3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
26 * @defgroup Language Language
29 use CLDRPluralRuleParser\Evaluator
;
30 use MediaWiki\MediaWikiServices
;
31 use Wikimedia\Assert\Assert
;
34 * Internationalisation code
39 * Return autonyms in fetchLanguageName(s).
42 const AS_AUTONYMS
= null;
45 * Return all known languages in fetchLanguageName(s).
51 * Return in fetchLanguageName(s) only the languages for which we have at
52 * least some localisation.
55 const SUPPORTED
= 'mwfile';
58 * @var LanguageConverter
62 public $mVariants, $mCode, $mLoaded = false;
63 public $mMagicExtensions = [];
64 private $mHtmlCode = null;
65 /** @var Language|false */
66 private $mParentLanguage = false;
68 public $dateFormatStrings = [];
69 public $mExtendedSpecialPageAliases;
71 /** @var array|null */
72 protected $namespaceNames;
73 protected $mNamespaceIds, $namespaceAliases;
76 * ReplacementArray object caches
78 public $transformData = [];
81 * @var LocalisationCache
83 public static $dataCache;
85 public static $mLangObjCache = [];
88 * Return a fallback chain for messages in getFallbacksFor
91 const MESSAGES_FALLBACKS
= 0;
94 * Return a strict fallback chain in getFallbacksFor
97 const STRICT_FALLBACKS
= 1;
99 public static $mWeekdayMsgs = [
100 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
104 public static $mWeekdayAbbrevMsgs = [
105 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
108 public static $mMonthMsgs = [
109 'january', 'february', 'march', 'april', 'may_long', 'june',
110 'july', 'august', 'september', 'october', 'november',
113 public static $mMonthGenMsgs = [
114 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
115 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
118 public static $mMonthAbbrevMsgs = [
119 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
120 'sep', 'oct', 'nov', 'dec'
123 public static $mIranianCalendarMonthMsgs = [
124 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
125 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
126 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
127 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
130 public static $mHebrewCalendarMonthMsgs = [
131 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
132 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
133 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
134 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
135 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
138 public static $mHebrewCalendarMonthGenMsgs = [
139 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
140 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
141 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
142 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
143 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
146 public static $mHijriCalendarMonthMsgs = [
147 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
148 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
149 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
150 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
157 public static $durationIntervals = [
158 'millennia' => 31556952000,
159 'centuries' => 3155695200,
160 'decades' => 315569520,
161 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
170 * Cache for language fallbacks.
171 * @see Language::getFallbacksIncludingSiteLanguage
175 private static $fallbackLanguageCache = [];
178 * Cache for grammar rules data
179 * @var MapCacheLRU|null
181 private static $grammarTransformations;
184 * Cache for language names
185 * @var HashBagOStuff|null
187 private static $languageNameCache;
190 * Unicode directional formatting characters, for embedBidi()
192 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
193 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
194 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
197 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
198 * - in group 1 if it is LTR
199 * - in group 2 if it is RTL
200 * Does not match if there is no strong directionality codepoint.
202 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
204 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
205 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
207 // @codeCoverageIgnoreStart
208 // phpcs:ignore Generic.Files.LineLength
209 private static $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';
210 // @codeCoverageIgnoreEnd
213 * Get a cached or new language object for a given language code
214 * @param string $code
215 * @throws MWException
218 static function factory( $code ) {
219 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
221 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
222 $code = $wgDummyLanguageCodes[$code];
225 // get the language object to process
226 $langObj = self
::$mLangObjCache[$code] ?? self
::newFromCode( $code );
228 // merge the language object in to get it up front in the cache
229 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
230 // get rid of the oldest ones in case we have an overflow
231 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
237 * Create a language object for a given language code
238 * @param string $code
239 * @param bool $fallback Whether we're going through language fallback chain
240 * @throws MWException
243 protected static function newFromCode( $code, $fallback = false ) {
244 if ( !self
::isValidCode( $code ) ) {
245 throw new MWException( "Invalid language code \"$code\"" );
248 if ( !self
::isValidBuiltInCode( $code ) ) {
249 // It's not possible to customise this code with class files, so
250 // just return a Language object. This is to support uselang= hacks.
251 $lang = new Language
;
252 $lang->mCode
= $code;
256 // Check if there is a language class for the code
257 $class = self
::classFromCode( $code, $fallback );
258 // LanguageCode does not inherit Language
259 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
264 // Keep trying the fallback list until we find an existing class
265 $fallbacks = self
::getFallbacksFor( $code );
266 foreach ( $fallbacks as $fallbackCode ) {
267 if ( !self
::isValidBuiltInCode( $fallbackCode ) ) {
268 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
271 $class = self
::classFromCode( $fallbackCode );
272 if ( class_exists( $class ) ) {
274 $lang->mCode
= $code;
279 throw new MWException( "Invalid fallback sequence for language '$code'" );
283 * Intended for tests that may change configuration in a way that invalidates caches.
287 public static function clearCaches() {
288 if ( !defined( 'MW_PHPUNIT_TEST' ) ) {
289 throw new MWException( __METHOD__
. ' must not be used outside tests' );
291 self
::$dataCache = null;
292 // Reinitialize $dataCache, since it's expected to always be available
293 self
::getLocalisationCache();
294 self
::$mLangObjCache = [];
295 self
::$fallbackLanguageCache = [];
296 self
::$grammarTransformations = null;
297 self
::$languageNameCache = null;
301 * Checks whether any localisation is available for that language tag
302 * in MediaWiki (MessagesXx.php exists).
304 * @param string $code Language tag (in lower case)
305 * @return bool Whether language is supported
308 public static function isSupportedLanguage( $code ) {
309 if ( !self
::isValidBuiltInCode( $code ) ) {
313 if ( $code === 'qqq' ) {
317 return is_readable( self
::getMessagesFileName( $code ) ) ||
318 is_readable( self
::getJsonMessagesFileName( $code ) );
322 * Returns true if a language code string is a well-formed language tag
323 * according to RFC 5646.
324 * This function only checks well-formedness; it doesn't check that
325 * language, script or variant codes actually exist in the repositories.
327 * Based on regexes by Mark Davis of the Unicode Consortium:
328 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
330 * @param string $code
331 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
336 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
339 $alphanum = '[a-z0-9]';
340 $x = 'x'; # private use singleton
341 $singleton = '[a-wy-z]'; # other singleton
342 $s = $lenient ?
'[-_]' : '-';
344 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
345 $script = "$alpha{4}"; # ISO 15924
346 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
347 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
348 $extension = "$singleton(?:$s$alphanum{2,8})+";
349 $privateUse = "$x(?:$s$alphanum{1,8})+";
351 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
352 # Since these are limited, this is safe even later changes to the registry --
353 # the only oddity is that it might change the type of the tag, and thus
354 # the results from the capturing groups.
355 # https://www.iana.org/assignments/language-subtag-registry
357 $grandfathered = "en{$s}GB{$s}oed"
358 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
359 . "|no{$s}(?:bok|nyn)"
360 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
361 . "|zh{$s}min{$s}nan";
363 $variantList = "$variant(?:$s$variant)*";
364 $extensionList = "$extension(?:$s$extension)*";
366 $langtag = "(?:($language)"
369 . "(?:$s$variantList)?"
370 . "(?:$s$extensionList)?"
371 . "(?:$s$privateUse)?)";
373 # The final breakdown, with capturing groups for each of these components
374 # The variants, extensions, grandfathered, and private-use may have interior '-'
376 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
378 return (bool)preg_match( "/$root/", strtolower( $code ) );
382 * Returns true if a language code string is of a valid form, whether or
383 * not it exists. This includes codes which are used solely for
384 * customisation via the MediaWiki namespace.
386 * @param string $code
390 public static function isValidCode( $code ) {
392 Assert
::parameterType( 'string', $code, '$code' );
393 if ( !isset( $cache[$code] ) ) {
394 // People think language codes are html safe, so enforce it.
395 // Ideally we should only allow a-zA-Z0-9-
396 // but, .+ and other chars are often used for {{int:}} hacks
397 // see bugs T39564, T39587, T38938
399 // Protect against path traversal
400 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
401 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
403 return $cache[$code];
407 * Returns true if a language code is of a valid form for the purposes of
408 * internal customisation of MediaWiki, via Messages*.php or *.json.
410 * @param string $code
415 public static function isValidBuiltInCode( $code ) {
416 Assert
::parameterType( 'string', $code, '$code' );
418 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
422 * Returns true if a language code is an IETF tag known to MediaWiki.
429 public static function isKnownLanguageTag( $tag ) {
430 // Quick escape for invalid input to avoid exceptions down the line
431 // when code tries to process tags which are not valid at all.
432 if ( !self
::isValidBuiltInCode( $tag ) ) {
436 if ( isset( MediaWiki\Languages\Data\Names
::$names[$tag] )
437 || self
::fetchLanguageName( $tag, $tag ) !== ''
446 * Get the LocalisationCache instance
448 * @return LocalisationCache
450 public static function getLocalisationCache() {
451 if ( is_null( self
::$dataCache ) ) {
452 global $wgLocalisationCacheConf;
453 $class = $wgLocalisationCacheConf['class'];
454 self
::$dataCache = new $class( $wgLocalisationCacheConf );
456 return self
::$dataCache;
459 function __construct() {
460 // @phan-suppress-next-line PhanTypeMismatchProperty
461 $this->mConverter
= new FakeConverter( $this );
462 // Set the code to the name of the descendant
463 if ( static::class === 'Language' ) {
466 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
468 self
::getLocalisationCache();
472 * Reduce memory usage
473 * @suppress PhanTypeSuspiciousNonTraversableForeach
475 function __destruct() {
476 foreach ( $this as $name => $value ) {
477 unset( $this->$name );
482 * Hook which will be called if this is the content language.
483 * Descendants can use this to register hook functions or modify globals
485 function initContLang() {
492 public function getFallbackLanguages() {
493 return self
::getFallbacksFor( $this->mCode
);
497 * Exports $wgBookstoreListEn
500 public function getBookstoreList() {
501 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
505 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
506 * available in localised form, it will be included in English.
508 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
510 public function getNamespaces() {
511 if ( is_null( $this->namespaceNames
) ) {
512 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
514 $validNamespaces = MediaWikiServices
::getInstance()->getNamespaceInfo()->
515 getCanonicalNamespaces();
517 $this->namespaceNames
= $wgExtraNamespaces +
518 self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
519 $this->namespaceNames +
= $validNamespaces;
521 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
522 if ( $wgMetaNamespaceTalk ) {
523 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
525 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
526 $this->namespaceNames
[NS_PROJECT_TALK
] =
527 $this->fixVariableInNamespace( $talk );
530 # Sometimes a language will be localised but not actually exist on this wiki.
531 // @phan-suppress-next-line PhanTypeMismatchForeach
532 foreach ( $this->namespaceNames
as $key => $text ) {
533 if ( !isset( $validNamespaces[$key] ) ) {
534 unset( $this->namespaceNames
[$key] );
538 # The above mixing may leave namespaces out of canonical order.
539 # Re-order by namespace ID number...
540 // @phan-suppress-next-line PhanTypeMismatchArgumentInternal
541 ksort( $this->namespaceNames
);
543 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
546 return $this->namespaceNames
;
550 * Arbitrarily set all of the namespace names at once. Mainly used for testing
551 * @param array $namespaces Array of namespaces (id => name)
553 public function setNamespaces( array $namespaces ) {
554 $this->namespaceNames
= $namespaces;
555 $this->mNamespaceIds
= null;
559 * Resets all of the namespace caches. Mainly used for testing
561 public function resetNamespaces() {
562 $this->namespaceNames
= null;
563 $this->mNamespaceIds
= null;
564 $this->namespaceAliases
= null;
568 * A convenience function that returns getNamespaces() with spaces instead of underscores
569 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
573 public function getFormattedNamespaces() {
574 $ns = $this->getNamespaces();
575 foreach ( $ns as $k => $v ) {
576 $ns[$k] = strtr( $v, '_', ' ' );
582 * Get a namespace value by key
585 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
586 * echo $mw_ns; // prints 'MediaWiki'
589 * @param int $index The array key of the namespace to return
590 * @return string|bool String if the namespace value exists, otherwise false
592 public function getNsText( $index ) {
593 $ns = $this->getNamespaces();
594 return $ns[$index] ??
false;
598 * A convenience function that returns the same thing as
599 * getNsText() except with '_' changed to ' ', useful for
603 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
604 * echo $mw_ns; // prints 'MediaWiki talk'
607 * @param int $index The array key of the namespace to return
608 * @return string Namespace name without underscores (empty string if namespace does not exist)
610 public function getFormattedNsText( $index ) {
611 $ns = $this->getNsText( $index );
612 return strtr( $ns, '_', ' ' );
616 * Returns gender-dependent namespace alias if available.
617 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
618 * @param int $index Namespace index
619 * @param string $gender Gender key (male, female... )
623 public function getGenderNsText( $index, $gender ) {
624 global $wgExtraGenderNamespaces;
626 $ns = $wgExtraGenderNamespaces +
627 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
629 return $ns[$index][$gender] ??
$this->getNsText( $index );
633 * Whether this language uses gender-dependent namespace aliases.
634 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
638 public function needsGenderDistinction() {
639 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
640 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
641 // $wgExtraGenderNamespaces overrides everything
643 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
644 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
645 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
648 // Check what is in i18n files
649 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
650 return count( $aliases ) > 0;
655 * Get a namespace key by value, case insensitive.
656 * Only matches namespace names for the current language, not the
657 * canonical ones defined in Namespace.php.
659 * @param string $text
660 * @return int|bool An integer if $text is a valid value otherwise false
662 function getLocalNsIndex( $text ) {
663 $lctext = $this->lc( $text );
664 $ids = $this->getNamespaceIds();
665 return $ids[$lctext] ??
false;
671 public function getNamespaceAliases() {
672 if ( is_null( $this->namespaceAliases
) ) {
673 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
677 foreach ( $aliases as $name => $index ) {
678 if ( $index === NS_PROJECT_TALK
) {
679 unset( $aliases[$name] );
680 $name = $this->fixVariableInNamespace( $name );
681 $aliases[$name] = $index;
686 global $wgExtraGenderNamespaces;
687 $genders = $wgExtraGenderNamespaces +
688 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
689 foreach ( $genders as $index => $forms ) {
690 foreach ( $forms as $alias ) {
691 $aliases[$alias] = $index;
695 # Also add converted namespace names as aliases, to avoid confusion.
696 $convertedNames = [];
697 foreach ( $this->getVariants() as $variant ) {
698 if ( $variant === $this->mCode
) {
701 foreach ( $this->getNamespaces() as $ns => $_ ) {
702 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
706 $this->namespaceAliases
= $aliases +
$convertedNames;
708 # Filter out aliases to namespaces that don't exist, e.g. from extensions
709 # that aren't loaded here but are included in the l10n cache.
710 # (array_intersect preserves keys from its first argument)
711 $this->namespaceAliases
= array_intersect(
712 $this->namespaceAliases
,
713 array_keys( $this->getNamespaces() )
717 return $this->namespaceAliases
;
723 public function getNamespaceIds() {
724 if ( is_null( $this->mNamespaceIds
) ) {
725 global $wgNamespaceAliases;
726 # Put namespace names and aliases into a hashtable.
727 # If this is too slow, then we should arrange it so that it is done
728 # before caching. The catch is that at pre-cache time, the above
729 # class-specific fixup hasn't been done.
730 $this->mNamespaceIds
= [];
731 foreach ( $this->getNamespaces() as $index => $name ) {
732 $this->mNamespaceIds
[$this->lc( $name )] = $index;
734 foreach ( $this->getNamespaceAliases() as $name => $index ) {
735 $this->mNamespaceIds
[$this->lc( $name )] = $index;
737 if ( $wgNamespaceAliases ) {
738 foreach ( $wgNamespaceAliases as $name => $index ) {
739 $this->mNamespaceIds
[$this->lc( $name )] = $index;
743 return $this->mNamespaceIds
;
747 * Get a namespace key by value, case insensitive. Canonical namespace
748 * names override custom ones defined for the current language.
750 * @param string $text
751 * @return int|bool An integer if $text is a valid value otherwise false
753 public function getNsIndex( $text ) {
754 $lctext = $this->lc( $text );
755 $ns = MediaWikiServices
::getInstance()->getNamespaceInfo()->
756 getCanonicalIndex( $lctext );
757 if ( $ns !== null ) {
760 $ids = $this->getNamespaceIds();
761 return $ids[$lctext] ??
false;
765 * short names for language variants used for language conversion links.
767 * @param string $code
768 * @param bool $usemsg Use the "variantname-xyz" message if it exists
771 public function getVariantname( $code, $usemsg = true ) {
772 $msg = "variantname-$code";
773 if ( $usemsg && wfMessage( $msg )->exists() ) {
774 return $this->getMessageFromDB( $msg );
776 $name = self
::fetchLanguageName( $code );
778 return $name; # if it's defined as a language name, show that
780 # otherwise, output the language code
786 * @return string[]|bool List of date format preference keys, or false if disabled.
788 public function getDatePreferences() {
789 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
795 function getDateFormats() {
796 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
800 * @return array|string
802 public function getDefaultDateFormat() {
803 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
804 if ( $df === 'dmy or mdy' ) {
805 global $wgAmericanDates;
806 return $wgAmericanDates ?
'mdy' : 'dmy';
815 public function getDatePreferenceMigrationMap() {
816 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
822 public function getExtraUserToggles() {
823 wfDeprecated( __METHOD__
, '1.34' );
831 function getUserToggle( $tog ) {
832 return $this->getMessageFromDB( "tog-$tog" );
836 * Get an array of language names, indexed by code.
837 * @param null|string $inLanguage Code of language in which to return the names
838 * Use self::AS_AUTONYMS for autonyms (native names)
839 * @param string $include One of:
840 * self::ALL all available languages
841 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
842 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
843 * @return array Language code => language name (sorted by key)
846 public static function fetchLanguageNames( $inLanguage = self
::AS_AUTONYMS
, $include = 'mw' ) {
847 $cacheKey = $inLanguage === self
::AS_AUTONYMS ?
'null' : $inLanguage;
848 $cacheKey .= ":$include";
849 if ( self
::$languageNameCache === null ) {
850 self
::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
853 $ret = self
::$languageNameCache->get( $cacheKey );
855 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
856 self
::$languageNameCache->set( $cacheKey, $ret );
862 * Uncached helper for fetchLanguageNames
863 * @param null|string $inLanguage Code of language in which to return the names
864 * Use self::AS_AUTONYMS for autonyms (native names)
865 * @param string $include One of:
866 * self::ALL all available languages
867 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
868 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
869 * @return array Language code => language name (sorted by key)
871 private static function fetchLanguageNamesUncached(
872 $inLanguage = self
::AS_AUTONYMS
,
875 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
877 // If passed an invalid language code to use, fallback to en
878 if ( $inLanguage !== self
::AS_AUTONYMS
&& !self
::isValidCode( $inLanguage ) ) {
885 # TODO: also include when $inLanguage is null, when this code is more efficient
886 Hooks
::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
889 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names
::$names;
890 if ( $wgUsePigLatinVariant ) {
891 // Pig Latin (for variant development)
892 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
895 foreach ( $mwNames as $mwCode => $mwName ) {
896 # - Prefer own MediaWiki native name when not using the hook
897 # - For other names just add if not added through the hook
898 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
899 $names[$mwCode] = $mwName;
903 if ( $include === self
::ALL
) {
909 $coreCodes = array_keys( $mwNames );
910 foreach ( $coreCodes as $coreCode ) {
911 $returnMw[$coreCode] = $names[$coreCode];
914 if ( $include === self
::SUPPORTED
) {
916 # We do this using a foreach over the codes instead of a directory
917 # loop so that messages files in extensions will work correctly.
918 foreach ( $returnMw as $code => $value ) {
919 if ( is_readable( self
::getMessagesFileName( $code ) )
920 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
922 $namesMwFile[$code] = $names[$code];
926 ksort( $namesMwFile );
931 # 'mw' option; default if it's not one of the other two options (all/mwfile)
936 * @param string $code The code of the language for which to get the name
937 * @param null|string $inLanguage Code of language in which to return the name
938 * (SELF::AS_AUTONYMS for autonyms)
939 * @param string $include See fetchLanguageNames()
940 * @return string Language name or empty
943 public static function fetchLanguageName(
945 $inLanguage = self
::AS_AUTONYMS
,
948 $code = strtolower( $code );
949 $array = self
::fetchLanguageNames( $inLanguage, $include );
950 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
954 * Get a message from the MediaWiki namespace.
956 * @param string $msg Message name
959 public function getMessageFromDB( $msg ) {
960 return $this->msg( $msg )->text();
964 * Get message object in this language. Only for use inside this class.
966 * @param string $msg Message name
969 protected function msg( $msg ) {
970 return wfMessage( $msg )->inLanguage( $this );
977 public function getMonthName( $key ) {
978 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
984 public function getMonthNamesArray() {
985 $monthNames = [ '' ];
986 for ( $i = 1; $i < 13; $i++
) {
987 $monthNames[] = $this->getMonthName( $i );
996 public function getMonthNameGen( $key ) {
997 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1001 * @param string $key
1004 public function getMonthAbbreviation( $key ) {
1005 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1011 public function getMonthAbbreviationsArray() {
1012 $monthNames = [ '' ];
1013 for ( $i = 1; $i < 13; $i++
) {
1014 $monthNames[] = $this->getMonthAbbreviation( $i );
1020 * @param string $key
1023 public function getWeekdayName( $key ) {
1024 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1028 * @param string $key
1031 function getWeekdayAbbreviation( $key ) {
1032 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1036 * @param string $key
1039 function getIranianCalendarMonthName( $key ) {
1040 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1044 * @param string $key
1047 function getHebrewCalendarMonthName( $key ) {
1048 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1052 * @param string $key
1055 function getHebrewCalendarMonthNameGen( $key ) {
1056 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1060 * @param string $key
1063 function getHijriCalendarMonthName( $key ) {
1064 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1068 * Pass through result from $dateTimeObj->format()
1069 * @param DateTime|bool|null &$dateTimeObj
1071 * @param DateTimeZone|bool|null $zone
1072 * @param string $code
1075 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1076 if ( !$dateTimeObj ) {
1077 $dateTimeObj = DateTime
::createFromFormat(
1078 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1081 return $dateTimeObj->format( $code );
1085 * This is a workalike of PHP's date() function, but with better
1086 * internationalisation, a reduced set of format characters, and a better
1089 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1090 * the PHP manual for definitions. There are a number of extensions, which
1093 * xn Do not translate digits of the next numeric format character
1094 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1095 * xr Use roman numerals for the next numeric format character
1096 * xh Use hebrew numerals for the next numeric format character
1098 * xg Genitive month name
1100 * xij j (day number) in Iranian calendar
1101 * xiF F (month name) in Iranian calendar
1102 * xin n (month number) in Iranian calendar
1103 * xiy y (two digit year) in Iranian calendar
1104 * xiY Y (full year) in Iranian calendar
1105 * xit t (days in month) in Iranian calendar
1106 * xiz z (day of the year) in Iranian calendar
1108 * xjj j (day number) in Hebrew calendar
1109 * xjF F (month name) in Hebrew calendar
1110 * xjt t (days in month) in Hebrew calendar
1111 * xjx xg (genitive month name) in Hebrew calendar
1112 * xjn n (month number) in Hebrew calendar
1113 * xjY Y (full year) in Hebrew calendar
1115 * xmj j (day number) in Hijri calendar
1116 * xmF F (month name) in Hijri calendar
1117 * xmn n (month number) in Hijri calendar
1118 * xmY Y (full year) in Hijri calendar
1120 * xkY Y (full year) in Thai solar calendar. Months and days are
1121 * identical to the Gregorian calendar
1122 * xoY Y (full year) in Minguo calendar or Juche year.
1123 * Months and days are identical to the
1124 * Gregorian calendar
1125 * xtY Y (full year) in Japanese nengo. Months and days are
1126 * identical to the Gregorian calendar
1128 * Characters enclosed in double quotes will be considered literal (with
1129 * the quotes themselves removed). Unmatched quotes will be considered
1130 * literal quotes. Example:
1132 * "The month is" F => The month is January
1135 * Backslash escaping is also supported.
1137 * Input timestamp is assumed to be pre-normalized to the desired local
1138 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1139 * $ts is UTC if $zone is not given.
1141 * @param string $format
1142 * @param string $ts 14-character timestamp
1145 * @param DateTimeZone|null $zone Timezone of $ts
1146 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1147 * Only makes sense if $ts is the current time.
1148 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1150 * @throws MWException
1153 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1158 $dateTimeObj = false;
1167 $usedSecond = false;
1168 $usedMinute = false;
1175 $usedISOYear = false;
1176 $usedIsLeapYear = false;
1178 $usedHebrewMonth = false;
1179 $usedIranianMonth = false;
1180 $usedHijriMonth = false;
1181 $usedHebrewYear = false;
1182 $usedIranianYear = false;
1183 $usedHijriYear = false;
1184 $usedTennoYear = false;
1186 if ( strlen( $ts ) !== 14 ) {
1187 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1190 if ( !ctype_digit( $ts ) ) {
1191 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1194 $formatLength = strlen( $format );
1195 for ( $p = 0; $p < $formatLength; $p++
) {
1197 $code = $format[$p];
1198 if ( $code == 'x' && $p < $formatLength - 1 ) {
1199 $code .= $format[++
$p];
1202 if ( ( $code === 'xi'
1208 && $p < $formatLength - 1 ) {
1209 $code .= $format[++
$p];
1220 $rawToggle = !$rawToggle;
1230 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1233 $usedHebrewMonth = true;
1235 $hebrew = self
::tsToHebrew( $ts );
1237 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1241 $num = substr( $ts, 6, 2 );
1245 $s .= $this->getWeekdayAbbreviation(
1246 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1251 $num = intval( substr( $ts, 6, 2 ) );
1256 $iranian = self
::tsToIranian( $ts );
1263 $hijri = self
::tsToHijri( $ts );
1270 $hebrew = self
::tsToHebrew( $ts );
1276 $s .= $this->getWeekdayName(
1277 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1282 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1285 $usedIranianMonth = true;
1287 $iranian = self
::tsToIranian( $ts );
1289 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1292 $usedHijriMonth = true;
1294 $hijri = self
::tsToHijri( $ts );
1296 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1299 $usedHebrewMonth = true;
1301 $hebrew = self
::tsToHebrew( $ts );
1303 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1307 $num = substr( $ts, 4, 2 );
1311 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1315 $num = intval( substr( $ts, 4, 2 ) );
1318 $usedIranianMonth = true;
1320 $iranian = self
::tsToIranian( $ts );
1325 $usedHijriMonth = true;
1327 $hijri = self
::tsToHijri( $ts );
1332 $usedHebrewMonth = true;
1334 $hebrew = self
::tsToHebrew( $ts );
1339 $usedHebrewMonth = true;
1341 $hebrew = self
::tsToHebrew( $ts );
1347 $num = substr( $ts, 0, 4 );
1350 $usedIranianYear = true;
1352 $iranian = self
::tsToIranian( $ts );
1357 $usedHijriYear = true;
1359 $hijri = self
::tsToHijri( $ts );
1364 $usedHebrewYear = true;
1366 $hebrew = self
::tsToHebrew( $ts );
1373 $thai = self
::tsToYear( $ts, 'thai' );
1380 $minguo = self
::tsToYear( $ts, 'minguo' );
1385 $usedTennoYear = true;
1387 $tenno = self
::tsToYear( $ts, 'tenno' );
1393 $num = substr( $ts, 2, 2 );
1396 $usedIranianYear = true;
1398 $iranian = self
::tsToIranian( $ts );
1400 $num = substr( $iranian[0], -2 );
1403 $usedIranianYear = true;
1405 $iranian = self
::tsToIranian( $ts );
1407 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1410 $usedIranianYear = true;
1412 $iranian = self
::tsToIranian( $ts );
1418 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1422 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1426 $h = substr( $ts, 8, 2 );
1427 $num = $h %
12 ?
: 12;
1431 $num = intval( substr( $ts, 8, 2 ) );
1435 $h = substr( $ts, 8, 2 );
1436 $num = sprintf( '%02d', $h %
12 ?
: 12 );
1440 $num = substr( $ts, 8, 2 );
1444 $num = substr( $ts, 10, 2 );
1448 $num = substr( $ts, 12, 2 );
1458 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1464 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1468 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1472 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1475 $usedIsLeapYear = true;
1476 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1479 $usedISOYear = true;
1480 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1487 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1490 # Backslash escaping
1491 if ( $p < $formatLength - 1 ) {
1492 $s .= $format[++
$p];
1499 if ( $p < $formatLength - 1 ) {
1500 $endQuote = strpos( $format, '"', $p +
1 );
1501 if ( $endQuote === false ) {
1502 # No terminating quote, assume literal "
1505 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1509 # Quote at end of string, assume literal "
1516 if ( $num !== false ) {
1517 if ( $rawToggle ||
$raw ) {
1520 } elseif ( $roman ) {
1521 $s .= self
::romanNumeral( $num );
1523 } elseif ( $hebrewNum ) {
1524 $s .= self
::hebrewNumeral( $num );
1527 $s .= $this->formatNum( $num, true );
1532 if ( $ttl === 'unused' ) {
1533 // No need to calculate the TTL, the caller wont use it anyway.
1534 } elseif ( $usedSecond ) {
1536 } elseif ( $usedMinute ) {
1537 $ttl = 60 - substr( $ts, 12, 2 );
1538 } elseif ( $usedHour ) {
1539 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1540 } elseif ( $usedAMPM ) {
1541 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1542 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1546 $usedIranianMonth ||
1553 // @todo Someone who understands the non-Gregorian calendars
1554 // should write proper logic for them so that they don't need purged every day.
1555 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1556 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1559 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1560 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1563 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1564 $timeRemainingInDay;
1565 } elseif ( $usedISOYear ) {
1566 // December 28th falls on the last ISO week of the year, every year.
1567 // The last ISO week of a year can be 52 or 53.
1568 $lastWeekOfISOYear = DateTime
::createFromFormat(
1570 substr( $ts, 0, 4 ) . '1228',
1571 $zone ?
: new DateTimeZone( 'UTC' )
1573 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1574 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1575 $timeRemainingInWeek =
1576 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1577 +
$timeRemainingInDay;
1578 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1583 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1584 substr( $ts, 6, 2 ) ) * 86400
1585 +
$timeRemainingInDay;
1586 } elseif ( $usedYear ) {
1588 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1589 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1590 +
$timeRemainingInDay;
1591 } elseif ( $usedIsLeapYear ) {
1592 $year = substr( $ts, 0, 4 );
1593 $timeRemainingInYear =
1594 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1595 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1596 +
$timeRemainingInDay;
1598 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1599 // this isn't a leap year. see when the next one starts
1600 $nextCandidate = $year - $mod +
4;
1601 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1602 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1603 $timeRemainingInYear;
1605 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1606 $timeRemainingInYear;
1609 // this is a leap year, so the next year isn't
1610 $possibleTtls[] = $timeRemainingInYear;
1614 if ( $possibleTtls ) {
1615 $ttl = min( $possibleTtls );
1622 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1623 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1626 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1627 * Gregorian dates to Iranian dates. Originally written in C, it
1628 * is released under the terms of GNU Lesser General Public
1629 * License. Conversion to PHP was performed by Niklas Laxström.
1631 * Link: http://www.farsiweb.info/jalali/jalali.c
1637 private static function tsToIranian( $ts ) {
1638 $gy = substr( $ts, 0, 4 ) - 1600;
1639 $gm = substr( $ts, 4, 2 ) - 1;
1640 $gd = substr( $ts, 6, 2 ) - 1;
1642 # Days passed from the beginning (including leap years)
1644 +
floor( ( $gy +
3 ) / 4 )
1645 - floor( ( $gy +
99 ) / 100 )
1646 +
floor( ( $gy +
399 ) / 400 );
1648 // Add days of the past months of this year
1649 for ( $i = 0; $i < $gm; $i++
) {
1650 $gDayNo +
= self
::$GREG_DAYS[$i];
1654 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1658 // Days passed in current month
1659 $gDayNo +
= (int)$gd;
1661 $jDayNo = $gDayNo - 79;
1663 $jNp = floor( $jDayNo / 12053 );
1666 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1669 if ( $jDayNo >= 366 ) {
1670 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1671 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1676 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1677 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1683 return [ $jy, $jm, $jd, $jz ];
1687 * Converting Gregorian dates to Hijri dates.
1689 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1691 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1697 private static function tsToHijri( $ts ) {
1698 $year = substr( $ts, 0, 4 );
1699 $month = substr( $ts, 4, 2 );
1700 $day = substr( $ts, 6, 2 );
1708 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1709 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1711 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1712 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1713 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1716 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1717 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1720 $zl = $zjd - 1948440 +
10632;
1721 $zn = (int)( ( $zl - 1 ) / 10631 );
1722 $zl = $zl - 10631 * $zn +
354;
1723 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1724 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1725 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1726 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1727 $zm = (int)( ( 24 * $zl ) / 709 );
1728 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1729 $zy = 30 * $zn +
$zj - 30;
1731 return [ $zy, $zm, $zd ];
1735 * Converting Gregorian dates to Hebrew dates.
1737 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1738 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1739 * to translate the relevant functions into PHP and release them under
1742 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1743 * and Adar II is 14. In a non-leap year, Adar is 6.
1749 private static function tsToHebrew( $ts ) {
1751 $year = substr( $ts, 0, 4 );
1752 $month = substr( $ts, 4, 2 );
1753 $day = substr( $ts, 6, 2 );
1755 # Calculate Hebrew year
1756 $hebrewYear = $year +
3760;
1758 # Month number when September = 1, August = 12
1760 if ( $month > 12 ) {
1767 # Calculate day of year from 1 September
1769 for ( $i = 1; $i < $month; $i++
) {
1773 # Check if the year is leap
1774 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1777 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1784 # Calculate the start of the Hebrew year
1785 $start = self
::hebrewYearStart( $hebrewYear );
1787 # Calculate next year's start
1788 if ( $dayOfYear <= $start ) {
1789 # Day is before the start of the year - it is the previous year
1791 $nextStart = $start;
1795 # Add days since previous year's 1 September
1797 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1801 # Start of the new (previous) year
1802 $start = self
::hebrewYearStart( $hebrewYear );
1805 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1808 # Calculate Hebrew day of year
1809 $hebrewDayOfYear = $dayOfYear - $start;
1811 # Difference between year's days
1812 $diff = $nextStart - $start;
1813 # Add 12 (or 13 for leap years) days to ignore the difference between
1814 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1815 # difference is only about the year type
1816 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1822 # Check the year pattern, and is leap year
1823 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1824 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1825 # and non-leap years
1826 $yearPattern = $diff %
30;
1827 # Check if leap year
1828 $isLeap = $diff >= 30;
1830 # Calculate day in the month from number of day in the Hebrew year
1831 # Don't check Adar - if the day is not in Adar, we will stop before;
1832 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1833 $hebrewDay = $hebrewDayOfYear;
1836 while ( $hebrewMonth <= 12 ) {
1837 # Calculate days in this month
1838 if ( $isLeap && $hebrewMonth == 6 ) {
1839 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1841 if ( $hebrewDay <= $days ) {
1845 # Subtract the days of Adar I
1846 $hebrewDay -= $days;
1849 if ( $hebrewDay <= $days ) {
1854 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1855 # Cheshvan in a complete year (otherwise as the rule below)
1857 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1858 # Kislev in an incomplete year (otherwise as the rule below)
1861 # Odd months have 30 days, even have 29
1862 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1864 if ( $hebrewDay <= $days ) {
1865 # In the current month
1868 # Subtract the days of the current month
1869 $hebrewDay -= $days;
1870 # Try in the next month
1875 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1879 * This calculates the Hebrew year start, as days since 1 September.
1880 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1881 * Used for Hebrew date.
1887 private static function hebrewYearStart( $year ) {
1888 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1889 $b = intval( ( $year - 1 ) %
4 );
1890 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1894 $Mar = intval( $m );
1900 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1901 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1903 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1905 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1909 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1914 * Algorithm to convert Gregorian dates to Thai solar dates,
1915 * Minguo dates or Minguo dates.
1917 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1918 * https://en.wikipedia.org/wiki/Minguo_calendar
1919 * https://en.wikipedia.org/wiki/Japanese_era_name
1921 * @param string $ts 14-character timestamp
1922 * @param string $cName Calender name
1923 * @return array Converted year, month, day
1925 private static function tsToYear( $ts, $cName ) {
1926 $gy = substr( $ts, 0, 4 );
1927 $gm = substr( $ts, 4, 2 );
1928 $gd = substr( $ts, 6, 2 );
1930 if ( !strcmp( $cName, 'thai' ) ) {
1932 # Add 543 years to the Gregorian calendar
1933 # Months and days are identical
1934 $gy_offset = $gy +
543;
1935 # fix for dates between 1912 and 1941
1936 # https://en.wikipedia.org/?oldid=836596673#New_year
1937 if ( $gy >= 1912 && $gy <= 1940 ) {
1941 $gm = ( $gm - 3 ) %
12;
1943 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1945 # Deduct 1911 years from the Gregorian calendar
1946 # Months and days are identical
1947 $gy_offset = $gy - 1911;
1948 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1949 # Nengō dates up to Meiji period
1950 # Deduct years from the Gregorian calendar
1951 # depending on the nengo periods
1952 # Months and days are identical
1954 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1955 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1958 $gy_gannen = $gy - 1868 +
1;
1959 $gy_offset = $gy_gannen;
1960 if ( $gy_gannen == 1 ) {
1963 $gy_offset = '明治' . $gy_offset;
1965 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1966 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1967 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1968 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1969 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1972 $gy_gannen = $gy - 1912 +
1;
1973 $gy_offset = $gy_gannen;
1974 if ( $gy_gannen == 1 ) {
1977 $gy_offset = '大正' . $gy_offset;
1979 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1980 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1981 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1984 $gy_gannen = $gy - 1926 +
1;
1985 $gy_offset = $gy_gannen;
1986 if ( $gy_gannen == 1 ) {
1989 $gy_offset = '昭和' . $gy_offset;
1991 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1992 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1993 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1996 $gy_gannen = $gy - 1989 +
1;
1997 $gy_offset = $gy_gannen;
1998 if ( $gy_gannen == 1 ) {
2001 $gy_offset = '平成' . $gy_offset;
2004 $gy_gannen = $gy - 2019 +
1;
2005 $gy_offset = $gy_gannen;
2006 if ( $gy_gannen == 1 ) {
2009 $gy_offset = '令和' . $gy_offset;
2015 return [ $gy_offset, $gm, $gd ];
2019 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2021 * This is the rule the BIDI algorithm uses to determine the directionality of
2022 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
2023 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2025 * TODO: Does not handle BIDI control characters inside the text.
2026 * TODO: Does not handle unallocated characters.
2028 * @param string $text Text to test
2029 * @return null|string Directionality ('ltr' or 'rtl') or null
2031 private static function strongDirFromContent( $text = '' ) {
2032 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
2035 if ( $matches[1] === '' ) {
2042 * Roman number formatting up to 10000
2048 static function romanNumeral( $num ) {
2050 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
2051 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
2052 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
2053 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2054 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2057 $num = intval( $num );
2058 if ( $num > 10000 ||
$num <= 0 ) {
2063 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2064 if ( $num >= $pow10 ) {
2065 $s .= $table[$i][(int)floor( $num / $pow10 )];
2067 $num = $num %
$pow10;
2073 * Hebrew Gematria number formatting up to 9999
2079 static function hebrewNumeral( $num ) {
2081 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2082 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2095 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2098 $num = intval( $num );
2099 if ( $num > 9999 ||
$num <= 0 ) {
2103 // Round thousands have special notations
2104 if ( $num === 1000 ) {
2106 } elseif ( $num %
1000 === 0 ) {
2107 return $table[0][$num / 1000] . "' אלפים";
2112 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2113 if ( $num >= $pow10 ) {
2114 if ( $num === 15 ||
$num === 16 ) {
2115 $letters[] = $table[0][9];
2116 $letters[] = $table[0][$num - 9];
2119 $letters = array_merge(
2121 (array)$table[$i][intval( $num / $pow10 )]
2124 if ( $pow10 === 1000 ) {
2130 $num = $num %
$pow10;
2133 $preTransformLength = count( $letters );
2134 if ( $preTransformLength === 1 ) {
2135 // Add geresh (single quote) to one-letter numbers
2138 $lastIndex = $preTransformLength - 1;
2139 $letters[$lastIndex] = str_replace(
2140 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2141 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2142 $letters[$lastIndex]
2145 // Add gershayim (double quote) to multiple-letter numbers,
2146 // but exclude numbers with only one letter after the thousands
2147 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2148 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2151 array_splice( $letters, -1, 0, '"' );
2155 return implode( $letters );
2159 * Used by date() and time() to adjust the time output.
2161 * @param string $ts The time in date('YmdHis') format
2162 * @param mixed $tz Adjust the time by this amount (default false, mean we
2163 * get user timecorrection setting)
2166 public function userAdjust( $ts, $tz = false ) {
2167 global $wgUser, $wgLocalTZoffset;
2169 if ( $tz === false ) {
2170 $tz = $wgUser->getOption( 'timecorrection' );
2173 $data = explode( '|', $tz, 3 );
2175 if ( $data[0] == 'ZoneInfo' ) {
2177 $userTZ = new DateTimeZone( $data[2] );
2178 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2179 $date->setTimezone( $userTZ );
2180 return $date->format( 'YmdHis' );
2181 } catch ( Exception
$e ) {
2182 // Unrecognized timezone, default to 'Offset' with the stored offset.
2183 $data[0] = 'Offset';
2187 if ( $data[0] == 'System' ||
$tz == '' ) {
2188 # Global offset in minutes.
2189 $minDiff = $wgLocalTZoffset;
2190 } elseif ( $data[0] == 'Offset' ) {
2191 $minDiff = intval( $data[1] );
2193 $data = explode( ':', $tz );
2194 if ( count( $data ) == 2 ) {
2195 $data[0] = intval( $data[0] );
2196 $data[1] = intval( $data[1] );
2197 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2198 if ( $data[0] < 0 ) {
2199 $minDiff = -$minDiff;
2202 $minDiff = intval( $data[0] ) * 60;
2206 # No difference ? Return time unchanged
2207 if ( $minDiff == 0 ) {
2211 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2212 # Generate an adjusted date; take advantage of the fact that mktime
2213 # will normalize out-of-range values so we don't have to split $minDiff
2214 # into hours and minutes.
2216 (int)substr( $ts, 8, 2 ) ), # Hours
2217 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2218 (int)substr( $ts, 12, 2 ), # Seconds
2219 (int)substr( $ts, 4, 2 ), # Month
2220 (int)substr( $ts, 6, 2 ), # Day
2221 (int)substr( $ts, 0, 4 ) ); # Year
2223 $date = date( 'YmdHis', $t );
2224 Wikimedia\restoreWarnings
();
2230 * This is meant to be used by time(), date(), and timeanddate() to get
2231 * the date preference they're supposed to use, it should be used in
2234 * function timeanddate([...], $format = true) {
2235 * $datePreference = $this->dateFormat($format);
2239 * @param int|string|bool $usePrefs If true, the user's preference is used
2240 * if false, the site/language default is used
2241 * if int/string, assumed to be a format.
2244 function dateFormat( $usePrefs = true ) {
2247 if ( is_bool( $usePrefs ) ) {
2249 $datePreference = $wgUser->getDatePreference();
2251 $datePreference = (string)User
::getDefaultOption( 'date' );
2254 $datePreference = (string)$usePrefs;
2258 if ( $datePreference == '' ) {
2262 return $datePreference;
2266 * Get a format string for a given type and preference
2267 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2268 * @param string $pref The format name as it appears in Messages*.php under
2271 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2275 function getDateFormatString( $type, $pref ) {
2276 $wasDefault = false;
2277 if ( $pref == 'default' ) {
2279 $pref = $this->getDefaultDateFormat();
2282 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2283 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2285 if ( $type === 'pretty' && $df === null ) {
2286 $df = $this->getDateFormatString( 'date', $pref );
2289 if ( !$wasDefault && $df === null ) {
2290 $pref = $this->getDefaultDateFormat();
2291 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2294 $this->dateFormatStrings
[$type][$pref] = $df;
2296 return $this->dateFormatStrings
[$type][$pref];
2300 * @param string $ts The time format which needs to be turned into a
2301 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2302 * @param bool $adj Whether to adjust the time output according to the
2303 * user configured offset ($timecorrection)
2304 * @param mixed $format True to use user's date format preference
2305 * @param string|bool $timecorrection The time offset as returned by
2306 * validateTimeZone() in Special:Preferences
2309 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2310 $ts = wfTimestamp( TS_MW
, $ts );
2312 $ts = $this->userAdjust( $ts, $timecorrection );
2314 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2315 return $this->sprintfDate( $df, $ts );
2319 * @param string $ts The time format which needs to be turned into a
2320 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2321 * @param bool $adj Whether to adjust the time output according to the
2322 * user configured offset ($timecorrection)
2323 * @param mixed $format True to use user's date format preference
2324 * @param string|bool $timecorrection The time offset as returned by
2325 * validateTimeZone() in Special:Preferences
2328 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2329 $ts = wfTimestamp( TS_MW
, $ts );
2331 $ts = $this->userAdjust( $ts, $timecorrection );
2333 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2334 return $this->sprintfDate( $df, $ts );
2338 * @param string $ts The time format which needs to be turned into a
2339 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2340 * @param bool $adj Whether to adjust the time output according to the
2341 * user configured offset ($timecorrection)
2342 * @param mixed $format What format to return, if it's false output the
2343 * default one (default true)
2344 * @param string|bool $timecorrection The time offset as returned by
2345 * validateTimeZone() in Special:Preferences
2348 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2349 $ts = wfTimestamp( TS_MW
, $ts );
2351 $ts = $this->userAdjust( $ts, $timecorrection );
2353 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2354 return $this->sprintfDate( $df, $ts );
2358 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2362 * @param int $seconds The amount of seconds.
2363 * @param array $chosenIntervals The intervals to enable.
2367 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2368 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2372 foreach ( $intervals as $intervalName => $intervalValue ) {
2373 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2374 // duration-years, duration-decades, duration-centuries, duration-millennia
2375 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2376 $segments[] = $message->inLanguage( $this )->escaped();
2379 return $this->listToText( $segments );
2383 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2384 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2388 * @param int $seconds The amount of seconds.
2389 * @param array $chosenIntervals The intervals to enable.
2393 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2394 if ( empty( $chosenIntervals ) ) {
2395 $chosenIntervals = [
2407 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2408 $sortedNames = array_keys( $intervals );
2409 $smallestInterval = array_pop( $sortedNames );
2413 foreach ( $intervals as $name => $length ) {
2414 $value = floor( $seconds / $length );
2416 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2417 $seconds -= $value * $length;
2418 $segments[$name] = $value;
2426 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2428 * @param string $type Can be 'date', 'time' or 'both'
2429 * @param string $ts The time format which needs to be turned into a
2430 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2431 * @param User $user User object used to get preferences for timezone and format
2432 * @param array $options Array, can contain the following keys:
2433 * - 'timecorrection': time correction, can have the following values:
2434 * - true: use user's preference
2435 * - false: don't use time correction
2436 * - int: value of time correction in minutes
2437 * - 'format': format to use, can have the following values:
2438 * - true: use user's preference
2439 * - false: use default preference
2440 * - string: format to use
2444 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2445 $ts = wfTimestamp( TS_MW
, $ts );
2446 $options +
= [ 'timecorrection' => true, 'format' => true ];
2447 if ( $options['timecorrection'] !== false ) {
2448 if ( $options['timecorrection'] === true ) {
2449 $offset = $user->getOption( 'timecorrection' );
2451 $offset = $options['timecorrection'];
2453 $ts = $this->userAdjust( $ts, $offset );
2455 if ( $options['format'] === true ) {
2456 $format = $user->getDatePreference();
2458 $format = $options['format'];
2460 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2461 return $this->sprintfDate( $df, $ts );
2465 * Get the formatted date for the given timestamp and formatted for
2468 * @param mixed $ts Mixed: the time format which needs to be turned into a
2469 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2470 * @param User $user User object used to get preferences for timezone and format
2471 * @param array $options Array, can contain the following keys:
2472 * - 'timecorrection': time correction, can have the following values:
2473 * - true: use user's preference
2474 * - false: don't use time correction
2475 * - int: value of time correction in minutes
2476 * - 'format': format to use, can have the following values:
2477 * - true: use user's preference
2478 * - false: use default preference
2479 * - string: format to use
2483 public function userDate( $ts, User
$user, array $options = [] ) {
2484 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2488 * Get the formatted time for the given timestamp and formatted for
2491 * @param mixed $ts The time format which needs to be turned into a
2492 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2493 * @param User $user User object used to get preferences for timezone and format
2494 * @param array $options Array, can contain the following keys:
2495 * - 'timecorrection': time correction, can have the following values:
2496 * - true: use user's preference
2497 * - false: don't use time correction
2498 * - int: value of time correction in minutes
2499 * - 'format': format to use, can have the following values:
2500 * - true: use user's preference
2501 * - false: use default preference
2502 * - string: format to use
2506 public function userTime( $ts, User
$user, array $options = [] ) {
2507 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2511 * Get the formatted date and time for the given timestamp and formatted for
2514 * @param mixed $ts The time format which needs to be turned into a
2515 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2516 * @param User $user User object used to get preferences for timezone and format
2517 * @param array $options Array, can contain the following keys:
2518 * - 'timecorrection': time correction, can have the following values:
2519 * - true: use user's preference
2520 * - false: don't use time correction
2521 * - int: value of time correction in minutes
2522 * - 'format': format to use, can have the following values:
2523 * - true: use user's preference
2524 * - false: use default preference
2525 * - string: format to use
2529 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2530 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2534 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2536 * Determine the difference between the timestamp and the current time, and
2537 * generate a readable timestamp by returning "<N> <units> ago", where the
2538 * largest possible unit is used.
2540 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2542 * @param MWTimestamp $time
2543 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2544 * @param User|null $user User the timestamp is being generated for
2545 * (or null to use main context's user)
2546 * @return string Formatted timestamp
2548 public function getHumanTimestamp(
2549 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2551 if ( $relativeTo === null ) {
2552 $relativeTo = new MWTimestamp();
2554 if ( $user === null ) {
2555 $user = RequestContext
::getMain()->getUser();
2558 // Adjust for the user's timezone.
2559 $offsetThis = $time->offsetForUser( $user );
2560 $offsetRel = $relativeTo->offsetForUser( $user );
2563 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2564 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2567 // Reset the timezone on the objects.
2568 $time->timestamp
->sub( $offsetThis );
2569 $relativeTo->timestamp
->sub( $offsetRel );
2575 * Convert an MWTimestamp into a pretty human-readable timestamp using
2576 * the given user preferences and relative base time.
2578 * @see Language::getHumanTimestamp
2579 * @param MWTimestamp $ts Timestamp to prettify
2580 * @param MWTimestamp $relativeTo Base timestamp
2581 * @param User $user User preferences to use
2582 * @return string Human timestamp
2585 private function getHumanTimestampInternal(
2586 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2588 $diff = $ts->diff( $relativeTo );
2589 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2590 (int)$relativeTo->timestamp
->format( 'w' ) );
2591 $days = $diff->days ?
: (int)$diffDay;
2592 if ( $diff->invert ||
$days > 5
2593 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2595 // Timestamps are in different years: use full timestamp
2596 // Also do full timestamp for future dates
2598 * @todo FIXME: Add better handling of future timestamps.
2600 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2601 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2602 } elseif ( $days > 5 ) {
2603 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2604 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2605 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2606 } elseif ( $days > 1 ) {
2607 // Timestamp within the past week: show the day of the week and time
2608 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2609 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2611 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2612 $ts = wfMessage( "$weekday-at" )
2613 ->inLanguage( $this )
2614 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2616 } elseif ( $days == 1 ) {
2617 // Timestamp was yesterday: say 'yesterday' and the time.
2618 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2619 $ts = wfMessage( 'yesterday-at' )
2620 ->inLanguage( $this )
2621 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2623 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2624 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2625 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2626 $ts = wfMessage( 'today-at' )
2627 ->inLanguage( $this )
2628 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2631 // From here on in, the timestamp was soon enough ago so that we can simply say
2632 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2633 } elseif ( $diff->h
== 1 ) {
2634 // Less than 90 minutes, but more than an hour ago.
2635 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2636 } elseif ( $diff->i
>= 1 ) {
2637 // A few minutes ago.
2638 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2639 } elseif ( $diff->s
>= 30 ) {
2640 // Less than a minute, but more than 30 sec ago.
2641 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2643 // Less than 30 seconds ago.
2644 $ts = wfMessage( 'just-now' )->text();
2651 * @param string $key
2652 * @return string|null
2654 public function getMessage( $key ) {
2655 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2661 function getAllMessages() {
2662 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2667 * @param string $out
2668 * @param string $string
2671 public function iconv( $in, $out, $string ) {
2672 # Even with //IGNORE iconv can whine about illegal characters in
2673 # *input* string. We just ignore those too.
2674 # REF: https://bugs.php.net/bug.php?id=37166
2675 # REF: https://phabricator.wikimedia.org/T18885
2676 Wikimedia\
suppressWarnings();
2677 $text = iconv( $in, $out . '//IGNORE', $string );
2678 Wikimedia\restoreWarnings
();
2682 // callback functions for ucwords(), ucwordbreaks()
2685 * @param array $matches
2686 * @return mixed|string
2688 function ucwordbreaksCallbackAscii( $matches ) {
2689 return $this->ucfirst( $matches[1] );
2693 * @param array $matches
2696 function ucwordbreaksCallbackMB( $matches ) {
2697 return mb_strtoupper( $matches[0] );
2701 * @param array $matches
2704 function ucwordsCallbackMB( $matches ) {
2705 return mb_strtoupper( $matches[0] );
2709 * Make a string's first character uppercase
2711 * @param string $str
2715 public function ucfirst( $str ) {
2717 if ( $o < 96 ) { // if already uppercase...
2719 } elseif ( $o < 128 ) {
2720 return ucfirst( $str ); // use PHP's ucfirst()
2722 // fall back to more complex logic in case of multibyte strings
2723 return $this->uc( $str, true );
2728 * Convert a string to uppercase
2730 * @param string $str
2731 * @param bool $first
2735 public function uc( $str, $first = false ) {
2737 if ( $this->isMultibyte( $str ) ) {
2738 return $this->mbUpperChar( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2740 return ucfirst( $str );
2743 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2748 * Convert character to uppercase, allowing overrides of the default mb_upper
2749 * behaviour, which is buggy in many ways. Having a conversion table can be
2750 * useful during transitions between PHP versions where unicode changes happen.
2751 * This can make some resources unreachable on-wiki, see discussion at T219279.
2752 * Providing such a conversion table can allow to manage the transition period.
2756 * @param string $char
2760 protected function mbUpperChar( $char ) {
2761 global $wgOverrideUcfirstCharacters;
2762 if ( array_key_exists( $char, $wgOverrideUcfirstCharacters ) ) {
2763 return $wgOverrideUcfirstCharacters[$char];
2765 return mb_strtoupper( $char );
2770 * @param string $str
2771 * @return mixed|string
2773 function lcfirst( $str ) {
2776 return strval( $str );
2777 } elseif ( $o >= 128 ) {
2778 return $this->lc( $str, true );
2779 } elseif ( $o > 96 ) {
2782 $str[0] = strtolower( $str[0] );
2788 * @param string $str
2789 * @param bool $first
2790 * @return mixed|string
2792 function lc( $str, $first = false ) {
2794 if ( $this->isMultibyte( $str ) ) {
2795 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2797 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2800 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2805 * @param string $str
2808 function isMultibyte( $str ) {
2809 return strlen( $str ) !== mb_strlen( $str );
2813 * @param string $str
2814 * @return mixed|string
2816 function ucwords( $str ) {
2817 if ( $this->isMultibyte( $str ) ) {
2818 $str = $this->lc( $str );
2820 // regexp to find first letter in each word (i.e. after each space)
2821 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2823 // function to use to capitalize a single char
2824 return preg_replace_callback(
2826 [ $this, 'ucwordsCallbackMB' ],
2830 return ucwords( strtolower( $str ) );
2835 * capitalize words at word breaks
2837 * @param string $str
2840 function ucwordbreaks( $str ) {
2841 if ( $this->isMultibyte( $str ) ) {
2842 $str = $this->lc( $str );
2844 // since \b doesn't work for UTF-8, we explicitely define word break chars
2845 $breaks = "[ \-\(\)\}\{\.,\?!]";
2847 // find first letter after word break
2848 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2849 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2851 return preg_replace_callback(
2853 [ $this, 'ucwordbreaksCallbackMB' ],
2857 return preg_replace_callback(
2858 '/\b([\w\x80-\xff]+)\b/',
2859 [ $this, 'ucwordbreaksCallbackAscii' ],
2866 * Return a case-folded representation of $s
2868 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2869 * and $s2 are the same except for the case of their characters. It is not
2870 * necessary for the value returned to make sense when displayed.
2872 * Do *not* perform any other normalisation in this function. If a caller
2873 * uses this function when it should be using a more general normalisation
2874 * function, then fix the caller.
2880 function caseFold( $s ) {
2881 return $this->uc( $s );
2885 * TODO: $s is not always a string per T218883
2889 function checkTitleEncoding( $s ) {
2890 if ( is_array( $s ) ) {
2891 throw new MWException( 'Given array to checkTitleEncoding.' );
2893 if ( StringUtils
::isUtf8( $s ) ) {
2897 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2903 function fallback8bitEncoding() {
2904 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2908 * Most writing systems use whitespace to break up words.
2909 * Some languages such as Chinese don't conventionally do this,
2910 * which requires special handling when breaking up words for
2915 function hasWordBreaks() {
2920 * Some languages such as Chinese require word segmentation,
2921 * Specify such segmentation when overridden in derived class.
2923 * @param string $string
2926 function segmentByWord( $string ) {
2931 * Some languages have special punctuation need to be normalized.
2932 * Make such changes here.
2934 * @param string $string
2937 function normalizeForSearch( $string ) {
2938 return self
::convertDoubleWidth( $string );
2942 * convert double-width roman characters to single-width.
2943 * range: ff00-ff5f ~= 0020-007f
2945 * @param string $string
2949 protected static function convertDoubleWidth( $string ) {
2950 static $full = null;
2951 static $half = null;
2953 if ( $full === null ) {
2954 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2955 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2956 $full = str_split( $fullWidth, 3 );
2957 $half = str_split( $halfWidth );
2960 $string = str_replace( $full, $half, $string );
2965 * @param string $string
2966 * @param string $pattern
2969 protected static function insertSpace( $string, $pattern ) {
2970 $string = preg_replace( $pattern, " $1 ", $string );
2971 $string = preg_replace( '/ +/', ' ', $string );
2976 * @param string[] $termsArray
2979 function convertForSearchResult( $termsArray ) {
2980 # some languages, e.g. Chinese, need to do a conversion
2981 # in order for search results to be displayed correctly
2986 * Get the first character of a string.
2991 function firstChar( $s ) {
2994 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2995 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
3000 if ( isset( $matches[1] ) ) {
3001 if ( strlen( $matches[1] ) != 3 ) {
3005 // Break down Hangul syllables to grab the first jamo
3006 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3007 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3009 } elseif ( $code < 0xb098 ) {
3011 } elseif ( $code < 0xb2e4 ) {
3013 } elseif ( $code < 0xb77c ) {
3015 } elseif ( $code < 0xb9c8 ) {
3017 } elseif ( $code < 0xbc14 ) {
3019 } elseif ( $code < 0xc0ac ) {
3021 } elseif ( $code < 0xc544 ) {
3023 } elseif ( $code < 0xc790 ) {
3025 } elseif ( $code < 0xcc28 ) {
3027 } elseif ( $code < 0xce74 ) {
3029 } elseif ( $code < 0xd0c0 ) {
3031 } elseif ( $code < 0xd30c ) {
3033 } elseif ( $code < 0xd558 ) {
3044 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3045 * also cleans up certain backwards-compatible sequences, converting them
3046 * to the modern Unicode equivalent.
3048 * This is language-specific for performance reasons only.
3054 public function normalize( $s ) {
3055 global $wgAllUnicodeFixes;
3056 $s = UtfNormal\Validator
::cleanUp( $s );
3057 if ( $wgAllUnicodeFixes ) {
3058 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
3059 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3066 * Transform a string using serialized data stored in the given file (which
3067 * must be in the serialized subdirectory of $IP). The file contains pairs
3068 * mapping source characters to destination characters.
3070 * The data is cached in process memory. This will go faster if you have the
3071 * FastStringSearch extension.
3073 * @param string $file
3074 * @param string $string
3076 * @throws MWException
3079 protected function transformUsingPairFile( $file, $string ) {
3080 if ( !isset( $this->transformData
[$file] ) ) {
3082 $data = require "$IP/languages/data/{$file}";
3083 $this->transformData
[$file] = new ReplacementArray( $data );
3085 return $this->transformData
[$file]->replace( $string );
3089 * For right-to-left language support
3094 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3098 * Return the correct HTML 'dir' attribute value for this language.
3102 return $this->isRTL() ?
'rtl' : 'ltr';
3106 * Return 'left' or 'right' as appropriate alignment for line-start
3107 * for this language's text direction.
3109 * Should be equivalent to CSS3 'start' text-align value....
3113 function alignStart() {
3114 return $this->isRTL() ?
'right' : 'left';
3118 * Return 'right' or 'left' as appropriate alignment for line-end
3119 * for this language's text direction.
3121 * Should be equivalent to CSS3 'end' text-align value....
3125 function alignEnd() {
3126 return $this->isRTL() ?
'left' : 'right';
3130 * A hidden direction mark (LRM or RLM), depending on the language direction.
3131 * Unlike getDirMark(), this function returns the character as an HTML entity.
3132 * This function should be used when the output is guaranteed to be HTML,
3133 * because it makes the output HTML source code more readable. When
3134 * the output is plain text or can be escaped, getDirMark() should be used.
3136 * @param bool $opposite Get the direction mark opposite to your language
3140 function getDirMarkEntity( $opposite = false ) {
3142 return $this->isRTL() ?
'‎' : '‏';
3144 return $this->isRTL() ?
'‏' : '‎';
3148 * A hidden direction mark (LRM or RLM), depending on the language direction.
3149 * This function produces them as invisible Unicode characters and
3150 * the output may be hard to read and debug, so it should only be used
3151 * when the output is plain text or can be escaped. When the output is
3152 * HTML, use getDirMarkEntity() instead.
3154 * @param bool $opposite Get the direction mark opposite to your language
3157 function getDirMark( $opposite = false ) {
3158 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3159 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3161 return $this->isRTL() ?
$lrm : $rlm;
3163 return $this->isRTL() ?
$rlm : $lrm;
3169 function capitalizeAllNouns() {
3170 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3174 * An arrow, depending on the language direction.
3176 * @param string $direction The direction of the arrow: forwards (default),
3177 * backwards, left, right, up, down.
3180 function getArrow( $direction = 'forwards' ) {
3181 switch ( $direction ) {
3183 return $this->isRTL() ?
'←' : '→';
3185 return $this->isRTL() ?
'→' : '←';
3198 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3202 function linkPrefixExtension() {
3203 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3207 * Get all magic words from cache.
3210 function getMagicWords() {
3211 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3215 * Fill a MagicWord object with data from here
3217 * @param MagicWord $mw
3219 function getMagic( $mw ) {
3220 $rawEntry = $this->mMagicExtensions
[$mw->mId
] ??
3221 self
::$dataCache->getSubitem( $this->mCode
, 'magicWords', $mw->mId
);
3223 if ( !is_array( $rawEntry ) ) {
3224 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3226 $mw->mCaseSensitive
= $rawEntry[0];
3227 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3232 * Add magic words to the extension array
3234 * @param array $newWords
3236 function addMagicWordsByLang( $newWords ) {
3237 $fallbackChain = $this->getFallbackLanguages();
3238 $fallbackChain = array_reverse( $fallbackChain );
3239 foreach ( $fallbackChain as $code ) {
3240 if ( isset( $newWords[$code] ) ) {
3241 // @phan-suppress-next-line PhanTypeMismatchProperty
3242 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3248 * Get special page names, as an associative array
3249 * canonical name => array of valid names, including aliases
3252 function getSpecialPageAliases() {
3253 // Cache aliases because it may be slow to load them
3254 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3256 $this->mExtendedSpecialPageAliases
=
3257 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3260 return $this->mExtendedSpecialPageAliases
;
3264 * Italic is unsuitable for some languages
3266 * @param string $text The text to be emphasized.
3269 function emphasize( $text ) {
3270 return "<em>$text</em>";
3274 * Normally we output all numbers in plain en_US style, that is
3275 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3276 * point twohundredthirtyfive. However this is not suitable for all
3277 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3278 * Icelandic just want to use commas instead of dots, and dots instead
3279 * of commas like "293.291,235".
3281 * An example of this function being called:
3283 * wfMessage( 'message' )->numParams( $num )->text()
3286 * See $separatorTransformTable on MessageIs.php for
3287 * the , => . and . => , implementation.
3289 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3290 * @param int|float $number The string to be formatted, should be an integer
3291 * or a floating point number.
3292 * @param bool $nocommafy Set to true for special numbers like dates
3295 public function formatNum( $number, $nocommafy = false ) {
3296 global $wgTranslateNumerals;
3297 if ( !$nocommafy ) {
3298 $number = $this->commafy( $number );
3299 $s = $this->separatorTransformTable();
3301 $number = strtr( $number, $s );
3305 if ( $wgTranslateNumerals ) {
3306 $s = $this->digitTransformTable();
3308 $number = strtr( $number, $s );
3312 return (string)$number;
3316 * Front-end for non-commafied formatNum
3318 * @param int|float $number The string to be formatted, should be an integer
3319 * or a floating point number.
3323 public function formatNumNoSeparators( $number ) {
3324 return $this->formatNum( $number, true );
3328 * @param string $number
3331 public function parseFormattedNumber( $number ) {
3332 $s = $this->digitTransformTable();
3334 // eliminate empty array values such as ''. (T66347)
3335 $s = array_filter( $s );
3336 $number = strtr( $number, array_flip( $s ) );
3339 $s = $this->separatorTransformTable();
3341 // eliminate empty array values such as ''. (T66347)
3342 $s = array_filter( $s );
3343 $number = strtr( $number, array_flip( $s ) );
3346 $number = strtr( $number, [ ',' => '' ] );
3351 * Adds commas to a given number
3353 * @param mixed $number
3356 function commafy( $number ) {
3357 $digitGroupingPattern = $this->digitGroupingPattern();
3358 $minimumGroupingDigits = $this->minimumGroupingDigits();
3359 if ( $number === null ) {
3363 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3364 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3365 // In some languages it's conventional not to insert a thousands separator
3366 // in numbers that are four digits long (1000-9999).
3367 if ( $minimumGroupingDigits ) {
3368 // Number of '#' characters after last comma in the grouping pattern.
3369 // The pattern is hardcoded here, but this would vary for different patterns.
3370 $primaryGroupingSize = 3;
3371 // Maximum length of a number to suppress digit grouping for.
3372 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3373 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3377 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3379 // Ref: http://cldr.unicode.org/translation/number-patterns
3381 if ( intval( $number ) < 0 ) {
3382 // For negative numbers apply the algorithm like positive number and add sign.
3384 $number = substr( $number, 1 );
3388 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3389 preg_match( "/\d+/", $number, $integerPart );
3390 preg_match( "/\.\d*/", $number, $decimalPart );
3391 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3392 if ( $groupedNumber === $number ) {
3393 // the string does not have any number part. Eg: .12345
3394 return $sign . $groupedNumber;
3396 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3397 while ( $start > 0 ) {
3398 $match = $matches[0][$numMatches - 1];
3399 $matchLen = strlen( $match );
3400 $start = $end - $matchLen;
3404 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3406 if ( $numMatches > 1 ) {
3407 // use the last pattern for the rest of the number
3411 $groupedNumber = "," . $groupedNumber;
3414 return $sign . $groupedNumber;
3421 function digitGroupingPattern() {
3422 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3428 function digitTransformTable() {
3429 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3435 function separatorTransformTable() {
3436 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3442 function minimumGroupingDigits() {
3443 return self
::$dataCache->getItem( $this->mCode
, 'minimumGroupingDigits' );
3447 * Take a list of strings and build a locale-friendly comma-separated
3448 * list, using the local comma-separator message.
3449 * The last two strings are chained with an "and".
3451 * @param string[] $list
3454 public function listToText( array $list ) {
3455 $itemCount = count( $list );
3456 if ( $itemCount < 1 ) {
3459 $text = array_pop( $list );
3460 if ( $itemCount > 1 ) {
3461 $and = $this->msg( 'and' )->escaped();
3462 $space = $this->msg( 'word-separator' )->escaped();
3464 if ( $itemCount > 2 ) {
3465 $comma = $this->msg( 'comma-separator' )->escaped();
3467 $text = implode( $comma, $list ) . $and . $space . $text;
3473 * Take a list of strings and build a locale-friendly comma-separated
3474 * list, using the local comma-separator message.
3475 * @param string[] $list Array of strings to put in a comma list
3478 function commaList( array $list ) {
3480 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3486 * Take a list of strings and build a locale-friendly semicolon-separated
3487 * list, using the local semicolon-separator message.
3488 * @param string[] $list Array of strings to put in a semicolon list
3491 function semicolonList( array $list ) {
3493 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3499 * Same as commaList, but separate it with the pipe instead.
3500 * @param string[] $list Array of strings to put in a pipe list
3503 function pipeList( array $list ) {
3505 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3511 * Truncate a string to a specified length in bytes, appending an optional
3512 * string (e.g. for ellipsis)
3514 * If $length is negative, the string will be truncated from the beginning
3518 * @param string $string String to truncate
3519 * @param int $length Maximum length in bytes
3520 * @param string $ellipsis String to append to the end of truncated text
3521 * @param bool $adjustLength Subtract length of ellipsis from $length
3525 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3526 return $this->truncateInternal(
3527 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3532 * Truncate a string to a specified number of characters, appending an optional
3533 * string (e.g. for ellipsis).
3535 * This provides multibyte version of truncateForDatabase() method of this class,
3536 * suitable for truncation based on number of characters, instead of number of bytes.
3538 * If $length is negative, the string will be truncated from the beginning.
3542 * @param string $string String to truncate
3543 * @param int $length Maximum number of characters
3544 * @param string $ellipsis String to append to the end of truncated text
3545 * @param bool $adjustLength Subtract length of ellipsis from $length
3549 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3550 // Passing encoding to mb_strlen and mb_substr is optional.
3551 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3552 // explicit specification of encoding is skipped.
3553 // Note: Both multibyte methods are callables invoked in truncateInternal.
3554 return $this->truncateInternal(
3555 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3560 * Internal method used for truncation. This method abstracts text truncation into
3561 * one common method, allowing users to provide length measurement function and
3562 * function for finding substring.
3564 * For usages, see truncateForDatabase and truncateForVisual.
3566 * @param string $string String to truncate
3567 * @param int $length Maximum length of final text
3568 * @param string $ellipsis String to append to the end of truncated text
3569 * @param bool $adjustLength Subtract length of ellipsis from $length
3570 * @param callable $measureLength Callable function used for determining the length of text
3571 * @param callable $getSubstring Callable function used for getting the substrings
3575 private function truncateInternal(
3576 $string, $length, $ellipsis, $adjustLength, callable
$measureLength, callable
$getSubstring
3578 # Check if there is no need to truncate
3579 if ( $measureLength( $string ) <= abs( $length ) ) {
3580 return $string; // no need to truncate
3583 # Use the localized ellipsis character
3584 if ( $ellipsis == '...' ) {
3585 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3587 if ( $length == 0 ) {
3588 return $ellipsis; // convention
3591 $stringOriginal = $string;
3592 # If ellipsis length is >= $length then we can't apply $adjustLength
3593 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3594 $string = $ellipsis; // this can be slightly unexpected
3595 # Otherwise, truncate and add ellipsis...
3597 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3598 if ( $length > 0 ) {
3599 $length -= $ellipsisLength;
3600 $string = $getSubstring( $string, 0, $length ); // xyz...
3601 $string = $this->removeBadCharLast( $string );
3602 $string = rtrim( $string ) . $ellipsis;
3604 $length +
= $ellipsisLength;
3605 $string = $getSubstring( $string, $length ); // ...xyz
3606 $string = $this->removeBadCharFirst( $string );
3607 $string = $ellipsis . ltrim( $string );
3611 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3612 # This check is *not* redundant if $adjustLength, due to the single case where
3613 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3614 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3617 return $stringOriginal;
3622 * Remove bytes that represent an incomplete Unicode character
3623 * at the end of string (e.g. bytes of the char are missing)
3625 * @param string $string
3628 protected function removeBadCharLast( $string ) {
3629 if ( $string != '' ) {
3630 $char = ord( $string[strlen( $string ) - 1] );
3632 if ( $char >= 0xc0 ) {
3633 # We got the first byte only of a multibyte char; remove it.
3634 $string = substr( $string, 0, -1 );
3635 } elseif ( $char >= 0x80 &&
3636 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3637 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3638 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3640 # We chopped in the middle of a character; remove it
3648 * Remove bytes that represent an incomplete Unicode character
3649 * at the start of string (e.g. bytes of the char are missing)
3651 * @param string $string
3654 protected function removeBadCharFirst( $string ) {
3655 if ( $string != '' ) {
3656 $char = ord( $string[0] );
3657 if ( $char >= 0x80 && $char < 0xc0 ) {
3658 # We chopped in the middle of a character; remove the whole thing
3659 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3666 * Truncate a string of valid HTML to a specified length in bytes,
3667 * appending an optional string (e.g. for ellipses), and return valid HTML
3669 * This is only intended for styled/linked text, such as HTML with
3670 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3671 * Also, this will not detect things like "display:none" CSS.
3673 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3675 * @param string $text HTML string to truncate
3676 * @param int $length (zero/positive) Maximum length (including ellipses)
3677 * @param string $ellipsis String to append to the truncated text
3680 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3681 # Use the localized ellipsis character
3682 if ( $ellipsis == '...' ) {
3683 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3685 # Check if there is clearly no need to truncate
3686 if ( $length <= 0 ) {
3687 return $ellipsis; // no text shown, nothing to format (convention)
3688 } elseif ( strlen( $text ) <= $length ) {
3689 return $text; // string short enough even *with* HTML (short-circuit)
3692 $dispLen = 0; // innerHTML legth so far
3693 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3694 $tagType = 0; // 0-open, 1-close
3695 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3696 $entityState = 0; // 0-not entity, 1-entity
3697 $tag = $ret = ''; // accumulated tag name, accumulated result string
3698 $openTags = []; // open tag stack
3699 $maybeState = null; // possible truncation state
3701 $textLen = strlen( $text );
3702 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3703 for ( $pos = 0; true; ++
$pos ) {
3704 # Consider truncation once the display length has reached the maximim.
3705 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3706 # Check that we're not in the middle of a bracket/entity...
3707 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3708 if ( !$testingEllipsis ) {
3709 $testingEllipsis = true;
3710 # Save where we are; we will truncate here unless there turn out to
3711 # be so few remaining characters that truncation is not necessary.
3712 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3713 $maybeState = [ $ret, $openTags ]; // save state
3715 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3716 # String in fact does need truncation, the truncation point was OK.
3717 // @phan-suppress-next-line PhanTypeInvalidExpressionArrayDestructuring
3718 list( $ret, $openTags ) = $maybeState; // reload state
3719 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3720 $ret .= $ellipsis; // add ellipsis
3724 if ( $pos >= $textLen ) {
3725 break; // extra iteration just for above checks
3728 # Read the next char...
3730 $lastCh = $pos ?
$text[$pos - 1] : '';
3731 $ret .= $ch; // add to result string
3733 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3734 $entityState = 0; // for bad HTML
3735 $bracketState = 1; // tag started (checking for backslash)
3736 } elseif ( $ch == '>' ) {
3737 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3738 $entityState = 0; // for bad HTML
3739 $bracketState = 0; // out of brackets
3740 } elseif ( $bracketState == 1 ) {
3742 $tagType = 1; // close tag (e.g. "</span>")
3744 $tagType = 0; // open tag (e.g. "<span>")
3747 $bracketState = 2; // building tag name
3748 } elseif ( $bracketState == 2 ) {
3752 // Name found (e.g. "<a href=..."), add on tag attributes...
3753 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3755 } elseif ( $bracketState == 0 ) {
3756 if ( $entityState ) {
3759 $dispLen++
; // entity is one displayed char
3762 if ( $neLength == 0 && !$maybeState ) {
3763 // Save state without $ch. We want to *hit* the first
3764 // display char (to get tags) but not *use* it if truncating.
3765 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3768 $entityState = 1; // entity found, (e.g. " ")
3770 $dispLen++
; // this char is displayed
3771 // Add the next $max display text chars after this in one swoop...
3772 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3773 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3774 $dispLen +
= $skipped;
3780 // Close the last tag if left unclosed by bad HTML
3781 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3782 while ( count( $openTags ) > 0 ) {
3783 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3789 * truncateHtml() helper function
3790 * like strcspn() but adds the skipped chars to $ret
3792 * @param string $ret
3793 * @param string $text
3794 * @param string $search
3796 * @param null|int $len
3799 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3800 if ( $len === null ) {
3801 $len = -1; // -1 means "no limit" for strcspn
3802 } elseif ( $len < 0 ) {
3806 if ( $start < strlen( $text ) ) {
3807 $skipCount = strcspn( $text, $search, $start, $len );
3808 $ret .= substr( $text, $start, $skipCount );
3814 * truncateHtml() helper function
3815 * (a) push or pop $tag from $openTags as needed
3816 * (b) clear $tag value
3817 * @param string &$tag Current HTML tag name we are looking at
3818 * @param int $tagType (0-open tag, 1-close tag)
3819 * @param string $lastCh Character before the '>' that ended this tag
3820 * @param array &$openTags Open tag stack (not accounting for $tag)
3822 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3823 $tag = ltrim( $tag );
3825 if ( $tagType == 0 && $lastCh != '/' ) {
3826 $openTags[] = $tag; // tag opened (didn't close itself)
3827 } elseif ( $tagType == 1 ) {
3828 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3829 array_pop( $openTags ); // tag closed
3837 * Grammatical transformations, needed for inflected languages
3838 * Invoked by putting {{grammar:case|word}} in a message
3840 * @param string $word
3841 * @param string $case
3844 function convertGrammar( $word, $case ) {
3845 global $wgGrammarForms;
3846 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3847 return $wgGrammarForms[$this->getCode()][$case][$word];
3850 $grammarTransformations = $this->getGrammarTransformations();
3852 if ( isset( $grammarTransformations[$case] ) ) {
3853 $forms = $grammarTransformations[$case];
3855 // Some names of grammar rules are aliases for other rules.
3856 // In such cases the value is a string rather than object,
3857 // so load the actual rules.
3858 if ( is_string( $forms ) ) {
3859 $forms = $grammarTransformations[$forms];
3862 foreach ( array_values( $forms ) as $rule ) {
3865 if ( $form === '@metadata' ) {
3869 $replacement = $rule[1];
3871 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3872 $patternMatches = preg_match( $regex, $word );
3874 if ( $patternMatches === false ) {
3876 'An error occurred while processing grammar. ' .
3877 "Word: '$word'. Regex: /$form/."
3879 } elseif ( $patternMatches === 1 ) {
3880 $word = preg_replace( $regex, $replacement, $word );
3891 * Get the grammar forms for the content language
3892 * @return array Array of grammar forms
3895 function getGrammarForms() {
3896 global $wgGrammarForms;
3897 if ( isset( $wgGrammarForms[$this->getCode()] )
3898 && is_array( $wgGrammarForms[$this->getCode()] )
3900 return $wgGrammarForms[$this->getCode()];
3907 * Get the grammar transformations data for the language.
3908 * Used like grammar forms, with {{GRAMMAR}} and cases,
3909 * but uses pairs of regexes and replacements instead of code.
3911 * @return array[] Array of grammar transformations.
3912 * @throws MWException
3915 public function getGrammarTransformations() {
3916 $languageCode = $this->getCode();
3918 if ( self
::$grammarTransformations === null ) {
3919 self
::$grammarTransformations = new MapCacheLRU( 10 );
3922 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3923 return self
::$grammarTransformations->get( $languageCode );
3928 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3929 if ( is_readable( $grammarDataFile ) ) {
3930 $data = FormatJson
::decode(
3931 file_get_contents( $grammarDataFile ),
3935 if ( $data === null ) {
3936 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3939 self
::$grammarTransformations->set( $languageCode, $data );
3946 * Provides an alternative text depending on specified gender.
3947 * Usage {{gender:username|masculine|feminine|unknown}}.
3948 * username is optional, in which case the gender of current user is used,
3949 * but only in (some) interface messages; otherwise default gender is used.
3951 * If no forms are given, an empty string is returned. If only one form is
3952 * given, it will be returned unconditionally. These details are implied by
3953 * the caller and cannot be overridden in subclasses.
3955 * If three forms are given, the default is to use the third (unknown) form.
3956 * If fewer than three forms are given, the default is to use the first (masculine) form.
3957 * These details can be overridden in subclasses.
3959 * @param string $gender
3960 * @param array $forms
3964 function gender( $gender, $forms ) {
3965 if ( !count( $forms ) ) {
3968 $forms = $this->preConvertPlural( $forms, 2 );
3969 if ( $gender === 'male' ) {
3972 if ( $gender === 'female' ) {
3975 return $forms[2] ??
$forms[0];
3979 * Plural form transformations, needed for some languages.
3980 * For example, there are 3 form of plural in Russian and Polish,
3981 * depending on "count mod 10". See [[w:Plural]]
3982 * For English it is pretty simple.
3984 * Invoked by putting {{plural:count|wordform1|wordform2}}
3985 * or {{plural:count|wordform1|wordform2|wordform3}}
3987 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3989 * @param int $count Non-localized number
3990 * @param array $forms Different plural forms
3991 * @return string Correct form of plural for $count in this language
3993 function convertPlural( $count, $forms ) {
3994 // Handle explicit n=pluralform cases
3995 $forms = $this->handleExplicitPluralForms( $count, $forms );
3996 if ( is_string( $forms ) ) {
3999 if ( !count( $forms ) ) {
4003 $pluralForm = $this->getPluralRuleIndexNumber( $count );
4004 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
4005 return $forms[$pluralForm];
4009 * Handles explicit plural forms for Language::convertPlural()
4011 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4012 * If an explicitly defined plural form matches the $count, then
4013 * string value returned, otherwise array returned for further consideration
4014 * by CLDR rules or overridden convertPlural().
4018 * @param int $count Non-localized number
4019 * @param array $forms Different plural forms
4021 * @return array|string
4023 protected function handleExplicitPluralForms( $count, array $forms ) {
4024 foreach ( $forms as $index => $form ) {
4025 if ( preg_match( '/\d+=/i', $form ) ) {
4026 $pos = strpos( $form, '=' );
4027 if ( substr( $form, 0, $pos ) === (string)$count ) {
4028 return substr( $form, $pos +
1 );
4030 unset( $forms[$index] );
4033 return array_values( $forms );
4037 * Checks that convertPlural was given an array and pads it to requested
4038 * amount of forms by copying the last one.
4040 * @param array $forms
4041 * @param int $count How many forms should there be at least
4042 * @return array Padded array of forms
4044 protected function preConvertPlural( /* Array */ $forms, $count ) {
4045 return array_pad( $forms, $count, end( $forms ) );
4049 * Wraps argument with unicode control characters for directionality safety
4051 * This solves the problem where directionality-neutral characters at the edge of
4052 * the argument string get interpreted with the wrong directionality from the
4053 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4055 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4056 * directionality of the argument string, using the BIDI algorithm's own "First
4057 * strong directional codepoint" rule. Essentially, this works round the fact that
4058 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4059 * direction inference). The latter is cleaner but still not widely supported.
4061 * @param string $text Text to wrap
4062 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4064 public function embedBidi( $text = '' ) {
4065 $dir = self
::strongDirFromContent( $text );
4066 if ( $dir === 'ltr' ) {
4067 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4068 return self
::$lre . $text . self
::$pdf;
4070 if ( $dir === 'rtl' ) {
4071 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4072 return self
::$rle . $text . self
::$pdf;
4074 // No strong directionality: do not wrap
4079 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4080 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4081 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4082 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4085 * @param string $str The validated block duration in English
4086 * @param User|null $user User object to use timezone from or null for $wgUser
4087 * @param int $now Current timestamp, for formatting relative block durations
4088 * @return string Somehow translated block duration
4089 * @see LanguageFi.php for example implementation
4091 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
4092 $duration = SpecialBlock
::getSuggestedDurations( $this );
4093 foreach ( $duration as $show => $value ) {
4094 if ( strcmp( $str, $value ) == 0 ) {
4095 return htmlspecialchars( trim( $show ) );
4099 if ( wfIsInfinity( $str ) ) {
4100 foreach ( $duration as $show => $value ) {
4101 if ( wfIsInfinity( $value ) ) {
4102 return htmlspecialchars( trim( $show ) );
4107 // If all else fails, return a standard duration or timestamp description.
4108 $time = strtotime( $str, $now );
4109 if ( $time === false ) { // Unknown format. Return it as-is in case.
4111 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
4112 // The result differs based on current time, so the difference
4113 // is a fixed duration length.
4114 return $this->formatDuration( $time - $now );
4115 } else { // It's an absolute timestamp.
4116 if ( $time === 0 ) {
4117 // wfTimestamp() handles 0 as current time instead of epoch.
4118 $time = '19700101000000';
4121 return $this->userTimeAndDate( $time, $user );
4123 return $this->timeanddate( $time );
4128 * languages like Chinese need to be segmented in order for the diff
4131 * @param string $text
4134 public function segmentForDiff( $text ) {
4139 * and unsegment to show the result
4141 * @param string $text
4144 public function unsegmentForDiff( $text ) {
4149 * Return the LanguageConverter used in the Language
4152 * @return LanguageConverter
4154 public function getConverter() {
4155 return $this->mConverter
;
4159 * convert text to a variant
4161 * @param string $text text to convert
4162 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4163 * variant (if logged in), or the project default variant
4164 * @return string the converted string
4166 public function autoConvert( $text, $variant = false ) {
4167 return $this->mConverter
->autoConvert( $text, $variant );
4171 * convert text to all supported variants
4173 * @param string $text
4176 public function autoConvertToAllVariants( $text ) {
4177 return $this->mConverter
->autoConvertToAllVariants( $text );
4181 * convert text to different variants of a language.
4183 * @warning Glossary state is maintained between calls. This means
4184 * if you pass unescaped text to this method it can cause an XSS
4185 * in later calls to this method, even if the later calls have properly
4186 * escaped the input. Never feed this method user controlled text that
4187 * is not properly escaped!
4188 * @param string $text Content that has been already escaped for use in HTML
4189 * @return string HTML
4191 public function convert( $text ) {
4192 return $this->mConverter
->convert( $text );
4196 * Convert a Title object to a string in the preferred variant
4198 * @param Title $title
4201 public function convertTitle( $title ) {
4202 return $this->mConverter
->convertTitle( $title );
4206 * Convert a namespace index to a string in the preferred variant
4208 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4209 * @param string|null $variant variant to convert to, or null to use the user's preferred
4210 * variant (if logged in), or the project default variant
4211 * @return string a string representation of the namespace
4213 public function convertNamespace( $ns, $variant = null ) {
4214 return $this->mConverter
->convertNamespace( $ns, $variant );
4218 * Check if this is a language with variants
4222 public function hasVariants() {
4223 return count( $this->getVariants() ) > 1;
4227 * Strict check if the language has the specific variant.
4229 * Compare to LanguageConverter::validateVariant() which does a more
4230 * lenient check and attempts to coerce the given code to a valid one.
4233 * @param string $variant
4236 public function hasVariant( $variant ) {
4237 return $variant && ( $variant === $this->mConverter
->validateVariant( $variant ) );
4241 * Perform output conversion on a string, and encode for safe HTML output.
4242 * @param string $text Text to be converted
4244 * @todo this should get integrated somewhere sane
4246 public function convertHtml( $text ) {
4247 return htmlspecialchars( $this->convert( $text ) );
4251 * @param string $key
4254 public function convertCategoryKey( $key ) {
4255 return $this->mConverter
->convertCategoryKey( $key );
4259 * Get the list of variants supported by this language
4260 * see sample implementation in LanguageZh.php
4262 * @return string[] An array of language codes
4264 public function getVariants() {
4265 return $this->mConverter
->getVariants();
4271 public function getPreferredVariant() {
4272 return $this->mConverter
->getPreferredVariant();
4278 public function getDefaultVariant() {
4279 return $this->mConverter
->getDefaultVariant();
4285 public function getURLVariant() {
4286 return $this->mConverter
->getURLVariant();
4290 * If a language supports multiple variants, it is
4291 * possible that non-existing link in one variant
4292 * actually exists in another variant. this function
4293 * tries to find it. See e.g. LanguageZh.php
4294 * The input parameters may be modified upon return
4296 * @param string &$link The name of the link
4297 * @param Title &$nt The title object of the link
4298 * @param bool $ignoreOtherCond To disable other conditions when
4299 * we need to transclude a template or update a category's link
4301 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4302 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4306 * returns language specific options used by User::getPageRenderHash()
4307 * for example, the preferred language variant
4311 function getExtraHashOptions() {
4312 return $this->mConverter
->getExtraHashOptions();
4316 * For languages that support multiple variants, the title of an
4317 * article may be displayed differently in different variants. this
4318 * function returns the apporiate title defined in the body of the article.
4322 public function getParsedTitle() {
4323 return $this->mConverter
->getParsedTitle();
4327 * Refresh the cache of conversion tables when
4328 * MediaWiki:Conversiontable* is updated.
4330 * @param Title $title The Title of the page being updated
4332 public function updateConversionTable( Title
$title ) {
4333 $this->mConverter
->updateConversionTable( $title );
4337 * A regular expression to match legal word-trailing characters
4338 * which should be merged onto a link of the form [[foo]]bar.
4342 public function linkTrail() {
4343 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4347 * A regular expression character set to match legal word-prefixing
4348 * characters which should be merged onto a link of the form foo[[bar]].
4352 public function linkPrefixCharset() {
4353 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4357 * Get the "parent" language which has a converter to convert a "compatible" language
4358 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4360 * @return Language|null
4363 public function getParentLanguage() {
4364 if ( $this->mParentLanguage
!== false ) {
4365 return $this->mParentLanguage
;
4368 $code = explode( '-', $this->getCode() )[0];
4369 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4370 $this->mParentLanguage
= null;
4373 $lang = self
::factory( $code );
4374 if ( !$lang->hasVariant( $this->getCode() ) ) {
4375 $this->mParentLanguage
= null;
4379 $this->mParentLanguage
= $lang;
4384 * Compare with an other language object
4387 * @param Language $lang
4390 public function equals( Language
$lang ) {
4391 return $lang === $this ||
$lang->getCode() === $this->mCode
;
4395 * Get the internal language code for this language object
4397 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4398 * htmlspecialchars() or similar
4402 public function getCode() {
4403 return $this->mCode
;
4407 * Get the code in BCP 47 format which we can use
4408 * inside of html lang="" tags.
4410 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4411 * htmlspecialchars() or similar.
4416 public function getHtmlCode() {
4417 if ( is_null( $this->mHtmlCode
) ) {
4418 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4420 return $this->mHtmlCode
;
4424 * Get the language code from a file name. Inverse of getFileName()
4425 * @param string $filename $prefix . $languageCode . $suffix
4426 * @param string $prefix Prefix before the language code
4427 * @param string $suffix Suffix after the language code
4428 * @return string Language code, or false if $prefix or $suffix isn't found
4430 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4432 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4433 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4434 if ( !count( $m ) ) {
4437 return str_replace( '_', '-', strtolower( $m[1] ) );
4441 * @param string $code
4442 * @param bool $fallback Whether we're going through language fallback chain
4443 * @return string Name of the language class
4445 public static function classFromCode( $code, $fallback = true ) {
4446 if ( $fallback && $code == 'en' ) {
4449 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4454 * Get the name of a file for a certain language code
4455 * @param string $prefix Prepend this to the filename
4456 * @param string $code Language code
4457 * @param string $suffix Append this to the filename
4458 * @throws MWException
4459 * @return string $prefix . $mangledCode . $suffix
4461 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4462 if ( !self
::isValidBuiltInCode( $code ) ) {
4463 throw new MWException( "Invalid language code \"$code\"" );
4466 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4470 * @param string $code
4473 public static function getMessagesFileName( $code ) {
4475 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4476 Hooks
::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4481 * @param string $code
4483 * @throws MWException
4486 public static function getJsonMessagesFileName( $code ) {
4489 if ( !self
::isValidBuiltInCode( $code ) ) {
4490 throw new MWException( "Invalid language code \"$code\"" );
4493 return "$IP/languages/i18n/$code.json";
4497 * Get the first fallback for a given language.
4499 * @param string $code
4501 * @return bool|string
4503 public static function getFallbackFor( $code ) {
4504 $fallbacks = self
::getFallbacksFor( $code );
4506 return $fallbacks[0];
4512 * Get the ordered list of fallback languages.
4515 * @param string $code Language code
4516 * @param int $mode Fallback mode, either MESSAGES_FALLBACKS (which always falls back to 'en'),
4517 * or STRICT_FALLBACKS (whic honly falls back to 'en' when explicitly defined)
4518 * @throws MWException
4519 * @return array List of language codes
4521 public static function getFallbacksFor( $code, $mode = self
::MESSAGES_FALLBACKS
) {
4522 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4526 case self
::MESSAGES_FALLBACKS
:
4527 // For unknown languages, fallbackSequence returns an empty array,
4528 // hardcode fallback to 'en' in that case as English messages are
4530 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4531 case self
::STRICT_FALLBACKS
:
4532 // Use this mode when you don't want to fallback to English unless
4533 // explicitly defined, for example when you have language-variant icons
4534 // and an international language-independent fallback.
4535 return self
::getLocalisationCache()->getItem( $code, 'originalFallbackSequence' );
4537 throw new MWException( "Invalid fallback mode \"$mode\"" );
4542 * Get the ordered list of fallback languages, ending with the fallback
4543 * language chain for the site language.
4546 * @param string $code Language code
4547 * @return array [ fallbacks, site fallbacks ]
4549 public static function getFallbacksIncludingSiteLanguage( $code ) {
4550 global $wgLanguageCode;
4552 // Usually, we will only store a tiny number of fallback chains, so we
4553 // keep them in static memory.
4554 $cacheKey = "{$code}-{$wgLanguageCode}";
4556 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4557 $fallbacks = self
::getFallbacksFor( $code );
4559 // Append the site's fallback chain, including the site language itself
4560 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4561 array_unshift( $siteFallbacks, $wgLanguageCode );
4563 // Eliminate any languages already included in the chain
4564 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4566 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4568 return self
::$fallbackLanguageCache[$cacheKey];
4572 * Get all messages for a given language
4573 * WARNING: this may take a long time. If you just need all message *keys*
4574 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4576 * @param string $code
4580 public static function getMessagesFor( $code ) {
4581 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4585 * Get a message for a given language
4587 * @param string $key
4588 * @param string $code
4592 public static function getMessageFor( $key, $code ) {
4593 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4597 * Get all message keys for a given language. This is a faster alternative to
4598 * array_keys( Language::getMessagesFor( $code ) )
4601 * @param string $code Language code
4602 * @return array Array of message keys (strings)
4604 public static function getMessageKeysFor( $code ) {
4605 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4609 * @param string $talk
4612 function fixVariableInNamespace( $talk ) {
4613 if ( strpos( $talk, '$1' ) === false ) {
4617 global $wgMetaNamespace;
4618 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4620 # Allow grammar transformations
4621 # Allowing full message-style parsing would make simple requests
4622 # such as action=raw much more expensive than they need to be.
4623 # This will hopefully cover most cases.
4624 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4625 [ $this, 'replaceGrammarInNamespace' ], $talk );
4626 return str_replace( ' ', '_', $talk );
4633 function replaceGrammarInNamespace( $m ) {
4634 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4638 * Decode an expiry (block, protection, etc) which has come from the DB
4640 * @param string $expiry Database expiry String
4641 * @param bool|int $format True to process using language functions, or TS_ constant
4642 * to return the expiry in a given timestamp
4643 * @param string $infinity If $format is not true, use this string for infinite expiry
4647 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4649 if ( $dbInfinity === null ) {
4650 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4653 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4654 return $format === true
4655 ?
$this->getMessageFromDB( 'infiniteblock' )
4658 return $format === true
4659 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4660 : wfTimestamp( $format, $expiry );
4665 * Formats a time given in seconds into a string representation of that time.
4667 * @param int|float $seconds
4668 * @param array $format An optional argument that formats the returned string in different ways:
4669 * If $format['avoid'] === 'avoidhours': don't show hours, just show days
4670 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4671 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4672 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4674 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4675 * or 'avoidminutes'.
4678 function formatTimePeriod( $seconds, $format = [] ) {
4679 if ( !is_array( $format ) ) {
4680 $format = [ 'avoid' => $format ]; // For backwards compatibility
4682 if ( !isset( $format['avoid'] ) ) {
4683 $format['avoid'] = false;
4685 if ( !isset( $format['noabbrevs'] ) ) {
4686 $format['noabbrevs'] = false;
4688 $secondsMsg = wfMessage(
4689 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4690 $minutesMsg = wfMessage(
4691 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4692 $hoursMsg = wfMessage(
4693 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4694 $daysMsg = wfMessage(
4695 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4697 if ( round( $seconds * 10 ) < 100 ) {
4698 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4699 $s = $secondsMsg->params( $s )->text();
4700 } elseif ( round( $seconds ) < 60 ) {
4701 $s = $this->formatNum( round( $seconds ) );
4702 $s = $secondsMsg->params( $s )->text();
4703 } elseif ( round( $seconds ) < 3600 ) {
4704 $minutes = floor( $seconds / 60 );
4705 $secondsPart = round( fmod( $seconds, 60 ) );
4706 if ( $secondsPart == 60 ) {
4710 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4712 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4713 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4714 $hours = floor( $seconds / 3600 );
4715 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4716 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4717 if ( $secondsPart == 60 ) {
4721 if ( $minutes == 60 ) {
4725 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4727 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4728 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes', 'avoidhours' ] ) ) {
4729 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4732 $days = floor( $seconds / 86400 );
4733 if ( $format['avoid'] === 'avoidhours' ) {
4734 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4735 if ( $hours == 24 ) {
4739 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4740 } elseif ( $format['avoid'] === 'avoidminutes' ) {
4741 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4742 if ( $hours == 24 ) {
4746 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4748 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4749 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4750 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4751 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4752 if ( $minutes == 60 ) {
4756 if ( $hours == 24 ) {
4760 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4762 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4764 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4766 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4768 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4775 * Format a bitrate for output, using an appropriate
4776 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4777 * the magnitude in question.
4779 * This use base 1000. For base 1024 use formatSize(), for another base
4780 * see formatComputingNumbers().
4785 function formatBitrate( $bps ) {
4786 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4790 * @param int $size Size of the unit
4791 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4792 * @param string $messageKey Message key to be uesd
4795 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4797 return str_replace( '$1', $this->formatNum( $size ),
4798 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4801 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4804 $maxIndex = count( $sizes ) - 1;
4805 while ( $size >= $boundary && $index < $maxIndex ) {
4810 // For small sizes no decimal places necessary
4813 // For MB and bigger two decimal places are smarter
4816 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4818 $size = round( $size, $round );
4819 $text = $this->getMessageFromDB( $msg );
4820 return str_replace( '$1', $this->formatNum( $size ), $text );
4824 * Format a size in bytes for output, using an appropriate
4825 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4827 * This method use base 1024. For base 1000 use formatBitrate(), for
4828 * another base see formatComputingNumbers()
4830 * @param int $size Size to format
4831 * @return string Plain text (not HTML)
4833 function formatSize( $size ) {
4834 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4838 * Make a list item, used by various special pages
4840 * @param string $page Page link
4841 * @param string $details HTML safe text between brackets
4842 * @param bool $oppositedm Add the direction mark opposite to your
4843 * language, to display text properly
4844 * @return string HTML escaped
4846 function specialList( $page, $details, $oppositedm = true ) {
4851 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4854 $this->msg( 'word-separator' )->escaped() .
4855 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4859 * Generate (prev x| next x) (20|50|100...) type links for paging
4861 * @param Title $title Title object to link
4862 * @param int $offset
4864 * @param array $query Optional URL query parameter string
4865 * @param bool $atend Optional param for specified if this is the last page
4867 * @deprecated since 1.34, use PrevNextNavigationRenderer::buildPrevNextNavigation()
4870 public function viewPrevNext( Title
$title, $offset, $limit,
4871 array $query = [], $atend = false
4873 wfDeprecated( __METHOD__
, '1.34' );
4874 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4876 # Make 'previous' link
4877 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4878 if ( $offset > 0 ) {
4879 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4880 $query, $prev, 'prevn-title', 'mw-prevlink' );
4882 $plink = htmlspecialchars( $prev );
4886 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4888 $nlink = htmlspecialchars( $next );
4890 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4891 $query, $next, 'nextn-title', 'mw-nextlink' );
4894 # Make links to set number of items per page
4896 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4897 $numLinks[] = $this->numLink( $title, $offset, $num,
4898 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4901 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4902 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4906 * Helper function for viewPrevNext() that generates links
4908 * @param Title $title Title object to link
4909 * @param int $offset
4911 * @param array $query Extra query parameters
4912 * @param string $link Text to use for the link; will be escaped
4913 * @param string $tooltipMsg Name of the message to use as tooltip
4914 * @param string $class Value of the "class" attribute of the link
4915 * @return string HTML fragment
4917 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4920 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4921 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4922 ->numParams( $limit )->text();
4924 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4925 'title' => $tooltip, 'class' => $class ], $link );
4929 * Get the conversion rule title, if any.
4933 public function getConvRuleTitle() {
4934 return $this->mConverter
->getConvRuleTitle();
4938 * Get the compiled plural rules for the language
4940 * @return array Associative array with plural form, and plural rule as key-value pairs
4942 public function getCompiledPluralRules() {
4943 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4944 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4945 if ( !$pluralRules ) {
4946 foreach ( $fallbacks as $fallbackCode ) {
4947 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4948 if ( $pluralRules ) {
4953 return $pluralRules;
4957 * Get the plural rules for the language
4959 * @return array Associative array with plural form number and plural rule as key-value pairs
4961 public function getPluralRules() {
4962 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4963 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4964 if ( !$pluralRules ) {
4965 foreach ( $fallbacks as $fallbackCode ) {
4966 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4967 if ( $pluralRules ) {
4972 return $pluralRules;
4976 * Get the plural rule types for the language
4978 * @return array Associative array with plural form number and plural rule type as key-value pairs
4980 public function getPluralRuleTypes() {
4981 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4982 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4983 if ( !$pluralRuleTypes ) {
4984 foreach ( $fallbacks as $fallbackCode ) {
4985 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4986 if ( $pluralRuleTypes ) {
4991 return $pluralRuleTypes;
4995 * Find the index number of the plural rule appropriate for the given number
4996 * @param int $number
4997 * @return int The index number of the plural rule
4999 public function getPluralRuleIndexNumber( $number ) {
5000 $pluralRules = $this->getCompiledPluralRules();
5001 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
5006 * Find the plural rule type appropriate for the given number
5007 * For example, if the language is set to Arabic, getPluralType(5) should
5010 * @param int $number
5011 * @return string The name of the plural rule type, e.g. one, two, few, many
5013 public function getPluralRuleType( $number ) {
5014 $index = $this->getPluralRuleIndexNumber( $number );
5015 $pluralRuleTypes = $this->getPluralRuleTypes();
5016 return $pluralRuleTypes[$index] ??
'other';