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' );
38 * Internationalisation code
43 * @var LanguageConverter
47 public $mVariants, $mCode, $mLoaded = false;
48 public $mMagicExtensions = array(), $mMagicHookDone = false;
49 private $mHtmlCode = null, $mParentLanguage = false;
51 public $dateFormatStrings = array();
52 public $mExtendedSpecialPageAliases;
54 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
57 * ReplacementArray object caches
59 public $transformData = array();
62 * @var LocalisationCache
64 static public $dataCache;
66 static public $mLangObjCache = array();
68 static public $mWeekdayMsgs = array(
69 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
73 static public $mWeekdayAbbrevMsgs = array(
74 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
77 static public $mMonthMsgs = array(
78 'january', 'february', 'march', 'april', 'may_long', 'june',
79 'july', 'august', 'september', 'october', 'november',
82 static public $mMonthGenMsgs = array(
83 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
84 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
87 static public $mMonthAbbrevMsgs = array(
88 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
89 'sep', 'oct', 'nov', 'dec'
92 static public $mIranianCalendarMonthMsgs = array(
93 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
94 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
95 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
96 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
99 static public $mHebrewCalendarMonthMsgs = array(
100 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
101 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
102 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
103 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
104 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
107 static public $mHebrewCalendarMonthGenMsgs = array(
108 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
109 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
110 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
111 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
112 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
115 static public $mHijriCalendarMonthMsgs = array(
116 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
117 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
118 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
119 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
126 static public $durationIntervals = array(
127 'millennia' => 31556952000,
128 'centuries' => 3155695200,
129 'decades' => 315569520,
130 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
139 * Cache for language fallbacks.
140 * @see Language::getFallbacksIncludingSiteLanguage
144 static private $fallbackLanguageCache = array();
147 * Cache for language names
148 * @var MapCacheLRU|null
150 static private $languageNameCache;
153 * Unicode directional formatting characters, for embedBidi()
155 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
156 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
157 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
160 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
161 * - in group 1 if it is LTR
162 * - in group 2 if it is RTL
163 * Does not match if there is no strong directionality codepoint.
165 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
167 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
168 * https://git.wikimedia.org/summary/unicodejs.git .
170 // @codingStandardsIgnoreStart
171 // @codeCoverageIgnoreStart
172 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';
173 // @codeCoverageIgnoreEnd
174 // @codingStandardsIgnoreEnd
177 * Get a cached or new language object for a given language code
178 * @param string $code
181 static function factory( $code ) {
182 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
184 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
185 $code = $wgDummyLanguageCodes[$code];
188 // get the language object to process
189 $langObj = isset( self
::$mLangObjCache[$code] )
190 ? self
::$mLangObjCache[$code]
191 : self
::newFromCode( $code );
193 // merge the language object in to get it up front in the cache
194 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
195 // get rid of the oldest ones in case we have an overflow
196 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
202 * Create a language object for a given language code
203 * @param string $code
204 * @throws MWException
207 protected static function newFromCode( $code ) {
208 // Protect against path traversal below
209 if ( !Language
::isValidCode( $code )
210 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code )
212 throw new MWException( "Invalid language code \"$code\"" );
215 if ( !Language
::isValidBuiltInCode( $code ) ) {
216 // It's not possible to customise this code with class files, so
217 // just return a Language object. This is to support uselang= hacks.
218 $lang = new Language
;
219 $lang->setCode( $code );
223 // Check if there is a language class for the code
224 $class = self
::classFromCode( $code );
225 self
::preloadLanguageClass( $class );
226 if ( class_exists( $class ) ) {
231 // Keep trying the fallback list until we find an existing class
232 $fallbacks = Language
::getFallbacksFor( $code );
233 foreach ( $fallbacks as $fallbackCode ) {
234 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
235 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
238 $class = self
::classFromCode( $fallbackCode );
239 self
::preloadLanguageClass( $class );
240 if ( class_exists( $class ) ) {
241 $lang = Language
::newFromCode( $fallbackCode );
242 $lang->setCode( $code );
247 throw new MWException( "Invalid fallback sequence for language '$code'" );
251 * Checks whether any localisation is available for that language tag
252 * in MediaWiki (MessagesXx.php exists).
254 * @param string $code Language tag (in lower case)
255 * @return bool Whether language is supported
258 public static function isSupportedLanguage( $code ) {
259 if ( !self
::isValidBuiltInCode( $code ) ) {
263 if ( $code === 'qqq' ) {
267 return is_readable( self
::getMessagesFileName( $code ) ) ||
268 is_readable( self
::getJsonMessagesFileName( $code ) );
272 * Returns true if a language code string is a well-formed language tag
273 * according to RFC 5646.
274 * This function only checks well-formedness; it doesn't check that
275 * language, script or variant codes actually exist in the repositories.
277 * Based on regexes by Mark Davis of the Unicode Consortium:
278 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
280 * @param string $code
281 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
286 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
289 $alphanum = '[a-z0-9]';
290 $x = 'x'; # private use singleton
291 $singleton = '[a-wy-z]'; # other singleton
292 $s = $lenient ?
'[-_]' : '-';
294 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
295 $script = "$alpha{4}"; # ISO 15924
296 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
297 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
298 $extension = "$singleton(?:$s$alphanum{2,8})+";
299 $privateUse = "$x(?:$s$alphanum{1,8})+";
301 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
302 # Since these are limited, this is safe even later changes to the registry --
303 # the only oddity is that it might change the type of the tag, and thus
304 # the results from the capturing groups.
305 # http://www.iana.org/assignments/language-subtag-registry
307 $grandfathered = "en{$s}GB{$s}oed"
308 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
309 . "|no{$s}(?:bok|nyn)"
310 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
311 . "|zh{$s}min{$s}nan";
313 $variantList = "$variant(?:$s$variant)*";
314 $extensionList = "$extension(?:$s$extension)*";
316 $langtag = "(?:($language)"
319 . "(?:$s$variantList)?"
320 . "(?:$s$extensionList)?"
321 . "(?:$s$privateUse)?)";
323 # The final breakdown, with capturing groups for each of these components
324 # The variants, extensions, grandfathered, and private-use may have interior '-'
326 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
328 return (bool)preg_match( "/$root/", strtolower( $code ) );
332 * Returns true if a language code string is of a valid form, whether or
333 * not it exists. This includes codes which are used solely for
334 * customisation via the MediaWiki namespace.
336 * @param string $code
340 public static function isValidCode( $code ) {
341 static $cache = array();
342 if ( isset( $cache[$code] ) ) {
343 return $cache[$code];
345 // People think language codes are html safe, so enforce it.
346 // Ideally we should only allow a-zA-Z0-9-
347 // but, .+ and other chars are often used for {{int:}} hacks
348 // see bugs 37564, 37587, 36938
350 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
351 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
353 return $cache[$code];
357 * Returns true if a language code is of a valid form for the purposes of
358 * internal customisation of MediaWiki, via Messages*.php or *.json.
360 * @param string $code
362 * @throws MWException
366 public static function isValidBuiltInCode( $code ) {
368 if ( !is_string( $code ) ) {
369 if ( is_object( $code ) ) {
370 $addmsg = " of class " . get_class( $code );
374 $type = gettype( $code );
375 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
378 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
382 * Returns true if a language code is an IETF tag known to MediaWiki.
389 public static function isKnownLanguageTag( $tag ) {
390 static $coreLanguageNames;
392 // Quick escape for invalid input to avoid exceptions down the line
393 // when code tries to process tags which are not valid at all.
394 if ( !self
::isValidBuiltInCode( $tag ) ) {
398 if ( $coreLanguageNames === null ) {
400 include "$IP/languages/Names.php";
403 if ( isset( $coreLanguageNames[$tag] )
404 || self
::fetchLanguageName( $tag, $tag ) !== ''
413 * @param string $code
414 * @return string Name of the language class
416 public static function classFromCode( $code ) {
417 if ( $code == 'en' ) {
420 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
425 * Includes language class files
427 * @param string $class Name of the language class
429 public static function preloadLanguageClass( $class ) {
432 if ( $class === 'Language' ) {
436 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
437 include_once "$IP/languages/classes/$class.php";
442 * Get the LocalisationCache instance
444 * @return LocalisationCache
446 public static function getLocalisationCache() {
447 if ( is_null( self
::$dataCache ) ) {
448 global $wgLocalisationCacheConf;
449 $class = $wgLocalisationCacheConf['class'];
450 self
::$dataCache = new $class( $wgLocalisationCacheConf );
452 return self
::$dataCache;
455 function __construct() {
456 $this->mConverter
= new FakeConverter( $this );
457 // Set the code to the name of the descendant
458 if ( get_class( $this ) == 'Language' ) {
461 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
463 self
::getLocalisationCache();
467 * Reduce memory usage
469 function __destruct() {
470 foreach ( $this as $name => $value ) {
471 unset( $this->$name );
476 * Hook which will be called if this is the content language.
477 * Descendants can use this to register hook functions or modify globals
479 function initContLang() {
486 function getFallbackLanguages() {
487 return self
::getFallbacksFor( $this->mCode
);
491 * Exports $wgBookstoreListEn
494 function getBookstoreList() {
495 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
499 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
500 * available in localised form, it will be included in English.
504 public function getNamespaces() {
505 if ( is_null( $this->namespaceNames
) ) {
506 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
508 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
509 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
511 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
513 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
514 if ( $wgMetaNamespaceTalk ) {
515 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
517 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
518 $this->namespaceNames
[NS_PROJECT_TALK
] =
519 $this->fixVariableInNamespace( $talk );
522 # Sometimes a language will be localised but not actually exist on this wiki.
523 foreach ( $this->namespaceNames
as $key => $text ) {
524 if ( !isset( $validNamespaces[$key] ) ) {
525 unset( $this->namespaceNames
[$key] );
529 # The above mixing may leave namespaces out of canonical order.
530 # Re-order by namespace ID number...
531 ksort( $this->namespaceNames
);
533 Hooks
::run( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
536 return $this->namespaceNames
;
540 * Arbitrarily set all of the namespace names at once. Mainly used for testing
541 * @param array $namespaces Array of namespaces (id => name)
543 public function setNamespaces( array $namespaces ) {
544 $this->namespaceNames
= $namespaces;
545 $this->mNamespaceIds
= null;
549 * Resets all of the namespace caches. Mainly used for testing
551 public function resetNamespaces() {
552 $this->namespaceNames
= null;
553 $this->mNamespaceIds
= null;
554 $this->namespaceAliases
= null;
558 * A convenience function that returns getNamespaces() with spaces instead of underscores
559 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
563 function getFormattedNamespaces() {
564 $ns = $this->getNamespaces();
565 foreach ( $ns as $k => $v ) {
566 $ns[$k] = strtr( $v, '_', ' ' );
572 * Get a namespace value by key
575 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
576 * echo $mw_ns; // prints 'MediaWiki'
579 * @param int $index The array key of the namespace to return
580 * @return string|bool String if the namespace value exists, otherwise false
582 function getNsText( $index ) {
583 $ns = $this->getNamespaces();
584 return isset( $ns[$index] ) ?
$ns[$index] : false;
588 * A convenience function that returns the same thing as
589 * getNsText() except with '_' changed to ' ', useful for
593 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
594 * echo $mw_ns; // prints 'MediaWiki talk'
597 * @param int $index The array key of the namespace to return
598 * @return string Namespace name without underscores (empty string if namespace does not exist)
600 function getFormattedNsText( $index ) {
601 $ns = $this->getNsText( $index );
602 return strtr( $ns, '_', ' ' );
606 * Returns gender-dependent namespace alias if available.
607 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
608 * @param int $index Namespace index
609 * @param string $gender Gender key (male, female... )
613 function getGenderNsText( $index, $gender ) {
614 global $wgExtraGenderNamespaces;
616 $ns = $wgExtraGenderNamespaces +
617 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
619 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
623 * Whether this language uses gender-dependent namespace aliases.
624 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
628 function needsGenderDistinction() {
629 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
630 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
631 // $wgExtraGenderNamespaces overrides everything
633 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
634 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
635 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
638 // Check what is in i18n files
639 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
640 return count( $aliases ) > 0;
645 * Get a namespace key by value, case insensitive.
646 * Only matches namespace names for the current language, not the
647 * canonical ones defined in Namespace.php.
649 * @param string $text
650 * @return int|bool An integer if $text is a valid value otherwise false
652 function getLocalNsIndex( $text ) {
653 $lctext = $this->lc( $text );
654 $ids = $this->getNamespaceIds();
655 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
661 function getNamespaceAliases() {
662 if ( is_null( $this->namespaceAliases
) ) {
663 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
667 foreach ( $aliases as $name => $index ) {
668 if ( $index === NS_PROJECT_TALK
) {
669 unset( $aliases[$name] );
670 $name = $this->fixVariableInNamespace( $name );
671 $aliases[$name] = $index;
676 global $wgExtraGenderNamespaces;
677 $genders = $wgExtraGenderNamespaces +
678 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
679 foreach ( $genders as $index => $forms ) {
680 foreach ( $forms as $alias ) {
681 $aliases[$alias] = $index;
685 # Also add converted namespace names as aliases, to avoid confusion.
686 $convertedNames = array();
687 foreach ( $this->getVariants() as $variant ) {
688 if ( $variant === $this->mCode
) {
691 foreach ( $this->getNamespaces() as $ns => $_ ) {
692 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
696 $this->namespaceAliases
= $aliases +
$convertedNames;
699 return $this->namespaceAliases
;
705 function getNamespaceIds() {
706 if ( is_null( $this->mNamespaceIds
) ) {
707 global $wgNamespaceAliases;
708 # Put namespace names and aliases into a hashtable.
709 # If this is too slow, then we should arrange it so that it is done
710 # before caching. The catch is that at pre-cache time, the above
711 # class-specific fixup hasn't been done.
712 $this->mNamespaceIds
= array();
713 foreach ( $this->getNamespaces() as $index => $name ) {
714 $this->mNamespaceIds
[$this->lc( $name )] = $index;
716 foreach ( $this->getNamespaceAliases() as $name => $index ) {
717 $this->mNamespaceIds
[$this->lc( $name )] = $index;
719 if ( $wgNamespaceAliases ) {
720 foreach ( $wgNamespaceAliases as $name => $index ) {
721 $this->mNamespaceIds
[$this->lc( $name )] = $index;
725 return $this->mNamespaceIds
;
729 * Get a namespace key by value, case insensitive. Canonical namespace
730 * names override custom ones defined for the current language.
732 * @param string $text
733 * @return int|bool An integer if $text is a valid value otherwise false
735 function getNsIndex( $text ) {
736 $lctext = $this->lc( $text );
737 $ns = MWNamespace
::getCanonicalIndex( $lctext );
738 if ( $ns !== null ) {
741 $ids = $this->getNamespaceIds();
742 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
746 * short names for language variants used for language conversion links.
748 * @param string $code
749 * @param bool $usemsg Use the "variantname-xyz" message if it exists
752 function getVariantname( $code, $usemsg = true ) {
753 $msg = "variantname-$code";
754 if ( $usemsg && wfMessage( $msg )->exists() ) {
755 return $this->getMessageFromDB( $msg );
757 $name = self
::fetchLanguageName( $code );
759 return $name; # if it's defined as a language name, show that
761 # otherwise, output the language code
767 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
768 * SpecialPageFactory::getLocalNameFor instead.
769 * @param string $name
772 function specialPage( $name ) {
773 $aliases = $this->getSpecialPageAliases();
774 if ( isset( $aliases[$name][0] ) ) {
775 $name = $aliases[$name][0];
777 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
783 function getDatePreferences() {
784 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
790 function getDateFormats() {
791 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
795 * @return array|string
797 function getDefaultDateFormat() {
798 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
799 if ( $df === 'dmy or mdy' ) {
800 global $wgAmericanDates;
801 return $wgAmericanDates ?
'mdy' : 'dmy';
810 function getDatePreferenceMigrationMap() {
811 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
815 * @param string $image
818 function getImageFile( $image ) {
819 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
826 function getImageFiles() {
827 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
833 function getExtraUserToggles() {
834 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
841 function getUserToggle( $tog ) {
842 return $this->getMessageFromDB( "tog-$tog" );
846 * Get native language names, indexed by code.
847 * Only those defined in MediaWiki, no other data like CLDR.
848 * If $customisedOnly is true, only returns codes with a messages file
850 * @param bool $customisedOnly
853 * @deprecated since 1.20, use fetchLanguageNames()
855 public static function getLanguageNames( $customisedOnly = false ) {
856 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
860 * Get translated language names. This is done on best effort and
861 * by default this is exactly the same as Language::getLanguageNames.
862 * The CLDR extension provides translated names.
863 * @param string $code Language code.
864 * @return array Language code => language name
866 * @deprecated since 1.20, use fetchLanguageNames()
868 public static function getTranslatedLanguageNames( $code ) {
869 return self
::fetchLanguageNames( $code, 'all' );
873 * Get an array of language names, indexed by code.
874 * @param null|string $inLanguage Code of language in which to return the names
875 * Use null for autonyms (native names)
876 * @param string $include One of:
877 * 'all' all available languages
878 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
879 * 'mwfile' only if the language is in 'mw' *and* has a message file
880 * @return array Language code => language name
883 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
884 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
885 $cacheKey .= ":$include";
886 if ( self
::$languageNameCache === null ) {
887 self
::$languageNameCache = new MapCacheLRU( 20 );
889 if ( self
::$languageNameCache->has( $cacheKey ) ) {
890 $ret = self
::$languageNameCache->get( $cacheKey );
892 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
893 self
::$languageNameCache->set( $cacheKey, $ret );
899 * Uncached helper for fetchLanguageNames
900 * @param null|string $inLanguage Code of language in which to return the names
901 * Use null for autonyms (native names)
902 * @param string $include One of:
903 * 'all' all available languages
904 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
905 * 'mwfile' only if the language is in 'mw' *and* has a message file
906 * @return array Language code => language name
908 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
909 global $wgExtraLanguageNames;
910 static $coreLanguageNames;
912 if ( $coreLanguageNames === null ) {
914 include "$IP/languages/Names.php";
917 // If passed an invalid language code to use, fallback to en
918 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
925 # TODO: also include when $inLanguage is null, when this code is more efficient
926 Hooks
::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
929 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
930 foreach ( $mwNames as $mwCode => $mwName ) {
931 # - Prefer own MediaWiki native name when not using the hook
932 # - For other names just add if not added through the hook
933 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
934 $names[$mwCode] = $mwName;
938 if ( $include === 'all' ) {
944 $coreCodes = array_keys( $mwNames );
945 foreach ( $coreCodes as $coreCode ) {
946 $returnMw[$coreCode] = $names[$coreCode];
949 if ( $include === 'mwfile' ) {
950 $namesMwFile = array();
951 # We do this using a foreach over the codes instead of a directory
952 # loop so that messages files in extensions will work correctly.
953 foreach ( $returnMw as $code => $value ) {
954 if ( is_readable( self
::getMessagesFileName( $code ) )
955 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
957 $namesMwFile[$code] = $names[$code];
961 ksort( $namesMwFile );
966 # 'mw' option; default if it's not one of the other two options (all/mwfile)
971 * @param string $code The code of the language for which to get the name
972 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
973 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
974 * @return string Language name or empty
977 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
978 $code = strtolower( $code );
979 $array = self
::fetchLanguageNames( $inLanguage, $include );
980 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
984 * Get a message from the MediaWiki namespace.
986 * @param string $msg Message name
989 function getMessageFromDB( $msg ) {
990 return $this->msg( $msg )->text();
994 * Get message object in this language. Only for use inside this class.
996 * @param string $msg Message name
999 protected function msg( $msg ) {
1000 return wfMessage( $msg )->inLanguage( $this );
1004 * Get the native language name of $code.
1005 * Only if defined in MediaWiki, no other data like CLDR.
1006 * @param string $code
1008 * @deprecated since 1.20, use fetchLanguageName()
1010 function getLanguageName( $code ) {
1011 return self
::fetchLanguageName( $code );
1015 * @param string $key
1018 function getMonthName( $key ) {
1019 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
1025 function getMonthNamesArray() {
1026 $monthNames = array( '' );
1027 for ( $i = 1; $i < 13; $i++
) {
1028 $monthNames[] = $this->getMonthName( $i );
1034 * @param string $key
1037 function getMonthNameGen( $key ) {
1038 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1042 * @param string $key
1045 function getMonthAbbreviation( $key ) {
1046 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1052 function getMonthAbbreviationsArray() {
1053 $monthNames = array( '' );
1054 for ( $i = 1; $i < 13; $i++
) {
1055 $monthNames[] = $this->getMonthAbbreviation( $i );
1061 * @param string $key
1064 function getWeekdayName( $key ) {
1065 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1069 * @param string $key
1072 function getWeekdayAbbreviation( $key ) {
1073 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1077 * @param string $key
1080 function getIranianCalendarMonthName( $key ) {
1081 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1085 * @param string $key
1088 function getHebrewCalendarMonthName( $key ) {
1089 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1093 * @param string $key
1096 function getHebrewCalendarMonthNameGen( $key ) {
1097 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1101 * @param string $key
1104 function getHijriCalendarMonthName( $key ) {
1105 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1109 * Pass through result from $dateTimeObj->format()
1110 * @param DateTime|bool|null &$dateTimeObj
1112 * @param DateTimeZone|bool|null $zone
1113 * @param string $code
1116 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1117 if ( !$dateTimeObj ) {
1118 $dateTimeObj = DateTime
::createFromFormat(
1119 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1122 return $dateTimeObj->format( $code );
1126 * This is a workalike of PHP's date() function, but with better
1127 * internationalisation, a reduced set of format characters, and a better
1130 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1131 * the PHP manual for definitions. There are a number of extensions, which
1134 * xn Do not translate digits of the next numeric format character
1135 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1136 * xr Use roman numerals for the next numeric format character
1137 * xh Use hebrew numerals for the next numeric format character
1139 * xg Genitive month name
1141 * xij j (day number) in Iranian calendar
1142 * xiF F (month name) in Iranian calendar
1143 * xin n (month number) in Iranian calendar
1144 * xiy y (two digit year) in Iranian calendar
1145 * xiY Y (full year) in Iranian calendar
1147 * xjj j (day number) in Hebrew calendar
1148 * xjF F (month name) in Hebrew calendar
1149 * xjt t (days in month) in Hebrew calendar
1150 * xjx xg (genitive month name) in Hebrew calendar
1151 * xjn n (month number) in Hebrew calendar
1152 * xjY Y (full year) in Hebrew calendar
1154 * xmj j (day number) in Hijri calendar
1155 * xmF F (month name) in Hijri calendar
1156 * xmn n (month number) in Hijri calendar
1157 * xmY Y (full year) in Hijri calendar
1159 * xkY Y (full year) in Thai solar calendar. Months and days are
1160 * identical to the Gregorian calendar
1161 * xoY Y (full year) in Minguo calendar or Juche year.
1162 * Months and days are identical to the
1163 * Gregorian calendar
1164 * xtY Y (full year) in Japanese nengo. Months and days are
1165 * identical to the Gregorian calendar
1167 * Characters enclosed in double quotes will be considered literal (with
1168 * the quotes themselves removed). Unmatched quotes will be considered
1169 * literal quotes. Example:
1171 * "The month is" F => The month is January
1174 * Backslash escaping is also supported.
1176 * Input timestamp is assumed to be pre-normalized to the desired local
1177 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1178 * $ts is UTC if $zone is not given.
1180 * @param string $format
1181 * @param string $ts 14-character timestamp
1184 * @param DateTimeZone $zone Timezone of $ts
1185 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1186 * Only makes sense if $ts is the current time.
1187 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1189 * @throws MWException
1192 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1197 $dateTimeObj = false;
1206 $usedSecond = false;
1207 $usedMinute = false;
1214 $usedISOYear = false;
1215 $usedIsLeapYear = false;
1217 $usedHebrewMonth = false;
1218 $usedIranianMonth = false;
1219 $usedHijriMonth = false;
1220 $usedHebrewYear = false;
1221 $usedIranianYear = false;
1222 $usedHijriYear = false;
1223 $usedTennoYear = false;
1225 if ( strlen( $ts ) !== 14 ) {
1226 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1229 if ( !ctype_digit( $ts ) ) {
1230 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1233 $formatLength = strlen( $format );
1234 for ( $p = 0; $p < $formatLength; $p++
) {
1236 $code = $format[$p];
1237 if ( $code == 'x' && $p < $formatLength - 1 ) {
1238 $code .= $format[++
$p];
1241 if ( ( $code === 'xi'
1247 && $p < $formatLength - 1 ) {
1248 $code .= $format[++
$p];
1259 $rawToggle = !$rawToggle;
1269 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1272 $usedHebrewMonth = true;
1274 $hebrew = self
::tsToHebrew( $ts );
1276 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1280 $num = substr( $ts, 6, 2 );
1284 $s .= $this->getWeekdayAbbreviation(
1285 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1290 $num = intval( substr( $ts, 6, 2 ) );
1295 $iranian = self
::tsToIranian( $ts );
1302 $hijri = self
::tsToHijri( $ts );
1309 $hebrew = self
::tsToHebrew( $ts );
1315 $s .= $this->getWeekdayName(
1316 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1321 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1324 $usedIranianMonth = true;
1326 $iranian = self
::tsToIranian( $ts );
1328 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1331 $usedHijriMonth = true;
1333 $hijri = self
::tsToHijri( $ts );
1335 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1338 $usedHebrewMonth = true;
1340 $hebrew = self
::tsToHebrew( $ts );
1342 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1346 $num = substr( $ts, 4, 2 );
1350 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1354 $num = intval( substr( $ts, 4, 2 ) );
1357 $usedIranianMonth = true;
1359 $iranian = self
::tsToIranian( $ts );
1364 $usedHijriMonth = true;
1366 $hijri = self
::tsToHijri( $ts );
1371 $usedHebrewMonth = true;
1373 $hebrew = self
::tsToHebrew( $ts );
1378 $usedHebrewMonth = true;
1380 $hebrew = self
::tsToHebrew( $ts );
1386 $num = substr( $ts, 0, 4 );
1389 $usedIranianYear = true;
1391 $iranian = self
::tsToIranian( $ts );
1396 $usedHijriYear = true;
1398 $hijri = self
::tsToHijri( $ts );
1403 $usedHebrewYear = true;
1405 $hebrew = self
::tsToHebrew( $ts );
1412 $thai = self
::tsToYear( $ts, 'thai' );
1419 $minguo = self
::tsToYear( $ts, 'minguo' );
1424 $usedTennoYear = true;
1426 $tenno = self
::tsToYear( $ts, 'tenno' );
1432 $num = substr( $ts, 2, 2 );
1435 $usedIranianYear = true;
1437 $iranian = self
::tsToIranian( $ts );
1439 $num = substr( $iranian[0], -2 );
1443 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1447 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1451 $h = substr( $ts, 8, 2 );
1452 $num = $h %
12 ?
$h %
12 : 12;
1456 $num = intval( substr( $ts, 8, 2 ) );
1460 $h = substr( $ts, 8, 2 );
1461 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1465 $num = substr( $ts, 8, 2 );
1469 $num = substr( $ts, 10, 2 );
1473 $num = substr( $ts, 12, 2 );
1483 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1489 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1493 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1497 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1500 $usedIsLeapYear = true;
1501 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1504 $usedISOYear = true;
1505 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1512 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1515 # Backslash escaping
1516 if ( $p < $formatLength - 1 ) {
1517 $s .= $format[++
$p];
1524 if ( $p < $formatLength - 1 ) {
1525 $endQuote = strpos( $format, '"', $p +
1 );
1526 if ( $endQuote === false ) {
1527 # No terminating quote, assume literal "
1530 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1534 # Quote at end of string, assume literal "
1541 if ( $num !== false ) {
1542 if ( $rawToggle ||
$raw ) {
1545 } elseif ( $roman ) {
1546 $s .= Language
::romanNumeral( $num );
1548 } elseif ( $hebrewNum ) {
1549 $s .= self
::hebrewNumeral( $num );
1552 $s .= $this->formatNum( $num, true );
1557 if ( $usedSecond ) {
1559 } elseif ( $usedMinute ) {
1560 $ttl = 60 - substr( $ts, 12, 2 );
1561 } elseif ( $usedHour ) {
1562 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1563 } elseif ( $usedAMPM ) {
1564 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1565 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1569 $usedIranianMonth ||
1576 // @todo Someone who understands the non-Gregorian calendars
1577 // should write proper logic for them so that they don't need purged every day.
1578 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1579 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1581 $possibleTtls = array();
1582 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1583 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1586 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1587 $timeRemainingInDay;
1588 } elseif ( $usedISOYear ) {
1589 // December 28th falls on the last ISO week of the year, every year.
1590 // The last ISO week of a year can be 52 or 53.
1591 $lastWeekOfISOYear = DateTime
::createFromFormat(
1593 substr( $ts, 0, 4 ) . '1228',
1594 $zone ?
: new DateTimeZone( 'UTC' )
1596 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1597 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1598 $timeRemainingInWeek =
1599 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1600 +
$timeRemainingInDay;
1601 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1606 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1607 substr( $ts, 6, 2 ) ) * 86400
1608 +
$timeRemainingInDay;
1609 } elseif ( $usedYear ) {
1611 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1612 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1613 +
$timeRemainingInDay;
1614 } elseif ( $usedIsLeapYear ) {
1615 $year = substr( $ts, 0, 4 );
1616 $timeRemainingInYear =
1617 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1618 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1619 +
$timeRemainingInDay;
1621 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1622 // this isn't a leap year. see when the next one starts
1623 $nextCandidate = $year - $mod +
4;
1624 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1625 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1626 $timeRemainingInYear;
1628 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1629 $timeRemainingInYear;
1632 // this is a leap year, so the next year isn't
1633 $possibleTtls[] = $timeRemainingInYear;
1637 if ( $possibleTtls ) {
1638 $ttl = min( $possibleTtls );
1645 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1646 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1649 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1650 * Gregorian dates to Iranian dates. Originally written in C, it
1651 * is released under the terms of GNU Lesser General Public
1652 * License. Conversion to PHP was performed by Niklas Laxström.
1654 * Link: http://www.farsiweb.info/jalali/jalali.c
1660 private static function tsToIranian( $ts ) {
1661 $gy = substr( $ts, 0, 4 ) -1600;
1662 $gm = substr( $ts, 4, 2 ) -1;
1663 $gd = substr( $ts, 6, 2 ) -1;
1665 # Days passed from the beginning (including leap years)
1667 +
floor( ( $gy +
3 ) / 4 )
1668 - floor( ( $gy +
99 ) / 100 )
1669 +
floor( ( $gy +
399 ) / 400 );
1671 // Add days of the past months of this year
1672 for ( $i = 0; $i < $gm; $i++
) {
1673 $gDayNo +
= self
::$GREG_DAYS[$i];
1677 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1681 // Days passed in current month
1682 $gDayNo +
= (int)$gd;
1684 $jDayNo = $gDayNo - 79;
1686 $jNp = floor( $jDayNo / 12053 );
1689 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1692 if ( $jDayNo >= 366 ) {
1693 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1694 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1697 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1698 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1704 return array( $jy, $jm, $jd );
1708 * Converting Gregorian dates to Hijri dates.
1710 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1712 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1718 private static function tsToHijri( $ts ) {
1719 $year = substr( $ts, 0, 4 );
1720 $month = substr( $ts, 4, 2 );
1721 $day = substr( $ts, 6, 2 );
1729 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1730 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1732 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1733 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1734 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1737 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1738 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1741 $zl = $zjd -1948440 +
10632;
1742 $zn = (int)( ( $zl - 1 ) / 10631 );
1743 $zl = $zl - 10631 * $zn +
354;
1744 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1745 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1746 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1747 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1748 $zm = (int)( ( 24 * $zl ) / 709 );
1749 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1750 $zy = 30 * $zn +
$zj - 30;
1752 return array( $zy, $zm, $zd );
1756 * Converting Gregorian dates to Hebrew dates.
1758 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1759 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1760 * to translate the relevant functions into PHP and release them under
1763 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1764 * and Adar II is 14. In a non-leap year, Adar is 6.
1770 private static function tsToHebrew( $ts ) {
1772 $year = substr( $ts, 0, 4 );
1773 $month = substr( $ts, 4, 2 );
1774 $day = substr( $ts, 6, 2 );
1776 # Calculate Hebrew year
1777 $hebrewYear = $year +
3760;
1779 # Month number when September = 1, August = 12
1781 if ( $month > 12 ) {
1788 # Calculate day of year from 1 September
1790 for ( $i = 1; $i < $month; $i++
) {
1794 # Check if the year is leap
1795 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1798 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1805 # Calculate the start of the Hebrew year
1806 $start = self
::hebrewYearStart( $hebrewYear );
1808 # Calculate next year's start
1809 if ( $dayOfYear <= $start ) {
1810 # Day is before the start of the year - it is the previous year
1812 $nextStart = $start;
1816 # Add days since previous year's 1 September
1818 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1822 # Start of the new (previous) year
1823 $start = self
::hebrewYearStart( $hebrewYear );
1826 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1829 # Calculate Hebrew day of year
1830 $hebrewDayOfYear = $dayOfYear - $start;
1832 # Difference between year's days
1833 $diff = $nextStart - $start;
1834 # Add 12 (or 13 for leap years) days to ignore the difference between
1835 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1836 # difference is only about the year type
1837 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1843 # Check the year pattern, and is leap year
1844 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1845 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1846 # and non-leap years
1847 $yearPattern = $diff %
30;
1848 # Check if leap year
1849 $isLeap = $diff >= 30;
1851 # Calculate day in the month from number of day in the Hebrew year
1852 # Don't check Adar - if the day is not in Adar, we will stop before;
1853 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1854 $hebrewDay = $hebrewDayOfYear;
1857 while ( $hebrewMonth <= 12 ) {
1858 # Calculate days in this month
1859 if ( $isLeap && $hebrewMonth == 6 ) {
1860 # Adar in a leap year
1862 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1864 if ( $hebrewDay <= $days ) {
1868 # Subtract the days of Adar I
1869 $hebrewDay -= $days;
1872 if ( $hebrewDay <= $days ) {
1878 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1879 # Cheshvan in a complete year (otherwise as the rule below)
1881 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1882 # Kislev in an incomplete year (otherwise as the rule below)
1885 # Odd months have 30 days, even have 29
1886 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1888 if ( $hebrewDay <= $days ) {
1889 # In the current month
1892 # Subtract the days of the current month
1893 $hebrewDay -= $days;
1894 # Try in the next month
1899 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1903 * This calculates the Hebrew year start, as days since 1 September.
1904 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1905 * Used for Hebrew date.
1911 private static function hebrewYearStart( $year ) {
1912 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1913 $b = intval( ( $year - 1 ) %
4 );
1914 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1918 $Mar = intval( $m );
1924 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1925 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1927 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1929 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1933 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1938 * Algorithm to convert Gregorian dates to Thai solar dates,
1939 * Minguo dates or Minguo dates.
1941 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1942 * http://en.wikipedia.org/wiki/Minguo_calendar
1943 * http://en.wikipedia.org/wiki/Japanese_era_name
1945 * @param string $ts 14-character timestamp
1946 * @param string $cName Calender name
1947 * @return array Converted year, month, day
1949 private static function tsToYear( $ts, $cName ) {
1950 $gy = substr( $ts, 0, 4 );
1951 $gm = substr( $ts, 4, 2 );
1952 $gd = substr( $ts, 6, 2 );
1954 if ( !strcmp( $cName, 'thai' ) ) {
1956 # Add 543 years to the Gregorian calendar
1957 # Months and days are identical
1958 $gy_offset = $gy +
543;
1959 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1961 # Deduct 1911 years from the Gregorian calendar
1962 # Months and days are identical
1963 $gy_offset = $gy - 1911;
1964 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1965 # Nengō dates up to Meiji period
1966 # Deduct years from the Gregorian calendar
1967 # depending on the nengo periods
1968 # Months and days are identical
1970 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1971 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1974 $gy_gannen = $gy - 1868 +
1;
1975 $gy_offset = $gy_gannen;
1976 if ( $gy_gannen == 1 ) {
1979 $gy_offset = '明治' . $gy_offset;
1981 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1982 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1983 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1984 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1985 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1988 $gy_gannen = $gy - 1912 +
1;
1989 $gy_offset = $gy_gannen;
1990 if ( $gy_gannen == 1 ) {
1993 $gy_offset = '大正' . $gy_offset;
1995 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1996 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1997 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
2000 $gy_gannen = $gy - 1926 +
1;
2001 $gy_offset = $gy_gannen;
2002 if ( $gy_gannen == 1 ) {
2005 $gy_offset = '昭和' . $gy_offset;
2008 $gy_gannen = $gy - 1989 +
1;
2009 $gy_offset = $gy_gannen;
2010 if ( $gy_gannen == 1 ) {
2013 $gy_offset = '平成' . $gy_offset;
2019 return array( $gy_offset, $gm, $gd );
2023 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2025 * This is the rule the BIDI algorithm uses to determine the directionality of
2026 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
2027 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2029 * TODO: Does not handle BIDI control characters inside the text.
2030 * TODO: Does not handle unallocated characters.
2032 * @param string $text Text to test
2033 * @return null|string Directionality ('ltr' or 'rtl') or null
2035 private static function strongDirFromContent( $text = '' ) {
2036 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2039 if ( $matches[1] === '' ) {
2046 * Roman number formatting up to 10000
2052 static function romanNumeral( $num ) {
2053 static $table = array(
2054 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
2055 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
2056 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
2057 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2058 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
2061 $num = intval( $num );
2062 if ( $num > 10000 ||
$num <= 0 ) {
2067 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2068 if ( $num >= $pow10 ) {
2069 $s .= $table[$i][(int)floor( $num / $pow10 )];
2071 $num = $num %
$pow10;
2077 * Hebrew Gematria number formatting up to 9999
2083 static function hebrewNumeral( $num ) {
2084 static $table = array(
2085 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2086 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2096 array( 'ת', 'ת', 'ק' ),
2097 array( 'ת', 'ת', 'ר' ),
2099 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2102 $num = intval( $num );
2103 if ( $num > 9999 ||
$num <= 0 ) {
2107 // Round thousands have special notations
2108 if ( $num === 1000 ) {
2110 } elseif ( $num %
1000 === 0 ) {
2111 return $table[0][$num / 1000] . "' אלפים";
2116 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2117 if ( $num >= $pow10 ) {
2118 if ( $num === 15 ||
$num === 16 ) {
2119 $letters[] = $table[0][9];
2120 $letters[] = $table[0][$num - 9];
2123 $letters = array_merge(
2125 (array)$table[$i][intval( $num / $pow10 )]
2128 if ( $pow10 === 1000 ) {
2134 $num = $num %
$pow10;
2137 $preTransformLength = count( $letters );
2138 if ( $preTransformLength === 1 ) {
2139 // Add geresh (single quote) to one-letter numbers
2142 $lastIndex = $preTransformLength - 1;
2143 $letters[$lastIndex] = str_replace(
2144 array( 'כ', 'מ', 'נ', 'פ', 'צ' ),
2145 array( 'ך', 'ם', 'ן', 'ף', 'ץ' ),
2146 $letters[$lastIndex]
2149 // Add gershayim (double quote) to multiple-letter numbers,
2150 // but exclude numbers with only one letter after the thousands
2151 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2152 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2155 array_splice( $letters, -1, 0, '"' );
2159 return implode( $letters );
2163 * Used by date() and time() to adjust the time output.
2165 * @param string $ts The time in date('YmdHis') format
2166 * @param mixed $tz Adjust the time by this amount (default false, mean we
2167 * get user timecorrection setting)
2170 function userAdjust( $ts, $tz = false ) {
2171 global $wgUser, $wgLocalTZoffset;
2173 if ( $tz === false ) {
2174 $tz = $wgUser->getOption( 'timecorrection' );
2177 $data = explode( '|', $tz, 3 );
2179 if ( $data[0] == 'ZoneInfo' ) {
2180 MediaWiki\
suppressWarnings();
2181 $userTZ = timezone_open( $data[2] );
2182 MediaWiki\restoreWarnings
();
2183 if ( $userTZ !== false ) {
2184 $date = date_create( $ts, timezone_open( 'UTC' ) );
2185 date_timezone_set( $date, $userTZ );
2186 $date = date_format( $date, 'YmdHis' );
2189 # Unrecognized timezone, default to 'Offset' with the stored offset.
2190 $data[0] = 'Offset';
2193 if ( $data[0] == 'System' ||
$tz == '' ) {
2194 # Global offset in minutes.
2195 $minDiff = $wgLocalTZoffset;
2196 } elseif ( $data[0] == 'Offset' ) {
2197 $minDiff = intval( $data[1] );
2199 $data = explode( ':', $tz );
2200 if ( count( $data ) == 2 ) {
2201 $data[0] = intval( $data[0] );
2202 $data[1] = intval( $data[1] );
2203 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2204 if ( $data[0] < 0 ) {
2205 $minDiff = -$minDiff;
2208 $minDiff = intval( $data[0] ) * 60;
2212 # No difference ? Return time unchanged
2213 if ( 0 == $minDiff ) {
2217 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2218 # Generate an adjusted date; take advantage of the fact that mktime
2219 # will normalize out-of-range values so we don't have to split $minDiff
2220 # into hours and minutes.
2222 (int)substr( $ts, 8, 2 ) ), # Hours
2223 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2224 (int)substr( $ts, 12, 2 ), # Seconds
2225 (int)substr( $ts, 4, 2 ), # Month
2226 (int)substr( $ts, 6, 2 ), # Day
2227 (int)substr( $ts, 0, 4 ) ); # Year
2229 $date = date( 'YmdHis', $t );
2230 MediaWiki\restoreWarnings
();
2236 * This is meant to be used by time(), date(), and timeanddate() to get
2237 * the date preference they're supposed to use, it should be used in
2241 * function timeanddate([...], $format = true) {
2242 * $datePreference = $this->dateFormat($format);
2247 * @param int|string|bool $usePrefs If true, the user's preference is used
2248 * if false, the site/language default is used
2249 * if int/string, assumed to be a format.
2252 function dateFormat( $usePrefs = true ) {
2255 if ( is_bool( $usePrefs ) ) {
2257 $datePreference = $wgUser->getDatePreference();
2259 $datePreference = (string)User
::getDefaultOption( 'date' );
2262 $datePreference = (string)$usePrefs;
2266 if ( $datePreference == '' ) {
2270 return $datePreference;
2274 * Get a format string for a given type and preference
2275 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2276 * @param string $pref The format name as it appears in Messages*.php under
2279 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2283 function getDateFormatString( $type, $pref ) {
2284 $wasDefault = false;
2285 if ( $pref == 'default' ) {
2287 $pref = $this->getDefaultDateFormat();
2290 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2291 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2293 if ( $type === 'pretty' && $df === null ) {
2294 $df = $this->getDateFormatString( 'date', $pref );
2297 if ( !$wasDefault && $df === null ) {
2298 $pref = $this->getDefaultDateFormat();
2299 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2302 $this->dateFormatStrings
[$type][$pref] = $df;
2304 return $this->dateFormatStrings
[$type][$pref];
2308 * @param string $ts The time format which needs to be turned into a
2309 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2310 * @param bool $adj Whether to adjust the time output according to the
2311 * user configured offset ($timecorrection)
2312 * @param mixed $format True to use user's date format preference
2313 * @param string|bool $timecorrection The time offset as returned by
2314 * validateTimeZone() in Special:Preferences
2317 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2318 $ts = wfTimestamp( TS_MW
, $ts );
2320 $ts = $this->userAdjust( $ts, $timecorrection );
2322 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2323 return $this->sprintfDate( $df, $ts );
2327 * @param string $ts The time format which needs to be turned into a
2328 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2329 * @param bool $adj Whether to adjust the time output according to the
2330 * user configured offset ($timecorrection)
2331 * @param mixed $format True to use user's date format preference
2332 * @param string|bool $timecorrection The time offset as returned by
2333 * validateTimeZone() in Special:Preferences
2336 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2337 $ts = wfTimestamp( TS_MW
, $ts );
2339 $ts = $this->userAdjust( $ts, $timecorrection );
2341 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2342 return $this->sprintfDate( $df, $ts );
2346 * @param string $ts The time format which needs to be turned into a
2347 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2348 * @param bool $adj Whether to adjust the time output according to the
2349 * user configured offset ($timecorrection)
2350 * @param mixed $format What format to return, if it's false output the
2351 * default one (default true)
2352 * @param string|bool $timecorrection The time offset as returned by
2353 * validateTimeZone() in Special:Preferences
2356 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2357 $ts = wfTimestamp( TS_MW
, $ts );
2359 $ts = $this->userAdjust( $ts, $timecorrection );
2361 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2362 return $this->sprintfDate( $df, $ts );
2366 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2370 * @param int $seconds The amount of seconds.
2371 * @param array $chosenIntervals The intervals to enable.
2375 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2376 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2378 $segments = array();
2380 foreach ( $intervals as $intervalName => $intervalValue ) {
2381 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2382 // duration-years, duration-decades, duration-centuries, duration-millennia
2383 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2384 $segments[] = $message->inLanguage( $this )->escaped();
2387 return $this->listToText( $segments );
2391 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2392 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2396 * @param int $seconds The amount of seconds.
2397 * @param array $chosenIntervals The intervals to enable.
2401 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2402 if ( empty( $chosenIntervals ) ) {
2403 $chosenIntervals = array(
2415 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2416 $sortedNames = array_keys( $intervals );
2417 $smallestInterval = array_pop( $sortedNames );
2419 $segments = array();
2421 foreach ( $intervals as $name => $length ) {
2422 $value = floor( $seconds / $length );
2424 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2425 $seconds -= $value * $length;
2426 $segments[$name] = $value;
2434 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2436 * @param string $type Can be 'date', 'time' or 'both'
2437 * @param string $ts The time format which needs to be turned into a
2438 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2439 * @param User $user User object used to get preferences for timezone and format
2440 * @param array $options Array, can contain the following keys:
2441 * - 'timecorrection': time correction, can have the following values:
2442 * - true: use user's preference
2443 * - false: don't use time correction
2444 * - int: value of time correction in minutes
2445 * - 'format': format to use, can have the following values:
2446 * - true: use user's preference
2447 * - false: use default preference
2448 * - string: format to use
2452 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2453 $ts = wfTimestamp( TS_MW
, $ts );
2454 $options +
= array( 'timecorrection' => true, 'format' => true );
2455 if ( $options['timecorrection'] !== false ) {
2456 if ( $options['timecorrection'] === true ) {
2457 $offset = $user->getOption( 'timecorrection' );
2459 $offset = $options['timecorrection'];
2461 $ts = $this->userAdjust( $ts, $offset );
2463 if ( $options['format'] === true ) {
2464 $format = $user->getDatePreference();
2466 $format = $options['format'];
2468 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2469 return $this->sprintfDate( $df, $ts );
2473 * Get the formatted date for the given timestamp and formatted for
2476 * @param mixed $ts Mixed: the time format which needs to be turned into a
2477 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2478 * @param User $user User object used to get preferences for timezone and format
2479 * @param array $options Array, can contain the following keys:
2480 * - 'timecorrection': time correction, can have the following values:
2481 * - true: use user's preference
2482 * - false: don't use time correction
2483 * - int: value of time correction in minutes
2484 * - 'format': format to use, can have the following values:
2485 * - true: use user's preference
2486 * - false: use default preference
2487 * - string: format to use
2491 public function userDate( $ts, User
$user, array $options = array() ) {
2492 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2496 * Get the formatted time for the given timestamp and formatted for
2499 * @param mixed $ts The time format which needs to be turned into a
2500 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2501 * @param User $user User object used to get preferences for timezone and format
2502 * @param array $options Array, can contain the following keys:
2503 * - 'timecorrection': time correction, can have the following values:
2504 * - true: use user's preference
2505 * - false: don't use time correction
2506 * - int: value of time correction in minutes
2507 * - 'format': format to use, can have the following values:
2508 * - true: use user's preference
2509 * - false: use default preference
2510 * - string: format to use
2514 public function userTime( $ts, User
$user, array $options = array() ) {
2515 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2519 * Get the formatted date and time for the given timestamp and formatted for
2522 * @param mixed $ts The time format which needs to be turned into a
2523 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2524 * @param User $user User object used to get preferences for timezone and format
2525 * @param array $options Array, can contain the following keys:
2526 * - 'timecorrection': time correction, can have the following values:
2527 * - true: use user's preference
2528 * - false: don't use time correction
2529 * - int: value of time correction in minutes
2530 * - 'format': format to use, can have the following values:
2531 * - true: use user's preference
2532 * - false: use default preference
2533 * - string: format to use
2537 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2538 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2542 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2544 * Determine the difference between the timestamp and the current time, and
2545 * generate a readable timestamp by returning "<N> <units> ago", where the
2546 * largest possible unit is used.
2548 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2550 * @param MWTimestamp $time
2551 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2552 * @param User|null $user User the timestamp is being generated for (or null to use main context's user)
2553 * @return string Formatted timestamp
2555 public function getHumanTimestamp( MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null ) {
2556 if ( $relativeTo === null ) {
2557 $relativeTo = new MWTimestamp();
2559 if ( $user === null ) {
2560 $user = RequestContext
::getMain()->getUser();
2563 // Adjust for the user's timezone.
2564 $offsetThis = $time->offsetForUser( $user );
2565 $offsetRel = $relativeTo->offsetForUser( $user );
2568 if ( Hooks
::run( 'GetHumanTimestamp', array( &$ts, $time, $relativeTo, $user, $this ) ) ) {
2569 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2572 // Reset the timezone on the objects.
2573 $time->timestamp
->sub( $offsetThis );
2574 $relativeTo->timestamp
->sub( $offsetRel );
2580 * Convert an MWTimestamp into a pretty human-readable timestamp using
2581 * the given user preferences and relative base time.
2583 * @see Language::getHumanTimestamp
2584 * @param MWTimestamp $ts Timestamp to prettify
2585 * @param MWTimestamp $relativeTo Base timestamp
2586 * @param User $user User preferences to use
2587 * @return string Human timestamp
2590 private function getHumanTimestampInternal( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2591 $diff = $ts->diff( $relativeTo );
2592 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2593 (int)$relativeTo->timestamp
->format( 'w' ) );
2594 $days = $diff->days ?
: (int)$diffDay;
2595 if ( $diff->invert ||
$days > 5
2596 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2598 // Timestamps are in different years: use full timestamp
2599 // Also do full timestamp for future dates
2601 * @todo FIXME: Add better handling of future timestamps.
2603 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2604 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2605 } elseif ( $days > 5 ) {
2606 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2607 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2608 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2609 } elseif ( $days > 1 ) {
2610 // Timestamp within the past week: show the day of the week and time
2611 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2612 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2614 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2615 $ts = wfMessage( "$weekday-at" )
2616 ->inLanguage( $this )
2617 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2619 } elseif ( $days == 1 ) {
2620 // Timestamp was yesterday: say 'yesterday' and the time.
2621 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2622 $ts = wfMessage( 'yesterday-at' )
2623 ->inLanguage( $this )
2624 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2626 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2627 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2628 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2629 $ts = wfMessage( 'today-at' )
2630 ->inLanguage( $this )
2631 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2634 // From here on in, the timestamp was soon enough ago so that we can simply say
2635 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2636 } elseif ( $diff->h
== 1 ) {
2637 // Less than 90 minutes, but more than an hour ago.
2638 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2639 } elseif ( $diff->i
>= 1 ) {
2640 // A few minutes ago.
2641 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2642 } elseif ( $diff->s
>= 30 ) {
2643 // Less than a minute, but more than 30 sec ago.
2644 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2646 // Less than 30 seconds ago.
2647 $ts = wfMessage( 'just-now' )->text();
2654 * @param string $key
2655 * @return array|null
2657 function getMessage( $key ) {
2658 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2664 function getAllMessages() {
2665 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2670 * @param string $out
2671 * @param string $string
2674 function iconv( $in, $out, $string ) {
2675 # This is a wrapper for iconv in all languages except esperanto,
2676 # which does some nasty x-conversions beforehand
2678 # Even with //IGNORE iconv can whine about illegal characters in
2679 # *input* string. We just ignore those too.
2680 # REF: http://bugs.php.net/bug.php?id=37166
2681 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2682 MediaWiki\
suppressWarnings();
2683 $text = iconv( $in, $out . '//IGNORE', $string );
2684 MediaWiki\restoreWarnings
();
2688 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2691 * @param array $matches
2692 * @return mixed|string
2694 function ucwordbreaksCallbackAscii( $matches ) {
2695 return $this->ucfirst( $matches[1] );
2699 * @param array $matches
2702 function ucwordbreaksCallbackMB( $matches ) {
2703 return mb_strtoupper( $matches[0] );
2707 * @param array $matches
2710 function ucCallback( $matches ) {
2711 list( $wikiUpperChars ) = self
::getCaseMaps();
2712 return strtr( $matches[1], $wikiUpperChars );
2716 * @param array $matches
2719 function lcCallback( $matches ) {
2720 list( , $wikiLowerChars ) = self
::getCaseMaps();
2721 return strtr( $matches[1], $wikiLowerChars );
2725 * @param array $matches
2728 function ucwordsCallbackMB( $matches ) {
2729 return mb_strtoupper( $matches[0] );
2733 * @param array $matches
2736 function ucwordsCallbackWiki( $matches ) {
2737 list( $wikiUpperChars ) = self
::getCaseMaps();
2738 return strtr( $matches[0], $wikiUpperChars );
2742 * Make a string's first character uppercase
2744 * @param string $str
2748 function ucfirst( $str ) {
2750 if ( $o < 96 ) { // if already uppercase...
2752 } elseif ( $o < 128 ) {
2753 return ucfirst( $str ); // use PHP's ucfirst()
2755 // fall back to more complex logic in case of multibyte strings
2756 return $this->uc( $str, true );
2761 * Convert a string to uppercase
2763 * @param string $str
2764 * @param bool $first
2768 function uc( $str, $first = false ) {
2769 if ( function_exists( 'mb_strtoupper' ) ) {
2771 if ( $this->isMultibyte( $str ) ) {
2772 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2774 return ucfirst( $str );
2777 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2780 if ( $this->isMultibyte( $str ) ) {
2781 $x = $first ?
'^' : '';
2782 return preg_replace_callback(
2783 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2784 array( $this, 'ucCallback' ),
2788 return $first ?
ucfirst( $str ) : strtoupper( $str );
2794 * @param string $str
2795 * @return mixed|string
2797 function lcfirst( $str ) {
2800 return strval( $str );
2801 } elseif ( $o >= 128 ) {
2802 return $this->lc( $str, true );
2803 } elseif ( $o > 96 ) {
2806 $str[0] = strtolower( $str[0] );
2812 * @param string $str
2813 * @param bool $first
2814 * @return mixed|string
2816 function lc( $str, $first = false ) {
2817 if ( function_exists( 'mb_strtolower' ) ) {
2819 if ( $this->isMultibyte( $str ) ) {
2820 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2822 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2825 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2828 if ( $this->isMultibyte( $str ) ) {
2829 $x = $first ?
'^' : '';
2830 return preg_replace_callback(
2831 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2832 array( $this, 'lcCallback' ),
2836 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2842 * @param string $str
2845 function isMultibyte( $str ) {
2846 return strlen( $str ) !== mb_strlen( $str );
2850 * @param string $str
2851 * @return mixed|string
2853 function ucwords( $str ) {
2854 if ( $this->isMultibyte( $str ) ) {
2855 $str = $this->lc( $str );
2857 // regexp to find first letter in each word (i.e. after each space)
2858 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2860 // function to use to capitalize a single char
2861 if ( function_exists( 'mb_strtoupper' ) ) {
2862 return preg_replace_callback(
2864 array( $this, 'ucwordsCallbackMB' ),
2868 return preg_replace_callback(
2870 array( $this, 'ucwordsCallbackWiki' ),
2875 return ucwords( strtolower( $str ) );
2880 * capitalize words at word breaks
2882 * @param string $str
2885 function ucwordbreaks( $str ) {
2886 if ( $this->isMultibyte( $str ) ) {
2887 $str = $this->lc( $str );
2889 // since \b doesn't work for UTF-8, we explicitely define word break chars
2890 $breaks = "[ \-\(\)\}\{\.,\?!]";
2892 // find first letter after word break
2893 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2894 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2896 if ( function_exists( 'mb_strtoupper' ) ) {
2897 return preg_replace_callback(
2899 array( $this, 'ucwordbreaksCallbackMB' ),
2903 return preg_replace_callback(
2905 array( $this, 'ucwordsCallbackWiki' ),
2910 return preg_replace_callback(
2911 '/\b([\w\x80-\xff]+)\b/',
2912 array( $this, 'ucwordbreaksCallbackAscii' ),
2919 * Return a case-folded representation of $s
2921 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2922 * and $s2 are the same except for the case of their characters. It is not
2923 * necessary for the value returned to make sense when displayed.
2925 * Do *not* perform any other normalisation in this function. If a caller
2926 * uses this function when it should be using a more general normalisation
2927 * function, then fix the caller.
2933 function caseFold( $s ) {
2934 return $this->uc( $s );
2941 function checkTitleEncoding( $s ) {
2942 if ( is_array( $s ) ) {
2943 throw new MWException( 'Given array to checkTitleEncoding.' );
2945 if ( StringUtils
::isUtf8( $s ) ) {
2949 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2955 function fallback8bitEncoding() {
2956 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2960 * Most writing systems use whitespace to break up words.
2961 * Some languages such as Chinese don't conventionally do this,
2962 * which requires special handling when breaking up words for
2967 function hasWordBreaks() {
2972 * Some languages such as Chinese require word segmentation,
2973 * Specify such segmentation when overridden in derived class.
2975 * @param string $string
2978 function segmentByWord( $string ) {
2983 * Some languages have special punctuation need to be normalized.
2984 * Make such changes here.
2986 * @param string $string
2989 function normalizeForSearch( $string ) {
2990 return self
::convertDoubleWidth( $string );
2994 * convert double-width roman characters to single-width.
2995 * range: ff00-ff5f ~= 0020-007f
2997 * @param string $string
3001 protected static function convertDoubleWidth( $string ) {
3002 static $full = null;
3003 static $half = null;
3005 if ( $full === null ) {
3006 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
3007 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
3008 $full = str_split( $fullWidth, 3 );
3009 $half = str_split( $halfWidth );
3012 $string = str_replace( $full, $half, $string );
3017 * @param string $string
3018 * @param string $pattern
3021 protected static function insertSpace( $string, $pattern ) {
3022 $string = preg_replace( $pattern, " $1 ", $string );
3023 $string = preg_replace( '/ +/', ' ', $string );
3028 * @param array $termsArray
3031 function convertForSearchResult( $termsArray ) {
3032 # some languages, e.g. Chinese, need to do a conversion
3033 # in order for search results to be displayed correctly
3038 * Get the first character of a string.
3043 function firstChar( $s ) {
3046 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
3047 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
3052 if ( isset( $matches[1] ) ) {
3053 if ( strlen( $matches[1] ) != 3 ) {
3057 // Break down Hangul syllables to grab the first jamo
3058 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3059 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3061 } elseif ( $code < 0xb098 ) {
3062 return "\xe3\x84\xb1";
3063 } elseif ( $code < 0xb2e4 ) {
3064 return "\xe3\x84\xb4";
3065 } elseif ( $code < 0xb77c ) {
3066 return "\xe3\x84\xb7";
3067 } elseif ( $code < 0xb9c8 ) {
3068 return "\xe3\x84\xb9";
3069 } elseif ( $code < 0xbc14 ) {
3070 return "\xe3\x85\x81";
3071 } elseif ( $code < 0xc0ac ) {
3072 return "\xe3\x85\x82";
3073 } elseif ( $code < 0xc544 ) {
3074 return "\xe3\x85\x85";
3075 } elseif ( $code < 0xc790 ) {
3076 return "\xe3\x85\x87";
3077 } elseif ( $code < 0xcc28 ) {
3078 return "\xe3\x85\x88";
3079 } elseif ( $code < 0xce74 ) {
3080 return "\xe3\x85\x8a";
3081 } elseif ( $code < 0xd0c0 ) {
3082 return "\xe3\x85\x8b";
3083 } elseif ( $code < 0xd30c ) {
3084 return "\xe3\x85\x8c";
3085 } elseif ( $code < 0xd558 ) {
3086 return "\xe3\x85\x8d";
3088 return "\xe3\x85\x8e";
3095 function initEncoding() {
3096 # Some languages may have an alternate char encoding option
3097 # (Esperanto X-coding, Japanese furigana conversion, etc)
3098 # If this language is used as the primary content language,
3099 # an override to the defaults can be set here on startup.
3106 function recodeForEdit( $s ) {
3107 # For some languages we'll want to explicitly specify
3108 # which characters make it into the edit box raw
3109 # or are converted in some way or another.
3110 global $wgEditEncoding;
3111 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
3114 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3122 function recodeInput( $s ) {
3123 # Take the previous into account.
3124 global $wgEditEncoding;
3125 if ( $wgEditEncoding != '' ) {
3126 $enc = $wgEditEncoding;
3130 if ( $enc == 'UTF-8' ) {
3133 return $this->iconv( $enc, 'UTF-8', $s );
3138 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3139 * also cleans up certain backwards-compatible sequences, converting them
3140 * to the modern Unicode equivalent.
3142 * This is language-specific for performance reasons only.
3148 function normalize( $s ) {
3149 global $wgAllUnicodeFixes;
3150 $s = UtfNormal\Validator
::cleanUp( $s );
3151 if ( $wgAllUnicodeFixes ) {
3152 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3153 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3160 * Transform a string using serialized data stored in the given file (which
3161 * must be in the serialized subdirectory of $IP). The file contains pairs
3162 * mapping source characters to destination characters.
3164 * The data is cached in process memory. This will go faster if you have the
3165 * FastStringSearch extension.
3167 * @param string $file
3168 * @param string $string
3170 * @throws MWException
3173 function transformUsingPairFile( $file, $string ) {
3174 if ( !isset( $this->transformData
[$file] ) ) {
3175 $data = wfGetPrecompiledData( $file );
3176 if ( $data === false ) {
3177 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3179 $this->transformData
[$file] = new ReplacementArray( $data );
3181 return $this->transformData
[$file]->replace( $string );
3185 * For right-to-left language support
3190 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3194 * Return the correct HTML 'dir' attribute value for this language.
3198 return $this->isRTL() ?
'rtl' : 'ltr';
3202 * Return 'left' or 'right' as appropriate alignment for line-start
3203 * for this language's text direction.
3205 * Should be equivalent to CSS3 'start' text-align value....
3209 function alignStart() {
3210 return $this->isRTL() ?
'right' : 'left';
3214 * Return 'right' or 'left' as appropriate alignment for line-end
3215 * for this language's text direction.
3217 * Should be equivalent to CSS3 'end' text-align value....
3221 function alignEnd() {
3222 return $this->isRTL() ?
'left' : 'right';
3226 * A hidden direction mark (LRM or RLM), depending on the language direction.
3227 * Unlike getDirMark(), this function returns the character as an HTML entity.
3228 * This function should be used when the output is guaranteed to be HTML,
3229 * because it makes the output HTML source code more readable. When
3230 * the output is plain text or can be escaped, getDirMark() should be used.
3232 * @param bool $opposite Get the direction mark opposite to your language
3236 function getDirMarkEntity( $opposite = false ) {
3238 return $this->isRTL() ?
'‎' : '‏';
3240 return $this->isRTL() ?
'‏' : '‎';
3244 * A hidden direction mark (LRM or RLM), depending on the language direction.
3245 * This function produces them as invisible Unicode characters and
3246 * the output may be hard to read and debug, so it should only be used
3247 * when the output is plain text or can be escaped. When the output is
3248 * HTML, use getDirMarkEntity() instead.
3250 * @param bool $opposite Get the direction mark opposite to your language
3253 function getDirMark( $opposite = false ) {
3254 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3255 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3257 return $this->isRTL() ?
$lrm : $rlm;
3259 return $this->isRTL() ?
$rlm : $lrm;
3265 function capitalizeAllNouns() {
3266 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3270 * An arrow, depending on the language direction.
3272 * @param string $direction The direction of the arrow: forwards (default),
3273 * backwards, left, right, up, down.
3276 function getArrow( $direction = 'forwards' ) {
3277 switch ( $direction ) {
3279 return $this->isRTL() ?
'←' : '→';
3281 return $this->isRTL() ?
'→' : '←';
3294 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3298 function linkPrefixExtension() {
3299 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3303 * Get all magic words from cache.
3306 function getMagicWords() {
3307 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3311 * Run the LanguageGetMagic hook once.
3313 protected function doMagicHook() {
3314 if ( $this->mMagicHookDone
) {
3317 $this->mMagicHookDone
= true;
3318 Hooks
::run( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3322 * Fill a MagicWord object with data from here
3324 * @param MagicWord $mw
3326 function getMagic( $mw ) {
3327 // Saves a function call
3328 if ( !$this->mMagicHookDone
) {
3329 $this->doMagicHook();
3332 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3333 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3335 $rawEntry = self
::$dataCache->getSubitem(
3336 $this->mCode
, 'magicWords', $mw->mId
);
3339 if ( !is_array( $rawEntry ) ) {
3340 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3342 $mw->mCaseSensitive
= $rawEntry[0];
3343 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3348 * Add magic words to the extension array
3350 * @param array $newWords
3352 function addMagicWordsByLang( $newWords ) {
3353 $fallbackChain = $this->getFallbackLanguages();
3354 $fallbackChain = array_reverse( $fallbackChain );
3355 foreach ( $fallbackChain as $code ) {
3356 if ( isset( $newWords[$code] ) ) {
3357 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3363 * Get special page names, as an associative array
3364 * canonical name => array of valid names, including aliases
3367 function getSpecialPageAliases() {
3368 // Cache aliases because it may be slow to load them
3369 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3371 $this->mExtendedSpecialPageAliases
=
3372 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3373 Hooks
::run( 'LanguageGetSpecialPageAliases',
3374 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3377 return $this->mExtendedSpecialPageAliases
;
3381 * Italic is unsuitable for some languages
3383 * @param string $text The text to be emphasized.
3386 function emphasize( $text ) {
3387 return "<em>$text</em>";
3391 * Normally we output all numbers in plain en_US style, that is
3392 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3393 * point twohundredthirtyfive. However this is not suitable for all
3394 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3395 * Icelandic just want to use commas instead of dots, and dots instead
3396 * of commas like "293.291,235".
3398 * An example of this function being called:
3400 * wfMessage( 'message' )->numParams( $num )->text()
3403 * See $separatorTransformTable on MessageIs.php for
3404 * the , => . and . => , implementation.
3406 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3407 * @param int|float $number The string to be formatted, should be an integer
3408 * or a floating point number.
3409 * @param bool $nocommafy Set to true for special numbers like dates
3412 public function formatNum( $number, $nocommafy = false ) {
3413 global $wgTranslateNumerals;
3414 if ( !$nocommafy ) {
3415 $number = $this->commafy( $number );
3416 $s = $this->separatorTransformTable();
3418 $number = strtr( $number, $s );
3422 if ( $wgTranslateNumerals ) {
3423 $s = $this->digitTransformTable();
3425 $number = strtr( $number, $s );
3433 * Front-end for non-commafied formatNum
3435 * @param int|float $number The string to be formatted, should be an integer
3436 * or a floating point number.
3440 public function formatNumNoSeparators( $number ) {
3441 return $this->formatNum( $number, true );
3445 * @param string $number
3448 public function parseFormattedNumber( $number ) {
3449 $s = $this->digitTransformTable();
3451 // eliminate empty array values such as ''. (bug 64347)
3452 $s = array_filter( $s );
3453 $number = strtr( $number, array_flip( $s ) );
3456 $s = $this->separatorTransformTable();
3458 // eliminate empty array values such as ''. (bug 64347)
3459 $s = array_filter( $s );
3460 $number = strtr( $number, array_flip( $s ) );
3463 $number = strtr( $number, array( ',' => '' ) );
3468 * Adds commas to a given number
3470 * @param mixed $number
3473 function commafy( $number ) {
3474 $digitGroupingPattern = $this->digitGroupingPattern();
3475 if ( $number === null ) {
3479 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3480 // default grouping is at thousands, use the same for ###,###,### pattern too.
3481 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3483 // Ref: http://cldr.unicode.org/translation/number-patterns
3485 if ( intval( $number ) < 0 ) {
3486 // For negative numbers apply the algorithm like positive number and add sign.
3488 $number = substr( $number, 1 );
3490 $integerPart = array();
3491 $decimalPart = array();
3492 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3493 preg_match( "/\d+/", $number, $integerPart );
3494 preg_match( "/\.\d*/", $number, $decimalPart );
3495 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3496 if ( $groupedNumber === $number ) {
3497 // the string does not have any number part. Eg: .12345
3498 return $sign . $groupedNumber;
3500 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3501 while ( $start > 0 ) {
3502 $match = $matches[0][$numMatches - 1];
3503 $matchLen = strlen( $match );
3504 $start = $end - $matchLen;
3508 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3510 if ( $numMatches > 1 ) {
3511 // use the last pattern for the rest of the number
3515 $groupedNumber = "," . $groupedNumber;
3518 return $sign . $groupedNumber;
3525 function digitGroupingPattern() {
3526 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3532 function digitTransformTable() {
3533 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3539 function separatorTransformTable() {
3540 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3544 * Take a list of strings and build a locale-friendly comma-separated
3545 * list, using the local comma-separator message.
3546 * The last two strings are chained with an "and".
3547 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3549 * @param string[] $l
3552 function listToText( array $l ) {
3553 $m = count( $l ) - 1;
3558 $and = $this->msg( 'and' )->escaped();
3559 $space = $this->msg( 'word-separator' )->escaped();
3561 $comma = $this->msg( 'comma-separator' )->escaped();
3565 for ( $i = $m - 1; $i >= 0; $i-- ) {
3566 if ( $i == $m - 1 ) {
3567 $s = $l[$i] . $and . $space . $s;
3569 $s = $l[$i] . $comma . $s;
3576 * Take a list of strings and build a locale-friendly comma-separated
3577 * list, using the local comma-separator message.
3578 * @param string[] $list Array of strings to put in a comma list
3581 function commaList( array $list ) {
3583 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3589 * Take a list of strings and build a locale-friendly semicolon-separated
3590 * list, using the local semicolon-separator message.
3591 * @param string[] $list Array of strings to put in a semicolon list
3594 function semicolonList( array $list ) {
3596 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3602 * Same as commaList, but separate it with the pipe instead.
3603 * @param string[] $list Array of strings to put in a pipe list
3606 function pipeList( array $list ) {
3608 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3614 * Truncate a string to a specified length in bytes, appending an optional
3615 * string (e.g. for ellipses)
3617 * The database offers limited byte lengths for some columns in the database;
3618 * multi-byte character sets mean we need to ensure that only whole characters
3619 * are included, otherwise broken characters can be passed to the user
3621 * If $length is negative, the string will be truncated from the beginning
3623 * @param string $string String to truncate
3624 * @param int $length Maximum length (including ellipses)
3625 * @param string $ellipsis String to append to the truncated text
3626 * @param bool $adjustLength Subtract length of ellipsis from $length.
3627 * $adjustLength was introduced in 1.18, before that behaved as if false.
3630 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3631 # Use the localized ellipsis character
3632 if ( $ellipsis == '...' ) {
3633 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3635 # Check if there is no need to truncate
3636 if ( $length == 0 ) {
3637 return $ellipsis; // convention
3638 } elseif ( strlen( $string ) <= abs( $length ) ) {
3639 return $string; // no need to truncate
3641 $stringOriginal = $string;
3642 # If ellipsis length is >= $length then we can't apply $adjustLength
3643 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3644 $string = $ellipsis; // this can be slightly unexpected
3645 # Otherwise, truncate and add ellipsis...
3647 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3648 if ( $length > 0 ) {
3649 $length -= $eLength;
3650 $string = substr( $string, 0, $length ); // xyz...
3651 $string = $this->removeBadCharLast( $string );
3652 $string = rtrim( $string );
3653 $string = $string . $ellipsis;
3655 $length +
= $eLength;
3656 $string = substr( $string, $length ); // ...xyz
3657 $string = $this->removeBadCharFirst( $string );
3658 $string = ltrim( $string );
3659 $string = $ellipsis . $string;
3662 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3663 # This check is *not* redundant if $adjustLength, due to the single case where
3664 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3665 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3668 return $stringOriginal;
3673 * Remove bytes that represent an incomplete Unicode character
3674 * at the end of string (e.g. bytes of the char are missing)
3676 * @param string $string
3679 protected function removeBadCharLast( $string ) {
3680 if ( $string != '' ) {
3681 $char = ord( $string[strlen( $string ) - 1] );
3683 if ( $char >= 0xc0 ) {
3684 # We got the first byte only of a multibyte char; remove it.
3685 $string = substr( $string, 0, -1 );
3686 } elseif ( $char >= 0x80 &&
3687 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3688 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3690 # We chopped in the middle of a character; remove it
3698 * Remove bytes that represent an incomplete Unicode character
3699 * at the start of string (e.g. bytes of the char are missing)
3701 * @param string $string
3704 protected function removeBadCharFirst( $string ) {
3705 if ( $string != '' ) {
3706 $char = ord( $string[0] );
3707 if ( $char >= 0x80 && $char < 0xc0 ) {
3708 # We chopped in the middle of a character; remove the whole thing
3709 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3716 * Truncate a string of valid HTML to a specified length in bytes,
3717 * appending an optional string (e.g. for ellipses), and return valid HTML
3719 * This is only intended for styled/linked text, such as HTML with
3720 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3721 * Also, this will not detect things like "display:none" CSS.
3723 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3725 * @param string $text HTML string to truncate
3726 * @param int $length (zero/positive) Maximum length (including ellipses)
3727 * @param string $ellipsis String to append to the truncated text
3730 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3731 # Use the localized ellipsis character
3732 if ( $ellipsis == '...' ) {
3733 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3735 # Check if there is clearly no need to truncate
3736 if ( $length <= 0 ) {
3737 return $ellipsis; // no text shown, nothing to format (convention)
3738 } elseif ( strlen( $text ) <= $length ) {
3739 return $text; // string short enough even *with* HTML (short-circuit)
3742 $dispLen = 0; // innerHTML legth so far
3743 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3744 $tagType = 0; // 0-open, 1-close
3745 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3746 $entityState = 0; // 0-not entity, 1-entity
3747 $tag = $ret = ''; // accumulated tag name, accumulated result string
3748 $openTags = array(); // open tag stack
3749 $maybeState = null; // possible truncation state
3751 $textLen = strlen( $text );
3752 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3753 for ( $pos = 0; true; ++
$pos ) {
3754 # Consider truncation once the display length has reached the maximim.
3755 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3756 # Check that we're not in the middle of a bracket/entity...
3757 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3758 if ( !$testingEllipsis ) {
3759 $testingEllipsis = true;
3760 # Save where we are; we will truncate here unless there turn out to
3761 # be so few remaining characters that truncation is not necessary.
3762 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3763 $maybeState = array( $ret, $openTags ); // save state
3765 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3766 # String in fact does need truncation, the truncation point was OK.
3767 list( $ret, $openTags ) = $maybeState; // reload state
3768 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3769 $ret .= $ellipsis; // add ellipsis
3773 if ( $pos >= $textLen ) {
3774 break; // extra iteration just for above checks
3777 # Read the next char...
3779 $lastCh = $pos ?
$text[$pos - 1] : '';
3780 $ret .= $ch; // add to result string
3782 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3783 $entityState = 0; // for bad HTML
3784 $bracketState = 1; // tag started (checking for backslash)
3785 } elseif ( $ch == '>' ) {
3786 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3787 $entityState = 0; // for bad HTML
3788 $bracketState = 0; // out of brackets
3789 } elseif ( $bracketState == 1 ) {
3791 $tagType = 1; // close tag (e.g. "</span>")
3793 $tagType = 0; // open tag (e.g. "<span>")
3796 $bracketState = 2; // building tag name
3797 } elseif ( $bracketState == 2 ) {
3801 // Name found (e.g. "<a href=..."), add on tag attributes...
3802 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3804 } elseif ( $bracketState == 0 ) {
3805 if ( $entityState ) {
3808 $dispLen++
; // entity is one displayed char
3811 if ( $neLength == 0 && !$maybeState ) {
3812 // Save state without $ch. We want to *hit* the first
3813 // display char (to get tags) but not *use* it if truncating.
3814 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3817 $entityState = 1; // entity found, (e.g. " ")
3819 $dispLen++
; // this char is displayed
3820 // Add the next $max display text chars after this in one swoop...
3821 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3822 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3823 $dispLen +
= $skipped;
3829 // Close the last tag if left unclosed by bad HTML
3830 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3831 while ( count( $openTags ) > 0 ) {
3832 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3838 * truncateHtml() helper function
3839 * like strcspn() but adds the skipped chars to $ret
3841 * @param string $ret
3842 * @param string $text
3843 * @param string $search
3845 * @param null|int $len
3848 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3849 if ( $len === null ) {
3850 $len = -1; // -1 means "no limit" for strcspn
3851 } elseif ( $len < 0 ) {
3855 if ( $start < strlen( $text ) ) {
3856 $skipCount = strcspn( $text, $search, $start, $len );
3857 $ret .= substr( $text, $start, $skipCount );
3863 * truncateHtml() helper function
3864 * (a) push or pop $tag from $openTags as needed
3865 * (b) clear $tag value
3866 * @param string &$tag Current HTML tag name we are looking at
3867 * @param int $tagType (0-open tag, 1-close tag)
3868 * @param string $lastCh Character before the '>' that ended this tag
3869 * @param array &$openTags Open tag stack (not accounting for $tag)
3871 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3872 $tag = ltrim( $tag );
3874 if ( $tagType == 0 && $lastCh != '/' ) {
3875 $openTags[] = $tag; // tag opened (didn't close itself)
3876 } elseif ( $tagType == 1 ) {
3877 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3878 array_pop( $openTags ); // tag closed
3886 * Grammatical transformations, needed for inflected languages
3887 * Invoked by putting {{grammar:case|word}} in a message
3889 * @param string $word
3890 * @param string $case
3893 function convertGrammar( $word, $case ) {
3894 global $wgGrammarForms;
3895 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3896 return $wgGrammarForms[$this->getCode()][$case][$word];
3902 * Get the grammar forms for the content language
3903 * @return array Array of grammar forms
3906 function getGrammarForms() {
3907 global $wgGrammarForms;
3908 if ( isset( $wgGrammarForms[$this->getCode()] )
3909 && is_array( $wgGrammarForms[$this->getCode()] )
3911 return $wgGrammarForms[$this->getCode()];
3917 * Provides an alternative text depending on specified gender.
3918 * Usage {{gender:username|masculine|feminine|unknown}}.
3919 * username is optional, in which case the gender of current user is used,
3920 * but only in (some) interface messages; otherwise default gender is used.
3922 * If no forms are given, an empty string is returned. If only one form is
3923 * given, it will be returned unconditionally. These details are implied by
3924 * the caller and cannot be overridden in subclasses.
3926 * If three forms are given, the default is to use the third (unknown) form.
3927 * If fewer than three forms are given, the default is to use the first (masculine) form.
3928 * These details can be overridden in subclasses.
3930 * @param string $gender
3931 * @param array $forms
3935 function gender( $gender, $forms ) {
3936 if ( !count( $forms ) ) {
3939 $forms = $this->preConvertPlural( $forms, 2 );
3940 if ( $gender === 'male' ) {
3943 if ( $gender === 'female' ) {
3946 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3950 * Plural form transformations, needed for some languages.
3951 * For example, there are 3 form of plural in Russian and Polish,
3952 * depending on "count mod 10". See [[w:Plural]]
3953 * For English it is pretty simple.
3955 * Invoked by putting {{plural:count|wordform1|wordform2}}
3956 * or {{plural:count|wordform1|wordform2|wordform3}}
3958 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3960 * @param int $count Non-localized number
3961 * @param array $forms Different plural forms
3962 * @return string Correct form of plural for $count in this language
3964 function convertPlural( $count, $forms ) {
3965 // Handle explicit n=pluralform cases
3966 $forms = $this->handleExplicitPluralForms( $count, $forms );
3967 if ( is_string( $forms ) ) {
3970 if ( !count( $forms ) ) {
3974 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3975 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3976 return $forms[$pluralForm];
3980 * Handles explicit plural forms for Language::convertPlural()
3982 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3983 * If an explicitly defined plural form matches the $count, then
3984 * string value returned, otherwise array returned for further consideration
3985 * by CLDR rules or overridden convertPlural().
3989 * @param int $count Non-localized number
3990 * @param array $forms Different plural forms
3992 * @return array|string
3994 protected function handleExplicitPluralForms( $count, array $forms ) {
3995 foreach ( $forms as $index => $form ) {
3996 if ( preg_match( '/\d+=/i', $form ) ) {
3997 $pos = strpos( $form, '=' );
3998 if ( substr( $form, 0, $pos ) === (string)$count ) {
3999 return substr( $form, $pos +
1 );
4001 unset( $forms[$index] );
4004 return array_values( $forms );
4008 * Checks that convertPlural was given an array and pads it to requested
4009 * amount of forms by copying the last one.
4011 * @param array $forms Array of forms given to convertPlural
4012 * @param int $count How many forms should there be at least
4013 * @return array Padded array of forms or an exception if not an array
4015 protected function preConvertPlural( /* Array */ $forms, $count ) {
4016 while ( count( $forms ) < $count ) {
4017 $forms[] = $forms[count( $forms ) - 1];
4023 * Wraps argument with unicode control characters for directionality safety
4025 * This solves the problem where directionality-neutral characters at the edge of
4026 * the argument string get interpreted with the wrong directionality from the
4027 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4029 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4030 * directionality of the argument string, using the BIDI algorithm's own "First
4031 * strong directional codepoint" rule. Essentially, this works round the fact that
4032 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4033 * direction inference). The latter is cleaner but still not widely supported.
4035 * @param string $text Text to wrap
4036 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4038 public function embedBidi( $text = '' ) {
4039 $dir = Language
::strongDirFromContent( $text );
4040 if ( $dir === 'ltr' ) {
4041 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4042 return self
::$lre . $text . self
::$pdf;
4044 if ( $dir === 'rtl' ) {
4045 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4046 return self
::$rle . $text . self
::$pdf;
4048 // No strong directionality: do not wrap
4053 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4054 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4055 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4056 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4059 * @param string $str The validated block duration in English
4060 * @return string Somehow translated block duration
4061 * @see LanguageFi.php for example implementation
4063 function translateBlockExpiry( $str ) {
4064 $duration = SpecialBlock
::getSuggestedDurations( $this );
4065 foreach ( $duration as $show => $value ) {
4066 if ( strcmp( $str, $value ) == 0 ) {
4067 return htmlspecialchars( trim( $show ) );
4071 if ( wfIsInfinity( $str ) ) {
4072 foreach ( $duration as $show => $value ) {
4073 if ( wfIsInfinity( $value ) ) {
4074 return htmlspecialchars( trim( $show ) );
4079 // If all else fails, return a standard duration or timestamp description.
4080 $time = strtotime( $str, 0 );
4081 if ( $time === false ) { // Unknown format. Return it as-is in case.
4083 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4084 // $time is relative to 0 so it's a duration length.
4085 return $this->formatDuration( $time );
4086 } else { // It's an absolute timestamp.
4087 if ( $time === 0 ) {
4088 // wfTimestamp() handles 0 as current time instead of epoch.
4089 return $this->timeanddate( '19700101000000' );
4091 return $this->timeanddate( $time );
4097 * languages like Chinese need to be segmented in order for the diff
4100 * @param string $text
4103 public function segmentForDiff( $text ) {
4108 * and unsegment to show the result
4110 * @param string $text
4113 public function unsegmentForDiff( $text ) {
4118 * Return the LanguageConverter used in the Language
4121 * @return LanguageConverter
4123 public function getConverter() {
4124 return $this->mConverter
;
4128 * convert text to all supported variants
4130 * @param string $text
4133 public function autoConvertToAllVariants( $text ) {
4134 return $this->mConverter
->autoConvertToAllVariants( $text );
4138 * convert text to different variants of a language.
4140 * @param string $text
4143 public function convert( $text ) {
4144 return $this->mConverter
->convert( $text );
4148 * Convert a Title object to a string in the preferred variant
4150 * @param Title $title
4153 public function convertTitle( $title ) {
4154 return $this->mConverter
->convertTitle( $title );
4158 * Convert a namespace index to a string in the preferred variant
4163 public function convertNamespace( $ns ) {
4164 return $this->mConverter
->convertNamespace( $ns );
4168 * Check if this is a language with variants
4172 public function hasVariants() {
4173 return count( $this->getVariants() ) > 1;
4177 * Check if the language has the specific variant
4180 * @param string $variant
4183 public function hasVariant( $variant ) {
4184 return (bool)$this->mConverter
->validateVariant( $variant );
4188 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4190 * @param string $text
4192 * @deprecated since 1.22 is no longer used
4194 public function armourMath( $text ) {
4195 return $this->mConverter
->armourMath( $text );
4199 * Perform output conversion on a string, and encode for safe HTML output.
4200 * @param string $text Text to be converted
4201 * @param bool $isTitle Whether this conversion is for the article title
4203 * @todo this should get integrated somewhere sane
4205 public function convertHtml( $text, $isTitle = false ) {
4206 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4210 * @param string $key
4213 public function convertCategoryKey( $key ) {
4214 return $this->mConverter
->convertCategoryKey( $key );
4218 * Get the list of variants supported by this language
4219 * see sample implementation in LanguageZh.php
4221 * @return array An array of language codes
4223 public function getVariants() {
4224 return $this->mConverter
->getVariants();
4230 public function getPreferredVariant() {
4231 return $this->mConverter
->getPreferredVariant();
4237 public function getDefaultVariant() {
4238 return $this->mConverter
->getDefaultVariant();
4244 public function getURLVariant() {
4245 return $this->mConverter
->getURLVariant();
4249 * If a language supports multiple variants, it is
4250 * possible that non-existing link in one variant
4251 * actually exists in another variant. this function
4252 * tries to find it. See e.g. LanguageZh.php
4253 * The input parameters may be modified upon return
4255 * @param string &$link The name of the link
4256 * @param Title &$nt The title object of the link
4257 * @param bool $ignoreOtherCond To disable other conditions when
4258 * we need to transclude a template or update a category's link
4260 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4261 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4265 * returns language specific options used by User::getPageRenderHash()
4266 * for example, the preferred language variant
4270 function getExtraHashOptions() {
4271 return $this->mConverter
->getExtraHashOptions();
4275 * For languages that support multiple variants, the title of an
4276 * article may be displayed differently in different variants. this
4277 * function returns the apporiate title defined in the body of the article.
4281 public function getParsedTitle() {
4282 return $this->mConverter
->getParsedTitle();
4286 * Prepare external link text for conversion. When the text is
4287 * a URL, it shouldn't be converted, and it'll be wrapped in
4288 * the "raw" tag (-{R| }-) to prevent conversion.
4290 * This function is called "markNoConversion" for historical
4293 * @param string $text Text to be used for external link
4294 * @param bool $noParse Wrap it without confirming it's a real URL first
4295 * @return string The tagged text
4297 public function markNoConversion( $text, $noParse = false ) {
4298 // Excluding protocal-relative URLs may avoid many false positives.
4299 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4300 return $this->mConverter
->markNoConversion( $text );
4307 * A regular expression to match legal word-trailing characters
4308 * which should be merged onto a link of the form [[foo]]bar.
4312 public function linkTrail() {
4313 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4317 * A regular expression character set to match legal word-prefixing
4318 * characters which should be merged onto a link of the form foo[[bar]].
4322 public function linkPrefixCharset() {
4323 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4327 * @deprecated since 1.24, will be removed in 1.25
4330 function getLangObj() {
4331 wfDeprecated( __METHOD__
, '1.24' );
4336 * Get the "parent" language which has a converter to convert a "compatible" language
4337 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4339 * @return Language|null
4342 public function getParentLanguage() {
4343 if ( $this->mParentLanguage
!== false ) {
4344 return $this->mParentLanguage
;
4347 $pieces = explode( '-', $this->getCode() );
4349 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4350 $this->mParentLanguage
= null;
4353 $lang = Language
::factory( $code );
4354 if ( !$lang->hasVariant( $this->getCode() ) ) {
4355 $this->mParentLanguage
= null;
4359 $this->mParentLanguage
= $lang;
4364 * Get the RFC 3066 code for this language object
4366 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4367 * htmlspecialchars() or similar
4371 public function getCode() {
4372 return $this->mCode
;
4376 * Get the code in Bcp47 format which we can use
4377 * inside of html lang="" tags.
4379 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4380 * htmlspecialchars() or similar.
4385 public function getHtmlCode() {
4386 if ( is_null( $this->mHtmlCode
) ) {
4387 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4389 return $this->mHtmlCode
;
4393 * @param string $code
4395 public function setCode( $code ) {
4396 $this->mCode
= $code;
4397 // Ensure we don't leave incorrect cached data lying around
4398 $this->mHtmlCode
= null;
4399 $this->mParentLanguage
= false;
4403 * Get the name of a file for a certain language code
4404 * @param string $prefix Prepend this to the filename
4405 * @param string $code Language code
4406 * @param string $suffix Append this to the filename
4407 * @throws MWException
4408 * @return string $prefix . $mangledCode . $suffix
4410 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4411 if ( !self
::isValidBuiltInCode( $code ) ) {
4412 throw new MWException( "Invalid language code \"$code\"" );
4415 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4419 * Get the language code from a file name. Inverse of getFileName()
4420 * @param string $filename $prefix . $languageCode . $suffix
4421 * @param string $prefix Prefix before the language code
4422 * @param string $suffix Suffix after the language code
4423 * @return string Language code, or false if $prefix or $suffix isn't found
4425 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4427 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4428 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4429 if ( !count( $m ) ) {
4432 return str_replace( '_', '-', strtolower( $m[1] ) );
4436 * @param string $code
4439 public static function getMessagesFileName( $code ) {
4441 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4442 Hooks
::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4447 * @param string $code
4451 public static function getJsonMessagesFileName( $code ) {
4454 if ( !self
::isValidBuiltInCode( $code ) ) {
4455 throw new MWException( "Invalid language code \"$code\"" );
4458 return "$IP/languages/i18n/$code.json";
4462 * @param string $code
4465 public static function getClassFileName( $code ) {
4467 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4471 * Get the first fallback for a given language.
4473 * @param string $code
4475 * @return bool|string
4477 public static function getFallbackFor( $code ) {
4478 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4481 $fallbacks = self
::getFallbacksFor( $code );
4482 return $fallbacks[0];
4487 * Get the ordered list of fallback languages.
4490 * @param string $code Language code
4491 * @return array Non-empty array, ending in "en"
4493 public static function getFallbacksFor( $code ) {
4494 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4497 // For unknown languages, fallbackSequence returns an empty array,
4498 // hardcode fallback to 'en' in that case.
4499 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: array( 'en' );
4503 * Get the ordered list of fallback languages, ending with the fallback
4504 * language chain for the site language.
4507 * @param string $code Language code
4508 * @return array Array( fallbacks, site fallbacks )
4510 public static function getFallbacksIncludingSiteLanguage( $code ) {
4511 global $wgLanguageCode;
4513 // Usually, we will only store a tiny number of fallback chains, so we
4514 // keep them in static memory.
4515 $cacheKey = "{$code}-{$wgLanguageCode}";
4517 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4518 $fallbacks = self
::getFallbacksFor( $code );
4520 // Append the site's fallback chain, including the site language itself
4521 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4522 array_unshift( $siteFallbacks, $wgLanguageCode );
4524 // Eliminate any languages already included in the chain
4525 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4527 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4529 return self
::$fallbackLanguageCache[$cacheKey];
4533 * Get all messages for a given language
4534 * WARNING: this may take a long time. If you just need all message *keys*
4535 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4537 * @param string $code
4541 public static function getMessagesFor( $code ) {
4542 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4546 * Get a message for a given language
4548 * @param string $key
4549 * @param string $code
4553 public static function getMessageFor( $key, $code ) {
4554 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4558 * Get all message keys for a given language. This is a faster alternative to
4559 * array_keys( Language::getMessagesFor( $code ) )
4562 * @param string $code Language code
4563 * @return array Array of message keys (strings)
4565 public static function getMessageKeysFor( $code ) {
4566 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4570 * @param string $talk
4573 function fixVariableInNamespace( $talk ) {
4574 if ( strpos( $talk, '$1' ) === false ) {
4578 global $wgMetaNamespace;
4579 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4581 # Allow grammar transformations
4582 # Allowing full message-style parsing would make simple requests
4583 # such as action=raw much more expensive than they need to be.
4584 # This will hopefully cover most cases.
4585 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4586 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4587 return str_replace( ' ', '_', $talk );
4594 function replaceGrammarInNamespace( $m ) {
4595 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4599 * @throws MWException
4602 static function getCaseMaps() {
4603 static $wikiUpperChars, $wikiLowerChars;
4604 if ( isset( $wikiUpperChars ) ) {
4605 return array( $wikiUpperChars, $wikiLowerChars );
4608 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4609 if ( $arr === false ) {
4610 throw new MWException(
4611 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4613 $wikiUpperChars = $arr['wikiUpperChars'];
4614 $wikiLowerChars = $arr['wikiLowerChars'];
4615 return array( $wikiUpperChars, $wikiLowerChars );
4619 * Decode an expiry (block, protection, etc) which has come from the DB
4621 * @param string $expiry Database expiry String
4622 * @param bool|int $format True to process using language functions, or TS_ constant
4623 * to return the expiry in a given timestamp
4624 * @param string $inifinity If $format is not true, use this string for infinite expiry
4628 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4630 if ( $dbInfinity === null ) {
4631 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4634 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4635 return $format === true
4636 ?
$this->getMessageFromDB( 'infiniteblock' )
4639 return $format === true
4640 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4641 : wfTimestamp( $format, $expiry );
4647 * @param int|float $seconds
4648 * @param array $format Optional
4649 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4650 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4651 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4653 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4654 * or 'avoidminutes'.
4657 function formatTimePeriod( $seconds, $format = array() ) {
4658 if ( !is_array( $format ) ) {
4659 $format = array( 'avoid' => $format ); // For backwards compatibility
4661 if ( !isset( $format['avoid'] ) ) {
4662 $format['avoid'] = false;
4664 if ( !isset( $format['noabbrevs'] ) ) {
4665 $format['noabbrevs'] = false;
4667 $secondsMsg = wfMessage(
4668 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4669 $minutesMsg = wfMessage(
4670 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4671 $hoursMsg = wfMessage(
4672 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4673 $daysMsg = wfMessage(
4674 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4676 if ( round( $seconds * 10 ) < 100 ) {
4677 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4678 $s = $secondsMsg->params( $s )->text();
4679 } elseif ( round( $seconds ) < 60 ) {
4680 $s = $this->formatNum( round( $seconds ) );
4681 $s = $secondsMsg->params( $s )->text();
4682 } elseif ( round( $seconds ) < 3600 ) {
4683 $minutes = floor( $seconds / 60 );
4684 $secondsPart = round( fmod( $seconds, 60 ) );
4685 if ( $secondsPart == 60 ) {
4689 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4691 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4692 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4693 $hours = floor( $seconds / 3600 );
4694 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4695 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4696 if ( $secondsPart == 60 ) {
4700 if ( $minutes == 60 ) {
4704 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4706 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4707 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4708 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4711 $days = floor( $seconds / 86400 );
4712 if ( $format['avoid'] === 'avoidminutes' ) {
4713 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4714 if ( $hours == 24 ) {
4718 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4720 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4721 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4722 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4723 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4724 if ( $minutes == 60 ) {
4728 if ( $hours == 24 ) {
4732 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4734 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4736 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4738 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4740 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4747 * Format a bitrate for output, using an appropriate
4748 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4749 * the magnitude in question.
4751 * This use base 1000. For base 1024 use formatSize(), for another base
4752 * see formatComputingNumbers().
4757 function formatBitrate( $bps ) {
4758 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4762 * @param int $size Size of the unit
4763 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4764 * @param string $messageKey Message key to be uesd
4767 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4769 return str_replace( '$1', $this->formatNum( $size ),
4770 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4773 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4776 $maxIndex = count( $sizes ) - 1;
4777 while ( $size >= $boundary && $index < $maxIndex ) {
4782 // For small sizes no decimal places necessary
4785 // For MB and bigger two decimal places are smarter
4788 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4790 $size = round( $size, $round );
4791 $text = $this->getMessageFromDB( $msg );
4792 return str_replace( '$1', $this->formatNum( $size ), $text );
4796 * Format a size in bytes for output, using an appropriate
4797 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4799 * This method use base 1024. For base 1000 use formatBitrate(), for
4800 * another base see formatComputingNumbers()
4802 * @param int $size Size to format
4803 * @return string Plain text (not HTML)
4805 function formatSize( $size ) {
4806 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4810 * Make a list item, used by various special pages
4812 * @param string $page Page link
4813 * @param string $details HTML safe text between brackets
4814 * @param bool $oppositedm Add the direction mark opposite to your
4815 * language, to display text properly
4816 * @return HTML escaped string
4818 function specialList( $page, $details, $oppositedm = true ) {
4823 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4827 $this->msg( 'word-separator' )->escaped() .
4828 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4832 * Generate (prev x| next x) (20|50|100...) type links for paging
4834 * @param Title $title Title object to link
4835 * @param int $offset
4837 * @param array $query Optional URL query parameter string
4838 * @param bool $atend Optional param for specified if this is the last page
4841 public function viewPrevNext( Title
$title, $offset, $limit,
4842 array $query = array(), $atend = false
4844 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4846 # Make 'previous' link
4847 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4848 if ( $offset > 0 ) {
4849 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4850 $query, $prev, 'prevn-title', 'mw-prevlink' );
4852 $plink = htmlspecialchars( $prev );
4856 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4858 $nlink = htmlspecialchars( $next );
4860 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4861 $query, $next, 'nextn-title', 'mw-nextlink' );
4864 # Make links to set number of items per page
4865 $numLinks = array();
4866 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4867 $numLinks[] = $this->numLink( $title, $offset, $num,
4868 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4871 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4872 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4876 * Helper function for viewPrevNext() that generates links
4878 * @param Title $title Title object to link
4879 * @param int $offset
4881 * @param array $query Extra query parameters
4882 * @param string $link Text to use for the link; will be escaped
4883 * @param string $tooltipMsg Name of the message to use as tooltip
4884 * @param string $class Value of the "class" attribute of the link
4885 * @return string HTML fragment
4887 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4890 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4891 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4892 ->numParams( $limit )->text();
4894 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4895 'title' => $tooltip, 'class' => $class ), $link );
4899 * Get the conversion rule title, if any.
4903 public function getConvRuleTitle() {
4904 return $this->mConverter
->getConvRuleTitle();
4908 * Get the compiled plural rules for the language
4910 * @return array Associative array with plural form, and plural rule as key-value pairs
4912 public function getCompiledPluralRules() {
4913 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4914 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4915 if ( !$pluralRules ) {
4916 foreach ( $fallbacks as $fallbackCode ) {
4917 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4918 if ( $pluralRules ) {
4923 return $pluralRules;
4927 * Get the plural rules for the language
4929 * @return array Associative array with plural form number and plural rule as key-value pairs
4931 public function getPluralRules() {
4932 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4933 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4934 if ( !$pluralRules ) {
4935 foreach ( $fallbacks as $fallbackCode ) {
4936 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4937 if ( $pluralRules ) {
4942 return $pluralRules;
4946 * Get the plural rule types for the language
4948 * @return array Associative array with plural form number and plural rule type as key-value pairs
4950 public function getPluralRuleTypes() {
4951 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4952 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4953 if ( !$pluralRuleTypes ) {
4954 foreach ( $fallbacks as $fallbackCode ) {
4955 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4956 if ( $pluralRuleTypes ) {
4961 return $pluralRuleTypes;
4965 * Find the index number of the plural rule appropriate for the given number
4966 * @param int $number
4967 * @return int The index number of the plural rule
4969 public function getPluralRuleIndexNumber( $number ) {
4970 $pluralRules = $this->getCompiledPluralRules();
4971 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4976 * Find the plural rule type appropriate for the given number
4977 * For example, if the language is set to Arabic, getPluralType(5) should
4980 * @param int $number
4981 * @return string The name of the plural rule type, e.g. one, two, few, many
4983 public function getPluralRuleType( $number ) {
4984 $index = $this->getPluralRuleIndexNumber( $number );
4985 $pluralRuleTypes = $this->getPluralRuleTypes();
4986 if ( isset( $pluralRuleTypes[$index] ) ) {
4987 return $pluralRuleTypes[$index];