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
;
32 * Internationalisation code
37 * @var LanguageConverter
41 public $mVariants, $mCode, $mLoaded = false;
42 public $mMagicExtensions = [], $mMagicHookDone = false;
43 private $mHtmlCode = null, $mParentLanguage = false;
45 public $dateFormatStrings = [];
46 public $mExtendedSpecialPageAliases;
48 /** @var array|null */
49 protected $namespaceNames;
50 protected $mNamespaceIds, $namespaceAliases;
53 * ReplacementArray object caches
55 public $transformData = [];
58 * @var LocalisationCache
60 static public $dataCache;
62 static public $mLangObjCache = [];
64 static public $mWeekdayMsgs = [
65 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
69 static public $mWeekdayAbbrevMsgs = [
70 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
73 static public $mMonthMsgs = [
74 'january', 'february', 'march', 'april', 'may_long', 'june',
75 'july', 'august', 'september', 'october', 'november',
78 static public $mMonthGenMsgs = [
79 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
80 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
83 static public $mMonthAbbrevMsgs = [
84 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
85 'sep', 'oct', 'nov', 'dec'
88 static public $mIranianCalendarMonthMsgs = [
89 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
90 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
91 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
92 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
95 static public $mHebrewCalendarMonthMsgs = [
96 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
97 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
98 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
99 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
100 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
103 static public $mHebrewCalendarMonthGenMsgs = [
104 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
105 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
106 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
107 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
108 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
111 static public $mHijriCalendarMonthMsgs = [
112 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
113 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
114 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
115 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
122 static public $durationIntervals = [
123 'millennia' => 31556952000,
124 'centuries' => 3155695200,
125 'decades' => 315569520,
126 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
135 * Cache for language fallbacks.
136 * @see Language::getFallbacksIncludingSiteLanguage
140 static private $fallbackLanguageCache = [];
143 * Cache for grammar rules data
144 * @var MapCacheLRU|null
146 static private $grammarTransformations;
149 * Cache for language names
150 * @var HashBagOStuff|null
152 static private $languageNameCache;
155 * Unicode directional formatting characters, for embedBidi()
157 static private $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
158 static private $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
159 static private $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
162 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
163 * - in group 1 if it is LTR
164 * - in group 2 if it is RTL
165 * Does not match if there is no strong directionality codepoint.
167 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
169 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
170 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
172 // @codeCoverageIgnoreStart
173 // phpcs:ignore Generic.Files.LineLength
174 static private $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';
175 // @codeCoverageIgnoreEnd
178 * Get a cached or new language object for a given language code
179 * @param string $code
180 * @throws MWException
183 static function factory( $code ) {
184 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
186 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
187 $code = $wgDummyLanguageCodes[$code];
190 // get the language object to process
191 $langObj = self
::$mLangObjCache[$code] ?? self
::newFromCode( $code );
193 // merge the language object in to get it up front in the cache
194 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
195 // get rid of the oldest ones in case we have an overflow
196 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
202 * Create a language object for a given language code
203 * @param string $code
204 * @param bool $fallback Whether we're going through language fallback chain
205 * @throws MWException
208 protected static function newFromCode( $code, $fallback = false ) {
209 if ( !self
::isValidCode( $code ) ) {
210 throw new MWException( "Invalid language code \"$code\"" );
213 if ( !self
::isValidBuiltInCode( $code ) ) {
214 // It's not possible to customise this code with class files, so
215 // just return a Language object. This is to support uselang= hacks.
216 $lang = new Language
;
217 $lang->setCode( $code );
221 // Check if there is a language class for the code
222 $class = self
::classFromCode( $code, $fallback );
223 // LanguageCode does not inherit Language
224 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
229 // Keep trying the fallback list until we find an existing class
230 $fallbacks = self
::getFallbacksFor( $code );
231 foreach ( $fallbacks as $fallbackCode ) {
232 if ( !self
::isValidBuiltInCode( $fallbackCode ) ) {
233 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
236 $class = self
::classFromCode( $fallbackCode );
237 if ( class_exists( $class ) ) {
239 $lang->setCode( $code );
244 throw new MWException( "Invalid fallback sequence for language '$code'" );
248 * Checks whether any localisation is available for that language tag
249 * in MediaWiki (MessagesXx.php exists).
251 * @param string $code Language tag (in lower case)
252 * @return bool Whether language is supported
255 public static function isSupportedLanguage( $code ) {
256 if ( !self
::isValidBuiltInCode( $code ) ) {
260 if ( $code === 'qqq' ) {
264 return is_readable( self
::getMessagesFileName( $code ) ) ||
265 is_readable( self
::getJsonMessagesFileName( $code ) );
269 * Returns true if a language code string is a well-formed language tag
270 * according to RFC 5646.
271 * This function only checks well-formedness; it doesn't check that
272 * language, script or variant codes actually exist in the repositories.
274 * Based on regexes by Mark Davis of the Unicode Consortium:
275 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
277 * @param string $code
278 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
283 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
286 $alphanum = '[a-z0-9]';
287 $x = 'x'; # private use singleton
288 $singleton = '[a-wy-z]'; # other singleton
289 $s = $lenient ?
'[-_]' : '-';
291 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
292 $script = "$alpha{4}"; # ISO 15924
293 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
294 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
295 $extension = "$singleton(?:$s$alphanum{2,8})+";
296 $privateUse = "$x(?:$s$alphanum{1,8})+";
298 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
299 # Since these are limited, this is safe even later changes to the registry --
300 # the only oddity is that it might change the type of the tag, and thus
301 # the results from the capturing groups.
302 # https://www.iana.org/assignments/language-subtag-registry
304 $grandfathered = "en{$s}GB{$s}oed"
305 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
306 . "|no{$s}(?:bok|nyn)"
307 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
308 . "|zh{$s}min{$s}nan";
310 $variantList = "$variant(?:$s$variant)*";
311 $extensionList = "$extension(?:$s$extension)*";
313 $langtag = "(?:($language)"
316 . "(?:$s$variantList)?"
317 . "(?:$s$extensionList)?"
318 . "(?:$s$privateUse)?)";
320 # The final breakdown, with capturing groups for each of these components
321 # The variants, extensions, grandfathered, and private-use may have interior '-'
323 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
325 return (bool)preg_match( "/$root/", strtolower( $code ) );
329 * Returns true if a language code string is of a valid form, whether or
330 * not it exists. This includes codes which are used solely for
331 * customisation via the MediaWiki namespace.
333 * @param string $code
337 public static function isValidCode( $code ) {
339 if ( !isset( $cache[$code] ) ) {
340 // People think language codes are html safe, so enforce it.
341 // Ideally we should only allow a-zA-Z0-9-
342 // but, .+ and other chars are often used for {{int:}} hacks
343 // see bugs T39564, T39587, T38938
345 // Protect against path traversal
346 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
347 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
349 return $cache[$code];
353 * Returns true if a language code is of a valid form for the purposes of
354 * internal customisation of MediaWiki, via Messages*.php or *.json.
356 * @param string $code
358 * @throws MWException
362 public static function isValidBuiltInCode( $code ) {
363 if ( !is_string( $code ) ) {
364 if ( is_object( $code ) ) {
365 $addmsg = " of class " . get_class( $code );
369 $type = gettype( $code );
370 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
373 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
377 * Returns true if a language code is an IETF tag known to MediaWiki.
384 public static function isKnownLanguageTag( $tag ) {
385 // Quick escape for invalid input to avoid exceptions down the line
386 // when code tries to process tags which are not valid at all.
387 if ( !self
::isValidBuiltInCode( $tag ) ) {
391 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
392 || self
::fetchLanguageName( $tag, $tag ) !== ''
401 * Get the LocalisationCache instance
403 * @return LocalisationCache
405 public static function getLocalisationCache() {
406 if ( is_null( self
::$dataCache ) ) {
407 global $wgLocalisationCacheConf;
408 $class = $wgLocalisationCacheConf['class'];
409 self
::$dataCache = new $class( $wgLocalisationCacheConf );
411 return self
::$dataCache;
414 function __construct() {
415 $this->mConverter
= new FakeConverter( $this );
416 // Set the code to the name of the descendant
417 if ( static::class === 'Language' ) {
420 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
422 self
::getLocalisationCache();
426 * Reduce memory usage
428 function __destruct() {
429 foreach ( $this as $name => $value ) {
430 unset( $this->$name );
435 * Hook which will be called if this is the content language.
436 * Descendants can use this to register hook functions or modify globals
438 function initContLang() {
445 public function getFallbackLanguages() {
446 return self
::getFallbacksFor( $this->mCode
);
450 * Exports $wgBookstoreListEn
453 public function getBookstoreList() {
454 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
458 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
459 * available in localised form, it will be included in English.
463 public function getNamespaces() {
464 if ( is_null( $this->namespaceNames
) ) {
465 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
467 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
469 $this->namespaceNames
= $wgExtraNamespaces +
470 self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
471 $this->namespaceNames +
= $validNamespaces;
473 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
474 if ( $wgMetaNamespaceTalk ) {
475 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
477 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
478 $this->namespaceNames
[NS_PROJECT_TALK
] =
479 $this->fixVariableInNamespace( $talk );
482 # Sometimes a language will be localised but not actually exist on this wiki.
483 foreach ( $this->namespaceNames
as $key => $text ) {
484 if ( !isset( $validNamespaces[$key] ) ) {
485 unset( $this->namespaceNames
[$key] );
489 # The above mixing may leave namespaces out of canonical order.
490 # Re-order by namespace ID number...
491 ksort( $this->namespaceNames
);
493 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
496 return $this->namespaceNames
;
500 * Arbitrarily set all of the namespace names at once. Mainly used for testing
501 * @param array $namespaces Array of namespaces (id => name)
503 public function setNamespaces( array $namespaces ) {
504 $this->namespaceNames
= $namespaces;
505 $this->mNamespaceIds
= null;
509 * Resets all of the namespace caches. Mainly used for testing
511 public function resetNamespaces() {
512 $this->namespaceNames
= null;
513 $this->mNamespaceIds
= null;
514 $this->namespaceAliases
= null;
518 * A convenience function that returns getNamespaces() with spaces instead of underscores
519 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
523 public function getFormattedNamespaces() {
524 $ns = $this->getNamespaces();
525 foreach ( $ns as $k => $v ) {
526 $ns[$k] = strtr( $v, '_', ' ' );
532 * Get a namespace value by key
535 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
536 * echo $mw_ns; // prints 'MediaWiki'
539 * @param int $index The array key of the namespace to return
540 * @return string|bool String if the namespace value exists, otherwise false
542 public function getNsText( $index ) {
543 $ns = $this->getNamespaces();
544 return $ns[$index] ??
false;
548 * A convenience function that returns the same thing as
549 * getNsText() except with '_' changed to ' ', useful for
553 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
554 * echo $mw_ns; // prints 'MediaWiki talk'
557 * @param int $index The array key of the namespace to return
558 * @return string Namespace name without underscores (empty string if namespace does not exist)
560 public function getFormattedNsText( $index ) {
561 $ns = $this->getNsText( $index );
562 return strtr( $ns, '_', ' ' );
566 * Returns gender-dependent namespace alias if available.
567 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
568 * @param int $index Namespace index
569 * @param string $gender Gender key (male, female... )
573 public function getGenderNsText( $index, $gender ) {
574 global $wgExtraGenderNamespaces;
576 $ns = $wgExtraGenderNamespaces +
577 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
579 return $ns[$index][$gender] ??
$this->getNsText( $index );
583 * Whether this language uses gender-dependent namespace aliases.
584 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
588 public function needsGenderDistinction() {
589 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
590 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
591 // $wgExtraGenderNamespaces overrides everything
593 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
594 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
595 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
598 // Check what is in i18n files
599 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
600 return count( $aliases ) > 0;
605 * Get a namespace key by value, case insensitive.
606 * Only matches namespace names for the current language, not the
607 * canonical ones defined in Namespace.php.
609 * @param string $text
610 * @return int|bool An integer if $text is a valid value otherwise false
612 function getLocalNsIndex( $text ) {
613 $lctext = $this->lc( $text );
614 $ids = $this->getNamespaceIds();
615 return $ids[$lctext] ??
false;
621 public function getNamespaceAliases() {
622 if ( is_null( $this->namespaceAliases
) ) {
623 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
627 foreach ( $aliases as $name => $index ) {
628 if ( $index === NS_PROJECT_TALK
) {
629 unset( $aliases[$name] );
630 $name = $this->fixVariableInNamespace( $name );
631 $aliases[$name] = $index;
636 global $wgExtraGenderNamespaces;
637 $genders = $wgExtraGenderNamespaces +
638 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
639 foreach ( $genders as $index => $forms ) {
640 foreach ( $forms as $alias ) {
641 $aliases[$alias] = $index;
645 # Also add converted namespace names as aliases, to avoid confusion.
646 $convertedNames = [];
647 foreach ( $this->getVariants() as $variant ) {
648 if ( $variant === $this->mCode
) {
651 foreach ( $this->getNamespaces() as $ns => $_ ) {
652 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
656 $this->namespaceAliases
= $aliases +
$convertedNames;
659 return $this->namespaceAliases
;
665 public function getNamespaceIds() {
666 if ( is_null( $this->mNamespaceIds
) ) {
667 global $wgNamespaceAliases;
668 # Put namespace names and aliases into a hashtable.
669 # If this is too slow, then we should arrange it so that it is done
670 # before caching. The catch is that at pre-cache time, the above
671 # class-specific fixup hasn't been done.
672 $this->mNamespaceIds
= [];
673 foreach ( $this->getNamespaces() as $index => $name ) {
674 $this->mNamespaceIds
[$this->lc( $name )] = $index;
676 foreach ( $this->getNamespaceAliases() as $name => $index ) {
677 $this->mNamespaceIds
[$this->lc( $name )] = $index;
679 if ( $wgNamespaceAliases ) {
680 foreach ( $wgNamespaceAliases as $name => $index ) {
681 $this->mNamespaceIds
[$this->lc( $name )] = $index;
685 return $this->mNamespaceIds
;
689 * Get a namespace key by value, case insensitive. Canonical namespace
690 * names override custom ones defined for the current language.
692 * @param string $text
693 * @return int|bool An integer if $text is a valid value otherwise false
695 public function getNsIndex( $text ) {
696 $lctext = $this->lc( $text );
697 $ns = MWNamespace
::getCanonicalIndex( $lctext );
698 if ( $ns !== null ) {
701 $ids = $this->getNamespaceIds();
702 return $ids[$lctext] ??
false;
706 * short names for language variants used for language conversion links.
708 * @param string $code
709 * @param bool $usemsg Use the "variantname-xyz" message if it exists
712 public function getVariantname( $code, $usemsg = true ) {
713 $msg = "variantname-$code";
714 if ( $usemsg && wfMessage( $msg )->exists() ) {
715 return $this->getMessageFromDB( $msg );
717 $name = self
::fetchLanguageName( $code );
719 return $name; # if it's defined as a language name, show that
721 # otherwise, output the language code
727 * @return string[]|bool List of date format preference keys, or false if disabled.
729 public function getDatePreferences() {
730 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
736 function getDateFormats() {
737 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
741 * @return array|string
743 public function getDefaultDateFormat() {
744 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
745 if ( $df === 'dmy or mdy' ) {
746 global $wgAmericanDates;
747 return $wgAmericanDates ?
'mdy' : 'dmy';
756 public function getDatePreferenceMigrationMap() {
757 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
761 * @param string $image
764 function getImageFile( $image ) {
765 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
772 public function getImageFiles() {
773 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
779 public function getExtraUserToggles() {
780 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
787 function getUserToggle( $tog ) {
788 return $this->getMessageFromDB( "tog-$tog" );
792 * Get an array of language names, indexed by code.
793 * @param null|string $inLanguage Code of language in which to return the names
794 * Use null for autonyms (native names)
795 * @param string $include One of:
796 * 'all' all available languages
797 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
798 * 'mwfile' only if the language is in 'mw' *and* has a message file
799 * @return array Language code => language name (sorted by key)
802 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
803 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
804 $cacheKey .= ":$include";
805 if ( self
::$languageNameCache === null ) {
806 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
809 $ret = self
::$languageNameCache->get( $cacheKey );
811 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
812 self
::$languageNameCache->set( $cacheKey, $ret );
818 * Uncached helper for fetchLanguageNames
819 * @param null|string $inLanguage Code of language in which to return the names
820 * Use null for autonyms (native names)
821 * @param string $include One of:
822 * 'all' all available languages
823 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
824 * 'mwfile' only if the language is in 'mw' *and* has a message file
825 * @return array Language code => language name (sorted by key)
827 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
828 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
830 // If passed an invalid language code to use, fallback to en
831 if ( $inLanguage !== null && !self
::isValidCode( $inLanguage ) ) {
838 # TODO: also include when $inLanguage is null, when this code is more efficient
839 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
842 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
843 if ( $wgUsePigLatinVariant ) {
844 // Pig Latin (for variant development)
845 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
848 foreach ( $mwNames as $mwCode => $mwName ) {
849 # - Prefer own MediaWiki native name when not using the hook
850 # - For other names just add if not added through the hook
851 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
852 $names[$mwCode] = $mwName;
856 if ( $include === 'all' ) {
862 $coreCodes = array_keys( $mwNames );
863 foreach ( $coreCodes as $coreCode ) {
864 $returnMw[$coreCode] = $names[$coreCode];
867 if ( $include === 'mwfile' ) {
869 # We do this using a foreach over the codes instead of a directory
870 # loop so that messages files in extensions will work correctly.
871 foreach ( $returnMw as $code => $value ) {
872 if ( is_readable( self
::getMessagesFileName( $code ) )
873 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
875 $namesMwFile[$code] = $names[$code];
879 ksort( $namesMwFile );
884 # 'mw' option; default if it's not one of the other two options (all/mwfile)
889 * @param string $code The code of the language for which to get the name
890 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
891 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
892 * @return string Language name or empty
895 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
896 $code = strtolower( $code );
897 $array = self
::fetchLanguageNames( $inLanguage, $include );
898 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
902 * Get a message from the MediaWiki namespace.
904 * @param string $msg Message name
907 public function getMessageFromDB( $msg ) {
908 return $this->msg( $msg )->text();
912 * Get message object in this language. Only for use inside this class.
914 * @param string $msg Message name
917 protected function msg( $msg ) {
918 return wfMessage( $msg )->inLanguage( $this );
925 public function getMonthName( $key ) {
926 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
932 public function getMonthNamesArray() {
933 $monthNames = [ '' ];
934 for ( $i = 1; $i < 13; $i++
) {
935 $monthNames[] = $this->getMonthName( $i );
944 public function getMonthNameGen( $key ) {
945 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
952 public function getMonthAbbreviation( $key ) {
953 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
959 public function getMonthAbbreviationsArray() {
960 $monthNames = [ '' ];
961 for ( $i = 1; $i < 13; $i++
) {
962 $monthNames[] = $this->getMonthAbbreviation( $i );
971 public function getWeekdayName( $key ) {
972 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
979 function getWeekdayAbbreviation( $key ) {
980 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
987 function getIranianCalendarMonthName( $key ) {
988 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
995 function getHebrewCalendarMonthName( $key ) {
996 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1000 * @param string $key
1003 function getHebrewCalendarMonthNameGen( $key ) {
1004 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1008 * @param string $key
1011 function getHijriCalendarMonthName( $key ) {
1012 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1016 * Pass through result from $dateTimeObj->format()
1017 * @param DateTime|bool|null &$dateTimeObj
1019 * @param DateTimeZone|bool|null $zone
1020 * @param string $code
1023 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1024 if ( !$dateTimeObj ) {
1025 $dateTimeObj = DateTime
::createFromFormat(
1026 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1029 return $dateTimeObj->format( $code );
1033 * This is a workalike of PHP's date() function, but with better
1034 * internationalisation, a reduced set of format characters, and a better
1037 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1038 * the PHP manual for definitions. There are a number of extensions, which
1041 * xn Do not translate digits of the next numeric format character
1042 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1043 * xr Use roman numerals for the next numeric format character
1044 * xh Use hebrew numerals for the next numeric format character
1046 * xg Genitive month name
1048 * xij j (day number) in Iranian calendar
1049 * xiF F (month name) in Iranian calendar
1050 * xin n (month number) in Iranian calendar
1051 * xiy y (two digit year) in Iranian calendar
1052 * xiY Y (full year) in Iranian calendar
1053 * xit t (days in month) in Iranian calendar
1054 * xiz z (day of the year) in Iranian calendar
1056 * xjj j (day number) in Hebrew calendar
1057 * xjF F (month name) in Hebrew calendar
1058 * xjt t (days in month) in Hebrew calendar
1059 * xjx xg (genitive month name) in Hebrew calendar
1060 * xjn n (month number) in Hebrew calendar
1061 * xjY Y (full year) in Hebrew calendar
1063 * xmj j (day number) in Hijri calendar
1064 * xmF F (month name) in Hijri calendar
1065 * xmn n (month number) in Hijri calendar
1066 * xmY Y (full year) in Hijri calendar
1068 * xkY Y (full year) in Thai solar calendar. Months and days are
1069 * identical to the Gregorian calendar
1070 * xoY Y (full year) in Minguo calendar or Juche year.
1071 * Months and days are identical to the
1072 * Gregorian calendar
1073 * xtY Y (full year) in Japanese nengo. Months and days are
1074 * identical to the Gregorian calendar
1076 * Characters enclosed in double quotes will be considered literal (with
1077 * the quotes themselves removed). Unmatched quotes will be considered
1078 * literal quotes. Example:
1080 * "The month is" F => The month is January
1083 * Backslash escaping is also supported.
1085 * Input timestamp is assumed to be pre-normalized to the desired local
1086 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1087 * $ts is UTC if $zone is not given.
1089 * @param string $format
1090 * @param string $ts 14-character timestamp
1093 * @param DateTimeZone|null $zone Timezone of $ts
1094 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1095 * Only makes sense if $ts is the current time.
1096 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1098 * @throws MWException
1101 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1106 $dateTimeObj = false;
1115 $usedSecond = false;
1116 $usedMinute = false;
1123 $usedISOYear = false;
1124 $usedIsLeapYear = false;
1126 $usedHebrewMonth = false;
1127 $usedIranianMonth = false;
1128 $usedHijriMonth = false;
1129 $usedHebrewYear = false;
1130 $usedIranianYear = false;
1131 $usedHijriYear = false;
1132 $usedTennoYear = false;
1134 if ( strlen( $ts ) !== 14 ) {
1135 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1138 if ( !ctype_digit( $ts ) ) {
1139 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1142 $formatLength = strlen( $format );
1143 for ( $p = 0; $p < $formatLength; $p++
) {
1145 $code = $format[$p];
1146 if ( $code == 'x' && $p < $formatLength - 1 ) {
1147 $code .= $format[++
$p];
1150 if ( ( $code === 'xi'
1156 && $p < $formatLength - 1 ) {
1157 $code .= $format[++
$p];
1168 $rawToggle = !$rawToggle;
1178 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1181 $usedHebrewMonth = true;
1183 $hebrew = self
::tsToHebrew( $ts );
1185 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1189 $num = substr( $ts, 6, 2 );
1193 $s .= $this->getWeekdayAbbreviation(
1194 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1199 $num = intval( substr( $ts, 6, 2 ) );
1204 $iranian = self
::tsToIranian( $ts );
1211 $hijri = self
::tsToHijri( $ts );
1218 $hebrew = self
::tsToHebrew( $ts );
1224 $s .= $this->getWeekdayName(
1225 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1230 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1233 $usedIranianMonth = true;
1235 $iranian = self
::tsToIranian( $ts );
1237 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1240 $usedHijriMonth = true;
1242 $hijri = self
::tsToHijri( $ts );
1244 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1247 $usedHebrewMonth = true;
1249 $hebrew = self
::tsToHebrew( $ts );
1251 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1255 $num = substr( $ts, 4, 2 );
1259 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1263 $num = intval( substr( $ts, 4, 2 ) );
1266 $usedIranianMonth = true;
1268 $iranian = self
::tsToIranian( $ts );
1273 $usedHijriMonth = true;
1275 $hijri = self
::tsToHijri( $ts );
1280 $usedHebrewMonth = true;
1282 $hebrew = self
::tsToHebrew( $ts );
1287 $usedHebrewMonth = true;
1289 $hebrew = self
::tsToHebrew( $ts );
1295 $num = substr( $ts, 0, 4 );
1298 $usedIranianYear = true;
1300 $iranian = self
::tsToIranian( $ts );
1305 $usedHijriYear = true;
1307 $hijri = self
::tsToHijri( $ts );
1312 $usedHebrewYear = true;
1314 $hebrew = self
::tsToHebrew( $ts );
1321 $thai = self
::tsToYear( $ts, 'thai' );
1328 $minguo = self
::tsToYear( $ts, 'minguo' );
1333 $usedTennoYear = true;
1335 $tenno = self
::tsToYear( $ts, 'tenno' );
1341 $num = substr( $ts, 2, 2 );
1344 $usedIranianYear = true;
1346 $iranian = self
::tsToIranian( $ts );
1348 $num = substr( $iranian[0], -2 );
1351 $usedIranianYear = true;
1353 $iranian = self
::tsToIranian( $ts );
1355 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1358 $usedIranianYear = true;
1360 $iranian = self
::tsToIranian( $ts );
1366 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1370 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1374 $h = substr( $ts, 8, 2 );
1375 $num = $h %
12 ?
$h %
12 : 12;
1379 $num = intval( substr( $ts, 8, 2 ) );
1383 $h = substr( $ts, 8, 2 );
1384 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1388 $num = substr( $ts, 8, 2 );
1392 $num = substr( $ts, 10, 2 );
1396 $num = substr( $ts, 12, 2 );
1406 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1412 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1416 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1420 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1423 $usedIsLeapYear = true;
1424 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1427 $usedISOYear = true;
1428 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1435 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1438 # Backslash escaping
1439 if ( $p < $formatLength - 1 ) {
1440 $s .= $format[++
$p];
1447 if ( $p < $formatLength - 1 ) {
1448 $endQuote = strpos( $format, '"', $p +
1 );
1449 if ( $endQuote === false ) {
1450 # No terminating quote, assume literal "
1453 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1457 # Quote at end of string, assume literal "
1464 if ( $num !== false ) {
1465 if ( $rawToggle ||
$raw ) {
1468 } elseif ( $roman ) {
1469 $s .= self
::romanNumeral( $num );
1471 } elseif ( $hebrewNum ) {
1472 $s .= self
::hebrewNumeral( $num );
1475 $s .= $this->formatNum( $num, true );
1480 if ( $ttl === 'unused' ) {
1481 // No need to calculate the TTL, the caller wont use it anyway.
1482 } elseif ( $usedSecond ) {
1484 } elseif ( $usedMinute ) {
1485 $ttl = 60 - substr( $ts, 12, 2 );
1486 } elseif ( $usedHour ) {
1487 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1488 } elseif ( $usedAMPM ) {
1489 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1490 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1494 $usedIranianMonth ||
1501 // @todo Someone who understands the non-Gregorian calendars
1502 // should write proper logic for them so that they don't need purged every day.
1503 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1504 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1507 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1508 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1511 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1512 $timeRemainingInDay;
1513 } elseif ( $usedISOYear ) {
1514 // December 28th falls on the last ISO week of the year, every year.
1515 // The last ISO week of a year can be 52 or 53.
1516 $lastWeekOfISOYear = DateTime
::createFromFormat(
1518 substr( $ts, 0, 4 ) . '1228',
1519 $zone ?
: new DateTimeZone( 'UTC' )
1521 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1522 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1523 $timeRemainingInWeek =
1524 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1525 +
$timeRemainingInDay;
1526 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1531 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1532 substr( $ts, 6, 2 ) ) * 86400
1533 +
$timeRemainingInDay;
1534 } elseif ( $usedYear ) {
1536 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1537 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1538 +
$timeRemainingInDay;
1539 } elseif ( $usedIsLeapYear ) {
1540 $year = substr( $ts, 0, 4 );
1541 $timeRemainingInYear =
1542 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1543 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1544 +
$timeRemainingInDay;
1546 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1547 // this isn't a leap year. see when the next one starts
1548 $nextCandidate = $year - $mod +
4;
1549 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1550 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1551 $timeRemainingInYear;
1553 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1554 $timeRemainingInYear;
1557 // this is a leap year, so the next year isn't
1558 $possibleTtls[] = $timeRemainingInYear;
1562 if ( $possibleTtls ) {
1563 $ttl = min( $possibleTtls );
1570 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1571 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1574 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1575 * Gregorian dates to Iranian dates. Originally written in C, it
1576 * is released under the terms of GNU Lesser General Public
1577 * License. Conversion to PHP was performed by Niklas Laxström.
1579 * Link: http://www.farsiweb.info/jalali/jalali.c
1585 private static function tsToIranian( $ts ) {
1586 $gy = substr( $ts, 0, 4 ) - 1600;
1587 $gm = substr( $ts, 4, 2 ) - 1;
1588 $gd = substr( $ts, 6, 2 ) - 1;
1590 # Days passed from the beginning (including leap years)
1592 +
floor( ( $gy +
3 ) / 4 )
1593 - floor( ( $gy +
99 ) / 100 )
1594 +
floor( ( $gy +
399 ) / 400 );
1596 // Add days of the past months of this year
1597 for ( $i = 0; $i < $gm; $i++
) {
1598 $gDayNo +
= self
::$GREG_DAYS[$i];
1602 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1606 // Days passed in current month
1607 $gDayNo +
= (int)$gd;
1609 $jDayNo = $gDayNo - 79;
1611 $jNp = floor( $jDayNo / 12053 );
1614 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1617 if ( $jDayNo >= 366 ) {
1618 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1619 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1624 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1625 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1631 return [ $jy, $jm, $jd, $jz ];
1635 * Converting Gregorian dates to Hijri dates.
1637 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1639 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1645 private static function tsToHijri( $ts ) {
1646 $year = substr( $ts, 0, 4 );
1647 $month = substr( $ts, 4, 2 );
1648 $day = substr( $ts, 6, 2 );
1656 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1657 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1659 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1660 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1661 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1664 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1665 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1668 $zl = $zjd - 1948440 +
10632;
1669 $zn = (int)( ( $zl - 1 ) / 10631 );
1670 $zl = $zl - 10631 * $zn +
354;
1671 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1672 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1673 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1674 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1675 $zm = (int)( ( 24 * $zl ) / 709 );
1676 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1677 $zy = 30 * $zn +
$zj - 30;
1679 return [ $zy, $zm, $zd ];
1683 * Converting Gregorian dates to Hebrew dates.
1685 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1686 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1687 * to translate the relevant functions into PHP and release them under
1690 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1691 * and Adar II is 14. In a non-leap year, Adar is 6.
1697 private static function tsToHebrew( $ts ) {
1699 $year = substr( $ts, 0, 4 );
1700 $month = substr( $ts, 4, 2 );
1701 $day = substr( $ts, 6, 2 );
1703 # Calculate Hebrew year
1704 $hebrewYear = $year +
3760;
1706 # Month number when September = 1, August = 12
1708 if ( $month > 12 ) {
1715 # Calculate day of year from 1 September
1717 for ( $i = 1; $i < $month; $i++
) {
1721 # Check if the year is leap
1722 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1725 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1732 # Calculate the start of the Hebrew year
1733 $start = self
::hebrewYearStart( $hebrewYear );
1735 # Calculate next year's start
1736 if ( $dayOfYear <= $start ) {
1737 # Day is before the start of the year - it is the previous year
1739 $nextStart = $start;
1743 # Add days since previous year's 1 September
1745 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1749 # Start of the new (previous) year
1750 $start = self
::hebrewYearStart( $hebrewYear );
1753 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1756 # Calculate Hebrew day of year
1757 $hebrewDayOfYear = $dayOfYear - $start;
1759 # Difference between year's days
1760 $diff = $nextStart - $start;
1761 # Add 12 (or 13 for leap years) days to ignore the difference between
1762 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1763 # difference is only about the year type
1764 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1770 # Check the year pattern, and is leap year
1771 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1772 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1773 # and non-leap years
1774 $yearPattern = $diff %
30;
1775 # Check if leap year
1776 $isLeap = $diff >= 30;
1778 # Calculate day in the month from number of day in the Hebrew year
1779 # Don't check Adar - if the day is not in Adar, we will stop before;
1780 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1781 $hebrewDay = $hebrewDayOfYear;
1784 while ( $hebrewMonth <= 12 ) {
1785 # Calculate days in this month
1786 if ( $isLeap && $hebrewMonth == 6 ) {
1787 # Adar in a leap year
1789 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1791 if ( $hebrewDay <= $days ) {
1795 # Subtract the days of Adar I
1796 $hebrewDay -= $days;
1799 if ( $hebrewDay <= $days ) {
1805 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1806 # Cheshvan in a complete year (otherwise as the rule below)
1808 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1809 # Kislev in an incomplete year (otherwise as the rule below)
1812 # Odd months have 30 days, even have 29
1813 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1815 if ( $hebrewDay <= $days ) {
1816 # In the current month
1819 # Subtract the days of the current month
1820 $hebrewDay -= $days;
1821 # Try in the next month
1826 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1830 * This calculates the Hebrew year start, as days since 1 September.
1831 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1832 * Used for Hebrew date.
1838 private static function hebrewYearStart( $year ) {
1839 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1840 $b = intval( ( $year - 1 ) %
4 );
1841 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1845 $Mar = intval( $m );
1851 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1852 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1854 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1856 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1860 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1865 * Algorithm to convert Gregorian dates to Thai solar dates,
1866 * Minguo dates or Minguo dates.
1868 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1869 * https://en.wikipedia.org/wiki/Minguo_calendar
1870 * https://en.wikipedia.org/wiki/Japanese_era_name
1872 * @param string $ts 14-character timestamp
1873 * @param string $cName Calender name
1874 * @return array Converted year, month, day
1876 private static function tsToYear( $ts, $cName ) {
1877 $gy = substr( $ts, 0, 4 );
1878 $gm = substr( $ts, 4, 2 );
1879 $gd = substr( $ts, 6, 2 );
1881 if ( !strcmp( $cName, 'thai' ) ) {
1883 # Add 543 years to the Gregorian calendar
1884 # Months and days are identical
1885 $gy_offset = $gy +
543;
1886 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1888 # Deduct 1911 years from the Gregorian calendar
1889 # Months and days are identical
1890 $gy_offset = $gy - 1911;
1891 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1892 # Nengō dates up to Meiji period
1893 # Deduct years from the Gregorian calendar
1894 # depending on the nengo periods
1895 # Months and days are identical
1897 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1898 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1901 $gy_gannen = $gy - 1868 +
1;
1902 $gy_offset = $gy_gannen;
1903 if ( $gy_gannen == 1 ) {
1906 $gy_offset = '明治' . $gy_offset;
1908 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1909 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1910 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1911 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1912 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1915 $gy_gannen = $gy - 1912 +
1;
1916 $gy_offset = $gy_gannen;
1917 if ( $gy_gannen == 1 ) {
1920 $gy_offset = '大正' . $gy_offset;
1922 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1923 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1924 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1927 $gy_gannen = $gy - 1926 +
1;
1928 $gy_offset = $gy_gannen;
1929 if ( $gy_gannen == 1 ) {
1932 $gy_offset = '昭和' . $gy_offset;
1935 $gy_gannen = $gy - 1989 +
1;
1936 $gy_offset = $gy_gannen;
1937 if ( $gy_gannen == 1 ) {
1940 $gy_offset = '平成' . $gy_offset;
1946 return [ $gy_offset, $gm, $gd ];
1950 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1952 * This is the rule the BIDI algorithm uses to determine the directionality of
1953 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
1954 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1956 * TODO: Does not handle BIDI control characters inside the text.
1957 * TODO: Does not handle unallocated characters.
1959 * @param string $text Text to test
1960 * @return null|string Directionality ('ltr' or 'rtl') or null
1962 private static function strongDirFromContent( $text = '' ) {
1963 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
1966 if ( $matches[1] === '' ) {
1973 * Roman number formatting up to 10000
1979 static function romanNumeral( $num ) {
1981 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1982 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1983 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1984 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1985 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1988 $num = intval( $num );
1989 if ( $num > 10000 ||
$num <= 0 ) {
1994 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1995 if ( $num >= $pow10 ) {
1996 $s .= $table[$i][(int)floor( $num / $pow10 )];
1998 $num = $num %
$pow10;
2004 * Hebrew Gematria number formatting up to 9999
2010 static function hebrewNumeral( $num ) {
2012 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2013 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2026 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2029 $num = intval( $num );
2030 if ( $num > 9999 ||
$num <= 0 ) {
2034 // Round thousands have special notations
2035 if ( $num === 1000 ) {
2037 } elseif ( $num %
1000 === 0 ) {
2038 return $table[0][$num / 1000] . "' אלפים";
2043 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2044 if ( $num >= $pow10 ) {
2045 if ( $num === 15 ||
$num === 16 ) {
2046 $letters[] = $table[0][9];
2047 $letters[] = $table[0][$num - 9];
2050 $letters = array_merge(
2052 (array)$table[$i][intval( $num / $pow10 )]
2055 if ( $pow10 === 1000 ) {
2061 $num = $num %
$pow10;
2064 $preTransformLength = count( $letters );
2065 if ( $preTransformLength === 1 ) {
2066 // Add geresh (single quote) to one-letter numbers
2069 $lastIndex = $preTransformLength - 1;
2070 $letters[$lastIndex] = str_replace(
2071 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2072 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2073 $letters[$lastIndex]
2076 // Add gershayim (double quote) to multiple-letter numbers,
2077 // but exclude numbers with only one letter after the thousands
2078 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2079 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2082 array_splice( $letters, -1, 0, '"' );
2086 return implode( $letters );
2090 * Used by date() and time() to adjust the time output.
2092 * @param string $ts The time in date('YmdHis') format
2093 * @param mixed $tz Adjust the time by this amount (default false, mean we
2094 * get user timecorrection setting)
2097 public function userAdjust( $ts, $tz = false ) {
2098 global $wgUser, $wgLocalTZoffset;
2100 if ( $tz === false ) {
2101 $tz = $wgUser->getOption( 'timecorrection' );
2104 $data = explode( '|', $tz, 3 );
2106 if ( $data[0] == 'ZoneInfo' ) {
2108 $userTZ = new DateTimeZone( $data[2] );
2109 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2110 $date->setTimezone( $userTZ );
2111 return $date->format( 'YmdHis' );
2112 } catch ( Exception
$e ) {
2113 // Unrecognized timezone, default to 'Offset' with the stored offset.
2114 $data[0] = 'Offset';
2118 if ( $data[0] == 'System' ||
$tz == '' ) {
2119 # Global offset in minutes.
2120 $minDiff = $wgLocalTZoffset;
2121 } elseif ( $data[0] == 'Offset' ) {
2122 $minDiff = intval( $data[1] );
2124 $data = explode( ':', $tz );
2125 if ( count( $data ) == 2 ) {
2126 $data[0] = intval( $data[0] );
2127 $data[1] = intval( $data[1] );
2128 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2129 if ( $data[0] < 0 ) {
2130 $minDiff = -$minDiff;
2133 $minDiff = intval( $data[0] ) * 60;
2137 # No difference ? Return time unchanged
2138 if ( 0 == $minDiff ) {
2142 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2143 # Generate an adjusted date; take advantage of the fact that mktime
2144 # will normalize out-of-range values so we don't have to split $minDiff
2145 # into hours and minutes.
2147 (int)substr( $ts, 8, 2 ) ), # Hours
2148 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2149 (int)substr( $ts, 12, 2 ), # Seconds
2150 (int)substr( $ts, 4, 2 ), # Month
2151 (int)substr( $ts, 6, 2 ), # Day
2152 (int)substr( $ts, 0, 4 ) ); # Year
2154 $date = date( 'YmdHis', $t );
2155 Wikimedia\restoreWarnings
();
2161 * This is meant to be used by time(), date(), and timeanddate() to get
2162 * the date preference they're supposed to use, it should be used in
2165 * function timeanddate([...], $format = true) {
2166 * $datePreference = $this->dateFormat($format);
2170 * @param int|string|bool $usePrefs If true, the user's preference is used
2171 * if false, the site/language default is used
2172 * if int/string, assumed to be a format.
2175 function dateFormat( $usePrefs = true ) {
2178 if ( is_bool( $usePrefs ) ) {
2180 $datePreference = $wgUser->getDatePreference();
2182 $datePreference = (string)User
::getDefaultOption( 'date' );
2185 $datePreference = (string)$usePrefs;
2189 if ( $datePreference == '' ) {
2193 return $datePreference;
2197 * Get a format string for a given type and preference
2198 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2199 * @param string $pref The format name as it appears in Messages*.php under
2202 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2206 function getDateFormatString( $type, $pref ) {
2207 $wasDefault = false;
2208 if ( $pref == 'default' ) {
2210 $pref = $this->getDefaultDateFormat();
2213 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2214 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2216 if ( $type === 'pretty' && $df === null ) {
2217 $df = $this->getDateFormatString( 'date', $pref );
2220 if ( !$wasDefault && $df === null ) {
2221 $pref = $this->getDefaultDateFormat();
2222 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2225 $this->dateFormatStrings
[$type][$pref] = $df;
2227 return $this->dateFormatStrings
[$type][$pref];
2231 * @param string $ts The time format which needs to be turned into a
2232 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2233 * @param bool $adj Whether to adjust the time output according to the
2234 * user configured offset ($timecorrection)
2235 * @param mixed $format True to use user's date format preference
2236 * @param string|bool $timecorrection The time offset as returned by
2237 * validateTimeZone() in Special:Preferences
2240 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2241 $ts = wfTimestamp( TS_MW
, $ts );
2243 $ts = $this->userAdjust( $ts, $timecorrection );
2245 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2246 return $this->sprintfDate( $df, $ts );
2250 * @param string $ts The time format which needs to be turned into a
2251 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2252 * @param bool $adj Whether to adjust the time output according to the
2253 * user configured offset ($timecorrection)
2254 * @param mixed $format True to use user's date format preference
2255 * @param string|bool $timecorrection The time offset as returned by
2256 * validateTimeZone() in Special:Preferences
2259 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2260 $ts = wfTimestamp( TS_MW
, $ts );
2262 $ts = $this->userAdjust( $ts, $timecorrection );
2264 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2265 return $this->sprintfDate( $df, $ts );
2269 * @param string $ts The time format which needs to be turned into a
2270 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2271 * @param bool $adj Whether to adjust the time output according to the
2272 * user configured offset ($timecorrection)
2273 * @param mixed $format What format to return, if it's false output the
2274 * default one (default true)
2275 * @param string|bool $timecorrection The time offset as returned by
2276 * validateTimeZone() in Special:Preferences
2279 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2280 $ts = wfTimestamp( TS_MW
, $ts );
2282 $ts = $this->userAdjust( $ts, $timecorrection );
2284 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2285 return $this->sprintfDate( $df, $ts );
2289 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2293 * @param int $seconds The amount of seconds.
2294 * @param array $chosenIntervals The intervals to enable.
2298 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2299 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2303 foreach ( $intervals as $intervalName => $intervalValue ) {
2304 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2305 // duration-years, duration-decades, duration-centuries, duration-millennia
2306 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2307 $segments[] = $message->inLanguage( $this )->escaped();
2310 return $this->listToText( $segments );
2314 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2315 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2319 * @param int $seconds The amount of seconds.
2320 * @param array $chosenIntervals The intervals to enable.
2324 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2325 if ( empty( $chosenIntervals ) ) {
2326 $chosenIntervals = [
2338 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2339 $sortedNames = array_keys( $intervals );
2340 $smallestInterval = array_pop( $sortedNames );
2344 foreach ( $intervals as $name => $length ) {
2345 $value = floor( $seconds / $length );
2347 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2348 $seconds -= $value * $length;
2349 $segments[$name] = $value;
2357 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2359 * @param string $type Can be 'date', 'time' or 'both'
2360 * @param string $ts The time format which needs to be turned into a
2361 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2362 * @param User $user User object used to get preferences for timezone and format
2363 * @param array $options Array, can contain the following keys:
2364 * - 'timecorrection': time correction, can have the following values:
2365 * - true: use user's preference
2366 * - false: don't use time correction
2367 * - int: value of time correction in minutes
2368 * - 'format': format to use, can have the following values:
2369 * - true: use user's preference
2370 * - false: use default preference
2371 * - string: format to use
2375 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2376 $ts = wfTimestamp( TS_MW
, $ts );
2377 $options +
= [ 'timecorrection' => true, 'format' => true ];
2378 if ( $options['timecorrection'] !== false ) {
2379 if ( $options['timecorrection'] === true ) {
2380 $offset = $user->getOption( 'timecorrection' );
2382 $offset = $options['timecorrection'];
2384 $ts = $this->userAdjust( $ts, $offset );
2386 if ( $options['format'] === true ) {
2387 $format = $user->getDatePreference();
2389 $format = $options['format'];
2391 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2392 return $this->sprintfDate( $df, $ts );
2396 * Get the formatted date for the given timestamp and formatted for
2399 * @param mixed $ts Mixed: the time format which needs to be turned into a
2400 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2401 * @param User $user User object used to get preferences for timezone and format
2402 * @param array $options Array, can contain the following keys:
2403 * - 'timecorrection': time correction, can have the following values:
2404 * - true: use user's preference
2405 * - false: don't use time correction
2406 * - int: value of time correction in minutes
2407 * - 'format': format to use, can have the following values:
2408 * - true: use user's preference
2409 * - false: use default preference
2410 * - string: format to use
2414 public function userDate( $ts, User
$user, array $options = [] ) {
2415 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2419 * Get the formatted time for the given timestamp and formatted for
2422 * @param mixed $ts The time format which needs to be turned into a
2423 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2424 * @param User $user User object used to get preferences for timezone and format
2425 * @param array $options Array, can contain the following keys:
2426 * - 'timecorrection': time correction, can have the following values:
2427 * - true: use user's preference
2428 * - false: don't use time correction
2429 * - int: value of time correction in minutes
2430 * - 'format': format to use, can have the following values:
2431 * - true: use user's preference
2432 * - false: use default preference
2433 * - string: format to use
2437 public function userTime( $ts, User
$user, array $options = [] ) {
2438 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2442 * Get the formatted date and time for the given timestamp and formatted for
2445 * @param mixed $ts The time format which needs to be turned into a
2446 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2447 * @param User $user User object used to get preferences for timezone and format
2448 * @param array $options Array, can contain the following keys:
2449 * - 'timecorrection': time correction, can have the following values:
2450 * - true: use user's preference
2451 * - false: don't use time correction
2452 * - int: value of time correction in minutes
2453 * - 'format': format to use, can have the following values:
2454 * - true: use user's preference
2455 * - false: use default preference
2456 * - string: format to use
2460 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2461 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2465 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2467 * Determine the difference between the timestamp and the current time, and
2468 * generate a readable timestamp by returning "<N> <units> ago", where the
2469 * largest possible unit is used.
2471 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2473 * @param MWTimestamp $time
2474 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2475 * @param User|null $user User the timestamp is being generated for
2476 * (or null to use main context's user)
2477 * @return string Formatted timestamp
2479 public function getHumanTimestamp(
2480 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2482 if ( $relativeTo === null ) {
2483 $relativeTo = new MWTimestamp();
2485 if ( $user === null ) {
2486 $user = RequestContext
::getMain()->getUser();
2489 // Adjust for the user's timezone.
2490 $offsetThis = $time->offsetForUser( $user );
2491 $offsetRel = $relativeTo->offsetForUser( $user );
2494 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2495 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2498 // Reset the timezone on the objects.
2499 $time->timestamp
->sub( $offsetThis );
2500 $relativeTo->timestamp
->sub( $offsetRel );
2506 * Convert an MWTimestamp into a pretty human-readable timestamp using
2507 * the given user preferences and relative base time.
2509 * @see Language::getHumanTimestamp
2510 * @param MWTimestamp $ts Timestamp to prettify
2511 * @param MWTimestamp $relativeTo Base timestamp
2512 * @param User $user User preferences to use
2513 * @return string Human timestamp
2516 private function getHumanTimestampInternal(
2517 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2519 $diff = $ts->diff( $relativeTo );
2520 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2521 (int)$relativeTo->timestamp
->format( 'w' ) );
2522 $days = $diff->days ?
: (int)$diffDay;
2523 if ( $diff->invert ||
$days > 5
2524 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2526 // Timestamps are in different years: use full timestamp
2527 // Also do full timestamp for future dates
2529 * @todo FIXME: Add better handling of future timestamps.
2531 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2532 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2533 } elseif ( $days > 5 ) {
2534 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2535 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2536 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2537 } elseif ( $days > 1 ) {
2538 // Timestamp within the past week: show the day of the week and time
2539 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2540 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2542 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2543 $ts = wfMessage( "$weekday-at" )
2544 ->inLanguage( $this )
2545 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2547 } elseif ( $days == 1 ) {
2548 // Timestamp was yesterday: say 'yesterday' and the time.
2549 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2550 $ts = wfMessage( 'yesterday-at' )
2551 ->inLanguage( $this )
2552 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2554 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2555 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2556 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2557 $ts = wfMessage( 'today-at' )
2558 ->inLanguage( $this )
2559 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2562 // From here on in, the timestamp was soon enough ago so that we can simply say
2563 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2564 } elseif ( $diff->h
== 1 ) {
2565 // Less than 90 minutes, but more than an hour ago.
2566 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2567 } elseif ( $diff->i
>= 1 ) {
2568 // A few minutes ago.
2569 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2570 } elseif ( $diff->s
>= 30 ) {
2571 // Less than a minute, but more than 30 sec ago.
2572 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2574 // Less than 30 seconds ago.
2575 $ts = wfMessage( 'just-now' )->text();
2582 * @param string $key
2583 * @return string|null
2585 public function getMessage( $key ) {
2586 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2592 function getAllMessages() {
2593 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2598 * @param string $out
2599 * @param string $string
2602 public function iconv( $in, $out, $string ) {
2603 # Even with //IGNORE iconv can whine about illegal characters in
2604 # *input* string. We just ignore those too.
2605 # REF: https://bugs.php.net/bug.php?id=37166
2606 # REF: https://phabricator.wikimedia.org/T18885
2607 Wikimedia\
suppressWarnings();
2608 $text = iconv( $in, $out . '//IGNORE', $string );
2609 Wikimedia\restoreWarnings
();
2613 // callback functions for ucwords(), ucwordbreaks()
2616 * @param array $matches
2617 * @return mixed|string
2619 function ucwordbreaksCallbackAscii( $matches ) {
2620 return $this->ucfirst( $matches[1] );
2624 * @param array $matches
2627 function ucwordbreaksCallbackMB( $matches ) {
2628 return mb_strtoupper( $matches[0] );
2632 * @param array $matches
2635 function ucwordsCallbackMB( $matches ) {
2636 return mb_strtoupper( $matches[0] );
2640 * Make a string's first character uppercase
2642 * @param string $str
2646 public function ucfirst( $str ) {
2648 if ( $o < 96 ) { // if already uppercase...
2650 } elseif ( $o < 128 ) {
2651 return ucfirst( $str ); // use PHP's ucfirst()
2653 // fall back to more complex logic in case of multibyte strings
2654 return $this->uc( $str, true );
2659 * Convert a string to uppercase
2661 * @param string $str
2662 * @param bool $first
2666 public function uc( $str, $first = false ) {
2668 if ( $this->isMultibyte( $str ) ) {
2669 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2671 return ucfirst( $str );
2674 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2679 * @param string $str
2680 * @return mixed|string
2682 function lcfirst( $str ) {
2685 return strval( $str );
2686 } elseif ( $o >= 128 ) {
2687 return $this->lc( $str, true );
2688 } elseif ( $o > 96 ) {
2691 $str[0] = strtolower( $str[0] );
2697 * @param string $str
2698 * @param bool $first
2699 * @return mixed|string
2701 function lc( $str, $first = false ) {
2703 if ( $this->isMultibyte( $str ) ) {
2704 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2706 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2709 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2714 * @param string $str
2717 function isMultibyte( $str ) {
2718 return strlen( $str ) !== mb_strlen( $str );
2722 * @param string $str
2723 * @return mixed|string
2725 function ucwords( $str ) {
2726 if ( $this->isMultibyte( $str ) ) {
2727 $str = $this->lc( $str );
2729 // regexp to find first letter in each word (i.e. after each space)
2730 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2732 // function to use to capitalize a single char
2733 return preg_replace_callback(
2735 [ $this, 'ucwordsCallbackMB' ],
2739 return ucwords( strtolower( $str ) );
2744 * capitalize words at word breaks
2746 * @param string $str
2749 function ucwordbreaks( $str ) {
2750 if ( $this->isMultibyte( $str ) ) {
2751 $str = $this->lc( $str );
2753 // since \b doesn't work for UTF-8, we explicitely define word break chars
2754 $breaks = "[ \-\(\)\}\{\.,\?!]";
2756 // find first letter after word break
2757 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2758 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2760 return preg_replace_callback(
2762 [ $this, 'ucwordbreaksCallbackMB' ],
2766 return preg_replace_callback(
2767 '/\b([\w\x80-\xff]+)\b/',
2768 [ $this, 'ucwordbreaksCallbackAscii' ],
2775 * Return a case-folded representation of $s
2777 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2778 * and $s2 are the same except for the case of their characters. It is not
2779 * necessary for the value returned to make sense when displayed.
2781 * Do *not* perform any other normalisation in this function. If a caller
2782 * uses this function when it should be using a more general normalisation
2783 * function, then fix the caller.
2789 function caseFold( $s ) {
2790 return $this->uc( $s );
2796 * @throws MWException
2798 function checkTitleEncoding( $s ) {
2799 if ( is_array( $s ) ) {
2800 throw new MWException( 'Given array to checkTitleEncoding.' );
2802 if ( StringUtils
::isUtf8( $s ) ) {
2806 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2812 function fallback8bitEncoding() {
2813 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2817 * Most writing systems use whitespace to break up words.
2818 * Some languages such as Chinese don't conventionally do this,
2819 * which requires special handling when breaking up words for
2824 function hasWordBreaks() {
2829 * Some languages such as Chinese require word segmentation,
2830 * Specify such segmentation when overridden in derived class.
2832 * @param string $string
2835 function segmentByWord( $string ) {
2840 * Some languages have special punctuation need to be normalized.
2841 * Make such changes here.
2843 * @param string $string
2846 function normalizeForSearch( $string ) {
2847 return self
::convertDoubleWidth( $string );
2851 * convert double-width roman characters to single-width.
2852 * range: ff00-ff5f ~= 0020-007f
2854 * @param string $string
2858 protected static function convertDoubleWidth( $string ) {
2859 static $full = null;
2860 static $half = null;
2862 if ( $full === null ) {
2863 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2864 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2865 $full = str_split( $fullWidth, 3 );
2866 $half = str_split( $halfWidth );
2869 $string = str_replace( $full, $half, $string );
2874 * @param string $string
2875 * @param string $pattern
2878 protected static function insertSpace( $string, $pattern ) {
2879 $string = preg_replace( $pattern, " $1 ", $string );
2880 $string = preg_replace( '/ +/', ' ', $string );
2885 * @param array $termsArray
2888 function convertForSearchResult( $termsArray ) {
2889 # some languages, e.g. Chinese, need to do a conversion
2890 # in order for search results to be displayed correctly
2895 * Get the first character of a string.
2900 function firstChar( $s ) {
2903 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2904 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2909 if ( isset( $matches[1] ) ) {
2910 if ( strlen( $matches[1] ) != 3 ) {
2914 // Break down Hangul syllables to grab the first jamo
2915 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
2916 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2918 } elseif ( $code < 0xb098 ) {
2920 } elseif ( $code < 0xb2e4 ) {
2922 } elseif ( $code < 0xb77c ) {
2924 } elseif ( $code < 0xb9c8 ) {
2926 } elseif ( $code < 0xbc14 ) {
2928 } elseif ( $code < 0xc0ac ) {
2930 } elseif ( $code < 0xc544 ) {
2932 } elseif ( $code < 0xc790 ) {
2934 } elseif ( $code < 0xcc28 ) {
2936 } elseif ( $code < 0xce74 ) {
2938 } elseif ( $code < 0xd0c0 ) {
2940 } elseif ( $code < 0xd30c ) {
2942 } elseif ( $code < 0xd558 ) {
2953 * @deprecated No-op since 1.28
2955 function initEncoding() {
2956 wfDeprecated( __METHOD__
, '1.28' );
2963 * @deprecated No-op since 1.28
2965 function recodeForEdit( $s ) {
2966 wfDeprecated( __METHOD__
, '1.28' );
2973 * @deprecated No-op since 1.28
2975 function recodeInput( $s ) {
2976 wfDeprecated( __METHOD__
, '1.28' );
2981 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2982 * also cleans up certain backwards-compatible sequences, converting them
2983 * to the modern Unicode equivalent.
2985 * This is language-specific for performance reasons only.
2991 function normalize( $s ) {
2992 global $wgAllUnicodeFixes;
2993 $s = UtfNormal\Validator
::cleanUp( $s );
2994 if ( $wgAllUnicodeFixes ) {
2995 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
2996 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3003 * Transform a string using serialized data stored in the given file (which
3004 * must be in the serialized subdirectory of $IP). The file contains pairs
3005 * mapping source characters to destination characters.
3007 * The data is cached in process memory. This will go faster if you have the
3008 * FastStringSearch extension.
3010 * @param string $file
3011 * @param string $string
3013 * @throws MWException
3016 protected function transformUsingPairFile( $file, $string ) {
3017 if ( !isset( $this->transformData
[$file] ) ) {
3019 $data = require "$IP/languages/data/{$file}";
3020 $this->transformData
[$file] = new ReplacementArray( $data );
3022 return $this->transformData
[$file]->replace( $string );
3026 * For right-to-left language support
3031 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3035 * Return the correct HTML 'dir' attribute value for this language.
3039 return $this->isRTL() ?
'rtl' : 'ltr';
3043 * Return 'left' or 'right' as appropriate alignment for line-start
3044 * for this language's text direction.
3046 * Should be equivalent to CSS3 'start' text-align value....
3050 function alignStart() {
3051 return $this->isRTL() ?
'right' : 'left';
3055 * Return 'right' or 'left' as appropriate alignment for line-end
3056 * for this language's text direction.
3058 * Should be equivalent to CSS3 'end' text-align value....
3062 function alignEnd() {
3063 return $this->isRTL() ?
'left' : 'right';
3067 * A hidden direction mark (LRM or RLM), depending on the language direction.
3068 * Unlike getDirMark(), this function returns the character as an HTML entity.
3069 * This function should be used when the output is guaranteed to be HTML,
3070 * because it makes the output HTML source code more readable. When
3071 * the output is plain text or can be escaped, getDirMark() should be used.
3073 * @param bool $opposite Get the direction mark opposite to your language
3077 function getDirMarkEntity( $opposite = false ) {
3079 return $this->isRTL() ?
'‎' : '‏';
3081 return $this->isRTL() ?
'‏' : '‎';
3085 * A hidden direction mark (LRM or RLM), depending on the language direction.
3086 * This function produces them as invisible Unicode characters and
3087 * the output may be hard to read and debug, so it should only be used
3088 * when the output is plain text or can be escaped. When the output is
3089 * HTML, use getDirMarkEntity() instead.
3091 * @param bool $opposite Get the direction mark opposite to your language
3094 function getDirMark( $opposite = false ) {
3095 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3096 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3098 return $this->isRTL() ?
$lrm : $rlm;
3100 return $this->isRTL() ?
$rlm : $lrm;
3106 function capitalizeAllNouns() {
3107 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3111 * An arrow, depending on the language direction.
3113 * @param string $direction The direction of the arrow: forwards (default),
3114 * backwards, left, right, up, down.
3117 function getArrow( $direction = 'forwards' ) {
3118 switch ( $direction ) {
3120 return $this->isRTL() ?
'←' : '→';
3122 return $this->isRTL() ?
'→' : '←';
3135 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3139 function linkPrefixExtension() {
3140 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3144 * Get all magic words from cache.
3147 function getMagicWords() {
3148 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3152 * Run the LanguageGetMagic hook once.
3154 protected function doMagicHook() {
3155 if ( $this->mMagicHookDone
) {
3158 $this->mMagicHookDone
= true;
3159 Hooks
::run( 'LanguageGetMagic', [ &$this->mMagicExtensions
, $this->getCode() ], '1.16' );
3163 * Fill a MagicWord object with data from here
3165 * @param MagicWord $mw
3167 function getMagic( $mw ) {
3168 // Saves a function call
3169 if ( !$this->mMagicHookDone
) {
3170 $this->doMagicHook();
3173 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3174 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3176 $rawEntry = self
::$dataCache->getSubitem(
3177 $this->mCode
, 'magicWords', $mw->mId
);
3180 if ( !is_array( $rawEntry ) ) {
3181 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3183 $mw->mCaseSensitive
= $rawEntry[0];
3184 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3189 * Add magic words to the extension array
3191 * @param array $newWords
3193 function addMagicWordsByLang( $newWords ) {
3194 $fallbackChain = $this->getFallbackLanguages();
3195 $fallbackChain = array_reverse( $fallbackChain );
3196 foreach ( $fallbackChain as $code ) {
3197 if ( isset( $newWords[$code] ) ) {
3198 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3204 * Get special page names, as an associative array
3205 * canonical name => array of valid names, including aliases
3208 function getSpecialPageAliases() {
3209 // Cache aliases because it may be slow to load them
3210 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3212 $this->mExtendedSpecialPageAliases
=
3213 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3214 Hooks
::run( 'LanguageGetSpecialPageAliases',
3215 [ &$this->mExtendedSpecialPageAliases
, $this->getCode() ], '1.16' );
3218 return $this->mExtendedSpecialPageAliases
;
3222 * Italic is unsuitable for some languages
3224 * @param string $text The text to be emphasized.
3227 function emphasize( $text ) {
3228 return "<em>$text</em>";
3232 * Normally we output all numbers in plain en_US style, that is
3233 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3234 * point twohundredthirtyfive. However this is not suitable for all
3235 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3236 * Icelandic just want to use commas instead of dots, and dots instead
3237 * of commas like "293.291,235".
3239 * An example of this function being called:
3241 * wfMessage( 'message' )->numParams( $num )->text()
3244 * See $separatorTransformTable on MessageIs.php for
3245 * the , => . and . => , implementation.
3247 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3248 * @param int|float $number The string to be formatted, should be an integer
3249 * or a floating point number.
3250 * @param bool $nocommafy Set to true for special numbers like dates
3253 public function formatNum( $number, $nocommafy = false ) {
3254 global $wgTranslateNumerals;
3255 if ( !$nocommafy ) {
3256 $number = $this->commafy( $number );
3257 $s = $this->separatorTransformTable();
3259 $number = strtr( $number, $s );
3263 if ( $wgTranslateNumerals ) {
3264 $s = $this->digitTransformTable();
3266 $number = strtr( $number, $s );
3270 return (string)$number;
3274 * Front-end for non-commafied formatNum
3276 * @param int|float $number The string to be formatted, should be an integer
3277 * or a floating point number.
3281 public function formatNumNoSeparators( $number ) {
3282 return $this->formatNum( $number, true );
3286 * @param string $number
3289 public function parseFormattedNumber( $number ) {
3290 $s = $this->digitTransformTable();
3292 // eliminate empty array values such as ''. (T66347)
3293 $s = array_filter( $s );
3294 $number = strtr( $number, array_flip( $s ) );
3297 $s = $this->separatorTransformTable();
3299 // eliminate empty array values such as ''. (T66347)
3300 $s = array_filter( $s );
3301 $number = strtr( $number, array_flip( $s ) );
3304 $number = strtr( $number, [ ',' => '' ] );
3309 * Adds commas to a given number
3311 * @param mixed $number
3314 function commafy( $number ) {
3315 $digitGroupingPattern = $this->digitGroupingPattern();
3316 $minimumGroupingDigits = $this->minimumGroupingDigits();
3317 if ( $number === null ) {
3321 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3322 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3323 // In some languages it's conventional not to insert a thousands separator
3324 // in numbers that are four digits long (1000-9999).
3325 if ( $minimumGroupingDigits ) {
3326 // Number of '#' characters after last comma in the grouping pattern.
3327 // The pattern is hardcoded here, but this would vary for different patterns.
3328 $primaryGroupingSize = 3;
3329 // Maximum length of a number to suppress digit grouping for.
3330 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3331 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3335 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3337 // Ref: http://cldr.unicode.org/translation/number-patterns
3339 if ( intval( $number ) < 0 ) {
3340 // For negative numbers apply the algorithm like positive number and add sign.
3342 $number = substr( $number, 1 );
3346 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3347 preg_match( "/\d+/", $number, $integerPart );
3348 preg_match( "/\.\d*/", $number, $decimalPart );
3349 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3350 if ( $groupedNumber === $number ) {
3351 // the string does not have any number part. Eg: .12345
3352 return $sign . $groupedNumber;
3354 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3355 while ( $start > 0 ) {
3356 $match = $matches[0][$numMatches - 1];
3357 $matchLen = strlen( $match );
3358 $start = $end - $matchLen;
3362 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3364 if ( $numMatches > 1 ) {
3365 // use the last pattern for the rest of the number
3369 $groupedNumber = "," . $groupedNumber;
3372 return $sign . $groupedNumber;
3379 function digitGroupingPattern() {
3380 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3386 function digitTransformTable() {
3387 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3393 function separatorTransformTable() {
3394 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3400 function minimumGroupingDigits() {
3401 return self
::$dataCache->getItem( $this->mCode
, 'minimumGroupingDigits' );
3405 * Take a list of strings and build a locale-friendly comma-separated
3406 * list, using the local comma-separator message.
3407 * The last two strings are chained with an "and".
3408 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3410 * @param string[] $l
3413 function listToText( array $l ) {
3414 $m = count( $l ) - 1;
3419 $and = $this->msg( 'and' )->escaped();
3420 $space = $this->msg( 'word-separator' )->escaped();
3422 $comma = $this->msg( 'comma-separator' )->escaped();
3426 for ( $i = $m - 1; $i >= 0; $i-- ) {
3427 if ( $i == $m - 1 ) {
3428 $s = $l[$i] . $and . $space . $s;
3430 $s = $l[$i] . $comma . $s;
3437 * Take a list of strings and build a locale-friendly comma-separated
3438 * list, using the local comma-separator message.
3439 * @param string[] $list Array of strings to put in a comma list
3442 function commaList( array $list ) {
3444 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3450 * Take a list of strings and build a locale-friendly semicolon-separated
3451 * list, using the local semicolon-separator message.
3452 * @param string[] $list Array of strings to put in a semicolon list
3455 function semicolonList( array $list ) {
3457 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3463 * Same as commaList, but separate it with the pipe instead.
3464 * @param string[] $list Array of strings to put in a pipe list
3467 function pipeList( array $list ) {
3469 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3475 * This method is deprecated since 1.31 and kept as alias for truncateForDatabase, which
3476 * has replaced it. This method provides truncation suitable for DB.
3478 * The database offers limited byte lengths for some columns in the database;
3479 * multi-byte character sets mean we need to ensure that only whole characters
3480 * are included, otherwise broken characters can be passed to the user.
3482 * @deprecated since 1.31, use truncateForDatabase or truncateForVisual as appropriate.
3484 * @param string $string String to truncate
3485 * @param int $length Maximum length (including ellipsis)
3486 * @param string $ellipsis String to append to the truncated text
3487 * @param bool $adjustLength Subtract length of ellipsis from $length.
3488 * $adjustLength was introduced in 1.18, before that behaved as if false.
3491 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3492 return $this->truncateForDatabase( $string, $length, $ellipsis, $adjustLength );
3496 * Truncate a string to a specified length in bytes, appending an optional
3497 * string (e.g. for ellipsis)
3499 * If $length is negative, the string will be truncated from the beginning
3503 * @param string $string String to truncate
3504 * @param int $length Maximum length in bytes
3505 * @param string $ellipsis String to append to the end of truncated text
3506 * @param bool $adjustLength Subtract length of ellipsis from $length
3510 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3511 return $this->truncateInternal(
3512 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3517 * Truncate a string to a specified number of characters, appending an optional
3518 * string (e.g. for ellipsis).
3520 * This provides multibyte version of truncate() method of this class, suitable for truncation
3521 * based on number of characters, instead of number of bytes.
3523 * If $length is negative, the string will be truncated from the beginning.
3527 * @param string $string String to truncate
3528 * @param int $length Maximum number of characters
3529 * @param string $ellipsis String to append to the end of truncated text
3530 * @param bool $adjustLength Subtract length of ellipsis from $length
3534 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3535 // Passing encoding to mb_strlen and mb_substr is optional.
3536 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3537 // explicit specification of encoding is skipped.
3538 // Note: Both multibyte methods are callables invoked in truncateInternal.
3539 return $this->truncateInternal(
3540 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3545 * Internal method used for truncation. This method abstracts text truncation into
3546 * one common method, allowing users to provide length measurement function and
3547 * function for finding substring.
3549 * For usages, see truncateForDatabase and truncateForVisual.
3551 * @param string $string String to truncate
3552 * @param int $length Maximum length of final text
3553 * @param string $ellipsis String to append to the end of truncated text
3554 * @param bool $adjustLength Subtract length of ellipsis from $length
3555 * @param callable $measureLength Callable function used for determining the length of text
3556 * @param callable $getSubstring Callable function used for getting the substrings
3560 private function truncateInternal(
3561 $string, $length, $ellipsis, $adjustLength, $measureLength, $getSubstring
3563 if ( !is_callable( $measureLength ) ||
!is_callable( $getSubstring ) ) {
3564 throw new InvalidArgumentException( 'Invalid callback provided' );
3567 # Check if there is no need to truncate
3568 if ( $measureLength( $string ) <= abs( $length ) ) {
3569 return $string; // no need to truncate
3572 # Use the localized ellipsis character
3573 if ( $ellipsis == '...' ) {
3574 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3576 if ( $length == 0 ) {
3577 return $ellipsis; // convention
3580 $stringOriginal = $string;
3581 # If ellipsis length is >= $length then we can't apply $adjustLength
3582 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3583 $string = $ellipsis; // this can be slightly unexpected
3584 # Otherwise, truncate and add ellipsis...
3586 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3587 if ( $length > 0 ) {
3588 $length -= $ellipsisLength;
3589 $string = $getSubstring( $string, 0, $length ); // xyz...
3590 $string = $this->removeBadCharLast( $string );
3591 $string = rtrim( $string );
3592 $string = $string . $ellipsis;
3594 $length +
= $ellipsisLength;
3595 $string = $getSubstring( $string, $length ); // ...xyz
3596 $string = $this->removeBadCharFirst( $string );
3597 $string = ltrim( $string );
3598 $string = $ellipsis . $string;
3602 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3603 # This check is *not* redundant if $adjustLength, due to the single case where
3604 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3605 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3608 return $stringOriginal;
3613 * Remove bytes that represent an incomplete Unicode character
3614 * at the end of string (e.g. bytes of the char are missing)
3616 * @param string $string
3619 protected function removeBadCharLast( $string ) {
3620 if ( $string != '' ) {
3621 $char = ord( $string[strlen( $string ) - 1] );
3623 if ( $char >= 0xc0 ) {
3624 # We got the first byte only of a multibyte char; remove it.
3625 $string = substr( $string, 0, -1 );
3626 } elseif ( $char >= 0x80 &&
3627 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3628 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3629 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3631 # We chopped in the middle of a character; remove it
3639 * Remove bytes that represent an incomplete Unicode character
3640 * at the start of string (e.g. bytes of the char are missing)
3642 * @param string $string
3645 protected function removeBadCharFirst( $string ) {
3646 if ( $string != '' ) {
3647 $char = ord( $string[0] );
3648 if ( $char >= 0x80 && $char < 0xc0 ) {
3649 # We chopped in the middle of a character; remove the whole thing
3650 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3657 * Truncate a string of valid HTML to a specified length in bytes,
3658 * appending an optional string (e.g. for ellipses), and return valid HTML
3660 * This is only intended for styled/linked text, such as HTML with
3661 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3662 * Also, this will not detect things like "display:none" CSS.
3664 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3666 * @param string $text HTML string to truncate
3667 * @param int $length (zero/positive) Maximum length (including ellipses)
3668 * @param string $ellipsis String to append to the truncated text
3671 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3672 # Use the localized ellipsis character
3673 if ( $ellipsis == '...' ) {
3674 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3676 # Check if there is clearly no need to truncate
3677 if ( $length <= 0 ) {
3678 return $ellipsis; // no text shown, nothing to format (convention)
3679 } elseif ( strlen( $text ) <= $length ) {
3680 return $text; // string short enough even *with* HTML (short-circuit)
3683 $dispLen = 0; // innerHTML legth so far
3684 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3685 $tagType = 0; // 0-open, 1-close
3686 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3687 $entityState = 0; // 0-not entity, 1-entity
3688 $tag = $ret = ''; // accumulated tag name, accumulated result string
3689 $openTags = []; // open tag stack
3690 $maybeState = null; // possible truncation state
3692 $textLen = strlen( $text );
3693 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3694 for ( $pos = 0; true; ++
$pos ) {
3695 # Consider truncation once the display length has reached the maximim.
3696 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3697 # Check that we're not in the middle of a bracket/entity...
3698 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3699 if ( !$testingEllipsis ) {
3700 $testingEllipsis = true;
3701 # Save where we are; we will truncate here unless there turn out to
3702 # be so few remaining characters that truncation is not necessary.
3703 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3704 $maybeState = [ $ret, $openTags ]; // save state
3706 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3707 # String in fact does need truncation, the truncation point was OK.
3708 list( $ret, $openTags ) = $maybeState; // reload state
3709 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3710 $ret .= $ellipsis; // add ellipsis
3714 if ( $pos >= $textLen ) {
3715 break; // extra iteration just for above checks
3718 # Read the next char...
3720 $lastCh = $pos ?
$text[$pos - 1] : '';
3721 $ret .= $ch; // add to result string
3723 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3724 $entityState = 0; // for bad HTML
3725 $bracketState = 1; // tag started (checking for backslash)
3726 } elseif ( $ch == '>' ) {
3727 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3728 $entityState = 0; // for bad HTML
3729 $bracketState = 0; // out of brackets
3730 } elseif ( $bracketState == 1 ) {
3732 $tagType = 1; // close tag (e.g. "</span>")
3734 $tagType = 0; // open tag (e.g. "<span>")
3737 $bracketState = 2; // building tag name
3738 } elseif ( $bracketState == 2 ) {
3742 // Name found (e.g. "<a href=..."), add on tag attributes...
3743 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3745 } elseif ( $bracketState == 0 ) {
3746 if ( $entityState ) {
3749 $dispLen++
; // entity is one displayed char
3752 if ( $neLength == 0 && !$maybeState ) {
3753 // Save state without $ch. We want to *hit* the first
3754 // display char (to get tags) but not *use* it if truncating.
3755 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3758 $entityState = 1; // entity found, (e.g. " ")
3760 $dispLen++
; // this char is displayed
3761 // Add the next $max display text chars after this in one swoop...
3762 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3763 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3764 $dispLen +
= $skipped;
3770 // Close the last tag if left unclosed by bad HTML
3771 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3772 while ( count( $openTags ) > 0 ) {
3773 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3779 * truncateHtml() helper function
3780 * like strcspn() but adds the skipped chars to $ret
3782 * @param string $ret
3783 * @param string $text
3784 * @param string $search
3786 * @param null|int $len
3789 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3790 if ( $len === null ) {
3791 $len = -1; // -1 means "no limit" for strcspn
3792 } elseif ( $len < 0 ) {
3796 if ( $start < strlen( $text ) ) {
3797 $skipCount = strcspn( $text, $search, $start, $len );
3798 $ret .= substr( $text, $start, $skipCount );
3804 * truncateHtml() helper function
3805 * (a) push or pop $tag from $openTags as needed
3806 * (b) clear $tag value
3807 * @param string &$tag Current HTML tag name we are looking at
3808 * @param int $tagType (0-open tag, 1-close tag)
3809 * @param string $lastCh Character before the '>' that ended this tag
3810 * @param array &$openTags Open tag stack (not accounting for $tag)
3812 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3813 $tag = ltrim( $tag );
3815 if ( $tagType == 0 && $lastCh != '/' ) {
3816 $openTags[] = $tag; // tag opened (didn't close itself)
3817 } elseif ( $tagType == 1 ) {
3818 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3819 array_pop( $openTags ); // tag closed
3827 * Grammatical transformations, needed for inflected languages
3828 * Invoked by putting {{grammar:case|word}} in a message
3830 * @param string $word
3831 * @param string $case
3834 function convertGrammar( $word, $case ) {
3835 global $wgGrammarForms;
3836 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3837 return $wgGrammarForms[$this->getCode()][$case][$word];
3840 $grammarTransformations = $this->getGrammarTransformations();
3842 if ( isset( $grammarTransformations[$case] ) ) {
3843 $forms = $grammarTransformations[$case];
3845 // Some names of grammar rules are aliases for other rules.
3846 // In such cases the value is a string rather than object,
3847 // so load the actual rules.
3848 if ( is_string( $forms ) ) {
3849 $forms = $grammarTransformations[$forms];
3852 foreach ( array_values( $forms ) as $rule ) {
3855 if ( $form === '@metadata' ) {
3859 $replacement = $rule[1];
3861 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3862 $patternMatches = preg_match( $regex, $word );
3864 if ( $patternMatches === false ) {
3866 'An error occurred while processing grammar. ' .
3867 "Word: '$word'. Regex: /$form/."
3869 } elseif ( $patternMatches === 1 ) {
3870 $word = preg_replace( $regex, $replacement, $word );
3881 * Get the grammar forms for the content language
3882 * @return array Array of grammar forms
3885 function getGrammarForms() {
3886 global $wgGrammarForms;
3887 if ( isset( $wgGrammarForms[$this->getCode()] )
3888 && is_array( $wgGrammarForms[$this->getCode()] )
3890 return $wgGrammarForms[$this->getCode()];
3897 * Get the grammar transformations data for the language.
3898 * Used like grammar forms, with {{GRAMMAR}} and cases,
3899 * but uses pairs of regexes and replacements instead of code.
3901 * @return array[] Array of grammar transformations.
3902 * @throws MWException
3905 public function getGrammarTransformations() {
3906 $languageCode = $this->getCode();
3908 if ( self
::$grammarTransformations === null ) {
3909 self
::$grammarTransformations = new MapCacheLRU( 10 );
3912 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3913 return self
::$grammarTransformations->get( $languageCode );
3918 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3919 if ( is_readable( $grammarDataFile ) ) {
3920 $data = FormatJson
::decode(
3921 file_get_contents( $grammarDataFile ),
3925 if ( $data === null ) {
3926 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3929 self
::$grammarTransformations->set( $languageCode, $data );
3936 * Provides an alternative text depending on specified gender.
3937 * Usage {{gender:username|masculine|feminine|unknown}}.
3938 * username is optional, in which case the gender of current user is used,
3939 * but only in (some) interface messages; otherwise default gender is used.
3941 * If no forms are given, an empty string is returned. If only one form is
3942 * given, it will be returned unconditionally. These details are implied by
3943 * the caller and cannot be overridden in subclasses.
3945 * If three forms are given, the default is to use the third (unknown) form.
3946 * If fewer than three forms are given, the default is to use the first (masculine) form.
3947 * These details can be overridden in subclasses.
3949 * @param string $gender
3950 * @param array $forms
3954 function gender( $gender, $forms ) {
3955 if ( !count( $forms ) ) {
3958 $forms = $this->preConvertPlural( $forms, 2 );
3959 if ( $gender === 'male' ) {
3962 if ( $gender === 'female' ) {
3965 return $forms[2] ??
$forms[0];
3969 * Plural form transformations, needed for some languages.
3970 * For example, there are 3 form of plural in Russian and Polish,
3971 * depending on "count mod 10". See [[w:Plural]]
3972 * For English it is pretty simple.
3974 * Invoked by putting {{plural:count|wordform1|wordform2}}
3975 * or {{plural:count|wordform1|wordform2|wordform3}}
3977 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3979 * @param int $count Non-localized number
3980 * @param array $forms Different plural forms
3981 * @return string Correct form of plural for $count in this language
3983 function convertPlural( $count, $forms ) {
3984 // Handle explicit n=pluralform cases
3985 $forms = $this->handleExplicitPluralForms( $count, $forms );
3986 if ( is_string( $forms ) ) {
3989 if ( !count( $forms ) ) {
3993 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3994 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3995 return $forms[$pluralForm];
3999 * Handles explicit plural forms for Language::convertPlural()
4001 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4002 * If an explicitly defined plural form matches the $count, then
4003 * string value returned, otherwise array returned for further consideration
4004 * by CLDR rules or overridden convertPlural().
4008 * @param int $count Non-localized number
4009 * @param array $forms Different plural forms
4011 * @return array|string
4013 protected function handleExplicitPluralForms( $count, array $forms ) {
4014 foreach ( $forms as $index => $form ) {
4015 if ( preg_match( '/\d+=/i', $form ) ) {
4016 $pos = strpos( $form, '=' );
4017 if ( substr( $form, 0, $pos ) === (string)$count ) {
4018 return substr( $form, $pos +
1 );
4020 unset( $forms[$index] );
4023 return array_values( $forms );
4027 * Checks that convertPlural was given an array and pads it to requested
4028 * amount of forms by copying the last one.
4030 * @param array $forms Array of forms given to convertPlural
4031 * @param int $count How many forms should there be at least
4032 * @return array Padded array of forms or an exception if not an array
4034 protected function preConvertPlural( /* Array */ $forms, $count ) {
4035 while ( count( $forms ) < $count ) {
4036 $forms[] = $forms[count( $forms ) - 1];
4042 * Wraps argument with unicode control characters for directionality safety
4044 * This solves the problem where directionality-neutral characters at the edge of
4045 * the argument string get interpreted with the wrong directionality from the
4046 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4048 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4049 * directionality of the argument string, using the BIDI algorithm's own "First
4050 * strong directional codepoint" rule. Essentially, this works round the fact that
4051 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4052 * direction inference). The latter is cleaner but still not widely supported.
4054 * @param string $text Text to wrap
4055 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4057 public function embedBidi( $text = '' ) {
4058 $dir = self
::strongDirFromContent( $text );
4059 if ( $dir === 'ltr' ) {
4060 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4061 return self
::$lre . $text . self
::$pdf;
4063 if ( $dir === 'rtl' ) {
4064 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4065 return self
::$rle . $text . self
::$pdf;
4067 // No strong directionality: do not wrap
4072 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4073 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4074 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4075 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4078 * @param string $str The validated block duration in English
4079 * @param User|null $user User object to use timezone from or null for $wgUser
4080 * @param int $now Current timestamp, for formatting relative block durations
4081 * @return string Somehow translated block duration
4082 * @see LanguageFi.php for example implementation
4084 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
4085 $duration = SpecialBlock
::getSuggestedDurations( $this );
4086 foreach ( $duration as $show => $value ) {
4087 if ( strcmp( $str, $value ) == 0 ) {
4088 return htmlspecialchars( trim( $show ) );
4092 if ( wfIsInfinity( $str ) ) {
4093 foreach ( $duration as $show => $value ) {
4094 if ( wfIsInfinity( $value ) ) {
4095 return htmlspecialchars( trim( $show ) );
4100 // If all else fails, return a standard duration or timestamp description.
4101 $time = strtotime( $str, $now );
4102 if ( $time === false ) { // Unknown format. Return it as-is in case.
4104 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
4105 // The result differs based on current time, so the difference
4106 // is a fixed duration length.
4107 return $this->formatDuration( $time - $now );
4108 } else { // It's an absolute timestamp.
4109 if ( $time === 0 ) {
4110 // wfTimestamp() handles 0 as current time instead of epoch.
4111 $time = '19700101000000';
4114 return $this->userTimeAndDate( $time, $user );
4116 return $this->timeanddate( $time );
4121 * languages like Chinese need to be segmented in order for the diff
4124 * @param string $text
4127 public function segmentForDiff( $text ) {
4132 * and unsegment to show the result
4134 * @param string $text
4137 public function unsegmentForDiff( $text ) {
4142 * Return the LanguageConverter used in the Language
4145 * @return LanguageConverter
4147 public function getConverter() {
4148 return $this->mConverter
;
4152 * convert text to a variant
4154 * @param string $text text to convert
4155 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4156 * variant (if logged in), or the project default variant
4157 * @return string the converted string
4159 public function autoConvert( $text, $variant = false ) {
4160 return $this->mConverter
->autoConvert( $text, $variant );
4164 * convert text to all supported variants
4166 * @param string $text
4169 public function autoConvertToAllVariants( $text ) {
4170 return $this->mConverter
->autoConvertToAllVariants( $text );
4174 * convert text to different variants of a language.
4176 * @param string $text
4179 public function convert( $text ) {
4180 return $this->mConverter
->convert( $text );
4184 * Convert a Title object to a string in the preferred variant
4186 * @param Title $title
4189 public function convertTitle( $title ) {
4190 return $this->mConverter
->convertTitle( $title );
4194 * Convert a namespace index to a string in the preferred variant
4196 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4197 * @param string|null $variant variant to convert to, or null to use the user's preferred
4198 * variant (if logged in), or the project default variant
4199 * @return string a string representation of the namespace
4201 public function convertNamespace( $ns, $variant = null ) {
4202 return $this->mConverter
->convertNamespace( $ns, $variant );
4206 * Check if this is a language with variants
4210 public function hasVariants() {
4211 return count( $this->getVariants() ) > 1;
4215 * Check if the language has the specific variant
4218 * @param string $variant
4221 public function hasVariant( $variant ) {
4222 return (bool)$this->mConverter
->validateVariant( $variant );
4226 * Perform output conversion on a string, and encode for safe HTML output.
4227 * @param string $text Text to be converted
4228 * @param bool $isTitle Whether this conversion is for the article title
4230 * @todo this should get integrated somewhere sane
4232 public function convertHtml( $text, $isTitle = false ) {
4233 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4237 * @param string $key
4240 public function convertCategoryKey( $key ) {
4241 return $this->mConverter
->convertCategoryKey( $key );
4245 * Get the list of variants supported by this language
4246 * see sample implementation in LanguageZh.php
4248 * @return string[] An array of language codes
4250 public function getVariants() {
4251 return $this->mConverter
->getVariants();
4257 public function getPreferredVariant() {
4258 return $this->mConverter
->getPreferredVariant();
4264 public function getDefaultVariant() {
4265 return $this->mConverter
->getDefaultVariant();
4271 public function getURLVariant() {
4272 return $this->mConverter
->getURLVariant();
4276 * If a language supports multiple variants, it is
4277 * possible that non-existing link in one variant
4278 * actually exists in another variant. this function
4279 * tries to find it. See e.g. LanguageZh.php
4280 * The input parameters may be modified upon return
4282 * @param string &$link The name of the link
4283 * @param Title &$nt The title object of the link
4284 * @param bool $ignoreOtherCond To disable other conditions when
4285 * we need to transclude a template or update a category's link
4287 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4288 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4292 * returns language specific options used by User::getPageRenderHash()
4293 * for example, the preferred language variant
4297 function getExtraHashOptions() {
4298 return $this->mConverter
->getExtraHashOptions();
4302 * For languages that support multiple variants, the title of an
4303 * article may be displayed differently in different variants. this
4304 * function returns the apporiate title defined in the body of the article.
4308 public function getParsedTitle() {
4309 return $this->mConverter
->getParsedTitle();
4313 * Refresh the cache of conversion tables when
4314 * MediaWiki:Conversiontable* is updated.
4316 * @param Title $title The Title of the page being updated
4318 public function updateConversionTable( Title
$title ) {
4319 $this->mConverter
->updateConversionTable( $title );
4323 * Prepare external link text for conversion. When the text is
4324 * a URL, it shouldn't be converted, and it'll be wrapped in
4325 * the "raw" tag (-{R| }-) to prevent conversion.
4327 * This function is called "markNoConversion" for historical
4328 * reasons *BUT DIFFERS SIGNIFICANTLY* from
4329 * LanguageConverter::markNoConversion(), with which it is easily
4332 * @param string $text Text to be used for external link
4333 * @param bool $noParse Wrap it without confirming it's a real URL first
4334 * @return string The tagged text
4335 * @deprecated since 1.32, use LanguageConverter::markNoConversion()
4338 public function markNoConversion( $text, $noParse = false ) {
4339 wfDeprecated( __METHOD__
, '1.32' );
4340 // Excluding protocal-relative URLs may avoid many false positives.
4341 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4342 return $this->mConverter
->markNoConversion( $text );
4349 * A regular expression to match legal word-trailing characters
4350 * which should be merged onto a link of the form [[foo]]bar.
4354 public function linkTrail() {
4355 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4359 * A regular expression character set to match legal word-prefixing
4360 * characters which should be merged onto a link of the form foo[[bar]].
4364 public function linkPrefixCharset() {
4365 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4369 * Get the "parent" language which has a converter to convert a "compatible" language
4370 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4372 * @return Language|null
4375 public function getParentLanguage() {
4376 if ( $this->mParentLanguage
!== false ) {
4377 return $this->mParentLanguage
;
4380 $code = explode( '-', $this->getCode() )[0];
4381 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4382 $this->mParentLanguage
= null;
4385 $lang = self
::factory( $code );
4386 if ( !$lang->hasVariant( $this->getCode() ) ) {
4387 $this->mParentLanguage
= null;
4391 $this->mParentLanguage
= $lang;
4396 * Compare with an other language object
4399 * @param Language $lang
4402 public function equals( Language
$lang ) {
4403 return $lang === $this ||
$lang->getCode() === $this->mCode
;
4407 * Get the internal language code for this language object
4409 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4410 * htmlspecialchars() or similar
4414 public function getCode() {
4415 return $this->mCode
;
4419 * Get the code in BCP 47 format which we can use
4420 * inside of html lang="" tags.
4422 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4423 * htmlspecialchars() or similar.
4428 public function getHtmlCode() {
4429 if ( is_null( $this->mHtmlCode
) ) {
4430 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4432 return $this->mHtmlCode
;
4436 * @param string $code
4438 public function setCode( $code ) {
4439 $this->mCode
= $code;
4440 // Ensure we don't leave incorrect cached data lying around
4441 $this->mHtmlCode
= null;
4442 $this->mParentLanguage
= false;
4446 * Get the language code from a file name. Inverse of getFileName()
4447 * @param string $filename $prefix . $languageCode . $suffix
4448 * @param string $prefix Prefix before the language code
4449 * @param string $suffix Suffix after the language code
4450 * @return string Language code, or false if $prefix or $suffix isn't found
4452 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4454 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4455 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4456 if ( !count( $m ) ) {
4459 return str_replace( '_', '-', strtolower( $m[1] ) );
4463 * @param string $code
4464 * @param bool $fallback Whether we're going through language fallback chain
4465 * @return string Name of the language class
4467 public static function classFromCode( $code, $fallback = true ) {
4468 if ( $fallback && $code == 'en' ) {
4471 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4476 * Get the name of a file for a certain language code
4477 * @param string $prefix Prepend this to the filename
4478 * @param string $code Language code
4479 * @param string $suffix Append this to the filename
4480 * @throws MWException
4481 * @return string $prefix . $mangledCode . $suffix
4483 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4484 if ( !self
::isValidBuiltInCode( $code ) ) {
4485 throw new MWException( "Invalid language code \"$code\"" );
4488 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4492 * @param string $code
4495 public static function getMessagesFileName( $code ) {
4497 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4498 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4503 * @param string $code
4505 * @throws MWException
4508 public static function getJsonMessagesFileName( $code ) {
4511 if ( !self
::isValidBuiltInCode( $code ) ) {
4512 throw new MWException( "Invalid language code \"$code\"" );
4515 return "$IP/languages/i18n/$code.json";
4519 * Get the first fallback for a given language.
4521 * @param string $code
4523 * @return bool|string
4525 public static function getFallbackFor( $code ) {
4526 $fallbacks = self
::getFallbacksFor( $code );
4528 return $fallbacks[0];
4534 * Get the ordered list of fallback languages.
4537 * @param string $code Language code
4538 * @return array Non-empty array, ending in "en"
4540 public static function getFallbacksFor( $code ) {
4541 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4544 // For unknown languages, fallbackSequence returns an empty array,
4545 // hardcode fallback to 'en' in that case.
4546 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4550 * Get the ordered list of fallback languages, ending with the fallback
4551 * language chain for the site language.
4554 * @param string $code Language code
4555 * @return array Array( fallbacks, site fallbacks )
4557 public static function getFallbacksIncludingSiteLanguage( $code ) {
4558 global $wgLanguageCode;
4560 // Usually, we will only store a tiny number of fallback chains, so we
4561 // keep them in static memory.
4562 $cacheKey = "{$code}-{$wgLanguageCode}";
4564 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4565 $fallbacks = self
::getFallbacksFor( $code );
4567 // Append the site's fallback chain, including the site language itself
4568 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4569 array_unshift( $siteFallbacks, $wgLanguageCode );
4571 // Eliminate any languages already included in the chain
4572 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4574 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4576 return self
::$fallbackLanguageCache[$cacheKey];
4580 * Get all messages for a given language
4581 * WARNING: this may take a long time. If you just need all message *keys*
4582 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4584 * @param string $code
4588 public static function getMessagesFor( $code ) {
4589 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4593 * Get a message for a given language
4595 * @param string $key
4596 * @param string $code
4600 public static function getMessageFor( $key, $code ) {
4601 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4605 * Get all message keys for a given language. This is a faster alternative to
4606 * array_keys( Language::getMessagesFor( $code ) )
4609 * @param string $code Language code
4610 * @return array Array of message keys (strings)
4612 public static function getMessageKeysFor( $code ) {
4613 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4617 * @param string $talk
4620 function fixVariableInNamespace( $talk ) {
4621 if ( strpos( $talk, '$1' ) === false ) {
4625 global $wgMetaNamespace;
4626 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4628 # Allow grammar transformations
4629 # Allowing full message-style parsing would make simple requests
4630 # such as action=raw much more expensive than they need to be.
4631 # This will hopefully cover most cases.
4632 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4633 [ $this, 'replaceGrammarInNamespace' ], $talk );
4634 return str_replace( ' ', '_', $talk );
4641 function replaceGrammarInNamespace( $m ) {
4642 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4646 * Decode an expiry (block, protection, etc) which has come from the DB
4648 * @param string $expiry Database expiry String
4649 * @param bool|int $format True to process using language functions, or TS_ constant
4650 * to return the expiry in a given timestamp
4651 * @param string $infinity If $format is not true, use this string for infinite expiry
4655 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4657 if ( $dbInfinity === null ) {
4658 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4661 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4662 return $format === true
4663 ?
$this->getMessageFromDB( 'infiniteblock' )
4666 return $format === true
4667 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4668 : wfTimestamp( $format, $expiry );
4673 * Formats a time given in seconds into a string representation of that time.
4675 * @param int|float $seconds
4676 * @param array $format An optional argument that formats the returned string in different ways:
4677 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4678 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4679 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4681 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4682 * or 'avoidminutes'.
4685 function formatTimePeriod( $seconds, $format = [] ) {
4686 if ( !is_array( $format ) ) {
4687 $format = [ 'avoid' => $format ]; // For backwards compatibility
4689 if ( !isset( $format['avoid'] ) ) {
4690 $format['avoid'] = false;
4692 if ( !isset( $format['noabbrevs'] ) ) {
4693 $format['noabbrevs'] = false;
4695 $secondsMsg = wfMessage(
4696 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4697 $minutesMsg = wfMessage(
4698 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4699 $hoursMsg = wfMessage(
4700 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4701 $daysMsg = wfMessage(
4702 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4704 if ( round( $seconds * 10 ) < 100 ) {
4705 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4706 $s = $secondsMsg->params( $s )->text();
4707 } elseif ( round( $seconds ) < 60 ) {
4708 $s = $this->formatNum( round( $seconds ) );
4709 $s = $secondsMsg->params( $s )->text();
4710 } elseif ( round( $seconds ) < 3600 ) {
4711 $minutes = floor( $seconds / 60 );
4712 $secondsPart = round( fmod( $seconds, 60 ) );
4713 if ( $secondsPart == 60 ) {
4717 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4719 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4720 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4721 $hours = floor( $seconds / 3600 );
4722 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4723 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4724 if ( $secondsPart == 60 ) {
4728 if ( $minutes == 60 ) {
4732 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4734 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4735 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4736 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4739 $days = floor( $seconds / 86400 );
4740 if ( $format['avoid'] === 'avoidminutes' ) {
4741 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4742 if ( $hours == 24 ) {
4746 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4748 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4749 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4750 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4751 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4752 if ( $minutes == 60 ) {
4756 if ( $hours == 24 ) {
4760 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4762 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4764 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4766 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4768 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4775 * Format a bitrate for output, using an appropriate
4776 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4777 * the magnitude in question.
4779 * This use base 1000. For base 1024 use formatSize(), for another base
4780 * see formatComputingNumbers().
4785 function formatBitrate( $bps ) {
4786 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4790 * @param int $size Size of the unit
4791 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4792 * @param string $messageKey Message key to be uesd
4795 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4797 return str_replace( '$1', $this->formatNum( $size ),
4798 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4801 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4804 $maxIndex = count( $sizes ) - 1;
4805 while ( $size >= $boundary && $index < $maxIndex ) {
4810 // For small sizes no decimal places necessary
4813 // For MB and bigger two decimal places are smarter
4816 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4818 $size = round( $size, $round );
4819 $text = $this->getMessageFromDB( $msg );
4820 return str_replace( '$1', $this->formatNum( $size ), $text );
4824 * Format a size in bytes for output, using an appropriate
4825 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4827 * This method use base 1024. For base 1000 use formatBitrate(), for
4828 * another base see formatComputingNumbers()
4830 * @param int $size Size to format
4831 * @return string Plain text (not HTML)
4833 function formatSize( $size ) {
4834 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4838 * Make a list item, used by various special pages
4840 * @param string $page Page link
4841 * @param string $details HTML safe text between brackets
4842 * @param bool $oppositedm Add the direction mark opposite to your
4843 * language, to display text properly
4844 * @return string HTML escaped
4846 function specialList( $page, $details, $oppositedm = true ) {
4851 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4854 $this->msg( 'word-separator' )->escaped() .
4855 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4859 * Generate (prev x| next x) (20|50|100...) type links for paging
4861 * @param Title $title Title object to link
4862 * @param int $offset
4864 * @param array $query Optional URL query parameter string
4865 * @param bool $atend Optional param for specified if this is the last page
4868 public function viewPrevNext( Title
$title, $offset, $limit,
4869 array $query = [], $atend = false
4871 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4873 # Make 'previous' link
4874 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4875 if ( $offset > 0 ) {
4876 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4877 $query, $prev, 'prevn-title', 'mw-prevlink' );
4879 $plink = htmlspecialchars( $prev );
4883 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4885 $nlink = htmlspecialchars( $next );
4887 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4888 $query, $next, 'nextn-title', 'mw-nextlink' );
4891 # Make links to set number of items per page
4893 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4894 $numLinks[] = $this->numLink( $title, $offset, $num,
4895 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4898 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4899 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4903 * Helper function for viewPrevNext() that generates links
4905 * @param Title $title Title object to link
4906 * @param int $offset
4908 * @param array $query Extra query parameters
4909 * @param string $link Text to use for the link; will be escaped
4910 * @param string $tooltipMsg Name of the message to use as tooltip
4911 * @param string $class Value of the "class" attribute of the link
4912 * @return string HTML fragment
4914 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4917 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4918 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4919 ->numParams( $limit )->text();
4921 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4922 'title' => $tooltip, 'class' => $class ], $link );
4926 * Get the conversion rule title, if any.
4930 public function getConvRuleTitle() {
4931 return $this->mConverter
->getConvRuleTitle();
4935 * Get the compiled plural rules for the language
4937 * @return array Associative array with plural form, and plural rule as key-value pairs
4939 public function getCompiledPluralRules() {
4940 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4941 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4942 if ( !$pluralRules ) {
4943 foreach ( $fallbacks as $fallbackCode ) {
4944 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4945 if ( $pluralRules ) {
4950 return $pluralRules;
4954 * Get the plural rules for the language
4956 * @return array Associative array with plural form number and plural rule as key-value pairs
4958 public function getPluralRules() {
4959 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4960 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4961 if ( !$pluralRules ) {
4962 foreach ( $fallbacks as $fallbackCode ) {
4963 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4964 if ( $pluralRules ) {
4969 return $pluralRules;
4973 * Get the plural rule types for the language
4975 * @return array Associative array with plural form number and plural rule type as key-value pairs
4977 public function getPluralRuleTypes() {
4978 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4979 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4980 if ( !$pluralRuleTypes ) {
4981 foreach ( $fallbacks as $fallbackCode ) {
4982 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4983 if ( $pluralRuleTypes ) {
4988 return $pluralRuleTypes;
4992 * Find the index number of the plural rule appropriate for the given number
4993 * @param int $number
4994 * @return int The index number of the plural rule
4996 public function getPluralRuleIndexNumber( $number ) {
4997 $pluralRules = $this->getCompiledPluralRules();
4998 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
5003 * Find the plural rule type appropriate for the given number
5004 * For example, if the language is set to Arabic, getPluralType(5) should
5007 * @param int $number
5008 * @return string The name of the plural rule type, e.g. one, two, few, many
5010 public function getPluralRuleType( $number ) {
5011 $index = $this->getPluralRuleIndexNumber( $number );
5012 $pluralRuleTypes = $this->getPluralRuleTypes();
5013 if ( isset( $pluralRuleTypes[$index] ) ) {
5014 return $pluralRuleTypes[$index];