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 return self
::isValidBuiltInCode( $code )
258 && ( is_readable( self
::getMessagesFileName( $code ) )
259 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
264 * Returns true if a language code string is a well-formed language tag
265 * according to RFC 5646.
266 * This function only checks well-formedness; it doesn't check that
267 * language, script or variant codes actually exist in the repositories.
269 * Based on regexes by Mark Davis of the Unicode Consortium:
270 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
272 * @param string $code
273 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
278 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
281 $alphanum = '[a-z0-9]';
282 $x = 'x'; # private use singleton
283 $singleton = '[a-wy-z]'; # other singleton
284 $s = $lenient ?
'[-_]' : '-';
286 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
287 $script = "$alpha{4}"; # ISO 15924
288 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
289 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
290 $extension = "$singleton(?:$s$alphanum{2,8})+";
291 $privateUse = "$x(?:$s$alphanum{1,8})+";
293 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
294 # Since these are limited, this is safe even later changes to the registry --
295 # the only oddity is that it might change the type of the tag, and thus
296 # the results from the capturing groups.
297 # http://www.iana.org/assignments/language-subtag-registry
299 $grandfathered = "en{$s}GB{$s}oed"
300 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
301 . "|no{$s}(?:bok|nyn)"
302 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
303 . "|zh{$s}min{$s}nan";
305 $variantList = "$variant(?:$s$variant)*";
306 $extensionList = "$extension(?:$s$extension)*";
308 $langtag = "(?:($language)"
311 . "(?:$s$variantList)?"
312 . "(?:$s$extensionList)?"
313 . "(?:$s$privateUse)?)";
315 # The final breakdown, with capturing groups for each of these components
316 # The variants, extensions, grandfathered, and private-use may have interior '-'
318 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
320 return (bool)preg_match( "/$root/", strtolower( $code ) );
324 * Returns true if a language code string is of a valid form, whether or
325 * not it exists. This includes codes which are used solely for
326 * customisation via the MediaWiki namespace.
328 * @param string $code
332 public static function isValidCode( $code ) {
333 static $cache = array();
334 if ( isset( $cache[$code] ) ) {
335 return $cache[$code];
337 // People think language codes are html safe, so enforce it.
338 // Ideally we should only allow a-zA-Z0-9-
339 // but, .+ and other chars are often used for {{int:}} hacks
340 // see bugs 37564, 37587, 36938
342 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
343 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
345 return $cache[$code];
349 * Returns true if a language code is of a valid form for the purposes of
350 * internal customisation of MediaWiki, via Messages*.php or *.json.
352 * @param string $code
354 * @throws MWException
358 public static function isValidBuiltInCode( $code ) {
360 if ( !is_string( $code ) ) {
361 if ( is_object( $code ) ) {
362 $addmsg = " of class " . get_class( $code );
366 $type = gettype( $code );
367 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
370 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
374 * Returns true if a language code is an IETF tag known to MediaWiki.
381 public static function isKnownLanguageTag( $tag ) {
382 static $coreLanguageNames;
384 // Quick escape for invalid input to avoid exceptions down the line
385 // when code tries to process tags which are not valid at all.
386 if ( !self
::isValidBuiltInCode( $tag ) ) {
390 if ( $coreLanguageNames === null ) {
392 include "$IP/languages/Names.php";
395 if ( isset( $coreLanguageNames[$tag] )
396 || self
::fetchLanguageName( $tag, $tag ) !== ''
405 * @param string $code
406 * @return string Name of the language class
408 public static function classFromCode( $code ) {
409 if ( $code == 'en' ) {
412 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
417 * Includes language class files
419 * @param string $class Name of the language class
421 public static function preloadLanguageClass( $class ) {
424 if ( $class === 'Language' ) {
428 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
429 include_once "$IP/languages/classes/$class.php";
434 * Get the LocalisationCache instance
436 * @return LocalisationCache
438 public static function getLocalisationCache() {
439 if ( is_null( self
::$dataCache ) ) {
440 global $wgLocalisationCacheConf;
441 $class = $wgLocalisationCacheConf['class'];
442 self
::$dataCache = new $class( $wgLocalisationCacheConf );
444 return self
::$dataCache;
447 function __construct() {
448 $this->mConverter
= new FakeConverter( $this );
449 // Set the code to the name of the descendant
450 if ( get_class( $this ) == 'Language' ) {
453 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
455 self
::getLocalisationCache();
459 * Reduce memory usage
461 function __destruct() {
462 foreach ( $this as $name => $value ) {
463 unset( $this->$name );
468 * Hook which will be called if this is the content language.
469 * Descendants can use this to register hook functions or modify globals
471 function initContLang() {
478 function getFallbackLanguages() {
479 return self
::getFallbacksFor( $this->mCode
);
483 * Exports $wgBookstoreListEn
486 function getBookstoreList() {
487 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
491 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
492 * available in localised form, it will be included in English.
496 public function getNamespaces() {
497 if ( is_null( $this->namespaceNames
) ) {
498 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
500 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
501 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
503 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
505 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
506 if ( $wgMetaNamespaceTalk ) {
507 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
509 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
510 $this->namespaceNames
[NS_PROJECT_TALK
] =
511 $this->fixVariableInNamespace( $talk );
514 # Sometimes a language will be localised but not actually exist on this wiki.
515 foreach ( $this->namespaceNames
as $key => $text ) {
516 if ( !isset( $validNamespaces[$key] ) ) {
517 unset( $this->namespaceNames
[$key] );
521 # The above mixing may leave namespaces out of canonical order.
522 # Re-order by namespace ID number...
523 ksort( $this->namespaceNames
);
525 Hooks
::run( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
528 return $this->namespaceNames
;
532 * Arbitrarily set all of the namespace names at once. Mainly used for testing
533 * @param array $namespaces Array of namespaces (id => name)
535 public function setNamespaces( array $namespaces ) {
536 $this->namespaceNames
= $namespaces;
537 $this->mNamespaceIds
= null;
541 * Resets all of the namespace caches. Mainly used for testing
543 public function resetNamespaces() {
544 $this->namespaceNames
= null;
545 $this->mNamespaceIds
= null;
546 $this->namespaceAliases
= null;
550 * A convenience function that returns getNamespaces() with spaces instead of underscores
551 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
555 function getFormattedNamespaces() {
556 $ns = $this->getNamespaces();
557 foreach ( $ns as $k => $v ) {
558 $ns[$k] = strtr( $v, '_', ' ' );
564 * Get a namespace value by key
567 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
568 * echo $mw_ns; // prints 'MediaWiki'
571 * @param int $index The array key of the namespace to return
572 * @return string|bool String if the namespace value exists, otherwise false
574 function getNsText( $index ) {
575 $ns = $this->getNamespaces();
576 return isset( $ns[$index] ) ?
$ns[$index] : false;
580 * A convenience function that returns the same thing as
581 * getNsText() except with '_' changed to ' ', useful for
585 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
586 * echo $mw_ns; // prints 'MediaWiki talk'
589 * @param int $index The array key of the namespace to return
590 * @return string Namespace name without underscores (empty string if namespace does not exist)
592 function getFormattedNsText( $index ) {
593 $ns = $this->getNsText( $index );
594 return strtr( $ns, '_', ' ' );
598 * Returns gender-dependent namespace alias if available.
599 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
600 * @param int $index Namespace index
601 * @param string $gender Gender key (male, female... )
605 function getGenderNsText( $index, $gender ) {
606 global $wgExtraGenderNamespaces;
608 $ns = $wgExtraGenderNamespaces +
609 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
611 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
615 * Whether this language uses gender-dependent namespace aliases.
616 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
620 function needsGenderDistinction() {
621 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
622 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
623 // $wgExtraGenderNamespaces overrides everything
625 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
626 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
627 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
630 // Check what is in i18n files
631 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
632 return count( $aliases ) > 0;
637 * Get a namespace key by value, case insensitive.
638 * Only matches namespace names for the current language, not the
639 * canonical ones defined in Namespace.php.
641 * @param string $text
642 * @return int|bool An integer if $text is a valid value otherwise false
644 function getLocalNsIndex( $text ) {
645 $lctext = $this->lc( $text );
646 $ids = $this->getNamespaceIds();
647 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
653 function getNamespaceAliases() {
654 if ( is_null( $this->namespaceAliases
) ) {
655 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
659 foreach ( $aliases as $name => $index ) {
660 if ( $index === NS_PROJECT_TALK
) {
661 unset( $aliases[$name] );
662 $name = $this->fixVariableInNamespace( $name );
663 $aliases[$name] = $index;
668 global $wgExtraGenderNamespaces;
669 $genders = $wgExtraGenderNamespaces +
670 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
671 foreach ( $genders as $index => $forms ) {
672 foreach ( $forms as $alias ) {
673 $aliases[$alias] = $index;
677 # Also add converted namespace names as aliases, to avoid confusion.
678 $convertedNames = array();
679 foreach ( $this->getVariants() as $variant ) {
680 if ( $variant === $this->mCode
) {
683 foreach ( $this->getNamespaces() as $ns => $_ ) {
684 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
688 $this->namespaceAliases
= $aliases +
$convertedNames;
691 return $this->namespaceAliases
;
697 function getNamespaceIds() {
698 if ( is_null( $this->mNamespaceIds
) ) {
699 global $wgNamespaceAliases;
700 # Put namespace names and aliases into a hashtable.
701 # If this is too slow, then we should arrange it so that it is done
702 # before caching. The catch is that at pre-cache time, the above
703 # class-specific fixup hasn't been done.
704 $this->mNamespaceIds
= array();
705 foreach ( $this->getNamespaces() as $index => $name ) {
706 $this->mNamespaceIds
[$this->lc( $name )] = $index;
708 foreach ( $this->getNamespaceAliases() as $name => $index ) {
709 $this->mNamespaceIds
[$this->lc( $name )] = $index;
711 if ( $wgNamespaceAliases ) {
712 foreach ( $wgNamespaceAliases as $name => $index ) {
713 $this->mNamespaceIds
[$this->lc( $name )] = $index;
717 return $this->mNamespaceIds
;
721 * Get a namespace key by value, case insensitive. Canonical namespace
722 * names override custom ones defined for the current language.
724 * @param string $text
725 * @return int|bool An integer if $text is a valid value otherwise false
727 function getNsIndex( $text ) {
728 $lctext = $this->lc( $text );
729 $ns = MWNamespace
::getCanonicalIndex( $lctext );
730 if ( $ns !== null ) {
733 $ids = $this->getNamespaceIds();
734 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
738 * short names for language variants used for language conversion links.
740 * @param string $code
741 * @param bool $usemsg Use the "variantname-xyz" message if it exists
744 function getVariantname( $code, $usemsg = true ) {
745 $msg = "variantname-$code";
746 if ( $usemsg && wfMessage( $msg )->exists() ) {
747 return $this->getMessageFromDB( $msg );
749 $name = self
::fetchLanguageName( $code );
751 return $name; # if it's defined as a language name, show that
753 # otherwise, output the language code
759 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
760 * SpecialPageFactory::getLocalNameFor instead.
761 * @param string $name
764 function specialPage( $name ) {
765 $aliases = $this->getSpecialPageAliases();
766 if ( isset( $aliases[$name][0] ) ) {
767 $name = $aliases[$name][0];
769 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
775 function getDatePreferences() {
776 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
782 function getDateFormats() {
783 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
787 * @return array|string
789 function getDefaultDateFormat() {
790 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
791 if ( $df === 'dmy or mdy' ) {
792 global $wgAmericanDates;
793 return $wgAmericanDates ?
'mdy' : 'dmy';
802 function getDatePreferenceMigrationMap() {
803 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
807 * @param string $image
810 function getImageFile( $image ) {
811 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
818 function getImageFiles() {
819 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
825 function getExtraUserToggles() {
826 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
833 function getUserToggle( $tog ) {
834 return $this->getMessageFromDB( "tog-$tog" );
838 * Get native language names, indexed by code.
839 * Only those defined in MediaWiki, no other data like CLDR.
840 * If $customisedOnly is true, only returns codes with a messages file
842 * @param bool $customisedOnly
845 * @deprecated since 1.20, use fetchLanguageNames()
847 public static function getLanguageNames( $customisedOnly = false ) {
848 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
852 * Get translated language names. This is done on best effort and
853 * by default this is exactly the same as Language::getLanguageNames.
854 * The CLDR extension provides translated names.
855 * @param string $code Language code.
856 * @return array Language code => language name
858 * @deprecated since 1.20, use fetchLanguageNames()
860 public static function getTranslatedLanguageNames( $code ) {
861 return self
::fetchLanguageNames( $code, 'all' );
865 * Get an array of language names, indexed by code.
866 * @param null|string $inLanguage Code of language in which to return the names
867 * Use null for autonyms (native names)
868 * @param string $include One of:
869 * 'all' all available languages
870 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
871 * 'mwfile' only if the language is in 'mw' *and* has a message file
872 * @return array Language code => language name
875 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
876 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
877 $cacheKey .= ":$include";
878 if ( self
::$languageNameCache === null ) {
879 self
::$languageNameCache = new MapCacheLRU( 20 );
881 if ( self
::$languageNameCache->has( $cacheKey ) ) {
882 $ret = self
::$languageNameCache->get( $cacheKey );
884 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
885 self
::$languageNameCache->set( $cacheKey, $ret );
891 * Uncached helper for fetchLanguageNames
892 * @param null|string $inLanguage Code of language in which to return the names
893 * Use null for autonyms (native names)
894 * @param string $include One of:
895 * 'all' all available languages
896 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
897 * 'mwfile' only if the language is in 'mw' *and* has a message file
898 * @return array Language code => language name
900 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
901 global $wgExtraLanguageNames;
902 static $coreLanguageNames;
904 if ( $coreLanguageNames === null ) {
906 include "$IP/languages/Names.php";
909 // If passed an invalid language code to use, fallback to en
910 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
917 # TODO: also include when $inLanguage is null, when this code is more efficient
918 Hooks
::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
921 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
922 foreach ( $mwNames as $mwCode => $mwName ) {
923 # - Prefer own MediaWiki native name when not using the hook
924 # - For other names just add if not added through the hook
925 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
926 $names[$mwCode] = $mwName;
930 if ( $include === 'all' ) {
936 $coreCodes = array_keys( $mwNames );
937 foreach ( $coreCodes as $coreCode ) {
938 $returnMw[$coreCode] = $names[$coreCode];
941 if ( $include === 'mwfile' ) {
942 $namesMwFile = array();
943 # We do this using a foreach over the codes instead of a directory
944 # loop so that messages files in extensions will work correctly.
945 foreach ( $returnMw as $code => $value ) {
946 if ( is_readable( self
::getMessagesFileName( $code ) )
947 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
949 $namesMwFile[$code] = $names[$code];
953 ksort( $namesMwFile );
958 # 'mw' option; default if it's not one of the other two options (all/mwfile)
963 * @param string $code The code of the language for which to get the name
964 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
965 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
966 * @return string Language name or empty
969 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
970 $code = strtolower( $code );
971 $array = self
::fetchLanguageNames( $inLanguage, $include );
972 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
976 * Get a message from the MediaWiki namespace.
978 * @param string $msg Message name
981 function getMessageFromDB( $msg ) {
982 return $this->msg( $msg )->text();
986 * Get message object in this language. Only for use inside this class.
988 * @param string $msg Message name
991 protected function msg( $msg ) {
992 return wfMessage( $msg )->inLanguage( $this );
996 * Get the native language name of $code.
997 * Only if defined in MediaWiki, no other data like CLDR.
998 * @param string $code
1000 * @deprecated since 1.20, use fetchLanguageName()
1002 function getLanguageName( $code ) {
1003 return self
::fetchLanguageName( $code );
1007 * @param string $key
1010 function getMonthName( $key ) {
1011 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
1017 function getMonthNamesArray() {
1018 $monthNames = array( '' );
1019 for ( $i = 1; $i < 13; $i++
) {
1020 $monthNames[] = $this->getMonthName( $i );
1026 * @param string $key
1029 function getMonthNameGen( $key ) {
1030 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1034 * @param string $key
1037 function getMonthAbbreviation( $key ) {
1038 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1044 function getMonthAbbreviationsArray() {
1045 $monthNames = array( '' );
1046 for ( $i = 1; $i < 13; $i++
) {
1047 $monthNames[] = $this->getMonthAbbreviation( $i );
1053 * @param string $key
1056 function getWeekdayName( $key ) {
1057 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1061 * @param string $key
1064 function getWeekdayAbbreviation( $key ) {
1065 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1069 * @param string $key
1072 function getIranianCalendarMonthName( $key ) {
1073 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1077 * @param string $key
1080 function getHebrewCalendarMonthName( $key ) {
1081 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1085 * @param string $key
1088 function getHebrewCalendarMonthNameGen( $key ) {
1089 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1093 * @param string $key
1096 function getHijriCalendarMonthName( $key ) {
1097 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1101 * Pass through result from $dateTimeObj->format()
1102 * @param DateTime|bool|null &$dateTimeObj
1104 * @param DateTimeZone|bool|null $zone
1105 * @param string $code
1108 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1109 if ( !$dateTimeObj ) {
1110 $dateTimeObj = DateTime
::createFromFormat(
1111 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1114 return $dateTimeObj->format( $code );
1118 * This is a workalike of PHP's date() function, but with better
1119 * internationalisation, a reduced set of format characters, and a better
1122 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1123 * the PHP manual for definitions. There are a number of extensions, which
1126 * xn Do not translate digits of the next numeric format character
1127 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1128 * xr Use roman numerals for the next numeric format character
1129 * xh Use hebrew numerals for the next numeric format character
1131 * xg Genitive month name
1133 * xij j (day number) in Iranian calendar
1134 * xiF F (month name) in Iranian calendar
1135 * xin n (month number) in Iranian calendar
1136 * xiy y (two digit year) in Iranian calendar
1137 * xiY Y (full year) in Iranian calendar
1139 * xjj j (day number) in Hebrew calendar
1140 * xjF F (month name) in Hebrew calendar
1141 * xjt t (days in month) in Hebrew calendar
1142 * xjx xg (genitive month name) in Hebrew calendar
1143 * xjn n (month number) in Hebrew calendar
1144 * xjY Y (full year) in Hebrew calendar
1146 * xmj j (day number) in Hijri calendar
1147 * xmF F (month name) in Hijri calendar
1148 * xmn n (month number) in Hijri calendar
1149 * xmY Y (full year) in Hijri calendar
1151 * xkY Y (full year) in Thai solar calendar. Months and days are
1152 * identical to the Gregorian calendar
1153 * xoY Y (full year) in Minguo calendar or Juche year.
1154 * Months and days are identical to the
1155 * Gregorian calendar
1156 * xtY Y (full year) in Japanese nengo. Months and days are
1157 * identical to the Gregorian calendar
1159 * Characters enclosed in double quotes will be considered literal (with
1160 * the quotes themselves removed). Unmatched quotes will be considered
1161 * literal quotes. Example:
1163 * "The month is" F => The month is January
1166 * Backslash escaping is also supported.
1168 * Input timestamp is assumed to be pre-normalized to the desired local
1169 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1170 * $ts is UTC if $zone is not given.
1172 * @param string $format
1173 * @param string $ts 14-character timestamp
1176 * @param DateTimeZone $zone Timezone of $ts
1177 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1178 * Only makes sense if $ts is the current time.
1179 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1181 * @throws MWException
1184 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1189 $dateTimeObj = false;
1198 $usedSecond = false;
1199 $usedMinute = false;
1206 $usedISOYear = false;
1207 $usedIsLeapYear = false;
1209 $usedHebrewMonth = false;
1210 $usedIranianMonth = false;
1211 $usedHijriMonth = false;
1212 $usedHebrewYear = false;
1213 $usedIranianYear = false;
1214 $usedHijriYear = false;
1215 $usedTennoYear = false;
1217 if ( strlen( $ts ) !== 14 ) {
1218 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1221 if ( !ctype_digit( $ts ) ) {
1222 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1225 $formatLength = strlen( $format );
1226 for ( $p = 0; $p < $formatLength; $p++
) {
1228 $code = $format[$p];
1229 if ( $code == 'x' && $p < $formatLength - 1 ) {
1230 $code .= $format[++
$p];
1233 if ( ( $code === 'xi'
1239 && $p < $formatLength - 1 ) {
1240 $code .= $format[++
$p];
1251 $rawToggle = !$rawToggle;
1261 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1264 $usedHebrewMonth = true;
1266 $hebrew = self
::tsToHebrew( $ts );
1268 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1272 $num = substr( $ts, 6, 2 );
1276 $s .= $this->getWeekdayAbbreviation(
1277 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1282 $num = intval( substr( $ts, 6, 2 ) );
1287 $iranian = self
::tsToIranian( $ts );
1294 $hijri = self
::tsToHijri( $ts );
1301 $hebrew = self
::tsToHebrew( $ts );
1307 $s .= $this->getWeekdayName(
1308 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1313 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1316 $usedIranianMonth = true;
1318 $iranian = self
::tsToIranian( $ts );
1320 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1323 $usedHijriMonth = true;
1325 $hijri = self
::tsToHijri( $ts );
1327 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1330 $usedHebrewMonth = true;
1332 $hebrew = self
::tsToHebrew( $ts );
1334 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1338 $num = substr( $ts, 4, 2 );
1342 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1346 $num = intval( substr( $ts, 4, 2 ) );
1349 $usedIranianMonth = true;
1351 $iranian = self
::tsToIranian( $ts );
1356 $usedHijriMonth = true;
1358 $hijri = self
::tsToHijri( $ts );
1363 $usedHebrewMonth = true;
1365 $hebrew = self
::tsToHebrew( $ts );
1370 $usedHebrewMonth = true;
1372 $hebrew = self
::tsToHebrew( $ts );
1378 $num = substr( $ts, 0, 4 );
1381 $usedIranianYear = true;
1383 $iranian = self
::tsToIranian( $ts );
1388 $usedHijriYear = true;
1390 $hijri = self
::tsToHijri( $ts );
1395 $usedHebrewYear = true;
1397 $hebrew = self
::tsToHebrew( $ts );
1404 $thai = self
::tsToYear( $ts, 'thai' );
1411 $minguo = self
::tsToYear( $ts, 'minguo' );
1416 $usedTennoYear = true;
1418 $tenno = self
::tsToYear( $ts, 'tenno' );
1424 $num = substr( $ts, 2, 2 );
1427 $usedIranianYear = true;
1429 $iranian = self
::tsToIranian( $ts );
1431 $num = substr( $iranian[0], -2 );
1435 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1439 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1443 $h = substr( $ts, 8, 2 );
1444 $num = $h %
12 ?
$h %
12 : 12;
1448 $num = intval( substr( $ts, 8, 2 ) );
1452 $h = substr( $ts, 8, 2 );
1453 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1457 $num = substr( $ts, 8, 2 );
1461 $num = substr( $ts, 10, 2 );
1465 $num = substr( $ts, 12, 2 );
1475 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1481 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1485 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1489 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1492 $usedIsLeapYear = true;
1493 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1496 $usedISOYear = true;
1497 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1504 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1507 # Backslash escaping
1508 if ( $p < $formatLength - 1 ) {
1509 $s .= $format[++
$p];
1516 if ( $p < $formatLength - 1 ) {
1517 $endQuote = strpos( $format, '"', $p +
1 );
1518 if ( $endQuote === false ) {
1519 # No terminating quote, assume literal "
1522 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1526 # Quote at end of string, assume literal "
1533 if ( $num !== false ) {
1534 if ( $rawToggle ||
$raw ) {
1537 } elseif ( $roman ) {
1538 $s .= Language
::romanNumeral( $num );
1540 } elseif ( $hebrewNum ) {
1541 $s .= self
::hebrewNumeral( $num );
1544 $s .= $this->formatNum( $num, true );
1549 if ( $usedSecond ) {
1551 } elseif ( $usedMinute ) {
1552 $ttl = 60 - substr( $ts, 12, 2 );
1553 } elseif ( $usedHour ) {
1554 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1555 } elseif ( $usedAMPM ) {
1556 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1557 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1561 $usedIranianMonth ||
1568 // @todo Someone who understands the non-Gregorian calendars
1569 // should write proper logic for them so that they don't need purged every day.
1570 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1571 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1573 $possibleTtls = array();
1574 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1575 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1578 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1579 $timeRemainingInDay;
1580 } elseif ( $usedISOYear ) {
1581 // December 28th falls on the last ISO week of the year, every year.
1582 // The last ISO week of a year can be 52 or 53.
1583 $lastWeekOfISOYear = DateTime
::createFromFormat(
1585 substr( $ts, 0, 4 ) . '1228',
1586 $zone ?
: new DateTimeZone( 'UTC' )
1588 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1589 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1590 $timeRemainingInWeek =
1591 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1592 +
$timeRemainingInDay;
1593 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1598 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1599 substr( $ts, 6, 2 ) ) * 86400
1600 +
$timeRemainingInDay;
1601 } elseif ( $usedYear ) {
1603 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1604 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1605 +
$timeRemainingInDay;
1606 } elseif ( $usedIsLeapYear ) {
1607 $year = substr( $ts, 0, 4 );
1608 $timeRemainingInYear =
1609 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1610 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1611 +
$timeRemainingInDay;
1613 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1614 // this isn't a leap year. see when the next one starts
1615 $nextCandidate = $year - $mod +
4;
1616 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1617 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1618 $timeRemainingInYear;
1620 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1621 $timeRemainingInYear;
1624 // this is a leap year, so the next year isn't
1625 $possibleTtls[] = $timeRemainingInYear;
1629 if ( $possibleTtls ) {
1630 $ttl = min( $possibleTtls );
1637 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1638 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1641 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1642 * Gregorian dates to Iranian dates. Originally written in C, it
1643 * is released under the terms of GNU Lesser General Public
1644 * License. Conversion to PHP was performed by Niklas Laxström.
1646 * Link: http://www.farsiweb.info/jalali/jalali.c
1652 private static function tsToIranian( $ts ) {
1653 $gy = substr( $ts, 0, 4 ) -1600;
1654 $gm = substr( $ts, 4, 2 ) -1;
1655 $gd = substr( $ts, 6, 2 ) -1;
1657 # Days passed from the beginning (including leap years)
1659 +
floor( ( $gy +
3 ) / 4 )
1660 - floor( ( $gy +
99 ) / 100 )
1661 +
floor( ( $gy +
399 ) / 400 );
1663 // Add days of the past months of this year
1664 for ( $i = 0; $i < $gm; $i++
) {
1665 $gDayNo +
= self
::$GREG_DAYS[$i];
1669 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1673 // Days passed in current month
1674 $gDayNo +
= (int)$gd;
1676 $jDayNo = $gDayNo - 79;
1678 $jNp = floor( $jDayNo / 12053 );
1681 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1684 if ( $jDayNo >= 366 ) {
1685 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1686 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1689 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1690 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1696 return array( $jy, $jm, $jd );
1700 * Converting Gregorian dates to Hijri dates.
1702 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1704 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1710 private static function tsToHijri( $ts ) {
1711 $year = substr( $ts, 0, 4 );
1712 $month = substr( $ts, 4, 2 );
1713 $day = substr( $ts, 6, 2 );
1721 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1722 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1724 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1725 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1726 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1729 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1730 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1733 $zl = $zjd -1948440 +
10632;
1734 $zn = (int)( ( $zl - 1 ) / 10631 );
1735 $zl = $zl - 10631 * $zn +
354;
1736 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1737 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1738 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1739 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1740 $zm = (int)( ( 24 * $zl ) / 709 );
1741 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1742 $zy = 30 * $zn +
$zj - 30;
1744 return array( $zy, $zm, $zd );
1748 * Converting Gregorian dates to Hebrew dates.
1750 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1751 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1752 * to translate the relevant functions into PHP and release them under
1755 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1756 * and Adar II is 14. In a non-leap year, Adar is 6.
1762 private static function tsToHebrew( $ts ) {
1764 $year = substr( $ts, 0, 4 );
1765 $month = substr( $ts, 4, 2 );
1766 $day = substr( $ts, 6, 2 );
1768 # Calculate Hebrew year
1769 $hebrewYear = $year +
3760;
1771 # Month number when September = 1, August = 12
1773 if ( $month > 12 ) {
1780 # Calculate day of year from 1 September
1782 for ( $i = 1; $i < $month; $i++
) {
1786 # Check if the year is leap
1787 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1790 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1797 # Calculate the start of the Hebrew year
1798 $start = self
::hebrewYearStart( $hebrewYear );
1800 # Calculate next year's start
1801 if ( $dayOfYear <= $start ) {
1802 # Day is before the start of the year - it is the previous year
1804 $nextStart = $start;
1808 # Add days since previous year's 1 September
1810 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1814 # Start of the new (previous) year
1815 $start = self
::hebrewYearStart( $hebrewYear );
1818 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1821 # Calculate Hebrew day of year
1822 $hebrewDayOfYear = $dayOfYear - $start;
1824 # Difference between year's days
1825 $diff = $nextStart - $start;
1826 # Add 12 (or 13 for leap years) days to ignore the difference between
1827 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1828 # difference is only about the year type
1829 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1835 # Check the year pattern, and is leap year
1836 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1837 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1838 # and non-leap years
1839 $yearPattern = $diff %
30;
1840 # Check if leap year
1841 $isLeap = $diff >= 30;
1843 # Calculate day in the month from number of day in the Hebrew year
1844 # Don't check Adar - if the day is not in Adar, we will stop before;
1845 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1846 $hebrewDay = $hebrewDayOfYear;
1849 while ( $hebrewMonth <= 12 ) {
1850 # Calculate days in this month
1851 if ( $isLeap && $hebrewMonth == 6 ) {
1852 # Adar in a leap year
1854 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1856 if ( $hebrewDay <= $days ) {
1860 # Subtract the days of Adar I
1861 $hebrewDay -= $days;
1864 if ( $hebrewDay <= $days ) {
1870 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1871 # Cheshvan in a complete year (otherwise as the rule below)
1873 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1874 # Kislev in an incomplete year (otherwise as the rule below)
1877 # Odd months have 30 days, even have 29
1878 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1880 if ( $hebrewDay <= $days ) {
1881 # In the current month
1884 # Subtract the days of the current month
1885 $hebrewDay -= $days;
1886 # Try in the next month
1891 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1895 * This calculates the Hebrew year start, as days since 1 September.
1896 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1897 * Used for Hebrew date.
1903 private static function hebrewYearStart( $year ) {
1904 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1905 $b = intval( ( $year - 1 ) %
4 );
1906 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1910 $Mar = intval( $m );
1916 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1917 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1919 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1921 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1925 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1930 * Algorithm to convert Gregorian dates to Thai solar dates,
1931 * Minguo dates or Minguo dates.
1933 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1934 * http://en.wikipedia.org/wiki/Minguo_calendar
1935 * http://en.wikipedia.org/wiki/Japanese_era_name
1937 * @param string $ts 14-character timestamp
1938 * @param string $cName Calender name
1939 * @return array Converted year, month, day
1941 private static function tsToYear( $ts, $cName ) {
1942 $gy = substr( $ts, 0, 4 );
1943 $gm = substr( $ts, 4, 2 );
1944 $gd = substr( $ts, 6, 2 );
1946 if ( !strcmp( $cName, 'thai' ) ) {
1948 # Add 543 years to the Gregorian calendar
1949 # Months and days are identical
1950 $gy_offset = $gy +
543;
1951 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1953 # Deduct 1911 years from the Gregorian calendar
1954 # Months and days are identical
1955 $gy_offset = $gy - 1911;
1956 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1957 # Nengō dates up to Meiji period
1958 # Deduct years from the Gregorian calendar
1959 # depending on the nengo periods
1960 # Months and days are identical
1962 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1963 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1966 $gy_gannen = $gy - 1868 +
1;
1967 $gy_offset = $gy_gannen;
1968 if ( $gy_gannen == 1 ) {
1971 $gy_offset = '明治' . $gy_offset;
1973 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1974 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1975 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1976 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1977 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1980 $gy_gannen = $gy - 1912 +
1;
1981 $gy_offset = $gy_gannen;
1982 if ( $gy_gannen == 1 ) {
1985 $gy_offset = '大正' . $gy_offset;
1987 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1988 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1989 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1992 $gy_gannen = $gy - 1926 +
1;
1993 $gy_offset = $gy_gannen;
1994 if ( $gy_gannen == 1 ) {
1997 $gy_offset = '昭和' . $gy_offset;
2000 $gy_gannen = $gy - 1989 +
1;
2001 $gy_offset = $gy_gannen;
2002 if ( $gy_gannen == 1 ) {
2005 $gy_offset = '平成' . $gy_offset;
2011 return array( $gy_offset, $gm, $gd );
2015 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2017 * This is the rule the BIDI algorithm uses to determine the directionality of
2018 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
2019 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2021 * TODO: Does not handle BIDI control characters inside the text.
2022 * TODO: Does not handle unallocated characters.
2024 * @param string $text Text to test
2025 * @return null|string Directionality ('ltr' or 'rtl') or null
2027 private static function strongDirFromContent( $text = '' ) {
2028 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2031 if ( $matches[1] === '' ) {
2038 * Roman number formatting up to 10000
2044 static function romanNumeral( $num ) {
2045 static $table = array(
2046 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
2047 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
2048 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
2049 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2050 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
2053 $num = intval( $num );
2054 if ( $num > 10000 ||
$num <= 0 ) {
2059 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2060 if ( $num >= $pow10 ) {
2061 $s .= $table[$i][(int)floor( $num / $pow10 )];
2063 $num = $num %
$pow10;
2069 * Hebrew Gematria number formatting up to 9999
2075 static function hebrewNumeral( $num ) {
2076 static $table = array(
2077 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2078 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2088 array( 'ת', 'ת', 'ק' ),
2089 array( 'ת', 'ת', 'ר' ),
2091 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2094 $num = intval( $num );
2095 if ( $num > 9999 ||
$num <= 0 ) {
2099 // Round thousands have special notations
2100 if ( $num === 1000 ) {
2102 } elseif ( $num %
1000 === 0 ) {
2103 return $table[0][$num / 1000] . "' אלפים";
2108 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2109 if ( $num >= $pow10 ) {
2110 if ( $num === 15 ||
$num === 16 ) {
2111 $letters[] = $table[0][9];
2112 $letters[] = $table[0][$num - 9];
2115 $letters = array_merge(
2117 (array)$table[$i][intval( $num / $pow10 )]
2120 if ( $pow10 === 1000 ) {
2126 $num = $num %
$pow10;
2129 $preTransformLength = count( $letters );
2130 if ( $preTransformLength === 1 ) {
2131 // Add geresh (single quote) to one-letter numbers
2134 $lastIndex = $preTransformLength - 1;
2135 $letters[$lastIndex] = str_replace(
2136 array( 'כ', 'מ', 'נ', 'פ', 'צ' ),
2137 array( 'ך', 'ם', 'ן', 'ף', 'ץ' ),
2138 $letters[$lastIndex]
2141 // Add gershayim (double quote) to multiple-letter numbers,
2142 // but exclude numbers with only one letter after the thousands
2143 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2144 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2147 array_splice( $letters, -1, 0, '"' );
2151 return implode( $letters );
2155 * Used by date() and time() to adjust the time output.
2157 * @param string $ts The time in date('YmdHis') format
2158 * @param mixed $tz Adjust the time by this amount (default false, mean we
2159 * get user timecorrection setting)
2162 function userAdjust( $ts, $tz = false ) {
2163 global $wgUser, $wgLocalTZoffset;
2165 if ( $tz === false ) {
2166 $tz = $wgUser->getOption( 'timecorrection' );
2169 $data = explode( '|', $tz, 3 );
2171 if ( $data[0] == 'ZoneInfo' ) {
2172 MediaWiki\
suppressWarnings();
2173 $userTZ = timezone_open( $data[2] );
2174 MediaWiki\restoreWarnings
();
2175 if ( $userTZ !== false ) {
2176 $date = date_create( $ts, timezone_open( 'UTC' ) );
2177 date_timezone_set( $date, $userTZ );
2178 $date = date_format( $date, 'YmdHis' );
2181 # Unrecognized timezone, default to 'Offset' with the stored offset.
2182 $data[0] = 'Offset';
2185 if ( $data[0] == 'System' ||
$tz == '' ) {
2186 # Global offset in minutes.
2187 $minDiff = $wgLocalTZoffset;
2188 } elseif ( $data[0] == 'Offset' ) {
2189 $minDiff = intval( $data[1] );
2191 $data = explode( ':', $tz );
2192 if ( count( $data ) == 2 ) {
2193 $data[0] = intval( $data[0] );
2194 $data[1] = intval( $data[1] );
2195 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2196 if ( $data[0] < 0 ) {
2197 $minDiff = -$minDiff;
2200 $minDiff = intval( $data[0] ) * 60;
2204 # No difference ? Return time unchanged
2205 if ( 0 == $minDiff ) {
2209 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2210 # Generate an adjusted date; take advantage of the fact that mktime
2211 # will normalize out-of-range values so we don't have to split $minDiff
2212 # into hours and minutes.
2214 (int)substr( $ts, 8, 2 ) ), # Hours
2215 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2216 (int)substr( $ts, 12, 2 ), # Seconds
2217 (int)substr( $ts, 4, 2 ), # Month
2218 (int)substr( $ts, 6, 2 ), # Day
2219 (int)substr( $ts, 0, 4 ) ); # Year
2221 $date = date( 'YmdHis', $t );
2222 MediaWiki\restoreWarnings
();
2228 * This is meant to be used by time(), date(), and timeanddate() to get
2229 * the date preference they're supposed to use, it should be used in
2233 * function timeanddate([...], $format = true) {
2234 * $datePreference = $this->dateFormat($format);
2239 * @param int|string|bool $usePrefs If true, the user's preference is used
2240 * if false, the site/language default is used
2241 * if int/string, assumed to be a format.
2244 function dateFormat( $usePrefs = true ) {
2247 if ( is_bool( $usePrefs ) ) {
2249 $datePreference = $wgUser->getDatePreference();
2251 $datePreference = (string)User
::getDefaultOption( 'date' );
2254 $datePreference = (string)$usePrefs;
2258 if ( $datePreference == '' ) {
2262 return $datePreference;
2266 * Get a format string for a given type and preference
2267 * @param string $type May be date, time or both
2268 * @param string $pref The format name as it appears in Messages*.php
2270 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2274 function getDateFormatString( $type, $pref ) {
2275 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2276 if ( $pref == 'default' ) {
2277 $pref = $this->getDefaultDateFormat();
2278 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2280 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2282 if ( $type === 'pretty' && $df === null ) {
2283 $df = $this->getDateFormatString( 'date', $pref );
2286 if ( $df === null ) {
2287 $pref = $this->getDefaultDateFormat();
2288 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2291 $this->dateFormatStrings
[$type][$pref] = $df;
2293 return $this->dateFormatStrings
[$type][$pref];
2297 * @param string $ts The time format which needs to be turned into a
2298 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2299 * @param bool $adj Whether to adjust the time output according to the
2300 * user configured offset ($timecorrection)
2301 * @param mixed $format True to use user's date format preference
2302 * @param string|bool $timecorrection The time offset as returned by
2303 * validateTimeZone() in Special:Preferences
2306 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2307 $ts = wfTimestamp( TS_MW
, $ts );
2309 $ts = $this->userAdjust( $ts, $timecorrection );
2311 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2312 return $this->sprintfDate( $df, $ts );
2316 * @param string $ts The time format which needs to be turned into a
2317 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2318 * @param bool $adj Whether to adjust the time output according to the
2319 * user configured offset ($timecorrection)
2320 * @param mixed $format True to use user's date format preference
2321 * @param string|bool $timecorrection The time offset as returned by
2322 * validateTimeZone() in Special:Preferences
2325 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2326 $ts = wfTimestamp( TS_MW
, $ts );
2328 $ts = $this->userAdjust( $ts, $timecorrection );
2330 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2331 return $this->sprintfDate( $df, $ts );
2335 * @param string $ts The time format which needs to be turned into a
2336 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2337 * @param bool $adj Whether to adjust the time output according to the
2338 * user configured offset ($timecorrection)
2339 * @param mixed $format What format to return, if it's false output the
2340 * default one (default true)
2341 * @param string|bool $timecorrection The time offset as returned by
2342 * validateTimeZone() in Special:Preferences
2345 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2346 $ts = wfTimestamp( TS_MW
, $ts );
2348 $ts = $this->userAdjust( $ts, $timecorrection );
2350 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2351 return $this->sprintfDate( $df, $ts );
2355 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2359 * @param int $seconds The amount of seconds.
2360 * @param array $chosenIntervals The intervals to enable.
2364 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2365 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2367 $segments = array();
2369 foreach ( $intervals as $intervalName => $intervalValue ) {
2370 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2371 // duration-years, duration-decades, duration-centuries, duration-millennia
2372 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2373 $segments[] = $message->inLanguage( $this )->escaped();
2376 return $this->listToText( $segments );
2380 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2381 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2385 * @param int $seconds The amount of seconds.
2386 * @param array $chosenIntervals The intervals to enable.
2390 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2391 if ( empty( $chosenIntervals ) ) {
2392 $chosenIntervals = array(
2404 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2405 $sortedNames = array_keys( $intervals );
2406 $smallestInterval = array_pop( $sortedNames );
2408 $segments = array();
2410 foreach ( $intervals as $name => $length ) {
2411 $value = floor( $seconds / $length );
2413 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2414 $seconds -= $value * $length;
2415 $segments[$name] = $value;
2423 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2425 * @param string $type Can be 'date', 'time' or 'both'
2426 * @param string $ts The time format which needs to be turned into a
2427 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2428 * @param User $user User object used to get preferences for timezone and format
2429 * @param array $options Array, can contain the following keys:
2430 * - 'timecorrection': time correction, can have the following values:
2431 * - true: use user's preference
2432 * - false: don't use time correction
2433 * - int: value of time correction in minutes
2434 * - 'format': format to use, can have the following values:
2435 * - true: use user's preference
2436 * - false: use default preference
2437 * - string: format to use
2441 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2442 $ts = wfTimestamp( TS_MW
, $ts );
2443 $options +
= array( 'timecorrection' => true, 'format' => true );
2444 if ( $options['timecorrection'] !== false ) {
2445 if ( $options['timecorrection'] === true ) {
2446 $offset = $user->getOption( 'timecorrection' );
2448 $offset = $options['timecorrection'];
2450 $ts = $this->userAdjust( $ts, $offset );
2452 if ( $options['format'] === true ) {
2453 $format = $user->getDatePreference();
2455 $format = $options['format'];
2457 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2458 return $this->sprintfDate( $df, $ts );
2462 * Get the formatted date for the given timestamp and formatted for
2465 * @param mixed $ts Mixed: the time format which needs to be turned into a
2466 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2467 * @param User $user User object used to get preferences for timezone and format
2468 * @param array $options Array, can contain the following keys:
2469 * - 'timecorrection': time correction, can have the following values:
2470 * - true: use user's preference
2471 * - false: don't use time correction
2472 * - int: value of time correction in minutes
2473 * - 'format': format to use, can have the following values:
2474 * - true: use user's preference
2475 * - false: use default preference
2476 * - string: format to use
2480 public function userDate( $ts, User
$user, array $options = array() ) {
2481 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2485 * Get the formatted time for the given timestamp and formatted for
2488 * @param mixed $ts The time format which needs to be turned into a
2489 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2490 * @param User $user User object used to get preferences for timezone and format
2491 * @param array $options Array, can contain the following keys:
2492 * - 'timecorrection': time correction, can have the following values:
2493 * - true: use user's preference
2494 * - false: don't use time correction
2495 * - int: value of time correction in minutes
2496 * - 'format': format to use, can have the following values:
2497 * - true: use user's preference
2498 * - false: use default preference
2499 * - string: format to use
2503 public function userTime( $ts, User
$user, array $options = array() ) {
2504 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2508 * Get the formatted date and time for the given timestamp and formatted for
2511 * @param mixed $ts The time format which needs to be turned into a
2512 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2513 * @param User $user User object used to get preferences for timezone and format
2514 * @param array $options Array, can contain the following keys:
2515 * - 'timecorrection': time correction, can have the following values:
2516 * - true: use user's preference
2517 * - false: don't use time correction
2518 * - int: value of time correction in minutes
2519 * - 'format': format to use, can have the following values:
2520 * - true: use user's preference
2521 * - false: use default preference
2522 * - string: format to use
2526 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2527 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2531 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2533 * Determine the difference between the timestamp and the current time, and
2534 * generate a readable timestamp by returning "<N> <units> ago", where the
2535 * largest possible unit is used.
2537 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2539 * @param MWTimestamp $time
2540 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2541 * @param User|null $user User the timestamp is being generated for (or null to use main context's user)
2542 * @return string Formatted timestamp
2544 public function getHumanTimestamp( MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null ) {
2545 if ( $relativeTo === null ) {
2546 $relativeTo = new MWTimestamp();
2548 if ( $user === null ) {
2549 $user = RequestContext
::getMain()->getUser();
2552 // Adjust for the user's timezone.
2553 $offsetThis = $time->offsetForUser( $user );
2554 $offsetRel = $relativeTo->offsetForUser( $user );
2557 if ( Hooks
::run( 'GetHumanTimestamp', array( &$ts, $time, $relativeTo, $user, $this ) ) ) {
2558 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2561 // Reset the timezone on the objects.
2562 $time->timestamp
->sub( $offsetThis );
2563 $relativeTo->timestamp
->sub( $offsetRel );
2569 * Convert an MWTimestamp into a pretty human-readable timestamp using
2570 * the given user preferences and relative base time.
2572 * @see Language::getHumanTimestamp
2573 * @param MWTimestamp $ts Timestamp to prettify
2574 * @param MWTimestamp $relativeTo Base timestamp
2575 * @param User $user User preferences to use
2576 * @return string Human timestamp
2579 private function getHumanTimestampInternal( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2580 $diff = $ts->diff( $relativeTo );
2581 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2582 (int)$relativeTo->timestamp
->format( 'w' ) );
2583 $days = $diff->days ?
: (int)$diffDay;
2584 if ( $diff->invert ||
$days > 5
2585 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2587 // Timestamps are in different years: use full timestamp
2588 // Also do full timestamp for future dates
2590 * @todo FIXME: Add better handling of future timestamps.
2592 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2593 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2594 } elseif ( $days > 5 ) {
2595 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2596 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2597 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2598 } elseif ( $days > 1 ) {
2599 // Timestamp within the past week: show the day of the week and time
2600 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2601 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2603 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2604 $ts = wfMessage( "$weekday-at" )
2605 ->inLanguage( $this )
2606 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2608 } elseif ( $days == 1 ) {
2609 // Timestamp was yesterday: say 'yesterday' and the time.
2610 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2611 $ts = wfMessage( 'yesterday-at' )
2612 ->inLanguage( $this )
2613 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2615 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2616 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2617 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2618 $ts = wfMessage( 'today-at' )
2619 ->inLanguage( $this )
2620 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2623 // From here on in, the timestamp was soon enough ago so that we can simply say
2624 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2625 } elseif ( $diff->h
== 1 ) {
2626 // Less than 90 minutes, but more than an hour ago.
2627 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2628 } elseif ( $diff->i
>= 1 ) {
2629 // A few minutes ago.
2630 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2631 } elseif ( $diff->s
>= 30 ) {
2632 // Less than a minute, but more than 30 sec ago.
2633 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2635 // Less than 30 seconds ago.
2636 $ts = wfMessage( 'just-now' )->text();
2643 * @param string $key
2644 * @return array|null
2646 function getMessage( $key ) {
2647 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2653 function getAllMessages() {
2654 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2659 * @param string $out
2660 * @param string $string
2663 function iconv( $in, $out, $string ) {
2664 # This is a wrapper for iconv in all languages except esperanto,
2665 # which does some nasty x-conversions beforehand
2667 # Even with //IGNORE iconv can whine about illegal characters in
2668 # *input* string. We just ignore those too.
2669 # REF: http://bugs.php.net/bug.php?id=37166
2670 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2671 MediaWiki\
suppressWarnings();
2672 $text = iconv( $in, $out . '//IGNORE', $string );
2673 MediaWiki\restoreWarnings
();
2677 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2680 * @param array $matches
2681 * @return mixed|string
2683 function ucwordbreaksCallbackAscii( $matches ) {
2684 return $this->ucfirst( $matches[1] );
2688 * @param array $matches
2691 function ucwordbreaksCallbackMB( $matches ) {
2692 return mb_strtoupper( $matches[0] );
2696 * @param array $matches
2699 function ucCallback( $matches ) {
2700 list( $wikiUpperChars ) = self
::getCaseMaps();
2701 return strtr( $matches[1], $wikiUpperChars );
2705 * @param array $matches
2708 function lcCallback( $matches ) {
2709 list( , $wikiLowerChars ) = self
::getCaseMaps();
2710 return strtr( $matches[1], $wikiLowerChars );
2714 * @param array $matches
2717 function ucwordsCallbackMB( $matches ) {
2718 return mb_strtoupper( $matches[0] );
2722 * @param array $matches
2725 function ucwordsCallbackWiki( $matches ) {
2726 list( $wikiUpperChars ) = self
::getCaseMaps();
2727 return strtr( $matches[0], $wikiUpperChars );
2731 * Make a string's first character uppercase
2733 * @param string $str
2737 function ucfirst( $str ) {
2739 if ( $o < 96 ) { // if already uppercase...
2741 } elseif ( $o < 128 ) {
2742 return ucfirst( $str ); // use PHP's ucfirst()
2744 // fall back to more complex logic in case of multibyte strings
2745 return $this->uc( $str, true );
2750 * Convert a string to uppercase
2752 * @param string $str
2753 * @param bool $first
2757 function uc( $str, $first = false ) {
2758 if ( function_exists( 'mb_strtoupper' ) ) {
2760 if ( $this->isMultibyte( $str ) ) {
2761 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2763 return ucfirst( $str );
2766 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2769 if ( $this->isMultibyte( $str ) ) {
2770 $x = $first ?
'^' : '';
2771 return preg_replace_callback(
2772 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2773 array( $this, 'ucCallback' ),
2777 return $first ?
ucfirst( $str ) : strtoupper( $str );
2783 * @param string $str
2784 * @return mixed|string
2786 function lcfirst( $str ) {
2789 return strval( $str );
2790 } elseif ( $o >= 128 ) {
2791 return $this->lc( $str, true );
2792 } elseif ( $o > 96 ) {
2795 $str[0] = strtolower( $str[0] );
2801 * @param string $str
2802 * @param bool $first
2803 * @return mixed|string
2805 function lc( $str, $first = false ) {
2806 if ( function_exists( 'mb_strtolower' ) ) {
2808 if ( $this->isMultibyte( $str ) ) {
2809 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2811 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2814 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2817 if ( $this->isMultibyte( $str ) ) {
2818 $x = $first ?
'^' : '';
2819 return preg_replace_callback(
2820 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2821 array( $this, 'lcCallback' ),
2825 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2831 * @param string $str
2834 function isMultibyte( $str ) {
2835 return (bool)preg_match( '/[\x80-\xff]/', $str );
2839 * @param string $str
2840 * @return mixed|string
2842 function ucwords( $str ) {
2843 if ( $this->isMultibyte( $str ) ) {
2844 $str = $this->lc( $str );
2846 // regexp to find first letter in each word (i.e. after each space)
2847 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2849 // function to use to capitalize a single char
2850 if ( function_exists( 'mb_strtoupper' ) ) {
2851 return preg_replace_callback(
2853 array( $this, 'ucwordsCallbackMB' ),
2857 return preg_replace_callback(
2859 array( $this, 'ucwordsCallbackWiki' ),
2864 return ucwords( strtolower( $str ) );
2869 * capitalize words at word breaks
2871 * @param string $str
2874 function ucwordbreaks( $str ) {
2875 if ( $this->isMultibyte( $str ) ) {
2876 $str = $this->lc( $str );
2878 // since \b doesn't work for UTF-8, we explicitely define word break chars
2879 $breaks = "[ \-\(\)\}\{\.,\?!]";
2881 // find first letter after word break
2882 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2883 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2885 if ( function_exists( 'mb_strtoupper' ) ) {
2886 return preg_replace_callback(
2888 array( $this, 'ucwordbreaksCallbackMB' ),
2892 return preg_replace_callback(
2894 array( $this, 'ucwordsCallbackWiki' ),
2899 return preg_replace_callback(
2900 '/\b([\w\x80-\xff]+)\b/',
2901 array( $this, 'ucwordbreaksCallbackAscii' ),
2908 * Return a case-folded representation of $s
2910 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2911 * and $s2 are the same except for the case of their characters. It is not
2912 * necessary for the value returned to make sense when displayed.
2914 * Do *not* perform any other normalisation in this function. If a caller
2915 * uses this function when it should be using a more general normalisation
2916 * function, then fix the caller.
2922 function caseFold( $s ) {
2923 return $this->uc( $s );
2930 function checkTitleEncoding( $s ) {
2931 if ( is_array( $s ) ) {
2932 throw new MWException( 'Given array to checkTitleEncoding.' );
2934 if ( StringUtils
::isUtf8( $s ) ) {
2938 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2944 function fallback8bitEncoding() {
2945 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2949 * Most writing systems use whitespace to break up words.
2950 * Some languages such as Chinese don't conventionally do this,
2951 * which requires special handling when breaking up words for
2956 function hasWordBreaks() {
2961 * Some languages such as Chinese require word segmentation,
2962 * Specify such segmentation when overridden in derived class.
2964 * @param string $string
2967 function segmentByWord( $string ) {
2972 * Some languages have special punctuation need to be normalized.
2973 * Make such changes here.
2975 * @param string $string
2978 function normalizeForSearch( $string ) {
2979 return self
::convertDoubleWidth( $string );
2983 * convert double-width roman characters to single-width.
2984 * range: ff00-ff5f ~= 0020-007f
2986 * @param string $string
2990 protected static function convertDoubleWidth( $string ) {
2991 static $full = null;
2992 static $half = null;
2994 if ( $full === null ) {
2995 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2996 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2997 $full = str_split( $fullWidth, 3 );
2998 $half = str_split( $halfWidth );
3001 $string = str_replace( $full, $half, $string );
3006 * @param string $string
3007 * @param string $pattern
3010 protected static function insertSpace( $string, $pattern ) {
3011 $string = preg_replace( $pattern, " $1 ", $string );
3012 $string = preg_replace( '/ +/', ' ', $string );
3017 * @param array $termsArray
3020 function convertForSearchResult( $termsArray ) {
3021 # some languages, e.g. Chinese, need to do a conversion
3022 # in order for search results to be displayed correctly
3027 * Get the first character of a string.
3032 function firstChar( $s ) {
3035 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
3036 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
3041 if ( isset( $matches[1] ) ) {
3042 if ( strlen( $matches[1] ) != 3 ) {
3046 // Break down Hangul syllables to grab the first jamo
3047 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3048 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3050 } elseif ( $code < 0xb098 ) {
3051 return "\xe3\x84\xb1";
3052 } elseif ( $code < 0xb2e4 ) {
3053 return "\xe3\x84\xb4";
3054 } elseif ( $code < 0xb77c ) {
3055 return "\xe3\x84\xb7";
3056 } elseif ( $code < 0xb9c8 ) {
3057 return "\xe3\x84\xb9";
3058 } elseif ( $code < 0xbc14 ) {
3059 return "\xe3\x85\x81";
3060 } elseif ( $code < 0xc0ac ) {
3061 return "\xe3\x85\x82";
3062 } elseif ( $code < 0xc544 ) {
3063 return "\xe3\x85\x85";
3064 } elseif ( $code < 0xc790 ) {
3065 return "\xe3\x85\x87";
3066 } elseif ( $code < 0xcc28 ) {
3067 return "\xe3\x85\x88";
3068 } elseif ( $code < 0xce74 ) {
3069 return "\xe3\x85\x8a";
3070 } elseif ( $code < 0xd0c0 ) {
3071 return "\xe3\x85\x8b";
3072 } elseif ( $code < 0xd30c ) {
3073 return "\xe3\x85\x8c";
3074 } elseif ( $code < 0xd558 ) {
3075 return "\xe3\x85\x8d";
3077 return "\xe3\x85\x8e";
3084 function initEncoding() {
3085 # Some languages may have an alternate char encoding option
3086 # (Esperanto X-coding, Japanese furigana conversion, etc)
3087 # If this language is used as the primary content language,
3088 # an override to the defaults can be set here on startup.
3095 function recodeForEdit( $s ) {
3096 # For some languages we'll want to explicitly specify
3097 # which characters make it into the edit box raw
3098 # or are converted in some way or another.
3099 global $wgEditEncoding;
3100 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
3103 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3111 function recodeInput( $s ) {
3112 # Take the previous into account.
3113 global $wgEditEncoding;
3114 if ( $wgEditEncoding != '' ) {
3115 $enc = $wgEditEncoding;
3119 if ( $enc == 'UTF-8' ) {
3122 return $this->iconv( $enc, 'UTF-8', $s );
3127 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3128 * also cleans up certain backwards-compatible sequences, converting them
3129 * to the modern Unicode equivalent.
3131 * This is language-specific for performance reasons only.
3137 function normalize( $s ) {
3138 global $wgAllUnicodeFixes;
3139 $s = UtfNormal\Validator
::cleanUp( $s );
3140 if ( $wgAllUnicodeFixes ) {
3141 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3142 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3149 * Transform a string using serialized data stored in the given file (which
3150 * must be in the serialized subdirectory of $IP). The file contains pairs
3151 * mapping source characters to destination characters.
3153 * The data is cached in process memory. This will go faster if you have the
3154 * FastStringSearch extension.
3156 * @param string $file
3157 * @param string $string
3159 * @throws MWException
3162 function transformUsingPairFile( $file, $string ) {
3163 if ( !isset( $this->transformData
[$file] ) ) {
3164 $data = wfGetPrecompiledData( $file );
3165 if ( $data === false ) {
3166 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3168 $this->transformData
[$file] = new ReplacementArray( $data );
3170 return $this->transformData
[$file]->replace( $string );
3174 * For right-to-left language support
3179 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3183 * Return the correct HTML 'dir' attribute value for this language.
3187 return $this->isRTL() ?
'rtl' : 'ltr';
3191 * Return 'left' or 'right' as appropriate alignment for line-start
3192 * for this language's text direction.
3194 * Should be equivalent to CSS3 'start' text-align value....
3198 function alignStart() {
3199 return $this->isRTL() ?
'right' : 'left';
3203 * Return 'right' or 'left' as appropriate alignment for line-end
3204 * for this language's text direction.
3206 * Should be equivalent to CSS3 'end' text-align value....
3210 function alignEnd() {
3211 return $this->isRTL() ?
'left' : 'right';
3215 * A hidden direction mark (LRM or RLM), depending on the language direction.
3216 * Unlike getDirMark(), this function returns the character as an HTML entity.
3217 * This function should be used when the output is guaranteed to be HTML,
3218 * because it makes the output HTML source code more readable. When
3219 * the output is plain text or can be escaped, getDirMark() should be used.
3221 * @param bool $opposite Get the direction mark opposite to your language
3225 function getDirMarkEntity( $opposite = false ) {
3227 return $this->isRTL() ?
'‎' : '‏';
3229 return $this->isRTL() ?
'‏' : '‎';
3233 * A hidden direction mark (LRM or RLM), depending on the language direction.
3234 * This function produces them as invisible Unicode characters and
3235 * the output may be hard to read and debug, so it should only be used
3236 * when the output is plain text or can be escaped. When the output is
3237 * HTML, use getDirMarkEntity() instead.
3239 * @param bool $opposite Get the direction mark opposite to your language
3242 function getDirMark( $opposite = false ) {
3243 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3244 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3246 return $this->isRTL() ?
$lrm : $rlm;
3248 return $this->isRTL() ?
$rlm : $lrm;
3254 function capitalizeAllNouns() {
3255 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3259 * An arrow, depending on the language direction.
3261 * @param string $direction The direction of the arrow: forwards (default),
3262 * backwards, left, right, up, down.
3265 function getArrow( $direction = 'forwards' ) {
3266 switch ( $direction ) {
3268 return $this->isRTL() ?
'←' : '→';
3270 return $this->isRTL() ?
'→' : '←';
3283 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3287 function linkPrefixExtension() {
3288 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3292 * Get all magic words from cache.
3295 function getMagicWords() {
3296 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3300 * Run the LanguageGetMagic hook once.
3302 protected function doMagicHook() {
3303 if ( $this->mMagicHookDone
) {
3306 $this->mMagicHookDone
= true;
3307 Hooks
::run( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3311 * Fill a MagicWord object with data from here
3313 * @param MagicWord $mw
3315 function getMagic( $mw ) {
3316 // Saves a function call
3317 if ( !$this->mMagicHookDone
) {
3318 $this->doMagicHook();
3321 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3322 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3324 $rawEntry = self
::$dataCache->getSubitem(
3325 $this->mCode
, 'magicWords', $mw->mId
);
3328 if ( !is_array( $rawEntry ) ) {
3329 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3331 $mw->mCaseSensitive
= $rawEntry[0];
3332 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3337 * Add magic words to the extension array
3339 * @param array $newWords
3341 function addMagicWordsByLang( $newWords ) {
3342 $fallbackChain = $this->getFallbackLanguages();
3343 $fallbackChain = array_reverse( $fallbackChain );
3344 foreach ( $fallbackChain as $code ) {
3345 if ( isset( $newWords[$code] ) ) {
3346 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3352 * Get special page names, as an associative array
3353 * canonical name => array of valid names, including aliases
3356 function getSpecialPageAliases() {
3357 // Cache aliases because it may be slow to load them
3358 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3360 $this->mExtendedSpecialPageAliases
=
3361 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3362 Hooks
::run( 'LanguageGetSpecialPageAliases',
3363 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3366 return $this->mExtendedSpecialPageAliases
;
3370 * Italic is unsuitable for some languages
3372 * @param string $text The text to be emphasized.
3375 function emphasize( $text ) {
3376 return "<em>$text</em>";
3380 * Normally we output all numbers in plain en_US style, that is
3381 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3382 * point twohundredthirtyfive. However this is not suitable for all
3383 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3384 * Icelandic just want to use commas instead of dots, and dots instead
3385 * of commas like "293.291,235".
3387 * An example of this function being called:
3389 * wfMessage( 'message' )->numParams( $num )->text()
3392 * See $separatorTransformTable on MessageIs.php for
3393 * the , => . and . => , implementation.
3395 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3396 * @param int|float $number The string to be formatted, should be an integer
3397 * or a floating point number.
3398 * @param bool $nocommafy Set to true for special numbers like dates
3401 public function formatNum( $number, $nocommafy = false ) {
3402 global $wgTranslateNumerals;
3403 if ( !$nocommafy ) {
3404 $number = $this->commafy( $number );
3405 $s = $this->separatorTransformTable();
3407 $number = strtr( $number, $s );
3411 if ( $wgTranslateNumerals ) {
3412 $s = $this->digitTransformTable();
3414 $number = strtr( $number, $s );
3422 * Front-end for non-commafied formatNum
3424 * @param int|float $number The string to be formatted, should be an integer
3425 * or a floating point number.
3429 public function formatNumNoSeparators( $number ) {
3430 return $this->formatNum( $number, true );
3434 * @param string $number
3437 public function parseFormattedNumber( $number ) {
3438 $s = $this->digitTransformTable();
3440 // eliminate empty array values such as ''. (bug 64347)
3441 $s = array_filter( $s );
3442 $number = strtr( $number, array_flip( $s ) );
3445 $s = $this->separatorTransformTable();
3447 // eliminate empty array values such as ''. (bug 64347)
3448 $s = array_filter( $s );
3449 $number = strtr( $number, array_flip( $s ) );
3452 $number = strtr( $number, array( ',' => '' ) );
3457 * Adds commas to a given number
3459 * @param mixed $number
3462 function commafy( $number ) {
3463 $digitGroupingPattern = $this->digitGroupingPattern();
3464 if ( $number === null ) {
3468 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3469 // default grouping is at thousands, use the same for ###,###,### pattern too.
3470 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3472 // Ref: http://cldr.unicode.org/translation/number-patterns
3474 if ( intval( $number ) < 0 ) {
3475 // For negative numbers apply the algorithm like positive number and add sign.
3477 $number = substr( $number, 1 );
3479 $integerPart = array();
3480 $decimalPart = array();
3481 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3482 preg_match( "/\d+/", $number, $integerPart );
3483 preg_match( "/\.\d*/", $number, $decimalPart );
3484 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3485 if ( $groupedNumber === $number ) {
3486 // the string does not have any number part. Eg: .12345
3487 return $sign . $groupedNumber;
3489 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3490 while ( $start > 0 ) {
3491 $match = $matches[0][$numMatches - 1];
3492 $matchLen = strlen( $match );
3493 $start = $end - $matchLen;
3497 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3499 if ( $numMatches > 1 ) {
3500 // use the last pattern for the rest of the number
3504 $groupedNumber = "," . $groupedNumber;
3507 return $sign . $groupedNumber;
3514 function digitGroupingPattern() {
3515 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3521 function digitTransformTable() {
3522 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3528 function separatorTransformTable() {
3529 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3533 * Take a list of strings and build a locale-friendly comma-separated
3534 * list, using the local comma-separator message.
3535 * The last two strings are chained with an "and".
3536 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3538 * @param string[] $l
3541 function listToText( array $l ) {
3542 $m = count( $l ) - 1;
3547 $and = $this->msg( 'and' )->escaped();
3548 $space = $this->msg( 'word-separator' )->escaped();
3550 $comma = $this->msg( 'comma-separator' )->escaped();
3554 for ( $i = $m - 1; $i >= 0; $i-- ) {
3555 if ( $i == $m - 1 ) {
3556 $s = $l[$i] . $and . $space . $s;
3558 $s = $l[$i] . $comma . $s;
3565 * Take a list of strings and build a locale-friendly comma-separated
3566 * list, using the local comma-separator message.
3567 * @param string[] $list Array of strings to put in a comma list
3570 function commaList( array $list ) {
3572 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3578 * Take a list of strings and build a locale-friendly semicolon-separated
3579 * list, using the local semicolon-separator message.
3580 * @param string[] $list Array of strings to put in a semicolon list
3583 function semicolonList( array $list ) {
3585 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3591 * Same as commaList, but separate it with the pipe instead.
3592 * @param string[] $list Array of strings to put in a pipe list
3595 function pipeList( array $list ) {
3597 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3603 * Truncate a string to a specified length in bytes, appending an optional
3604 * string (e.g. for ellipses)
3606 * The database offers limited byte lengths for some columns in the database;
3607 * multi-byte character sets mean we need to ensure that only whole characters
3608 * are included, otherwise broken characters can be passed to the user
3610 * If $length is negative, the string will be truncated from the beginning
3612 * @param string $string String to truncate
3613 * @param int $length Maximum length (including ellipses)
3614 * @param string $ellipsis String to append to the truncated text
3615 * @param bool $adjustLength Subtract length of ellipsis from $length.
3616 * $adjustLength was introduced in 1.18, before that behaved as if false.
3619 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3620 # Use the localized ellipsis character
3621 if ( $ellipsis == '...' ) {
3622 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3624 # Check if there is no need to truncate
3625 if ( $length == 0 ) {
3626 return $ellipsis; // convention
3627 } elseif ( strlen( $string ) <= abs( $length ) ) {
3628 return $string; // no need to truncate
3630 $stringOriginal = $string;
3631 # If ellipsis length is >= $length then we can't apply $adjustLength
3632 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3633 $string = $ellipsis; // this can be slightly unexpected
3634 # Otherwise, truncate and add ellipsis...
3636 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3637 if ( $length > 0 ) {
3638 $length -= $eLength;
3639 $string = substr( $string, 0, $length ); // xyz...
3640 $string = $this->removeBadCharLast( $string );
3641 $string = rtrim( $string );
3642 $string = $string . $ellipsis;
3644 $length +
= $eLength;
3645 $string = substr( $string, $length ); // ...xyz
3646 $string = $this->removeBadCharFirst( $string );
3647 $string = ltrim( $string );
3648 $string = $ellipsis . $string;
3651 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3652 # This check is *not* redundant if $adjustLength, due to the single case where
3653 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3654 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3657 return $stringOriginal;
3662 * Remove bytes that represent an incomplete Unicode character
3663 * at the end of string (e.g. bytes of the char are missing)
3665 * @param string $string
3668 protected function removeBadCharLast( $string ) {
3669 if ( $string != '' ) {
3670 $char = ord( $string[strlen( $string ) - 1] );
3672 if ( $char >= 0xc0 ) {
3673 # We got the first byte only of a multibyte char; remove it.
3674 $string = substr( $string, 0, -1 );
3675 } elseif ( $char >= 0x80 &&
3676 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3677 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3679 # We chopped in the middle of a character; remove it
3687 * Remove bytes that represent an incomplete Unicode character
3688 * at the start of string (e.g. bytes of the char are missing)
3690 * @param string $string
3693 protected function removeBadCharFirst( $string ) {
3694 if ( $string != '' ) {
3695 $char = ord( $string[0] );
3696 if ( $char >= 0x80 && $char < 0xc0 ) {
3697 # We chopped in the middle of a character; remove the whole thing
3698 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3705 * Truncate a string of valid HTML to a specified length in bytes,
3706 * appending an optional string (e.g. for ellipses), and return valid HTML
3708 * This is only intended for styled/linked text, such as HTML with
3709 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3710 * Also, this will not detect things like "display:none" CSS.
3712 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3714 * @param string $text HTML string to truncate
3715 * @param int $length (zero/positive) Maximum length (including ellipses)
3716 * @param string $ellipsis String to append to the truncated text
3719 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3720 # Use the localized ellipsis character
3721 if ( $ellipsis == '...' ) {
3722 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3724 # Check if there is clearly no need to truncate
3725 if ( $length <= 0 ) {
3726 return $ellipsis; // no text shown, nothing to format (convention)
3727 } elseif ( strlen( $text ) <= $length ) {
3728 return $text; // string short enough even *with* HTML (short-circuit)
3731 $dispLen = 0; // innerHTML legth so far
3732 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3733 $tagType = 0; // 0-open, 1-close
3734 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3735 $entityState = 0; // 0-not entity, 1-entity
3736 $tag = $ret = ''; // accumulated tag name, accumulated result string
3737 $openTags = array(); // open tag stack
3738 $maybeState = null; // possible truncation state
3740 $textLen = strlen( $text );
3741 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3742 for ( $pos = 0; true; ++
$pos ) {
3743 # Consider truncation once the display length has reached the maximim.
3744 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3745 # Check that we're not in the middle of a bracket/entity...
3746 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3747 if ( !$testingEllipsis ) {
3748 $testingEllipsis = true;
3749 # Save where we are; we will truncate here unless there turn out to
3750 # be so few remaining characters that truncation is not necessary.
3751 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3752 $maybeState = array( $ret, $openTags ); // save state
3754 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3755 # String in fact does need truncation, the truncation point was OK.
3756 list( $ret, $openTags ) = $maybeState; // reload state
3757 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3758 $ret .= $ellipsis; // add ellipsis
3762 if ( $pos >= $textLen ) {
3763 break; // extra iteration just for above checks
3766 # Read the next char...
3768 $lastCh = $pos ?
$text[$pos - 1] : '';
3769 $ret .= $ch; // add to result string
3771 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3772 $entityState = 0; // for bad HTML
3773 $bracketState = 1; // tag started (checking for backslash)
3774 } elseif ( $ch == '>' ) {
3775 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3776 $entityState = 0; // for bad HTML
3777 $bracketState = 0; // out of brackets
3778 } elseif ( $bracketState == 1 ) {
3780 $tagType = 1; // close tag (e.g. "</span>")
3782 $tagType = 0; // open tag (e.g. "<span>")
3785 $bracketState = 2; // building tag name
3786 } elseif ( $bracketState == 2 ) {
3790 // Name found (e.g. "<a href=..."), add on tag attributes...
3791 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3793 } elseif ( $bracketState == 0 ) {
3794 if ( $entityState ) {
3797 $dispLen++
; // entity is one displayed char
3800 if ( $neLength == 0 && !$maybeState ) {
3801 // Save state without $ch. We want to *hit* the first
3802 // display char (to get tags) but not *use* it if truncating.
3803 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3806 $entityState = 1; // entity found, (e.g. " ")
3808 $dispLen++
; // this char is displayed
3809 // Add the next $max display text chars after this in one swoop...
3810 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3811 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3812 $dispLen +
= $skipped;
3818 // Close the last tag if left unclosed by bad HTML
3819 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3820 while ( count( $openTags ) > 0 ) {
3821 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3827 * truncateHtml() helper function
3828 * like strcspn() but adds the skipped chars to $ret
3830 * @param string $ret
3831 * @param string $text
3832 * @param string $search
3834 * @param null|int $len
3837 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3838 if ( $len === null ) {
3839 $len = -1; // -1 means "no limit" for strcspn
3840 } elseif ( $len < 0 ) {
3844 if ( $start < strlen( $text ) ) {
3845 $skipCount = strcspn( $text, $search, $start, $len );
3846 $ret .= substr( $text, $start, $skipCount );
3852 * truncateHtml() helper function
3853 * (a) push or pop $tag from $openTags as needed
3854 * (b) clear $tag value
3855 * @param string &$tag Current HTML tag name we are looking at
3856 * @param int $tagType (0-open tag, 1-close tag)
3857 * @param string $lastCh Character before the '>' that ended this tag
3858 * @param array &$openTags Open tag stack (not accounting for $tag)
3860 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3861 $tag = ltrim( $tag );
3863 if ( $tagType == 0 && $lastCh != '/' ) {
3864 $openTags[] = $tag; // tag opened (didn't close itself)
3865 } elseif ( $tagType == 1 ) {
3866 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3867 array_pop( $openTags ); // tag closed
3875 * Grammatical transformations, needed for inflected languages
3876 * Invoked by putting {{grammar:case|word}} in a message
3878 * @param string $word
3879 * @param string $case
3882 function convertGrammar( $word, $case ) {
3883 global $wgGrammarForms;
3884 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3885 return $wgGrammarForms[$this->getCode()][$case][$word];
3891 * Get the grammar forms for the content language
3892 * @return array Array of grammar forms
3895 function getGrammarForms() {
3896 global $wgGrammarForms;
3897 if ( isset( $wgGrammarForms[$this->getCode()] )
3898 && is_array( $wgGrammarForms[$this->getCode()] )
3900 return $wgGrammarForms[$this->getCode()];
3906 * Provides an alternative text depending on specified gender.
3907 * Usage {{gender:username|masculine|feminine|unknown}}.
3908 * username is optional, in which case the gender of current user is used,
3909 * but only in (some) interface messages; otherwise default gender is used.
3911 * If no forms are given, an empty string is returned. If only one form is
3912 * given, it will be returned unconditionally. These details are implied by
3913 * the caller and cannot be overridden in subclasses.
3915 * If three forms are given, the default is to use the third (unknown) form.
3916 * If fewer than three forms are given, the default is to use the first (masculine) form.
3917 * These details can be overridden in subclasses.
3919 * @param string $gender
3920 * @param array $forms
3924 function gender( $gender, $forms ) {
3925 if ( !count( $forms ) ) {
3928 $forms = $this->preConvertPlural( $forms, 2 );
3929 if ( $gender === 'male' ) {
3932 if ( $gender === 'female' ) {
3935 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3939 * Plural form transformations, needed for some languages.
3940 * For example, there are 3 form of plural in Russian and Polish,
3941 * depending on "count mod 10". See [[w:Plural]]
3942 * For English it is pretty simple.
3944 * Invoked by putting {{plural:count|wordform1|wordform2}}
3945 * or {{plural:count|wordform1|wordform2|wordform3}}
3947 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3949 * @param int $count Non-localized number
3950 * @param array $forms Different plural forms
3951 * @return string Correct form of plural for $count in this language
3953 function convertPlural( $count, $forms ) {
3954 // Handle explicit n=pluralform cases
3955 $forms = $this->handleExplicitPluralForms( $count, $forms );
3956 if ( is_string( $forms ) ) {
3959 if ( !count( $forms ) ) {
3963 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3964 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3965 return $forms[$pluralForm];
3969 * Handles explicit plural forms for Language::convertPlural()
3971 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3972 * If an explicitly defined plural form matches the $count, then
3973 * string value returned, otherwise array returned for further consideration
3974 * by CLDR rules or overridden convertPlural().
3978 * @param int $count Non-localized number
3979 * @param array $forms Different plural forms
3981 * @return array|string
3983 protected function handleExplicitPluralForms( $count, array $forms ) {
3984 foreach ( $forms as $index => $form ) {
3985 if ( preg_match( '/\d+=/i', $form ) ) {
3986 $pos = strpos( $form, '=' );
3987 if ( substr( $form, 0, $pos ) === (string)$count ) {
3988 return substr( $form, $pos +
1 );
3990 unset( $forms[$index] );
3993 return array_values( $forms );
3997 * Checks that convertPlural was given an array and pads it to requested
3998 * amount of forms by copying the last one.
4000 * @param array $forms Array of forms given to convertPlural
4001 * @param int $count How many forms should there be at least
4002 * @return array Padded array of forms or an exception if not an array
4004 protected function preConvertPlural( /* Array */ $forms, $count ) {
4005 while ( count( $forms ) < $count ) {
4006 $forms[] = $forms[count( $forms ) - 1];
4012 * Wraps argument with unicode control characters for directionality safety
4014 * This solves the problem where directionality-neutral characters at the edge of
4015 * the argument string get interpreted with the wrong directionality from the
4016 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4018 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4019 * directionality of the argument string, using the BIDI algorithm's own "First
4020 * strong directional codepoint" rule. Essentially, this works round the fact that
4021 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4022 * direction inference). The latter is cleaner but still not widely supported.
4024 * @param string $text Text to wrap
4025 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4027 public function embedBidi( $text = '' ) {
4028 $dir = Language
::strongDirFromContent( $text );
4029 if ( $dir === 'ltr' ) {
4030 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4031 return self
::$lre . $text . self
::$pdf;
4033 if ( $dir === 'rtl' ) {
4034 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4035 return self
::$rle . $text . self
::$pdf;
4037 // No strong directionality: do not wrap
4042 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4043 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4044 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4045 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4048 * @param string $str The validated block duration in English
4049 * @return string Somehow translated block duration
4050 * @see LanguageFi.php for example implementation
4052 function translateBlockExpiry( $str ) {
4053 $duration = SpecialBlock
::getSuggestedDurations( $this );
4054 foreach ( $duration as $show => $value ) {
4055 if ( strcmp( $str, $value ) == 0 ) {
4056 return htmlspecialchars( trim( $show ) );
4060 if ( wfIsInfinity( $str ) ) {
4061 foreach ( $duration as $show => $value ) {
4062 if ( wfIsInfinity( $value ) ) {
4063 return htmlspecialchars( trim( $show ) );
4068 // If all else fails, return a standard duration or timestamp description.
4069 $time = strtotime( $str, 0 );
4070 if ( $time === false ) { // Unknown format. Return it as-is in case.
4072 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4073 // $time is relative to 0 so it's a duration length.
4074 return $this->formatDuration( $time );
4075 } else { // It's an absolute timestamp.
4076 if ( $time === 0 ) {
4077 // wfTimestamp() handles 0 as current time instead of epoch.
4078 return $this->timeanddate( '19700101000000' );
4080 return $this->timeanddate( $time );
4086 * languages like Chinese need to be segmented in order for the diff
4089 * @param string $text
4092 public function segmentForDiff( $text ) {
4097 * and unsegment to show the result
4099 * @param string $text
4102 public function unsegmentForDiff( $text ) {
4107 * Return the LanguageConverter used in the Language
4110 * @return LanguageConverter
4112 public function getConverter() {
4113 return $this->mConverter
;
4117 * convert text to all supported variants
4119 * @param string $text
4122 public function autoConvertToAllVariants( $text ) {
4123 return $this->mConverter
->autoConvertToAllVariants( $text );
4127 * convert text to different variants of a language.
4129 * @param string $text
4132 public function convert( $text ) {
4133 return $this->mConverter
->convert( $text );
4137 * Convert a Title object to a string in the preferred variant
4139 * @param Title $title
4142 public function convertTitle( $title ) {
4143 return $this->mConverter
->convertTitle( $title );
4147 * Convert a namespace index to a string in the preferred variant
4152 public function convertNamespace( $ns ) {
4153 return $this->mConverter
->convertNamespace( $ns );
4157 * Check if this is a language with variants
4161 public function hasVariants() {
4162 return count( $this->getVariants() ) > 1;
4166 * Check if the language has the specific variant
4169 * @param string $variant
4172 public function hasVariant( $variant ) {
4173 return (bool)$this->mConverter
->validateVariant( $variant );
4177 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4179 * @param string $text
4181 * @deprecated since 1.22 is no longer used
4183 public function armourMath( $text ) {
4184 return $this->mConverter
->armourMath( $text );
4188 * Perform output conversion on a string, and encode for safe HTML output.
4189 * @param string $text Text to be converted
4190 * @param bool $isTitle Whether this conversion is for the article title
4192 * @todo this should get integrated somewhere sane
4194 public function convertHtml( $text, $isTitle = false ) {
4195 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4199 * @param string $key
4202 public function convertCategoryKey( $key ) {
4203 return $this->mConverter
->convertCategoryKey( $key );
4207 * Get the list of variants supported by this language
4208 * see sample implementation in LanguageZh.php
4210 * @return array An array of language codes
4212 public function getVariants() {
4213 return $this->mConverter
->getVariants();
4219 public function getPreferredVariant() {
4220 return $this->mConverter
->getPreferredVariant();
4226 public function getDefaultVariant() {
4227 return $this->mConverter
->getDefaultVariant();
4233 public function getURLVariant() {
4234 return $this->mConverter
->getURLVariant();
4238 * If a language supports multiple variants, it is
4239 * possible that non-existing link in one variant
4240 * actually exists in another variant. this function
4241 * tries to find it. See e.g. LanguageZh.php
4242 * The input parameters may be modified upon return
4244 * @param string &$link The name of the link
4245 * @param Title &$nt The title object of the link
4246 * @param bool $ignoreOtherCond To disable other conditions when
4247 * we need to transclude a template or update a category's link
4249 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4250 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4254 * returns language specific options used by User::getPageRenderHash()
4255 * for example, the preferred language variant
4259 function getExtraHashOptions() {
4260 return $this->mConverter
->getExtraHashOptions();
4264 * For languages that support multiple variants, the title of an
4265 * article may be displayed differently in different variants. this
4266 * function returns the apporiate title defined in the body of the article.
4270 public function getParsedTitle() {
4271 return $this->mConverter
->getParsedTitle();
4275 * Prepare external link text for conversion. When the text is
4276 * a URL, it shouldn't be converted, and it'll be wrapped in
4277 * the "raw" tag (-{R| }-) to prevent conversion.
4279 * This function is called "markNoConversion" for historical
4282 * @param string $text Text to be used for external link
4283 * @param bool $noParse Wrap it without confirming it's a real URL first
4284 * @return string The tagged text
4286 public function markNoConversion( $text, $noParse = false ) {
4287 // Excluding protocal-relative URLs may avoid many false positives.
4288 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4289 return $this->mConverter
->markNoConversion( $text );
4296 * A regular expression to match legal word-trailing characters
4297 * which should be merged onto a link of the form [[foo]]bar.
4301 public function linkTrail() {
4302 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4306 * A regular expression character set to match legal word-prefixing
4307 * characters which should be merged onto a link of the form foo[[bar]].
4311 public function linkPrefixCharset() {
4312 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4316 * @deprecated since 1.24, will be removed in 1.25
4319 function getLangObj() {
4320 wfDeprecated( __METHOD__
, '1.24' );
4325 * Get the "parent" language which has a converter to convert a "compatible" language
4326 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4328 * @return Language|null
4331 public function getParentLanguage() {
4332 if ( $this->mParentLanguage
!== false ) {
4333 return $this->mParentLanguage
;
4336 $pieces = explode( '-', $this->getCode() );
4338 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4339 $this->mParentLanguage
= null;
4342 $lang = Language
::factory( $code );
4343 if ( !$lang->hasVariant( $this->getCode() ) ) {
4344 $this->mParentLanguage
= null;
4348 $this->mParentLanguage
= $lang;
4353 * Get the RFC 3066 code for this language object
4355 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4356 * htmlspecialchars() or similar
4360 public function getCode() {
4361 return $this->mCode
;
4365 * Get the code in Bcp47 format which we can use
4366 * inside of html lang="" tags.
4368 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4369 * htmlspecialchars() or similar.
4374 public function getHtmlCode() {
4375 if ( is_null( $this->mHtmlCode
) ) {
4376 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4378 return $this->mHtmlCode
;
4382 * @param string $code
4384 public function setCode( $code ) {
4385 $this->mCode
= $code;
4386 // Ensure we don't leave incorrect cached data lying around
4387 $this->mHtmlCode
= null;
4388 $this->mParentLanguage
= false;
4392 * Get the name of a file for a certain language code
4393 * @param string $prefix Prepend this to the filename
4394 * @param string $code Language code
4395 * @param string $suffix Append this to the filename
4396 * @throws MWException
4397 * @return string $prefix . $mangledCode . $suffix
4399 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4400 if ( !self
::isValidBuiltInCode( $code ) ) {
4401 throw new MWException( "Invalid language code \"$code\"" );
4404 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4408 * Get the language code from a file name. Inverse of getFileName()
4409 * @param string $filename $prefix . $languageCode . $suffix
4410 * @param string $prefix Prefix before the language code
4411 * @param string $suffix Suffix after the language code
4412 * @return string Language code, or false if $prefix or $suffix isn't found
4414 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4416 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4417 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4418 if ( !count( $m ) ) {
4421 return str_replace( '_', '-', strtolower( $m[1] ) );
4425 * @param string $code
4428 public static function getMessagesFileName( $code ) {
4430 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4431 Hooks
::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4436 * @param string $code
4440 public static function getJsonMessagesFileName( $code ) {
4443 if ( !self
::isValidBuiltInCode( $code ) ) {
4444 throw new MWException( "Invalid language code \"$code\"" );
4447 return "$IP/languages/i18n/$code.json";
4451 * @param string $code
4454 public static function getClassFileName( $code ) {
4456 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4460 * Get the first fallback for a given language.
4462 * @param string $code
4464 * @return bool|string
4466 public static function getFallbackFor( $code ) {
4467 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4470 $fallbacks = self
::getFallbacksFor( $code );
4471 return $fallbacks[0];
4476 * Get the ordered list of fallback languages.
4479 * @param string $code Language code
4480 * @return array Non-empty array, ending in "en"
4482 public static function getFallbacksFor( $code ) {
4483 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4486 // For unknown languages, fallbackSequence returns an empty array,
4487 // hardcode fallback to 'en' in that case.
4488 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: array( 'en' );
4492 * Get the ordered list of fallback languages, ending with the fallback
4493 * language chain for the site language.
4496 * @param string $code Language code
4497 * @return array Array( fallbacks, site fallbacks )
4499 public static function getFallbacksIncludingSiteLanguage( $code ) {
4500 global $wgLanguageCode;
4502 // Usually, we will only store a tiny number of fallback chains, so we
4503 // keep them in static memory.
4504 $cacheKey = "{$code}-{$wgLanguageCode}";
4506 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4507 $fallbacks = self
::getFallbacksFor( $code );
4509 // Append the site's fallback chain, including the site language itself
4510 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4511 array_unshift( $siteFallbacks, $wgLanguageCode );
4513 // Eliminate any languages already included in the chain
4514 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4516 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4518 return self
::$fallbackLanguageCache[$cacheKey];
4522 * Get all messages for a given language
4523 * WARNING: this may take a long time. If you just need all message *keys*
4524 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4526 * @param string $code
4530 public static function getMessagesFor( $code ) {
4531 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4535 * Get a message for a given language
4537 * @param string $key
4538 * @param string $code
4542 public static function getMessageFor( $key, $code ) {
4543 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4547 * Get all message keys for a given language. This is a faster alternative to
4548 * array_keys( Language::getMessagesFor( $code ) )
4551 * @param string $code Language code
4552 * @return array Array of message keys (strings)
4554 public static function getMessageKeysFor( $code ) {
4555 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4559 * @param string $talk
4562 function fixVariableInNamespace( $talk ) {
4563 if ( strpos( $talk, '$1' ) === false ) {
4567 global $wgMetaNamespace;
4568 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4570 # Allow grammar transformations
4571 # Allowing full message-style parsing would make simple requests
4572 # such as action=raw much more expensive than they need to be.
4573 # This will hopefully cover most cases.
4574 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4575 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4576 return str_replace( ' ', '_', $talk );
4583 function replaceGrammarInNamespace( $m ) {
4584 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4588 * @throws MWException
4591 static function getCaseMaps() {
4592 static $wikiUpperChars, $wikiLowerChars;
4593 if ( isset( $wikiUpperChars ) ) {
4594 return array( $wikiUpperChars, $wikiLowerChars );
4597 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4598 if ( $arr === false ) {
4599 throw new MWException(
4600 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4602 $wikiUpperChars = $arr['wikiUpperChars'];
4603 $wikiLowerChars = $arr['wikiLowerChars'];
4604 return array( $wikiUpperChars, $wikiLowerChars );
4608 * Decode an expiry (block, protection, etc) which has come from the DB
4610 * @param string $expiry Database expiry String
4611 * @param bool|int $format True to process using language functions, or TS_ constant
4612 * to return the expiry in a given timestamp
4613 * @param string $inifinity If $format is not true, use this string for infinite expiry
4617 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4619 if ( $dbInfinity === null ) {
4620 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4623 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4624 return $format === true
4625 ?
$this->getMessageFromDB( 'infiniteblock' )
4628 return $format === true
4629 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4630 : wfTimestamp( $format, $expiry );
4636 * @param int|float $seconds
4637 * @param array $format Optional
4638 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4639 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4640 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4642 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4643 * or 'avoidminutes'.
4646 function formatTimePeriod( $seconds, $format = array() ) {
4647 if ( !is_array( $format ) ) {
4648 $format = array( 'avoid' => $format ); // For backwards compatibility
4650 if ( !isset( $format['avoid'] ) ) {
4651 $format['avoid'] = false;
4653 if ( !isset( $format['noabbrevs'] ) ) {
4654 $format['noabbrevs'] = false;
4656 $secondsMsg = wfMessage(
4657 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4658 $minutesMsg = wfMessage(
4659 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4660 $hoursMsg = wfMessage(
4661 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4662 $daysMsg = wfMessage(
4663 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4665 if ( round( $seconds * 10 ) < 100 ) {
4666 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4667 $s = $secondsMsg->params( $s )->text();
4668 } elseif ( round( $seconds ) < 60 ) {
4669 $s = $this->formatNum( round( $seconds ) );
4670 $s = $secondsMsg->params( $s )->text();
4671 } elseif ( round( $seconds ) < 3600 ) {
4672 $minutes = floor( $seconds / 60 );
4673 $secondsPart = round( fmod( $seconds, 60 ) );
4674 if ( $secondsPart == 60 ) {
4678 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4680 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4681 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4682 $hours = floor( $seconds / 3600 );
4683 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4684 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4685 if ( $secondsPart == 60 ) {
4689 if ( $minutes == 60 ) {
4693 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4695 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4696 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4697 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4700 $days = floor( $seconds / 86400 );
4701 if ( $format['avoid'] === 'avoidminutes' ) {
4702 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4703 if ( $hours == 24 ) {
4707 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4709 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4710 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4711 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4712 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4713 if ( $minutes == 60 ) {
4717 if ( $hours == 24 ) {
4721 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4723 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4725 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4727 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4729 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4736 * Format a bitrate for output, using an appropriate
4737 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4738 * the magnitude in question.
4740 * This use base 1000. For base 1024 use formatSize(), for another base
4741 * see formatComputingNumbers().
4746 function formatBitrate( $bps ) {
4747 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4751 * @param int $size Size of the unit
4752 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4753 * @param string $messageKey Message key to be uesd
4756 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4758 return str_replace( '$1', $this->formatNum( $size ),
4759 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4762 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4765 $maxIndex = count( $sizes ) - 1;
4766 while ( $size >= $boundary && $index < $maxIndex ) {
4771 // For small sizes no decimal places necessary
4774 // For MB and bigger two decimal places are smarter
4777 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4779 $size = round( $size, $round );
4780 $text = $this->getMessageFromDB( $msg );
4781 return str_replace( '$1', $this->formatNum( $size ), $text );
4785 * Format a size in bytes for output, using an appropriate
4786 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4788 * This method use base 1024. For base 1000 use formatBitrate(), for
4789 * another base see formatComputingNumbers()
4791 * @param int $size Size to format
4792 * @return string Plain text (not HTML)
4794 function formatSize( $size ) {
4795 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4799 * Make a list item, used by various special pages
4801 * @param string $page Page link
4802 * @param string $details HTML safe text between brackets
4803 * @param bool $oppositedm Add the direction mark opposite to your
4804 * language, to display text properly
4805 * @return HTML escaped string
4807 function specialList( $page, $details, $oppositedm = true ) {
4812 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4816 $this->msg( 'word-separator' )->escaped() .
4817 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4821 * Generate (prev x| next x) (20|50|100...) type links for paging
4823 * @param Title $title Title object to link
4824 * @param int $offset
4826 * @param array $query Optional URL query parameter string
4827 * @param bool $atend Optional param for specified if this is the last page
4830 public function viewPrevNext( Title
$title, $offset, $limit,
4831 array $query = array(), $atend = false
4833 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4835 # Make 'previous' link
4836 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4837 if ( $offset > 0 ) {
4838 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4839 $query, $prev, 'prevn-title', 'mw-prevlink' );
4841 $plink = htmlspecialchars( $prev );
4845 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4847 $nlink = htmlspecialchars( $next );
4849 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4850 $query, $next, 'nextn-title', 'mw-nextlink' );
4853 # Make links to set number of items per page
4854 $numLinks = array();
4855 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4856 $numLinks[] = $this->numLink( $title, $offset, $num,
4857 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4860 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4861 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4865 * Helper function for viewPrevNext() that generates links
4867 * @param Title $title Title object to link
4868 * @param int $offset
4870 * @param array $query Extra query parameters
4871 * @param string $link Text to use for the link; will be escaped
4872 * @param string $tooltipMsg Name of the message to use as tooltip
4873 * @param string $class Value of the "class" attribute of the link
4874 * @return string HTML fragment
4876 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4879 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4880 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4881 ->numParams( $limit )->text();
4883 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4884 'title' => $tooltip, 'class' => $class ), $link );
4888 * Get the conversion rule title, if any.
4892 public function getConvRuleTitle() {
4893 return $this->mConverter
->getConvRuleTitle();
4897 * Get the compiled plural rules for the language
4899 * @return array Associative array with plural form, and plural rule as key-value pairs
4901 public function getCompiledPluralRules() {
4902 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4903 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4904 if ( !$pluralRules ) {
4905 foreach ( $fallbacks as $fallbackCode ) {
4906 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4907 if ( $pluralRules ) {
4912 return $pluralRules;
4916 * Get the plural rules for the language
4918 * @return array Associative array with plural form number and plural rule as key-value pairs
4920 public function getPluralRules() {
4921 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4922 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4923 if ( !$pluralRules ) {
4924 foreach ( $fallbacks as $fallbackCode ) {
4925 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4926 if ( $pluralRules ) {
4931 return $pluralRules;
4935 * Get the plural rule types for the language
4937 * @return array Associative array with plural form number and plural rule type as key-value pairs
4939 public function getPluralRuleTypes() {
4940 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4941 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4942 if ( !$pluralRuleTypes ) {
4943 foreach ( $fallbacks as $fallbackCode ) {
4944 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4945 if ( $pluralRuleTypes ) {
4950 return $pluralRuleTypes;
4954 * Find the index number of the plural rule appropriate for the given number
4955 * @param int $number
4956 * @return int The index number of the plural rule
4958 public function getPluralRuleIndexNumber( $number ) {
4959 $pluralRules = $this->getCompiledPluralRules();
4960 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4965 * Find the plural rule type appropriate for the given number
4966 * For example, if the language is set to Arabic, getPluralType(5) should
4969 * @param int $number
4970 * @return string The name of the plural rule type, e.g. one, two, few, many
4972 public function getPluralRuleType( $number ) {
4973 $index = $this->getPluralRuleIndexNumber( $number );
4974 $pluralRuleTypes = $this->getPluralRuleTypes();
4975 if ( isset( $pluralRuleTypes[$index] ) ) {
4976 return $pluralRuleTypes[$index];