3 * Internationalisation code.
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
25 * @defgroup Language Language
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
37 use CLDRPluralRuleParser\Evaluator
;
40 * Internationalisation code
45 * @var LanguageConverter
49 public $mVariants, $mCode, $mLoaded = false;
50 public $mMagicExtensions = [], $mMagicHookDone = false;
51 private $mHtmlCode = null, $mParentLanguage = false;
53 public $dateFormatStrings = [];
54 public $mExtendedSpecialPageAliases;
56 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
59 * ReplacementArray object caches
61 public $transformData = [];
64 * @var LocalisationCache
66 static public $dataCache;
68 static public $mLangObjCache = [];
70 static public $mWeekdayMsgs = [
71 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
75 static public $mWeekdayAbbrevMsgs = [
76 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
79 static public $mMonthMsgs = [
80 'january', 'february', 'march', 'april', 'may_long', 'june',
81 'july', 'august', 'september', 'october', 'november',
84 static public $mMonthGenMsgs = [
85 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
86 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
89 static public $mMonthAbbrevMsgs = [
90 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
91 'sep', 'oct', 'nov', 'dec'
94 static public $mIranianCalendarMonthMsgs = [
95 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
96 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
97 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
98 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
101 static public $mHebrewCalendarMonthMsgs = [
102 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
103 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
104 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
105 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
106 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
109 static public $mHebrewCalendarMonthGenMsgs = [
110 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
111 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
112 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
113 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
114 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
117 static public $mHijriCalendarMonthMsgs = [
118 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
119 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
120 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
121 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
128 static public $durationIntervals = [
129 'millennia' => 31556952000,
130 'centuries' => 3155695200,
131 'decades' => 315569520,
132 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
141 * Cache for language fallbacks.
142 * @see Language::getFallbacksIncludingSiteLanguage
146 static private $fallbackLanguageCache = [];
149 * Cache for language names
150 * @var HashBagOStuff|null
152 static private $languageNameCache;
155 * Unicode directional formatting characters, for embedBidi()
157 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
158 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
159 static private $pdf = "\xE2\x80\xAC"; // 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://git.wikimedia.org/summary/unicodejs.git .
172 // @codingStandardsIgnoreStart
173 // @codeCoverageIgnoreStart
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
176 // @codingStandardsIgnoreEnd
179 * Get a cached or new language object for a given language code
180 * @param string $code
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 $langObj = isset( self
::$mLangObjCache[$code] )
192 ? self
::$mLangObjCache[$code]
193 : self
::newFromCode( $code );
195 // merge the language object in to get it up front in the cache
196 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
197 // get rid of the oldest ones in case we have an overflow
198 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
204 * Create a language object for a given language code
205 * @param string $code
206 * @throws MWException
209 protected static function newFromCode( $code ) {
210 if ( !Language
::isValidCode( $code ) ) {
211 throw new MWException( "Invalid language code \"$code\"" );
214 if ( !Language
::isValidBuiltInCode( $code ) ) {
215 // It's not possible to customise this code with class files, so
216 // just return a Language object. This is to support uselang= hacks.
217 $lang = new Language
;
218 $lang->setCode( $code );
222 // Check if there is a language class for the code
223 $class = self
::classFromCode( $code );
224 if ( class_exists( $class ) ) {
229 // Keep trying the fallback list until we find an existing class
230 $fallbacks = Language
::getFallbacksFor( $code );
231 foreach ( $fallbacks as $fallbackCode ) {
232 if ( !Language
::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 * http://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 # http://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 37564, 37587, 36938
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 ) {
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
370 $type = gettype( $code );
371 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
378 * Returns true if a language code is an IETF tag known to MediaWiki.
385 public static function isKnownLanguageTag( $tag ) {
386 // Quick escape for invalid input to avoid exceptions down the line
387 // when code tries to process tags which are not valid at all.
388 if ( !self
::isValidBuiltInCode( $tag ) ) {
392 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
393 || self
::fetchLanguageName( $tag, $tag ) !== ''
402 * Get the LocalisationCache instance
404 * @return LocalisationCache
406 public static function getLocalisationCache() {
407 if ( is_null( self
::$dataCache ) ) {
408 global $wgLocalisationCacheConf;
409 $class = $wgLocalisationCacheConf['class'];
410 self
::$dataCache = new $class( $wgLocalisationCacheConf );
412 return self
::$dataCache;
415 function __construct() {
416 $this->mConverter
= new FakeConverter( $this );
417 // Set the code to the name of the descendant
418 if ( get_class( $this ) == 'Language' ) {
421 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
423 self
::getLocalisationCache();
427 * Reduce memory usage
429 function __destruct() {
430 foreach ( $this as $name => $value ) {
431 unset( $this->$name );
436 * Hook which will be called if this is the content language.
437 * Descendants can use this to register hook functions or modify globals
439 function initContLang() {
446 function getFallbackLanguages() {
447 return self
::getFallbacksFor( $this->mCode
);
451 * Exports $wgBookstoreListEn
454 function getBookstoreList() {
455 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
459 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
460 * available in localised form, it will be included in English.
464 public function getNamespaces() {
465 if ( is_null( $this->namespaceNames
) ) {
466 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
468 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
469 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
471 $this->namespaceNames
= $wgExtraNamespaces +
$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 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 function getNsText( $index ) {
543 $ns = $this->getNamespaces();
544 return isset( $ns[$index] ) ?
$ns[$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 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 function getGenderNsText( $index, $gender ) {
574 global $wgExtraGenderNamespaces;
576 $ns = $wgExtraGenderNamespaces +
577 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
579 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
583 * Whether this language uses gender-dependent namespace aliases.
584 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
588 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 isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
621 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 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 function getNsIndex( $text ) {
696 $lctext = $this->lc( $text );
697 $ns = MWNamespace
::getCanonicalIndex( $lctext );
698 if ( $ns !== null ) {
701 $ids = $this->getNamespaceIds();
702 return isset( $ids[$lctext] ) ?
$ids[$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 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 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
728 * SpecialPageFactory::getLocalNameFor instead.
729 * @param string $name
732 function specialPage( $name ) {
733 $aliases = $this->getSpecialPageAliases();
734 if ( isset( $aliases[$name][0] ) ) {
735 $name = $aliases[$name][0];
737 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
743 function getDatePreferences() {
744 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
750 function getDateFormats() {
751 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
755 * @return array|string
757 function getDefaultDateFormat() {
758 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
759 if ( $df === 'dmy or mdy' ) {
760 global $wgAmericanDates;
761 return $wgAmericanDates ?
'mdy' : 'dmy';
770 function getDatePreferenceMigrationMap() {
771 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
775 * @param string $image
778 function getImageFile( $image ) {
779 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
786 function getImageFiles() {
787 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
793 function getExtraUserToggles() {
794 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
801 function getUserToggle( $tog ) {
802 return $this->getMessageFromDB( "tog-$tog" );
806 * Get an array of language names, indexed by code.
807 * @param null|string $inLanguage Code of language in which to return the names
808 * Use null for autonyms (native names)
809 * @param string $include One of:
810 * 'all' all available languages
811 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
812 * 'mwfile' only if the language is in 'mw' *and* has a message file
813 * @return array Language code => language name
816 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
817 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
818 $cacheKey .= ":$include";
819 if ( self
::$languageNameCache === null ) {
820 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
823 $ret = self
::$languageNameCache->get( $cacheKey );
825 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
826 self
::$languageNameCache->set( $cacheKey, $ret );
832 * Uncached helper for fetchLanguageNames
833 * @param null|string $inLanguage Code of language in which to return the names
834 * Use null for autonyms (native names)
835 * @param string $include One of:
836 * 'all' all available languages
837 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
838 * 'mwfile' only if the language is in 'mw' *and* has a message file
839 * @return array Language code => language name
841 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
842 global $wgExtraLanguageNames;
844 // If passed an invalid language code to use, fallback to en
845 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
852 # TODO: also include when $inLanguage is null, when this code is more efficient
853 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
856 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
857 foreach ( $mwNames as $mwCode => $mwName ) {
858 # - Prefer own MediaWiki native name when not using the hook
859 # - For other names just add if not added through the hook
860 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
861 $names[$mwCode] = $mwName;
865 if ( $include === 'all' ) {
871 $coreCodes = array_keys( $mwNames );
872 foreach ( $coreCodes as $coreCode ) {
873 $returnMw[$coreCode] = $names[$coreCode];
876 if ( $include === 'mwfile' ) {
878 # We do this using a foreach over the codes instead of a directory
879 # loop so that messages files in extensions will work correctly.
880 foreach ( $returnMw as $code => $value ) {
881 if ( is_readable( self
::getMessagesFileName( $code ) )
882 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
884 $namesMwFile[$code] = $names[$code];
888 ksort( $namesMwFile );
893 # 'mw' option; default if it's not one of the other two options (all/mwfile)
898 * @param string $code The code of the language for which to get the name
899 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
900 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
901 * @return string Language name or empty
904 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
905 $code = strtolower( $code );
906 $array = self
::fetchLanguageNames( $inLanguage, $include );
907 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
911 * Get a message from the MediaWiki namespace.
913 * @param string $msg Message name
916 function getMessageFromDB( $msg ) {
917 return $this->msg( $msg )->text();
921 * Get message object in this language. Only for use inside this class.
923 * @param string $msg Message name
926 protected function msg( $msg ) {
927 return wfMessage( $msg )->inLanguage( $this );
934 function getMonthName( $key ) {
935 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
941 function getMonthNamesArray() {
942 $monthNames = [ '' ];
943 for ( $i = 1; $i < 13; $i++
) {
944 $monthNames[] = $this->getMonthName( $i );
953 function getMonthNameGen( $key ) {
954 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
961 function getMonthAbbreviation( $key ) {
962 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
968 function getMonthAbbreviationsArray() {
969 $monthNames = [ '' ];
970 for ( $i = 1; $i < 13; $i++
) {
971 $monthNames[] = $this->getMonthAbbreviation( $i );
980 function getWeekdayName( $key ) {
981 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
988 function getWeekdayAbbreviation( $key ) {
989 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
996 function getIranianCalendarMonthName( $key ) {
997 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1001 * @param string $key
1004 function getHebrewCalendarMonthName( $key ) {
1005 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1009 * @param string $key
1012 function getHebrewCalendarMonthNameGen( $key ) {
1013 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1017 * @param string $key
1020 function getHijriCalendarMonthName( $key ) {
1021 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1025 * Pass through result from $dateTimeObj->format()
1026 * @param DateTime|bool|null &$dateTimeObj
1028 * @param DateTimeZone|bool|null $zone
1029 * @param string $code
1032 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1033 if ( !$dateTimeObj ) {
1034 $dateTimeObj = DateTime
::createFromFormat(
1035 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1038 return $dateTimeObj->format( $code );
1042 * This is a workalike of PHP's date() function, but with better
1043 * internationalisation, a reduced set of format characters, and a better
1046 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1047 * the PHP manual for definitions. There are a number of extensions, which
1050 * xn Do not translate digits of the next numeric format character
1051 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1052 * xr Use roman numerals for the next numeric format character
1053 * xh Use hebrew numerals for the next numeric format character
1055 * xg Genitive month name
1057 * xij j (day number) in Iranian calendar
1058 * xiF F (month name) in Iranian calendar
1059 * xin n (month number) in Iranian calendar
1060 * xiy y (two digit year) in Iranian calendar
1061 * xiY Y (full year) in Iranian calendar
1063 * xjj j (day number) in Hebrew calendar
1064 * xjF F (month name) in Hebrew calendar
1065 * xjt t (days in month) in Hebrew calendar
1066 * xjx xg (genitive month name) in Hebrew calendar
1067 * xjn n (month number) in Hebrew calendar
1068 * xjY Y (full year) in Hebrew calendar
1070 * xmj j (day number) in Hijri calendar
1071 * xmF F (month name) in Hijri calendar
1072 * xmn n (month number) in Hijri calendar
1073 * xmY Y (full year) in Hijri calendar
1075 * xkY Y (full year) in Thai solar calendar. Months and days are
1076 * identical to the Gregorian calendar
1077 * xoY Y (full year) in Minguo calendar or Juche year.
1078 * Months and days are identical to the
1079 * Gregorian calendar
1080 * xtY Y (full year) in Japanese nengo. Months and days are
1081 * identical to the Gregorian calendar
1083 * Characters enclosed in double quotes will be considered literal (with
1084 * the quotes themselves removed). Unmatched quotes will be considered
1085 * literal quotes. Example:
1087 * "The month is" F => The month is January
1090 * Backslash escaping is also supported.
1092 * Input timestamp is assumed to be pre-normalized to the desired local
1093 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1094 * $ts is UTC if $zone is not given.
1096 * @param string $format
1097 * @param string $ts 14-character timestamp
1100 * @param DateTimeZone $zone Timezone of $ts
1101 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1102 * Only makes sense if $ts is the current time.
1103 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1105 * @throws MWException
1108 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1113 $dateTimeObj = false;
1122 $usedSecond = false;
1123 $usedMinute = false;
1130 $usedISOYear = false;
1131 $usedIsLeapYear = false;
1133 $usedHebrewMonth = false;
1134 $usedIranianMonth = false;
1135 $usedHijriMonth = false;
1136 $usedHebrewYear = false;
1137 $usedIranianYear = false;
1138 $usedHijriYear = false;
1139 $usedTennoYear = false;
1141 if ( strlen( $ts ) !== 14 ) {
1142 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1145 if ( !ctype_digit( $ts ) ) {
1146 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1149 $formatLength = strlen( $format );
1150 for ( $p = 0; $p < $formatLength; $p++
) {
1152 $code = $format[$p];
1153 if ( $code == 'x' && $p < $formatLength - 1 ) {
1154 $code .= $format[++
$p];
1157 if ( ( $code === 'xi'
1163 && $p < $formatLength - 1 ) {
1164 $code .= $format[++
$p];
1175 $rawToggle = !$rawToggle;
1185 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1188 $usedHebrewMonth = true;
1190 $hebrew = self
::tsToHebrew( $ts );
1192 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1196 $num = substr( $ts, 6, 2 );
1200 $s .= $this->getWeekdayAbbreviation(
1201 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1206 $num = intval( substr( $ts, 6, 2 ) );
1211 $iranian = self
::tsToIranian( $ts );
1218 $hijri = self
::tsToHijri( $ts );
1225 $hebrew = self
::tsToHebrew( $ts );
1231 $s .= $this->getWeekdayName(
1232 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1237 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1240 $usedIranianMonth = true;
1242 $iranian = self
::tsToIranian( $ts );
1244 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1247 $usedHijriMonth = true;
1249 $hijri = self
::tsToHijri( $ts );
1251 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1254 $usedHebrewMonth = true;
1256 $hebrew = self
::tsToHebrew( $ts );
1258 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1262 $num = substr( $ts, 4, 2 );
1266 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1270 $num = intval( substr( $ts, 4, 2 ) );
1273 $usedIranianMonth = true;
1275 $iranian = self
::tsToIranian( $ts );
1280 $usedHijriMonth = true;
1282 $hijri = self
::tsToHijri( $ts );
1287 $usedHebrewMonth = true;
1289 $hebrew = self
::tsToHebrew( $ts );
1294 $usedHebrewMonth = true;
1296 $hebrew = self
::tsToHebrew( $ts );
1302 $num = substr( $ts, 0, 4 );
1305 $usedIranianYear = true;
1307 $iranian = self
::tsToIranian( $ts );
1312 $usedHijriYear = true;
1314 $hijri = self
::tsToHijri( $ts );
1319 $usedHebrewYear = true;
1321 $hebrew = self
::tsToHebrew( $ts );
1328 $thai = self
::tsToYear( $ts, 'thai' );
1335 $minguo = self
::tsToYear( $ts, 'minguo' );
1340 $usedTennoYear = true;
1342 $tenno = self
::tsToYear( $ts, 'tenno' );
1348 $num = substr( $ts, 2, 2 );
1351 $usedIranianYear = true;
1353 $iranian = self
::tsToIranian( $ts );
1355 $num = substr( $iranian[0], -2 );
1359 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1363 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1367 $h = substr( $ts, 8, 2 );
1368 $num = $h %
12 ?
$h %
12 : 12;
1372 $num = intval( substr( $ts, 8, 2 ) );
1376 $h = substr( $ts, 8, 2 );
1377 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1381 $num = substr( $ts, 8, 2 );
1385 $num = substr( $ts, 10, 2 );
1389 $num = substr( $ts, 12, 2 );
1399 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1405 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1409 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1413 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1416 $usedIsLeapYear = true;
1417 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1420 $usedISOYear = true;
1421 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1428 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1431 # Backslash escaping
1432 if ( $p < $formatLength - 1 ) {
1433 $s .= $format[++
$p];
1440 if ( $p < $formatLength - 1 ) {
1441 $endQuote = strpos( $format, '"', $p +
1 );
1442 if ( $endQuote === false ) {
1443 # No terminating quote, assume literal "
1446 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1450 # Quote at end of string, assume literal "
1457 if ( $num !== false ) {
1458 if ( $rawToggle ||
$raw ) {
1461 } elseif ( $roman ) {
1462 $s .= Language
::romanNumeral( $num );
1464 } elseif ( $hebrewNum ) {
1465 $s .= self
::hebrewNumeral( $num );
1468 $s .= $this->formatNum( $num, true );
1473 if ( $usedSecond ) {
1475 } elseif ( $usedMinute ) {
1476 $ttl = 60 - substr( $ts, 12, 2 );
1477 } elseif ( $usedHour ) {
1478 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1479 } elseif ( $usedAMPM ) {
1480 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1481 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1485 $usedIranianMonth ||
1492 // @todo Someone who understands the non-Gregorian calendars
1493 // should write proper logic for them so that they don't need purged every day.
1494 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1495 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1498 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1499 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1502 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1503 $timeRemainingInDay;
1504 } elseif ( $usedISOYear ) {
1505 // December 28th falls on the last ISO week of the year, every year.
1506 // The last ISO week of a year can be 52 or 53.
1507 $lastWeekOfISOYear = DateTime
::createFromFormat(
1509 substr( $ts, 0, 4 ) . '1228',
1510 $zone ?
: new DateTimeZone( 'UTC' )
1512 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1513 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1514 $timeRemainingInWeek =
1515 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1516 +
$timeRemainingInDay;
1517 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1522 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1523 substr( $ts, 6, 2 ) ) * 86400
1524 +
$timeRemainingInDay;
1525 } elseif ( $usedYear ) {
1527 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1528 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1529 +
$timeRemainingInDay;
1530 } elseif ( $usedIsLeapYear ) {
1531 $year = substr( $ts, 0, 4 );
1532 $timeRemainingInYear =
1533 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1534 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1535 +
$timeRemainingInDay;
1537 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1538 // this isn't a leap year. see when the next one starts
1539 $nextCandidate = $year - $mod +
4;
1540 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1541 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1542 $timeRemainingInYear;
1544 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1545 $timeRemainingInYear;
1548 // this is a leap year, so the next year isn't
1549 $possibleTtls[] = $timeRemainingInYear;
1553 if ( $possibleTtls ) {
1554 $ttl = min( $possibleTtls );
1561 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1562 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1565 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1566 * Gregorian dates to Iranian dates. Originally written in C, it
1567 * is released under the terms of GNU Lesser General Public
1568 * License. Conversion to PHP was performed by Niklas Laxström.
1570 * Link: http://www.farsiweb.info/jalali/jalali.c
1576 private static function tsToIranian( $ts ) {
1577 $gy = substr( $ts, 0, 4 ) -1600;
1578 $gm = substr( $ts, 4, 2 ) -1;
1579 $gd = substr( $ts, 6, 2 ) -1;
1581 # Days passed from the beginning (including leap years)
1583 +
floor( ( $gy +
3 ) / 4 )
1584 - floor( ( $gy +
99 ) / 100 )
1585 +
floor( ( $gy +
399 ) / 400 );
1587 // Add days of the past months of this year
1588 for ( $i = 0; $i < $gm; $i++
) {
1589 $gDayNo +
= self
::$GREG_DAYS[$i];
1593 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1597 // Days passed in current month
1598 $gDayNo +
= (int)$gd;
1600 $jDayNo = $gDayNo - 79;
1602 $jNp = floor( $jDayNo / 12053 );
1605 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1608 if ( $jDayNo >= 366 ) {
1609 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1610 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1613 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1614 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1620 return [ $jy, $jm, $jd ];
1624 * Converting Gregorian dates to Hijri dates.
1626 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1628 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1634 private static function tsToHijri( $ts ) {
1635 $year = substr( $ts, 0, 4 );
1636 $month = substr( $ts, 4, 2 );
1637 $day = substr( $ts, 6, 2 );
1645 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1646 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1648 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1649 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1650 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1653 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1654 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1657 $zl = $zjd -1948440 +
10632;
1658 $zn = (int)( ( $zl - 1 ) / 10631 );
1659 $zl = $zl - 10631 * $zn +
354;
1660 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1661 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1662 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1663 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1664 $zm = (int)( ( 24 * $zl ) / 709 );
1665 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1666 $zy = 30 * $zn +
$zj - 30;
1668 return [ $zy, $zm, $zd ];
1672 * Converting Gregorian dates to Hebrew dates.
1674 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1675 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1676 * to translate the relevant functions into PHP and release them under
1679 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1680 * and Adar II is 14. In a non-leap year, Adar is 6.
1686 private static function tsToHebrew( $ts ) {
1688 $year = substr( $ts, 0, 4 );
1689 $month = substr( $ts, 4, 2 );
1690 $day = substr( $ts, 6, 2 );
1692 # Calculate Hebrew year
1693 $hebrewYear = $year +
3760;
1695 # Month number when September = 1, August = 12
1697 if ( $month > 12 ) {
1704 # Calculate day of year from 1 September
1706 for ( $i = 1; $i < $month; $i++
) {
1710 # Check if the year is leap
1711 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1714 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1721 # Calculate the start of the Hebrew year
1722 $start = self
::hebrewYearStart( $hebrewYear );
1724 # Calculate next year's start
1725 if ( $dayOfYear <= $start ) {
1726 # Day is before the start of the year - it is the previous year
1728 $nextStart = $start;
1732 # Add days since previous year's 1 September
1734 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1738 # Start of the new (previous) year
1739 $start = self
::hebrewYearStart( $hebrewYear );
1742 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1745 # Calculate Hebrew day of year
1746 $hebrewDayOfYear = $dayOfYear - $start;
1748 # Difference between year's days
1749 $diff = $nextStart - $start;
1750 # Add 12 (or 13 for leap years) days to ignore the difference between
1751 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1752 # difference is only about the year type
1753 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1759 # Check the year pattern, and is leap year
1760 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1761 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1762 # and non-leap years
1763 $yearPattern = $diff %
30;
1764 # Check if leap year
1765 $isLeap = $diff >= 30;
1767 # Calculate day in the month from number of day in the Hebrew year
1768 # Don't check Adar - if the day is not in Adar, we will stop before;
1769 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1770 $hebrewDay = $hebrewDayOfYear;
1773 while ( $hebrewMonth <= 12 ) {
1774 # Calculate days in this month
1775 if ( $isLeap && $hebrewMonth == 6 ) {
1776 # Adar in a leap year
1778 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1780 if ( $hebrewDay <= $days ) {
1784 # Subtract the days of Adar I
1785 $hebrewDay -= $days;
1788 if ( $hebrewDay <= $days ) {
1794 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1795 # Cheshvan in a complete year (otherwise as the rule below)
1797 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1798 # Kislev in an incomplete year (otherwise as the rule below)
1801 # Odd months have 30 days, even have 29
1802 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1804 if ( $hebrewDay <= $days ) {
1805 # In the current month
1808 # Subtract the days of the current month
1809 $hebrewDay -= $days;
1810 # Try in the next month
1815 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1819 * This calculates the Hebrew year start, as days since 1 September.
1820 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1821 * Used for Hebrew date.
1827 private static function hebrewYearStart( $year ) {
1828 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1829 $b = intval( ( $year - 1 ) %
4 );
1830 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1834 $Mar = intval( $m );
1840 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1841 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1843 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1845 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1849 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1854 * Algorithm to convert Gregorian dates to Thai solar dates,
1855 * Minguo dates or Minguo dates.
1857 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1858 * http://en.wikipedia.org/wiki/Minguo_calendar
1859 * http://en.wikipedia.org/wiki/Japanese_era_name
1861 * @param string $ts 14-character timestamp
1862 * @param string $cName Calender name
1863 * @return array Converted year, month, day
1865 private static function tsToYear( $ts, $cName ) {
1866 $gy = substr( $ts, 0, 4 );
1867 $gm = substr( $ts, 4, 2 );
1868 $gd = substr( $ts, 6, 2 );
1870 if ( !strcmp( $cName, 'thai' ) ) {
1872 # Add 543 years to the Gregorian calendar
1873 # Months and days are identical
1874 $gy_offset = $gy +
543;
1875 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1877 # Deduct 1911 years from the Gregorian calendar
1878 # Months and days are identical
1879 $gy_offset = $gy - 1911;
1880 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1881 # Nengō dates up to Meiji period
1882 # Deduct years from the Gregorian calendar
1883 # depending on the nengo periods
1884 # Months and days are identical
1886 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1887 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1890 $gy_gannen = $gy - 1868 +
1;
1891 $gy_offset = $gy_gannen;
1892 if ( $gy_gannen == 1 ) {
1895 $gy_offset = '明治' . $gy_offset;
1897 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1898 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1899 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1900 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1901 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1904 $gy_gannen = $gy - 1912 +
1;
1905 $gy_offset = $gy_gannen;
1906 if ( $gy_gannen == 1 ) {
1909 $gy_offset = '大正' . $gy_offset;
1911 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1912 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1913 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1916 $gy_gannen = $gy - 1926 +
1;
1917 $gy_offset = $gy_gannen;
1918 if ( $gy_gannen == 1 ) {
1921 $gy_offset = '昭和' . $gy_offset;
1924 $gy_gannen = $gy - 1989 +
1;
1925 $gy_offset = $gy_gannen;
1926 if ( $gy_gannen == 1 ) {
1929 $gy_offset = '平成' . $gy_offset;
1935 return [ $gy_offset, $gm, $gd ];
1939 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1941 * This is the rule the BIDI algorithm uses to determine the directionality of
1942 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1943 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1945 * TODO: Does not handle BIDI control characters inside the text.
1946 * TODO: Does not handle unallocated characters.
1948 * @param string $text Text to test
1949 * @return null|string Directionality ('ltr' or 'rtl') or null
1951 private static function strongDirFromContent( $text = '' ) {
1952 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
1955 if ( $matches[1] === '' ) {
1962 * Roman number formatting up to 10000
1968 static function romanNumeral( $num ) {
1970 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1971 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1972 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1973 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1974 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1977 $num = intval( $num );
1978 if ( $num > 10000 ||
$num <= 0 ) {
1983 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1984 if ( $num >= $pow10 ) {
1985 $s .= $table[$i][(int)floor( $num / $pow10 )];
1987 $num = $num %
$pow10;
1993 * Hebrew Gematria number formatting up to 9999
1999 static function hebrewNumeral( $num ) {
2001 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2002 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2015 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2018 $num = intval( $num );
2019 if ( $num > 9999 ||
$num <= 0 ) {
2023 // Round thousands have special notations
2024 if ( $num === 1000 ) {
2026 } elseif ( $num %
1000 === 0 ) {
2027 return $table[0][$num / 1000] . "' אלפים";
2032 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2033 if ( $num >= $pow10 ) {
2034 if ( $num === 15 ||
$num === 16 ) {
2035 $letters[] = $table[0][9];
2036 $letters[] = $table[0][$num - 9];
2039 $letters = array_merge(
2041 (array)$table[$i][intval( $num / $pow10 )]
2044 if ( $pow10 === 1000 ) {
2050 $num = $num %
$pow10;
2053 $preTransformLength = count( $letters );
2054 if ( $preTransformLength === 1 ) {
2055 // Add geresh (single quote) to one-letter numbers
2058 $lastIndex = $preTransformLength - 1;
2059 $letters[$lastIndex] = str_replace(
2060 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2061 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2062 $letters[$lastIndex]
2065 // Add gershayim (double quote) to multiple-letter numbers,
2066 // but exclude numbers with only one letter after the thousands
2067 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2068 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2071 array_splice( $letters, -1, 0, '"' );
2075 return implode( $letters );
2079 * Used by date() and time() to adjust the time output.
2081 * @param string $ts The time in date('YmdHis') format
2082 * @param mixed $tz Adjust the time by this amount (default false, mean we
2083 * get user timecorrection setting)
2086 function userAdjust( $ts, $tz = false ) {
2087 global $wgUser, $wgLocalTZoffset;
2089 if ( $tz === false ) {
2090 $tz = $wgUser->getOption( 'timecorrection' );
2093 $data = explode( '|', $tz, 3 );
2095 if ( $data[0] == 'ZoneInfo' ) {
2096 MediaWiki\
suppressWarnings();
2097 $userTZ = timezone_open( $data[2] );
2098 MediaWiki\restoreWarnings
();
2099 if ( $userTZ !== false ) {
2100 $date = date_create( $ts, timezone_open( 'UTC' ) );
2101 date_timezone_set( $date, $userTZ );
2102 $date = date_format( $date, 'YmdHis' );
2105 # Unrecognized timezone, default to 'Offset' with the stored offset.
2106 $data[0] = 'Offset';
2109 if ( $data[0] == 'System' ||
$tz == '' ) {
2110 # Global offset in minutes.
2111 $minDiff = $wgLocalTZoffset;
2112 } elseif ( $data[0] == 'Offset' ) {
2113 $minDiff = intval( $data[1] );
2115 $data = explode( ':', $tz );
2116 if ( count( $data ) == 2 ) {
2117 $data[0] = intval( $data[0] );
2118 $data[1] = intval( $data[1] );
2119 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2120 if ( $data[0] < 0 ) {
2121 $minDiff = -$minDiff;
2124 $minDiff = intval( $data[0] ) * 60;
2128 # No difference ? Return time unchanged
2129 if ( 0 == $minDiff ) {
2133 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2134 # Generate an adjusted date; take advantage of the fact that mktime
2135 # will normalize out-of-range values so we don't have to split $minDiff
2136 # into hours and minutes.
2138 (int)substr( $ts, 8, 2 ) ), # Hours
2139 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2140 (int)substr( $ts, 12, 2 ), # Seconds
2141 (int)substr( $ts, 4, 2 ), # Month
2142 (int)substr( $ts, 6, 2 ), # Day
2143 (int)substr( $ts, 0, 4 ) ); # Year
2145 $date = date( 'YmdHis', $t );
2146 MediaWiki\restoreWarnings
();
2152 * This is meant to be used by time(), date(), and timeanddate() to get
2153 * the date preference they're supposed to use, it should be used in
2157 * function timeanddate([...], $format = true) {
2158 * $datePreference = $this->dateFormat($format);
2163 * @param int|string|bool $usePrefs If true, the user's preference is used
2164 * if false, the site/language default is used
2165 * if int/string, assumed to be a format.
2168 function dateFormat( $usePrefs = true ) {
2171 if ( is_bool( $usePrefs ) ) {
2173 $datePreference = $wgUser->getDatePreference();
2175 $datePreference = (string)User
::getDefaultOption( 'date' );
2178 $datePreference = (string)$usePrefs;
2182 if ( $datePreference == '' ) {
2186 return $datePreference;
2190 * Get a format string for a given type and preference
2191 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2192 * @param string $pref The format name as it appears in Messages*.php under
2195 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2199 function getDateFormatString( $type, $pref ) {
2200 $wasDefault = false;
2201 if ( $pref == 'default' ) {
2203 $pref = $this->getDefaultDateFormat();
2206 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2207 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2209 if ( $type === 'pretty' && $df === null ) {
2210 $df = $this->getDateFormatString( 'date', $pref );
2213 if ( !$wasDefault && $df === null ) {
2214 $pref = $this->getDefaultDateFormat();
2215 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2218 $this->dateFormatStrings
[$type][$pref] = $df;
2220 return $this->dateFormatStrings
[$type][$pref];
2224 * @param string $ts The time format which needs to be turned into a
2225 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2226 * @param bool $adj Whether to adjust the time output according to the
2227 * user configured offset ($timecorrection)
2228 * @param mixed $format True to use user's date format preference
2229 * @param string|bool $timecorrection The time offset as returned by
2230 * validateTimeZone() in Special:Preferences
2233 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2234 $ts = wfTimestamp( TS_MW
, $ts );
2236 $ts = $this->userAdjust( $ts, $timecorrection );
2238 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2239 return $this->sprintfDate( $df, $ts );
2243 * @param string $ts The time format which needs to be turned into a
2244 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2245 * @param bool $adj Whether to adjust the time output according to the
2246 * user configured offset ($timecorrection)
2247 * @param mixed $format True to use user's date format preference
2248 * @param string|bool $timecorrection The time offset as returned by
2249 * validateTimeZone() in Special:Preferences
2252 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2253 $ts = wfTimestamp( TS_MW
, $ts );
2255 $ts = $this->userAdjust( $ts, $timecorrection );
2257 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2258 return $this->sprintfDate( $df, $ts );
2262 * @param string $ts The time format which needs to be turned into a
2263 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2264 * @param bool $adj Whether to adjust the time output according to the
2265 * user configured offset ($timecorrection)
2266 * @param mixed $format What format to return, if it's false output the
2267 * default one (default true)
2268 * @param string|bool $timecorrection The time offset as returned by
2269 * validateTimeZone() in Special:Preferences
2272 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2273 $ts = wfTimestamp( TS_MW
, $ts );
2275 $ts = $this->userAdjust( $ts, $timecorrection );
2277 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2278 return $this->sprintfDate( $df, $ts );
2282 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2286 * @param int $seconds The amount of seconds.
2287 * @param array $chosenIntervals The intervals to enable.
2291 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2292 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2296 foreach ( $intervals as $intervalName => $intervalValue ) {
2297 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2298 // duration-years, duration-decades, duration-centuries, duration-millennia
2299 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2300 $segments[] = $message->inLanguage( $this )->escaped();
2303 return $this->listToText( $segments );
2307 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2308 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2312 * @param int $seconds The amount of seconds.
2313 * @param array $chosenIntervals The intervals to enable.
2317 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2318 if ( empty( $chosenIntervals ) ) {
2319 $chosenIntervals = [
2331 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2332 $sortedNames = array_keys( $intervals );
2333 $smallestInterval = array_pop( $sortedNames );
2337 foreach ( $intervals as $name => $length ) {
2338 $value = floor( $seconds / $length );
2340 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2341 $seconds -= $value * $length;
2342 $segments[$name] = $value;
2350 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2352 * @param string $type Can be 'date', 'time' or 'both'
2353 * @param string $ts The time format which needs to be turned into a
2354 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2355 * @param User $user User object used to get preferences for timezone and format
2356 * @param array $options Array, can contain the following keys:
2357 * - 'timecorrection': time correction, can have the following values:
2358 * - true: use user's preference
2359 * - false: don't use time correction
2360 * - int: value of time correction in minutes
2361 * - 'format': format to use, can have the following values:
2362 * - true: use user's preference
2363 * - false: use default preference
2364 * - string: format to use
2368 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2369 $ts = wfTimestamp( TS_MW
, $ts );
2370 $options +
= [ 'timecorrection' => true, 'format' => true ];
2371 if ( $options['timecorrection'] !== false ) {
2372 if ( $options['timecorrection'] === true ) {
2373 $offset = $user->getOption( 'timecorrection' );
2375 $offset = $options['timecorrection'];
2377 $ts = $this->userAdjust( $ts, $offset );
2379 if ( $options['format'] === true ) {
2380 $format = $user->getDatePreference();
2382 $format = $options['format'];
2384 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2385 return $this->sprintfDate( $df, $ts );
2389 * Get the formatted date for the given timestamp and formatted for
2392 * @param mixed $ts Mixed: the time format which needs to be turned into a
2393 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2394 * @param User $user User object used to get preferences for timezone and format
2395 * @param array $options Array, can contain the following keys:
2396 * - 'timecorrection': time correction, can have the following values:
2397 * - true: use user's preference
2398 * - false: don't use time correction
2399 * - int: value of time correction in minutes
2400 * - 'format': format to use, can have the following values:
2401 * - true: use user's preference
2402 * - false: use default preference
2403 * - string: format to use
2407 public function userDate( $ts, User
$user, array $options = [] ) {
2408 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2412 * Get the formatted time for the given timestamp and formatted for
2415 * @param mixed $ts The time format which needs to be turned into a
2416 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2417 * @param User $user User object used to get preferences for timezone and format
2418 * @param array $options Array, can contain the following keys:
2419 * - 'timecorrection': time correction, can have the following values:
2420 * - true: use user's preference
2421 * - false: don't use time correction
2422 * - int: value of time correction in minutes
2423 * - 'format': format to use, can have the following values:
2424 * - true: use user's preference
2425 * - false: use default preference
2426 * - string: format to use
2430 public function userTime( $ts, User
$user, array $options = [] ) {
2431 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2435 * Get the formatted date and time for the given timestamp and formatted for
2438 * @param mixed $ts The time format which needs to be turned into a
2439 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2440 * @param User $user User object used to get preferences for timezone and format
2441 * @param array $options Array, can contain the following keys:
2442 * - 'timecorrection': time correction, can have the following values:
2443 * - true: use user's preference
2444 * - false: don't use time correction
2445 * - int: value of time correction in minutes
2446 * - 'format': format to use, can have the following values:
2447 * - true: use user's preference
2448 * - false: use default preference
2449 * - string: format to use
2453 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2454 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2458 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2460 * Determine the difference between the timestamp and the current time, and
2461 * generate a readable timestamp by returning "<N> <units> ago", where the
2462 * largest possible unit is used.
2464 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2466 * @param MWTimestamp $time
2467 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2468 * @param User|null $user User the timestamp is being generated for
2469 * (or null to use main context's user)
2470 * @return string Formatted timestamp
2472 public function getHumanTimestamp(
2473 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2475 if ( $relativeTo === null ) {
2476 $relativeTo = new MWTimestamp();
2478 if ( $user === null ) {
2479 $user = RequestContext
::getMain()->getUser();
2482 // Adjust for the user's timezone.
2483 $offsetThis = $time->offsetForUser( $user );
2484 $offsetRel = $relativeTo->offsetForUser( $user );
2487 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2488 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2491 // Reset the timezone on the objects.
2492 $time->timestamp
->sub( $offsetThis );
2493 $relativeTo->timestamp
->sub( $offsetRel );
2499 * Convert an MWTimestamp into a pretty human-readable timestamp using
2500 * the given user preferences and relative base time.
2502 * @see Language::getHumanTimestamp
2503 * @param MWTimestamp $ts Timestamp to prettify
2504 * @param MWTimestamp $relativeTo Base timestamp
2505 * @param User $user User preferences to use
2506 * @return string Human timestamp
2509 private function getHumanTimestampInternal(
2510 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2512 $diff = $ts->diff( $relativeTo );
2513 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2514 (int)$relativeTo->timestamp
->format( 'w' ) );
2515 $days = $diff->days ?
: (int)$diffDay;
2516 if ( $diff->invert ||
$days > 5
2517 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2519 // Timestamps are in different years: use full timestamp
2520 // Also do full timestamp for future dates
2522 * @todo FIXME: Add better handling of future timestamps.
2524 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2525 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2526 } elseif ( $days > 5 ) {
2527 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2528 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2529 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2530 } elseif ( $days > 1 ) {
2531 // Timestamp within the past week: show the day of the week and time
2532 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2533 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2535 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2536 $ts = wfMessage( "$weekday-at" )
2537 ->inLanguage( $this )
2538 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2540 } elseif ( $days == 1 ) {
2541 // Timestamp was yesterday: say 'yesterday' and the time.
2542 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2543 $ts = wfMessage( 'yesterday-at' )
2544 ->inLanguage( $this )
2545 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2547 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2548 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2549 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2550 $ts = wfMessage( 'today-at' )
2551 ->inLanguage( $this )
2552 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2555 // From here on in, the timestamp was soon enough ago so that we can simply say
2556 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2557 } elseif ( $diff->h
== 1 ) {
2558 // Less than 90 minutes, but more than an hour ago.
2559 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2560 } elseif ( $diff->i
>= 1 ) {
2561 // A few minutes ago.
2562 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2563 } elseif ( $diff->s
>= 30 ) {
2564 // Less than a minute, but more than 30 sec ago.
2565 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2567 // Less than 30 seconds ago.
2568 $ts = wfMessage( 'just-now' )->text();
2575 * @param string $key
2576 * @return array|null
2578 function getMessage( $key ) {
2579 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2585 function getAllMessages() {
2586 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2591 * @param string $out
2592 * @param string $string
2595 function iconv( $in, $out, $string ) {
2596 # This is a wrapper for iconv in all languages except esperanto,
2597 # which does some nasty x-conversions beforehand
2599 # Even with //IGNORE iconv can whine about illegal characters in
2600 # *input* string. We just ignore those too.
2601 # REF: http://bugs.php.net/bug.php?id=37166
2602 # REF: https://phabricator.wikimedia.org/T18885
2603 MediaWiki\
suppressWarnings();
2604 $text = iconv( $in, $out . '//IGNORE', $string );
2605 MediaWiki\restoreWarnings
();
2609 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2612 * @param array $matches
2613 * @return mixed|string
2615 function ucwordbreaksCallbackAscii( $matches ) {
2616 return $this->ucfirst( $matches[1] );
2620 * @param array $matches
2623 function ucwordbreaksCallbackMB( $matches ) {
2624 return mb_strtoupper( $matches[0] );
2628 * @param array $matches
2631 function ucCallback( $matches ) {
2632 list( $wikiUpperChars ) = self
::getCaseMaps();
2633 return strtr( $matches[1], $wikiUpperChars );
2637 * @param array $matches
2640 function lcCallback( $matches ) {
2641 list( , $wikiLowerChars ) = self
::getCaseMaps();
2642 return strtr( $matches[1], $wikiLowerChars );
2646 * @param array $matches
2649 function ucwordsCallbackMB( $matches ) {
2650 return mb_strtoupper( $matches[0] );
2654 * @param array $matches
2657 function ucwordsCallbackWiki( $matches ) {
2658 list( $wikiUpperChars ) = self
::getCaseMaps();
2659 return strtr( $matches[0], $wikiUpperChars );
2663 * Make a string's first character uppercase
2665 * @param string $str
2669 function ucfirst( $str ) {
2671 if ( $o < 96 ) { // if already uppercase...
2673 } elseif ( $o < 128 ) {
2674 return ucfirst( $str ); // use PHP's ucfirst()
2676 // fall back to more complex logic in case of multibyte strings
2677 return $this->uc( $str, true );
2682 * Convert a string to uppercase
2684 * @param string $str
2685 * @param bool $first
2689 function uc( $str, $first = false ) {
2690 if ( function_exists( 'mb_strtoupper' ) ) {
2692 if ( $this->isMultibyte( $str ) ) {
2693 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2695 return ucfirst( $str );
2698 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2701 if ( $this->isMultibyte( $str ) ) {
2702 $x = $first ?
'^' : '';
2703 return preg_replace_callback(
2704 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2705 [ $this, 'ucCallback' ],
2709 return $first ?
ucfirst( $str ) : strtoupper( $str );
2715 * @param string $str
2716 * @return mixed|string
2718 function lcfirst( $str ) {
2721 return strval( $str );
2722 } elseif ( $o >= 128 ) {
2723 return $this->lc( $str, true );
2724 } elseif ( $o > 96 ) {
2727 $str[0] = strtolower( $str[0] );
2733 * @param string $str
2734 * @param bool $first
2735 * @return mixed|string
2737 function lc( $str, $first = false ) {
2738 if ( function_exists( 'mb_strtolower' ) ) {
2740 if ( $this->isMultibyte( $str ) ) {
2741 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2743 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2746 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2749 if ( $this->isMultibyte( $str ) ) {
2750 $x = $first ?
'^' : '';
2751 return preg_replace_callback(
2752 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2753 [ $this, 'lcCallback' ],
2757 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2763 * @param string $str
2766 function isMultibyte( $str ) {
2767 return strlen( $str ) !== mb_strlen( $str );
2771 * @param string $str
2772 * @return mixed|string
2774 function ucwords( $str ) {
2775 if ( $this->isMultibyte( $str ) ) {
2776 $str = $this->lc( $str );
2778 // regexp to find first letter in each word (i.e. after each space)
2779 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2781 // function to use to capitalize a single char
2782 if ( function_exists( 'mb_strtoupper' ) ) {
2783 return preg_replace_callback(
2785 [ $this, 'ucwordsCallbackMB' ],
2789 return preg_replace_callback(
2791 [ $this, 'ucwordsCallbackWiki' ],
2796 return ucwords( strtolower( $str ) );
2801 * capitalize words at word breaks
2803 * @param string $str
2806 function ucwordbreaks( $str ) {
2807 if ( $this->isMultibyte( $str ) ) {
2808 $str = $this->lc( $str );
2810 // since \b doesn't work for UTF-8, we explicitely define word break chars
2811 $breaks = "[ \-\(\)\}\{\.,\?!]";
2813 // find first letter after word break
2814 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2815 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2817 if ( function_exists( 'mb_strtoupper' ) ) {
2818 return preg_replace_callback(
2820 [ $this, 'ucwordbreaksCallbackMB' ],
2824 return preg_replace_callback(
2826 [ $this, 'ucwordsCallbackWiki' ],
2831 return preg_replace_callback(
2832 '/\b([\w\x80-\xff]+)\b/',
2833 [ $this, 'ucwordbreaksCallbackAscii' ],
2840 * Return a case-folded representation of $s
2842 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2843 * and $s2 are the same except for the case of their characters. It is not
2844 * necessary for the value returned to make sense when displayed.
2846 * Do *not* perform any other normalisation in this function. If a caller
2847 * uses this function when it should be using a more general normalisation
2848 * function, then fix the caller.
2854 function caseFold( $s ) {
2855 return $this->uc( $s );
2862 function checkTitleEncoding( $s ) {
2863 if ( is_array( $s ) ) {
2864 throw new MWException( 'Given array to checkTitleEncoding.' );
2866 if ( StringUtils
::isUtf8( $s ) ) {
2870 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2876 function fallback8bitEncoding() {
2877 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2881 * Most writing systems use whitespace to break up words.
2882 * Some languages such as Chinese don't conventionally do this,
2883 * which requires special handling when breaking up words for
2888 function hasWordBreaks() {
2893 * Some languages such as Chinese require word segmentation,
2894 * Specify such segmentation when overridden in derived class.
2896 * @param string $string
2899 function segmentByWord( $string ) {
2904 * Some languages have special punctuation need to be normalized.
2905 * Make such changes here.
2907 * @param string $string
2910 function normalizeForSearch( $string ) {
2911 return self
::convertDoubleWidth( $string );
2915 * convert double-width roman characters to single-width.
2916 * range: ff00-ff5f ~= 0020-007f
2918 * @param string $string
2922 protected static function convertDoubleWidth( $string ) {
2923 static $full = null;
2924 static $half = null;
2926 if ( $full === null ) {
2927 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2928 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2929 $full = str_split( $fullWidth, 3 );
2930 $half = str_split( $halfWidth );
2933 $string = str_replace( $full, $half, $string );
2938 * @param string $string
2939 * @param string $pattern
2942 protected static function insertSpace( $string, $pattern ) {
2943 $string = preg_replace( $pattern, " $1 ", $string );
2944 $string = preg_replace( '/ +/', ' ', $string );
2949 * @param array $termsArray
2952 function convertForSearchResult( $termsArray ) {
2953 # some languages, e.g. Chinese, need to do a conversion
2954 # in order for search results to be displayed correctly
2959 * Get the first character of a string.
2964 function firstChar( $s ) {
2967 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2968 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2973 if ( isset( $matches[1] ) ) {
2974 if ( strlen( $matches[1] ) != 3 ) {
2978 // Break down Hangul syllables to grab the first jamo
2979 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
2980 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2982 } elseif ( $code < 0xb098 ) {
2983 return "\xe3\x84\xb1";
2984 } elseif ( $code < 0xb2e4 ) {
2985 return "\xe3\x84\xb4";
2986 } elseif ( $code < 0xb77c ) {
2987 return "\xe3\x84\xb7";
2988 } elseif ( $code < 0xb9c8 ) {
2989 return "\xe3\x84\xb9";
2990 } elseif ( $code < 0xbc14 ) {
2991 return "\xe3\x85\x81";
2992 } elseif ( $code < 0xc0ac ) {
2993 return "\xe3\x85\x82";
2994 } elseif ( $code < 0xc544 ) {
2995 return "\xe3\x85\x85";
2996 } elseif ( $code < 0xc790 ) {
2997 return "\xe3\x85\x87";
2998 } elseif ( $code < 0xcc28 ) {
2999 return "\xe3\x85\x88";
3000 } elseif ( $code < 0xce74 ) {
3001 return "\xe3\x85\x8a";
3002 } elseif ( $code < 0xd0c0 ) {
3003 return "\xe3\x85\x8b";
3004 } elseif ( $code < 0xd30c ) {
3005 return "\xe3\x85\x8c";
3006 } elseif ( $code < 0xd558 ) {
3007 return "\xe3\x85\x8d";
3009 return "\xe3\x85\x8e";
3016 function initEncoding() {
3017 # Some languages may have an alternate char encoding option
3018 # (Esperanto X-coding, Japanese furigana conversion, etc)
3019 # If this language is used as the primary content language,
3020 # an override to the defaults can be set here on startup.
3027 function recodeForEdit( $s ) {
3028 # For some languages we'll want to explicitly specify
3029 # which characters make it into the edit box raw
3030 # or are converted in some way or another.
3031 global $wgEditEncoding;
3032 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
3035 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3043 function recodeInput( $s ) {
3044 # Take the previous into account.
3045 global $wgEditEncoding;
3046 if ( $wgEditEncoding != '' ) {
3047 $enc = $wgEditEncoding;
3051 if ( $enc == 'UTF-8' ) {
3054 return $this->iconv( $enc, 'UTF-8', $s );
3059 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3060 * also cleans up certain backwards-compatible sequences, converting them
3061 * to the modern Unicode equivalent.
3063 * This is language-specific for performance reasons only.
3069 function normalize( $s ) {
3070 global $wgAllUnicodeFixes;
3071 $s = UtfNormal\Validator
::cleanUp( $s );
3072 if ( $wgAllUnicodeFixes ) {
3073 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3074 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3081 * Transform a string using serialized data stored in the given file (which
3082 * must be in the serialized subdirectory of $IP). The file contains pairs
3083 * mapping source characters to destination characters.
3085 * The data is cached in process memory. This will go faster if you have the
3086 * FastStringSearch extension.
3088 * @param string $file
3089 * @param string $string
3091 * @throws MWException
3094 function transformUsingPairFile( $file, $string ) {
3095 if ( !isset( $this->transformData
[$file] ) ) {
3096 $data = wfGetPrecompiledData( $file );
3097 if ( $data === false ) {
3098 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3100 $this->transformData
[$file] = new ReplacementArray( $data );
3102 return $this->transformData
[$file]->replace( $string );
3106 * For right-to-left language support
3111 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3115 * Return the correct HTML 'dir' attribute value for this language.
3119 return $this->isRTL() ?
'rtl' : 'ltr';
3123 * Return 'left' or 'right' as appropriate alignment for line-start
3124 * for this language's text direction.
3126 * Should be equivalent to CSS3 'start' text-align value....
3130 function alignStart() {
3131 return $this->isRTL() ?
'right' : 'left';
3135 * Return 'right' or 'left' as appropriate alignment for line-end
3136 * for this language's text direction.
3138 * Should be equivalent to CSS3 'end' text-align value....
3142 function alignEnd() {
3143 return $this->isRTL() ?
'left' : 'right';
3147 * A hidden direction mark (LRM or RLM), depending on the language direction.
3148 * Unlike getDirMark(), this function returns the character as an HTML entity.
3149 * This function should be used when the output is guaranteed to be HTML,
3150 * because it makes the output HTML source code more readable. When
3151 * the output is plain text or can be escaped, getDirMark() should be used.
3153 * @param bool $opposite Get the direction mark opposite to your language
3157 function getDirMarkEntity( $opposite = false ) {
3159 return $this->isRTL() ?
'‎' : '‏';
3161 return $this->isRTL() ?
'‏' : '‎';
3165 * A hidden direction mark (LRM or RLM), depending on the language direction.
3166 * This function produces them as invisible Unicode characters and
3167 * the output may be hard to read and debug, so it should only be used
3168 * when the output is plain text or can be escaped. When the output is
3169 * HTML, use getDirMarkEntity() instead.
3171 * @param bool $opposite Get the direction mark opposite to your language
3174 function getDirMark( $opposite = false ) {
3175 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3176 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3178 return $this->isRTL() ?
$lrm : $rlm;
3180 return $this->isRTL() ?
$rlm : $lrm;
3186 function capitalizeAllNouns() {
3187 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3191 * An arrow, depending on the language direction.
3193 * @param string $direction The direction of the arrow: forwards (default),
3194 * backwards, left, right, up, down.
3197 function getArrow( $direction = 'forwards' ) {
3198 switch ( $direction ) {
3200 return $this->isRTL() ?
'←' : '→';
3202 return $this->isRTL() ?
'→' : '←';
3215 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3219 function linkPrefixExtension() {
3220 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3224 * Get all magic words from cache.
3227 function getMagicWords() {
3228 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3232 * Run the LanguageGetMagic hook once.
3234 protected function doMagicHook() {
3235 if ( $this->mMagicHookDone
) {
3238 $this->mMagicHookDone
= true;
3239 Hooks
::run( 'LanguageGetMagic', [ &$this->mMagicExtensions
, $this->getCode() ] );
3243 * Fill a MagicWord object with data from here
3245 * @param MagicWord $mw
3247 function getMagic( $mw ) {
3248 // Saves a function call
3249 if ( !$this->mMagicHookDone
) {
3250 $this->doMagicHook();
3253 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3254 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3256 $rawEntry = self
::$dataCache->getSubitem(
3257 $this->mCode
, 'magicWords', $mw->mId
);
3260 if ( !is_array( $rawEntry ) ) {
3261 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3263 $mw->mCaseSensitive
= $rawEntry[0];
3264 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3269 * Add magic words to the extension array
3271 * @param array $newWords
3273 function addMagicWordsByLang( $newWords ) {
3274 $fallbackChain = $this->getFallbackLanguages();
3275 $fallbackChain = array_reverse( $fallbackChain );
3276 foreach ( $fallbackChain as $code ) {
3277 if ( isset( $newWords[$code] ) ) {
3278 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3284 * Get special page names, as an associative array
3285 * canonical name => array of valid names, including aliases
3288 function getSpecialPageAliases() {
3289 // Cache aliases because it may be slow to load them
3290 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3292 $this->mExtendedSpecialPageAliases
=
3293 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3294 Hooks
::run( 'LanguageGetSpecialPageAliases',
3295 [ &$this->mExtendedSpecialPageAliases
, $this->getCode() ] );
3298 return $this->mExtendedSpecialPageAliases
;
3302 * Italic is unsuitable for some languages
3304 * @param string $text The text to be emphasized.
3307 function emphasize( $text ) {
3308 return "<em>$text</em>";
3312 * Normally we output all numbers in plain en_US style, that is
3313 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3314 * point twohundredthirtyfive. However this is not suitable for all
3315 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3316 * Icelandic just want to use commas instead of dots, and dots instead
3317 * of commas like "293.291,235".
3319 * An example of this function being called:
3321 * wfMessage( 'message' )->numParams( $num )->text()
3324 * See $separatorTransformTable on MessageIs.php for
3325 * the , => . and . => , implementation.
3327 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3328 * @param int|float $number The string to be formatted, should be an integer
3329 * or a floating point number.
3330 * @param bool $nocommafy Set to true for special numbers like dates
3333 public function formatNum( $number, $nocommafy = false ) {
3334 global $wgTranslateNumerals;
3335 if ( !$nocommafy ) {
3336 $number = $this->commafy( $number );
3337 $s = $this->separatorTransformTable();
3339 $number = strtr( $number, $s );
3343 if ( $wgTranslateNumerals ) {
3344 $s = $this->digitTransformTable();
3346 $number = strtr( $number, $s );
3354 * Front-end for non-commafied formatNum
3356 * @param int|float $number The string to be formatted, should be an integer
3357 * or a floating point number.
3361 public function formatNumNoSeparators( $number ) {
3362 return $this->formatNum( $number, true );
3366 * @param string $number
3369 public function parseFormattedNumber( $number ) {
3370 $s = $this->digitTransformTable();
3372 // eliminate empty array values such as ''. (bug 64347)
3373 $s = array_filter( $s );
3374 $number = strtr( $number, array_flip( $s ) );
3377 $s = $this->separatorTransformTable();
3379 // eliminate empty array values such as ''. (bug 64347)
3380 $s = array_filter( $s );
3381 $number = strtr( $number, array_flip( $s ) );
3384 $number = strtr( $number, [ ',' => '' ] );
3389 * Adds commas to a given number
3391 * @param mixed $number
3394 function commafy( $number ) {
3395 $digitGroupingPattern = $this->digitGroupingPattern();
3396 if ( $number === null ) {
3400 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3401 // default grouping is at thousands, use the same for ###,###,### pattern too.
3402 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3404 // Ref: http://cldr.unicode.org/translation/number-patterns
3406 if ( intval( $number ) < 0 ) {
3407 // For negative numbers apply the algorithm like positive number and add sign.
3409 $number = substr( $number, 1 );
3413 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3414 preg_match( "/\d+/", $number, $integerPart );
3415 preg_match( "/\.\d*/", $number, $decimalPart );
3416 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3417 if ( $groupedNumber === $number ) {
3418 // the string does not have any number part. Eg: .12345
3419 return $sign . $groupedNumber;
3421 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3422 while ( $start > 0 ) {
3423 $match = $matches[0][$numMatches - 1];
3424 $matchLen = strlen( $match );
3425 $start = $end - $matchLen;
3429 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3431 if ( $numMatches > 1 ) {
3432 // use the last pattern for the rest of the number
3436 $groupedNumber = "," . $groupedNumber;
3439 return $sign . $groupedNumber;
3446 function digitGroupingPattern() {
3447 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3453 function digitTransformTable() {
3454 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3460 function separatorTransformTable() {
3461 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3465 * Take a list of strings and build a locale-friendly comma-separated
3466 * list, using the local comma-separator message.
3467 * The last two strings are chained with an "and".
3468 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3470 * @param string[] $l
3473 function listToText( array $l ) {
3474 $m = count( $l ) - 1;
3479 $and = $this->msg( 'and' )->escaped();
3480 $space = $this->msg( 'word-separator' )->escaped();
3482 $comma = $this->msg( 'comma-separator' )->escaped();
3486 for ( $i = $m - 1; $i >= 0; $i-- ) {
3487 if ( $i == $m - 1 ) {
3488 $s = $l[$i] . $and . $space . $s;
3490 $s = $l[$i] . $comma . $s;
3497 * Take a list of strings and build a locale-friendly comma-separated
3498 * list, using the local comma-separator message.
3499 * @param string[] $list Array of strings to put in a comma list
3502 function commaList( array $list ) {
3504 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3510 * Take a list of strings and build a locale-friendly semicolon-separated
3511 * list, using the local semicolon-separator message.
3512 * @param string[] $list Array of strings to put in a semicolon list
3515 function semicolonList( array $list ) {
3517 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3523 * Same as commaList, but separate it with the pipe instead.
3524 * @param string[] $list Array of strings to put in a pipe list
3527 function pipeList( array $list ) {
3529 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3535 * Truncate a string to a specified length in bytes, appending an optional
3536 * string (e.g. for ellipses)
3538 * The database offers limited byte lengths for some columns in the database;
3539 * multi-byte character sets mean we need to ensure that only whole characters
3540 * are included, otherwise broken characters can be passed to the user
3542 * If $length is negative, the string will be truncated from the beginning
3544 * @param string $string String to truncate
3545 * @param int $length Maximum length (including ellipses)
3546 * @param string $ellipsis String to append to the truncated text
3547 * @param bool $adjustLength Subtract length of ellipsis from $length.
3548 * $adjustLength was introduced in 1.18, before that behaved as if false.
3551 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3552 # Use the localized ellipsis character
3553 if ( $ellipsis == '...' ) {
3554 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3556 # Check if there is no need to truncate
3557 if ( $length == 0 ) {
3558 return $ellipsis; // convention
3559 } elseif ( strlen( $string ) <= abs( $length ) ) {
3560 return $string; // no need to truncate
3562 $stringOriginal = $string;
3563 # If ellipsis length is >= $length then we can't apply $adjustLength
3564 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3565 $string = $ellipsis; // this can be slightly unexpected
3566 # Otherwise, truncate and add ellipsis...
3568 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3569 if ( $length > 0 ) {
3570 $length -= $eLength;
3571 $string = substr( $string, 0, $length ); // xyz...
3572 $string = $this->removeBadCharLast( $string );
3573 $string = rtrim( $string );
3574 $string = $string . $ellipsis;
3576 $length +
= $eLength;
3577 $string = substr( $string, $length ); // ...xyz
3578 $string = $this->removeBadCharFirst( $string );
3579 $string = ltrim( $string );
3580 $string = $ellipsis . $string;
3583 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3584 # This check is *not* redundant if $adjustLength, due to the single case where
3585 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3586 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3589 return $stringOriginal;
3594 * Remove bytes that represent an incomplete Unicode character
3595 * at the end of string (e.g. bytes of the char are missing)
3597 * @param string $string
3600 protected function removeBadCharLast( $string ) {
3601 if ( $string != '' ) {
3602 $char = ord( $string[strlen( $string ) - 1] );
3604 if ( $char >= 0xc0 ) {
3605 # We got the first byte only of a multibyte char; remove it.
3606 $string = substr( $string, 0, -1 );
3607 } elseif ( $char >= 0x80 &&
3608 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3609 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3610 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3612 # We chopped in the middle of a character; remove it
3620 * Remove bytes that represent an incomplete Unicode character
3621 * at the start of string (e.g. bytes of the char are missing)
3623 * @param string $string
3626 protected function removeBadCharFirst( $string ) {
3627 if ( $string != '' ) {
3628 $char = ord( $string[0] );
3629 if ( $char >= 0x80 && $char < 0xc0 ) {
3630 # We chopped in the middle of a character; remove the whole thing
3631 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3638 * Truncate a string of valid HTML to a specified length in bytes,
3639 * appending an optional string (e.g. for ellipses), and return valid HTML
3641 * This is only intended for styled/linked text, such as HTML with
3642 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3643 * Also, this will not detect things like "display:none" CSS.
3645 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3647 * @param string $text HTML string to truncate
3648 * @param int $length (zero/positive) Maximum length (including ellipses)
3649 * @param string $ellipsis String to append to the truncated text
3652 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3653 # Use the localized ellipsis character
3654 if ( $ellipsis == '...' ) {
3655 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3657 # Check if there is clearly no need to truncate
3658 if ( $length <= 0 ) {
3659 return $ellipsis; // no text shown, nothing to format (convention)
3660 } elseif ( strlen( $text ) <= $length ) {
3661 return $text; // string short enough even *with* HTML (short-circuit)
3664 $dispLen = 0; // innerHTML legth so far
3665 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3666 $tagType = 0; // 0-open, 1-close
3667 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3668 $entityState = 0; // 0-not entity, 1-entity
3669 $tag = $ret = ''; // accumulated tag name, accumulated result string
3670 $openTags = []; // open tag stack
3671 $maybeState = null; // possible truncation state
3673 $textLen = strlen( $text );
3674 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3675 for ( $pos = 0; true; ++
$pos ) {
3676 # Consider truncation once the display length has reached the maximim.
3677 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3678 # Check that we're not in the middle of a bracket/entity...
3679 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3680 if ( !$testingEllipsis ) {
3681 $testingEllipsis = true;
3682 # Save where we are; we will truncate here unless there turn out to
3683 # be so few remaining characters that truncation is not necessary.
3684 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3685 $maybeState = [ $ret, $openTags ]; // save state
3687 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3688 # String in fact does need truncation, the truncation point was OK.
3689 list( $ret, $openTags ) = $maybeState; // reload state
3690 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3691 $ret .= $ellipsis; // add ellipsis
3695 if ( $pos >= $textLen ) {
3696 break; // extra iteration just for above checks
3699 # Read the next char...
3701 $lastCh = $pos ?
$text[$pos - 1] : '';
3702 $ret .= $ch; // add to result string
3704 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3705 $entityState = 0; // for bad HTML
3706 $bracketState = 1; // tag started (checking for backslash)
3707 } elseif ( $ch == '>' ) {
3708 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3709 $entityState = 0; // for bad HTML
3710 $bracketState = 0; // out of brackets
3711 } elseif ( $bracketState == 1 ) {
3713 $tagType = 1; // close tag (e.g. "</span>")
3715 $tagType = 0; // open tag (e.g. "<span>")
3718 $bracketState = 2; // building tag name
3719 } elseif ( $bracketState == 2 ) {
3723 // Name found (e.g. "<a href=..."), add on tag attributes...
3724 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3726 } elseif ( $bracketState == 0 ) {
3727 if ( $entityState ) {
3730 $dispLen++
; // entity is one displayed char
3733 if ( $neLength == 0 && !$maybeState ) {
3734 // Save state without $ch. We want to *hit* the first
3735 // display char (to get tags) but not *use* it if truncating.
3736 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3739 $entityState = 1; // entity found, (e.g. " ")
3741 $dispLen++
; // this char is displayed
3742 // Add the next $max display text chars after this in one swoop...
3743 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3744 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3745 $dispLen +
= $skipped;
3751 // Close the last tag if left unclosed by bad HTML
3752 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3753 while ( count( $openTags ) > 0 ) {
3754 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3760 * truncateHtml() helper function
3761 * like strcspn() but adds the skipped chars to $ret
3763 * @param string $ret
3764 * @param string $text
3765 * @param string $search
3767 * @param null|int $len
3770 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3771 if ( $len === null ) {
3772 $len = -1; // -1 means "no limit" for strcspn
3773 } elseif ( $len < 0 ) {
3777 if ( $start < strlen( $text ) ) {
3778 $skipCount = strcspn( $text, $search, $start, $len );
3779 $ret .= substr( $text, $start, $skipCount );
3785 * truncateHtml() helper function
3786 * (a) push or pop $tag from $openTags as needed
3787 * (b) clear $tag value
3788 * @param string &$tag Current HTML tag name we are looking at
3789 * @param int $tagType (0-open tag, 1-close tag)
3790 * @param string $lastCh Character before the '>' that ended this tag
3791 * @param array &$openTags Open tag stack (not accounting for $tag)
3793 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3794 $tag = ltrim( $tag );
3796 if ( $tagType == 0 && $lastCh != '/' ) {
3797 $openTags[] = $tag; // tag opened (didn't close itself)
3798 } elseif ( $tagType == 1 ) {
3799 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3800 array_pop( $openTags ); // tag closed
3808 * Grammatical transformations, needed for inflected languages
3809 * Invoked by putting {{grammar:case|word}} in a message
3811 * @param string $word
3812 * @param string $case
3815 function convertGrammar( $word, $case ) {
3816 global $wgGrammarForms;
3817 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3818 return $wgGrammarForms[$this->getCode()][$case][$word];
3824 * Get the grammar forms for the content language
3825 * @return array Array of grammar forms
3828 function getGrammarForms() {
3829 global $wgGrammarForms;
3830 if ( isset( $wgGrammarForms[$this->getCode()] )
3831 && is_array( $wgGrammarForms[$this->getCode()] )
3833 return $wgGrammarForms[$this->getCode()];
3839 * Provides an alternative text depending on specified gender.
3840 * Usage {{gender:username|masculine|feminine|unknown}}.
3841 * username is optional, in which case the gender of current user is used,
3842 * but only in (some) interface messages; otherwise default gender is used.
3844 * If no forms are given, an empty string is returned. If only one form is
3845 * given, it will be returned unconditionally. These details are implied by
3846 * the caller and cannot be overridden in subclasses.
3848 * If three forms are given, the default is to use the third (unknown) form.
3849 * If fewer than three forms are given, the default is to use the first (masculine) form.
3850 * These details can be overridden in subclasses.
3852 * @param string $gender
3853 * @param array $forms
3857 function gender( $gender, $forms ) {
3858 if ( !count( $forms ) ) {
3861 $forms = $this->preConvertPlural( $forms, 2 );
3862 if ( $gender === 'male' ) {
3865 if ( $gender === 'female' ) {
3868 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3872 * Plural form transformations, needed for some languages.
3873 * For example, there are 3 form of plural in Russian and Polish,
3874 * depending on "count mod 10". See [[w:Plural]]
3875 * For English it is pretty simple.
3877 * Invoked by putting {{plural:count|wordform1|wordform2}}
3878 * or {{plural:count|wordform1|wordform2|wordform3}}
3880 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3882 * @param int $count Non-localized number
3883 * @param array $forms Different plural forms
3884 * @return string Correct form of plural for $count in this language
3886 function convertPlural( $count, $forms ) {
3887 // Handle explicit n=pluralform cases
3888 $forms = $this->handleExplicitPluralForms( $count, $forms );
3889 if ( is_string( $forms ) ) {
3892 if ( !count( $forms ) ) {
3896 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3897 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3898 return $forms[$pluralForm];
3902 * Handles explicit plural forms for Language::convertPlural()
3904 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3905 * If an explicitly defined plural form matches the $count, then
3906 * string value returned, otherwise array returned for further consideration
3907 * by CLDR rules or overridden convertPlural().
3911 * @param int $count Non-localized number
3912 * @param array $forms Different plural forms
3914 * @return array|string
3916 protected function handleExplicitPluralForms( $count, array $forms ) {
3917 foreach ( $forms as $index => $form ) {
3918 if ( preg_match( '/\d+=/i', $form ) ) {
3919 $pos = strpos( $form, '=' );
3920 if ( substr( $form, 0, $pos ) === (string)$count ) {
3921 return substr( $form, $pos +
1 );
3923 unset( $forms[$index] );
3926 return array_values( $forms );
3930 * Checks that convertPlural was given an array and pads it to requested
3931 * amount of forms by copying the last one.
3933 * @param array $forms Array of forms given to convertPlural
3934 * @param int $count How many forms should there be at least
3935 * @return array Padded array of forms or an exception if not an array
3937 protected function preConvertPlural( /* Array */ $forms, $count ) {
3938 while ( count( $forms ) < $count ) {
3939 $forms[] = $forms[count( $forms ) - 1];
3945 * Wraps argument with unicode control characters for directionality safety
3947 * This solves the problem where directionality-neutral characters at the edge of
3948 * the argument string get interpreted with the wrong directionality from the
3949 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3951 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3952 * directionality of the argument string, using the BIDI algorithm's own "First
3953 * strong directional codepoint" rule. Essentially, this works round the fact that
3954 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3955 * direction inference). The latter is cleaner but still not widely supported.
3957 * @param string $text Text to wrap
3958 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3960 public function embedBidi( $text = '' ) {
3961 $dir = Language
::strongDirFromContent( $text );
3962 if ( $dir === 'ltr' ) {
3963 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3964 return self
::$lre . $text . self
::$pdf;
3966 if ( $dir === 'rtl' ) {
3967 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3968 return self
::$rle . $text . self
::$pdf;
3970 // No strong directionality: do not wrap
3975 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3976 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3977 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3978 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3981 * @param string $str The validated block duration in English
3982 * @return string Somehow translated block duration
3983 * @see LanguageFi.php for example implementation
3985 function translateBlockExpiry( $str ) {
3986 $duration = SpecialBlock
::getSuggestedDurations( $this );
3987 foreach ( $duration as $show => $value ) {
3988 if ( strcmp( $str, $value ) == 0 ) {
3989 return htmlspecialchars( trim( $show ) );
3993 if ( wfIsInfinity( $str ) ) {
3994 foreach ( $duration as $show => $value ) {
3995 if ( wfIsInfinity( $value ) ) {
3996 return htmlspecialchars( trim( $show ) );
4001 // If all else fails, return a standard duration or timestamp description.
4002 $time = strtotime( $str, 0 );
4003 if ( $time === false ) { // Unknown format. Return it as-is in case.
4005 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4006 // $time is relative to 0 so it's a duration length.
4007 return $this->formatDuration( $time );
4008 } else { // It's an absolute timestamp.
4009 if ( $time === 0 ) {
4010 // wfTimestamp() handles 0 as current time instead of epoch.
4011 return $this->timeanddate( '19700101000000' );
4013 return $this->timeanddate( $time );
4019 * languages like Chinese need to be segmented in order for the diff
4022 * @param string $text
4025 public function segmentForDiff( $text ) {
4030 * and unsegment to show the result
4032 * @param string $text
4035 public function unsegmentForDiff( $text ) {
4040 * Return the LanguageConverter used in the Language
4043 * @return LanguageConverter
4045 public function getConverter() {
4046 return $this->mConverter
;
4050 * convert text to all supported variants
4052 * @param string $text
4055 public function autoConvertToAllVariants( $text ) {
4056 return $this->mConverter
->autoConvertToAllVariants( $text );
4060 * convert text to different variants of a language.
4062 * @param string $text
4065 public function convert( $text ) {
4066 return $this->mConverter
->convert( $text );
4070 * Convert a Title object to a string in the preferred variant
4072 * @param Title $title
4075 public function convertTitle( $title ) {
4076 return $this->mConverter
->convertTitle( $title );
4080 * Convert a namespace index to a string in the preferred variant
4085 public function convertNamespace( $ns ) {
4086 return $this->mConverter
->convertNamespace( $ns );
4090 * Check if this is a language with variants
4094 public function hasVariants() {
4095 return count( $this->getVariants() ) > 1;
4099 * Check if the language has the specific variant
4102 * @param string $variant
4105 public function hasVariant( $variant ) {
4106 return (bool)$this->mConverter
->validateVariant( $variant );
4110 * Perform output conversion on a string, and encode for safe HTML output.
4111 * @param string $text Text to be converted
4112 * @param bool $isTitle Whether this conversion is for the article title
4114 * @todo this should get integrated somewhere sane
4116 public function convertHtml( $text, $isTitle = false ) {
4117 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4121 * @param string $key
4124 public function convertCategoryKey( $key ) {
4125 return $this->mConverter
->convertCategoryKey( $key );
4129 * Get the list of variants supported by this language
4130 * see sample implementation in LanguageZh.php
4132 * @return array An array of language codes
4134 public function getVariants() {
4135 return $this->mConverter
->getVariants();
4141 public function getPreferredVariant() {
4142 return $this->mConverter
->getPreferredVariant();
4148 public function getDefaultVariant() {
4149 return $this->mConverter
->getDefaultVariant();
4155 public function getURLVariant() {
4156 return $this->mConverter
->getURLVariant();
4160 * If a language supports multiple variants, it is
4161 * possible that non-existing link in one variant
4162 * actually exists in another variant. this function
4163 * tries to find it. See e.g. LanguageZh.php
4164 * The input parameters may be modified upon return
4166 * @param string &$link The name of the link
4167 * @param Title &$nt The title object of the link
4168 * @param bool $ignoreOtherCond To disable other conditions when
4169 * we need to transclude a template or update a category's link
4171 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4172 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4176 * returns language specific options used by User::getPageRenderHash()
4177 * for example, the preferred language variant
4181 function getExtraHashOptions() {
4182 return $this->mConverter
->getExtraHashOptions();
4186 * For languages that support multiple variants, the title of an
4187 * article may be displayed differently in different variants. this
4188 * function returns the apporiate title defined in the body of the article.
4192 public function getParsedTitle() {
4193 return $this->mConverter
->getParsedTitle();
4197 * Refresh the cache of conversion tables when
4198 * MediaWiki:Conversiontable* is updated.
4200 * @param Title $titleobj The Title of the page being updated
4202 public function updateConversionTable( Title
$title ) {
4203 $this->mConverter
->updateConversionTable( $title );
4207 * Prepare external link text for conversion. When the text is
4208 * a URL, it shouldn't be converted, and it'll be wrapped in
4209 * the "raw" tag (-{R| }-) to prevent conversion.
4211 * This function is called "markNoConversion" for historical
4214 * @param string $text Text to be used for external link
4215 * @param bool $noParse Wrap it without confirming it's a real URL first
4216 * @return string The tagged text
4218 public function markNoConversion( $text, $noParse = false ) {
4219 // Excluding protocal-relative URLs may avoid many false positives.
4220 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4221 return $this->mConverter
->markNoConversion( $text );
4228 * A regular expression to match legal word-trailing characters
4229 * which should be merged onto a link of the form [[foo]]bar.
4233 public function linkTrail() {
4234 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4238 * A regular expression character set to match legal word-prefixing
4239 * characters which should be merged onto a link of the form foo[[bar]].
4243 public function linkPrefixCharset() {
4244 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4248 * @deprecated since 1.24, will be removed in 1.25
4251 function getLangObj() {
4252 wfDeprecated( __METHOD__
, '1.24' );
4257 * Get the "parent" language which has a converter to convert a "compatible" language
4258 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4260 * @return Language|null
4263 public function getParentLanguage() {
4264 if ( $this->mParentLanguage
!== false ) {
4265 return $this->mParentLanguage
;
4268 $code = explode( '-', $this->getCode() )[0];
4269 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4270 $this->mParentLanguage
= null;
4273 $lang = Language
::factory( $code );
4274 if ( !$lang->hasVariant( $this->getCode() ) ) {
4275 $this->mParentLanguage
= null;
4279 $this->mParentLanguage
= $lang;
4284 * Get the internal language code for this language object
4286 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4287 * htmlspecialchars() or similar
4291 public function getCode() {
4292 return $this->mCode
;
4296 * Get the code in BCP 47 format which we can use
4297 * inside of html lang="" tags.
4299 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4300 * htmlspecialchars() or similar.
4305 public function getHtmlCode() {
4306 if ( is_null( $this->mHtmlCode
) ) {
4307 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4309 return $this->mHtmlCode
;
4313 * @param string $code
4315 public function setCode( $code ) {
4316 $this->mCode
= $code;
4317 // Ensure we don't leave incorrect cached data lying around
4318 $this->mHtmlCode
= null;
4319 $this->mParentLanguage
= false;
4323 * Get the language code from a file name. Inverse of getFileName()
4324 * @param string $filename $prefix . $languageCode . $suffix
4325 * @param string $prefix Prefix before the language code
4326 * @param string $suffix Suffix after the language code
4327 * @return string Language code, or false if $prefix or $suffix isn't found
4329 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4331 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4332 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4333 if ( !count( $m ) ) {
4336 return str_replace( '_', '-', strtolower( $m[1] ) );
4340 * @param string $code
4341 * @return string Name of the language class
4343 public static function classFromCode( $code ) {
4344 if ( $code == 'en' ) {
4347 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4352 * Get the name of a file for a certain language code
4353 * @param string $prefix Prepend this to the filename
4354 * @param string $code Language code
4355 * @param string $suffix Append this to the filename
4356 * @throws MWException
4357 * @return string $prefix . $mangledCode . $suffix
4359 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4360 if ( !self
::isValidBuiltInCode( $code ) ) {
4361 throw new MWException( "Invalid language code \"$code\"" );
4364 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4368 * @param string $code
4371 public static function getMessagesFileName( $code ) {
4373 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4374 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4379 * @param string $code
4383 public static function getJsonMessagesFileName( $code ) {
4386 if ( !self
::isValidBuiltInCode( $code ) ) {
4387 throw new MWException( "Invalid language code \"$code\"" );
4390 return "$IP/languages/i18n/$code.json";
4394 * Get the first fallback for a given language.
4396 * @param string $code
4398 * @return bool|string
4400 public static function getFallbackFor( $code ) {
4401 $fallbacks = self
::getFallbacksFor( $code );
4403 return $fallbacks[0];
4409 * Get the ordered list of fallback languages.
4412 * @param string $code Language code
4413 * @return array Non-empty array, ending in "en"
4415 public static function getFallbacksFor( $code ) {
4416 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4419 // For unknown languages, fallbackSequence returns an empty array,
4420 // hardcode fallback to 'en' in that case.
4421 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4425 * Get the ordered list of fallback languages, ending with the fallback
4426 * language chain for the site language.
4429 * @param string $code Language code
4430 * @return array Array( fallbacks, site fallbacks )
4432 public static function getFallbacksIncludingSiteLanguage( $code ) {
4433 global $wgLanguageCode;
4435 // Usually, we will only store a tiny number of fallback chains, so we
4436 // keep them in static memory.
4437 $cacheKey = "{$code}-{$wgLanguageCode}";
4439 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4440 $fallbacks = self
::getFallbacksFor( $code );
4442 // Append the site's fallback chain, including the site language itself
4443 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4444 array_unshift( $siteFallbacks, $wgLanguageCode );
4446 // Eliminate any languages already included in the chain
4447 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4449 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4451 return self
::$fallbackLanguageCache[$cacheKey];
4455 * Get all messages for a given language
4456 * WARNING: this may take a long time. If you just need all message *keys*
4457 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4459 * @param string $code
4463 public static function getMessagesFor( $code ) {
4464 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4468 * Get a message for a given language
4470 * @param string $key
4471 * @param string $code
4475 public static function getMessageFor( $key, $code ) {
4476 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4480 * Get all message keys for a given language. This is a faster alternative to
4481 * array_keys( Language::getMessagesFor( $code ) )
4484 * @param string $code Language code
4485 * @return array Array of message keys (strings)
4487 public static function getMessageKeysFor( $code ) {
4488 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4492 * @param string $talk
4495 function fixVariableInNamespace( $talk ) {
4496 if ( strpos( $talk, '$1' ) === false ) {
4500 global $wgMetaNamespace;
4501 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4503 # Allow grammar transformations
4504 # Allowing full message-style parsing would make simple requests
4505 # such as action=raw much more expensive than they need to be.
4506 # This will hopefully cover most cases.
4507 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4508 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4509 return str_replace( ' ', '_', $talk );
4516 function replaceGrammarInNamespace( $m ) {
4517 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4521 * @throws MWException
4524 static function getCaseMaps() {
4525 static $wikiUpperChars, $wikiLowerChars;
4526 if ( isset( $wikiUpperChars ) ) {
4527 return [ $wikiUpperChars, $wikiLowerChars ];
4530 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4531 if ( $arr === false ) {
4532 throw new MWException(
4533 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4535 $wikiUpperChars = $arr['wikiUpperChars'];
4536 $wikiLowerChars = $arr['wikiLowerChars'];
4537 return [ $wikiUpperChars, $wikiLowerChars ];
4541 * Decode an expiry (block, protection, etc) which has come from the DB
4543 * @param string $expiry Database expiry String
4544 * @param bool|int $format True to process using language functions, or TS_ constant
4545 * to return the expiry in a given timestamp
4546 * @param string $inifinity If $format is not true, use this string for infinite expiry
4550 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4552 if ( $dbInfinity === null ) {
4553 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4556 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4557 return $format === true
4558 ?
$this->getMessageFromDB( 'infiniteblock' )
4561 return $format === true
4562 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4563 : wfTimestamp( $format, $expiry );
4569 * @param int|float $seconds
4570 * @param array $format Optional
4571 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4572 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4573 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4575 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4576 * or 'avoidminutes'.
4579 function formatTimePeriod( $seconds, $format = [] ) {
4580 if ( !is_array( $format ) ) {
4581 $format = [ 'avoid' => $format ]; // For backwards compatibility
4583 if ( !isset( $format['avoid'] ) ) {
4584 $format['avoid'] = false;
4586 if ( !isset( $format['noabbrevs'] ) ) {
4587 $format['noabbrevs'] = false;
4589 $secondsMsg = wfMessage(
4590 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4591 $minutesMsg = wfMessage(
4592 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4593 $hoursMsg = wfMessage(
4594 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4595 $daysMsg = wfMessage(
4596 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4598 if ( round( $seconds * 10 ) < 100 ) {
4599 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4600 $s = $secondsMsg->params( $s )->text();
4601 } elseif ( round( $seconds ) < 60 ) {
4602 $s = $this->formatNum( round( $seconds ) );
4603 $s = $secondsMsg->params( $s )->text();
4604 } elseif ( round( $seconds ) < 3600 ) {
4605 $minutes = floor( $seconds / 60 );
4606 $secondsPart = round( fmod( $seconds, 60 ) );
4607 if ( $secondsPart == 60 ) {
4611 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4613 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4614 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4615 $hours = floor( $seconds / 3600 );
4616 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4617 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4618 if ( $secondsPart == 60 ) {
4622 if ( $minutes == 60 ) {
4626 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4628 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4629 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4630 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4633 $days = floor( $seconds / 86400 );
4634 if ( $format['avoid'] === 'avoidminutes' ) {
4635 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4636 if ( $hours == 24 ) {
4640 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4642 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4643 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4644 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4645 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4646 if ( $minutes == 60 ) {
4650 if ( $hours == 24 ) {
4654 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4656 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4658 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4660 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4662 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4669 * Format a bitrate for output, using an appropriate
4670 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4671 * the magnitude in question.
4673 * This use base 1000. For base 1024 use formatSize(), for another base
4674 * see formatComputingNumbers().
4679 function formatBitrate( $bps ) {
4680 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4684 * @param int $size Size of the unit
4685 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4686 * @param string $messageKey Message key to be uesd
4689 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4691 return str_replace( '$1', $this->formatNum( $size ),
4692 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4695 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4698 $maxIndex = count( $sizes ) - 1;
4699 while ( $size >= $boundary && $index < $maxIndex ) {
4704 // For small sizes no decimal places necessary
4707 // For MB and bigger two decimal places are smarter
4710 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4712 $size = round( $size, $round );
4713 $text = $this->getMessageFromDB( $msg );
4714 return str_replace( '$1', $this->formatNum( $size ), $text );
4718 * Format a size in bytes for output, using an appropriate
4719 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4721 * This method use base 1024. For base 1000 use formatBitrate(), for
4722 * another base see formatComputingNumbers()
4724 * @param int $size Size to format
4725 * @return string Plain text (not HTML)
4727 function formatSize( $size ) {
4728 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4732 * Make a list item, used by various special pages
4734 * @param string $page Page link
4735 * @param string $details HTML safe text between brackets
4736 * @param bool $oppositedm Add the direction mark opposite to your
4737 * language, to display text properly
4738 * @return HTML escaped string
4740 function specialList( $page, $details, $oppositedm = true ) {
4745 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4749 $this->msg( 'word-separator' )->escaped() .
4750 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4754 * Generate (prev x| next x) (20|50|100...) type links for paging
4756 * @param Title $title Title object to link
4757 * @param int $offset
4759 * @param array $query Optional URL query parameter string
4760 * @param bool $atend Optional param for specified if this is the last page
4763 public function viewPrevNext( Title
$title, $offset, $limit,
4764 array $query = [], $atend = false
4766 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4768 # Make 'previous' link
4769 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4770 if ( $offset > 0 ) {
4771 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4772 $query, $prev, 'prevn-title', 'mw-prevlink' );
4774 $plink = htmlspecialchars( $prev );
4778 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4780 $nlink = htmlspecialchars( $next );
4782 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4783 $query, $next, 'nextn-title', 'mw-nextlink' );
4786 # Make links to set number of items per page
4788 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4789 $numLinks[] = $this->numLink( $title, $offset, $num,
4790 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4793 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4794 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4798 * Helper function for viewPrevNext() that generates links
4800 * @param Title $title Title object to link
4801 * @param int $offset
4803 * @param array $query Extra query parameters
4804 * @param string $link Text to use for the link; will be escaped
4805 * @param string $tooltipMsg Name of the message to use as tooltip
4806 * @param string $class Value of the "class" attribute of the link
4807 * @return string HTML fragment
4809 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4812 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4813 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4814 ->numParams( $limit )->text();
4816 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4817 'title' => $tooltip, 'class' => $class ], $link );
4821 * Get the conversion rule title, if any.
4825 public function getConvRuleTitle() {
4826 return $this->mConverter
->getConvRuleTitle();
4830 * Get the compiled plural rules for the language
4832 * @return array Associative array with plural form, and plural rule as key-value pairs
4834 public function getCompiledPluralRules() {
4835 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4836 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4837 if ( !$pluralRules ) {
4838 foreach ( $fallbacks as $fallbackCode ) {
4839 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4840 if ( $pluralRules ) {
4845 return $pluralRules;
4849 * Get the plural rules for the language
4851 * @return array Associative array with plural form number and plural rule as key-value pairs
4853 public function getPluralRules() {
4854 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4855 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4856 if ( !$pluralRules ) {
4857 foreach ( $fallbacks as $fallbackCode ) {
4858 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4859 if ( $pluralRules ) {
4864 return $pluralRules;
4868 * Get the plural rule types for the language
4870 * @return array Associative array with plural form number and plural rule type as key-value pairs
4872 public function getPluralRuleTypes() {
4873 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4874 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4875 if ( !$pluralRuleTypes ) {
4876 foreach ( $fallbacks as $fallbackCode ) {
4877 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4878 if ( $pluralRuleTypes ) {
4883 return $pluralRuleTypes;
4887 * Find the index number of the plural rule appropriate for the given number
4888 * @param int $number
4889 * @return int The index number of the plural rule
4891 public function getPluralRuleIndexNumber( $number ) {
4892 $pluralRules = $this->getCompiledPluralRules();
4893 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
4898 * Find the plural rule type appropriate for the given number
4899 * For example, if the language is set to Arabic, getPluralType(5) should
4902 * @param int $number
4903 * @return string The name of the plural rule type, e.g. one, two, few, many
4905 public function getPluralRuleType( $number ) {
4906 $index = $this->getPluralRuleIndexNumber( $number );
4907 $pluralRuleTypes = $this->getPluralRuleTypes();
4908 if ( isset( $pluralRuleTypes[$index] ) ) {
4909 return $pluralRuleTypes[$index];