returned unconditionally. These details are implied by
3952 * the caller and cannot be overridden in subclasses.
3954 * If three forms are given, the default is to use the third (unknown) form.
3955 * If fewer than three forms are given, the default is to use the first (masculine) form.
3956 * These details can be overridden in subclasses.
3958 * @param string $gender
3959 * @param array $forms
3963 function gender( $gender, $forms ) {
3964 if ( !count( $forms ) ) {
3967 $forms = $this->preConvertPlural( $forms, 2 );
3968 if ( $gender === 'male' ) {
3971 if ( $gender === 'female' ) {
3974 return $forms[2] ??
$forms[0];
3978 * Plural form transformations, needed for some languages.
3979 * For example, there are 3 form of plural in Russian and Polish,
3980 * depending on "count mod 10". See [[w:Plural]]
3981 * For English it is pretty simple.
3983 * Invoked by putting {{plural:count|wordform1|wordform2}}
3984 * or {{plural:count|wordform1|wordform2|wordform3}}
3986 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3988 * @param int $count Non-localized number
3989 * @param array $forms Different plural forms
3990 * @return string Correct form of plural for $count in this language
3992 function convertPlural( $count, $forms ) {
3993 // Handle explicit n=pluralform cases
3994 $forms = $this->handleExplicitPluralForms( $count, $forms );
3995 if ( is_string( $forms ) ) {
3998 if ( !count( $forms ) ) {
4002 $pluralForm = $this->getPluralRuleIndexNumber( $count );
4003 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
4004 return $forms[$pluralForm];
4008 * Handles explicit plural forms for Language::convertPlural()
4010 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4011 * If an explicitly defined plural form matches the $count, then
4012 * string value returned, otherwise array returned for further consideration
4013 * by CLDR rules or overridden convertPlural().
4017 * @param int $count Non-localized number
4018 * @param array $forms Different plural forms
4020 * @return array|string
4022 protected function handleExplicitPluralForms( $count, array $forms ) {
4023 foreach ( $forms as $index => $form ) {
4024 if ( preg_match( '/\d+=/i', $form ) ) {
4025 $pos = strpos( $form, '=' );
4026 if ( substr( $form, 0, $pos ) === (string)$count ) {
4027 return substr( $form, $pos +
1 );
4029 unset( $forms[$index] );
4032 return array_values( $forms );
4036 * Checks that convertPlural was given an array and pads it to requested
4037 * amount of forms by copying the last one.
4039 * @param array $forms
4040 * @param int $count How many forms should there be at least
4041 * @return array Padded array of forms
4043 protected function preConvertPlural( /* Array */ $forms, $count ) {
4044 return array_pad( $forms, $count, end( $forms ) );
4048 * Wraps argument with unicode control characters for directionality safety
4050 * This solves the problem where directionality-neutral characters at the edge of
4051 * the argument string get interpreted with the wrong directionality from the
4052 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4054 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4055 * directionality of the argument string, using the BIDI algorithm's own "First
4056 * strong directional codepoint" rule. Essentially, this works round the fact that
4057 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4058 * direction inference). The latter is cleaner but still not widely supported.
4060 * @param string $text Text to wrap
4061 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4063 public function embedBidi( $text = '' ) {
4064 $dir = self
::strongDirFromContent( $text );
4065 if ( $dir === 'ltr' ) {
4066 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4067 return self
::$lre . $text . self
::$pdf;
4069 if ( $dir === 'rtl' ) {
4070 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4071 return self
::$rle . $text . self
::$pdf;
4073 // No strong directionality: do not wrap
4078 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4079 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4080 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4081 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4084 * @param string $str The validated block duration in English
4085 * @param User|null $user User object to use timezone from or null for $wgUser
4086 * @param int $now Current timestamp, for formatting relative block durations
4087 * @return string Somehow translated block duration
4088 * @see LanguageFi.php for example implementation
4090 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
4091 $duration = SpecialBlock
::getSuggestedDurations( $this );
4092 foreach ( $duration as $show => $value ) {
4093 if ( strcmp( $str, $value ) == 0 ) {
4094 return htmlspecialchars( trim( $show ) );
4098 if ( wfIsInfinity( $str ) ) {
4099 foreach ( $duration as $show => $value ) {
4100 if ( wfIsInfinity( $value ) ) {
4101 return htmlspecialchars( trim( $show ) );
4106 // If all else fails, return a standard duration or timestamp description.
4107 $time = strtotime( $str, $now );
4108 if ( $time === false ) { // Unknown format. Return it as-is in case.
4110 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
4111 // The result differs based on current time, so the difference
4112 // is a fixed duration length.
4113 return $this->formatDuration( $time - $now );
4114 } else { // It's an absolute timestamp.
4115 if ( $time === 0 ) {
4116 // wfTimestamp() handles 0 as current time instead of epoch.
4117 $time = '19700101000000';
4120 return $this->userTimeAndDate( $time, $user );
4122 return $this->timeanddate( $time );
4127 * languages like Chinese need to be segmented in order for the diff
4130 * @param string $text
4133 public function segmentForDiff( $text ) {
4138 * and unsegment to show the result
4140 * @param string $text
4143 public function unsegmentForDiff( $text ) {
4148 * Return the LanguageConverter used in the Language
4151 * @return LanguageConverter
4153 public function getConverter() {
4154 return $this->mConverter
;
4158 * convert text to a variant
4160 * @param string $text text to convert
4161 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4162 * variant (if logged in), or the project default variant
4163 * @return string the converted string
4165 public function autoConvert( $text, $variant = false ) {
4166 return $this->mConverter
->autoConvert( $text, $variant );
4170 * convert text to all supported variants
4172 * @param string $text
4175 public function autoConvertToAllVariants( $text ) {
4176 return $this->mConverter
->autoConvertToAllVariants( $text );
4180 * convert text to different variants of a language.
4182 * @warning Glossary state is maintained between calls. This means
4183 * if you pass unescaped text to this method it can cause an XSS
4184 * in later calls to this method, even if the later calls have properly
4185 * escaped the input. Never feed this method user controlled text that
4186 * is not properly escaped!
4187 * @param string $text Content that has been already escaped for use in HTML
4188 * @return string HTML
4190 public function convert( $text ) {
4191 return $this->mConverter
->convert( $text );
4195 * Convert a Title object to a string in the preferred variant
4197 * @param Title $title
4200 public function convertTitle( $title ) {
4201 return $this->mConverter
->convertTitle( $title );
4205 * Convert a namespace index to a string in the preferred variant
4207 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4208 * @param string|null $variant variant to convert to, or null to use the user's preferred
4209 * variant (if logged in), or the project default variant
4210 * @return string a string representation of the namespace
4212 public function convertNamespace( $ns, $variant = null ) {
4213 return $this->mConverter
->convertNamespace( $ns, $variant );
4217 * Check if this is a language with variants
4221 public function hasVariants() {
4222 return count( $this->getVariants() ) > 1;
4226 * Strict check if the language has the specific variant.
4228 * Compare to LanguageConverter::validateVariant() which does a more
4229 * lenient check and attempts to coerce the given code to a valid one.
4232 * @param string $variant
4235 public function hasVariant( $variant ) {
4236 return $variant && ( $variant === $this->mConverter
->validateVariant( $variant ) );
4240 * Perform output conversion on a string, and encode for safe HTML output.
4241 * @param string $text Text to be converted
4243 * @todo this should get integrated somewhere sane
4245 public function convertHtml( $text ) {
4246 return htmlspecialchars( $this->convert( $text ) );
4250 * @param string $key
4253 public function convertCategoryKey( $key ) {
4254 return $this->mConverter
->convertCategoryKey( $key );
4258 * Get the list of variants supported by this language
4259 * see sample implementation in LanguageZh.php
4261 * @return string[] An array of language codes
4263 public function getVariants() {
4264 return $this->mConverter
->getVariants();
4270 public function getPreferredVariant() {
4271 return $this->mConverter
->getPreferredVariant();
4277 public function getDefaultVariant() {
4278 return $this->mConverter
->getDefaultVariant();
4284 public function getURLVariant() {
4285 return $this->mConverter
->getURLVariant();
4289 * If a language supports multiple variants, it is
4290 * possible that non-existing link in one variant
4291 * actually exists in another variant. this function
4292 * tries to find it. See e.g. LanguageZh.php
4293 * The input parameters may be modified upon return
4295 * @param string &$link The name of the link
4296 * @param Title &$nt The title object of the link
4297 * @param bool $ignoreOtherCond To disable other conditions when
4298 * we need to transclude a template or update a category's link
4300 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4301 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4305 * returns language specific options used by User::getPageRenderHash()
4306 * for example, the preferred language variant
4310 function getExtraHashOptions() {
4311 return $this->mConverter
->getExtraHashOptions();
4315 * For languages that support multiple variants, the title of an
4316 * article may be displayed differently in different variants. this
4317 * function returns the apporiate title defined in the body of the article.
4321 public function getParsedTitle() {
4322 return $this->mConverter
->getParsedTitle();
4326 * Refresh the cache of conversion tables when
4327 * MediaWiki:Conversiontable* is updated.
4329 * @param Title $title The Title of the page being updated
4331 public function updateConversionTable( Title
$title ) {
4332 $this->mConverter
->updateConversionTable( $title );
4336 * A regular expression to match legal word-trailing characters
4337 * which should be merged onto a link of the form [[foo]]bar.
4341 public function linkTrail() {
4342 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4346 * A regular expression character set to match legal word-prefixing
4347 * characters which should be merged onto a link of the form foo[[bar]].
4351 public function linkPrefixCharset() {
4352 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4356 * Get the "parent" language which has a converter to convert a "compatible" language
4357 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4359 * @return Language|null
4362 public function getParentLanguage() {
4363 if ( $this->mParentLanguage
!== false ) {
4364 return $this->mParentLanguage
;
4367 $code = explode( '-', $this->getCode() )[0];
4368 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4369 $this->mParentLanguage
= null;
4372 $lang = self
::factory( $code );
4373 if ( !$lang->hasVariant( $this->getCode() ) ) {
4374 $this->mParentLanguage
= null;
4378 $this->mParentLanguage
= $lang;
4383 * Compare with an other language object
4386 * @param Language $lang
4389 public function equals( Language
$lang ) {
4390 return $lang === $this ||
$lang->getCode() === $this->mCode
;
4394 * Get the internal language code for this language object
4396 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4397 * htmlspecialchars() or similar
4401 public function getCode() {
4402 return $this->mCode
;
4406 * Get the code in BCP 47 format which we can use
4407 * inside of html lang="" tags.
4409 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4410 * htmlspecialchars() or similar.
4415 public function getHtmlCode() {
4416 if ( is_null( $this->mHtmlCode
) ) {
4417 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4419 return $this->mHtmlCode
;
4423 * Get the language code from a file name. Inverse of getFileName()
4424 * @param string $filename $prefix . $languageCode . $suffix
4425 * @param string $prefix Prefix before the language code
4426 * @param string $suffix Suffix after the language code
4427 * @return string Language code, or false if $prefix or $suffix isn't found
4429 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4431 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4432 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4433 if ( !count( $m ) ) {
4436 return str_replace( '_', '-', strtolower( $m[1] ) );
4440 * @param string $code
4441 * @param bool $fallback Whether we're going through language fallback chain
4442 * @return string Name of the language class
4444 public static function classFromCode( $code, $fallback = true ) {
4445 if ( $fallback && $code == 'en' ) {
4448 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4453 * Get the name of a file for a certain language code
4454 * @param string $prefix Prepend this to the filename
4455 * @param string $code Language code
4456 * @param string $suffix Append this to the filename
4457 * @throws MWException
4458 * @return string $prefix . $mangledCode . $suffix
4460 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4461 if ( !self
::isValidBuiltInCode( $code ) ) {
4462 throw new MWException( "Invalid language code \"$code\"" );
4465 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4469 * @param string $code
4472 public static function getMessagesFileName( $code ) {
4474 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4475 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4480 * @param string $code
4482 * @throws MWException
4485 public static function getJsonMessagesFileName( $code ) {
4488 if ( !self
::isValidBuiltInCode( $code ) ) {
4489 throw new MWException( "Invalid language code \"$code\"" );
4492 return "$IP/languages/i18n/$code.json";
4496 * Get the first fallback for a given language.
4498 * @param string $code
4500 * @return bool|string
4502 public static function getFallbackFor( $code ) {
4503 $fallbacks = self
::getFallbacksFor( $code );
4505 return $fallbacks[0];
4511 * Get the ordered list of fallback languages.
4514 * @param string $code Language code
4515 * @param int $mode Fallback mode, either MESSAGES_FALLBACKS (which always falls back to 'en'),
4516 * or STRICT_FALLBACKS (whic honly falls back to 'en' when explicitly defined)
4517 * @throws MWException
4518 * @return array List of language codes
4520 public static function getFallbacksFor( $code, $mode = self
::MESSAGES_FALLBACKS
) {
4521 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4525 case self
::MESSAGES_FALLBACKS
:
4526 // For unknown languages, fallbackSequence returns an empty array,
4527 // hardcode fallback to 'en' in that case as English messages are
4529 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4530 case self
::STRICT_FALLBACKS
:
4531 // Use this mode when you don't want to fallback to English unless
4532 // explicitly defined, for example when you have language-variant icons
4533 // and an international language-independent fallback.
4534 return self
::getLocalisationCache()->getItem( $code, 'originalFallbackSequence' );
4536 throw new MWException( "Invalid fallback mode \"$mode\"" );
4541 * Get the ordered list of fallback languages, ending with the fallback
4542 * language chain for the site language.
4545 * @param string $code Language code
4546 * @return array [ fallbacks, site fallbacks ]
4548 public static function getFallbacksIncludingSiteLanguage( $code ) {
4549 global $wgLanguageCode;
4551 // Usually, we will only store a tiny number of fallback chains, so we
4552 // keep them in static memory.
4553 $cacheKey = "{$code}-{$wgLanguageCode}";
4555 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4556 $fallbacks = self
::getFallbacksFor( $code );
4558 // Append the site's fallback chain, including the site language itself
4559 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4560 array_unshift( $siteFallbacks, $wgLanguageCode );
4562 // Eliminate any languages already included in the chain
4563 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4565 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4567 return self
::$fallbackLanguageCache[$cacheKey];
4571 * Get all messages for a given language
4572 * WARNING: this may take a long time. If you just need all message *keys*
4573 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4575 * @param string $code
4579 public static function getMessagesFor( $code ) {
4580 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4584 * Get a message for a given language
4586 * @param string $key
4587 * @param string $code
4591 public static function getMessageFor( $key, $code ) {
4592 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4596 * Get all message keys for a given language. This is a faster alternative to
4597 * array_keys( Language::getMessagesFor( $code ) )
4600 * @param string $code Language code
4601 * @return array Array of message keys (strings)
4603 public static function getMessageKeysFor( $code ) {
4604 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4608 * @param string $talk
4611 function fixVariableInNamespace( $talk ) {
4612 if ( strpos( $talk, '$1' ) === false ) {
4616 global $wgMetaNamespace;
4617 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4619 # Allow grammar transformations
4620 # Allowing full message-style parsing would make simple requests
4621 # such as action=raw much more expensive than they need to be.
4622 # This will hopefully cover most cases.
4623 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4624 [ $this, 'replaceGrammarInNamespace' ], $talk );
4625 return str_replace( ' ', '_', $talk );
4632 function replaceGrammarInNamespace( $m ) {
4633 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4637 * Decode an expiry (block, protection, etc) which has come from the DB
4639 * @param string $expiry Database expiry String
4640 * @param bool|int $format True to process using language functions, or TS_ constant
4641 * to return the expiry in a given timestamp
4642 * @param string $infinity If $format is not true, use this string for infinite expiry
4646 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4648 if ( $dbInfinity === null ) {
4649 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4652 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4653 return $format === true
4654 ?
$this->getMessageFromDB( 'infiniteblock' )
4657 return $format === true
4658 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4659 : wfTimestamp( $format, $expiry );
4664 * Formats a time given in seconds into a string representation of that time.
4666 * @param int|float $seconds
4667 * @param array $format An optional argument that formats the returned string in different ways:
4668 * If $format['avoid'] === 'avoidhours': don't show hours, just show days
4669 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4670 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4671 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4673 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4674 * or 'avoidminutes'.
4677 function formatTimePeriod( $seconds, $format = [] ) {
4678 if ( !is_array( $format ) ) {
4679 $format = [ 'avoid' => $format ]; // For backwards compatibility
4681 if ( !isset( $format['avoid'] ) ) {
4682 $format['avoid'] = false;
4684 if ( !isset( $format['noabbrevs'] ) ) {
4685 $format['noabbrevs'] = false;
4687 $secondsMsg = wfMessage(
4688 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4689 $minutesMsg = wfMessage(
4690 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4691 $hoursMsg = wfMessage(
4692 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4693 $daysMsg = wfMessage(
4694 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4696 if ( round( $seconds * 10 ) < 100 ) {
4697 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4698 $s = $secondsMsg->params( $s )->text();
4699 } elseif ( round( $seconds ) < 60 ) {
4700 $s = $this->formatNum( round( $seconds ) );
4701 $s = $secondsMsg->params( $s )->text();
4702 } elseif ( round( $seconds ) < 3600 ) {
4703 $minutes = floor( $seconds / 60 );
4704 $secondsPart = round( fmod( $seconds, 60 ) );
4705 if ( $secondsPart == 60 ) {
4709 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4711 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4712 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4713 $hours = floor( $seconds / 3600 );
4714 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4715 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4716 if ( $secondsPart == 60 ) {
4720 if ( $minutes == 60 ) {
4724 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4726 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4727 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes', 'avoidhours' ] ) ) {
4728 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4731 $days = floor( $seconds / 86400 );
4732 if ( $format['avoid'] === 'avoidhours' ) {
4733 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4734 if ( $hours == 24 ) {
4738 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4739 } elseif ( $format['avoid'] === 'avoidminutes' ) {
4740 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4741 if ( $hours == 24 ) {
4745 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4747 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4748 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4749 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4750 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4751 if ( $minutes == 60 ) {
4755 if ( $hours == 24 ) {
4759 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4761 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4763 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4765 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4767 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4774 * Format a bitrate for output, using an appropriate
4775 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4776 * the magnitude in question.
4778 * This use base 1000. For base 1024 use formatSize(), for another base
4779 * see formatComputingNumbers().
4784 function formatBitrate( $bps ) {
4785 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4789 * @param int $size Size of the unit
4790 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4791 * @param string $messageKey Message key to be uesd
4794 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4796 return str_replace( '$1', $this->formatNum( $size ),
4797 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4800 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4803 $maxIndex = count( $sizes ) - 1;
4804 while ( $size >= $boundary && $index < $maxIndex ) {
4809 // For small sizes no decimal places necessary
4812 // For MB and bigger two decimal places are smarter
4815 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4817 $size = round( $size, $round );
4818 $text = $this->getMessageFromDB( $msg );
4819 return str_replace( '$1', $this->formatNum( $size ), $text );
4823 * Format a size in bytes for output, using an appropriate
4824 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4826 * This method use base 1024. For base 1000 use formatBitrate(), for
4827 * another base see formatComputingNumbers()
4829 * @param int $size Size to format
4830 * @return string Plain text (not HTML)
4832 function formatSize( $size ) {
4833 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4837 * Make a list item, used by various special pages
4839 * @param string $page Page link
4840 * @param string $details HTML safe text between brackets
4841 * @param bool $oppositedm Add the direction mark opposite to your
4842 * language, to display text properly
4843 * @return string HTML escaped
4845 function specialList( $page, $details, $oppositedm = true ) {
4850 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4853 $this->msg( 'word-separator' )->escaped() .
4854 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4858 * Generate (prev x| next x) (20|50|100...) type links for paging
4860 * @param Title $title Title object to link
4861 * @param int $offset
4863 * @param array $query Optional URL query parameter string
4864 * @param bool $atend Optional param for specified if this is the last page
4866 * @deprecated since 1.34, use PrevNextNavigationRenderer::buildPrevNextNavigation()
4869 public function viewPrevNext( Title
$title, $offset, $limit,
4870 array $query = [], $atend = false
4872 wfDeprecated( __METHOD__
, '1.34' );
4873 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4875 # Make 'previous' link
4876 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4877 if ( $offset > 0 ) {
4878 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4879 $query, $prev, 'prevn-title', 'mw-prevlink' );
4881 $plink = htmlspecialchars( $prev );
4885 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4887 $nlink = htmlspecialchars( $next );
4889 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4890 $query, $next, 'nextn-title', 'mw-nextlink' );
4893 # Make links to set number of items per page
4895 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4896 $numLinks[] = $this->numLink( $title, $offset, $num,
4897 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4900 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4901 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4905 * Helper function for viewPrevNext() that generates links
4907 * @param Title $title Title object to link
4908 * @param int $offset
4910 * @param array $query Extra query parameters
4911 * @param string $link Text to use for the link; will be escaped
4912 * @param string $tooltipMsg Name of the message to use as tooltip
4913 * @param string $class Value of the "class" attribute of the link
4914 * @return string HTML fragment
4916 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4919 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4920 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4921 ->numParams( $limit )->text();
4923 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4924 'title' => $tooltip, 'class' => $class ], $link );
4928 * Get the conversion rule title, if any.
4932 public function getConvRuleTitle() {
4933 return $this->mConverter
->getConvRuleTitle();
4937 * Get the compiled plural rules for the language
4939 * @return array Associative array with plural form, and plural rule as key-value pairs
4941 public function getCompiledPluralRules() {
4942 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4943 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4944 if ( !$pluralRules ) {
4945 foreach ( $fallbacks as $fallbackCode ) {
4946 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4947 if ( $pluralRules ) {
4952 return $pluralRules;
4956 * Get the plural rules for the language
4958 * @return array Associative array with plural form number and plural rule as key-value pairs
4960 public function getPluralRules() {
4961 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4962 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4963 if ( !$pluralRules ) {
4964 foreach ( $fallbacks as $fallbackCode ) {
4965 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4966 if ( $pluralRules ) {
4971 return $pluralRules;
4975 * Get the plural rule types for the language
4977 * @return array Associative array with plural form number and plural rule type as key-value pairs
4979 public function getPluralRuleTypes() {
4980 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4981 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4982 if ( !$pluralRuleTypes ) {
4983 foreach ( $fallbacks as $fallbackCode ) {
4984 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4985 if ( $pluralRuleTypes ) {
4990 return $pluralRuleTypes;
4994 * Find the index number of the plural rule appropriate for the given number
4995 * @param int $number
4996 * @return int The index number of the plural rule
4998 public function getPluralRuleIndexNumber( $number ) {
4999 $pluralRules = $this->getCompiledPluralRules();
5000 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
5005 * Find the plural rule type appropriate for the given number
5006 * For example, if the language is set to Arabic, getPluralType(5) should
5009 * @param int $number
5010 * @return string The name of the plural rule type, e.g. one, two, few, many
5012 public function getPluralRuleType( $number ) {
5013 $index = $this->getPluralRuleIndexNumber( $number );
5014 $pluralRuleTypes = $this->getPluralRuleTypes();
5015 return $pluralRuleTypes[$index] ??
'other';