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 use CLDRPluralRuleParser\Evaluator
;
36 * Internationalisation code
41 * @var LanguageConverter
45 public $mVariants, $mCode, $mLoaded = false;
46 public $mMagicExtensions = [], $mMagicHookDone = false;
47 private $mHtmlCode = null, $mParentLanguage = false;
49 public $dateFormatStrings = [];
50 public $mExtendedSpecialPageAliases;
52 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
55 * ReplacementArray object caches
57 public $transformData = [];
60 * @var LocalisationCache
62 static public $dataCache;
64 static public $mLangObjCache = [];
66 static public $mWeekdayMsgs = [
67 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
71 static public $mWeekdayAbbrevMsgs = [
72 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
75 static public $mMonthMsgs = [
76 'january', 'february', 'march', 'april', 'may_long', 'june',
77 'july', 'august', 'september', 'october', 'november',
80 static public $mMonthGenMsgs = [
81 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
82 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
85 static public $mMonthAbbrevMsgs = [
86 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
87 'sep', 'oct', 'nov', 'dec'
90 static public $mIranianCalendarMonthMsgs = [
91 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
92 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
93 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
94 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
97 static public $mHebrewCalendarMonthMsgs = [
98 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
99 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
100 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
101 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
102 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
105 static public $mHebrewCalendarMonthGenMsgs = [
106 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
107 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
108 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
109 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
110 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
113 static public $mHijriCalendarMonthMsgs = [
114 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
115 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
116 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
117 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
124 static public $durationIntervals = [
125 'millennia' => 31556952000,
126 'centuries' => 3155695200,
127 'decades' => 315569520,
128 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
137 * Cache for language fallbacks.
138 * @see Language::getFallbacksIncludingSiteLanguage
142 static private $fallbackLanguageCache = [];
145 * Cache for language names
146 * @var HashBagOStuff|null
148 static private $languageNameCache;
151 * Unicode directional formatting characters, for embedBidi()
153 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
154 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
155 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
158 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
159 * - in group 1 if it is LTR
160 * - in group 2 if it is RTL
161 * Does not match if there is no strong directionality codepoint.
163 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
165 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
166 * https://git.wikimedia.org/summary/unicodejs.git .
168 // @codingStandardsIgnoreStart
169 // @codeCoverageIgnoreStart
170 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';
171 // @codeCoverageIgnoreEnd
172 // @codingStandardsIgnoreEnd
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( [ $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 if ( !Language
::isValidCode( $code ) ) {
207 throw new MWException( "Invalid language code \"$code\"" );
210 if ( !Language
::isValidBuiltInCode( $code ) ) {
211 // It's not possible to customise this code with class files, so
212 // just return a Language object. This is to support uselang= hacks.
213 $lang = new Language
;
214 $lang->setCode( $code );
218 // Check if there is a language class for the code
219 $class = self
::classFromCode( $code );
220 if ( class_exists( $class ) ) {
225 // Keep trying the fallback list until we find an existing class
226 $fallbacks = Language
::getFallbacksFor( $code );
227 foreach ( $fallbacks as $fallbackCode ) {
228 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
229 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
232 $class = self
::classFromCode( $fallbackCode );
233 if ( class_exists( $class ) ) {
235 $lang->setCode( $code );
240 throw new MWException( "Invalid fallback sequence for language '$code'" );
244 * Checks whether any localisation is available for that language tag
245 * in MediaWiki (MessagesXx.php exists).
247 * @param string $code Language tag (in lower case)
248 * @return bool Whether language is supported
251 public static function isSupportedLanguage( $code ) {
252 if ( !self
::isValidBuiltInCode( $code ) ) {
256 if ( $code === 'qqq' ) {
260 return is_readable( self
::getMessagesFileName( $code ) ) ||
261 is_readable( self
::getJsonMessagesFileName( $code ) );
265 * Returns true if a language code string is a well-formed language tag
266 * according to RFC 5646.
267 * This function only checks well-formedness; it doesn't check that
268 * language, script or variant codes actually exist in the repositories.
270 * Based on regexes by Mark Davis of the Unicode Consortium:
271 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
273 * @param string $code
274 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
279 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
282 $alphanum = '[a-z0-9]';
283 $x = 'x'; # private use singleton
284 $singleton = '[a-wy-z]'; # other singleton
285 $s = $lenient ?
'[-_]' : '-';
287 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
288 $script = "$alpha{4}"; # ISO 15924
289 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
290 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
291 $extension = "$singleton(?:$s$alphanum{2,8})+";
292 $privateUse = "$x(?:$s$alphanum{1,8})+";
294 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
295 # Since these are limited, this is safe even later changes to the registry --
296 # the only oddity is that it might change the type of the tag, and thus
297 # the results from the capturing groups.
298 # http://www.iana.org/assignments/language-subtag-registry
300 $grandfathered = "en{$s}GB{$s}oed"
301 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
302 . "|no{$s}(?:bok|nyn)"
303 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
304 . "|zh{$s}min{$s}nan";
306 $variantList = "$variant(?:$s$variant)*";
307 $extensionList = "$extension(?:$s$extension)*";
309 $langtag = "(?:($language)"
312 . "(?:$s$variantList)?"
313 . "(?:$s$extensionList)?"
314 . "(?:$s$privateUse)?)";
316 # The final breakdown, with capturing groups for each of these components
317 # The variants, extensions, grandfathered, and private-use may have interior '-'
319 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
321 return (bool)preg_match( "/$root/", strtolower( $code ) );
325 * Returns true if a language code string is of a valid form, whether or
326 * not it exists. This includes codes which are used solely for
327 * customisation via the MediaWiki namespace.
329 * @param string $code
333 public static function isValidCode( $code ) {
335 if ( !isset( $cache[$code] ) ) {
336 // People think language codes are html safe, so enforce it.
337 // Ideally we should only allow a-zA-Z0-9-
338 // but, .+ and other chars are often used for {{int:}} hacks
339 // see bugs 37564, 37587, 36938
341 // Protect against path traversal
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 // Quick escape for invalid input to avoid exceptions down the line
383 // when code tries to process tags which are not valid at all.
384 if ( !self
::isValidBuiltInCode( $tag ) ) {
388 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
389 || self
::fetchLanguageName( $tag, $tag ) !== ''
398 * Get the LocalisationCache instance
400 * @return LocalisationCache
402 public static function getLocalisationCache() {
403 if ( is_null( self
::$dataCache ) ) {
404 global $wgLocalisationCacheConf;
405 $class = $wgLocalisationCacheConf['class'];
406 self
::$dataCache = new $class( $wgLocalisationCacheConf );
408 return self
::$dataCache;
411 function __construct() {
412 $this->mConverter
= new FakeConverter( $this );
413 // Set the code to the name of the descendant
414 if ( get_class( $this ) == 'Language' ) {
417 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
419 self
::getLocalisationCache();
423 * Reduce memory usage
425 function __destruct() {
426 foreach ( $this as $name => $value ) {
427 unset( $this->$name );
432 * Hook which will be called if this is the content language.
433 * Descendants can use this to register hook functions or modify globals
435 function initContLang() {
442 public function getFallbackLanguages() {
443 return self
::getFallbacksFor( $this->mCode
);
447 * Exports $wgBookstoreListEn
450 public function getBookstoreList() {
451 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
455 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
456 * available in localised form, it will be included in English.
460 public function getNamespaces() {
461 if ( is_null( $this->namespaceNames
) ) {
462 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
464 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
465 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
467 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
469 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
470 if ( $wgMetaNamespaceTalk ) {
471 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
473 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
474 $this->namespaceNames
[NS_PROJECT_TALK
] =
475 $this->fixVariableInNamespace( $talk );
478 # Sometimes a language will be localised but not actually exist on this wiki.
479 foreach ( $this->namespaceNames
as $key => $text ) {
480 if ( !isset( $validNamespaces[$key] ) ) {
481 unset( $this->namespaceNames
[$key] );
485 # The above mixing may leave namespaces out of canonical order.
486 # Re-order by namespace ID number...
487 ksort( $this->namespaceNames
);
489 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
492 return $this->namespaceNames
;
496 * Arbitrarily set all of the namespace names at once. Mainly used for testing
497 * @param array $namespaces Array of namespaces (id => name)
499 public function setNamespaces( array $namespaces ) {
500 $this->namespaceNames
= $namespaces;
501 $this->mNamespaceIds
= null;
505 * Resets all of the namespace caches. Mainly used for testing
507 public function resetNamespaces() {
508 $this->namespaceNames
= null;
509 $this->mNamespaceIds
= null;
510 $this->namespaceAliases
= null;
514 * A convenience function that returns getNamespaces() with spaces instead of underscores
515 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
519 public function getFormattedNamespaces() {
520 $ns = $this->getNamespaces();
521 foreach ( $ns as $k => $v ) {
522 $ns[$k] = strtr( $v, '_', ' ' );
528 * Get a namespace value by key
531 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
532 * echo $mw_ns; // prints 'MediaWiki'
535 * @param int $index The array key of the namespace to return
536 * @return string|bool String if the namespace value exists, otherwise false
538 public function getNsText( $index ) {
539 $ns = $this->getNamespaces();
540 return isset( $ns[$index] ) ?
$ns[$index] : false;
544 * A convenience function that returns the same thing as
545 * getNsText() except with '_' changed to ' ', useful for
549 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
550 * echo $mw_ns; // prints 'MediaWiki talk'
553 * @param int $index The array key of the namespace to return
554 * @return string Namespace name without underscores (empty string if namespace does not exist)
556 public function getFormattedNsText( $index ) {
557 $ns = $this->getNsText( $index );
558 return strtr( $ns, '_', ' ' );
562 * Returns gender-dependent namespace alias if available.
563 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
564 * @param int $index Namespace index
565 * @param string $gender Gender key (male, female... )
569 public function getGenderNsText( $index, $gender ) {
570 global $wgExtraGenderNamespaces;
572 $ns = $wgExtraGenderNamespaces +
573 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
575 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
579 * Whether this language uses gender-dependent namespace aliases.
580 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
584 public function needsGenderDistinction() {
585 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
586 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
587 // $wgExtraGenderNamespaces overrides everything
589 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
590 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
591 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
594 // Check what is in i18n files
595 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
596 return count( $aliases ) > 0;
601 * Get a namespace key by value, case insensitive.
602 * Only matches namespace names for the current language, not the
603 * canonical ones defined in Namespace.php.
605 * @param string $text
606 * @return int|bool An integer if $text is a valid value otherwise false
608 function getLocalNsIndex( $text ) {
609 $lctext = $this->lc( $text );
610 $ids = $this->getNamespaceIds();
611 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
617 public function getNamespaceAliases() {
618 if ( is_null( $this->namespaceAliases
) ) {
619 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
623 foreach ( $aliases as $name => $index ) {
624 if ( $index === NS_PROJECT_TALK
) {
625 unset( $aliases[$name] );
626 $name = $this->fixVariableInNamespace( $name );
627 $aliases[$name] = $index;
632 global $wgExtraGenderNamespaces;
633 $genders = $wgExtraGenderNamespaces +
634 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
635 foreach ( $genders as $index => $forms ) {
636 foreach ( $forms as $alias ) {
637 $aliases[$alias] = $index;
641 # Also add converted namespace names as aliases, to avoid confusion.
642 $convertedNames = [];
643 foreach ( $this->getVariants() as $variant ) {
644 if ( $variant === $this->mCode
) {
647 foreach ( $this->getNamespaces() as $ns => $_ ) {
648 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
652 $this->namespaceAliases
= $aliases +
$convertedNames;
655 return $this->namespaceAliases
;
661 public function getNamespaceIds() {
662 if ( is_null( $this->mNamespaceIds
) ) {
663 global $wgNamespaceAliases;
664 # Put namespace names and aliases into a hashtable.
665 # If this is too slow, then we should arrange it so that it is done
666 # before caching. The catch is that at pre-cache time, the above
667 # class-specific fixup hasn't been done.
668 $this->mNamespaceIds
= [];
669 foreach ( $this->getNamespaces() as $index => $name ) {
670 $this->mNamespaceIds
[$this->lc( $name )] = $index;
672 foreach ( $this->getNamespaceAliases() as $name => $index ) {
673 $this->mNamespaceIds
[$this->lc( $name )] = $index;
675 if ( $wgNamespaceAliases ) {
676 foreach ( $wgNamespaceAliases as $name => $index ) {
677 $this->mNamespaceIds
[$this->lc( $name )] = $index;
681 return $this->mNamespaceIds
;
685 * Get a namespace key by value, case insensitive. Canonical namespace
686 * names override custom ones defined for the current language.
688 * @param string $text
689 * @return int|bool An integer if $text is a valid value otherwise false
691 public function getNsIndex( $text ) {
692 $lctext = $this->lc( $text );
693 $ns = MWNamespace
::getCanonicalIndex( $lctext );
694 if ( $ns !== null ) {
697 $ids = $this->getNamespaceIds();
698 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
702 * short names for language variants used for language conversion links.
704 * @param string $code
705 * @param bool $usemsg Use the "variantname-xyz" message if it exists
708 public function getVariantname( $code, $usemsg = true ) {
709 $msg = "variantname-$code";
710 if ( $usemsg && wfMessage( $msg )->exists() ) {
711 return $this->getMessageFromDB( $msg );
713 $name = self
::fetchLanguageName( $code );
715 return $name; # if it's defined as a language name, show that
717 # otherwise, output the language code
725 public function getDatePreferences() {
726 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
732 function getDateFormats() {
733 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
737 * @return array|string
739 public function getDefaultDateFormat() {
740 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
741 if ( $df === 'dmy or mdy' ) {
742 global $wgAmericanDates;
743 return $wgAmericanDates ?
'mdy' : 'dmy';
752 public function getDatePreferenceMigrationMap() {
753 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
757 * @param string $image
760 function getImageFile( $image ) {
761 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
768 public function getImageFiles() {
769 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
775 public function getExtraUserToggles() {
776 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
783 function getUserToggle( $tog ) {
784 return $this->getMessageFromDB( "tog-$tog" );
788 * Get an array of language names, indexed by code.
789 * @param null|string $inLanguage Code of language in which to return the names
790 * Use null for autonyms (native names)
791 * @param string $include One of:
792 * 'all' all available languages
793 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
794 * 'mwfile' only if the language is in 'mw' *and* has a message file
795 * @return array Language code => language name
798 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
799 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
800 $cacheKey .= ":$include";
801 if ( self
::$languageNameCache === null ) {
802 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
805 $ret = self
::$languageNameCache->get( $cacheKey );
807 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
808 self
::$languageNameCache->set( $cacheKey, $ret );
814 * Uncached helper for fetchLanguageNames
815 * @param null|string $inLanguage Code of language in which to return the names
816 * Use null for autonyms (native names)
817 * @param string $include One of:
818 * 'all' all available languages
819 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
820 * 'mwfile' only if the language is in 'mw' *and* has a message file
821 * @return array Language code => language name
823 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
824 global $wgExtraLanguageNames;
826 // If passed an invalid language code to use, fallback to en
827 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
834 # TODO: also include when $inLanguage is null, when this code is more efficient
835 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
838 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
839 foreach ( $mwNames as $mwCode => $mwName ) {
840 # - Prefer own MediaWiki native name when not using the hook
841 # - For other names just add if not added through the hook
842 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
843 $names[$mwCode] = $mwName;
847 if ( $include === 'all' ) {
853 $coreCodes = array_keys( $mwNames );
854 foreach ( $coreCodes as $coreCode ) {
855 $returnMw[$coreCode] = $names[$coreCode];
858 if ( $include === 'mwfile' ) {
860 # We do this using a foreach over the codes instead of a directory
861 # loop so that messages files in extensions will work correctly.
862 foreach ( $returnMw as $code => $value ) {
863 if ( is_readable( self
::getMessagesFileName( $code ) )
864 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
866 $namesMwFile[$code] = $names[$code];
870 ksort( $namesMwFile );
875 # 'mw' option; default if it's not one of the other two options (all/mwfile)
880 * @param string $code The code of the language for which to get the name
881 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
882 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
883 * @return string Language name or empty
886 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
887 $code = strtolower( $code );
888 $array = self
::fetchLanguageNames( $inLanguage, $include );
889 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
893 * Get a message from the MediaWiki namespace.
895 * @param string $msg Message name
898 public function getMessageFromDB( $msg ) {
899 return $this->msg( $msg )->text();
903 * Get message object in this language. Only for use inside this class.
905 * @param string $msg Message name
908 protected function msg( $msg ) {
909 return wfMessage( $msg )->inLanguage( $this );
916 public function getMonthName( $key ) {
917 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
923 public function getMonthNamesArray() {
924 $monthNames = [ '' ];
925 for ( $i = 1; $i < 13; $i++
) {
926 $monthNames[] = $this->getMonthName( $i );
935 public function getMonthNameGen( $key ) {
936 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
943 public function getMonthAbbreviation( $key ) {
944 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
950 public function getMonthAbbreviationsArray() {
951 $monthNames = [ '' ];
952 for ( $i = 1; $i < 13; $i++
) {
953 $monthNames[] = $this->getMonthAbbreviation( $i );
962 public function getWeekdayName( $key ) {
963 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
970 function getWeekdayAbbreviation( $key ) {
971 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
978 function getIranianCalendarMonthName( $key ) {
979 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
986 function getHebrewCalendarMonthName( $key ) {
987 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
994 function getHebrewCalendarMonthNameGen( $key ) {
995 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1002 function getHijriCalendarMonthName( $key ) {
1003 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1007 * Pass through result from $dateTimeObj->format()
1008 * @param DateTime|bool|null &$dateTimeObj
1010 * @param DateTimeZone|bool|null $zone
1011 * @param string $code
1014 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1015 if ( !$dateTimeObj ) {
1016 $dateTimeObj = DateTime
::createFromFormat(
1017 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1020 return $dateTimeObj->format( $code );
1024 * This is a workalike of PHP's date() function, but with better
1025 * internationalisation, a reduced set of format characters, and a better
1028 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1029 * the PHP manual for definitions. There are a number of extensions, which
1032 * xn Do not translate digits of the next numeric format character
1033 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1034 * xr Use roman numerals for the next numeric format character
1035 * xh Use hebrew numerals for the next numeric format character
1037 * xg Genitive month name
1039 * xij j (day number) in Iranian calendar
1040 * xiF F (month name) in Iranian calendar
1041 * xin n (month number) in Iranian calendar
1042 * xiy y (two digit year) in Iranian calendar
1043 * xiY Y (full year) in Iranian calendar
1044 * xit t (days in month) in Iranian calendar
1046 * xjj j (day number) in Hebrew calendar
1047 * xjF F (month name) in Hebrew calendar
1048 * xjt t (days in month) in Hebrew calendar
1049 * xjx xg (genitive month name) in Hebrew calendar
1050 * xjn n (month number) in Hebrew calendar
1051 * xjY Y (full year) in Hebrew calendar
1053 * xmj j (day number) in Hijri calendar
1054 * xmF F (month name) in Hijri calendar
1055 * xmn n (month number) in Hijri calendar
1056 * xmY Y (full year) in Hijri calendar
1058 * xkY Y (full year) in Thai solar calendar. Months and days are
1059 * identical to the Gregorian calendar
1060 * xoY Y (full year) in Minguo calendar or Juche year.
1061 * Months and days are identical to the
1062 * Gregorian calendar
1063 * xtY Y (full year) in Japanese nengo. Months and days are
1064 * identical to the Gregorian calendar
1066 * Characters enclosed in double quotes will be considered literal (with
1067 * the quotes themselves removed). Unmatched quotes will be considered
1068 * literal quotes. Example:
1070 * "The month is" F => The month is January
1073 * Backslash escaping is also supported.
1075 * Input timestamp is assumed to be pre-normalized to the desired local
1076 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1077 * $ts is UTC if $zone is not given.
1079 * @param string $format
1080 * @param string $ts 14-character timestamp
1083 * @param DateTimeZone $zone Timezone of $ts
1084 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1085 * Only makes sense if $ts is the current time.
1086 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1088 * @throws MWException
1091 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1096 $dateTimeObj = false;
1105 $usedSecond = false;
1106 $usedMinute = false;
1113 $usedISOYear = false;
1114 $usedIsLeapYear = false;
1116 $usedHebrewMonth = false;
1117 $usedIranianMonth = false;
1118 $usedHijriMonth = false;
1119 $usedHebrewYear = false;
1120 $usedIranianYear = false;
1121 $usedHijriYear = false;
1122 $usedTennoYear = false;
1124 if ( strlen( $ts ) !== 14 ) {
1125 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1128 if ( !ctype_digit( $ts ) ) {
1129 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1132 $formatLength = strlen( $format );
1133 for ( $p = 0; $p < $formatLength; $p++
) {
1135 $code = $format[$p];
1136 if ( $code == 'x' && $p < $formatLength - 1 ) {
1137 $code .= $format[++
$p];
1140 if ( ( $code === 'xi'
1146 && $p < $formatLength - 1 ) {
1147 $code .= $format[++
$p];
1158 $rawToggle = !$rawToggle;
1168 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1171 $usedHebrewMonth = true;
1173 $hebrew = self
::tsToHebrew( $ts );
1175 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1179 $num = substr( $ts, 6, 2 );
1183 $s .= $this->getWeekdayAbbreviation(
1184 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1189 $num = intval( substr( $ts, 6, 2 ) );
1194 $iranian = self
::tsToIranian( $ts );
1201 $hijri = self
::tsToHijri( $ts );
1208 $hebrew = self
::tsToHebrew( $ts );
1214 $s .= $this->getWeekdayName(
1215 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1220 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1223 $usedIranianMonth = true;
1225 $iranian = self
::tsToIranian( $ts );
1227 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1230 $usedHijriMonth = true;
1232 $hijri = self
::tsToHijri( $ts );
1234 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1237 $usedHebrewMonth = true;
1239 $hebrew = self
::tsToHebrew( $ts );
1241 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1245 $num = substr( $ts, 4, 2 );
1249 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1253 $num = intval( substr( $ts, 4, 2 ) );
1256 $usedIranianMonth = true;
1258 $iranian = self
::tsToIranian( $ts );
1263 $usedHijriMonth = true;
1265 $hijri = self
::tsToHijri( $ts );
1270 $usedHebrewMonth = true;
1272 $hebrew = self
::tsToHebrew( $ts );
1277 $usedHebrewMonth = true;
1279 $hebrew = self
::tsToHebrew( $ts );
1285 $num = substr( $ts, 0, 4 );
1288 $usedIranianYear = true;
1290 $iranian = self
::tsToIranian( $ts );
1295 $usedHijriYear = true;
1297 $hijri = self
::tsToHijri( $ts );
1302 $usedHebrewYear = true;
1304 $hebrew = self
::tsToHebrew( $ts );
1311 $thai = self
::tsToYear( $ts, 'thai' );
1318 $minguo = self
::tsToYear( $ts, 'minguo' );
1323 $usedTennoYear = true;
1325 $tenno = self
::tsToYear( $ts, 'tenno' );
1331 $num = substr( $ts, 2, 2 );
1334 $usedIranianYear = true;
1336 $iranian = self
::tsToIranian( $ts );
1338 $num = substr( $iranian[0], -2 );
1341 $usedIranianYear = true;
1343 $iranian = self
::tsToIranian( $ts );
1345 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1349 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1353 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1357 $h = substr( $ts, 8, 2 );
1358 $num = $h %
12 ?
$h %
12 : 12;
1362 $num = intval( substr( $ts, 8, 2 ) );
1366 $h = substr( $ts, 8, 2 );
1367 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1371 $num = substr( $ts, 8, 2 );
1375 $num = substr( $ts, 10, 2 );
1379 $num = substr( $ts, 12, 2 );
1389 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1395 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1399 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1403 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1406 $usedIsLeapYear = true;
1407 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1410 $usedISOYear = true;
1411 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1418 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1421 # Backslash escaping
1422 if ( $p < $formatLength - 1 ) {
1423 $s .= $format[++
$p];
1430 if ( $p < $formatLength - 1 ) {
1431 $endQuote = strpos( $format, '"', $p +
1 );
1432 if ( $endQuote === false ) {
1433 # No terminating quote, assume literal "
1436 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1440 # Quote at end of string, assume literal "
1447 if ( $num !== false ) {
1448 if ( $rawToggle ||
$raw ) {
1451 } elseif ( $roman ) {
1452 $s .= Language
::romanNumeral( $num );
1454 } elseif ( $hebrewNum ) {
1455 $s .= self
::hebrewNumeral( $num );
1458 $s .= $this->formatNum( $num, true );
1463 if ( $ttl === 'unused' ) {
1464 // No need to calculate the TTL, the caller wont use it anyway.
1465 } elseif ( $usedSecond ) {
1467 } elseif ( $usedMinute ) {
1468 $ttl = 60 - substr( $ts, 12, 2 );
1469 } elseif ( $usedHour ) {
1470 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1471 } elseif ( $usedAMPM ) {
1472 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1473 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1477 $usedIranianMonth ||
1484 // @todo Someone who understands the non-Gregorian calendars
1485 // should write proper logic for them so that they don't need purged every day.
1486 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1487 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1490 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1491 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1494 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1495 $timeRemainingInDay;
1496 } elseif ( $usedISOYear ) {
1497 // December 28th falls on the last ISO week of the year, every year.
1498 // The last ISO week of a year can be 52 or 53.
1499 $lastWeekOfISOYear = DateTime
::createFromFormat(
1501 substr( $ts, 0, 4 ) . '1228',
1502 $zone ?
: new DateTimeZone( 'UTC' )
1504 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1505 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1506 $timeRemainingInWeek =
1507 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1508 +
$timeRemainingInDay;
1509 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1514 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1515 substr( $ts, 6, 2 ) ) * 86400
1516 +
$timeRemainingInDay;
1517 } elseif ( $usedYear ) {
1519 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1520 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1521 +
$timeRemainingInDay;
1522 } elseif ( $usedIsLeapYear ) {
1523 $year = substr( $ts, 0, 4 );
1524 $timeRemainingInYear =
1525 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1526 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1527 +
$timeRemainingInDay;
1529 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1530 // this isn't a leap year. see when the next one starts
1531 $nextCandidate = $year - $mod +
4;
1532 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1533 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1534 $timeRemainingInYear;
1536 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1537 $timeRemainingInYear;
1540 // this is a leap year, so the next year isn't
1541 $possibleTtls[] = $timeRemainingInYear;
1545 if ( $possibleTtls ) {
1546 $ttl = min( $possibleTtls );
1553 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1554 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1557 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1558 * Gregorian dates to Iranian dates. Originally written in C, it
1559 * is released under the terms of GNU Lesser General Public
1560 * License. Conversion to PHP was performed by Niklas Laxström.
1562 * Link: http://www.farsiweb.info/jalali/jalali.c
1568 private static function tsToIranian( $ts ) {
1569 $gy = substr( $ts, 0, 4 ) -1600;
1570 $gm = substr( $ts, 4, 2 ) -1;
1571 $gd = substr( $ts, 6, 2 ) -1;
1573 # Days passed from the beginning (including leap years)
1575 +
floor( ( $gy +
3 ) / 4 )
1576 - floor( ( $gy +
99 ) / 100 )
1577 +
floor( ( $gy +
399 ) / 400 );
1579 // Add days of the past months of this year
1580 for ( $i = 0; $i < $gm; $i++
) {
1581 $gDayNo +
= self
::$GREG_DAYS[$i];
1585 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1589 // Days passed in current month
1590 $gDayNo +
= (int)$gd;
1592 $jDayNo = $gDayNo - 79;
1594 $jNp = floor( $jDayNo / 12053 );
1597 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1600 if ( $jDayNo >= 366 ) {
1601 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1602 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1605 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1606 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1612 return [ $jy, $jm, $jd ];
1616 * Converting Gregorian dates to Hijri dates.
1618 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1620 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1626 private static function tsToHijri( $ts ) {
1627 $year = substr( $ts, 0, 4 );
1628 $month = substr( $ts, 4, 2 );
1629 $day = substr( $ts, 6, 2 );
1637 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1638 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1640 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1641 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1642 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1645 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1646 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1649 $zl = $zjd -1948440 +
10632;
1650 $zn = (int)( ( $zl - 1 ) / 10631 );
1651 $zl = $zl - 10631 * $zn +
354;
1652 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1653 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1654 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1655 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1656 $zm = (int)( ( 24 * $zl ) / 709 );
1657 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1658 $zy = 30 * $zn +
$zj - 30;
1660 return [ $zy, $zm, $zd ];
1664 * Converting Gregorian dates to Hebrew dates.
1666 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1667 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1668 * to translate the relevant functions into PHP and release them under
1671 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1672 * and Adar II is 14. In a non-leap year, Adar is 6.
1678 private static function tsToHebrew( $ts ) {
1680 $year = substr( $ts, 0, 4 );
1681 $month = substr( $ts, 4, 2 );
1682 $day = substr( $ts, 6, 2 );
1684 # Calculate Hebrew year
1685 $hebrewYear = $year +
3760;
1687 # Month number when September = 1, August = 12
1689 if ( $month > 12 ) {
1696 # Calculate day of year from 1 September
1698 for ( $i = 1; $i < $month; $i++
) {
1702 # Check if the year is leap
1703 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1706 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1713 # Calculate the start of the Hebrew year
1714 $start = self
::hebrewYearStart( $hebrewYear );
1716 # Calculate next year's start
1717 if ( $dayOfYear <= $start ) {
1718 # Day is before the start of the year - it is the previous year
1720 $nextStart = $start;
1724 # Add days since previous year's 1 September
1726 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1730 # Start of the new (previous) year
1731 $start = self
::hebrewYearStart( $hebrewYear );
1734 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1737 # Calculate Hebrew day of year
1738 $hebrewDayOfYear = $dayOfYear - $start;
1740 # Difference between year's days
1741 $diff = $nextStart - $start;
1742 # Add 12 (or 13 for leap years) days to ignore the difference between
1743 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1744 # difference is only about the year type
1745 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1751 # Check the year pattern, and is leap year
1752 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1753 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1754 # and non-leap years
1755 $yearPattern = $diff %
30;
1756 # Check if leap year
1757 $isLeap = $diff >= 30;
1759 # Calculate day in the month from number of day in the Hebrew year
1760 # Don't check Adar - if the day is not in Adar, we will stop before;
1761 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1762 $hebrewDay = $hebrewDayOfYear;
1765 while ( $hebrewMonth <= 12 ) {
1766 # Calculate days in this month
1767 if ( $isLeap && $hebrewMonth == 6 ) {
1768 # Adar in a leap year
1770 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1772 if ( $hebrewDay <= $days ) {
1776 # Subtract the days of Adar I
1777 $hebrewDay -= $days;
1780 if ( $hebrewDay <= $days ) {
1786 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1787 # Cheshvan in a complete year (otherwise as the rule below)
1789 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1790 # Kislev in an incomplete year (otherwise as the rule below)
1793 # Odd months have 30 days, even have 29
1794 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1796 if ( $hebrewDay <= $days ) {
1797 # In the current month
1800 # Subtract the days of the current month
1801 $hebrewDay -= $days;
1802 # Try in the next month
1807 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1811 * This calculates the Hebrew year start, as days since 1 September.
1812 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1813 * Used for Hebrew date.
1819 private static function hebrewYearStart( $year ) {
1820 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1821 $b = intval( ( $year - 1 ) %
4 );
1822 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1826 $Mar = intval( $m );
1832 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1833 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1835 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1837 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1841 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1846 * Algorithm to convert Gregorian dates to Thai solar dates,
1847 * Minguo dates or Minguo dates.
1849 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1850 * http://en.wikipedia.org/wiki/Minguo_calendar
1851 * http://en.wikipedia.org/wiki/Japanese_era_name
1853 * @param string $ts 14-character timestamp
1854 * @param string $cName Calender name
1855 * @return array Converted year, month, day
1857 private static function tsToYear( $ts, $cName ) {
1858 $gy = substr( $ts, 0, 4 );
1859 $gm = substr( $ts, 4, 2 );
1860 $gd = substr( $ts, 6, 2 );
1862 if ( !strcmp( $cName, 'thai' ) ) {
1864 # Add 543 years to the Gregorian calendar
1865 # Months and days are identical
1866 $gy_offset = $gy +
543;
1867 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1869 # Deduct 1911 years from the Gregorian calendar
1870 # Months and days are identical
1871 $gy_offset = $gy - 1911;
1872 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1873 # Nengō dates up to Meiji period
1874 # Deduct years from the Gregorian calendar
1875 # depending on the nengo periods
1876 # Months and days are identical
1878 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1879 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1882 $gy_gannen = $gy - 1868 +
1;
1883 $gy_offset = $gy_gannen;
1884 if ( $gy_gannen == 1 ) {
1887 $gy_offset = '明治' . $gy_offset;
1889 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1890 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1891 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1892 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1893 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1896 $gy_gannen = $gy - 1912 +
1;
1897 $gy_offset = $gy_gannen;
1898 if ( $gy_gannen == 1 ) {
1901 $gy_offset = '大正' . $gy_offset;
1903 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1904 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1905 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1908 $gy_gannen = $gy - 1926 +
1;
1909 $gy_offset = $gy_gannen;
1910 if ( $gy_gannen == 1 ) {
1913 $gy_offset = '昭和' . $gy_offset;
1916 $gy_gannen = $gy - 1989 +
1;
1917 $gy_offset = $gy_gannen;
1918 if ( $gy_gannen == 1 ) {
1921 $gy_offset = '平成' . $gy_offset;
1927 return [ $gy_offset, $gm, $gd ];
1931 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1933 * This is the rule the BIDI algorithm uses to determine the directionality of
1934 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1935 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1937 * TODO: Does not handle BIDI control characters inside the text.
1938 * TODO: Does not handle unallocated characters.
1940 * @param string $text Text to test
1941 * @return null|string Directionality ('ltr' or 'rtl') or null
1943 private static function strongDirFromContent( $text = '' ) {
1944 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
1947 if ( $matches[1] === '' ) {
1954 * Roman number formatting up to 10000
1960 static function romanNumeral( $num ) {
1962 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1963 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1964 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1965 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1966 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1969 $num = intval( $num );
1970 if ( $num > 10000 ||
$num <= 0 ) {
1975 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1976 if ( $num >= $pow10 ) {
1977 $s .= $table[$i][(int)floor( $num / $pow10 )];
1979 $num = $num %
$pow10;
1985 * Hebrew Gematria number formatting up to 9999
1991 static function hebrewNumeral( $num ) {
1993 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
1994 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2007 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2010 $num = intval( $num );
2011 if ( $num > 9999 ||
$num <= 0 ) {
2015 // Round thousands have special notations
2016 if ( $num === 1000 ) {
2018 } elseif ( $num %
1000 === 0 ) {
2019 return $table[0][$num / 1000] . "' אלפים";
2024 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2025 if ( $num >= $pow10 ) {
2026 if ( $num === 15 ||
$num === 16 ) {
2027 $letters[] = $table[0][9];
2028 $letters[] = $table[0][$num - 9];
2031 $letters = array_merge(
2033 (array)$table[$i][intval( $num / $pow10 )]
2036 if ( $pow10 === 1000 ) {
2042 $num = $num %
$pow10;
2045 $preTransformLength = count( $letters );
2046 if ( $preTransformLength === 1 ) {
2047 // Add geresh (single quote) to one-letter numbers
2050 $lastIndex = $preTransformLength - 1;
2051 $letters[$lastIndex] = str_replace(
2052 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2053 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2054 $letters[$lastIndex]
2057 // Add gershayim (double quote) to multiple-letter numbers,
2058 // but exclude numbers with only one letter after the thousands
2059 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2060 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2063 array_splice( $letters, -1, 0, '"' );
2067 return implode( $letters );
2071 * Used by date() and time() to adjust the time output.
2073 * @param string $ts The time in date('YmdHis') format
2074 * @param mixed $tz Adjust the time by this amount (default false, mean we
2075 * get user timecorrection setting)
2078 public function userAdjust( $ts, $tz = false ) {
2079 global $wgUser, $wgLocalTZoffset;
2081 if ( $tz === false ) {
2082 $tz = $wgUser->getOption( 'timecorrection' );
2085 $data = explode( '|', $tz, 3 );
2087 if ( $data[0] == 'ZoneInfo' ) {
2088 MediaWiki\
suppressWarnings();
2089 $userTZ = timezone_open( $data[2] );
2090 MediaWiki\restoreWarnings
();
2091 if ( $userTZ !== false ) {
2092 $date = date_create( $ts, timezone_open( 'UTC' ) );
2093 date_timezone_set( $date, $userTZ );
2094 $date = date_format( $date, 'YmdHis' );
2097 # Unrecognized timezone, default to 'Offset' with the stored offset.
2098 $data[0] = 'Offset';
2101 if ( $data[0] == 'System' ||
$tz == '' ) {
2102 # Global offset in minutes.
2103 $minDiff = $wgLocalTZoffset;
2104 } elseif ( $data[0] == 'Offset' ) {
2105 $minDiff = intval( $data[1] );
2107 $data = explode( ':', $tz );
2108 if ( count( $data ) == 2 ) {
2109 $data[0] = intval( $data[0] );
2110 $data[1] = intval( $data[1] );
2111 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2112 if ( $data[0] < 0 ) {
2113 $minDiff = -$minDiff;
2116 $minDiff = intval( $data[0] ) * 60;
2120 # No difference ? Return time unchanged
2121 if ( 0 == $minDiff ) {
2125 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2126 # Generate an adjusted date; take advantage of the fact that mktime
2127 # will normalize out-of-range values so we don't have to split $minDiff
2128 # into hours and minutes.
2130 (int)substr( $ts, 8, 2 ) ), # Hours
2131 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2132 (int)substr( $ts, 12, 2 ), # Seconds
2133 (int)substr( $ts, 4, 2 ), # Month
2134 (int)substr( $ts, 6, 2 ), # Day
2135 (int)substr( $ts, 0, 4 ) ); # Year
2137 $date = date( 'YmdHis', $t );
2138 MediaWiki\restoreWarnings
();
2144 * This is meant to be used by time(), date(), and timeanddate() to get
2145 * the date preference they're supposed to use, it should be used in
2149 * function timeanddate([...], $format = true) {
2150 * $datePreference = $this->dateFormat($format);
2155 * @param int|string|bool $usePrefs If true, the user's preference is used
2156 * if false, the site/language default is used
2157 * if int/string, assumed to be a format.
2160 function dateFormat( $usePrefs = true ) {
2163 if ( is_bool( $usePrefs ) ) {
2165 $datePreference = $wgUser->getDatePreference();
2167 $datePreference = (string)User
::getDefaultOption( 'date' );
2170 $datePreference = (string)$usePrefs;
2174 if ( $datePreference == '' ) {
2178 return $datePreference;
2182 * Get a format string for a given type and preference
2183 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2184 * @param string $pref The format name as it appears in Messages*.php under
2187 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2191 function getDateFormatString( $type, $pref ) {
2192 $wasDefault = false;
2193 if ( $pref == 'default' ) {
2195 $pref = $this->getDefaultDateFormat();
2198 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2199 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2201 if ( $type === 'pretty' && $df === null ) {
2202 $df = $this->getDateFormatString( 'date', $pref );
2205 if ( !$wasDefault && $df === null ) {
2206 $pref = $this->getDefaultDateFormat();
2207 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2210 $this->dateFormatStrings
[$type][$pref] = $df;
2212 return $this->dateFormatStrings
[$type][$pref];
2216 * @param string $ts The time format which needs to be turned into a
2217 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2218 * @param bool $adj Whether to adjust the time output according to the
2219 * user configured offset ($timecorrection)
2220 * @param mixed $format True to use user's date format preference
2221 * @param string|bool $timecorrection The time offset as returned by
2222 * validateTimeZone() in Special:Preferences
2225 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2226 $ts = wfTimestamp( TS_MW
, $ts );
2228 $ts = $this->userAdjust( $ts, $timecorrection );
2230 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2231 return $this->sprintfDate( $df, $ts );
2235 * @param string $ts The time format which needs to be turned into a
2236 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2237 * @param bool $adj Whether to adjust the time output according to the
2238 * user configured offset ($timecorrection)
2239 * @param mixed $format True to use user's date format preference
2240 * @param string|bool $timecorrection The time offset as returned by
2241 * validateTimeZone() in Special:Preferences
2244 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2245 $ts = wfTimestamp( TS_MW
, $ts );
2247 $ts = $this->userAdjust( $ts, $timecorrection );
2249 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2250 return $this->sprintfDate( $df, $ts );
2254 * @param string $ts The time format which needs to be turned into a
2255 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2256 * @param bool $adj Whether to adjust the time output according to the
2257 * user configured offset ($timecorrection)
2258 * @param mixed $format What format to return, if it's false output the
2259 * default one (default true)
2260 * @param string|bool $timecorrection The time offset as returned by
2261 * validateTimeZone() in Special:Preferences
2264 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2265 $ts = wfTimestamp( TS_MW
, $ts );
2267 $ts = $this->userAdjust( $ts, $timecorrection );
2269 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2270 return $this->sprintfDate( $df, $ts );
2274 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2278 * @param int $seconds The amount of seconds.
2279 * @param array $chosenIntervals The intervals to enable.
2283 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2284 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2288 foreach ( $intervals as $intervalName => $intervalValue ) {
2289 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2290 // duration-years, duration-decades, duration-centuries, duration-millennia
2291 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2292 $segments[] = $message->inLanguage( $this )->escaped();
2295 return $this->listToText( $segments );
2299 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2300 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2304 * @param int $seconds The amount of seconds.
2305 * @param array $chosenIntervals The intervals to enable.
2309 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2310 if ( empty( $chosenIntervals ) ) {
2311 $chosenIntervals = [
2323 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2324 $sortedNames = array_keys( $intervals );
2325 $smallestInterval = array_pop( $sortedNames );
2329 foreach ( $intervals as $name => $length ) {
2330 $value = floor( $seconds / $length );
2332 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2333 $seconds -= $value * $length;
2334 $segments[$name] = $value;
2342 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2344 * @param string $type Can be 'date', 'time' or 'both'
2345 * @param string $ts The time format which needs to be turned into a
2346 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2347 * @param User $user User object used to get preferences for timezone and format
2348 * @param array $options Array, can contain the following keys:
2349 * - 'timecorrection': time correction, can have the following values:
2350 * - true: use user's preference
2351 * - false: don't use time correction
2352 * - int: value of time correction in minutes
2353 * - 'format': format to use, can have the following values:
2354 * - true: use user's preference
2355 * - false: use default preference
2356 * - string: format to use
2360 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2361 $ts = wfTimestamp( TS_MW
, $ts );
2362 $options +
= [ 'timecorrection' => true, 'format' => true ];
2363 if ( $options['timecorrection'] !== false ) {
2364 if ( $options['timecorrection'] === true ) {
2365 $offset = $user->getOption( 'timecorrection' );
2367 $offset = $options['timecorrection'];
2369 $ts = $this->userAdjust( $ts, $offset );
2371 if ( $options['format'] === true ) {
2372 $format = $user->getDatePreference();
2374 $format = $options['format'];
2376 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2377 return $this->sprintfDate( $df, $ts );
2381 * Get the formatted date for the given timestamp and formatted for
2384 * @param mixed $ts Mixed: the time format which needs to be turned into a
2385 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2386 * @param User $user User object used to get preferences for timezone and format
2387 * @param array $options Array, can contain the following keys:
2388 * - 'timecorrection': time correction, can have the following values:
2389 * - true: use user's preference
2390 * - false: don't use time correction
2391 * - int: value of time correction in minutes
2392 * - 'format': format to use, can have the following values:
2393 * - true: use user's preference
2394 * - false: use default preference
2395 * - string: format to use
2399 public function userDate( $ts, User
$user, array $options = [] ) {
2400 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2404 * Get the formatted time for the given timestamp and formatted for
2407 * @param mixed $ts The time format which needs to be turned into a
2408 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2409 * @param User $user User object used to get preferences for timezone and format
2410 * @param array $options Array, can contain the following keys:
2411 * - 'timecorrection': time correction, can have the following values:
2412 * - true: use user's preference
2413 * - false: don't use time correction
2414 * - int: value of time correction in minutes
2415 * - 'format': format to use, can have the following values:
2416 * - true: use user's preference
2417 * - false: use default preference
2418 * - string: format to use
2422 public function userTime( $ts, User
$user, array $options = [] ) {
2423 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2427 * Get the formatted date and time for the given timestamp and formatted for
2430 * @param mixed $ts The time format which needs to be turned into a
2431 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2432 * @param User $user User object used to get preferences for timezone and format
2433 * @param array $options Array, can contain the following keys:
2434 * - 'timecorrection': time correction, can have the following values:
2435 * - true: use user's preference
2436 * - false: don't use time correction
2437 * - int: value of time correction in minutes
2438 * - 'format': format to use, can have the following values:
2439 * - true: use user's preference
2440 * - false: use default preference
2441 * - string: format to use
2445 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2446 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2450 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2452 * Determine the difference between the timestamp and the current time, and
2453 * generate a readable timestamp by returning "<N> <units> ago", where the
2454 * largest possible unit is used.
2456 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2458 * @param MWTimestamp $time
2459 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2460 * @param User|null $user User the timestamp is being generated for
2461 * (or null to use main context's user)
2462 * @return string Formatted timestamp
2464 public function getHumanTimestamp(
2465 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2467 if ( $relativeTo === null ) {
2468 $relativeTo = new MWTimestamp();
2470 if ( $user === null ) {
2471 $user = RequestContext
::getMain()->getUser();
2474 // Adjust for the user's timezone.
2475 $offsetThis = $time->offsetForUser( $user );
2476 $offsetRel = $relativeTo->offsetForUser( $user );
2479 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2480 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2483 // Reset the timezone on the objects.
2484 $time->timestamp
->sub( $offsetThis );
2485 $relativeTo->timestamp
->sub( $offsetRel );
2491 * Convert an MWTimestamp into a pretty human-readable timestamp using
2492 * the given user preferences and relative base time.
2494 * @see Language::getHumanTimestamp
2495 * @param MWTimestamp $ts Timestamp to prettify
2496 * @param MWTimestamp $relativeTo Base timestamp
2497 * @param User $user User preferences to use
2498 * @return string Human timestamp
2501 private function getHumanTimestampInternal(
2502 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2504 $diff = $ts->diff( $relativeTo );
2505 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2506 (int)$relativeTo->timestamp
->format( 'w' ) );
2507 $days = $diff->days ?
: (int)$diffDay;
2508 if ( $diff->invert ||
$days > 5
2509 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2511 // Timestamps are in different years: use full timestamp
2512 // Also do full timestamp for future dates
2514 * @todo FIXME: Add better handling of future timestamps.
2516 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2517 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2518 } elseif ( $days > 5 ) {
2519 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2520 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2521 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2522 } elseif ( $days > 1 ) {
2523 // Timestamp within the past week: show the day of the week and time
2524 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2525 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2527 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2528 $ts = wfMessage( "$weekday-at" )
2529 ->inLanguage( $this )
2530 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2532 } elseif ( $days == 1 ) {
2533 // Timestamp was yesterday: say 'yesterday' and the time.
2534 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2535 $ts = wfMessage( 'yesterday-at' )
2536 ->inLanguage( $this )
2537 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2539 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2540 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2541 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2542 $ts = wfMessage( 'today-at' )
2543 ->inLanguage( $this )
2544 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2547 // From here on in, the timestamp was soon enough ago so that we can simply say
2548 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2549 } elseif ( $diff->h
== 1 ) {
2550 // Less than 90 minutes, but more than an hour ago.
2551 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2552 } elseif ( $diff->i
>= 1 ) {
2553 // A few minutes ago.
2554 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2555 } elseif ( $diff->s
>= 30 ) {
2556 // Less than a minute, but more than 30 sec ago.
2557 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2559 // Less than 30 seconds ago.
2560 $ts = wfMessage( 'just-now' )->text();
2567 * @param string $key
2568 * @return array|null
2570 public function getMessage( $key ) {
2571 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2577 function getAllMessages() {
2578 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2583 * @param string $out
2584 * @param string $string
2587 public function iconv( $in, $out, $string ) {
2588 # Even with //IGNORE iconv can whine about illegal characters in
2589 # *input* string. We just ignore those too.
2590 # REF: http://bugs.php.net/bug.php?id=37166
2591 # REF: https://phabricator.wikimedia.org/T18885
2592 MediaWiki\
suppressWarnings();
2593 $text = iconv( $in, $out . '//IGNORE', $string );
2594 MediaWiki\restoreWarnings
();
2598 // callback functions for ucwords(), ucwordbreaks()
2601 * @param array $matches
2602 * @return mixed|string
2604 function ucwordbreaksCallbackAscii( $matches ) {
2605 return $this->ucfirst( $matches[1] );
2609 * @param array $matches
2612 function ucwordbreaksCallbackMB( $matches ) {
2613 return mb_strtoupper( $matches[0] );
2617 * @param array $matches
2620 function ucwordsCallbackMB( $matches ) {
2621 return mb_strtoupper( $matches[0] );
2625 * Make a string's first character uppercase
2627 * @param string $str
2631 public function ucfirst( $str ) {
2633 if ( $o < 96 ) { // if already uppercase...
2635 } elseif ( $o < 128 ) {
2636 return ucfirst( $str ); // use PHP's ucfirst()
2638 // fall back to more complex logic in case of multibyte strings
2639 return $this->uc( $str, true );
2644 * Convert a string to uppercase
2646 * @param string $str
2647 * @param bool $first
2651 public function uc( $str, $first = false ) {
2653 if ( $this->isMultibyte( $str ) ) {
2654 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2656 return ucfirst( $str );
2659 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2664 * @param string $str
2665 * @return mixed|string
2667 function lcfirst( $str ) {
2670 return strval( $str );
2671 } elseif ( $o >= 128 ) {
2672 return $this->lc( $str, true );
2673 } elseif ( $o > 96 ) {
2676 $str[0] = strtolower( $str[0] );
2682 * @param string $str
2683 * @param bool $first
2684 * @return mixed|string
2686 function lc( $str, $first = false ) {
2688 if ( $this->isMultibyte( $str ) ) {
2689 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2691 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2694 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2699 * @param string $str
2702 function isMultibyte( $str ) {
2703 return strlen( $str ) !== mb_strlen( $str );
2707 * @param string $str
2708 * @return mixed|string
2710 function ucwords( $str ) {
2711 if ( $this->isMultibyte( $str ) ) {
2712 $str = $this->lc( $str );
2714 // regexp to find first letter in each word (i.e. after each space)
2715 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2717 // function to use to capitalize a single char
2718 return preg_replace_callback(
2720 [ $this, 'ucwordsCallbackMB' ],
2724 return ucwords( strtolower( $str ) );
2729 * capitalize words at word breaks
2731 * @param string $str
2734 function ucwordbreaks( $str ) {
2735 if ( $this->isMultibyte( $str ) ) {
2736 $str = $this->lc( $str );
2738 // since \b doesn't work for UTF-8, we explicitely define word break chars
2739 $breaks = "[ \-\(\)\}\{\.,\?!]";
2741 // find first letter after word break
2742 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2743 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2745 return preg_replace_callback(
2747 [ $this, 'ucwordbreaksCallbackMB' ],
2751 return preg_replace_callback(
2752 '/\b([\w\x80-\xff]+)\b/',
2753 [ $this, 'ucwordbreaksCallbackAscii' ],
2760 * Return a case-folded representation of $s
2762 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2763 * and $s2 are the same except for the case of their characters. It is not
2764 * necessary for the value returned to make sense when displayed.
2766 * Do *not* perform any other normalisation in this function. If a caller
2767 * uses this function when it should be using a more general normalisation
2768 * function, then fix the caller.
2774 function caseFold( $s ) {
2775 return $this->uc( $s );
2781 * @throws MWException
2783 function checkTitleEncoding( $s ) {
2784 if ( is_array( $s ) ) {
2785 throw new MWException( 'Given array to checkTitleEncoding.' );
2787 if ( StringUtils
::isUtf8( $s ) ) {
2791 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2797 function fallback8bitEncoding() {
2798 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2802 * Most writing systems use whitespace to break up words.
2803 * Some languages such as Chinese don't conventionally do this,
2804 * which requires special handling when breaking up words for
2809 function hasWordBreaks() {
2814 * Some languages such as Chinese require word segmentation,
2815 * Specify such segmentation when overridden in derived class.
2817 * @param string $string
2820 function segmentByWord( $string ) {
2825 * Some languages have special punctuation need to be normalized.
2826 * Make such changes here.
2828 * @param string $string
2831 function normalizeForSearch( $string ) {
2832 return self
::convertDoubleWidth( $string );
2836 * convert double-width roman characters to single-width.
2837 * range: ff00-ff5f ~= 0020-007f
2839 * @param string $string
2843 protected static function convertDoubleWidth( $string ) {
2844 static $full = null;
2845 static $half = null;
2847 if ( $full === null ) {
2848 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2849 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2850 $full = str_split( $fullWidth, 3 );
2851 $half = str_split( $halfWidth );
2854 $string = str_replace( $full, $half, $string );
2859 * @param string $string
2860 * @param string $pattern
2863 protected static function insertSpace( $string, $pattern ) {
2864 $string = preg_replace( $pattern, " $1 ", $string );
2865 $string = preg_replace( '/ +/', ' ', $string );
2870 * @param array $termsArray
2873 function convertForSearchResult( $termsArray ) {
2874 # some languages, e.g. Chinese, need to do a conversion
2875 # in order for search results to be displayed correctly
2880 * Get the first character of a string.
2885 function firstChar( $s ) {
2888 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2889 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2894 if ( isset( $matches[1] ) ) {
2895 if ( strlen( $matches[1] ) != 3 ) {
2899 // Break down Hangul syllables to grab the first jamo
2900 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
2901 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2903 } elseif ( $code < 0xb098 ) {
2904 return "\xe3\x84\xb1";
2905 } elseif ( $code < 0xb2e4 ) {
2906 return "\xe3\x84\xb4";
2907 } elseif ( $code < 0xb77c ) {
2908 return "\xe3\x84\xb7";
2909 } elseif ( $code < 0xb9c8 ) {
2910 return "\xe3\x84\xb9";
2911 } elseif ( $code < 0xbc14 ) {
2912 return "\xe3\x85\x81";
2913 } elseif ( $code < 0xc0ac ) {
2914 return "\xe3\x85\x82";
2915 } elseif ( $code < 0xc544 ) {
2916 return "\xe3\x85\x85";
2917 } elseif ( $code < 0xc790 ) {
2918 return "\xe3\x85\x87";
2919 } elseif ( $code < 0xcc28 ) {
2920 return "\xe3\x85\x88";
2921 } elseif ( $code < 0xce74 ) {
2922 return "\xe3\x85\x8a";
2923 } elseif ( $code < 0xd0c0 ) {
2924 return "\xe3\x85\x8b";
2925 } elseif ( $code < 0xd30c ) {
2926 return "\xe3\x85\x8c";
2927 } elseif ( $code < 0xd558 ) {
2928 return "\xe3\x85\x8d";
2930 return "\xe3\x85\x8e";
2938 * @deprecated No-op since 1.28
2940 function initEncoding() {
2947 * @deprecated No-op since 1.28
2949 function recodeForEdit( $s ) {
2956 * @deprecated No-op since 1.28
2958 function recodeInput( $s ) {
2963 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2964 * also cleans up certain backwards-compatible sequences, converting them
2965 * to the modern Unicode equivalent.
2967 * This is language-specific for performance reasons only.
2973 function normalize( $s ) {
2974 global $wgAllUnicodeFixes;
2975 $s = UtfNormal\Validator
::cleanUp( $s );
2976 if ( $wgAllUnicodeFixes ) {
2977 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2978 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2985 * Transform a string using serialized data stored in the given file (which
2986 * must be in the serialized subdirectory of $IP). The file contains pairs
2987 * mapping source characters to destination characters.
2989 * The data is cached in process memory. This will go faster if you have the
2990 * FastStringSearch extension.
2992 * @param string $file
2993 * @param string $string
2995 * @throws MWException
2998 function transformUsingPairFile( $file, $string ) {
2999 if ( !isset( $this->transformData
[$file] ) ) {
3000 $data = wfGetPrecompiledData( $file );
3001 if ( $data === false ) {
3002 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3004 $this->transformData
[$file] = new ReplacementArray( $data );
3006 return $this->transformData
[$file]->replace( $string );
3010 * For right-to-left language support
3015 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3019 * Return the correct HTML 'dir' attribute value for this language.
3023 return $this->isRTL() ?
'rtl' : 'ltr';
3027 * Return 'left' or 'right' as appropriate alignment for line-start
3028 * for this language's text direction.
3030 * Should be equivalent to CSS3 'start' text-align value....
3034 function alignStart() {
3035 return $this->isRTL() ?
'right' : 'left';
3039 * Return 'right' or 'left' as appropriate alignment for line-end
3040 * for this language's text direction.
3042 * Should be equivalent to CSS3 'end' text-align value....
3046 function alignEnd() {
3047 return $this->isRTL() ?
'left' : 'right';
3051 * A hidden direction mark (LRM or RLM), depending on the language direction.
3052 * Unlike getDirMark(), this function returns the character as an HTML entity.
3053 * This function should be used when the output is guaranteed to be HTML,
3054 * because it makes the output HTML source code more readable. When
3055 * the output is plain text or can be escaped, getDirMark() should be used.
3057 * @param bool $opposite Get the direction mark opposite to your language
3061 function getDirMarkEntity( $opposite = false ) {
3063 return $this->isRTL() ?
'‎' : '‏';
3065 return $this->isRTL() ?
'‏' : '‎';
3069 * A hidden direction mark (LRM or RLM), depending on the language direction.
3070 * This function produces them as invisible Unicode characters and
3071 * the output may be hard to read and debug, so it should only be used
3072 * when the output is plain text or can be escaped. When the output is
3073 * HTML, use getDirMarkEntity() instead.
3075 * @param bool $opposite Get the direction mark opposite to your language
3078 function getDirMark( $opposite = false ) {
3079 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3080 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3082 return $this->isRTL() ?
$lrm : $rlm;
3084 return $this->isRTL() ?
$rlm : $lrm;
3090 function capitalizeAllNouns() {
3091 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3095 * An arrow, depending on the language direction.
3097 * @param string $direction The direction of the arrow: forwards (default),
3098 * backwards, left, right, up, down.
3101 function getArrow( $direction = 'forwards' ) {
3102 switch ( $direction ) {
3104 return $this->isRTL() ?
'←' : '→';
3106 return $this->isRTL() ?
'→' : '←';
3119 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3123 function linkPrefixExtension() {
3124 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3128 * Get all magic words from cache.
3131 function getMagicWords() {
3132 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3136 * Run the LanguageGetMagic hook once.
3138 protected function doMagicHook() {
3139 if ( $this->mMagicHookDone
) {
3142 $this->mMagicHookDone
= true;
3143 Hooks
::run( 'LanguageGetMagic', [ &$this->mMagicExtensions
, $this->getCode() ] );
3147 * Fill a MagicWord object with data from here
3149 * @param MagicWord $mw
3151 function getMagic( $mw ) {
3152 // Saves a function call
3153 if ( !$this->mMagicHookDone
) {
3154 $this->doMagicHook();
3157 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3158 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3160 $rawEntry = self
::$dataCache->getSubitem(
3161 $this->mCode
, 'magicWords', $mw->mId
);
3164 if ( !is_array( $rawEntry ) ) {
3165 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3167 $mw->mCaseSensitive
= $rawEntry[0];
3168 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3173 * Add magic words to the extension array
3175 * @param array $newWords
3177 function addMagicWordsByLang( $newWords ) {
3178 $fallbackChain = $this->getFallbackLanguages();
3179 $fallbackChain = array_reverse( $fallbackChain );
3180 foreach ( $fallbackChain as $code ) {
3181 if ( isset( $newWords[$code] ) ) {
3182 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3188 * Get special page names, as an associative array
3189 * canonical name => array of valid names, including aliases
3192 function getSpecialPageAliases() {
3193 // Cache aliases because it may be slow to load them
3194 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3196 $this->mExtendedSpecialPageAliases
=
3197 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3198 Hooks
::run( 'LanguageGetSpecialPageAliases',
3199 [ &$this->mExtendedSpecialPageAliases
, $this->getCode() ] );
3202 return $this->mExtendedSpecialPageAliases
;
3206 * Italic is unsuitable for some languages
3208 * @param string $text The text to be emphasized.
3211 function emphasize( $text ) {
3212 return "<em>$text</em>";
3216 * Normally we output all numbers in plain en_US style, that is
3217 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3218 * point twohundredthirtyfive. However this is not suitable for all
3219 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3220 * Icelandic just want to use commas instead of dots, and dots instead
3221 * of commas like "293.291,235".
3223 * An example of this function being called:
3225 * wfMessage( 'message' )->numParams( $num )->text()
3228 * See $separatorTransformTable on MessageIs.php for
3229 * the , => . and . => , implementation.
3231 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3232 * @param int|float $number The string to be formatted, should be an integer
3233 * or a floating point number.
3234 * @param bool $nocommafy Set to true for special numbers like dates
3237 public function formatNum( $number, $nocommafy = false ) {
3238 global $wgTranslateNumerals;
3239 if ( !$nocommafy ) {
3240 $number = $this->commafy( $number );
3241 $s = $this->separatorTransformTable();
3243 $number = strtr( $number, $s );
3247 if ( $wgTranslateNumerals ) {
3248 $s = $this->digitTransformTable();
3250 $number = strtr( $number, $s );
3258 * Front-end for non-commafied formatNum
3260 * @param int|float $number The string to be formatted, should be an integer
3261 * or a floating point number.
3265 public function formatNumNoSeparators( $number ) {
3266 return $this->formatNum( $number, true );
3270 * @param string $number
3273 public function parseFormattedNumber( $number ) {
3274 $s = $this->digitTransformTable();
3276 // eliminate empty array values such as ''. (bug 64347)
3277 $s = array_filter( $s );
3278 $number = strtr( $number, array_flip( $s ) );
3281 $s = $this->separatorTransformTable();
3283 // eliminate empty array values such as ''. (bug 64347)
3284 $s = array_filter( $s );
3285 $number = strtr( $number, array_flip( $s ) );
3288 $number = strtr( $number, [ ',' => '' ] );
3293 * Adds commas to a given number
3295 * @param mixed $number
3298 function commafy( $number ) {
3299 $digitGroupingPattern = $this->digitGroupingPattern();
3300 if ( $number === null ) {
3304 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3305 // default grouping is at thousands, use the same for ###,###,### pattern too.
3306 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3308 // Ref: http://cldr.unicode.org/translation/number-patterns
3310 if ( intval( $number ) < 0 ) {
3311 // For negative numbers apply the algorithm like positive number and add sign.
3313 $number = substr( $number, 1 );
3317 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3318 preg_match( "/\d+/", $number, $integerPart );
3319 preg_match( "/\.\d*/", $number, $decimalPart );
3320 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3321 if ( $groupedNumber === $number ) {
3322 // the string does not have any number part. Eg: .12345
3323 return $sign . $groupedNumber;
3325 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3326 while ( $start > 0 ) {
3327 $match = $matches[0][$numMatches - 1];
3328 $matchLen = strlen( $match );
3329 $start = $end - $matchLen;
3333 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3335 if ( $numMatches > 1 ) {
3336 // use the last pattern for the rest of the number
3340 $groupedNumber = "," . $groupedNumber;
3343 return $sign . $groupedNumber;
3350 function digitGroupingPattern() {
3351 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3357 function digitTransformTable() {
3358 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3364 function separatorTransformTable() {
3365 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3369 * Take a list of strings and build a locale-friendly comma-separated
3370 * list, using the local comma-separator message.
3371 * The last two strings are chained with an "and".
3372 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3374 * @param string[] $l
3377 function listToText( array $l ) {
3378 $m = count( $l ) - 1;
3383 $and = $this->msg( 'and' )->escaped();
3384 $space = $this->msg( 'word-separator' )->escaped();
3386 $comma = $this->msg( 'comma-separator' )->escaped();
3390 for ( $i = $m - 1; $i >= 0; $i-- ) {
3391 if ( $i == $m - 1 ) {
3392 $s = $l[$i] . $and . $space . $s;
3394 $s = $l[$i] . $comma . $s;
3401 * Take a list of strings and build a locale-friendly comma-separated
3402 * list, using the local comma-separator message.
3403 * @param string[] $list Array of strings to put in a comma list
3406 function commaList( array $list ) {
3408 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3414 * Take a list of strings and build a locale-friendly semicolon-separated
3415 * list, using the local semicolon-separator message.
3416 * @param string[] $list Array of strings to put in a semicolon list
3419 function semicolonList( array $list ) {
3421 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3427 * Same as commaList, but separate it with the pipe instead.
3428 * @param string[] $list Array of strings to put in a pipe list
3431 function pipeList( array $list ) {
3433 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3439 * Truncate a string to a specified length in bytes, appending an optional
3440 * string (e.g. for ellipses)
3442 * The database offers limited byte lengths for some columns in the database;
3443 * multi-byte character sets mean we need to ensure that only whole characters
3444 * are included, otherwise broken characters can be passed to the user
3446 * If $length is negative, the string will be truncated from the beginning
3448 * @param string $string String to truncate
3449 * @param int $length Maximum length (including ellipses)
3450 * @param string $ellipsis String to append to the truncated text
3451 * @param bool $adjustLength Subtract length of ellipsis from $length.
3452 * $adjustLength was introduced in 1.18, before that behaved as if false.
3455 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3456 # Use the localized ellipsis character
3457 if ( $ellipsis == '...' ) {
3458 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3460 # Check if there is no need to truncate
3461 if ( $length == 0 ) {
3462 return $ellipsis; // convention
3463 } elseif ( strlen( $string ) <= abs( $length ) ) {
3464 return $string; // no need to truncate
3466 $stringOriginal = $string;
3467 # If ellipsis length is >= $length then we can't apply $adjustLength
3468 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3469 $string = $ellipsis; // this can be slightly unexpected
3470 # Otherwise, truncate and add ellipsis...
3472 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3473 if ( $length > 0 ) {
3474 $length -= $eLength;
3475 $string = substr( $string, 0, $length ); // xyz...
3476 $string = $this->removeBadCharLast( $string );
3477 $string = rtrim( $string );
3478 $string = $string . $ellipsis;
3480 $length +
= $eLength;
3481 $string = substr( $string, $length ); // ...xyz
3482 $string = $this->removeBadCharFirst( $string );
3483 $string = ltrim( $string );
3484 $string = $ellipsis . $string;
3487 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3488 # This check is *not* redundant if $adjustLength, due to the single case where
3489 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3490 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3493 return $stringOriginal;
3498 * Remove bytes that represent an incomplete Unicode character
3499 * at the end of string (e.g. bytes of the char are missing)
3501 * @param string $string
3504 protected function removeBadCharLast( $string ) {
3505 if ( $string != '' ) {
3506 $char = ord( $string[strlen( $string ) - 1] );
3508 if ( $char >= 0xc0 ) {
3509 # We got the first byte only of a multibyte char; remove it.
3510 $string = substr( $string, 0, -1 );
3511 } elseif ( $char >= 0x80 &&
3512 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3513 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3514 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3516 # We chopped in the middle of a character; remove it
3524 * Remove bytes that represent an incomplete Unicode character
3525 * at the start of string (e.g. bytes of the char are missing)
3527 * @param string $string
3530 protected function removeBadCharFirst( $string ) {
3531 if ( $string != '' ) {
3532 $char = ord( $string[0] );
3533 if ( $char >= 0x80 && $char < 0xc0 ) {
3534 # We chopped in the middle of a character; remove the whole thing
3535 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3542 * Truncate a string of valid HTML to a specified length in bytes,
3543 * appending an optional string (e.g. for ellipses), and return valid HTML
3545 * This is only intended for styled/linked text, such as HTML with
3546 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3547 * Also, this will not detect things like "display:none" CSS.
3549 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3551 * @param string $text HTML string to truncate
3552 * @param int $length (zero/positive) Maximum length (including ellipses)
3553 * @param string $ellipsis String to append to the truncated text
3556 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3557 # Use the localized ellipsis character
3558 if ( $ellipsis == '...' ) {
3559 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3561 # Check if there is clearly no need to truncate
3562 if ( $length <= 0 ) {
3563 return $ellipsis; // no text shown, nothing to format (convention)
3564 } elseif ( strlen( $text ) <= $length ) {
3565 return $text; // string short enough even *with* HTML (short-circuit)
3568 $dispLen = 0; // innerHTML legth so far
3569 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3570 $tagType = 0; // 0-open, 1-close
3571 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3572 $entityState = 0; // 0-not entity, 1-entity
3573 $tag = $ret = ''; // accumulated tag name, accumulated result string
3574 $openTags = []; // open tag stack
3575 $maybeState = null; // possible truncation state
3577 $textLen = strlen( $text );
3578 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3579 for ( $pos = 0; true; ++
$pos ) {
3580 # Consider truncation once the display length has reached the maximim.
3581 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3582 # Check that we're not in the middle of a bracket/entity...
3583 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3584 if ( !$testingEllipsis ) {
3585 $testingEllipsis = true;
3586 # Save where we are; we will truncate here unless there turn out to
3587 # be so few remaining characters that truncation is not necessary.
3588 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3589 $maybeState = [ $ret, $openTags ]; // save state
3591 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3592 # String in fact does need truncation, the truncation point was OK.
3593 list( $ret, $openTags ) = $maybeState; // reload state
3594 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3595 $ret .= $ellipsis; // add ellipsis
3599 if ( $pos >= $textLen ) {
3600 break; // extra iteration just for above checks
3603 # Read the next char...
3605 $lastCh = $pos ?
$text[$pos - 1] : '';
3606 $ret .= $ch; // add to result string
3608 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3609 $entityState = 0; // for bad HTML
3610 $bracketState = 1; // tag started (checking for backslash)
3611 } elseif ( $ch == '>' ) {
3612 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3613 $entityState = 0; // for bad HTML
3614 $bracketState = 0; // out of brackets
3615 } elseif ( $bracketState == 1 ) {
3617 $tagType = 1; // close tag (e.g. "</span>")
3619 $tagType = 0; // open tag (e.g. "<span>")
3622 $bracketState = 2; // building tag name
3623 } elseif ( $bracketState == 2 ) {
3627 // Name found (e.g. "<a href=..."), add on tag attributes...
3628 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3630 } elseif ( $bracketState == 0 ) {
3631 if ( $entityState ) {
3634 $dispLen++
; // entity is one displayed char
3637 if ( $neLength == 0 && !$maybeState ) {
3638 // Save state without $ch. We want to *hit* the first
3639 // display char (to get tags) but not *use* it if truncating.
3640 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3643 $entityState = 1; // entity found, (e.g. " ")
3645 $dispLen++
; // this char is displayed
3646 // Add the next $max display text chars after this in one swoop...
3647 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3648 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3649 $dispLen +
= $skipped;
3655 // Close the last tag if left unclosed by bad HTML
3656 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3657 while ( count( $openTags ) > 0 ) {
3658 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3664 * truncateHtml() helper function
3665 * like strcspn() but adds the skipped chars to $ret
3667 * @param string $ret
3668 * @param string $text
3669 * @param string $search
3671 * @param null|int $len
3674 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3675 if ( $len === null ) {
3676 $len = -1; // -1 means "no limit" for strcspn
3677 } elseif ( $len < 0 ) {
3681 if ( $start < strlen( $text ) ) {
3682 $skipCount = strcspn( $text, $search, $start, $len );
3683 $ret .= substr( $text, $start, $skipCount );
3689 * truncateHtml() helper function
3690 * (a) push or pop $tag from $openTags as needed
3691 * (b) clear $tag value
3692 * @param string &$tag Current HTML tag name we are looking at
3693 * @param int $tagType (0-open tag, 1-close tag)
3694 * @param string $lastCh Character before the '>' that ended this tag
3695 * @param array &$openTags Open tag stack (not accounting for $tag)
3697 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3698 $tag = ltrim( $tag );
3700 if ( $tagType == 0 && $lastCh != '/' ) {
3701 $openTags[] = $tag; // tag opened (didn't close itself)
3702 } elseif ( $tagType == 1 ) {
3703 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3704 array_pop( $openTags ); // tag closed
3712 * Grammatical transformations, needed for inflected languages
3713 * Invoked by putting {{grammar:case|word}} in a message
3715 * @param string $word
3716 * @param string $case
3719 function convertGrammar( $word, $case ) {
3720 global $wgGrammarForms;
3721 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3722 return $wgGrammarForms[$this->getCode()][$case][$word];
3728 * Get the grammar forms for the content language
3729 * @return array Array of grammar forms
3732 function getGrammarForms() {
3733 global $wgGrammarForms;
3734 if ( isset( $wgGrammarForms[$this->getCode()] )
3735 && is_array( $wgGrammarForms[$this->getCode()] )
3737 return $wgGrammarForms[$this->getCode()];
3743 * Provides an alternative text depending on specified gender.
3744 * Usage {{gender:username|masculine|feminine|unknown}}.
3745 * username is optional, in which case the gender of current user is used,
3746 * but only in (some) interface messages; otherwise default gender is used.
3748 * If no forms are given, an empty string is returned. If only one form is
3749 * given, it will be returned unconditionally. These details are implied by
3750 * the caller and cannot be overridden in subclasses.
3752 * If three forms are given, the default is to use the third (unknown) form.
3753 * If fewer than three forms are given, the default is to use the first (masculine) form.
3754 * These details can be overridden in subclasses.
3756 * @param string $gender
3757 * @param array $forms
3761 function gender( $gender, $forms ) {
3762 if ( !count( $forms ) ) {
3765 $forms = $this->preConvertPlural( $forms, 2 );
3766 if ( $gender === 'male' ) {
3769 if ( $gender === 'female' ) {
3772 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3776 * Plural form transformations, needed for some languages.
3777 * For example, there are 3 form of plural in Russian and Polish,
3778 * depending on "count mod 10". See [[w:Plural]]
3779 * For English it is pretty simple.
3781 * Invoked by putting {{plural:count|wordform1|wordform2}}
3782 * or {{plural:count|wordform1|wordform2|wordform3}}
3784 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3786 * @param int $count Non-localized number
3787 * @param array $forms Different plural forms
3788 * @return string Correct form of plural for $count in this language
3790 function convertPlural( $count, $forms ) {
3791 // Handle explicit n=pluralform cases
3792 $forms = $this->handleExplicitPluralForms( $count, $forms );
3793 if ( is_string( $forms ) ) {
3796 if ( !count( $forms ) ) {
3800 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3801 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3802 return $forms[$pluralForm];
3806 * Handles explicit plural forms for Language::convertPlural()
3808 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3809 * If an explicitly defined plural form matches the $count, then
3810 * string value returned, otherwise array returned for further consideration
3811 * by CLDR rules or overridden convertPlural().
3815 * @param int $count Non-localized number
3816 * @param array $forms Different plural forms
3818 * @return array|string
3820 protected function handleExplicitPluralForms( $count, array $forms ) {
3821 foreach ( $forms as $index => $form ) {
3822 if ( preg_match( '/\d+=/i', $form ) ) {
3823 $pos = strpos( $form, '=' );
3824 if ( substr( $form, 0, $pos ) === (string)$count ) {
3825 return substr( $form, $pos +
1 );
3827 unset( $forms[$index] );
3830 return array_values( $forms );
3834 * Checks that convertPlural was given an array and pads it to requested
3835 * amount of forms by copying the last one.
3837 * @param array $forms Array of forms given to convertPlural
3838 * @param int $count How many forms should there be at least
3839 * @return array Padded array of forms or an exception if not an array
3841 protected function preConvertPlural( /* Array */ $forms, $count ) {
3842 while ( count( $forms ) < $count ) {
3843 $forms[] = $forms[count( $forms ) - 1];
3849 * Wraps argument with unicode control characters for directionality safety
3851 * This solves the problem where directionality-neutral characters at the edge of
3852 * the argument string get interpreted with the wrong directionality from the
3853 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3855 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3856 * directionality of the argument string, using the BIDI algorithm's own "First
3857 * strong directional codepoint" rule. Essentially, this works round the fact that
3858 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3859 * direction inference). The latter is cleaner but still not widely supported.
3861 * @param string $text Text to wrap
3862 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3864 public function embedBidi( $text = '' ) {
3865 $dir = Language
::strongDirFromContent( $text );
3866 if ( $dir === 'ltr' ) {
3867 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3868 return self
::$lre . $text . self
::$pdf;
3870 if ( $dir === 'rtl' ) {
3871 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3872 return self
::$rle . $text . self
::$pdf;
3874 // No strong directionality: do not wrap
3879 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3880 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3881 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3882 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3885 * @param string $str The validated block duration in English
3886 * @param User $user User object to use timezone from or null for $wgUser
3887 * @return string Somehow translated block duration
3888 * @see LanguageFi.php for example implementation
3890 function translateBlockExpiry( $str, User
$user = null ) {
3891 $duration = SpecialBlock
::getSuggestedDurations( $this );
3892 foreach ( $duration as $show => $value ) {
3893 if ( strcmp( $str, $value ) == 0 ) {
3894 return htmlspecialchars( trim( $show ) );
3898 if ( wfIsInfinity( $str ) ) {
3899 foreach ( $duration as $show => $value ) {
3900 if ( wfIsInfinity( $value ) ) {
3901 return htmlspecialchars( trim( $show ) );
3906 // If all else fails, return a standard duration or timestamp description.
3907 $time = strtotime( $str, 0 );
3908 if ( $time === false ) { // Unknown format. Return it as-is in case.
3910 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3911 // $time is relative to 0 so it's a duration length.
3912 return $this->formatDuration( $time );
3913 } else { // It's an absolute timestamp.
3914 if ( $time === 0 ) {
3915 // wfTimestamp() handles 0 as current time instead of epoch.
3916 $time = '19700101000000';
3919 return $this->userTimeAndDate( $time, $user );
3921 return $this->timeanddate( $time );
3926 * languages like Chinese need to be segmented in order for the diff
3929 * @param string $text
3932 public function segmentForDiff( $text ) {
3937 * and unsegment to show the result
3939 * @param string $text
3942 public function unsegmentForDiff( $text ) {
3947 * Return the LanguageConverter used in the Language
3950 * @return LanguageConverter
3952 public function getConverter() {
3953 return $this->mConverter
;
3957 * convert text to all supported variants
3959 * @param string $text
3962 public function autoConvertToAllVariants( $text ) {
3963 return $this->mConverter
->autoConvertToAllVariants( $text );
3967 * convert text to different variants of a language.
3969 * @param string $text
3972 public function convert( $text ) {
3973 return $this->mConverter
->convert( $text );
3977 * Convert a Title object to a string in the preferred variant
3979 * @param Title $title
3982 public function convertTitle( $title ) {
3983 return $this->mConverter
->convertTitle( $title );
3987 * Convert a namespace index to a string in the preferred variant
3992 public function convertNamespace( $ns ) {
3993 return $this->mConverter
->convertNamespace( $ns );
3997 * Check if this is a language with variants
4001 public function hasVariants() {
4002 return count( $this->getVariants() ) > 1;
4006 * Check if the language has the specific variant
4009 * @param string $variant
4012 public function hasVariant( $variant ) {
4013 return (bool)$this->mConverter
->validateVariant( $variant );
4017 * Perform output conversion on a string, and encode for safe HTML output.
4018 * @param string $text Text to be converted
4019 * @param bool $isTitle Whether this conversion is for the article title
4021 * @todo this should get integrated somewhere sane
4023 public function convertHtml( $text, $isTitle = false ) {
4024 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4028 * @param string $key
4031 public function convertCategoryKey( $key ) {
4032 return $this->mConverter
->convertCategoryKey( $key );
4036 * Get the list of variants supported by this language
4037 * see sample implementation in LanguageZh.php
4039 * @return array An array of language codes
4041 public function getVariants() {
4042 return $this->mConverter
->getVariants();
4048 public function getPreferredVariant() {
4049 return $this->mConverter
->getPreferredVariant();
4055 public function getDefaultVariant() {
4056 return $this->mConverter
->getDefaultVariant();
4062 public function getURLVariant() {
4063 return $this->mConverter
->getURLVariant();
4067 * If a language supports multiple variants, it is
4068 * possible that non-existing link in one variant
4069 * actually exists in another variant. this function
4070 * tries to find it. See e.g. LanguageZh.php
4071 * The input parameters may be modified upon return
4073 * @param string &$link The name of the link
4074 * @param Title &$nt The title object of the link
4075 * @param bool $ignoreOtherCond To disable other conditions when
4076 * we need to transclude a template or update a category's link
4078 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4079 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4083 * returns language specific options used by User::getPageRenderHash()
4084 * for example, the preferred language variant
4088 function getExtraHashOptions() {
4089 return $this->mConverter
->getExtraHashOptions();
4093 * For languages that support multiple variants, the title of an
4094 * article may be displayed differently in different variants. this
4095 * function returns the apporiate title defined in the body of the article.
4099 public function getParsedTitle() {
4100 return $this->mConverter
->getParsedTitle();
4104 * Refresh the cache of conversion tables when
4105 * MediaWiki:Conversiontable* is updated.
4107 * @param Title $title The Title of the page being updated
4109 public function updateConversionTable( Title
$title ) {
4110 $this->mConverter
->updateConversionTable( $title );
4114 * Prepare external link text for conversion. When the text is
4115 * a URL, it shouldn't be converted, and it'll be wrapped in
4116 * the "raw" tag (-{R| }-) to prevent conversion.
4118 * This function is called "markNoConversion" for historical
4121 * @param string $text Text to be used for external link
4122 * @param bool $noParse Wrap it without confirming it's a real URL first
4123 * @return string The tagged text
4125 public function markNoConversion( $text, $noParse = false ) {
4126 // Excluding protocal-relative URLs may avoid many false positives.
4127 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4128 return $this->mConverter
->markNoConversion( $text );
4135 * A regular expression to match legal word-trailing characters
4136 * which should be merged onto a link of the form [[foo]]bar.
4140 public function linkTrail() {
4141 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4145 * A regular expression character set to match legal word-prefixing
4146 * characters which should be merged onto a link of the form foo[[bar]].
4150 public function linkPrefixCharset() {
4151 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4155 * Get the "parent" language which has a converter to convert a "compatible" language
4156 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4158 * @return Language|null
4161 public function getParentLanguage() {
4162 if ( $this->mParentLanguage
!== false ) {
4163 return $this->mParentLanguage
;
4166 $code = explode( '-', $this->getCode() )[0];
4167 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4168 $this->mParentLanguage
= null;
4171 $lang = Language
::factory( $code );
4172 if ( !$lang->hasVariant( $this->getCode() ) ) {
4173 $this->mParentLanguage
= null;
4177 $this->mParentLanguage
= $lang;
4182 * Compare with an other language object
4185 * @param Language $lang
4188 public function equals( Language
$lang ) {
4189 return $lang->getCode() === $this->mCode
;
4193 * Get the internal language code for this language object
4195 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4196 * htmlspecialchars() or similar
4200 public function getCode() {
4201 return $this->mCode
;
4205 * Get the code in BCP 47 format which we can use
4206 * inside of html lang="" tags.
4208 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4209 * htmlspecialchars() or similar.
4214 public function getHtmlCode() {
4215 if ( is_null( $this->mHtmlCode
) ) {
4216 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4218 return $this->mHtmlCode
;
4222 * @param string $code
4224 public function setCode( $code ) {
4225 $this->mCode
= $code;
4226 // Ensure we don't leave incorrect cached data lying around
4227 $this->mHtmlCode
= null;
4228 $this->mParentLanguage
= false;
4232 * Get the language code from a file name. Inverse of getFileName()
4233 * @param string $filename $prefix . $languageCode . $suffix
4234 * @param string $prefix Prefix before the language code
4235 * @param string $suffix Suffix after the language code
4236 * @return string Language code, or false if $prefix or $suffix isn't found
4238 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4240 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4241 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4242 if ( !count( $m ) ) {
4245 return str_replace( '_', '-', strtolower( $m[1] ) );
4249 * @param string $code
4250 * @return string Name of the language class
4252 public static function classFromCode( $code ) {
4253 if ( $code == 'en' ) {
4256 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4261 * Get the name of a file for a certain language code
4262 * @param string $prefix Prepend this to the filename
4263 * @param string $code Language code
4264 * @param string $suffix Append this to the filename
4265 * @throws MWException
4266 * @return string $prefix . $mangledCode . $suffix
4268 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4269 if ( !self
::isValidBuiltInCode( $code ) ) {
4270 throw new MWException( "Invalid language code \"$code\"" );
4273 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4277 * @param string $code
4280 public static function getMessagesFileName( $code ) {
4282 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4283 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4288 * @param string $code
4290 * @throws MWException
4293 public static function getJsonMessagesFileName( $code ) {
4296 if ( !self
::isValidBuiltInCode( $code ) ) {
4297 throw new MWException( "Invalid language code \"$code\"" );
4300 return "$IP/languages/i18n/$code.json";
4304 * Get the first fallback for a given language.
4306 * @param string $code
4308 * @return bool|string
4310 public static function getFallbackFor( $code ) {
4311 $fallbacks = self
::getFallbacksFor( $code );
4313 return $fallbacks[0];
4319 * Get the ordered list of fallback languages.
4322 * @param string $code Language code
4323 * @return array Non-empty array, ending in "en"
4325 public static function getFallbacksFor( $code ) {
4326 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4329 // For unknown languages, fallbackSequence returns an empty array,
4330 // hardcode fallback to 'en' in that case.
4331 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4335 * Get the ordered list of fallback languages, ending with the fallback
4336 * language chain for the site language.
4339 * @param string $code Language code
4340 * @return array Array( fallbacks, site fallbacks )
4342 public static function getFallbacksIncludingSiteLanguage( $code ) {
4343 global $wgLanguageCode;
4345 // Usually, we will only store a tiny number of fallback chains, so we
4346 // keep them in static memory.
4347 $cacheKey = "{$code}-{$wgLanguageCode}";
4349 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4350 $fallbacks = self
::getFallbacksFor( $code );
4352 // Append the site's fallback chain, including the site language itself
4353 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4354 array_unshift( $siteFallbacks, $wgLanguageCode );
4356 // Eliminate any languages already included in the chain
4357 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4359 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4361 return self
::$fallbackLanguageCache[$cacheKey];
4365 * Get all messages for a given language
4366 * WARNING: this may take a long time. If you just need all message *keys*
4367 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4369 * @param string $code
4373 public static function getMessagesFor( $code ) {
4374 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4378 * Get a message for a given language
4380 * @param string $key
4381 * @param string $code
4385 public static function getMessageFor( $key, $code ) {
4386 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4390 * Get all message keys for a given language. This is a faster alternative to
4391 * array_keys( Language::getMessagesFor( $code ) )
4394 * @param string $code Language code
4395 * @return array Array of message keys (strings)
4397 public static function getMessageKeysFor( $code ) {
4398 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4402 * @param string $talk
4405 function fixVariableInNamespace( $talk ) {
4406 if ( strpos( $talk, '$1' ) === false ) {
4410 global $wgMetaNamespace;
4411 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4413 # Allow grammar transformations
4414 # Allowing full message-style parsing would make simple requests
4415 # such as action=raw much more expensive than they need to be.
4416 # This will hopefully cover most cases.
4417 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4418 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4419 return str_replace( ' ', '_', $talk );
4426 function replaceGrammarInNamespace( $m ) {
4427 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4431 * Decode an expiry (block, protection, etc) which has come from the DB
4433 * @param string $expiry Database expiry String
4434 * @param bool|int $format True to process using language functions, or TS_ constant
4435 * to return the expiry in a given timestamp
4436 * @param string $infinity If $format is not true, use this string for infinite expiry
4440 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4442 if ( $dbInfinity === null ) {
4443 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4446 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4447 return $format === true
4448 ?
$this->getMessageFromDB( 'infiniteblock' )
4451 return $format === true
4452 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4453 : wfTimestamp( $format, $expiry );
4459 * @param int|float $seconds
4460 * @param array $format Optional
4461 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4462 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4463 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4465 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4466 * or 'avoidminutes'.
4469 function formatTimePeriod( $seconds, $format = [] ) {
4470 if ( !is_array( $format ) ) {
4471 $format = [ 'avoid' => $format ]; // For backwards compatibility
4473 if ( !isset( $format['avoid'] ) ) {
4474 $format['avoid'] = false;
4476 if ( !isset( $format['noabbrevs'] ) ) {
4477 $format['noabbrevs'] = false;
4479 $secondsMsg = wfMessage(
4480 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4481 $minutesMsg = wfMessage(
4482 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4483 $hoursMsg = wfMessage(
4484 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4485 $daysMsg = wfMessage(
4486 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4488 if ( round( $seconds * 10 ) < 100 ) {
4489 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4490 $s = $secondsMsg->params( $s )->text();
4491 } elseif ( round( $seconds ) < 60 ) {
4492 $s = $this->formatNum( round( $seconds ) );
4493 $s = $secondsMsg->params( $s )->text();
4494 } elseif ( round( $seconds ) < 3600 ) {
4495 $minutes = floor( $seconds / 60 );
4496 $secondsPart = round( fmod( $seconds, 60 ) );
4497 if ( $secondsPart == 60 ) {
4501 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4503 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4504 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4505 $hours = floor( $seconds / 3600 );
4506 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4507 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4508 if ( $secondsPart == 60 ) {
4512 if ( $minutes == 60 ) {
4516 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4518 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4519 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4520 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4523 $days = floor( $seconds / 86400 );
4524 if ( $format['avoid'] === 'avoidminutes' ) {
4525 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4526 if ( $hours == 24 ) {
4530 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4532 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4533 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4534 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4535 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4536 if ( $minutes == 60 ) {
4540 if ( $hours == 24 ) {
4544 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4546 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4548 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4550 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4552 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4559 * Format a bitrate for output, using an appropriate
4560 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4561 * the magnitude in question.
4563 * This use base 1000. For base 1024 use formatSize(), for another base
4564 * see formatComputingNumbers().
4569 function formatBitrate( $bps ) {
4570 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4574 * @param int $size Size of the unit
4575 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4576 * @param string $messageKey Message key to be uesd
4579 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4581 return str_replace( '$1', $this->formatNum( $size ),
4582 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4585 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4588 $maxIndex = count( $sizes ) - 1;
4589 while ( $size >= $boundary && $index < $maxIndex ) {
4594 // For small sizes no decimal places necessary
4597 // For MB and bigger two decimal places are smarter
4600 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4602 $size = round( $size, $round );
4603 $text = $this->getMessageFromDB( $msg );
4604 return str_replace( '$1', $this->formatNum( $size ), $text );
4608 * Format a size in bytes for output, using an appropriate
4609 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4611 * This method use base 1024. For base 1000 use formatBitrate(), for
4612 * another base see formatComputingNumbers()
4614 * @param int $size Size to format
4615 * @return string Plain text (not HTML)
4617 function formatSize( $size ) {
4618 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4622 * Make a list item, used by various special pages
4624 * @param string $page Page link
4625 * @param string $details HTML safe text between brackets
4626 * @param bool $oppositedm Add the direction mark opposite to your
4627 * language, to display text properly
4628 * @return HTML escaped string
4630 function specialList( $page, $details, $oppositedm = true ) {
4635 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4639 $this->msg( 'word-separator' )->escaped() .
4640 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4644 * Generate (prev x| next x) (20|50|100...) type links for paging
4646 * @param Title $title Title object to link
4647 * @param int $offset
4649 * @param array $query Optional URL query parameter string
4650 * @param bool $atend Optional param for specified if this is the last page
4653 public function viewPrevNext( Title
$title, $offset, $limit,
4654 array $query = [], $atend = false
4656 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4658 # Make 'previous' link
4659 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4660 if ( $offset > 0 ) {
4661 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4662 $query, $prev, 'prevn-title', 'mw-prevlink' );
4664 $plink = htmlspecialchars( $prev );
4668 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4670 $nlink = htmlspecialchars( $next );
4672 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4673 $query, $next, 'nextn-title', 'mw-nextlink' );
4676 # Make links to set number of items per page
4678 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4679 $numLinks[] = $this->numLink( $title, $offset, $num,
4680 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4683 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4684 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4688 * Helper function for viewPrevNext() that generates links
4690 * @param Title $title Title object to link
4691 * @param int $offset
4693 * @param array $query Extra query parameters
4694 * @param string $link Text to use for the link; will be escaped
4695 * @param string $tooltipMsg Name of the message to use as tooltip
4696 * @param string $class Value of the "class" attribute of the link
4697 * @return string HTML fragment
4699 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4702 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4703 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4704 ->numParams( $limit )->text();
4706 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4707 'title' => $tooltip, 'class' => $class ], $link );
4711 * Get the conversion rule title, if any.
4715 public function getConvRuleTitle() {
4716 return $this->mConverter
->getConvRuleTitle();
4720 * Get the compiled plural rules for the language
4722 * @return array Associative array with plural form, and plural rule as key-value pairs
4724 public function getCompiledPluralRules() {
4725 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4726 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4727 if ( !$pluralRules ) {
4728 foreach ( $fallbacks as $fallbackCode ) {
4729 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4730 if ( $pluralRules ) {
4735 return $pluralRules;
4739 * Get the plural rules for the language
4741 * @return array Associative array with plural form number and plural rule as key-value pairs
4743 public function getPluralRules() {
4744 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4745 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4746 if ( !$pluralRules ) {
4747 foreach ( $fallbacks as $fallbackCode ) {
4748 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4749 if ( $pluralRules ) {
4754 return $pluralRules;
4758 * Get the plural rule types for the language
4760 * @return array Associative array with plural form number and plural rule type as key-value pairs
4762 public function getPluralRuleTypes() {
4763 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4764 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4765 if ( !$pluralRuleTypes ) {
4766 foreach ( $fallbacks as $fallbackCode ) {
4767 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4768 if ( $pluralRuleTypes ) {
4773 return $pluralRuleTypes;
4777 * Find the index number of the plural rule appropriate for the given number
4778 * @param int $number
4779 * @return int The index number of the plural rule
4781 public function getPluralRuleIndexNumber( $number ) {
4782 $pluralRules = $this->getCompiledPluralRules();
4783 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
4788 * Find the plural rule type appropriate for the given number
4789 * For example, if the language is set to Arabic, getPluralType(5) should
4792 * @param int $number
4793 * @return string The name of the plural rule type, e.g. one, two, few, many
4795 public function getPluralRuleType( $number ) {
4796 $index = $this->getPluralRuleIndexNumber( $number );
4797 $pluralRuleTypes = $this->getPluralRuleTypes();
4798 if ( isset( $pluralRuleTypes[$index] ) ) {
4799 return $pluralRuleTypes[$index];