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 = array(), $mMagicHookDone = false;
51 private $mHtmlCode = null, $mParentLanguage = false;
53 public $dateFormatStrings = array();
54 public $mExtendedSpecialPageAliases;
56 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
59 * ReplacementArray object caches
61 public $transformData = array();
64 * @var LocalisationCache
66 static public $dataCache;
68 static public $mLangObjCache = array();
70 static public $mWeekdayMsgs = array(
71 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
75 static public $mWeekdayAbbrevMsgs = array(
76 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
79 static public $mMonthMsgs = array(
80 'january', 'february', 'march', 'april', 'may_long', 'june',
81 'july', 'august', 'september', 'october', 'november',
84 static public $mMonthGenMsgs = array(
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 = array(
90 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
91 'sep', 'oct', 'nov', 'dec'
94 static public $mIranianCalendarMonthMsgs = array(
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 = array(
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 = array(
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 = array(
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 = array(
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 = array();
149 * Cache for language names
150 * @var MapCacheLRU|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( array( $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 // Protect against path traversal below
211 if ( !Language
::isValidCode( $code )
212 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code )
214 throw new MWException( "Invalid language code \"$code\"" );
217 if ( !Language
::isValidBuiltInCode( $code ) ) {
218 // It's not possible to customise this code with class files, so
219 // just return a Language object. This is to support uselang= hacks.
220 $lang = new Language
;
221 $lang->setCode( $code );
225 // Check if there is a language class for the code
226 $class = self
::classFromCode( $code );
227 self
::preloadLanguageClass( $class );
228 if ( class_exists( $class ) ) {
233 // Keep trying the fallback list until we find an existing class
234 $fallbacks = Language
::getFallbacksFor( $code );
235 foreach ( $fallbacks as $fallbackCode ) {
236 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
237 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
240 $class = self
::classFromCode( $fallbackCode );
241 self
::preloadLanguageClass( $class );
242 if ( class_exists( $class ) ) {
243 $lang = Language
::newFromCode( $fallbackCode );
244 $lang->setCode( $code );
249 throw new MWException( "Invalid fallback sequence for language '$code'" );
253 * Checks whether any localisation is available for that language tag
254 * in MediaWiki (MessagesXx.php exists).
256 * @param string $code Language tag (in lower case)
257 * @return bool Whether language is supported
260 public static function isSupportedLanguage( $code ) {
261 if ( !self
::isValidBuiltInCode( $code ) ) {
265 if ( $code === 'qqq' ) {
269 return is_readable( self
::getMessagesFileName( $code ) ) ||
270 is_readable( self
::getJsonMessagesFileName( $code ) );
274 * Returns true if a language code string is a well-formed language tag
275 * according to RFC 5646.
276 * This function only checks well-formedness; it doesn't check that
277 * language, script or variant codes actually exist in the repositories.
279 * Based on regexes by Mark Davis of the Unicode Consortium:
280 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
282 * @param string $code
283 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
288 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
291 $alphanum = '[a-z0-9]';
292 $x = 'x'; # private use singleton
293 $singleton = '[a-wy-z]'; # other singleton
294 $s = $lenient ?
'[-_]' : '-';
296 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
297 $script = "$alpha{4}"; # ISO 15924
298 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
299 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
300 $extension = "$singleton(?:$s$alphanum{2,8})+";
301 $privateUse = "$x(?:$s$alphanum{1,8})+";
303 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
304 # Since these are limited, this is safe even later changes to the registry --
305 # the only oddity is that it might change the type of the tag, and thus
306 # the results from the capturing groups.
307 # http://www.iana.org/assignments/language-subtag-registry
309 $grandfathered = "en{$s}GB{$s}oed"
310 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
311 . "|no{$s}(?:bok|nyn)"
312 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
313 . "|zh{$s}min{$s}nan";
315 $variantList = "$variant(?:$s$variant)*";
316 $extensionList = "$extension(?:$s$extension)*";
318 $langtag = "(?:($language)"
321 . "(?:$s$variantList)?"
322 . "(?:$s$extensionList)?"
323 . "(?:$s$privateUse)?)";
325 # The final breakdown, with capturing groups for each of these components
326 # The variants, extensions, grandfathered, and private-use may have interior '-'
328 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
330 return (bool)preg_match( "/$root/", strtolower( $code ) );
334 * Returns true if a language code string is of a valid form, whether or
335 * not it exists. This includes codes which are used solely for
336 * customisation via the MediaWiki namespace.
338 * @param string $code
342 public static function isValidCode( $code ) {
343 static $cache = array();
344 if ( isset( $cache[$code] ) ) {
345 return $cache[$code];
347 // People think language codes are html safe, so enforce it.
348 // Ideally we should only allow a-zA-Z0-9-
349 // but, .+ and other chars are often used for {{int:}} hacks
350 // see bugs 37564, 37587, 36938
352 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
353 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
355 return $cache[$code];
359 * Returns true if a language code is of a valid form for the purposes of
360 * internal customisation of MediaWiki, via Messages*.php or *.json.
362 * @param string $code
364 * @throws MWException
368 public static function isValidBuiltInCode( $code ) {
370 if ( !is_string( $code ) ) {
371 if ( is_object( $code ) ) {
372 $addmsg = " of class " . get_class( $code );
376 $type = gettype( $code );
377 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
380 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
384 * Returns true if a language code is an IETF tag known to MediaWiki.
391 public static function isKnownLanguageTag( $tag ) {
392 static $coreLanguageNames;
394 // Quick escape for invalid input to avoid exceptions down the line
395 // when code tries to process tags which are not valid at all.
396 if ( !self
::isValidBuiltInCode( $tag ) ) {
400 if ( $coreLanguageNames === null ) {
402 include "$IP/languages/Names.php";
405 if ( isset( $coreLanguageNames[$tag] )
406 || self
::fetchLanguageName( $tag, $tag ) !== ''
415 * @param string $code
416 * @return string Name of the language class
418 public static function classFromCode( $code ) {
419 if ( $code == 'en' ) {
422 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
427 * Includes language class files
429 * @param string $class Name of the language class
431 public static function preloadLanguageClass( $class ) {
434 if ( $class === 'Language' ) {
438 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
439 include_once "$IP/languages/classes/$class.php";
444 * Get the LocalisationCache instance
446 * @return LocalisationCache
448 public static function getLocalisationCache() {
449 if ( is_null( self
::$dataCache ) ) {
450 global $wgLocalisationCacheConf;
451 $class = $wgLocalisationCacheConf['class'];
452 self
::$dataCache = new $class( $wgLocalisationCacheConf );
454 return self
::$dataCache;
457 function __construct() {
458 $this->mConverter
= new FakeConverter( $this );
459 // Set the code to the name of the descendant
460 if ( get_class( $this ) == 'Language' ) {
463 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
465 self
::getLocalisationCache();
469 * Reduce memory usage
471 function __destruct() {
472 foreach ( $this as $name => $value ) {
473 unset( $this->$name );
478 * Hook which will be called if this is the content language.
479 * Descendants can use this to register hook functions or modify globals
481 function initContLang() {
488 function getFallbackLanguages() {
489 return self
::getFallbacksFor( $this->mCode
);
493 * Exports $wgBookstoreListEn
496 function getBookstoreList() {
497 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
501 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
502 * available in localised form, it will be included in English.
506 public function getNamespaces() {
507 if ( is_null( $this->namespaceNames
) ) {
508 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
510 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
511 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
513 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
515 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
516 if ( $wgMetaNamespaceTalk ) {
517 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
519 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
520 $this->namespaceNames
[NS_PROJECT_TALK
] =
521 $this->fixVariableInNamespace( $talk );
524 # Sometimes a language will be localised but not actually exist on this wiki.
525 foreach ( $this->namespaceNames
as $key => $text ) {
526 if ( !isset( $validNamespaces[$key] ) ) {
527 unset( $this->namespaceNames
[$key] );
531 # The above mixing may leave namespaces out of canonical order.
532 # Re-order by namespace ID number...
533 ksort( $this->namespaceNames
);
535 Hooks
::run( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
538 return $this->namespaceNames
;
542 * Arbitrarily set all of the namespace names at once. Mainly used for testing
543 * @param array $namespaces Array of namespaces (id => name)
545 public function setNamespaces( array $namespaces ) {
546 $this->namespaceNames
= $namespaces;
547 $this->mNamespaceIds
= null;
551 * Resets all of the namespace caches. Mainly used for testing
553 public function resetNamespaces() {
554 $this->namespaceNames
= null;
555 $this->mNamespaceIds
= null;
556 $this->namespaceAliases
= null;
560 * A convenience function that returns getNamespaces() with spaces instead of underscores
561 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
565 function getFormattedNamespaces() {
566 $ns = $this->getNamespaces();
567 foreach ( $ns as $k => $v ) {
568 $ns[$k] = strtr( $v, '_', ' ' );
574 * Get a namespace value by key
577 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
578 * echo $mw_ns; // prints 'MediaWiki'
581 * @param int $index The array key of the namespace to return
582 * @return string|bool String if the namespace value exists, otherwise false
584 function getNsText( $index ) {
585 $ns = $this->getNamespaces();
586 return isset( $ns[$index] ) ?
$ns[$index] : false;
590 * A convenience function that returns the same thing as
591 * getNsText() except with '_' changed to ' ', useful for
595 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
596 * echo $mw_ns; // prints 'MediaWiki talk'
599 * @param int $index The array key of the namespace to return
600 * @return string Namespace name without underscores (empty string if namespace does not exist)
602 function getFormattedNsText( $index ) {
603 $ns = $this->getNsText( $index );
604 return strtr( $ns, '_', ' ' );
608 * Returns gender-dependent namespace alias if available.
609 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
610 * @param int $index Namespace index
611 * @param string $gender Gender key (male, female... )
615 function getGenderNsText( $index, $gender ) {
616 global $wgExtraGenderNamespaces;
618 $ns = $wgExtraGenderNamespaces +
619 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
621 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
625 * Whether this language uses gender-dependent namespace aliases.
626 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
630 function needsGenderDistinction() {
631 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
632 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
633 // $wgExtraGenderNamespaces overrides everything
635 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
636 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
637 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
640 // Check what is in i18n files
641 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
642 return count( $aliases ) > 0;
647 * Get a namespace key by value, case insensitive.
648 * Only matches namespace names for the current language, not the
649 * canonical ones defined in Namespace.php.
651 * @param string $text
652 * @return int|bool An integer if $text is a valid value otherwise false
654 function getLocalNsIndex( $text ) {
655 $lctext = $this->lc( $text );
656 $ids = $this->getNamespaceIds();
657 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
663 function getNamespaceAliases() {
664 if ( is_null( $this->namespaceAliases
) ) {
665 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
669 foreach ( $aliases as $name => $index ) {
670 if ( $index === NS_PROJECT_TALK
) {
671 unset( $aliases[$name] );
672 $name = $this->fixVariableInNamespace( $name );
673 $aliases[$name] = $index;
678 global $wgExtraGenderNamespaces;
679 $genders = $wgExtraGenderNamespaces +
680 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
681 foreach ( $genders as $index => $forms ) {
682 foreach ( $forms as $alias ) {
683 $aliases[$alias] = $index;
687 # Also add converted namespace names as aliases, to avoid confusion.
688 $convertedNames = array();
689 foreach ( $this->getVariants() as $variant ) {
690 if ( $variant === $this->mCode
) {
693 foreach ( $this->getNamespaces() as $ns => $_ ) {
694 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
698 $this->namespaceAliases
= $aliases +
$convertedNames;
701 return $this->namespaceAliases
;
707 function getNamespaceIds() {
708 if ( is_null( $this->mNamespaceIds
) ) {
709 global $wgNamespaceAliases;
710 # Put namespace names and aliases into a hashtable.
711 # If this is too slow, then we should arrange it so that it is done
712 # before caching. The catch is that at pre-cache time, the above
713 # class-specific fixup hasn't been done.
714 $this->mNamespaceIds
= array();
715 foreach ( $this->getNamespaces() as $index => $name ) {
716 $this->mNamespaceIds
[$this->lc( $name )] = $index;
718 foreach ( $this->getNamespaceAliases() as $name => $index ) {
719 $this->mNamespaceIds
[$this->lc( $name )] = $index;
721 if ( $wgNamespaceAliases ) {
722 foreach ( $wgNamespaceAliases as $name => $index ) {
723 $this->mNamespaceIds
[$this->lc( $name )] = $index;
727 return $this->mNamespaceIds
;
731 * Get a namespace key by value, case insensitive. Canonical namespace
732 * names override custom ones defined for the current language.
734 * @param string $text
735 * @return int|bool An integer if $text is a valid value otherwise false
737 function getNsIndex( $text ) {
738 $lctext = $this->lc( $text );
739 $ns = MWNamespace
::getCanonicalIndex( $lctext );
740 if ( $ns !== null ) {
743 $ids = $this->getNamespaceIds();
744 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
748 * short names for language variants used for language conversion links.
750 * @param string $code
751 * @param bool $usemsg Use the "variantname-xyz" message if it exists
754 function getVariantname( $code, $usemsg = true ) {
755 $msg = "variantname-$code";
756 if ( $usemsg && wfMessage( $msg )->exists() ) {
757 return $this->getMessageFromDB( $msg );
759 $name = self
::fetchLanguageName( $code );
761 return $name; # if it's defined as a language name, show that
763 # otherwise, output the language code
769 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
770 * SpecialPageFactory::getLocalNameFor instead.
771 * @param string $name
774 function specialPage( $name ) {
775 $aliases = $this->getSpecialPageAliases();
776 if ( isset( $aliases[$name][0] ) ) {
777 $name = $aliases[$name][0];
779 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
785 function getDatePreferences() {
786 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
792 function getDateFormats() {
793 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
797 * @return array|string
799 function getDefaultDateFormat() {
800 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
801 if ( $df === 'dmy or mdy' ) {
802 global $wgAmericanDates;
803 return $wgAmericanDates ?
'mdy' : 'dmy';
812 function getDatePreferenceMigrationMap() {
813 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
817 * @param string $image
820 function getImageFile( $image ) {
821 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
828 function getImageFiles() {
829 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
835 function getExtraUserToggles() {
836 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
843 function getUserToggle( $tog ) {
844 return $this->getMessageFromDB( "tog-$tog" );
848 * Get native language names, indexed by code.
849 * Only those defined in MediaWiki, no other data like CLDR.
850 * If $customisedOnly is true, only returns codes with a messages file
852 * @param bool $customisedOnly
855 * @deprecated since 1.20, use fetchLanguageNames()
857 public static function getLanguageNames( $customisedOnly = false ) {
858 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
862 * Get translated language names. This is done on best effort and
863 * by default this is exactly the same as Language::getLanguageNames.
864 * The CLDR extension provides translated names.
865 * @param string $code Language code.
866 * @return array Language code => language name
868 * @deprecated since 1.20, use fetchLanguageNames()
870 public static function getTranslatedLanguageNames( $code ) {
871 return self
::fetchLanguageNames( $code, 'all' );
875 * Get an array of language names, indexed by code.
876 * @param null|string $inLanguage Code of language in which to return the names
877 * Use null for autonyms (native names)
878 * @param string $include One of:
879 * 'all' all available languages
880 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
881 * 'mwfile' only if the language is in 'mw' *and* has a message file
882 * @return array Language code => language name
885 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
886 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
887 $cacheKey .= ":$include";
888 if ( self
::$languageNameCache === null ) {
889 self
::$languageNameCache = new MapCacheLRU( 20 );
891 if ( self
::$languageNameCache->has( $cacheKey ) ) {
892 $ret = self
::$languageNameCache->get( $cacheKey );
894 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
895 self
::$languageNameCache->set( $cacheKey, $ret );
901 * Uncached helper for fetchLanguageNames
902 * @param null|string $inLanguage Code of language in which to return the names
903 * Use null for autonyms (native names)
904 * @param string $include One of:
905 * 'all' all available languages
906 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
907 * 'mwfile' only if the language is in 'mw' *and* has a message file
908 * @return array Language code => language name
910 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
911 global $wgExtraLanguageNames;
912 static $coreLanguageNames;
914 if ( $coreLanguageNames === null ) {
916 include "$IP/languages/Names.php";
919 // If passed an invalid language code to use, fallback to en
920 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
927 # TODO: also include when $inLanguage is null, when this code is more efficient
928 Hooks
::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
931 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
932 foreach ( $mwNames as $mwCode => $mwName ) {
933 # - Prefer own MediaWiki native name when not using the hook
934 # - For other names just add if not added through the hook
935 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
936 $names[$mwCode] = $mwName;
940 if ( $include === 'all' ) {
946 $coreCodes = array_keys( $mwNames );
947 foreach ( $coreCodes as $coreCode ) {
948 $returnMw[$coreCode] = $names[$coreCode];
951 if ( $include === 'mwfile' ) {
952 $namesMwFile = array();
953 # We do this using a foreach over the codes instead of a directory
954 # loop so that messages files in extensions will work correctly.
955 foreach ( $returnMw as $code => $value ) {
956 if ( is_readable( self
::getMessagesFileName( $code ) )
957 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
959 $namesMwFile[$code] = $names[$code];
963 ksort( $namesMwFile );
968 # 'mw' option; default if it's not one of the other two options (all/mwfile)
973 * @param string $code The code of the language for which to get the name
974 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
975 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
976 * @return string Language name or empty
979 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
980 $code = strtolower( $code );
981 $array = self
::fetchLanguageNames( $inLanguage, $include );
982 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
986 * Get a message from the MediaWiki namespace.
988 * @param string $msg Message name
991 function getMessageFromDB( $msg ) {
992 return $this->msg( $msg )->text();
996 * Get message object in this language. Only for use inside this class.
998 * @param string $msg Message name
1001 protected function msg( $msg ) {
1002 return wfMessage( $msg )->inLanguage( $this );
1006 * Get the native language name of $code.
1007 * Only if defined in MediaWiki, no other data like CLDR.
1008 * @param string $code
1010 * @deprecated since 1.20, use fetchLanguageName()
1012 function getLanguageName( $code ) {
1013 return self
::fetchLanguageName( $code );
1017 * @param string $key
1020 function getMonthName( $key ) {
1021 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
1027 function getMonthNamesArray() {
1028 $monthNames = array( '' );
1029 for ( $i = 1; $i < 13; $i++
) {
1030 $monthNames[] = $this->getMonthName( $i );
1036 * @param string $key
1039 function getMonthNameGen( $key ) {
1040 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1044 * @param string $key
1047 function getMonthAbbreviation( $key ) {
1048 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1054 function getMonthAbbreviationsArray() {
1055 $monthNames = array( '' );
1056 for ( $i = 1; $i < 13; $i++
) {
1057 $monthNames[] = $this->getMonthAbbreviation( $i );
1063 * @param string $key
1066 function getWeekdayName( $key ) {
1067 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1071 * @param string $key
1074 function getWeekdayAbbreviation( $key ) {
1075 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1079 * @param string $key
1082 function getIranianCalendarMonthName( $key ) {
1083 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1087 * @param string $key
1090 function getHebrewCalendarMonthName( $key ) {
1091 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1095 * @param string $key
1098 function getHebrewCalendarMonthNameGen( $key ) {
1099 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1103 * @param string $key
1106 function getHijriCalendarMonthName( $key ) {
1107 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1111 * Pass through result from $dateTimeObj->format()
1112 * @param DateTime|bool|null &$dateTimeObj
1114 * @param DateTimeZone|bool|null $zone
1115 * @param string $code
1118 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1119 if ( !$dateTimeObj ) {
1120 $dateTimeObj = DateTime
::createFromFormat(
1121 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1124 return $dateTimeObj->format( $code );
1128 * This is a workalike of PHP's date() function, but with better
1129 * internationalisation, a reduced set of format characters, and a better
1132 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1133 * the PHP manual for definitions. There are a number of extensions, which
1136 * xn Do not translate digits of the next numeric format character
1137 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1138 * xr Use roman numerals for the next numeric format character
1139 * xh Use hebrew numerals for the next numeric format character
1141 * xg Genitive month name
1143 * xij j (day number) in Iranian calendar
1144 * xiF F (month name) in Iranian calendar
1145 * xin n (month number) in Iranian calendar
1146 * xiy y (two digit year) in Iranian calendar
1147 * xiY Y (full year) in Iranian calendar
1149 * xjj j (day number) in Hebrew calendar
1150 * xjF F (month name) in Hebrew calendar
1151 * xjt t (days in month) in Hebrew calendar
1152 * xjx xg (genitive month name) in Hebrew calendar
1153 * xjn n (month number) in Hebrew calendar
1154 * xjY Y (full year) in Hebrew calendar
1156 * xmj j (day number) in Hijri calendar
1157 * xmF F (month name) in Hijri calendar
1158 * xmn n (month number) in Hijri calendar
1159 * xmY Y (full year) in Hijri calendar
1161 * xkY Y (full year) in Thai solar calendar. Months and days are
1162 * identical to the Gregorian calendar
1163 * xoY Y (full year) in Minguo calendar or Juche year.
1164 * Months and days are identical to the
1165 * Gregorian calendar
1166 * xtY Y (full year) in Japanese nengo. Months and days are
1167 * identical to the Gregorian calendar
1169 * Characters enclosed in double quotes will be considered literal (with
1170 * the quotes themselves removed). Unmatched quotes will be considered
1171 * literal quotes. Example:
1173 * "The month is" F => The month is January
1176 * Backslash escaping is also supported.
1178 * Input timestamp is assumed to be pre-normalized to the desired local
1179 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1180 * $ts is UTC if $zone is not given.
1182 * @param string $format
1183 * @param string $ts 14-character timestamp
1186 * @param DateTimeZone $zone Timezone of $ts
1187 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1188 * Only makes sense if $ts is the current time.
1189 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1191 * @throws MWException
1194 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1199 $dateTimeObj = false;
1208 $usedSecond = false;
1209 $usedMinute = false;
1216 $usedISOYear = false;
1217 $usedIsLeapYear = false;
1219 $usedHebrewMonth = false;
1220 $usedIranianMonth = false;
1221 $usedHijriMonth = false;
1222 $usedHebrewYear = false;
1223 $usedIranianYear = false;
1224 $usedHijriYear = false;
1225 $usedTennoYear = false;
1227 if ( strlen( $ts ) !== 14 ) {
1228 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1231 if ( !ctype_digit( $ts ) ) {
1232 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1235 $formatLength = strlen( $format );
1236 for ( $p = 0; $p < $formatLength; $p++
) {
1238 $code = $format[$p];
1239 if ( $code == 'x' && $p < $formatLength - 1 ) {
1240 $code .= $format[++
$p];
1243 if ( ( $code === 'xi'
1249 && $p < $formatLength - 1 ) {
1250 $code .= $format[++
$p];
1261 $rawToggle = !$rawToggle;
1271 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1274 $usedHebrewMonth = true;
1276 $hebrew = self
::tsToHebrew( $ts );
1278 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1282 $num = substr( $ts, 6, 2 );
1286 $s .= $this->getWeekdayAbbreviation(
1287 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1292 $num = intval( substr( $ts, 6, 2 ) );
1297 $iranian = self
::tsToIranian( $ts );
1304 $hijri = self
::tsToHijri( $ts );
1311 $hebrew = self
::tsToHebrew( $ts );
1317 $s .= $this->getWeekdayName(
1318 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1323 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1326 $usedIranianMonth = true;
1328 $iranian = self
::tsToIranian( $ts );
1330 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1333 $usedHijriMonth = true;
1335 $hijri = self
::tsToHijri( $ts );
1337 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1340 $usedHebrewMonth = true;
1342 $hebrew = self
::tsToHebrew( $ts );
1344 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1348 $num = substr( $ts, 4, 2 );
1352 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1356 $num = intval( substr( $ts, 4, 2 ) );
1359 $usedIranianMonth = true;
1361 $iranian = self
::tsToIranian( $ts );
1366 $usedHijriMonth = true;
1368 $hijri = self
::tsToHijri( $ts );
1373 $usedHebrewMonth = true;
1375 $hebrew = self
::tsToHebrew( $ts );
1380 $usedHebrewMonth = true;
1382 $hebrew = self
::tsToHebrew( $ts );
1388 $num = substr( $ts, 0, 4 );
1391 $usedIranianYear = true;
1393 $iranian = self
::tsToIranian( $ts );
1398 $usedHijriYear = true;
1400 $hijri = self
::tsToHijri( $ts );
1405 $usedHebrewYear = true;
1407 $hebrew = self
::tsToHebrew( $ts );
1414 $thai = self
::tsToYear( $ts, 'thai' );
1421 $minguo = self
::tsToYear( $ts, 'minguo' );
1426 $usedTennoYear = true;
1428 $tenno = self
::tsToYear( $ts, 'tenno' );
1434 $num = substr( $ts, 2, 2 );
1437 $usedIranianYear = true;
1439 $iranian = self
::tsToIranian( $ts );
1441 $num = substr( $iranian[0], -2 );
1445 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1449 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1453 $h = substr( $ts, 8, 2 );
1454 $num = $h %
12 ?
$h %
12 : 12;
1458 $num = intval( substr( $ts, 8, 2 ) );
1462 $h = substr( $ts, 8, 2 );
1463 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1467 $num = substr( $ts, 8, 2 );
1471 $num = substr( $ts, 10, 2 );
1475 $num = substr( $ts, 12, 2 );
1485 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1491 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1495 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1499 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1502 $usedIsLeapYear = true;
1503 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1506 $usedISOYear = true;
1507 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1514 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1517 # Backslash escaping
1518 if ( $p < $formatLength - 1 ) {
1519 $s .= $format[++
$p];
1526 if ( $p < $formatLength - 1 ) {
1527 $endQuote = strpos( $format, '"', $p +
1 );
1528 if ( $endQuote === false ) {
1529 # No terminating quote, assume literal "
1532 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1536 # Quote at end of string, assume literal "
1543 if ( $num !== false ) {
1544 if ( $rawToggle ||
$raw ) {
1547 } elseif ( $roman ) {
1548 $s .= Language
::romanNumeral( $num );
1550 } elseif ( $hebrewNum ) {
1551 $s .= self
::hebrewNumeral( $num );
1554 $s .= $this->formatNum( $num, true );
1559 if ( $usedSecond ) {
1561 } elseif ( $usedMinute ) {
1562 $ttl = 60 - substr( $ts, 12, 2 );
1563 } elseif ( $usedHour ) {
1564 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1565 } elseif ( $usedAMPM ) {
1566 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1567 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1571 $usedIranianMonth ||
1578 // @todo Someone who understands the non-Gregorian calendars
1579 // should write proper logic for them so that they don't need purged every day.
1580 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1581 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1583 $possibleTtls = array();
1584 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1585 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1588 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1589 $timeRemainingInDay;
1590 } elseif ( $usedISOYear ) {
1591 // December 28th falls on the last ISO week of the year, every year.
1592 // The last ISO week of a year can be 52 or 53.
1593 $lastWeekOfISOYear = DateTime
::createFromFormat(
1595 substr( $ts, 0, 4 ) . '1228',
1596 $zone ?
: new DateTimeZone( 'UTC' )
1598 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1599 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1600 $timeRemainingInWeek =
1601 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1602 +
$timeRemainingInDay;
1603 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1608 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1609 substr( $ts, 6, 2 ) ) * 86400
1610 +
$timeRemainingInDay;
1611 } elseif ( $usedYear ) {
1613 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1614 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1615 +
$timeRemainingInDay;
1616 } elseif ( $usedIsLeapYear ) {
1617 $year = substr( $ts, 0, 4 );
1618 $timeRemainingInYear =
1619 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1620 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1621 +
$timeRemainingInDay;
1623 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1624 // this isn't a leap year. see when the next one starts
1625 $nextCandidate = $year - $mod +
4;
1626 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1627 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1628 $timeRemainingInYear;
1630 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1631 $timeRemainingInYear;
1634 // this is a leap year, so the next year isn't
1635 $possibleTtls[] = $timeRemainingInYear;
1639 if ( $possibleTtls ) {
1640 $ttl = min( $possibleTtls );
1647 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1648 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1651 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1652 * Gregorian dates to Iranian dates. Originally written in C, it
1653 * is released under the terms of GNU Lesser General Public
1654 * License. Conversion to PHP was performed by Niklas Laxström.
1656 * Link: http://www.farsiweb.info/jalali/jalali.c
1662 private static function tsToIranian( $ts ) {
1663 $gy = substr( $ts, 0, 4 ) -1600;
1664 $gm = substr( $ts, 4, 2 ) -1;
1665 $gd = substr( $ts, 6, 2 ) -1;
1667 # Days passed from the beginning (including leap years)
1669 +
floor( ( $gy +
3 ) / 4 )
1670 - floor( ( $gy +
99 ) / 100 )
1671 +
floor( ( $gy +
399 ) / 400 );
1673 // Add days of the past months of this year
1674 for ( $i = 0; $i < $gm; $i++
) {
1675 $gDayNo +
= self
::$GREG_DAYS[$i];
1679 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1683 // Days passed in current month
1684 $gDayNo +
= (int)$gd;
1686 $jDayNo = $gDayNo - 79;
1688 $jNp = floor( $jDayNo / 12053 );
1691 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1694 if ( $jDayNo >= 366 ) {
1695 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1696 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1699 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1700 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1706 return array( $jy, $jm, $jd );
1710 * Converting Gregorian dates to Hijri dates.
1712 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1714 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1720 private static function tsToHijri( $ts ) {
1721 $year = substr( $ts, 0, 4 );
1722 $month = substr( $ts, 4, 2 );
1723 $day = substr( $ts, 6, 2 );
1731 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1732 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1734 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1735 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1736 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1739 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1740 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1743 $zl = $zjd -1948440 +
10632;
1744 $zn = (int)( ( $zl - 1 ) / 10631 );
1745 $zl = $zl - 10631 * $zn +
354;
1746 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1747 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1748 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1749 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1750 $zm = (int)( ( 24 * $zl ) / 709 );
1751 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1752 $zy = 30 * $zn +
$zj - 30;
1754 return array( $zy, $zm, $zd );
1758 * Converting Gregorian dates to Hebrew dates.
1760 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1761 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1762 * to translate the relevant functions into PHP and release them under
1765 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1766 * and Adar II is 14. In a non-leap year, Adar is 6.
1772 private static function tsToHebrew( $ts ) {
1774 $year = substr( $ts, 0, 4 );
1775 $month = substr( $ts, 4, 2 );
1776 $day = substr( $ts, 6, 2 );
1778 # Calculate Hebrew year
1779 $hebrewYear = $year +
3760;
1781 # Month number when September = 1, August = 12
1783 if ( $month > 12 ) {
1790 # Calculate day of year from 1 September
1792 for ( $i = 1; $i < $month; $i++
) {
1796 # Check if the year is leap
1797 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1800 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1807 # Calculate the start of the Hebrew year
1808 $start = self
::hebrewYearStart( $hebrewYear );
1810 # Calculate next year's start
1811 if ( $dayOfYear <= $start ) {
1812 # Day is before the start of the year - it is the previous year
1814 $nextStart = $start;
1818 # Add days since previous year's 1 September
1820 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1824 # Start of the new (previous) year
1825 $start = self
::hebrewYearStart( $hebrewYear );
1828 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1831 # Calculate Hebrew day of year
1832 $hebrewDayOfYear = $dayOfYear - $start;
1834 # Difference between year's days
1835 $diff = $nextStart - $start;
1836 # Add 12 (or 13 for leap years) days to ignore the difference between
1837 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1838 # difference is only about the year type
1839 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1845 # Check the year pattern, and is leap year
1846 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1847 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1848 # and non-leap years
1849 $yearPattern = $diff %
30;
1850 # Check if leap year
1851 $isLeap = $diff >= 30;
1853 # Calculate day in the month from number of day in the Hebrew year
1854 # Don't check Adar - if the day is not in Adar, we will stop before;
1855 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1856 $hebrewDay = $hebrewDayOfYear;
1859 while ( $hebrewMonth <= 12 ) {
1860 # Calculate days in this month
1861 if ( $isLeap && $hebrewMonth == 6 ) {
1862 # Adar in a leap year
1864 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1866 if ( $hebrewDay <= $days ) {
1870 # Subtract the days of Adar I
1871 $hebrewDay -= $days;
1874 if ( $hebrewDay <= $days ) {
1880 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1881 # Cheshvan in a complete year (otherwise as the rule below)
1883 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1884 # Kislev in an incomplete year (otherwise as the rule below)
1887 # Odd months have 30 days, even have 29
1888 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1890 if ( $hebrewDay <= $days ) {
1891 # In the current month
1894 # Subtract the days of the current month
1895 $hebrewDay -= $days;
1896 # Try in the next month
1901 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1905 * This calculates the Hebrew year start, as days since 1 September.
1906 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1907 * Used for Hebrew date.
1913 private static function hebrewYearStart( $year ) {
1914 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1915 $b = intval( ( $year - 1 ) %
4 );
1916 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1920 $Mar = intval( $m );
1926 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1927 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1929 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1931 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1935 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1940 * Algorithm to convert Gregorian dates to Thai solar dates,
1941 * Minguo dates or Minguo dates.
1943 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1944 * http://en.wikipedia.org/wiki/Minguo_calendar
1945 * http://en.wikipedia.org/wiki/Japanese_era_name
1947 * @param string $ts 14-character timestamp
1948 * @param string $cName Calender name
1949 * @return array Converted year, month, day
1951 private static function tsToYear( $ts, $cName ) {
1952 $gy = substr( $ts, 0, 4 );
1953 $gm = substr( $ts, 4, 2 );
1954 $gd = substr( $ts, 6, 2 );
1956 if ( !strcmp( $cName, 'thai' ) ) {
1958 # Add 543 years to the Gregorian calendar
1959 # Months and days are identical
1960 $gy_offset = $gy +
543;
1961 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1963 # Deduct 1911 years from the Gregorian calendar
1964 # Months and days are identical
1965 $gy_offset = $gy - 1911;
1966 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1967 # Nengō dates up to Meiji period
1968 # Deduct years from the Gregorian calendar
1969 # depending on the nengo periods
1970 # Months and days are identical
1972 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1973 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1976 $gy_gannen = $gy - 1868 +
1;
1977 $gy_offset = $gy_gannen;
1978 if ( $gy_gannen == 1 ) {
1981 $gy_offset = '明治' . $gy_offset;
1983 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1984 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1985 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1986 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1987 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1990 $gy_gannen = $gy - 1912 +
1;
1991 $gy_offset = $gy_gannen;
1992 if ( $gy_gannen == 1 ) {
1995 $gy_offset = '大正' . $gy_offset;
1997 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1998 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1999 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
2002 $gy_gannen = $gy - 1926 +
1;
2003 $gy_offset = $gy_gannen;
2004 if ( $gy_gannen == 1 ) {
2007 $gy_offset = '昭和' . $gy_offset;
2010 $gy_gannen = $gy - 1989 +
1;
2011 $gy_offset = $gy_gannen;
2012 if ( $gy_gannen == 1 ) {
2015 $gy_offset = '平成' . $gy_offset;
2021 return array( $gy_offset, $gm, $gd );
2025 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2027 * This is the rule the BIDI algorithm uses to determine the directionality of
2028 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
2029 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2031 * TODO: Does not handle BIDI control characters inside the text.
2032 * TODO: Does not handle unallocated characters.
2034 * @param string $text Text to test
2035 * @return null|string Directionality ('ltr' or 'rtl') or null
2037 private static function strongDirFromContent( $text = '' ) {
2038 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2041 if ( $matches[1] === '' ) {
2048 * Roman number formatting up to 10000
2054 static function romanNumeral( $num ) {
2055 static $table = array(
2056 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
2057 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
2058 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
2059 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2060 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
2063 $num = intval( $num );
2064 if ( $num > 10000 ||
$num <= 0 ) {
2069 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2070 if ( $num >= $pow10 ) {
2071 $s .= $table[$i][(int)floor( $num / $pow10 )];
2073 $num = $num %
$pow10;
2079 * Hebrew Gematria number formatting up to 9999
2085 static function hebrewNumeral( $num ) {
2086 static $table = array(
2087 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2088 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2098 array( 'ת', 'ת', 'ק' ),
2099 array( 'ת', 'ת', 'ר' ),
2101 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2104 $num = intval( $num );
2105 if ( $num > 9999 ||
$num <= 0 ) {
2109 // Round thousands have special notations
2110 if ( $num === 1000 ) {
2112 } elseif ( $num %
1000 === 0 ) {
2113 return $table[0][$num / 1000] . "' אלפים";
2118 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2119 if ( $num >= $pow10 ) {
2120 if ( $num === 15 ||
$num === 16 ) {
2121 $letters[] = $table[0][9];
2122 $letters[] = $table[0][$num - 9];
2125 $letters = array_merge(
2127 (array)$table[$i][intval( $num / $pow10 )]
2130 if ( $pow10 === 1000 ) {
2136 $num = $num %
$pow10;
2139 $preTransformLength = count( $letters );
2140 if ( $preTransformLength === 1 ) {
2141 // Add geresh (single quote) to one-letter numbers
2144 $lastIndex = $preTransformLength - 1;
2145 $letters[$lastIndex] = str_replace(
2146 array( 'כ', 'מ', 'נ', 'פ', 'צ' ),
2147 array( 'ך', 'ם', 'ן', 'ף', 'ץ' ),
2148 $letters[$lastIndex]
2151 // Add gershayim (double quote) to multiple-letter numbers,
2152 // but exclude numbers with only one letter after the thousands
2153 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2154 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2157 array_splice( $letters, -1, 0, '"' );
2161 return implode( $letters );
2165 * Used by date() and time() to adjust the time output.
2167 * @param string $ts The time in date('YmdHis') format
2168 * @param mixed $tz Adjust the time by this amount (default false, mean we
2169 * get user timecorrection setting)
2172 function userAdjust( $ts, $tz = false ) {
2173 global $wgUser, $wgLocalTZoffset;
2175 if ( $tz === false ) {
2176 $tz = $wgUser->getOption( 'timecorrection' );
2179 $data = explode( '|', $tz, 3 );
2181 if ( $data[0] == 'ZoneInfo' ) {
2182 MediaWiki\
suppressWarnings();
2183 $userTZ = timezone_open( $data[2] );
2184 MediaWiki\restoreWarnings
();
2185 if ( $userTZ !== false ) {
2186 $date = date_create( $ts, timezone_open( 'UTC' ) );
2187 date_timezone_set( $date, $userTZ );
2188 $date = date_format( $date, 'YmdHis' );
2191 # Unrecognized timezone, default to 'Offset' with the stored offset.
2192 $data[0] = 'Offset';
2195 if ( $data[0] == 'System' ||
$tz == '' ) {
2196 # Global offset in minutes.
2197 $minDiff = $wgLocalTZoffset;
2198 } elseif ( $data[0] == 'Offset' ) {
2199 $minDiff = intval( $data[1] );
2201 $data = explode( ':', $tz );
2202 if ( count( $data ) == 2 ) {
2203 $data[0] = intval( $data[0] );
2204 $data[1] = intval( $data[1] );
2205 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2206 if ( $data[0] < 0 ) {
2207 $minDiff = -$minDiff;
2210 $minDiff = intval( $data[0] ) * 60;
2214 # No difference ? Return time unchanged
2215 if ( 0 == $minDiff ) {
2219 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2220 # Generate an adjusted date; take advantage of the fact that mktime
2221 # will normalize out-of-range values so we don't have to split $minDiff
2222 # into hours and minutes.
2224 (int)substr( $ts, 8, 2 ) ), # Hours
2225 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2226 (int)substr( $ts, 12, 2 ), # Seconds
2227 (int)substr( $ts, 4, 2 ), # Month
2228 (int)substr( $ts, 6, 2 ), # Day
2229 (int)substr( $ts, 0, 4 ) ); # Year
2231 $date = date( 'YmdHis', $t );
2232 MediaWiki\restoreWarnings
();
2238 * This is meant to be used by time(), date(), and timeanddate() to get
2239 * the date preference they're supposed to use, it should be used in
2243 * function timeanddate([...], $format = true) {
2244 * $datePreference = $this->dateFormat($format);
2249 * @param int|string|bool $usePrefs If true, the user's preference is used
2250 * if false, the site/language default is used
2251 * if int/string, assumed to be a format.
2254 function dateFormat( $usePrefs = true ) {
2257 if ( is_bool( $usePrefs ) ) {
2259 $datePreference = $wgUser->getDatePreference();
2261 $datePreference = (string)User
::getDefaultOption( 'date' );
2264 $datePreference = (string)$usePrefs;
2268 if ( $datePreference == '' ) {
2272 return $datePreference;
2276 * Get a format string for a given type and preference
2277 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2278 * @param string $pref The format name as it appears in Messages*.php under
2281 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2285 function getDateFormatString( $type, $pref ) {
2286 $wasDefault = false;
2287 if ( $pref == 'default' ) {
2289 $pref = $this->getDefaultDateFormat();
2292 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2293 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2295 if ( $type === 'pretty' && $df === null ) {
2296 $df = $this->getDateFormatString( 'date', $pref );
2299 if ( !$wasDefault && $df === null ) {
2300 $pref = $this->getDefaultDateFormat();
2301 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2304 $this->dateFormatStrings
[$type][$pref] = $df;
2306 return $this->dateFormatStrings
[$type][$pref];
2310 * @param string $ts The time format which needs to be turned into a
2311 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2312 * @param bool $adj Whether to adjust the time output according to the
2313 * user configured offset ($timecorrection)
2314 * @param mixed $format True to use user's date format preference
2315 * @param string|bool $timecorrection The time offset as returned by
2316 * validateTimeZone() in Special:Preferences
2319 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2320 $ts = wfTimestamp( TS_MW
, $ts );
2322 $ts = $this->userAdjust( $ts, $timecorrection );
2324 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2325 return $this->sprintfDate( $df, $ts );
2329 * @param string $ts The time format which needs to be turned into a
2330 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2331 * @param bool $adj Whether to adjust the time output according to the
2332 * user configured offset ($timecorrection)
2333 * @param mixed $format True to use user's date format preference
2334 * @param string|bool $timecorrection The time offset as returned by
2335 * validateTimeZone() in Special:Preferences
2338 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2339 $ts = wfTimestamp( TS_MW
, $ts );
2341 $ts = $this->userAdjust( $ts, $timecorrection );
2343 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2344 return $this->sprintfDate( $df, $ts );
2348 * @param string $ts The time format which needs to be turned into a
2349 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2350 * @param bool $adj Whether to adjust the time output according to the
2351 * user configured offset ($timecorrection)
2352 * @param mixed $format What format to return, if it's false output the
2353 * default one (default true)
2354 * @param string|bool $timecorrection The time offset as returned by
2355 * validateTimeZone() in Special:Preferences
2358 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2359 $ts = wfTimestamp( TS_MW
, $ts );
2361 $ts = $this->userAdjust( $ts, $timecorrection );
2363 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2364 return $this->sprintfDate( $df, $ts );
2368 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2372 * @param int $seconds The amount of seconds.
2373 * @param array $chosenIntervals The intervals to enable.
2377 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2378 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2380 $segments = array();
2382 foreach ( $intervals as $intervalName => $intervalValue ) {
2383 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2384 // duration-years, duration-decades, duration-centuries, duration-millennia
2385 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2386 $segments[] = $message->inLanguage( $this )->escaped();
2389 return $this->listToText( $segments );
2393 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2394 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2398 * @param int $seconds The amount of seconds.
2399 * @param array $chosenIntervals The intervals to enable.
2403 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2404 if ( empty( $chosenIntervals ) ) {
2405 $chosenIntervals = array(
2417 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2418 $sortedNames = array_keys( $intervals );
2419 $smallestInterval = array_pop( $sortedNames );
2421 $segments = array();
2423 foreach ( $intervals as $name => $length ) {
2424 $value = floor( $seconds / $length );
2426 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2427 $seconds -= $value * $length;
2428 $segments[$name] = $value;
2436 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2438 * @param string $type Can be 'date', 'time' or 'both'
2439 * @param string $ts The time format which needs to be turned into a
2440 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2441 * @param User $user User object used to get preferences for timezone and format
2442 * @param array $options Array, can contain the following keys:
2443 * - 'timecorrection': time correction, can have the following values:
2444 * - true: use user's preference
2445 * - false: don't use time correction
2446 * - int: value of time correction in minutes
2447 * - 'format': format to use, can have the following values:
2448 * - true: use user's preference
2449 * - false: use default preference
2450 * - string: format to use
2454 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2455 $ts = wfTimestamp( TS_MW
, $ts );
2456 $options +
= array( 'timecorrection' => true, 'format' => true );
2457 if ( $options['timecorrection'] !== false ) {
2458 if ( $options['timecorrection'] === true ) {
2459 $offset = $user->getOption( 'timecorrection' );
2461 $offset = $options['timecorrection'];
2463 $ts = $this->userAdjust( $ts, $offset );
2465 if ( $options['format'] === true ) {
2466 $format = $user->getDatePreference();
2468 $format = $options['format'];
2470 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2471 return $this->sprintfDate( $df, $ts );
2475 * Get the formatted date for the given timestamp and formatted for
2478 * @param mixed $ts Mixed: the time format which needs to be turned into a
2479 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2480 * @param User $user User object used to get preferences for timezone and format
2481 * @param array $options Array, can contain the following keys:
2482 * - 'timecorrection': time correction, can have the following values:
2483 * - true: use user's preference
2484 * - false: don't use time correction
2485 * - int: value of time correction in minutes
2486 * - 'format': format to use, can have the following values:
2487 * - true: use user's preference
2488 * - false: use default preference
2489 * - string: format to use
2493 public function userDate( $ts, User
$user, array $options = array() ) {
2494 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2498 * Get the formatted time for the given timestamp and formatted for
2501 * @param mixed $ts The time format which needs to be turned into a
2502 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2503 * @param User $user User object used to get preferences for timezone and format
2504 * @param array $options Array, can contain the following keys:
2505 * - 'timecorrection': time correction, can have the following values:
2506 * - true: use user's preference
2507 * - false: don't use time correction
2508 * - int: value of time correction in minutes
2509 * - 'format': format to use, can have the following values:
2510 * - true: use user's preference
2511 * - false: use default preference
2512 * - string: format to use
2516 public function userTime( $ts, User
$user, array $options = array() ) {
2517 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2521 * Get the formatted date and time for the given timestamp and formatted for
2524 * @param mixed $ts The time format which needs to be turned into a
2525 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2526 * @param User $user User object used to get preferences for timezone and format
2527 * @param array $options Array, can contain the following keys:
2528 * - 'timecorrection': time correction, can have the following values:
2529 * - true: use user's preference
2530 * - false: don't use time correction
2531 * - int: value of time correction in minutes
2532 * - 'format': format to use, can have the following values:
2533 * - true: use user's preference
2534 * - false: use default preference
2535 * - string: format to use
2539 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2540 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2544 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2546 * Determine the difference between the timestamp and the current time, and
2547 * generate a readable timestamp by returning "<N> <units> ago", where the
2548 * largest possible unit is used.
2550 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2552 * @param MWTimestamp $time
2553 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2554 * @param User|null $user User the timestamp is being generated for (or null to use main context's user)
2555 * @return string Formatted timestamp
2557 public function getHumanTimestamp( MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null ) {
2558 if ( $relativeTo === null ) {
2559 $relativeTo = new MWTimestamp();
2561 if ( $user === null ) {
2562 $user = RequestContext
::getMain()->getUser();
2565 // Adjust for the user's timezone.
2566 $offsetThis = $time->offsetForUser( $user );
2567 $offsetRel = $relativeTo->offsetForUser( $user );
2570 if ( Hooks
::run( 'GetHumanTimestamp', array( &$ts, $time, $relativeTo, $user, $this ) ) ) {
2571 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2574 // Reset the timezone on the objects.
2575 $time->timestamp
->sub( $offsetThis );
2576 $relativeTo->timestamp
->sub( $offsetRel );
2582 * Convert an MWTimestamp into a pretty human-readable timestamp using
2583 * the given user preferences and relative base time.
2585 * @see Language::getHumanTimestamp
2586 * @param MWTimestamp $ts Timestamp to prettify
2587 * @param MWTimestamp $relativeTo Base timestamp
2588 * @param User $user User preferences to use
2589 * @return string Human timestamp
2592 private function getHumanTimestampInternal( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2593 $diff = $ts->diff( $relativeTo );
2594 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2595 (int)$relativeTo->timestamp
->format( 'w' ) );
2596 $days = $diff->days ?
: (int)$diffDay;
2597 if ( $diff->invert ||
$days > 5
2598 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2600 // Timestamps are in different years: use full timestamp
2601 // Also do full timestamp for future dates
2603 * @todo FIXME: Add better handling of future timestamps.
2605 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2606 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2607 } elseif ( $days > 5 ) {
2608 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2609 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2610 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2611 } elseif ( $days > 1 ) {
2612 // Timestamp within the past week: show the day of the week and time
2613 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2614 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2616 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2617 $ts = wfMessage( "$weekday-at" )
2618 ->inLanguage( $this )
2619 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2621 } elseif ( $days == 1 ) {
2622 // Timestamp was yesterday: say 'yesterday' and the time.
2623 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2624 $ts = wfMessage( 'yesterday-at' )
2625 ->inLanguage( $this )
2626 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2628 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2629 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2630 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2631 $ts = wfMessage( 'today-at' )
2632 ->inLanguage( $this )
2633 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2636 // From here on in, the timestamp was soon enough ago so that we can simply say
2637 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2638 } elseif ( $diff->h
== 1 ) {
2639 // Less than 90 minutes, but more than an hour ago.
2640 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2641 } elseif ( $diff->i
>= 1 ) {
2642 // A few minutes ago.
2643 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2644 } elseif ( $diff->s
>= 30 ) {
2645 // Less than a minute, but more than 30 sec ago.
2646 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2648 // Less than 30 seconds ago.
2649 $ts = wfMessage( 'just-now' )->text();
2656 * @param string $key
2657 * @return array|null
2659 function getMessage( $key ) {
2660 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2666 function getAllMessages() {
2667 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2672 * @param string $out
2673 * @param string $string
2676 function iconv( $in, $out, $string ) {
2677 # This is a wrapper for iconv in all languages except esperanto,
2678 # which does some nasty x-conversions beforehand
2680 # Even with //IGNORE iconv can whine about illegal characters in
2681 # *input* string. We just ignore those too.
2682 # REF: http://bugs.php.net/bug.php?id=37166
2683 # REF: https://phabricator.wikimedia.org/T18885
2684 MediaWiki\
suppressWarnings();
2685 $text = iconv( $in, $out . '//IGNORE', $string );
2686 MediaWiki\restoreWarnings
();
2690 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2693 * @param array $matches
2694 * @return mixed|string
2696 function ucwordbreaksCallbackAscii( $matches ) {
2697 return $this->ucfirst( $matches[1] );
2701 * @param array $matches
2704 function ucwordbreaksCallbackMB( $matches ) {
2705 return mb_strtoupper( $matches[0] );
2709 * @param array $matches
2712 function ucCallback( $matches ) {
2713 list( $wikiUpperChars ) = self
::getCaseMaps();
2714 return strtr( $matches[1], $wikiUpperChars );
2718 * @param array $matches
2721 function lcCallback( $matches ) {
2722 list( , $wikiLowerChars ) = self
::getCaseMaps();
2723 return strtr( $matches[1], $wikiLowerChars );
2727 * @param array $matches
2730 function ucwordsCallbackMB( $matches ) {
2731 return mb_strtoupper( $matches[0] );
2735 * @param array $matches
2738 function ucwordsCallbackWiki( $matches ) {
2739 list( $wikiUpperChars ) = self
::getCaseMaps();
2740 return strtr( $matches[0], $wikiUpperChars );
2744 * Make a string's first character uppercase
2746 * @param string $str
2750 function ucfirst( $str ) {
2752 if ( $o < 96 ) { // if already uppercase...
2754 } elseif ( $o < 128 ) {
2755 return ucfirst( $str ); // use PHP's ucfirst()
2757 // fall back to more complex logic in case of multibyte strings
2758 return $this->uc( $str, true );
2763 * Convert a string to uppercase
2765 * @param string $str
2766 * @param bool $first
2770 function uc( $str, $first = false ) {
2771 if ( function_exists( 'mb_strtoupper' ) ) {
2773 if ( $this->isMultibyte( $str ) ) {
2774 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2776 return ucfirst( $str );
2779 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2782 if ( $this->isMultibyte( $str ) ) {
2783 $x = $first ?
'^' : '';
2784 return preg_replace_callback(
2785 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2786 array( $this, 'ucCallback' ),
2790 return $first ?
ucfirst( $str ) : strtoupper( $str );
2796 * @param string $str
2797 * @return mixed|string
2799 function lcfirst( $str ) {
2802 return strval( $str );
2803 } elseif ( $o >= 128 ) {
2804 return $this->lc( $str, true );
2805 } elseif ( $o > 96 ) {
2808 $str[0] = strtolower( $str[0] );
2814 * @param string $str
2815 * @param bool $first
2816 * @return mixed|string
2818 function lc( $str, $first = false ) {
2819 if ( function_exists( 'mb_strtolower' ) ) {
2821 if ( $this->isMultibyte( $str ) ) {
2822 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2824 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2827 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2830 if ( $this->isMultibyte( $str ) ) {
2831 $x = $first ?
'^' : '';
2832 return preg_replace_callback(
2833 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2834 array( $this, 'lcCallback' ),
2838 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2844 * @param string $str
2847 function isMultibyte( $str ) {
2848 return strlen( $str ) !== mb_strlen( $str );
2852 * @param string $str
2853 * @return mixed|string
2855 function ucwords( $str ) {
2856 if ( $this->isMultibyte( $str ) ) {
2857 $str = $this->lc( $str );
2859 // regexp to find first letter in each word (i.e. after each space)
2860 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2862 // function to use to capitalize a single char
2863 if ( function_exists( 'mb_strtoupper' ) ) {
2864 return preg_replace_callback(
2866 array( $this, 'ucwordsCallbackMB' ),
2870 return preg_replace_callback(
2872 array( $this, 'ucwordsCallbackWiki' ),
2877 return ucwords( strtolower( $str ) );
2882 * capitalize words at word breaks
2884 * @param string $str
2887 function ucwordbreaks( $str ) {
2888 if ( $this->isMultibyte( $str ) ) {
2889 $str = $this->lc( $str );
2891 // since \b doesn't work for UTF-8, we explicitely define word break chars
2892 $breaks = "[ \-\(\)\}\{\.,\?!]";
2894 // find first letter after word break
2895 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2896 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2898 if ( function_exists( 'mb_strtoupper' ) ) {
2899 return preg_replace_callback(
2901 array( $this, 'ucwordbreaksCallbackMB' ),
2905 return preg_replace_callback(
2907 array( $this, 'ucwordsCallbackWiki' ),
2912 return preg_replace_callback(
2913 '/\b([\w\x80-\xff]+)\b/',
2914 array( $this, 'ucwordbreaksCallbackAscii' ),
2921 * Return a case-folded representation of $s
2923 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2924 * and $s2 are the same except for the case of their characters. It is not
2925 * necessary for the value returned to make sense when displayed.
2927 * Do *not* perform any other normalisation in this function. If a caller
2928 * uses this function when it should be using a more general normalisation
2929 * function, then fix the caller.
2935 function caseFold( $s ) {
2936 return $this->uc( $s );
2943 function checkTitleEncoding( $s ) {
2944 if ( is_array( $s ) ) {
2945 throw new MWException( 'Given array to checkTitleEncoding.' );
2947 if ( StringUtils
::isUtf8( $s ) ) {
2951 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2957 function fallback8bitEncoding() {
2958 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2962 * Most writing systems use whitespace to break up words.
2963 * Some languages such as Chinese don't conventionally do this,
2964 * which requires special handling when breaking up words for
2969 function hasWordBreaks() {
2974 * Some languages such as Chinese require word segmentation,
2975 * Specify such segmentation when overridden in derived class.
2977 * @param string $string
2980 function segmentByWord( $string ) {
2985 * Some languages have special punctuation need to be normalized.
2986 * Make such changes here.
2988 * @param string $string
2991 function normalizeForSearch( $string ) {
2992 return self
::convertDoubleWidth( $string );
2996 * convert double-width roman characters to single-width.
2997 * range: ff00-ff5f ~= 0020-007f
2999 * @param string $string
3003 protected static function convertDoubleWidth( $string ) {
3004 static $full = null;
3005 static $half = null;
3007 if ( $full === null ) {
3008 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
3009 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
3010 $full = str_split( $fullWidth, 3 );
3011 $half = str_split( $halfWidth );
3014 $string = str_replace( $full, $half, $string );
3019 * @param string $string
3020 * @param string $pattern
3023 protected static function insertSpace( $string, $pattern ) {
3024 $string = preg_replace( $pattern, " $1 ", $string );
3025 $string = preg_replace( '/ +/', ' ', $string );
3030 * @param array $termsArray
3033 function convertForSearchResult( $termsArray ) {
3034 # some languages, e.g. Chinese, need to do a conversion
3035 # in order for search results to be displayed correctly
3040 * Get the first character of a string.
3045 function firstChar( $s ) {
3048 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
3049 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
3054 if ( isset( $matches[1] ) ) {
3055 if ( strlen( $matches[1] ) != 3 ) {
3059 // Break down Hangul syllables to grab the first jamo
3060 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3061 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3063 } elseif ( $code < 0xb098 ) {
3064 return "\xe3\x84\xb1";
3065 } elseif ( $code < 0xb2e4 ) {
3066 return "\xe3\x84\xb4";
3067 } elseif ( $code < 0xb77c ) {
3068 return "\xe3\x84\xb7";
3069 } elseif ( $code < 0xb9c8 ) {
3070 return "\xe3\x84\xb9";
3071 } elseif ( $code < 0xbc14 ) {
3072 return "\xe3\x85\x81";
3073 } elseif ( $code < 0xc0ac ) {
3074 return "\xe3\x85\x82";
3075 } elseif ( $code < 0xc544 ) {
3076 return "\xe3\x85\x85";
3077 } elseif ( $code < 0xc790 ) {
3078 return "\xe3\x85\x87";
3079 } elseif ( $code < 0xcc28 ) {
3080 return "\xe3\x85\x88";
3081 } elseif ( $code < 0xce74 ) {
3082 return "\xe3\x85\x8a";
3083 } elseif ( $code < 0xd0c0 ) {
3084 return "\xe3\x85\x8b";
3085 } elseif ( $code < 0xd30c ) {
3086 return "\xe3\x85\x8c";
3087 } elseif ( $code < 0xd558 ) {
3088 return "\xe3\x85\x8d";
3090 return "\xe3\x85\x8e";
3097 function initEncoding() {
3098 # Some languages may have an alternate char encoding option
3099 # (Esperanto X-coding, Japanese furigana conversion, etc)
3100 # If this language is used as the primary content language,
3101 # an override to the defaults can be set here on startup.
3108 function recodeForEdit( $s ) {
3109 # For some languages we'll want to explicitly specify
3110 # which characters make it into the edit box raw
3111 # or are converted in some way or another.
3112 global $wgEditEncoding;
3113 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
3116 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3124 function recodeInput( $s ) {
3125 # Take the previous into account.
3126 global $wgEditEncoding;
3127 if ( $wgEditEncoding != '' ) {
3128 $enc = $wgEditEncoding;
3132 if ( $enc == 'UTF-8' ) {
3135 return $this->iconv( $enc, 'UTF-8', $s );
3140 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3141 * also cleans up certain backwards-compatible sequences, converting them
3142 * to the modern Unicode equivalent.
3144 * This is language-specific for performance reasons only.
3150 function normalize( $s ) {
3151 global $wgAllUnicodeFixes;
3152 $s = UtfNormal\Validator
::cleanUp( $s );
3153 if ( $wgAllUnicodeFixes ) {
3154 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3155 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3162 * Transform a string using serialized data stored in the given file (which
3163 * must be in the serialized subdirectory of $IP). The file contains pairs
3164 * mapping source characters to destination characters.
3166 * The data is cached in process memory. This will go faster if you have the
3167 * FastStringSearch extension.
3169 * @param string $file
3170 * @param string $string
3172 * @throws MWException
3175 function transformUsingPairFile( $file, $string ) {
3176 if ( !isset( $this->transformData
[$file] ) ) {
3177 $data = wfGetPrecompiledData( $file );
3178 if ( $data === false ) {
3179 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3181 $this->transformData
[$file] = new ReplacementArray( $data );
3183 return $this->transformData
[$file]->replace( $string );
3187 * For right-to-left language support
3192 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3196 * Return the correct HTML 'dir' attribute value for this language.
3200 return $this->isRTL() ?
'rtl' : 'ltr';
3204 * Return 'left' or 'right' as appropriate alignment for line-start
3205 * for this language's text direction.
3207 * Should be equivalent to CSS3 'start' text-align value....
3211 function alignStart() {
3212 return $this->isRTL() ?
'right' : 'left';
3216 * Return 'right' or 'left' as appropriate alignment for line-end
3217 * for this language's text direction.
3219 * Should be equivalent to CSS3 'end' text-align value....
3223 function alignEnd() {
3224 return $this->isRTL() ?
'left' : 'right';
3228 * A hidden direction mark (LRM or RLM), depending on the language direction.
3229 * Unlike getDirMark(), this function returns the character as an HTML entity.
3230 * This function should be used when the output is guaranteed to be HTML,
3231 * because it makes the output HTML source code more readable. When
3232 * the output is plain text or can be escaped, getDirMark() should be used.
3234 * @param bool $opposite Get the direction mark opposite to your language
3238 function getDirMarkEntity( $opposite = false ) {
3240 return $this->isRTL() ?
'‎' : '‏';
3242 return $this->isRTL() ?
'‏' : '‎';
3246 * A hidden direction mark (LRM or RLM), depending on the language direction.
3247 * This function produces them as invisible Unicode characters and
3248 * the output may be hard to read and debug, so it should only be used
3249 * when the output is plain text or can be escaped. When the output is
3250 * HTML, use getDirMarkEntity() instead.
3252 * @param bool $opposite Get the direction mark opposite to your language
3255 function getDirMark( $opposite = false ) {
3256 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3257 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3259 return $this->isRTL() ?
$lrm : $rlm;
3261 return $this->isRTL() ?
$rlm : $lrm;
3267 function capitalizeAllNouns() {
3268 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3272 * An arrow, depending on the language direction.
3274 * @param string $direction The direction of the arrow: forwards (default),
3275 * backwards, left, right, up, down.
3278 function getArrow( $direction = 'forwards' ) {
3279 switch ( $direction ) {
3281 return $this->isRTL() ?
'←' : '→';
3283 return $this->isRTL() ?
'→' : '←';
3296 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3300 function linkPrefixExtension() {
3301 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3305 * Get all magic words from cache.
3308 function getMagicWords() {
3309 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3313 * Run the LanguageGetMagic hook once.
3315 protected function doMagicHook() {
3316 if ( $this->mMagicHookDone
) {
3319 $this->mMagicHookDone
= true;
3320 Hooks
::run( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3324 * Fill a MagicWord object with data from here
3326 * @param MagicWord $mw
3328 function getMagic( $mw ) {
3329 // Saves a function call
3330 if ( !$this->mMagicHookDone
) {
3331 $this->doMagicHook();
3334 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3335 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3337 $rawEntry = self
::$dataCache->getSubitem(
3338 $this->mCode
, 'magicWords', $mw->mId
);
3341 if ( !is_array( $rawEntry ) ) {
3342 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3344 $mw->mCaseSensitive
= $rawEntry[0];
3345 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3350 * Add magic words to the extension array
3352 * @param array $newWords
3354 function addMagicWordsByLang( $newWords ) {
3355 $fallbackChain = $this->getFallbackLanguages();
3356 $fallbackChain = array_reverse( $fallbackChain );
3357 foreach ( $fallbackChain as $code ) {
3358 if ( isset( $newWords[$code] ) ) {
3359 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3365 * Get special page names, as an associative array
3366 * canonical name => array of valid names, including aliases
3369 function getSpecialPageAliases() {
3370 // Cache aliases because it may be slow to load them
3371 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3373 $this->mExtendedSpecialPageAliases
=
3374 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3375 Hooks
::run( 'LanguageGetSpecialPageAliases',
3376 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3379 return $this->mExtendedSpecialPageAliases
;
3383 * Italic is unsuitable for some languages
3385 * @param string $text The text to be emphasized.
3388 function emphasize( $text ) {
3389 return "<em>$text</em>";
3393 * Normally we output all numbers in plain en_US style, that is
3394 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3395 * point twohundredthirtyfive. However this is not suitable for all
3396 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3397 * Icelandic just want to use commas instead of dots, and dots instead
3398 * of commas like "293.291,235".
3400 * An example of this function being called:
3402 * wfMessage( 'message' )->numParams( $num )->text()
3405 * See $separatorTransformTable on MessageIs.php for
3406 * the , => . and . => , implementation.
3408 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3409 * @param int|float $number The string to be formatted, should be an integer
3410 * or a floating point number.
3411 * @param bool $nocommafy Set to true for special numbers like dates
3414 public function formatNum( $number, $nocommafy = false ) {
3415 global $wgTranslateNumerals;
3416 if ( !$nocommafy ) {
3417 $number = $this->commafy( $number );
3418 $s = $this->separatorTransformTable();
3420 $number = strtr( $number, $s );
3424 if ( $wgTranslateNumerals ) {
3425 $s = $this->digitTransformTable();
3427 $number = strtr( $number, $s );
3435 * Front-end for non-commafied formatNum
3437 * @param int|float $number The string to be formatted, should be an integer
3438 * or a floating point number.
3442 public function formatNumNoSeparators( $number ) {
3443 return $this->formatNum( $number, true );
3447 * @param string $number
3450 public function parseFormattedNumber( $number ) {
3451 $s = $this->digitTransformTable();
3453 // eliminate empty array values such as ''. (bug 64347)
3454 $s = array_filter( $s );
3455 $number = strtr( $number, array_flip( $s ) );
3458 $s = $this->separatorTransformTable();
3460 // eliminate empty array values such as ''. (bug 64347)
3461 $s = array_filter( $s );
3462 $number = strtr( $number, array_flip( $s ) );
3465 $number = strtr( $number, array( ',' => '' ) );
3470 * Adds commas to a given number
3472 * @param mixed $number
3475 function commafy( $number ) {
3476 $digitGroupingPattern = $this->digitGroupingPattern();
3477 if ( $number === null ) {
3481 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3482 // default grouping is at thousands, use the same for ###,###,### pattern too.
3483 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3485 // Ref: http://cldr.unicode.org/translation/number-patterns
3487 if ( intval( $number ) < 0 ) {
3488 // For negative numbers apply the algorithm like positive number and add sign.
3490 $number = substr( $number, 1 );
3492 $integerPart = array();
3493 $decimalPart = array();
3494 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3495 preg_match( "/\d+/", $number, $integerPart );
3496 preg_match( "/\.\d*/", $number, $decimalPart );
3497 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3498 if ( $groupedNumber === $number ) {
3499 // the string does not have any number part. Eg: .12345
3500 return $sign . $groupedNumber;
3502 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3503 while ( $start > 0 ) {
3504 $match = $matches[0][$numMatches - 1];
3505 $matchLen = strlen( $match );
3506 $start = $end - $matchLen;
3510 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3512 if ( $numMatches > 1 ) {
3513 // use the last pattern for the rest of the number
3517 $groupedNumber = "," . $groupedNumber;
3520 return $sign . $groupedNumber;
3527 function digitGroupingPattern() {
3528 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3534 function digitTransformTable() {
3535 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3541 function separatorTransformTable() {
3542 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3546 * Take a list of strings and build a locale-friendly comma-separated
3547 * list, using the local comma-separator message.
3548 * The last two strings are chained with an "and".
3549 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3551 * @param string[] $l
3554 function listToText( array $l ) {
3555 $m = count( $l ) - 1;
3560 $and = $this->msg( 'and' )->escaped();
3561 $space = $this->msg( 'word-separator' )->escaped();
3563 $comma = $this->msg( 'comma-separator' )->escaped();
3567 for ( $i = $m - 1; $i >= 0; $i-- ) {
3568 if ( $i == $m - 1 ) {
3569 $s = $l[$i] . $and . $space . $s;
3571 $s = $l[$i] . $comma . $s;
3578 * Take a list of strings and build a locale-friendly comma-separated
3579 * list, using the local comma-separator message.
3580 * @param string[] $list Array of strings to put in a comma list
3583 function commaList( array $list ) {
3585 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3591 * Take a list of strings and build a locale-friendly semicolon-separated
3592 * list, using the local semicolon-separator message.
3593 * @param string[] $list Array of strings to put in a semicolon list
3596 function semicolonList( array $list ) {
3598 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3604 * Same as commaList, but separate it with the pipe instead.
3605 * @param string[] $list Array of strings to put in a pipe list
3608 function pipeList( array $list ) {
3610 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3616 * Truncate a string to a specified length in bytes, appending an optional
3617 * string (e.g. for ellipses)
3619 * The database offers limited byte lengths for some columns in the database;
3620 * multi-byte character sets mean we need to ensure that only whole characters
3621 * are included, otherwise broken characters can be passed to the user
3623 * If $length is negative, the string will be truncated from the beginning
3625 * @param string $string String to truncate
3626 * @param int $length Maximum length (including ellipses)
3627 * @param string $ellipsis String to append to the truncated text
3628 * @param bool $adjustLength Subtract length of ellipsis from $length.
3629 * $adjustLength was introduced in 1.18, before that behaved as if false.
3632 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3633 # Use the localized ellipsis character
3634 if ( $ellipsis == '...' ) {
3635 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3637 # Check if there is no need to truncate
3638 if ( $length == 0 ) {
3639 return $ellipsis; // convention
3640 } elseif ( strlen( $string ) <= abs( $length ) ) {
3641 return $string; // no need to truncate
3643 $stringOriginal = $string;
3644 # If ellipsis length is >= $length then we can't apply $adjustLength
3645 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3646 $string = $ellipsis; // this can be slightly unexpected
3647 # Otherwise, truncate and add ellipsis...
3649 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3650 if ( $length > 0 ) {
3651 $length -= $eLength;
3652 $string = substr( $string, 0, $length ); // xyz...
3653 $string = $this->removeBadCharLast( $string );
3654 $string = rtrim( $string );
3655 $string = $string . $ellipsis;
3657 $length +
= $eLength;
3658 $string = substr( $string, $length ); // ...xyz
3659 $string = $this->removeBadCharFirst( $string );
3660 $string = ltrim( $string );
3661 $string = $ellipsis . $string;
3664 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3665 # This check is *not* redundant if $adjustLength, due to the single case where
3666 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3667 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3670 return $stringOriginal;
3675 * Remove bytes that represent an incomplete Unicode character
3676 * at the end of string (e.g. bytes of the char are missing)
3678 * @param string $string
3681 protected function removeBadCharLast( $string ) {
3682 if ( $string != '' ) {
3683 $char = ord( $string[strlen( $string ) - 1] );
3685 if ( $char >= 0xc0 ) {
3686 # We got the first byte only of a multibyte char; remove it.
3687 $string = substr( $string, 0, -1 );
3688 } elseif ( $char >= 0x80 &&
3689 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3690 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3692 # We chopped in the middle of a character; remove it
3700 * Remove bytes that represent an incomplete Unicode character
3701 * at the start of string (e.g. bytes of the char are missing)
3703 * @param string $string
3706 protected function removeBadCharFirst( $string ) {
3707 if ( $string != '' ) {
3708 $char = ord( $string[0] );
3709 if ( $char >= 0x80 && $char < 0xc0 ) {
3710 # We chopped in the middle of a character; remove the whole thing
3711 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3718 * Truncate a string of valid HTML to a specified length in bytes,
3719 * appending an optional string (e.g. for ellipses), and return valid HTML
3721 * This is only intended for styled/linked text, such as HTML with
3722 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3723 * Also, this will not detect things like "display:none" CSS.
3725 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3727 * @param string $text HTML string to truncate
3728 * @param int $length (zero/positive) Maximum length (including ellipses)
3729 * @param string $ellipsis String to append to the truncated text
3732 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3733 # Use the localized ellipsis character
3734 if ( $ellipsis == '...' ) {
3735 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3737 # Check if there is clearly no need to truncate
3738 if ( $length <= 0 ) {
3739 return $ellipsis; // no text shown, nothing to format (convention)
3740 } elseif ( strlen( $text ) <= $length ) {
3741 return $text; // string short enough even *with* HTML (short-circuit)
3744 $dispLen = 0; // innerHTML legth so far
3745 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3746 $tagType = 0; // 0-open, 1-close
3747 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3748 $entityState = 0; // 0-not entity, 1-entity
3749 $tag = $ret = ''; // accumulated tag name, accumulated result string
3750 $openTags = array(); // open tag stack
3751 $maybeState = null; // possible truncation state
3753 $textLen = strlen( $text );
3754 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3755 for ( $pos = 0; true; ++
$pos ) {
3756 # Consider truncation once the display length has reached the maximim.
3757 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3758 # Check that we're not in the middle of a bracket/entity...
3759 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3760 if ( !$testingEllipsis ) {
3761 $testingEllipsis = true;
3762 # Save where we are; we will truncate here unless there turn out to
3763 # be so few remaining characters that truncation is not necessary.
3764 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3765 $maybeState = array( $ret, $openTags ); // save state
3767 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3768 # String in fact does need truncation, the truncation point was OK.
3769 list( $ret, $openTags ) = $maybeState; // reload state
3770 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3771 $ret .= $ellipsis; // add ellipsis
3775 if ( $pos >= $textLen ) {
3776 break; // extra iteration just for above checks
3779 # Read the next char...
3781 $lastCh = $pos ?
$text[$pos - 1] : '';
3782 $ret .= $ch; // add to result string
3784 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3785 $entityState = 0; // for bad HTML
3786 $bracketState = 1; // tag started (checking for backslash)
3787 } elseif ( $ch == '>' ) {
3788 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3789 $entityState = 0; // for bad HTML
3790 $bracketState = 0; // out of brackets
3791 } elseif ( $bracketState == 1 ) {
3793 $tagType = 1; // close tag (e.g. "</span>")
3795 $tagType = 0; // open tag (e.g. "<span>")
3798 $bracketState = 2; // building tag name
3799 } elseif ( $bracketState == 2 ) {
3803 // Name found (e.g. "<a href=..."), add on tag attributes...
3804 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3806 } elseif ( $bracketState == 0 ) {
3807 if ( $entityState ) {
3810 $dispLen++
; // entity is one displayed char
3813 if ( $neLength == 0 && !$maybeState ) {
3814 // Save state without $ch. We want to *hit* the first
3815 // display char (to get tags) but not *use* it if truncating.
3816 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3819 $entityState = 1; // entity found, (e.g. " ")
3821 $dispLen++
; // this char is displayed
3822 // Add the next $max display text chars after this in one swoop...
3823 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3824 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3825 $dispLen +
= $skipped;
3831 // Close the last tag if left unclosed by bad HTML
3832 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3833 while ( count( $openTags ) > 0 ) {
3834 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3840 * truncateHtml() helper function
3841 * like strcspn() but adds the skipped chars to $ret
3843 * @param string $ret
3844 * @param string $text
3845 * @param string $search
3847 * @param null|int $len
3850 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3851 if ( $len === null ) {
3852 $len = -1; // -1 means "no limit" for strcspn
3853 } elseif ( $len < 0 ) {
3857 if ( $start < strlen( $text ) ) {
3858 $skipCount = strcspn( $text, $search, $start, $len );
3859 $ret .= substr( $text, $start, $skipCount );
3865 * truncateHtml() helper function
3866 * (a) push or pop $tag from $openTags as needed
3867 * (b) clear $tag value
3868 * @param string &$tag Current HTML tag name we are looking at
3869 * @param int $tagType (0-open tag, 1-close tag)
3870 * @param string $lastCh Character before the '>' that ended this tag
3871 * @param array &$openTags Open tag stack (not accounting for $tag)
3873 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3874 $tag = ltrim( $tag );
3876 if ( $tagType == 0 && $lastCh != '/' ) {
3877 $openTags[] = $tag; // tag opened (didn't close itself)
3878 } elseif ( $tagType == 1 ) {
3879 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3880 array_pop( $openTags ); // tag closed
3888 * Grammatical transformations, needed for inflected languages
3889 * Invoked by putting {{grammar:case|word}} in a message
3891 * @param string $word
3892 * @param string $case
3895 function convertGrammar( $word, $case ) {
3896 global $wgGrammarForms;
3897 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3898 return $wgGrammarForms[$this->getCode()][$case][$word];
3904 * Get the grammar forms for the content language
3905 * @return array Array of grammar forms
3908 function getGrammarForms() {
3909 global $wgGrammarForms;
3910 if ( isset( $wgGrammarForms[$this->getCode()] )
3911 && is_array( $wgGrammarForms[$this->getCode()] )
3913 return $wgGrammarForms[$this->getCode()];
3919 * Provides an alternative text depending on specified gender.
3920 * Usage {{gender:username|masculine|feminine|unknown}}.
3921 * username is optional, in which case the gender of current user is used,
3922 * but only in (some) interface messages; otherwise default gender is used.
3924 * If no forms are given, an empty string is returned. If only one form is
3925 * given, it will be returned unconditionally. These details are implied by
3926 * the caller and cannot be overridden in subclasses.
3928 * If three forms are given, the default is to use the third (unknown) form.
3929 * If fewer than three forms are given, the default is to use the first (masculine) form.
3930 * These details can be overridden in subclasses.
3932 * @param string $gender
3933 * @param array $forms
3937 function gender( $gender, $forms ) {
3938 if ( !count( $forms ) ) {
3941 $forms = $this->preConvertPlural( $forms, 2 );
3942 if ( $gender === 'male' ) {
3945 if ( $gender === 'female' ) {
3948 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3952 * Plural form transformations, needed for some languages.
3953 * For example, there are 3 form of plural in Russian and Polish,
3954 * depending on "count mod 10". See [[w:Plural]]
3955 * For English it is pretty simple.
3957 * Invoked by putting {{plural:count|wordform1|wordform2}}
3958 * or {{plural:count|wordform1|wordform2|wordform3}}
3960 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3962 * @param int $count Non-localized number
3963 * @param array $forms Different plural forms
3964 * @return string Correct form of plural for $count in this language
3966 function convertPlural( $count, $forms ) {
3967 // Handle explicit n=pluralform cases
3968 $forms = $this->handleExplicitPluralForms( $count, $forms );
3969 if ( is_string( $forms ) ) {
3972 if ( !count( $forms ) ) {
3976 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3977 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3978 return $forms[$pluralForm];
3982 * Handles explicit plural forms for Language::convertPlural()
3984 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3985 * If an explicitly defined plural form matches the $count, then
3986 * string value returned, otherwise array returned for further consideration
3987 * by CLDR rules or overridden convertPlural().
3991 * @param int $count Non-localized number
3992 * @param array $forms Different plural forms
3994 * @return array|string
3996 protected function handleExplicitPluralForms( $count, array $forms ) {
3997 foreach ( $forms as $index => $form ) {
3998 if ( preg_match( '/\d+=/i', $form ) ) {
3999 $pos = strpos( $form, '=' );
4000 if ( substr( $form, 0, $pos ) === (string)$count ) {
4001 return substr( $form, $pos +
1 );
4003 unset( $forms[$index] );
4006 return array_values( $forms );
4010 * Checks that convertPlural was given an array and pads it to requested
4011 * amount of forms by copying the last one.
4013 * @param array $forms Array of forms given to convertPlural
4014 * @param int $count How many forms should there be at least
4015 * @return array Padded array of forms or an exception if not an array
4017 protected function preConvertPlural( /* Array */ $forms, $count ) {
4018 while ( count( $forms ) < $count ) {
4019 $forms[] = $forms[count( $forms ) - 1];
4025 * Wraps argument with unicode control characters for directionality safety
4027 * This solves the problem where directionality-neutral characters at the edge of
4028 * the argument string get interpreted with the wrong directionality from the
4029 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4031 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4032 * directionality of the argument string, using the BIDI algorithm's own "First
4033 * strong directional codepoint" rule. Essentially, this works round the fact that
4034 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4035 * direction inference). The latter is cleaner but still not widely supported.
4037 * @param string $text Text to wrap
4038 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4040 public function embedBidi( $text = '' ) {
4041 $dir = Language
::strongDirFromContent( $text );
4042 if ( $dir === 'ltr' ) {
4043 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4044 return self
::$lre . $text . self
::$pdf;
4046 if ( $dir === 'rtl' ) {
4047 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4048 return self
::$rle . $text . self
::$pdf;
4050 // No strong directionality: do not wrap
4055 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4056 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4057 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4058 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4061 * @param string $str The validated block duration in English
4062 * @return string Somehow translated block duration
4063 * @see LanguageFi.php for example implementation
4065 function translateBlockExpiry( $str ) {
4066 $duration = SpecialBlock
::getSuggestedDurations( $this );
4067 foreach ( $duration as $show => $value ) {
4068 if ( strcmp( $str, $value ) == 0 ) {
4069 return htmlspecialchars( trim( $show ) );
4073 if ( wfIsInfinity( $str ) ) {
4074 foreach ( $duration as $show => $value ) {
4075 if ( wfIsInfinity( $value ) ) {
4076 return htmlspecialchars( trim( $show ) );
4081 // If all else fails, return a standard duration or timestamp description.
4082 $time = strtotime( $str, 0 );
4083 if ( $time === false ) { // Unknown format. Return it as-is in case.
4085 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4086 // $time is relative to 0 so it's a duration length.
4087 return $this->formatDuration( $time );
4088 } else { // It's an absolute timestamp.
4089 if ( $time === 0 ) {
4090 // wfTimestamp() handles 0 as current time instead of epoch.
4091 return $this->timeanddate( '19700101000000' );
4093 return $this->timeanddate( $time );
4099 * languages like Chinese need to be segmented in order for the diff
4102 * @param string $text
4105 public function segmentForDiff( $text ) {
4110 * and unsegment to show the result
4112 * @param string $text
4115 public function unsegmentForDiff( $text ) {
4120 * Return the LanguageConverter used in the Language
4123 * @return LanguageConverter
4125 public function getConverter() {
4126 return $this->mConverter
;
4130 * convert text to all supported variants
4132 * @param string $text
4135 public function autoConvertToAllVariants( $text ) {
4136 return $this->mConverter
->autoConvertToAllVariants( $text );
4140 * convert text to different variants of a language.
4142 * @param string $text
4145 public function convert( $text ) {
4146 return $this->mConverter
->convert( $text );
4150 * Convert a Title object to a string in the preferred variant
4152 * @param Title $title
4155 public function convertTitle( $title ) {
4156 return $this->mConverter
->convertTitle( $title );
4160 * Convert a namespace index to a string in the preferred variant
4165 public function convertNamespace( $ns ) {
4166 return $this->mConverter
->convertNamespace( $ns );
4170 * Check if this is a language with variants
4174 public function hasVariants() {
4175 return count( $this->getVariants() ) > 1;
4179 * Check if the language has the specific variant
4182 * @param string $variant
4185 public function hasVariant( $variant ) {
4186 return (bool)$this->mConverter
->validateVariant( $variant );
4190 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4192 * @param string $text
4194 * @deprecated since 1.22 is no longer used
4196 public function armourMath( $text ) {
4197 return $this->mConverter
->armourMath( $text );
4201 * Perform output conversion on a string, and encode for safe HTML output.
4202 * @param string $text Text to be converted
4203 * @param bool $isTitle Whether this conversion is for the article title
4205 * @todo this should get integrated somewhere sane
4207 public function convertHtml( $text, $isTitle = false ) {
4208 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4212 * @param string $key
4215 public function convertCategoryKey( $key ) {
4216 return $this->mConverter
->convertCategoryKey( $key );
4220 * Get the list of variants supported by this language
4221 * see sample implementation in LanguageZh.php
4223 * @return array An array of language codes
4225 public function getVariants() {
4226 return $this->mConverter
->getVariants();
4232 public function getPreferredVariant() {
4233 return $this->mConverter
->getPreferredVariant();
4239 public function getDefaultVariant() {
4240 return $this->mConverter
->getDefaultVariant();
4246 public function getURLVariant() {
4247 return $this->mConverter
->getURLVariant();
4251 * If a language supports multiple variants, it is
4252 * possible that non-existing link in one variant
4253 * actually exists in another variant. this function
4254 * tries to find it. See e.g. LanguageZh.php
4255 * The input parameters may be modified upon return
4257 * @param string &$link The name of the link
4258 * @param Title &$nt The title object of the link
4259 * @param bool $ignoreOtherCond To disable other conditions when
4260 * we need to transclude a template or update a category's link
4262 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4263 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4267 * returns language specific options used by User::getPageRenderHash()
4268 * for example, the preferred language variant
4272 function getExtraHashOptions() {
4273 return $this->mConverter
->getExtraHashOptions();
4277 * For languages that support multiple variants, the title of an
4278 * article may be displayed differently in different variants. this
4279 * function returns the apporiate title defined in the body of the article.
4283 public function getParsedTitle() {
4284 return $this->mConverter
->getParsedTitle();
4288 * Prepare external link text for conversion. When the text is
4289 * a URL, it shouldn't be converted, and it'll be wrapped in
4290 * the "raw" tag (-{R| }-) to prevent conversion.
4292 * This function is called "markNoConversion" for historical
4295 * @param string $text Text to be used for external link
4296 * @param bool $noParse Wrap it without confirming it's a real URL first
4297 * @return string The tagged text
4299 public function markNoConversion( $text, $noParse = false ) {
4300 // Excluding protocal-relative URLs may avoid many false positives.
4301 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4302 return $this->mConverter
->markNoConversion( $text );
4309 * A regular expression to match legal word-trailing characters
4310 * which should be merged onto a link of the form [[foo]]bar.
4314 public function linkTrail() {
4315 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4319 * A regular expression character set to match legal word-prefixing
4320 * characters which should be merged onto a link of the form foo[[bar]].
4324 public function linkPrefixCharset() {
4325 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4329 * @deprecated since 1.24, will be removed in 1.25
4332 function getLangObj() {
4333 wfDeprecated( __METHOD__
, '1.24' );
4338 * Get the "parent" language which has a converter to convert a "compatible" language
4339 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4341 * @return Language|null
4344 public function getParentLanguage() {
4345 if ( $this->mParentLanguage
!== false ) {
4346 return $this->mParentLanguage
;
4349 $pieces = explode( '-', $this->getCode() );
4351 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4352 $this->mParentLanguage
= null;
4355 $lang = Language
::factory( $code );
4356 if ( !$lang->hasVariant( $this->getCode() ) ) {
4357 $this->mParentLanguage
= null;
4361 $this->mParentLanguage
= $lang;
4366 * Get the RFC 3066 code for this language object
4368 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4369 * htmlspecialchars() or similar
4373 public function getCode() {
4374 return $this->mCode
;
4378 * Get the code in Bcp47 format which we can use
4379 * inside of html lang="" tags.
4381 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4382 * htmlspecialchars() or similar.
4387 public function getHtmlCode() {
4388 if ( is_null( $this->mHtmlCode
) ) {
4389 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4391 return $this->mHtmlCode
;
4395 * @param string $code
4397 public function setCode( $code ) {
4398 $this->mCode
= $code;
4399 // Ensure we don't leave incorrect cached data lying around
4400 $this->mHtmlCode
= null;
4401 $this->mParentLanguage
= false;
4405 * Get the name of a file for a certain language code
4406 * @param string $prefix Prepend this to the filename
4407 * @param string $code Language code
4408 * @param string $suffix Append this to the filename
4409 * @throws MWException
4410 * @return string $prefix . $mangledCode . $suffix
4412 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4413 if ( !self
::isValidBuiltInCode( $code ) ) {
4414 throw new MWException( "Invalid language code \"$code\"" );
4417 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4421 * Get the language code from a file name. Inverse of getFileName()
4422 * @param string $filename $prefix . $languageCode . $suffix
4423 * @param string $prefix Prefix before the language code
4424 * @param string $suffix Suffix after the language code
4425 * @return string Language code, or false if $prefix or $suffix isn't found
4427 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4429 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4430 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4431 if ( !count( $m ) ) {
4434 return str_replace( '_', '-', strtolower( $m[1] ) );
4438 * @param string $code
4441 public static function getMessagesFileName( $code ) {
4443 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4444 Hooks
::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4449 * @param string $code
4453 public static function getJsonMessagesFileName( $code ) {
4456 if ( !self
::isValidBuiltInCode( $code ) ) {
4457 throw new MWException( "Invalid language code \"$code\"" );
4460 return "$IP/languages/i18n/$code.json";
4464 * @param string $code
4467 public static function getClassFileName( $code ) {
4469 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4473 * Get the first fallback for a given language.
4475 * @param string $code
4477 * @return bool|string
4479 public static function getFallbackFor( $code ) {
4480 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4483 $fallbacks = self
::getFallbacksFor( $code );
4484 return $fallbacks[0];
4489 * Get the ordered list of fallback languages.
4492 * @param string $code Language code
4493 * @return array Non-empty array, ending in "en"
4495 public static function getFallbacksFor( $code ) {
4496 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4499 // For unknown languages, fallbackSequence returns an empty array,
4500 // hardcode fallback to 'en' in that case.
4501 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: array( 'en' );
4505 * Get the ordered list of fallback languages, ending with the fallback
4506 * language chain for the site language.
4509 * @param string $code Language code
4510 * @return array Array( fallbacks, site fallbacks )
4512 public static function getFallbacksIncludingSiteLanguage( $code ) {
4513 global $wgLanguageCode;
4515 // Usually, we will only store a tiny number of fallback chains, so we
4516 // keep them in static memory.
4517 $cacheKey = "{$code}-{$wgLanguageCode}";
4519 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4520 $fallbacks = self
::getFallbacksFor( $code );
4522 // Append the site's fallback chain, including the site language itself
4523 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4524 array_unshift( $siteFallbacks, $wgLanguageCode );
4526 // Eliminate any languages already included in the chain
4527 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4529 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4531 return self
::$fallbackLanguageCache[$cacheKey];
4535 * Get all messages for a given language
4536 * WARNING: this may take a long time. If you just need all message *keys*
4537 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4539 * @param string $code
4543 public static function getMessagesFor( $code ) {
4544 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4548 * Get a message for a given language
4550 * @param string $key
4551 * @param string $code
4555 public static function getMessageFor( $key, $code ) {
4556 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4560 * Get all message keys for a given language. This is a faster alternative to
4561 * array_keys( Language::getMessagesFor( $code ) )
4564 * @param string $code Language code
4565 * @return array Array of message keys (strings)
4567 public static function getMessageKeysFor( $code ) {
4568 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4572 * @param string $talk
4575 function fixVariableInNamespace( $talk ) {
4576 if ( strpos( $talk, '$1' ) === false ) {
4580 global $wgMetaNamespace;
4581 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4583 # Allow grammar transformations
4584 # Allowing full message-style parsing would make simple requests
4585 # such as action=raw much more expensive than they need to be.
4586 # This will hopefully cover most cases.
4587 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4588 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4589 return str_replace( ' ', '_', $talk );
4596 function replaceGrammarInNamespace( $m ) {
4597 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4601 * @throws MWException
4604 static function getCaseMaps() {
4605 static $wikiUpperChars, $wikiLowerChars;
4606 if ( isset( $wikiUpperChars ) ) {
4607 return array( $wikiUpperChars, $wikiLowerChars );
4610 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4611 if ( $arr === false ) {
4612 throw new MWException(
4613 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4615 $wikiUpperChars = $arr['wikiUpperChars'];
4616 $wikiLowerChars = $arr['wikiLowerChars'];
4617 return array( $wikiUpperChars, $wikiLowerChars );
4621 * Decode an expiry (block, protection, etc) which has come from the DB
4623 * @param string $expiry Database expiry String
4624 * @param bool|int $format True to process using language functions, or TS_ constant
4625 * to return the expiry in a given timestamp
4626 * @param string $inifinity If $format is not true, use this string for infinite expiry
4630 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4632 if ( $dbInfinity === null ) {
4633 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4636 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4637 return $format === true
4638 ?
$this->getMessageFromDB( 'infiniteblock' )
4641 return $format === true
4642 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4643 : wfTimestamp( $format, $expiry );
4649 * @param int|float $seconds
4650 * @param array $format Optional
4651 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4652 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4653 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4655 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4656 * or 'avoidminutes'.
4659 function formatTimePeriod( $seconds, $format = array() ) {
4660 if ( !is_array( $format ) ) {
4661 $format = array( 'avoid' => $format ); // For backwards compatibility
4663 if ( !isset( $format['avoid'] ) ) {
4664 $format['avoid'] = false;
4666 if ( !isset( $format['noabbrevs'] ) ) {
4667 $format['noabbrevs'] = false;
4669 $secondsMsg = wfMessage(
4670 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4671 $minutesMsg = wfMessage(
4672 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4673 $hoursMsg = wfMessage(
4674 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4675 $daysMsg = wfMessage(
4676 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4678 if ( round( $seconds * 10 ) < 100 ) {
4679 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4680 $s = $secondsMsg->params( $s )->text();
4681 } elseif ( round( $seconds ) < 60 ) {
4682 $s = $this->formatNum( round( $seconds ) );
4683 $s = $secondsMsg->params( $s )->text();
4684 } elseif ( round( $seconds ) < 3600 ) {
4685 $minutes = floor( $seconds / 60 );
4686 $secondsPart = round( fmod( $seconds, 60 ) );
4687 if ( $secondsPart == 60 ) {
4691 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4693 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4694 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4695 $hours = floor( $seconds / 3600 );
4696 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4697 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4698 if ( $secondsPart == 60 ) {
4702 if ( $minutes == 60 ) {
4706 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4708 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4709 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4710 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4713 $days = floor( $seconds / 86400 );
4714 if ( $format['avoid'] === 'avoidminutes' ) {
4715 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4716 if ( $hours == 24 ) {
4720 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4722 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4723 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4724 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4725 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4726 if ( $minutes == 60 ) {
4730 if ( $hours == 24 ) {
4734 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4736 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4738 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4740 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4742 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4749 * Format a bitrate for output, using an appropriate
4750 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4751 * the magnitude in question.
4753 * This use base 1000. For base 1024 use formatSize(), for another base
4754 * see formatComputingNumbers().
4759 function formatBitrate( $bps ) {
4760 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4764 * @param int $size Size of the unit
4765 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4766 * @param string $messageKey Message key to be uesd
4769 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4771 return str_replace( '$1', $this->formatNum( $size ),
4772 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4775 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4778 $maxIndex = count( $sizes ) - 1;
4779 while ( $size >= $boundary && $index < $maxIndex ) {
4784 // For small sizes no decimal places necessary
4787 // For MB and bigger two decimal places are smarter
4790 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4792 $size = round( $size, $round );
4793 $text = $this->getMessageFromDB( $msg );
4794 return str_replace( '$1', $this->formatNum( $size ), $text );
4798 * Format a size in bytes for output, using an appropriate
4799 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4801 * This method use base 1024. For base 1000 use formatBitrate(), for
4802 * another base see formatComputingNumbers()
4804 * @param int $size Size to format
4805 * @return string Plain text (not HTML)
4807 function formatSize( $size ) {
4808 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4812 * Make a list item, used by various special pages
4814 * @param string $page Page link
4815 * @param string $details HTML safe text between brackets
4816 * @param bool $oppositedm Add the direction mark opposite to your
4817 * language, to display text properly
4818 * @return HTML escaped string
4820 function specialList( $page, $details, $oppositedm = true ) {
4825 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4829 $this->msg( 'word-separator' )->escaped() .
4830 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4834 * Generate (prev x| next x) (20|50|100...) type links for paging
4836 * @param Title $title Title object to link
4837 * @param int $offset
4839 * @param array $query Optional URL query parameter string
4840 * @param bool $atend Optional param for specified if this is the last page
4843 public function viewPrevNext( Title
$title, $offset, $limit,
4844 array $query = array(), $atend = false
4846 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4848 # Make 'previous' link
4849 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4850 if ( $offset > 0 ) {
4851 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4852 $query, $prev, 'prevn-title', 'mw-prevlink' );
4854 $plink = htmlspecialchars( $prev );
4858 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4860 $nlink = htmlspecialchars( $next );
4862 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4863 $query, $next, 'nextn-title', 'mw-nextlink' );
4866 # Make links to set number of items per page
4867 $numLinks = array();
4868 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4869 $numLinks[] = $this->numLink( $title, $offset, $num,
4870 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4873 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4874 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4878 * Helper function for viewPrevNext() that generates links
4880 * @param Title $title Title object to link
4881 * @param int $offset
4883 * @param array $query Extra query parameters
4884 * @param string $link Text to use for the link; will be escaped
4885 * @param string $tooltipMsg Name of the message to use as tooltip
4886 * @param string $class Value of the "class" attribute of the link
4887 * @return string HTML fragment
4889 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4892 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4893 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4894 ->numParams( $limit )->text();
4896 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4897 'title' => $tooltip, 'class' => $class ), $link );
4901 * Get the conversion rule title, if any.
4905 public function getConvRuleTitle() {
4906 return $this->mConverter
->getConvRuleTitle();
4910 * Get the compiled plural rules for the language
4912 * @return array Associative array with plural form, and plural rule as key-value pairs
4914 public function getCompiledPluralRules() {
4915 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4916 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4917 if ( !$pluralRules ) {
4918 foreach ( $fallbacks as $fallbackCode ) {
4919 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4920 if ( $pluralRules ) {
4925 return $pluralRules;
4929 * Get the plural rules for the language
4931 * @return array Associative array with plural form number and plural rule as key-value pairs
4933 public function getPluralRules() {
4934 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4935 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4936 if ( !$pluralRules ) {
4937 foreach ( $fallbacks as $fallbackCode ) {
4938 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4939 if ( $pluralRules ) {
4944 return $pluralRules;
4948 * Get the plural rule types for the language
4950 * @return array Associative array with plural form number and plural rule type as key-value pairs
4952 public function getPluralRuleTypes() {
4953 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4954 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4955 if ( !$pluralRuleTypes ) {
4956 foreach ( $fallbacks as $fallbackCode ) {
4957 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4958 if ( $pluralRuleTypes ) {
4963 return $pluralRuleTypes;
4967 * Find the index number of the plural rule appropriate for the given number
4968 * @param int $number
4969 * @return int The index number of the plural rule
4971 public function getPluralRuleIndexNumber( $number ) {
4972 $pluralRules = $this->getCompiledPluralRules();
4973 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
4978 * Find the plural rule type appropriate for the given number
4979 * For example, if the language is set to Arabic, getPluralType(5) should
4982 * @param int $number
4983 * @return string The name of the plural rule type, e.g. one, two, few, many
4985 public function getPluralRuleType( $number ) {
4986 $index = $this->getPluralRuleIndexNumber( $number );
4987 $pluralRuleTypes = $this->getPluralRuleTypes();
4988 if ( isset( $pluralRuleTypes[$index] ) ) {
4989 return $pluralRuleTypes[$index];