3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
26 * @defgroup Language Language
29 use CLDRPluralRuleParser\Evaluator
;
32 * Internationalisation code
37 * @var LanguageConverter
41 public $mVariants, $mCode, $mLoaded = false;
42 public $mMagicExtensions = [], $mMagicHookDone = false;
43 private $mHtmlCode = null, $mParentLanguage = false;
45 public $dateFormatStrings = [];
46 public $mExtendedSpecialPageAliases;
48 /** @var array|null */
49 protected $namespaceNames;
50 protected $mNamespaceIds, $namespaceAliases;
53 * ReplacementArray object caches
55 public $transformData = [];
58 * @var LocalisationCache
60 static public $dataCache;
62 static public $mLangObjCache = [];
64 static public $mWeekdayMsgs = [
65 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
69 static public $mWeekdayAbbrevMsgs = [
70 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
73 static public $mMonthMsgs = [
74 'january', 'february', 'march', 'april', 'may_long', 'june',
75 'july', 'august', 'september', 'october', 'november',
78 static public $mMonthGenMsgs = [
79 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
80 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
83 static public $mMonthAbbrevMsgs = [
84 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
85 'sep', 'oct', 'nov', 'dec'
88 static public $mIranianCalendarMonthMsgs = [
89 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
90 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
91 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
92 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
95 static public $mHebrewCalendarMonthMsgs = [
96 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
97 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
98 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
99 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
100 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
103 static public $mHebrewCalendarMonthGenMsgs = [
104 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
105 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
106 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
107 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
108 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
111 static public $mHijriCalendarMonthMsgs = [
112 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
113 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
114 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
115 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
122 static public $durationIntervals = [
123 'millennia' => 31556952000,
124 'centuries' => 3155695200,
125 'decades' => 315569520,
126 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
135 * Cache for language fallbacks.
136 * @see Language::getFallbacksIncludingSiteLanguage
140 static private $fallbackLanguageCache = [];
143 * Cache for grammar rules data
144 * @var MapCacheLRU|null
146 static private $grammarTransformations;
149 * Cache for language names
150 * @var HashBagOStuff|null
152 static private $languageNameCache;
155 * Unicode directional formatting characters, for embedBidi()
157 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
158 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
159 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
162 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
163 * - in group 1 if it is LTR
164 * - in group 2 if it is RTL
165 * Does not match if there is no strong directionality codepoint.
167 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
169 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
170 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
172 // @codeCoverageIgnoreStart
173 // phpcs:ignore Generic.Files.LineLength
174 static private $strongDirRegex = '/(?:([\x{41}-\x{5a}\x{61}-\x{7a}\x{aa}\x{b5}\x{ba}\x{c0}-\x{d6}\x{d8}-\x{f6}\x{f8}-\x{2b8}\x{2bb}-\x{2c1}\x{2d0}\x{2d1}\x{2e0}-\x{2e4}\x{2ee}\x{370}-\x{373}\x{376}\x{377}\x{37a}-\x{37d}\x{37f}\x{386}\x{388}-\x{38a}\x{38c}\x{38e}-\x{3a1}\x{3a3}-\x{3f5}\x{3f7}-\x{482}\x{48a}-\x{52f}\x{531}-\x{556}\x{559}-\x{55f}\x{561}-\x{587}\x{589}\x{903}-\x{939}\x{93b}\x{93d}-\x{940}\x{949}-\x{94c}\x{94e}-\x{950}\x{958}-\x{961}\x{964}-\x{980}\x{982}\x{983}\x{985}-\x{98c}\x{98f}\x{990}\x{993}-\x{9a8}\x{9aa}-\x{9b0}\x{9b2}\x{9b6}-\x{9b9}\x{9bd}-\x{9c0}\x{9c7}\x{9c8}\x{9cb}\x{9cc}\x{9ce}\x{9d7}\x{9dc}\x{9dd}\x{9df}-\x{9e1}\x{9e6}-\x{9f1}\x{9f4}-\x{9fa}\x{a03}\x{a05}-\x{a0a}\x{a0f}\x{a10}\x{a13}-\x{a28}\x{a2a}-\x{a30}\x{a32}\x{a33}\x{a35}\x{a36}\x{a38}\x{a39}\x{a3e}-\x{a40}\x{a59}-\x{a5c}\x{a5e}\x{a66}-\x{a6f}\x{a72}-\x{a74}\x{a83}\x{a85}-\x{a8d}\x{a8f}-\x{a91}\x{a93}-\x{aa8}\x{aaa}-\x{ab0}\x{ab2}\x{ab3}\x{ab5}-\x{ab9}\x{abd}-\x{ac0}\x{ac9}\x{acb}\x{acc}\x{ad0}\x{ae0}\x{ae1}\x{ae6}-\x{af0}\x{af9}\x{b02}\x{b03}\x{b05}-\x{b0c}\x{b0f}\x{b10}\x{b13}-\x{b28}\x{b2a}-\x{b30}\x{b32}\x{b33}\x{b35}-\x{b39}\x{b3d}\x{b3e}\x{b40}\x{b47}\x{b48}\x{b4b}\x{b4c}\x{b57}\x{b5c}\x{b5d}\x{b5f}-\x{b61}\x{b66}-\x{b77}\x{b83}\x{b85}-\x{b8a}\x{b8e}-\x{b90}\x{b92}-\x{b95}\x{b99}\x{b9a}\x{b9c}\x{b9e}\x{b9f}\x{ba3}\x{ba4}\x{ba8}-\x{baa}\x{bae}-\x{bb9}\x{bbe}\x{bbf}\x{bc1}\x{bc2}\x{bc6}-\x{bc8}\x{bca}-\x{bcc}\x{bd0}\x{bd7}\x{be6}-\x{bf2}\x{c01}-\x{c03}\x{c05}-\x{c0c}\x{c0e}-\x{c10}\x{c12}-\x{c28}\x{c2a}-\x{c39}\x{c3d}\x{c41}-\x{c44}\x{c58}-\x{c5a}\x{c60}\x{c61}\x{c66}-\x{c6f}\x{c7f}\x{c82}\x{c83}\x{c85}-\x{c8c}\x{c8e}-\x{c90}\x{c92}-\x{ca8}\x{caa}-\x{cb3}\x{cb5}-\x{cb9}\x{cbd}-\x{cc4}\x{cc6}-\x{cc8}\x{cca}\x{ccb}\x{cd5}\x{cd6}\x{cde}\x{ce0}\x{ce1}\x{ce6}-\x{cef}\x{cf1}\x{cf2}\x{d02}\x{d03}\x{d05}-\x{d0c}\x{d0e}-\x{d10}\x{d12}-\x{d3a}\x{d3d}-\x{d40}\x{d46}-\x{d48}\x{d4a}-\x{d4c}\x{d4e}\x{d57}\x{d5f}-\x{d61}\x{d66}-\x{d75}\x{d79}-\x{d7f}\x{d82}\x{d83}\x{d85}-\x{d96}\x{d9a}-\x{db1}\x{db3}-\x{dbb}\x{dbd}\x{dc0}-\x{dc6}\x{dcf}-\x{dd1}\x{dd8}-\x{ddf}\x{de6}-\x{def}\x{df2}-\x{df4}\x{e01}-\x{e30}\x{e32}\x{e33}\x{e40}-\x{e46}\x{e4f}-\x{e5b}\x{e81}\x{e82}\x{e84}\x{e87}\x{e88}\x{e8a}\x{e8d}\x{e94}-\x{e97}\x{e99}-\x{e9f}\x{ea1}-\x{ea3}\x{ea5}\x{ea7}\x{eaa}\x{eab}\x{ead}-\x{eb0}\x{eb2}\x{eb3}\x{ebd}\x{ec0}-\x{ec4}\x{ec6}\x{ed0}-\x{ed9}\x{edc}-\x{edf}\x{f00}-\x{f17}\x{f1a}-\x{f34}\x{f36}\x{f38}\x{f3e}-\x{f47}\x{f49}-\x{f6c}\x{f7f}\x{f85}\x{f88}-\x{f8c}\x{fbe}-\x{fc5}\x{fc7}-\x{fcc}\x{fce}-\x{fda}\x{1000}-\x{102c}\x{1031}\x{1038}\x{103b}\x{103c}\x{103f}-\x{1057}\x{105a}-\x{105d}\x{1061}-\x{1070}\x{1075}-\x{1081}\x{1083}\x{1084}\x{1087}-\x{108c}\x{108e}-\x{109c}\x{109e}-\x{10c5}\x{10c7}\x{10cd}\x{10d0}-\x{1248}\x{124a}-\x{124d}\x{1250}-\x{1256}\x{1258}\x{125a}-\x{125d}\x{1260}-\x{1288}\x{128a}-\x{128d}\x{1290}-\x{12b0}\x{12b2}-\x{12b5}\x{12b8}-\x{12be}\x{12c0}\x{12c2}-\x{12c5}\x{12c8}-\x{12d6}\x{12d8}-\x{1310}\x{1312}-\x{1315}\x{1318}-\x{135a}\x{1360}-\x{137c}\x{1380}-\x{138f}\x{13a0}-\x{13f5}\x{13f8}-\x{13fd}\x{1401}-\x{167f}\x{1681}-\x{169a}\x{16a0}-\x{16f8}\x{1700}-\x{170c}\x{170e}-\x{1711}\x{1720}-\x{1731}\x{1735}\x{1736}\x{1740}-\x{1751}\x{1760}-\x{176c}\x{176e}-\x{1770}\x{1780}-\x{17b3}\x{17b6}\x{17be}-\x{17c5}\x{17c7}\x{17c8}\x{17d4}-\x{17da}\x{17dc}\x{17e0}-\x{17e9}\x{1810}-\x{1819}\x{1820}-\x{1877}\x{1880}-\x{18a8}\x{18aa}\x{18b0}-\x{18f5}\x{1900}-\x{191e}\x{1923}-\x{1926}\x{1929}-\x{192b}\x{1930}\x{1931}\x{1933}-\x{1938}\x{1946}-\x{196d}\x{1970}-\x{1974}\x{1980}-\x{19ab}\x{19b0}-\x{19c9}\x{19d0}-\x{19da}\x{1a00}-\x{1a16}\x{1a19}\x{1a1a}\x{1a1e}-\x{1a55}\x{1a57}\x{1a61}\x{1a63}\x{1a64}\x{1a6d}-\x{1a72}\x{1a80}-\x{1a89}\x{1a90}-\x{1a99}\x{1aa0}-\x{1aad}\x{1b04}-\x{1b33}\x{1b35}\x{1b3b}\x{1b3d}-\x{1b41}\x{1b43}-\x{1b4b}\x{1b50}-\x{1b6a}\x{1b74}-\x{1b7c}\x{1b82}-\x{1ba1}\x{1ba6}\x{1ba7}\x{1baa}\x{1bae}-\x{1be5}\x{1be7}\x{1bea}-\x{1bec}\x{1bee}\x{1bf2}\x{1bf3}\x{1bfc}-\x{1c2b}\x{1c34}\x{1c35}\x{1c3b}-\x{1c49}\x{1c4d}-\x{1c7f}\x{1cc0}-\x{1cc7}\x{1cd3}\x{1ce1}\x{1ce9}-\x{1cec}\x{1cee}-\x{1cf3}\x{1cf5}\x{1cf6}\x{1d00}-\x{1dbf}\x{1e00}-\x{1f15}\x{1f18}-\x{1f1d}\x{1f20}-\x{1f45}\x{1f48}-\x{1f4d}\x{1f50}-\x{1f57}\x{1f59}\x{1f5b}\x{1f5d}\x{1f5f}-\x{1f7d}\x{1f80}-\x{1fb4}\x{1fb6}-\x{1fbc}\x{1fbe}\x{1fc2}-\x{1fc4}\x{1fc6}-\x{1fcc}\x{1fd0}-\x{1fd3}\x{1fd6}-\x{1fdb}\x{1fe0}-\x{1fec}\x{1ff2}-\x{1ff4}\x{1ff6}-\x{1ffc}\x{200e}\x{2071}\x{207f}\x{2090}-\x{209c}\x{2102}\x{2107}\x{210a}-\x{2113}\x{2115}\x{2119}-\x{211d}\x{2124}\x{2126}\x{2128}\x{212a}-\x{212d}\x{212f}-\x{2139}\x{213c}-\x{213f}\x{2145}-\x{2149}\x{214e}\x{214f}\x{2160}-\x{2188}\x{2336}-\x{237a}\x{2395}\x{249c}-\x{24e9}\x{26ac}\x{2800}-\x{28ff}\x{2c00}-\x{2c2e}\x{2c30}-\x{2c5e}\x{2c60}-\x{2ce4}\x{2ceb}-\x{2cee}\x{2cf2}\x{2cf3}\x{2d00}-\x{2d25}\x{2d27}\x{2d2d}\x{2d30}-\x{2d67}\x{2d6f}\x{2d70}\x{2d80}-\x{2d96}\x{2da0}-\x{2da6}\x{2da8}-\x{2dae}\x{2db0}-\x{2db6}\x{2db8}-\x{2dbe}\x{2dc0}-\x{2dc6}\x{2dc8}-\x{2dce}\x{2dd0}-\x{2dd6}\x{2dd8}-\x{2dde}\x{3005}-\x{3007}\x{3021}-\x{3029}\x{302e}\x{302f}\x{3031}-\x{3035}\x{3038}-\x{303c}\x{3041}-\x{3096}\x{309d}-\x{309f}\x{30a1}-\x{30fa}\x{30fc}-\x{30ff}\x{3105}-\x{312d}\x{3131}-\x{318e}\x{3190}-\x{31ba}\x{31f0}-\x{321c}\x{3220}-\x{324f}\x{3260}-\x{327b}\x{327f}-\x{32b0}\x{32c0}-\x{32cb}\x{32d0}-\x{32fe}\x{3300}-\x{3376}\x{337b}-\x{33dd}\x{33e0}-\x{33fe}\x{3400}-\x{4db5}\x{4e00}-\x{9fd5}\x{a000}-\x{a48c}\x{a4d0}-\x{a60c}\x{a610}-\x{a62b}\x{a640}-\x{a66e}\x{a680}-\x{a69d}\x{a6a0}-\x{a6ef}\x{a6f2}-\x{a6f7}\x{a722}-\x{a787}\x{a789}-\x{a7ad}\x{a7b0}-\x{a7b7}\x{a7f7}-\x{a801}\x{a803}-\x{a805}\x{a807}-\x{a80a}\x{a80c}-\x{a824}\x{a827}\x{a830}-\x{a837}\x{a840}-\x{a873}\x{a880}-\x{a8c3}\x{a8ce}-\x{a8d9}\x{a8f2}-\x{a8fd}\x{a900}-\x{a925}\x{a92e}-\x{a946}\x{a952}\x{a953}\x{a95f}-\x{a97c}\x{a983}-\x{a9b2}\x{a9b4}\x{a9b5}\x{a9ba}\x{a9bb}\x{a9bd}-\x{a9cd}\x{a9cf}-\x{a9d9}\x{a9de}-\x{a9e4}\x{a9e6}-\x{a9fe}\x{aa00}-\x{aa28}\x{aa2f}\x{aa30}\x{aa33}\x{aa34}\x{aa40}-\x{aa42}\x{aa44}-\x{aa4b}\x{aa4d}\x{aa50}-\x{aa59}\x{aa5c}-\x{aa7b}\x{aa7d}-\x{aaaf}\x{aab1}\x{aab5}\x{aab6}\x{aab9}-\x{aabd}\x{aac0}\x{aac2}\x{aadb}-\x{aaeb}\x{aaee}-\x{aaf5}\x{ab01}-\x{ab06}\x{ab09}-\x{ab0e}\x{ab11}-\x{ab16}\x{ab20}-\x{ab26}\x{ab28}-\x{ab2e}\x{ab30}-\x{ab65}\x{ab70}-\x{abe4}\x{abe6}\x{abe7}\x{abe9}-\x{abec}\x{abf0}-\x{abf9}\x{ac00}-\x{d7a3}\x{d7b0}-\x{d7c6}\x{d7cb}-\x{d7fb}\x{e000}-\x{fa6d}\x{fa70}-\x{fad9}\x{fb00}-\x{fb06}\x{fb13}-\x{fb17}\x{ff21}-\x{ff3a}\x{ff41}-\x{ff5a}\x{ff66}-\x{ffbe}\x{ffc2}-\x{ffc7}\x{ffca}-\x{ffcf}\x{ffd2}-\x{ffd7}\x{ffda}-\x{ffdc}\x{10000}-\x{1000b}\x{1000d}-\x{10026}\x{10028}-\x{1003a}\x{1003c}\x{1003d}\x{1003f}-\x{1004d}\x{10050}-\x{1005d}\x{10080}-\x{100fa}\x{10100}\x{10102}\x{10107}-\x{10133}\x{10137}-\x{1013f}\x{101d0}-\x{101fc}\x{10280}-\x{1029c}\x{102a0}-\x{102d0}\x{10300}-\x{10323}\x{10330}-\x{1034a}\x{10350}-\x{10375}\x{10380}-\x{1039d}\x{1039f}-\x{103c3}\x{103c8}-\x{103d5}\x{10400}-\x{1049d}\x{104a0}-\x{104a9}\x{10500}-\x{10527}\x{10530}-\x{10563}\x{1056f}\x{10600}-\x{10736}\x{10740}-\x{10755}\x{10760}-\x{10767}\x{11000}\x{11002}-\x{11037}\x{11047}-\x{1104d}\x{11066}-\x{1106f}\x{11082}-\x{110b2}\x{110b7}\x{110b8}\x{110bb}-\x{110c1}\x{110d0}-\x{110e8}\x{110f0}-\x{110f9}\x{11103}-\x{11126}\x{1112c}\x{11136}-\x{11143}\x{11150}-\x{11172}\x{11174}-\x{11176}\x{11182}-\x{111b5}\x{111bf}-\x{111c9}\x{111cd}\x{111d0}-\x{111df}\x{111e1}-\x{111f4}\x{11200}-\x{11211}\x{11213}-\x{1122e}\x{11232}\x{11233}\x{11235}\x{11238}-\x{1123d}\x{11280}-\x{11286}\x{11288}\x{1128a}-\x{1128d}\x{1128f}-\x{1129d}\x{1129f}-\x{112a9}\x{112b0}-\x{112de}\x{112e0}-\x{112e2}\x{112f0}-\x{112f9}\x{11302}\x{11303}\x{11305}-\x{1130c}\x{1130f}\x{11310}\x{11313}-\x{11328}\x{1132a}-\x{11330}\x{11332}\x{11333}\x{11335}-\x{11339}\x{1133d}-\x{1133f}\x{11341}-\x{11344}\x{11347}\x{11348}\x{1134b}-\x{1134d}\x{11350}\x{11357}\x{1135d}-\x{11363}\x{11480}-\x{114b2}\x{114b9}\x{114bb}-\x{114be}\x{114c1}\x{114c4}-\x{114c7}\x{114d0}-\x{114d9}\x{11580}-\x{115b1}\x{115b8}-\x{115bb}\x{115be}\x{115c1}-\x{115db}\x{11600}-\x{11632}\x{1163b}\x{1163c}\x{1163e}\x{11641}-\x{11644}\x{11650}-\x{11659}\x{11680}-\x{116aa}\x{116ac}\x{116ae}\x{116af}\x{116b6}\x{116c0}-\x{116c9}\x{11700}-\x{11719}\x{11720}\x{11721}\x{11726}\x{11730}-\x{1173f}\x{118a0}-\x{118f2}\x{118ff}\x{11ac0}-\x{11af8}\x{12000}-\x{12399}\x{12400}-\x{1246e}\x{12470}-\x{12474}\x{12480}-\x{12543}\x{13000}-\x{1342e}\x{14400}-\x{14646}\x{16800}-\x{16a38}\x{16a40}-\x{16a5e}\x{16a60}-\x{16a69}\x{16a6e}\x{16a6f}\x{16ad0}-\x{16aed}\x{16af5}\x{16b00}-\x{16b2f}\x{16b37}-\x{16b45}\x{16b50}-\x{16b59}\x{16b5b}-\x{16b61}\x{16b63}-\x{16b77}\x{16b7d}-\x{16b8f}\x{16f00}-\x{16f44}\x{16f50}-\x{16f7e}\x{16f93}-\x{16f9f}\x{1b000}\x{1b001}\x{1bc00}-\x{1bc6a}\x{1bc70}-\x{1bc7c}\x{1bc80}-\x{1bc88}\x{1bc90}-\x{1bc99}\x{1bc9c}\x{1bc9f}\x{1d000}-\x{1d0f5}\x{1d100}-\x{1d126}\x{1d129}-\x{1d166}\x{1d16a}-\x{1d172}\x{1d183}\x{1d184}\x{1d18c}-\x{1d1a9}\x{1d1ae}-\x{1d1e8}\x{1d360}-\x{1d371}\x{1d400}-\x{1d454}\x{1d456}-\x{1d49c}\x{1d49e}\x{1d49f}\x{1d4a2}\x{1d4a5}\x{1d4a6}\x{1d4a9}-\x{1d4ac}\x{1d4ae}-\x{1d4b9}\x{1d4bb}\x{1d4bd}-\x{1d4c3}\x{1d4c5}-\x{1d505}\x{1d507}-\x{1d50a}\x{1d50d}-\x{1d514}\x{1d516}-\x{1d51c}\x{1d51e}-\x{1d539}\x{1d53b}-\x{1d53e}\x{1d540}-\x{1d544}\x{1d546}\x{1d54a}-\x{1d550}\x{1d552}-\x{1d6a5}\x{1d6a8}-\x{1d6da}\x{1d6dc}-\x{1d714}\x{1d716}-\x{1d74e}\x{1d750}-\x{1d788}\x{1d78a}-\x{1d7c2}\x{1d7c4}-\x{1d7cb}\x{1d800}-\x{1d9ff}\x{1da37}-\x{1da3a}\x{1da6d}-\x{1da74}\x{1da76}-\x{1da83}\x{1da85}-\x{1da8b}\x{1f110}-\x{1f12e}\x{1f130}-\x{1f169}\x{1f170}-\x{1f19a}\x{1f1e6}-\x{1f202}\x{1f210}-\x{1f23a}\x{1f240}-\x{1f248}\x{1f250}\x{1f251}\x{20000}-\x{2a6d6}\x{2a700}-\x{2b734}\x{2b740}-\x{2b81d}\x{2b820}-\x{2cea1}\x{2f800}-\x{2fa1d}\x{f0000}-\x{ffffd}\x{100000}-\x{10fffd}])|([\x{590}\x{5be}\x{5c0}\x{5c3}\x{5c6}\x{5c8}-\x{5ff}\x{7c0}-\x{7ea}\x{7f4}\x{7f5}\x{7fa}-\x{815}\x{81a}\x{824}\x{828}\x{82e}-\x{858}\x{85c}-\x{89f}\x{200f}\x{fb1d}\x{fb1f}-\x{fb28}\x{fb2a}-\x{fb4f}\x{10800}-\x{1091e}\x{10920}-\x{10a00}\x{10a04}\x{10a07}-\x{10a0b}\x{10a10}-\x{10a37}\x{10a3b}-\x{10a3e}\x{10a40}-\x{10ae4}\x{10ae7}-\x{10b38}\x{10b40}-\x{10e5f}\x{10e7f}-\x{10fff}\x{1e800}-\x{1e8cf}\x{1e8d7}-\x{1edff}\x{1ef00}-\x{1efff}\x{608}\x{60b}\x{60d}\x{61b}-\x{64a}\x{66d}-\x{66f}\x{671}-\x{6d5}\x{6e5}\x{6e6}\x{6ee}\x{6ef}\x{6fa}-\x{710}\x{712}-\x{72f}\x{74b}-\x{7a5}\x{7b1}-\x{7bf}\x{8a0}-\x{8e2}\x{fb50}-\x{fd3d}\x{fd40}-\x{fdcf}\x{fdf0}-\x{fdfc}\x{fdfe}\x{fdff}\x{fe70}-\x{fefe}\x{1ee00}-\x{1eeef}\x{1eef2}-\x{1eeff}]))/u';
175 // @codeCoverageIgnoreEnd
178 * Get a cached or new language object for a given language code
179 * @param string $code
180 * @throws MWException
183 static function factory( $code ) {
184 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
186 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
187 $code = $wgDummyLanguageCodes[$code];
190 // get the language object to process
191 $langObj = isset( self
::$mLangObjCache[$code] )
192 ? self
::$mLangObjCache[$code]
193 : self
::newFromCode( $code );
195 // merge the language object in to get it up front in the cache
196 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
197 // get rid of the oldest ones in case we have an overflow
198 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
204 * Create a language object for a given language code
205 * @param string $code
206 * @param bool $fallback Whether we're going through language fallback chain
207 * @throws MWException
210 protected static function newFromCode( $code, $fallback = false ) {
211 if ( !self
::isValidCode( $code ) ) {
212 throw new MWException( "Invalid language code \"$code\"" );
215 if ( !self
::isValidBuiltInCode( $code ) ) {
216 // It's not possible to customise this code with class files, so
217 // just return a Language object. This is to support uselang= hacks.
218 $lang = new Language
;
219 $lang->setCode( $code );
223 // Check if there is a language class for the code
224 $class = self
::classFromCode( $code, $fallback );
225 if ( class_exists( $class ) ) {
230 // Keep trying the fallback list until we find an existing class
231 $fallbacks = self
::getFallbacksFor( $code );
232 foreach ( $fallbacks as $fallbackCode ) {
233 if ( !self
::isValidBuiltInCode( $fallbackCode ) ) {
234 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
237 $class = self
::classFromCode( $fallbackCode );
238 if ( class_exists( $class ) ) {
240 $lang->setCode( $code );
245 throw new MWException( "Invalid fallback sequence for language '$code'" );
249 * Checks whether any localisation is available for that language tag
250 * in MediaWiki (MessagesXx.php exists).
252 * @param string $code Language tag (in lower case)
253 * @return bool Whether language is supported
256 public static function isSupportedLanguage( $code ) {
257 if ( !self
::isValidBuiltInCode( $code ) ) {
261 if ( $code === 'qqq' ) {
265 return is_readable( self
::getMessagesFileName( $code ) ) ||
266 is_readable( self
::getJsonMessagesFileName( $code ) );
270 * Returns true if a language code string is a well-formed language tag
271 * according to RFC 5646.
272 * This function only checks well-formedness; it doesn't check that
273 * language, script or variant codes actually exist in the repositories.
275 * Based on regexes by Mark Davis of the Unicode Consortium:
276 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
278 * @param string $code
279 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
284 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
287 $alphanum = '[a-z0-9]';
288 $x = 'x'; # private use singleton
289 $singleton = '[a-wy-z]'; # other singleton
290 $s = $lenient ?
'[-_]' : '-';
292 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
293 $script = "$alpha{4}"; # ISO 15924
294 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
295 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
296 $extension = "$singleton(?:$s$alphanum{2,8})+";
297 $privateUse = "$x(?:$s$alphanum{1,8})+";
299 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
300 # Since these are limited, this is safe even later changes to the registry --
301 # the only oddity is that it might change the type of the tag, and thus
302 # the results from the capturing groups.
303 # https://www.iana.org/assignments/language-subtag-registry
305 $grandfathered = "en{$s}GB{$s}oed"
306 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
307 . "|no{$s}(?:bok|nyn)"
308 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
309 . "|zh{$s}min{$s}nan";
311 $variantList = "$variant(?:$s$variant)*";
312 $extensionList = "$extension(?:$s$extension)*";
314 $langtag = "(?:($language)"
317 . "(?:$s$variantList)?"
318 . "(?:$s$extensionList)?"
319 . "(?:$s$privateUse)?)";
321 # The final breakdown, with capturing groups for each of these components
322 # The variants, extensions, grandfathered, and private-use may have interior '-'
324 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
326 return (bool)preg_match( "/$root/", strtolower( $code ) );
330 * Returns true if a language code string is of a valid form, whether or
331 * not it exists. This includes codes which are used solely for
332 * customisation via the MediaWiki namespace.
334 * @param string $code
338 public static function isValidCode( $code ) {
340 if ( !isset( $cache[$code] ) ) {
341 // People think language codes are html safe, so enforce it.
342 // Ideally we should only allow a-zA-Z0-9-
343 // but, .+ and other chars are often used for {{int:}} hacks
344 // see bugs T39564, T39587, T38938
346 // Protect against path traversal
347 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
348 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
350 return $cache[$code];
354 * Returns true if a language code is of a valid form for the purposes of
355 * internal customisation of MediaWiki, via Messages*.php or *.json.
357 * @param string $code
359 * @throws MWException
363 public static function isValidBuiltInCode( $code ) {
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
370 $type = gettype( $code );
371 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
378 * Returns true if a language code is an IETF tag known to MediaWiki.
385 public static function isKnownLanguageTag( $tag ) {
386 // Quick escape for invalid input to avoid exceptions down the line
387 // when code tries to process tags which are not valid at all.
388 if ( !self
::isValidBuiltInCode( $tag ) ) {
392 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
393 || self
::fetchLanguageName( $tag, $tag ) !== ''
402 * Get the LocalisationCache instance
404 * @return LocalisationCache
406 public static function getLocalisationCache() {
407 if ( is_null( self
::$dataCache ) ) {
408 global $wgLocalisationCacheConf;
409 $class = $wgLocalisationCacheConf['class'];
410 self
::$dataCache = new $class( $wgLocalisationCacheConf );
412 return self
::$dataCache;
415 function __construct() {
416 $this->mConverter
= new FakeConverter( $this );
417 // Set the code to the name of the descendant
418 if ( static::class === 'Language' ) {
421 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
423 self
::getLocalisationCache();
427 * Reduce memory usage
429 function __destruct() {
430 foreach ( $this as $name => $value ) {
431 unset( $this->$name );
436 * Hook which will be called if this is the content language.
437 * Descendants can use this to register hook functions or modify globals
439 function initContLang() {
446 public function getFallbackLanguages() {
447 return self
::getFallbacksFor( $this->mCode
);
451 * Exports $wgBookstoreListEn
454 public function getBookstoreList() {
455 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
459 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
460 * available in localised form, it will be included in English.
464 public function getNamespaces() {
465 if ( is_null( $this->namespaceNames
) ) {
466 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
468 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
470 $this->namespaceNames
= $wgExtraNamespaces +
471 self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
472 $this->namespaceNames +
= $validNamespaces;
474 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
475 if ( $wgMetaNamespaceTalk ) {
476 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
478 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
479 $this->namespaceNames
[NS_PROJECT_TALK
] =
480 $this->fixVariableInNamespace( $talk );
483 # Sometimes a language will be localised but not actually exist on this wiki.
484 foreach ( $this->namespaceNames
as $key => $text ) {
485 if ( !isset( $validNamespaces[$key] ) ) {
486 unset( $this->namespaceNames
[$key] );
490 # The above mixing may leave namespaces out of canonical order.
491 # Re-order by namespace ID number...
492 ksort( $this->namespaceNames
);
494 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
497 return $this->namespaceNames
;
501 * Arbitrarily set all of the namespace names at once. Mainly used for testing
502 * @param array $namespaces Array of namespaces (id => name)
504 public function setNamespaces( array $namespaces ) {
505 $this->namespaceNames
= $namespaces;
506 $this->mNamespaceIds
= null;
510 * Resets all of the namespace caches. Mainly used for testing
512 public function resetNamespaces() {
513 $this->namespaceNames
= null;
514 $this->mNamespaceIds
= null;
515 $this->namespaceAliases
= null;
519 * A convenience function that returns getNamespaces() with spaces instead of underscores
520 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
524 public function getFormattedNamespaces() {
525 $ns = $this->getNamespaces();
526 foreach ( $ns as $k => $v ) {
527 $ns[$k] = strtr( $v, '_', ' ' );
533 * Get a namespace value by key
536 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
537 * echo $mw_ns; // prints 'MediaWiki'
540 * @param int $index The array key of the namespace to return
541 * @return string|bool String if the namespace value exists, otherwise false
543 public function getNsText( $index ) {
544 $ns = $this->getNamespaces();
545 return isset( $ns[$index] ) ?
$ns[$index] : false;
549 * A convenience function that returns the same thing as
550 * getNsText() except with '_' changed to ' ', useful for
554 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
555 * echo $mw_ns; // prints 'MediaWiki talk'
558 * @param int $index The array key of the namespace to return
559 * @return string Namespace name without underscores (empty string if namespace does not exist)
561 public function getFormattedNsText( $index ) {
562 $ns = $this->getNsText( $index );
563 return strtr( $ns, '_', ' ' );
567 * Returns gender-dependent namespace alias if available.
568 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
569 * @param int $index Namespace index
570 * @param string $gender Gender key (male, female... )
574 public function getGenderNsText( $index, $gender ) {
575 global $wgExtraGenderNamespaces;
577 $ns = $wgExtraGenderNamespaces +
578 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
580 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
584 * Whether this language uses gender-dependent namespace aliases.
585 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
589 public function needsGenderDistinction() {
590 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
591 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
592 // $wgExtraGenderNamespaces overrides everything
594 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
595 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
596 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
599 // Check what is in i18n files
600 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
601 return count( $aliases ) > 0;
606 * Get a namespace key by value, case insensitive.
607 * Only matches namespace names for the current language, not the
608 * canonical ones defined in Namespace.php.
610 * @param string $text
611 * @return int|bool An integer if $text is a valid value otherwise false
613 function getLocalNsIndex( $text ) {
614 $lctext = $this->lc( $text );
615 $ids = $this->getNamespaceIds();
616 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
622 public function getNamespaceAliases() {
623 if ( is_null( $this->namespaceAliases
) ) {
624 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
628 foreach ( $aliases as $name => $index ) {
629 if ( $index === NS_PROJECT_TALK
) {
630 unset( $aliases[$name] );
631 $name = $this->fixVariableInNamespace( $name );
632 $aliases[$name] = $index;
637 global $wgExtraGenderNamespaces;
638 $genders = $wgExtraGenderNamespaces +
639 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
640 foreach ( $genders as $index => $forms ) {
641 foreach ( $forms as $alias ) {
642 $aliases[$alias] = $index;
646 # Also add converted namespace names as aliases, to avoid confusion.
647 $convertedNames = [];
648 foreach ( $this->getVariants() as $variant ) {
649 if ( $variant === $this->mCode
) {
652 foreach ( $this->getNamespaces() as $ns => $_ ) {
653 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
657 $this->namespaceAliases
= $aliases +
$convertedNames;
660 return $this->namespaceAliases
;
666 public function getNamespaceIds() {
667 if ( is_null( $this->mNamespaceIds
) ) {
668 global $wgNamespaceAliases;
669 # Put namespace names and aliases into a hashtable.
670 # If this is too slow, then we should arrange it so that it is done
671 # before caching. The catch is that at pre-cache time, the above
672 # class-specific fixup hasn't been done.
673 $this->mNamespaceIds
= [];
674 foreach ( $this->getNamespaces() as $index => $name ) {
675 $this->mNamespaceIds
[$this->lc( $name )] = $index;
677 foreach ( $this->getNamespaceAliases() as $name => $index ) {
678 $this->mNamespaceIds
[$this->lc( $name )] = $index;
680 if ( $wgNamespaceAliases ) {
681 foreach ( $wgNamespaceAliases as $name => $index ) {
682 $this->mNamespaceIds
[$this->lc( $name )] = $index;
686 return $this->mNamespaceIds
;
690 * Get a namespace key by value, case insensitive. Canonical namespace
691 * names override custom ones defined for the current language.
693 * @param string $text
694 * @return int|bool An integer if $text is a valid value otherwise false
696 public function getNsIndex( $text ) {
697 $lctext = $this->lc( $text );
698 $ns = MWNamespace
::getCanonicalIndex( $lctext );
699 if ( $ns !== null ) {
702 $ids = $this->getNamespaceIds();
703 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
707 * short names for language variants used for language conversion links.
709 * @param string $code
710 * @param bool $usemsg Use the "variantname-xyz" message if it exists
713 public function getVariantname( $code, $usemsg = true ) {
714 $msg = "variantname-$code";
715 if ( $usemsg && wfMessage( $msg )->exists() ) {
716 return $this->getMessageFromDB( $msg );
718 $name = self
::fetchLanguageName( $code );
720 return $name; # if it's defined as a language name, show that
722 # otherwise, output the language code
728 * @return string[]|bool List of date format preference keys, or false if disabled.
730 public function getDatePreferences() {
731 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
737 function getDateFormats() {
738 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
742 * @return array|string
744 public function getDefaultDateFormat() {
745 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
746 if ( $df === 'dmy or mdy' ) {
747 global $wgAmericanDates;
748 return $wgAmericanDates ?
'mdy' : 'dmy';
757 public function getDatePreferenceMigrationMap() {
758 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
762 * @param string $image
765 function getImageFile( $image ) {
766 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
773 public function getImageFiles() {
774 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
780 public function getExtraUserToggles() {
781 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
788 function getUserToggle( $tog ) {
789 return $this->getMessageFromDB( "tog-$tog" );
793 * Get an array of language names, indexed by code.
794 * @param null|string $inLanguage Code of language in which to return the names
795 * Use null for autonyms (native names)
796 * @param string $include One of:
797 * 'all' all available languages
798 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
799 * 'mwfile' only if the language is in 'mw' *and* has a message file
800 * @return array Language code => language name
803 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
804 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
805 $cacheKey .= ":$include";
806 if ( self
::$languageNameCache === null ) {
807 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
810 $ret = self
::$languageNameCache->get( $cacheKey );
812 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
813 self
::$languageNameCache->set( $cacheKey, $ret );
819 * Uncached helper for fetchLanguageNames
820 * @param null|string $inLanguage Code of language in which to return the names
821 * Use null for autonyms (native names)
822 * @param string $include One of:
823 * 'all' all available languages
824 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
825 * 'mwfile' only if the language is in 'mw' *and* has a message file
826 * @return array Language code => language name
828 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
829 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
831 // If passed an invalid language code to use, fallback to en
832 if ( $inLanguage !== null && !self
::isValidCode( $inLanguage ) ) {
839 # TODO: also include when $inLanguage is null, when this code is more efficient
840 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
843 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
844 if ( $wgUsePigLatinVariant ) {
845 // Pig Latin (for variant development)
846 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
849 foreach ( $mwNames as $mwCode => $mwName ) {
850 # - Prefer own MediaWiki native name when not using the hook
851 # - For other names just add if not added through the hook
852 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
853 $names[$mwCode] = $mwName;
857 if ( $include === 'all' ) {
863 $coreCodes = array_keys( $mwNames );
864 foreach ( $coreCodes as $coreCode ) {
865 $returnMw[$coreCode] = $names[$coreCode];
868 if ( $include === 'mwfile' ) {
870 # We do this using a foreach over the codes instead of a directory
871 # loop so that messages files in extensions will work correctly.
872 foreach ( $returnMw as $code => $value ) {
873 if ( is_readable( self
::getMessagesFileName( $code ) )
874 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
876 $namesMwFile[$code] = $names[$code];
880 ksort( $namesMwFile );
885 # 'mw' option; default if it's not one of the other two options (all/mwfile)
890 * @param string $code The code of the language for which to get the name
891 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
892 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
893 * @return string Language name or empty
896 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
897 $code = strtolower( $code );
898 $array = self
::fetchLanguageNames( $inLanguage, $include );
899 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
903 * Get a message from the MediaWiki namespace.
905 * @param string $msg Message name
908 public function getMessageFromDB( $msg ) {
909 return $this->msg( $msg )->text();
913 * Get message object in this language. Only for use inside this class.
915 * @param string $msg Message name
918 protected function msg( $msg ) {
919 return wfMessage( $msg )->inLanguage( $this );
926 public function getMonthName( $key ) {
927 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
933 public function getMonthNamesArray() {
934 $monthNames = [ '' ];
935 for ( $i = 1; $i < 13; $i++
) {
936 $monthNames[] = $this->getMonthName( $i );
945 public function getMonthNameGen( $key ) {
946 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
953 public function getMonthAbbreviation( $key ) {
954 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
960 public function getMonthAbbreviationsArray() {
961 $monthNames = [ '' ];
962 for ( $i = 1; $i < 13; $i++
) {
963 $monthNames[] = $this->getMonthAbbreviation( $i );
972 public function getWeekdayName( $key ) {
973 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
980 function getWeekdayAbbreviation( $key ) {
981 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
988 function getIranianCalendarMonthName( $key ) {
989 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
996 function getHebrewCalendarMonthName( $key ) {
997 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1001 * @param string $key
1004 function getHebrewCalendarMonthNameGen( $key ) {
1005 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1009 * @param string $key
1012 function getHijriCalendarMonthName( $key ) {
1013 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1017 * Pass through result from $dateTimeObj->format()
1018 * @param DateTime|bool|null &$dateTimeObj
1020 * @param DateTimeZone|bool|null $zone
1021 * @param string $code
1024 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1025 if ( !$dateTimeObj ) {
1026 $dateTimeObj = DateTime
::createFromFormat(
1027 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1030 return $dateTimeObj->format( $code );
1034 * This is a workalike of PHP's date() function, but with better
1035 * internationalisation, a reduced set of format characters, and a better
1038 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1039 * the PHP manual for definitions. There are a number of extensions, which
1042 * xn Do not translate digits of the next numeric format character
1043 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1044 * xr Use roman numerals for the next numeric format character
1045 * xh Use hebrew numerals for the next numeric format character
1047 * xg Genitive month name
1049 * xij j (day number) in Iranian calendar
1050 * xiF F (month name) in Iranian calendar
1051 * xin n (month number) in Iranian calendar
1052 * xiy y (two digit year) in Iranian calendar
1053 * xiY Y (full year) in Iranian calendar
1054 * xit t (days in month) in Iranian calendar
1055 * xiz z (day of the year) in Iranian calendar
1057 * xjj j (day number) in Hebrew calendar
1058 * xjF F (month name) in Hebrew calendar
1059 * xjt t (days in month) in Hebrew calendar
1060 * xjx xg (genitive month name) in Hebrew calendar
1061 * xjn n (month number) in Hebrew calendar
1062 * xjY Y (full year) in Hebrew calendar
1064 * xmj j (day number) in Hijri calendar
1065 * xmF F (month name) in Hijri calendar
1066 * xmn n (month number) in Hijri calendar
1067 * xmY Y (full year) in Hijri calendar
1069 * xkY Y (full year) in Thai solar calendar. Months and days are
1070 * identical to the Gregorian calendar
1071 * xoY Y (full year) in Minguo calendar or Juche year.
1072 * Months and days are identical to the
1073 * Gregorian calendar
1074 * xtY Y (full year) in Japanese nengo. Months and days are
1075 * identical to the Gregorian calendar
1077 * Characters enclosed in double quotes will be considered literal (with
1078 * the quotes themselves removed). Unmatched quotes will be considered
1079 * literal quotes. Example:
1081 * "The month is" F => The month is January
1084 * Backslash escaping is also supported.
1086 * Input timestamp is assumed to be pre-normalized to the desired local
1087 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1088 * $ts is UTC if $zone is not given.
1090 * @param string $format
1091 * @param string $ts 14-character timestamp
1094 * @param DateTimeZone $zone Timezone of $ts
1095 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1096 * Only makes sense if $ts is the current time.
1097 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1099 * @throws MWException
1102 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1107 $dateTimeObj = false;
1116 $usedSecond = false;
1117 $usedMinute = false;
1124 $usedISOYear = false;
1125 $usedIsLeapYear = false;
1127 $usedHebrewMonth = false;
1128 $usedIranianMonth = false;
1129 $usedHijriMonth = false;
1130 $usedHebrewYear = false;
1131 $usedIranianYear = false;
1132 $usedHijriYear = false;
1133 $usedTennoYear = false;
1135 if ( strlen( $ts ) !== 14 ) {
1136 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1139 if ( !ctype_digit( $ts ) ) {
1140 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1143 $formatLength = strlen( $format );
1144 for ( $p = 0; $p < $formatLength; $p++
) {
1146 $code = $format[$p];
1147 if ( $code == 'x' && $p < $formatLength - 1 ) {
1148 $code .= $format[++
$p];
1151 if ( ( $code === 'xi'
1157 && $p < $formatLength - 1 ) {
1158 $code .= $format[++
$p];
1169 $rawToggle = !$rawToggle;
1179 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1182 $usedHebrewMonth = true;
1184 $hebrew = self
::tsToHebrew( $ts );
1186 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1190 $num = substr( $ts, 6, 2 );
1194 $s .= $this->getWeekdayAbbreviation(
1195 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1200 $num = intval( substr( $ts, 6, 2 ) );
1205 $iranian = self
::tsToIranian( $ts );
1212 $hijri = self
::tsToHijri( $ts );
1219 $hebrew = self
::tsToHebrew( $ts );
1225 $s .= $this->getWeekdayName(
1226 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1231 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1234 $usedIranianMonth = true;
1236 $iranian = self
::tsToIranian( $ts );
1238 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1241 $usedHijriMonth = true;
1243 $hijri = self
::tsToHijri( $ts );
1245 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1248 $usedHebrewMonth = true;
1250 $hebrew = self
::tsToHebrew( $ts );
1252 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1256 $num = substr( $ts, 4, 2 );
1260 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1264 $num = intval( substr( $ts, 4, 2 ) );
1267 $usedIranianMonth = true;
1269 $iranian = self
::tsToIranian( $ts );
1274 $usedHijriMonth = true;
1276 $hijri = self
::tsToHijri( $ts );
1281 $usedHebrewMonth = true;
1283 $hebrew = self
::tsToHebrew( $ts );
1288 $usedHebrewMonth = true;
1290 $hebrew = self
::tsToHebrew( $ts );
1296 $num = substr( $ts, 0, 4 );
1299 $usedIranianYear = true;
1301 $iranian = self
::tsToIranian( $ts );
1306 $usedHijriYear = true;
1308 $hijri = self
::tsToHijri( $ts );
1313 $usedHebrewYear = true;
1315 $hebrew = self
::tsToHebrew( $ts );
1322 $thai = self
::tsToYear( $ts, 'thai' );
1329 $minguo = self
::tsToYear( $ts, 'minguo' );
1334 $usedTennoYear = true;
1336 $tenno = self
::tsToYear( $ts, 'tenno' );
1342 $num = substr( $ts, 2, 2 );
1345 $usedIranianYear = true;
1347 $iranian = self
::tsToIranian( $ts );
1349 $num = substr( $iranian[0], -2 );
1352 $usedIranianYear = true;
1354 $iranian = self
::tsToIranian( $ts );
1356 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1359 $usedIranianYear = true;
1361 $iranian = self
::tsToIranian( $ts );
1367 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1371 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1375 $h = substr( $ts, 8, 2 );
1376 $num = $h %
12 ?
$h %
12 : 12;
1380 $num = intval( substr( $ts, 8, 2 ) );
1384 $h = substr( $ts, 8, 2 );
1385 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1389 $num = substr( $ts, 8, 2 );
1393 $num = substr( $ts, 10, 2 );
1397 $num = substr( $ts, 12, 2 );
1407 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1413 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1417 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1421 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1424 $usedIsLeapYear = true;
1425 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1428 $usedISOYear = true;
1429 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1436 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1439 # Backslash escaping
1440 if ( $p < $formatLength - 1 ) {
1441 $s .= $format[++
$p];
1448 if ( $p < $formatLength - 1 ) {
1449 $endQuote = strpos( $format, '"', $p +
1 );
1450 if ( $endQuote === false ) {
1451 # No terminating quote, assume literal "
1454 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1458 # Quote at end of string, assume literal "
1465 if ( $num !== false ) {
1466 if ( $rawToggle ||
$raw ) {
1469 } elseif ( $roman ) {
1470 $s .= self
::romanNumeral( $num );
1472 } elseif ( $hebrewNum ) {
1473 $s .= self
::hebrewNumeral( $num );
1476 $s .= $this->formatNum( $num, true );
1481 if ( $ttl === 'unused' ) {
1482 // No need to calculate the TTL, the caller wont use it anyway.
1483 } elseif ( $usedSecond ) {
1485 } elseif ( $usedMinute ) {
1486 $ttl = 60 - substr( $ts, 12, 2 );
1487 } elseif ( $usedHour ) {
1488 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1489 } elseif ( $usedAMPM ) {
1490 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1491 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1495 $usedIranianMonth ||
1502 // @todo Someone who understands the non-Gregorian calendars
1503 // should write proper logic for them so that they don't need purged every day.
1504 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1505 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1508 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1509 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1512 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1513 $timeRemainingInDay;
1514 } elseif ( $usedISOYear ) {
1515 // December 28th falls on the last ISO week of the year, every year.
1516 // The last ISO week of a year can be 52 or 53.
1517 $lastWeekOfISOYear = DateTime
::createFromFormat(
1519 substr( $ts, 0, 4 ) . '1228',
1520 $zone ?
: new DateTimeZone( 'UTC' )
1522 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1523 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1524 $timeRemainingInWeek =
1525 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1526 +
$timeRemainingInDay;
1527 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1532 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1533 substr( $ts, 6, 2 ) ) * 86400
1534 +
$timeRemainingInDay;
1535 } elseif ( $usedYear ) {
1537 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1538 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1539 +
$timeRemainingInDay;
1540 } elseif ( $usedIsLeapYear ) {
1541 $year = substr( $ts, 0, 4 );
1542 $timeRemainingInYear =
1543 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1544 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1545 +
$timeRemainingInDay;
1547 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1548 // this isn't a leap year. see when the next one starts
1549 $nextCandidate = $year - $mod +
4;
1550 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1551 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1552 $timeRemainingInYear;
1554 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1555 $timeRemainingInYear;
1558 // this is a leap year, so the next year isn't
1559 $possibleTtls[] = $timeRemainingInYear;
1563 if ( $possibleTtls ) {
1564 $ttl = min( $possibleTtls );
1571 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1572 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1575 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1576 * Gregorian dates to Iranian dates. Originally written in C, it
1577 * is released under the terms of GNU Lesser General Public
1578 * License. Conversion to PHP was performed by Niklas Laxström.
1580 * Link: http://www.farsiweb.info/jalali/jalali.c
1586 private static function tsToIranian( $ts ) {
1587 $gy = substr( $ts, 0, 4 ) - 1600;
1588 $gm = substr( $ts, 4, 2 ) - 1;
1589 $gd = substr( $ts, 6, 2 ) - 1;
1591 # Days passed from the beginning (including leap years)
1593 +
floor( ( $gy +
3 ) / 4 )
1594 - floor( ( $gy +
99 ) / 100 )
1595 +
floor( ( $gy +
399 ) / 400 );
1597 // Add days of the past months of this year
1598 for ( $i = 0; $i < $gm; $i++
) {
1599 $gDayNo +
= self
::$GREG_DAYS[$i];
1603 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1607 // Days passed in current month
1608 $gDayNo +
= (int)$gd;
1610 $jDayNo = $gDayNo - 79;
1612 $jNp = floor( $jDayNo / 12053 );
1615 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1618 if ( $jDayNo >= 366 ) {
1619 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1620 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1625 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1626 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1632 return [ $jy, $jm, $jd, $jz ];
1636 * Converting Gregorian dates to Hijri dates.
1638 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1640 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1646 private static function tsToHijri( $ts ) {
1647 $year = substr( $ts, 0, 4 );
1648 $month = substr( $ts, 4, 2 );
1649 $day = substr( $ts, 6, 2 );
1657 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1658 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1660 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1661 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1662 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1665 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1666 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1669 $zl = $zjd - 1948440 +
10632;
1670 $zn = (int)( ( $zl - 1 ) / 10631 );
1671 $zl = $zl - 10631 * $zn +
354;
1672 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1673 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1674 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1675 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1676 $zm = (int)( ( 24 * $zl ) / 709 );
1677 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1678 $zy = 30 * $zn +
$zj - 30;
1680 return [ $zy, $zm, $zd ];
1684 * Converting Gregorian dates to Hebrew dates.
1686 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1687 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1688 * to translate the relevant functions into PHP and release them under
1691 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1692 * and Adar II is 14. In a non-leap year, Adar is 6.
1698 private static function tsToHebrew( $ts ) {
1700 $year = substr( $ts, 0, 4 );
1701 $month = substr( $ts, 4, 2 );
1702 $day = substr( $ts, 6, 2 );
1704 # Calculate Hebrew year
1705 $hebrewYear = $year +
3760;
1707 # Month number when September = 1, August = 12
1709 if ( $month > 12 ) {
1716 # Calculate day of year from 1 September
1718 for ( $i = 1; $i < $month; $i++
) {
1722 # Check if the year is leap
1723 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1726 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1733 # Calculate the start of the Hebrew year
1734 $start = self
::hebrewYearStart( $hebrewYear );
1736 # Calculate next year's start
1737 if ( $dayOfYear <= $start ) {
1738 # Day is before the start of the year - it is the previous year
1740 $nextStart = $start;
1744 # Add days since previous year's 1 September
1746 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1750 # Start of the new (previous) year
1751 $start = self
::hebrewYearStart( $hebrewYear );
1754 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1757 # Calculate Hebrew day of year
1758 $hebrewDayOfYear = $dayOfYear - $start;
1760 # Difference between year's days
1761 $diff = $nextStart - $start;
1762 # Add 12 (or 13 for leap years) days to ignore the difference between
1763 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1764 # difference is only about the year type
1765 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1771 # Check the year pattern, and is leap year
1772 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1773 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1774 # and non-leap years
1775 $yearPattern = $diff %
30;
1776 # Check if leap year
1777 $isLeap = $diff >= 30;
1779 # Calculate day in the month from number of day in the Hebrew year
1780 # Don't check Adar - if the day is not in Adar, we will stop before;
1781 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1782 $hebrewDay = $hebrewDayOfYear;
1785 while ( $hebrewMonth <= 12 ) {
1786 # Calculate days in this month
1787 if ( $isLeap && $hebrewMonth == 6 ) {
1788 # Adar in a leap year
1790 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1792 if ( $hebrewDay <= $days ) {
1796 # Subtract the days of Adar I
1797 $hebrewDay -= $days;
1800 if ( $hebrewDay <= $days ) {
1806 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1807 # Cheshvan in a complete year (otherwise as the rule below)
1809 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1810 # Kislev in an incomplete year (otherwise as the rule below)
1813 # Odd months have 30 days, even have 29
1814 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1816 if ( $hebrewDay <= $days ) {
1817 # In the current month
1820 # Subtract the days of the current month
1821 $hebrewDay -= $days;
1822 # Try in the next month
1827 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1831 * This calculates the Hebrew year start, as days since 1 September.
1832 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1833 * Used for Hebrew date.
1839 private static function hebrewYearStart( $year ) {
1840 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1841 $b = intval( ( $year - 1 ) %
4 );
1842 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1846 $Mar = intval( $m );
1852 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1853 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1855 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1857 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1861 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1866 * Algorithm to convert Gregorian dates to Thai solar dates,
1867 * Minguo dates or Minguo dates.
1869 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1870 * https://en.wikipedia.org/wiki/Minguo_calendar
1871 * https://en.wikipedia.org/wiki/Japanese_era_name
1873 * @param string $ts 14-character timestamp
1874 * @param string $cName Calender name
1875 * @return array Converted year, month, day
1877 private static function tsToYear( $ts, $cName ) {
1878 $gy = substr( $ts, 0, 4 );
1879 $gm = substr( $ts, 4, 2 );
1880 $gd = substr( $ts, 6, 2 );
1882 if ( !strcmp( $cName, 'thai' ) ) {
1884 # Add 543 years to the Gregorian calendar
1885 # Months and days are identical
1886 $gy_offset = $gy +
543;
1887 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1889 # Deduct 1911 years from the Gregorian calendar
1890 # Months and days are identical
1891 $gy_offset = $gy - 1911;
1892 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1893 # Nengō dates up to Meiji period
1894 # Deduct years from the Gregorian calendar
1895 # depending on the nengo periods
1896 # Months and days are identical
1898 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1899 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1902 $gy_gannen = $gy - 1868 +
1;
1903 $gy_offset = $gy_gannen;
1904 if ( $gy_gannen == 1 ) {
1907 $gy_offset = '明治' . $gy_offset;
1909 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1910 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1911 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1912 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1913 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1916 $gy_gannen = $gy - 1912 +
1;
1917 $gy_offset = $gy_gannen;
1918 if ( $gy_gannen == 1 ) {
1921 $gy_offset = '大正' . $gy_offset;
1923 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1924 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1925 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1928 $gy_gannen = $gy - 1926 +
1;
1929 $gy_offset = $gy_gannen;
1930 if ( $gy_gannen == 1 ) {
1933 $gy_offset = '昭和' . $gy_offset;
1935 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1936 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1937 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1940 $gy_gannen = $gy - 1989 +
1;
1941 $gy_offset = $gy_gannen;
1942 if ( $gy_gannen == 1 ) {
1945 $gy_offset = '平成' . $gy_offset;
1948 $gy_gannen = $gy - 2019 +
1;
1949 $gy_offset = $gy_gannen;
1950 if ( $gy_gannen == 1 ) {
1953 $gy_offset = '令和' . $gy_offset;
1959 return [ $gy_offset, $gm, $gd ];
1963 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1965 * This is the rule the BIDI algorithm uses to determine the directionality of
1966 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1967 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1969 * TODO: Does not handle BIDI control characters inside the text.
1970 * TODO: Does not handle unallocated characters.
1972 * @param string $text Text to test
1973 * @return null|string Directionality ('ltr' or 'rtl') or null
1975 private static function strongDirFromContent( $text = '' ) {
1976 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
1979 if ( $matches[1] === '' ) {
1986 * Roman number formatting up to 10000
1992 static function romanNumeral( $num ) {
1994 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1995 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1996 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1997 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1998 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2001 $num = intval( $num );
2002 if ( $num > 10000 ||
$num <= 0 ) {
2007 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2008 if ( $num >= $pow10 ) {
2009 $s .= $table[$i][(int)floor( $num / $pow10 )];
2011 $num = $num %
$pow10;
2017 * Hebrew Gematria number formatting up to 9999
2023 static function hebrewNumeral( $num ) {
2025 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2026 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2039 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2042 $num = intval( $num );
2043 if ( $num > 9999 ||
$num <= 0 ) {
2047 // Round thousands have special notations
2048 if ( $num === 1000 ) {
2050 } elseif ( $num %
1000 === 0 ) {
2051 return $table[0][$num / 1000] . "' אלפים";
2056 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2057 if ( $num >= $pow10 ) {
2058 if ( $num === 15 ||
$num === 16 ) {
2059 $letters[] = $table[0][9];
2060 $letters[] = $table[0][$num - 9];
2063 $letters = array_merge(
2065 (array)$table[$i][intval( $num / $pow10 )]
2068 if ( $pow10 === 1000 ) {
2074 $num = $num %
$pow10;
2077 $preTransformLength = count( $letters );
2078 if ( $preTransformLength === 1 ) {
2079 // Add geresh (single quote) to one-letter numbers
2082 $lastIndex = $preTransformLength - 1;
2083 $letters[$lastIndex] = str_replace(
2084 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2085 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2086 $letters[$lastIndex]
2089 // Add gershayim (double quote) to multiple-letter numbers,
2090 // but exclude numbers with only one letter after the thousands
2091 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2092 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2095 array_splice( $letters, -1, 0, '"' );
2099 return implode( $letters );
2103 * Used by date() and time() to adjust the time output.
2105 * @param string $ts The time in date('YmdHis') format
2106 * @param mixed $tz Adjust the time by this amount (default false, mean we
2107 * get user timecorrection setting)
2110 public function userAdjust( $ts, $tz = false ) {
2111 global $wgUser, $wgLocalTZoffset;
2113 if ( $tz === false ) {
2114 $tz = $wgUser->getOption( 'timecorrection' );
2117 $data = explode( '|', $tz, 3 );
2119 if ( $data[0] == 'ZoneInfo' ) {
2121 $userTZ = new DateTimeZone( $data[2] );
2122 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2123 $date->setTimezone( $userTZ );
2124 return $date->format( 'YmdHis' );
2125 } catch ( Exception
$e ) {
2126 // Unrecognized timezone, default to 'Offset' with the stored offset.
2127 $data[0] = 'Offset';
2131 if ( $data[0] == 'System' ||
$tz == '' ) {
2132 # Global offset in minutes.
2133 $minDiff = $wgLocalTZoffset;
2134 } elseif ( $data[0] == 'Offset' ) {
2135 $minDiff = intval( $data[1] );
2137 $data = explode( ':', $tz );
2138 if ( count( $data ) == 2 ) {
2139 $data[0] = intval( $data[0] );
2140 $data[1] = intval( $data[1] );
2141 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2142 if ( $data[0] < 0 ) {
2143 $minDiff = -$minDiff;
2146 $minDiff = intval( $data[0] ) * 60;
2150 # No difference ? Return time unchanged
2151 if ( 0 == $minDiff ) {
2155 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2156 # Generate an adjusted date; take advantage of the fact that mktime
2157 # will normalize out-of-range values so we don't have to split $minDiff
2158 # into hours and minutes.
2160 (int)substr( $ts, 8, 2 ) ), # Hours
2161 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2162 (int)substr( $ts, 12, 2 ), # Seconds
2163 (int)substr( $ts, 4, 2 ), # Month
2164 (int)substr( $ts, 6, 2 ), # Day
2165 (int)substr( $ts, 0, 4 ) ); # Year
2167 $date = date( 'YmdHis', $t );
2168 Wikimedia\restoreWarnings
();
2174 * This is meant to be used by time(), date(), and timeanddate() to get
2175 * the date preference they're supposed to use, it should be used in
2178 * function timeanddate([...], $format = true) {
2179 * $datePreference = $this->dateFormat($format);
2183 * @param int|string|bool $usePrefs If true, the user's preference is used
2184 * if false, the site/language default is used
2185 * if int/string, assumed to be a format.
2188 function dateFormat( $usePrefs = true ) {
2191 if ( is_bool( $usePrefs ) ) {
2193 $datePreference = $wgUser->getDatePreference();
2195 $datePreference = (string)User
::getDefaultOption( 'date' );
2198 $datePreference = (string)$usePrefs;
2202 if ( $datePreference == '' ) {
2206 return $datePreference;
2210 * Get a format string for a given type and preference
2211 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2212 * @param string $pref The format name as it appears in Messages*.php under
2215 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2219 function getDateFormatString( $type, $pref ) {
2220 $wasDefault = false;
2221 if ( $pref == 'default' ) {
2223 $pref = $this->getDefaultDateFormat();
2226 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2227 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2229 if ( $type === 'pretty' && $df === null ) {
2230 $df = $this->getDateFormatString( 'date', $pref );
2233 if ( !$wasDefault && $df === null ) {
2234 $pref = $this->getDefaultDateFormat();
2235 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2238 $this->dateFormatStrings
[$type][$pref] = $df;
2240 return $this->dateFormatStrings
[$type][$pref];
2244 * @param string $ts The time format which needs to be turned into a
2245 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2246 * @param bool $adj Whether to adjust the time output according to the
2247 * user configured offset ($timecorrection)
2248 * @param mixed $format True to use user's date format preference
2249 * @param string|bool $timecorrection The time offset as returned by
2250 * validateTimeZone() in Special:Preferences
2253 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2254 $ts = wfTimestamp( TS_MW
, $ts );
2256 $ts = $this->userAdjust( $ts, $timecorrection );
2258 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2259 return $this->sprintfDate( $df, $ts );
2263 * @param string $ts The time format which needs to be turned into a
2264 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2265 * @param bool $adj Whether to adjust the time output according to the
2266 * user configured offset ($timecorrection)
2267 * @param mixed $format True to use user's date format preference
2268 * @param string|bool $timecorrection The time offset as returned by
2269 * validateTimeZone() in Special:Preferences
2272 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2273 $ts = wfTimestamp( TS_MW
, $ts );
2275 $ts = $this->userAdjust( $ts, $timecorrection );
2277 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2278 return $this->sprintfDate( $df, $ts );
2282 * @param string $ts The time format which needs to be turned into a
2283 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2284 * @param bool $adj Whether to adjust the time output according to the
2285 * user configured offset ($timecorrection)
2286 * @param mixed $format What format to return, if it's false output the
2287 * default one (default true)
2288 * @param string|bool $timecorrection The time offset as returned by
2289 * validateTimeZone() in Special:Preferences
2292 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2293 $ts = wfTimestamp( TS_MW
, $ts );
2295 $ts = $this->userAdjust( $ts, $timecorrection );
2297 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2298 return $this->sprintfDate( $df, $ts );
2302 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2306 * @param int $seconds The amount of seconds.
2307 * @param array $chosenIntervals The intervals to enable.
2311 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2312 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2316 foreach ( $intervals as $intervalName => $intervalValue ) {
2317 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2318 // duration-years, duration-decades, duration-centuries, duration-millennia
2319 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2320 $segments[] = $message->inLanguage( $this )->escaped();
2323 return $this->listToText( $segments );
2327 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2328 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2332 * @param int $seconds The amount of seconds.
2333 * @param array $chosenIntervals The intervals to enable.
2337 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2338 if ( empty( $chosenIntervals ) ) {
2339 $chosenIntervals = [
2351 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2352 $sortedNames = array_keys( $intervals );
2353 $smallestInterval = array_pop( $sortedNames );
2357 foreach ( $intervals as $name => $length ) {
2358 $value = floor( $seconds / $length );
2360 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2361 $seconds -= $value * $length;
2362 $segments[$name] = $value;
2370 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2372 * @param string $type Can be 'date', 'time' or 'both'
2373 * @param string $ts The time format which needs to be turned into a
2374 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2375 * @param User $user User object used to get preferences for timezone and format
2376 * @param array $options Array, can contain the following keys:
2377 * - 'timecorrection': time correction, can have the following values:
2378 * - true: use user's preference
2379 * - false: don't use time correction
2380 * - int: value of time correction in minutes
2381 * - 'format': format to use, can have the following values:
2382 * - true: use user's preference
2383 * - false: use default preference
2384 * - string: format to use
2388 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2389 $ts = wfTimestamp( TS_MW
, $ts );
2390 $options +
= [ 'timecorrection' => true, 'format' => true ];
2391 if ( $options['timecorrection'] !== false ) {
2392 if ( $options['timecorrection'] === true ) {
2393 $offset = $user->getOption( 'timecorrection' );
2395 $offset = $options['timecorrection'];
2397 $ts = $this->userAdjust( $ts, $offset );
2399 if ( $options['format'] === true ) {
2400 $format = $user->getDatePreference();
2402 $format = $options['format'];
2404 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2405 return $this->sprintfDate( $df, $ts );
2409 * Get the formatted date for the given timestamp and formatted for
2412 * @param mixed $ts Mixed: the time format which needs to be turned into a
2413 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2414 * @param User $user User object used to get preferences for timezone and format
2415 * @param array $options Array, can contain the following keys:
2416 * - 'timecorrection': time correction, can have the following values:
2417 * - true: use user's preference
2418 * - false: don't use time correction
2419 * - int: value of time correction in minutes
2420 * - 'format': format to use, can have the following values:
2421 * - true: use user's preference
2422 * - false: use default preference
2423 * - string: format to use
2427 public function userDate( $ts, User
$user, array $options = [] ) {
2428 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2432 * Get the formatted time for the given timestamp and formatted for
2435 * @param mixed $ts The time format which needs to be turned into a
2436 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2437 * @param User $user User object used to get preferences for timezone and format
2438 * @param array $options Array, can contain the following keys:
2439 * - 'timecorrection': time correction, can have the following values:
2440 * - true: use user's preference
2441 * - false: don't use time correction
2442 * - int: value of time correction in minutes
2443 * - 'format': format to use, can have the following values:
2444 * - true: use user's preference
2445 * - false: use default preference
2446 * - string: format to use
2450 public function userTime( $ts, User
$user, array $options = [] ) {
2451 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2455 * Get the formatted date and time for the given timestamp and formatted for
2458 * @param mixed $ts The time format which needs to be turned into a
2459 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2460 * @param User $user User object used to get preferences for timezone and format
2461 * @param array $options Array, can contain the following keys:
2462 * - 'timecorrection': time correction, can have the following values:
2463 * - true: use user's preference
2464 * - false: don't use time correction
2465 * - int: value of time correction in minutes
2466 * - 'format': format to use, can have the following values:
2467 * - true: use user's preference
2468 * - false: use default preference
2469 * - string: format to use
2473 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2474 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2478 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2480 * Determine the difference between the timestamp and the current time, and
2481 * generate a readable timestamp by returning "<N> <units> ago", where the
2482 * largest possible unit is used.
2484 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2486 * @param MWTimestamp $time
2487 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2488 * @param User|null $user User the timestamp is being generated for
2489 * (or null to use main context's user)
2490 * @return string Formatted timestamp
2492 public function getHumanTimestamp(
2493 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2495 if ( $relativeTo === null ) {
2496 $relativeTo = new MWTimestamp();
2498 if ( $user === null ) {
2499 $user = RequestContext
::getMain()->getUser();
2502 // Adjust for the user's timezone.
2503 $offsetThis = $time->offsetForUser( $user );
2504 $offsetRel = $relativeTo->offsetForUser( $user );
2507 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2508 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2511 // Reset the timezone on the objects.
2512 $time->timestamp
->sub( $offsetThis );
2513 $relativeTo->timestamp
->sub( $offsetRel );
2519 * Convert an MWTimestamp into a pretty human-readable timestamp using
2520 * the given user preferences and relative base time.
2522 * @see Language::getHumanTimestamp
2523 * @param MWTimestamp $ts Timestamp to prettify
2524 * @param MWTimestamp $relativeTo Base timestamp
2525 * @param User $user User preferences to use
2526 * @return string Human timestamp
2529 private function getHumanTimestampInternal(
2530 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2532 $diff = $ts->diff( $relativeTo );
2533 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2534 (int)$relativeTo->timestamp
->format( 'w' ) );
2535 $days = $diff->days ?
: (int)$diffDay;
2536 if ( $diff->invert ||
$days > 5
2537 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2539 // Timestamps are in different years: use full timestamp
2540 // Also do full timestamp for future dates
2542 * @todo FIXME: Add better handling of future timestamps.
2544 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2545 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2546 } elseif ( $days > 5 ) {
2547 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2548 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2549 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2550 } elseif ( $days > 1 ) {
2551 // Timestamp within the past week: show the day of the week and time
2552 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2553 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2555 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2556 $ts = wfMessage( "$weekday-at" )
2557 ->inLanguage( $this )
2558 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2560 } elseif ( $days == 1 ) {
2561 // Timestamp was yesterday: say 'yesterday' and the time.
2562 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2563 $ts = wfMessage( 'yesterday-at' )
2564 ->inLanguage( $this )
2565 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2567 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2568 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2569 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2570 $ts = wfMessage( 'today-at' )
2571 ->inLanguage( $this )
2572 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2575 // From here on in, the timestamp was soon enough ago so that we can simply say
2576 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2577 } elseif ( $diff->h
== 1 ) {
2578 // Less than 90 minutes, but more than an hour ago.
2579 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2580 } elseif ( $diff->i
>= 1 ) {
2581 // A few minutes ago.
2582 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2583 } elseif ( $diff->s
>= 30 ) {
2584 // Less than a minute, but more than 30 sec ago.
2585 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2587 // Less than 30 seconds ago.
2588 $ts = wfMessage( 'just-now' )->text();
2595 * @param string $key
2596 * @return string|null
2598 public function getMessage( $key ) {
2599 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2605 function getAllMessages() {
2606 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2611 * @param string $out
2612 * @param string $string
2615 public function iconv( $in, $out, $string ) {
2616 # Even with //IGNORE iconv can whine about illegal characters in
2617 # *input* string. We just ignore those too.
2618 # REF: https://bugs.php.net/bug.php?id=37166
2619 # REF: https://phabricator.wikimedia.org/T18885
2620 Wikimedia\
suppressWarnings();
2621 $text = iconv( $in, $out . '//IGNORE', $string );
2622 Wikimedia\restoreWarnings
();
2626 // callback functions for ucwords(), ucwordbreaks()
2629 * @param array $matches
2630 * @return mixed|string
2632 function ucwordbreaksCallbackAscii( $matches ) {
2633 return $this->ucfirst( $matches[1] );
2637 * @param array $matches
2640 function ucwordbreaksCallbackMB( $matches ) {
2641 return mb_strtoupper( $matches[0] );
2645 * @param array $matches
2648 function ucwordsCallbackMB( $matches ) {
2649 return mb_strtoupper( $matches[0] );
2653 * Make a string's first character uppercase
2655 * @param string $str
2659 public function ucfirst( $str ) {
2661 if ( $o < 96 ) { // if already uppercase...
2663 } elseif ( $o < 128 ) {
2664 return ucfirst( $str ); // use PHP's ucfirst()
2666 // fall back to more complex logic in case of multibyte strings
2667 return $this->uc( $str, true );
2672 * Convert a string to uppercase
2674 * @param string $str
2675 * @param bool $first
2679 public function uc( $str, $first = false ) {
2681 if ( $this->isMultibyte( $str ) ) {
2682 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2684 return ucfirst( $str );
2687 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2692 * @param string $str
2693 * @return mixed|string
2695 function lcfirst( $str ) {
2698 return strval( $str );
2699 } elseif ( $o >= 128 ) {
2700 return $this->lc( $str, true );
2701 } elseif ( $o > 96 ) {
2704 $str[0] = strtolower( $str[0] );
2710 * @param string $str
2711 * @param bool $first
2712 * @return mixed|string
2714 function lc( $str, $first = false ) {
2716 if ( $this->isMultibyte( $str ) ) {
2717 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2719 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2722 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2727 * @param string $str
2730 function isMultibyte( $str ) {
2731 return strlen( $str ) !== mb_strlen( $str );
2735 * @param string $str
2736 * @return mixed|string
2738 function ucwords( $str ) {
2739 if ( $this->isMultibyte( $str ) ) {
2740 $str = $this->lc( $str );
2742 // regexp to find first letter in each word (i.e. after each space)
2743 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2745 // function to use to capitalize a single char
2746 return preg_replace_callback(
2748 [ $this, 'ucwordsCallbackMB' ],
2752 return ucwords( strtolower( $str ) );
2757 * capitalize words at word breaks
2759 * @param string $str
2762 function ucwordbreaks( $str ) {
2763 if ( $this->isMultibyte( $str ) ) {
2764 $str = $this->lc( $str );
2766 // since \b doesn't work for UTF-8, we explicitely define word break chars
2767 $breaks = "[ \-\(\)\}\{\.,\?!]";
2769 // find first letter after word break
2770 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2771 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2773 return preg_replace_callback(
2775 [ $this, 'ucwordbreaksCallbackMB' ],
2779 return preg_replace_callback(
2780 '/\b([\w\x80-\xff]+)\b/',
2781 [ $this, 'ucwordbreaksCallbackAscii' ],
2788 * Return a case-folded representation of $s
2790 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2791 * and $s2 are the same except for the case of their characters. It is not
2792 * necessary for the value returned to make sense when displayed.
2794 * Do *not* perform any other normalisation in this function. If a caller
2795 * uses this function when it should be using a more general normalisation
2796 * function, then fix the caller.
2802 function caseFold( $s ) {
2803 return $this->uc( $s );
2809 * @throws MWException
2811 function checkTitleEncoding( $s ) {
2812 if ( is_array( $s ) ) {
2813 throw new MWException( 'Given array to checkTitleEncoding.' );
2815 if ( StringUtils
::isUtf8( $s ) ) {
2819 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2825 function fallback8bitEncoding() {
2826 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2830 * Most writing systems use whitespace to break up words.
2831 * Some languages such as Chinese don't conventionally do this,
2832 * which requires special handling when breaking up words for
2837 function hasWordBreaks() {
2842 * Some languages such as Chinese require word segmentation,
2843 * Specify such segmentation when overridden in derived class.
2845 * @param string $string
2848 function segmentByWord( $string ) {
2853 * Some languages have special punctuation need to be normalized.
2854 * Make such changes here.
2856 * @param string $string
2859 function normalizeForSearch( $string ) {
2860 return self
::convertDoubleWidth( $string );
2864 * convert double-width roman characters to single-width.
2865 * range: ff00-ff5f ~= 0020-007f
2867 * @param string $string
2871 protected static function convertDoubleWidth( $string ) {
2872 static $full = null;
2873 static $half = null;
2875 if ( $full === null ) {
2876 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2877 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2878 $full = str_split( $fullWidth, 3 );
2879 $half = str_split( $halfWidth );
2882 $string = str_replace( $full, $half, $string );
2887 * @param string $string
2888 * @param string $pattern
2891 protected static function insertSpace( $string, $pattern ) {
2892 $string = preg_replace( $pattern, " $1 ", $string );
2893 $string = preg_replace( '/ +/', ' ', $string );
2898 * @param array $termsArray
2901 function convertForSearchResult( $termsArray ) {
2902 # some languages, e.g. Chinese, need to do a conversion
2903 # in order for search results to be displayed correctly
2908 * Get the first character of a string.
2913 function firstChar( $s ) {
2916 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2917 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2922 if ( isset( $matches[1] ) ) {
2923 if ( strlen( $matches[1] ) != 3 ) {
2927 // Break down Hangul syllables to grab the first jamo
2928 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
2929 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2931 } elseif ( $code < 0xb098 ) {
2932 return "\xe3\x84\xb1";
2933 } elseif ( $code < 0xb2e4 ) {
2934 return "\xe3\x84\xb4";
2935 } elseif ( $code < 0xb77c ) {
2936 return "\xe3\x84\xb7";
2937 } elseif ( $code < 0xb9c8 ) {
2938 return "\xe3\x84\xb9";
2939 } elseif ( $code < 0xbc14 ) {
2940 return "\xe3\x85\x81";
2941 } elseif ( $code < 0xc0ac ) {
2942 return "\xe3\x85\x82";
2943 } elseif ( $code < 0xc544 ) {
2944 return "\xe3\x85\x85";
2945 } elseif ( $code < 0xc790 ) {
2946 return "\xe3\x85\x87";
2947 } elseif ( $code < 0xcc28 ) {
2948 return "\xe3\x85\x88";
2949 } elseif ( $code < 0xce74 ) {
2950 return "\xe3\x85\x8a";
2951 } elseif ( $code < 0xd0c0 ) {
2952 return "\xe3\x85\x8b";
2953 } elseif ( $code < 0xd30c ) {
2954 return "\xe3\x85\x8c";
2955 } elseif ( $code < 0xd558 ) {
2956 return "\xe3\x85\x8d";
2958 return "\xe3\x85\x8e";
2966 * @deprecated No-op since 1.28
2968 function initEncoding() {
2975 * @deprecated No-op since 1.28
2977 function recodeForEdit( $s ) {
2984 * @deprecated No-op since 1.28
2986 function recodeInput( $s ) {
2991 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2992 * also cleans up certain backwards-compatible sequences, converting them
2993 * to the modern Unicode equivalent.
2995 * This is language-specific for performance reasons only.
3001 function normalize( $s ) {
3002 global $wgAllUnicodeFixes;
3003 $s = UtfNormal\Validator
::cleanUp( $s );
3004 if ( $wgAllUnicodeFixes ) {
3005 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3006 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3013 * Transform a string using serialized data stored in the given file (which
3014 * must be in the serialized subdirectory of $IP). The file contains pairs
3015 * mapping source characters to destination characters.
3017 * The data is cached in process memory. This will go faster if you have the
3018 * FastStringSearch extension.
3020 * @param string $file
3021 * @param string $string
3023 * @throws MWException
3026 function transformUsingPairFile( $file, $string ) {
3027 if ( !isset( $this->transformData
[$file] ) ) {
3028 $data = wfGetPrecompiledData( $file );
3029 if ( $data === false ) {
3030 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3032 $this->transformData
[$file] = new ReplacementArray( $data );
3034 return $this->transformData
[$file]->replace( $string );
3038 * For right-to-left language support
3043 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3047 * Return the correct HTML 'dir' attribute value for this language.
3051 return $this->isRTL() ?
'rtl' : 'ltr';
3055 * Return 'left' or 'right' as appropriate alignment for line-start
3056 * for this language's text direction.
3058 * Should be equivalent to CSS3 'start' text-align value....
3062 function alignStart() {
3063 return $this->isRTL() ?
'right' : 'left';
3067 * Return 'right' or 'left' as appropriate alignment for line-end
3068 * for this language's text direction.
3070 * Should be equivalent to CSS3 'end' text-align value....
3074 function alignEnd() {
3075 return $this->isRTL() ?
'left' : 'right';
3079 * A hidden direction mark (LRM or RLM), depending on the language direction.
3080 * Unlike getDirMark(), this function returns the character as an HTML entity.
3081 * This function should be used when the output is guaranteed to be HTML,
3082 * because it makes the output HTML source code more readable. When
3083 * the output is plain text or can be escaped, getDirMark() should be used.
3085 * @param bool $opposite Get the direction mark opposite to your language
3089 function getDirMarkEntity( $opposite = false ) {
3091 return $this->isRTL() ?
'‎' : '‏';
3093 return $this->isRTL() ?
'‏' : '‎';
3097 * A hidden direction mark (LRM or RLM), depending on the language direction.
3098 * This function produces them as invisible Unicode characters and
3099 * the output may be hard to read and debug, so it should only be used
3100 * when the output is plain text or can be escaped. When the output is
3101 * HTML, use getDirMarkEntity() instead.
3103 * @param bool $opposite Get the direction mark opposite to your language
3106 function getDirMark( $opposite = false ) {
3107 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3108 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3110 return $this->isRTL() ?
$lrm : $rlm;
3112 return $this->isRTL() ?
$rlm : $lrm;
3118 function capitalizeAllNouns() {
3119 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3123 * An arrow, depending on the language direction.
3125 * @param string $direction The direction of the arrow: forwards (default),
3126 * backwards, left, right, up, down.
3129 function getArrow( $direction = 'forwards' ) {
3130 switch ( $direction ) {
3132 return $this->isRTL() ?
'←' : '→';
3134 return $this->isRTL() ?
'→' : '←';
3147 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3151 function linkPrefixExtension() {
3152 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3156 * Get all magic words from cache.
3159 function getMagicWords() {
3160 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3164 * Run the LanguageGetMagic hook once.
3166 protected function doMagicHook() {
3167 if ( $this->mMagicHookDone
) {
3170 $this->mMagicHookDone
= true;
3171 Hooks
::run( 'LanguageGetMagic', [ &$this->mMagicExtensions
, $this->getCode() ] );
3175 * Fill a MagicWord object with data from here
3177 * @param MagicWord $mw
3179 function getMagic( $mw ) {
3180 // Saves a function call
3181 if ( !$this->mMagicHookDone
) {
3182 $this->doMagicHook();
3185 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3186 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3188 $rawEntry = self
::$dataCache->getSubitem(
3189 $this->mCode
, 'magicWords', $mw->mId
);
3192 if ( !is_array( $rawEntry ) ) {
3193 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3195 $mw->mCaseSensitive
= $rawEntry[0];
3196 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3201 * Add magic words to the extension array
3203 * @param array $newWords
3205 function addMagicWordsByLang( $newWords ) {
3206 $fallbackChain = $this->getFallbackLanguages();
3207 $fallbackChain = array_reverse( $fallbackChain );
3208 foreach ( $fallbackChain as $code ) {
3209 if ( isset( $newWords[$code] ) ) {
3210 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3216 * Get special page names, as an associative array
3217 * canonical name => array of valid names, including aliases
3220 function getSpecialPageAliases() {
3221 // Cache aliases because it may be slow to load them
3222 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3224 $this->mExtendedSpecialPageAliases
=
3225 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3226 Hooks
::run( 'LanguageGetSpecialPageAliases',
3227 [ &$this->mExtendedSpecialPageAliases
, $this->getCode() ] );
3230 return $this->mExtendedSpecialPageAliases
;
3234 * Italic is unsuitable for some languages
3236 * @param string $text The text to be emphasized.
3239 function emphasize( $text ) {
3240 return "<em>$text</em>";
3244 * Normally we output all numbers in plain en_US style, that is
3245 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3246 * point twohundredthirtyfive. However this is not suitable for all
3247 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3248 * Icelandic just want to use commas instead of dots, and dots instead
3249 * of commas like "293.291,235".
3251 * An example of this function being called:
3253 * wfMessage( 'message' )->numParams( $num )->text()
3256 * See $separatorTransformTable on MessageIs.php for
3257 * the , => . and . => , implementation.
3259 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3260 * @param int|float $number The string to be formatted, should be an integer
3261 * or a floating point number.
3262 * @param bool $nocommafy Set to true for special numbers like dates
3265 public function formatNum( $number, $nocommafy = false ) {
3266 global $wgTranslateNumerals;
3267 if ( !$nocommafy ) {
3268 $number = $this->commafy( $number );
3269 $s = $this->separatorTransformTable();
3271 $number = strtr( $number, $s );
3275 if ( $wgTranslateNumerals ) {
3276 $s = $this->digitTransformTable();
3278 $number = strtr( $number, $s );
3282 return (string)$number;
3286 * Front-end for non-commafied formatNum
3288 * @param int|float $number The string to be formatted, should be an integer
3289 * or a floating point number.
3293 public function formatNumNoSeparators( $number ) {
3294 return $this->formatNum( $number, true );
3298 * @param string $number
3301 public function parseFormattedNumber( $number ) {
3302 $s = $this->digitTransformTable();
3304 // eliminate empty array values such as ''. (T66347)
3305 $s = array_filter( $s );
3306 $number = strtr( $number, array_flip( $s ) );
3309 $s = $this->separatorTransformTable();
3311 // eliminate empty array values such as ''. (T66347)
3312 $s = array_filter( $s );
3313 $number = strtr( $number, array_flip( $s ) );
3316 $number = strtr( $number, [ ',' => '' ] );
3321 * Adds commas to a given number
3323 * @param mixed $number
3326 function commafy( $number ) {
3327 $digitGroupingPattern = $this->digitGroupingPattern();
3328 $minimumGroupingDigits = $this->minimumGroupingDigits();
3329 if ( $number === null ) {
3333 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3334 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3335 // In some languages it's conventional not to insert a thousands separator
3336 // in numbers that are four digits long (1000-9999).
3337 if ( $minimumGroupingDigits ) {
3338 // Number of '#' characters after last comma in the grouping pattern.
3339 // The pattern is hardcoded here, but this would vary for different patterns.
3340 $primaryGroupingSize = 3;
3341 // Maximum length of a number to suppress digit grouping for.
3342 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3343 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3347 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3349 // Ref: http://cldr.unicode.org/translation/number-patterns
3351 if ( intval( $number ) < 0 ) {
3352 // For negative numbers apply the algorithm like positive number and add sign.
3354 $number = substr( $number, 1 );
3358 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3359 preg_match( "/\d+/", $number, $integerPart );
3360 preg_match( "/\.\d*/", $number, $decimalPart );
3361 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3362 if ( $groupedNumber === $number ) {
3363 // the string does not have any number part. Eg: .12345
3364 return $sign . $groupedNumber;
3366 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3367 while ( $start > 0 ) {
3368 $match = $matches[0][$numMatches - 1];
3369 $matchLen = strlen( $match );
3370 $start = $end - $matchLen;
3374 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3376 if ( $numMatches > 1 ) {
3377 // use the last pattern for the rest of the number
3381 $groupedNumber = "," . $groupedNumber;
3384 return $sign . $groupedNumber;
3391 function digitGroupingPattern() {
3392 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3398 function digitTransformTable() {
3399 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3405 function separatorTransformTable() {
3406 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3412 function minimumGroupingDigits() {
3413 return self
::$dataCache->getItem( $this->mCode
, 'minimumGroupingDigits' );
3417 * Take a list of strings and build a locale-friendly comma-separated
3418 * list, using the local comma-separator message.
3419 * The last two strings are chained with an "and".
3420 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3422 * @param string[] $l
3425 function listToText( array $l ) {
3426 $m = count( $l ) - 1;
3431 $and = $this->msg( 'and' )->escaped();
3432 $space = $this->msg( 'word-separator' )->escaped();
3434 $comma = $this->msg( 'comma-separator' )->escaped();
3438 for ( $i = $m - 1; $i >= 0; $i-- ) {
3439 if ( $i == $m - 1 ) {
3440 $s = $l[$i] . $and . $space . $s;
3442 $s = $l[$i] . $comma . $s;
3449 * Take a list of strings and build a locale-friendly comma-separated
3450 * list, using the local comma-separator message.
3451 * @param string[] $list Array of strings to put in a comma list
3454 function commaList( array $list ) {
3456 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3462 * Take a list of strings and build a locale-friendly semicolon-separated
3463 * list, using the local semicolon-separator message.
3464 * @param string[] $list Array of strings to put in a semicolon list
3467 function semicolonList( array $list ) {
3469 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3475 * Same as commaList, but separate it with the pipe instead.
3476 * @param string[] $list Array of strings to put in a pipe list
3479 function pipeList( array $list ) {
3481 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3487 * This method is deprecated since 1.31 and kept as alias for truncateForDatabase, which
3488 * has replaced it. This method provides truncation suitable for DB.
3490 * The database offers limited byte lengths for some columns in the database;
3491 * multi-byte character sets mean we need to ensure that only whole characters
3492 * are included, otherwise broken characters can be passed to the user.
3494 * @deprecated since 1.31, use truncateForDatabase or truncateForVisual as appropriate.
3496 * @param string $string String to truncate
3497 * @param int $length Maximum length (including ellipsis)
3498 * @param string $ellipsis String to append to the truncated text
3499 * @param bool $adjustLength Subtract length of ellipsis from $length.
3500 * $adjustLength was introduced in 1.18, before that behaved as if false.
3503 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3504 return $this->truncateForDatabase( $string, $length, $ellipsis, $adjustLength );
3508 * Truncate a string to a specified length in bytes, appending an optional
3509 * string (e.g. for ellipsis)
3511 * If $length is negative, the string will be truncated from the beginning
3515 * @param string $string String to truncate
3516 * @param int $length Maximum length in bytes
3517 * @param string $ellipsis String to append to the end of truncated text
3518 * @param bool $adjustLength Subtract length of ellipsis from $length
3522 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3523 return $this->truncateInternal(
3524 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3529 * Truncate a string to a specified number of characters, appending an optional
3530 * string (e.g. for ellipsis).
3532 * This provides multibyte version of truncate() method of this class, suitable for truncation
3533 * based on number of characters, instead of number of bytes.
3535 * If $length is negative, the string will be truncated from the beginning.
3539 * @param string $string String to truncate
3540 * @param int $length Maximum number of characters
3541 * @param string $ellipsis String to append to the end of truncated text
3542 * @param bool $adjustLength Subtract length of ellipsis from $length
3546 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3547 // Passing encoding to mb_strlen and mb_substr is optional.
3548 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3549 // explicit specification of encoding is skipped.
3550 // Note: Both multibyte methods are callables invoked in truncateInternal.
3551 return $this->truncateInternal(
3552 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3557 * Internal method used for truncation. This method abstracts text truncation into
3558 * one common method, allowing users to provide length measurement function and
3559 * function for finding substring.
3561 * For usages, see truncateForDatabase and truncateForVisual.
3563 * @param string $string String to truncate
3564 * @param int $length Maximum length of final text
3565 * @param string $ellipsis String to append to the end of truncated text
3566 * @param bool $adjustLength Subtract length of ellipsis from $length
3567 * @param callable $measureLength Callable function used for determining the length of text
3568 * @param callable $getSubstring Callable function used for getting the substrings
3572 private function truncateInternal(
3573 $string, $length, $ellipsis = '...', $adjustLength = true, $measureLength, $getSubstring
3575 if ( !is_callable( $measureLength ) ||
!is_callable( $getSubstring ) ) {
3576 throw new InvalidArgumentException( 'Invalid callback provided' );
3579 # Check if there is no need to truncate
3580 if ( $measureLength( $string ) <= abs( $length ) ) {
3581 return $string; // no need to truncate
3584 # Use the localized ellipsis character
3585 if ( $ellipsis == '...' ) {
3586 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3588 if ( $length == 0 ) {
3589 return $ellipsis; // convention
3592 $stringOriginal = $string;
3593 # If ellipsis length is >= $length then we can't apply $adjustLength
3594 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3595 $string = $ellipsis; // this can be slightly unexpected
3596 # Otherwise, truncate and add ellipsis...
3598 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3599 if ( $length > 0 ) {
3600 $length -= $ellipsisLength;
3601 $string = $getSubstring( $string, 0, $length ); // xyz...
3602 $string = $this->removeBadCharLast( $string );
3603 $string = rtrim( $string );
3604 $string = $string . $ellipsis;
3606 $length +
= $ellipsisLength;
3607 $string = $getSubstring( $string, $length ); // ...xyz
3608 $string = $this->removeBadCharFirst( $string );
3609 $string = ltrim( $string );
3610 $string = $ellipsis . $string;
3614 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3615 # This check is *not* redundant if $adjustLength, due to the single case where
3616 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3617 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3620 return $stringOriginal;
3625 * Remove bytes that represent an incomplete Unicode character
3626 * at the end of string (e.g. bytes of the char are missing)
3628 * @param string $string
3631 protected function removeBadCharLast( $string ) {
3632 if ( $string != '' ) {
3633 $char = ord( $string[strlen( $string ) - 1] );
3635 if ( $char >= 0xc0 ) {
3636 # We got the first byte only of a multibyte char; remove it.
3637 $string = substr( $string, 0, -1 );
3638 } elseif ( $char >= 0x80 &&
3639 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3640 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3641 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3643 # We chopped in the middle of a character; remove it
3651 * Remove bytes that represent an incomplete Unicode character
3652 * at the start of string (e.g. bytes of the char are missing)
3654 * @param string $string
3657 protected function removeBadCharFirst( $string ) {
3658 if ( $string != '' ) {
3659 $char = ord( $string[0] );
3660 if ( $char >= 0x80 && $char < 0xc0 ) {
3661 # We chopped in the middle of a character; remove the whole thing
3662 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3669 * Truncate a string of valid HTML to a specified length in bytes,
3670 * appending an optional string (e.g. for ellipses), and return valid HTML
3672 * This is only intended for styled/linked text, such as HTML with
3673 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3674 * Also, this will not detect things like "display:none" CSS.
3676 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3678 * @param string $text HTML string to truncate
3679 * @param int $length (zero/positive) Maximum length (including ellipses)
3680 * @param string $ellipsis String to append to the truncated text
3683 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3684 # Use the localized ellipsis character
3685 if ( $ellipsis == '...' ) {
3686 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3688 # Check if there is clearly no need to truncate
3689 if ( $length <= 0 ) {
3690 return $ellipsis; // no text shown, nothing to format (convention)
3691 } elseif ( strlen( $text ) <= $length ) {
3692 return $text; // string short enough even *with* HTML (short-circuit)
3695 $dispLen = 0; // innerHTML legth so far
3696 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3697 $tagType = 0; // 0-open, 1-close
3698 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3699 $entityState = 0; // 0-not entity, 1-entity
3700 $tag = $ret = ''; // accumulated tag name, accumulated result string
3701 $openTags = []; // open tag stack
3702 $maybeState = null; // possible truncation state
3704 $textLen = strlen( $text );
3705 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3706 for ( $pos = 0; true; ++
$pos ) {
3707 # Consider truncation once the display length has reached the maximim.
3708 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3709 # Check that we're not in the middle of a bracket/entity...
3710 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3711 if ( !$testingEllipsis ) {
3712 $testingEllipsis = true;
3713 # Save where we are; we will truncate here unless there turn out to
3714 # be so few remaining characters that truncation is not necessary.
3715 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3716 $maybeState = [ $ret, $openTags ]; // save state
3718 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3719 # String in fact does need truncation, the truncation point was OK.
3720 list( $ret, $openTags ) = $maybeState; // reload state
3721 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3722 $ret .= $ellipsis; // add ellipsis
3726 if ( $pos >= $textLen ) {
3727 break; // extra iteration just for above checks
3730 # Read the next char...
3732 $lastCh = $pos ?
$text[$pos - 1] : '';
3733 $ret .= $ch; // add to result string
3735 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3736 $entityState = 0; // for bad HTML
3737 $bracketState = 1; // tag started (checking for backslash)
3738 } elseif ( $ch == '>' ) {
3739 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3740 $entityState = 0; // for bad HTML
3741 $bracketState = 0; // out of brackets
3742 } elseif ( $bracketState == 1 ) {
3744 $tagType = 1; // close tag (e.g. "</span>")
3746 $tagType = 0; // open tag (e.g. "<span>")
3749 $bracketState = 2; // building tag name
3750 } elseif ( $bracketState == 2 ) {
3754 // Name found (e.g. "<a href=..."), add on tag attributes...
3755 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3757 } elseif ( $bracketState == 0 ) {
3758 if ( $entityState ) {
3761 $dispLen++
; // entity is one displayed char
3764 if ( $neLength == 0 && !$maybeState ) {
3765 // Save state without $ch. We want to *hit* the first
3766 // display char (to get tags) but not *use* it if truncating.
3767 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3770 $entityState = 1; // entity found, (e.g. " ")
3772 $dispLen++
; // this char is displayed
3773 // Add the next $max display text chars after this in one swoop...
3774 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3775 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3776 $dispLen +
= $skipped;
3782 // Close the last tag if left unclosed by bad HTML
3783 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3784 while ( count( $openTags ) > 0 ) {
3785 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3791 * truncateHtml() helper function
3792 * like strcspn() but adds the skipped chars to $ret
3794 * @param string $ret
3795 * @param string $text
3796 * @param string $search
3798 * @param null|int $len
3801 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3802 if ( $len === null ) {
3803 $len = -1; // -1 means "no limit" for strcspn
3804 } elseif ( $len < 0 ) {
3808 if ( $start < strlen( $text ) ) {
3809 $skipCount = strcspn( $text, $search, $start, $len );
3810 $ret .= substr( $text, $start, $skipCount );
3816 * truncateHtml() helper function
3817 * (a) push or pop $tag from $openTags as needed
3818 * (b) clear $tag value
3819 * @param string &$tag Current HTML tag name we are looking at
3820 * @param int $tagType (0-open tag, 1-close tag)
3821 * @param string $lastCh Character before the '>' that ended this tag
3822 * @param array &$openTags Open tag stack (not accounting for $tag)
3824 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3825 $tag = ltrim( $tag );
3827 if ( $tagType == 0 && $lastCh != '/' ) {
3828 $openTags[] = $tag; // tag opened (didn't close itself)
3829 } elseif ( $tagType == 1 ) {
3830 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3831 array_pop( $openTags ); // tag closed
3839 * Grammatical transformations, needed for inflected languages
3840 * Invoked by putting {{grammar:case|word}} in a message
3842 * @param string $word
3843 * @param string $case
3846 function convertGrammar( $word, $case ) {
3847 global $wgGrammarForms;
3848 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3849 return $wgGrammarForms[$this->getCode()][$case][$word];
3852 $grammarTransformations = $this->getGrammarTransformations();
3854 if ( isset( $grammarTransformations[$case] ) ) {
3855 $forms = $grammarTransformations[$case];
3857 // Some names of grammar rules are aliases for other rules.
3858 // In such cases the value is a string rather than object,
3859 // so load the actual rules.
3860 if ( is_string( $forms ) ) {
3861 $forms = $grammarTransformations[$forms];
3864 foreach ( array_values( $forms ) as $rule ) {
3867 if ( $form === '@metadata' ) {
3871 $replacement = $rule[1];
3873 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3874 $patternMatches = preg_match( $regex, $word );
3876 if ( $patternMatches === false ) {
3878 'An error occurred while processing grammar. ' .
3879 "Word: '$word'. Regex: /$form/."
3881 } elseif ( $patternMatches === 1 ) {
3882 $word = preg_replace( $regex, $replacement, $word );
3893 * Get the grammar forms for the content language
3894 * @return array Array of grammar forms
3897 function getGrammarForms() {
3898 global $wgGrammarForms;
3899 if ( isset( $wgGrammarForms[$this->getCode()] )
3900 && is_array( $wgGrammarForms[$this->getCode()] )
3902 return $wgGrammarForms[$this->getCode()];
3909 * Get the grammar transformations data for the language.
3910 * Used like grammar forms, with {{GRAMMAR}} and cases,
3911 * but uses pairs of regexes and replacements instead of code.
3913 * @return array[] Array of grammar transformations.
3914 * @throws MWException
3917 public function getGrammarTransformations() {
3918 $languageCode = $this->getCode();
3920 if ( self
::$grammarTransformations === null ) {
3921 self
::$grammarTransformations = new MapCacheLRU( 10 );
3924 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3925 return self
::$grammarTransformations->get( $languageCode );
3930 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3931 if ( is_readable( $grammarDataFile ) ) {
3932 $data = FormatJson
::decode(
3933 file_get_contents( $grammarDataFile ),
3937 if ( $data === null ) {
3938 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3941 self
::$grammarTransformations->set( $languageCode, $data );
3948 * Provides an alternative text depending on specified gender.
3949 * Usage {{gender:username|masculine|feminine|unknown}}.
3950 * username is optional, in which case the gender of current user is used,
3951 * but only in (some) interface messages; otherwise default gender is used.
3953 * If no forms are given, an empty string is returned. If only one form is
3954 * given, it will be returned unconditionally. These details are implied by
3955 * the caller and cannot be overridden in subclasses.
3957 * If three forms are given, the default is to use the third (unknown) form.
3958 * If fewer than three forms are given, the default is to use the first (masculine) form.
3959 * These details can be overridden in subclasses.
3961 * @param string $gender
3962 * @param array $forms
3966 function gender( $gender, $forms ) {
3967 if ( !count( $forms ) ) {
3970 $forms = $this->preConvertPlural( $forms, 2 );
3971 if ( $gender === 'male' ) {
3974 if ( $gender === 'female' ) {
3977 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3981 * Plural form transformations, needed for some languages.
3982 * For example, there are 3 form of plural in Russian and Polish,
3983 * depending on "count mod 10". See [[w:Plural]]
3984 * For English it is pretty simple.
3986 * Invoked by putting {{plural:count|wordform1|wordform2}}
3987 * or {{plural:count|wordform1|wordform2|wordform3}}
3989 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3991 * @param int $count Non-localized number
3992 * @param array $forms Different plural forms
3993 * @return string Correct form of plural for $count in this language
3995 function convertPlural( $count, $forms ) {
3996 // Handle explicit n=pluralform cases
3997 $forms = $this->handleExplicitPluralForms( $count, $forms );
3998 if ( is_string( $forms ) ) {
4001 if ( !count( $forms ) ) {
4005 $pluralForm = $this->getPluralRuleIndexNumber( $count );
4006 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
4007 return $forms[$pluralForm];
4011 * Handles explicit plural forms for Language::convertPlural()
4013 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4014 * If an explicitly defined plural form matches the $count, then
4015 * string value returned, otherwise array returned for further consideration
4016 * by CLDR rules or overridden convertPlural().
4020 * @param int $count Non-localized number
4021 * @param array $forms Different plural forms
4023 * @return array|string
4025 protected function handleExplicitPluralForms( $count, array $forms ) {
4026 foreach ( $forms as $index => $form ) {
4027 if ( preg_match( '/\d+=/i', $form ) ) {
4028 $pos = strpos( $form, '=' );
4029 if ( substr( $form, 0, $pos ) === (string)$count ) {
4030 return substr( $form, $pos +
1 );
4032 unset( $forms[$index] );
4035 return array_values( $forms );
4039 * Checks that convertPlural was given an array and pads it to requested
4040 * amount of forms by copying the last one.
4042 * @param array $forms Array of forms given to convertPlural
4043 * @param int $count How many forms should there be at least
4044 * @return array Padded array of forms or an exception if not an array
4046 protected function preConvertPlural( /* Array */ $forms, $count ) {
4047 while ( count( $forms ) < $count ) {
4048 $forms[] = $forms[count( $forms ) - 1];
4054 * Wraps argument with unicode control characters for directionality safety
4056 * This solves the problem where directionality-neutral characters at the edge of
4057 * the argument string get interpreted with the wrong directionality from the
4058 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4060 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4061 * directionality of the argument string, using the BIDI algorithm's own "First
4062 * strong directional codepoint" rule. Essentially, this works round the fact that
4063 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4064 * direction inference). The latter is cleaner but still not widely supported.
4066 * @param string $text Text to wrap
4067 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4069 public function embedBidi( $text = '' ) {
4070 $dir = self
::strongDirFromContent( $text );
4071 if ( $dir === 'ltr' ) {
4072 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4073 return self
::$lre . $text . self
::$pdf;
4075 if ( $dir === 'rtl' ) {
4076 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4077 return self
::$rle . $text . self
::$pdf;
4079 // No strong directionality: do not wrap
4084 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4085 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4086 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4087 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4090 * @param string $str The validated block duration in English
4091 * @param User $user User object to use timezone from or null for $wgUser
4092 * @param int $now Current timestamp, for formatting relative block durations
4093 * @return string Somehow translated block duration
4094 * @see LanguageFi.php for example implementation
4096 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
4097 $duration = SpecialBlock
::getSuggestedDurations( $this );
4098 foreach ( $duration as $show => $value ) {
4099 if ( strcmp( $str, $value ) == 0 ) {
4100 return htmlspecialchars( trim( $show ) );
4104 if ( wfIsInfinity( $str ) ) {
4105 foreach ( $duration as $show => $value ) {
4106 if ( wfIsInfinity( $value ) ) {
4107 return htmlspecialchars( trim( $show ) );
4112 // If all else fails, return a standard duration or timestamp description.
4113 $time = strtotime( $str, $now );
4114 if ( $time === false ) { // Unknown format. Return it as-is in case.
4116 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
4117 // The result differs based on current time, so the difference
4118 // is a fixed duration length.
4119 return $this->formatDuration( $time - $now );
4120 } else { // It's an absolute timestamp.
4121 if ( $time === 0 ) {
4122 // wfTimestamp() handles 0 as current time instead of epoch.
4123 $time = '19700101000000';
4126 return $this->userTimeAndDate( $time, $user );
4128 return $this->timeanddate( $time );
4133 * languages like Chinese need to be segmented in order for the diff
4136 * @param string $text
4139 public function segmentForDiff( $text ) {
4144 * and unsegment to show the result
4146 * @param string $text
4149 public function unsegmentForDiff( $text ) {
4154 * Return the LanguageConverter used in the Language
4157 * @return LanguageConverter
4159 public function getConverter() {
4160 return $this->mConverter
;
4164 * convert text to a variant
4166 * @param string $text text to convert
4167 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4168 * variant (if logged in), or the project default variant
4169 * @return string the converted string
4171 public function autoConvert( $text, $variant = false ) {
4172 return $this->mConverter
->autoConvert( $text, $variant );
4176 * convert text to all supported variants
4178 * @param string $text
4181 public function autoConvertToAllVariants( $text ) {
4182 return $this->mConverter
->autoConvertToAllVariants( $text );
4186 * convert text to different variants of a language.
4188 * @param string $text
4191 public function convert( $text ) {
4192 return $this->mConverter
->convert( $text );
4196 * Convert a Title object to a string in the preferred variant
4198 * @param Title $title
4201 public function convertTitle( $title ) {
4202 return $this->mConverter
->convertTitle( $title );
4206 * Convert a namespace index to a string in the preferred variant
4208 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4209 * @param string|null $variant variant to convert to, or null to use the user's preferred
4210 * variant (if logged in), or the project default variant
4211 * @return string a string representation of the namespace
4213 public function convertNamespace( $ns, $variant = null ) {
4214 return $this->mConverter
->convertNamespace( $ns, $variant );
4218 * Check if this is a language with variants
4222 public function hasVariants() {
4223 return count( $this->getVariants() ) > 1;
4227 * Check if the language has the specific variant
4230 * @param string $variant
4233 public function hasVariant( $variant ) {
4234 return (bool)$this->mConverter
->validateVariant( $variant );
4238 * Perform output conversion on a string, and encode for safe HTML output.
4239 * @param string $text Text to be converted
4240 * @param bool $isTitle Whether this conversion is for the article title
4242 * @todo this should get integrated somewhere sane
4244 public function convertHtml( $text, $isTitle = false ) {
4245 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4249 * @param string $key
4252 public function convertCategoryKey( $key ) {
4253 return $this->mConverter
->convertCategoryKey( $key );
4257 * Get the list of variants supported by this language
4258 * see sample implementation in LanguageZh.php
4260 * @return string[] An array of language codes
4262 public function getVariants() {
4263 return $this->mConverter
->getVariants();
4269 public function getPreferredVariant() {
4270 return $this->mConverter
->getPreferredVariant();
4276 public function getDefaultVariant() {
4277 return $this->mConverter
->getDefaultVariant();
4283 public function getURLVariant() {
4284 return $this->mConverter
->getURLVariant();
4288 * If a language supports multiple variants, it is
4289 * possible that non-existing link in one variant
4290 * actually exists in another variant. this function
4291 * tries to find it. See e.g. LanguageZh.php
4292 * The input parameters may be modified upon return
4294 * @param string &$link The name of the link
4295 * @param Title &$nt The title object of the link
4296 * @param bool $ignoreOtherCond To disable other conditions when
4297 * we need to transclude a template or update a category's link
4299 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4300 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4304 * returns language specific options used by User::getPageRenderHash()
4305 * for example, the preferred language variant
4309 function getExtraHashOptions() {
4310 return $this->mConverter
->getExtraHashOptions();
4314 * For languages that support multiple variants, the title of an
4315 * article may be displayed differently in different variants. this
4316 * function returns the apporiate title defined in the body of the article.
4320 public function getParsedTitle() {
4321 return $this->mConverter
->getParsedTitle();
4325 * Refresh the cache of conversion tables when
4326 * MediaWiki:Conversiontable* is updated.
4328 * @param Title $title The Title of the page being updated
4330 public function updateConversionTable( Title
$title ) {
4331 $this->mConverter
->updateConversionTable( $title );
4335 * Prepare external link text for conversion. When the text is
4336 * a URL, it shouldn't be converted, and it'll be wrapped in
4337 * the "raw" tag (-{R| }-) to prevent conversion.
4339 * This function is called "markNoConversion" for historical
4342 * @param string $text Text to be used for external link
4343 * @param bool $noParse Wrap it without confirming it's a real URL first
4344 * @return string The tagged text
4346 public function markNoConversion( $text, $noParse = false ) {
4347 // Excluding protocal-relative URLs may avoid many false positives.
4348 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4349 return $this->mConverter
->markNoConversion( $text );
4356 * A regular expression to match legal word-trailing characters
4357 * which should be merged onto a link of the form [[foo]]bar.
4361 public function linkTrail() {
4362 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4366 * A regular expression character set to match legal word-prefixing
4367 * characters which should be merged onto a link of the form foo[[bar]].
4371 public function linkPrefixCharset() {
4372 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4376 * Get the "parent" language which has a converter to convert a "compatible" language
4377 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4379 * @return Language|null
4382 public function getParentLanguage() {
4383 if ( $this->mParentLanguage
!== false ) {
4384 return $this->mParentLanguage
;
4387 $code = explode( '-', $this->getCode() )[0];
4388 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4389 $this->mParentLanguage
= null;
4392 $lang = self
::factory( $code );
4393 if ( !$lang->hasVariant( $this->getCode() ) ) {
4394 $this->mParentLanguage
= null;
4398 $this->mParentLanguage
= $lang;
4403 * Compare with an other language object
4406 * @param Language $lang
4409 public function equals( Language
$lang ) {
4410 return $lang->getCode() === $this->mCode
;
4414 * Get the internal language code for this language object
4416 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4417 * htmlspecialchars() or similar
4421 public function getCode() {
4422 return $this->mCode
;
4426 * Get the code in BCP 47 format which we can use
4427 * inside of html lang="" tags.
4429 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4430 * htmlspecialchars() or similar.
4435 public function getHtmlCode() {
4436 if ( is_null( $this->mHtmlCode
) ) {
4437 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4439 return $this->mHtmlCode
;
4443 * @param string $code
4445 public function setCode( $code ) {
4446 $this->mCode
= $code;
4447 // Ensure we don't leave incorrect cached data lying around
4448 $this->mHtmlCode
= null;
4449 $this->mParentLanguage
= false;
4453 * Get the language code from a file name. Inverse of getFileName()
4454 * @param string $filename $prefix . $languageCode . $suffix
4455 * @param string $prefix Prefix before the language code
4456 * @param string $suffix Suffix after the language code
4457 * @return string Language code, or false if $prefix or $suffix isn't found
4459 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4461 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4462 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4463 if ( !count( $m ) ) {
4466 return str_replace( '_', '-', strtolower( $m[1] ) );
4470 * @param string $code
4471 * @param bool $fallback Whether we're going through language fallback chain
4472 * @return string Name of the language class
4474 public static function classFromCode( $code, $fallback = true ) {
4475 if ( $fallback && $code == 'en' ) {
4478 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4483 * Get the name of a file for a certain language code
4484 * @param string $prefix Prepend this to the filename
4485 * @param string $code Language code
4486 * @param string $suffix Append this to the filename
4487 * @throws MWException
4488 * @return string $prefix . $mangledCode . $suffix
4490 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4491 if ( !self
::isValidBuiltInCode( $code ) ) {
4492 throw new MWException( "Invalid language code \"$code\"" );
4495 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4499 * @param string $code
4502 public static function getMessagesFileName( $code ) {
4504 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4505 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4510 * @param string $code
4512 * @throws MWException
4515 public static function getJsonMessagesFileName( $code ) {
4518 if ( !self
::isValidBuiltInCode( $code ) ) {
4519 throw new MWException( "Invalid language code \"$code\"" );
4522 return "$IP/languages/i18n/$code.json";
4526 * Get the first fallback for a given language.
4528 * @param string $code
4530 * @return bool|string
4532 public static function getFallbackFor( $code ) {
4533 $fallbacks = self
::getFallbacksFor( $code );
4535 return $fallbacks[0];
4541 * Get the ordered list of fallback languages.
4544 * @param string $code Language code
4545 * @return array Non-empty array, ending in "en"
4547 public static function getFallbacksFor( $code ) {
4548 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4551 // For unknown languages, fallbackSequence returns an empty array,
4552 // hardcode fallback to 'en' in that case.
4553 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4557 * Get the ordered list of fallback languages, ending with the fallback
4558 * language chain for the site language.
4561 * @param string $code Language code
4562 * @return array Array( fallbacks, site fallbacks )
4564 public static function getFallbacksIncludingSiteLanguage( $code ) {
4565 global $wgLanguageCode;
4567 // Usually, we will only store a tiny number of fallback chains, so we
4568 // keep them in static memory.
4569 $cacheKey = "{$code}-{$wgLanguageCode}";
4571 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4572 $fallbacks = self
::getFallbacksFor( $code );
4574 // Append the site's fallback chain, including the site language itself
4575 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4576 array_unshift( $siteFallbacks, $wgLanguageCode );
4578 // Eliminate any languages already included in the chain
4579 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4581 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4583 return self
::$fallbackLanguageCache[$cacheKey];
4587 * Get all messages for a given language
4588 * WARNING: this may take a long time. If you just need all message *keys*
4589 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4591 * @param string $code
4595 public static function getMessagesFor( $code ) {
4596 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4600 * Get a message for a given language
4602 * @param string $key
4603 * @param string $code
4607 public static function getMessageFor( $key, $code ) {
4608 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4612 * Get all message keys for a given language. This is a faster alternative to
4613 * array_keys( Language::getMessagesFor( $code ) )
4616 * @param string $code Language code
4617 * @return array Array of message keys (strings)
4619 public static function getMessageKeysFor( $code ) {
4620 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4624 * @param string $talk
4627 function fixVariableInNamespace( $talk ) {
4628 if ( strpos( $talk, '$1' ) === false ) {
4632 global $wgMetaNamespace;
4633 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4635 # Allow grammar transformations
4636 # Allowing full message-style parsing would make simple requests
4637 # such as action=raw much more expensive than they need to be.
4638 # This will hopefully cover most cases.
4639 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4640 [ $this, 'replaceGrammarInNamespace' ], $talk );
4641 return str_replace( ' ', '_', $talk );
4648 function replaceGrammarInNamespace( $m ) {
4649 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4653 * Decode an expiry (block, protection, etc) which has come from the DB
4655 * @param string $expiry Database expiry String
4656 * @param bool|int $format True to process using language functions, or TS_ constant
4657 * to return the expiry in a given timestamp
4658 * @param string $infinity If $format is not true, use this string for infinite expiry
4662 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4664 if ( $dbInfinity === null ) {
4665 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4668 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4669 return $format === true
4670 ?
$this->getMessageFromDB( 'infiniteblock' )
4673 return $format === true
4674 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4675 : wfTimestamp( $format, $expiry );
4680 * Formats a time given in seconds into a string representation of that time.
4682 * @param int|float $seconds
4683 * @param array $format An optional argument that formats the returned string in different ways:
4684 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4685 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4686 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4688 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4689 * or 'avoidminutes'.
4692 function formatTimePeriod( $seconds, $format = [] ) {
4693 if ( !is_array( $format ) ) {
4694 $format = [ 'avoid' => $format ]; // For backwards compatibility
4696 if ( !isset( $format['avoid'] ) ) {
4697 $format['avoid'] = false;
4699 if ( !isset( $format['noabbrevs'] ) ) {
4700 $format['noabbrevs'] = false;
4702 $secondsMsg = wfMessage(
4703 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4704 $minutesMsg = wfMessage(
4705 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4706 $hoursMsg = wfMessage(
4707 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4708 $daysMsg = wfMessage(
4709 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4711 if ( round( $seconds * 10 ) < 100 ) {
4712 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4713 $s = $secondsMsg->params( $s )->text();
4714 } elseif ( round( $seconds ) < 60 ) {
4715 $s = $this->formatNum( round( $seconds ) );
4716 $s = $secondsMsg->params( $s )->text();
4717 } elseif ( round( $seconds ) < 3600 ) {
4718 $minutes = floor( $seconds / 60 );
4719 $secondsPart = round( fmod( $seconds, 60 ) );
4720 if ( $secondsPart == 60 ) {
4724 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4726 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4727 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4728 $hours = floor( $seconds / 3600 );
4729 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4730 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4731 if ( $secondsPart == 60 ) {
4735 if ( $minutes == 60 ) {
4739 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4741 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4742 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4743 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4746 $days = floor( $seconds / 86400 );
4747 if ( $format['avoid'] === 'avoidminutes' ) {
4748 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4749 if ( $hours == 24 ) {
4753 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4755 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4756 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4757 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4758 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4759 if ( $minutes == 60 ) {
4763 if ( $hours == 24 ) {
4767 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4769 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4771 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4773 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4775 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4782 * Format a bitrate for output, using an appropriate
4783 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4784 * the magnitude in question.
4786 * This use base 1000. For base 1024 use formatSize(), for another base
4787 * see formatComputingNumbers().
4792 function formatBitrate( $bps ) {
4793 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4797 * @param int $size Size of the unit
4798 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4799 * @param string $messageKey Message key to be uesd
4802 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4804 return str_replace( '$1', $this->formatNum( $size ),
4805 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4808 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4811 $maxIndex = count( $sizes ) - 1;
4812 while ( $size >= $boundary && $index < $maxIndex ) {
4817 // For small sizes no decimal places necessary
4820 // For MB and bigger two decimal places are smarter
4823 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4825 $size = round( $size, $round );
4826 $text = $this->getMessageFromDB( $msg );
4827 return str_replace( '$1', $this->formatNum( $size ), $text );
4831 * Format a size in bytes for output, using an appropriate
4832 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4834 * This method use base 1024. For base 1000 use formatBitrate(), for
4835 * another base see formatComputingNumbers()
4837 * @param int $size Size to format
4838 * @return string Plain text (not HTML)
4840 function formatSize( $size ) {
4841 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4845 * Make a list item, used by various special pages
4847 * @param string $page Page link
4848 * @param string $details HTML safe text between brackets
4849 * @param bool $oppositedm Add the direction mark opposite to your
4850 * language, to display text properly
4851 * @return string HTML escaped
4853 function specialList( $page, $details, $oppositedm = true ) {
4858 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4861 $this->msg( 'word-separator' )->escaped() .
4862 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4866 * Generate (prev x| next x) (20|50|100...) type links for paging
4868 * @param Title $title Title object to link
4869 * @param int $offset
4871 * @param array $query Optional URL query parameter string
4872 * @param bool $atend Optional param for specified if this is the last page
4875 public function viewPrevNext( Title
$title, $offset, $limit,
4876 array $query = [], $atend = false
4878 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4880 # Make 'previous' link
4881 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4882 if ( $offset > 0 ) {
4883 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4884 $query, $prev, 'prevn-title', 'mw-prevlink' );
4886 $plink = htmlspecialchars( $prev );
4890 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4892 $nlink = htmlspecialchars( $next );
4894 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4895 $query, $next, 'nextn-title', 'mw-nextlink' );
4898 # Make links to set number of items per page
4900 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4901 $numLinks[] = $this->numLink( $title, $offset, $num,
4902 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4905 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4906 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4910 * Helper function for viewPrevNext() that generates links
4912 * @param Title $title Title object to link
4913 * @param int $offset
4915 * @param array $query Extra query parameters
4916 * @param string $link Text to use for the link; will be escaped
4917 * @param string $tooltipMsg Name of the message to use as tooltip
4918 * @param string $class Value of the "class" attribute of the link
4919 * @return string HTML fragment
4921 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4924 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4925 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4926 ->numParams( $limit )->text();
4928 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4929 'title' => $tooltip, 'class' => $class ], $link );
4933 * Get the conversion rule title, if any.
4937 public function getConvRuleTitle() {
4938 return $this->mConverter
->getConvRuleTitle();
4942 * Get the compiled plural rules for the language
4944 * @return array Associative array with plural form, and plural rule as key-value pairs
4946 public function getCompiledPluralRules() {
4947 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4948 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4949 if ( !$pluralRules ) {
4950 foreach ( $fallbacks as $fallbackCode ) {
4951 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4952 if ( $pluralRules ) {
4957 return $pluralRules;
4961 * Get the plural rules for the language
4963 * @return array Associative array with plural form number and plural rule as key-value pairs
4965 public function getPluralRules() {
4966 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4967 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4968 if ( !$pluralRules ) {
4969 foreach ( $fallbacks as $fallbackCode ) {
4970 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4971 if ( $pluralRules ) {
4976 return $pluralRules;
4980 * Get the plural rule types for the language
4982 * @return array Associative array with plural form number and plural rule type as key-value pairs
4984 public function getPluralRuleTypes() {
4985 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4986 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4987 if ( !$pluralRuleTypes ) {
4988 foreach ( $fallbacks as $fallbackCode ) {
4989 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4990 if ( $pluralRuleTypes ) {
4995 return $pluralRuleTypes;
4999 * Find the index number of the plural rule appropriate for the given number
5000 * @param int $number
5001 * @return int The index number of the plural rule
5003 public function getPluralRuleIndexNumber( $number ) {
5004 $pluralRules = $this->getCompiledPluralRules();
5005 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
5010 * Find the plural rule type appropriate for the given number
5011 * For example, if the language is set to Arabic, getPluralType(5) should
5014 * @param int $number
5015 * @return string The name of the plural rule type, e.g. one, two, few, many
5017 public function getPluralRuleType( $number ) {
5018 $index = $this->getPluralRuleIndexNumber( $number );
5019 $pluralRuleTypes = $this->getPluralRuleTypes();
5020 if ( isset( $pluralRuleTypes[$index] ) ) {
5021 return $pluralRuleTypes[$index];