3 * Internationalisation code.
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
25 * @defgroup Language Language
28 use CLDRPluralRuleParser\Evaluator
;
31 * Internationalisation code
36 * @var LanguageConverter
40 public $mVariants, $mCode, $mLoaded = false;
41 public $mMagicExtensions = [], $mMagicHookDone = false;
42 private $mHtmlCode = null, $mParentLanguage = false;
44 public $dateFormatStrings = [];
45 public $mExtendedSpecialPageAliases;
47 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
50 * ReplacementArray object caches
52 public $transformData = [];
55 * @var LocalisationCache
57 static public $dataCache;
59 static public $mLangObjCache = [];
61 static public $mWeekdayMsgs = [
62 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
66 static public $mWeekdayAbbrevMsgs = [
67 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
70 static public $mMonthMsgs = [
71 'january', 'february', 'march', 'april', 'may_long', 'june',
72 'july', 'august', 'september', 'october', 'november',
75 static public $mMonthGenMsgs = [
76 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
77 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
80 static public $mMonthAbbrevMsgs = [
81 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
82 'sep', 'oct', 'nov', 'dec'
85 static public $mIranianCalendarMonthMsgs = [
86 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
87 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
88 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
89 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
92 static public $mHebrewCalendarMonthMsgs = [
93 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
94 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
95 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
96 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
97 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
100 static public $mHebrewCalendarMonthGenMsgs = [
101 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
102 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
103 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
104 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
105 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
108 static public $mHijriCalendarMonthMsgs = [
109 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
110 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
111 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
112 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
119 static public $durationIntervals = [
120 'millennia' => 31556952000,
121 'centuries' => 3155695200,
122 'decades' => 315569520,
123 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
132 * Cache for language fallbacks.
133 * @see Language::getFallbacksIncludingSiteLanguage
137 static private $fallbackLanguageCache = [];
140 * Cache for language names
141 * @var HashBagOStuff|null
143 static private $languageNameCache;
146 * Unicode directional formatting characters, for embedBidi()
148 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
149 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
150 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
153 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
154 * - in group 1 if it is LTR
155 * - in group 2 if it is RTL
156 * Does not match if there is no strong directionality codepoint.
158 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
160 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
161 * https://git.wikimedia.org/summary/unicodejs.git .
163 // @codingStandardsIgnoreStart
164 // @codeCoverageIgnoreStart
165 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';
166 // @codeCoverageIgnoreEnd
167 // @codingStandardsIgnoreEnd
170 * Get a cached or new language object for a given language code
171 * @param string $code
174 static function factory( $code ) {
175 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
177 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
178 $code = $wgDummyLanguageCodes[$code];
181 // get the language object to process
182 $langObj = isset( self
::$mLangObjCache[$code] )
183 ? self
::$mLangObjCache[$code]
184 : self
::newFromCode( $code );
186 // merge the language object in to get it up front in the cache
187 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
188 // get rid of the oldest ones in case we have an overflow
189 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
195 * Create a language object for a given language code
196 * @param string $code
197 * @throws MWException
200 protected static function newFromCode( $code ) {
201 if ( !Language
::isValidCode( $code ) ) {
202 throw new MWException( "Invalid language code \"$code\"" );
205 if ( !Language
::isValidBuiltInCode( $code ) ) {
206 // It's not possible to customise this code with class files, so
207 // just return a Language object. This is to support uselang= hacks.
208 $lang = new Language
;
209 $lang->setCode( $code );
213 // Check if there is a language class for the code
214 $class = self
::classFromCode( $code );
215 if ( class_exists( $class ) ) {
220 // Keep trying the fallback list until we find an existing class
221 $fallbacks = Language
::getFallbacksFor( $code );
222 foreach ( $fallbacks as $fallbackCode ) {
223 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
224 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
227 $class = self
::classFromCode( $fallbackCode );
228 if ( class_exists( $class ) ) {
230 $lang->setCode( $code );
235 throw new MWException( "Invalid fallback sequence for language '$code'" );
239 * Checks whether any localisation is available for that language tag
240 * in MediaWiki (MessagesXx.php exists).
242 * @param string $code Language tag (in lower case)
243 * @return bool Whether language is supported
246 public static function isSupportedLanguage( $code ) {
247 if ( !self
::isValidBuiltInCode( $code ) ) {
251 if ( $code === 'qqq' ) {
255 return is_readable( self
::getMessagesFileName( $code ) ) ||
256 is_readable( self
::getJsonMessagesFileName( $code ) );
260 * Returns true if a language code string is a well-formed language tag
261 * according to RFC 5646.
262 * This function only checks well-formedness; it doesn't check that
263 * language, script or variant codes actually exist in the repositories.
265 * Based on regexes by Mark Davis of the Unicode Consortium:
266 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
268 * @param string $code
269 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
274 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
277 $alphanum = '[a-z0-9]';
278 $x = 'x'; # private use singleton
279 $singleton = '[a-wy-z]'; # other singleton
280 $s = $lenient ?
'[-_]' : '-';
282 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
283 $script = "$alpha{4}"; # ISO 15924
284 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
285 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
286 $extension = "$singleton(?:$s$alphanum{2,8})+";
287 $privateUse = "$x(?:$s$alphanum{1,8})+";
289 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
290 # Since these are limited, this is safe even later changes to the registry --
291 # the only oddity is that it might change the type of the tag, and thus
292 # the results from the capturing groups.
293 # http://www.iana.org/assignments/language-subtag-registry
295 $grandfathered = "en{$s}GB{$s}oed"
296 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
297 . "|no{$s}(?:bok|nyn)"
298 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
299 . "|zh{$s}min{$s}nan";
301 $variantList = "$variant(?:$s$variant)*";
302 $extensionList = "$extension(?:$s$extension)*";
304 $langtag = "(?:($language)"
307 . "(?:$s$variantList)?"
308 . "(?:$s$extensionList)?"
309 . "(?:$s$privateUse)?)";
311 # The final breakdown, with capturing groups for each of these components
312 # The variants, extensions, grandfathered, and private-use may have interior '-'
314 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
316 return (bool)preg_match( "/$root/", strtolower( $code ) );
320 * Returns true if a language code string is of a valid form, whether or
321 * not it exists. This includes codes which are used solely for
322 * customisation via the MediaWiki namespace.
324 * @param string $code
328 public static function isValidCode( $code ) {
330 if ( !isset( $cache[$code] ) ) {
331 // People think language codes are html safe, so enforce it.
332 // Ideally we should only allow a-zA-Z0-9-
333 // but, .+ and other chars are often used for {{int:}} hacks
334 // see bugs 37564, 37587, 36938
336 // Protect against path traversal
337 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
338 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
340 return $cache[$code];
344 * Returns true if a language code is of a valid form for the purposes of
345 * internal customisation of MediaWiki, via Messages*.php or *.json.
347 * @param string $code
349 * @throws MWException
353 public static function isValidBuiltInCode( $code ) {
355 if ( !is_string( $code ) ) {
356 if ( is_object( $code ) ) {
357 $addmsg = " of class " . get_class( $code );
361 $type = gettype( $code );
362 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
365 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
369 * Returns true if a language code is an IETF tag known to MediaWiki.
376 public static function isKnownLanguageTag( $tag ) {
377 // Quick escape for invalid input to avoid exceptions down the line
378 // when code tries to process tags which are not valid at all.
379 if ( !self
::isValidBuiltInCode( $tag ) ) {
383 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
384 || self
::fetchLanguageName( $tag, $tag ) !== ''
393 * Get the LocalisationCache instance
395 * @return LocalisationCache
397 public static function getLocalisationCache() {
398 if ( is_null( self
::$dataCache ) ) {
399 global $wgLocalisationCacheConf;
400 $class = $wgLocalisationCacheConf['class'];
401 self
::$dataCache = new $class( $wgLocalisationCacheConf );
403 return self
::$dataCache;
406 function __construct() {
407 $this->mConverter
= new FakeConverter( $this );
408 // Set the code to the name of the descendant
409 if ( get_class( $this ) == 'Language' ) {
412 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
414 self
::getLocalisationCache();
418 * Reduce memory usage
420 function __destruct() {
421 foreach ( $this as $name => $value ) {
422 unset( $this->$name );
427 * Hook which will be called if this is the content language.
428 * Descendants can use this to register hook functions or modify globals
430 function initContLang() {
437 public function getFallbackLanguages() {
438 return self
::getFallbacksFor( $this->mCode
);
442 * Exports $wgBookstoreListEn
445 public function getBookstoreList() {
446 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
450 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
451 * available in localised form, it will be included in English.
455 public function getNamespaces() {
456 if ( is_null( $this->namespaceNames
) ) {
457 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
459 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
460 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
462 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
464 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
465 if ( $wgMetaNamespaceTalk ) {
466 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
468 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
469 $this->namespaceNames
[NS_PROJECT_TALK
] =
470 $this->fixVariableInNamespace( $talk );
473 # Sometimes a language will be localised but not actually exist on this wiki.
474 foreach ( $this->namespaceNames
as $key => $text ) {
475 if ( !isset( $validNamespaces[$key] ) ) {
476 unset( $this->namespaceNames
[$key] );
480 # The above mixing may leave namespaces out of canonical order.
481 # Re-order by namespace ID number...
482 ksort( $this->namespaceNames
);
484 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
487 return $this->namespaceNames
;
491 * Arbitrarily set all of the namespace names at once. Mainly used for testing
492 * @param array $namespaces Array of namespaces (id => name)
494 public function setNamespaces( array $namespaces ) {
495 $this->namespaceNames
= $namespaces;
496 $this->mNamespaceIds
= null;
500 * Resets all of the namespace caches. Mainly used for testing
502 public function resetNamespaces() {
503 $this->namespaceNames
= null;
504 $this->mNamespaceIds
= null;
505 $this->namespaceAliases
= null;
509 * A convenience function that returns getNamespaces() with spaces instead of underscores
510 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
514 public function getFormattedNamespaces() {
515 $ns = $this->getNamespaces();
516 foreach ( $ns as $k => $v ) {
517 $ns[$k] = strtr( $v, '_', ' ' );
523 * Get a namespace value by key
526 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
527 * echo $mw_ns; // prints 'MediaWiki'
530 * @param int $index The array key of the namespace to return
531 * @return string|bool String if the namespace value exists, otherwise false
533 public function getNsText( $index ) {
534 $ns = $this->getNamespaces();
535 return isset( $ns[$index] ) ?
$ns[$index] : false;
539 * A convenience function that returns the same thing as
540 * getNsText() except with '_' changed to ' ', useful for
544 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
545 * echo $mw_ns; // prints 'MediaWiki talk'
548 * @param int $index The array key of the namespace to return
549 * @return string Namespace name without underscores (empty string if namespace does not exist)
551 public function getFormattedNsText( $index ) {
552 $ns = $this->getNsText( $index );
553 return strtr( $ns, '_', ' ' );
557 * Returns gender-dependent namespace alias if available.
558 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
559 * @param int $index Namespace index
560 * @param string $gender Gender key (male, female... )
564 public function getGenderNsText( $index, $gender ) {
565 global $wgExtraGenderNamespaces;
567 $ns = $wgExtraGenderNamespaces +
568 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
570 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
574 * Whether this language uses gender-dependent namespace aliases.
575 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
579 public function needsGenderDistinction() {
580 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
581 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
582 // $wgExtraGenderNamespaces overrides everything
584 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
585 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
586 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
589 // Check what is in i18n files
590 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
591 return count( $aliases ) > 0;
596 * Get a namespace key by value, case insensitive.
597 * Only matches namespace names for the current language, not the
598 * canonical ones defined in Namespace.php.
600 * @param string $text
601 * @return int|bool An integer if $text is a valid value otherwise false
603 function getLocalNsIndex( $text ) {
604 $lctext = $this->lc( $text );
605 $ids = $this->getNamespaceIds();
606 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
612 public function getNamespaceAliases() {
613 if ( is_null( $this->namespaceAliases
) ) {
614 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
618 foreach ( $aliases as $name => $index ) {
619 if ( $index === NS_PROJECT_TALK
) {
620 unset( $aliases[$name] );
621 $name = $this->fixVariableInNamespace( $name );
622 $aliases[$name] = $index;
627 global $wgExtraGenderNamespaces;
628 $genders = $wgExtraGenderNamespaces +
629 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
630 foreach ( $genders as $index => $forms ) {
631 foreach ( $forms as $alias ) {
632 $aliases[$alias] = $index;
636 # Also add converted namespace names as aliases, to avoid confusion.
637 $convertedNames = [];
638 foreach ( $this->getVariants() as $variant ) {
639 if ( $variant === $this->mCode
) {
642 foreach ( $this->getNamespaces() as $ns => $_ ) {
643 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
647 $this->namespaceAliases
= $aliases +
$convertedNames;
650 return $this->namespaceAliases
;
656 public function getNamespaceIds() {
657 if ( is_null( $this->mNamespaceIds
) ) {
658 global $wgNamespaceAliases;
659 # Put namespace names and aliases into a hashtable.
660 # If this is too slow, then we should arrange it so that it is done
661 # before caching. The catch is that at pre-cache time, the above
662 # class-specific fixup hasn't been done.
663 $this->mNamespaceIds
= [];
664 foreach ( $this->getNamespaces() as $index => $name ) {
665 $this->mNamespaceIds
[$this->lc( $name )] = $index;
667 foreach ( $this->getNamespaceAliases() as $name => $index ) {
668 $this->mNamespaceIds
[$this->lc( $name )] = $index;
670 if ( $wgNamespaceAliases ) {
671 foreach ( $wgNamespaceAliases as $name => $index ) {
672 $this->mNamespaceIds
[$this->lc( $name )] = $index;
676 return $this->mNamespaceIds
;
680 * Get a namespace key by value, case insensitive. Canonical namespace
681 * names override custom ones defined for the current language.
683 * @param string $text
684 * @return int|bool An integer if $text is a valid value otherwise false
686 public function getNsIndex( $text ) {
687 $lctext = $this->lc( $text );
688 $ns = MWNamespace
::getCanonicalIndex( $lctext );
689 if ( $ns !== null ) {
692 $ids = $this->getNamespaceIds();
693 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
697 * short names for language variants used for language conversion links.
699 * @param string $code
700 * @param bool $usemsg Use the "variantname-xyz" message if it exists
703 public function getVariantname( $code, $usemsg = true ) {
704 $msg = "variantname-$code";
705 if ( $usemsg && wfMessage( $msg )->exists() ) {
706 return $this->getMessageFromDB( $msg );
708 $name = self
::fetchLanguageName( $code );
710 return $name; # if it's defined as a language name, show that
712 # otherwise, output the language code
720 public function getDatePreferences() {
721 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
727 function getDateFormats() {
728 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
732 * @return array|string
734 public function getDefaultDateFormat() {
735 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
736 if ( $df === 'dmy or mdy' ) {
737 global $wgAmericanDates;
738 return $wgAmericanDates ?
'mdy' : 'dmy';
747 public function getDatePreferenceMigrationMap() {
748 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
752 * @param string $image
755 function getImageFile( $image ) {
756 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
763 public function getImageFiles() {
764 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
770 public function getExtraUserToggles() {
771 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
778 function getUserToggle( $tog ) {
779 return $this->getMessageFromDB( "tog-$tog" );
783 * Get an array of language names, indexed by code.
784 * @param null|string $inLanguage Code of language in which to return the names
785 * Use null for autonyms (native names)
786 * @param string $include One of:
787 * 'all' all available languages
788 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
789 * 'mwfile' only if the language is in 'mw' *and* has a message file
790 * @return array Language code => language name
793 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
794 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
795 $cacheKey .= ":$include";
796 if ( self
::$languageNameCache === null ) {
797 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
800 $ret = self
::$languageNameCache->get( $cacheKey );
802 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
803 self
::$languageNameCache->set( $cacheKey, $ret );
809 * Uncached helper for fetchLanguageNames
810 * @param null|string $inLanguage Code of language in which to return the names
811 * Use null for autonyms (native names)
812 * @param string $include One of:
813 * 'all' all available languages
814 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
815 * 'mwfile' only if the language is in 'mw' *and* has a message file
816 * @return array Language code => language name
818 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
819 global $wgExtraLanguageNames;
821 // If passed an invalid language code to use, fallback to en
822 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
829 # TODO: also include when $inLanguage is null, when this code is more efficient
830 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
833 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
834 foreach ( $mwNames as $mwCode => $mwName ) {
835 # - Prefer own MediaWiki native name when not using the hook
836 # - For other names just add if not added through the hook
837 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
838 $names[$mwCode] = $mwName;
842 if ( $include === 'all' ) {
848 $coreCodes = array_keys( $mwNames );
849 foreach ( $coreCodes as $coreCode ) {
850 $returnMw[$coreCode] = $names[$coreCode];
853 if ( $include === 'mwfile' ) {
855 # We do this using a foreach over the codes instead of a directory
856 # loop so that messages files in extensions will work correctly.
857 foreach ( $returnMw as $code => $value ) {
858 if ( is_readable( self
::getMessagesFileName( $code ) )
859 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
861 $namesMwFile[$code] = $names[$code];
865 ksort( $namesMwFile );
870 # 'mw' option; default if it's not one of the other two options (all/mwfile)
875 * @param string $code The code of the language for which to get the name
876 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
877 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
878 * @return string Language name or empty
881 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
882 $code = strtolower( $code );
883 $array = self
::fetchLanguageNames( $inLanguage, $include );
884 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
888 * Get a message from the MediaWiki namespace.
890 * @param string $msg Message name
893 public function getMessageFromDB( $msg ) {
894 return $this->msg( $msg )->text();
898 * Get message object in this language. Only for use inside this class.
900 * @param string $msg Message name
903 protected function msg( $msg ) {
904 return wfMessage( $msg )->inLanguage( $this );
911 public function getMonthName( $key ) {
912 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
918 public function getMonthNamesArray() {
919 $monthNames = [ '' ];
920 for ( $i = 1; $i < 13; $i++
) {
921 $monthNames[] = $this->getMonthName( $i );
930 public function getMonthNameGen( $key ) {
931 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
938 public function getMonthAbbreviation( $key ) {
939 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
945 public function getMonthAbbreviationsArray() {
946 $monthNames = [ '' ];
947 for ( $i = 1; $i < 13; $i++
) {
948 $monthNames[] = $this->getMonthAbbreviation( $i );
957 public function getWeekdayName( $key ) {
958 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
965 function getWeekdayAbbreviation( $key ) {
966 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
973 function getIranianCalendarMonthName( $key ) {
974 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
981 function getHebrewCalendarMonthName( $key ) {
982 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
989 function getHebrewCalendarMonthNameGen( $key ) {
990 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
997 function getHijriCalendarMonthName( $key ) {
998 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1002 * Pass through result from $dateTimeObj->format()
1003 * @param DateTime|bool|null &$dateTimeObj
1005 * @param DateTimeZone|bool|null $zone
1006 * @param string $code
1009 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1010 if ( !$dateTimeObj ) {
1011 $dateTimeObj = DateTime
::createFromFormat(
1012 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1015 return $dateTimeObj->format( $code );
1019 * This is a workalike of PHP's date() function, but with better
1020 * internationalisation, a reduced set of format characters, and a better
1023 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1024 * the PHP manual for definitions. There are a number of extensions, which
1027 * xn Do not translate digits of the next numeric format character
1028 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1029 * xr Use roman numerals for the next numeric format character
1030 * xh Use hebrew numerals for the next numeric format character
1032 * xg Genitive month name
1034 * xij j (day number) in Iranian calendar
1035 * xiF F (month name) in Iranian calendar
1036 * xin n (month number) in Iranian calendar
1037 * xiy y (two digit year) in Iranian calendar
1038 * xiY Y (full year) in Iranian calendar
1040 * xjj j (day number) in Hebrew calendar
1041 * xjF F (month name) in Hebrew calendar
1042 * xjt t (days in month) in Hebrew calendar
1043 * xjx xg (genitive month name) in Hebrew calendar
1044 * xjn n (month number) in Hebrew calendar
1045 * xjY Y (full year) in Hebrew calendar
1047 * xmj j (day number) in Hijri calendar
1048 * xmF F (month name) in Hijri calendar
1049 * xmn n (month number) in Hijri calendar
1050 * xmY Y (full year) in Hijri calendar
1052 * xkY Y (full year) in Thai solar calendar. Months and days are
1053 * identical to the Gregorian calendar
1054 * xoY Y (full year) in Minguo calendar or Juche year.
1055 * Months and days are identical to the
1056 * Gregorian calendar
1057 * xtY Y (full year) in Japanese nengo. Months and days are
1058 * identical to the Gregorian calendar
1060 * Characters enclosed in double quotes will be considered literal (with
1061 * the quotes themselves removed). Unmatched quotes will be considered
1062 * literal quotes. Example:
1064 * "The month is" F => The month is January
1067 * Backslash escaping is also supported.
1069 * Input timestamp is assumed to be pre-normalized to the desired local
1070 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1071 * $ts is UTC if $zone is not given.
1073 * @param string $format
1074 * @param string $ts 14-character timestamp
1077 * @param DateTimeZone $zone Timezone of $ts
1078 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1079 * Only makes sense if $ts is the current time.
1080 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1082 * @throws MWException
1085 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1090 $dateTimeObj = false;
1099 $usedSecond = false;
1100 $usedMinute = false;
1107 $usedISOYear = false;
1108 $usedIsLeapYear = false;
1110 $usedHebrewMonth = false;
1111 $usedIranianMonth = false;
1112 $usedHijriMonth = false;
1113 $usedHebrewYear = false;
1114 $usedIranianYear = false;
1115 $usedHijriYear = false;
1116 $usedTennoYear = false;
1118 if ( strlen( $ts ) !== 14 ) {
1119 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1122 if ( !ctype_digit( $ts ) ) {
1123 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1126 $formatLength = strlen( $format );
1127 for ( $p = 0; $p < $formatLength; $p++
) {
1129 $code = $format[$p];
1130 if ( $code == 'x' && $p < $formatLength - 1 ) {
1131 $code .= $format[++
$p];
1134 if ( ( $code === 'xi'
1140 && $p < $formatLength - 1 ) {
1141 $code .= $format[++
$p];
1152 $rawToggle = !$rawToggle;
1162 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1165 $usedHebrewMonth = true;
1167 $hebrew = self
::tsToHebrew( $ts );
1169 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1173 $num = substr( $ts, 6, 2 );
1177 $s .= $this->getWeekdayAbbreviation(
1178 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1183 $num = intval( substr( $ts, 6, 2 ) );
1188 $iranian = self
::tsToIranian( $ts );
1195 $hijri = self
::tsToHijri( $ts );
1202 $hebrew = self
::tsToHebrew( $ts );
1208 $s .= $this->getWeekdayName(
1209 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1214 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1217 $usedIranianMonth = true;
1219 $iranian = self
::tsToIranian( $ts );
1221 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1224 $usedHijriMonth = true;
1226 $hijri = self
::tsToHijri( $ts );
1228 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1231 $usedHebrewMonth = true;
1233 $hebrew = self
::tsToHebrew( $ts );
1235 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1239 $num = substr( $ts, 4, 2 );
1243 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1247 $num = intval( substr( $ts, 4, 2 ) );
1250 $usedIranianMonth = true;
1252 $iranian = self
::tsToIranian( $ts );
1257 $usedHijriMonth = true;
1259 $hijri = self
::tsToHijri( $ts );
1264 $usedHebrewMonth = true;
1266 $hebrew = self
::tsToHebrew( $ts );
1271 $usedHebrewMonth = true;
1273 $hebrew = self
::tsToHebrew( $ts );
1279 $num = substr( $ts, 0, 4 );
1282 $usedIranianYear = true;
1284 $iranian = self
::tsToIranian( $ts );
1289 $usedHijriYear = true;
1291 $hijri = self
::tsToHijri( $ts );
1296 $usedHebrewYear = true;
1298 $hebrew = self
::tsToHebrew( $ts );
1305 $thai = self
::tsToYear( $ts, 'thai' );
1312 $minguo = self
::tsToYear( $ts, 'minguo' );
1317 $usedTennoYear = true;
1319 $tenno = self
::tsToYear( $ts, 'tenno' );
1325 $num = substr( $ts, 2, 2 );
1328 $usedIranianYear = true;
1330 $iranian = self
::tsToIranian( $ts );
1332 $num = substr( $iranian[0], -2 );
1336 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1340 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1344 $h = substr( $ts, 8, 2 );
1345 $num = $h %
12 ?
$h %
12 : 12;
1349 $num = intval( substr( $ts, 8, 2 ) );
1353 $h = substr( $ts, 8, 2 );
1354 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1358 $num = substr( $ts, 8, 2 );
1362 $num = substr( $ts, 10, 2 );
1366 $num = substr( $ts, 12, 2 );
1376 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1382 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1386 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1390 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1393 $usedIsLeapYear = true;
1394 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1397 $usedISOYear = true;
1398 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1405 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1408 # Backslash escaping
1409 if ( $p < $formatLength - 1 ) {
1410 $s .= $format[++
$p];
1417 if ( $p < $formatLength - 1 ) {
1418 $endQuote = strpos( $format, '"', $p +
1 );
1419 if ( $endQuote === false ) {
1420 # No terminating quote, assume literal "
1423 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1427 # Quote at end of string, assume literal "
1434 if ( $num !== false ) {
1435 if ( $rawToggle ||
$raw ) {
1438 } elseif ( $roman ) {
1439 $s .= Language
::romanNumeral( $num );
1441 } elseif ( $hebrewNum ) {
1442 $s .= self
::hebrewNumeral( $num );
1445 $s .= $this->formatNum( $num, true );
1450 if ( $ttl === 'unused' ) {
1451 // No need to calculate the TTL, the caller wont use it anyway.
1452 } elseif ( $usedSecond ) {
1454 } elseif ( $usedMinute ) {
1455 $ttl = 60 - substr( $ts, 12, 2 );
1456 } elseif ( $usedHour ) {
1457 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1458 } elseif ( $usedAMPM ) {
1459 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1460 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1464 $usedIranianMonth ||
1471 // @todo Someone who understands the non-Gregorian calendars
1472 // should write proper logic for them so that they don't need purged every day.
1473 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1474 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1477 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1478 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1481 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1482 $timeRemainingInDay;
1483 } elseif ( $usedISOYear ) {
1484 // December 28th falls on the last ISO week of the year, every year.
1485 // The last ISO week of a year can be 52 or 53.
1486 $lastWeekOfISOYear = DateTime
::createFromFormat(
1488 substr( $ts, 0, 4 ) . '1228',
1489 $zone ?
: new DateTimeZone( 'UTC' )
1491 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1492 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1493 $timeRemainingInWeek =
1494 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1495 +
$timeRemainingInDay;
1496 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1501 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1502 substr( $ts, 6, 2 ) ) * 86400
1503 +
$timeRemainingInDay;
1504 } elseif ( $usedYear ) {
1506 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1507 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1508 +
$timeRemainingInDay;
1509 } elseif ( $usedIsLeapYear ) {
1510 $year = substr( $ts, 0, 4 );
1511 $timeRemainingInYear =
1512 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1513 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1514 +
$timeRemainingInDay;
1516 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1517 // this isn't a leap year. see when the next one starts
1518 $nextCandidate = $year - $mod +
4;
1519 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1520 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1521 $timeRemainingInYear;
1523 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1524 $timeRemainingInYear;
1527 // this is a leap year, so the next year isn't
1528 $possibleTtls[] = $timeRemainingInYear;
1532 if ( $possibleTtls ) {
1533 $ttl = min( $possibleTtls );
1540 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1541 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1544 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1545 * Gregorian dates to Iranian dates. Originally written in C, it
1546 * is released under the terms of GNU Lesser General Public
1547 * License. Conversion to PHP was performed by Niklas Laxström.
1549 * Link: http://www.farsiweb.info/jalali/jalali.c
1555 private static function tsToIranian( $ts ) {
1556 $gy = substr( $ts, 0, 4 ) -1600;
1557 $gm = substr( $ts, 4, 2 ) -1;
1558 $gd = substr( $ts, 6, 2 ) -1;
1560 # Days passed from the beginning (including leap years)
1562 +
floor( ( $gy +
3 ) / 4 )
1563 - floor( ( $gy +
99 ) / 100 )
1564 +
floor( ( $gy +
399 ) / 400 );
1566 // Add days of the past months of this year
1567 for ( $i = 0; $i < $gm; $i++
) {
1568 $gDayNo +
= self
::$GREG_DAYS[$i];
1572 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1576 // Days passed in current month
1577 $gDayNo +
= (int)$gd;
1579 $jDayNo = $gDayNo - 79;
1581 $jNp = floor( $jDayNo / 12053 );
1584 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1587 if ( $jDayNo >= 366 ) {
1588 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1589 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1592 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1593 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1599 return [ $jy, $jm, $jd ];
1603 * Converting Gregorian dates to Hijri dates.
1605 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1607 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1613 private static function tsToHijri( $ts ) {
1614 $year = substr( $ts, 0, 4 );
1615 $month = substr( $ts, 4, 2 );
1616 $day = substr( $ts, 6, 2 );
1624 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1625 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1627 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1628 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1629 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1632 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1633 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1636 $zl = $zjd -1948440 +
10632;
1637 $zn = (int)( ( $zl - 1 ) / 10631 );
1638 $zl = $zl - 10631 * $zn +
354;
1639 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1640 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1641 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1642 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1643 $zm = (int)( ( 24 * $zl ) / 709 );
1644 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1645 $zy = 30 * $zn +
$zj - 30;
1647 return [ $zy, $zm, $zd ];
1651 * Converting Gregorian dates to Hebrew dates.
1653 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1654 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1655 * to translate the relevant functions into PHP and release them under
1658 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1659 * and Adar II is 14. In a non-leap year, Adar is 6.
1665 private static function tsToHebrew( $ts ) {
1667 $year = substr( $ts, 0, 4 );
1668 $month = substr( $ts, 4, 2 );
1669 $day = substr( $ts, 6, 2 );
1671 # Calculate Hebrew year
1672 $hebrewYear = $year +
3760;
1674 # Month number when September = 1, August = 12
1676 if ( $month > 12 ) {
1683 # Calculate day of year from 1 September
1685 for ( $i = 1; $i < $month; $i++
) {
1689 # Check if the year is leap
1690 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1693 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1700 # Calculate the start of the Hebrew year
1701 $start = self
::hebrewYearStart( $hebrewYear );
1703 # Calculate next year's start
1704 if ( $dayOfYear <= $start ) {
1705 # Day is before the start of the year - it is the previous year
1707 $nextStart = $start;
1711 # Add days since previous year's 1 September
1713 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1717 # Start of the new (previous) year
1718 $start = self
::hebrewYearStart( $hebrewYear );
1721 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1724 # Calculate Hebrew day of year
1725 $hebrewDayOfYear = $dayOfYear - $start;
1727 # Difference between year's days
1728 $diff = $nextStart - $start;
1729 # Add 12 (or 13 for leap years) days to ignore the difference between
1730 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1731 # difference is only about the year type
1732 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1738 # Check the year pattern, and is leap year
1739 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1740 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1741 # and non-leap years
1742 $yearPattern = $diff %
30;
1743 # Check if leap year
1744 $isLeap = $diff >= 30;
1746 # Calculate day in the month from number of day in the Hebrew year
1747 # Don't check Adar - if the day is not in Adar, we will stop before;
1748 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1749 $hebrewDay = $hebrewDayOfYear;
1752 while ( $hebrewMonth <= 12 ) {
1753 # Calculate days in this month
1754 if ( $isLeap && $hebrewMonth == 6 ) {
1755 # Adar in a leap year
1757 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1759 if ( $hebrewDay <= $days ) {
1763 # Subtract the days of Adar I
1764 $hebrewDay -= $days;
1767 if ( $hebrewDay <= $days ) {
1773 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1774 # Cheshvan in a complete year (otherwise as the rule below)
1776 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1777 # Kislev in an incomplete year (otherwise as the rule below)
1780 # Odd months have 30 days, even have 29
1781 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1783 if ( $hebrewDay <= $days ) {
1784 # In the current month
1787 # Subtract the days of the current month
1788 $hebrewDay -= $days;
1789 # Try in the next month
1794 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1798 * This calculates the Hebrew year start, as days since 1 September.
1799 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1800 * Used for Hebrew date.
1806 private static function hebrewYearStart( $year ) {
1807 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1808 $b = intval( ( $year - 1 ) %
4 );
1809 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1813 $Mar = intval( $m );
1819 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1820 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1822 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1824 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1828 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1833 * Algorithm to convert Gregorian dates to Thai solar dates,
1834 * Minguo dates or Minguo dates.
1836 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1837 * http://en.wikipedia.org/wiki/Minguo_calendar
1838 * http://en.wikipedia.org/wiki/Japanese_era_name
1840 * @param string $ts 14-character timestamp
1841 * @param string $cName Calender name
1842 * @return array Converted year, month, day
1844 private static function tsToYear( $ts, $cName ) {
1845 $gy = substr( $ts, 0, 4 );
1846 $gm = substr( $ts, 4, 2 );
1847 $gd = substr( $ts, 6, 2 );
1849 if ( !strcmp( $cName, 'thai' ) ) {
1851 # Add 543 years to the Gregorian calendar
1852 # Months and days are identical
1853 $gy_offset = $gy +
543;
1854 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1856 # Deduct 1911 years from the Gregorian calendar
1857 # Months and days are identical
1858 $gy_offset = $gy - 1911;
1859 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1860 # Nengō dates up to Meiji period
1861 # Deduct years from the Gregorian calendar
1862 # depending on the nengo periods
1863 # Months and days are identical
1865 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1866 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1869 $gy_gannen = $gy - 1868 +
1;
1870 $gy_offset = $gy_gannen;
1871 if ( $gy_gannen == 1 ) {
1874 $gy_offset = '明治' . $gy_offset;
1876 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1877 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1878 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1879 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1880 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1883 $gy_gannen = $gy - 1912 +
1;
1884 $gy_offset = $gy_gannen;
1885 if ( $gy_gannen == 1 ) {
1888 $gy_offset = '大正' . $gy_offset;
1890 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1891 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1892 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1895 $gy_gannen = $gy - 1926 +
1;
1896 $gy_offset = $gy_gannen;
1897 if ( $gy_gannen == 1 ) {
1900 $gy_offset = '昭和' . $gy_offset;
1903 $gy_gannen = $gy - 1989 +
1;
1904 $gy_offset = $gy_gannen;
1905 if ( $gy_gannen == 1 ) {
1908 $gy_offset = '平成' . $gy_offset;
1914 return [ $gy_offset, $gm, $gd ];
1918 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1920 * This is the rule the BIDI algorithm uses to determine the directionality of
1921 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1922 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1924 * TODO: Does not handle BIDI control characters inside the text.
1925 * TODO: Does not handle unallocated characters.
1927 * @param string $text Text to test
1928 * @return null|string Directionality ('ltr' or 'rtl') or null
1930 private static function strongDirFromContent( $text = '' ) {
1931 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
1934 if ( $matches[1] === '' ) {
1941 * Roman number formatting up to 10000
1947 static function romanNumeral( $num ) {
1949 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1950 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1951 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1952 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1953 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1956 $num = intval( $num );
1957 if ( $num > 10000 ||
$num <= 0 ) {
1962 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1963 if ( $num >= $pow10 ) {
1964 $s .= $table[$i][(int)floor( $num / $pow10 )];
1966 $num = $num %
$pow10;
1972 * Hebrew Gematria number formatting up to 9999
1978 static function hebrewNumeral( $num ) {
1980 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
1981 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
1994 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
1997 $num = intval( $num );
1998 if ( $num > 9999 ||
$num <= 0 ) {
2002 // Round thousands have special notations
2003 if ( $num === 1000 ) {
2005 } elseif ( $num %
1000 === 0 ) {
2006 return $table[0][$num / 1000] . "' אלפים";
2011 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2012 if ( $num >= $pow10 ) {
2013 if ( $num === 15 ||
$num === 16 ) {
2014 $letters[] = $table[0][9];
2015 $letters[] = $table[0][$num - 9];
2018 $letters = array_merge(
2020 (array)$table[$i][intval( $num / $pow10 )]
2023 if ( $pow10 === 1000 ) {
2029 $num = $num %
$pow10;
2032 $preTransformLength = count( $letters );
2033 if ( $preTransformLength === 1 ) {
2034 // Add geresh (single quote) to one-letter numbers
2037 $lastIndex = $preTransformLength - 1;
2038 $letters[$lastIndex] = str_replace(
2039 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2040 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2041 $letters[$lastIndex]
2044 // Add gershayim (double quote) to multiple-letter numbers,
2045 // but exclude numbers with only one letter after the thousands
2046 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2047 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2050 array_splice( $letters, -1, 0, '"' );
2054 return implode( $letters );
2058 * Used by date() and time() to adjust the time output.
2060 * @param string $ts The time in date('YmdHis') format
2061 * @param mixed $tz Adjust the time by this amount (default false, mean we
2062 * get user timecorrection setting)
2065 public function userAdjust( $ts, $tz = false ) {
2066 global $wgUser, $wgLocalTZoffset;
2068 if ( $tz === false ) {
2069 $tz = $wgUser->getOption( 'timecorrection' );
2072 $data = explode( '|', $tz, 3 );
2074 if ( $data[0] == 'ZoneInfo' ) {
2075 MediaWiki\
suppressWarnings();
2076 $userTZ = timezone_open( $data[2] );
2077 MediaWiki\restoreWarnings
();
2078 if ( $userTZ !== false ) {
2079 $date = date_create( $ts, timezone_open( 'UTC' ) );
2080 date_timezone_set( $date, $userTZ );
2081 $date = date_format( $date, 'YmdHis' );
2084 # Unrecognized timezone, default to 'Offset' with the stored offset.
2085 $data[0] = 'Offset';
2088 if ( $data[0] == 'System' ||
$tz == '' ) {
2089 # Global offset in minutes.
2090 $minDiff = $wgLocalTZoffset;
2091 } elseif ( $data[0] == 'Offset' ) {
2092 $minDiff = intval( $data[1] );
2094 $data = explode( ':', $tz );
2095 if ( count( $data ) == 2 ) {
2096 $data[0] = intval( $data[0] );
2097 $data[1] = intval( $data[1] );
2098 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2099 if ( $data[0] < 0 ) {
2100 $minDiff = -$minDiff;
2103 $minDiff = intval( $data[0] ) * 60;
2107 # No difference ? Return time unchanged
2108 if ( 0 == $minDiff ) {
2112 MediaWiki\
suppressWarnings(); // E_STRICT system time bitching
2113 # Generate an adjusted date; take advantage of the fact that mktime
2114 # will normalize out-of-range values so we don't have to split $minDiff
2115 # into hours and minutes.
2117 (int)substr( $ts, 8, 2 ) ), # Hours
2118 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2119 (int)substr( $ts, 12, 2 ), # Seconds
2120 (int)substr( $ts, 4, 2 ), # Month
2121 (int)substr( $ts, 6, 2 ), # Day
2122 (int)substr( $ts, 0, 4 ) ); # Year
2124 $date = date( 'YmdHis', $t );
2125 MediaWiki\restoreWarnings
();
2131 * This is meant to be used by time(), date(), and timeanddate() to get
2132 * the date preference they're supposed to use, it should be used in
2136 * function timeanddate([...], $format = true) {
2137 * $datePreference = $this->dateFormat($format);
2142 * @param int|string|bool $usePrefs If true, the user's preference is used
2143 * if false, the site/language default is used
2144 * if int/string, assumed to be a format.
2147 function dateFormat( $usePrefs = true ) {
2150 if ( is_bool( $usePrefs ) ) {
2152 $datePreference = $wgUser->getDatePreference();
2154 $datePreference = (string)User
::getDefaultOption( 'date' );
2157 $datePreference = (string)$usePrefs;
2161 if ( $datePreference == '' ) {
2165 return $datePreference;
2169 * Get a format string for a given type and preference
2170 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2171 * @param string $pref The format name as it appears in Messages*.php under
2174 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2178 function getDateFormatString( $type, $pref ) {
2179 $wasDefault = false;
2180 if ( $pref == 'default' ) {
2182 $pref = $this->getDefaultDateFormat();
2185 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2186 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2188 if ( $type === 'pretty' && $df === null ) {
2189 $df = $this->getDateFormatString( 'date', $pref );
2192 if ( !$wasDefault && $df === null ) {
2193 $pref = $this->getDefaultDateFormat();
2194 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2197 $this->dateFormatStrings
[$type][$pref] = $df;
2199 return $this->dateFormatStrings
[$type][$pref];
2203 * @param string $ts The time format which needs to be turned into a
2204 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2205 * @param bool $adj Whether to adjust the time output according to the
2206 * user configured offset ($timecorrection)
2207 * @param mixed $format True to use user's date format preference
2208 * @param string|bool $timecorrection The time offset as returned by
2209 * validateTimeZone() in Special:Preferences
2212 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2213 $ts = wfTimestamp( TS_MW
, $ts );
2215 $ts = $this->userAdjust( $ts, $timecorrection );
2217 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2218 return $this->sprintfDate( $df, $ts );
2222 * @param string $ts The time format which needs to be turned into a
2223 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2224 * @param bool $adj Whether to adjust the time output according to the
2225 * user configured offset ($timecorrection)
2226 * @param mixed $format True to use user's date format preference
2227 * @param string|bool $timecorrection The time offset as returned by
2228 * validateTimeZone() in Special:Preferences
2231 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2232 $ts = wfTimestamp( TS_MW
, $ts );
2234 $ts = $this->userAdjust( $ts, $timecorrection );
2236 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2237 return $this->sprintfDate( $df, $ts );
2241 * @param string $ts The time format which needs to be turned into a
2242 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2243 * @param bool $adj Whether to adjust the time output according to the
2244 * user configured offset ($timecorrection)
2245 * @param mixed $format What format to return, if it's false output the
2246 * default one (default true)
2247 * @param string|bool $timecorrection The time offset as returned by
2248 * validateTimeZone() in Special:Preferences
2251 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2252 $ts = wfTimestamp( TS_MW
, $ts );
2254 $ts = $this->userAdjust( $ts, $timecorrection );
2256 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2257 return $this->sprintfDate( $df, $ts );
2261 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2265 * @param int $seconds The amount of seconds.
2266 * @param array $chosenIntervals The intervals to enable.
2270 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2271 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2275 foreach ( $intervals as $intervalName => $intervalValue ) {
2276 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2277 // duration-years, duration-decades, duration-centuries, duration-millennia
2278 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2279 $segments[] = $message->inLanguage( $this )->escaped();
2282 return $this->listToText( $segments );
2286 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2287 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2291 * @param int $seconds The amount of seconds.
2292 * @param array $chosenIntervals The intervals to enable.
2296 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2297 if ( empty( $chosenIntervals ) ) {
2298 $chosenIntervals = [
2310 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2311 $sortedNames = array_keys( $intervals );
2312 $smallestInterval = array_pop( $sortedNames );
2316 foreach ( $intervals as $name => $length ) {
2317 $value = floor( $seconds / $length );
2319 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2320 $seconds -= $value * $length;
2321 $segments[$name] = $value;
2329 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2331 * @param string $type Can be 'date', 'time' or 'both'
2332 * @param string $ts The time format which needs to be turned into a
2333 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2334 * @param User $user User object used to get preferences for timezone and format
2335 * @param array $options Array, can contain the following keys:
2336 * - 'timecorrection': time correction, can have the following values:
2337 * - true: use user's preference
2338 * - false: don't use time correction
2339 * - int: value of time correction in minutes
2340 * - 'format': format to use, can have the following values:
2341 * - true: use user's preference
2342 * - false: use default preference
2343 * - string: format to use
2347 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2348 $ts = wfTimestamp( TS_MW
, $ts );
2349 $options +
= [ 'timecorrection' => true, 'format' => true ];
2350 if ( $options['timecorrection'] !== false ) {
2351 if ( $options['timecorrection'] === true ) {
2352 $offset = $user->getOption( 'timecorrection' );
2354 $offset = $options['timecorrection'];
2356 $ts = $this->userAdjust( $ts, $offset );
2358 if ( $options['format'] === true ) {
2359 $format = $user->getDatePreference();
2361 $format = $options['format'];
2363 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2364 return $this->sprintfDate( $df, $ts );
2368 * Get the formatted date for the given timestamp and formatted for
2371 * @param mixed $ts Mixed: the time format which needs to be turned into a
2372 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2373 * @param User $user User object used to get preferences for timezone and format
2374 * @param array $options Array, can contain the following keys:
2375 * - 'timecorrection': time correction, can have the following values:
2376 * - true: use user's preference
2377 * - false: don't use time correction
2378 * - int: value of time correction in minutes
2379 * - 'format': format to use, can have the following values:
2380 * - true: use user's preference
2381 * - false: use default preference
2382 * - string: format to use
2386 public function userDate( $ts, User
$user, array $options = [] ) {
2387 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2391 * Get the formatted time for the given timestamp and formatted for
2394 * @param mixed $ts The time format which needs to be turned into a
2395 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2396 * @param User $user User object used to get preferences for timezone and format
2397 * @param array $options Array, can contain the following keys:
2398 * - 'timecorrection': time correction, can have the following values:
2399 * - true: use user's preference
2400 * - false: don't use time correction
2401 * - int: value of time correction in minutes
2402 * - 'format': format to use, can have the following values:
2403 * - true: use user's preference
2404 * - false: use default preference
2405 * - string: format to use
2409 public function userTime( $ts, User
$user, array $options = [] ) {
2410 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2414 * Get the formatted date and time for the given timestamp and formatted for
2417 * @param mixed $ts The time format which needs to be turned into a
2418 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2419 * @param User $user User object used to get preferences for timezone and format
2420 * @param array $options Array, can contain the following keys:
2421 * - 'timecorrection': time correction, can have the following values:
2422 * - true: use user's preference
2423 * - false: don't use time correction
2424 * - int: value of time correction in minutes
2425 * - 'format': format to use, can have the following values:
2426 * - true: use user's preference
2427 * - false: use default preference
2428 * - string: format to use
2432 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2433 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2437 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2439 * Determine the difference between the timestamp and the current time, and
2440 * generate a readable timestamp by returning "<N> <units> ago", where the
2441 * largest possible unit is used.
2443 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2445 * @param MWTimestamp $time
2446 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2447 * @param User|null $user User the timestamp is being generated for
2448 * (or null to use main context's user)
2449 * @return string Formatted timestamp
2451 public function getHumanTimestamp(
2452 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2454 if ( $relativeTo === null ) {
2455 $relativeTo = new MWTimestamp();
2457 if ( $user === null ) {
2458 $user = RequestContext
::getMain()->getUser();
2461 // Adjust for the user's timezone.
2462 $offsetThis = $time->offsetForUser( $user );
2463 $offsetRel = $relativeTo->offsetForUser( $user );
2466 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2467 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2470 // Reset the timezone on the objects.
2471 $time->timestamp
->sub( $offsetThis );
2472 $relativeTo->timestamp
->sub( $offsetRel );
2478 * Convert an MWTimestamp into a pretty human-readable timestamp using
2479 * the given user preferences and relative base time.
2481 * @see Language::getHumanTimestamp
2482 * @param MWTimestamp $ts Timestamp to prettify
2483 * @param MWTimestamp $relativeTo Base timestamp
2484 * @param User $user User preferences to use
2485 * @return string Human timestamp
2488 private function getHumanTimestampInternal(
2489 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2491 $diff = $ts->diff( $relativeTo );
2492 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2493 (int)$relativeTo->timestamp
->format( 'w' ) );
2494 $days = $diff->days ?
: (int)$diffDay;
2495 if ( $diff->invert ||
$days > 5
2496 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2498 // Timestamps are in different years: use full timestamp
2499 // Also do full timestamp for future dates
2501 * @todo FIXME: Add better handling of future timestamps.
2503 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2504 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2505 } elseif ( $days > 5 ) {
2506 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2507 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2508 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2509 } elseif ( $days > 1 ) {
2510 // Timestamp within the past week: show the day of the week and time
2511 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2512 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2514 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2515 $ts = wfMessage( "$weekday-at" )
2516 ->inLanguage( $this )
2517 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2519 } elseif ( $days == 1 ) {
2520 // Timestamp was yesterday: say 'yesterday' and the time.
2521 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2522 $ts = wfMessage( 'yesterday-at' )
2523 ->inLanguage( $this )
2524 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2526 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2527 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2528 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2529 $ts = wfMessage( 'today-at' )
2530 ->inLanguage( $this )
2531 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2534 // From here on in, the timestamp was soon enough ago so that we can simply say
2535 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2536 } elseif ( $diff->h
== 1 ) {
2537 // Less than 90 minutes, but more than an hour ago.
2538 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2539 } elseif ( $diff->i
>= 1 ) {
2540 // A few minutes ago.
2541 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2542 } elseif ( $diff->s
>= 30 ) {
2543 // Less than a minute, but more than 30 sec ago.
2544 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2546 // Less than 30 seconds ago.
2547 $ts = wfMessage( 'just-now' )->text();
2554 * @param string $key
2555 * @return array|null
2557 public function getMessage( $key ) {
2558 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2564 function getAllMessages() {
2565 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2570 * @param string $out
2571 * @param string $string
2574 public function iconv( $in, $out, $string ) {
2575 # Even with //IGNORE iconv can whine about illegal characters in
2576 # *input* string. We just ignore those too.
2577 # REF: http://bugs.php.net/bug.php?id=37166
2578 # REF: https://phabricator.wikimedia.org/T18885
2579 MediaWiki\
suppressWarnings();
2580 $text = iconv( $in, $out . '//IGNORE', $string );
2581 MediaWiki\restoreWarnings
();
2585 // callback functions for ucwords(), ucwordbreaks()
2588 * @param array $matches
2589 * @return mixed|string
2591 function ucwordbreaksCallbackAscii( $matches ) {
2592 return $this->ucfirst( $matches[1] );
2596 * @param array $matches
2599 function ucwordbreaksCallbackMB( $matches ) {
2600 return mb_strtoupper( $matches[0] );
2604 * @param array $matches
2607 function ucwordsCallbackMB( $matches ) {
2608 return mb_strtoupper( $matches[0] );
2612 * Make a string's first character uppercase
2614 * @param string $str
2618 public function ucfirst( $str ) {
2620 if ( $o < 96 ) { // if already uppercase...
2622 } elseif ( $o < 128 ) {
2623 return ucfirst( $str ); // use PHP's ucfirst()
2625 // fall back to more complex logic in case of multibyte strings
2626 return $this->uc( $str, true );
2631 * Convert a string to uppercase
2633 * @param string $str
2634 * @param bool $first
2638 public function uc( $str, $first = false ) {
2640 if ( $this->isMultibyte( $str ) ) {
2641 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2643 return ucfirst( $str );
2646 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2651 * @param string $str
2652 * @return mixed|string
2654 function lcfirst( $str ) {
2657 return strval( $str );
2658 } elseif ( $o >= 128 ) {
2659 return $this->lc( $str, true );
2660 } elseif ( $o > 96 ) {
2663 $str[0] = strtolower( $str[0] );
2669 * @param string $str
2670 * @param bool $first
2671 * @return mixed|string
2673 function lc( $str, $first = false ) {
2675 if ( $this->isMultibyte( $str ) ) {
2676 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2678 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2681 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2686 * @param string $str
2689 function isMultibyte( $str ) {
2690 return strlen( $str ) !== mb_strlen( $str );
2694 * @param string $str
2695 * @return mixed|string
2697 function ucwords( $str ) {
2698 if ( $this->isMultibyte( $str ) ) {
2699 $str = $this->lc( $str );
2701 // regexp to find first letter in each word (i.e. after each space)
2702 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2704 // function to use to capitalize a single char
2705 return preg_replace_callback(
2707 [ $this, 'ucwordsCallbackMB' ],
2711 return ucwords( strtolower( $str ) );
2716 * capitalize words at word breaks
2718 * @param string $str
2721 function ucwordbreaks( $str ) {
2722 if ( $this->isMultibyte( $str ) ) {
2723 $str = $this->lc( $str );
2725 // since \b doesn't work for UTF-8, we explicitely define word break chars
2726 $breaks = "[ \-\(\)\}\{\.,\?!]";
2728 // find first letter after word break
2729 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2730 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2732 return preg_replace_callback(
2734 [ $this, 'ucwordbreaksCallbackMB' ],
2738 return preg_replace_callback(
2739 '/\b([\w\x80-\xff]+)\b/',
2740 [ $this, 'ucwordbreaksCallbackAscii' ],
2747 * Return a case-folded representation of $s
2749 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2750 * and $s2 are the same except for the case of their characters. It is not
2751 * necessary for the value returned to make sense when displayed.
2753 * Do *not* perform any other normalisation in this function. If a caller
2754 * uses this function when it should be using a more general normalisation
2755 * function, then fix the caller.
2761 function caseFold( $s ) {
2762 return $this->uc( $s );
2768 * @throws MWException
2770 function checkTitleEncoding( $s ) {
2771 if ( is_array( $s ) ) {
2772 throw new MWException( 'Given array to checkTitleEncoding.' );
2774 if ( StringUtils
::isUtf8( $s ) ) {
2778 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2784 function fallback8bitEncoding() {
2785 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2789 * Most writing systems use whitespace to break up words.
2790 * Some languages such as Chinese don't conventionally do this,
2791 * which requires special handling when breaking up words for
2796 function hasWordBreaks() {
2801 * Some languages such as Chinese require word segmentation,
2802 * Specify such segmentation when overridden in derived class.
2804 * @param string $string
2807 function segmentByWord( $string ) {
2812 * Some languages have special punctuation need to be normalized.
2813 * Make such changes here.
2815 * @param string $string
2818 function normalizeForSearch( $string ) {
2819 return self
::convertDoubleWidth( $string );
2823 * convert double-width roman characters to single-width.
2824 * range: ff00-ff5f ~= 0020-007f
2826 * @param string $string
2830 protected static function convertDoubleWidth( $string ) {
2831 static $full = null;
2832 static $half = null;
2834 if ( $full === null ) {
2835 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2836 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2837 $full = str_split( $fullWidth, 3 );
2838 $half = str_split( $halfWidth );
2841 $string = str_replace( $full, $half, $string );
2846 * @param string $string
2847 * @param string $pattern
2850 protected static function insertSpace( $string, $pattern ) {
2851 $string = preg_replace( $pattern, " $1 ", $string );
2852 $string = preg_replace( '/ +/', ' ', $string );
2857 * @param array $termsArray
2860 function convertForSearchResult( $termsArray ) {
2861 # some languages, e.g. Chinese, need to do a conversion
2862 # in order for search results to be displayed correctly
2867 * Get the first character of a string.
2872 function firstChar( $s ) {
2875 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2876 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2881 if ( isset( $matches[1] ) ) {
2882 if ( strlen( $matches[1] ) != 3 ) {
2886 // Break down Hangul syllables to grab the first jamo
2887 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
2888 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2890 } elseif ( $code < 0xb098 ) {
2891 return "\xe3\x84\xb1";
2892 } elseif ( $code < 0xb2e4 ) {
2893 return "\xe3\x84\xb4";
2894 } elseif ( $code < 0xb77c ) {
2895 return "\xe3\x84\xb7";
2896 } elseif ( $code < 0xb9c8 ) {
2897 return "\xe3\x84\xb9";
2898 } elseif ( $code < 0xbc14 ) {
2899 return "\xe3\x85\x81";
2900 } elseif ( $code < 0xc0ac ) {
2901 return "\xe3\x85\x82";
2902 } elseif ( $code < 0xc544 ) {
2903 return "\xe3\x85\x85";
2904 } elseif ( $code < 0xc790 ) {
2905 return "\xe3\x85\x87";
2906 } elseif ( $code < 0xcc28 ) {
2907 return "\xe3\x85\x88";
2908 } elseif ( $code < 0xce74 ) {
2909 return "\xe3\x85\x8a";
2910 } elseif ( $code < 0xd0c0 ) {
2911 return "\xe3\x85\x8b";
2912 } elseif ( $code < 0xd30c ) {
2913 return "\xe3\x85\x8c";
2914 } elseif ( $code < 0xd558 ) {
2915 return "\xe3\x85\x8d";
2917 return "\xe3\x85\x8e";
2925 * @deprecated No-op since 1.28
2927 function initEncoding() {
2934 * @deprecated No-op since 1.28
2936 function recodeForEdit( $s ) {
2943 * @deprecated No-op since 1.28
2945 function recodeInput( $s ) {
2950 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2951 * also cleans up certain backwards-compatible sequences, converting them
2952 * to the modern Unicode equivalent.
2954 * This is language-specific for performance reasons only.
2960 function normalize( $s ) {
2961 global $wgAllUnicodeFixes;
2962 $s = UtfNormal\Validator
::cleanUp( $s );
2963 if ( $wgAllUnicodeFixes ) {
2964 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2965 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2972 * Transform a string using serialized data stored in the given file (which
2973 * must be in the serialized subdirectory of $IP). The file contains pairs
2974 * mapping source characters to destination characters.
2976 * The data is cached in process memory. This will go faster if you have the
2977 * FastStringSearch extension.
2979 * @param string $file
2980 * @param string $string
2982 * @throws MWException
2985 function transformUsingPairFile( $file, $string ) {
2986 if ( !isset( $this->transformData
[$file] ) ) {
2987 $data = wfGetPrecompiledData( $file );
2988 if ( $data === false ) {
2989 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2991 $this->transformData
[$file] = new ReplacementArray( $data );
2993 return $this->transformData
[$file]->replace( $string );
2997 * For right-to-left language support
3002 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3006 * Return the correct HTML 'dir' attribute value for this language.
3010 return $this->isRTL() ?
'rtl' : 'ltr';
3014 * Return 'left' or 'right' as appropriate alignment for line-start
3015 * for this language's text direction.
3017 * Should be equivalent to CSS3 'start' text-align value....
3021 function alignStart() {
3022 return $this->isRTL() ?
'right' : 'left';
3026 * Return 'right' or 'left' as appropriate alignment for line-end
3027 * for this language's text direction.
3029 * Should be equivalent to CSS3 'end' text-align value....
3033 function alignEnd() {
3034 return $this->isRTL() ?
'left' : 'right';
3038 * A hidden direction mark (LRM or RLM), depending on the language direction.
3039 * Unlike getDirMark(), this function returns the character as an HTML entity.
3040 * This function should be used when the output is guaranteed to be HTML,
3041 * because it makes the output HTML source code more readable. When
3042 * the output is plain text or can be escaped, getDirMark() should be used.
3044 * @param bool $opposite Get the direction mark opposite to your language
3048 function getDirMarkEntity( $opposite = false ) {
3050 return $this->isRTL() ?
'‎' : '‏';
3052 return $this->isRTL() ?
'‏' : '‎';
3056 * A hidden direction mark (LRM or RLM), depending on the language direction.
3057 * This function produces them as invisible Unicode characters and
3058 * the output may be hard to read and debug, so it should only be used
3059 * when the output is plain text or can be escaped. When the output is
3060 * HTML, use getDirMarkEntity() instead.
3062 * @param bool $opposite Get the direction mark opposite to your language
3065 function getDirMark( $opposite = false ) {
3066 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3067 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3069 return $this->isRTL() ?
$lrm : $rlm;
3071 return $this->isRTL() ?
$rlm : $lrm;
3077 function capitalizeAllNouns() {
3078 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3082 * An arrow, depending on the language direction.
3084 * @param string $direction The direction of the arrow: forwards (default),
3085 * backwards, left, right, up, down.
3088 function getArrow( $direction = 'forwards' ) {
3089 switch ( $direction ) {
3091 return $this->isRTL() ?
'←' : '→';
3093 return $this->isRTL() ?
'→' : '←';
3106 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3110 function linkPrefixExtension() {
3111 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3115 * Get all magic words from cache.
3118 function getMagicWords() {
3119 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3123 * Run the LanguageGetMagic hook once.
3125 protected function doMagicHook() {
3126 if ( $this->mMagicHookDone
) {
3129 $this->mMagicHookDone
= true;
3130 Hooks
::run( 'LanguageGetMagic', [ &$this->mMagicExtensions
, $this->getCode() ] );
3134 * Fill a MagicWord object with data from here
3136 * @param MagicWord $mw
3138 function getMagic( $mw ) {
3139 // Saves a function call
3140 if ( !$this->mMagicHookDone
) {
3141 $this->doMagicHook();
3144 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3145 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3147 $rawEntry = self
::$dataCache->getSubitem(
3148 $this->mCode
, 'magicWords', $mw->mId
);
3151 if ( !is_array( $rawEntry ) ) {
3152 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3154 $mw->mCaseSensitive
= $rawEntry[0];
3155 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3160 * Add magic words to the extension array
3162 * @param array $newWords
3164 function addMagicWordsByLang( $newWords ) {
3165 $fallbackChain = $this->getFallbackLanguages();
3166 $fallbackChain = array_reverse( $fallbackChain );
3167 foreach ( $fallbackChain as $code ) {
3168 if ( isset( $newWords[$code] ) ) {
3169 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3175 * Get special page names, as an associative array
3176 * canonical name => array of valid names, including aliases
3179 function getSpecialPageAliases() {
3180 // Cache aliases because it may be slow to load them
3181 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3183 $this->mExtendedSpecialPageAliases
=
3184 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3185 Hooks
::run( 'LanguageGetSpecialPageAliases',
3186 [ &$this->mExtendedSpecialPageAliases
, $this->getCode() ] );
3189 return $this->mExtendedSpecialPageAliases
;
3193 * Italic is unsuitable for some languages
3195 * @param string $text The text to be emphasized.
3198 function emphasize( $text ) {
3199 return "<em>$text</em>";
3203 * Normally we output all numbers in plain en_US style, that is
3204 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3205 * point twohundredthirtyfive. However this is not suitable for all
3206 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3207 * Icelandic just want to use commas instead of dots, and dots instead
3208 * of commas like "293.291,235".
3210 * An example of this function being called:
3212 * wfMessage( 'message' )->numParams( $num )->text()
3215 * See $separatorTransformTable on MessageIs.php for
3216 * the , => . and . => , implementation.
3218 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3219 * @param int|float $number The string to be formatted, should be an integer
3220 * or a floating point number.
3221 * @param bool $nocommafy Set to true for special numbers like dates
3224 public function formatNum( $number, $nocommafy = false ) {
3225 global $wgTranslateNumerals;
3226 if ( !$nocommafy ) {
3227 $number = $this->commafy( $number );
3228 $s = $this->separatorTransformTable();
3230 $number = strtr( $number, $s );
3234 if ( $wgTranslateNumerals ) {
3235 $s = $this->digitTransformTable();
3237 $number = strtr( $number, $s );
3245 * Front-end for non-commafied formatNum
3247 * @param int|float $number The string to be formatted, should be an integer
3248 * or a floating point number.
3252 public function formatNumNoSeparators( $number ) {
3253 return $this->formatNum( $number, true );
3257 * @param string $number
3260 public function parseFormattedNumber( $number ) {
3261 $s = $this->digitTransformTable();
3263 // eliminate empty array values such as ''. (bug 64347)
3264 $s = array_filter( $s );
3265 $number = strtr( $number, array_flip( $s ) );
3268 $s = $this->separatorTransformTable();
3270 // eliminate empty array values such as ''. (bug 64347)
3271 $s = array_filter( $s );
3272 $number = strtr( $number, array_flip( $s ) );
3275 $number = strtr( $number, [ ',' => '' ] );
3280 * Adds commas to a given number
3282 * @param mixed $number
3285 function commafy( $number ) {
3286 $digitGroupingPattern = $this->digitGroupingPattern();
3287 if ( $number === null ) {
3291 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3292 // default grouping is at thousands, use the same for ###,###,### pattern too.
3293 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3295 // Ref: http://cldr.unicode.org/translation/number-patterns
3297 if ( intval( $number ) < 0 ) {
3298 // For negative numbers apply the algorithm like positive number and add sign.
3300 $number = substr( $number, 1 );
3304 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3305 preg_match( "/\d+/", $number, $integerPart );
3306 preg_match( "/\.\d*/", $number, $decimalPart );
3307 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3308 if ( $groupedNumber === $number ) {
3309 // the string does not have any number part. Eg: .12345
3310 return $sign . $groupedNumber;
3312 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3313 while ( $start > 0 ) {
3314 $match = $matches[0][$numMatches - 1];
3315 $matchLen = strlen( $match );
3316 $start = $end - $matchLen;
3320 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3322 if ( $numMatches > 1 ) {
3323 // use the last pattern for the rest of the number
3327 $groupedNumber = "," . $groupedNumber;
3330 return $sign . $groupedNumber;
3337 function digitGroupingPattern() {
3338 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3344 function digitTransformTable() {
3345 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3351 function separatorTransformTable() {
3352 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3356 * Take a list of strings and build a locale-friendly comma-separated
3357 * list, using the local comma-separator message.
3358 * The last two strings are chained with an "and".
3359 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3361 * @param string[] $l
3364 function listToText( array $l ) {
3365 $m = count( $l ) - 1;
3370 $and = $this->msg( 'and' )->escaped();
3371 $space = $this->msg( 'word-separator' )->escaped();
3373 $comma = $this->msg( 'comma-separator' )->escaped();
3377 for ( $i = $m - 1; $i >= 0; $i-- ) {
3378 if ( $i == $m - 1 ) {
3379 $s = $l[$i] . $and . $space . $s;
3381 $s = $l[$i] . $comma . $s;
3388 * Take a list of strings and build a locale-friendly comma-separated
3389 * list, using the local comma-separator message.
3390 * @param string[] $list Array of strings to put in a comma list
3393 function commaList( array $list ) {
3395 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3401 * Take a list of strings and build a locale-friendly semicolon-separated
3402 * list, using the local semicolon-separator message.
3403 * @param string[] $list Array of strings to put in a semicolon list
3406 function semicolonList( array $list ) {
3408 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3414 * Same as commaList, but separate it with the pipe instead.
3415 * @param string[] $list Array of strings to put in a pipe list
3418 function pipeList( array $list ) {
3420 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3426 * Truncate a string to a specified length in bytes, appending an optional
3427 * string (e.g. for ellipses)
3429 * The database offers limited byte lengths for some columns in the database;
3430 * multi-byte character sets mean we need to ensure that only whole characters
3431 * are included, otherwise broken characters can be passed to the user
3433 * If $length is negative, the string will be truncated from the beginning
3435 * @param string $string String to truncate
3436 * @param int $length Maximum length (including ellipses)
3437 * @param string $ellipsis String to append to the truncated text
3438 * @param bool $adjustLength Subtract length of ellipsis from $length.
3439 * $adjustLength was introduced in 1.18, before that behaved as if false.
3442 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3443 # Use the localized ellipsis character
3444 if ( $ellipsis == '...' ) {
3445 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3447 # Check if there is no need to truncate
3448 if ( $length == 0 ) {
3449 return $ellipsis; // convention
3450 } elseif ( strlen( $string ) <= abs( $length ) ) {
3451 return $string; // no need to truncate
3453 $stringOriginal = $string;
3454 # If ellipsis length is >= $length then we can't apply $adjustLength
3455 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3456 $string = $ellipsis; // this can be slightly unexpected
3457 # Otherwise, truncate and add ellipsis...
3459 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3460 if ( $length > 0 ) {
3461 $length -= $eLength;
3462 $string = substr( $string, 0, $length ); // xyz...
3463 $string = $this->removeBadCharLast( $string );
3464 $string = rtrim( $string );
3465 $string = $string . $ellipsis;
3467 $length +
= $eLength;
3468 $string = substr( $string, $length ); // ...xyz
3469 $string = $this->removeBadCharFirst( $string );
3470 $string = ltrim( $string );
3471 $string = $ellipsis . $string;
3474 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3475 # This check is *not* redundant if $adjustLength, due to the single case where
3476 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3477 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3480 return $stringOriginal;
3485 * Remove bytes that represent an incomplete Unicode character
3486 * at the end of string (e.g. bytes of the char are missing)
3488 * @param string $string
3491 protected function removeBadCharLast( $string ) {
3492 if ( $string != '' ) {
3493 $char = ord( $string[strlen( $string ) - 1] );
3495 if ( $char >= 0xc0 ) {
3496 # We got the first byte only of a multibyte char; remove it.
3497 $string = substr( $string, 0, -1 );
3498 } elseif ( $char >= 0x80 &&
3499 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3500 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3501 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3503 # We chopped in the middle of a character; remove it
3511 * Remove bytes that represent an incomplete Unicode character
3512 * at the start of string (e.g. bytes of the char are missing)
3514 * @param string $string
3517 protected function removeBadCharFirst( $string ) {
3518 if ( $string != '' ) {
3519 $char = ord( $string[0] );
3520 if ( $char >= 0x80 && $char < 0xc0 ) {
3521 # We chopped in the middle of a character; remove the whole thing
3522 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3529 * Truncate a string of valid HTML to a specified length in bytes,
3530 * appending an optional string (e.g. for ellipses), and return valid HTML
3532 * This is only intended for styled/linked text, such as HTML with
3533 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3534 * Also, this will not detect things like "display:none" CSS.
3536 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3538 * @param string $text HTML string to truncate
3539 * @param int $length (zero/positive) Maximum length (including ellipses)
3540 * @param string $ellipsis String to append to the truncated text
3543 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3544 # Use the localized ellipsis character
3545 if ( $ellipsis == '...' ) {
3546 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3548 # Check if there is clearly no need to truncate
3549 if ( $length <= 0 ) {
3550 return $ellipsis; // no text shown, nothing to format (convention)
3551 } elseif ( strlen( $text ) <= $length ) {
3552 return $text; // string short enough even *with* HTML (short-circuit)
3555 $dispLen = 0; // innerHTML legth so far
3556 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3557 $tagType = 0; // 0-open, 1-close
3558 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3559 $entityState = 0; // 0-not entity, 1-entity
3560 $tag = $ret = ''; // accumulated tag name, accumulated result string
3561 $openTags = []; // open tag stack
3562 $maybeState = null; // possible truncation state
3564 $textLen = strlen( $text );
3565 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3566 for ( $pos = 0; true; ++
$pos ) {
3567 # Consider truncation once the display length has reached the maximim.
3568 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3569 # Check that we're not in the middle of a bracket/entity...
3570 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3571 if ( !$testingEllipsis ) {
3572 $testingEllipsis = true;
3573 # Save where we are; we will truncate here unless there turn out to
3574 # be so few remaining characters that truncation is not necessary.
3575 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3576 $maybeState = [ $ret, $openTags ]; // save state
3578 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3579 # String in fact does need truncation, the truncation point was OK.
3580 list( $ret, $openTags ) = $maybeState; // reload state
3581 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3582 $ret .= $ellipsis; // add ellipsis
3586 if ( $pos >= $textLen ) {
3587 break; // extra iteration just for above checks
3590 # Read the next char...
3592 $lastCh = $pos ?
$text[$pos - 1] : '';
3593 $ret .= $ch; // add to result string
3595 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3596 $entityState = 0; // for bad HTML
3597 $bracketState = 1; // tag started (checking for backslash)
3598 } elseif ( $ch == '>' ) {
3599 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3600 $entityState = 0; // for bad HTML
3601 $bracketState = 0; // out of brackets
3602 } elseif ( $bracketState == 1 ) {
3604 $tagType = 1; // close tag (e.g. "</span>")
3606 $tagType = 0; // open tag (e.g. "<span>")
3609 $bracketState = 2; // building tag name
3610 } elseif ( $bracketState == 2 ) {
3614 // Name found (e.g. "<a href=..."), add on tag attributes...
3615 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3617 } elseif ( $bracketState == 0 ) {
3618 if ( $entityState ) {
3621 $dispLen++
; // entity is one displayed char
3624 if ( $neLength == 0 && !$maybeState ) {
3625 // Save state without $ch. We want to *hit* the first
3626 // display char (to get tags) but not *use* it if truncating.
3627 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3630 $entityState = 1; // entity found, (e.g. " ")
3632 $dispLen++
; // this char is displayed
3633 // Add the next $max display text chars after this in one swoop...
3634 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3635 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3636 $dispLen +
= $skipped;
3642 // Close the last tag if left unclosed by bad HTML
3643 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3644 while ( count( $openTags ) > 0 ) {
3645 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3651 * truncateHtml() helper function
3652 * like strcspn() but adds the skipped chars to $ret
3654 * @param string $ret
3655 * @param string $text
3656 * @param string $search
3658 * @param null|int $len
3661 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3662 if ( $len === null ) {
3663 $len = -1; // -1 means "no limit" for strcspn
3664 } elseif ( $len < 0 ) {
3668 if ( $start < strlen( $text ) ) {
3669 $skipCount = strcspn( $text, $search, $start, $len );
3670 $ret .= substr( $text, $start, $skipCount );
3676 * truncateHtml() helper function
3677 * (a) push or pop $tag from $openTags as needed
3678 * (b) clear $tag value
3679 * @param string &$tag Current HTML tag name we are looking at
3680 * @param int $tagType (0-open tag, 1-close tag)
3681 * @param string $lastCh Character before the '>' that ended this tag
3682 * @param array &$openTags Open tag stack (not accounting for $tag)
3684 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3685 $tag = ltrim( $tag );
3687 if ( $tagType == 0 && $lastCh != '/' ) {
3688 $openTags[] = $tag; // tag opened (didn't close itself)
3689 } elseif ( $tagType == 1 ) {
3690 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3691 array_pop( $openTags ); // tag closed
3699 * Grammatical transformations, needed for inflected languages
3700 * Invoked by putting {{grammar:case|word}} in a message
3702 * @param string $word
3703 * @param string $case
3706 function convertGrammar( $word, $case ) {
3707 global $wgGrammarForms;
3708 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3709 return $wgGrammarForms[$this->getCode()][$case][$word];
3715 * Get the grammar forms for the content language
3716 * @return array Array of grammar forms
3719 function getGrammarForms() {
3720 global $wgGrammarForms;
3721 if ( isset( $wgGrammarForms[$this->getCode()] )
3722 && is_array( $wgGrammarForms[$this->getCode()] )
3724 return $wgGrammarForms[$this->getCode()];
3730 * Provides an alternative text depending on specified gender.
3731 * Usage {{gender:username|masculine|feminine|unknown}}.
3732 * username is optional, in which case the gender of current user is used,
3733 * but only in (some) interface messages; otherwise default gender is used.
3735 * If no forms are given, an empty string is returned. If only one form is
3736 * given, it will be returned unconditionally. These details are implied by
3737 * the caller and cannot be overridden in subclasses.
3739 * If three forms are given, the default is to use the third (unknown) form.
3740 * If fewer than three forms are given, the default is to use the first (masculine) form.
3741 * These details can be overridden in subclasses.
3743 * @param string $gender
3744 * @param array $forms
3748 function gender( $gender, $forms ) {
3749 if ( !count( $forms ) ) {
3752 $forms = $this->preConvertPlural( $forms, 2 );
3753 if ( $gender === 'male' ) {
3756 if ( $gender === 'female' ) {
3759 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3763 * Plural form transformations, needed for some languages.
3764 * For example, there are 3 form of plural in Russian and Polish,
3765 * depending on "count mod 10". See [[w:Plural]]
3766 * For English it is pretty simple.
3768 * Invoked by putting {{plural:count|wordform1|wordform2}}
3769 * or {{plural:count|wordform1|wordform2|wordform3}}
3771 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3773 * @param int $count Non-localized number
3774 * @param array $forms Different plural forms
3775 * @return string Correct form of plural for $count in this language
3777 function convertPlural( $count, $forms ) {
3778 // Handle explicit n=pluralform cases
3779 $forms = $this->handleExplicitPluralForms( $count, $forms );
3780 if ( is_string( $forms ) ) {
3783 if ( !count( $forms ) ) {
3787 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3788 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3789 return $forms[$pluralForm];
3793 * Handles explicit plural forms for Language::convertPlural()
3795 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3796 * If an explicitly defined plural form matches the $count, then
3797 * string value returned, otherwise array returned for further consideration
3798 * by CLDR rules or overridden convertPlural().
3802 * @param int $count Non-localized number
3803 * @param array $forms Different plural forms
3805 * @return array|string
3807 protected function handleExplicitPluralForms( $count, array $forms ) {
3808 foreach ( $forms as $index => $form ) {
3809 if ( preg_match( '/\d+=/i', $form ) ) {
3810 $pos = strpos( $form, '=' );
3811 if ( substr( $form, 0, $pos ) === (string)$count ) {
3812 return substr( $form, $pos +
1 );
3814 unset( $forms[$index] );
3817 return array_values( $forms );
3821 * Checks that convertPlural was given an array and pads it to requested
3822 * amount of forms by copying the last one.
3824 * @param array $forms Array of forms given to convertPlural
3825 * @param int $count How many forms should there be at least
3826 * @return array Padded array of forms or an exception if not an array
3828 protected function preConvertPlural( /* Array */ $forms, $count ) {
3829 while ( count( $forms ) < $count ) {
3830 $forms[] = $forms[count( $forms ) - 1];
3836 * Wraps argument with unicode control characters for directionality safety
3838 * This solves the problem where directionality-neutral characters at the edge of
3839 * the argument string get interpreted with the wrong directionality from the
3840 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3842 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3843 * directionality of the argument string, using the BIDI algorithm's own "First
3844 * strong directional codepoint" rule. Essentially, this works round the fact that
3845 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3846 * direction inference). The latter is cleaner but still not widely supported.
3848 * @param string $text Text to wrap
3849 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3851 public function embedBidi( $text = '' ) {
3852 $dir = Language
::strongDirFromContent( $text );
3853 if ( $dir === 'ltr' ) {
3854 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3855 return self
::$lre . $text . self
::$pdf;
3857 if ( $dir === 'rtl' ) {
3858 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3859 return self
::$rle . $text . self
::$pdf;
3861 // No strong directionality: do not wrap
3866 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3867 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3868 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3869 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3872 * @param string $str The validated block duration in English
3873 * @param User $user User object to use timezone from or null for $wgUser
3874 * @return string Somehow translated block duration
3875 * @see LanguageFi.php for example implementation
3877 function translateBlockExpiry( $str, User
$user = null ) {
3878 $duration = SpecialBlock
::getSuggestedDurations( $this );
3879 foreach ( $duration as $show => $value ) {
3880 if ( strcmp( $str, $value ) == 0 ) {
3881 return htmlspecialchars( trim( $show ) );
3885 if ( wfIsInfinity( $str ) ) {
3886 foreach ( $duration as $show => $value ) {
3887 if ( wfIsInfinity( $value ) ) {
3888 return htmlspecialchars( trim( $show ) );
3893 // If all else fails, return a standard duration or timestamp description.
3894 $time = strtotime( $str, 0 );
3895 if ( $time === false ) { // Unknown format. Return it as-is in case.
3897 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3898 // $time is relative to 0 so it's a duration length.
3899 return $this->formatDuration( $time );
3900 } else { // It's an absolute timestamp.
3901 if ( $time === 0 ) {
3902 // wfTimestamp() handles 0 as current time instead of epoch.
3903 $time = '19700101000000';
3906 return $this->userTimeAndDate( $time, $user );
3908 return $this->timeanddate( $time );
3913 * languages like Chinese need to be segmented in order for the diff
3916 * @param string $text
3919 public function segmentForDiff( $text ) {
3924 * and unsegment to show the result
3926 * @param string $text
3929 public function unsegmentForDiff( $text ) {
3934 * Return the LanguageConverter used in the Language
3937 * @return LanguageConverter
3939 public function getConverter() {
3940 return $this->mConverter
;
3944 * convert text to all supported variants
3946 * @param string $text
3949 public function autoConvertToAllVariants( $text ) {
3950 return $this->mConverter
->autoConvertToAllVariants( $text );
3954 * convert text to different variants of a language.
3956 * @param string $text
3959 public function convert( $text ) {
3960 return $this->mConverter
->convert( $text );
3964 * Convert a Title object to a string in the preferred variant
3966 * @param Title $title
3969 public function convertTitle( $title ) {
3970 return $this->mConverter
->convertTitle( $title );
3974 * Convert a namespace index to a string in the preferred variant
3979 public function convertNamespace( $ns ) {
3980 return $this->mConverter
->convertNamespace( $ns );
3984 * Check if this is a language with variants
3988 public function hasVariants() {
3989 return count( $this->getVariants() ) > 1;
3993 * Check if the language has the specific variant
3996 * @param string $variant
3999 public function hasVariant( $variant ) {
4000 return (bool)$this->mConverter
->validateVariant( $variant );
4004 * Perform output conversion on a string, and encode for safe HTML output.
4005 * @param string $text Text to be converted
4006 * @param bool $isTitle Whether this conversion is for the article title
4008 * @todo this should get integrated somewhere sane
4010 public function convertHtml( $text, $isTitle = false ) {
4011 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4015 * @param string $key
4018 public function convertCategoryKey( $key ) {
4019 return $this->mConverter
->convertCategoryKey( $key );
4023 * Get the list of variants supported by this language
4024 * see sample implementation in LanguageZh.php
4026 * @return array An array of language codes
4028 public function getVariants() {
4029 return $this->mConverter
->getVariants();
4035 public function getPreferredVariant() {
4036 return $this->mConverter
->getPreferredVariant();
4042 public function getDefaultVariant() {
4043 return $this->mConverter
->getDefaultVariant();
4049 public function getURLVariant() {
4050 return $this->mConverter
->getURLVariant();
4054 * If a language supports multiple variants, it is
4055 * possible that non-existing link in one variant
4056 * actually exists in another variant. this function
4057 * tries to find it. See e.g. LanguageZh.php
4058 * The input parameters may be modified upon return
4060 * @param string &$link The name of the link
4061 * @param Title &$nt The title object of the link
4062 * @param bool $ignoreOtherCond To disable other conditions when
4063 * we need to transclude a template or update a category's link
4065 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4066 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4070 * returns language specific options used by User::getPageRenderHash()
4071 * for example, the preferred language variant
4075 function getExtraHashOptions() {
4076 return $this->mConverter
->getExtraHashOptions();
4080 * For languages that support multiple variants, the title of an
4081 * article may be displayed differently in different variants. this
4082 * function returns the apporiate title defined in the body of the article.
4086 public function getParsedTitle() {
4087 return $this->mConverter
->getParsedTitle();
4091 * Refresh the cache of conversion tables when
4092 * MediaWiki:Conversiontable* is updated.
4094 * @param Title $title The Title of the page being updated
4096 public function updateConversionTable( Title
$title ) {
4097 $this->mConverter
->updateConversionTable( $title );
4101 * Prepare external link text for conversion. When the text is
4102 * a URL, it shouldn't be converted, and it'll be wrapped in
4103 * the "raw" tag (-{R| }-) to prevent conversion.
4105 * This function is called "markNoConversion" for historical
4108 * @param string $text Text to be used for external link
4109 * @param bool $noParse Wrap it without confirming it's a real URL first
4110 * @return string The tagged text
4112 public function markNoConversion( $text, $noParse = false ) {
4113 // Excluding protocal-relative URLs may avoid many false positives.
4114 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4115 return $this->mConverter
->markNoConversion( $text );
4122 * A regular expression to match legal word-trailing characters
4123 * which should be merged onto a link of the form [[foo]]bar.
4127 public function linkTrail() {
4128 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4132 * A regular expression character set to match legal word-prefixing
4133 * characters which should be merged onto a link of the form foo[[bar]].
4137 public function linkPrefixCharset() {
4138 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4142 * Get the "parent" language which has a converter to convert a "compatible" language
4143 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4145 * @return Language|null
4148 public function getParentLanguage() {
4149 if ( $this->mParentLanguage
!== false ) {
4150 return $this->mParentLanguage
;
4153 $code = explode( '-', $this->getCode() )[0];
4154 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4155 $this->mParentLanguage
= null;
4158 $lang = Language
::factory( $code );
4159 if ( !$lang->hasVariant( $this->getCode() ) ) {
4160 $this->mParentLanguage
= null;
4164 $this->mParentLanguage
= $lang;
4169 * Compare with an other language object
4172 * @param Language $lang
4175 public function equals( Language
$lang ) {
4176 return $lang->getCode() === $this->mCode
;
4180 * Get the internal language code for this language object
4182 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4183 * htmlspecialchars() or similar
4187 public function getCode() {
4188 return $this->mCode
;
4192 * Get the code in BCP 47 format which we can use
4193 * inside of html lang="" tags.
4195 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4196 * htmlspecialchars() or similar.
4201 public function getHtmlCode() {
4202 if ( is_null( $this->mHtmlCode
) ) {
4203 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4205 return $this->mHtmlCode
;
4209 * @param string $code
4211 public function setCode( $code ) {
4212 $this->mCode
= $code;
4213 // Ensure we don't leave incorrect cached data lying around
4214 $this->mHtmlCode
= null;
4215 $this->mParentLanguage
= false;
4219 * Get the language code from a file name. Inverse of getFileName()
4220 * @param string $filename $prefix . $languageCode . $suffix
4221 * @param string $prefix Prefix before the language code
4222 * @param string $suffix Suffix after the language code
4223 * @return string Language code, or false if $prefix or $suffix isn't found
4225 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4227 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4228 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4229 if ( !count( $m ) ) {
4232 return str_replace( '_', '-', strtolower( $m[1] ) );
4236 * @param string $code
4237 * @return string Name of the language class
4239 public static function classFromCode( $code ) {
4240 if ( $code == 'en' ) {
4243 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4248 * Get the name of a file for a certain language code
4249 * @param string $prefix Prepend this to the filename
4250 * @param string $code Language code
4251 * @param string $suffix Append this to the filename
4252 * @throws MWException
4253 * @return string $prefix . $mangledCode . $suffix
4255 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4256 if ( !self
::isValidBuiltInCode( $code ) ) {
4257 throw new MWException( "Invalid language code \"$code\"" );
4260 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4264 * @param string $code
4267 public static function getMessagesFileName( $code ) {
4269 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4270 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4275 * @param string $code
4277 * @throws MWException
4280 public static function getJsonMessagesFileName( $code ) {
4283 if ( !self
::isValidBuiltInCode( $code ) ) {
4284 throw new MWException( "Invalid language code \"$code\"" );
4287 return "$IP/languages/i18n/$code.json";
4291 * Get the first fallback for a given language.
4293 * @param string $code
4295 * @return bool|string
4297 public static function getFallbackFor( $code ) {
4298 $fallbacks = self
::getFallbacksFor( $code );
4300 return $fallbacks[0];
4306 * Get the ordered list of fallback languages.
4309 * @param string $code Language code
4310 * @return array Non-empty array, ending in "en"
4312 public static function getFallbacksFor( $code ) {
4313 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4316 // For unknown languages, fallbackSequence returns an empty array,
4317 // hardcode fallback to 'en' in that case.
4318 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4322 * Get the ordered list of fallback languages, ending with the fallback
4323 * language chain for the site language.
4326 * @param string $code Language code
4327 * @return array Array( fallbacks, site fallbacks )
4329 public static function getFallbacksIncludingSiteLanguage( $code ) {
4330 global $wgLanguageCode;
4332 // Usually, we will only store a tiny number of fallback chains, so we
4333 // keep them in static memory.
4334 $cacheKey = "{$code}-{$wgLanguageCode}";
4336 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4337 $fallbacks = self
::getFallbacksFor( $code );
4339 // Append the site's fallback chain, including the site language itself
4340 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4341 array_unshift( $siteFallbacks, $wgLanguageCode );
4343 // Eliminate any languages already included in the chain
4344 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4346 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4348 return self
::$fallbackLanguageCache[$cacheKey];
4352 * Get all messages for a given language
4353 * WARNING: this may take a long time. If you just need all message *keys*
4354 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4356 * @param string $code
4360 public static function getMessagesFor( $code ) {
4361 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4365 * Get a message for a given language
4367 * @param string $key
4368 * @param string $code
4372 public static function getMessageFor( $key, $code ) {
4373 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4377 * Get all message keys for a given language. This is a faster alternative to
4378 * array_keys( Language::getMessagesFor( $code ) )
4381 * @param string $code Language code
4382 * @return array Array of message keys (strings)
4384 public static function getMessageKeysFor( $code ) {
4385 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4389 * @param string $talk
4392 function fixVariableInNamespace( $talk ) {
4393 if ( strpos( $talk, '$1' ) === false ) {
4397 global $wgMetaNamespace;
4398 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4400 # Allow grammar transformations
4401 # Allowing full message-style parsing would make simple requests
4402 # such as action=raw much more expensive than they need to be.
4403 # This will hopefully cover most cases.
4404 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4405 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4406 return str_replace( ' ', '_', $talk );
4413 function replaceGrammarInNamespace( $m ) {
4414 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4418 * Decode an expiry (block, protection, etc) which has come from the DB
4420 * @param string $expiry Database expiry String
4421 * @param bool|int $format True to process using language functions, or TS_ constant
4422 * to return the expiry in a given timestamp
4423 * @param string $infinity If $format is not true, use this string for infinite expiry
4427 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4429 if ( $dbInfinity === null ) {
4430 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4433 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4434 return $format === true
4435 ?
$this->getMessageFromDB( 'infiniteblock' )
4438 return $format === true
4439 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4440 : wfTimestamp( $format, $expiry );
4446 * @param int|float $seconds
4447 * @param array $format Optional
4448 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4449 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4450 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4452 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4453 * or 'avoidminutes'.
4456 function formatTimePeriod( $seconds, $format = [] ) {
4457 if ( !is_array( $format ) ) {
4458 $format = [ 'avoid' => $format ]; // For backwards compatibility
4460 if ( !isset( $format['avoid'] ) ) {
4461 $format['avoid'] = false;
4463 if ( !isset( $format['noabbrevs'] ) ) {
4464 $format['noabbrevs'] = false;
4466 $secondsMsg = wfMessage(
4467 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4468 $minutesMsg = wfMessage(
4469 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4470 $hoursMsg = wfMessage(
4471 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4472 $daysMsg = wfMessage(
4473 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4475 if ( round( $seconds * 10 ) < 100 ) {
4476 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4477 $s = $secondsMsg->params( $s )->text();
4478 } elseif ( round( $seconds ) < 60 ) {
4479 $s = $this->formatNum( round( $seconds ) );
4480 $s = $secondsMsg->params( $s )->text();
4481 } elseif ( round( $seconds ) < 3600 ) {
4482 $minutes = floor( $seconds / 60 );
4483 $secondsPart = round( fmod( $seconds, 60 ) );
4484 if ( $secondsPart == 60 ) {
4488 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4490 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4491 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4492 $hours = floor( $seconds / 3600 );
4493 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4494 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4495 if ( $secondsPart == 60 ) {
4499 if ( $minutes == 60 ) {
4503 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4505 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4506 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4507 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4510 $days = floor( $seconds / 86400 );
4511 if ( $format['avoid'] === 'avoidminutes' ) {
4512 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4513 if ( $hours == 24 ) {
4517 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4519 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4520 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4521 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4522 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4523 if ( $minutes == 60 ) {
4527 if ( $hours == 24 ) {
4531 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4533 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4535 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4537 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4539 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4546 * Format a bitrate for output, using an appropriate
4547 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4548 * the magnitude in question.
4550 * This use base 1000. For base 1024 use formatSize(), for another base
4551 * see formatComputingNumbers().
4556 function formatBitrate( $bps ) {
4557 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4561 * @param int $size Size of the unit
4562 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4563 * @param string $messageKey Message key to be uesd
4566 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4568 return str_replace( '$1', $this->formatNum( $size ),
4569 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4572 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4575 $maxIndex = count( $sizes ) - 1;
4576 while ( $size >= $boundary && $index < $maxIndex ) {
4581 // For small sizes no decimal places necessary
4584 // For MB and bigger two decimal places are smarter
4587 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4589 $size = round( $size, $round );
4590 $text = $this->getMessageFromDB( $msg );
4591 return str_replace( '$1', $this->formatNum( $size ), $text );
4595 * Format a size in bytes for output, using an appropriate
4596 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4598 * This method use base 1024. For base 1000 use formatBitrate(), for
4599 * another base see formatComputingNumbers()
4601 * @param int $size Size to format
4602 * @return string Plain text (not HTML)
4604 function formatSize( $size ) {
4605 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4609 * Make a list item, used by various special pages
4611 * @param string $page Page link
4612 * @param string $details HTML safe text between brackets
4613 * @param bool $oppositedm Add the direction mark opposite to your
4614 * language, to display text properly
4615 * @return HTML escaped string
4617 function specialList( $page, $details, $oppositedm = true ) {
4622 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4626 $this->msg( 'word-separator' )->escaped() .
4627 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4631 * Generate (prev x| next x) (20|50|100...) type links for paging
4633 * @param Title $title Title object to link
4634 * @param int $offset
4636 * @param array $query Optional URL query parameter string
4637 * @param bool $atend Optional param for specified if this is the last page
4640 public function viewPrevNext( Title
$title, $offset, $limit,
4641 array $query = [], $atend = false
4643 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4645 # Make 'previous' link
4646 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4647 if ( $offset > 0 ) {
4648 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4649 $query, $prev, 'prevn-title', 'mw-prevlink' );
4651 $plink = htmlspecialchars( $prev );
4655 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4657 $nlink = htmlspecialchars( $next );
4659 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4660 $query, $next, 'nextn-title', 'mw-nextlink' );
4663 # Make links to set number of items per page
4665 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4666 $numLinks[] = $this->numLink( $title, $offset, $num,
4667 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4670 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4671 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4675 * Helper function for viewPrevNext() that generates links
4677 * @param Title $title Title object to link
4678 * @param int $offset
4680 * @param array $query Extra query parameters
4681 * @param string $link Text to use for the link; will be escaped
4682 * @param string $tooltipMsg Name of the message to use as tooltip
4683 * @param string $class Value of the "class" attribute of the link
4684 * @return string HTML fragment
4686 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4689 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4690 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4691 ->numParams( $limit )->text();
4693 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4694 'title' => $tooltip, 'class' => $class ], $link );
4698 * Get the conversion rule title, if any.
4702 public function getConvRuleTitle() {
4703 return $this->mConverter
->getConvRuleTitle();
4707 * Get the compiled plural rules for the language
4709 * @return array Associative array with plural form, and plural rule as key-value pairs
4711 public function getCompiledPluralRules() {
4712 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4713 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4714 if ( !$pluralRules ) {
4715 foreach ( $fallbacks as $fallbackCode ) {
4716 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4717 if ( $pluralRules ) {
4722 return $pluralRules;
4726 * Get the plural rules for the language
4728 * @return array Associative array with plural form number and plural rule as key-value pairs
4730 public function getPluralRules() {
4731 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4732 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4733 if ( !$pluralRules ) {
4734 foreach ( $fallbacks as $fallbackCode ) {
4735 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4736 if ( $pluralRules ) {
4741 return $pluralRules;
4745 * Get the plural rule types for the language
4747 * @return array Associative array with plural form number and plural rule type as key-value pairs
4749 public function getPluralRuleTypes() {
4750 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4751 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4752 if ( !$pluralRuleTypes ) {
4753 foreach ( $fallbacks as $fallbackCode ) {
4754 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4755 if ( $pluralRuleTypes ) {
4760 return $pluralRuleTypes;
4764 * Find the index number of the plural rule appropriate for the given number
4765 * @param int $number
4766 * @return int The index number of the plural rule
4768 public function getPluralRuleIndexNumber( $number ) {
4769 $pluralRules = $this->getCompiledPluralRules();
4770 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
4775 * Find the plural rule type appropriate for the given number
4776 * For example, if the language is set to Arabic, getPluralType(5) should
4779 * @param int $number
4780 * @return string The name of the plural rule type, e.g. one, two, few, many
4782 public function getPluralRuleType( $number ) {
4783 $index = $this->getPluralRuleIndexNumber( $number );
4784 $pluralRuleTypes = $this->getPluralRuleTypes();
4785 if ( isset( $pluralRuleTypes[$index] ) ) {
4786 return $pluralRuleTypes[$index];