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 // @codeCoverageIgnoreStart
171 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';
172 // @codeCoverageIgnoreEnd
175 * Get a cached or new language object for a given language code
176 * @param string $code
179 static function factory( $code ) {
180 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
182 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
183 $code = $wgDummyLanguageCodes[$code];
186 // get the language object to process
187 $langObj = isset( self
::$mLangObjCache[$code] )
188 ? self
::$mLangObjCache[$code]
189 : self
::newFromCode( $code );
191 // merge the language object in to get it up front in the cache
192 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
193 // get rid of the oldest ones in case we have an overflow
194 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
200 * Create a language object for a given language code
201 * @param string $code
202 * @throws MWException
205 protected static function newFromCode( $code ) {
206 // Protect against path traversal below
207 if ( !Language
::isValidCode( $code )
208 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code )
210 throw new MWException( "Invalid language code \"$code\"" );
213 if ( !Language
::isValidBuiltInCode( $code ) ) {
214 // It's not possible to customise this code with class files, so
215 // just return a Language object. This is to support uselang= hacks.
216 $lang = new Language
;
217 $lang->setCode( $code );
221 // Check if there is a language class for the code
222 $class = self
::classFromCode( $code );
223 self
::preloadLanguageClass( $class );
224 if ( class_exists( $class ) ) {
229 // Keep trying the fallback list until we find an existing class
230 $fallbacks = Language
::getFallbacksFor( $code );
231 foreach ( $fallbacks as $fallbackCode ) {
232 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
233 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
236 $class = self
::classFromCode( $fallbackCode );
237 self
::preloadLanguageClass( $class );
238 if ( class_exists( $class ) ) {
239 $lang = Language
::newFromCode( $fallbackCode );
240 $lang->setCode( $code );
245 throw new MWException( "Invalid fallback sequence for language '$code'" );
249 * Checks whether any localisation is available for that language tag
250 * in MediaWiki (MessagesXx.php exists).
252 * @param string $code Language tag (in lower case)
253 * @return bool Whether language is supported
256 public static function isSupportedLanguage( $code ) {
257 if ( !self
::isValidBuiltInCode( $code ) ) {
261 if ( $code === 'qqq' ) {
265 return is_readable( self
::getMessagesFileName( $code ) ) ||
266 is_readable( self
::getJsonMessagesFileName( $code ) );
270 * Returns true if a language code string is a well-formed language tag
271 * according to RFC 5646.
272 * This function only checks well-formedness; it doesn't check that
273 * language, script or variant codes actually exist in the repositories.
275 * Based on regexes by Mark Davis of the Unicode Consortium:
276 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
278 * @param string $code
279 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
284 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
287 $alphanum = '[a-z0-9]';
288 $x = 'x'; # private use singleton
289 $singleton = '[a-wy-z]'; # other singleton
290 $s = $lenient ?
'[-_]' : '-';
292 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
293 $script = "$alpha{4}"; # ISO 15924
294 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
295 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
296 $extension = "$singleton(?:$s$alphanum{2,8})+";
297 $privateUse = "$x(?:$s$alphanum{1,8})+";
299 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
300 # Since these are limited, this is safe even later changes to the registry --
301 # the only oddity is that it might change the type of the tag, and thus
302 # the results from the capturing groups.
303 # http://www.iana.org/assignments/language-subtag-registry
305 $grandfathered = "en{$s}GB{$s}oed"
306 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
307 . "|no{$s}(?:bok|nyn)"
308 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
309 . "|zh{$s}min{$s}nan";
311 $variantList = "$variant(?:$s$variant)*";
312 $extensionList = "$extension(?:$s$extension)*";
314 $langtag = "(?:($language)"
317 . "(?:$s$variantList)?"
318 . "(?:$s$extensionList)?"
319 . "(?:$s$privateUse)?)";
321 # The final breakdown, with capturing groups for each of these components
322 # The variants, extensions, grandfathered, and private-use may have interior '-'
324 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
326 return (bool)preg_match( "/$root/", strtolower( $code ) );
330 * Returns true if a language code string is of a valid form, whether or
331 * not it exists. This includes codes which are used solely for
332 * customisation via the MediaWiki namespace.
334 * @param string $code
338 public static function isValidCode( $code ) {
339 static $cache = array();
340 if ( isset( $cache[$code] ) ) {
341 return $cache[$code];
343 // People think language codes are html safe, so enforce it.
344 // Ideally we should only allow a-zA-Z0-9-
345 // but, .+ and other chars are often used for {{int:}} hacks
346 // see bugs 37564, 37587, 36938
348 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
349 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
351 return $cache[$code];
355 * Returns true if a language code is of a valid form for the purposes of
356 * internal customisation of MediaWiki, via Messages*.php or *.json.
358 * @param string $code
360 * @throws MWException
364 public static function isValidBuiltInCode( $code ) {
366 if ( !is_string( $code ) ) {
367 if ( is_object( $code ) ) {
368 $addmsg = " of class " . get_class( $code );
372 $type = gettype( $code );
373 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
376 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
380 * Returns true if a language code is an IETF tag known to MediaWiki.
387 public static function isKnownLanguageTag( $tag ) {
388 static $coreLanguageNames;
390 // Quick escape for invalid input to avoid exceptions down the line
391 // when code tries to process tags which are not valid at all.
392 if ( !self
::isValidBuiltInCode( $tag ) ) {
396 if ( $coreLanguageNames === null ) {
398 include "$IP/languages/Names.php";
401 if ( isset( $coreLanguageNames[$tag] )
402 || self
::fetchLanguageName( $tag, $tag ) !== ''
411 * @param string $code
412 * @return string Name of the language class
414 public static function classFromCode( $code ) {
415 if ( $code == 'en' ) {
418 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
423 * Includes language class files
425 * @param string $class Name of the language class
427 public static function preloadLanguageClass( $class ) {
430 if ( $class === 'Language' ) {
434 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
435 include_once "$IP/languages/classes/$class.php";
440 * Get the LocalisationCache instance
442 * @return LocalisationCache
444 public static function getLocalisationCache() {
445 if ( is_null( self
::$dataCache ) ) {
446 global $wgLocalisationCacheConf;
447 $class = $wgLocalisationCacheConf['class'];
448 self
::$dataCache = new $class( $wgLocalisationCacheConf );
450 return self
::$dataCache;
453 function __construct() {
454 $this->mConverter
= new FakeConverter( $this );
455 // Set the code to the name of the descendant
456 if ( get_class( $this ) == 'Language' ) {
459 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
461 self
::getLocalisationCache();
465 * Reduce memory usage
467 function __destruct() {
468 foreach ( $this as $name => $value ) {
469 unset( $this->$name );
474 * Hook which will be called if this is the content language.
475 * Descendants can use this to register hook functions or modify globals
477 function initContLang() {
484 function getFallbackLanguages() {
485 return self
::getFallbacksFor( $this->mCode
);
489 * Exports $wgBookstoreListEn
492 function getBookstoreList() {
493 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
497 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
498 * available in localised form, it will be included in English.
502 public function getNamespaces() {
503 if ( is_null( $this->namespaceNames
) ) {
504 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
506 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
507 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
509 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
511 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
512 if ( $wgMetaNamespaceTalk ) {
513 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
515 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
516 $this->namespaceNames
[NS_PROJECT_TALK
] =
517 $this->fixVariableInNamespace( $talk );
520 # Sometimes a language will be localised but not actually exist on this wiki.
521 foreach ( $this->namespaceNames
as $key => $text ) {
522 if ( !isset( $validNamespaces[$key] ) ) {
523 unset( $this->namespaceNames
[$key] );
527 # The above mixing may leave namespaces out of canonical order.
528 # Re-order by namespace ID number...
529 ksort( $this->namespaceNames
);
531 Hooks
::run( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
534 return $this->namespaceNames
;
538 * Arbitrarily set all of the namespace names at once. Mainly used for testing
539 * @param array $namespaces Array of namespaces (id => name)
541 public function setNamespaces( array $namespaces ) {
542 $this->namespaceNames
= $namespaces;
543 $this->mNamespaceIds
= null;
547 * Resets all of the namespace caches. Mainly used for testing
549 public function resetNamespaces() {
550 $this->namespaceNames
= null;
551 $this->mNamespaceIds
= null;
552 $this->namespaceAliases
= null;
556 * A convenience function that returns getNamespaces() with spaces instead of underscores
557 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
561 function getFormattedNamespaces() {
562 $ns = $this->getNamespaces();
563 foreach ( $ns as $k => $v ) {
564 $ns[$k] = strtr( $v, '_', ' ' );
570 * Get a namespace value by key
573 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
574 * echo $mw_ns; // prints 'MediaWiki'
577 * @param int $index The array key of the namespace to return
578 * @return string|bool String if the namespace value exists, otherwise false
580 function getNsText( $index ) {
581 $ns = $this->getNamespaces();
582 return isset( $ns[$index] ) ?
$ns[$index] : false;
586 * A convenience function that returns the same thing as
587 * getNsText() except with '_' changed to ' ', useful for
591 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
592 * echo $mw_ns; // prints 'MediaWiki talk'
595 * @param int $index The array key of the namespace to return
596 * @return string Namespace name without underscores (empty string if namespace does not exist)
598 function getFormattedNsText( $index ) {
599 $ns = $this->getNsText( $index );
600 return strtr( $ns, '_', ' ' );
604 * Returns gender-dependent namespace alias if available.
605 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
606 * @param int $index Namespace index
607 * @param string $gender Gender key (male, female... )
611 function getGenderNsText( $index, $gender ) {
612 global $wgExtraGenderNamespaces;
614 $ns = $wgExtraGenderNamespaces +
615 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
617 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
621 * Whether this language uses gender-dependent namespace aliases.
622 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
626 function needsGenderDistinction() {
627 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
628 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
629 // $wgExtraGenderNamespaces overrides everything
631 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
632 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
633 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
636 // Check what is in i18n files
637 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
638 return count( $aliases ) > 0;
643 * Get a namespace key by value, case insensitive.
644 * Only matches namespace names for the current language, not the
645 * canonical ones defined in Namespace.php.
647 * @param string $text
648 * @return int|bool An integer if $text is a valid value otherwise false
650 function getLocalNsIndex( $text ) {
651 $lctext = $this->lc( $text );
652 $ids = $this->getNamespaceIds();
653 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
659 function getNamespaceAliases() {
660 if ( is_null( $this->namespaceAliases
) ) {
661 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
665 foreach ( $aliases as $name => $index ) {
666 if ( $index === NS_PROJECT_TALK
) {
667 unset( $aliases[$name] );
668 $name = $this->fixVariableInNamespace( $name );
669 $aliases[$name] = $index;
674 global $wgExtraGenderNamespaces;
675 $genders = $wgExtraGenderNamespaces +
676 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
677 foreach ( $genders as $index => $forms ) {
678 foreach ( $forms as $alias ) {
679 $aliases[$alias] = $index;
683 # Also add converted namespace names as aliases, to avoid confusion.
684 $convertedNames = array();
685 foreach ( $this->getVariants() as $variant ) {
686 if ( $variant === $this->mCode
) {
689 foreach ( $this->getNamespaces() as $ns => $_ ) {
690 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
694 $this->namespaceAliases
= $aliases +
$convertedNames;
697 return $this->namespaceAliases
;
703 function getNamespaceIds() {
704 if ( is_null( $this->mNamespaceIds
) ) {
705 global $wgNamespaceAliases;
706 # Put namespace names and aliases into a hashtable.
707 # If this is too slow, then we should arrange it so that it is done
708 # before caching. The catch is that at pre-cache time, the above
709 # class-specific fixup hasn't been done.
710 $this->mNamespaceIds
= array();
711 foreach ( $this->getNamespaces() as $index => $name ) {
712 $this->mNamespaceIds
[$this->lc( $name )] = $index;
714 foreach ( $this->getNamespaceAliases() as $name => $index ) {
715 $this->mNamespaceIds
[$this->lc( $name )] = $index;
717 if ( $wgNamespaceAliases ) {
718 foreach ( $wgNamespaceAliases as $name => $index ) {
719 $this->mNamespaceIds
[$this->lc( $name )] = $index;
723 return $this->mNamespaceIds
;
727 * Get a namespace key by value, case insensitive. Canonical namespace
728 * names override custom ones defined for the current language.
730 * @param string $text
731 * @return int|bool An integer if $text is a valid value otherwise false
733 function getNsIndex( $text ) {
734 $lctext = $this->lc( $text );
735 $ns = MWNamespace
::getCanonicalIndex( $lctext );
736 if ( $ns !== null ) {
739 $ids = $this->getNamespaceIds();
740 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
744 * short names for language variants used for language conversion links.
746 * @param string $code
747 * @param bool $usemsg Use the "variantname-xyz" message if it exists
750 function getVariantname( $code, $usemsg = true ) {
751 $msg = "variantname-$code";
752 if ( $usemsg && wfMessage( $msg )->exists() ) {
753 return $this->getMessageFromDB( $msg );
755 $name = self
::fetchLanguageName( $code );
757 return $name; # if it's defined as a language name, show that
759 # otherwise, output the language code
765 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
766 * SpecialPageFactory::getLocalNameFor instead.
767 * @param string $name
770 function specialPage( $name ) {
771 $aliases = $this->getSpecialPageAliases();
772 if ( isset( $aliases[$name][0] ) ) {
773 $name = $aliases[$name][0];
775 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
781 function getDatePreferences() {
782 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
788 function getDateFormats() {
789 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
793 * @return array|string
795 function getDefaultDateFormat() {
796 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
797 if ( $df === 'dmy or mdy' ) {
798 global $wgAmericanDates;
799 return $wgAmericanDates ?
'mdy' : 'dmy';
808 function getDatePreferenceMigrationMap() {
809 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
813 * @param string $image
816 function getImageFile( $image ) {
817 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
824 function getImageFiles() {
825 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
831 function getExtraUserToggles() {
832 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
839 function getUserToggle( $tog ) {
840 return $this->getMessageFromDB( "tog-$tog" );
844 * Get native language names, indexed by code.
845 * Only those defined in MediaWiki, no other data like CLDR.
846 * If $customisedOnly is true, only returns codes with a messages file
848 * @param bool $customisedOnly
851 * @deprecated since 1.20, use fetchLanguageNames()
853 public static function getLanguageNames( $customisedOnly = false ) {
854 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
858 * Get translated language names. This is done on best effort and
859 * by default this is exactly the same as Language::getLanguageNames.
860 * The CLDR extension provides translated names.
861 * @param string $code Language code.
862 * @return array Language code => language name
864 * @deprecated since 1.20, use fetchLanguageNames()
866 public static function getTranslatedLanguageNames( $code ) {
867 return self
::fetchLanguageNames( $code, 'all' );
871 * Get an array of language names, indexed by code.
872 * @param null|string $inLanguage Code of language in which to return the names
873 * Use null for autonyms (native names)
874 * @param string $include One of:
875 * 'all' all available languages
876 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
877 * 'mwfile' only if the language is in 'mw' *and* has a message file
878 * @return array Language code => language name
881 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
882 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
883 $cacheKey .= ":$include";
884 if ( self
::$languageNameCache === null ) {
885 self
::$languageNameCache = new MapCacheLRU( 20 );
887 if ( self
::$languageNameCache->has( $cacheKey ) ) {
888 $ret = self
::$languageNameCache->get( $cacheKey );
890 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
891 self
::$languageNameCache->set( $cacheKey, $ret );
897 * Uncached helper for fetchLanguageNames
898 * @param null|string $inLanguage Code of language in which to return the names
899 * Use null for autonyms (native names)
900 * @param string $include One of:
901 * 'all' all available languages
902 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
903 * 'mwfile' only if the language is in 'mw' *and* has a message file
904 * @return array Language code => language name
906 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
907 global $wgExtraLanguageNames;
908 static $coreLanguageNames;
910 if ( $coreLanguageNames === null ) {
912 include "$IP/languages/Names.php";
915 // If passed an invalid language code to use, fallback to en
916 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
923 # TODO: also include when $inLanguage is null, when this code is more efficient
924 Hooks
::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
927 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
928 foreach ( $mwNames as $mwCode => $mwName ) {
929 # - Prefer own MediaWiki native name when not using the hook
930 # - For other names just add if not added through the hook
931 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
932 $names[$mwCode] = $mwName;
936 if ( $include === 'all' ) {
942 $coreCodes = array_keys( $mwNames );
943 foreach ( $coreCodes as $coreCode ) {
944 $returnMw[$coreCode] = $names[$coreCode];
947 if ( $include === 'mwfile' ) {
948 $namesMwFile = array();
949 # We do this using a foreach over the codes instead of a directory
950 # loop so that messages files in extensions will work correctly.
951 foreach ( $returnMw as $code => $value ) {
952 if ( is_readable( self
::getMessagesFileName( $code ) )
953 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
955 $namesMwFile[$code] = $names[$code];
959 ksort( $namesMwFile );
964 # 'mw' option; default if it's not one of the other two options (all/mwfile)
969 * @param string $code The code of the language for which to get the name
970 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
971 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
972 * @return string Language name or empty
975 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
976 $code = strtolower( $code );
977 $array = self
::fetchLanguageNames( $inLanguage, $include );
978 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
982 * Get a message from the MediaWiki namespace.
984 * @param string $msg Message name
987 function getMessageFromDB( $msg ) {
988 return $this->msg( $msg )->text();
992 * Get message object in this language. Only for use inside this class.
994 * @param string $msg Message name
997 protected function msg( $msg ) {
998 return wfMessage( $msg )->inLanguage( $this );
1002 * Get the native language name of $code.
1003 * Only if defined in MediaWiki, no other data like CLDR.
1004 * @param string $code
1006 * @deprecated since 1.20, use fetchLanguageName()
1008 function getLanguageName( $code ) {
1009 return self
::fetchLanguageName( $code );
1013 * @param string $key
1016 function getMonthName( $key ) {
1017 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
1023 function getMonthNamesArray() {
1024 $monthNames = array( '' );
1025 for ( $i = 1; $i < 13; $i++
) {
1026 $monthNames[] = $this->getMonthName( $i );
1032 * @param string $key
1035 function getMonthNameGen( $key ) {
1036 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1040 * @param string $key
1043 function getMonthAbbreviation( $key ) {
1044 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1050 function getMonthAbbreviationsArray() {
1051 $monthNames = array( '' );
1052 for ( $i = 1; $i < 13; $i++
) {
1053 $monthNames[] = $this->getMonthAbbreviation( $i );
1059 * @param string $key
1062 function getWeekdayName( $key ) {
1063 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1067 * @param string $key
1070 function getWeekdayAbbreviation( $key ) {
1071 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1075 * @param string $key
1078 function getIranianCalendarMonthName( $key ) {
1079 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1083 * @param string $key
1086 function getHebrewCalendarMonthName( $key ) {
1087 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1091 * @param string $key
1094 function getHebrewCalendarMonthNameGen( $key ) {
1095 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1099 * @param string $key
1102 function getHijriCalendarMonthName( $key ) {
1103 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1107 * Pass through result from $dateTimeObj->format()
1108 * @param DateTime|bool|null &$dateTimeObj
1110 * @param DateTimeZone|bool|null $zone
1111 * @param string $code
1114 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1115 if ( !$dateTimeObj ) {
1116 $dateTimeObj = DateTime
::createFromFormat(
1117 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1120 return $dateTimeObj->format( $code );
1124 * This is a workalike of PHP's date() function, but with better
1125 * internationalisation, a reduced set of format characters, and a better
1128 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1129 * the PHP manual for definitions. There are a number of extensions, which
1132 * xn Do not translate digits of the next numeric format character
1133 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1134 * xr Use roman numerals for the next numeric format character
1135 * xh Use hebrew numerals for the next numeric format character
1137 * xg Genitive month name
1139 * xij j (day number) in Iranian calendar
1140 * xiF F (month name) in Iranian calendar
1141 * xin n (month number) in Iranian calendar
1142 * xiy y (two digit year) in Iranian calendar
1143 * xiY Y (full year) in Iranian calendar
1145 * xjj j (day number) in Hebrew calendar
1146 * xjF F (month name) in Hebrew calendar
1147 * xjt t (days in month) in Hebrew calendar
1148 * xjx xg (genitive month name) in Hebrew calendar
1149 * xjn n (month number) in Hebrew calendar
1150 * xjY Y (full year) in Hebrew calendar
1152 * xmj j (day number) in Hijri calendar
1153 * xmF F (month name) in Hijri calendar
1154 * xmn n (month number) in Hijri calendar
1155 * xmY Y (full year) in Hijri calendar
1157 * xkY Y (full year) in Thai solar calendar. Months and days are
1158 * identical to the Gregorian calendar
1159 * xoY Y (full year) in Minguo calendar or Juche year.
1160 * Months and days are identical to the
1161 * Gregorian calendar
1162 * xtY Y (full year) in Japanese nengo. Months and days are
1163 * identical to the Gregorian calendar
1165 * Characters enclosed in double quotes will be considered literal (with
1166 * the quotes themselves removed). Unmatched quotes will be considered
1167 * literal quotes. Example:
1169 * "The month is" F => The month is January
1172 * Backslash escaping is also supported.
1174 * Input timestamp is assumed to be pre-normalized to the desired local
1175 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1176 * $ts is UTC if $zone is not given.
1178 * @param string $format
1179 * @param string $ts 14-character timestamp
1182 * @param DateTimeZone $zone Timezone of $ts
1183 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1184 * Only makes sense if $ts is the current time.
1185 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1187 * @throws MWException
1190 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1195 $dateTimeObj = false;
1204 $usedSecond = false;
1205 $usedMinute = false;
1212 $usedISOYear = false;
1213 $usedIsLeapYear = false;
1215 $usedHebrewMonth = false;
1216 $usedIranianMonth = false;
1217 $usedHijriMonth = false;
1218 $usedHebrewYear = false;
1219 $usedIranianYear = false;
1220 $usedHijriYear = false;
1221 $usedTennoYear = false;
1223 if ( strlen( $ts ) !== 14 ) {
1224 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1227 if ( !ctype_digit( $ts ) ) {
1228 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1231 $formatLength = strlen( $format );
1232 for ( $p = 0; $p < $formatLength; $p++
) {
1234 $code = $format[$p];
1235 if ( $code == 'x' && $p < $formatLength - 1 ) {
1236 $code .= $format[++
$p];
1239 if ( ( $code === 'xi'
1245 && $p < $formatLength - 1 ) {
1246 $code .= $format[++
$p];
1257 $rawToggle = !$rawToggle;
1267 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1270 $usedHebrewMonth = true;
1272 $hebrew = self
::tsToHebrew( $ts );
1274 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1278 $num = substr( $ts, 6, 2 );
1282 $s .= $this->getWeekdayAbbreviation(
1283 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1288 $num = intval( substr( $ts, 6, 2 ) );
1293 $iranian = self
::tsToIranian( $ts );
1300 $hijri = self
::tsToHijri( $ts );
1307 $hebrew = self
::tsToHebrew( $ts );
1313 $s .= $this->getWeekdayName(
1314 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1319 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1322 $usedIranianMonth = true;
1324 $iranian = self
::tsToIranian( $ts );
1326 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1329 $usedHijriMonth = true;
1331 $hijri = self
::tsToHijri( $ts );
1333 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1336 $usedHebrewMonth = true;
1338 $hebrew = self
::tsToHebrew( $ts );
1340 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1344 $num = substr( $ts, 4, 2 );
1348 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1352 $num = intval( substr( $ts, 4, 2 ) );
1355 $usedIranianMonth = true;
1357 $iranian = self
::tsToIranian( $ts );
1362 $usedHijriMonth = true;
1364 $hijri = self
::tsToHijri( $ts );
1369 $usedHebrewMonth = true;
1371 $hebrew = self
::tsToHebrew( $ts );
1376 $usedHebrewMonth = true;
1378 $hebrew = self
::tsToHebrew( $ts );
1384 $num = substr( $ts, 0, 4 );
1387 $usedIranianYear = true;
1389 $iranian = self
::tsToIranian( $ts );
1394 $usedHijriYear = true;
1396 $hijri = self
::tsToHijri( $ts );
1401 $usedHebrewYear = true;
1403 $hebrew = self
::tsToHebrew( $ts );
1410 $thai = self
::tsToYear( $ts, 'thai' );
1417 $minguo = self
::tsToYear( $ts, 'minguo' );
1422 $usedTennoYear = true;
1424 $tenno = self
::tsToYear( $ts, 'tenno' );
1430 $num = substr( $ts, 2, 2 );
1433 $usedIranianYear = true;
1435 $iranian = self
::tsToIranian( $ts );
1437 $num = substr( $iranian[0], -2 );
1441 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1445 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1449 $h = substr( $ts, 8, 2 );
1450 $num = $h %
12 ?
$h %
12 : 12;
1454 $num = intval( substr( $ts, 8, 2 ) );
1458 $h = substr( $ts, 8, 2 );
1459 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1463 $num = substr( $ts, 8, 2 );
1467 $num = substr( $ts, 10, 2 );
1471 $num = substr( $ts, 12, 2 );
1481 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1487 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1491 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1495 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1498 $usedIsLeapYear = true;
1499 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1502 $usedISOYear = true;
1503 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1510 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1513 # Backslash escaping
1514 if ( $p < $formatLength - 1 ) {
1515 $s .= $format[++
$p];
1522 if ( $p < $formatLength - 1 ) {
1523 $endQuote = strpos( $format, '"', $p +
1 );
1524 if ( $endQuote === false ) {
1525 # No terminating quote, assume literal "
1528 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1532 # Quote at end of string, assume literal "
1539 if ( $num !== false ) {
1540 if ( $rawToggle ||
$raw ) {
1543 } elseif ( $roman ) {
1544 $s .= Language
::romanNumeral( $num );
1546 } elseif ( $hebrewNum ) {
1547 $s .= self
::hebrewNumeral( $num );
1550 $s .= $this->formatNum( $num, true );
1555 if ( $usedSecond ) {
1557 } elseif ( $usedMinute ) {
1558 $ttl = 60 - substr( $ts, 12, 2 );
1559 } elseif ( $usedHour ) {
1560 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1561 } elseif ( $usedAMPM ) {
1562 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1563 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1567 $usedIranianMonth ||
1574 // @todo Someone who understands the non-Gregorian calendars
1575 // should write proper logic for them so that they don't need purged every day.
1576 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1577 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1579 $possibleTtls = array();
1580 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1581 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1584 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1585 $timeRemainingInDay;
1586 } elseif ( $usedISOYear ) {
1587 // December 28th falls on the last ISO week of the year, every year.
1588 // The last ISO week of a year can be 52 or 53.
1589 $lastWeekOfISOYear = DateTime
::createFromFormat(
1591 substr( $ts, 0, 4 ) . '1228',
1592 $zone ?
: new DateTimeZone( 'UTC' )
1594 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1595 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1596 $timeRemainingInWeek =
1597 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1598 +
$timeRemainingInDay;
1599 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1604 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1605 substr( $ts, 6, 2 ) ) * 86400
1606 +
$timeRemainingInDay;
1607 } elseif ( $usedYear ) {
1609 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1610 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1611 +
$timeRemainingInDay;
1612 } elseif ( $usedIsLeapYear ) {
1613 $year = substr( $ts, 0, 4 );
1614 $timeRemainingInYear =
1615 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1616 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1617 +
$timeRemainingInDay;
1619 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1620 // this isn't a leap year. see when the next one starts
1621 $nextCandidate = $year - $mod +
4;
1622 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1623 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1624 $timeRemainingInYear;
1626 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1627 $timeRemainingInYear;
1630 // this is a leap year, so the next year isn't
1631 $possibleTtls[] = $timeRemainingInYear;
1635 if ( $possibleTtls ) {
1636 $ttl = min( $possibleTtls );
1643 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1644 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1647 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1648 * Gregorian dates to Iranian dates. Originally written in C, it
1649 * is released under the terms of GNU Lesser General Public
1650 * License. Conversion to PHP was performed by Niklas Laxström.
1652 * Link: http://www.farsiweb.info/jalali/jalali.c
1658 private static function tsToIranian( $ts ) {
1659 $gy = substr( $ts, 0, 4 ) -1600;
1660 $gm = substr( $ts, 4, 2 ) -1;
1661 $gd = substr( $ts, 6, 2 ) -1;
1663 # Days passed from the beginning (including leap years)
1665 +
floor( ( $gy +
3 ) / 4 )
1666 - floor( ( $gy +
99 ) / 100 )
1667 +
floor( ( $gy +
399 ) / 400 );
1669 // Add days of the past months of this year
1670 for ( $i = 0; $i < $gm; $i++
) {
1671 $gDayNo +
= self
::$GREG_DAYS[$i];
1675 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1679 // Days passed in current month
1680 $gDayNo +
= (int)$gd;
1682 $jDayNo = $gDayNo - 79;
1684 $jNp = floor( $jDayNo / 12053 );
1687 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1690 if ( $jDayNo >= 366 ) {
1691 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1692 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1695 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1696 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1702 return array( $jy, $jm, $jd );
1706 * Converting Gregorian dates to Hijri dates.
1708 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1710 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1716 private static function tsToHijri( $ts ) {
1717 $year = substr( $ts, 0, 4 );
1718 $month = substr( $ts, 4, 2 );
1719 $day = substr( $ts, 6, 2 );
1727 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1728 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1730 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1731 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1732 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1735 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1736 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1739 $zl = $zjd -1948440 +
10632;
1740 $zn = (int)( ( $zl - 1 ) / 10631 );
1741 $zl = $zl - 10631 * $zn +
354;
1742 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1743 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1744 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1745 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1746 $zm = (int)( ( 24 * $zl ) / 709 );
1747 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1748 $zy = 30 * $zn +
$zj - 30;
1750 return array( $zy, $zm, $zd );
1754 * Converting Gregorian dates to Hebrew dates.
1756 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1757 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1758 * to translate the relevant functions into PHP and release them under
1761 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1762 * and Adar II is 14. In a non-leap year, Adar is 6.
1768 private static function tsToHebrew( $ts ) {
1770 $year = substr( $ts, 0, 4 );
1771 $month = substr( $ts, 4, 2 );
1772 $day = substr( $ts, 6, 2 );
1774 # Calculate Hebrew year
1775 $hebrewYear = $year +
3760;
1777 # Month number when September = 1, August = 12
1779 if ( $month > 12 ) {
1786 # Calculate day of year from 1 September
1788 for ( $i = 1; $i < $month; $i++
) {
1792 # Check if the year is leap
1793 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1796 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1803 # Calculate the start of the Hebrew year
1804 $start = self
::hebrewYearStart( $hebrewYear );
1806 # Calculate next year's start
1807 if ( $dayOfYear <= $start ) {
1808 # Day is before the start of the year - it is the previous year
1810 $nextStart = $start;
1814 # Add days since previous year's 1 September
1816 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1820 # Start of the new (previous) year
1821 $start = self
::hebrewYearStart( $hebrewYear );
1824 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1827 # Calculate Hebrew day of year
1828 $hebrewDayOfYear = $dayOfYear - $start;
1830 # Difference between year's days
1831 $diff = $nextStart - $start;
1832 # Add 12 (or 13 for leap years) days to ignore the difference between
1833 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1834 # difference is only about the year type
1835 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1841 # Check the year pattern, and is leap year
1842 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1843 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1844 # and non-leap years
1845 $yearPattern = $diff %
30;
1846 # Check if leap year
1847 $isLeap = $diff >= 30;
1849 # Calculate day in the month from number of day in the Hebrew year
1850 # Don't check Adar - if the day is not in Adar, we will stop before;
1851 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1852 $hebrewDay = $hebrewDayOfYear;
1855 while ( $hebrewMonth <= 12 ) {
1856 # Calculate days in this month
1857 if ( $isLeap && $hebrewMonth == 6 ) {
1858 # Adar in a leap year
1860 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1862 if ( $hebrewDay <= $days ) {
1866 # Subtract the days of Adar I
1867 $hebrewDay -= $days;
1870 if ( $hebrewDay <= $days ) {
1876 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1877 # Cheshvan in a complete year (otherwise as the rule below)
1879 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1880 # Kislev in an incomplete year (otherwise as the rule below)
1883 # Odd months have 30 days, even have 29
1884 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1886 if ( $hebrewDay <= $days ) {
1887 # In the current month
1890 # Subtract the days of the current month
1891 $hebrewDay -= $days;
1892 # Try in the next month
1897 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1901 * This calculates the Hebrew year start, as days since 1 September.
1902 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1903 * Used for Hebrew date.
1909 private static function hebrewYearStart( $year ) {
1910 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1911 $b = intval( ( $year - 1 ) %
4 );
1912 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1916 $Mar = intval( $m );
1922 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1923 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1925 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1927 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1931 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1936 * Algorithm to convert Gregorian dates to Thai solar dates,
1937 * Minguo dates or Minguo dates.
1939 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1940 * http://en.wikipedia.org/wiki/Minguo_calendar
1941 * http://en.wikipedia.org/wiki/Japanese_era_name
1943 * @param string $ts 14-character timestamp
1944 * @param string $cName Calender name
1945 * @return array Converted year, month, day
1947 private static function tsToYear( $ts, $cName ) {
1948 $gy = substr( $ts, 0, 4 );
1949 $gm = substr( $ts, 4, 2 );
1950 $gd = substr( $ts, 6, 2 );
1952 if ( !strcmp( $cName, 'thai' ) ) {
1954 # Add 543 years to the Gregorian calendar
1955 # Months and days are identical
1956 $gy_offset = $gy +
543;
1957 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1959 # Deduct 1911 years from the Gregorian calendar
1960 # Months and days are identical
1961 $gy_offset = $gy - 1911;
1962 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1963 # Nengō dates up to Meiji period
1964 # Deduct years from the Gregorian calendar
1965 # depending on the nengo periods
1966 # Months and days are identical
1968 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1969 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1972 $gy_gannen = $gy - 1868 +
1;
1973 $gy_offset = $gy_gannen;
1974 if ( $gy_gannen == 1 ) {
1977 $gy_offset = '明治' . $gy_offset;
1979 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1980 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1981 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1982 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1983 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1986 $gy_gannen = $gy - 1912 +
1;
1987 $gy_offset = $gy_gannen;
1988 if ( $gy_gannen == 1 ) {
1991 $gy_offset = '大正' . $gy_offset;
1993 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1994 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1995 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1998 $gy_gannen = $gy - 1926 +
1;
1999 $gy_offset = $gy_gannen;
2000 if ( $gy_gannen == 1 ) {
2003 $gy_offset = '昭和' . $gy_offset;
2006 $gy_gannen = $gy - 1989 +
1;
2007 $gy_offset = $gy_gannen;
2008 if ( $gy_gannen == 1 ) {
2011 $gy_offset = '平成' . $gy_offset;
2017 return array( $gy_offset, $gm, $gd );
2021 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2023 * This is the rule the BIDI algorithm uses to determine the directionality of
2024 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
2025 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2027 * TODO: Does not handle BIDI control characters inside the text.
2028 * TODO: Does not handle unallocated characters.
2030 * @param string $text Text to test
2031 * @return null|string Directionality ('ltr' or 'rtl') or null
2033 private static function strongDirFromContent( $text = '' ) {
2034 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2037 if ( $matches[1] === '' ) {
2044 * Roman number formatting up to 10000
2050 static function romanNumeral( $num ) {
2051 static $table = array(
2052 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
2053 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
2054 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
2055 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2056 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
2059 $num = intval( $num );
2060 if ( $num > 10000 ||
$num <= 0 ) {
2065 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2066 if ( $num >= $pow10 ) {
2067 $s .= $table[$i][(int)floor( $num / $pow10 )];
2069 $num = $num %
$pow10;
2075 * Hebrew Gematria number formatting up to 9999
2081 static function hebrewNumeral( $num ) {
2082 static $table = array(
2083 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2084 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2094 array( 'ת', 'ת', 'ק' ),
2095 array( 'ת', 'ת', 'ר' ),
2097 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2100 $num = intval( $num );
2101 if ( $num > 9999 ||
$num <= 0 ) {
2105 // Round thousands have special notations
2106 if ( $num === 1000 ) {
2108 } elseif ( $num %
1000 === 0 ) {
2109 return $table[0][$num / 1000] . "' אלפים";
2114 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2115 if ( $num >= $pow10 ) {
2116 if ( $num === 15 ||
$num === 16 ) {
2117 $letters[] = $table[0][9];
2118 $letters[] = $table[0][$num - 9];
2121 $letters = array_merge(
2123 (array)$table[$i][intval( $num / $pow10 )]
2126 if ( $pow10 === 1000 ) {
2132 $num = $num %
$pow10;
2135 $preTransformLength = count( $letters );
2136 if ( $preTransformLength === 1 ) {
2137 // Add geresh (single quote) to one-letter numbers
2140 $lastIndex = $preTransformLength - 1;
2141 $letters[$lastIndex] = str_replace(
2142 array( 'כ', 'מ', 'נ', 'פ', 'צ' ),
2143 array( 'ך', 'ם', 'ן', 'ף', 'ץ' ),
2144 $letters[$lastIndex]
2147 // Add gershayim (double quote) to multiple-letter numbers,
2148 // but exclude numbers with only one letter after the thousands
2149 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2150 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2153 array_splice( $letters, -1, 0, '"' );
2157 return implode( $letters );
2161 * Used by date() and time() to adjust the time output.
2163 * @param string $ts The time in date('YmdHis') format
2164 * @param mixed $tz Adjust the time by this amount (default false, mean we
2165 * get user timecorrection setting)
2168 function userAdjust( $ts, $tz = false ) {
2169 global $wgUser, $wgLocalTZoffset;
2171 if ( $tz === false ) {
2172 $tz = $wgUser->getOption( 'timecorrection' );
2175 $data = explode( '|', $tz, 3 );
2177 if ( $data[0] == 'ZoneInfo' ) {
2178 MediaWiki\
suppressWarnings();
2179 $userTZ = timezone_open( $data[2] );
2180 MediaWiki\restoreWarnings
();
2181 if ( $userTZ !== false ) {
2182 $date = date_create( $ts, timezone_open( 'UTC' ) );
2183 date_timezone_set( $date, $userTZ );
2184 $date = date_format( $date, 'YmdHis' );
2187 # Unrecognized timezone, default to 'Offset' with the stored offset.
2188 $data[0] = 'Offset';
2191 if ( $data[0] == 'System' ||
$tz == '' ) {
2192 # Global offset in minutes.
2193 $minDiff = $wgLocalTZoffset;
2194 } elseif ( $data[0] == 'Offset' ) {
2195 $minDiff = intval( $data[1] );
2197 $data = explode( ':', $tz );
2198 if ( count( $data ) == 2 ) {
2199 $data[0] = intval( $data[0] );
2200 $data[1] = intval( $data[1] );
2201 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2202 if ( $data[0] < 0 ) {
2203 $minDiff = -$minDiff;
2206 $minDiff = intval( $data[0] ) * 60;
2210 # No difference ? Return time unchanged
2211 if ( 0 == $minDiff ) {
2215 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2216 # Generate an adjusted date; take advantage of the fact that mktime
2217 # will normalize out-of-range values so we don't have to split $minDiff
2218 # into hours and minutes.
2220 (int)substr( $ts, 8, 2 ) ), # Hours
2221 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2222 (int)substr( $ts, 12, 2 ), # Seconds
2223 (int)substr( $ts, 4, 2 ), # Month
2224 (int)substr( $ts, 6, 2 ), # Day
2225 (int)substr( $ts, 0, 4 ) ); # Year
2227 $date = date( 'YmdHis', $t );
2228 MediaWiki\restoreWarnings
();
2234 * This is meant to be used by time(), date(), and timeanddate() to get
2235 * the date preference they're supposed to use, it should be used in
2239 * function timeanddate([...], $format = true) {
2240 * $datePreference = $this->dateFormat($format);
2245 * @param int|string|bool $usePrefs If true, the user's preference is used
2246 * if false, the site/language default is used
2247 * if int/string, assumed to be a format.
2250 function dateFormat( $usePrefs = true ) {
2253 if ( is_bool( $usePrefs ) ) {
2255 $datePreference = $wgUser->getDatePreference();
2257 $datePreference = (string)User
::getDefaultOption( 'date' );
2260 $datePreference = (string)$usePrefs;
2264 if ( $datePreference == '' ) {
2268 return $datePreference;
2272 * Get a format string for a given type and preference
2273 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2274 * @param string $pref The format name as it appears in Messages*.php under
2277 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2281 function getDateFormatString( $type, $pref ) {
2282 $wasDefault = false;
2283 if ( $pref == 'default' ) {
2285 $pref = $this->getDefaultDateFormat();
2288 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2289 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2291 if ( $type === 'pretty' && $df === null ) {
2292 $df = $this->getDateFormatString( 'date', $pref );
2295 if ( !$wasDefault && $df === null ) {
2296 $pref = $this->getDefaultDateFormat();
2297 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2300 $this->dateFormatStrings
[$type][$pref] = $df;
2302 return $this->dateFormatStrings
[$type][$pref];
2306 * @param string $ts The time format which needs to be turned into a
2307 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2308 * @param bool $adj Whether to adjust the time output according to the
2309 * user configured offset ($timecorrection)
2310 * @param mixed $format True to use user's date format preference
2311 * @param string|bool $timecorrection The time offset as returned by
2312 * validateTimeZone() in Special:Preferences
2315 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2316 $ts = wfTimestamp( TS_MW
, $ts );
2318 $ts = $this->userAdjust( $ts, $timecorrection );
2320 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2321 return $this->sprintfDate( $df, $ts );
2325 * @param string $ts The time format which needs to be turned into a
2326 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2327 * @param bool $adj Whether to adjust the time output according to the
2328 * user configured offset ($timecorrection)
2329 * @param mixed $format True to use user's date format preference
2330 * @param string|bool $timecorrection The time offset as returned by
2331 * validateTimeZone() in Special:Preferences
2334 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2335 $ts = wfTimestamp( TS_MW
, $ts );
2337 $ts = $this->userAdjust( $ts, $timecorrection );
2339 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2340 return $this->sprintfDate( $df, $ts );
2344 * @param string $ts The time format which needs to be turned into a
2345 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2346 * @param bool $adj Whether to adjust the time output according to the
2347 * user configured offset ($timecorrection)
2348 * @param mixed $format What format to return, if it's false output the
2349 * default one (default true)
2350 * @param string|bool $timecorrection The time offset as returned by
2351 * validateTimeZone() in Special:Preferences
2354 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2355 $ts = wfTimestamp( TS_MW
, $ts );
2357 $ts = $this->userAdjust( $ts, $timecorrection );
2359 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2360 return $this->sprintfDate( $df, $ts );
2364 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2368 * @param int $seconds The amount of seconds.
2369 * @param array $chosenIntervals The intervals to enable.
2373 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2374 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2376 $segments = array();
2378 foreach ( $intervals as $intervalName => $intervalValue ) {
2379 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2380 // duration-years, duration-decades, duration-centuries, duration-millennia
2381 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2382 $segments[] = $message->inLanguage( $this )->escaped();
2385 return $this->listToText( $segments );
2389 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2390 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2394 * @param int $seconds The amount of seconds.
2395 * @param array $chosenIntervals The intervals to enable.
2399 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2400 if ( empty( $chosenIntervals ) ) {
2401 $chosenIntervals = array(
2413 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2414 $sortedNames = array_keys( $intervals );
2415 $smallestInterval = array_pop( $sortedNames );
2417 $segments = array();
2419 foreach ( $intervals as $name => $length ) {
2420 $value = floor( $seconds / $length );
2422 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2423 $seconds -= $value * $length;
2424 $segments[$name] = $value;
2432 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2434 * @param string $type Can be 'date', 'time' or 'both'
2435 * @param string $ts The time format which needs to be turned into a
2436 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2437 * @param User $user User object used to get preferences for timezone and format
2438 * @param array $options Array, can contain the following keys:
2439 * - 'timecorrection': time correction, can have the following values:
2440 * - true: use user's preference
2441 * - false: don't use time correction
2442 * - int: value of time correction in minutes
2443 * - 'format': format to use, can have the following values:
2444 * - true: use user's preference
2445 * - false: use default preference
2446 * - string: format to use
2450 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2451 $ts = wfTimestamp( TS_MW
, $ts );
2452 $options +
= array( 'timecorrection' => true, 'format' => true );
2453 if ( $options['timecorrection'] !== false ) {
2454 if ( $options['timecorrection'] === true ) {
2455 $offset = $user->getOption( 'timecorrection' );
2457 $offset = $options['timecorrection'];
2459 $ts = $this->userAdjust( $ts, $offset );
2461 if ( $options['format'] === true ) {
2462 $format = $user->getDatePreference();
2464 $format = $options['format'];
2466 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2467 return $this->sprintfDate( $df, $ts );
2471 * Get the formatted date for the given timestamp and formatted for
2474 * @param mixed $ts Mixed: the time format which needs to be turned into a
2475 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2476 * @param User $user User object used to get preferences for timezone and format
2477 * @param array $options Array, can contain the following keys:
2478 * - 'timecorrection': time correction, can have the following values:
2479 * - true: use user's preference
2480 * - false: don't use time correction
2481 * - int: value of time correction in minutes
2482 * - 'format': format to use, can have the following values:
2483 * - true: use user's preference
2484 * - false: use default preference
2485 * - string: format to use
2489 public function userDate( $ts, User
$user, array $options = array() ) {
2490 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2494 * Get the formatted time for the given timestamp and formatted for
2497 * @param mixed $ts The time format which needs to be turned into a
2498 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2499 * @param User $user User object used to get preferences for timezone and format
2500 * @param array $options Array, can contain the following keys:
2501 * - 'timecorrection': time correction, can have the following values:
2502 * - true: use user's preference
2503 * - false: don't use time correction
2504 * - int: value of time correction in minutes
2505 * - 'format': format to use, can have the following values:
2506 * - true: use user's preference
2507 * - false: use default preference
2508 * - string: format to use
2512 public function userTime( $ts, User
$user, array $options = array() ) {
2513 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2517 * Get the formatted date and time for the given timestamp and formatted for
2520 * @param mixed $ts The time format which needs to be turned into a
2521 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2522 * @param User $user User object used to get preferences for timezone and format
2523 * @param array $options Array, can contain the following keys:
2524 * - 'timecorrection': time correction, can have the following values:
2525 * - true: use user's preference
2526 * - false: don't use time correction
2527 * - int: value of time correction in minutes
2528 * - 'format': format to use, can have the following values:
2529 * - true: use user's preference
2530 * - false: use default preference
2531 * - string: format to use
2535 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2536 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2540 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2542 * Determine the difference between the timestamp and the current time, and
2543 * generate a readable timestamp by returning "<N> <units> ago", where the
2544 * largest possible unit is used.
2546 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2548 * @param MWTimestamp $time
2549 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2550 * @param User|null $user User the timestamp is being generated for (or null to use main context's user)
2551 * @return string Formatted timestamp
2553 public function getHumanTimestamp( MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null ) {
2554 if ( $relativeTo === null ) {
2555 $relativeTo = new MWTimestamp();
2557 if ( $user === null ) {
2558 $user = RequestContext
::getMain()->getUser();
2561 // Adjust for the user's timezone.
2562 $offsetThis = $time->offsetForUser( $user );
2563 $offsetRel = $relativeTo->offsetForUser( $user );
2566 if ( Hooks
::run( 'GetHumanTimestamp', array( &$ts, $time, $relativeTo, $user, $this ) ) ) {
2567 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2570 // Reset the timezone on the objects.
2571 $time->timestamp
->sub( $offsetThis );
2572 $relativeTo->timestamp
->sub( $offsetRel );
2578 * Convert an MWTimestamp into a pretty human-readable timestamp using
2579 * the given user preferences and relative base time.
2581 * @see Language::getHumanTimestamp
2582 * @param MWTimestamp $ts Timestamp to prettify
2583 * @param MWTimestamp $relativeTo Base timestamp
2584 * @param User $user User preferences to use
2585 * @return string Human timestamp
2588 private function getHumanTimestampInternal( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2589 $diff = $ts->diff( $relativeTo );
2590 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2591 (int)$relativeTo->timestamp
->format( 'w' ) );
2592 $days = $diff->days ?
: (int)$diffDay;
2593 if ( $diff->invert ||
$days > 5
2594 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2596 // Timestamps are in different years: use full timestamp
2597 // Also do full timestamp for future dates
2599 * @todo FIXME: Add better handling of future timestamps.
2601 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2602 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2603 } elseif ( $days > 5 ) {
2604 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2605 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2606 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2607 } elseif ( $days > 1 ) {
2608 // Timestamp within the past week: show the day of the week and time
2609 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2610 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2612 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2613 $ts = wfMessage( "$weekday-at" )
2614 ->inLanguage( $this )
2615 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2617 } elseif ( $days == 1 ) {
2618 // Timestamp was yesterday: say 'yesterday' and the time.
2619 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2620 $ts = wfMessage( 'yesterday-at' )
2621 ->inLanguage( $this )
2622 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2624 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2625 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2626 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2627 $ts = wfMessage( 'today-at' )
2628 ->inLanguage( $this )
2629 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2632 // From here on in, the timestamp was soon enough ago so that we can simply say
2633 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2634 } elseif ( $diff->h
== 1 ) {
2635 // Less than 90 minutes, but more than an hour ago.
2636 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2637 } elseif ( $diff->i
>= 1 ) {
2638 // A few minutes ago.
2639 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2640 } elseif ( $diff->s
>= 30 ) {
2641 // Less than a minute, but more than 30 sec ago.
2642 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2644 // Less than 30 seconds ago.
2645 $ts = wfMessage( 'just-now' )->text();
2652 * @param string $key
2653 * @return array|null
2655 function getMessage( $key ) {
2656 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2662 function getAllMessages() {
2663 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2668 * @param string $out
2669 * @param string $string
2672 function iconv( $in, $out, $string ) {
2673 # This is a wrapper for iconv in all languages except esperanto,
2674 # which does some nasty x-conversions beforehand
2676 # Even with //IGNORE iconv can whine about illegal characters in
2677 # *input* string. We just ignore those too.
2678 # REF: http://bugs.php.net/bug.php?id=37166
2679 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2680 MediaWiki\
suppressWarnings();
2681 $text = iconv( $in, $out . '//IGNORE', $string );
2682 MediaWiki\restoreWarnings
();
2686 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2689 * @param array $matches
2690 * @return mixed|string
2692 function ucwordbreaksCallbackAscii( $matches ) {
2693 return $this->ucfirst( $matches[1] );
2697 * @param array $matches
2700 function ucwordbreaksCallbackMB( $matches ) {
2701 return mb_strtoupper( $matches[0] );
2705 * @param array $matches
2708 function ucCallback( $matches ) {
2709 list( $wikiUpperChars ) = self
::getCaseMaps();
2710 return strtr( $matches[1], $wikiUpperChars );
2714 * @param array $matches
2717 function lcCallback( $matches ) {
2718 list( , $wikiLowerChars ) = self
::getCaseMaps();
2719 return strtr( $matches[1], $wikiLowerChars );
2723 * @param array $matches
2726 function ucwordsCallbackMB( $matches ) {
2727 return mb_strtoupper( $matches[0] );
2731 * @param array $matches
2734 function ucwordsCallbackWiki( $matches ) {
2735 list( $wikiUpperChars ) = self
::getCaseMaps();
2736 return strtr( $matches[0], $wikiUpperChars );
2740 * Make a string's first character uppercase
2742 * @param string $str
2746 function ucfirst( $str ) {
2748 if ( $o < 96 ) { // if already uppercase...
2750 } elseif ( $o < 128 ) {
2751 return ucfirst( $str ); // use PHP's ucfirst()
2753 // fall back to more complex logic in case of multibyte strings
2754 return $this->uc( $str, true );
2759 * Convert a string to uppercase
2761 * @param string $str
2762 * @param bool $first
2766 function uc( $str, $first = false ) {
2767 if ( function_exists( 'mb_strtoupper' ) ) {
2769 if ( $this->isMultibyte( $str ) ) {
2770 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2772 return ucfirst( $str );
2775 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2778 if ( $this->isMultibyte( $str ) ) {
2779 $x = $first ?
'^' : '';
2780 return preg_replace_callback(
2781 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2782 array( $this, 'ucCallback' ),
2786 return $first ?
ucfirst( $str ) : strtoupper( $str );
2792 * @param string $str
2793 * @return mixed|string
2795 function lcfirst( $str ) {
2798 return strval( $str );
2799 } elseif ( $o >= 128 ) {
2800 return $this->lc( $str, true );
2801 } elseif ( $o > 96 ) {
2804 $str[0] = strtolower( $str[0] );
2810 * @param string $str
2811 * @param bool $first
2812 * @return mixed|string
2814 function lc( $str, $first = false ) {
2815 if ( function_exists( 'mb_strtolower' ) ) {
2817 if ( $this->isMultibyte( $str ) ) {
2818 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2820 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2823 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2826 if ( $this->isMultibyte( $str ) ) {
2827 $x = $first ?
'^' : '';
2828 return preg_replace_callback(
2829 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2830 array( $this, 'lcCallback' ),
2834 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2840 * @param string $str
2843 function isMultibyte( $str ) {
2844 return strlen( $str ) !== mb_strlen( $str );
2848 * @param string $str
2849 * @return mixed|string
2851 function ucwords( $str ) {
2852 if ( $this->isMultibyte( $str ) ) {
2853 $str = $this->lc( $str );
2855 // regexp to find first letter in each word (i.e. after each space)
2856 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2858 // function to use to capitalize a single char
2859 if ( function_exists( 'mb_strtoupper' ) ) {
2860 return preg_replace_callback(
2862 array( $this, 'ucwordsCallbackMB' ),
2866 return preg_replace_callback(
2868 array( $this, 'ucwordsCallbackWiki' ),
2873 return ucwords( strtolower( $str ) );
2878 * capitalize words at word breaks
2880 * @param string $str
2883 function ucwordbreaks( $str ) {
2884 if ( $this->isMultibyte( $str ) ) {
2885 $str = $this->lc( $str );
2887 // since \b doesn't work for UTF-8, we explicitely define word break chars
2888 $breaks = "[ \-\(\)\}\{\.,\?!]";
2890 // find first letter after word break
2891 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2892 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2894 if ( function_exists( 'mb_strtoupper' ) ) {
2895 return preg_replace_callback(
2897 array( $this, 'ucwordbreaksCallbackMB' ),
2901 return preg_replace_callback(
2903 array( $this, 'ucwordsCallbackWiki' ),
2908 return preg_replace_callback(
2909 '/\b([\w\x80-\xff]+)\b/',
2910 array( $this, 'ucwordbreaksCallbackAscii' ),
2917 * Return a case-folded representation of $s
2919 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2920 * and $s2 are the same except for the case of their characters. It is not
2921 * necessary for the value returned to make sense when displayed.
2923 * Do *not* perform any other normalisation in this function. If a caller
2924 * uses this function when it should be using a more general normalisation
2925 * function, then fix the caller.
2931 function caseFold( $s ) {
2932 return $this->uc( $s );
2939 function checkTitleEncoding( $s ) {
2940 if ( is_array( $s ) ) {
2941 throw new MWException( 'Given array to checkTitleEncoding.' );
2943 if ( StringUtils
::isUtf8( $s ) ) {
2947 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2953 function fallback8bitEncoding() {
2954 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2958 * Most writing systems use whitespace to break up words.
2959 * Some languages such as Chinese don't conventionally do this,
2960 * which requires special handling when breaking up words for
2965 function hasWordBreaks() {
2970 * Some languages such as Chinese require word segmentation,
2971 * Specify such segmentation when overridden in derived class.
2973 * @param string $string
2976 function segmentByWord( $string ) {
2981 * Some languages have special punctuation need to be normalized.
2982 * Make such changes here.
2984 * @param string $string
2987 function normalizeForSearch( $string ) {
2988 return self
::convertDoubleWidth( $string );
2992 * convert double-width roman characters to single-width.
2993 * range: ff00-ff5f ~= 0020-007f
2995 * @param string $string
2999 protected static function convertDoubleWidth( $string ) {
3000 static $full = null;
3001 static $half = null;
3003 if ( $full === null ) {
3004 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
3005 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
3006 $full = str_split( $fullWidth, 3 );
3007 $half = str_split( $halfWidth );
3010 $string = str_replace( $full, $half, $string );
3015 * @param string $string
3016 * @param string $pattern
3019 protected static function insertSpace( $string, $pattern ) {
3020 $string = preg_replace( $pattern, " $1 ", $string );
3021 $string = preg_replace( '/ +/', ' ', $string );
3026 * @param array $termsArray
3029 function convertForSearchResult( $termsArray ) {
3030 # some languages, e.g. Chinese, need to do a conversion
3031 # in order for search results to be displayed correctly
3036 * Get the first character of a string.
3041 function firstChar( $s ) {
3044 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
3045 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
3050 if ( isset( $matches[1] ) ) {
3051 if ( strlen( $matches[1] ) != 3 ) {
3055 // Break down Hangul syllables to grab the first jamo
3056 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3057 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3059 } elseif ( $code < 0xb098 ) {
3060 return "\xe3\x84\xb1";
3061 } elseif ( $code < 0xb2e4 ) {
3062 return "\xe3\x84\xb4";
3063 } elseif ( $code < 0xb77c ) {
3064 return "\xe3\x84\xb7";
3065 } elseif ( $code < 0xb9c8 ) {
3066 return "\xe3\x84\xb9";
3067 } elseif ( $code < 0xbc14 ) {
3068 return "\xe3\x85\x81";
3069 } elseif ( $code < 0xc0ac ) {
3070 return "\xe3\x85\x82";
3071 } elseif ( $code < 0xc544 ) {
3072 return "\xe3\x85\x85";
3073 } elseif ( $code < 0xc790 ) {
3074 return "\xe3\x85\x87";
3075 } elseif ( $code < 0xcc28 ) {
3076 return "\xe3\x85\x88";
3077 } elseif ( $code < 0xce74 ) {
3078 return "\xe3\x85\x8a";
3079 } elseif ( $code < 0xd0c0 ) {
3080 return "\xe3\x85\x8b";
3081 } elseif ( $code < 0xd30c ) {
3082 return "\xe3\x85\x8c";
3083 } elseif ( $code < 0xd558 ) {
3084 return "\xe3\x85\x8d";
3086 return "\xe3\x85\x8e";
3093 function initEncoding() {
3094 # Some languages may have an alternate char encoding option
3095 # (Esperanto X-coding, Japanese furigana conversion, etc)
3096 # If this language is used as the primary content language,
3097 # an override to the defaults can be set here on startup.
3104 function recodeForEdit( $s ) {
3105 # For some languages we'll want to explicitly specify
3106 # which characters make it into the edit box raw
3107 # or are converted in some way or another.
3108 global $wgEditEncoding;
3109 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
3112 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3120 function recodeInput( $s ) {
3121 # Take the previous into account.
3122 global $wgEditEncoding;
3123 if ( $wgEditEncoding != '' ) {
3124 $enc = $wgEditEncoding;
3128 if ( $enc == 'UTF-8' ) {
3131 return $this->iconv( $enc, 'UTF-8', $s );
3136 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3137 * also cleans up certain backwards-compatible sequences, converting them
3138 * to the modern Unicode equivalent.
3140 * This is language-specific for performance reasons only.
3146 function normalize( $s ) {
3147 global $wgAllUnicodeFixes;
3148 $s = UtfNormal\Validator
::cleanUp( $s );
3149 if ( $wgAllUnicodeFixes ) {
3150 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3151 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3158 * Transform a string using serialized data stored in the given file (which
3159 * must be in the serialized subdirectory of $IP). The file contains pairs
3160 * mapping source characters to destination characters.
3162 * The data is cached in process memory. This will go faster if you have the
3163 * FastStringSearch extension.
3165 * @param string $file
3166 * @param string $string
3168 * @throws MWException
3171 function transformUsingPairFile( $file, $string ) {
3172 if ( !isset( $this->transformData
[$file] ) ) {
3173 $data = wfGetPrecompiledData( $file );
3174 if ( $data === false ) {
3175 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3177 $this->transformData
[$file] = new ReplacementArray( $data );
3179 return $this->transformData
[$file]->replace( $string );
3183 * For right-to-left language support
3188 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3192 * Return the correct HTML 'dir' attribute value for this language.
3196 return $this->isRTL() ?
'rtl' : 'ltr';
3200 * Return 'left' or 'right' as appropriate alignment for line-start
3201 * for this language's text direction.
3203 * Should be equivalent to CSS3 'start' text-align value....
3207 function alignStart() {
3208 return $this->isRTL() ?
'right' : 'left';
3212 * Return 'right' or 'left' as appropriate alignment for line-end
3213 * for this language's text direction.
3215 * Should be equivalent to CSS3 'end' text-align value....
3219 function alignEnd() {
3220 return $this->isRTL() ?
'left' : 'right';
3224 * A hidden direction mark (LRM or RLM), depending on the language direction.
3225 * Unlike getDirMark(), this function returns the character as an HTML entity.
3226 * This function should be used when the output is guaranteed to be HTML,
3227 * because it makes the output HTML source code more readable. When
3228 * the output is plain text or can be escaped, getDirMark() should be used.
3230 * @param bool $opposite Get the direction mark opposite to your language
3234 function getDirMarkEntity( $opposite = false ) {
3236 return $this->isRTL() ?
'‎' : '‏';
3238 return $this->isRTL() ?
'‏' : '‎';
3242 * A hidden direction mark (LRM or RLM), depending on the language direction.
3243 * This function produces them as invisible Unicode characters and
3244 * the output may be hard to read and debug, so it should only be used
3245 * when the output is plain text or can be escaped. When the output is
3246 * HTML, use getDirMarkEntity() instead.
3248 * @param bool $opposite Get the direction mark opposite to your language
3251 function getDirMark( $opposite = false ) {
3252 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3253 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3255 return $this->isRTL() ?
$lrm : $rlm;
3257 return $this->isRTL() ?
$rlm : $lrm;
3263 function capitalizeAllNouns() {
3264 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3268 * An arrow, depending on the language direction.
3270 * @param string $direction The direction of the arrow: forwards (default),
3271 * backwards, left, right, up, down.
3274 function getArrow( $direction = 'forwards' ) {
3275 switch ( $direction ) {
3277 return $this->isRTL() ?
'←' : '→';
3279 return $this->isRTL() ?
'→' : '←';
3292 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3296 function linkPrefixExtension() {
3297 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3301 * Get all magic words from cache.
3304 function getMagicWords() {
3305 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3309 * Run the LanguageGetMagic hook once.
3311 protected function doMagicHook() {
3312 if ( $this->mMagicHookDone
) {
3315 $this->mMagicHookDone
= true;
3316 Hooks
::run( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3320 * Fill a MagicWord object with data from here
3322 * @param MagicWord $mw
3324 function getMagic( $mw ) {
3325 // Saves a function call
3326 if ( !$this->mMagicHookDone
) {
3327 $this->doMagicHook();
3330 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3331 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3333 $rawEntry = self
::$dataCache->getSubitem(
3334 $this->mCode
, 'magicWords', $mw->mId
);
3337 if ( !is_array( $rawEntry ) ) {
3338 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3340 $mw->mCaseSensitive
= $rawEntry[0];
3341 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3346 * Add magic words to the extension array
3348 * @param array $newWords
3350 function addMagicWordsByLang( $newWords ) {
3351 $fallbackChain = $this->getFallbackLanguages();
3352 $fallbackChain = array_reverse( $fallbackChain );
3353 foreach ( $fallbackChain as $code ) {
3354 if ( isset( $newWords[$code] ) ) {
3355 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3361 * Get special page names, as an associative array
3362 * canonical name => array of valid names, including aliases
3365 function getSpecialPageAliases() {
3366 // Cache aliases because it may be slow to load them
3367 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3369 $this->mExtendedSpecialPageAliases
=
3370 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3371 Hooks
::run( 'LanguageGetSpecialPageAliases',
3372 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3375 return $this->mExtendedSpecialPageAliases
;
3379 * Italic is unsuitable for some languages
3381 * @param string $text The text to be emphasized.
3384 function emphasize( $text ) {
3385 return "<em>$text</em>";
3389 * Normally we output all numbers in plain en_US style, that is
3390 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3391 * point twohundredthirtyfive. However this is not suitable for all
3392 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3393 * Icelandic just want to use commas instead of dots, and dots instead
3394 * of commas like "293.291,235".
3396 * An example of this function being called:
3398 * wfMessage( 'message' )->numParams( $num )->text()
3401 * See $separatorTransformTable on MessageIs.php for
3402 * the , => . and . => , implementation.
3404 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3405 * @param int|float $number The string to be formatted, should be an integer
3406 * or a floating point number.
3407 * @param bool $nocommafy Set to true for special numbers like dates
3410 public function formatNum( $number, $nocommafy = false ) {
3411 global $wgTranslateNumerals;
3412 if ( !$nocommafy ) {
3413 $number = $this->commafy( $number );
3414 $s = $this->separatorTransformTable();
3416 $number = strtr( $number, $s );
3420 if ( $wgTranslateNumerals ) {
3421 $s = $this->digitTransformTable();
3423 $number = strtr( $number, $s );
3431 * Front-end for non-commafied formatNum
3433 * @param int|float $number The string to be formatted, should be an integer
3434 * or a floating point number.
3438 public function formatNumNoSeparators( $number ) {
3439 return $this->formatNum( $number, true );
3443 * @param string $number
3446 public function parseFormattedNumber( $number ) {
3447 $s = $this->digitTransformTable();
3449 // eliminate empty array values such as ''. (bug 64347)
3450 $s = array_filter( $s );
3451 $number = strtr( $number, array_flip( $s ) );
3454 $s = $this->separatorTransformTable();
3456 // eliminate empty array values such as ''. (bug 64347)
3457 $s = array_filter( $s );
3458 $number = strtr( $number, array_flip( $s ) );
3461 $number = strtr( $number, array( ',' => '' ) );
3466 * Adds commas to a given number
3468 * @param mixed $number
3471 function commafy( $number ) {
3472 $digitGroupingPattern = $this->digitGroupingPattern();
3473 if ( $number === null ) {
3477 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3478 // default grouping is at thousands, use the same for ###,###,### pattern too.
3479 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3481 // Ref: http://cldr.unicode.org/translation/number-patterns
3483 if ( intval( $number ) < 0 ) {
3484 // For negative numbers apply the algorithm like positive number and add sign.
3486 $number = substr( $number, 1 );
3488 $integerPart = array();
3489 $decimalPart = array();
3490 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3491 preg_match( "/\d+/", $number, $integerPart );
3492 preg_match( "/\.\d*/", $number, $decimalPart );
3493 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3494 if ( $groupedNumber === $number ) {
3495 // the string does not have any number part. Eg: .12345
3496 return $sign . $groupedNumber;
3498 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3499 while ( $start > 0 ) {
3500 $match = $matches[0][$numMatches - 1];
3501 $matchLen = strlen( $match );
3502 $start = $end - $matchLen;
3506 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3508 if ( $numMatches > 1 ) {
3509 // use the last pattern for the rest of the number
3513 $groupedNumber = "," . $groupedNumber;
3516 return $sign . $groupedNumber;
3523 function digitGroupingPattern() {
3524 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3530 function digitTransformTable() {
3531 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3537 function separatorTransformTable() {
3538 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3542 * Take a list of strings and build a locale-friendly comma-separated
3543 * list, using the local comma-separator message.
3544 * The last two strings are chained with an "and".
3545 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3547 * @param string[] $l
3550 function listToText( array $l ) {
3551 $m = count( $l ) - 1;
3556 $and = $this->msg( 'and' )->escaped();
3557 $space = $this->msg( 'word-separator' )->escaped();
3559 $comma = $this->msg( 'comma-separator' )->escaped();
3563 for ( $i = $m - 1; $i >= 0; $i-- ) {
3564 if ( $i == $m - 1 ) {
3565 $s = $l[$i] . $and . $space . $s;
3567 $s = $l[$i] . $comma . $s;
3574 * Take a list of strings and build a locale-friendly comma-separated
3575 * list, using the local comma-separator message.
3576 * @param string[] $list Array of strings to put in a comma list
3579 function commaList( array $list ) {
3581 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3587 * Take a list of strings and build a locale-friendly semicolon-separated
3588 * list, using the local semicolon-separator message.
3589 * @param string[] $list Array of strings to put in a semicolon list
3592 function semicolonList( array $list ) {
3594 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3600 * Same as commaList, but separate it with the pipe instead.
3601 * @param string[] $list Array of strings to put in a pipe list
3604 function pipeList( array $list ) {
3606 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3612 * Truncate a string to a specified length in bytes, appending an optional
3613 * string (e.g. for ellipses)
3615 * The database offers limited byte lengths for some columns in the database;
3616 * multi-byte character sets mean we need to ensure that only whole characters
3617 * are included, otherwise broken characters can be passed to the user
3619 * If $length is negative, the string will be truncated from the beginning
3621 * @param string $string String to truncate
3622 * @param int $length Maximum length (including ellipses)
3623 * @param string $ellipsis String to append to the truncated text
3624 * @param bool $adjustLength Subtract length of ellipsis from $length.
3625 * $adjustLength was introduced in 1.18, before that behaved as if false.
3628 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3629 # Use the localized ellipsis character
3630 if ( $ellipsis == '...' ) {
3631 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3633 # Check if there is no need to truncate
3634 if ( $length == 0 ) {
3635 return $ellipsis; // convention
3636 } elseif ( strlen( $string ) <= abs( $length ) ) {
3637 return $string; // no need to truncate
3639 $stringOriginal = $string;
3640 # If ellipsis length is >= $length then we can't apply $adjustLength
3641 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3642 $string = $ellipsis; // this can be slightly unexpected
3643 # Otherwise, truncate and add ellipsis...
3645 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3646 if ( $length > 0 ) {
3647 $length -= $eLength;
3648 $string = substr( $string, 0, $length ); // xyz...
3649 $string = $this->removeBadCharLast( $string );
3650 $string = rtrim( $string );
3651 $string = $string . $ellipsis;
3653 $length +
= $eLength;
3654 $string = substr( $string, $length ); // ...xyz
3655 $string = $this->removeBadCharFirst( $string );
3656 $string = ltrim( $string );
3657 $string = $ellipsis . $string;
3660 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3661 # This check is *not* redundant if $adjustLength, due to the single case where
3662 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3663 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3666 return $stringOriginal;
3671 * Remove bytes that represent an incomplete Unicode character
3672 * at the end of string (e.g. bytes of the char are missing)
3674 * @param string $string
3677 protected function removeBadCharLast( $string ) {
3678 if ( $string != '' ) {
3679 $char = ord( $string[strlen( $string ) - 1] );
3681 if ( $char >= 0xc0 ) {
3682 # We got the first byte only of a multibyte char; remove it.
3683 $string = substr( $string, 0, -1 );
3684 } elseif ( $char >= 0x80 &&
3685 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3686 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3688 # We chopped in the middle of a character; remove it
3696 * Remove bytes that represent an incomplete Unicode character
3697 * at the start of string (e.g. bytes of the char are missing)
3699 * @param string $string
3702 protected function removeBadCharFirst( $string ) {
3703 if ( $string != '' ) {
3704 $char = ord( $string[0] );
3705 if ( $char >= 0x80 && $char < 0xc0 ) {
3706 # We chopped in the middle of a character; remove the whole thing
3707 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3714 * Truncate a string of valid HTML to a specified length in bytes,
3715 * appending an optional string (e.g. for ellipses), and return valid HTML
3717 * This is only intended for styled/linked text, such as HTML with
3718 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3719 * Also, this will not detect things like "display:none" CSS.
3721 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3723 * @param string $text HTML string to truncate
3724 * @param int $length (zero/positive) Maximum length (including ellipses)
3725 * @param string $ellipsis String to append to the truncated text
3728 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3729 # Use the localized ellipsis character
3730 if ( $ellipsis == '...' ) {
3731 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3733 # Check if there is clearly no need to truncate
3734 if ( $length <= 0 ) {
3735 return $ellipsis; // no text shown, nothing to format (convention)
3736 } elseif ( strlen( $text ) <= $length ) {
3737 return $text; // string short enough even *with* HTML (short-circuit)
3740 $dispLen = 0; // innerHTML legth so far
3741 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3742 $tagType = 0; // 0-open, 1-close
3743 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3744 $entityState = 0; // 0-not entity, 1-entity
3745 $tag = $ret = ''; // accumulated tag name, accumulated result string
3746 $openTags = array(); // open tag stack
3747 $maybeState = null; // possible truncation state
3749 $textLen = strlen( $text );
3750 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3751 for ( $pos = 0; true; ++
$pos ) {
3752 # Consider truncation once the display length has reached the maximim.
3753 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3754 # Check that we're not in the middle of a bracket/entity...
3755 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3756 if ( !$testingEllipsis ) {
3757 $testingEllipsis = true;
3758 # Save where we are; we will truncate here unless there turn out to
3759 # be so few remaining characters that truncation is not necessary.
3760 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3761 $maybeState = array( $ret, $openTags ); // save state
3763 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3764 # String in fact does need truncation, the truncation point was OK.
3765 list( $ret, $openTags ) = $maybeState; // reload state
3766 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3767 $ret .= $ellipsis; // add ellipsis
3771 if ( $pos >= $textLen ) {
3772 break; // extra iteration just for above checks
3775 # Read the next char...
3777 $lastCh = $pos ?
$text[$pos - 1] : '';
3778 $ret .= $ch; // add to result string
3780 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3781 $entityState = 0; // for bad HTML
3782 $bracketState = 1; // tag started (checking for backslash)
3783 } elseif ( $ch == '>' ) {
3784 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3785 $entityState = 0; // for bad HTML
3786 $bracketState = 0; // out of brackets
3787 } elseif ( $bracketState == 1 ) {
3789 $tagType = 1; // close tag (e.g. "</span>")
3791 $tagType = 0; // open tag (e.g. "<span>")
3794 $bracketState = 2; // building tag name
3795 } elseif ( $bracketState == 2 ) {
3799 // Name found (e.g. "<a href=..."), add on tag attributes...
3800 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3802 } elseif ( $bracketState == 0 ) {
3803 if ( $entityState ) {
3806 $dispLen++
; // entity is one displayed char
3809 if ( $neLength == 0 && !$maybeState ) {
3810 // Save state without $ch. We want to *hit* the first
3811 // display char (to get tags) but not *use* it if truncating.
3812 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3815 $entityState = 1; // entity found, (e.g. " ")
3817 $dispLen++
; // this char is displayed
3818 // Add the next $max display text chars after this in one swoop...
3819 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3820 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3821 $dispLen +
= $skipped;
3827 // Close the last tag if left unclosed by bad HTML
3828 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3829 while ( count( $openTags ) > 0 ) {
3830 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3836 * truncateHtml() helper function
3837 * like strcspn() but adds the skipped chars to $ret
3839 * @param string $ret
3840 * @param string $text
3841 * @param string $search
3843 * @param null|int $len
3846 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3847 if ( $len === null ) {
3848 $len = -1; // -1 means "no limit" for strcspn
3849 } elseif ( $len < 0 ) {
3853 if ( $start < strlen( $text ) ) {
3854 $skipCount = strcspn( $text, $search, $start, $len );
3855 $ret .= substr( $text, $start, $skipCount );
3861 * truncateHtml() helper function
3862 * (a) push or pop $tag from $openTags as needed
3863 * (b) clear $tag value
3864 * @param string &$tag Current HTML tag name we are looking at
3865 * @param int $tagType (0-open tag, 1-close tag)
3866 * @param string $lastCh Character before the '>' that ended this tag
3867 * @param array &$openTags Open tag stack (not accounting for $tag)
3869 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3870 $tag = ltrim( $tag );
3872 if ( $tagType == 0 && $lastCh != '/' ) {
3873 $openTags[] = $tag; // tag opened (didn't close itself)
3874 } elseif ( $tagType == 1 ) {
3875 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3876 array_pop( $openTags ); // tag closed
3884 * Grammatical transformations, needed for inflected languages
3885 * Invoked by putting {{grammar:case|word}} in a message
3887 * @param string $word
3888 * @param string $case
3891 function convertGrammar( $word, $case ) {
3892 global $wgGrammarForms;
3893 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3894 return $wgGrammarForms[$this->getCode()][$case][$word];
3900 * Get the grammar forms for the content language
3901 * @return array Array of grammar forms
3904 function getGrammarForms() {
3905 global $wgGrammarForms;
3906 if ( isset( $wgGrammarForms[$this->getCode()] )
3907 && is_array( $wgGrammarForms[$this->getCode()] )
3909 return $wgGrammarForms[$this->getCode()];
3915 * Provides an alternative text depending on specified gender.
3916 * Usage {{gender:username|masculine|feminine|unknown}}.
3917 * username is optional, in which case the gender of current user is used,
3918 * but only in (some) interface messages; otherwise default gender is used.
3920 * If no forms are given, an empty string is returned. If only one form is
3921 * given, it will be returned unconditionally. These details are implied by
3922 * the caller and cannot be overridden in subclasses.
3924 * If three forms are given, the default is to use the third (unknown) form.
3925 * If fewer than three forms are given, the default is to use the first (masculine) form.
3926 * These details can be overridden in subclasses.
3928 * @param string $gender
3929 * @param array $forms
3933 function gender( $gender, $forms ) {
3934 if ( !count( $forms ) ) {
3937 $forms = $this->preConvertPlural( $forms, 2 );
3938 if ( $gender === 'male' ) {
3941 if ( $gender === 'female' ) {
3944 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3948 * Plural form transformations, needed for some languages.
3949 * For example, there are 3 form of plural in Russian and Polish,
3950 * depending on "count mod 10". See [[w:Plural]]
3951 * For English it is pretty simple.
3953 * Invoked by putting {{plural:count|wordform1|wordform2}}
3954 * or {{plural:count|wordform1|wordform2|wordform3}}
3956 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3958 * @param int $count Non-localized number
3959 * @param array $forms Different plural forms
3960 * @return string Correct form of plural for $count in this language
3962 function convertPlural( $count, $forms ) {
3963 // Handle explicit n=pluralform cases
3964 $forms = $this->handleExplicitPluralForms( $count, $forms );
3965 if ( is_string( $forms ) ) {
3968 if ( !count( $forms ) ) {
3972 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3973 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3974 return $forms[$pluralForm];
3978 * Handles explicit plural forms for Language::convertPlural()
3980 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3981 * If an explicitly defined plural form matches the $count, then
3982 * string value returned, otherwise array returned for further consideration
3983 * by CLDR rules or overridden convertPlural().
3987 * @param int $count Non-localized number
3988 * @param array $forms Different plural forms
3990 * @return array|string
3992 protected function handleExplicitPluralForms( $count, array $forms ) {
3993 foreach ( $forms as $index => $form ) {
3994 if ( preg_match( '/\d+=/i', $form ) ) {
3995 $pos = strpos( $form, '=' );
3996 if ( substr( $form, 0, $pos ) === (string)$count ) {
3997 return substr( $form, $pos +
1 );
3999 unset( $forms[$index] );
4002 return array_values( $forms );
4006 * Checks that convertPlural was given an array and pads it to requested
4007 * amount of forms by copying the last one.
4009 * @param array $forms Array of forms given to convertPlural
4010 * @param int $count How many forms should there be at least
4011 * @return array Padded array of forms or an exception if not an array
4013 protected function preConvertPlural( /* Array */ $forms, $count ) {
4014 while ( count( $forms ) < $count ) {
4015 $forms[] = $forms[count( $forms ) - 1];
4021 * Wraps argument with unicode control characters for directionality safety
4023 * This solves the problem where directionality-neutral characters at the edge of
4024 * the argument string get interpreted with the wrong directionality from the
4025 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4027 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4028 * directionality of the argument string, using the BIDI algorithm's own "First
4029 * strong directional codepoint" rule. Essentially, this works round the fact that
4030 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4031 * direction inference). The latter is cleaner but still not widely supported.
4033 * @param string $text Text to wrap
4034 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4036 public function embedBidi( $text = '' ) {
4037 $dir = Language
::strongDirFromContent( $text );
4038 if ( $dir === 'ltr' ) {
4039 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4040 return self
::$lre . $text . self
::$pdf;
4042 if ( $dir === 'rtl' ) {
4043 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4044 return self
::$rle . $text . self
::$pdf;
4046 // No strong directionality: do not wrap
4051 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4052 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4053 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4054 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4057 * @param string $str The validated block duration in English
4058 * @return string Somehow translated block duration
4059 * @see LanguageFi.php for example implementation
4061 function translateBlockExpiry( $str ) {
4062 $duration = SpecialBlock
::getSuggestedDurations( $this );
4063 foreach ( $duration as $show => $value ) {
4064 if ( strcmp( $str, $value ) == 0 ) {
4065 return htmlspecialchars( trim( $show ) );
4069 if ( wfIsInfinity( $str ) ) {
4070 foreach ( $duration as $show => $value ) {
4071 if ( wfIsInfinity( $value ) ) {
4072 return htmlspecialchars( trim( $show ) );
4077 // If all else fails, return a standard duration or timestamp description.
4078 $time = strtotime( $str, 0 );
4079 if ( $time === false ) { // Unknown format. Return it as-is in case.
4081 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4082 // $time is relative to 0 so it's a duration length.
4083 return $this->formatDuration( $time );
4084 } else { // It's an absolute timestamp.
4085 if ( $time === 0 ) {
4086 // wfTimestamp() handles 0 as current time instead of epoch.
4087 return $this->timeanddate( '19700101000000' );
4089 return $this->timeanddate( $time );
4095 * languages like Chinese need to be segmented in order for the diff
4098 * @param string $text
4101 public function segmentForDiff( $text ) {
4106 * and unsegment to show the result
4108 * @param string $text
4111 public function unsegmentForDiff( $text ) {
4116 * Return the LanguageConverter used in the Language
4119 * @return LanguageConverter
4121 public function getConverter() {
4122 return $this->mConverter
;
4126 * convert text to all supported variants
4128 * @param string $text
4131 public function autoConvertToAllVariants( $text ) {
4132 return $this->mConverter
->autoConvertToAllVariants( $text );
4136 * convert text to different variants of a language.
4138 * @param string $text
4141 public function convert( $text ) {
4142 return $this->mConverter
->convert( $text );
4146 * Convert a Title object to a string in the preferred variant
4148 * @param Title $title
4151 public function convertTitle( $title ) {
4152 return $this->mConverter
->convertTitle( $title );
4156 * Convert a namespace index to a string in the preferred variant
4161 public function convertNamespace( $ns ) {
4162 return $this->mConverter
->convertNamespace( $ns );
4166 * Check if this is a language with variants
4170 public function hasVariants() {
4171 return count( $this->getVariants() ) > 1;
4175 * Check if the language has the specific variant
4178 * @param string $variant
4181 public function hasVariant( $variant ) {
4182 return (bool)$this->mConverter
->validateVariant( $variant );
4186 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4188 * @param string $text
4190 * @deprecated since 1.22 is no longer used
4192 public function armourMath( $text ) {
4193 return $this->mConverter
->armourMath( $text );
4197 * Perform output conversion on a string, and encode for safe HTML output.
4198 * @param string $text Text to be converted
4199 * @param bool $isTitle Whether this conversion is for the article title
4201 * @todo this should get integrated somewhere sane
4203 public function convertHtml( $text, $isTitle = false ) {
4204 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4208 * @param string $key
4211 public function convertCategoryKey( $key ) {
4212 return $this->mConverter
->convertCategoryKey( $key );
4216 * Get the list of variants supported by this language
4217 * see sample implementation in LanguageZh.php
4219 * @return array An array of language codes
4221 public function getVariants() {
4222 return $this->mConverter
->getVariants();
4228 public function getPreferredVariant() {
4229 return $this->mConverter
->getPreferredVariant();
4235 public function getDefaultVariant() {
4236 return $this->mConverter
->getDefaultVariant();
4242 public function getURLVariant() {
4243 return $this->mConverter
->getURLVariant();
4247 * If a language supports multiple variants, it is
4248 * possible that non-existing link in one variant
4249 * actually exists in another variant. this function
4250 * tries to find it. See e.g. LanguageZh.php
4251 * The input parameters may be modified upon return
4253 * @param string &$link The name of the link
4254 * @param Title &$nt The title object of the link
4255 * @param bool $ignoreOtherCond To disable other conditions when
4256 * we need to transclude a template or update a category's link
4258 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4259 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4263 * returns language specific options used by User::getPageRenderHash()
4264 * for example, the preferred language variant
4268 function getExtraHashOptions() {
4269 return $this->mConverter
->getExtraHashOptions();
4273 * For languages that support multiple variants, the title of an
4274 * article may be displayed differently in different variants. this
4275 * function returns the apporiate title defined in the body of the article.
4279 public function getParsedTitle() {
4280 return $this->mConverter
->getParsedTitle();
4284 * Prepare external link text for conversion. When the text is
4285 * a URL, it shouldn't be converted, and it'll be wrapped in
4286 * the "raw" tag (-{R| }-) to prevent conversion.
4288 * This function is called "markNoConversion" for historical
4291 * @param string $text Text to be used for external link
4292 * @param bool $noParse Wrap it without confirming it's a real URL first
4293 * @return string The tagged text
4295 public function markNoConversion( $text, $noParse = false ) {
4296 // Excluding protocal-relative URLs may avoid many false positives.
4297 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4298 return $this->mConverter
->markNoConversion( $text );
4305 * A regular expression to match legal word-trailing characters
4306 * which should be merged onto a link of the form [[foo]]bar.
4310 public function linkTrail() {
4311 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4315 * A regular expression character set to match legal word-prefixing
4316 * characters which should be merged onto a link of the form foo[[bar]].
4320 public function linkPrefixCharset() {
4321 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4325 * @deprecated since 1.24, will be removed in 1.25
4328 function getLangObj() {
4329 wfDeprecated( __METHOD__
, '1.24' );
4334 * Get the "parent" language which has a converter to convert a "compatible" language
4335 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4337 * @return Language|null
4340 public function getParentLanguage() {
4341 if ( $this->mParentLanguage
!== false ) {
4342 return $this->mParentLanguage
;
4345 $pieces = explode( '-', $this->getCode() );
4347 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4348 $this->mParentLanguage
= null;
4351 $lang = Language
::factory( $code );
4352 if ( !$lang->hasVariant( $this->getCode() ) ) {
4353 $this->mParentLanguage
= null;
4357 $this->mParentLanguage
= $lang;
4362 * Get the RFC 3066 code for this language object
4364 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4365 * htmlspecialchars() or similar
4369 public function getCode() {
4370 return $this->mCode
;
4374 * Get the code in Bcp47 format which we can use
4375 * inside of html lang="" tags.
4377 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4378 * htmlspecialchars() or similar.
4383 public function getHtmlCode() {
4384 if ( is_null( $this->mHtmlCode
) ) {
4385 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4387 return $this->mHtmlCode
;
4391 * @param string $code
4393 public function setCode( $code ) {
4394 $this->mCode
= $code;
4395 // Ensure we don't leave incorrect cached data lying around
4396 $this->mHtmlCode
= null;
4397 $this->mParentLanguage
= false;
4401 * Get the name of a file for a certain language code
4402 * @param string $prefix Prepend this to the filename
4403 * @param string $code Language code
4404 * @param string $suffix Append this to the filename
4405 * @throws MWException
4406 * @return string $prefix . $mangledCode . $suffix
4408 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4409 if ( !self
::isValidBuiltInCode( $code ) ) {
4410 throw new MWException( "Invalid language code \"$code\"" );
4413 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4417 * Get the language code from a file name. Inverse of getFileName()
4418 * @param string $filename $prefix . $languageCode . $suffix
4419 * @param string $prefix Prefix before the language code
4420 * @param string $suffix Suffix after the language code
4421 * @return string Language code, or false if $prefix or $suffix isn't found
4423 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4425 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4426 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4427 if ( !count( $m ) ) {
4430 return str_replace( '_', '-', strtolower( $m[1] ) );
4434 * @param string $code
4437 public static function getMessagesFileName( $code ) {
4439 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4440 Hooks
::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4445 * @param string $code
4449 public static function getJsonMessagesFileName( $code ) {
4452 if ( !self
::isValidBuiltInCode( $code ) ) {
4453 throw new MWException( "Invalid language code \"$code\"" );
4456 return "$IP/languages/i18n/$code.json";
4460 * @param string $code
4463 public static function getClassFileName( $code ) {
4465 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4469 * Get the first fallback for a given language.
4471 * @param string $code
4473 * @return bool|string
4475 public static function getFallbackFor( $code ) {
4476 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4479 $fallbacks = self
::getFallbacksFor( $code );
4480 return $fallbacks[0];
4485 * Get the ordered list of fallback languages.
4488 * @param string $code Language code
4489 * @return array Non-empty array, ending in "en"
4491 public static function getFallbacksFor( $code ) {
4492 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4495 // For unknown languages, fallbackSequence returns an empty array,
4496 // hardcode fallback to 'en' in that case.
4497 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: array( 'en' );
4501 * Get the ordered list of fallback languages, ending with the fallback
4502 * language chain for the site language.
4505 * @param string $code Language code
4506 * @return array Array( fallbacks, site fallbacks )
4508 public static function getFallbacksIncludingSiteLanguage( $code ) {
4509 global $wgLanguageCode;
4511 // Usually, we will only store a tiny number of fallback chains, so we
4512 // keep them in static memory.
4513 $cacheKey = "{$code}-{$wgLanguageCode}";
4515 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4516 $fallbacks = self
::getFallbacksFor( $code );
4518 // Append the site's fallback chain, including the site language itself
4519 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4520 array_unshift( $siteFallbacks, $wgLanguageCode );
4522 // Eliminate any languages already included in the chain
4523 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4525 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4527 return self
::$fallbackLanguageCache[$cacheKey];
4531 * Get all messages for a given language
4532 * WARNING: this may take a long time. If you just need all message *keys*
4533 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4535 * @param string $code
4539 public static function getMessagesFor( $code ) {
4540 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4544 * Get a message for a given language
4546 * @param string $key
4547 * @param string $code
4551 public static function getMessageFor( $key, $code ) {
4552 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4556 * Get all message keys for a given language. This is a faster alternative to
4557 * array_keys( Language::getMessagesFor( $code ) )
4560 * @param string $code Language code
4561 * @return array Array of message keys (strings)
4563 public static function getMessageKeysFor( $code ) {
4564 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4568 * @param string $talk
4571 function fixVariableInNamespace( $talk ) {
4572 if ( strpos( $talk, '$1' ) === false ) {
4576 global $wgMetaNamespace;
4577 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4579 # Allow grammar transformations
4580 # Allowing full message-style parsing would make simple requests
4581 # such as action=raw much more expensive than they need to be.
4582 # This will hopefully cover most cases.
4583 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4584 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4585 return str_replace( ' ', '_', $talk );
4592 function replaceGrammarInNamespace( $m ) {
4593 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4597 * @throws MWException
4600 static function getCaseMaps() {
4601 static $wikiUpperChars, $wikiLowerChars;
4602 if ( isset( $wikiUpperChars ) ) {
4603 return array( $wikiUpperChars, $wikiLowerChars );
4606 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4607 if ( $arr === false ) {
4608 throw new MWException(
4609 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4611 $wikiUpperChars = $arr['wikiUpperChars'];
4612 $wikiLowerChars = $arr['wikiLowerChars'];
4613 return array( $wikiUpperChars, $wikiLowerChars );
4617 * Decode an expiry (block, protection, etc) which has come from the DB
4619 * @param string $expiry Database expiry String
4620 * @param bool|int $format True to process using language functions, or TS_ constant
4621 * to return the expiry in a given timestamp
4622 * @param string $inifinity If $format is not true, use this string for infinite expiry
4626 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4628 if ( $dbInfinity === null ) {
4629 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4632 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4633 return $format === true
4634 ?
$this->getMessageFromDB( 'infiniteblock' )
4637 return $format === true
4638 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4639 : wfTimestamp( $format, $expiry );
4645 * @param int|float $seconds
4646 * @param array $format Optional
4647 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4648 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4649 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4651 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4652 * or 'avoidminutes'.
4655 function formatTimePeriod( $seconds, $format = array() ) {
4656 if ( !is_array( $format ) ) {
4657 $format = array( 'avoid' => $format ); // For backwards compatibility
4659 if ( !isset( $format['avoid'] ) ) {
4660 $format['avoid'] = false;
4662 if ( !isset( $format['noabbrevs'] ) ) {
4663 $format['noabbrevs'] = false;
4665 $secondsMsg = wfMessage(
4666 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4667 $minutesMsg = wfMessage(
4668 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4669 $hoursMsg = wfMessage(
4670 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4671 $daysMsg = wfMessage(
4672 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4674 if ( round( $seconds * 10 ) < 100 ) {
4675 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4676 $s = $secondsMsg->params( $s )->text();
4677 } elseif ( round( $seconds ) < 60 ) {
4678 $s = $this->formatNum( round( $seconds ) );
4679 $s = $secondsMsg->params( $s )->text();
4680 } elseif ( round( $seconds ) < 3600 ) {
4681 $minutes = floor( $seconds / 60 );
4682 $secondsPart = round( fmod( $seconds, 60 ) );
4683 if ( $secondsPart == 60 ) {
4687 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4689 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4690 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4691 $hours = floor( $seconds / 3600 );
4692 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4693 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4694 if ( $secondsPart == 60 ) {
4698 if ( $minutes == 60 ) {
4702 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4704 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4705 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4706 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4709 $days = floor( $seconds / 86400 );
4710 if ( $format['avoid'] === 'avoidminutes' ) {
4711 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4712 if ( $hours == 24 ) {
4716 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4718 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4719 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4720 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4721 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4722 if ( $minutes == 60 ) {
4726 if ( $hours == 24 ) {
4730 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4732 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4734 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4736 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4738 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4745 * Format a bitrate for output, using an appropriate
4746 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4747 * the magnitude in question.
4749 * This use base 1000. For base 1024 use formatSize(), for another base
4750 * see formatComputingNumbers().
4755 function formatBitrate( $bps ) {
4756 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4760 * @param int $size Size of the unit
4761 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4762 * @param string $messageKey Message key to be uesd
4765 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4767 return str_replace( '$1', $this->formatNum( $size ),
4768 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4771 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4774 $maxIndex = count( $sizes ) - 1;
4775 while ( $size >= $boundary && $index < $maxIndex ) {
4780 // For small sizes no decimal places necessary
4783 // For MB and bigger two decimal places are smarter
4786 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4788 $size = round( $size, $round );
4789 $text = $this->getMessageFromDB( $msg );
4790 return str_replace( '$1', $this->formatNum( $size ), $text );
4794 * Format a size in bytes for output, using an appropriate
4795 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4797 * This method use base 1024. For base 1000 use formatBitrate(), for
4798 * another base see formatComputingNumbers()
4800 * @param int $size Size to format
4801 * @return string Plain text (not HTML)
4803 function formatSize( $size ) {
4804 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4808 * Make a list item, used by various special pages
4810 * @param string $page Page link
4811 * @param string $details HTML safe text between brackets
4812 * @param bool $oppositedm Add the direction mark opposite to your
4813 * language, to display text properly
4814 * @return HTML escaped string
4816 function specialList( $page, $details, $oppositedm = true ) {
4821 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4825 $this->msg( 'word-separator' )->escaped() .
4826 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4830 * Generate (prev x| next x) (20|50|100...) type links for paging
4832 * @param Title $title Title object to link
4833 * @param int $offset
4835 * @param array $query Optional URL query parameter string
4836 * @param bool $atend Optional param for specified if this is the last page
4839 public function viewPrevNext( Title
$title, $offset, $limit,
4840 array $query = array(), $atend = false
4842 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4844 # Make 'previous' link
4845 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4846 if ( $offset > 0 ) {
4847 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4848 $query, $prev, 'prevn-title', 'mw-prevlink' );
4850 $plink = htmlspecialchars( $prev );
4854 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4856 $nlink = htmlspecialchars( $next );
4858 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4859 $query, $next, 'nextn-title', 'mw-nextlink' );
4862 # Make links to set number of items per page
4863 $numLinks = array();
4864 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4865 $numLinks[] = $this->numLink( $title, $offset, $num,
4866 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4869 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4870 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4874 * Helper function for viewPrevNext() that generates links
4876 * @param Title $title Title object to link
4877 * @param int $offset
4879 * @param array $query Extra query parameters
4880 * @param string $link Text to use for the link; will be escaped
4881 * @param string $tooltipMsg Name of the message to use as tooltip
4882 * @param string $class Value of the "class" attribute of the link
4883 * @return string HTML fragment
4885 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4888 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4889 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4890 ->numParams( $limit )->text();
4892 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4893 'title' => $tooltip, 'class' => $class ), $link );
4897 * Get the conversion rule title, if any.
4901 public function getConvRuleTitle() {
4902 return $this->mConverter
->getConvRuleTitle();
4906 * Get the compiled plural rules for the language
4908 * @return array Associative array with plural form, and plural rule as key-value pairs
4910 public function getCompiledPluralRules() {
4911 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4912 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4913 if ( !$pluralRules ) {
4914 foreach ( $fallbacks as $fallbackCode ) {
4915 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4916 if ( $pluralRules ) {
4921 return $pluralRules;
4925 * Get the plural rules for the language
4927 * @return array Associative array with plural form number and plural rule as key-value pairs
4929 public function getPluralRules() {
4930 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4931 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4932 if ( !$pluralRules ) {
4933 foreach ( $fallbacks as $fallbackCode ) {
4934 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4935 if ( $pluralRules ) {
4940 return $pluralRules;
4944 * Get the plural rule types for the language
4946 * @return array Associative array with plural form number and plural rule type as key-value pairs
4948 public function getPluralRuleTypes() {
4949 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4950 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4951 if ( !$pluralRuleTypes ) {
4952 foreach ( $fallbacks as $fallbackCode ) {
4953 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4954 if ( $pluralRuleTypes ) {
4959 return $pluralRuleTypes;
4963 * Find the index number of the plural rule appropriate for the given number
4964 * @param int $number
4965 * @return int The index number of the plural rule
4967 public function getPluralRuleIndexNumber( $number ) {
4968 $pluralRules = $this->getCompiledPluralRules();
4969 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4974 * Find the plural rule type appropriate for the given number
4975 * For example, if the language is set to Arabic, getPluralType(5) should
4978 * @param int $number
4979 * @return string The name of the plural rule type, e.g. one, two, few, many
4981 public function getPluralRuleType( $number ) {
4982 $index = $this->getPluralRuleIndexNumber( $number );
4983 $pluralRuleTypes = $this->getPluralRuleTypes();
4984 if ( isset( $pluralRuleTypes[$index] ) ) {
4985 return $pluralRuleTypes[$index];