Merge "Fix and make some types in PHPDoc and JSDoc tags more specific"
[lhc/web/wiklou.git] / languages / Language.php
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 /** @var array|null */
49 protected $namespaceNames;
50 protected $mNamespaceIds, $namespaceAliases;
51
52 /**
53 * ReplacementArray object caches
54 */
55 public $transformData = [];
56
57 /**
58 * @var LocalisationCache
59 */
60 static public $dataCache;
61
62 static public $mLangObjCache = [];
63
64 static public $mWeekdayMsgs = [
65 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
66 'friday', 'saturday'
67 ];
68
69 static public $mWeekdayAbbrevMsgs = [
70 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
71 ];
72
73 static public $mMonthMsgs = [
74 'january', 'february', 'march', 'april', 'may_long', 'june',
75 'july', 'august', 'september', 'october', 'november',
76 'december'
77 ];
78 static public $mMonthGenMsgs = [
79 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
80 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
81 'december-gen'
82 ];
83 static public $mMonthAbbrevMsgs = [
84 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
85 'sep', 'oct', 'nov', 'dec'
86 ];
87
88 static public $mIranianCalendarMonthMsgs = [
89 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
90 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
91 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
92 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
93 ];
94
95 static public $mHebrewCalendarMonthMsgs = [
96 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
97 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
98 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
99 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
100 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
101 ];
102
103 static public $mHebrewCalendarMonthGenMsgs = [
104 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
105 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
106 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
107 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
108 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
109 ];
110
111 static public $mHijriCalendarMonthMsgs = [
112 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
113 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
114 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
115 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
116 ];
117
118 /**
119 * @since 1.20
120 * @var array
121 */
122 static public $durationIntervals = [
123 'millennia' => 31556952000,
124 'centuries' => 3155695200,
125 'decades' => 315569520,
126 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
127 'weeks' => 604800,
128 'days' => 86400,
129 'hours' => 3600,
130 'minutes' => 60,
131 'seconds' => 1,
132 ];
133
134 /**
135 * Cache for language fallbacks.
136 * @see Language::getFallbacksIncludingSiteLanguage
137 * @since 1.21
138 * @var array
139 */
140 static private $fallbackLanguageCache = [];
141
142 /**
143 * Cache for grammar rules data
144 * @var MapCacheLRU|null
145 */
146 static private $grammarTransformations;
147
148 /**
149 * Cache for language names
150 * @var HashBagOStuff|null
151 */
152 static private $languageNameCache;
153
154 /**
155 * Unicode directional formatting characters, for embedBidi()
156 */
157 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
158 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
159 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
160
161 /**
162 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
163 * - in group 1 if it is LTR
164 * - in group 2 if it is RTL
165 * Does not match if there is no strong directionality codepoint.
166 *
167 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
168 *
169 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
170 * https://git.wikimedia.org/summary/unicodejs.git .
171 */
172 // @codingStandardsIgnoreStart
173 // @codeCoverageIgnoreStart
174 static private $strongDirRegex = '/(?:([\x{41}-\x{5a}\x{61}-\x{7a}\x{aa}\x{b5}\x{ba}\x{c0}-\x{d6}\x{d8}-\x{f6}\x{f8}-\x{2b8}\x{2bb}-\x{2c1}\x{2d0}\x{2d1}\x{2e0}-\x{2e4}\x{2ee}\x{370}-\x{373}\x{376}\x{377}\x{37a}-\x{37d}\x{37f}\x{386}\x{388}-\x{38a}\x{38c}\x{38e}-\x{3a1}\x{3a3}-\x{3f5}\x{3f7}-\x{482}\x{48a}-\x{52f}\x{531}-\x{556}\x{559}-\x{55f}\x{561}-\x{587}\x{589}\x{903}-\x{939}\x{93b}\x{93d}-\x{940}\x{949}-\x{94c}\x{94e}-\x{950}\x{958}-\x{961}\x{964}-\x{980}\x{982}\x{983}\x{985}-\x{98c}\x{98f}\x{990}\x{993}-\x{9a8}\x{9aa}-\x{9b0}\x{9b2}\x{9b6}-\x{9b9}\x{9bd}-\x{9c0}\x{9c7}\x{9c8}\x{9cb}\x{9cc}\x{9ce}\x{9d7}\x{9dc}\x{9dd}\x{9df}-\x{9e1}\x{9e6}-\x{9f1}\x{9f4}-\x{9fa}\x{a03}\x{a05}-\x{a0a}\x{a0f}\x{a10}\x{a13}-\x{a28}\x{a2a}-\x{a30}\x{a32}\x{a33}\x{a35}\x{a36}\x{a38}\x{a39}\x{a3e}-\x{a40}\x{a59}-\x{a5c}\x{a5e}\x{a66}-\x{a6f}\x{a72}-\x{a74}\x{a83}\x{a85}-\x{a8d}\x{a8f}-\x{a91}\x{a93}-\x{aa8}\x{aaa}-\x{ab0}\x{ab2}\x{ab3}\x{ab5}-\x{ab9}\x{abd}-\x{ac0}\x{ac9}\x{acb}\x{acc}\x{ad0}\x{ae0}\x{ae1}\x{ae6}-\x{af0}\x{af9}\x{b02}\x{b03}\x{b05}-\x{b0c}\x{b0f}\x{b10}\x{b13}-\x{b28}\x{b2a}-\x{b30}\x{b32}\x{b33}\x{b35}-\x{b39}\x{b3d}\x{b3e}\x{b40}\x{b47}\x{b48}\x{b4b}\x{b4c}\x{b57}\x{b5c}\x{b5d}\x{b5f}-\x{b61}\x{b66}-\x{b77}\x{b83}\x{b85}-\x{b8a}\x{b8e}-\x{b90}\x{b92}-\x{b95}\x{b99}\x{b9a}\x{b9c}\x{b9e}\x{b9f}\x{ba3}\x{ba4}\x{ba8}-\x{baa}\x{bae}-\x{bb9}\x{bbe}\x{bbf}\x{bc1}\x{bc2}\x{bc6}-\x{bc8}\x{bca}-\x{bcc}\x{bd0}\x{bd7}\x{be6}-\x{bf2}\x{c01}-\x{c03}\x{c05}-\x{c0c}\x{c0e}-\x{c10}\x{c12}-\x{c28}\x{c2a}-\x{c39}\x{c3d}\x{c41}-\x{c44}\x{c58}-\x{c5a}\x{c60}\x{c61}\x{c66}-\x{c6f}\x{c7f}\x{c82}\x{c83}\x{c85}-\x{c8c}\x{c8e}-\x{c90}\x{c92}-\x{ca8}\x{caa}-\x{cb3}\x{cb5}-\x{cb9}\x{cbd}-\x{cc4}\x{cc6}-\x{cc8}\x{cca}\x{ccb}\x{cd5}\x{cd6}\x{cde}\x{ce0}\x{ce1}\x{ce6}-\x{cef}\x{cf1}\x{cf2}\x{d02}\x{d03}\x{d05}-\x{d0c}\x{d0e}-\x{d10}\x{d12}-\x{d3a}\x{d3d}-\x{d40}\x{d46}-\x{d48}\x{d4a}-\x{d4c}\x{d4e}\x{d57}\x{d5f}-\x{d61}\x{d66}-\x{d75}\x{d79}-\x{d7f}\x{d82}\x{d83}\x{d85}-\x{d96}\x{d9a}-\x{db1}\x{db3}-\x{dbb}\x{dbd}\x{dc0}-\x{dc6}\x{dcf}-\x{dd1}\x{dd8}-\x{ddf}\x{de6}-\x{def}\x{df2}-\x{df4}\x{e01}-\x{e30}\x{e32}\x{e33}\x{e40}-\x{e46}\x{e4f}-\x{e5b}\x{e81}\x{e82}\x{e84}\x{e87}\x{e88}\x{e8a}\x{e8d}\x{e94}-\x{e97}\x{e99}-\x{e9f}\x{ea1}-\x{ea3}\x{ea5}\x{ea7}\x{eaa}\x{eab}\x{ead}-\x{eb0}\x{eb2}\x{eb3}\x{ebd}\x{ec0}-\x{ec4}\x{ec6}\x{ed0}-\x{ed9}\x{edc}-\x{edf}\x{f00}-\x{f17}\x{f1a}-\x{f34}\x{f36}\x{f38}\x{f3e}-\x{f47}\x{f49}-\x{f6c}\x{f7f}\x{f85}\x{f88}-\x{f8c}\x{fbe}-\x{fc5}\x{fc7}-\x{fcc}\x{fce}-\x{fda}\x{1000}-\x{102c}\x{1031}\x{1038}\x{103b}\x{103c}\x{103f}-\x{1057}\x{105a}-\x{105d}\x{1061}-\x{1070}\x{1075}-\x{1081}\x{1083}\x{1084}\x{1087}-\x{108c}\x{108e}-\x{109c}\x{109e}-\x{10c5}\x{10c7}\x{10cd}\x{10d0}-\x{1248}\x{124a}-\x{124d}\x{1250}-\x{1256}\x{1258}\x{125a}-\x{125d}\x{1260}-\x{1288}\x{128a}-\x{128d}\x{1290}-\x{12b0}\x{12b2}-\x{12b5}\x{12b8}-\x{12be}\x{12c0}\x{12c2}-\x{12c5}\x{12c8}-\x{12d6}\x{12d8}-\x{1310}\x{1312}-\x{1315}\x{1318}-\x{135a}\x{1360}-\x{137c}\x{1380}-\x{138f}\x{13a0}-\x{13f5}\x{13f8}-\x{13fd}\x{1401}-\x{167f}\x{1681}-\x{169a}\x{16a0}-\x{16f8}\x{1700}-\x{170c}\x{170e}-\x{1711}\x{1720}-\x{1731}\x{1735}\x{1736}\x{1740}-\x{1751}\x{1760}-\x{176c}\x{176e}-\x{1770}\x{1780}-\x{17b3}\x{17b6}\x{17be}-\x{17c5}\x{17c7}\x{17c8}\x{17d4}-\x{17da}\x{17dc}\x{17e0}-\x{17e9}\x{1810}-\x{1819}\x{1820}-\x{1877}\x{1880}-\x{18a8}\x{18aa}\x{18b0}-\x{18f5}\x{1900}-\x{191e}\x{1923}-\x{1926}\x{1929}-\x{192b}\x{1930}\x{1931}\x{1933}-\x{1938}\x{1946}-\x{196d}\x{1970}-\x{1974}\x{1980}-\x{19ab}\x{19b0}-\x{19c9}\x{19d0}-\x{19da}\x{1a00}-\x{1a16}\x{1a19}\x{1a1a}\x{1a1e}-\x{1a55}\x{1a57}\x{1a61}\x{1a63}\x{1a64}\x{1a6d}-\x{1a72}\x{1a80}-\x{1a89}\x{1a90}-\x{1a99}\x{1aa0}-\x{1aad}\x{1b04}-\x{1b33}\x{1b35}\x{1b3b}\x{1b3d}-\x{1b41}\x{1b43}-\x{1b4b}\x{1b50}-\x{1b6a}\x{1b74}-\x{1b7c}\x{1b82}-\x{1ba1}\x{1ba6}\x{1ba7}\x{1baa}\x{1bae}-\x{1be5}\x{1be7}\x{1bea}-\x{1bec}\x{1bee}\x{1bf2}\x{1bf3}\x{1bfc}-\x{1c2b}\x{1c34}\x{1c35}\x{1c3b}-\x{1c49}\x{1c4d}-\x{1c7f}\x{1cc0}-\x{1cc7}\x{1cd3}\x{1ce1}\x{1ce9}-\x{1cec}\x{1cee}-\x{1cf3}\x{1cf5}\x{1cf6}\x{1d00}-\x{1dbf}\x{1e00}-\x{1f15}\x{1f18}-\x{1f1d}\x{1f20}-\x{1f45}\x{1f48}-\x{1f4d}\x{1f50}-\x{1f57}\x{1f59}\x{1f5b}\x{1f5d}\x{1f5f}-\x{1f7d}\x{1f80}-\x{1fb4}\x{1fb6}-\x{1fbc}\x{1fbe}\x{1fc2}-\x{1fc4}\x{1fc6}-\x{1fcc}\x{1fd0}-\x{1fd3}\x{1fd6}-\x{1fdb}\x{1fe0}-\x{1fec}\x{1ff2}-\x{1ff4}\x{1ff6}-\x{1ffc}\x{200e}\x{2071}\x{207f}\x{2090}-\x{209c}\x{2102}\x{2107}\x{210a}-\x{2113}\x{2115}\x{2119}-\x{211d}\x{2124}\x{2126}\x{2128}\x{212a}-\x{212d}\x{212f}-\x{2139}\x{213c}-\x{213f}\x{2145}-\x{2149}\x{214e}\x{214f}\x{2160}-\x{2188}\x{2336}-\x{237a}\x{2395}\x{249c}-\x{24e9}\x{26ac}\x{2800}-\x{28ff}\x{2c00}-\x{2c2e}\x{2c30}-\x{2c5e}\x{2c60}-\x{2ce4}\x{2ceb}-\x{2cee}\x{2cf2}\x{2cf3}\x{2d00}-\x{2d25}\x{2d27}\x{2d2d}\x{2d30}-\x{2d67}\x{2d6f}\x{2d70}\x{2d80}-\x{2d96}\x{2da0}-\x{2da6}\x{2da8}-\x{2dae}\x{2db0}-\x{2db6}\x{2db8}-\x{2dbe}\x{2dc0}-\x{2dc6}\x{2dc8}-\x{2dce}\x{2dd0}-\x{2dd6}\x{2dd8}-\x{2dde}\x{3005}-\x{3007}\x{3021}-\x{3029}\x{302e}\x{302f}\x{3031}-\x{3035}\x{3038}-\x{303c}\x{3041}-\x{3096}\x{309d}-\x{309f}\x{30a1}-\x{30fa}\x{30fc}-\x{30ff}\x{3105}-\x{312d}\x{3131}-\x{318e}\x{3190}-\x{31ba}\x{31f0}-\x{321c}\x{3220}-\x{324f}\x{3260}-\x{327b}\x{327f}-\x{32b0}\x{32c0}-\x{32cb}\x{32d0}-\x{32fe}\x{3300}-\x{3376}\x{337b}-\x{33dd}\x{33e0}-\x{33fe}\x{3400}-\x{4db5}\x{4e00}-\x{9fd5}\x{a000}-\x{a48c}\x{a4d0}-\x{a60c}\x{a610}-\x{a62b}\x{a640}-\x{a66e}\x{a680}-\x{a69d}\x{a6a0}-\x{a6ef}\x{a6f2}-\x{a6f7}\x{a722}-\x{a787}\x{a789}-\x{a7ad}\x{a7b0}-\x{a7b7}\x{a7f7}-\x{a801}\x{a803}-\x{a805}\x{a807}-\x{a80a}\x{a80c}-\x{a824}\x{a827}\x{a830}-\x{a837}\x{a840}-\x{a873}\x{a880}-\x{a8c3}\x{a8ce}-\x{a8d9}\x{a8f2}-\x{a8fd}\x{a900}-\x{a925}\x{a92e}-\x{a946}\x{a952}\x{a953}\x{a95f}-\x{a97c}\x{a983}-\x{a9b2}\x{a9b4}\x{a9b5}\x{a9ba}\x{a9bb}\x{a9bd}-\x{a9cd}\x{a9cf}-\x{a9d9}\x{a9de}-\x{a9e4}\x{a9e6}-\x{a9fe}\x{aa00}-\x{aa28}\x{aa2f}\x{aa30}\x{aa33}\x{aa34}\x{aa40}-\x{aa42}\x{aa44}-\x{aa4b}\x{aa4d}\x{aa50}-\x{aa59}\x{aa5c}-\x{aa7b}\x{aa7d}-\x{aaaf}\x{aab1}\x{aab5}\x{aab6}\x{aab9}-\x{aabd}\x{aac0}\x{aac2}\x{aadb}-\x{aaeb}\x{aaee}-\x{aaf5}\x{ab01}-\x{ab06}\x{ab09}-\x{ab0e}\x{ab11}-\x{ab16}\x{ab20}-\x{ab26}\x{ab28}-\x{ab2e}\x{ab30}-\x{ab65}\x{ab70}-\x{abe4}\x{abe6}\x{abe7}\x{abe9}-\x{abec}\x{abf0}-\x{abf9}\x{ac00}-\x{d7a3}\x{d7b0}-\x{d7c6}\x{d7cb}-\x{d7fb}\x{e000}-\x{fa6d}\x{fa70}-\x{fad9}\x{fb00}-\x{fb06}\x{fb13}-\x{fb17}\x{ff21}-\x{ff3a}\x{ff41}-\x{ff5a}\x{ff66}-\x{ffbe}\x{ffc2}-\x{ffc7}\x{ffca}-\x{ffcf}\x{ffd2}-\x{ffd7}\x{ffda}-\x{ffdc}\x{10000}-\x{1000b}\x{1000d}-\x{10026}\x{10028}-\x{1003a}\x{1003c}\x{1003d}\x{1003f}-\x{1004d}\x{10050}-\x{1005d}\x{10080}-\x{100fa}\x{10100}\x{10102}\x{10107}-\x{10133}\x{10137}-\x{1013f}\x{101d0}-\x{101fc}\x{10280}-\x{1029c}\x{102a0}-\x{102d0}\x{10300}-\x{10323}\x{10330}-\x{1034a}\x{10350}-\x{10375}\x{10380}-\x{1039d}\x{1039f}-\x{103c3}\x{103c8}-\x{103d5}\x{10400}-\x{1049d}\x{104a0}-\x{104a9}\x{10500}-\x{10527}\x{10530}-\x{10563}\x{1056f}\x{10600}-\x{10736}\x{10740}-\x{10755}\x{10760}-\x{10767}\x{11000}\x{11002}-\x{11037}\x{11047}-\x{1104d}\x{11066}-\x{1106f}\x{11082}-\x{110b2}\x{110b7}\x{110b8}\x{110bb}-\x{110c1}\x{110d0}-\x{110e8}\x{110f0}-\x{110f9}\x{11103}-\x{11126}\x{1112c}\x{11136}-\x{11143}\x{11150}-\x{11172}\x{11174}-\x{11176}\x{11182}-\x{111b5}\x{111bf}-\x{111c9}\x{111cd}\x{111d0}-\x{111df}\x{111e1}-\x{111f4}\x{11200}-\x{11211}\x{11213}-\x{1122e}\x{11232}\x{11233}\x{11235}\x{11238}-\x{1123d}\x{11280}-\x{11286}\x{11288}\x{1128a}-\x{1128d}\x{1128f}-\x{1129d}\x{1129f}-\x{112a9}\x{112b0}-\x{112de}\x{112e0}-\x{112e2}\x{112f0}-\x{112f9}\x{11302}\x{11303}\x{11305}-\x{1130c}\x{1130f}\x{11310}\x{11313}-\x{11328}\x{1132a}-\x{11330}\x{11332}\x{11333}\x{11335}-\x{11339}\x{1133d}-\x{1133f}\x{11341}-\x{11344}\x{11347}\x{11348}\x{1134b}-\x{1134d}\x{11350}\x{11357}\x{1135d}-\x{11363}\x{11480}-\x{114b2}\x{114b9}\x{114bb}-\x{114be}\x{114c1}\x{114c4}-\x{114c7}\x{114d0}-\x{114d9}\x{11580}-\x{115b1}\x{115b8}-\x{115bb}\x{115be}\x{115c1}-\x{115db}\x{11600}-\x{11632}\x{1163b}\x{1163c}\x{1163e}\x{11641}-\x{11644}\x{11650}-\x{11659}\x{11680}-\x{116aa}\x{116ac}\x{116ae}\x{116af}\x{116b6}\x{116c0}-\x{116c9}\x{11700}-\x{11719}\x{11720}\x{11721}\x{11726}\x{11730}-\x{1173f}\x{118a0}-\x{118f2}\x{118ff}\x{11ac0}-\x{11af8}\x{12000}-\x{12399}\x{12400}-\x{1246e}\x{12470}-\x{12474}\x{12480}-\x{12543}\x{13000}-\x{1342e}\x{14400}-\x{14646}\x{16800}-\x{16a38}\x{16a40}-\x{16a5e}\x{16a60}-\x{16a69}\x{16a6e}\x{16a6f}\x{16ad0}-\x{16aed}\x{16af5}\x{16b00}-\x{16b2f}\x{16b37}-\x{16b45}\x{16b50}-\x{16b59}\x{16b5b}-\x{16b61}\x{16b63}-\x{16b77}\x{16b7d}-\x{16b8f}\x{16f00}-\x{16f44}\x{16f50}-\x{16f7e}\x{16f93}-\x{16f9f}\x{1b000}\x{1b001}\x{1bc00}-\x{1bc6a}\x{1bc70}-\x{1bc7c}\x{1bc80}-\x{1bc88}\x{1bc90}-\x{1bc99}\x{1bc9c}\x{1bc9f}\x{1d000}-\x{1d0f5}\x{1d100}-\x{1d126}\x{1d129}-\x{1d166}\x{1d16a}-\x{1d172}\x{1d183}\x{1d184}\x{1d18c}-\x{1d1a9}\x{1d1ae}-\x{1d1e8}\x{1d360}-\x{1d371}\x{1d400}-\x{1d454}\x{1d456}-\x{1d49c}\x{1d49e}\x{1d49f}\x{1d4a2}\x{1d4a5}\x{1d4a6}\x{1d4a9}-\x{1d4ac}\x{1d4ae}-\x{1d4b9}\x{1d4bb}\x{1d4bd}-\x{1d4c3}\x{1d4c5}-\x{1d505}\x{1d507}-\x{1d50a}\x{1d50d}-\x{1d514}\x{1d516}-\x{1d51c}\x{1d51e}-\x{1d539}\x{1d53b}-\x{1d53e}\x{1d540}-\x{1d544}\x{1d546}\x{1d54a}-\x{1d550}\x{1d552}-\x{1d6a5}\x{1d6a8}-\x{1d6da}\x{1d6dc}-\x{1d714}\x{1d716}-\x{1d74e}\x{1d750}-\x{1d788}\x{1d78a}-\x{1d7c2}\x{1d7c4}-\x{1d7cb}\x{1d800}-\x{1d9ff}\x{1da37}-\x{1da3a}\x{1da6d}-\x{1da74}\x{1da76}-\x{1da83}\x{1da85}-\x{1da8b}\x{1f110}-\x{1f12e}\x{1f130}-\x{1f169}\x{1f170}-\x{1f19a}\x{1f1e6}-\x{1f202}\x{1f210}-\x{1f23a}\x{1f240}-\x{1f248}\x{1f250}\x{1f251}\x{20000}-\x{2a6d6}\x{2a700}-\x{2b734}\x{2b740}-\x{2b81d}\x{2b820}-\x{2cea1}\x{2f800}-\x{2fa1d}\x{f0000}-\x{ffffd}\x{100000}-\x{10fffd}])|([\x{590}\x{5be}\x{5c0}\x{5c3}\x{5c6}\x{5c8}-\x{5ff}\x{7c0}-\x{7ea}\x{7f4}\x{7f5}\x{7fa}-\x{815}\x{81a}\x{824}\x{828}\x{82e}-\x{858}\x{85c}-\x{89f}\x{200f}\x{fb1d}\x{fb1f}-\x{fb28}\x{fb2a}-\x{fb4f}\x{10800}-\x{1091e}\x{10920}-\x{10a00}\x{10a04}\x{10a07}-\x{10a0b}\x{10a10}-\x{10a37}\x{10a3b}-\x{10a3e}\x{10a40}-\x{10ae4}\x{10ae7}-\x{10b38}\x{10b40}-\x{10e5f}\x{10e7f}-\x{10fff}\x{1e800}-\x{1e8cf}\x{1e8d7}-\x{1edff}\x{1ef00}-\x{1efff}\x{608}\x{60b}\x{60d}\x{61b}-\x{64a}\x{66d}-\x{66f}\x{671}-\x{6d5}\x{6e5}\x{6e6}\x{6ee}\x{6ef}\x{6fa}-\x{710}\x{712}-\x{72f}\x{74b}-\x{7a5}\x{7b1}-\x{7bf}\x{8a0}-\x{8e2}\x{fb50}-\x{fd3d}\x{fd40}-\x{fdcf}\x{fdf0}-\x{fdfc}\x{fdfe}\x{fdff}\x{fe70}-\x{fefe}\x{1ee00}-\x{1eeef}\x{1eef2}-\x{1eeff}]))/u';
175 // @codeCoverageIgnoreEnd
176 // @codingStandardsIgnoreEnd
177
178 /**
179 * Get a cached or new language object for a given language code
180 * @param string $code
181 * @return Language
182 */
183 static function factory( $code ) {
184 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
185
186 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
187 $code = $wgDummyLanguageCodes[$code];
188 }
189
190 // get the language object to process
191 $langObj = isset( self::$mLangObjCache[$code] )
192 ? self::$mLangObjCache[$code]
193 : self::newFromCode( $code );
194
195 // merge the language object in to get it up front in the cache
196 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
197 // get rid of the oldest ones in case we have an overflow
198 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
199
200 return $langObj;
201 }
202
203 /**
204 * Create a language object for a given language code
205 * @param string $code
206 * @param $fallback boolean Whether we're going through language fallback chain
207 * @throws MWException
208 * @return Language
209 */
210 protected static function newFromCode( $code, $fallback = false ) {
211 if ( !self::isValidCode( $code ) ) {
212 throw new MWException( "Invalid language code \"$code\"" );
213 }
214
215 if ( !self::isValidBuiltInCode( $code ) ) {
216 // It's not possible to customise this code with class files, so
217 // just return a Language object. This is to support uselang= hacks.
218 $lang = new Language;
219 $lang->setCode( $code );
220 return $lang;
221 }
222
223 // Check if there is a language class for the code
224 $class = self::classFromCode( $code, $fallback );
225 if ( class_exists( $class ) ) {
226 $lang = new $class;
227 return $lang;
228 }
229
230 // Keep trying the fallback list until we find an existing class
231 $fallbacks = self::getFallbacksFor( $code );
232 foreach ( $fallbacks as $fallbackCode ) {
233 if ( !self::isValidBuiltInCode( $fallbackCode ) ) {
234 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
235 }
236
237 $class = self::classFromCode( $fallbackCode );
238 if ( class_exists( $class ) ) {
239 $lang = new $class;
240 $lang->setCode( $code );
241 return $lang;
242 }
243 }
244
245 throw new MWException( "Invalid fallback sequence for language '$code'" );
246 }
247
248 /**
249 * Checks whether any localisation is available for that language tag
250 * in MediaWiki (MessagesXx.php exists).
251 *
252 * @param string $code Language tag (in lower case)
253 * @return bool Whether language is supported
254 * @since 1.21
255 */
256 public static function isSupportedLanguage( $code ) {
257 if ( !self::isValidBuiltInCode( $code ) ) {
258 return false;
259 }
260
261 if ( $code === 'qqq' ) {
262 return false;
263 }
264
265 return is_readable( self::getMessagesFileName( $code ) ) ||
266 is_readable( self::getJsonMessagesFileName( $code ) );
267 }
268
269 /**
270 * Returns true if a language code string is a well-formed language tag
271 * according to RFC 5646.
272 * This function only checks well-formedness; it doesn't check that
273 * language, script or variant codes actually exist in the repositories.
274 *
275 * Based on regexes by Mark Davis of the Unicode Consortium:
276 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
277 *
278 * @param string $code
279 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
280 *
281 * @return bool
282 * @since 1.21
283 */
284 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
285 $alpha = '[a-z]';
286 $digit = '[0-9]';
287 $alphanum = '[a-z0-9]';
288 $x = 'x'; # private use singleton
289 $singleton = '[a-wy-z]'; # other singleton
290 $s = $lenient ? '[-_]' : '-';
291
292 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
293 $script = "$alpha{4}"; # ISO 15924
294 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
295 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
296 $extension = "$singleton(?:$s$alphanum{2,8})+";
297 $privateUse = "$x(?:$s$alphanum{1,8})+";
298
299 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
300 # Since these are limited, this is safe even later changes to the registry --
301 # the only oddity is that it might change the type of the tag, and thus
302 # the results from the capturing groups.
303 # https://www.iana.org/assignments/language-subtag-registry
304
305 $grandfathered = "en{$s}GB{$s}oed"
306 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
307 . "|no{$s}(?:bok|nyn)"
308 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
309 . "|zh{$s}min{$s}nan";
310
311 $variantList = "$variant(?:$s$variant)*";
312 $extensionList = "$extension(?:$s$extension)*";
313
314 $langtag = "(?:($language)"
315 . "(?:$s$script)?"
316 . "(?:$s$region)?"
317 . "(?:$s$variantList)?"
318 . "(?:$s$extensionList)?"
319 . "(?:$s$privateUse)?)";
320
321 # The final breakdown, with capturing groups for each of these components
322 # The variants, extensions, grandfathered, and private-use may have interior '-'
323
324 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
325
326 return (bool)preg_match( "/$root/", strtolower( $code ) );
327 }
328
329 /**
330 * Returns true if a language code string is of a valid form, whether or
331 * not it exists. This includes codes which are used solely for
332 * customisation via the MediaWiki namespace.
333 *
334 * @param string $code
335 *
336 * @return bool
337 */
338 public static function isValidCode( $code ) {
339 static $cache = [];
340 if ( !isset( $cache[$code] ) ) {
341 // People think language codes are html safe, so enforce it.
342 // Ideally we should only allow a-zA-Z0-9-
343 // but, .+ and other chars are often used for {{int:}} hacks
344 // see bugs T39564, T39587, T38938
345 $cache[$code] =
346 // Protect against path traversal
347 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
348 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
349 }
350 return $cache[$code];
351 }
352
353 /**
354 * Returns true if a language code is of a valid form for the purposes of
355 * internal customisation of MediaWiki, via Messages*.php or *.json.
356 *
357 * @param string $code
358 *
359 * @throws MWException
360 * @since 1.18
361 * @return bool
362 */
363 public static function isValidBuiltInCode( $code ) {
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
367 } else {
368 $addmsg = '';
369 }
370 $type = gettype( $code );
371 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
372 }
373
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
375 }
376
377 /**
378 * Returns true if a language code is an IETF tag known to MediaWiki.
379 *
380 * @param string $tag
381 *
382 * @since 1.21
383 * @return bool
384 */
385 public static function isKnownLanguageTag( $tag ) {
386 // Quick escape for invalid input to avoid exceptions down the line
387 // when code tries to process tags which are not valid at all.
388 if ( !self::isValidBuiltInCode( $tag ) ) {
389 return false;
390 }
391
392 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
393 || self::fetchLanguageName( $tag, $tag ) !== ''
394 ) {
395 return true;
396 }
397
398 return false;
399 }
400
401 /**
402 * Get the LocalisationCache instance
403 *
404 * @return LocalisationCache
405 */
406 public static function getLocalisationCache() {
407 if ( is_null( self::$dataCache ) ) {
408 global $wgLocalisationCacheConf;
409 $class = $wgLocalisationCacheConf['class'];
410 self::$dataCache = new $class( $wgLocalisationCacheConf );
411 }
412 return self::$dataCache;
413 }
414
415 function __construct() {
416 $this->mConverter = new FakeConverter( $this );
417 // Set the code to the name of the descendant
418 if ( static::class === 'Language' ) {
419 $this->mCode = 'en';
420 } else {
421 $this->mCode = str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
422 }
423 self::getLocalisationCache();
424 }
425
426 /**
427 * Reduce memory usage
428 */
429 function __destruct() {
430 foreach ( $this as $name => $value ) {
431 unset( $this->$name );
432 }
433 }
434
435 /**
436 * Hook which will be called if this is the content language.
437 * Descendants can use this to register hook functions or modify globals
438 */
439 function initContLang() {
440 }
441
442 /**
443 * @return array
444 * @since 1.19
445 */
446 public function getFallbackLanguages() {
447 return self::getFallbacksFor( $this->mCode );
448 }
449
450 /**
451 * Exports $wgBookstoreListEn
452 * @return array
453 */
454 public function getBookstoreList() {
455 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
456 }
457
458 /**
459 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
460 * available in localised form, it will be included in English.
461 *
462 * @return array
463 */
464 public function getNamespaces() {
465 if ( is_null( $this->namespaceNames ) ) {
466 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
467
468 $validNamespaces = MWNamespace::getCanonicalNamespaces();
469
470 $this->namespaceNames = $wgExtraNamespaces +
471 self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
472 $this->namespaceNames += $validNamespaces;
473
474 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
475 if ( $wgMetaNamespaceTalk ) {
476 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
477 } else {
478 $talk = $this->namespaceNames[NS_PROJECT_TALK];
479 $this->namespaceNames[NS_PROJECT_TALK] =
480 $this->fixVariableInNamespace( $talk );
481 }
482
483 # Sometimes a language will be localised but not actually exist on this wiki.
484 foreach ( $this->namespaceNames as $key => $text ) {
485 if ( !isset( $validNamespaces[$key] ) ) {
486 unset( $this->namespaceNames[$key] );
487 }
488 }
489
490 # The above mixing may leave namespaces out of canonical order.
491 # Re-order by namespace ID number...
492 ksort( $this->namespaceNames );
493
494 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
495 }
496
497 return $this->namespaceNames;
498 }
499
500 /**
501 * Arbitrarily set all of the namespace names at once. Mainly used for testing
502 * @param array $namespaces Array of namespaces (id => name)
503 */
504 public function setNamespaces( array $namespaces ) {
505 $this->namespaceNames = $namespaces;
506 $this->mNamespaceIds = null;
507 }
508
509 /**
510 * Resets all of the namespace caches. Mainly used for testing
511 */
512 public function resetNamespaces() {
513 $this->namespaceNames = null;
514 $this->mNamespaceIds = null;
515 $this->namespaceAliases = null;
516 }
517
518 /**
519 * A convenience function that returns getNamespaces() with spaces instead of underscores
520 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
521 *
522 * @return array
523 */
524 public function getFormattedNamespaces() {
525 $ns = $this->getNamespaces();
526 foreach ( $ns as $k => $v ) {
527 $ns[$k] = strtr( $v, '_', ' ' );
528 }
529 return $ns;
530 }
531
532 /**
533 * Get a namespace value by key
534 *
535 * <code>
536 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
537 * echo $mw_ns; // prints 'MediaWiki'
538 * </code>
539 *
540 * @param int $index The array key of the namespace to return
541 * @return string|bool String if the namespace value exists, otherwise false
542 */
543 public function getNsText( $index ) {
544 $ns = $this->getNamespaces();
545 return isset( $ns[$index] ) ? $ns[$index] : false;
546 }
547
548 /**
549 * A convenience function that returns the same thing as
550 * getNsText() except with '_' changed to ' ', useful for
551 * producing output.
552 *
553 * <code>
554 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
555 * echo $mw_ns; // prints 'MediaWiki talk'
556 * </code>
557 *
558 * @param int $index The array key of the namespace to return
559 * @return string Namespace name without underscores (empty string if namespace does not exist)
560 */
561 public function getFormattedNsText( $index ) {
562 $ns = $this->getNsText( $index );
563 return strtr( $ns, '_', ' ' );
564 }
565
566 /**
567 * Returns gender-dependent namespace alias if available.
568 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
569 * @param int $index Namespace index
570 * @param string $gender Gender key (male, female... )
571 * @return string
572 * @since 1.18
573 */
574 public function getGenderNsText( $index, $gender ) {
575 global $wgExtraGenderNamespaces;
576
577 $ns = $wgExtraGenderNamespaces +
578 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
579
580 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
581 }
582
583 /**
584 * Whether this language uses gender-dependent namespace aliases.
585 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
586 * @return bool
587 * @since 1.18
588 */
589 public function needsGenderDistinction() {
590 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
591 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
592 // $wgExtraGenderNamespaces overrides everything
593 return true;
594 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
595 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
596 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
597 return false;
598 } else {
599 // Check what is in i18n files
600 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
601 return count( $aliases ) > 0;
602 }
603 }
604
605 /**
606 * Get a namespace key by value, case insensitive.
607 * Only matches namespace names for the current language, not the
608 * canonical ones defined in Namespace.php.
609 *
610 * @param string $text
611 * @return int|bool An integer if $text is a valid value otherwise false
612 */
613 function getLocalNsIndex( $text ) {
614 $lctext = $this->lc( $text );
615 $ids = $this->getNamespaceIds();
616 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
617 }
618
619 /**
620 * @return array
621 */
622 public function getNamespaceAliases() {
623 if ( is_null( $this->namespaceAliases ) ) {
624 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
625 if ( !$aliases ) {
626 $aliases = [];
627 } else {
628 foreach ( $aliases as $name => $index ) {
629 if ( $index === NS_PROJECT_TALK ) {
630 unset( $aliases[$name] );
631 $name = $this->fixVariableInNamespace( $name );
632 $aliases[$name] = $index;
633 }
634 }
635 }
636
637 global $wgExtraGenderNamespaces;
638 $genders = $wgExtraGenderNamespaces +
639 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
640 foreach ( $genders as $index => $forms ) {
641 foreach ( $forms as $alias ) {
642 $aliases[$alias] = $index;
643 }
644 }
645
646 # Also add converted namespace names as aliases, to avoid confusion.
647 $convertedNames = [];
648 foreach ( $this->getVariants() as $variant ) {
649 if ( $variant === $this->mCode ) {
650 continue;
651 }
652 foreach ( $this->getNamespaces() as $ns => $_ ) {
653 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
654 }
655 }
656
657 $this->namespaceAliases = $aliases + $convertedNames;
658 }
659
660 return $this->namespaceAliases;
661 }
662
663 /**
664 * @return array
665 */
666 public function getNamespaceIds() {
667 if ( is_null( $this->mNamespaceIds ) ) {
668 global $wgNamespaceAliases;
669 # Put namespace names and aliases into a hashtable.
670 # If this is too slow, then we should arrange it so that it is done
671 # before caching. The catch is that at pre-cache time, the above
672 # class-specific fixup hasn't been done.
673 $this->mNamespaceIds = [];
674 foreach ( $this->getNamespaces() as $index => $name ) {
675 $this->mNamespaceIds[$this->lc( $name )] = $index;
676 }
677 foreach ( $this->getNamespaceAliases() as $name => $index ) {
678 $this->mNamespaceIds[$this->lc( $name )] = $index;
679 }
680 if ( $wgNamespaceAliases ) {
681 foreach ( $wgNamespaceAliases as $name => $index ) {
682 $this->mNamespaceIds[$this->lc( $name )] = $index;
683 }
684 }
685 }
686 return $this->mNamespaceIds;
687 }
688
689 /**
690 * Get a namespace key by value, case insensitive. Canonical namespace
691 * names override custom ones defined for the current language.
692 *
693 * @param string $text
694 * @return int|bool An integer if $text is a valid value otherwise false
695 */
696 public function getNsIndex( $text ) {
697 $lctext = $this->lc( $text );
698 $ns = MWNamespace::getCanonicalIndex( $lctext );
699 if ( $ns !== null ) {
700 return $ns;
701 }
702 $ids = $this->getNamespaceIds();
703 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
704 }
705
706 /**
707 * short names for language variants used for language conversion links.
708 *
709 * @param string $code
710 * @param bool $usemsg Use the "variantname-xyz" message if it exists
711 * @return string
712 */
713 public function getVariantname( $code, $usemsg = true ) {
714 $msg = "variantname-$code";
715 if ( $usemsg && wfMessage( $msg )->exists() ) {
716 return $this->getMessageFromDB( $msg );
717 }
718 $name = self::fetchLanguageName( $code );
719 if ( $name ) {
720 return $name; # if it's defined as a language name, show that
721 } else {
722 # otherwise, output the language code
723 return $code;
724 }
725 }
726
727 /**
728 * @return string[]|bool List of date format preference keys, or false if disabled.
729 */
730 public function getDatePreferences() {
731 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
732 }
733
734 /**
735 * @return array
736 */
737 function getDateFormats() {
738 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
739 }
740
741 /**
742 * @return array|string
743 */
744 public function getDefaultDateFormat() {
745 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
746 if ( $df === 'dmy or mdy' ) {
747 global $wgAmericanDates;
748 return $wgAmericanDates ? 'mdy' : 'dmy';
749 } else {
750 return $df;
751 }
752 }
753
754 /**
755 * @return array
756 */
757 public function getDatePreferenceMigrationMap() {
758 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
759 }
760
761 /**
762 * @param string $image
763 * @return array|null
764 */
765 function getImageFile( $image ) {
766 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
767 }
768
769 /**
770 * @return array
771 * @since 1.24
772 */
773 public function getImageFiles() {
774 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
775 }
776
777 /**
778 * @return array
779 */
780 public function getExtraUserToggles() {
781 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
782 }
783
784 /**
785 * @param string $tog
786 * @return string
787 */
788 function getUserToggle( $tog ) {
789 return $this->getMessageFromDB( "tog-$tog" );
790 }
791
792 /**
793 * Get an array of language names, indexed by code.
794 * @param null|string $inLanguage Code of language in which to return the names
795 * Use null for autonyms (native names)
796 * @param string $include One of:
797 * 'all' all available languages
798 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
799 * 'mwfile' only if the language is in 'mw' *and* has a message file
800 * @return array Language code => language name
801 * @since 1.20
802 */
803 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
804 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
805 $cacheKey .= ":$include";
806 if ( self::$languageNameCache === null ) {
807 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
808 }
809
810 $ret = self::$languageNameCache->get( $cacheKey );
811 if ( !$ret ) {
812 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
813 self::$languageNameCache->set( $cacheKey, $ret );
814 }
815 return $ret;
816 }
817
818 /**
819 * Uncached helper for fetchLanguageNames
820 * @param null|string $inLanguage Code of language in which to return the names
821 * Use null for autonyms (native names)
822 * @param string $include One of:
823 * 'all' all available languages
824 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
825 * 'mwfile' only if the language is in 'mw' *and* has a message file
826 * @return array Language code => language name
827 */
828 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
829 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
830
831 // If passed an invalid language code to use, fallback to en
832 if ( $inLanguage !== null && !self::isValidCode( $inLanguage ) ) {
833 $inLanguage = 'en';
834 }
835
836 $names = [];
837
838 if ( $inLanguage ) {
839 # TODO: also include when $inLanguage is null, when this code is more efficient
840 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
841 }
842
843 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
844 if ( $wgUsePigLatinVariant ) {
845 // Pig Latin (for variant development)
846 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
847 }
848
849 foreach ( $mwNames as $mwCode => $mwName ) {
850 # - Prefer own MediaWiki native name when not using the hook
851 # - For other names just add if not added through the hook
852 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
853 $names[$mwCode] = $mwName;
854 }
855 }
856
857 if ( $include === 'all' ) {
858 ksort( $names );
859 return $names;
860 }
861
862 $returnMw = [];
863 $coreCodes = array_keys( $mwNames );
864 foreach ( $coreCodes as $coreCode ) {
865 $returnMw[$coreCode] = $names[$coreCode];
866 }
867
868 if ( $include === 'mwfile' ) {
869 $namesMwFile = [];
870 # We do this using a foreach over the codes instead of a directory
871 # loop so that messages files in extensions will work correctly.
872 foreach ( $returnMw as $code => $value ) {
873 if ( is_readable( self::getMessagesFileName( $code ) )
874 || is_readable( self::getJsonMessagesFileName( $code ) )
875 ) {
876 $namesMwFile[$code] = $names[$code];
877 }
878 }
879
880 ksort( $namesMwFile );
881 return $namesMwFile;
882 }
883
884 ksort( $returnMw );
885 # 'mw' option; default if it's not one of the other two options (all/mwfile)
886 return $returnMw;
887 }
888
889 /**
890 * @param string $code The code of the language for which to get the name
891 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
892 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
893 * @return string Language name or empty
894 * @since 1.20
895 */
896 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
897 $code = strtolower( $code );
898 $array = self::fetchLanguageNames( $inLanguage, $include );
899 return !array_key_exists( $code, $array ) ? '' : $array[$code];
900 }
901
902 /**
903 * Get a message from the MediaWiki namespace.
904 *
905 * @param string $msg Message name
906 * @return string
907 */
908 public function getMessageFromDB( $msg ) {
909 return $this->msg( $msg )->text();
910 }
911
912 /**
913 * Get message object in this language. Only for use inside this class.
914 *
915 * @param string $msg Message name
916 * @return Message
917 */
918 protected function msg( $msg ) {
919 return wfMessage( $msg )->inLanguage( $this );
920 }
921
922 /**
923 * @param string $key
924 * @return string
925 */
926 public function getMonthName( $key ) {
927 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
928 }
929
930 /**
931 * @return array
932 */
933 public function getMonthNamesArray() {
934 $monthNames = [ '' ];
935 for ( $i = 1; $i < 13; $i++ ) {
936 $monthNames[] = $this->getMonthName( $i );
937 }
938 return $monthNames;
939 }
940
941 /**
942 * @param string $key
943 * @return string
944 */
945 public function getMonthNameGen( $key ) {
946 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
947 }
948
949 /**
950 * @param string $key
951 * @return string
952 */
953 public function getMonthAbbreviation( $key ) {
954 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
955 }
956
957 /**
958 * @return array
959 */
960 public function getMonthAbbreviationsArray() {
961 $monthNames = [ '' ];
962 for ( $i = 1; $i < 13; $i++ ) {
963 $monthNames[] = $this->getMonthAbbreviation( $i );
964 }
965 return $monthNames;
966 }
967
968 /**
969 * @param string $key
970 * @return string
971 */
972 public function getWeekdayName( $key ) {
973 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
974 }
975
976 /**
977 * @param string $key
978 * @return string
979 */
980 function getWeekdayAbbreviation( $key ) {
981 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
982 }
983
984 /**
985 * @param string $key
986 * @return string
987 */
988 function getIranianCalendarMonthName( $key ) {
989 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
990 }
991
992 /**
993 * @param string $key
994 * @return string
995 */
996 function getHebrewCalendarMonthName( $key ) {
997 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
998 }
999
1000 /**
1001 * @param string $key
1002 * @return string
1003 */
1004 function getHebrewCalendarMonthNameGen( $key ) {
1005 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1006 }
1007
1008 /**
1009 * @param string $key
1010 * @return string
1011 */
1012 function getHijriCalendarMonthName( $key ) {
1013 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1014 }
1015
1016 /**
1017 * Pass through result from $dateTimeObj->format()
1018 * @param DateTime|bool|null &$dateTimeObj
1019 * @param string $ts
1020 * @param DateTimeZone|bool|null $zone
1021 * @param string $code
1022 * @return string
1023 */
1024 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1025 if ( !$dateTimeObj ) {
1026 $dateTimeObj = DateTime::createFromFormat(
1027 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1028 );
1029 }
1030 return $dateTimeObj->format( $code );
1031 }
1032
1033 /**
1034 * This is a workalike of PHP's date() function, but with better
1035 * internationalisation, a reduced set of format characters, and a better
1036 * escaping format.
1037 *
1038 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1039 * the PHP manual for definitions. There are a number of extensions, which
1040 * start with "x":
1041 *
1042 * xn Do not translate digits of the next numeric format character
1043 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1044 * xr Use roman numerals for the next numeric format character
1045 * xh Use hebrew numerals for the next numeric format character
1046 * xx Literal x
1047 * xg Genitive month name
1048 *
1049 * xij j (day number) in Iranian calendar
1050 * xiF F (month name) in Iranian calendar
1051 * xin n (month number) in Iranian calendar
1052 * xiy y (two digit year) in Iranian calendar
1053 * xiY Y (full year) in Iranian calendar
1054 * xit t (days in month) in Iranian calendar
1055 * xiz z (day of the year) in Iranian calendar
1056 *
1057 * xjj j (day number) in Hebrew calendar
1058 * xjF F (month name) in Hebrew calendar
1059 * xjt t (days in month) in Hebrew calendar
1060 * xjx xg (genitive month name) in Hebrew calendar
1061 * xjn n (month number) in Hebrew calendar
1062 * xjY Y (full year) in Hebrew calendar
1063 *
1064 * xmj j (day number) in Hijri calendar
1065 * xmF F (month name) in Hijri calendar
1066 * xmn n (month number) in Hijri calendar
1067 * xmY Y (full year) in Hijri calendar
1068 *
1069 * xkY Y (full year) in Thai solar calendar. Months and days are
1070 * identical to the Gregorian calendar
1071 * xoY Y (full year) in Minguo calendar or Juche year.
1072 * Months and days are identical to the
1073 * Gregorian calendar
1074 * xtY Y (full year) in Japanese nengo. Months and days are
1075 * identical to the Gregorian calendar
1076 *
1077 * Characters enclosed in double quotes will be considered literal (with
1078 * the quotes themselves removed). Unmatched quotes will be considered
1079 * literal quotes. Example:
1080 *
1081 * "The month is" F => The month is January
1082 * i's" => 20'11"
1083 *
1084 * Backslash escaping is also supported.
1085 *
1086 * Input timestamp is assumed to be pre-normalized to the desired local
1087 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1088 * $ts is UTC if $zone is not given.
1089 *
1090 * @param string $format
1091 * @param string $ts 14-character timestamp
1092 * YYYYMMDDHHMMSS
1093 * 01234567890123
1094 * @param DateTimeZone $zone Timezone of $ts
1095 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1096 * Only makes sense if $ts is the current time.
1097 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1098 *
1099 * @throws MWException
1100 * @return string
1101 */
1102 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1103 $s = '';
1104 $raw = false;
1105 $roman = false;
1106 $hebrewNum = false;
1107 $dateTimeObj = false;
1108 $rawToggle = false;
1109 $iranian = false;
1110 $hebrew = false;
1111 $hijri = false;
1112 $thai = false;
1113 $minguo = false;
1114 $tenno = false;
1115
1116 $usedSecond = false;
1117 $usedMinute = false;
1118 $usedHour = false;
1119 $usedAMPM = false;
1120 $usedDay = false;
1121 $usedWeek = false;
1122 $usedMonth = false;
1123 $usedYear = false;
1124 $usedISOYear = false;
1125 $usedIsLeapYear = false;
1126
1127 $usedHebrewMonth = false;
1128 $usedIranianMonth = false;
1129 $usedHijriMonth = false;
1130 $usedHebrewYear = false;
1131 $usedIranianYear = false;
1132 $usedHijriYear = false;
1133 $usedTennoYear = false;
1134
1135 if ( strlen( $ts ) !== 14 ) {
1136 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1137 }
1138
1139 if ( !ctype_digit( $ts ) ) {
1140 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1141 }
1142
1143 $formatLength = strlen( $format );
1144 for ( $p = 0; $p < $formatLength; $p++ ) {
1145 $num = false;
1146 $code = $format[$p];
1147 if ( $code == 'x' && $p < $formatLength - 1 ) {
1148 $code .= $format[++$p];
1149 }
1150
1151 if ( ( $code === 'xi'
1152 || $code === 'xj'
1153 || $code === 'xk'
1154 || $code === 'xm'
1155 || $code === 'xo'
1156 || $code === 'xt' )
1157 && $p < $formatLength - 1 ) {
1158 $code .= $format[++$p];
1159 }
1160
1161 switch ( $code ) {
1162 case 'xx':
1163 $s .= 'x';
1164 break;
1165 case 'xn':
1166 $raw = true;
1167 break;
1168 case 'xN':
1169 $rawToggle = !$rawToggle;
1170 break;
1171 case 'xr':
1172 $roman = true;
1173 break;
1174 case 'xh':
1175 $hebrewNum = true;
1176 break;
1177 case 'xg':
1178 $usedMonth = true;
1179 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1180 break;
1181 case 'xjx':
1182 $usedHebrewMonth = true;
1183 if ( !$hebrew ) {
1184 $hebrew = self::tsToHebrew( $ts );
1185 }
1186 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1187 break;
1188 case 'd':
1189 $usedDay = true;
1190 $num = substr( $ts, 6, 2 );
1191 break;
1192 case 'D':
1193 $usedDay = true;
1194 $s .= $this->getWeekdayAbbreviation(
1195 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1196 );
1197 break;
1198 case 'j':
1199 $usedDay = true;
1200 $num = intval( substr( $ts, 6, 2 ) );
1201 break;
1202 case 'xij':
1203 $usedDay = true;
1204 if ( !$iranian ) {
1205 $iranian = self::tsToIranian( $ts );
1206 }
1207 $num = $iranian[2];
1208 break;
1209 case 'xmj':
1210 $usedDay = true;
1211 if ( !$hijri ) {
1212 $hijri = self::tsToHijri( $ts );
1213 }
1214 $num = $hijri[2];
1215 break;
1216 case 'xjj':
1217 $usedDay = true;
1218 if ( !$hebrew ) {
1219 $hebrew = self::tsToHebrew( $ts );
1220 }
1221 $num = $hebrew[2];
1222 break;
1223 case 'l':
1224 $usedDay = true;
1225 $s .= $this->getWeekdayName(
1226 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1227 );
1228 break;
1229 case 'F':
1230 $usedMonth = true;
1231 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1232 break;
1233 case 'xiF':
1234 $usedIranianMonth = true;
1235 if ( !$iranian ) {
1236 $iranian = self::tsToIranian( $ts );
1237 }
1238 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1239 break;
1240 case 'xmF':
1241 $usedHijriMonth = true;
1242 if ( !$hijri ) {
1243 $hijri = self::tsToHijri( $ts );
1244 }
1245 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1246 break;
1247 case 'xjF':
1248 $usedHebrewMonth = true;
1249 if ( !$hebrew ) {
1250 $hebrew = self::tsToHebrew( $ts );
1251 }
1252 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1253 break;
1254 case 'm':
1255 $usedMonth = true;
1256 $num = substr( $ts, 4, 2 );
1257 break;
1258 case 'M':
1259 $usedMonth = true;
1260 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1261 break;
1262 case 'n':
1263 $usedMonth = true;
1264 $num = intval( substr( $ts, 4, 2 ) );
1265 break;
1266 case 'xin':
1267 $usedIranianMonth = true;
1268 if ( !$iranian ) {
1269 $iranian = self::tsToIranian( $ts );
1270 }
1271 $num = $iranian[1];
1272 break;
1273 case 'xmn':
1274 $usedHijriMonth = true;
1275 if ( !$hijri ) {
1276 $hijri = self::tsToHijri( $ts );
1277 }
1278 $num = $hijri[1];
1279 break;
1280 case 'xjn':
1281 $usedHebrewMonth = true;
1282 if ( !$hebrew ) {
1283 $hebrew = self::tsToHebrew( $ts );
1284 }
1285 $num = $hebrew[1];
1286 break;
1287 case 'xjt':
1288 $usedHebrewMonth = true;
1289 if ( !$hebrew ) {
1290 $hebrew = self::tsToHebrew( $ts );
1291 }
1292 $num = $hebrew[3];
1293 break;
1294 case 'Y':
1295 $usedYear = true;
1296 $num = substr( $ts, 0, 4 );
1297 break;
1298 case 'xiY':
1299 $usedIranianYear = true;
1300 if ( !$iranian ) {
1301 $iranian = self::tsToIranian( $ts );
1302 }
1303 $num = $iranian[0];
1304 break;
1305 case 'xmY':
1306 $usedHijriYear = true;
1307 if ( !$hijri ) {
1308 $hijri = self::tsToHijri( $ts );
1309 }
1310 $num = $hijri[0];
1311 break;
1312 case 'xjY':
1313 $usedHebrewYear = true;
1314 if ( !$hebrew ) {
1315 $hebrew = self::tsToHebrew( $ts );
1316 }
1317 $num = $hebrew[0];
1318 break;
1319 case 'xkY':
1320 $usedYear = true;
1321 if ( !$thai ) {
1322 $thai = self::tsToYear( $ts, 'thai' );
1323 }
1324 $num = $thai[0];
1325 break;
1326 case 'xoY':
1327 $usedYear = true;
1328 if ( !$minguo ) {
1329 $minguo = self::tsToYear( $ts, 'minguo' );
1330 }
1331 $num = $minguo[0];
1332 break;
1333 case 'xtY':
1334 $usedTennoYear = true;
1335 if ( !$tenno ) {
1336 $tenno = self::tsToYear( $ts, 'tenno' );
1337 }
1338 $num = $tenno[0];
1339 break;
1340 case 'y':
1341 $usedYear = true;
1342 $num = substr( $ts, 2, 2 );
1343 break;
1344 case 'xiy':
1345 $usedIranianYear = true;
1346 if ( !$iranian ) {
1347 $iranian = self::tsToIranian( $ts );
1348 }
1349 $num = substr( $iranian[0], -2 );
1350 break;
1351 case 'xit':
1352 $usedIranianYear = true;
1353 if ( !$iranian ) {
1354 $iranian = self::tsToIranian( $ts );
1355 }
1356 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1357 break;
1358 case 'xiz':
1359 $usedIranianYear = true;
1360 if ( !$iranian ) {
1361 $iranian = self::tsToIranian( $ts );
1362 }
1363 $num = $iranian[3];
1364 break;
1365 case 'a':
1366 $usedAMPM = true;
1367 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1368 break;
1369 case 'A':
1370 $usedAMPM = true;
1371 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1372 break;
1373 case 'g':
1374 $usedHour = true;
1375 $h = substr( $ts, 8, 2 );
1376 $num = $h % 12 ? $h % 12 : 12;
1377 break;
1378 case 'G':
1379 $usedHour = true;
1380 $num = intval( substr( $ts, 8, 2 ) );
1381 break;
1382 case 'h':
1383 $usedHour = true;
1384 $h = substr( $ts, 8, 2 );
1385 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1386 break;
1387 case 'H':
1388 $usedHour = true;
1389 $num = substr( $ts, 8, 2 );
1390 break;
1391 case 'i':
1392 $usedMinute = true;
1393 $num = substr( $ts, 10, 2 );
1394 break;
1395 case 's':
1396 $usedSecond = true;
1397 $num = substr( $ts, 12, 2 );
1398 break;
1399 case 'c':
1400 case 'r':
1401 $usedSecond = true;
1402 // fall through
1403 case 'e':
1404 case 'O':
1405 case 'P':
1406 case 'T':
1407 $s .= self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1408 break;
1409 case 'w':
1410 case 'N':
1411 case 'z':
1412 $usedDay = true;
1413 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1414 break;
1415 case 'W':
1416 $usedWeek = true;
1417 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1418 break;
1419 case 't':
1420 $usedMonth = true;
1421 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1422 break;
1423 case 'L':
1424 $usedIsLeapYear = true;
1425 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1426 break;
1427 case 'o':
1428 $usedISOYear = true;
1429 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1430 break;
1431 case 'U':
1432 $usedSecond = true;
1433 // fall through
1434 case 'I':
1435 case 'Z':
1436 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1437 break;
1438 case '\\':
1439 # Backslash escaping
1440 if ( $p < $formatLength - 1 ) {
1441 $s .= $format[++$p];
1442 } else {
1443 $s .= '\\';
1444 }
1445 break;
1446 case '"':
1447 # Quoted literal
1448 if ( $p < $formatLength - 1 ) {
1449 $endQuote = strpos( $format, '"', $p + 1 );
1450 if ( $endQuote === false ) {
1451 # No terminating quote, assume literal "
1452 $s .= '"';
1453 } else {
1454 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1455 $p = $endQuote;
1456 }
1457 } else {
1458 # Quote at end of string, assume literal "
1459 $s .= '"';
1460 }
1461 break;
1462 default:
1463 $s .= $format[$p];
1464 }
1465 if ( $num !== false ) {
1466 if ( $rawToggle || $raw ) {
1467 $s .= $num;
1468 $raw = false;
1469 } elseif ( $roman ) {
1470 $s .= self::romanNumeral( $num );
1471 $roman = false;
1472 } elseif ( $hebrewNum ) {
1473 $s .= self::hebrewNumeral( $num );
1474 $hebrewNum = false;
1475 } else {
1476 $s .= $this->formatNum( $num, true );
1477 }
1478 }
1479 }
1480
1481 if ( $ttl === 'unused' ) {
1482 // No need to calculate the TTL, the caller wont use it anyway.
1483 } elseif ( $usedSecond ) {
1484 $ttl = 1;
1485 } elseif ( $usedMinute ) {
1486 $ttl = 60 - substr( $ts, 12, 2 );
1487 } elseif ( $usedHour ) {
1488 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1489 } elseif ( $usedAMPM ) {
1490 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1491 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1492 } elseif (
1493 $usedDay ||
1494 $usedHebrewMonth ||
1495 $usedIranianMonth ||
1496 $usedHijriMonth ||
1497 $usedHebrewYear ||
1498 $usedIranianYear ||
1499 $usedHijriYear ||
1500 $usedTennoYear
1501 ) {
1502 // @todo Someone who understands the non-Gregorian calendars
1503 // should write proper logic for them so that they don't need purged every day.
1504 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1505 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1506 } else {
1507 $possibleTtls = [];
1508 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1509 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1510 if ( $usedWeek ) {
1511 $possibleTtls[] =
1512 ( 7 - self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1513 $timeRemainingInDay;
1514 } elseif ( $usedISOYear ) {
1515 // December 28th falls on the last ISO week of the year, every year.
1516 // The last ISO week of a year can be 52 or 53.
1517 $lastWeekOfISOYear = DateTime::createFromFormat(
1518 'Ymd',
1519 substr( $ts, 0, 4 ) . '1228',
1520 $zone ?: new DateTimeZone( 'UTC' )
1521 )->format( 'W' );
1522 $currentISOWeek = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1523 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1524 $timeRemainingInWeek =
1525 ( 7 - self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1526 + $timeRemainingInDay;
1527 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1528 }
1529
1530 if ( $usedMonth ) {
1531 $possibleTtls[] =
1532 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1533 substr( $ts, 6, 2 ) ) * 86400
1534 + $timeRemainingInDay;
1535 } elseif ( $usedYear ) {
1536 $possibleTtls[] =
1537 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1538 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1539 + $timeRemainingInDay;
1540 } elseif ( $usedIsLeapYear ) {
1541 $year = substr( $ts, 0, 4 );
1542 $timeRemainingInYear =
1543 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1544 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1545 + $timeRemainingInDay;
1546 $mod = $year % 4;
1547 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1548 // this isn't a leap year. see when the next one starts
1549 $nextCandidate = $year - $mod + 4;
1550 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1551 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1552 $timeRemainingInYear;
1553 } else {
1554 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1555 $timeRemainingInYear;
1556 }
1557 } else {
1558 // this is a leap year, so the next year isn't
1559 $possibleTtls[] = $timeRemainingInYear;
1560 }
1561 }
1562
1563 if ( $possibleTtls ) {
1564 $ttl = min( $possibleTtls );
1565 }
1566 }
1567
1568 return $s;
1569 }
1570
1571 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1572 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1573
1574 /**
1575 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1576 * Gregorian dates to Iranian dates. Originally written in C, it
1577 * is released under the terms of GNU Lesser General Public
1578 * License. Conversion to PHP was performed by Niklas Laxström.
1579 *
1580 * Link: http://www.farsiweb.info/jalali/jalali.c
1581 *
1582 * @param string $ts
1583 *
1584 * @return int[]
1585 */
1586 private static function tsToIranian( $ts ) {
1587 $gy = substr( $ts, 0, 4 ) -1600;
1588 $gm = substr( $ts, 4, 2 ) -1;
1589 $gd = substr( $ts, 6, 2 ) -1;
1590
1591 # Days passed from the beginning (including leap years)
1592 $gDayNo = 365 * $gy
1593 + floor( ( $gy + 3 ) / 4 )
1594 - floor( ( $gy + 99 ) / 100 )
1595 + floor( ( $gy + 399 ) / 400 );
1596
1597 // Add days of the past months of this year
1598 for ( $i = 0; $i < $gm; $i++ ) {
1599 $gDayNo += self::$GREG_DAYS[$i];
1600 }
1601
1602 // Leap years
1603 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1604 $gDayNo++;
1605 }
1606
1607 // Days passed in current month
1608 $gDayNo += (int)$gd;
1609
1610 $jDayNo = $gDayNo - 79;
1611
1612 $jNp = floor( $jDayNo / 12053 );
1613 $jDayNo %= 12053;
1614
1615 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1616 $jDayNo %= 1461;
1617
1618 if ( $jDayNo >= 366 ) {
1619 $jy += floor( ( $jDayNo - 1 ) / 365 );
1620 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1621 }
1622
1623 $jz = $jDayNo;
1624
1625 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1626 $jDayNo -= self::$IRANIAN_DAYS[$i];
1627 }
1628
1629 $jm = $i + 1;
1630 $jd = $jDayNo + 1;
1631
1632 return [ $jy, $jm, $jd, $jz ];
1633 }
1634
1635 /**
1636 * Converting Gregorian dates to Hijri dates.
1637 *
1638 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1639 *
1640 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1641 *
1642 * @param string $ts
1643 *
1644 * @return int[]
1645 */
1646 private static function tsToHijri( $ts ) {
1647 $year = substr( $ts, 0, 4 );
1648 $month = substr( $ts, 4, 2 );
1649 $day = substr( $ts, 6, 2 );
1650
1651 $zyr = $year;
1652 $zd = $day;
1653 $zm = $month;
1654 $zy = $zyr;
1655
1656 if (
1657 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1658 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1659 ) {
1660 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1661 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1662 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1663 $zd - 32075;
1664 } else {
1665 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1666 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1667 }
1668
1669 $zl = $zjd -1948440 + 10632;
1670 $zn = (int)( ( $zl - 1 ) / 10631 );
1671 $zl = $zl - 10631 * $zn + 354;
1672 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1673 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1674 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1675 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1676 $zm = (int)( ( 24 * $zl ) / 709 );
1677 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1678 $zy = 30 * $zn + $zj - 30;
1679
1680 return [ $zy, $zm, $zd ];
1681 }
1682
1683 /**
1684 * Converting Gregorian dates to Hebrew dates.
1685 *
1686 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1687 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1688 * to translate the relevant functions into PHP and release them under
1689 * GNU GPL.
1690 *
1691 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1692 * and Adar II is 14. In a non-leap year, Adar is 6.
1693 *
1694 * @param string $ts
1695 *
1696 * @return int[]
1697 */
1698 private static function tsToHebrew( $ts ) {
1699 # Parse date
1700 $year = substr( $ts, 0, 4 );
1701 $month = substr( $ts, 4, 2 );
1702 $day = substr( $ts, 6, 2 );
1703
1704 # Calculate Hebrew year
1705 $hebrewYear = $year + 3760;
1706
1707 # Month number when September = 1, August = 12
1708 $month += 4;
1709 if ( $month > 12 ) {
1710 # Next year
1711 $month -= 12;
1712 $year++;
1713 $hebrewYear++;
1714 }
1715
1716 # Calculate day of year from 1 September
1717 $dayOfYear = $day;
1718 for ( $i = 1; $i < $month; $i++ ) {
1719 if ( $i == 6 ) {
1720 # February
1721 $dayOfYear += 28;
1722 # Check if the year is leap
1723 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1724 $dayOfYear++;
1725 }
1726 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1727 $dayOfYear += 30;
1728 } else {
1729 $dayOfYear += 31;
1730 }
1731 }
1732
1733 # Calculate the start of the Hebrew year
1734 $start = self::hebrewYearStart( $hebrewYear );
1735
1736 # Calculate next year's start
1737 if ( $dayOfYear <= $start ) {
1738 # Day is before the start of the year - it is the previous year
1739 # Next year's start
1740 $nextStart = $start;
1741 # Previous year
1742 $year--;
1743 $hebrewYear--;
1744 # Add days since previous year's 1 September
1745 $dayOfYear += 365;
1746 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1747 # Leap year
1748 $dayOfYear++;
1749 }
1750 # Start of the new (previous) year
1751 $start = self::hebrewYearStart( $hebrewYear );
1752 } else {
1753 # Next year's start
1754 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1755 }
1756
1757 # Calculate Hebrew day of year
1758 $hebrewDayOfYear = $dayOfYear - $start;
1759
1760 # Difference between year's days
1761 $diff = $nextStart - $start;
1762 # Add 12 (or 13 for leap years) days to ignore the difference between
1763 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1764 # difference is only about the year type
1765 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1766 $diff += 13;
1767 } else {
1768 $diff += 12;
1769 }
1770
1771 # Check the year pattern, and is leap year
1772 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1773 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1774 # and non-leap years
1775 $yearPattern = $diff % 30;
1776 # Check if leap year
1777 $isLeap = $diff >= 30;
1778
1779 # Calculate day in the month from number of day in the Hebrew year
1780 # Don't check Adar - if the day is not in Adar, we will stop before;
1781 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1782 $hebrewDay = $hebrewDayOfYear;
1783 $hebrewMonth = 1;
1784 $days = 0;
1785 while ( $hebrewMonth <= 12 ) {
1786 # Calculate days in this month
1787 if ( $isLeap && $hebrewMonth == 6 ) {
1788 # Adar in a leap year
1789 if ( $isLeap ) {
1790 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1791 $days = 30;
1792 if ( $hebrewDay <= $days ) {
1793 # Day in Adar I
1794 $hebrewMonth = 13;
1795 } else {
1796 # Subtract the days of Adar I
1797 $hebrewDay -= $days;
1798 # Try Adar II
1799 $days = 29;
1800 if ( $hebrewDay <= $days ) {
1801 # Day in Adar II
1802 $hebrewMonth = 14;
1803 }
1804 }
1805 }
1806 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1807 # Cheshvan in a complete year (otherwise as the rule below)
1808 $days = 30;
1809 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1810 # Kislev in an incomplete year (otherwise as the rule below)
1811 $days = 29;
1812 } else {
1813 # Odd months have 30 days, even have 29
1814 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1815 }
1816 if ( $hebrewDay <= $days ) {
1817 # In the current month
1818 break;
1819 } else {
1820 # Subtract the days of the current month
1821 $hebrewDay -= $days;
1822 # Try in the next month
1823 $hebrewMonth++;
1824 }
1825 }
1826
1827 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1828 }
1829
1830 /**
1831 * This calculates the Hebrew year start, as days since 1 September.
1832 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1833 * Used for Hebrew date.
1834 *
1835 * @param int $year
1836 *
1837 * @return string
1838 */
1839 private static function hebrewYearStart( $year ) {
1840 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1841 $b = intval( ( $year - 1 ) % 4 );
1842 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1843 if ( $m < 0 ) {
1844 $m--;
1845 }
1846 $Mar = intval( $m );
1847 if ( $m < 0 ) {
1848 $m++;
1849 }
1850 $m -= $Mar;
1851
1852 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1853 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1854 $Mar++;
1855 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1856 $Mar += 2;
1857 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1858 $Mar++;
1859 }
1860
1861 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1862 return $Mar;
1863 }
1864
1865 /**
1866 * Algorithm to convert Gregorian dates to Thai solar dates,
1867 * Minguo dates or Minguo dates.
1868 *
1869 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1870 * https://en.wikipedia.org/wiki/Minguo_calendar
1871 * https://en.wikipedia.org/wiki/Japanese_era_name
1872 *
1873 * @param string $ts 14-character timestamp
1874 * @param string $cName Calender name
1875 * @return array Converted year, month, day
1876 */
1877 private static function tsToYear( $ts, $cName ) {
1878 $gy = substr( $ts, 0, 4 );
1879 $gm = substr( $ts, 4, 2 );
1880 $gd = substr( $ts, 6, 2 );
1881
1882 if ( !strcmp( $cName, 'thai' ) ) {
1883 # Thai solar dates
1884 # Add 543 years to the Gregorian calendar
1885 # Months and days are identical
1886 $gy_offset = $gy + 543;
1887 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1888 # Minguo dates
1889 # Deduct 1911 years from the Gregorian calendar
1890 # Months and days are identical
1891 $gy_offset = $gy - 1911;
1892 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1893 # Nengō dates up to Meiji period
1894 # Deduct years from the Gregorian calendar
1895 # depending on the nengo periods
1896 # Months and days are identical
1897 if ( ( $gy < 1912 )
1898 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1899 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1900 ) {
1901 # Meiji period
1902 $gy_gannen = $gy - 1868 + 1;
1903 $gy_offset = $gy_gannen;
1904 if ( $gy_gannen == 1 ) {
1905 $gy_offset = '元';
1906 }
1907 $gy_offset = '明治' . $gy_offset;
1908 } elseif (
1909 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1910 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1911 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1912 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1913 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1914 ) {
1915 # Taishō period
1916 $gy_gannen = $gy - 1912 + 1;
1917 $gy_offset = $gy_gannen;
1918 if ( $gy_gannen == 1 ) {
1919 $gy_offset = '元';
1920 }
1921 $gy_offset = '大正' . $gy_offset;
1922 } elseif (
1923 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1924 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1925 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1926 ) {
1927 # Shōwa period
1928 $gy_gannen = $gy - 1926 + 1;
1929 $gy_offset = $gy_gannen;
1930 if ( $gy_gannen == 1 ) {
1931 $gy_offset = '元';
1932 }
1933 $gy_offset = '昭和' . $gy_offset;
1934 } else {
1935 # Heisei period
1936 $gy_gannen = $gy - 1989 + 1;
1937 $gy_offset = $gy_gannen;
1938 if ( $gy_gannen == 1 ) {
1939 $gy_offset = '元';
1940 }
1941 $gy_offset = '平成' . $gy_offset;
1942 }
1943 } else {
1944 $gy_offset = $gy;
1945 }
1946
1947 return [ $gy_offset, $gm, $gd ];
1948 }
1949
1950 /**
1951 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1952 *
1953 * This is the rule the BIDI algorithm uses to determine the directionality of
1954 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1955 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1956 *
1957 * TODO: Does not handle BIDI control characters inside the text.
1958 * TODO: Does not handle unallocated characters.
1959 *
1960 * @param string $text Text to test
1961 * @return null|string Directionality ('ltr' or 'rtl') or null
1962 */
1963 private static function strongDirFromContent( $text = '' ) {
1964 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1965 return null;
1966 }
1967 if ( $matches[1] === '' ) {
1968 return 'rtl';
1969 }
1970 return 'ltr';
1971 }
1972
1973 /**
1974 * Roman number formatting up to 10000
1975 *
1976 * @param int $num
1977 *
1978 * @return string
1979 */
1980 static function romanNumeral( $num ) {
1981 static $table = [
1982 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1983 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1984 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1985 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1986 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1987 ];
1988
1989 $num = intval( $num );
1990 if ( $num > 10000 || $num <= 0 ) {
1991 return $num;
1992 }
1993
1994 $s = '';
1995 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1996 if ( $num >= $pow10 ) {
1997 $s .= $table[$i][(int)floor( $num / $pow10 )];
1998 }
1999 $num = $num % $pow10;
2000 }
2001 return $s;
2002 }
2003
2004 /**
2005 * Hebrew Gematria number formatting up to 9999
2006 *
2007 * @param int $num
2008 *
2009 * @return string
2010 */
2011 static function hebrewNumeral( $num ) {
2012 static $table = [
2013 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2014 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2015 [ '',
2016 [ 'ק' ],
2017 [ 'ר' ],
2018 [ 'ש' ],
2019 [ 'ת' ],
2020 [ 'ת', 'ק' ],
2021 [ 'ת', 'ר' ],
2022 [ 'ת', 'ש' ],
2023 [ 'ת', 'ת' ],
2024 [ 'ת', 'ת', 'ק' ],
2025 [ 'ת', 'ת', 'ר' ],
2026 ],
2027 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2028 ];
2029
2030 $num = intval( $num );
2031 if ( $num > 9999 || $num <= 0 ) {
2032 return $num;
2033 }
2034
2035 // Round thousands have special notations
2036 if ( $num === 1000 ) {
2037 return "א' אלף";
2038 } elseif ( $num % 1000 === 0 ) {
2039 return $table[0][$num / 1000] . "' אלפים";
2040 }
2041
2042 $letters = [];
2043
2044 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2045 if ( $num >= $pow10 ) {
2046 if ( $num === 15 || $num === 16 ) {
2047 $letters[] = $table[0][9];
2048 $letters[] = $table[0][$num - 9];
2049 $num = 0;
2050 } else {
2051 $letters = array_merge(
2052 $letters,
2053 (array)$table[$i][intval( $num / $pow10 )]
2054 );
2055
2056 if ( $pow10 === 1000 ) {
2057 $letters[] = "'";
2058 }
2059 }
2060 }
2061
2062 $num = $num % $pow10;
2063 }
2064
2065 $preTransformLength = count( $letters );
2066 if ( $preTransformLength === 1 ) {
2067 // Add geresh (single quote) to one-letter numbers
2068 $letters[] = "'";
2069 } else {
2070 $lastIndex = $preTransformLength - 1;
2071 $letters[$lastIndex] = str_replace(
2072 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2073 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2074 $letters[$lastIndex]
2075 );
2076
2077 // Add gershayim (double quote) to multiple-letter numbers,
2078 // but exclude numbers with only one letter after the thousands
2079 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2080 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2081 $letters[] = "'";
2082 } else {
2083 array_splice( $letters, -1, 0, '"' );
2084 }
2085 }
2086
2087 return implode( $letters );
2088 }
2089
2090 /**
2091 * Used by date() and time() to adjust the time output.
2092 *
2093 * @param string $ts The time in date('YmdHis') format
2094 * @param mixed $tz Adjust the time by this amount (default false, mean we
2095 * get user timecorrection setting)
2096 * @return int
2097 */
2098 public function userAdjust( $ts, $tz = false ) {
2099 global $wgUser, $wgLocalTZoffset;
2100
2101 if ( $tz === false ) {
2102 $tz = $wgUser->getOption( 'timecorrection' );
2103 }
2104
2105 $data = explode( '|', $tz, 3 );
2106
2107 if ( $data[0] == 'ZoneInfo' ) {
2108 try {
2109 $userTZ = new DateTimeZone( $data[2] );
2110 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2111 $date->setTimezone( $userTZ );
2112 return $date->format( 'YmdHis' );
2113 } catch ( Exception $e ) {
2114 // Unrecognized timezone, default to 'Offset' with the stored offset.
2115 $data[0] = 'Offset';
2116 }
2117 }
2118
2119 if ( $data[0] == 'System' || $tz == '' ) {
2120 # Global offset in minutes.
2121 $minDiff = $wgLocalTZoffset;
2122 } elseif ( $data[0] == 'Offset' ) {
2123 $minDiff = intval( $data[1] );
2124 } else {
2125 $data = explode( ':', $tz );
2126 if ( count( $data ) == 2 ) {
2127 $data[0] = intval( $data[0] );
2128 $data[1] = intval( $data[1] );
2129 $minDiff = abs( $data[0] ) * 60 + $data[1];
2130 if ( $data[0] < 0 ) {
2131 $minDiff = -$minDiff;
2132 }
2133 } else {
2134 $minDiff = intval( $data[0] ) * 60;
2135 }
2136 }
2137
2138 # No difference ? Return time unchanged
2139 if ( 0 == $minDiff ) {
2140 return $ts;
2141 }
2142
2143 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2144 # Generate an adjusted date; take advantage of the fact that mktime
2145 # will normalize out-of-range values so we don't have to split $minDiff
2146 # into hours and minutes.
2147 $t = mktime( (
2148 (int)substr( $ts, 8, 2 ) ), # Hours
2149 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2150 (int)substr( $ts, 12, 2 ), # Seconds
2151 (int)substr( $ts, 4, 2 ), # Month
2152 (int)substr( $ts, 6, 2 ), # Day
2153 (int)substr( $ts, 0, 4 ) ); # Year
2154
2155 $date = date( 'YmdHis', $t );
2156 MediaWiki\restoreWarnings();
2157
2158 return $date;
2159 }
2160
2161 /**
2162 * This is meant to be used by time(), date(), and timeanddate() to get
2163 * the date preference they're supposed to use, it should be used in
2164 * all children.
2165 *
2166 * function timeanddate([...], $format = true) {
2167 * $datePreference = $this->dateFormat($format);
2168 * [...]
2169 * }
2170 *
2171 * @param int|string|bool $usePrefs If true, the user's preference is used
2172 * if false, the site/language default is used
2173 * if int/string, assumed to be a format.
2174 * @return string
2175 */
2176 function dateFormat( $usePrefs = true ) {
2177 global $wgUser;
2178
2179 if ( is_bool( $usePrefs ) ) {
2180 if ( $usePrefs ) {
2181 $datePreference = $wgUser->getDatePreference();
2182 } else {
2183 $datePreference = (string)User::getDefaultOption( 'date' );
2184 }
2185 } else {
2186 $datePreference = (string)$usePrefs;
2187 }
2188
2189 // return int
2190 if ( $datePreference == '' ) {
2191 return 'default';
2192 }
2193
2194 return $datePreference;
2195 }
2196
2197 /**
2198 * Get a format string for a given type and preference
2199 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2200 * @param string $pref The format name as it appears in Messages*.php under
2201 * $datePreferences.
2202 *
2203 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2204 *
2205 * @return string
2206 */
2207 function getDateFormatString( $type, $pref ) {
2208 $wasDefault = false;
2209 if ( $pref == 'default' ) {
2210 $wasDefault = true;
2211 $pref = $this->getDefaultDateFormat();
2212 }
2213
2214 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2215 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2216
2217 if ( $type === 'pretty' && $df === null ) {
2218 $df = $this->getDateFormatString( 'date', $pref );
2219 }
2220
2221 if ( !$wasDefault && $df === null ) {
2222 $pref = $this->getDefaultDateFormat();
2223 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2224 }
2225
2226 $this->dateFormatStrings[$type][$pref] = $df;
2227 }
2228 return $this->dateFormatStrings[$type][$pref];
2229 }
2230
2231 /**
2232 * @param string $ts The time format which needs to be turned into a
2233 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2234 * @param bool $adj Whether to adjust the time output according to the
2235 * user configured offset ($timecorrection)
2236 * @param mixed $format True to use user's date format preference
2237 * @param string|bool $timecorrection The time offset as returned by
2238 * validateTimeZone() in Special:Preferences
2239 * @return string
2240 */
2241 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2242 $ts = wfTimestamp( TS_MW, $ts );
2243 if ( $adj ) {
2244 $ts = $this->userAdjust( $ts, $timecorrection );
2245 }
2246 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2247 return $this->sprintfDate( $df, $ts );
2248 }
2249
2250 /**
2251 * @param string $ts The time format which needs to be turned into a
2252 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2253 * @param bool $adj Whether to adjust the time output according to the
2254 * user configured offset ($timecorrection)
2255 * @param mixed $format True to use user's date format preference
2256 * @param string|bool $timecorrection The time offset as returned by
2257 * validateTimeZone() in Special:Preferences
2258 * @return string
2259 */
2260 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2261 $ts = wfTimestamp( TS_MW, $ts );
2262 if ( $adj ) {
2263 $ts = $this->userAdjust( $ts, $timecorrection );
2264 }
2265 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2266 return $this->sprintfDate( $df, $ts );
2267 }
2268
2269 /**
2270 * @param string $ts The time format which needs to be turned into a
2271 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2272 * @param bool $adj Whether to adjust the time output according to the
2273 * user configured offset ($timecorrection)
2274 * @param mixed $format What format to return, if it's false output the
2275 * default one (default true)
2276 * @param string|bool $timecorrection The time offset as returned by
2277 * validateTimeZone() in Special:Preferences
2278 * @return string
2279 */
2280 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2281 $ts = wfTimestamp( TS_MW, $ts );
2282 if ( $adj ) {
2283 $ts = $this->userAdjust( $ts, $timecorrection );
2284 }
2285 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2286 return $this->sprintfDate( $df, $ts );
2287 }
2288
2289 /**
2290 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2291 *
2292 * @since 1.20
2293 *
2294 * @param int $seconds The amount of seconds.
2295 * @param array $chosenIntervals The intervals to enable.
2296 *
2297 * @return string
2298 */
2299 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2300 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2301
2302 $segments = [];
2303
2304 foreach ( $intervals as $intervalName => $intervalValue ) {
2305 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2306 // duration-years, duration-decades, duration-centuries, duration-millennia
2307 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2308 $segments[] = $message->inLanguage( $this )->escaped();
2309 }
2310
2311 return $this->listToText( $segments );
2312 }
2313
2314 /**
2315 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2316 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2317 *
2318 * @since 1.20
2319 *
2320 * @param int $seconds The amount of seconds.
2321 * @param array $chosenIntervals The intervals to enable.
2322 *
2323 * @return array
2324 */
2325 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2326 if ( empty( $chosenIntervals ) ) {
2327 $chosenIntervals = [
2328 'millennia',
2329 'centuries',
2330 'decades',
2331 'years',
2332 'days',
2333 'hours',
2334 'minutes',
2335 'seconds'
2336 ];
2337 }
2338
2339 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2340 $sortedNames = array_keys( $intervals );
2341 $smallestInterval = array_pop( $sortedNames );
2342
2343 $segments = [];
2344
2345 foreach ( $intervals as $name => $length ) {
2346 $value = floor( $seconds / $length );
2347
2348 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2349 $seconds -= $value * $length;
2350 $segments[$name] = $value;
2351 }
2352 }
2353
2354 return $segments;
2355 }
2356
2357 /**
2358 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2359 *
2360 * @param string $type Can be 'date', 'time' or 'both'
2361 * @param string $ts The time format which needs to be turned into a
2362 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2363 * @param User $user User object used to get preferences for timezone and format
2364 * @param array $options Array, can contain the following keys:
2365 * - 'timecorrection': time correction, can have the following values:
2366 * - true: use user's preference
2367 * - false: don't use time correction
2368 * - int: value of time correction in minutes
2369 * - 'format': format to use, can have the following values:
2370 * - true: use user's preference
2371 * - false: use default preference
2372 * - string: format to use
2373 * @since 1.19
2374 * @return string
2375 */
2376 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2377 $ts = wfTimestamp( TS_MW, $ts );
2378 $options += [ 'timecorrection' => true, 'format' => true ];
2379 if ( $options['timecorrection'] !== false ) {
2380 if ( $options['timecorrection'] === true ) {
2381 $offset = $user->getOption( 'timecorrection' );
2382 } else {
2383 $offset = $options['timecorrection'];
2384 }
2385 $ts = $this->userAdjust( $ts, $offset );
2386 }
2387 if ( $options['format'] === true ) {
2388 $format = $user->getDatePreference();
2389 } else {
2390 $format = $options['format'];
2391 }
2392 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2393 return $this->sprintfDate( $df, $ts );
2394 }
2395
2396 /**
2397 * Get the formatted date for the given timestamp and formatted for
2398 * the given user.
2399 *
2400 * @param mixed $ts Mixed: the time format which needs to be turned into a
2401 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2402 * @param User $user User object used to get preferences for timezone and format
2403 * @param array $options Array, can contain the following keys:
2404 * - 'timecorrection': time correction, can have the following values:
2405 * - true: use user's preference
2406 * - false: don't use time correction
2407 * - int: value of time correction in minutes
2408 * - 'format': format to use, can have the following values:
2409 * - true: use user's preference
2410 * - false: use default preference
2411 * - string: format to use
2412 * @since 1.19
2413 * @return string
2414 */
2415 public function userDate( $ts, User $user, array $options = [] ) {
2416 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2417 }
2418
2419 /**
2420 * Get the formatted time for the given timestamp and formatted for
2421 * the given user.
2422 *
2423 * @param mixed $ts The time format which needs to be turned into a
2424 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2425 * @param User $user User object used to get preferences for timezone and format
2426 * @param array $options Array, can contain the following keys:
2427 * - 'timecorrection': time correction, can have the following values:
2428 * - true: use user's preference
2429 * - false: don't use time correction
2430 * - int: value of time correction in minutes
2431 * - 'format': format to use, can have the following values:
2432 * - true: use user's preference
2433 * - false: use default preference
2434 * - string: format to use
2435 * @since 1.19
2436 * @return string
2437 */
2438 public function userTime( $ts, User $user, array $options = [] ) {
2439 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2440 }
2441
2442 /**
2443 * Get the formatted date and time for the given timestamp and formatted for
2444 * the given user.
2445 *
2446 * @param mixed $ts The time format which needs to be turned into a
2447 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2448 * @param User $user User object used to get preferences for timezone and format
2449 * @param array $options Array, can contain the following keys:
2450 * - 'timecorrection': time correction, can have the following values:
2451 * - true: use user's preference
2452 * - false: don't use time correction
2453 * - int: value of time correction in minutes
2454 * - 'format': format to use, can have the following values:
2455 * - true: use user's preference
2456 * - false: use default preference
2457 * - string: format to use
2458 * @since 1.19
2459 * @return string
2460 */
2461 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2462 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2463 }
2464
2465 /**
2466 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2467 *
2468 * Determine the difference between the timestamp and the current time, and
2469 * generate a readable timestamp by returning "<N> <units> ago", where the
2470 * largest possible unit is used.
2471 *
2472 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2473 *
2474 * @param MWTimestamp $time
2475 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2476 * @param User|null $user User the timestamp is being generated for
2477 * (or null to use main context's user)
2478 * @return string Formatted timestamp
2479 */
2480 public function getHumanTimestamp(
2481 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2482 ) {
2483 if ( $relativeTo === null ) {
2484 $relativeTo = new MWTimestamp();
2485 }
2486 if ( $user === null ) {
2487 $user = RequestContext::getMain()->getUser();
2488 }
2489
2490 // Adjust for the user's timezone.
2491 $offsetThis = $time->offsetForUser( $user );
2492 $offsetRel = $relativeTo->offsetForUser( $user );
2493
2494 $ts = '';
2495 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2496 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2497 }
2498
2499 // Reset the timezone on the objects.
2500 $time->timestamp->sub( $offsetThis );
2501 $relativeTo->timestamp->sub( $offsetRel );
2502
2503 return $ts;
2504 }
2505
2506 /**
2507 * Convert an MWTimestamp into a pretty human-readable timestamp using
2508 * the given user preferences and relative base time.
2509 *
2510 * @see Language::getHumanTimestamp
2511 * @param MWTimestamp $ts Timestamp to prettify
2512 * @param MWTimestamp $relativeTo Base timestamp
2513 * @param User $user User preferences to use
2514 * @return string Human timestamp
2515 * @since 1.26
2516 */
2517 private function getHumanTimestampInternal(
2518 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2519 ) {
2520 $diff = $ts->diff( $relativeTo );
2521 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2522 (int)$relativeTo->timestamp->format( 'w' ) );
2523 $days = $diff->days ?: (int)$diffDay;
2524 if ( $diff->invert || $days > 5
2525 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2526 ) {
2527 // Timestamps are in different years: use full timestamp
2528 // Also do full timestamp for future dates
2529 /**
2530 * @todo FIXME: Add better handling of future timestamps.
2531 */
2532 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2533 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2534 } elseif ( $days > 5 ) {
2535 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2536 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2537 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2538 } elseif ( $days > 1 ) {
2539 // Timestamp within the past week: show the day of the week and time
2540 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2541 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2542 // Messages:
2543 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2544 $ts = wfMessage( "$weekday-at" )
2545 ->inLanguage( $this )
2546 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2547 ->text();
2548 } elseif ( $days == 1 ) {
2549 // Timestamp was yesterday: say 'yesterday' and the time.
2550 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2551 $ts = wfMessage( 'yesterday-at' )
2552 ->inLanguage( $this )
2553 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2554 ->text();
2555 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2556 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2557 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2558 $ts = wfMessage( 'today-at' )
2559 ->inLanguage( $this )
2560 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2561 ->text();
2562
2563 // From here on in, the timestamp was soon enough ago so that we can simply say
2564 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2565 } elseif ( $diff->h == 1 ) {
2566 // Less than 90 minutes, but more than an hour ago.
2567 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2568 } elseif ( $diff->i >= 1 ) {
2569 // A few minutes ago.
2570 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2571 } elseif ( $diff->s >= 30 ) {
2572 // Less than a minute, but more than 30 sec ago.
2573 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2574 } else {
2575 // Less than 30 seconds ago.
2576 $ts = wfMessage( 'just-now' )->text();
2577 }
2578
2579 return $ts;
2580 }
2581
2582 /**
2583 * @param string $key
2584 * @return string|null
2585 */
2586 public function getMessage( $key ) {
2587 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2588 }
2589
2590 /**
2591 * @return array
2592 */
2593 function getAllMessages() {
2594 return self::$dataCache->getItem( $this->mCode, 'messages' );
2595 }
2596
2597 /**
2598 * @param string $in
2599 * @param string $out
2600 * @param string $string
2601 * @return string
2602 */
2603 public function iconv( $in, $out, $string ) {
2604 # Even with //IGNORE iconv can whine about illegal characters in
2605 # *input* string. We just ignore those too.
2606 # REF: https://bugs.php.net/bug.php?id=37166
2607 # REF: https://phabricator.wikimedia.org/T18885
2608 MediaWiki\suppressWarnings();
2609 $text = iconv( $in, $out . '//IGNORE', $string );
2610 MediaWiki\restoreWarnings();
2611 return $text;
2612 }
2613
2614 // callback functions for ucwords(), ucwordbreaks()
2615
2616 /**
2617 * @param array $matches
2618 * @return mixed|string
2619 */
2620 function ucwordbreaksCallbackAscii( $matches ) {
2621 return $this->ucfirst( $matches[1] );
2622 }
2623
2624 /**
2625 * @param array $matches
2626 * @return string
2627 */
2628 function ucwordbreaksCallbackMB( $matches ) {
2629 return mb_strtoupper( $matches[0] );
2630 }
2631
2632 /**
2633 * @param array $matches
2634 * @return string
2635 */
2636 function ucwordsCallbackMB( $matches ) {
2637 return mb_strtoupper( $matches[0] );
2638 }
2639
2640 /**
2641 * Make a string's first character uppercase
2642 *
2643 * @param string $str
2644 *
2645 * @return string
2646 */
2647 public function ucfirst( $str ) {
2648 $o = ord( $str );
2649 if ( $o < 96 ) { // if already uppercase...
2650 return $str;
2651 } elseif ( $o < 128 ) {
2652 return ucfirst( $str ); // use PHP's ucfirst()
2653 } else {
2654 // fall back to more complex logic in case of multibyte strings
2655 return $this->uc( $str, true );
2656 }
2657 }
2658
2659 /**
2660 * Convert a string to uppercase
2661 *
2662 * @param string $str
2663 * @param bool $first
2664 *
2665 * @return string
2666 */
2667 public function uc( $str, $first = false ) {
2668 if ( $first ) {
2669 if ( $this->isMultibyte( $str ) ) {
2670 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2671 } else {
2672 return ucfirst( $str );
2673 }
2674 } else {
2675 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2676 }
2677 }
2678
2679 /**
2680 * @param string $str
2681 * @return mixed|string
2682 */
2683 function lcfirst( $str ) {
2684 $o = ord( $str );
2685 if ( !$o ) {
2686 return strval( $str );
2687 } elseif ( $o >= 128 ) {
2688 return $this->lc( $str, true );
2689 } elseif ( $o > 96 ) {
2690 return $str;
2691 } else {
2692 $str[0] = strtolower( $str[0] );
2693 return $str;
2694 }
2695 }
2696
2697 /**
2698 * @param string $str
2699 * @param bool $first
2700 * @return mixed|string
2701 */
2702 function lc( $str, $first = false ) {
2703 if ( $first ) {
2704 if ( $this->isMultibyte( $str ) ) {
2705 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2706 } else {
2707 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2708 }
2709 } else {
2710 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2711 }
2712 }
2713
2714 /**
2715 * @param string $str
2716 * @return bool
2717 */
2718 function isMultibyte( $str ) {
2719 return strlen( $str ) !== mb_strlen( $str );
2720 }
2721
2722 /**
2723 * @param string $str
2724 * @return mixed|string
2725 */
2726 function ucwords( $str ) {
2727 if ( $this->isMultibyte( $str ) ) {
2728 $str = $this->lc( $str );
2729
2730 // regexp to find first letter in each word (i.e. after each space)
2731 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2732
2733 // function to use to capitalize a single char
2734 return preg_replace_callback(
2735 $replaceRegexp,
2736 [ $this, 'ucwordsCallbackMB' ],
2737 $str
2738 );
2739 } else {
2740 return ucwords( strtolower( $str ) );
2741 }
2742 }
2743
2744 /**
2745 * capitalize words at word breaks
2746 *
2747 * @param string $str
2748 * @return mixed
2749 */
2750 function ucwordbreaks( $str ) {
2751 if ( $this->isMultibyte( $str ) ) {
2752 $str = $this->lc( $str );
2753
2754 // since \b doesn't work for UTF-8, we explicitely define word break chars
2755 $breaks = "[ \-\(\)\}\{\.,\?!]";
2756
2757 // find first letter after word break
2758 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2759 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2760
2761 return preg_replace_callback(
2762 $replaceRegexp,
2763 [ $this, 'ucwordbreaksCallbackMB' ],
2764 $str
2765 );
2766 } else {
2767 return preg_replace_callback(
2768 '/\b([\w\x80-\xff]+)\b/',
2769 [ $this, 'ucwordbreaksCallbackAscii' ],
2770 $str
2771 );
2772 }
2773 }
2774
2775 /**
2776 * Return a case-folded representation of $s
2777 *
2778 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2779 * and $s2 are the same except for the case of their characters. It is not
2780 * necessary for the value returned to make sense when displayed.
2781 *
2782 * Do *not* perform any other normalisation in this function. If a caller
2783 * uses this function when it should be using a more general normalisation
2784 * function, then fix the caller.
2785 *
2786 * @param string $s
2787 *
2788 * @return string
2789 */
2790 function caseFold( $s ) {
2791 return $this->uc( $s );
2792 }
2793
2794 /**
2795 * @param string $s
2796 * @return string
2797 * @throws MWException
2798 */
2799 function checkTitleEncoding( $s ) {
2800 if ( is_array( $s ) ) {
2801 throw new MWException( 'Given array to checkTitleEncoding.' );
2802 }
2803 if ( StringUtils::isUtf8( $s ) ) {
2804 return $s;
2805 }
2806
2807 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2808 }
2809
2810 /**
2811 * @return array
2812 */
2813 function fallback8bitEncoding() {
2814 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2815 }
2816
2817 /**
2818 * Most writing systems use whitespace to break up words.
2819 * Some languages such as Chinese don't conventionally do this,
2820 * which requires special handling when breaking up words for
2821 * searching etc.
2822 *
2823 * @return bool
2824 */
2825 function hasWordBreaks() {
2826 return true;
2827 }
2828
2829 /**
2830 * Some languages such as Chinese require word segmentation,
2831 * Specify such segmentation when overridden in derived class.
2832 *
2833 * @param string $string
2834 * @return string
2835 */
2836 function segmentByWord( $string ) {
2837 return $string;
2838 }
2839
2840 /**
2841 * Some languages have special punctuation need to be normalized.
2842 * Make such changes here.
2843 *
2844 * @param string $string
2845 * @return string
2846 */
2847 function normalizeForSearch( $string ) {
2848 return self::convertDoubleWidth( $string );
2849 }
2850
2851 /**
2852 * convert double-width roman characters to single-width.
2853 * range: ff00-ff5f ~= 0020-007f
2854 *
2855 * @param string $string
2856 *
2857 * @return string
2858 */
2859 protected static function convertDoubleWidth( $string ) {
2860 static $full = null;
2861 static $half = null;
2862
2863 if ( $full === null ) {
2864 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2865 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2866 $full = str_split( $fullWidth, 3 );
2867 $half = str_split( $halfWidth );
2868 }
2869
2870 $string = str_replace( $full, $half, $string );
2871 return $string;
2872 }
2873
2874 /**
2875 * @param string $string
2876 * @param string $pattern
2877 * @return string
2878 */
2879 protected static function insertSpace( $string, $pattern ) {
2880