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\Languages\LanguageNameUtils
;
31 use MediaWiki\MediaWikiServices
;
34 * Internationalisation code
39 * Return autonyms in fetchLanguageName(s).
41 * @deprecated since 1.34, LanguageNameUtils::AUTONYMS
43 const AS_AUTONYMS
= LanguageNameUtils
::AUTONYMS
;
46 * Return all known languages in fetchLanguageName(s).
48 * @deprecated since 1.34, use LanguageNameUtils::ALL
50 const ALL
= LanguageNameUtils
::ALL
;
53 * Return in fetchLanguageName(s) only the languages for which we have at
54 * least some localisation.
56 * @deprecated since 1.34, use LanguageNameUtils::SUPPORTED
58 const SUPPORTED
= LanguageNameUtils
::SUPPORTED
;
61 * @var LanguageConverter
65 public $mVariants, $mCode, $mLoaded = false;
66 public $mMagicExtensions = [];
67 private $mHtmlCode = null, $mParentLanguage = false;
69 public $dateFormatStrings = [];
70 public $mExtendedSpecialPageAliases;
72 /** @var array|null */
73 protected $namespaceNames;
74 protected $mNamespaceIds, $namespaceAliases;
77 * ReplacementArray object caches
79 public $transformData = [];
81 /** @var LocalisationCache */
82 private $localisationCache;
84 /** @var LanguageNameUtils */
85 private $langNameUtils;
87 public static $mLangObjCache = [];
90 * Return a fallback chain for messages in getFallbacksFor
93 const MESSAGES_FALLBACKS
= 0;
96 * Return a strict fallback chain in getFallbacksFor
99 const STRICT_FALLBACKS
= 1;
101 // TODO Make these const once we drop HHVM support (T192166)
102 public static $mWeekdayMsgs = [
103 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
107 public static $mWeekdayAbbrevMsgs = [
108 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
111 public static $mMonthMsgs = [
112 'january', 'february', 'march', 'april', 'may_long', 'june',
113 'july', 'august', 'september', 'october', 'november',
116 public static $mMonthGenMsgs = [
117 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
118 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
121 public static $mMonthAbbrevMsgs = [
122 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
123 'sep', 'oct', 'nov', 'dec'
126 public static $mIranianCalendarMonthMsgs = [
127 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
128 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
129 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
130 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
133 public static $mHebrewCalendarMonthMsgs = [
134 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
135 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
136 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
137 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
138 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
141 public static $mHebrewCalendarMonthGenMsgs = [
142 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
143 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
144 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
145 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
146 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
149 public static $mHijriCalendarMonthMsgs = [
150 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
151 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
152 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
153 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
160 public static $durationIntervals = [
161 'millennia' => 31556952000,
162 'centuries' => 3155695200,
163 'decades' => 315569520,
164 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
173 * Cache for language fallbacks.
174 * @see Language::getFallbacksIncludingSiteLanguage
178 private static $fallbackLanguageCache = [];
181 * Cache for grammar rules data
182 * @var MapCacheLRU|null
184 private static $grammarTransformations;
187 * Unicode directional formatting characters, for embedBidi()
189 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
190 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
191 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
194 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
195 * - in group 1 if it is LTR
196 * - in group 2 if it is RTL
197 * Does not match if there is no strong directionality codepoint.
199 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
201 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
202 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
204 // @codeCoverageIgnoreStart
205 // phpcs:ignore Generic.Files.LineLength
206 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';
207 // @codeCoverageIgnoreEnd
210 * Get a cached or new language object for a given language code
211 * @param string $code
212 * @throws MWException
215 static function factory( $code ) {
216 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
218 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
219 $code = $wgDummyLanguageCodes[$code];
222 // get the language object to process
223 $langObj = self
::$mLangObjCache[$code] ?? self
::newFromCode( $code );
225 // merge the language object in to get it up front in the cache
226 self
::$mLangObjCache = array_merge( [ $code => $langObj ], self
::$mLangObjCache );
227 // get rid of the oldest ones in case we have an overflow
228 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
234 * Create a language object for a given language code
235 * @param string $code
236 * @param bool $fallback Whether we're going through language fallback chain
237 * @throws MWException
240 protected static function newFromCode( $code, $fallback = false ) {
241 $langNameUtils = MediaWikiServices
::getInstance()->getLanguageNameUtils();
242 if ( !$langNameUtils->isValidCode( $code ) ) {
243 throw new MWException( "Invalid language code \"$code\"" );
246 if ( !$langNameUtils->isValidBuiltInCode( $code ) ) {
247 // It's not possible to customise this code with class files, so
248 // just return a Language object. This is to support uselang= hacks.
249 $lang = new Language
;
250 $lang->mCode
= $code;
254 // Check if there is a language class for the code
255 $class = self
::classFromCode( $code, $fallback );
256 // LanguageCode does not inherit Language
257 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
262 // Keep trying the fallback list until we find an existing class
263 $fallbacks = self
::getFallbacksFor( $code );
264 foreach ( $fallbacks as $fallbackCode ) {
265 if ( !$langNameUtils->isValidBuiltInCode( $fallbackCode ) ) {
266 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
269 $class = self
::classFromCode( $fallbackCode );
270 if ( class_exists( $class ) ) {
272 $lang->mCode
= $code;
277 throw new MWException( "Invalid fallback sequence for language '$code'" );
281 * Intended for tests that may change configuration in a way that invalidates caches.
285 public static function clearCaches() {
286 if ( !defined( 'MW_PHPUNIT_TEST' ) && !defined( 'MEDIAWIKI_INSTALL' ) ) {
287 throw new MWException( __METHOD__
. ' must not be used outside tests/installer' );
289 if ( !defined( 'MEDIAWIKI_INSTALL' ) ) {
290 MediaWikiServices
::getInstance()->resetServiceForTesting( 'LocalisationCache' );
291 MediaWikiServices
::getInstance()->resetServiceForTesting( 'LanguageNameUtils' );
293 self
::$mLangObjCache = [];
294 self
::$fallbackLanguageCache = [];
295 self
::$grammarTransformations = null;
299 * Checks whether any localisation is available for that language tag
300 * in MediaWiki (MessagesXx.php exists).
302 * @deprecated since 1.34, use LanguageNameUtils
303 * @param string $code Language tag (in lower case)
304 * @return bool Whether language is supported
307 public static function isSupportedLanguage( $code ) {
308 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
309 ->isSupportedLanguage( $code );
313 * Returns true if a language code string is a well-formed language tag
314 * according to RFC 5646.
315 * This function only checks well-formedness; it doesn't check that
316 * language, script or variant codes actually exist in the repositories.
318 * Based on regexes by Mark Davis of the Unicode Consortium:
319 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
321 * @param string $code
322 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
327 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
330 $alphanum = '[a-z0-9]';
331 $x = 'x'; # private use singleton
332 $singleton = '[a-wy-z]'; # other singleton
333 $s = $lenient ?
'[-_]' : '-';
335 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
336 $script = "$alpha{4}"; # ISO 15924
337 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
338 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
339 $extension = "$singleton(?:$s$alphanum{2,8})+";
340 $privateUse = "$x(?:$s$alphanum{1,8})+";
342 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
343 # Since these are limited, this is safe even later changes to the registry --
344 # the only oddity is that it might change the type of the tag, and thus
345 # the results from the capturing groups.
346 # https://www.iana.org/assignments/language-subtag-registry
348 $grandfathered = "en{$s}GB{$s}oed"
349 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
350 . "|no{$s}(?:bok|nyn)"
351 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
352 . "|zh{$s}min{$s}nan";
354 $variantList = "$variant(?:$s$variant)*";
355 $extensionList = "$extension(?:$s$extension)*";
357 $langtag = "(?:($language)"
360 . "(?:$s$variantList)?"
361 . "(?:$s$extensionList)?"
362 . "(?:$s$privateUse)?)";
364 # The final breakdown, with capturing groups for each of these components
365 # The variants, extensions, grandfathered, and private-use may have interior '-'
367 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
369 return (bool)preg_match( "/$root/", strtolower( $code ) );
373 * Returns true if a language code string is of a valid form, whether or
374 * not it exists. This includes codes which are used solely for
375 * customisation via the MediaWiki namespace.
377 * @deprecated since 1.34, use LanguageNameUtils
379 * @param string $code
383 public static function isValidCode( $code ) {
384 return MediaWikiServices
::getInstance()->getLanguageNameUtils()->isValidCode( $code );
388 * Returns true if a language code is of a valid form for the purposes of
389 * internal customisation of MediaWiki, via Messages*.php or *.json.
391 * @deprecated since 1.34, use LanguageNameUtils
393 * @param string $code
398 public static function isValidBuiltInCode( $code ) {
399 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
400 ->isValidBuiltInCode( $code );
404 * Returns true if a language code is an IETF tag known to MediaWiki.
406 * @deprecated since 1.34, use LanguageNameUtils
413 public static function isKnownLanguageTag( $tag ) {
414 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
415 ->isKnownLanguageTag( $tag );
419 * Get the LocalisationCache instance
421 * @deprecated since 1.34, use MediaWikiServices
422 * @return LocalisationCache
424 public static function getLocalisationCache() {
425 return MediaWikiServices
::getInstance()->getLocalisationCache();
428 function __construct() {
429 $this->mConverter
= new FakeConverter( $this );
430 // Set the code to the name of the descendant
431 if ( static::class === 'Language' ) {
434 $this->mCode
= str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
436 $services = MediaWikiServices
::getInstance();
437 $this->localisationCache
= $services->getLocalisationCache();
438 $this->langNameUtils
= $services->getLanguageNameUtils();
442 * Reduce memory usage
444 function __destruct() {
445 foreach ( $this as $name => $value ) {
446 unset( $this->$name );
451 * Hook which will be called if this is the content language.
452 * Descendants can use this to register hook functions or modify globals
454 function initContLang() {
461 public function getFallbackLanguages() {
462 return self
::getFallbacksFor( $this->mCode
);
466 * Exports $wgBookstoreListEn
469 public function getBookstoreList() {
470 return $this->localisationCache
->getItem( $this->mCode
, 'bookstoreList' );
474 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
475 * available in localised form, it will be included in English.
477 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
479 public function getNamespaces() {
480 if ( is_null( $this->namespaceNames
) ) {
481 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
483 $validNamespaces = MediaWikiServices
::getInstance()->getNamespaceInfo()->
484 getCanonicalNamespaces();
486 $this->namespaceNames
= $wgExtraNamespaces +
487 $this->localisationCache
->getItem( $this->mCode
, 'namespaceNames' );
488 $this->namespaceNames +
= $validNamespaces;
490 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
491 if ( $wgMetaNamespaceTalk ) {
492 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
494 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
495 $this->namespaceNames
[NS_PROJECT_TALK
] =
496 $this->fixVariableInNamespace( $talk );
499 # Sometimes a language will be localised but not actually exist on this wiki.
500 foreach ( $this->namespaceNames
as $key => $text ) {
501 if ( !isset( $validNamespaces[$key] ) ) {
502 unset( $this->namespaceNames
[$key] );
506 # The above mixing may leave namespaces out of canonical order.
507 # Re-order by namespace ID number...
508 ksort( $this->namespaceNames
);
510 Hooks
::run( 'LanguageGetNamespaces', [ &$this->namespaceNames
] );
513 return $this->namespaceNames
;
517 * Arbitrarily set all of the namespace names at once. Mainly used for testing
518 * @param array $namespaces Array of namespaces (id => name)
520 public function setNamespaces( array $namespaces ) {
521 $this->namespaceNames
= $namespaces;
522 $this->mNamespaceIds
= null;
526 * Resets all of the namespace caches. Mainly used for testing
528 public function resetNamespaces() {
529 $this->namespaceNames
= null;
530 $this->mNamespaceIds
= null;
531 $this->namespaceAliases
= null;
535 * A convenience function that returns getNamespaces() with spaces instead of underscores
536 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
540 public function getFormattedNamespaces() {
541 $ns = $this->getNamespaces();
542 foreach ( $ns as $k => $v ) {
543 $ns[$k] = strtr( $v, '_', ' ' );
549 * Get a namespace value by key
552 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
553 * echo $mw_ns; // prints 'MediaWiki'
556 * @param int $index The array key of the namespace to return
557 * @return string|bool String if the namespace value exists, otherwise false
559 public function getNsText( $index ) {
560 $ns = $this->getNamespaces();
561 return $ns[$index] ??
false;
565 * A convenience function that returns the same thing as
566 * getNsText() except with '_' changed to ' ', useful for
570 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
571 * echo $mw_ns; // prints 'MediaWiki talk'
574 * @param int $index The array key of the namespace to return
575 * @return string Namespace name without underscores (empty string if namespace does not exist)
577 public function getFormattedNsText( $index ) {
578 $ns = $this->getNsText( $index );
579 return strtr( $ns, '_', ' ' );
583 * Returns gender-dependent namespace alias if available.
584 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
585 * @param int $index Namespace index
586 * @param string $gender Gender key (male, female... )
590 public function getGenderNsText( $index, $gender ) {
591 global $wgExtraGenderNamespaces;
593 $ns = $wgExtraGenderNamespaces +
594 (array)$this->localisationCache
->getItem( $this->mCode
, 'namespaceGenderAliases' );
596 return $ns[$index][$gender] ??
$this->getNsText( $index );
600 * Whether this language uses gender-dependent namespace aliases.
601 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
605 public function needsGenderDistinction() {
606 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
607 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
608 // $wgExtraGenderNamespaces overrides everything
610 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
611 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
612 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
615 // Check what is in i18n files
616 $aliases = $this->localisationCache
->getItem( $this->mCode
, 'namespaceGenderAliases' );
617 return count( $aliases ) > 0;
622 * Get a namespace key by value, case insensitive.
623 * Only matches namespace names for the current language, not the
624 * canonical ones defined in Namespace.php.
626 * @param string $text
627 * @return int|bool An integer if $text is a valid value otherwise false
629 function getLocalNsIndex( $text ) {
630 $lctext = $this->lc( $text );
631 $ids = $this->getNamespaceIds();
632 return $ids[$lctext] ??
false;
638 public function getNamespaceAliases() {
639 if ( is_null( $this->namespaceAliases
) ) {
640 $aliases = $this->localisationCache
->getItem( $this->mCode
, 'namespaceAliases' );
644 foreach ( $aliases as $name => $index ) {
645 if ( $index === NS_PROJECT_TALK
) {
646 unset( $aliases[$name] );
647 $name = $this->fixVariableInNamespace( $name );
648 $aliases[$name] = $index;
653 global $wgExtraGenderNamespaces;
654 $genders = $wgExtraGenderNamespaces +
(array)$this->localisationCache
655 ->getItem( $this->mCode
, 'namespaceGenderAliases' );
656 foreach ( $genders as $index => $forms ) {
657 foreach ( $forms as $alias ) {
658 $aliases[$alias] = $index;
662 # Also add converted namespace names as aliases, to avoid confusion.
663 $convertedNames = [];
664 foreach ( $this->getVariants() as $variant ) {
665 if ( $variant === $this->mCode
) {
668 foreach ( $this->getNamespaces() as $ns => $_ ) {
669 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
673 $this->namespaceAliases
= $aliases +
$convertedNames;
675 # Filter out aliases to namespaces that don't exist, e.g. from extensions
676 # that aren't loaded here but are included in the l10n cache.
677 # (array_intersect preserves keys from its first argument)
678 $this->namespaceAliases
= array_intersect(
679 $this->namespaceAliases
,
680 array_keys( $this->getNamespaces() )
684 return $this->namespaceAliases
;
690 public function getNamespaceIds() {
691 if ( is_null( $this->mNamespaceIds
) ) {
692 global $wgNamespaceAliases;
693 # Put namespace names and aliases into a hashtable.
694 # If this is too slow, then we should arrange it so that it is done
695 # before caching. The catch is that at pre-cache time, the above
696 # class-specific fixup hasn't been done.
697 $this->mNamespaceIds
= [];
698 foreach ( $this->getNamespaces() as $index => $name ) {
699 $this->mNamespaceIds
[$this->lc( $name )] = $index;
701 foreach ( $this->getNamespaceAliases() as $name => $index ) {
702 $this->mNamespaceIds
[$this->lc( $name )] = $index;
704 if ( $wgNamespaceAliases ) {
705 foreach ( $wgNamespaceAliases as $name => $index ) {
706 $this->mNamespaceIds
[$this->lc( $name )] = $index;
710 return $this->mNamespaceIds
;
714 * Get a namespace key by value, case insensitive. Canonical namespace
715 * names override custom ones defined for the current language.
717 * @param string $text
718 * @return int|bool An integer if $text is a valid value otherwise false
720 public function getNsIndex( $text ) {
721 $lctext = $this->lc( $text );
722 $ns = MediaWikiServices
::getInstance()->getNamespaceInfo()->
723 getCanonicalIndex( $lctext );
724 if ( $ns !== null ) {
727 $ids = $this->getNamespaceIds();
728 return $ids[$lctext] ??
false;
732 * short names for language variants used for language conversion links.
734 * @param string $code
735 * @param bool $usemsg Use the "variantname-xyz" message if it exists
738 public function getVariantname( $code, $usemsg = true ) {
739 $msg = "variantname-$code";
740 if ( $usemsg && wfMessage( $msg )->exists() ) {
741 return $this->getMessageFromDB( $msg );
743 $name = $this->langNameUtils
->getLanguageName( $code );
745 return $name; # if it's defined as a language name, show that
747 # otherwise, output the language code
753 * @return string[]|bool List of date format preference keys, or false if disabled.
755 public function getDatePreferences() {
756 return $this->localisationCache
->getItem( $this->mCode
, 'datePreferences' );
762 function getDateFormats() {
763 return $this->localisationCache
->getItem( $this->mCode
, 'dateFormats' );
767 * @return array|string
769 public function getDefaultDateFormat() {
770 $df = $this->localisationCache
->getItem( $this->mCode
, 'defaultDateFormat' );
771 if ( $df === 'dmy or mdy' ) {
772 global $wgAmericanDates;
773 return $wgAmericanDates ?
'mdy' : 'dmy';
782 public function getDatePreferenceMigrationMap() {
783 return $this->localisationCache
->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
789 public function getExtraUserToggles() {
790 wfDeprecated( __METHOD__
, '1.34' );
798 function getUserToggle( $tog ) {
799 return $this->getMessageFromDB( "tog-$tog" );
803 * Get an array of language names, indexed by code.
805 * @deprecated since 1.34, use LanguageNameUtils::getLanguageNames
806 * @param null|string $inLanguage Code of language in which to return the names
807 * Use self::AS_AUTONYMS for autonyms (native names)
808 * @param string $include One of:
809 * self::ALL all available languages
810 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
811 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
812 * @return array Language code => language name (sorted by key)
815 public static function fetchLanguageNames( $inLanguage = self
::AS_AUTONYMS
, $include = 'mw' ) {
816 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
817 ->getLanguageNames( $inLanguage, $include );
821 * @deprecated since 1.34, use LanguageNameUtils::getLanguageName
822 * @param string $code The code of the language for which to get the name
823 * @param null|string $inLanguage Code of language in which to return the name
824 * (SELF::AS_AUTONYMS for autonyms)
825 * @param string $include See fetchLanguageNames()
826 * @return string Language name or empty
829 public static function fetchLanguageName(
831 $inLanguage = self
::AS_AUTONYMS
,
834 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
835 ->getLanguageName( $code, $inLanguage, $include );
839 * Get a message from the MediaWiki namespace.
841 * @param string $msg Message name
844 public function getMessageFromDB( $msg ) {
845 return $this->msg( $msg )->text();
849 * Get message object in this language. Only for use inside this class.
851 * @param string $msg Message name
854 protected function msg( $msg ) {
855 return wfMessage( $msg )->inLanguage( $this );
862 public function getMonthName( $key ) {
863 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
869 public function getMonthNamesArray() {
870 $monthNames = [ '' ];
871 for ( $i = 1; $i < 13; $i++
) {
872 $monthNames[] = $this->getMonthName( $i );
881 public function getMonthNameGen( $key ) {
882 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
889 public function getMonthAbbreviation( $key ) {
890 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
896 public function getMonthAbbreviationsArray() {
897 $monthNames = [ '' ];
898 for ( $i = 1; $i < 13; $i++
) {
899 $monthNames[] = $this->getMonthAbbreviation( $i );
908 public function getWeekdayName( $key ) {
909 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
916 function getWeekdayAbbreviation( $key ) {
917 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
924 function getIranianCalendarMonthName( $key ) {
925 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
932 function getHebrewCalendarMonthName( $key ) {
933 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
940 function getHebrewCalendarMonthNameGen( $key ) {
941 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
948 function getHijriCalendarMonthName( $key ) {
949 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
953 * Pass through result from $dateTimeObj->format()
954 * @param DateTime|bool|null &$dateTimeObj
956 * @param DateTimeZone|bool|null $zone
957 * @param string $code
960 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
961 if ( !$dateTimeObj ) {
962 $dateTimeObj = DateTime
::createFromFormat(
963 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
966 return $dateTimeObj->format( $code );
970 * This is a workalike of PHP's date() function, but with better
971 * internationalisation, a reduced set of format characters, and a better
974 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
975 * the PHP manual for definitions. There are a number of extensions, which
978 * xn Do not translate digits of the next numeric format character
979 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
980 * xr Use roman numerals for the next numeric format character
981 * xh Use hebrew numerals for the next numeric format character
983 * xg Genitive month name
985 * xij j (day number) in Iranian calendar
986 * xiF F (month name) in Iranian calendar
987 * xin n (month number) in Iranian calendar
988 * xiy y (two digit year) in Iranian calendar
989 * xiY Y (full year) in Iranian calendar
990 * xit t (days in month) in Iranian calendar
991 * xiz z (day of the year) in Iranian calendar
993 * xjj j (day number) in Hebrew calendar
994 * xjF F (month name) in Hebrew calendar
995 * xjt t (days in month) in Hebrew calendar
996 * xjx xg (genitive month name) in Hebrew calendar
997 * xjn n (month number) in Hebrew calendar
998 * xjY Y (full year) in Hebrew calendar
1000 * xmj j (day number) in Hijri calendar
1001 * xmF F (month name) in Hijri calendar
1002 * xmn n (month number) in Hijri calendar
1003 * xmY Y (full year) in Hijri calendar
1005 * xkY Y (full year) in Thai solar calendar. Months and days are
1006 * identical to the Gregorian calendar
1007 * xoY Y (full year) in Minguo calendar or Juche year.
1008 * Months and days are identical to the
1009 * Gregorian calendar
1010 * xtY Y (full year) in Japanese nengo. Months and days are
1011 * identical to the Gregorian calendar
1013 * Characters enclosed in double quotes will be considered literal (with
1014 * the quotes themselves removed). Unmatched quotes will be considered
1015 * literal quotes. Example:
1017 * "The month is" F => The month is January
1020 * Backslash escaping is also supported.
1022 * Input timestamp is assumed to be pre-normalized to the desired local
1023 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1024 * $ts is UTC if $zone is not given.
1026 * @param string $format
1027 * @param string $ts 14-character timestamp
1030 * @param DateTimeZone|null $zone Timezone of $ts
1031 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1032 * Only makes sense if $ts is the current time.
1033 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1035 * @throws MWException
1038 public function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = 'unused' ) {
1043 $dateTimeObj = false;
1052 $usedSecond = false;
1053 $usedMinute = false;
1060 $usedISOYear = false;
1061 $usedIsLeapYear = false;
1063 $usedHebrewMonth = false;
1064 $usedIranianMonth = false;
1065 $usedHijriMonth = false;
1066 $usedHebrewYear = false;
1067 $usedIranianYear = false;
1068 $usedHijriYear = false;
1069 $usedTennoYear = false;
1071 if ( strlen( $ts ) !== 14 ) {
1072 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1075 if ( !ctype_digit( $ts ) ) {
1076 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1079 $formatLength = strlen( $format );
1080 for ( $p = 0; $p < $formatLength; $p++
) {
1082 $code = $format[$p];
1083 if ( $code == 'x' && $p < $formatLength - 1 ) {
1084 $code .= $format[++
$p];
1087 if ( ( $code === 'xi'
1093 && $p < $formatLength - 1 ) {
1094 $code .= $format[++
$p];
1105 $rawToggle = !$rawToggle;
1115 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1118 $usedHebrewMonth = true;
1120 $hebrew = self
::tsToHebrew( $ts );
1122 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1126 $num = substr( $ts, 6, 2 );
1130 $s .= $this->getWeekdayAbbreviation(
1131 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1136 $num = intval( substr( $ts, 6, 2 ) );
1141 $iranian = self
::tsToIranian( $ts );
1148 $hijri = self
::tsToHijri( $ts );
1155 $hebrew = self
::tsToHebrew( $ts );
1161 $s .= $this->getWeekdayName(
1162 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1167 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1170 $usedIranianMonth = true;
1172 $iranian = self
::tsToIranian( $ts );
1174 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1177 $usedHijriMonth = true;
1179 $hijri = self
::tsToHijri( $ts );
1181 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1184 $usedHebrewMonth = true;
1186 $hebrew = self
::tsToHebrew( $ts );
1188 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1192 $num = substr( $ts, 4, 2 );
1196 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1200 $num = intval( substr( $ts, 4, 2 ) );
1203 $usedIranianMonth = true;
1205 $iranian = self
::tsToIranian( $ts );
1210 $usedHijriMonth = true;
1212 $hijri = self
::tsToHijri( $ts );
1217 $usedHebrewMonth = true;
1219 $hebrew = self
::tsToHebrew( $ts );
1224 $usedHebrewMonth = true;
1226 $hebrew = self
::tsToHebrew( $ts );
1232 $num = substr( $ts, 0, 4 );
1235 $usedIranianYear = true;
1237 $iranian = self
::tsToIranian( $ts );
1242 $usedHijriYear = true;
1244 $hijri = self
::tsToHijri( $ts );
1249 $usedHebrewYear = true;
1251 $hebrew = self
::tsToHebrew( $ts );
1258 $thai = self
::tsToYear( $ts, 'thai' );
1265 $minguo = self
::tsToYear( $ts, 'minguo' );
1270 $usedTennoYear = true;
1272 $tenno = self
::tsToYear( $ts, 'tenno' );
1278 $num = substr( $ts, 2, 2 );
1281 $usedIranianYear = true;
1283 $iranian = self
::tsToIranian( $ts );
1285 $num = substr( $iranian[0], -2 );
1288 $usedIranianYear = true;
1290 $iranian = self
::tsToIranian( $ts );
1292 $num = self
::$IRANIAN_DAYS[$iranian[1] - 1];
1295 $usedIranianYear = true;
1297 $iranian = self
::tsToIranian( $ts );
1303 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1307 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1311 $h = substr( $ts, 8, 2 );
1312 $num = $h %
12 ?
: 12;
1316 $num = intval( substr( $ts, 8, 2 ) );
1320 $h = substr( $ts, 8, 2 );
1321 $num = sprintf( '%02d', $h %
12 ?
: 12 );
1325 $num = substr( $ts, 8, 2 );
1329 $num = substr( $ts, 10, 2 );
1333 $num = substr( $ts, 12, 2 );
1343 $s .= self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1349 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1353 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1357 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1360 $usedIsLeapYear = true;
1361 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1364 $usedISOYear = true;
1365 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1372 $num = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1375 # Backslash escaping
1376 if ( $p < $formatLength - 1 ) {
1377 $s .= $format[++
$p];
1384 if ( $p < $formatLength - 1 ) {
1385 $endQuote = strpos( $format, '"', $p +
1 );
1386 if ( $endQuote === false ) {
1387 # No terminating quote, assume literal "
1390 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1394 # Quote at end of string, assume literal "
1401 if ( $num !== false ) {
1402 if ( $rawToggle ||
$raw ) {
1405 } elseif ( $roman ) {
1406 $s .= self
::romanNumeral( $num );
1408 } elseif ( $hebrewNum ) {
1409 $s .= self
::hebrewNumeral( $num );
1412 $s .= $this->formatNum( $num, true );
1417 if ( $ttl === 'unused' ) {
1418 // No need to calculate the TTL, the caller wont use it anyway.
1419 } elseif ( $usedSecond ) {
1421 } elseif ( $usedMinute ) {
1422 $ttl = 60 - substr( $ts, 12, 2 );
1423 } elseif ( $usedHour ) {
1424 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1425 } elseif ( $usedAMPM ) {
1426 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1427 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1431 $usedIranianMonth ||
1438 // @todo Someone who understands the non-Gregorian calendars
1439 // should write proper logic for them so that they don't need purged every day.
1440 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1441 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1444 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1445 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1448 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1449 $timeRemainingInDay;
1450 } elseif ( $usedISOYear ) {
1451 // December 28th falls on the last ISO week of the year, every year.
1452 // The last ISO week of a year can be 52 or 53.
1453 $lastWeekOfISOYear = DateTime
::createFromFormat(
1455 substr( $ts, 0, 4 ) . '1228',
1456 $zone ?
: new DateTimeZone( 'UTC' )
1458 $currentISOWeek = self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1459 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1460 $timeRemainingInWeek =
1461 ( 7 - self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1462 +
$timeRemainingInDay;
1463 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1468 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1469 substr( $ts, 6, 2 ) ) * 86400
1470 +
$timeRemainingInDay;
1471 } elseif ( $usedYear ) {
1473 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1474 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1475 +
$timeRemainingInDay;
1476 } elseif ( $usedIsLeapYear ) {
1477 $year = substr( $ts, 0, 4 );
1478 $timeRemainingInYear =
1479 ( self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1480 self
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1481 +
$timeRemainingInDay;
1483 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1484 // this isn't a leap year. see when the next one starts
1485 $nextCandidate = $year - $mod +
4;
1486 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1487 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1488 $timeRemainingInYear;
1490 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1491 $timeRemainingInYear;
1494 // this is a leap year, so the next year isn't
1495 $possibleTtls[] = $timeRemainingInYear;
1499 if ( $possibleTtls ) {
1500 $ttl = min( $possibleTtls );
1507 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1508 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1511 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1512 * Gregorian dates to Iranian dates. Originally written in C, it
1513 * is released under the terms of GNU Lesser General Public
1514 * License. Conversion to PHP was performed by Niklas Laxström.
1516 * Link: http://www.farsiweb.info/jalali/jalali.c
1522 private static function tsToIranian( $ts ) {
1523 $gy = substr( $ts, 0, 4 ) - 1600;
1524 $gm = substr( $ts, 4, 2 ) - 1;
1525 $gd = substr( $ts, 6, 2 ) - 1;
1527 # Days passed from the beginning (including leap years)
1529 +
floor( ( $gy +
3 ) / 4 )
1530 - floor( ( $gy +
99 ) / 100 )
1531 +
floor( ( $gy +
399 ) / 400 );
1533 // Add days of the past months of this year
1534 for ( $i = 0; $i < $gm; $i++
) {
1535 $gDayNo +
= self
::$GREG_DAYS[$i];
1539 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1543 // Days passed in current month
1544 $gDayNo +
= (int)$gd;
1546 $jDayNo = $gDayNo - 79;
1548 $jNp = floor( $jDayNo / 12053 );
1551 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1554 if ( $jDayNo >= 366 ) {
1555 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1556 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1561 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1562 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1568 return [ $jy, $jm, $jd, $jz ];
1572 * Converting Gregorian dates to Hijri dates.
1574 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1576 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1582 private static function tsToHijri( $ts ) {
1583 $year = substr( $ts, 0, 4 );
1584 $month = substr( $ts, 4, 2 );
1585 $day = substr( $ts, 6, 2 );
1593 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1594 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1596 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1597 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1598 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1601 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1602 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1605 $zl = $zjd - 1948440 +
10632;
1606 $zn = (int)( ( $zl - 1 ) / 10631 );
1607 $zl = $zl - 10631 * $zn +
354;
1608 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1609 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1610 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1611 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1612 $zm = (int)( ( 24 * $zl ) / 709 );
1613 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1614 $zy = 30 * $zn +
$zj - 30;
1616 return [ $zy, $zm, $zd ];
1620 * Converting Gregorian dates to Hebrew dates.
1622 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1623 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1624 * to translate the relevant functions into PHP and release them under
1627 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1628 * and Adar II is 14. In a non-leap year, Adar is 6.
1634 private static function tsToHebrew( $ts ) {
1636 $year = substr( $ts, 0, 4 );
1637 $month = substr( $ts, 4, 2 );
1638 $day = substr( $ts, 6, 2 );
1640 # Calculate Hebrew year
1641 $hebrewYear = $year +
3760;
1643 # Month number when September = 1, August = 12
1645 if ( $month > 12 ) {
1652 # Calculate day of year from 1 September
1654 for ( $i = 1; $i < $month; $i++
) {
1658 # Check if the year is leap
1659 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1662 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1669 # Calculate the start of the Hebrew year
1670 $start = self
::hebrewYearStart( $hebrewYear );
1672 # Calculate next year's start
1673 if ( $dayOfYear <= $start ) {
1674 # Day is before the start of the year - it is the previous year
1676 $nextStart = $start;
1680 # Add days since previous year's 1 September
1682 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1686 # Start of the new (previous) year
1687 $start = self
::hebrewYearStart( $hebrewYear );
1690 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1693 # Calculate Hebrew day of year
1694 $hebrewDayOfYear = $dayOfYear - $start;
1696 # Difference between year's days
1697 $diff = $nextStart - $start;
1698 # Add 12 (or 13 for leap years) days to ignore the difference between
1699 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1700 # difference is only about the year type
1701 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1707 # Check the year pattern, and is leap year
1708 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1709 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1710 # and non-leap years
1711 $yearPattern = $diff %
30;
1712 # Check if leap year
1713 $isLeap = $diff >= 30;
1715 # Calculate day in the month from number of day in the Hebrew year
1716 # Don't check Adar - if the day is not in Adar, we will stop before;
1717 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1718 $hebrewDay = $hebrewDayOfYear;
1721 while ( $hebrewMonth <= 12 ) {
1722 # Calculate days in this month
1723 if ( $isLeap && $hebrewMonth == 6 ) {
1724 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1726 if ( $hebrewDay <= $days ) {
1730 # Subtract the days of Adar I
1731 $hebrewDay -= $days;
1734 if ( $hebrewDay <= $days ) {
1739 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1740 # Cheshvan in a complete year (otherwise as the rule below)
1742 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1743 # Kislev in an incomplete year (otherwise as the rule below)
1746 # Odd months have 30 days, even have 29
1747 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1749 if ( $hebrewDay <= $days ) {
1750 # In the current month
1753 # Subtract the days of the current month
1754 $hebrewDay -= $days;
1755 # Try in the next month
1760 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1764 * This calculates the Hebrew year start, as days since 1 September.
1765 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1766 * Used for Hebrew date.
1772 private static function hebrewYearStart( $year ) {
1773 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1774 $b = intval( ( $year - 1 ) %
4 );
1775 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1779 $Mar = intval( $m );
1785 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1786 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1788 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1790 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1794 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1799 * Algorithm to convert Gregorian dates to Thai solar dates,
1800 * Minguo dates or Minguo dates.
1802 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1803 * https://en.wikipedia.org/wiki/Minguo_calendar
1804 * https://en.wikipedia.org/wiki/Japanese_era_name
1806 * @param string $ts 14-character timestamp
1807 * @param string $cName Calender name
1808 * @return array Converted year, month, day
1810 private static function tsToYear( $ts, $cName ) {
1811 $gy = substr( $ts, 0, 4 );
1812 $gm = substr( $ts, 4, 2 );
1813 $gd = substr( $ts, 6, 2 );
1815 if ( !strcmp( $cName, 'thai' ) ) {
1817 # Add 543 years to the Gregorian calendar
1818 # Months and days are identical
1819 $gy_offset = $gy +
543;
1820 # fix for dates between 1912 and 1941
1821 # https://en.wikipedia.org/?oldid=836596673#New_year
1822 if ( $gy >= 1912 && $gy <= 1940 ) {
1826 $gm = ( $gm - 3 ) %
12;
1828 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1830 # Deduct 1911 years from the Gregorian calendar
1831 # Months and days are identical
1832 $gy_offset = $gy - 1911;
1833 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1834 # Nengō dates up to Meiji period
1835 # Deduct years from the Gregorian calendar
1836 # depending on the nengo periods
1837 # Months and days are identical
1839 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1840 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1843 $gy_gannen = $gy - 1868 +
1;
1844 $gy_offset = $gy_gannen;
1845 if ( $gy_gannen == 1 ) {
1848 $gy_offset = '明治' . $gy_offset;
1850 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1851 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1852 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1853 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1854 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1857 $gy_gannen = $gy - 1912 +
1;
1858 $gy_offset = $gy_gannen;
1859 if ( $gy_gannen == 1 ) {
1862 $gy_offset = '大正' . $gy_offset;
1864 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1865 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1866 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1869 $gy_gannen = $gy - 1926 +
1;
1870 $gy_offset = $gy_gannen;
1871 if ( $gy_gannen == 1 ) {
1874 $gy_offset = '昭和' . $gy_offset;
1876 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1877 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1878 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1881 $gy_gannen = $gy - 1989 +
1;
1882 $gy_offset = $gy_gannen;
1883 if ( $gy_gannen == 1 ) {
1886 $gy_offset = '平成' . $gy_offset;
1889 $gy_gannen = $gy - 2019 +
1;
1890 $gy_offset = $gy_gannen;
1891 if ( $gy_gannen == 1 ) {
1894 $gy_offset = '令和' . $gy_offset;
1900 return [ $gy_offset, $gm, $gd ];
1904 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1906 * This is the rule the BIDI algorithm uses to determine the directionality of
1907 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
1908 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1910 * TODO: Does not handle BIDI control characters inside the text.
1911 * TODO: Does not handle unallocated characters.
1913 * @param string $text Text to test
1914 * @return null|string Directionality ('ltr' or 'rtl') or null
1916 private static function strongDirFromContent( $text = '' ) {
1917 if ( !preg_match( self
::$strongDirRegex, $text, $matches ) ) {
1920 if ( $matches[1] === '' ) {
1927 * Roman number formatting up to 10000
1933 static function romanNumeral( $num ) {
1935 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1936 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1937 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1938 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1939 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1942 $num = intval( $num );
1943 if ( $num > 10000 ||
$num <= 0 ) {
1948 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1949 if ( $num >= $pow10 ) {
1950 $s .= $table[$i][(int)floor( $num / $pow10 )];
1952 $num = $num %
$pow10;
1958 * Hebrew Gematria number formatting up to 9999
1964 static function hebrewNumeral( $num ) {
1966 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
1967 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
1980 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
1983 $num = intval( $num );
1984 if ( $num > 9999 ||
$num <= 0 ) {
1988 // Round thousands have special notations
1989 if ( $num === 1000 ) {
1991 } elseif ( $num %
1000 === 0 ) {
1992 return $table[0][$num / 1000] . "' אלפים";
1997 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1998 if ( $num >= $pow10 ) {
1999 if ( $num === 15 ||
$num === 16 ) {
2000 $letters[] = $table[0][9];
2001 $letters[] = $table[0][$num - 9];
2004 $letters = array_merge(
2006 (array)$table[$i][intval( $num / $pow10 )]
2009 if ( $pow10 === 1000 ) {
2015 $num = $num %
$pow10;
2018 $preTransformLength = count( $letters );
2019 if ( $preTransformLength === 1 ) {
2020 // Add geresh (single quote) to one-letter numbers
2023 $lastIndex = $preTransformLength - 1;
2024 $letters[$lastIndex] = str_replace(
2025 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2026 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2027 $letters[$lastIndex]
2030 // Add gershayim (double quote) to multiple-letter numbers,
2031 // but exclude numbers with only one letter after the thousands
2032 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2033 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2036 array_splice( $letters, -1, 0, '"' );
2040 return implode( $letters );
2044 * Used by date() and time() to adjust the time output.
2046 * @param string $ts The time in date('YmdHis') format
2047 * @param mixed $tz Adjust the time by this amount (default false, mean we
2048 * get user timecorrection setting)
2051 public function userAdjust( $ts, $tz = false ) {
2052 global $wgUser, $wgLocalTZoffset;
2054 if ( $tz === false ) {
2055 $tz = $wgUser->getOption( 'timecorrection' );
2058 $data = explode( '|', $tz, 3 );
2060 if ( $data[0] == 'ZoneInfo' ) {
2062 $userTZ = new DateTimeZone( $data[2] );
2063 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2064 $date->setTimezone( $userTZ );
2065 return $date->format( 'YmdHis' );
2066 } catch ( Exception
$e ) {
2067 // Unrecognized timezone, default to 'Offset' with the stored offset.
2068 $data[0] = 'Offset';
2072 if ( $data[0] == 'System' ||
$tz == '' ) {
2073 # Global offset in minutes.
2074 $minDiff = $wgLocalTZoffset;
2075 } elseif ( $data[0] == 'Offset' ) {
2076 $minDiff = intval( $data[1] );
2078 $data = explode( ':', $tz );
2079 if ( count( $data ) == 2 ) {
2080 $data[0] = intval( $data[0] );
2081 $data[1] = intval( $data[1] );
2082 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2083 if ( $data[0] < 0 ) {
2084 $minDiff = -$minDiff;
2087 $minDiff = intval( $data[0] ) * 60;
2091 # No difference ? Return time unchanged
2092 if ( $minDiff == 0 ) {
2096 Wikimedia\
suppressWarnings(); // E_STRICT system time bitching
2097 # Generate an adjusted date; take advantage of the fact that mktime
2098 # will normalize out-of-range values so we don't have to split $minDiff
2099 # into hours and minutes.
2101 (int)substr( $ts, 8, 2 ) ), # Hours
2102 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2103 (int)substr( $ts, 12, 2 ), # Seconds
2104 (int)substr( $ts, 4, 2 ), # Month
2105 (int)substr( $ts, 6, 2 ), # Day
2106 (int)substr( $ts, 0, 4 ) ); # Year
2108 $date = date( 'YmdHis', $t );
2109 Wikimedia\restoreWarnings
();
2115 * This is meant to be used by time(), date(), and timeanddate() to get
2116 * the date preference they're supposed to use, it should be used in
2119 * function timeanddate([...], $format = true) {
2120 * $datePreference = $this->dateFormat($format);
2124 * @param int|string|bool $usePrefs If true, the user's preference is used
2125 * if false, the site/language default is used
2126 * if int/string, assumed to be a format.
2129 function dateFormat( $usePrefs = true ) {
2132 if ( is_bool( $usePrefs ) ) {
2134 $datePreference = $wgUser->getDatePreference();
2136 $datePreference = (string)User
::getDefaultOption( 'date' );
2139 $datePreference = (string)$usePrefs;
2143 if ( $datePreference == '' ) {
2147 return $datePreference;
2151 * Get a format string for a given type and preference
2152 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2153 * @param string $pref The format name as it appears in Messages*.php under
2156 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2160 function getDateFormatString( $type, $pref ) {
2161 $wasDefault = false;
2162 if ( $pref == 'default' ) {
2164 $pref = $this->getDefaultDateFormat();
2167 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2169 $this->localisationCache
->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2171 if ( $type === 'pretty' && $df === null ) {
2172 $df = $this->getDateFormatString( 'date', $pref );
2175 if ( !$wasDefault && $df === null ) {
2176 $pref = $this->getDefaultDateFormat();
2177 $df = $this->getLocalisationCache()
2178 ->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2181 $this->dateFormatStrings
[$type][$pref] = $df;
2183 return $this->dateFormatStrings
[$type][$pref];
2187 * @param string $ts The time format which needs to be turned into a
2188 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2189 * @param bool $adj Whether to adjust the time output according to the
2190 * user configured offset ($timecorrection)
2191 * @param mixed $format True to use user's date format preference
2192 * @param string|bool $timecorrection The time offset as returned by
2193 * validateTimeZone() in Special:Preferences
2196 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2197 $ts = wfTimestamp( TS_MW
, $ts );
2199 $ts = $this->userAdjust( $ts, $timecorrection );
2201 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2202 return $this->sprintfDate( $df, $ts );
2206 * @param string $ts The time format which needs to be turned into a
2207 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2208 * @param bool $adj Whether to adjust the time output according to the
2209 * user configured offset ($timecorrection)
2210 * @param mixed $format True to use user's date format preference
2211 * @param string|bool $timecorrection The time offset as returned by
2212 * validateTimeZone() in Special:Preferences
2215 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2216 $ts = wfTimestamp( TS_MW
, $ts );
2218 $ts = $this->userAdjust( $ts, $timecorrection );
2220 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2221 return $this->sprintfDate( $df, $ts );
2225 * @param string $ts The time format which needs to be turned into a
2226 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2227 * @param bool $adj Whether to adjust the time output according to the
2228 * user configured offset ($timecorrection)
2229 * @param mixed $format What format to return, if it's false output the
2230 * default one (default true)
2231 * @param string|bool $timecorrection The time offset as returned by
2232 * validateTimeZone() in Special:Preferences
2235 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2236 $ts = wfTimestamp( TS_MW
, $ts );
2238 $ts = $this->userAdjust( $ts, $timecorrection );
2240 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2241 return $this->sprintfDate( $df, $ts );
2245 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2249 * @param int $seconds The amount of seconds.
2250 * @param array $chosenIntervals The intervals to enable.
2254 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2255 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2259 foreach ( $intervals as $intervalName => $intervalValue ) {
2260 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2261 // duration-years, duration-decades, duration-centuries, duration-millennia
2262 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2263 $segments[] = $message->inLanguage( $this )->escaped();
2266 return $this->listToText( $segments );
2270 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2271 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2275 * @param int $seconds The amount of seconds.
2276 * @param array $chosenIntervals The intervals to enable.
2280 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2281 if ( empty( $chosenIntervals ) ) {
2282 $chosenIntervals = [
2294 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2295 $sortedNames = array_keys( $intervals );
2296 $smallestInterval = array_pop( $sortedNames );
2300 foreach ( $intervals as $name => $length ) {
2301 $value = floor( $seconds / $length );
2303 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2304 $seconds -= $value * $length;
2305 $segments[$name] = $value;
2313 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2315 * @param string $type Can be 'date', 'time' or 'both'
2316 * @param string $ts The time format which needs to be turned into a
2317 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2318 * @param User $user User object used to get preferences for timezone and format
2319 * @param array $options Array, can contain the following keys:
2320 * - 'timecorrection': time correction, can have the following values:
2321 * - true: use user's preference
2322 * - false: don't use time correction
2323 * - int: value of time correction in minutes
2324 * - 'format': format to use, can have the following values:
2325 * - true: use user's preference
2326 * - false: use default preference
2327 * - string: format to use
2331 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2332 $ts = wfTimestamp( TS_MW
, $ts );
2333 $options +
= [ 'timecorrection' => true, 'format' => true ];
2334 if ( $options['timecorrection'] !== false ) {
2335 if ( $options['timecorrection'] === true ) {
2336 $offset = $user->getOption( 'timecorrection' );
2338 $offset = $options['timecorrection'];
2340 $ts = $this->userAdjust( $ts, $offset );
2342 if ( $options['format'] === true ) {
2343 $format = $user->getDatePreference();
2345 $format = $options['format'];
2347 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2348 return $this->sprintfDate( $df, $ts );
2352 * Get the formatted date for the given timestamp and formatted for
2355 * @param mixed $ts Mixed: the time format which needs to be turned into a
2356 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2357 * @param User $user User object used to get preferences for timezone and format
2358 * @param array $options Array, can contain the following keys:
2359 * - 'timecorrection': time correction, can have the following values:
2360 * - true: use user's preference
2361 * - false: don't use time correction
2362 * - int: value of time correction in minutes
2363 * - 'format': format to use, can have the following values:
2364 * - true: use user's preference
2365 * - false: use default preference
2366 * - string: format to use
2370 public function userDate( $ts, User
$user, array $options = [] ) {
2371 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2375 * Get the formatted time for the given timestamp and formatted for
2378 * @param mixed $ts The time format which needs to be turned into a
2379 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2380 * @param User $user User object used to get preferences for timezone and format
2381 * @param array $options Array, can contain the following keys:
2382 * - 'timecorrection': time correction, can have the following values:
2383 * - true: use user's preference
2384 * - false: don't use time correction
2385 * - int: value of time correction in minutes
2386 * - 'format': format to use, can have the following values:
2387 * - true: use user's preference
2388 * - false: use default preference
2389 * - string: format to use
2393 public function userTime( $ts, User
$user, array $options = [] ) {
2394 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2398 * Get the formatted date and time for the given timestamp and formatted for
2401 * @param mixed $ts The time format which needs to be turned into a
2402 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2403 * @param User $user User object used to get preferences for timezone and format
2404 * @param array $options Array, can contain the following keys:
2405 * - 'timecorrection': time correction, can have the following values:
2406 * - true: use user's preference
2407 * - false: don't use time correction
2408 * - int: value of time correction in minutes
2409 * - 'format': format to use, can have the following values:
2410 * - true: use user's preference
2411 * - false: use default preference
2412 * - string: format to use
2416 public function userTimeAndDate( $ts, User
$user, array $options = [] ) {
2417 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2421 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2423 * Determine the difference between the timestamp and the current time, and
2424 * generate a readable timestamp by returning "<N> <units> ago", where the
2425 * largest possible unit is used.
2427 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2429 * @param MWTimestamp $time
2430 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2431 * @param User|null $user User the timestamp is being generated for
2432 * (or null to use main context's user)
2433 * @return string Formatted timestamp
2435 public function getHumanTimestamp(
2436 MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null
2438 if ( $relativeTo === null ) {
2439 $relativeTo = new MWTimestamp();
2441 if ( $user === null ) {
2442 $user = RequestContext
::getMain()->getUser();
2445 // Adjust for the user's timezone.
2446 $offsetThis = $time->offsetForUser( $user );
2447 $offsetRel = $relativeTo->offsetForUser( $user );
2450 if ( Hooks
::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2451 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2454 // Reset the timezone on the objects.
2455 $time->timestamp
->sub( $offsetThis );
2456 $relativeTo->timestamp
->sub( $offsetRel );
2462 * Convert an MWTimestamp into a pretty human-readable timestamp using
2463 * the given user preferences and relative base time.
2465 * @see Language::getHumanTimestamp
2466 * @param MWTimestamp $ts Timestamp to prettify
2467 * @param MWTimestamp $relativeTo Base timestamp
2468 * @param User $user User preferences to use
2469 * @return string Human timestamp
2472 private function getHumanTimestampInternal(
2473 MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user
2475 $diff = $ts->diff( $relativeTo );
2476 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2477 (int)$relativeTo->timestamp
->format( 'w' ) );
2478 $days = $diff->days ?
: (int)$diffDay;
2479 if ( $diff->invert ||
$days > 5
2480 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2482 // Timestamps are in different years: use full timestamp
2483 // Also do full timestamp for future dates
2485 * @todo FIXME: Add better handling of future timestamps.
2487 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2488 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2489 } elseif ( $days > 5 ) {
2490 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2491 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2492 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2493 } elseif ( $days > 1 ) {
2494 // Timestamp within the past week: show the day of the week and time
2495 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2496 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2498 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2499 $ts = wfMessage( "$weekday-at" )
2500 ->inLanguage( $this )
2501 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2503 } elseif ( $days == 1 ) {
2504 // Timestamp was yesterday: say 'yesterday' and the time.
2505 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2506 $ts = wfMessage( 'yesterday-at' )
2507 ->inLanguage( $this )
2508 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2510 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2511 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2512 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2513 $ts = wfMessage( 'today-at' )
2514 ->inLanguage( $this )
2515 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2518 // From here on in, the timestamp was soon enough ago so that we can simply say
2519 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2520 } elseif ( $diff->h
== 1 ) {
2521 // Less than 90 minutes, but more than an hour ago.
2522 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2523 } elseif ( $diff->i
>= 1 ) {
2524 // A few minutes ago.
2525 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2526 } elseif ( $diff->s
>= 30 ) {
2527 // Less than a minute, but more than 30 sec ago.
2528 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2530 // Less than 30 seconds ago.
2531 $ts = wfMessage( 'just-now' )->text();
2538 * @param string $key
2539 * @return string|null
2541 public function getMessage( $key ) {
2542 return $this->localisationCache
->getSubitem( $this->mCode
, 'messages', $key );
2548 function getAllMessages() {
2549 return $this->localisationCache
->getItem( $this->mCode
, 'messages' );
2554 * @param string $out
2555 * @param string $string
2558 public function iconv( $in, $out, $string ) {
2559 # Even with //IGNORE iconv can whine about illegal characters in
2560 # *input* string. We just ignore those too.
2561 # REF: https://bugs.php.net/bug.php?id=37166
2562 # REF: https://phabricator.wikimedia.org/T18885
2563 Wikimedia\
suppressWarnings();
2564 $text = iconv( $in, $out . '//IGNORE', $string );
2565 Wikimedia\restoreWarnings
();
2569 // callback functions for ucwords(), ucwordbreaks()
2572 * @param array $matches
2573 * @return mixed|string
2575 function ucwordbreaksCallbackAscii( $matches ) {
2576 return $this->ucfirst( $matches[1] );
2580 * @param array $matches
2583 function ucwordbreaksCallbackMB( $matches ) {
2584 return mb_strtoupper( $matches[0] );
2588 * @param array $matches
2591 function ucwordsCallbackMB( $matches ) {
2592 return mb_strtoupper( $matches[0] );
2596 * Make a string's first character uppercase
2598 * @param string $str
2602 public function ucfirst( $str ) {
2604 if ( $o < 96 ) { // if already uppercase...
2606 } elseif ( $o < 128 ) {
2607 return ucfirst( $str ); // use PHP's ucfirst()
2609 // fall back to more complex logic in case of multibyte strings
2610 return $this->uc( $str, true );
2615 * Convert a string to uppercase
2617 * @param string $str
2618 * @param bool $first
2622 public function uc( $str, $first = false ) {
2624 if ( $this->isMultibyte( $str ) ) {
2625 return $this->mbUpperChar( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2627 return ucfirst( $str );
2630 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2635 * Convert character to uppercase, allowing overrides of the default mb_upper
2636 * behaviour, which is buggy in many ways. Having a conversion table can be
2637 * useful during transitions between PHP versions where unicode changes happen.
2638 * This can make some resources unreachable on-wiki, see discussion at T219279.
2639 * Providing such a conversion table can allow to manage the transition period.
2643 * @param string $char
2647 protected function mbUpperChar( $char ) {
2648 global $wgOverrideUcfirstCharacters;
2649 if ( array_key_exists( $char, $wgOverrideUcfirstCharacters ) ) {
2650 return $wgOverrideUcfirstCharacters[$char];
2652 return mb_strtoupper( $char );
2657 * @param string $str
2658 * @return mixed|string
2660 function lcfirst( $str ) {
2663 return strval( $str );
2664 } elseif ( $o >= 128 ) {
2665 return $this->lc( $str, true );
2666 } elseif ( $o > 96 ) {
2669 $str[0] = strtolower( $str[0] );
2675 * @param string $str
2676 * @param bool $first
2677 * @return mixed|string
2679 function lc( $str, $first = false ) {
2681 if ( $this->isMultibyte( $str ) ) {
2682 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2684 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2687 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2692 * @param string $str
2695 function isMultibyte( $str ) {
2696 return strlen( $str ) !== mb_strlen( $str );
2700 * @param string $str
2701 * @return mixed|string
2703 function ucwords( $str ) {
2704 if ( $this->isMultibyte( $str ) ) {
2705 $str = $this->lc( $str );
2707 // regexp to find first letter in each word (i.e. after each space)
2708 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2710 // function to use to capitalize a single char
2711 return preg_replace_callback(
2713 [ $this, 'ucwordsCallbackMB' ],
2717 return ucwords( strtolower( $str ) );
2722 * capitalize words at word breaks
2724 * @param string $str
2727 function ucwordbreaks( $str ) {
2728 if ( $this->isMultibyte( $str ) ) {
2729 $str = $this->lc( $str );
2731 // since \b doesn't work for UTF-8, we explicitely define word break chars
2732 $breaks = "[ \-\(\)\}\{\.,\?!]";
2734 // find first letter after word break
2735 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2736 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2738 return preg_replace_callback(
2740 [ $this, 'ucwordbreaksCallbackMB' ],
2744 return preg_replace_callback(
2745 '/\b([\w\x80-\xff]+)\b/',
2746 [ $this, 'ucwordbreaksCallbackAscii' ],
2753 * Return a case-folded representation of $s
2755 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2756 * and $s2 are the same except for the case of their characters. It is not
2757 * necessary for the value returned to make sense when displayed.
2759 * Do *not* perform any other normalisation in this function. If a caller
2760 * uses this function when it should be using a more general normalisation
2761 * function, then fix the caller.
2767 function caseFold( $s ) {
2768 return $this->uc( $s );
2772 * TODO: $s is not always a string per T218883
2776 function checkTitleEncoding( $s ) {
2777 if ( is_array( $s ) ) {
2778 throw new MWException( 'Given array to checkTitleEncoding.' );
2780 if ( StringUtils
::isUtf8( $s ) ) {
2784 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2790 function fallback8bitEncoding() {
2791 return $this->localisationCache
->getItem( $this->mCode
, 'fallback8bitEncoding' );
2795 * Most writing systems use whitespace to break up words.
2796 * Some languages such as Chinese don't conventionally do this,
2797 * which requires special handling when breaking up words for
2802 function hasWordBreaks() {
2807 * Some languages such as Chinese require word segmentation,
2808 * Specify such segmentation when overridden in derived class.
2810 * @param string $string
2813 function segmentByWord( $string ) {
2818 * Some languages have special punctuation need to be normalized.
2819 * Make such changes here.
2821 * @param string $string
2824 function normalizeForSearch( $string ) {
2825 return self
::convertDoubleWidth( $string );
2829 * convert double-width roman characters to single-width.
2830 * range: ff00-ff5f ~= 0020-007f
2832 * @param string $string
2836 protected static function convertDoubleWidth( $string ) {
2837 static $full = null;
2838 static $half = null;
2840 if ( $full === null ) {
2841 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2842 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2843 $full = str_split( $fullWidth, 3 );
2844 $half = str_split( $halfWidth );
2847 $string = str_replace( $full, $half, $string );
2852 * @param string $string
2853 * @param string $pattern
2856 protected static function insertSpace( $string, $pattern ) {
2857 $string = preg_replace( $pattern, " $1 ", $string );
2858 $string = preg_replace( '/ +/', ' ', $string );
2863 * @param string[] $termsArray
2866 function convertForSearchResult( $termsArray ) {
2867 # some languages, e.g. Chinese, need to do a conversion
2868 # in order for search results to be displayed correctly
2873 * Get the first character of a string.
2878 function firstChar( $s ) {
2881 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2882 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2887 if ( isset( $matches[1] ) ) {
2888 if ( strlen( $matches[1] ) != 3 ) {
2892 // Break down Hangul syllables to grab the first jamo
2893 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
2894 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2896 } elseif ( $code < 0xb098 ) {
2898 } elseif ( $code < 0xb2e4 ) {
2900 } elseif ( $code < 0xb77c ) {
2902 } elseif ( $code < 0xb9c8 ) {
2904 } elseif ( $code < 0xbc14 ) {
2906 } elseif ( $code < 0xc0ac ) {
2908 } elseif ( $code < 0xc544 ) {
2910 } elseif ( $code < 0xc790 ) {
2912 } elseif ( $code < 0xcc28 ) {
2914 } elseif ( $code < 0xce74 ) {
2916 } elseif ( $code < 0xd0c0 ) {
2918 } elseif ( $code < 0xd30c ) {
2920 } elseif ( $code < 0xd558 ) {
2931 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2932 * also cleans up certain backwards-compatible sequences, converting them
2933 * to the modern Unicode equivalent.
2935 * This is language-specific for performance reasons only.
2941 public function normalize( $s ) {
2942 global $wgAllUnicodeFixes;
2943 $s = UtfNormal\Validator
::cleanUp( $s );
2944 if ( $wgAllUnicodeFixes ) {
2945 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
2946 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
2953 * Transform a string using serialized data stored in the given file (which
2954 * must be in the serialized subdirectory of $IP). The file contains pairs
2955 * mapping source characters to destination characters.
2957 * The data is cached in process memory. This will go faster if you have the
2958 * FastStringSearch extension.
2960 * @param string $file
2961 * @param string $string
2963 * @throws MWException
2966 protected function transformUsingPairFile( $file, $string ) {
2967 if ( !isset( $this->transformData
[$file] ) ) {
2969 $data = require "$IP/languages/data/{$file}";
2970 $this->transformData
[$file] = new ReplacementArray( $data );
2972 return $this->transformData
[$file]->replace( $string );
2976 * For right-to-left language support
2981 return $this->localisationCache
->getItem( $this->mCode
, 'rtl' );
2985 * Return the correct HTML 'dir' attribute value for this language.
2989 return $this->isRTL() ?
'rtl' : 'ltr';
2993 * Return 'left' or 'right' as appropriate alignment for line-start
2994 * for this language's text direction.
2996 * Should be equivalent to CSS3 'start' text-align value....
3000 function alignStart() {
3001 return $this->isRTL() ?
'right' : 'left';
3005 * Return 'right' or 'left' as appropriate alignment for line-end
3006 * for this language's text direction.
3008 * Should be equivalent to CSS3 'end' text-align value....
3012 function alignEnd() {
3013 return $this->isRTL() ?
'left' : 'right';
3017 * A hidden direction mark (LRM or RLM), depending on the language direction.
3018 * Unlike getDirMark(), this function returns the character as an HTML entity.
3019 * This function should be used when the output is guaranteed to be HTML,
3020 * because it makes the output HTML source code more readable. When
3021 * the output is plain text or can be escaped, getDirMark() should be used.
3023 * @param bool $opposite Get the direction mark opposite to your language
3027 function getDirMarkEntity( $opposite = false ) {
3029 return $this->isRTL() ?
'‎' : '‏';
3031 return $this->isRTL() ?
'‏' : '‎';
3035 * A hidden direction mark (LRM or RLM), depending on the language direction.
3036 * This function produces them as invisible Unicode characters and
3037 * the output may be hard to read and debug, so it should only be used
3038 * when the output is plain text or can be escaped. When the output is
3039 * HTML, use getDirMarkEntity() instead.
3041 * @param bool $opposite Get the direction mark opposite to your language
3044 function getDirMark( $opposite = false ) {
3045 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3046 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3048 return $this->isRTL() ?
$lrm : $rlm;
3050 return $this->isRTL() ?
$rlm : $lrm;
3056 function capitalizeAllNouns() {
3057 return $this->localisationCache
->getItem( $this->mCode
, 'capitalizeAllNouns' );
3061 * An arrow, depending on the language direction.
3063 * @param string $direction The direction of the arrow: forwards (default),
3064 * backwards, left, right, up, down.
3067 function getArrow( $direction = 'forwards' ) {
3068 switch ( $direction ) {
3070 return $this->isRTL() ?
'←' : '→';
3072 return $this->isRTL() ?
'→' : '←';
3085 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3089 function linkPrefixExtension() {
3090 return $this->localisationCache
->getItem( $this->mCode
, 'linkPrefixExtension' );
3094 * Get all magic words from cache.
3097 function getMagicWords() {
3098 return $this->localisationCache
->getItem( $this->mCode
, 'magicWords' );
3102 * Fill a MagicWord object with data from here
3104 * @param MagicWord $mw
3106 function getMagic( $mw ) {
3107 $rawEntry = $this->mMagicExtensions
[$mw->mId
] ??
3108 $this->localisationCache
->getSubitem( $this->mCode
, 'magicWords', $mw->mId
);
3110 if ( !is_array( $rawEntry ) ) {
3111 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3113 $mw->mCaseSensitive
= $rawEntry[0];
3114 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3119 * Add magic words to the extension array
3121 * @param array $newWords
3123 function addMagicWordsByLang( $newWords ) {
3124 $fallbackChain = $this->getFallbackLanguages();
3125 $fallbackChain = array_reverse( $fallbackChain );
3126 foreach ( $fallbackChain as $code ) {
3127 if ( isset( $newWords[$code] ) ) {
3128 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3134 * Get special page names, as an associative array
3135 * canonical name => array of valid names, including aliases
3138 function getSpecialPageAliases() {
3139 // Cache aliases because it may be slow to load them
3140 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3142 $this->mExtendedSpecialPageAliases
=
3143 $this->localisationCache
->getItem( $this->mCode
, 'specialPageAliases' );
3146 return $this->mExtendedSpecialPageAliases
;
3150 * Italic is unsuitable for some languages
3152 * @param string $text The text to be emphasized.
3155 function emphasize( $text ) {
3156 return "<em>$text</em>";
3160 * Normally we output all numbers in plain en_US style, that is
3161 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3162 * point twohundredthirtyfive. However this is not suitable for all
3163 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3164 * Icelandic just want to use commas instead of dots, and dots instead
3165 * of commas like "293.291,235".
3167 * An example of this function being called:
3169 * wfMessage( 'message' )->numParams( $num )->text()
3172 * See $separatorTransformTable on MessageIs.php for
3173 * the , => . and . => , implementation.
3175 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3176 * @param int|float $number The string to be formatted, should be an integer
3177 * or a floating point number.
3178 * @param bool $nocommafy Set to true for special numbers like dates
3181 public function formatNum( $number, $nocommafy = false ) {
3182 global $wgTranslateNumerals;
3183 if ( !$nocommafy ) {
3184 $number = $this->commafy( $number );
3185 $s = $this->separatorTransformTable();
3187 $number = strtr( $number, $s );
3191 if ( $wgTranslateNumerals ) {
3192 $s = $this->digitTransformTable();
3194 $number = strtr( $number, $s );
3198 return (string)$number;
3202 * Front-end for non-commafied formatNum
3204 * @param int|float $number The string to be formatted, should be an integer
3205 * or a floating point number.
3209 public function formatNumNoSeparators( $number ) {
3210 return $this->formatNum( $number, true );
3214 * @param string $number
3217 public function parseFormattedNumber( $number ) {
3218 $s = $this->digitTransformTable();
3220 // eliminate empty array values such as ''. (T66347)
3221 $s = array_filter( $s );
3222 $number = strtr( $number, array_flip( $s ) );
3225 $s = $this->separatorTransformTable();
3227 // eliminate empty array values such as ''. (T66347)
3228 $s = array_filter( $s );
3229 $number = strtr( $number, array_flip( $s ) );
3232 $number = strtr( $number, [ ',' => '' ] );
3237 * Adds commas to a given number
3239 * @param mixed $number
3242 function commafy( $number ) {
3243 $digitGroupingPattern = $this->digitGroupingPattern();
3244 $minimumGroupingDigits = $this->minimumGroupingDigits();
3245 if ( $number === null ) {
3249 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3250 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3251 // In some languages it's conventional not to insert a thousands separator
3252 // in numbers that are four digits long (1000-9999).
3253 if ( $minimumGroupingDigits ) {
3254 // Number of '#' characters after last comma in the grouping pattern.
3255 // The pattern is hardcoded here, but this would vary for different patterns.
3256 $primaryGroupingSize = 3;
3257 // Maximum length of a number to suppress digit grouping for.
3258 $maximumLength = $minimumGroupingDigits +
$primaryGroupingSize - 1;
3259 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3263 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3265 // Ref: http://cldr.unicode.org/translation/number-patterns
3267 if ( intval( $number ) < 0 ) {
3268 // For negative numbers apply the algorithm like positive number and add sign.
3270 $number = substr( $number, 1 );
3274 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3275 preg_match( "/\d+/", $number, $integerPart );
3276 preg_match( "/\.\d*/", $number, $decimalPart );
3277 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3278 if ( $groupedNumber === $number ) {
3279 // the string does not have any number part. Eg: .12345
3280 return $sign . $groupedNumber;
3282 $start = $end = ( $integerPart ) ?
strlen( $integerPart[0] ) : 0;
3283 while ( $start > 0 ) {
3284 $match = $matches[0][$numMatches - 1];
3285 $matchLen = strlen( $match );
3286 $start = $end - $matchLen;
3290 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3292 if ( $numMatches > 1 ) {
3293 // use the last pattern for the rest of the number
3297 $groupedNumber = "," . $groupedNumber;
3300 return $sign . $groupedNumber;
3307 function digitGroupingPattern() {
3308 return $this->localisationCache
->getItem( $this->mCode
, 'digitGroupingPattern' );
3314 function digitTransformTable() {
3315 return $this->localisationCache
->getItem( $this->mCode
, 'digitTransformTable' );
3321 function separatorTransformTable() {
3322 return $this->localisationCache
->getItem( $this->mCode
, 'separatorTransformTable' );
3328 function minimumGroupingDigits() {
3329 return $this->localisationCache
->getItem( $this->mCode
, 'minimumGroupingDigits' );
3333 * Take a list of strings and build a locale-friendly comma-separated
3334 * list, using the local comma-separator message.
3335 * The last two strings are chained with an "and".
3337 * @param string[] $list
3340 public function listToText( array $list ) {
3341 $itemCount = count( $list );
3342 if ( $itemCount < 1 ) {
3345 $text = array_pop( $list );
3346 if ( $itemCount > 1 ) {
3347 $and = $this->msg( 'and' )->escaped();
3348 $space = $this->msg( 'word-separator' )->escaped();
3350 if ( $itemCount > 2 ) {
3351 $comma = $this->msg( 'comma-separator' )->escaped();
3353 $text = implode( $comma, $list ) . $and . $space . $text;
3359 * Take a list of strings and build a locale-friendly comma-separated
3360 * list, using the local comma-separator message.
3361 * @param string[] $list Array of strings to put in a comma list
3364 function commaList( array $list ) {
3366 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3372 * Take a list of strings and build a locale-friendly semicolon-separated
3373 * list, using the local semicolon-separator message.
3374 * @param string[] $list Array of strings to put in a semicolon list
3377 function semicolonList( array $list ) {
3379 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3385 * Same as commaList, but separate it with the pipe instead.
3386 * @param string[] $list Array of strings to put in a pipe list
3389 function pipeList( array $list ) {
3391 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3397 * Truncate a string to a specified length in bytes, appending an optional
3398 * string (e.g. for ellipsis)
3400 * If $length is negative, the string will be truncated from the beginning
3404 * @param string $string String to truncate
3405 * @param int $length Maximum length in bytes
3406 * @param string $ellipsis String to append to the end of truncated text
3407 * @param bool $adjustLength Subtract length of ellipsis from $length
3411 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3412 return $this->truncateInternal(
3413 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3418 * Truncate a string to a specified number of characters, appending an optional
3419 * string (e.g. for ellipsis).
3421 * This provides multibyte version of truncateForDatabase() method of this class,
3422 * suitable for truncation based on number of characters, instead of number of bytes.
3424 * If $length is negative, the string will be truncated from the beginning.
3428 * @param string $string String to truncate
3429 * @param int $length Maximum number of characters
3430 * @param string $ellipsis String to append to the end of truncated text
3431 * @param bool $adjustLength Subtract length of ellipsis from $length
3435 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3436 // Passing encoding to mb_strlen and mb_substr is optional.
3437 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3438 // explicit specification of encoding is skipped.
3439 // Note: Both multibyte methods are callables invoked in truncateInternal.
3440 return $this->truncateInternal(
3441 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3446 * Internal method used for truncation. This method abstracts text truncation into
3447 * one common method, allowing users to provide length measurement function and
3448 * function for finding substring.
3450 * For usages, see truncateForDatabase and truncateForVisual.
3452 * @param string $string String to truncate
3453 * @param int $length Maximum length of final text
3454 * @param string $ellipsis String to append to the end of truncated text
3455 * @param bool $adjustLength Subtract length of ellipsis from $length
3456 * @param callable $measureLength Callable function used for determining the length of text
3457 * @param callable $getSubstring Callable function used for getting the substrings
3461 private function truncateInternal(
3462 $string, $length, $ellipsis, $adjustLength, callable
$measureLength, callable
$getSubstring
3464 # Check if there is no need to truncate
3465 if ( $measureLength( $string ) <= abs( $length ) ) {
3466 return $string; // no need to truncate
3469 # Use the localized ellipsis character
3470 if ( $ellipsis == '...' ) {
3471 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3473 if ( $length == 0 ) {
3474 return $ellipsis; // convention
3477 $stringOriginal = $string;
3478 # If ellipsis length is >= $length then we can't apply $adjustLength
3479 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3480 $string = $ellipsis; // this can be slightly unexpected
3481 # Otherwise, truncate and add ellipsis...
3483 $ellipsisLength = $adjustLength ?
$measureLength( $ellipsis ) : 0;
3484 if ( $length > 0 ) {
3485 $length -= $ellipsisLength;
3486 $string = $getSubstring( $string, 0, $length ); // xyz...
3487 $string = $this->removeBadCharLast( $string );
3488 $string = rtrim( $string ) . $ellipsis;
3490 $length +
= $ellipsisLength;
3491 $string = $getSubstring( $string, $length ); // ...xyz
3492 $string = $this->removeBadCharFirst( $string );
3493 $string = $ellipsis . ltrim( $string );
3497 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3498 # This check is *not* redundant if $adjustLength, due to the single case where
3499 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3500 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3503 return $stringOriginal;
3508 * Remove bytes that represent an incomplete Unicode character
3509 * at the end of string (e.g. bytes of the char are missing)
3511 * @param string $string
3514 protected function removeBadCharLast( $string ) {
3515 if ( $string != '' ) {
3516 $char = ord( $string[strlen( $string ) - 1] );
3518 if ( $char >= 0xc0 ) {
3519 # We got the first byte only of a multibyte char; remove it.
3520 $string = substr( $string, 0, -1 );
3521 } elseif ( $char >= 0x80 &&
3522 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3523 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3524 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3526 # We chopped in the middle of a character; remove it
3534 * Remove bytes that represent an incomplete Unicode character
3535 * at the start of string (e.g. bytes of the char are missing)
3537 * @param string $string
3540 protected function removeBadCharFirst( $string ) {
3541 if ( $string != '' ) {
3542 $char = ord( $string[0] );
3543 if ( $char >= 0x80 && $char < 0xc0 ) {
3544 # We chopped in the middle of a character; remove the whole thing
3545 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3552 * Truncate a string of valid HTML to a specified length in bytes,
3553 * appending an optional string (e.g. for ellipses), and return valid HTML
3555 * This is only intended for styled/linked text, such as HTML with
3556 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3557 * Also, this will not detect things like "display:none" CSS.
3559 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3561 * @param string $text HTML string to truncate
3562 * @param int $length (zero/positive) Maximum length (including ellipses)
3563 * @param string $ellipsis String to append to the truncated text
3566 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3567 # Use the localized ellipsis character
3568 if ( $ellipsis == '...' ) {
3569 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3571 # Check if there is clearly no need to truncate
3572 if ( $length <= 0 ) {
3573 return $ellipsis; // no text shown, nothing to format (convention)
3574 } elseif ( strlen( $text ) <= $length ) {
3575 return $text; // string short enough even *with* HTML (short-circuit)
3578 $dispLen = 0; // innerHTML legth so far
3579 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3580 $tagType = 0; // 0-open, 1-close
3581 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3582 $entityState = 0; // 0-not entity, 1-entity
3583 $tag = $ret = ''; // accumulated tag name, accumulated result string
3584 $openTags = []; // open tag stack
3585 $maybeState = null; // possible truncation state
3587 $textLen = strlen( $text );
3588 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3589 for ( $pos = 0; true; ++
$pos ) {
3590 # Consider truncation once the display length has reached the maximim.
3591 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3592 # Check that we're not in the middle of a bracket/entity...
3593 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3594 if ( !$testingEllipsis ) {
3595 $testingEllipsis = true;
3596 # Save where we are; we will truncate here unless there turn out to
3597 # be so few remaining characters that truncation is not necessary.
3598 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3599 $maybeState = [ $ret, $openTags ]; // save state
3601 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3602 # String in fact does need truncation, the truncation point was OK.
3603 // @phan-suppress-next-line PhanTypeInvalidExpressionArrayDestructuring
3604 list( $ret, $openTags ) = $maybeState; // reload state
3605 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3606 $ret .= $ellipsis; // add ellipsis
3610 if ( $pos >= $textLen ) {
3611 break; // extra iteration just for above checks
3614 # Read the next char...
3616 $lastCh = $pos ?
$text[$pos - 1] : '';
3617 $ret .= $ch; // add to result string
3619 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3620 $entityState = 0; // for bad HTML
3621 $bracketState = 1; // tag started (checking for backslash)
3622 } elseif ( $ch == '>' ) {
3623 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3624 $entityState = 0; // for bad HTML
3625 $bracketState = 0; // out of brackets
3626 } elseif ( $bracketState == 1 ) {
3628 $tagType = 1; // close tag (e.g. "</span>")
3630 $tagType = 0; // open tag (e.g. "<span>")
3633 $bracketState = 2; // building tag name
3634 } elseif ( $bracketState == 2 ) {
3638 // Name found (e.g. "<a href=..."), add on tag attributes...
3639 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3641 } elseif ( $bracketState == 0 ) {
3642 if ( $entityState ) {
3645 $dispLen++
; // entity is one displayed char
3648 if ( $neLength == 0 && !$maybeState ) {
3649 // Save state without $ch. We want to *hit* the first
3650 // display char (to get tags) but not *use* it if truncating.
3651 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3654 $entityState = 1; // entity found, (e.g. " ")
3656 $dispLen++
; // this char is displayed
3657 // Add the next $max display text chars after this in one swoop...
3658 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3659 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3660 $dispLen +
= $skipped;
3666 // Close the last tag if left unclosed by bad HTML
3667 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3668 while ( count( $openTags ) > 0 ) {
3669 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3675 * truncateHtml() helper function
3676 * like strcspn() but adds the skipped chars to $ret
3678 * @param string $ret
3679 * @param string $text
3680 * @param string $search
3682 * @param null|int $len
3685 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3686 if ( $len === null ) {
3687 $len = -1; // -1 means "no limit" for strcspn
3688 } elseif ( $len < 0 ) {
3692 if ( $start < strlen( $text ) ) {
3693 $skipCount = strcspn( $text, $search, $start, $len );
3694 $ret .= substr( $text, $start, $skipCount );
3700 * truncateHtml() helper function
3701 * (a) push or pop $tag from $openTags as needed
3702 * (b) clear $tag value
3703 * @param string &$tag Current HTML tag name we are looking at
3704 * @param int $tagType (0-open tag, 1-close tag)
3705 * @param string $lastCh Character before the '>' that ended this tag
3706 * @param array &$openTags Open tag stack (not accounting for $tag)
3708 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3709 $tag = ltrim( $tag );
3711 if ( $tagType == 0 && $lastCh != '/' ) {
3712 $openTags[] = $tag; // tag opened (didn't close itself)
3713 } elseif ( $tagType == 1 ) {
3714 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3715 array_pop( $openTags ); // tag closed
3723 * Grammatical transformations, needed for inflected languages
3724 * Invoked by putting {{grammar:case|word}} in a message
3726 * @param string $word
3727 * @param string $case
3730 function convertGrammar( $word, $case ) {
3731 global $wgGrammarForms;
3732 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3733 return $wgGrammarForms[$this->getCode()][$case][$word];
3736 $grammarTransformations = $this->getGrammarTransformations();
3738 if ( isset( $grammarTransformations[$case] ) ) {
3739 $forms = $grammarTransformations[$case];
3741 // Some names of grammar rules are aliases for other rules.
3742 // In such cases the value is a string rather than object,
3743 // so load the actual rules.
3744 if ( is_string( $forms ) ) {
3745 $forms = $grammarTransformations[$forms];
3748 foreach ( array_values( $forms ) as $rule ) {
3751 if ( $form === '@metadata' ) {
3755 $replacement = $rule[1];
3757 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3758 $patternMatches = preg_match( $regex, $word );
3760 if ( $patternMatches === false ) {
3762 'An error occurred while processing grammar. ' .
3763 "Word: '$word'. Regex: /$form/."
3765 } elseif ( $patternMatches === 1 ) {
3766 $word = preg_replace( $regex, $replacement, $word );
3777 * Get the grammar forms for the content language
3778 * @return array Array of grammar forms
3781 function getGrammarForms() {
3782 global $wgGrammarForms;
3783 if ( isset( $wgGrammarForms[$this->getCode()] )
3784 && is_array( $wgGrammarForms[$this->getCode()] )
3786 return $wgGrammarForms[$this->getCode()];
3793 * Get the grammar transformations data for the language.
3794 * Used like grammar forms, with {{GRAMMAR}} and cases,
3795 * but uses pairs of regexes and replacements instead of code.
3797 * @return array[] Array of grammar transformations.
3798 * @throws MWException
3801 public function getGrammarTransformations() {
3802 $languageCode = $this->getCode();
3804 if ( self
::$grammarTransformations === null ) {
3805 self
::$grammarTransformations = new MapCacheLRU( 10 );
3808 if ( self
::$grammarTransformations->has( $languageCode ) ) {
3809 return self
::$grammarTransformations->get( $languageCode );
3814 $grammarDataFile = __DIR__
. "/data/grammarTransformations/$languageCode.json";
3815 if ( is_readable( $grammarDataFile ) ) {
3816 $data = FormatJson
::decode(
3817 file_get_contents( $grammarDataFile ),
3821 if ( $data === null ) {
3822 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3825 self
::$grammarTransformations->set( $languageCode, $data );
3832 * Provides an alternative text depending on specified gender.
3833 * Usage {{gender:username|masculine|feminine|unknown}}.
3834 * username is optional, in which case the gender of current user is used,
3835 * but only in (some) interface messages; otherwise default gender is used.
3837 * If no forms are given, an empty string is returned. If only one form is
3838 * given, it will be returned unconditionally. These details are implied by
3839 * the caller and cannot be overridden in subclasses.
3841 * If three forms are given, the default is to use the third (unknown) form.
3842 * If fewer than three forms are given, the default is to use the first (masculine) form.
3843 * These details can be overridden in subclasses.
3845 * @param string $gender
3846 * @param array $forms
3850 function gender( $gender, $forms ) {
3851 if ( !count( $forms ) ) {
3854 $forms = $this->preConvertPlural( $forms, 2 );
3855 if ( $gender === 'male' ) {
3858 if ( $gender === 'female' ) {
3861 return $forms[2] ??
$forms[0];
3865 * Plural form transformations, needed for some languages.
3866 * For example, there are 3 form of plural in Russian and Polish,
3867 * depending on "count mod 10". See [[w:Plural]]
3868 * For English it is pretty simple.
3870 * Invoked by putting {{plural:count|wordform1|wordform2}}
3871 * or {{plural:count|wordform1|wordform2|wordform3}}
3873 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3875 * @param int $count Non-localized number
3876 * @param array $forms Different plural forms
3877 * @return string Correct form of plural for $count in this language
3879 function convertPlural( $count, $forms ) {
3880 // Handle explicit n=pluralform cases
3881 $forms = $this->handleExplicitPluralForms( $count, $forms );
3882 if ( is_string( $forms ) ) {
3885 if ( !count( $forms ) ) {
3889 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3890 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3891 return $forms[$pluralForm];
3895 * Handles explicit plural forms for Language::convertPlural()
3897 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3898 * If an explicitly defined plural form matches the $count, then
3899 * string value returned, otherwise array returned for further consideration
3900 * by CLDR rules or overridden convertPlural().
3904 * @param int $count Non-localized number
3905 * @param array $forms Different plural forms
3907 * @return array|string
3909 protected function handleExplicitPluralForms( $count, array $forms ) {
3910 foreach ( $forms as $index => $form ) {
3911 if ( preg_match( '/\d+=/i', $form ) ) {
3912 $pos = strpos( $form, '=' );
3913 if ( substr( $form, 0, $pos ) === (string)$count ) {
3914 return substr( $form, $pos +
1 );
3916 unset( $forms[$index] );
3919 return array_values( $forms );
3923 * Checks that convertPlural was given an array and pads it to requested
3924 * amount of forms by copying the last one.
3926 * @param array $forms
3927 * @param int $count How many forms should there be at least
3928 * @return array Padded array of forms
3930 protected function preConvertPlural( /* Array */ $forms, $count ) {
3931 return array_pad( $forms, $count, end( $forms ) );
3935 * Wraps argument with unicode control characters for directionality safety
3937 * This solves the problem where directionality-neutral characters at the edge of
3938 * the argument string get interpreted with the wrong directionality from the
3939 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3941 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3942 * directionality of the argument string, using the BIDI algorithm's own "First
3943 * strong directional codepoint" rule. Essentially, this works round the fact that
3944 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3945 * direction inference). The latter is cleaner but still not widely supported.
3947 * @param string $text Text to wrap
3948 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3950 public function embedBidi( $text = '' ) {
3951 $dir = self
::strongDirFromContent( $text );
3952 if ( $dir === 'ltr' ) {
3953 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3954 return self
::$lre . $text . self
::$pdf;
3956 if ( $dir === 'rtl' ) {
3957 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3958 return self
::$rle . $text . self
::$pdf;
3960 // No strong directionality: do not wrap
3965 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3966 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3967 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3968 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3971 * @param string $str The validated block duration in English
3972 * @param User|null $user User object to use timezone from or null for $wgUser
3973 * @param int $now Current timestamp, for formatting relative block durations
3974 * @return string Somehow translated block duration
3975 * @see LanguageFi.php for example implementation
3977 function translateBlockExpiry( $str, User
$user = null, $now = 0 ) {
3978 $duration = SpecialBlock
::getSuggestedDurations( $this );
3979 foreach ( $duration as $show => $value ) {
3980 if ( strcmp( $str, $value ) == 0 ) {
3981 return htmlspecialchars( trim( $show ) );
3985 if ( wfIsInfinity( $str ) ) {
3986 foreach ( $duration as $show => $value ) {
3987 if ( wfIsInfinity( $value ) ) {
3988 return htmlspecialchars( trim( $show ) );
3993 // If all else fails, return a standard duration or timestamp description.
3994 $time = strtotime( $str, $now );
3995 if ( $time === false ) { // Unknown format. Return it as-is in case.
3997 } elseif ( $time !== strtotime( $str, $now +
1 ) ) { // It's a relative timestamp.
3998 // The result differs based on current time, so the difference
3999 // is a fixed duration length.
4000 return $this->formatDuration( $time - $now );
4001 } else { // It's an absolute timestamp.
4002 if ( $time === 0 ) {
4003 // wfTimestamp() handles 0 as current time instead of epoch.
4004 $time = '19700101000000';
4007 return $this->userTimeAndDate( $time, $user );
4009 return $this->timeanddate( $time );
4014 * languages like Chinese need to be segmented in order for the diff
4017 * @param string $text
4020 public function segmentForDiff( $text ) {
4025 * and unsegment to show the result
4027 * @param string $text
4030 public function unsegmentForDiff( $text ) {
4035 * Return the LanguageConverter used in the Language
4038 * @return LanguageConverter
4040 public function getConverter() {
4041 return $this->mConverter
;
4045 * convert text to a variant
4047 * @param string $text text to convert
4048 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4049 * variant (if logged in), or the project default variant
4050 * @return string the converted string
4052 public function autoConvert( $text, $variant = false ) {
4053 return $this->mConverter
->autoConvert( $text, $variant );
4057 * convert text to all supported variants
4059 * @param string $text
4062 public function autoConvertToAllVariants( $text ) {
4063 return $this->mConverter
->autoConvertToAllVariants( $text );
4067 * convert text to different variants of a language.
4069 * @warning Glossary state is maintained between calls. This means
4070 * if you pass unescaped text to this method it can cause an XSS
4071 * in later calls to this method, even if the later calls have properly
4072 * escaped the input. Never feed this method user controlled text that
4073 * is not properly escaped!
4074 * @param string $text Content that has been already escaped for use in HTML
4075 * @return string HTML
4077 public function convert( $text ) {
4078 return $this->mConverter
->convert( $text );
4082 * Convert a Title object to a string in the preferred variant
4084 * @param Title $title
4087 public function convertTitle( $title ) {
4088 return $this->mConverter
->convertTitle( $title );
4092 * Convert a namespace index to a string in the preferred variant
4094 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4095 * @param string|null $variant variant to convert to, or null to use the user's preferred
4096 * variant (if logged in), or the project default variant
4097 * @return string a string representation of the namespace
4099 public function convertNamespace( $ns, $variant = null ) {
4100 return $this->mConverter
->convertNamespace( $ns, $variant );
4104 * Check if this is a language with variants
4108 public function hasVariants() {
4109 return count( $this->getVariants() ) > 1;
4113 * Strict check if the language has the specific variant.
4115 * Compare to LanguageConverter::validateVariant() which does a more
4116 * lenient check and attempts to coerce the given code to a valid one.
4119 * @param string $variant
4122 public function hasVariant( $variant ) {
4123 return $variant && ( $variant === $this->mConverter
->validateVariant( $variant ) );
4127 * Perform output conversion on a string, and encode for safe HTML output.
4128 * @param string $text Text to be converted
4130 * @todo this should get integrated somewhere sane
4132 public function convertHtml( $text ) {
4133 return htmlspecialchars( $this->convert( $text ) );
4137 * @param string $key
4140 public function convertCategoryKey( $key ) {
4141 return $this->mConverter
->convertCategoryKey( $key );
4145 * Get the list of variants supported by this language
4146 * see sample implementation in LanguageZh.php
4148 * @return string[] An array of language codes
4150 public function getVariants() {
4151 return $this->mConverter
->getVariants();
4157 public function getPreferredVariant() {
4158 return $this->mConverter
->getPreferredVariant();
4164 public function getDefaultVariant() {
4165 return $this->mConverter
->getDefaultVariant();
4171 public function getURLVariant() {
4172 return $this->mConverter
->getURLVariant();
4176 * If a language supports multiple variants, it is
4177 * possible that non-existing link in one variant
4178 * actually exists in another variant. this function
4179 * tries to find it. See e.g. LanguageZh.php
4180 * The input parameters may be modified upon return
4182 * @param string &$link The name of the link
4183 * @param Title &$nt The title object of the link
4184 * @param bool $ignoreOtherCond To disable other conditions when
4185 * we need to transclude a template or update a category's link
4187 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4188 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4192 * returns language specific options used by User::getPageRenderHash()
4193 * for example, the preferred language variant
4197 function getExtraHashOptions() {
4198 return $this->mConverter
->getExtraHashOptions();
4202 * For languages that support multiple variants, the title of an
4203 * article may be displayed differently in different variants. this
4204 * function returns the apporiate title defined in the body of the article.
4208 public function getParsedTitle() {
4209 return $this->mConverter
->getParsedTitle();
4213 * Refresh the cache of conversion tables when
4214 * MediaWiki:Conversiontable* is updated.
4216 * @param Title $title The Title of the page being updated
4218 public function updateConversionTable( Title
$title ) {
4219 $this->mConverter
->updateConversionTable( $title );
4223 * A regular expression to match legal word-trailing characters
4224 * which should be merged onto a link of the form [[foo]]bar.
4228 public function linkTrail() {
4229 return $this->localisationCache
->getItem( $this->mCode
, 'linkTrail' );
4233 * A regular expression character set to match legal word-prefixing
4234 * characters which should be merged onto a link of the form foo[[bar]].
4238 public function linkPrefixCharset() {
4239 return $this->localisationCache
->getItem( $this->mCode
, 'linkPrefixCharset' );
4243 * Get the "parent" language which has a converter to convert a "compatible" language
4244 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4246 * @return Language|null
4249 public function getParentLanguage() {
4250 if ( $this->mParentLanguage
!== false ) {
4251 return $this->mParentLanguage
;
4254 $code = explode( '-', $this->getCode() )[0];
4255 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4256 $this->mParentLanguage
= null;
4259 $lang = self
::factory( $code );
4260 if ( !$lang->hasVariant( $this->getCode() ) ) {
4261 $this->mParentLanguage
= null;
4265 $this->mParentLanguage
= $lang;
4270 * Compare with an other language object
4273 * @param Language $lang
4276 public function equals( Language
$lang ) {
4277 return $lang === $this ||
$lang->getCode() === $this->mCode
;
4281 * Get the internal language code for this language object
4283 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4284 * htmlspecialchars() or similar
4288 public function getCode() {
4289 return $this->mCode
;
4293 * Get the code in BCP 47 format which we can use
4294 * inside of html lang="" tags.
4296 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4297 * htmlspecialchars() or similar.
4302 public function getHtmlCode() {
4303 if ( is_null( $this->mHtmlCode
) ) {
4304 $this->mHtmlCode
= LanguageCode
::bcp47( $this->getCode() );
4306 return $this->mHtmlCode
;
4310 * Get the language code from a file name. Inverse of getFileName()
4311 * @param string $filename $prefix . $languageCode . $suffix
4312 * @param string $prefix Prefix before the language code
4313 * @param string $suffix Suffix after the language code
4314 * @return string Language code, or false if $prefix or $suffix isn't found
4316 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4318 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4319 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4320 if ( !count( $m ) ) {
4323 return str_replace( '_', '-', strtolower( $m[1] ) );
4327 * @param string $code
4328 * @param bool $fallback Whether we're going through language fallback chain
4329 * @return string Name of the language class
4331 public static function classFromCode( $code, $fallback = true ) {
4332 if ( $fallback && $code == 'en' ) {
4335 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4340 * Get the name of a file for a certain language code
4342 * @deprecated since 1.34, use LanguageNameUtils
4343 * @param string $prefix Prepend this to the filename
4344 * @param string $code Language code
4345 * @param string $suffix Append this to the filename
4346 * @throws MWException
4347 * @return string $prefix . $mangledCode . $suffix
4349 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4350 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
4351 ->getFileName( $prefix, $code, $suffix );
4355 * @deprecated since 1.34, use LanguageNameUtils
4356 * @param string $code
4359 public static function getMessagesFileName( $code ) {
4360 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
4361 ->getMessagesFileName( $code );
4365 * @deprecated since 1.34, use LanguageNameUtils
4366 * @param string $code
4368 * @throws MWException
4371 public static function getJsonMessagesFileName( $code ) {
4372 return MediaWikiServices
::getInstance()->getLanguageNameUtils()
4373 ->getJsonMessagesFileName( $code );
4377 * Get the first fallback for a given language.
4379 * @param string $code
4381 * @return bool|string
4383 public static function getFallbackFor( $code ) {
4384 $fallbacks = self
::getFallbacksFor( $code );
4386 return $fallbacks[0];
4392 * Get the ordered list of fallback languages.
4395 * @param string $code Language code
4396 * @param int $mode Fallback mode, either MESSAGES_FALLBACKS (which always falls back to 'en'),
4397 * or STRICT_FALLBACKS (whic honly falls back to 'en' when explicitly defined)
4398 * @throws MWException
4399 * @return array List of language codes
4401 public static function getFallbacksFor( $code, $mode = self
::MESSAGES_FALLBACKS
) {
4402 if ( $code === 'en' ||
!self
::isValidBuiltInCode( $code ) ) {
4406 case self
::MESSAGES_FALLBACKS
:
4407 // For unknown languages, fallbackSequence returns an empty array,
4408 // hardcode fallback to 'en' in that case as English messages are
4410 return self
::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?
: [ 'en' ];
4411 case self
::STRICT_FALLBACKS
:
4412 // Use this mode when you don't want to fallback to English unless
4413 // explicitly defined, for example when you have language-variant icons
4414 // and an international language-independent fallback.
4415 return self
::getLocalisationCache()->getItem( $code, 'originalFallbackSequence' );
4417 throw new MWException( "Invalid fallback mode \"$mode\"" );
4422 * Get the ordered list of fallback languages, ending with the fallback
4423 * language chain for the site language.
4426 * @param string $code Language code
4427 * @return array [ fallbacks, site fallbacks ]
4429 public static function getFallbacksIncludingSiteLanguage( $code ) {
4430 global $wgLanguageCode;
4432 // Usually, we will only store a tiny number of fallback chains, so we
4433 // keep them in static memory.
4434 $cacheKey = "{$code}-{$wgLanguageCode}";
4436 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4437 $fallbacks = self
::getFallbacksFor( $code );
4439 // Append the site's fallback chain, including the site language itself
4440 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4441 array_unshift( $siteFallbacks, $wgLanguageCode );
4443 // Eliminate any languages already included in the chain
4444 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4446 self
::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4448 return self
::$fallbackLanguageCache[$cacheKey];
4452 * Get all messages for a given language
4453 * WARNING: this may take a long time. If you just need all message *keys*
4454 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4456 * @param string $code
4460 public static function getMessagesFor( $code ) {
4461 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4465 * Get a message for a given language
4467 * @param string $key
4468 * @param string $code
4472 public static function getMessageFor( $key, $code ) {
4473 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4477 * Get all message keys for a given language. This is a faster alternative to
4478 * array_keys( Language::getMessagesFor( $code ) )
4481 * @param string $code Language code
4482 * @return array Array of message keys (strings)
4484 public static function getMessageKeysFor( $code ) {
4485 return self
::getLocalisationCache()->getSubitemList( $code, 'messages' );
4489 * @param string $talk
4492 function fixVariableInNamespace( $talk ) {
4493 if ( strpos( $talk, '$1' ) === false ) {
4497 global $wgMetaNamespace;
4498 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4500 # Allow grammar transformations
4501 # Allowing full message-style parsing would make simple requests
4502 # such as action=raw much more expensive than they need to be.
4503 # This will hopefully cover most cases.
4504 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4505 [ $this, 'replaceGrammarInNamespace' ], $talk );
4506 return str_replace( ' ', '_', $talk );
4513 function replaceGrammarInNamespace( $m ) {
4514 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4518 * Decode an expiry (block, protection, etc) which has come from the DB
4520 * @param string $expiry Database expiry String
4521 * @param bool|int $format True to process using language functions, or TS_ constant
4522 * to return the expiry in a given timestamp
4523 * @param string $infinity If $format is not true, use this string for infinite expiry
4527 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4529 if ( $dbInfinity === null ) {
4530 $dbInfinity = wfGetDB( DB_REPLICA
)->getInfinity();
4533 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4534 return $format === true
4535 ?
$this->getMessageFromDB( 'infiniteblock' )
4538 return $format === true
4539 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4540 : wfTimestamp( $format, $expiry );
4545 * Formats a time given in seconds into a string representation of that time.
4547 * @param int|float $seconds
4548 * @param array $format An optional argument that formats the returned string in different ways:
4549 * If $format['avoid'] === 'avoidhours': don't show hours, just show days
4550 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4551 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4552 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4554 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4555 * or 'avoidminutes'.
4558 function formatTimePeriod( $seconds, $format = [] ) {
4559 if ( !is_array( $format ) ) {
4560 $format = [ 'avoid' => $format ]; // For backwards compatibility
4562 if ( !isset( $format['avoid'] ) ) {
4563 $format['avoid'] = false;
4565 if ( !isset( $format['noabbrevs'] ) ) {
4566 $format['noabbrevs'] = false;
4568 $secondsMsg = wfMessage(
4569 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4570 $minutesMsg = wfMessage(
4571 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4572 $hoursMsg = wfMessage(
4573 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4574 $daysMsg = wfMessage(
4575 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4577 if ( round( $seconds * 10 ) < 100 ) {
4578 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4579 $s = $secondsMsg->params( $s )->text();
4580 } elseif ( round( $seconds ) < 60 ) {
4581 $s = $this->formatNum( round( $seconds ) );
4582 $s = $secondsMsg->params( $s )->text();
4583 } elseif ( round( $seconds ) < 3600 ) {
4584 $minutes = floor( $seconds / 60 );
4585 $secondsPart = round( fmod( $seconds, 60 ) );
4586 if ( $secondsPart == 60 ) {
4590 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4592 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4593 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4594 $hours = floor( $seconds / 3600 );
4595 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4596 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4597 if ( $secondsPart == 60 ) {
4601 if ( $minutes == 60 ) {
4605 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4607 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4608 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes', 'avoidhours' ] ) ) {
4609 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4612 $days = floor( $seconds / 86400 );
4613 if ( $format['avoid'] === 'avoidhours' ) {
4614 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4615 if ( $hours == 24 ) {
4619 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4620 } elseif ( $format['avoid'] === 'avoidminutes' ) {
4621 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4622 if ( $hours == 24 ) {
4626 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4628 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4629 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4630 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4631 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4632 if ( $minutes == 60 ) {
4636 if ( $hours == 24 ) {
4640 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4642 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4644 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4646 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4648 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4655 * Format a bitrate for output, using an appropriate
4656 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4657 * the magnitude in question.
4659 * This use base 1000. For base 1024 use formatSize(), for another base
4660 * see formatComputingNumbers().
4665 function formatBitrate( $bps ) {
4666 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4670 * @param int $size Size of the unit
4671 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4672 * @param string $messageKey Message key to be uesd
4675 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4677 return str_replace( '$1', $this->formatNum( $size ),
4678 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4681 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4684 $maxIndex = count( $sizes ) - 1;
4685 while ( $size >= $boundary && $index < $maxIndex ) {
4690 // For small sizes no decimal places necessary
4693 // For MB and bigger two decimal places are smarter
4696 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4698 $size = round( $size, $round );
4699 $text = $this->getMessageFromDB( $msg );
4700 return str_replace( '$1', $this->formatNum( $size ), $text );
4704 * Format a size in bytes for output, using an appropriate
4705 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4707 * This method use base 1024. For base 1000 use formatBitrate(), for
4708 * another base see formatComputingNumbers()
4710 * @param int $size Size to format
4711 * @return string Plain text (not HTML)
4713 function formatSize( $size ) {
4714 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4718 * Make a list item, used by various special pages
4720 * @param string $page Page link
4721 * @param string $details HTML safe text between brackets
4722 * @param bool $oppositedm Add the direction mark opposite to your
4723 * language, to display text properly
4724 * @return string HTML escaped
4726 function specialList( $page, $details, $oppositedm = true ) {
4731 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4734 $this->msg( 'word-separator' )->escaped() .
4735 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4739 * Generate (prev x| next x) (20|50|100...) type links for paging
4741 * @param Title $title Title object to link
4742 * @param int $offset
4744 * @param array $query Optional URL query parameter string
4745 * @param bool $atend Optional param for specified if this is the last page
4747 * @deprecated since 1.34, use PrevNextNavigationRenderer::buildPrevNextNavigation()
4750 public function viewPrevNext( Title
$title, $offset, $limit,
4751 array $query = [], $atend = false
4753 wfDeprecated( __METHOD__
, '1.34' );
4754 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4756 # Make 'previous' link
4757 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4758 if ( $offset > 0 ) {
4759 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4760 $query, $prev, 'prevn-title', 'mw-prevlink' );
4762 $plink = htmlspecialchars( $prev );
4766 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4768 $nlink = htmlspecialchars( $next );
4770 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4771 $query, $next, 'nextn-title', 'mw-nextlink' );
4774 # Make links to set number of items per page
4776 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4777 $numLinks[] = $this->numLink( $title, $offset, $num,
4778 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4781 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4782 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4786 * Helper function for viewPrevNext() that generates links
4788 * @param Title $title Title object to link
4789 * @param int $offset
4791 * @param array $query Extra query parameters
4792 * @param string $link Text to use for the link; will be escaped
4793 * @param string $tooltipMsg Name of the message to use as tooltip
4794 * @param string $class Value of the "class" attribute of the link
4795 * @return string HTML fragment
4797 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4800 $query = [ 'limit' => $limit, 'offset' => $offset ] +
$query;
4801 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4802 ->numParams( $limit )->text();
4804 return Html
::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4805 'title' => $tooltip, 'class' => $class ], $link );
4809 * Get the conversion rule title, if any.
4813 public function getConvRuleTitle() {
4814 return $this->mConverter
->getConvRuleTitle();
4818 * Get the compiled plural rules for the language
4820 * @return array Associative array with plural form, and plural rule as key-value pairs
4822 public function getCompiledPluralRules() {
4824 $this->localisationCache
->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4825 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4826 if ( !$pluralRules ) {
4827 foreach ( $fallbacks as $fallbackCode ) {
4828 $pluralRules = $this->localisationCache
4829 ->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4830 if ( $pluralRules ) {
4835 return $pluralRules;
4839 * Get the plural rules for the language
4841 * @return array Associative array with plural form number and plural rule as key-value pairs
4843 public function getPluralRules() {
4845 $this->localisationCache
->getItem( strtolower( $this->mCode
), 'pluralRules' );
4846 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4847 if ( !$pluralRules ) {
4848 foreach ( $fallbacks as $fallbackCode ) {
4849 $pluralRules = $this->localisationCache
4850 ->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4851 if ( $pluralRules ) {
4856 return $pluralRules;
4860 * Get the plural rule types for the language
4862 * @return array Associative array with plural form number and plural rule type as key-value pairs
4864 public function getPluralRuleTypes() {
4866 $this->localisationCache
->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4867 $fallbacks = self
::getFallbacksFor( $this->mCode
);
4868 if ( !$pluralRuleTypes ) {
4869 foreach ( $fallbacks as $fallbackCode ) {
4870 $pluralRuleTypes = $this->localisationCache
4871 ->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4872 if ( $pluralRuleTypes ) {
4877 return $pluralRuleTypes;
4881 * Find the index number of the plural rule appropriate for the given number
4882 * @param int $number
4883 * @return int The index number of the plural rule
4885 public function getPluralRuleIndexNumber( $number ) {
4886 $pluralRules = $this->getCompiledPluralRules();
4887 $form = Evaluator
::evaluateCompiled( $number, $pluralRules );
4892 * Find the plural rule type appropriate for the given number
4893 * For example, if the language is set to Arabic, getPluralType(5) should
4896 * @param int $number
4897 * @return string The name of the plural rule type, e.g. one, two, few, many
4899 public function getPluralRuleType( $number ) {
4900 $index = $this->getPluralRuleIndexNumber( $number );
4901 $pluralRuleTypes = $this->getPluralRuleTypes();
4902 return $pluralRuleTypes[$index] ??
'other';