?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Make MessageCache::load() require a language code
[mediawiki.git] / languages / Language.php
blob7ef2effb985cce462196fa0c280e3871dee61a95
1 <?php
2 /**
3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
5 *
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.
10 *
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.
15 *
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
20 *
21 * @file
22 * @ingroup Language
23 */
24
25 /**
26 * @defgroup Language Language
27 */
28
29 use CLDRPluralRuleParser\Evaluator;
30
31 /**
32 * Internationalisation code
33 * @ingroup Language
34 */
35 class Language {
36 /**
37 * @var LanguageConverter
38 */
39 public $mConverter;
40
41 public $mVariants, $mCode, $mLoaded = false;
42 public $mMagicExtensions = [], $mMagicHookDone = false;
43 private $mHtmlCode = null, $mParentLanguage = false;
44
45 public $dateFormatStrings = [];
46 public $mExtendedSpecialPageAliases;
47
48 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
49
50 /**
51 * ReplacementArray object caches
52 */
53 public $transformData = [];
54
55 /**
56 * @var LocalisationCache
57 */
58 static public $dataCache;
59
60 static public $mLangObjCache = [];
61
62 static public $mWeekdayMsgs = [
63 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
64 'friday', 'saturday'
65 ];
66
67 static public $mWeekdayAbbrevMsgs = [
68 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
69 ];
70
71 static public $mMonthMsgs = [
72 'january', 'february', 'march', 'april', 'may_long', 'june',
73 'july', 'august', 'september', 'october', 'november',
74 'december'
75 ];
76 static public $mMonthGenMsgs = [
77 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
78 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
79 'december-gen'
80 ];
81 static public $mMonthAbbrevMsgs = [
82 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
83 'sep', 'oct', 'nov', 'dec'
84 ];
85
86 static public $mIranianCalendarMonthMsgs = [
87 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
88 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
89 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
90 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
91 ];
92
93 static public $mHebrewCalendarMonthMsgs = [
94 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
95 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
96 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
97 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
98 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
99 ];
100
101 static public $mHebrewCalendarMonthGenMsgs = [
102 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
103 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
104 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
105 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
106 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
107 ];
108
109 static public $mHijriCalendarMonthMsgs = [
110 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
111 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
112 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
113 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
114 ];
115
116 /**
117 * @since 1.20
118 * @var array
119 */
120 static public $durationIntervals = [
121 'millennia' => 31556952000,
122 'centuries' => 3155695200,
123 'decades' => 315569520,
124 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
125 'weeks' => 604800,
126 'days' => 86400,
127 'hours' => 3600,
128 'minutes' => 60,
129 'seconds' => 1,
130 ];
131
132 /**
133 * Cache for language fallbacks.
134 * @see Language::getFallbacksIncludingSiteLanguage
135 * @since 1.21
136 * @var array
137 */
138 static private $fallbackLanguageCache = [];
139
140 /**
141 * Cache for language names
142 * @var HashBagOStuff|null
143 */
144 static private $languageNameCache;
145
146 /**
147 * Unicode directional formatting characters, for embedBidi()
148 */
149 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
150 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
151 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
152
153 /**
154 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
155 * - in group 1 if it is LTR
156 * - in group 2 if it is RTL
157 * Does not match if there is no strong directionality codepoint.
158 *
159 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
160 *
161 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
162 * https://git.wikimedia.org/summary/unicodejs.git .
163 */
164 // @codingStandardsIgnoreStart
165 // @codeCoverageIgnoreStart
166 static private $strongDirRegex = '/(?:([\x{41}-\x{5a}\x{61}-\x{7a}\x{aa}\x{b5}\x{ba}\x{c0}-\x{d6}\x{d8}-\x{f6}\x{f8}-\x{2b8}\x{2bb}-\x{2c1}\x{2d0}\x{2d1}\x{2e0}-\x{2e4}\x{2ee}\x{370}-\x{373}\x{376}\x{377}\x{37a}-\x{37d}\x{37f}\x{386}\x{388}-\x{38a}\x{38c}\x{38e}-\x{3a1}\x{3a3}-\x{3f5}\x{3f7}-\x{482}\x{48a}-\x{52f}\x{531}-\x{556}\x{559}-\x{55f}\x{561}-\x{587}\x{589}\x{903}-\x{939}\x{93b}\x{93d}-\x{940}\x{949}-\x{94c}\x{94e}-\x{950}\x{958}-\x{961}\x{964}-\x{980}\x{982}\x{983}\x{985}-\x{98c}\x{98f}\x{990}\x{993}-\x{9a8}\x{9aa}-\x{9b0}\x{9b2}\x{9b6}-\x{9b9}\x{9bd}-\x{9c0}\x{9c7}\x{9c8}\x{9cb}\x{9cc}\x{9ce}\x{9d7}\x{9dc}\x{9dd}\x{9df}-\x{9e1}\x{9e6}-\x{9f1}\x{9f4}-\x{9fa}\x{a03}\x{a05}-\x{a0a}\x{a0f}\x{a10}\x{a13}-\x{a28}\x{a2a}-\x{a30}\x{a32}\x{a33}\x{a35}\x{a36}\x{a38}\x{a39}\x{a3e}-\x{a40}\x{a59}-\x{a5c}\x{a5e}\x{a66}-\x{a6f}\x{a72}-\x{a74}\x{a83}\x{a85}-\x{a8d}\x{a8f}-\x{a91}\x{a93}-\x{aa8}\x{aaa}-\x{ab0}\x{ab2}\x{ab3}\x{ab5}-\x{ab9}\x{abd}-\x{ac0}\x{ac9}\x{acb}\x{acc}\x{ad0}\x{ae0}\x{ae1}\x{ae6}-\x{af0}\x{af9}\x{b02}\x{b03}\x{b05}-\x{b0c}\x{b0f}\x{b10}\x{b13}-\x{b28}\x{b2a}-\x{b30}\x{b32}\x{b33}\x{b35}-\x{b39}\x{b3d}\x{b3e}\x{b40}\x{b47}\x{b48}\x{b4b}\x{b4c}\x{b57}\x{b5c}\x{b5d}\x{b5f}-\x{b61}\x{b66}-\x{b77}\x{b83}\x{b85}-\x{b8a}\x{b8e}-\x{b90}\x{b92}-\x{b95}\x{b99}\x{b9a}\x{b9c}\x{b9e}\x{b9f}\x{ba3}\x{ba4}\x{ba8}-\x{baa}\x{bae}-\x{bb9}\x{bbe}\x{bbf}\x{bc1}\x{bc2}\x{bc6}-\x{bc8}\x{bca}-\x{bcc}\x{bd0}\x{bd7}\x{be6}-\x{bf2}\x{c01}-\x{c03}\x{c05}-\x{c0c}\x{c0e}-\x{c10}\x{c12}-\x{c28}\x{c2a}-\x{c39}\x{c3d}\x{c41}-\x{c44}\x{c58}-\x{c5a}\x{c60}\x{c61}\x{c66}-\x{c6f}\x{c7f}\x{c82}\x{c83}\x{c85}-\x{c8c}\x{c8e}-\x{c90}\x{c92}-\x{ca8}\x{caa}-\x{cb3}\x{cb5}-\x{cb9}\x{cbd}-\x{cc4}\x{cc6}-\x{cc8}\x{cca}\x{ccb}\x{cd5}\x{cd6}\x{cde}\x{ce0}\x{ce1}\x{ce6}-\x{cef}\x{cf1}\x{cf2}\x{d02}\x{d03}\x{d05}-\x{d0c}\x{d0e}-\x{d10}\x{d12}-\x{d3a}\x{d3d}-\x{d40}\x{d46}-\x{d48}\x{d4a}-\x{d4c}\x{d4e}\x{d57}\x{d5f}-\x{d61}\x{d66}-\x{d75}\x{d79}-\x{d7f}\x{d82}\x{d83}\x{d85}-\x{d96}\x{d9a}-\x{db1}\x{db3}-\x{dbb}\x{dbd}\x{dc0}-\x{dc6}\x{dcf}-\x{dd1}\x{dd8}-\x{ddf}\x{de6}-\x{def}\x{df2}-\x{df4}\x{e01}-\x{e30}\x{e32}\x{e33}\x{e40}-\x{e46}\x{e4f}-\x{e5b}\x{e81}\x{e82}\x{e84}\x{e87}\x{e88}\x{e8a}\x{e8d}\x{e94}-\x{e97}\x{e99}-\x{e9f}\x{ea1}-\x{ea3}\x{ea5}\x{ea7}\x{eaa}\x{eab}\x{ead}-\x{eb0}\x{eb2}\x{eb3}\x{ebd}\x{ec0}-\x{ec4}\x{ec6}\x{ed0}-\x{ed9}\x{edc}-\x{edf}\x{f00}-\x{f17}\x{f1a}-\x{f34}\x{f36}\x{f38}\x{f3e}-\x{f47}\x{f49}-\x{f6c}\x{f7f}\x{f85}\x{f88}-\x{f8c}\x{fbe}-\x{fc5}\x{fc7}-\x{fcc}\x{fce}-\x{fda}\x{1000}-\x{102c}\x{1031}\x{1038}\x{103b}\x{103c}\x{103f}-\x{1057}\x{105a}-\x{105d}\x{1061}-\x{1070}\x{1075}-\x{1081}\x{1083}\x{1084}\x{1087}-\x{108c}\x{108e}-\x{109c}\x{109e}-\x{10c5}\x{10c7}\x{10cd}\x{10d0}-\x{1248}\x{124a}-\x{124d}\x{1250}-\x{1256}\x{1258}\x{125a}-\x{125d}\x{1260}-\x{1288}\x{128a}-\x{128d}\x{1290}-\x{12b0}\x{12b2}-\x{12b5}\x{12b8}-\x{12be}\x{12c0}\x{12c2}-\x{12c5}\x{12c8}-\x{12d6}\x{12d8}-\x{1310}\x{1312}-\x{1315}\x{1318}-\x{135a}\x{1360}-\x{137c}\x{1380}-\x{138f}\x{13a0}-\x{13f5}\x{13f8}-\x{13fd}\x{1401}-\x{167f}\x{1681}-\x{169a}\x{16a0}-\x{16f8}\x{1700}-\x{170c}\x{170e}-\x{1711}\x{1720}-\x{1731}\x{1735}\x{1736}\x{1740}-\x{1751}\x{1760}-\x{176c}\x{176e}-\x{1770}\x{1780}-\x{17b3}\x{17b6}\x{17be}-\x{17c5}\x{17c7}\x{17c8}\x{17d4}-\x{17da}\x{17dc}\x{17e0}-\x{17e9}\x{1810}-\x{1819}\x{1820}-\x{1877}\x{1880}-\x{18a8}\x{18aa}\x{18b0}-\x{18f5}\x{1900}-\x{191e}\x{1923}-\x{1926}\x{1929}-\x{192b}\x{1930}\x{1931}\x{1933}-\x{1938}\x{1946}-\x{196d}\x{1970}-\x{1974}\x{1980}-\x{19ab}\x{19b0}-\x{19c9}\x{19d0}-\x{19da}\x{1a00}-\x{1a16}\x{1a19}\x{1a1a}\x{1a1e}-\x{1a55}\x{1a57}\x{1a61}\x{1a63}\x{1a64}\x{1a6d}-\x{1a72}\x{1a80}-\x{1a89}\x{1a90}-\x{1a99}\x{1aa0}-\x{1aad}\x{1b04}-\x{1b33}\x{1b35}\x{1b3b}\x{1b3d}-\x{1b41}\x{1b43}-\x{1b4b}\x{1b50}-\x{1b6a}\x{1b74}-\x{1b7c}\x{1b82}-\x{1ba1}\x{1ba6}\x{1ba7}\x{1baa}\x{1bae}-\x{1be5}\x{1be7}\x{1bea}-\x{1bec}\x{1bee}\x{1bf2}\x{1bf3}\x{1bfc}-\x{1c2b}\x{1c34}\x{1c35}\x{1c3b}-\x{1c49}\x{1c4d}-\x{1c7f}\x{1cc0}-\x{1cc7}\x{1cd3}\x{1ce1}\x{1ce9}-\x{1cec}\x{1cee}-\x{1cf3}\x{1cf5}\x{1cf6}\x{1d00}-\x{1dbf}\x{1e00}-\x{1f15}\x{1f18}-\x{1f1d}\x{1f20}-\x{1f45}\x{1f48}-\x{1f4d}\x{1f50}-\x{1f57}\x{1f59}\x{1f5b}\x{1f5d}\x{1f5f}-\x{1f7d}\x{1f80}-\x{1fb4}\x{1fb6}-\x{1fbc}\x{1fbe}\x{1fc2}-\x{1fc4}\x{1fc6}-\x{1fcc}\x{1fd0}-\x{1fd3}\x{1fd6}-\x{1fdb}\x{1fe0}-\x{1fec}\x{1ff2}-\x{1ff4}\x{1ff6}-\x{1ffc}\x{200e}\x{2071}\x{207f}\x{2090}-\x{209c}\x{2102}\x{2107}\x{210a}-\x{2113}\x{2115}\x{2119}-\x{211d}\x{2124}\x{2126}\x{2128}\x{212a}-\x{212d}\x{212f}-\x{2139}\x{213c}-\x{213f}\x{2145}-\x{2149}\x{214e}\x{214f}\x{2160}-\x{2188}\x{2336}-\x{237a}\x{2395}\x{249c}-\x{24e9}\x{26ac}\x{2800}-\x{28ff}\x{2c00}-\x{2c2e}\x{2c30}-\x{2c5e}\x{2c60}-\x{2ce4}\x{2ceb}-\x{2cee}\x{2cf2}\x{2cf3}\x{2d00}-\x{2d25}\x{2d27}\x{2d2d}\x{2d30}-\x{2d67}\x{2d6f}\x{2d70}\x{2d80}-\x{2d96}\x{2da0}-\x{2da6}\x{2da8}-\x{2dae}\x{2db0}-\x{2db6}\x{2db8}-\x{2dbe}\x{2dc0}-\x{2dc6}\x{2dc8}-\x{2dce}\x{2dd0}-\x{2dd6}\x{2dd8}-\x{2dde}\x{3005}-\x{3007}\x{3021}-\x{3029}\x{302e}\x{302f}\x{3031}-\x{3035}\x{3038}-\x{303c}\x{3041}-\x{3096}\x{309d}-\x{309f}\x{30a1}-\x{30fa}\x{30fc}-\x{30ff}\x{3105}-\x{312d}\x{3131}-\x{318e}\x{3190}-\x{31ba}\x{31f0}-\x{321c}\x{3220}-\x{324f}\x{3260}-\x{327b}\x{327f}-\x{32b0}\x{32c0}-\x{32cb}\x{32d0}-\x{32fe}\x{3300}-\x{3376}\x{337b}-\x{33dd}\x{33e0}-\x{33fe}\x{3400}-\x{4db5}\x{4e00}-\x{9fd5}\x{a000}-\x{a48c}\x{a4d0}-\x{a60c}\x{a610}-\x{a62b}\x{a640}-\x{a66e}\x{a680}-\x{a69d}\x{a6a0}-\x{a6ef}\x{a6f2}-\x{a6f7}\x{a722}-\x{a787}\x{a789}-\x{a7ad}\x{a7b0}-\x{a7b7}\x{a7f7}-\x{a801}\x{a803}-\x{a805}\x{a807}-\x{a80a}\x{a80c}-\x{a824}\x{a827}\x{a830}-\x{a837}\x{a840}-\x{a873}\x{a880}-\x{a8c3}\x{a8ce}-\x{a8d9}\x{a8f2}-\x{a8fd}\x{a900}-\x{a925}\x{a92e}-\x{a946}\x{a952}\x{a953}\x{a95f}-\x{a97c}\x{a983}-\x{a9b2}\x{a9b4}\x{a9b5}\x{a9ba}\x{a9bb}\x{a9bd}-\x{a9cd}\x{a9cf}-\x{a9d9}\x{a9de}-\x{a9e4}\x{a9e6}-\x{a9fe}\x{aa00}-\x{aa28}\x{aa2f}\x{aa30}\x{aa33}\x{aa34}\x{aa40}-\x{aa42}\x{aa44}-\x{aa4b}\x{aa4d}\x{aa50}-\x{aa59}\x{aa5c}-\x{aa7b}\x{aa7d}-\x{aaaf}\x{aab1}\x{aab5}\x{aab6}\x{aab9}-\x{aabd}\x{aac0}\x{aac2}\x{aadb}-\x{aaeb}\x{aaee}-\x{aaf5}\x{ab01}-\x{ab06}\x{ab09}-\x{ab0e}\x{ab11}-\x{ab16}\x{ab20}-\x{ab26}\x{ab28}-\x{ab2e}\x{ab30}-\x{ab65}\x{ab70}-\x{abe4}\x{abe6}\x{abe7}\x{abe9}-\x{abec}\x{abf0}-\x{abf9}\x{ac00}-\x{d7a3}\x{d7b0}-\x{d7c6}\x{d7cb}-\x{d7fb}\x{e000}-\x{fa6d}\x{fa70}-\x{fad9}\x{fb00}-\x{fb06}\x{fb13}-\x{fb17}\x{ff21}-\x{ff3a}\x{ff41}-\x{ff5a}\x{ff66}-\x{ffbe}\x{ffc2}-\x{ffc7}\x{ffca}-\x{ffcf}\x{ffd2}-\x{ffd7}\x{ffda}-\x{ffdc}\x{10000}-\x{1000b}\x{1000d}-\x{10026}\x{10028}-\x{1003a}\x{1003c}\x{1003d}\x{1003f}-\x{1004d}\x{10050}-\x{1005d}\x{10080}-\x{100fa}\x{10100}\x{10102}\x{10107}-\x{10133}\x{10137}-\x{1013f}\x{101d0}-\x{101fc}\x{10280}-\x{1029c}\x{102a0}-\x{102d0}\x{10300}-\x{10323}\x{10330}-\x{1034a}\x{10350}-\x{10375}\x{10380}-\x{1039d}\x{1039f}-\x{103c3}\x{103c8}-\x{103d5}\x{10400}-\x{1049d}\x{104a0}-\x{104a9}\x{10500}-\x{10527}\x{10530}-\x{10563}\x{1056f}\x{10600}-\x{10736}\x{10740}-\x{10755}\x{10760}-\x{10767}\x{11000}\x{11002}-\x{11037}\x{11047}-\x{1104d}\x{11066}-\x{1106f}\x{11082}-\x{110b2}\x{110b7}\x{110b8}\x{110bb}-\x{110c1}\x{110d0}-\x{110e8}\x{110f0}-\x{110f9}\x{11103}-\x{11126}\x{1112c}\x{11136}-\x{11143}\x{11150}-\x{11172}\x{11174}-\x{11176}\x{11182}-\x{111b5}\x{111bf}-\x{111c9}\x{111cd}\x{111d0}-\x{111df}\x{111e1}-\x{111f4}\x{11200}-\x{11211}\x{11213}-\x{1122e}\x{11232}\x{11233}\x{11235}\x{11238}-\x{1123d}\x{11280}-\x{11286}\x{11288}\x{1128a}-\x{1128d}\x{1128f}-\x{1129d}\x{1129f}-\x{112a9}\x{112b0}-\x{112de}\x{112e0}-\x{112e2}\x{112f0}-\x{112f9}\x{11302}\x{11303}\x{11305}-\x{1130c}\x{1130f}\x{11310}\x{11313}-\x{11328}\x{1132a}-\x{11330}\x{11332}\x{11333}\x{11335}-\x{11339}\x{1133d}-\x{1133f}\x{11341}-\x{11344}\x{11347}\x{11348}\x{1134b}-\x{1134d}\x{11350}\x{11357}\x{1135d}-\x{11363}\x{11480}-\x{114b2}\x{114b9}\x{114bb}-\x{114be}\x{114c1}\x{114c4}-\x{114c7}\x{114d0}-\x{114d9}\x{11580}-\x{115b1}\x{115b8}-\x{115bb}\x{115be}\x{115c1}-\x{115db}\x{11600}-\x{11632}\x{1163b}\x{1163c}\x{1163e}\x{11641}-\x{11644}\x{11650}-\x{11659}\x{11680}-\x{116aa}\x{116ac}\x{116ae}\x{116af}\x{116b6}\x{116c0}-\x{116c9}\x{11700}-\x{11719}\x{11720}\x{11721}\x{11726}\x{11730}-\x{1173f}\x{118a0}-\x{118f2}\x{118ff}\x{11ac0}-\x{11af8}\x{12000}-\x{12399}\x{12400}-\x{1246e}\x{12470}-\x{12474}\x{12480}-\x{12543}\x{13000}-\x{1342e}\x{14400}-\x{14646}\x{16800}-\x{16a38}\x{16a40}-\x{16a5e}\x{16a60}-\x{16a69}\x{16a6e}\x{16a6f}\x{16ad0}-\x{16aed}\x{16af5}\x{16b00}-\x{16b2f}\x{16b37}-\x{16b45}\x{16b50}-\x{16b59}\x{16b5b}-\x{16b61}\x{16b63}-\x{16b77}\x{16b7d}-\x{16b8f}\x{16f00}-\x{16f44}\x{16f50}-\x{16f7e}\x{16f93}-\x{16f9f}\x{1b000}\x{1b001}\x{1bc00}-\x{1bc6a}\x{1bc70}-\x{1bc7c}\x{1bc80}-\x{1bc88}\x{1bc90}-\x{1bc99}\x{1bc9c}\x{1bc9f}\x{1d000}-\x{1d0f5}\x{1d100}-\x{1d126}\x{1d129}-\x{1d166}\x{1d16a}-\x{1d172}\x{1d183}\x{1d184}\x{1d18c}-\x{1d1a9}\x{1d1ae}-\x{1d1e8}\x{1d360}-\x{1d371}\x{1d400}-\x{1d454}\x{1d456}-\x{1d49c}\x{1d49e}\x{1d49f}\x{1d4a2}\x{1d4a5}\x{1d4a6}\x{1d4a9}-\x{1d4ac}\x{1d4ae}-\x{1d4b9}\x{1d4bb}\x{1d4bd}-\x{1d4c3}\x{1d4c5}-\x{1d505}\x{1d507}-\x{1d50a}\x{1d50d}-\x{1d514}\x{1d516}-\x{1d51c}\x{1d51e}-\x{1d539}\x{1d53b}-\x{1d53e}\x{1d540}-\x{1d544}\x{1d546}\x{1d54a}-\x{1d550}\x{1d552}-\x{1d6a5}\x{1d6a8}-\x{1d6da}\x{1d6dc}-\x{1d714}\x{1d716}-\x{1d74e}\x{1d750}-\x{1d788}\x{1d78a}-\x{1d7c2}\x{1d7c4}-\x{1d7cb}\x{1d800}-\x{1d9ff}\x{1da37}-\x{1da3a}\x{1da6d}-\x{1da74}\x{1da76}-\x{1da83}\x{1da85}-\x{1da8b}\x{1f110}-\x{1f12e}\x{1f130}-\x{1f169}\x{1f170}-\x{1f19a}\x{1f1e6}-\x{1f202}\x{1f210}-\x{1f23a}\x{1f240}-\x{1f248}\x{1f250}\x{1f251}\x{20000}-\x{2a6d6}\x{2a700}-\x{2b734}\x{2b740}-\x{2b81d}\x{2b820}-\x{2cea1}\x{2f800}-\x{2fa1d}\x{f0000}-\x{ffffd}\x{100000}-\x{10fffd}])|([\x{590}\x{5be}\x{5c0}\x{5c3}\x{5c6}\x{5c8}-\x{5ff}\x{7c0}-\x{7ea}\x{7f4}\x{7f5}\x{7fa}-\x{815}\x{81a}\x{824}\x{828}\x{82e}-\x{858}\x{85c}-\x{89f}\x{200f}\x{fb1d}\x{fb1f}-\x{fb28}\x{fb2a}-\x{fb4f}\x{10800}-\x{1091e}\x{10920}-\x{10a00}\x{10a04}\x{10a07}-\x{10a0b}\x{10a10}-\x{10a37}\x{10a3b}-\x{10a3e}\x{10a40}-\x{10ae4}\x{10ae7}-\x{10b38}\x{10b40}-\x{10e5f}\x{10e7f}-\x{10fff}\x{1e800}-\x{1e8cf}\x{1e8d7}-\x{1edff}\x{1ef00}-\x{1efff}\x{608}\x{60b}\x{60d}\x{61b}-\x{64a}\x{66d}-\x{66f}\x{671}-\x{6d5}\x{6e5}\x{6e6}\x{6ee}\x{6ef}\x{6fa}-\x{710}\x{712}-\x{72f}\x{74b}-\x{7a5}\x{7b1}-\x{7bf}\x{8a0}-\x{8e2}\x{fb50}-\x{fd3d}\x{fd40}-\x{fdcf}\x{fdf0}-\x{fdfc}\x{fdfe}\x{fdff}\x{fe70}-\x{fefe}\x{1ee00}-\x{1eeef}\x{1eef2}-\x{1eeff}]))/u';
167 // @codeCoverageIgnoreEnd
168 // @codingStandardsIgnoreEnd
169
170 /**
171 * Get a cached or new language object for a given language code
172 * @param string $code
173 * @return Language
174 */
175 static function factory( $code ) {
176 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
177
178 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
179 $code = $wgDummyLanguageCodes[$code];
180 }
181
182 // get the language object to process
183 $langObj = isset( self::$mLangObjCache[$code] )
184 ? self::$mLangObjCache[$code]
185 : self::newFromCode( $code );
186
187 // merge the language object in to get it up front in the cache
188 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
189 // get rid of the oldest ones in case we have an overflow
190 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
191
192 return $langObj;
193 }
194
195 /**
196 * Create a language object for a given language code
197 * @param string $code
198 * @throws MWException
199 * @return Language
200 */
201 protected static function newFromCode( $code ) {
202 if ( !Language::isValidCode( $code ) ) {
203 throw new MWException( "Invalid language code \"$code\"" );
204 }
205
206 if ( !Language::isValidBuiltInCode( $code ) ) {
207 // It's not possible to customise this code with class files, so
208 // just return a Language object. This is to support uselang= hacks.
209 $lang = new Language;
210 $lang->setCode( $code );
211 return $lang;
212 }
213
214 // Check if there is a language class for the code
215 $class = self::classFromCode( $code );
216 if ( class_exists( $class ) ) {
217 $lang = new $class;
218 return $lang;
219 }
220
221 // Keep trying the fallback list until we find an existing class
222 $fallbacks = Language::getFallbacksFor( $code );
223 foreach ( $fallbacks as $fallbackCode ) {
224 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
225 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
226 }
227
228 $class = self::classFromCode( $fallbackCode );
229 if ( class_exists( $class ) ) {
230 $lang = new $class;
231 $lang->setCode( $code );
232 return $lang;
233 }
234 }
235
236 throw new MWException( "Invalid fallback sequence for language '$code'" );
237 }
238
239 /**
240 * Checks whether any localisation is available for that language tag
241 * in MediaWiki (MessagesXx.php exists).
242 *
243 * @param string $code Language tag (in lower case)
244 * @return bool Whether language is supported
245 * @since 1.21
246 */
247 public static function isSupportedLanguage( $code ) {
248 if ( !self::isValidBuiltInCode( $code ) ) {
249 return false;
250 }
251
252 if ( $code === 'qqq' ) {
253 return false;
254 }
255
256 return is_readable( self::getMessagesFileName( $code ) ) ||
257 is_readable( self::getJsonMessagesFileName( $code ) );
258 }
259
260 /**
261 * Returns true if a language code string is a well-formed language tag
262 * according to RFC 5646.
263 * This function only checks well-formedness; it doesn't check that
264 * language, script or variant codes actually exist in the repositories.
265 *
266 * Based on regexes by Mark Davis of the Unicode Consortium:
267 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
268 *
269 * @param string $code
270 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
271 *
272 * @return bool
273 * @since 1.21
274 */
275 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
276 $alpha = '[a-z]';
277 $digit = '[0-9]';
278 $alphanum = '[a-z0-9]';
279 $x = 'x'; # private use singleton
280 $singleton = '[a-wy-z]'; # other singleton
281 $s = $lenient ? '[-_]' : '-';
282
283 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
284 $script = "$alpha{4}"; # ISO 15924
285 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
286 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
287 $extension = "$singleton(?:$s$alphanum{2,8})+";
288 $privateUse = "$x(?:$s$alphanum{1,8})+";
289
290 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
291 # Since these are limited, this is safe even later changes to the registry --
292 # the only oddity is that it might change the type of the tag, and thus
293 # the results from the capturing groups.
294 # https://www.iana.org/assignments/language-subtag-registry
295
296 $grandfathered = "en{$s}GB{$s}oed"
297 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
298 . "|no{$s}(?:bok|nyn)"
299 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
300 . "|zh{$s}min{$s}nan";
301
302 $variantList = "$variant(?:$s$variant)*";
303 $extensionList = "$extension(?:$s$extension)*";
304
305 $langtag = "(?:($language)"
306 . "(?:$s$script)?"
307 . "(?:$s$region)?"
308 . "(?:$s$variantList)?"
309 . "(?:$s$extensionList)?"
310 . "(?:$s$privateUse)?)";
311
312 # The final breakdown, with capturing groups for each of these components
313 # The variants, extensions, grandfathered, and private-use may have interior '-'
314
315 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
316
317 return (bool)preg_match( "/$root/", strtolower( $code ) );
318 }
319
320 /**
321 * Returns true if a language code string is of a valid form, whether or
322 * not it exists. This includes codes which are used solely for
323 * customisation via the MediaWiki namespace.
324 *
325 * @param string $code
326 *
327 * @return bool
328 */
329 public static function isValidCode( $code ) {
330 static $cache = [];
331 if ( !isset( $cache[$code] ) ) {
332 // People think language codes are html safe, so enforce it.
333 // Ideally we should only allow a-zA-Z0-9-
334 // but, .+ and other chars are often used for {{int:}} hacks
335 // see bugs 37564, 37587, 36938
336 $cache[$code] =
337 // Protect against path traversal
338 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
339 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
340 }
341 return $cache[$code];
342 }
343
344 /**
345 * Returns true if a language code is of a valid form for the purposes of
346 * internal customisation of MediaWiki, via Messages*.php or *.json.
347 *
348 * @param string $code
349 *
350 * @throws MWException
351 * @since 1.18
352 * @return bool
353 */
354 public static function isValidBuiltInCode( $code ) {
355
356 if ( !is_string( $code ) ) {
357 if ( is_object( $code ) ) {
358 $addmsg = " of class " . get_class( $code );
359 } else {
360 $addmsg = '';
361 }
362 $type = gettype( $code );
363 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
364 }
365
366 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
367 }
368
369 /**
370 * Returns true if a language code is an IETF tag known to MediaWiki.
371 *
372 * @param string $tag
373 *
374 * @since 1.21
375 * @return bool
376 */
377 public static function isKnownLanguageTag( $tag ) {
378 // Quick escape for invalid input to avoid exceptions down the line
379 // when code tries to process tags which are not valid at all.
380 if ( !self::isValidBuiltInCode( $tag ) ) {
381 return false;
382 }
383
384 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
385 || self::fetchLanguageName( $tag, $tag ) !== ''
386 ) {
387 return true;
388 }
389
390 return false;
391 }
392
393 /**
394 * Get the LocalisationCache instance
395 *
396 * @return LocalisationCache
397 */
398 public static function getLocalisationCache() {
399 if ( is_null( self::$dataCache ) ) {
400 global $wgLocalisationCacheConf;
401 $class = $wgLocalisationCacheConf['class'];
402 self::$dataCache = new $class( $wgLocalisationCacheConf );
403 }
404 return self::$dataCache;
405 }
406
407 function __construct() {
408 $this->mConverter = new FakeConverter( $this );
409 // Set the code to the name of the descendant
410 if ( get_class( $this ) == 'Language' ) {
411 $this->mCode = 'en';
412 } else {
413 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
414 }
415 self::getLocalisationCache();
416 }
417
418 /**
419 * Reduce memory usage
420 */
421 function __destruct() {
422 foreach ( $this as $name => $value ) {
423 unset( $this->$name );
424 }
425 }
426
427 /**
428 * Hook which will be called if this is the content language.
429 * Descendants can use this to register hook functions or modify globals
430 */
431 function initContLang() {
432 }
433
434 /**
435 * @return array
436 * @since 1.19
437 */
438 public function getFallbackLanguages() {
439 return self::getFallbacksFor( $this->mCode );
440 }
441
442 /**
443 * Exports $wgBookstoreListEn
444 * @return array
445 */
446 public function getBookstoreList() {
447 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
448 }
449
450 /**
451 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
452 * available in localised form, it will be included in English.
453 *
454 * @return array
455 */
456 public function getNamespaces() {
457 if ( is_null( $this->namespaceNames ) ) {
458 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
459
460 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
461 $validNamespaces = MWNamespace::getCanonicalNamespaces();
462
463 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
464
465 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
466 if ( $wgMetaNamespaceTalk ) {
467 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
468 } else {
469 $talk = $this->namespaceNames[NS_PROJECT_TALK];
470 $this->namespaceNames[NS_PROJECT_TALK] =
471 $this->fixVariableInNamespace( $talk );
472 }
473
474 # Sometimes a language will be localised but not actually exist on this wiki.
475 foreach ( $this->namespaceNames as $key => $text ) {
476 if ( !isset( $validNamespaces[$key] ) ) {
477 unset( $this->namespaceNames[$key] );
478 }
479 }
480
481 # The above mixing may leave namespaces out of canonical order.
482 # Re-order by namespace ID number...
483 ksort( $this->namespaceNames );
484
485 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
486 }
487
488 return $this->namespaceNames;
489 }
490
491 /**
492 * Arbitrarily set all of the namespace names at once. Mainly used for testing
493 * @param array $namespaces Array of namespaces (id => name)
494 */
495 public function setNamespaces( array $namespaces ) {
496 $this->namespaceNames = $namespaces;
497 $this->mNamespaceIds = null;
498 }
499
500 /**
501 * Resets all of the namespace caches. Mainly used for testing
502 */
503 public function resetNamespaces() {
504 $this->namespaceNames = null;
505 $this->mNamespaceIds = null;
506 $this->namespaceAliases = null;
507 }
508
509 /**
510 * A convenience function that returns getNamespaces() with spaces instead of underscores
511 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
512 *
513 * @return array
514 */
515 public function getFormattedNamespaces() {
516 $ns = $this->getNamespaces();
517 foreach ( $ns as $k => $v ) {
518 $ns[$k] = strtr( $v, '_', ' ' );
519 }
520 return $ns;
521 }
522
523 /**
524 * Get a namespace value by key
525 *
526 * <code>
527 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
528 * echo $mw_ns; // prints 'MediaWiki'
529 * </code>
530 *
531 * @param int $index The array key of the namespace to return
532 * @return string|bool String if the namespace value exists, otherwise false
533 */
534 public function getNsText( $index ) {
535 $ns = $this->getNamespaces();
536 return isset( $ns[$index] ) ? $ns[$index] : false;
537 }
538
539 /**
540 * A convenience function that returns the same thing as
541 * getNsText() except with '_' changed to ' ', useful for
542 * producing output.
543 *
544 * <code>
545 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
546 * echo $mw_ns; // prints 'MediaWiki talk'
547 * </code>
548 *
549 * @param int $index The array key of the namespace to return
550 * @return string Namespace name without underscores (empty string if namespace does not exist)
551 */
552 public function getFormattedNsText( $index ) {
553 $ns = $this->getNsText( $index );
554 return strtr( $ns, '_', ' ' );
555 }
556
557 /**
558 * Returns gender-dependent namespace alias if available.
559 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
560 * @param int $index Namespace index
561 * @param string $gender Gender key (male, female... )
562 * @return string
563 * @since 1.18
564 */
565 public function getGenderNsText( $index, $gender ) {
566 global $wgExtraGenderNamespaces;
567
568 $ns = $wgExtraGenderNamespaces +
569 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
570
571 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
572 }
573
574 /**
575 * Whether this language uses gender-dependent namespace aliases.
576 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
577 * @return bool
578 * @since 1.18
579 */
580 public function needsGenderDistinction() {
581 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
582 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
583 // $wgExtraGenderNamespaces overrides everything
584 return true;
585 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
586 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
587 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
588 return false;
589 } else {
590 // Check what is in i18n files
591 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
592 return count( $aliases ) > 0;
593 }
594 }
595
596 /**
597 * Get a namespace key by value, case insensitive.
598 * Only matches namespace names for the current language, not the
599 * canonical ones defined in Namespace.php.
600 *
601 * @param string $text
602 * @return int|bool An integer if $text is a valid value otherwise false
603 */
604 function getLocalNsIndex( $text ) {
605 $lctext = $this->lc( $text );
606 $ids = $this->getNamespaceIds();
607 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
608 }
609
610 /**
611 * @return array
612 */
613 public function getNamespaceAliases() {
614 if ( is_null( $this->namespaceAliases ) ) {
615 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
616 if ( !$aliases ) {
617 $aliases = [];
618 } else {
619 foreach ( $aliases as $name => $index ) {
620 if ( $index === NS_PROJECT_TALK ) {
621 unset( $aliases[$name] );
622 $name = $this->fixVariableInNamespace( $name );
623 $aliases[$name] = $index;
624 }
625 }
626 }
627
628 global $wgExtraGenderNamespaces;
629 $genders = $wgExtraGenderNamespaces +
630 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
631 foreach ( $genders as $index => $forms ) {
632 foreach ( $forms as $alias ) {
633 $aliases[$alias] = $index;
634 }
635 }
636
637 # Also add converted namespace names as aliases, to avoid confusion.
638 $convertedNames = [];
639 foreach ( $this->getVariants() as $variant ) {
640 if ( $variant === $this->mCode ) {
641 continue;
642 }
643 foreach ( $this->getNamespaces() as $ns => $_ ) {
644 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
645 }
646 }
647
648 $this->namespaceAliases = $aliases + $convertedNames;
649 }
650
651 return $this->namespaceAliases;
652 }
653
654 /**
655 * @return array
656 */
657 public function getNamespaceIds() {
658 if ( is_null( $this->mNamespaceIds ) ) {
659 global $wgNamespaceAliases;
660 # Put namespace names and aliases into a hashtable.
661 # If this is too slow, then we should arrange it so that it is done
662 # before caching. The catch is that at pre-cache time, the above
663 # class-specific fixup hasn't been done.
664 $this->mNamespaceIds = [];
665 foreach ( $this->getNamespaces() as $index => $name ) {
666 $this->mNamespaceIds[$this->lc( $name )] = $index;
667 }
668 foreach ( $this->getNamespaceAliases() as $name => $index ) {
669 $this->mNamespaceIds[$this->lc( $name )] = $index;
670 }
671 if ( $wgNamespaceAliases ) {
672 foreach ( $wgNamespaceAliases as $name => $index ) {
673 $this->mNamespaceIds[$this->lc( $name )] = $index;
674 }
675 }
676 }
677 return $this->mNamespaceIds;
678 }
679
680 /**
681 * Get a namespace key by value, case insensitive. Canonical namespace
682 * names override custom ones defined for the current language.
683 *
684 * @param string $text
685 * @return int|bool An integer if $text is a valid value otherwise false
686 */
687 public function getNsIndex( $text ) {
688 $lctext = $this->lc( $text );
689 $ns = MWNamespace::getCanonicalIndex( $lctext );
690 if ( $ns !== null ) {
691 return $ns;
692 }
693 $ids = $this->getNamespaceIds();
694 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
695 }
696
697 /**
698 * short names for language variants used for language conversion links.
699 *
700 * @param string $code
701 * @param bool $usemsg Use the "variantname-xyz" message if it exists
702 * @return string
703 */
704 public function getVariantname( $code, $usemsg = true ) {
705 $msg = "variantname-$code";
706 if ( $usemsg && wfMessage( $msg )->exists() ) {
707 return $this->getMessageFromDB( $msg );
708 }
709 $name = self::fetchLanguageName( $code );
710 if ( $name ) {
711 return $name; # if it's defined as a language name, show that
712 } else {
713 # otherwise, output the language code
714 return $code;
715 }
716 }
717
718 /**
719 * @return array
720 */
721 public function getDatePreferences() {
722 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
723 }
724
725 /**
726 * @return array
727 */
728 function getDateFormats() {
729 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
730 }
731
732 /**
733 * @return array|string
734 */
735 public function getDefaultDateFormat() {
736 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
737 if ( $df === 'dmy or mdy' ) {
738 global $wgAmericanDates;
739 return $wgAmericanDates ? 'mdy' : 'dmy';
740 } else {
741 return $df;
742 }
743 }
744
745 /**
746 * @return array
747 */
748 public function getDatePreferenceMigrationMap() {
749 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
750 }
751
752 /**
753 * @param string $image
754 * @return array|null
755 */
756 function getImageFile( $image ) {
757 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
758 }
759
760 /**
761 * @return array
762 * @since 1.24
763 */
764 public function getImageFiles() {
765 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
766 }
767
768 /**
769 * @return array
770 */
771 public function getExtraUserToggles() {
772 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
773 }
774
775 /**
776 * @param string $tog
777 * @return string
778 */
779 function getUserToggle( $tog ) {
780 return $this->getMessageFromDB( "tog-$tog" );
781 }
782
783 /**
784 * Get an array of language names, indexed by code.
785 * @param null|string $inLanguage Code of language in which to return the names
786 * Use null for autonyms (native names)
787 * @param string $include One of:
788 * 'all' all available languages
789 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
790 * 'mwfile' only if the language is in 'mw' *and* has a message file
791 * @return array Language code => language name
792 * @since 1.20
793 */
794 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
795 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
796 $cacheKey .= ":$include";
797 if ( self::$languageNameCache === null ) {
798 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
799 }
800
801 $ret = self::$languageNameCache->get( $cacheKey );
802 if ( !$ret ) {
803 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
804 self::$languageNameCache->set( $cacheKey, $ret );
805 }
806 return $ret;
807 }
808
809 /**
810 * Uncached helper for fetchLanguageNames
811 * @param null|string $inLanguage Code of language in which to return the names
812 * Use null for autonyms (native names)
813 * @param string $include One of:
814 * 'all' all available languages
815 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
816 * 'mwfile' only if the language is in 'mw' *and* has a message file
817 * @return array Language code => language name
818 */
819 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
820 global $wgExtraLanguageNames;
821
822 // If passed an invalid language code to use, fallback to en
823 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
824 $inLanguage = 'en';
825 }
826
827 $names = [];
828
829 if ( $inLanguage ) {
830 # TODO: also include when $inLanguage is null, when this code is more efficient
831 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
832 }
833
834 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
835 foreach ( $mwNames as $mwCode => $mwName ) {
836 # - Prefer own MediaWiki native name when not using the hook
837 # - For other names just add if not added through the hook
838 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
839 $names[$mwCode] = $mwName;
840 }
841 }
842
843 if ( $include === 'all' ) {
844 ksort( $names );
845 return $names;
846 }
847
848 $returnMw = [];
849 $coreCodes = array_keys( $mwNames );
850 foreach ( $coreCodes as $coreCode ) {
851 $returnMw[$coreCode] = $names[$coreCode];
852 }
853
854 if ( $include === 'mwfile' ) {
855 $namesMwFile = [];
856 # We do this using a foreach over the codes instead of a directory
857 # loop so that messages files in extensions will work correctly.
858 foreach ( $returnMw as $code => $value ) {
859 if ( is_readable( self::getMessagesFileName( $code ) )
860 || is_readable( self::getJsonMessagesFileName( $code ) )
861 ) {
862 $namesMwFile[$code] = $names[$code];
863 }
864 }
865
866 ksort( $namesMwFile );
867 return $namesMwFile;
868 }
869
870 ksort( $returnMw );
871 # 'mw' option; default if it's not one of the other two options (all/mwfile)
872 return $returnMw;
873 }
874
875 /**
876 * @param string $code The code of the language for which to get the name
877 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
878 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
879 * @return string Language name or empty
880 * @since 1.20
881 */
882 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
883 $code = strtolower( $code );
884 $array = self::fetchLanguageNames( $inLanguage, $include );
885 return !array_key_exists( $code, $array ) ? '' : $array[$code];
886 }
887
888 /**
889 * Get a message from the MediaWiki namespace.
890 *
891 * @param string $msg Message name
892 * @return string
893 */
894 public function getMessageFromDB( $msg ) {
895 return $this->msg( $msg )->text();
896 }
897
898 /**
899 * Get message object in this language. Only for use inside this class.
900 *
901 * @param string $msg Message name
902 * @return Message
903 */
904 protected function msg( $msg ) {
905 return wfMessage( $msg )->inLanguage( $this );
906 }
907
908 /**
909 * @param string $key
910 * @return string
911 */
912 public function getMonthName( $key ) {
913 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
914 }
915
916 /**
917 * @return array
918 */
919 public function getMonthNamesArray() {
920 $monthNames = [ '' ];
921 for ( $i = 1; $i < 13; $i++ ) {
922 $monthNames[] = $this->getMonthName( $i );
923 }
924 return $monthNames;
925 }
926
927 /**
928 * @param string $key
929 * @return string
930 */
931 public function getMonthNameGen( $key ) {
932 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
933 }
934
935 /**
936 * @param string $key
937 * @return string
938 */
939 public function getMonthAbbreviation( $key ) {
940 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
941 }
942
943 /**
944 * @return array
945 */
946 public function getMonthAbbreviationsArray() {
947 $monthNames = [ '' ];
948 for ( $i = 1; $i < 13; $i++ ) {
949 $monthNames[] = $this->getMonthAbbreviation( $i );
950 }
951 return $monthNames;
952 }
953
954 /**
955 * @param string $key
956 * @return string
957 */
958 public function getWeekdayName( $key ) {
959 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
960 }
961
962 /**
963 * @param string $key
964 * @return string
965 */
966 function getWeekdayAbbreviation( $key ) {
967 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
968 }
969
970 /**
971 * @param string $key
972 * @return string
973 */
974 function getIranianCalendarMonthName( $key ) {
975 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
976 }
977
978 /**
979 * @param string $key
980 * @return string
981 */
982 function getHebrewCalendarMonthName( $key ) {
983 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
984 }
985
986 /**
987 * @param string $key
988 * @return string
989 */
990 function getHebrewCalendarMonthNameGen( $key ) {
991 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
992 }
993
994 /**
995 * @param string $key
996 * @return string
997 */
998 function getHijriCalendarMonthName( $key ) {
999 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1000 }
1001
1002 /**
1003 * Pass through result from $dateTimeObj->format()
1004 * @param DateTime|bool|null &$dateTimeObj
1005 * @param string $ts
1006 * @param DateTimeZone|bool|null $zone
1007 * @param string $code
1008 * @return string
1009 */
1010 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1011 if ( !$dateTimeObj ) {
1012 $dateTimeObj = DateTime::createFromFormat(
1013 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1014 );
1015 }
1016 return $dateTimeObj->format( $code );
1017 }
1018
1019 /**
1020 * This is a workalike of PHP's date() function, but with better
1021 * internationalisation, a reduced set of format characters, and a better
1022 * escaping format.
1023 *
1024 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1025 * the PHP manual for definitions. There are a number of extensions, which
1026 * start with "x":
1027 *
1028 * xn Do not translate digits of the next numeric format character
1029 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1030 * xr Use roman numerals for the next numeric format character
1031 * xh Use hebrew numerals for the next numeric format character
1032 * xx Literal x
1033 * xg Genitive month name
1034 *
1035 * xij j (day number) in Iranian calendar
1036 * xiF F (month name) in Iranian calendar
1037 * xin n (month number) in Iranian calendar
1038 * xiy y (two digit year) in Iranian calendar
1039 * xiY Y (full year) in Iranian calendar
1040 * xit t (days in month) in Iranian calendar
1041 * xiz z (day of the year) in Iranian calendar
1042 *
1043 * xjj j (day number) in Hebrew calendar
1044 * xjF F (month name) in Hebrew calendar
1045 * xjt t (days in month) in Hebrew calendar
1046 * xjx xg (genitive month name) in Hebrew calendar
1047 * xjn n (month number) in Hebrew calendar
1048 * xjY Y (full year) in Hebrew calendar
1049 *
1050 * xmj j (day number) in Hijri calendar
1051 * xmF F (month name) in Hijri calendar
1052 * xmn n (month number) in Hijri calendar
1053 * xmY Y (full year) in Hijri calendar
1054 *
1055 * xkY Y (full year) in Thai solar calendar. Months and days are
1056 * identical to the Gregorian calendar
1057 * xoY Y (full year) in Minguo calendar or Juche year.
1058 * Months and days are identical to the
1059 * Gregorian calendar
1060 * xtY Y (full year) in Japanese nengo. Months and days are
1061 * identical to the Gregorian calendar
1062 *
1063 * Characters enclosed in double quotes will be considered literal (with
1064 * the quotes themselves removed). Unmatched quotes will be considered
1065 * literal quotes. Example:
1066 *
1067 * "The month is" F => The month is January
1068 * i's" => 20'11"
1069 *
1070 * Backslash escaping is also supported.
1071 *
1072 * Input timestamp is assumed to be pre-normalized to the desired local
1073 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1074 * $ts is UTC if $zone is not given.
1075 *
1076 * @param string $format
1077 * @param string $ts 14-character timestamp
1078 * YYYYMMDDHHMMSS
1079 * 01234567890123
1080 * @param DateTimeZone $zone Timezone of $ts
1081 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1082 * Only makes sense if $ts is the current time.
1083 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1084 *
1085 * @throws MWException
1086 * @return string
1087 */
1088 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1089 $s = '';
1090 $raw = false;
1091 $roman = false;
1092 $hebrewNum = false;
1093 $dateTimeObj = false;
1094 $rawToggle = false;
1095 $iranian = false;
1096 $hebrew = false;
1097 $hijri = false;
1098 $thai = false;
1099 $minguo = false;
1100 $tenno = false;
1101
1102 $usedSecond = false;
1103 $usedMinute = false;
1104 $usedHour = false;
1105 $usedAMPM = false;
1106 $usedDay = false;
1107 $usedWeek = false;
1108 $usedMonth = false;
1109 $usedYear = false;
1110 $usedISOYear = false;
1111 $usedIsLeapYear = false;
1112
1113 $usedHebrewMonth = false;
1114 $usedIranianMonth = false;
1115 $usedHijriMonth = false;
1116 $usedHebrewYear = false;
1117 $usedIranianYear = false;
1118 $usedHijriYear = false;
1119 $usedTennoYear = false;
1120
1121 if ( strlen( $ts ) !== 14 ) {
1122 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1123 }
1124
1125 if ( !ctype_digit( $ts ) ) {
1126 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1127 }
1128
1129 $formatLength = strlen( $format );
1130 for ( $p = 0; $p < $formatLength; $p++ ) {
1131 $num = false;
1132 $code = $format[$p];
1133 if ( $code == 'x' && $p < $formatLength - 1 ) {
1134 $code .= $format[++$p];
1135 }
1136
1137 if ( ( $code === 'xi'
1138 || $code === 'xj'
1139 || $code === 'xk'
1140 || $code === 'xm'
1141 || $code === 'xo'
1142 || $code === 'xt' )
1143 && $p < $formatLength - 1 ) {
1144 $code .= $format[++$p];
1145 }
1146
1147 switch ( $code ) {
1148 case 'xx':
1149 $s .= 'x';
1150 break;
1151 case 'xn':
1152 $raw = true;
1153 break;
1154 case 'xN':
1155 $rawToggle = !$rawToggle;
1156 break;
1157 case 'xr':
1158 $roman = true;
1159 break;
1160 case 'xh':
1161 $hebrewNum = true;
1162 break;
1163 case 'xg':
1164 $usedMonth = true;
1165 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1166 break;
1167 case 'xjx':
1168 $usedHebrewMonth = true;
1169 if ( !$hebrew ) {
1170 $hebrew = self::tsToHebrew( $ts );
1171 }
1172 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1173 break;
1174 case 'd':
1175 $usedDay = true;
1176 $num = substr( $ts, 6, 2 );
1177 break;
1178 case 'D':
1179 $usedDay = true;
1180 $s .= $this->getWeekdayAbbreviation(
1181 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1182 );
1183 break;
1184 case 'j':
1185 $usedDay = true;
1186 $num = intval( substr( $ts, 6, 2 ) );
1187 break;
1188 case 'xij':
1189 $usedDay = true;
1190 if ( !$iranian ) {
1191 $iranian = self::tsToIranian( $ts );
1192 }
1193 $num = $iranian[2];
1194 break;
1195 case 'xmj':
1196 $usedDay = true;
1197 if ( !$hijri ) {
1198 $hijri = self::tsToHijri( $ts );
1199 }
1200 $num = $hijri[2];
1201 break;
1202 case 'xjj':
1203 $usedDay = true;
1204 if ( !$hebrew ) {
1205 $hebrew = self::tsToHebrew( $ts );
1206 }
1207 $num = $hebrew[2];
1208 break;
1209 case 'l':
1210 $usedDay = true;
1211 $s .= $this->getWeekdayName(
1212 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1213 );
1214 break;
1215 case 'F':
1216 $usedMonth = true;
1217 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1218 break;
1219 case 'xiF':
1220 $usedIranianMonth = true;
1221 if ( !$iranian ) {
1222 $iranian = self::tsToIranian( $ts );
1223 }
1224 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1225 break;
1226 case 'xmF':
1227 $usedHijriMonth = true;
1228 if ( !$hijri ) {
1229 $hijri = self::tsToHijri( $ts );
1230 }
1231 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1232 break;
1233 case 'xjF':
1234 $usedHebrewMonth = true;
1235 if ( !$hebrew ) {
1236 $hebrew = self::tsToHebrew( $ts );
1237 }
1238 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1239 break;
1240 case 'm':
1241 $usedMonth = true;
1242 $num = substr( $ts, 4, 2 );
1243 break;
1244 case 'M':
1245 $usedMonth = true;
1246 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1247 break;
1248 case 'n':
1249 $usedMonth = true;
1250 $num = intval( substr( $ts, 4, 2 ) );
1251 break;
1252 case 'xin':
1253 $usedIranianMonth = true;
1254 if ( !$iranian ) {
1255 $iranian = self::tsToIranian( $ts );
1256 }
1257 $num = $iranian[1];
1258 break;
1259 case 'xmn':
1260 $usedHijriMonth = true;
1261 if ( !$hijri ) {
1262 $hijri = self::tsToHijri( $ts );
1263 }
1264 $num = $hijri[1];
1265 break;
1266 case 'xjn':
1267 $usedHebrewMonth = true;
1268 if ( !$hebrew ) {
1269 $hebrew = self::tsToHebrew( $ts );
1270 }
1271 $num = $hebrew[1];
1272 break;
1273 case 'xjt':
1274 $usedHebrewMonth = true;
1275 if ( !$hebrew ) {
1276 $hebrew = self::tsToHebrew( $ts );
1277 }
1278 $num = $hebrew[3];
1279 break;
1280 case 'Y':
1281 $usedYear = true;
1282 $num = substr( $ts, 0, 4 );
1283 break;
1284 case 'xiY':
1285 $usedIranianYear = true;
1286 if ( !$iranian ) {
1287 $iranian = self::tsToIranian( $ts );
1288 }
1289 $num = $iranian[0];
1290 break;
1291 case 'xmY':
1292 $usedHijriYear = true;
1293 if ( !$hijri ) {
1294 $hijri = self::tsToHijri( $ts );
1295 }
1296 $num = $hijri[0];
1297 break;
1298 case 'xjY':
1299 $usedHebrewYear = true;
1300 if ( !$hebrew ) {
1301 $hebrew = self::tsToHebrew( $ts );
1302 }
1303 $num = $hebrew[0];
1304 break;
1305 case 'xkY':
1306 $usedYear = true;
1307 if ( !$thai ) {
1308 $thai = self::tsToYear( $ts, 'thai' );
1309 }
1310 $num = $thai[0];
1311 break;
1312 case 'xoY':
1313 $usedYear = true;
1314 if ( !$minguo ) {
1315 $minguo = self::tsToYear( $ts, 'minguo' );
1316 }
1317 $num = $minguo[0];
1318 break;
1319 case 'xtY':
1320 $usedTennoYear = true;
1321 if ( !$tenno ) {
1322 $tenno = self::tsToYear( $ts, 'tenno' );
1323 }
1324 $num = $tenno[0];
1325 break;
1326 case 'y':
1327 $usedYear = true;
1328 $num = substr( $ts, 2, 2 );
1329 break;
1330 case 'xiy':
1331 $usedIranianYear = true;
1332 if ( !$iranian ) {
1333 $iranian = self::tsToIranian( $ts );
1334 }
1335 $num = substr( $iranian[0], -2 );
1336 break;
1337 case 'xit':
1338 $usedIranianYear = true;
1339 if ( !$iranian ) {
1340 $iranian = self::tsToIranian( $ts );
1341 }
1342 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1343 break;
1344 case 'xiz':
1345 $usedIranianYear = true;
1346 if ( !$iranian ) {
1347 $iranian = self::tsToIranian( $ts );
1348 }
1349 $num = $iranian[3];
1350 break;
1351 case 'a':
1352 $usedAMPM = true;
1353 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1354 break;
1355 case 'A':
1356 $usedAMPM = true;
1357 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1358 break;
1359 case 'g':
1360 $usedHour = true;
1361 $h = substr( $ts, 8, 2 );
1362 $num = $h % 12 ? $h % 12 : 12;
1363 break;
1364 case 'G':
1365 $usedHour = true;
1366 $num = intval( substr( $ts, 8, 2 ) );
1367 break;
1368 case 'h':
1369 $usedHour = true;
1370 $h = substr( $ts, 8, 2 );
1371 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1372 break;
1373 case 'H':
1374 $usedHour = true;
1375 $num = substr( $ts, 8, 2 );
1376 break;
1377 case 'i':
1378 $usedMinute = true;
1379 $num = substr( $ts, 10, 2 );
1380 break;
1381 case 's':
1382 $usedSecond = true;
1383 $num = substr( $ts, 12, 2 );
1384 break;
1385 case 'c':
1386 case 'r':
1387 $usedSecond = true;
1388 // fall through
1389 case 'e':
1390 case 'O':
1391 case 'P':
1392 case 'T':
1393 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1394 break;
1395 case 'w':
1396 case 'N':
1397 case 'z':
1398 $usedDay = true;
1399 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1400 break;
1401 case 'W':
1402 $usedWeek = true;
1403 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1404 break;
1405 case 't':
1406 $usedMonth = true;
1407 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1408 break;
1409 case 'L':
1410 $usedIsLeapYear = true;
1411 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1412 break;
1413 case 'o':
1414 $usedISOYear = true;
1415 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1416 break;
1417 case 'U':
1418 $usedSecond = true;
1419 // fall through
1420 case 'I':
1421 case 'Z':
1422 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1423 break;
1424 case '\\':
1425 # Backslash escaping
1426 if ( $p < $formatLength - 1 ) {
1427 $s .= $format[++$p];
1428 } else {
1429 $s .= '\\';
1430 }
1431 break;
1432 case '"':
1433 # Quoted literal
1434 if ( $p < $formatLength - 1 ) {
1435 $endQuote = strpos( $format, '"', $p + 1 );
1436 if ( $endQuote === false ) {
1437 # No terminating quote, assume literal "
1438 $s .= '"';
1439 } else {
1440 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1441 $p = $endQuote;
1442 }
1443 } else {
1444 # Quote at end of string, assume literal "
1445 $s .= '"';
1446 }
1447 break;
1448 default:
1449 $s .= $format[$p];
1450 }
1451 if ( $num !== false ) {
1452 if ( $rawToggle || $raw ) {
1453 $s .= $num;
1454 $raw = false;
1455 } elseif ( $roman ) {
1456 $s .= Language::romanNumeral( $num );
1457 $roman = false;
1458 } elseif ( $hebrewNum ) {
1459 $s .= self::hebrewNumeral( $num );
1460 $hebrewNum = false;
1461 } else {
1462 $s .= $this->formatNum( $num, true );
1463 }
1464 }
1465 }
1466
1467 if ( $ttl === 'unused' ) {
1468 // No need to calculate the TTL, the caller wont use it anyway.
1469 } elseif ( $usedSecond ) {
1470 $ttl = 1;
1471 } elseif ( $usedMinute ) {
1472 $ttl = 60 - substr( $ts, 12, 2 );
1473 } elseif ( $usedHour ) {
1474 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1475 } elseif ( $usedAMPM ) {
1476 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1477 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1478 } elseif (
1479 $usedDay ||
1480 $usedHebrewMonth ||
1481 $usedIranianMonth ||
1482 $usedHijriMonth ||
1483 $usedHebrewYear ||
1484 $usedIranianYear ||
1485 $usedHijriYear ||
1486 $usedTennoYear
1487 ) {
1488 // @todo Someone who understands the non-Gregorian calendars
1489 // should write proper logic for them so that they don't need purged every day.
1490 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1491 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1492 } else {
1493 $possibleTtls = [];
1494 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1495 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1496 if ( $usedWeek ) {
1497 $possibleTtls[] =
1498 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1499 $timeRemainingInDay;
1500 } elseif ( $usedISOYear ) {
1501 // December 28th falls on the last ISO week of the year, every year.
1502 // The last ISO week of a year can be 52 or 53.
1503 $lastWeekOfISOYear = DateTime::createFromFormat(
1504 'Ymd',
1505 substr( $ts, 0, 4 ) . '1228',
1506 $zone ?: new DateTimeZone( 'UTC' )
1507 )->format( 'W' );
1508 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1509 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1510 $timeRemainingInWeek =
1511 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1512 + $timeRemainingInDay;
1513 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1514 }
1515
1516 if ( $usedMonth ) {
1517 $possibleTtls[] =
1518 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1519 substr( $ts, 6, 2 ) ) * 86400
1520 + $timeRemainingInDay;
1521 } elseif ( $usedYear ) {
1522 $possibleTtls[] =
1523 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1524 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1525 + $timeRemainingInDay;
1526 } elseif ( $usedIsLeapYear ) {
1527 $year = substr( $ts, 0, 4 );
1528 $timeRemainingInYear =
1529 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1530 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1531 + $timeRemainingInDay;
1532 $mod = $year % 4;
1533 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1534 // this isn't a leap year. see when the next one starts
1535 $nextCandidate = $year - $mod + 4;
1536 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1537 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1538 $timeRemainingInYear;
1539 } else {
1540 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1541 $timeRemainingInYear;
1542 }
1543 } else {
1544 // this is a leap year, so the next year isn't
1545 $possibleTtls[] = $timeRemainingInYear;
1546 }
1547 }
1548
1549 if ( $possibleTtls ) {
1550 $ttl = min( $possibleTtls );
1551 }
1552 }
1553
1554 return $s;
1555 }
1556
1557 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1558 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1559
1560 /**
1561 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1562 * Gregorian dates to Iranian dates. Originally written in C, it
1563 * is released under the terms of GNU Lesser General Public
1564 * License. Conversion to PHP was performed by Niklas Laxström.
1565 *
1566 * Link: http://www.farsiweb.info/jalali/jalali.c
1567 *
1568 * @param string $ts
1569 *
1570 * @return int[]
1571 */
1572 private static function tsToIranian( $ts ) {
1573 $gy = substr( $ts, 0, 4 ) -1600;
1574 $gm = substr( $ts, 4, 2 ) -1;
1575 $gd = substr( $ts, 6, 2 ) -1;
1576
1577 # Days passed from the beginning (including leap years)
1578 $gDayNo = 365 * $gy
1579 + floor( ( $gy + 3 ) / 4 )
1580 - floor( ( $gy + 99 ) / 100 )
1581 + floor( ( $gy + 399 ) / 400 );
1582
1583 // Add days of the past months of this year
1584 for ( $i = 0; $i < $gm; $i++ ) {
1585 $gDayNo += self::$GREG_DAYS[$i];
1586 }
1587
1588 // Leap years
1589 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1590 $gDayNo++;
1591 }
1592
1593 // Days passed in current month
1594 $gDayNo += (int)$gd;
1595
1596 $jDayNo = $gDayNo - 79;
1597
1598 $jNp = floor( $jDayNo / 12053 );
1599 $jDayNo %= 12053;
1600
1601 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1602 $jDayNo %= 1461;
1603
1604 if ( $jDayNo >= 366 ) {
1605 $jy += floor( ( $jDayNo - 1 ) / 365 );
1606 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1607 }
1608
1609 $jz = $jDayNo;
1610
1611 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1612 $jDayNo -= self::$IRANIAN_DAYS[$i];
1613 }
1614
1615 $jm = $i + 1;
1616 $jd = $jDayNo + 1;
1617
1618 return [ $jy, $jm, $jd, $jz ];
1619 }
1620
1621 /**
1622 * Converting Gregorian dates to Hijri dates.
1623 *
1624 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1625 *
1626 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1627 *
1628 * @param string $ts
1629 *
1630 * @return int[]
1631 */
1632 private static function tsToHijri( $ts ) {
1633 $year = substr( $ts, 0, 4 );
1634 $month = substr( $ts, 4, 2 );
1635 $day = substr( $ts, 6, 2 );
1636
1637 $zyr = $year;
1638 $zd = $day;
1639 $zm = $month;
1640 $zy = $zyr;
1641
1642 if (
1643 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1644 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1645 ) {
1646 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1647 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1648 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1649 $zd - 32075;
1650 } else {
1651 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1652 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1653 }
1654
1655 $zl = $zjd -1948440 + 10632;
1656 $zn = (int)( ( $zl - 1 ) / 10631 );
1657 $zl = $zl - 10631 * $zn + 354;
1658 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1659 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1660 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1661 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1662 $zm = (int)( ( 24 * $zl ) / 709 );
1663 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1664 $zy = 30 * $zn + $zj - 30;
1665
1666 return [ $zy, $zm, $zd ];
1667 }
1668
1669 /**
1670 * Converting Gregorian dates to Hebrew dates.
1671 *
1672 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1673 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1674 * to translate the relevant functions into PHP and release them under
1675 * GNU GPL.
1676 *
1677 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1678 * and Adar II is 14. In a non-leap year, Adar is 6.
1679 *
1680 * @param string $ts
1681 *
1682 * @return int[]
1683 */
1684 private static function tsToHebrew( $ts ) {
1685 # Parse date
1686 $year = substr( $ts, 0, 4 );
1687 $month = substr( $ts, 4, 2 );
1688 $day = substr( $ts, 6, 2 );
1689
1690 # Calculate Hebrew year
1691 $hebrewYear = $year + 3760;
1692
1693 # Month number when September = 1, August = 12
1694 $month += 4;
1695 if ( $month > 12 ) {
1696 # Next year
1697 $month -= 12;
1698 $year++;
1699 $hebrewYear++;
1700 }
1701
1702 # Calculate day of year from 1 September
1703 $dayOfYear = $day;
1704 for ( $i = 1; $i < $month; $i++ ) {
1705 if ( $i == 6 ) {
1706 # February
1707 $dayOfYear += 28;
1708 # Check if the year is leap
1709 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1710 $dayOfYear++;
1711 }
1712 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1713 $dayOfYear += 30;
1714 } else {
1715 $dayOfYear += 31;
1716 }
1717 }
1718
1719 # Calculate the start of the Hebrew year
1720 $start = self::hebrewYearStart( $hebrewYear );
1721
1722 # Calculate next year's start
1723 if ( $dayOfYear <= $start ) {
1724 # Day is before the start of the year - it is the previous year
1725 # Next year's start
1726 $nextStart = $start;
1727 # Previous year
1728 $year--;
1729 $hebrewYear--;
1730 # Add days since previous year's 1 September
1731 $dayOfYear += 365;
1732 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1733 # Leap year
1734 $dayOfYear++;
1735 }
1736 # Start of the new (previous) year
1737 $start = self::hebrewYearStart( $hebrewYear );
1738 } else {
1739 # Next year's start
1740 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1741 }
1742
1743 # Calculate Hebrew day of year
1744 $hebrewDayOfYear = $dayOfYear - $start;
1745
1746 # Difference between year's days
1747 $diff = $nextStart - $start;
1748 # Add 12 (or 13 for leap years) days to ignore the difference between
1749 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1750 # difference is only about the year type
1751 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1752 $diff += 13;
1753 } else {
1754 $diff += 12;
1755 }
1756
1757 # Check the year pattern, and is leap year
1758 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1759 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1760 # and non-leap years
1761 $yearPattern = $diff % 30;
1762 # Check if leap year
1763 $isLeap = $diff >= 30;
1764
1765 # Calculate day in the month from number of day in the Hebrew year
1766 # Don't check Adar - if the day is not in Adar, we will stop before;
1767 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1768 $hebrewDay = $hebrewDayOfYear;
1769 $hebrewMonth = 1;
1770 $days = 0;
1771 while ( $hebrewMonth <= 12 ) {
1772 # Calculate days in this month
1773 if ( $isLeap && $hebrewMonth == 6 ) {
1774 # Adar in a leap year
1775 if ( $isLeap ) {
1776 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1777 $days = 30;
1778 if ( $hebrewDay <= $days ) {
1779 # Day in Adar I
1780 $hebrewMonth = 13;
1781 } else {
1782 # Subtract the days of Adar I
1783 $hebrewDay -= $days;
1784 # Try Adar II
1785 $days = 29;
1786 if ( $hebrewDay <= $days ) {
1787 # Day in Adar II
1788 $hebrewMonth = 14;
1789 }
1790 }
1791 }
1792 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1793 # Cheshvan in a complete year (otherwise as the rule below)
1794 $days = 30;
1795 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1796 # Kislev in an incomplete year (otherwise as the rule below)
1797 $days = 29;
1798 } else {
1799 # Odd months have 30 days, even have 29
1800 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1801 }
1802 if ( $hebrewDay <= $days ) {
1803 # In the current month
1804 break;
1805 } else {
1806 # Subtract the days of the current month
1807 $hebrewDay -= $days;
1808 # Try in the next month
1809 $hebrewMonth++;
1810 }
1811 }
1812
1813 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1814 }
1815
1816 /**
1817 * This calculates the Hebrew year start, as days since 1 September.
1818 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1819 * Used for Hebrew date.
1820 *
1821 * @param int $year
1822 *
1823 * @return string
1824 */
1825 private static function hebrewYearStart( $year ) {
1826 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1827 $b = intval( ( $year - 1 ) % 4 );
1828 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1829 if ( $m < 0 ) {
1830 $m--;
1831 }
1832 $Mar = intval( $m );
1833 if ( $m < 0 ) {
1834 $m++;
1835 }
1836 $m -= $Mar;
1837
1838 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1839 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1840 $Mar++;
1841 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1842 $Mar += 2;
1843 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1844 $Mar++;
1845 }
1846
1847 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1848 return $Mar;
1849 }
1850
1851 /**
1852 * Algorithm to convert Gregorian dates to Thai solar dates,
1853 * Minguo dates or Minguo dates.
1854 *
1855 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1856 * https://en.wikipedia.org/wiki/Minguo_calendar
1857 * https://en.wikipedia.org/wiki/Japanese_era_name
1858 *
1859 * @param string $ts 14-character timestamp
1860 * @param string $cName Calender name
1861 * @return array Converted year, month, day
1862 */
1863 private static function tsToYear( $ts, $cName ) {
1864 $gy = substr( $ts, 0, 4 );
1865 $gm = substr( $ts, 4, 2 );
1866 $gd = substr( $ts, 6, 2 );
1867
1868 if ( !strcmp( $cName, 'thai' ) ) {
1869 # Thai solar dates
1870 # Add 543 years to the Gregorian calendar
1871 # Months and days are identical
1872 $gy_offset = $gy + 543;
1873 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1874 # Minguo dates
1875 # Deduct 1911 years from the Gregorian calendar
1876 # Months and days are identical
1877 $gy_offset = $gy - 1911;
1878 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1879 # Nengō dates up to Meiji period
1880 # Deduct years from the Gregorian calendar
1881 # depending on the nengo periods
1882 # Months and days are identical
1883 if ( ( $gy < 1912 )
1884 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1885 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1886 ) {
1887 # Meiji period
1888 $gy_gannen = $gy - 1868 + 1;
1889 $gy_offset = $gy_gannen;
1890 if ( $gy_gannen == 1 ) {
1891 $gy_offset = '元';
1892 }
1893 $gy_offset = '明治' . $gy_offset;
1894 } elseif (
1895 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1896 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1897 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1898 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1899 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1900 ) {
1901 # Taishō period
1902 $gy_gannen = $gy - 1912 + 1;
1903 $gy_offset = $gy_gannen;
1904 if ( $gy_gannen == 1 ) {
1905 $gy_offset = '元';
1906 }
1907 $gy_offset = '大正' . $gy_offset;
1908 } elseif (
1909 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1910 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1911 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1912 ) {
1913 # Shōwa period
1914 $gy_gannen = $gy - 1926 + 1;
1915 $gy_offset = $gy_gannen;
1916 if ( $gy_gannen == 1 ) {
1917 $gy_offset = '元';
1918 }
1919 $gy_offset = '昭和' . $gy_offset;
1920 } else {
1921 # Heisei period
1922 $gy_gannen = $gy - 1989 + 1;
1923 $gy_offset = $gy_gannen;
1924 if ( $gy_gannen == 1 ) {
1925 $gy_offset = '元';
1926 }
1927 $gy_offset = '平成' . $gy_offset;
1928 }
1929 } else {
1930 $gy_offset = $gy;
1931 }
1932
1933 return [ $gy_offset, $gm, $gd ];
1934 }
1935
1936 /**
1937 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1938 *
1939 * This is the rule the BIDI algorithm uses to determine the directionality of
1940 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1941 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1942 *
1943 * TODO: Does not handle BIDI control characters inside the text.
1944 * TODO: Does not handle unallocated characters.
1945 *
1946 * @param string $text Text to test
1947 * @return null|string Directionality ('ltr' or 'rtl') or null
1948 */
1949 private static function strongDirFromContent( $text = '' ) {
1950 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1951 return null;
1952 }
1953 if ( $matches[1] === '' ) {
1954 return 'rtl';
1955 }
1956 return 'ltr';
1957 }
1958
1959 /**
1960 * Roman number formatting up to 10000
1961 *
1962 * @param int $num
1963 *
1964 * @return string
1965 */
1966 static function romanNumeral( $num ) {
1967 static $table = [
1968 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1969 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1970 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1971 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1972 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1973 ];
1974
1975 $num = intval( $num );
1976 if ( $num > 10000 || $num <= 0 ) {
1977 return $num;
1978 }
1979
1980 $s = '';
1981 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1982 if ( $num >= $pow10 ) {
1983 $s .= $table[$i][(int)floor( $num / $pow10 )];
1984 }
1985 $num = $num % $pow10;
1986 }
1987 return $s;
1988 }
1989
1990 /**
1991 * Hebrew Gematria number formatting up to 9999
1992 *
1993 * @param int $num
1994 *
1995 * @return string
1996 */
1997 static function hebrewNumeral( $num ) {
1998 static $table = [
1999 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2000 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2001 [ '',
2002 [ 'ק' ],
2003 [ 'ר' ],
2004 [ 'ש' ],
2005 [ 'ת' ],
2006 [ 'ת', 'ק' ],
2007 [ 'ת', 'ר' ],
2008 [ 'ת', 'ש' ],
2009 [ 'ת', 'ת' ],
2010 [ 'ת', 'ת', 'ק' ],
2011 [ 'ת', 'ת', 'ר' ],
2012 ],
2013 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2014 ];
2015
2016 $num = intval( $num );
2017 if ( $num > 9999 || $num <= 0 ) {
2018 return $num;
2019 }
2020
2021 // Round thousands have special notations
2022 if ( $num === 1000 ) {
2023 return "א' אלף";
2024 } elseif ( $num % 1000 === 0 ) {
2025 return $table[0][$num / 1000] . "' אלפים";
2026 }
2027
2028 $letters = [];
2029
2030 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2031 if ( $num >= $pow10 ) {
2032 if ( $num === 15 || $num === 16 ) {
2033 $letters[] = $table[0][9];
2034 $letters[] = $table[0][$num - 9];
2035 $num = 0;
2036 } else {
2037 $letters = array_merge(
2038 $letters,
2039 (array)$table[$i][intval( $num / $pow10 )]
2040 );
2041
2042 if ( $pow10 === 1000 ) {
2043 $letters[] = "'";
2044 }
2045 }
2046 }
2047
2048 $num = $num % $pow10;
2049 }
2050
2051 $preTransformLength = count( $letters );
2052 if ( $preTransformLength === 1 ) {
2053 // Add geresh (single quote) to one-letter numbers
2054 $letters[] = "'";
2055 } else {
2056 $lastIndex = $preTransformLength - 1;
2057 $letters[$lastIndex] = str_replace(
2058 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2059 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2060 $letters[$lastIndex]
2061 );
2062
2063 // Add gershayim (double quote) to multiple-letter numbers,
2064 // but exclude numbers with only one letter after the thousands
2065 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2066 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2067 $letters[] = "'";
2068 } else {
2069 array_splice( $letters, -1, 0, '"' );
2070 }
2071 }
2072
2073 return implode( $letters );
2074 }
2075
2076 /**
2077 * Used by date() and time() to adjust the time output.
2078 *
2079 * @param string $ts The time in date('YmdHis') format
2080 * @param mixed $tz Adjust the time by this amount (default false, mean we
2081 * get user timecorrection setting)
2082 * @return int
2083 */
2084 public function userAdjust( $ts, $tz = false ) {
2085 global $wgUser, $wgLocalTZoffset;
2086
2087 if ( $tz === false ) {
2088 $tz = $wgUser->getOption( 'timecorrection' );
2089 }
2090
2091 $data = explode( '|', $tz, 3 );
2092
2093 if ( $data[0] == 'ZoneInfo' ) {
2094 MediaWiki\suppressWarnings();
2095 $userTZ = timezone_open( $data[2] );
2096 MediaWiki\restoreWarnings();
2097 if ( $userTZ !== false ) {
2098 $date = date_create( $ts, timezone_open( 'UTC' ) );
2099 date_timezone_set( $date, $userTZ );
2100 $date = date_format( $date, 'YmdHis' );
2101 return $date;
2102 }
2103 # Unrecognized timezone, default to 'Offset' with the stored offset.
2104 $data[0] = 'Offset';
2105 }
2106
2107 if ( $data[0] == 'System' || $tz == '' ) {
2108 # Global offset in minutes.
2109 $minDiff = $wgLocalTZoffset;
2110 } elseif ( $data[0] == 'Offset' ) {
2111 $minDiff = intval( $data[1] );
2112 } else {
2113 $data = explode( ':', $tz );
2114 if ( count( $data ) == 2 ) {
2115 $data[0] = intval( $data[0] );
2116 $data[1] = intval( $data[1] );
2117 $minDiff = abs( $data[0] ) * 60 + $data[1];
2118 if ( $data[0] < 0 ) {
2119 $minDiff = -$minDiff;
2120 }
2121 } else {
2122 $minDiff = intval( $data[0] ) * 60;
2123 }
2124 }
2125
2126 # No difference ? Return time unchanged
2127 if ( 0 == $minDiff ) {
2128 return $ts;
2129 }
2130
2131 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2132 # Generate an adjusted date; take advantage of the fact that mktime
2133 # will normalize out-of-range values so we don't have to split $minDiff
2134 # into hours and minutes.
2135 $t = mktime( (
2136 (int)substr( $ts, 8, 2 ) ), # Hours
2137 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2138 (int)substr( $ts, 12, 2 ), # Seconds
2139 (int)substr( $ts, 4, 2 ), # Month
2140 (int)substr( $ts, 6, 2 ), # Day
2141 (int)substr( $ts, 0, 4 ) ); # Year
2142
2143 $date = date( 'YmdHis', $t );
2144 MediaWiki\restoreWarnings();
2145
2146 return $date;
2147 }
2148
2149 /**
2150 * This is meant to be used by time(), date(), and timeanddate() to get
2151 * the date preference they're supposed to use, it should be used in
2152 * all children.
2153 *
2154 *<code>
2155 * function timeanddate([...], $format = true) {
2156 * $datePreference = $this->dateFormat($format);
2157 * [...]
2158 * }
2159 *</code>
2160 *
2161 * @param int|string|bool $usePrefs If true, the user's preference is used
2162 * if false, the site/language default is used
2163 * if int/string, assumed to be a format.
2164 * @return string
2165 */
2166 function dateFormat( $usePrefs = true ) {
2167 global $wgUser;
2168
2169 if ( is_bool( $usePrefs ) ) {
2170 if ( $usePrefs ) {
2171 $datePreference = $wgUser->getDatePreference();
2172 } else {
2173 $datePreference = (string)User::getDefaultOption( 'date' );
2174 }
2175 } else {
2176 $datePreference = (string)$usePrefs;
2177 }
2178
2179 // return int
2180 if ( $datePreference == '' ) {
2181 return 'default';
2182 }
2183
2184 return $datePreference;
2185 }
2186
2187 /**
2188 * Get a format string for a given type and preference
2189 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2190 * @param string $pref The format name as it appears in Messages*.php under
2191 * $datePreferences.
2192 *
2193 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2194 *
2195 * @return string
2196 */
2197 function getDateFormatString( $type, $pref ) {
2198 $wasDefault = false;
2199 if ( $pref == 'default' ) {
2200 $wasDefault = true;
2201 $pref = $this->getDefaultDateFormat();
2202 }
2203
2204 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2205 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2206
2207 if ( $type === 'pretty' && $df === null ) {
2208 $df = $this->getDateFormatString( 'date', $pref );
2209 }
2210
2211 if ( !$wasDefault && $df === null ) {
2212 $pref = $this->getDefaultDateFormat();
2213 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2214 }
2215
2216 $this->dateFormatStrings[$type][$pref] = $df;
2217 }
2218 return $this->dateFormatStrings[$type][$pref];
2219 }
2220
2221 /**
2222 * @param string $ts The time format which needs to be turned into a
2223 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2224 * @param bool $adj Whether to adjust the time output according to the
2225 * user configured offset ($timecorrection)
2226 * @param mixed $format True to use user's date format preference
2227 * @param string|bool $timecorrection The time offset as returned by
2228 * validateTimeZone() in Special:Preferences
2229 * @return string
2230 */
2231 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2232 $ts = wfTimestamp( TS_MW, $ts );
2233 if ( $adj ) {
2234 $ts = $this->userAdjust( $ts, $timecorrection );
2235 }
2236 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2237 return $this->sprintfDate( $df, $ts );
2238 }
2239
2240 /**
2241 * @param string $ts The time format which needs to be turned into a
2242 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2243 * @param bool $adj Whether to adjust the time output according to the
2244 * user configured offset ($timecorrection)
2245 * @param mixed $format True to use user's date format preference
2246 * @param string|bool $timecorrection The time offset as returned by
2247 * validateTimeZone() in Special:Preferences
2248 * @return string
2249 */
2250 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2251 $ts = wfTimestamp( TS_MW, $ts );
2252 if ( $adj ) {
2253 $ts = $this->userAdjust( $ts, $timecorrection );
2254 }
2255 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2256 return $this->sprintfDate( $df, $ts );
2257 }
2258
2259 /**
2260 * @param string $ts The time format which needs to be turned into a
2261 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2262 * @param bool $adj Whether to adjust the time output according to the
2263 * user configured offset ($timecorrection)
2264 * @param mixed $format What format to return, if it's false output the
2265 * default one (default true)
2266 * @param string|bool $timecorrection The time offset as returned by
2267 * validateTimeZone() in Special:Preferences
2268 * @return string
2269 */
2270 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2271 $ts = wfTimestamp( TS_MW, $ts );
2272 if ( $adj ) {
2273 $ts = $this->userAdjust( $ts, $timecorrection );
2274 }
2275 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2276 return $this->sprintfDate( $df, $ts );
2277 }
2278
2279 /**
2280 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2281 *
2282 * @since 1.20
2283 *
2284 * @param int $seconds The amount of seconds.
2285 * @param array $chosenIntervals The intervals to enable.
2286 *
2287 * @return string
2288 */
2289 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2290 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2291
2292 $segments = [];
2293
2294 foreach ( $intervals as $intervalName => $intervalValue ) {
2295 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2296 // duration-years, duration-decades, duration-centuries, duration-millennia
2297 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2298 $segments[] = $message->inLanguage( $this )->escaped();
2299 }
2300
2301 return $this->listToText( $segments );
2302 }
2303
2304 /**
2305 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2306 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2307 *
2308 * @since 1.20
2309 *
2310 * @param int $seconds The amount of seconds.
2311 * @param array $chosenIntervals The intervals to enable.
2312 *
2313 * @return array
2314 */
2315 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2316 if ( empty( $chosenIntervals ) ) {
2317 $chosenIntervals = [
2318 'millennia',
2319 'centuries',
2320 'decades',
2321 'years',
2322 'days',
2323 'hours',
2324 'minutes',
2325 'seconds'
2326 ];
2327 }
2328
2329 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2330 $sortedNames = array_keys( $intervals );
2331 $smallestInterval = array_pop( $sortedNames );
2332
2333 $segments = [];
2334
2335 foreach ( $intervals as $name => $length ) {
2336 $value = floor( $seconds / $length );
2337
2338 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2339 $seconds -= $value * $length;
2340 $segments[$name] = $value;
2341 }
2342 }
2343
2344 return $segments;
2345 }
2346
2347 /**
2348 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2349 *
2350 * @param string $type Can be 'date', 'time' or 'both'
2351 * @param string $ts The time format which needs to be turned into a
2352 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2353 * @param User $user User object used to get preferences for timezone and format
2354 * @param array $options Array, can contain the following keys:
2355 * - 'timecorrection': time correction, can have the following values:
2356 * - true: use user's preference
2357 * - false: don't use time correction
2358 * - int: value of time correction in minutes
2359 * - 'format': format to use, can have the following values:
2360 * - true: use user's preference
2361 * - false: use default preference
2362 * - string: format to use
2363 * @since 1.19
2364 * @return string
2365 */
2366 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2367 $ts = wfTimestamp( TS_MW, $ts );
2368 $options += [ 'timecorrection' => true, 'format' => true ];
2369 if ( $options['timecorrection'] !== false ) {
2370 if ( $options['timecorrection'] === true ) {
2371 $offset = $user->getOption( 'timecorrection' );
2372 } else {
2373 $offset = $options['timecorrection'];
2374 }
2375 $ts = $this->userAdjust( $ts, $offset );
2376 }
2377 if ( $options['format'] === true ) {
2378 $format = $user->getDatePreference();
2379 } else {
2380 $format = $options['format'];
2381 }
2382 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2383 return $this->sprintfDate( $df, $ts );
2384 }
2385
2386 /**
2387 * Get the formatted date for the given timestamp and formatted for
2388 * the given user.
2389 *
2390 * @param mixed $ts Mixed: the time format which needs to be turned into a
2391 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2392 * @param User $user User object used to get preferences for timezone and format
2393 * @param array $options Array, can contain the following keys:
2394 * - 'timecorrection': time correction, can have the following values:
2395 * - true: use user's preference
2396 * - false: don't use time correction
2397 * - int: value of time correction in minutes
2398 * - 'format': format to use, can have the following values:
2399 * - true: use user's preference
2400 * - false: use default preference
2401 * - string: format to use
2402 * @since 1.19
2403 * @return string
2404 */
2405 public function userDate( $ts, User $user, array $options = [] ) {
2406 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2407 }
2408
2409 /**
2410 * Get the formatted time for the given timestamp and formatted for
2411 * the given user.
2412 *
2413 * @param mixed $ts The time format which needs to be turned into a
2414 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2415 * @param User $user User object used to get preferences for timezone and format
2416 * @param array $options Array, can contain the following keys:
2417 * - 'timecorrection': time correction, can have the following values:
2418 * - true: use user's preference
2419 * - false: don't use time correction
2420 * - int: value of time correction in minutes
2421 * - 'format': format to use, can have the following values:
2422 * - true: use user's preference
2423 * - false: use default preference
2424 * - string: format to use
2425 * @since 1.19
2426 * @return string
2427 */
2428 public function userTime( $ts, User $user, array $options = [] ) {
2429 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2430 }
2431
2432 /**
2433 * Get the formatted date and time for the given timestamp and formatted for
2434 * the given user.
2435 *
2436 * @param mixed $ts The time format which needs to be turned into a
2437 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2438 * @param User $user User object used to get preferences for timezone and format
2439 * @param array $options Array, can contain the following keys:
2440 * - 'timecorrection': time correction, can have the following values:
2441 * - true: use user's preference
2442 * - false: don't use time correction
2443 * - int: value of time correction in minutes
2444 * - 'format': format to use, can have the following values:
2445 * - true: use user's preference
2446 * - false: use default preference
2447 * - string: format to use
2448 * @since 1.19
2449 * @return string
2450 */
2451 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2452 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2453 }
2454
2455 /**
2456 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2457 *
2458 * Determine the difference between the timestamp and the current time, and
2459 * generate a readable timestamp by returning "<N> <units> ago", where the
2460 * largest possible unit is used.
2461 *
2462 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2463 *
2464 * @param MWTimestamp $time
2465 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2466 * @param User|null $user User the timestamp is being generated for
2467 * (or null to use main context's user)
2468 * @return string Formatted timestamp
2469 */
2470 public function getHumanTimestamp(
2471 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2472 ) {
2473 if ( $relativeTo === null ) {
2474 $relativeTo = new MWTimestamp();
2475 }
2476 if ( $user === null ) {
2477 $user = RequestContext::getMain()->getUser();
2478 }
2479
2480 // Adjust for the user's timezone.
2481 $offsetThis = $time->offsetForUser( $user );
2482 $offsetRel = $relativeTo->offsetForUser( $user );
2483
2484 $ts = '';
2485 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2486 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2487 }
2488
2489 // Reset the timezone on the objects.
2490 $time->timestamp->sub( $offsetThis );
2491 $relativeTo->timestamp->sub( $offsetRel );
2492
2493 return $ts;
2494 }
2495
2496 /**
2497 * Convert an MWTimestamp into a pretty human-readable timestamp using
2498 * the given user preferences and relative base time.
2499 *
2500 * @see Language::getHumanTimestamp
2501 * @param MWTimestamp $ts Timestamp to prettify
2502 * @param MWTimestamp $relativeTo Base timestamp
2503 * @param User $user User preferences to use
2504 * @return string Human timestamp
2505 * @since 1.26
2506 */
2507 private function getHumanTimestampInternal(
2508 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2509 ) {
2510 $diff = $ts->diff( $relativeTo );
2511 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2512 (int)$relativeTo->timestamp->format( 'w' ) );
2513 $days = $diff->days ?: (int)$diffDay;
2514 if ( $diff->invert || $days > 5
2515 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2516 ) {
2517 // Timestamps are in different years: use full timestamp
2518 // Also do full timestamp for future dates
2519 /**
2520 * @todo FIXME: Add better handling of future timestamps.
2521 */
2522 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2523 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2524 } elseif ( $days > 5 ) {
2525 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2526 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2527 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2528 } elseif ( $days > 1 ) {
2529 // Timestamp within the past week: show the day of the week and time
2530 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2531 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2532 // Messages:
2533 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2534 $ts = wfMessage( "$weekday-at" )
2535 ->inLanguage( $this )
2536 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2537 ->text();
2538 } elseif ( $days == 1 ) {
2539 // Timestamp was yesterday: say 'yesterday' and the time.
2540 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2541 $ts = wfMessage( 'yesterday-at' )
2542 ->inLanguage( $this )
2543 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2544 ->text();
2545 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2546 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2547 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2548 $ts = wfMessage( 'today-at' )
2549 ->inLanguage( $this )
2550 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2551 ->text();
2552
2553 // From here on in, the timestamp was soon enough ago so that we can simply say
2554 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2555 } elseif ( $diff->h == 1 ) {
2556 // Less than 90 minutes, but more than an hour ago.
2557 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2558 } elseif ( $diff->i >= 1 ) {
2559 // A few minutes ago.
2560 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2561 } elseif ( $diff->s >= 30 ) {
2562 // Less than a minute, but more than 30 sec ago.
2563 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2564 } else {
2565 // Less than 30 seconds ago.
2566 $ts = wfMessage( 'just-now' )->text();
2567 }
2568
2569 return $ts;
2570 }
2571
2572 /**
2573 * @param string $key
2574 * @return array|null
2575 */
2576 public function getMessage( $key ) {
2577 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2578 }
2579
2580 /**
2581 * @return array
2582 */
2583 function getAllMessages() {
2584 return self::$dataCache->getItem( $this->mCode, 'messages' );
2585 }
2586
2587 /**
2588 * @param string $in
2589 * @param string $out
2590 * @param string $string
2591 * @return string
2592 */
2593 public function iconv( $in, $out, $string ) {
2594 # Even with //IGNORE iconv can whine about illegal characters in
2595 # *input* string. We just ignore those too.
2596 # REF: https://bugs.php.net/bug.php?id=37166
2597 # REF: https://phabricator.wikimedia.org/T18885
2598 MediaWiki\suppressWarnings();
2599 $text = iconv( $in, $out . '//IGNORE', $string );
2600 MediaWiki\restoreWarnings();
2601 return $text;
2602 }
2603
2604 // callback functions for ucwords(), ucwordbreaks()
2605
2606 /**
2607 * @param array $matches
2608 * @return mixed|string
2609 */
2610 function ucwordbreaksCallbackAscii( $matches ) {
2611 return $this->ucfirst( $matches[1] );
2612 }
2613
2614 /**
2615 * @param array $matches
2616 * @return string
2617 */
2618 function ucwordbreaksCallbackMB( $matches ) {
2619 return mb_strtoupper( $matches[0] );
2620 }
2621
2622 /**
2623 * @param array $matches
2624 * @return string
2625 */
2626 function ucwordsCallbackMB( $matches ) {
2627 return mb_strtoupper( $matches[0] );
2628 }
2629
2630 /**
2631 * Make a string's first character uppercase
2632 *
2633 * @param string $str
2634 *
2635 * @return string
2636 */
2637 public function ucfirst( $str ) {
2638 $o = ord( $str );
2639 if ( $o < 96 ) { // if already uppercase...
2640 return $str;
2641 } elseif ( $o < 128 ) {
2642 return ucfirst( $str ); // use PHP's ucfirst()
2643 } else {
2644 // fall back to more complex logic in case of multibyte strings
2645 return $this->uc( $str, true );
2646 }
2647 }
2648
2649 /**
2650 * Convert a string to uppercase
2651 *
2652 * @param string $str
2653 * @param bool $first
2654 *
2655 * @return string
2656 */
2657 public function uc( $str, $first = false ) {
2658 if ( $first ) {
2659 if ( $this->isMultibyte( $str ) ) {
2660 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2661 } else {
2662 return ucfirst( $str );
2663 }
2664 } else {
2665 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2666 }
2667 }
2668
2669 /**
2670 * @param string $str
2671 * @return mixed|string
2672 */
2673 function lcfirst( $str ) {
2674 $o = ord( $str );
2675 if ( !$o ) {
2676 return strval( $str );
2677 } elseif ( $o >= 128 ) {
2678 return $this->lc( $str, true );
2679 } elseif ( $o > 96 ) {
2680 return $str;
2681 } else {
2682 $str[0] = strtolower( $str[0] );
2683 return $str;
2684 }
2685 }
2686
2687 /**
2688 * @param string $str
2689 * @param bool $first
2690 * @return mixed|string
2691 */
2692 function lc( $str, $first = false ) {
2693 if ( $first ) {
2694 if ( $this->isMultibyte( $str ) ) {
2695 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2696 } else {
2697 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2698 }
2699 } else {
2700 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2701 }
2702 }
2703
2704 /**
2705 * @param string $str
2706 * @return bool
2707 */
2708 function isMultibyte( $str ) {
2709 return strlen( $str ) !== mb_strlen( $str );
2710 }
2711
2712 /**
2713 * @param string $str
2714 * @return mixed|string
2715 */
2716 function ucwords( $str ) {
2717 if ( $this->isMultibyte( $str ) ) {
2718 $str = $this->lc( $str );
2719
2720 // regexp to find first letter in each word (i.e. after each space)
2721 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2722
2723 // function to use to capitalize a single char
2724 return preg_replace_callback(
2725 $replaceRegexp,
2726 [ $this, 'ucwordsCallbackMB' ],
2727 $str
2728 );
2729 } else {
2730 return ucwords( strtolower( $str ) );
2731 }
2732 }
2733
2734 /**
2735 * capitalize words at word breaks
2736 *
2737 * @param string $str
2738 * @return mixed
2739 */
2740 function ucwordbreaks( $str ) {
2741 if ( $this->isMultibyte( $str ) ) {
2742 $str = $this->lc( $str );
2743
2744 // since \b doesn't work for UTF-8, we explicitely define word break chars
2745 $breaks = "[ \-\(\)\}\{\.,\?!]";
2746
2747 // find first letter after word break
2748 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2749 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2750
2751 return preg_replace_callback(
2752 $replaceRegexp,
2753 [ $this, 'ucwordbreaksCallbackMB' ],
2754 $str
2755 );
2756 } else {
2757 return preg_replace_callback(
2758 '/\b([\w\x80-\xff]+)\b/',
2759 [ $this, 'ucwordbreaksCallbackAscii' ],
2760 $str
2761 );
2762 }
2763 }
2764
2765 /**
2766 * Return a case-folded representation of $s
2767 *
2768 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2769 * and $s2 are the same except for the case of their characters. It is not
2770 * necessary for the value returned to make sense when displayed.
2771 *
2772 * Do *not* perform any other normalisation in this function. If a caller
2773 * uses this function when it should be using a more general normalisation
2774 * function, then fix the caller.
2775 *
2776 * @param string $s
2777 *
2778 * @return string
2779 */
2780 function caseFold( $s ) {
2781 return $this->uc( $s );
2782 }
2783
2784 /**
2785 * @param string $s
2786 * @return string
2787 * @throws MWException
2788 */
2789 function checkTitleEncoding( $s ) {
2790 if ( is_array( $s ) ) {
2791 throw new MWException( 'Given array to checkTitleEncoding.' );
2792 }
2793 if ( StringUtils::isUtf8( $s ) ) {
2794 return $s;
2795 }
2796
2797 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2798 }
2799
2800 /**
2801 * @return array
2802 */
2803 function fallback8bitEncoding() {
2804 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2805 }
2806
2807 /**
2808 * Most writing systems use whitespace to break up words.
2809 * Some languages such as Chinese don't conventionally do this,
2810 * which requires special handling when breaking up words for
2811 * searching etc.
2812 *
2813 * @return bool
2814 */
2815 function hasWordBreaks() {
2816 return true;
2817 }
2818
2819 /**
2820 * Some languages such as Chinese require word segmentation,
2821 * Specify such segmentation when overridden in derived class.
2822 *
2823 * @param string $string
2824 * @return string
2825 */
2826 function segmentByWord( $string ) {
2827 return $string;
2828 }
2829
2830 /**
2831 * Some languages have special punctuation need to be normalized.
2832 * Make such changes here.
2833 *
2834 * @param string $string
2835 * @return string
2836 */
2837 function normalizeForSearch( $string ) {
2838 return self::convertDoubleWidth( $string );
2839 }
2840
2841 /**
2842 * convert double-width roman characters to single-width.
2843 * range: ff00-ff5f ~= 0020-007f
2844 *
2845 * @param string $string
2846 *
2847 * @return string
2848 */
2849 protected static function convertDoubleWidth( $string ) {
2850 static $full = null;
2851 static $half = null;
2852
2853 if ( $full === null ) {
2854 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2855 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2856 $full = str_split( $fullWidth, 3 );
2857 $half = str_split( $halfWidth );
2858 }
2859
2860 $string = str_replace( $full, $half, $string );
2861 return $string;
2862 }
2863
2864 /**
2865 * @param string $string
2866 * @param string $pattern
2867 * @return string
2868 */
2869 protected static function insertSpace( $string, $pattern ) {
2870 $string = preg_replace( $pattern, " $1 ", $string );
2871 $string = preg_replace( '/ +/', ' ', $string );
2872 return $string;
2873 }
2874
2875 /**
2876 * @param array $termsArray
2877 * @return array
2878 */
2879 function convertForSearchResult( $termsArray ) {
2880 # some languages, e.g. Chinese, need to do a conversion
2881 # in order for search results to be displayed correctly
2882 return $termsArray;
2883 }
2884
2885 /**
2886 * Get the first character of a string.
2887 *
2888 * @param string $s
2889 * @return string
2890 */
2891 function firstChar( $s ) {
2892 $matches = [];
2893 preg_match(
2894 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2895 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2896 $s,
2897 $matches
2898 );
2899
2900 if ( isset( $matches[1] ) ) {
2901 if ( strlen( $matches[1] ) != 3 ) {
2902 return $matches[1];
2903 }
2904
2905 // Break down Hangul syllables to grab the first jamo
2906 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2907 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2908 return $matches[1];
2909 } elseif ( $code < 0xb098 ) {
2910 return "\xe3\x84\xb1";
2911 } elseif ( $code < 0xb2e4 ) {
2912 return "\xe3\x84\xb4";
2913 } elseif ( $code < 0xb77c ) {
2914 return "\xe3\x84\xb7";
2915 } elseif ( $code < 0xb9c8 ) {
2916 return "\xe3\x84\xb9";
2917 } elseif ( $code < 0xbc14 ) {
2918 return "\xe3\x85\x81";
2919 } elseif ( $code < 0xc0ac ) {
2920 return "\xe3\x85\x82";
2921 } elseif ( $code < 0xc544 ) {
2922 return "\xe3\x85\x85";
2923 } elseif ( $code < 0xc790 ) {
2924 return "\xe3\x85\x87";
2925 } elseif ( $code < 0xcc28 ) {
2926 return "\xe3\x85\x88";
2927 } elseif ( $code < 0xce74 ) {
2928 return "\xe3\x85\x8a";
2929 } elseif ( $code < 0xd0c0 ) {
2930 return "\xe3\x85\x8b";
2931 } elseif ( $code < 0xd30c ) {
2932 return "\xe3\x85\x8c";
2933 } elseif ( $code < 0xd558 ) {
2934 return "\xe3\x85\x8d";
2935 } else {
2936 return "\xe3\x85\x8e";
2937 }
2938 } else {
2939 return '';
2940 }
2941 }
2942
2943 /**
2944 * @deprecated No-op since 1.28
2945 */
2946 function initEncoding() {
2947 // No-op.
2948 }
2949
2950 /**
2951 * @param string $s
2952 * @return string
2953 * @deprecated No-op since 1.28
2954 */
2955 function recodeForEdit( $s ) {
2956 return $s;
2957 }
2958
2959 /**
2960 * @param string $s
2961 * @return string
2962 * @deprecated No-op since 1.28
2963 */
2964 function recodeInput( $s ) {
2965 return $s;
2966 }
2967
2968 /**
2969 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2970 * also cleans up certain backwards-compatible sequences, converting them
2971 * to the modern Unicode equivalent.
2972 *
2973 * This is language-specific for performance reasons only.
2974 *
2975 * @param string $s
2976 *
2977 * @return string
2978 */
2979 function normalize( $s ) {
2980 global $wgAllUnicodeFixes;
2981 $s = UtfNormal\Validator::cleanUp( $s );
2982 if ( $wgAllUnicodeFixes ) {
2983 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2984 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2985 }
2986
2987 return $s;
2988 }
2989
2990 /**
2991 * Transform a string using serialized data stored in the given file (which
2992 * must be in the serialized subdirectory of $IP). The file contains pairs
2993 * mapping source characters to destination characters.
2994 *
2995 * The data is cached in process memory. This will go faster if you have the
2996 * FastStringSearch extension.
2997 *
2998 * @param string $file
2999 * @param string $string
3000 *
3001 * @throws MWException
3002 * @return string
3003 */
3004 function transformUsingPairFile( $file, $string ) {
3005 if ( !isset( $this->transformData[$file] ) ) {
3006 $data = wfGetPrecompiledData( $file );
3007 if ( $data === false ) {
3008 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
3009 }
3010 $this->transformData[$file] = new ReplacementArray( $data );
3011 }
3012 return $this->transformData[$file]->replace( $string );
3013 }
3014
3015 /**
3016 * For right-to-left language support
3017 *
3018 * @return bool
3019 */
3020 function isRTL() {
3021 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3022 }
3023
3024 /**
3025 * Return the correct HTML 'dir' attribute value for this language.
3026 * @return string
3027 */
3028 function getDir() {
3029 return $this->isRTL() ? 'rtl' : 'ltr';
3030 }
3031
3032 /**
3033 * Return 'left' or 'right' as appropriate alignment for line-start
3034 * for this language's text direction.
3035 *
3036 * Should be equivalent to CSS3 'start' text-align value....
3037 *
3038 * @return string
3039 */
3040 function alignStart() {
3041 return $this->isRTL() ? 'right' : 'left';
3042 }
3043
3044 /**
3045 * Return 'right' or 'left' as appropriate alignment for line-end
3046 * for this language's text direction.
3047 *
3048 * Should be equivalent to CSS3 'end' text-align value....
3049 *
3050 * @return string
3051 */
3052 function alignEnd() {
3053 return $this->isRTL() ? 'left' : 'right';
3054 }
3055
3056 /**
3057 * A hidden direction mark (LRM or RLM), depending on the language direction.
3058 * Unlike getDirMark(), this function returns the character as an HTML entity.
3059 * This function should be used when the output is guaranteed to be HTML,
3060 * because it makes the output HTML source code more readable. When
3061 * the output is plain text or can be escaped, getDirMark() should be used.
3062 *
3063 * @param bool $opposite Get the direction mark opposite to your language
3064 * @return string
3065 * @since 1.20
3066 */
3067 function getDirMarkEntity( $opposite = false ) {
3068 if ( $opposite ) {
3069 return $this->isRTL() ? '&lrm;' : '&rlm;';
3070 }
3071 return $this->isRTL() ? '&rlm;' : '&lrm;';
3072 }
3073
3074 /**
3075 * A hidden direction mark (LRM or RLM), depending on the language direction.
3076 * This function produces them as invisible Unicode characters and
3077 * the output may be hard to read and debug, so it should only be used
3078 * when the output is plain text or can be escaped. When the output is
3079 * HTML, use getDirMarkEntity() instead.
3080 *
3081 * @param bool $opposite Get the direction mark opposite to your language
3082 * @return string
3083 */
3084 function getDirMark( $opposite = false ) {
3085 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3086 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3087 if ( $opposite ) {
3088 return $this->isRTL() ? $lrm : $rlm;
3089 }
3090 return $this->isRTL() ? $rlm : $lrm;
3091 }
3092
3093 /**
3094 * @return array
3095 */
3096 function capitalizeAllNouns() {
3097 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3098 }
3099
3100 /**
3101 * An arrow, depending on the language direction.
3102 *
3103 * @param string $direction The direction of the arrow: forwards (default),
3104 * backwards, left, right, up, down.
3105 * @return string
3106 */
3107 function getArrow( $direction = 'forwards' ) {
3108 switch ( $direction ) {
3109 case 'forwards':
3110 return $this->isRTL() ? '←' : '→';
3111 case 'backwards':
3112 return $this->isRTL() ? '→' : '←';
3113 case 'left':
3114 return '←';
3115 case 'right':
3116 return '→';
3117 case 'up':
3118 return '↑';
3119 case 'down':
3120 return '↓';
3121 }
3122 }
3123
3124 /**
3125 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3126 *
3127 * @return bool
3128 */
3129 function linkPrefixExtension() {
3130 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3131 }
3132
3133 /**
3134 * Get all magic words from cache.
3135 * @return array
3136 */
3137 function getMagicWords() {
3138 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3139 }
3140
3141 /**
3142 * Run the LanguageGetMagic hook once.
3143 */
3144 protected function doMagicHook() {
3145 if ( $this->mMagicHookDone ) {
3146 return;
3147 }
3148 $this->mMagicHookDone = true;
3149 Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
3150 }
3151
3152 /**
3153 * Fill a MagicWord object with data from here
3154 *
3155 * @param MagicWord $mw
3156 */
3157 function getMagic( $mw ) {
3158 // Saves a function call
3159 if ( !$this->mMagicHookDone ) {
3160 $this->doMagicHook();
3161 }
3162
3163 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3164 $rawEntry = $this->mMagicExtensions[$mw->mId];
3165 } else {
3166 $rawEntry = self::$dataCache->getSubitem(
3167 $this->mCode, 'magicWords', $mw->mId );
3168 }
3169
3170 if ( !is_array( $rawEntry ) ) {
3171 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3172 } else {
3173 $mw->mCaseSensitive = $rawEntry[0];
3174 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3175 }
3176 }
3177
3178 /**
3179 * Add magic words to the extension array
3180 *
3181 * @param array $newWords
3182 */
3183 function addMagicWordsByLang( $newWords ) {
3184 $fallbackChain = $this->getFallbackLanguages();
3185 $fallbackChain = array_reverse( $fallbackChain );
3186 foreach ( $fallbackChain as $code ) {
3187 if ( isset( $newWords[$code] ) ) {
3188 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3189 }
3190 }
3191 }
3192
3193 /**
3194 * Get special page names, as an associative array
3195 * canonical name => array of valid names, including aliases
3196 * @return array
3197 */
3198 function getSpecialPageAliases() {
3199 // Cache aliases because it may be slow to load them
3200 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3201 // Initialise array
3202 $this->mExtendedSpecialPageAliases =
3203 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3204 Hooks::run( 'LanguageGetSpecialPageAliases',
3205 [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
3206 }
3207
3208 return $this->mExtendedSpecialPageAliases;
3209 }
3210
3211 /**
3212 * Italic is unsuitable for some languages
3213 *
3214 * @param string $text The text to be emphasized.
3215 * @return string
3216 */
3217 function emphasize( $text ) {
3218 return "<em>$text</em>";
3219 }
3220
3221 /**
3222 * Normally we output all numbers in plain en_US style, that is
3223 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3224 * point twohundredthirtyfive. However this is not suitable for all
3225 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3226 * Icelandic just want to use commas instead of dots, and dots instead
3227 * of commas like "293.291,235".
3228 *
3229 * An example of this function being called:
3230 * <code>
3231 * wfMessage( 'message' )->numParams( $num )->text()
3232 * </code>
3233 *
3234 * See $separatorTransformTable on MessageIs.php for
3235 * the , => . and . => , implementation.
3236 *
3237 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3238 * @param int|float $number The string to be formatted, should be an integer
3239 * or a floating point number.
3240 * @param bool $nocommafy Set to true for special numbers like dates
3241 * @return string
3242 */
3243 public function formatNum( $number, $nocommafy = false ) {
3244 global $wgTranslateNumerals;
3245 if ( !$nocommafy ) {
3246 $number = $this->commafy( $number );
3247 $s = $this->separatorTransformTable();
3248 if ( $s ) {
3249 $number = strtr( $number, $s );
3250 }
3251 }
3252
3253 if ( $wgTranslateNumerals ) {
3254 $s = $this->digitTransformTable();
3255 if ( $s ) {
3256 $number = strtr( $number, $s );
3257 }
3258 }
3259
3260 return $number;
3261 }
3262
3263 /**
3264 * Front-end for non-commafied formatNum
3265 *
3266 * @param int|float $number The string to be formatted, should be an integer
3267 * or a floating point number.
3268 * @since 1.21
3269 * @return string
3270 */
3271 public function formatNumNoSeparators( $number ) {
3272 return $this->formatNum( $number, true );
3273 }
3274
3275 /**
3276 * @param string $number
3277 * @return string
3278 */
3279 public function parseFormattedNumber( $number ) {
3280 $s = $this->digitTransformTable();
3281 if ( $s ) {
3282 // eliminate empty array values such as ''. (bug 64347)
3283 $s = array_filter( $s );
3284 $number = strtr( $number, array_flip( $s ) );
3285 }
3286
3287 $s = $this->separatorTransformTable();
3288 if ( $s ) {
3289 // eliminate empty array values such as ''. (bug 64347)
3290 $s = array_filter( $s );
3291 $number = strtr( $number, array_flip( $s ) );
3292 }
3293
3294 $number = strtr( $number, [ ',' => '' ] );
3295 return $number;
3296 }
3297
3298 /**
3299 * Adds commas to a given number
3300 * @since 1.19
3301 * @param mixed $number
3302 * @return string
3303 */
3304 function commafy( $number ) {
3305 $digitGroupingPattern = $this->digitGroupingPattern();
3306 if ( $number === null ) {
3307 return '';
3308 }
3309
3310 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3311 // default grouping is at thousands, use the same for ###,###,### pattern too.
3312 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3313 } else {
3314 // Ref: http://cldr.unicode.org/translation/number-patterns
3315 $sign = "";
3316 if ( intval( $number ) < 0 ) {
3317 // For negative numbers apply the algorithm like positive number and add sign.
3318 $sign = "-";
3319 $number = substr( $number, 1 );
3320 }
3321 $integerPart = [];
3322 $decimalPart = [];
3323 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3324 preg_match( "/\d+/", $number, $integerPart );
3325 preg_match( "/\.\d*/", $number, $decimalPart );
3326 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3327 if ( $groupedNumber === $number ) {
3328 // the string does not have any number part. Eg: .12345
3329 return $sign . $groupedNumber;
3330 }
3331 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3332 while ( $start > 0 ) {
3333 $match = $matches[0][$numMatches - 1];
3334 $matchLen = strlen( $match );
3335 $start = $end - $matchLen;
3336 if ( $start < 0 ) {
3337 $start = 0;
3338 }
3339 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3340 $end = $start;
3341 if ( $numMatches > 1 ) {
3342 // use the last pattern for the rest of the number
3343 $numMatches--;
3344 }
3345 if ( $start > 0 ) {
3346 $groupedNumber = "," . $groupedNumber;
3347 }
3348 }
3349 return $sign . $groupedNumber;
3350 }
3351 }
3352
3353 /**
3354 * @return string
3355 */
3356 function digitGroupingPattern() {
3357 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3358 }
3359
3360 /**
3361 * @return array
3362 */
3363 function digitTransformTable() {
3364 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3365 }
3366
3367 /**
3368 * @return array
3369 */
3370 function separatorTransformTable() {
3371 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3372 }
3373
3374 /**
3375 * Take a list of strings and build a locale-friendly comma-separated
3376 * list, using the local comma-separator message.
3377 * The last two strings are chained with an "and".
3378 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3379 *
3380 * @param string[] $l
3381 * @return string
3382 */
3383 function listToText( array $l ) {
3384 $m = count( $l ) - 1;
3385 if ( $m < 0 ) {
3386 return '';
3387 }
3388 if ( $m > 0 ) {
3389 $and = $this->msg( 'and' )->escaped();
3390 $space = $this->msg( 'word-separator' )->escaped();
3391 if ( $m > 1 ) {
3392 $comma = $this->msg( 'comma-separator' )->escaped();
3393 }
3394 }
3395 $s = $l[$m];
3396 for ( $i = $m - 1; $i >= 0; $i-- ) {
3397 if ( $i == $m - 1 ) {
3398 $s = $l[$i] . $and . $space . $s;
3399 } else {
3400 $s = $l[$i] . $comma . $s;
3401 }
3402 }
3403 return $s;
3404 }
3405
3406 /**
3407 * Take a list of strings and build a locale-friendly comma-separated
3408 * list, using the local comma-separator message.
3409 * @param string[] $list Array of strings to put in a comma list
3410 * @return string
3411 */
3412 function commaList( array $list ) {
3413 return implode(
3414 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3415 $list
3416 );
3417 }
3418
3419 /**
3420 * Take a list of strings and build a locale-friendly semicolon-separated
3421 * list, using the local semicolon-separator message.
3422 * @param string[] $list Array of strings to put in a semicolon list
3423 * @return string
3424 */
3425 function semicolonList( array $list ) {
3426 return implode(
3427 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3428 $list
3429 );
3430 }
3431
3432 /**
3433 * Same as commaList, but separate it with the pipe instead.
3434 * @param string[] $list Array of strings to put in a pipe list
3435 * @return string
3436 */
3437 function pipeList( array $list ) {
3438 return implode(
3439 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3440 $list
3441 );
3442 }
3443
3444 /**
3445 * Truncate a string to a specified length in bytes, appending an optional
3446 * string (e.g. for ellipses)
3447 *
3448 * The database offers limited byte lengths for some columns in the database;
3449 * multi-byte character sets mean we need to ensure that only whole characters
3450 * are included, otherwise broken characters can be passed to the user
3451 *
3452 * If $length is negative, the string will be truncated from the beginning
3453 *
3454 * @param string $string String to truncate
3455 * @param int $length Maximum length (including ellipses)
3456 * @param string $ellipsis String to append to the truncated text
3457 * @param bool $adjustLength Subtract length of ellipsis from $length.
3458 * $adjustLength was introduced in 1.18, before that behaved as if false.
3459 * @return string
3460 */
3461 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3462 # Use the localized ellipsis character
3463 if ( $ellipsis == '...' ) {
3464 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3465 }
3466 # Check if there is no need to truncate
3467 if ( $length == 0 ) {
3468 return $ellipsis; // convention
3469 } elseif ( strlen( $string ) <= abs( $length ) ) {
3470 return $string; // no need to truncate
3471 }
3472 $stringOriginal = $string;
3473 # If ellipsis length is >= $length then we can't apply $adjustLength
3474 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3475 $string = $ellipsis; // this can be slightly unexpected
3476 # Otherwise, truncate and add ellipsis...
3477 } else {
3478 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3479 if ( $length > 0 ) {
3480 $length -= $eLength;
3481 $string = substr( $string, 0, $length ); // xyz...
3482 $string = $this->removeBadCharLast( $string );
3483 $string = rtrim( $string );
3484 $string = $string . $ellipsis;
3485 } else {
3486 $length += $eLength;
3487 $string = substr( $string, $length ); // ...xyz
3488 $string = $this->removeBadCharFirst( $string );
3489 $string = ltrim( $string );
3490 $string = $ellipsis . $string;
3491 }
3492 }
3493 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3494 # This check is *not* redundant if $adjustLength, due to the single case where
3495 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3496 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3497 return $string;
3498 } else {
3499 return $stringOriginal;
3500 }
3501 }
3502
3503 /**
3504 * Remove bytes that represent an incomplete Unicode character
3505 * at the end of string (e.g. bytes of the char are missing)
3506 *
3507 * @param string $string
3508 * @return string
3509 */
3510 protected function removeBadCharLast( $string ) {
3511 if ( $string != '' ) {
3512 $char = ord( $string[strlen( $string ) - 1] );
3513 $m = [];
3514 if ( $char >= 0xc0 ) {
3515 # We got the first byte only of a multibyte char; remove it.
3516 $string = substr( $string, 0, -1 );
3517 } elseif ( $char >= 0x80 &&
3518 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3519 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3520 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3521 ) {
3522 # We chopped in the middle of a character; remove it
3523 $string = $m[1];
3524 }
3525 }
3526 return $string;
3527 }
3528
3529 /**
3530 * Remove bytes that represent an incomplete Unicode character
3531 * at the start of string (e.g. bytes of the char are missing)
3532 *
3533 * @param string $string
3534 * @return string
3535 */
3536 protected function removeBadCharFirst( $string ) {
3537 if ( $string != '' ) {
3538 $char = ord( $string[0] );
3539 if ( $char >= 0x80 && $char < 0xc0 ) {
3540 # We chopped in the middle of a character; remove the whole thing
3541 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3542 }
3543 }
3544 return $string;
3545 }
3546
3547 /**
3548 * Truncate a string of valid HTML to a specified length in bytes,
3549 * appending an optional string (e.g. for ellipses), and return valid HTML
3550 *
3551 * This is only intended for styled/linked text, such as HTML with
3552 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3553 * Also, this will not detect things like "display:none" CSS.
3554 *
3555 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3556 *
3557 * @param string $text HTML string to truncate
3558 * @param int $length (zero/positive) Maximum length (including ellipses)
3559 * @param string $ellipsis String to append to the truncated text
3560 * @return string
3561 */
3562 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3563 # Use the localized ellipsis character
3564 if ( $ellipsis == '...' ) {
3565 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3566 }
3567 # Check if there is clearly no need to truncate
3568 if ( $length <= 0 ) {
3569 return $ellipsis; // no text shown, nothing to format (convention)
3570 } elseif ( strlen( $text ) <= $length ) {
3571 return $text; // string short enough even *with* HTML (short-circuit)
3572 }
3573
3574 $dispLen = 0; // innerHTML legth so far
3575 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3576 $tagType = 0; // 0-open, 1-close
3577 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3578 $entityState = 0; // 0-not entity, 1-entity
3579 $tag = $ret = ''; // accumulated tag name, accumulated result string
3580 $openTags = []; // open tag stack
3581 $maybeState = null; // possible truncation state
3582
3583 $textLen = strlen( $text );
3584 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3585 for ( $pos = 0; true; ++$pos ) {
3586 # Consider truncation once the display length has reached the maximim.
3587 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3588 # Check that we're not in the middle of a bracket/entity...
3589 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3590 if ( !$testingEllipsis ) {
3591 $testingEllipsis = true;
3592 # Save where we are; we will truncate here unless there turn out to
3593 # be so few remaining characters that truncation is not necessary.
3594 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3595 $maybeState = [ $ret, $openTags ]; // save state
3596 }
3597 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3598 # String in fact does need truncation, the truncation point was OK.
3599 list( $ret, $openTags ) = $maybeState; // reload state
3600 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3601 $ret .= $ellipsis; // add ellipsis
3602 break;
3603 }
3604 }
3605 if ( $pos >= $textLen ) {
3606 break; // extra iteration just for above checks
3607 }
3608
3609 # Read the next char...
3610 $ch = $text[$pos];
3611 $lastCh = $pos ? $text[$pos - 1] : '';
3612 $ret .= $ch; // add to result string
3613 if ( $ch == '<' ) {
3614 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3615 $entityState = 0; // for bad HTML
3616 $bracketState = 1; // tag started (checking for backslash)
3617 } elseif ( $ch == '>' ) {
3618 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3619 $entityState = 0; // for bad HTML
3620 $bracketState = 0; // out of brackets
3621 } elseif ( $bracketState == 1 ) {
3622 if ( $ch == '/' ) {
3623 $tagType = 1; // close tag (e.g. "</span>")
3624 } else {
3625 $tagType = 0; // open tag (e.g. "<span>")
3626 $tag .= $ch;
3627 }
3628 $bracketState = 2; // building tag name
3629 } elseif ( $bracketState == 2 ) {
3630 if ( $ch != ' ' ) {
3631 $tag .= $ch;
3632 } else {
3633 // Name found (e.g. "<a href=..."), add on tag attributes...
3634 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3635 }
3636 } elseif ( $bracketState == 0 ) {
3637 if ( $entityState ) {
3638 if ( $ch == ';' ) {
3639 $entityState = 0;
3640 $dispLen++; // entity is one displayed char
3641 }
3642 } else {
3643 if ( $neLength == 0 && !$maybeState ) {
3644 // Save state without $ch. We want to *hit* the first
3645 // display char (to get tags) but not *use* it if truncating.
3646 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3647 }
3648 if ( $ch == '&' ) {
3649 $entityState = 1; // entity found, (e.g. "&#160;")
3650 } else {
3651 $dispLen++; // this char is displayed
3652 // Add the next $max display text chars after this in one swoop...
3653 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3654 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3655 $dispLen += $skipped;
3656 $pos += $skipped;
3657 }
3658 }
3659 }
3660 }
3661 // Close the last tag if left unclosed by bad HTML
3662 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3663 while ( count( $openTags ) > 0 ) {
3664 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3665 }
3666 return $ret;
3667 }
3668
3669 /**
3670 * truncateHtml() helper function
3671 * like strcspn() but adds the skipped chars to $ret
3672 *
3673 * @param string $ret
3674 * @param string $text
3675 * @param string $search
3676 * @param int $start
3677 * @param null|int $len
3678 * @return int
3679 */
3680 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3681 if ( $len === null ) {
3682 $len = -1; // -1 means "no limit" for strcspn
3683 } elseif ( $len < 0 ) {
3684 $len = 0; // sanity
3685 }
3686 $skipCount = 0;
3687 if ( $start < strlen( $text ) ) {
3688 $skipCount = strcspn( $text, $search, $start, $len );
3689 $ret .= substr( $text, $start, $skipCount );
3690 }
3691 return $skipCount;
3692 }
3693
3694 /**
3695 * truncateHtml() helper function
3696 * (a) push or pop $tag from $openTags as needed
3697 * (b) clear $tag value
3698 * @param string &$tag Current HTML tag name we are looking at
3699 * @param int $tagType (0-open tag, 1-close tag)
3700 * @param string $lastCh Character before the '>' that ended this tag
3701 * @param array &$openTags Open tag stack (not accounting for $tag)
3702 */
3703 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3704 $tag = ltrim( $tag );
3705 if ( $tag != '' ) {
3706 if ( $tagType == 0 && $lastCh != '/' ) {
3707 $openTags[] = $tag; // tag opened (didn't close itself)
3708 } elseif ( $tagType == 1 ) {
3709 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3710 array_pop( $openTags ); // tag closed
3711 }
3712 }
3713 $tag = '';
3714 }
3715 }
3716
3717 /**
3718 * Grammatical transformations, needed for inflected languages
3719 * Invoked by putting {{grammar:case|word}} in a message
3720 *
3721 * @param string $word
3722 * @param string $case
3723 * @return string
3724 */
3725 function convertGrammar( $word, $case ) {
3726 global $wgGrammarForms;
3727 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3728 return $wgGrammarForms[$this->getCode()][$case][$word];
3729 }
3730
3731 return $word;
3732 }
3733 /**
3734 * Get the grammar forms for the content language
3735 * @return array Array of grammar forms
3736 * @since 1.20
3737 */
3738 function getGrammarForms() {
3739 global $wgGrammarForms;
3740 if ( isset( $wgGrammarForms[$this->getCode()] )
3741 && is_array( $wgGrammarForms[$this->getCode()] )
3742 ) {
3743 return $wgGrammarForms[$this->getCode()];
3744 }
3745
3746 return [];
3747 }
3748 /**
3749 * Provides an alternative text depending on specified gender.
3750 * Usage {{gender:username|masculine|feminine|unknown}}.
3751 * username is optional, in which case the gender of current user is used,
3752 * but only in (some) interface messages; otherwise default gender is used.
3753 *
3754 * If no forms are given, an empty string is returned. If only one form is
3755 * given, it will be returned unconditionally. These details are implied by
3756 * the caller and cannot be overridden in subclasses.
3757 *
3758 * If three forms are given, the default is to use the third (unknown) form.
3759 * If fewer than three forms are given, the default is to use the first (masculine) form.
3760 * These details can be overridden in subclasses.
3761 *
3762 * @param string $gender
3763 * @param array $forms
3764 *
3765 * @return string
3766 */
3767 function gender( $gender, $forms ) {
3768 if ( !count( $forms ) ) {
3769 return '';
3770 }
3771 $forms = $this->preConvertPlural( $forms, 2 );
3772 if ( $gender === 'male' ) {
3773 return $forms[0];
3774 }
3775 if ( $gender === 'female' ) {
3776 return $forms[1];
3777 }
3778 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3779 }
3780
3781 /**
3782 * Plural form transformations, needed for some languages.
3783 * For example, there are 3 form of plural in Russian and Polish,
3784 * depending on "count mod 10". See [[w:Plural]]
3785 * For English it is pretty simple.
3786 *
3787 * Invoked by putting {{plural:count|wordform1|wordform2}}
3788 * or {{plural:count|wordform1|wordform2|wordform3}}
3789 *
3790 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3791 *
3792 * @param int $count Non-localized number
3793 * @param array $forms Different plural forms
3794 * @return string Correct form of plural for $count in this language
3795 */
3796 function convertPlural( $count, $forms ) {
3797 // Handle explicit n=pluralform cases
3798 $forms = $this->handleExplicitPluralForms( $count, $forms );
3799 if ( is_string( $forms ) ) {
3800 return $forms;
3801 }
3802 if ( !count( $forms ) ) {
3803 return '';
3804 }
3805
3806 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3807 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3808 return $forms[$pluralForm];
3809 }
3810
3811 /**
3812 * Handles explicit plural forms for Language::convertPlural()
3813 *
3814 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3815 * If an explicitly defined plural form matches the $count, then
3816 * string value returned, otherwise array returned for further consideration
3817 * by CLDR rules or overridden convertPlural().
3818 *
3819 * @since 1.23
3820 *
3821 * @param int $count Non-localized number
3822 * @param array $forms Different plural forms
3823 *
3824 * @return array|string
3825 */
3826 protected function handleExplicitPluralForms( $count, array $forms ) {
3827 foreach ( $forms as $index => $form ) {
3828 if ( preg_match( '/\d+=/i', $form ) ) {
3829 $pos = strpos( $form, '=' );
3830 if ( substr( $form, 0, $pos ) === (string)$count ) {
3831 return substr( $form, $pos + 1 );
3832 }
3833 unset( $forms[$index] );
3834 }
3835 }
3836 return array_values( $forms );
3837 }
3838
3839 /**
3840 * Checks that convertPlural was given an array and pads it to requested
3841 * amount of forms by copying the last one.
3842 *
3843 * @param array $forms Array of forms given to convertPlural
3844 * @param int $count How many forms should there be at least
3845 * @return array Padded array of forms or an exception if not an array
3846 */
3847 protected function preConvertPlural( /* Array */ $forms, $count ) {
3848 while ( count( $forms ) < $count ) {
3849 $forms[] = $forms[count( $forms ) - 1];
3850 }
3851 return $forms;
3852 }
3853
3854 /**
3855 * Wraps argument with unicode control characters for directionality safety
3856 *
3857 * This solves the problem where directionality-neutral characters at the edge of
3858 * the argument string get interpreted with the wrong directionality from the
3859 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3860 *
3861 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3862 * directionality of the argument string, using the BIDI algorithm's own "First
3863 * strong directional codepoint" rule. Essentially, this works round the fact that
3864 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3865 * direction inference). The latter is cleaner but still not widely supported.
3866 *
3867 * @param string $text Text to wrap
3868 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3869 */
3870 public function embedBidi( $text = '' ) {
3871 $dir = Language::strongDirFromContent( $text );
3872 if ( $dir === 'ltr' ) {
3873 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3874 return self::$lre . $text . self::$pdf;
3875 }
3876 if ( $dir === 'rtl' ) {
3877 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3878 return self::$rle . $text . self::$pdf;
3879 }
3880 // No strong directionality: do not wrap
3881 return $text;
3882 }
3883
3884 /**
3885 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3886 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3887 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3888 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3889 * match up with it.
3890 *
3891 * @param string $str The validated block duration in English
3892 * @param User $user User object to use timezone from or null for $wgUser
3893 * @return string Somehow translated block duration
3894 * @see LanguageFi.php for example implementation
3895 */
3896 function translateBlockExpiry( $str, User $user = null ) {
3897 $duration = SpecialBlock::getSuggestedDurations( $this );
3898 foreach ( $duration as $show => $value ) {
3899 if ( strcmp( $str, $value ) == 0 ) {
3900 return htmlspecialchars( trim( $show ) );
3901 }
3902 }
3903
3904 if ( wfIsInfinity( $str ) ) {
3905 foreach ( $duration as $show => $value ) {
3906 if ( wfIsInfinity( $value ) ) {
3907 return htmlspecialchars( trim( $show ) );
3908 }
3909 }
3910 }
3911
3912 // If all else fails, return a standard duration or timestamp description.
3913 $time = strtotime( $str, 0 );
3914 if ( $time === false ) { // Unknown format. Return it as-is in case.
3915 return $str;
3916 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3917 // $time is relative to 0 so it's a duration length.
3918 return $this->formatDuration( $time );
3919 } else { // It's an absolute timestamp.
3920 if ( $time === 0 ) {
3921 // wfTimestamp() handles 0 as current time instead of epoch.
3922 $time = '19700101000000';
3923 }
3924 if ( $user ) {
3925 return $this->userTimeAndDate( $time, $user );
3926 }
3927 return $this->timeanddate( $time );
3928 }
3929 }
3930
3931 /**
3932 * languages like Chinese need to be segmented in order for the diff
3933 * to be of any use
3934 *
3935 * @param string $text
3936 * @return string
3937 */
3938 public function segmentForDiff( $text ) {
3939 return $text;
3940 }
3941
3942 /**
3943 * and unsegment to show the result
3944 *
3945 * @param string $text
3946 * @return string
3947 */
3948 public function unsegmentForDiff( $text ) {
3949 return $text;
3950 }
3951
3952 /**
3953 * Return the LanguageConverter used in the Language
3954 *
3955 * @since 1.19
3956 * @return LanguageConverter
3957 */
3958 public function getConverter() {
3959 return $this->mConverter;
3960 }
3961
3962 /**
3963 * convert text to all supported variants
3964 *
3965 * @param string $text
3966 * @return array
3967 */
3968 public function autoConvertToAllVariants( $text ) {
3969 return $this->mConverter->autoConvertToAllVariants( $text );
3970 }
3971
3972 /**
3973 * convert text to different variants of a language.
3974 *
3975 * @param string $text
3976 * @return string
3977 */
3978 public function convert( $text ) {
3979 return $this->mConverter->convert( $text );
3980 }
3981
3982 /**
3983 * Convert a Title object to a string in the preferred variant
3984 *
3985 * @param Title $title
3986 * @return string
3987 */
3988 public function convertTitle( $title ) {
3989 return $this->mConverter->convertTitle( $title );
3990 }
3991
3992 /**
3993 * Convert a namespace index to a string in the preferred variant
3994 *
3995 * @param int $ns
3996 * @return string
3997 */
3998 public function convertNamespace( $ns ) {
3999 return $this->mConverter->convertNamespace( $ns );
4000 }
4001
4002 /**
4003 * Check if this is a language with variants
4004 *
4005 * @return bool
4006 */
4007 public function hasVariants() {
4008 return count( $this->getVariants() ) > 1;
4009 }
4010
4011 /**
4012 * Check if the language has the specific variant
4013 *
4014 * @since 1.19
4015 * @param string $variant
4016 * @return bool
4017 */
4018 public function hasVariant( $variant ) {
4019 return (bool)$this->mConverter->validateVariant( $variant );
4020 }
4021
4022 /**
4023 * Perform output conversion on a string, and encode for safe HTML output.
4024 * @param string $text Text to be converted
4025 * @param bool $isTitle Whether this conversion is for the article title
4026 * @return string
4027 * @todo this should get integrated somewhere sane
4028 */
4029 public function convertHtml( $text, $isTitle = false ) {
4030 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4031 }
4032
4033 /**
4034 * @param string $key
4035 * @return string
4036 */
4037 public function convertCategoryKey( $key ) {
4038 return $this->mConverter->convertCategoryKey( $key );
4039 }
4040
4041 /**
4042 * Get the list of variants supported by this language
4043 * see sample implementation in LanguageZh.php
4044 *
4045 * @return array An array of language codes
4046 */
4047 public function getVariants() {
4048 return $this->mConverter->getVariants();
4049 }
4050
4051 /**
4052 * @return string
4053 */
4054 public function getPreferredVariant() {
4055 return $this->mConverter->getPreferredVariant();
4056 }
4057
4058 /**
4059 * @return string
4060 */
4061 public function getDefaultVariant() {
4062 return $this->mConverter->getDefaultVariant();
4063 }
4064
4065 /**
4066 * @return string
4067 */
4068 public function getURLVariant() {
4069 return $this->mConverter->getURLVariant();
4070 }
4071
4072 /**
4073 * If a language supports multiple variants, it is
4074 * possible that non-existing link in one variant
4075 * actually exists in another variant. this function
4076 * tries to find it. See e.g. LanguageZh.php
4077 * The input parameters may be modified upon return
4078 *
4079 * @param string &$link The name of the link
4080 * @param Title &$nt The title object of the link
4081 * @param bool $ignoreOtherCond To disable other conditions when
4082 * we need to transclude a template or update a category's link
4083 */
4084 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4085 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4086 }
4087
4088 /**
4089 * returns language specific options used by User::getPageRenderHash()
4090 * for example, the preferred language variant
4091 *
4092 * @return string
4093 */
4094 function getExtraHashOptions() {
4095 return $this->mConverter->getExtraHashOptions();
4096 }
4097
4098 /**
4099 * For languages that support multiple variants, the title of an
4100 * article may be displayed differently in different variants. this
4101 * function returns the apporiate title defined in the body of the article.
4102 *
4103 * @return string
4104 */
4105 public function getParsedTitle() {
4106 return $this->mConverter->getParsedTitle();
4107 }
4108
4109 /**
4110 * Refresh the cache of conversion tables when
4111 * MediaWiki:Conversiontable* is updated.
4112 *
4113 * @param Title $title The Title of the page being updated
4114 */
4115 public function updateConversionTable( Title $title ) {
4116 $this->mConverter->updateConversionTable( $title );
4117 }
4118
4119 /**
4120 * Prepare external link text for conversion. When the text is
4121 * a URL, it shouldn't be converted, and it'll be wrapped in
4122 * the "raw" tag (-{R| }-) to prevent conversion.
4123 *
4124 * This function is called "markNoConversion" for historical
4125 * reasons.
4126 *
4127 * @param string $text Text to be used for external link
4128 * @param bool $noParse Wrap it without confirming it's a real URL first
4129 * @return string The tagged text
4130 */
4131 public function markNoConversion( $text, $noParse = false ) {
4132 // Excluding protocal-relative URLs may avoid many false positives.
4133 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4134 return $this->mConverter->markNoConversion( $text );
4135 } else {
4136 return $text;
4137 }
4138 }
4139
4140 /**
4141 * A regular expression to match legal word-trailing characters
4142 * which should be merged onto a link of the form [[foo]]bar.
4143 *
4144 * @return string
4145 */
4146 public function linkTrail() {
4147 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4148 }
4149
4150 /**
4151 * A regular expression character set to match legal word-prefixing
4152 * characters which should be merged onto a link of the form foo[[bar]].
4153 *
4154 * @return string
4155 */
4156 public function linkPrefixCharset() {
4157 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4158 }
4159
4160 /**
4161 * Get the "parent" language which has a converter to convert a "compatible" language
4162 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4163 *
4164 * @return Language|null
4165 * @since 1.22
4166 */
4167 public function getParentLanguage() {
4168 if ( $this->mParentLanguage !== false ) {
4169 return $this->mParentLanguage;
4170 }
4171
4172 $code = explode( '-', $this->getCode() )[0];
4173 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4174 $this->mParentLanguage = null;
4175 return null;
4176 }
4177 $lang = Language::factory( $code );
4178 if ( !$lang->hasVariant( $this->getCode() ) ) {
4179 $this->mParentLanguage = null;
4180 return null;
4181 }
4182
4183 $this->mParentLanguage = $lang;
4184 return $lang;
4185 }
4186
4187 /**
4188 * Compare with an other language object
4189 *
4190 * @since 1.28
4191 * @param Language $lang
4192 * @return boolean
4193 */
4194 public function equals( Language $lang ) {
4195 return $lang->getCode() === $this->mCode;
4196 }
4197
4198 /**
4199 * Get the internal language code for this language object
4200 *
4201 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4202 * htmlspecialchars() or similar
4203 *
4204 * @return string
4205 */
4206 public function getCode() {
4207 return $this->mCode;
4208 }
4209
4210 /**
4211 * Get the code in BCP 47 format which we can use
4212 * inside of html lang="" tags.
4213 *
4214 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4215 * htmlspecialchars() or similar.
4216 *
4217 * @since 1.19
4218 * @return string
4219 */
4220 public function getHtmlCode() {
4221 if ( is_null( $this->mHtmlCode ) ) {
4222 $this->mHtmlCode = wfBCP47( $this->getCode() );
4223 }
4224 return $this->mHtmlCode;
4225 }
4226
4227 /**
4228 * @param string $code
4229 */
4230 public function setCode( $code ) {
4231 $this->mCode = $code;
4232 // Ensure we don't leave incorrect cached data lying around
4233 $this->mHtmlCode = null;
4234 $this->mParentLanguage = false;
4235 }
4236
4237 /**
4238 * Get the language code from a file name. Inverse of getFileName()
4239 * @param string $filename $prefix . $languageCode . $suffix
4240 * @param string $prefix Prefix before the language code
4241 * @param string $suffix Suffix after the language code
4242 * @return string Language code, or false if $prefix or $suffix isn't found
4243 */
4244 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4245 $m = null;
4246 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4247 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4248 if ( !count( $m ) ) {
4249 return false;
4250 }
4251 return str_replace( '_', '-', strtolower( $m[1] ) );
4252 }
4253
4254 /**
4255 * @param string $code
4256 * @return string Name of the language class
4257 */
4258 public static function classFromCode( $code ) {
4259 if ( $code == 'en' ) {
4260 return 'Language';
4261 } else {
4262 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4263 }
4264 }
4265
4266 /**
4267 * Get the name of a file for a certain language code
4268 * @param string $prefix Prepend this to the filename
4269 * @param string $code Language code
4270 * @param string $suffix Append this to the filename
4271 * @throws MWException
4272 * @return string $prefix . $mangledCode . $suffix
4273 */
4274 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4275 if ( !self::isValidBuiltInCode( $code ) ) {
4276 throw new MWException( "Invalid language code \"$code\"" );
4277 }
4278
4279 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4280 }
4281
4282 /**
4283 * @param string $code
4284 * @return string
4285 */
4286 public static function getMessagesFileName( $code ) {
4287 global $IP;
4288 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4289 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4290 return $file;
4291 }
4292
4293 /**
4294 * @param string $code
4295 * @return string
4296 * @throws MWException
4297 * @since 1.23
4298 */
4299 public static function getJsonMessagesFileName( $code ) {
4300 global $IP;
4301
4302 if ( !self::isValidBuiltInCode( $code ) ) {
4303 throw new MWException( "Invalid language code \"$code\"" );
4304 }
4305
4306 return "$IP/languages/i18n/$code.json";
4307 }
4308
4309 /**
4310 * Get the first fallback for a given language.
4311 *
4312 * @param string $code
4313 *
4314 * @return bool|string
4315 */
4316 public static function getFallbackFor( $code ) {
4317 $fallbacks = self::getFallbacksFor( $code );
4318 if ( $fallbacks ) {
4319 return $fallbacks[0];
4320 }
4321 return false;
4322 }
4323
4324 /**
4325 * Get the ordered list of fallback languages.
4326 *
4327 * @since 1.19
4328 * @param string $code Language code
4329 * @return array Non-empty array, ending in "en"
4330 */
4331 public static function getFallbacksFor( $code ) {
4332 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4333 return [];
4334 }
4335 // For unknown languages, fallbackSequence returns an empty array,
4336 // hardcode fallback to 'en' in that case.
4337 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4338 }
4339
4340 /**
4341 * Get the ordered list of fallback languages, ending with the fallback
4342 * language chain for the site language.
4343 *
4344 * @since 1.22
4345 * @param string $code Language code
4346 * @return array Array( fallbacks, site fallbacks )
4347 */
4348 public static function getFallbacksIncludingSiteLanguage( $code ) {
4349 global $wgLanguageCode;
4350
4351 // Usually, we will only store a tiny number of fallback chains, so we
4352 // keep them in static memory.
4353 $cacheKey = "{$code}-{$wgLanguageCode}";
4354
4355 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4356 $fallbacks = self::getFallbacksFor( $code );
4357
4358 // Append the site's fallback chain, including the site language itself
4359 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4360 array_unshift( $siteFallbacks, $wgLanguageCode );
4361
4362 // Eliminate any languages already included in the chain
4363 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4364
4365 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4366 }
4367 return self::$fallbackLanguageCache[$cacheKey];
4368 }
4369
4370 /**
4371 * Get all messages for a given language
4372 * WARNING: this may take a long time. If you just need all message *keys*
4373 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4374 *
4375 * @param string $code
4376 *
4377 * @return array
4378 */
4379 public static function getMessagesFor( $code ) {
4380 return self::getLocalisationCache()->getItem( $code, 'messages' );
4381 }
4382
4383 /**
4384 * Get a message for a given language
4385 *
4386 * @param string $key
4387 * @param string $code
4388 *
4389 * @return string
4390 */
4391 public static function getMessageFor( $key, $code ) {
4392 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4393 }
4394
4395 /**
4396 * Get all message keys for a given language. This is a faster alternative to
4397 * array_keys( Language::getMessagesFor( $code ) )
4398 *
4399 * @since 1.19
4400 * @param string $code Language code
4401 * @return array Array of message keys (strings)
4402 */
4403 public static function getMessageKeysFor( $code ) {
4404 return self::getLocalisationCache()->getSubitemList( $code, 'messages' );
4405 }
4406
4407 /**
4408 * @param string $talk
4409 * @return mixed
4410 */
4411 function fixVariableInNamespace( $talk ) {
4412 if ( strpos( $talk, '$1' ) === false ) {
4413 return $talk;
4414 }
4415
4416 global $wgMetaNamespace;
4417 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4418
4419 # Allow grammar transformations
4420 # Allowing full message-style parsing would make simple requests
4421 # such as action=raw much more expensive than they need to be.
4422 # This will hopefully cover most cases.
4423 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4424 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4425 return str_replace( ' ', '_', $talk );
4426 }
4427
4428 /**
4429 * @param string $m
4430 * @return string
4431 */
4432 function replaceGrammarInNamespace( $m ) {
4433 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4434 }
4435
4436 /**
4437 * Decode an expiry (block, protection, etc) which has come from the DB
4438 *
4439 * @param string $expiry Database expiry String
4440 * @param bool|int $format True to process using language functions, or TS_ constant
4441 * to return the expiry in a given timestamp
4442 * @param string $infinity If $format is not true, use this string for infinite expiry
4443 * @return string
4444 * @since 1.18
4445 */
4446 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4447 static $dbInfinity;
4448 if ( $dbInfinity === null ) {
4449 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4450 }
4451
4452 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4453 return $format === true
4454 ? $this->getMessageFromDB( 'infiniteblock' )
4455 : $infinity;
4456 } else {
4457 return $format === true
4458 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4459 : wfTimestamp( $format, $expiry );
4460 }
4461 }
4462
4463 /**
4464 * Formats a time given in seconds into a string representation of that time.
4465 *
4466 * @param int|float $seconds
4467 * @param array $format An optional argument that formats the returned string in different ways:
4468 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4469 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4470 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4471 * and friends.
4472 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4473 * or 'avoidminutes'.
4474 * @return string
4475 */
4476 function formatTimePeriod( $seconds, $format = [] ) {
4477 if ( !is_array( $format ) ) {
4478 $format = [ 'avoid' => $format ]; // For backwards compatibility
4479 }
4480 if ( !isset( $format['avoid'] ) ) {
4481 $format['avoid'] = false;
4482 }
4483 if ( !isset( $format['noabbrevs'] ) ) {
4484 $format['noabbrevs'] = false;
4485 }
4486 $secondsMsg = wfMessage(
4487 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4488 $minutesMsg = wfMessage(
4489 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4490 $hoursMsg = wfMessage(
4491 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4492 $daysMsg = wfMessage(
4493 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4494
4495 if ( round( $seconds * 10 ) < 100 ) {
4496 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4497 $s = $secondsMsg->params( $s )->text();
4498 } elseif ( round( $seconds ) < 60 ) {
4499 $s = $this->formatNum( round( $seconds ) );
4500 $s = $secondsMsg->params( $s )->text();
4501 } elseif ( round( $seconds ) < 3600 ) {
4502 $minutes = floor( $seconds / 60 );
4503 $secondsPart = round( fmod( $seconds, 60 ) );
4504 if ( $secondsPart == 60 ) {
4505 $secondsPart = 0;
4506 $minutes++;
4507 }
4508 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4509 $s .= ' ';
4510 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4511 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4512 $hours = floor( $seconds / 3600 );
4513 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4514 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4515 if ( $secondsPart == 60 ) {
4516 $secondsPart = 0;
4517 $minutes++;
4518 }
4519 if ( $minutes == 60 ) {
4520 $minutes = 0;
4521 $hours++;
4522 }
4523 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4524 $s .= ' ';
4525 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4526 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4527 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4528 }
4529 } else {
4530 $days = floor( $seconds / 86400 );
4531 if ( $format['avoid'] === 'avoidminutes' ) {
4532 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4533 if ( $hours == 24 ) {
4534 $hours = 0;
4535 $days++;
4536 }
4537 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4538 $s .= ' ';
4539 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4540 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4541 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4542 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4543 if ( $minutes == 60 ) {
4544 $minutes = 0;
4545 $hours++;
4546 }
4547 if ( $hours == 24 ) {
4548 $hours = 0;
4549 $days++;
4550 }
4551 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4552 $s .= ' ';
4553 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4554 $s .= ' ';
4555 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4556 } else {
4557 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4558 $s .= ' ';
4559 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4560 }
4561 }
4562 return $s;
4563 }
4564
4565 /**
4566 * Format a bitrate for output, using an appropriate
4567 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4568 * the magnitude in question.
4569 *
4570 * This use base 1000. For base 1024 use formatSize(), for another base
4571 * see formatComputingNumbers().
4572 *
4573 * @param int $bps
4574 * @return string
4575 */
4576 function formatBitrate( $bps ) {
4577 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4578 }
4579
4580 /**
4581 * @param int $size Size of the unit
4582 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4583 * @param string $messageKey Message key to be uesd
4584 * @return string
4585 */
4586 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4587 if ( $size <= 0 ) {
4588 return str_replace( '$1', $this->formatNum( $size ),
4589 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4590 );
4591 }
4592 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4593 $index = 0;
4594
4595 $maxIndex = count( $sizes ) - 1;
4596 while ( $size >= $boundary && $index < $maxIndex ) {
4597 $index++;
4598 $size /= $boundary;
4599 }
4600
4601 // For small sizes no decimal places necessary
4602 $round = 0;
4603 if ( $index > 1 ) {
4604 // For MB and bigger two decimal places are smarter
4605 $round = 2;
4606 }
4607 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4608
4609 $size = round( $size, $round );
4610 $text = $this->getMessageFromDB( $msg );
4611 return str_replace( '$1', $this->formatNum( $size ), $text );
4612 }
4613
4614 /**
4615 * Format a size in bytes for output, using an appropriate
4616 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4617 *
4618 * This method use base 1024. For base 1000 use formatBitrate(), for
4619 * another base see formatComputingNumbers()
4620 *
4621 * @param int $size Size to format
4622 * @return string Plain text (not HTML)
4623 */
4624 function formatSize( $size ) {
4625 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4626 }
4627
4628 /**
4629 * Make a list item, used by various special pages
4630 *
4631 * @param string $page Page link
4632 * @param string $details HTML safe text between brackets
4633 * @param bool $oppositedm Add the direction mark opposite to your
4634 * language, to display text properly
4635 * @return HTML escaped string
4636 */
4637 function specialList( $page, $details, $oppositedm = true ) {
4638 if ( !$details ) {
4639 return $page;
4640 }
4641
4642 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4643 return
4644 $page .
4645 $dirmark .
4646 $this->msg( 'word-separator' )->escaped() .
4647 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4648 }
4649
4650 /**
4651 * Generate (prev x| next x) (20|50|100...) type links for paging
4652 *
4653 * @param Title $title Title object to link
4654 * @param int $offset
4655 * @param int $limit
4656 * @param array $query Optional URL query parameter string
4657 * @param bool $atend Optional param for specified if this is the last page
4658 * @return string
4659 */
4660 public function viewPrevNext( Title $title, $offset, $limit,
4661 array $query = [], $atend = false
4662 ) {
4663 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4664
4665 # Make 'previous' link
4666 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4667 if ( $offset > 0 ) {
4668 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4669 $query, $prev, 'prevn-title', 'mw-prevlink' );
4670 } else {
4671 $plink = htmlspecialchars( $prev );
4672 }
4673
4674 # Make 'next' link
4675 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4676 if ( $atend ) {
4677 $nlink = htmlspecialchars( $next );
4678 } else {
4679 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4680 $query, $next, 'nextn-title', 'mw-nextlink' );
4681 }
4682
4683 # Make links to set number of items per page
4684 $numLinks = [];
4685 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4686 $numLinks[] = $this->numLink( $title, $offset, $num,
4687 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4688 }
4689
4690 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4691 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4692 }
4693
4694 /**
4695 * Helper function for viewPrevNext() that generates links
4696 *
4697 * @param Title $title Title object to link
4698 * @param int $offset
4699 * @param int $limit
4700 * @param array $query Extra query parameters
4701 * @param string $link Text to use for the link; will be escaped
4702 * @param string $tooltipMsg Name of the message to use as tooltip
4703 * @param string $class Value of the "class" attribute of the link
4704 * @return string HTML fragment
4705 */
4706 private function numLink( Title $title, $offset, $limit, array $query, $link,
4707 $tooltipMsg, $class
4708 ) {
4709 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4710 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4711 ->numParams( $limit )->text();
4712
4713 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4714 'title' => $tooltip, 'class' => $class ], $link );
4715 }
4716
4717 /**
4718 * Get the conversion rule title, if any.
4719 *
4720 * @return string
4721 */
4722 public function getConvRuleTitle() {
4723 return $this->mConverter->getConvRuleTitle();
4724 }
4725
4726 /**
4727 * Get the compiled plural rules for the language
4728 * @since 1.20
4729 * @return array Associative array with plural form, and plural rule as key-value pairs
4730 */
4731 public function getCompiledPluralRules() {
4732 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4733 $fallbacks = Language::getFallbacksFor( $this->mCode );
4734 if ( !$pluralRules ) {
4735 foreach ( $fallbacks as $fallbackCode ) {
4736 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4737 if ( $pluralRules ) {
4738 break;
4739 }
4740 }
4741 }
4742 return $pluralRules;
4743 }
4744
4745 /**
4746 * Get the plural rules for the language
4747 * @since 1.20
4748 * @return array Associative array with plural form number and plural rule as key-value pairs
4749 */
4750 public function getPluralRules() {
4751 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4752 $fallbacks = Language::getFallbacksFor( $this->mCode );
4753 if ( !$pluralRules ) {
4754 foreach ( $fallbacks as $fallbackCode ) {
4755 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4756 if ( $pluralRules ) {
4757 break;
4758 }
4759 }
4760 }
4761 return $pluralRules;
4762 }
4763
4764 /**
4765 * Get the plural rule types for the language
4766 * @since 1.22
4767 * @return array Associative array with plural form number and plural rule type as key-value pairs
4768 */
4769 public function getPluralRuleTypes() {
4770 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4771 $fallbacks = Language::getFallbacksFor( $this->mCode );
4772 if ( !$pluralRuleTypes ) {
4773 foreach ( $fallbacks as $fallbackCode ) {
4774 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4775 if ( $pluralRuleTypes ) {
4776 break;
4777 }
4778 }
4779 }
4780 return $pluralRuleTypes;
4781 }
4782
4783 /**
4784 * Find the index number of the plural rule appropriate for the given number
4785 * @param int $number
4786 * @return int The index number of the plural rule
4787 */
4788 public function getPluralRuleIndexNumber( $number ) {
4789 $pluralRules = $this->getCompiledPluralRules();
4790 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4791 return $form;
4792 }
4793
4794 /**
4795 * Find the plural rule type appropriate for the given number
4796 * For example, if the language is set to Arabic, getPluralType(5) should
4797 * return 'few'.
4798 * @since 1.22
4799 * @param int $number
4800 * @return string The name of the plural rule type, e.g. one, two, few, many
4801 */
4802 public function getPluralRuleType( $number ) {
4803 $index = $this->getPluralRuleIndexNumber( $number );
4804 $pluralRuleTypes = $this->getPluralRuleTypes();
4805 if ( isset( $pluralRuleTypes[$index] ) ) {
4806 return $pluralRuleTypes[$index];
4807 } else {
4808 return 'other';
4809 }
4810 }
4811 }
MediaWiki Core