?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
API: Allow fetching login token from action=query&meta=tokens on private wikis
[mediawiki.git] / languages / Language.php
bloba14a376a6ab7ef8c05751b60e4b6c0178bde680c
1 <?php
2 /**
3 * Internationalisation code.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Language
22 */
23
24 /**
25 * @defgroup Language Language
26 */
27
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
30 exit( 1 );
31 }
32
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
35 }
36
37 use CLDRPluralRuleParser\Evaluator;
38
39 /**
40 * Internationalisation code
41 * @ingroup Language
42 */
43 class Language {
44 /**
45 * @var LanguageConverter
46 */
47 public $mConverter;
48
49 public $mVariants, $mCode, $mLoaded = false;
50 public $mMagicExtensions = [], $mMagicHookDone = false;
51 private $mHtmlCode = null, $mParentLanguage = false;
52
53 public $dateFormatStrings = [];
54 public $mExtendedSpecialPageAliases;
55
56 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
57
58 /**
59 * ReplacementArray object caches
60 */
61 public $transformData = [];
62
63 /**
64 * @var LocalisationCache
65 */
66 static public $dataCache;
67
68 static public $mLangObjCache = [];
69
70 static public $mWeekdayMsgs = [
71 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
72 'friday', 'saturday'
73 ];
74
75 static public $mWeekdayAbbrevMsgs = [
76 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
77 ];
78
79 static public $mMonthMsgs = [
80 'january', 'february', 'march', 'april', 'may_long', 'june',
81 'july', 'august', 'september', 'october', 'november',
82 'december'
83 ];
84 static public $mMonthGenMsgs = [
85 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
86 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
87 'december-gen'
88 ];
89 static public $mMonthAbbrevMsgs = [
90 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
91 'sep', 'oct', 'nov', 'dec'
92 ];
93
94 static public $mIranianCalendarMonthMsgs = [
95 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
96 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
97 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
98 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
99 ];
100
101 static public $mHebrewCalendarMonthMsgs = [
102 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
103 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
104 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
105 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
106 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
107 ];
108
109 static public $mHebrewCalendarMonthGenMsgs = [
110 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
111 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
112 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
113 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
114 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
115 ];
116
117 static public $mHijriCalendarMonthMsgs = [
118 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
119 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
120 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
121 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
122 ];
123
124 /**
125 * @since 1.20
126 * @var array
127 */
128 static public $durationIntervals = [
129 'millennia' => 31556952000,
130 'centuries' => 3155695200,
131 'decades' => 315569520,
132 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
133 'weeks' => 604800,
134 'days' => 86400,
135 'hours' => 3600,
136 'minutes' => 60,
137 'seconds' => 1,
138 ];
139
140 /**
141 * Cache for language fallbacks.
142 * @see Language::getFallbacksIncludingSiteLanguage
143 * @since 1.21
144 * @var array
145 */
146 static private $fallbackLanguageCache = [];
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 * @throws MWException
207 * @return Language
208 */
209 protected static function newFromCode( $code ) {
210 if ( !Language::isValidCode( $code ) ) {
211 throw new MWException( "Invalid language code \"$code\"" );
212 }
213
214 if ( !Language::isValidBuiltInCode( $code ) ) {
215 // It's not possible to customise this code with class files, so
216 // just return a Language object. This is to support uselang= hacks.
217 $lang = new Language;
218 $lang->setCode( $code );
219 return $lang;
220 }
221
222 // Check if there is a language class for the code
223 $class = self::classFromCode( $code );
224 if ( class_exists( $class ) ) {
225 $lang = new $class;
226 return $lang;
227 }
228
229 // Keep trying the fallback list until we find an existing class
230 $fallbacks = Language::getFallbacksFor( $code );
231 foreach ( $fallbacks as $fallbackCode ) {
232 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
233 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
234 }
235
236 $class = self::classFromCode( $fallbackCode );
237 if ( class_exists( $class ) ) {
238 $lang = new $class;
239 $lang->setCode( $code );
240 return $lang;
241 }
242 }
243
244 throw new MWException( "Invalid fallback sequence for language '$code'" );
245 }
246
247 /**
248 * Checks whether any localisation is available for that language tag
249 * in MediaWiki (MessagesXx.php exists).
250 *
251 * @param string $code Language tag (in lower case)
252 * @return bool Whether language is supported
253 * @since 1.21
254 */
255 public static function isSupportedLanguage( $code ) {
256 if ( !self::isValidBuiltInCode( $code ) ) {
257 return false;
258 }
259
260 if ( $code === 'qqq' ) {
261 return false;
262 }
263
264 return is_readable( self::getMessagesFileName( $code ) ) ||
265 is_readable( self::getJsonMessagesFileName( $code ) );
266 }
267
268 /**
269 * Returns true if a language code string is a well-formed language tag
270 * according to RFC 5646.
271 * This function only checks well-formedness; it doesn't check that
272 * language, script or variant codes actually exist in the repositories.
273 *
274 * Based on regexes by Mark Davis of the Unicode Consortium:
275 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
276 *
277 * @param string $code
278 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
279 *
280 * @return bool
281 * @since 1.21
282 */
283 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
284 $alpha = '[a-z]';
285 $digit = '[0-9]';
286 $alphanum = '[a-z0-9]';
287 $x = 'x'; # private use singleton
288 $singleton = '[a-wy-z]'; # other singleton
289 $s = $lenient ? '[-_]' : '-';
290
291 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
292 $script = "$alpha{4}"; # ISO 15924
293 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
294 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
295 $extension = "$singleton(?:$s$alphanum{2,8})+";
296 $privateUse = "$x(?:$s$alphanum{1,8})+";
297
298 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
299 # Since these are limited, this is safe even later changes to the registry --
300 # the only oddity is that it might change the type of the tag, and thus
301 # the results from the capturing groups.
302 # http://www.iana.org/assignments/language-subtag-registry
303
304 $grandfathered = "en{$s}GB{$s}oed"
305 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
306 . "|no{$s}(?:bok|nyn)"
307 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
308 . "|zh{$s}min{$s}nan";
309
310 $variantList = "$variant(?:$s$variant)*";
311 $extensionList = "$extension(?:$s$extension)*";
312
313 $langtag = "(?:($language)"
314 . "(?:$s$script)?"
315 . "(?:$s$region)?"
316 . "(?:$s$variantList)?"
317 . "(?:$s$extensionList)?"
318 . "(?:$s$privateUse)?)";
319
320 # The final breakdown, with capturing groups for each of these components
321 # The variants, extensions, grandfathered, and private-use may have interior '-'
322
323 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
324
325 return (bool)preg_match( "/$root/", strtolower( $code ) );
326 }
327
328 /**
329 * Returns true if a language code string is of a valid form, whether or
330 * not it exists. This includes codes which are used solely for
331 * customisation via the MediaWiki namespace.
332 *
333 * @param string $code
334 *
335 * @return bool
336 */
337 public static function isValidCode( $code ) {
338 static $cache = [];
339 if ( !isset( $cache[$code] ) ) {
340 // People think language codes are html safe, so enforce it.
341 // Ideally we should only allow a-zA-Z0-9-
342 // but, .+ and other chars are often used for {{int:}} hacks
343 // see bugs 37564, 37587, 36938
344 $cache[$code] =
345 // Protect against path traversal
346 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
347 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
348 }
349 return $cache[$code];
350 }
351
352 /**
353 * Returns true if a language code is of a valid form for the purposes of
354 * internal customisation of MediaWiki, via Messages*.php or *.json.
355 *
356 * @param string $code
357 *
358 * @throws MWException
359 * @since 1.18
360 * @return bool
361 */
362 public static function isValidBuiltInCode( $code ) {
363
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 ( get_class( $this ) == 'Language' ) {
419 $this->mCode = 'en';
420 } else {
421 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 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 function getFallbackLanguages() {
447 return self::getFallbacksFor( $this->mCode );
448 }
449
450 /**
451 * Exports $wgBookstoreListEn
452 * @return array
453 */
454 function getBookstoreList() {
455 return self::$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 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
469 $validNamespaces = MWNamespace::getCanonicalNamespaces();
470
471 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
472
473 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
474 if ( $wgMetaNamespaceTalk ) {
475 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
476 } else {
477 $talk = $this->namespaceNames[NS_PROJECT_TALK];
478 $this->namespaceNames[NS_PROJECT_TALK] =
479 $this->fixVariableInNamespace( $talk );
480 }
481
482 # Sometimes a language will be localised but not actually exist on this wiki.
483 foreach ( $this->namespaceNames as $key => $text ) {
484 if ( !isset( $validNamespaces[$key] ) ) {
485 unset( $this->namespaceNames[$key] );
486 }
487 }
488
489 # The above mixing may leave namespaces out of canonical order.
490 # Re-order by namespace ID number...
491 ksort( $this->namespaceNames );
492
493 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
494 }
495
496 return $this->namespaceNames;
497 }
498
499 /**
500 * Arbitrarily set all of the namespace names at once. Mainly used for testing
501 * @param array $namespaces Array of namespaces (id => name)
502 */
503 public function setNamespaces( array $namespaces ) {
504 $this->namespaceNames = $namespaces;
505 $this->mNamespaceIds = null;
506 }
507
508 /**
509 * Resets all of the namespace caches. Mainly used for testing
510 */
511 public function resetNamespaces() {
512 $this->namespaceNames = null;
513 $this->mNamespaceIds = null;
514 $this->namespaceAliases = null;
515 }
516
517 /**
518 * A convenience function that returns getNamespaces() with spaces instead of underscores
519 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
520 *
521 * @return array
522 */
523 function getFormattedNamespaces() {
524 $ns = $this->getNamespaces();
525 foreach ( $ns as $k => $v ) {
526 $ns[$k] = strtr( $v, '_', ' ' );
527 }
528 return $ns;
529 }
530
531 /**
532 * Get a namespace value by key
533 *
534 * <code>
535 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
536 * echo $mw_ns; // prints 'MediaWiki'
537 * </code>
538 *
539 * @param int $index The array key of the namespace to return
540 * @return string|bool String if the namespace value exists, otherwise false
541 */
542 function getNsText( $index ) {
543 $ns = $this->getNamespaces();
544 return isset( $ns[$index] ) ? $ns[$index] : false;
545 }
546
547 /**
548 * A convenience function that returns the same thing as
549 * getNsText() except with '_' changed to ' ', useful for
550 * producing output.
551 *
552 * <code>
553 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
554 * echo $mw_ns; // prints 'MediaWiki talk'
555 * </code>
556 *
557 * @param int $index The array key of the namespace to return
558 * @return string Namespace name without underscores (empty string if namespace does not exist)
559 */
560 function getFormattedNsText( $index ) {
561 $ns = $this->getNsText( $index );
562 return strtr( $ns, '_', ' ' );
563 }
564
565 /**
566 * Returns gender-dependent namespace alias if available.
567 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
568 * @param int $index Namespace index
569 * @param string $gender Gender key (male, female... )
570 * @return string
571 * @since 1.18
572 */
573 function getGenderNsText( $index, $gender ) {
574 global $wgExtraGenderNamespaces;
575
576 $ns = $wgExtraGenderNamespaces +
577 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
578
579 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
580 }
581
582 /**
583 * Whether this language uses gender-dependent namespace aliases.
584 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
585 * @return bool
586 * @since 1.18
587 */
588 function needsGenderDistinction() {
589 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
590 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
591 // $wgExtraGenderNamespaces overrides everything
592 return true;
593 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
594 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
595 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
596 return false;
597 } else {
598 // Check what is in i18n files
599 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
600 return count( $aliases ) > 0;
601 }
602 }
603
604 /**
605 * Get a namespace key by value, case insensitive.
606 * Only matches namespace names for the current language, not the
607 * canonical ones defined in Namespace.php.
608 *
609 * @param string $text
610 * @return int|bool An integer if $text is a valid value otherwise false
611 */
612 function getLocalNsIndex( $text ) {
613 $lctext = $this->lc( $text );
614 $ids = $this->getNamespaceIds();
615 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
616 }
617
618 /**
619 * @return array
620 */
621 function getNamespaceAliases() {
622 if ( is_null( $this->namespaceAliases ) ) {
623 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
624 if ( !$aliases ) {
625 $aliases = [];
626 } else {
627 foreach ( $aliases as $name => $index ) {
628 if ( $index === NS_PROJECT_TALK ) {
629 unset( $aliases[$name] );
630 $name = $this->fixVariableInNamespace( $name );
631 $aliases[$name] = $index;
632 }
633 }
634 }
635
636 global $wgExtraGenderNamespaces;
637 $genders = $wgExtraGenderNamespaces +
638 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
639 foreach ( $genders as $index => $forms ) {
640 foreach ( $forms as $alias ) {
641 $aliases[$alias] = $index;
642 }
643 }
644
645 # Also add converted namespace names as aliases, to avoid confusion.
646 $convertedNames = [];
647 foreach ( $this->getVariants() as $variant ) {
648 if ( $variant === $this->mCode ) {
649 continue;
650 }
651 foreach ( $this->getNamespaces() as $ns => $_ ) {
652 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
653 }
654 }
655
656 $this->namespaceAliases = $aliases + $convertedNames;
657 }
658
659 return $this->namespaceAliases;
660 }
661
662 /**
663 * @return array
664 */
665 function getNamespaceIds() {
666 if ( is_null( $this->mNamespaceIds ) ) {
667 global $wgNamespaceAliases;
668 # Put namespace names and aliases into a hashtable.
669 # If this is too slow, then we should arrange it so that it is done
670 # before caching. The catch is that at pre-cache time, the above
671 # class-specific fixup hasn't been done.
672 $this->mNamespaceIds = [];
673 foreach ( $this->getNamespaces() as $index => $name ) {
674 $this->mNamespaceIds[$this->lc( $name )] = $index;
675 }
676 foreach ( $this->getNamespaceAliases() as $name => $index ) {
677 $this->mNamespaceIds[$this->lc( $name )] = $index;
678 }
679 if ( $wgNamespaceAliases ) {
680 foreach ( $wgNamespaceAliases as $name => $index ) {
681 $this->mNamespaceIds[$this->lc( $name )] = $index;
682 }
683 }
684 }
685 return $this->mNamespaceIds;
686 }
687
688 /**
689 * Get a namespace key by value, case insensitive. Canonical namespace
690 * names override custom ones defined for the current language.
691 *
692 * @param string $text
693 * @return int|bool An integer if $text is a valid value otherwise false
694 */
695 function getNsIndex( $text ) {
696 $lctext = $this->lc( $text );
697 $ns = MWNamespace::getCanonicalIndex( $lctext );
698 if ( $ns !== null ) {
699 return $ns;
700 }
701 $ids = $this->getNamespaceIds();
702 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
703 }
704
705 /**
706 * short names for language variants used for language conversion links.
707 *
708 * @param string $code
709 * @param bool $usemsg Use the "variantname-xyz" message if it exists
710 * @return string
711 */
712 function getVariantname( $code, $usemsg = true ) {
713 $msg = "variantname-$code";
714 if ( $usemsg && wfMessage( $msg )->exists() ) {
715 return $this->getMessageFromDB( $msg );
716 }
717 $name = self::fetchLanguageName( $code );
718 if ( $name ) {
719 return $name; # if it's defined as a language name, show that
720 } else {
721 # otherwise, output the language code
722 return $code;
723 }
724 }
725
726 /**
727 * @return array
728 */
729 function getDatePreferences() {
730 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
731 }
732
733 /**
734 * @return array
735 */
736 function getDateFormats() {
737 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
738 }
739
740 /**
741 * @return array|string
742 */
743 function getDefaultDateFormat() {
744 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
745 if ( $df === 'dmy or mdy' ) {
746 global $wgAmericanDates;
747 return $wgAmericanDates ? 'mdy' : 'dmy';
748 } else {
749 return $df;
750 }
751 }
752
753 /**
754 * @return array
755 */
756 function getDatePreferenceMigrationMap() {
757 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
758 }
759
760 /**
761 * @param string $image
762 * @return array|null
763 */
764 function getImageFile( $image ) {
765 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
766 }
767
768 /**
769 * @return array
770 * @since 1.24
771 */
772 function getImageFiles() {
773 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
774 }
775
776 /**
777 * @return array
778 */
779 function getExtraUserToggles() {
780 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
781 }
782
783 /**
784 * @param string $tog
785 * @return string
786 */
787 function getUserToggle( $tog ) {
788 return $this->getMessageFromDB( "tog-$tog" );
789 }
790
791 /**
792 * Get an array of language names, indexed by code.
793 * @param null|string $inLanguage Code of language in which to return the names
794 * Use null for autonyms (native names)
795 * @param string $include One of:
796 * 'all' all available languages
797 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
798 * 'mwfile' only if the language is in 'mw' *and* has a message file
799 * @return array Language code => language name
800 * @since 1.20
801 */
802 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
803 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
804 $cacheKey .= ":$include";
805 if ( self::$languageNameCache === null ) {
806 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
807 }
808
809 $ret = self::$languageNameCache->get( $cacheKey );
810 if ( !$ret ) {
811 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
812 self::$languageNameCache->set( $cacheKey, $ret );
813 }
814 return $ret;
815 }
816
817 /**
818 * Uncached helper for fetchLanguageNames
819 * @param null|string $inLanguage Code of language in which to return the names
820 * Use null for autonyms (native names)
821 * @param string $include One of:
822 * 'all' all available languages
823 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
824 * 'mwfile' only if the language is in 'mw' *and* has a message file
825 * @return array Language code => language name
826 */
827 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
828 global $wgExtraLanguageNames;
829
830 // If passed an invalid language code to use, fallback to en
831 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
832 $inLanguage = 'en';
833 }
834
835 $names = [];
836
837 if ( $inLanguage ) {
838 # TODO: also include when $inLanguage is null, when this code is more efficient
839 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
840 }
841
842 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
843 foreach ( $mwNames as $mwCode => $mwName ) {
844 # - Prefer own MediaWiki native name when not using the hook
845 # - For other names just add if not added through the hook
846 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
847 $names[$mwCode] = $mwName;
848 }
849 }
850
851 if ( $include === 'all' ) {
852 ksort( $names );
853 return $names;
854 }
855
856 $returnMw = [];
857 $coreCodes = array_keys( $mwNames );
858 foreach ( $coreCodes as $coreCode ) {
859 $returnMw[$coreCode] = $names[$coreCode];
860 }
861
862 if ( $include === 'mwfile' ) {
863 $namesMwFile = [];
864 # We do this using a foreach over the codes instead of a directory
865 # loop so that messages files in extensions will work correctly.
866 foreach ( $returnMw as $code => $value ) {
867 if ( is_readable( self::getMessagesFileName( $code ) )
868 || is_readable( self::getJsonMessagesFileName( $code ) )
869 ) {
870 $namesMwFile[$code] = $names[$code];
871 }
872 }
873
874 ksort( $namesMwFile );
875 return $namesMwFile;
876 }
877
878 ksort( $returnMw );
879 # 'mw' option; default if it's not one of the other two options (all/mwfile)
880 return $returnMw;
881 }
882
883 /**
884 * @param string $code The code of the language for which to get the name
885 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
886 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
887 * @return string Language name or empty
888 * @since 1.20
889 */
890 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
891 $code = strtolower( $code );
892 $array = self::fetchLanguageNames( $inLanguage, $include );
893 return !array_key_exists( $code, $array ) ? '' : $array[$code];
894 }
895
896 /**
897 * Get a message from the MediaWiki namespace.
898 *
899 * @param string $msg Message name
900 * @return string
901 */
902 function getMessageFromDB( $msg ) {
903 return $this->msg( $msg )->text();
904 }
905
906 /**
907 * Get message object in this language. Only for use inside this class.
908 *
909 * @param string $msg Message name
910 * @return Message
911 */
912 protected function msg( $msg ) {
913 return wfMessage( $msg )->inLanguage( $this );
914 }
915
916 /**
917 * @param string $key
918 * @return string
919 */
920 function getMonthName( $key ) {
921 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
922 }
923
924 /**
925 * @return array
926 */
927 function getMonthNamesArray() {
928 $monthNames = [ '' ];
929 for ( $i = 1; $i < 13; $i++ ) {
930 $monthNames[] = $this->getMonthName( $i );
931 }
932 return $monthNames;
933 }
934
935 /**
936 * @param string $key
937 * @return string
938 */
939 function getMonthNameGen( $key ) {
940 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
941 }
942
943 /**
944 * @param string $key
945 * @return string
946 */
947 function getMonthAbbreviation( $key ) {
948 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
949 }
950
951 /**
952 * @return array
953 */
954 function getMonthAbbreviationsArray() {
955 $monthNames = [ '' ];
956 for ( $i = 1; $i < 13; $i++ ) {
957 $monthNames[] = $this->getMonthAbbreviation( $i );
958 }
959 return $monthNames;
960 }
961
962 /**
963 * @param string $key
964 * @return string
965 */
966 function getWeekdayName( $key ) {
967 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
968 }
969
970 /**
971 * @param string $key
972 * @return string
973 */
974 function getWeekdayAbbreviation( $key ) {
975 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
976 }
977
978 /**
979 * @param string $key
980 * @return string
981 */
982 function getIranianCalendarMonthName( $key ) {
983 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
984 }
985
986 /**
987 * @param string $key
988 * @return string
989 */
990 function getHebrewCalendarMonthName( $key ) {
991 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
992 }
993
994 /**
995 * @param string $key
996 * @return string
997 */
998 function getHebrewCalendarMonthNameGen( $key ) {
999 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1000 }
1001
1002 /**
1003 * @param string $key
1004 * @return string
1005 */
1006 function getHijriCalendarMonthName( $key ) {
1007 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1008 }
1009
1010 /**
1011 * Pass through result from $dateTimeObj->format()
1012 * @param DateTime|bool|null &$dateTimeObj
1013 * @param string $ts
1014 * @param DateTimeZone|bool|null $zone
1015 * @param string $code
1016 * @return string
1017 */
1018 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1019 if ( !$dateTimeObj ) {
1020 $dateTimeObj = DateTime::createFromFormat(
1021 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1022 );
1023 }
1024 return $dateTimeObj->format( $code );
1025 }
1026
1027 /**
1028 * This is a workalike of PHP's date() function, but with better
1029 * internationalisation, a reduced set of format characters, and a better
1030 * escaping format.
1031 *
1032 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1033 * the PHP manual for definitions. There are a number of extensions, which
1034 * start with "x":
1035 *
1036 * xn Do not translate digits of the next numeric format character
1037 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1038 * xr Use roman numerals for the next numeric format character
1039 * xh Use hebrew numerals for the next numeric format character
1040 * xx Literal x
1041 * xg Genitive month name
1042 *
1043 * xij j (day number) in Iranian calendar
1044 * xiF F (month name) in Iranian calendar
1045 * xin n (month number) in Iranian calendar
1046 * xiy y (two digit year) in Iranian calendar
1047 * xiY Y (full year) in Iranian calendar
1048 *
1049 * xjj j (day number) in Hebrew calendar
1050 * xjF F (month name) in Hebrew calendar
1051 * xjt t (days in month) in Hebrew calendar
1052 * xjx xg (genitive month name) in Hebrew calendar
1053 * xjn n (month number) in Hebrew calendar
1054 * xjY Y (full year) in Hebrew calendar
1055 *
1056 * xmj j (day number) in Hijri calendar
1057 * xmF F (month name) in Hijri calendar
1058 * xmn n (month number) in Hijri calendar
1059 * xmY Y (full year) in Hijri calendar
1060 *
1061 * xkY Y (full year) in Thai solar calendar. Months and days are
1062 * identical to the Gregorian calendar
1063 * xoY Y (full year) in Minguo calendar or Juche year.
1064 * Months and days are identical to the
1065 * Gregorian calendar
1066 * xtY Y (full year) in Japanese nengo. Months and days are
1067 * identical to the Gregorian calendar
1068 *
1069 * Characters enclosed in double quotes will be considered literal (with
1070 * the quotes themselves removed). Unmatched quotes will be considered
1071 * literal quotes. Example:
1072 *
1073 * "The month is" F => The month is January
1074 * i's" => 20'11"
1075 *
1076 * Backslash escaping is also supported.
1077 *
1078 * Input timestamp is assumed to be pre-normalized to the desired local
1079 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1080 * $ts is UTC if $zone is not given.
1081 *
1082 * @param string $format
1083 * @param string $ts 14-character timestamp
1084 * YYYYMMDDHHMMSS
1085 * 01234567890123
1086 * @param DateTimeZone $zone Timezone of $ts
1087 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1088 * Only makes sense if $ts is the current time.
1089 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1090 *
1091 * @throws MWException
1092 * @return string
1093 */
1094 function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = null ) {
1095 $s = '';
1096 $raw = false;
1097 $roman = false;
1098 $hebrewNum = false;
1099 $dateTimeObj = false;
1100 $rawToggle = false;
1101 $iranian = false;
1102 $hebrew = false;
1103 $hijri = false;
1104 $thai = false;
1105 $minguo = false;
1106 $tenno = false;
1107
1108 $usedSecond = false;
1109 $usedMinute = false;
1110 $usedHour = false;
1111 $usedAMPM = false;
1112 $usedDay = false;
1113 $usedWeek = false;
1114 $usedMonth = false;
1115 $usedYear = false;
1116 $usedISOYear = false;
1117 $usedIsLeapYear = false;
1118
1119 $usedHebrewMonth = false;
1120 $usedIranianMonth = false;
1121 $usedHijriMonth = false;
1122 $usedHebrewYear = false;
1123 $usedIranianYear = false;
1124 $usedHijriYear = false;
1125 $usedTennoYear = false;
1126
1127 if ( strlen( $ts ) !== 14 ) {
1128 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1129 }
1130
1131 if ( !ctype_digit( $ts ) ) {
1132 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1133 }
1134
1135 $formatLength = strlen( $format );
1136 for ( $p = 0; $p < $formatLength; $p++ ) {
1137 $num = false;
1138 $code = $format[$p];
1139 if ( $code == 'x' && $p < $formatLength - 1 ) {
1140 $code .= $format[++$p];
1141 }
1142
1143 if ( ( $code === 'xi'
1144 || $code === 'xj'
1145 || $code === 'xk'
1146 || $code === 'xm'
1147 || $code === 'xo'
1148 || $code === 'xt' )
1149 && $p < $formatLength - 1 ) {
1150 $code .= $format[++$p];
1151 }
1152
1153 switch ( $code ) {
1154 case 'xx':
1155 $s .= 'x';
1156 break;
1157 case 'xn':
1158 $raw = true;
1159 break;
1160 case 'xN':
1161 $rawToggle = !$rawToggle;
1162 break;
1163 case 'xr':
1164 $roman = true;
1165 break;
1166 case 'xh':
1167 $hebrewNum = true;
1168 break;
1169 case 'xg':
1170 $usedMonth = true;
1171 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1172 break;
1173 case 'xjx':
1174 $usedHebrewMonth = true;
1175 if ( !$hebrew ) {
1176 $hebrew = self::tsToHebrew( $ts );
1177 }
1178 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1179 break;
1180 case 'd':
1181 $usedDay = true;
1182 $num = substr( $ts, 6, 2 );
1183 break;
1184 case 'D':
1185 $usedDay = true;
1186 $s .= $this->getWeekdayAbbreviation(
1187 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1188 );
1189 break;
1190 case 'j':
1191 $usedDay = true;
1192 $num = intval( substr( $ts, 6, 2 ) );
1193 break;
1194 case 'xij':
1195 $usedDay = true;
1196 if ( !$iranian ) {
1197 $iranian = self::tsToIranian( $ts );
1198 }
1199 $num = $iranian[2];
1200 break;
1201 case 'xmj':
1202 $usedDay = true;
1203 if ( !$hijri ) {
1204 $hijri = self::tsToHijri( $ts );
1205 }
1206 $num = $hijri[2];
1207 break;
1208 case 'xjj':
1209 $usedDay = true;
1210 if ( !$hebrew ) {
1211 $hebrew = self::tsToHebrew( $ts );
1212 }
1213 $num = $hebrew[2];
1214 break;
1215 case 'l':
1216 $usedDay = true;
1217 $s .= $this->getWeekdayName(
1218 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1219 );
1220 break;
1221 case 'F':
1222 $usedMonth = true;
1223 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1224 break;
1225 case 'xiF':
1226 $usedIranianMonth = true;
1227 if ( !$iranian ) {
1228 $iranian = self::tsToIranian( $ts );
1229 }
1230 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1231 break;
1232 case 'xmF':
1233 $usedHijriMonth = true;
1234 if ( !$hijri ) {
1235 $hijri = self::tsToHijri( $ts );
1236 }
1237 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1238 break;
1239 case 'xjF':
1240 $usedHebrewMonth = true;
1241 if ( !$hebrew ) {
1242 $hebrew = self::tsToHebrew( $ts );
1243 }
1244 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1245 break;
1246 case 'm':
1247 $usedMonth = true;
1248 $num = substr( $ts, 4, 2 );
1249 break;
1250 case 'M':
1251 $usedMonth = true;
1252 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1253 break;
1254 case 'n':
1255 $usedMonth = true;
1256 $num = intval( substr( $ts, 4, 2 ) );
1257 break;
1258 case 'xin':
1259 $usedIranianMonth = true;
1260 if ( !$iranian ) {
1261 $iranian = self::tsToIranian( $ts );
1262 }
1263 $num = $iranian[1];
1264 break;
1265 case 'xmn':
1266 $usedHijriMonth = true;
1267 if ( !$hijri ) {
1268 $hijri = self::tsToHijri( $ts );
1269 }
1270 $num = $hijri[1];
1271 break;
1272 case 'xjn':
1273 $usedHebrewMonth = true;
1274 if ( !$hebrew ) {
1275 $hebrew = self::tsToHebrew( $ts );
1276 }
1277 $num = $hebrew[1];
1278 break;
1279 case 'xjt':
1280 $usedHebrewMonth = true;
1281 if ( !$hebrew ) {
1282 $hebrew = self::tsToHebrew( $ts );
1283 }
1284 $num = $hebrew[3];
1285 break;
1286 case 'Y':
1287 $usedYear = true;
1288 $num = substr( $ts, 0, 4 );
1289 break;
1290 case 'xiY':
1291 $usedIranianYear = true;
1292 if ( !$iranian ) {
1293 $iranian = self::tsToIranian( $ts );
1294 }
1295 $num = $iranian[0];
1296 break;
1297 case 'xmY':
1298 $usedHijriYear = true;
1299 if ( !$hijri ) {
1300 $hijri = self::tsToHijri( $ts );
1301 }
1302 $num = $hijri[0];
1303 break;
1304 case 'xjY':
1305 $usedHebrewYear = true;
1306 if ( !$hebrew ) {
1307 $hebrew = self::tsToHebrew( $ts );
1308 }
1309 $num = $hebrew[0];
1310 break;
1311 case 'xkY':
1312 $usedYear = true;
1313 if ( !$thai ) {
1314 $thai = self::tsToYear( $ts, 'thai' );
1315 }
1316 $num = $thai[0];
1317 break;
1318 case 'xoY':
1319 $usedYear = true;
1320 if ( !$minguo ) {
1321 $minguo = self::tsToYear( $ts, 'minguo' );
1322 }
1323 $num = $minguo[0];
1324 break;
1325 case 'xtY':
1326 $usedTennoYear = true;
1327 if ( !$tenno ) {
1328 $tenno = self::tsToYear( $ts, 'tenno' );
1329 }
1330 $num = $tenno[0];
1331 break;
1332 case 'y':
1333 $usedYear = true;
1334 $num = substr( $ts, 2, 2 );
1335 break;
1336 case 'xiy':
1337 $usedIranianYear = true;
1338 if ( !$iranian ) {
1339 $iranian = self::tsToIranian( $ts );
1340 }
1341 $num = substr( $iranian[0], -2 );
1342 break;
1343 case 'a':
1344 $usedAMPM = true;
1345 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1346 break;
1347 case 'A':
1348 $usedAMPM = true;
1349 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1350 break;
1351 case 'g':
1352 $usedHour = true;
1353 $h = substr( $ts, 8, 2 );
1354 $num = $h % 12 ? $h % 12 : 12;
1355 break;
1356 case 'G':
1357 $usedHour = true;
1358 $num = intval( substr( $ts, 8, 2 ) );
1359 break;
1360 case 'h':
1361 $usedHour = true;
1362 $h = substr( $ts, 8, 2 );
1363 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1364 break;
1365 case 'H':
1366 $usedHour = true;
1367 $num = substr( $ts, 8, 2 );
1368 break;
1369 case 'i':
1370 $usedMinute = true;
1371 $num = substr( $ts, 10, 2 );
1372 break;
1373 case 's':
1374 $usedSecond = true;
1375 $num = substr( $ts, 12, 2 );
1376 break;
1377 case 'c':
1378 case 'r':
1379 $usedSecond = true;
1380 // fall through
1381 case 'e':
1382 case 'O':
1383 case 'P':
1384 case 'T':
1385 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1386 break;
1387 case 'w':
1388 case 'N':
1389 case 'z':
1390 $usedDay = true;
1391 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1392 break;
1393 case 'W':
1394 $usedWeek = true;
1395 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1396 break;
1397 case 't':
1398 $usedMonth = true;
1399 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1400 break;
1401 case 'L':
1402 $usedIsLeapYear = true;
1403 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1404 break;
1405 case 'o':
1406 $usedISOYear = true;
1407 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1408 break;
1409 case 'U':
1410 $usedSecond = true;
1411 // fall through
1412 case 'I':
1413 case 'Z':
1414 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1415 break;
1416 case '\\':
1417 # Backslash escaping
1418 if ( $p < $formatLength - 1 ) {
1419 $s .= $format[++$p];
1420 } else {
1421 $s .= '\\';
1422 }
1423 break;
1424 case '"':
1425 # Quoted literal
1426 if ( $p < $formatLength - 1 ) {
1427 $endQuote = strpos( $format, '"', $p + 1 );
1428 if ( $endQuote === false ) {
1429 # No terminating quote, assume literal "
1430 $s .= '"';
1431 } else {
1432 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1433 $p = $endQuote;
1434 }
1435 } else {
1436 # Quote at end of string, assume literal "
1437 $s .= '"';
1438 }
1439 break;
1440 default:
1441 $s .= $format[$p];
1442 }
1443 if ( $num !== false ) {
1444 if ( $rawToggle || $raw ) {
1445 $s .= $num;
1446 $raw = false;
1447 } elseif ( $roman ) {
1448 $s .= Language::romanNumeral( $num );
1449 $roman = false;
1450 } elseif ( $hebrewNum ) {
1451 $s .= self::hebrewNumeral( $num );
1452 $hebrewNum = false;
1453 } else {
1454 $s .= $this->formatNum( $num, true );
1455 }
1456 }
1457 }
1458
1459 if ( $usedSecond ) {
1460 $ttl = 1;
1461 } elseif ( $usedMinute ) {
1462 $ttl = 60 - substr( $ts, 12, 2 );
1463 } elseif ( $usedHour ) {
1464 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1465 } elseif ( $usedAMPM ) {
1466 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1467 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1468 } elseif (
1469 $usedDay ||
1470 $usedHebrewMonth ||
1471 $usedIranianMonth ||
1472 $usedHijriMonth ||
1473 $usedHebrewYear ||
1474 $usedIranianYear ||
1475 $usedHijriYear ||
1476 $usedTennoYear
1477 ) {
1478 // @todo Someone who understands the non-Gregorian calendars
1479 // should write proper logic for them so that they don't need purged every day.
1480 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1481 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1482 } else {
1483 $possibleTtls = [];
1484 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1485 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1486 if ( $usedWeek ) {
1487 $possibleTtls[] =
1488 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1489 $timeRemainingInDay;
1490 } elseif ( $usedISOYear ) {
1491 // December 28th falls on the last ISO week of the year, every year.
1492 // The last ISO week of a year can be 52 or 53.
1493 $lastWeekOfISOYear = DateTime::createFromFormat(
1494 'Ymd',
1495 substr( $ts, 0, 4 ) . '1228',
1496 $zone ?: new DateTimeZone( 'UTC' )
1497 )->format( 'W' );
1498 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1499 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1500 $timeRemainingInWeek =
1501 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1502 + $timeRemainingInDay;
1503 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1504 }
1505
1506 if ( $usedMonth ) {
1507 $possibleTtls[] =
1508 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1509 substr( $ts, 6, 2 ) ) * 86400
1510 + $timeRemainingInDay;
1511 } elseif ( $usedYear ) {
1512 $possibleTtls[] =
1513 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1514 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1515 + $timeRemainingInDay;
1516 } elseif ( $usedIsLeapYear ) {
1517 $year = substr( $ts, 0, 4 );
1518 $timeRemainingInYear =
1519 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1520 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1521 + $timeRemainingInDay;
1522 $mod = $year % 4;
1523 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1524 // this isn't a leap year. see when the next one starts
1525 $nextCandidate = $year - $mod + 4;
1526 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1527 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1528 $timeRemainingInYear;
1529 } else {
1530 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1531 $timeRemainingInYear;
1532 }
1533 } else {
1534 // this is a leap year, so the next year isn't
1535 $possibleTtls[] = $timeRemainingInYear;
1536 }
1537 }
1538
1539 if ( $possibleTtls ) {
1540 $ttl = min( $possibleTtls );
1541 }
1542 }
1543
1544 return $s;
1545 }
1546
1547 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1548 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1549
1550 /**
1551 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1552 * Gregorian dates to Iranian dates. Originally written in C, it
1553 * is released under the terms of GNU Lesser General Public
1554 * License. Conversion to PHP was performed by Niklas Laxström.
1555 *
1556 * Link: http://www.farsiweb.info/jalali/jalali.c
1557 *
1558 * @param string $ts
1559 *
1560 * @return string
1561 */
1562 private static function tsToIranian( $ts ) {
1563 $gy = substr( $ts, 0, 4 ) -1600;
1564 $gm = substr( $ts, 4, 2 ) -1;
1565 $gd = substr( $ts, 6, 2 ) -1;
1566
1567 # Days passed from the beginning (including leap years)
1568 $gDayNo = 365 * $gy
1569 + floor( ( $gy + 3 ) / 4 )
1570 - floor( ( $gy + 99 ) / 100 )
1571 + floor( ( $gy + 399 ) / 400 );
1572
1573 // Add days of the past months of this year
1574 for ( $i = 0; $i < $gm; $i++ ) {
1575 $gDayNo += self::$GREG_DAYS[$i];
1576 }
1577
1578 // Leap years
1579 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1580 $gDayNo++;
1581 }
1582
1583 // Days passed in current month
1584 $gDayNo += (int)$gd;
1585
1586 $jDayNo = $gDayNo - 79;
1587
1588 $jNp = floor( $jDayNo / 12053 );
1589 $jDayNo %= 12053;
1590
1591 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1592 $jDayNo %= 1461;
1593
1594 if ( $jDayNo >= 366 ) {
1595 $jy += floor( ( $jDayNo - 1 ) / 365 );
1596 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1597 }
1598
1599 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1600 $jDayNo -= self::$IRANIAN_DAYS[$i];
1601 }
1602
1603 $jm = $i + 1;
1604 $jd = $jDayNo + 1;
1605
1606 return [ $jy, $jm, $jd ];
1607 }
1608
1609 /**
1610 * Converting Gregorian dates to Hijri dates.
1611 *
1612 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1613 *
1614 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1615 *
1616 * @param string $ts
1617 *
1618 * @return string
1619 */
1620 private static function tsToHijri( $ts ) {
1621 $year = substr( $ts, 0, 4 );
1622 $month = substr( $ts, 4, 2 );
1623 $day = substr( $ts, 6, 2 );
1624
1625 $zyr = $year;
1626 $zd = $day;
1627 $zm = $month;
1628 $zy = $zyr;
1629
1630 if (
1631 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1632 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1633 ) {
1634 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1635 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1636 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1637 $zd - 32075;
1638 } else {
1639 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1640 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1641 }
1642
1643 $zl = $zjd -1948440 + 10632;
1644 $zn = (int)( ( $zl - 1 ) / 10631 );
1645 $zl = $zl - 10631 * $zn + 354;
1646 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1647 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1648 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1649 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1650 $zm = (int)( ( 24 * $zl ) / 709 );
1651 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1652 $zy = 30 * $zn + $zj - 30;
1653
1654 return [ $zy, $zm, $zd ];
1655 }
1656
1657 /**
1658 * Converting Gregorian dates to Hebrew dates.
1659 *
1660 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1661 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1662 * to translate the relevant functions into PHP and release them under
1663 * GNU GPL.
1664 *
1665 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1666 * and Adar II is 14. In a non-leap year, Adar is 6.
1667 *
1668 * @param string $ts
1669 *
1670 * @return string
1671 */
1672 private static function tsToHebrew( $ts ) {
1673 # Parse date
1674 $year = substr( $ts, 0, 4 );
1675 $month = substr( $ts, 4, 2 );
1676 $day = substr( $ts, 6, 2 );
1677
1678 # Calculate Hebrew year
1679 $hebrewYear = $year + 3760;
1680
1681 # Month number when September = 1, August = 12
1682 $month += 4;
1683 if ( $month > 12 ) {
1684 # Next year
1685 $month -= 12;
1686 $year++;
1687 $hebrewYear++;
1688 }
1689
1690 # Calculate day of year from 1 September
1691 $dayOfYear = $day;
1692 for ( $i = 1; $i < $month; $i++ ) {
1693 if ( $i == 6 ) {
1694 # February
1695 $dayOfYear += 28;
1696 # Check if the year is leap
1697 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1698 $dayOfYear++;
1699 }
1700 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1701 $dayOfYear += 30;
1702 } else {
1703 $dayOfYear += 31;
1704 }
1705 }
1706
1707 # Calculate the start of the Hebrew year
1708 $start = self::hebrewYearStart( $hebrewYear );
1709
1710 # Calculate next year's start
1711 if ( $dayOfYear <= $start ) {
1712 # Day is before the start of the year - it is the previous year
1713 # Next year's start
1714 $nextStart = $start;
1715 # Previous year
1716 $year--;
1717 $hebrewYear--;
1718 # Add days since previous year's 1 September
1719 $dayOfYear += 365;
1720 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1721 # Leap year
1722 $dayOfYear++;
1723 }
1724 # Start of the new (previous) year
1725 $start = self::hebrewYearStart( $hebrewYear );
1726 } else {
1727 # Next year's start
1728 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1729 }
1730
1731 # Calculate Hebrew day of year
1732 $hebrewDayOfYear = $dayOfYear - $start;
1733
1734 # Difference between year's days
1735 $diff = $nextStart - $start;
1736 # Add 12 (or 13 for leap years) days to ignore the difference between
1737 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1738 # difference is only about the year type
1739 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1740 $diff += 13;
1741 } else {
1742 $diff += 12;
1743 }
1744
1745 # Check the year pattern, and is leap year
1746 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1747 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1748 # and non-leap years
1749 $yearPattern = $diff % 30;
1750 # Check if leap year
1751 $isLeap = $diff >= 30;
1752
1753 # Calculate day in the month from number of day in the Hebrew year
1754 # Don't check Adar - if the day is not in Adar, we will stop before;
1755 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1756 $hebrewDay = $hebrewDayOfYear;
1757 $hebrewMonth = 1;
1758 $days = 0;
1759 while ( $hebrewMonth <= 12 ) {
1760 # Calculate days in this month
1761 if ( $isLeap && $hebrewMonth == 6 ) {
1762 # Adar in a leap year
1763 if ( $isLeap ) {
1764 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1765 $days = 30;
1766 if ( $hebrewDay <= $days ) {
1767 # Day in Adar I
1768 $hebrewMonth = 13;
1769 } else {
1770 # Subtract the days of Adar I
1771 $hebrewDay -= $days;
1772 # Try Adar II
1773 $days = 29;
1774 if ( $hebrewDay <= $days ) {
1775 # Day in Adar II
1776 $hebrewMonth = 14;
1777 }
1778 }
1779 }
1780 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1781 # Cheshvan in a complete year (otherwise as the rule below)
1782 $days = 30;
1783 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1784 # Kislev in an incomplete year (otherwise as the rule below)
1785 $days = 29;
1786 } else {
1787 # Odd months have 30 days, even have 29
1788 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1789 }
1790 if ( $hebrewDay <= $days ) {
1791 # In the current month
1792 break;
1793 } else {
1794 # Subtract the days of the current month
1795 $hebrewDay -= $days;
1796 # Try in the next month
1797 $hebrewMonth++;
1798 }
1799 }
1800
1801 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1802 }
1803
1804 /**
1805 * This calculates the Hebrew year start, as days since 1 September.
1806 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1807 * Used for Hebrew date.
1808 *
1809 * @param int $year
1810 *
1811 * @return string
1812 */
1813 private static function hebrewYearStart( $year ) {
1814 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1815 $b = intval( ( $year - 1 ) % 4 );
1816 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1817 if ( $m < 0 ) {
1818 $m--;
1819 }
1820 $Mar = intval( $m );
1821 if ( $m < 0 ) {
1822 $m++;
1823 }
1824 $m -= $Mar;
1825
1826 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1827 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1828 $Mar++;
1829 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1830 $Mar += 2;
1831 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1832 $Mar++;
1833 }
1834
1835 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1836 return $Mar;
1837 }
1838
1839 /**
1840 * Algorithm to convert Gregorian dates to Thai solar dates,
1841 * Minguo dates or Minguo dates.
1842 *
1843 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1844 * http://en.wikipedia.org/wiki/Minguo_calendar
1845 * http://en.wikipedia.org/wiki/Japanese_era_name
1846 *
1847 * @param string $ts 14-character timestamp
1848 * @param string $cName Calender name
1849 * @return array Converted year, month, day
1850 */
1851 private static function tsToYear( $ts, $cName ) {
1852 $gy = substr( $ts, 0, 4 );
1853 $gm = substr( $ts, 4, 2 );
1854 $gd = substr( $ts, 6, 2 );
1855
1856 if ( !strcmp( $cName, 'thai' ) ) {
1857 # Thai solar dates
1858 # Add 543 years to the Gregorian calendar
1859 # Months and days are identical
1860 $gy_offset = $gy + 543;
1861 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1862 # Minguo dates
1863 # Deduct 1911 years from the Gregorian calendar
1864 # Months and days are identical
1865 $gy_offset = $gy - 1911;
1866 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1867 # Nengō dates up to Meiji period
1868 # Deduct years from the Gregorian calendar
1869 # depending on the nengo periods
1870 # Months and days are identical
1871 if ( ( $gy < 1912 )
1872 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1873 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1874 ) {
1875 # Meiji period
1876 $gy_gannen = $gy - 1868 + 1;
1877 $gy_offset = $gy_gannen;
1878 if ( $gy_gannen == 1 ) {
1879 $gy_offset = '元';
1880 }
1881 $gy_offset = '明治' . $gy_offset;
1882 } elseif (
1883 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1884 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1885 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1886 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1887 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1888 ) {
1889 # Taishō period
1890 $gy_gannen = $gy - 1912 + 1;
1891 $gy_offset = $gy_gannen;
1892 if ( $gy_gannen == 1 ) {
1893 $gy_offset = '元';
1894 }
1895 $gy_offset = '大正' . $gy_offset;
1896 } elseif (
1897 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1898 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1899 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1900 ) {
1901 # Shōwa period
1902 $gy_gannen = $gy - 1926 + 1;
1903 $gy_offset = $gy_gannen;
1904 if ( $gy_gannen == 1 ) {
1905 $gy_offset = '元';
1906 }
1907 $gy_offset = '昭和' . $gy_offset;
1908 } else {
1909 # Heisei period
1910 $gy_gannen = $gy - 1989 + 1;
1911 $gy_offset = $gy_gannen;
1912 if ( $gy_gannen == 1 ) {
1913 $gy_offset = '元';
1914 }
1915 $gy_offset = '平成' . $gy_offset;
1916 }
1917 } else {
1918 $gy_offset = $gy;
1919 }
1920
1921 return [ $gy_offset, $gm, $gd ];
1922 }
1923
1924 /**
1925 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1926 *
1927 * This is the rule the BIDI algorithm uses to determine the directionality of
1928 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1929 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1930 *
1931 * TODO: Does not handle BIDI control characters inside the text.
1932 * TODO: Does not handle unallocated characters.
1933 *
1934 * @param string $text Text to test
1935 * @return null|string Directionality ('ltr' or 'rtl') or null
1936 */
1937 private static function strongDirFromContent( $text = '' ) {
1938 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1939 return null;
1940 }
1941 if ( $matches[1] === '' ) {
1942 return 'rtl';
1943 }
1944 return 'ltr';
1945 }
1946
1947 /**
1948 * Roman number formatting up to 10000
1949 *
1950 * @param int $num
1951 *
1952 * @return string
1953 */
1954 static function romanNumeral( $num ) {
1955 static $table = [
1956 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1957 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1958 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1959 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1960 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1961 ];
1962
1963 $num = intval( $num );
1964 if ( $num > 10000 || $num <= 0 ) {
1965 return $num;
1966 }
1967
1968 $s = '';
1969 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1970 if ( $num >= $pow10 ) {
1971 $s .= $table[$i][(int)floor( $num / $pow10 )];
1972 }
1973 $num = $num % $pow10;
1974 }
1975 return $s;
1976 }
1977
1978 /**
1979 * Hebrew Gematria number formatting up to 9999
1980 *
1981 * @param int $num
1982 *
1983 * @return string
1984 */
1985 static function hebrewNumeral( $num ) {
1986 static $table = [
1987 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
1988 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
1989 [ '',
1990 [ 'ק' ],
1991 [ 'ר' ],
1992 [ 'ש' ],
1993 [ 'ת' ],
1994 [ 'ת', 'ק' ],
1995 [ 'ת', 'ר' ],
1996 [ 'ת', 'ש' ],
1997 [ 'ת', 'ת' ],
1998 [ 'ת', 'ת', 'ק' ],
1999 [ 'ת', 'ת', 'ר' ],
2000 ],
2001 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2002 ];
2003
2004 $num = intval( $num );
2005 if ( $num > 9999 || $num <= 0 ) {
2006 return $num;
2007 }
2008
2009 // Round thousands have special notations
2010 if ( $num === 1000 ) {
2011 return "א' אלף";
2012 } elseif ( $num % 1000 === 0 ) {
2013 return $table[0][$num / 1000] . "' אלפים";
2014 }
2015
2016 $letters = [];
2017
2018 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2019 if ( $num >= $pow10 ) {
2020 if ( $num === 15 || $num === 16 ) {
2021 $letters[] = $table[0][9];
2022 $letters[] = $table[0][$num - 9];
2023 $num = 0;
2024 } else {
2025 $letters = array_merge(
2026 $letters,
2027 (array)$table[$i][intval( $num / $pow10 )]
2028 );
2029
2030 if ( $pow10 === 1000 ) {
2031 $letters[] = "'";
2032 }
2033 }
2034 }
2035
2036 $num = $num % $pow10;
2037 }
2038
2039 $preTransformLength = count( $letters );
2040 if ( $preTransformLength === 1 ) {
2041 // Add geresh (single quote) to one-letter numbers
2042 $letters[] = "'";
2043 } else {
2044 $lastIndex = $preTransformLength - 1;
2045 $letters[$lastIndex] = str_replace(
2046 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2047 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2048 $letters[$lastIndex]
2049 );
2050
2051 // Add gershayim (double quote) to multiple-letter numbers,
2052 // but exclude numbers with only one letter after the thousands
2053 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2054 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2055 $letters[] = "'";
2056 } else {
2057 array_splice( $letters, -1, 0, '"' );
2058 }
2059 }
2060
2061 return implode( $letters );
2062 }
2063
2064 /**
2065 * Used by date() and time() to adjust the time output.
2066 *
2067 * @param string $ts The time in date('YmdHis') format
2068 * @param mixed $tz Adjust the time by this amount (default false, mean we
2069 * get user timecorrection setting)
2070 * @return int
2071 */
2072 function userAdjust( $ts, $tz = false ) {
2073 global $wgUser, $wgLocalTZoffset;
2074
2075 if ( $tz === false ) {
2076 $tz = $wgUser->getOption( 'timecorrection' );
2077 }
2078
2079 $data = explode( '|', $tz, 3 );
2080
2081 if ( $data[0] == 'ZoneInfo' ) {
2082 MediaWiki\suppressWarnings();
2083 $userTZ = timezone_open( $data[2] );
2084 MediaWiki\restoreWarnings();
2085 if ( $userTZ !== false ) {
2086 $date = date_create( $ts, timezone_open( 'UTC' ) );
2087 date_timezone_set( $date, $userTZ );
2088 $date = date_format( $date, 'YmdHis' );
2089 return $date;
2090 }
2091 # Unrecognized timezone, default to 'Offset' with the stored offset.
2092 $data[0] = 'Offset';
2093 }
2094
2095 if ( $data[0] == 'System' || $tz == '' ) {
2096 # Global offset in minutes.
2097 $minDiff = $wgLocalTZoffset;
2098 } elseif ( $data[0] == 'Offset' ) {
2099 $minDiff = intval( $data[1] );
2100 } else {
2101 $data = explode( ':', $tz );
2102 if ( count( $data ) == 2 ) {
2103 $data[0] = intval( $data[0] );
2104 $data[1] = intval( $data[1] );
2105 $minDiff = abs( $data[0] ) * 60 + $data[1];
2106 if ( $data[0] < 0 ) {
2107 $minDiff = -$minDiff;
2108 }
2109 } else {
2110 $minDiff = intval( $data[0] ) * 60;
2111 }
2112 }
2113
2114 # No difference ? Return time unchanged
2115 if ( 0 == $minDiff ) {
2116 return $ts;
2117 }
2118
2119 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2120 # Generate an adjusted date; take advantage of the fact that mktime
2121 # will normalize out-of-range values so we don't have to split $minDiff
2122 # into hours and minutes.
2123 $t = mktime( (
2124 (int)substr( $ts, 8, 2 ) ), # Hours
2125 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2126 (int)substr( $ts, 12, 2 ), # Seconds
2127 (int)substr( $ts, 4, 2 ), # Month
2128 (int)substr( $ts, 6, 2 ), # Day
2129 (int)substr( $ts, 0, 4 ) ); # Year
2130
2131 $date = date( 'YmdHis', $t );
2132 MediaWiki\restoreWarnings();
2133
2134 return $date;
2135 }
2136
2137 /**
2138 * This is meant to be used by time(), date(), and timeanddate() to get
2139 * the date preference they're supposed to use, it should be used in
2140 * all children.
2141 *
2142 *<code>
2143 * function timeanddate([...], $format = true) {
2144 * $datePreference = $this->dateFormat($format);
2145 * [...]
2146 * }
2147 *</code>
2148 *
2149 * @param int|string|bool $usePrefs If true, the user's preference is used
2150 * if false, the site/language default is used
2151 * if int/string, assumed to be a format.
2152 * @return string
2153 */
2154 function dateFormat( $usePrefs = true ) {
2155 global $wgUser;
2156
2157 if ( is_bool( $usePrefs ) ) {
2158 if ( $usePrefs ) {
2159 $datePreference = $wgUser->getDatePreference();
2160 } else {
2161 $datePreference = (string)User::getDefaultOption( 'date' );
2162 }
2163 } else {
2164 $datePreference = (string)$usePrefs;
2165 }
2166
2167 // return int
2168 if ( $datePreference == '' ) {
2169 return 'default';
2170 }
2171
2172 return $datePreference;
2173 }
2174
2175 /**
2176 * Get a format string for a given type and preference
2177 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2178 * @param string $pref The format name as it appears in Messages*.php under
2179 * $datePreferences.
2180 *
2181 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2182 *
2183 * @return string
2184 */
2185 function getDateFormatString( $type, $pref ) {
2186 $wasDefault = false;
2187 if ( $pref == 'default' ) {
2188 $wasDefault = true;
2189 $pref = $this->getDefaultDateFormat();
2190 }
2191
2192 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2193 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2194
2195 if ( $type === 'pretty' && $df === null ) {
2196 $df = $this->getDateFormatString( 'date', $pref );
2197 }
2198
2199 if ( !$wasDefault && $df === null ) {
2200 $pref = $this->getDefaultDateFormat();
2201 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2202 }
2203
2204 $this->dateFormatStrings[$type][$pref] = $df;
2205 }
2206 return $this->dateFormatStrings[$type][$pref];
2207 }
2208
2209 /**
2210 * @param string $ts The time format which needs to be turned into a
2211 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2212 * @param bool $adj Whether to adjust the time output according to the
2213 * user configured offset ($timecorrection)
2214 * @param mixed $format True to use user's date format preference
2215 * @param string|bool $timecorrection The time offset as returned by
2216 * validateTimeZone() in Special:Preferences
2217 * @return string
2218 */
2219 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2220 $ts = wfTimestamp( TS_MW, $ts );
2221 if ( $adj ) {
2222 $ts = $this->userAdjust( $ts, $timecorrection );
2223 }
2224 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2225 return $this->sprintfDate( $df, $ts );
2226 }
2227
2228 /**
2229 * @param string $ts The time format which needs to be turned into a
2230 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2231 * @param bool $adj Whether to adjust the time output according to the
2232 * user configured offset ($timecorrection)
2233 * @param mixed $format True to use user's date format preference
2234 * @param string|bool $timecorrection The time offset as returned by
2235 * validateTimeZone() in Special:Preferences
2236 * @return string
2237 */
2238 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2239 $ts = wfTimestamp( TS_MW, $ts );
2240 if ( $adj ) {
2241 $ts = $this->userAdjust( $ts, $timecorrection );
2242 }
2243 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2244 return $this->sprintfDate( $df, $ts );
2245 }
2246
2247 /**
2248 * @param string $ts The time format which needs to be turned into a
2249 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2250 * @param bool $adj Whether to adjust the time output according to the
2251 * user configured offset ($timecorrection)
2252 * @param mixed $format What format to return, if it's false output the
2253 * default one (default true)
2254 * @param string|bool $timecorrection The time offset as returned by
2255 * validateTimeZone() in Special:Preferences
2256 * @return string
2257 */
2258 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2259 $ts = wfTimestamp( TS_MW, $ts );
2260 if ( $adj ) {
2261 $ts = $this->userAdjust( $ts, $timecorrection );
2262 }
2263 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2264 return $this->sprintfDate( $df, $ts );
2265 }
2266
2267 /**
2268 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2269 *
2270 * @since 1.20
2271 *
2272 * @param int $seconds The amount of seconds.
2273 * @param array $chosenIntervals The intervals to enable.
2274 *
2275 * @return string
2276 */
2277 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2278 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2279
2280 $segments = [];
2281
2282 foreach ( $intervals as $intervalName => $intervalValue ) {
2283 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2284 // duration-years, duration-decades, duration-centuries, duration-millennia
2285 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2286 $segments[] = $message->inLanguage( $this )->escaped();
2287 }
2288
2289 return $this->listToText( $segments );
2290 }
2291
2292 /**
2293 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2294 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2295 *
2296 * @since 1.20
2297 *
2298 * @param int $seconds The amount of seconds.
2299 * @param array $chosenIntervals The intervals to enable.
2300 *
2301 * @return array
2302 */
2303 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2304 if ( empty( $chosenIntervals ) ) {
2305 $chosenIntervals = [
2306 'millennia',
2307 'centuries',
2308 'decades',
2309 'years',
2310 'days',
2311 'hours',
2312 'minutes',
2313 'seconds'
2314 ];
2315 }
2316
2317 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2318 $sortedNames = array_keys( $intervals );
2319 $smallestInterval = array_pop( $sortedNames );
2320
2321 $segments = [];
2322
2323 foreach ( $intervals as $name => $length ) {
2324 $value = floor( $seconds / $length );
2325
2326 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2327 $seconds -= $value * $length;
2328 $segments[$name] = $value;
2329 }
2330 }
2331
2332 return $segments;
2333 }
2334
2335 /**
2336 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2337 *
2338 * @param string $type Can be 'date', 'time' or 'both'
2339 * @param string $ts The time format which needs to be turned into a
2340 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2341 * @param User $user User object used to get preferences for timezone and format
2342 * @param array $options Array, can contain the following keys:
2343 * - 'timecorrection': time correction, can have the following values:
2344 * - true: use user's preference
2345 * - false: don't use time correction
2346 * - int: value of time correction in minutes
2347 * - 'format': format to use, can have the following values:
2348 * - true: use user's preference
2349 * - false: use default preference
2350 * - string: format to use
2351 * @since 1.19
2352 * @return string
2353 */
2354 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2355 $ts = wfTimestamp( TS_MW, $ts );
2356 $options += [ 'timecorrection' => true, 'format' => true ];
2357 if ( $options['timecorrection'] !== false ) {
2358 if ( $options['timecorrection'] === true ) {
2359 $offset = $user->getOption( 'timecorrection' );
2360 } else {
2361 $offset = $options['timecorrection'];
2362 }
2363 $ts = $this->userAdjust( $ts, $offset );
2364 }
2365 if ( $options['format'] === true ) {
2366 $format = $user->getDatePreference();
2367 } else {
2368 $format = $options['format'];
2369 }
2370 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2371 return $this->sprintfDate( $df, $ts );
2372 }
2373
2374 /**
2375 * Get the formatted date for the given timestamp and formatted for
2376 * the given user.
2377 *
2378 * @param mixed $ts Mixed: the time format which needs to be turned into a
2379 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2380 * @param User $user User object used to get preferences for timezone and format
2381 * @param array $options Array, can contain the following keys:
2382 * - 'timecorrection': time correction, can have the following values:
2383 * - true: use user's preference
2384 * - false: don't use time correction
2385 * - int: value of time correction in minutes
2386 * - 'format': format to use, can have the following values:
2387 * - true: use user's preference
2388 * - false: use default preference
2389 * - string: format to use
2390 * @since 1.19
2391 * @return string
2392 */
2393 public function userDate( $ts, User $user, array $options = [] ) {
2394 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2395 }
2396
2397 /**
2398 * Get the formatted time for the given timestamp and formatted for
2399 * the given user.
2400 *
2401 * @param mixed $ts The time format which needs to be turned into a
2402 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2403 * @param User $user User object used to get preferences for timezone and format
2404 * @param array $options Array, can contain the following keys:
2405 * - 'timecorrection': time correction, can have the following values:
2406 * - true: use user's preference
2407 * - false: don't use time correction
2408 * - int: value of time correction in minutes
2409 * - 'format': format to use, can have the following values:
2410 * - true: use user's preference
2411 * - false: use default preference
2412 * - string: format to use
2413 * @since 1.19
2414 * @return string
2415 */
2416 public function userTime( $ts, User $user, array $options = [] ) {
2417 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2418 }
2419
2420 /**
2421 * Get the formatted date and time for the given timestamp and formatted for
2422 * the given user.
2423 *
2424 * @param mixed $ts The time format which needs to be turned into a
2425 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2426 * @param User $user User object used to get preferences for timezone and format
2427 * @param array $options Array, can contain the following keys:
2428 * - 'timecorrection': time correction, can have the following values:
2429 * - true: use user's preference
2430 * - false: don't use time correction
2431 * - int: value of time correction in minutes
2432 * - 'format': format to use, can have the following values:
2433 * - true: use user's preference
2434 * - false: use default preference
2435 * - string: format to use
2436 * @since 1.19
2437 * @return string
2438 */
2439 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2440 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2441 }
2442
2443 /**
2444 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2445 *
2446 * Determine the difference between the timestamp and the current time, and
2447 * generate a readable timestamp by returning "<N> <units> ago", where the
2448 * largest possible unit is used.
2449 *
2450 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2451 *
2452 * @param MWTimestamp $time
2453 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2454 * @param User|null $user User the timestamp is being generated for
2455 * (or null to use main context's user)
2456 * @return string Formatted timestamp
2457 */
2458 public function getHumanTimestamp(
2459 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2460 ) {
2461 if ( $relativeTo === null ) {
2462 $relativeTo = new MWTimestamp();
2463 }
2464 if ( $user === null ) {
2465 $user = RequestContext::getMain()->getUser();
2466 }
2467
2468 // Adjust for the user's timezone.
2469 $offsetThis = $time->offsetForUser( $user );
2470 $offsetRel = $relativeTo->offsetForUser( $user );
2471
2472 $ts = '';
2473 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2474 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2475 }
2476
2477 // Reset the timezone on the objects.
2478 $time->timestamp->sub( $offsetThis );
2479 $relativeTo->timestamp->sub( $offsetRel );
2480
2481 return $ts;
2482 }
2483
2484 /**
2485 * Convert an MWTimestamp into a pretty human-readable timestamp using
2486 * the given user preferences and relative base time.
2487 *
2488 * @see Language::getHumanTimestamp
2489 * @param MWTimestamp $ts Timestamp to prettify
2490 * @param MWTimestamp $relativeTo Base timestamp
2491 * @param User $user User preferences to use
2492 * @return string Human timestamp
2493 * @since 1.26
2494 */
2495 private function getHumanTimestampInternal(
2496 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2497 ) {
2498 $diff = $ts->diff( $relativeTo );
2499 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2500 (int)$relativeTo->timestamp->format( 'w' ) );
2501 $days = $diff->days ?: (int)$diffDay;
2502 if ( $diff->invert || $days > 5
2503 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2504 ) {
2505 // Timestamps are in different years: use full timestamp
2506 // Also do full timestamp for future dates
2507 /**
2508 * @todo FIXME: Add better handling of future timestamps.
2509 */
2510 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2511 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2512 } elseif ( $days > 5 ) {
2513 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2514 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2515 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2516 } elseif ( $days > 1 ) {
2517 // Timestamp within the past week: show the day of the week and time
2518 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2519 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2520 // Messages:
2521 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2522 $ts = wfMessage( "$weekday-at" )
2523 ->inLanguage( $this )
2524 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2525 ->text();
2526 } elseif ( $days == 1 ) {
2527 // Timestamp was yesterday: say 'yesterday' and the time.
2528 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2529 $ts = wfMessage( 'yesterday-at' )
2530 ->inLanguage( $this )
2531 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2532 ->text();
2533 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2534 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2535 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2536 $ts = wfMessage( 'today-at' )
2537 ->inLanguage( $this )
2538 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2539 ->text();
2540
2541 // From here on in, the timestamp was soon enough ago so that we can simply say
2542 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2543 } elseif ( $diff->h == 1 ) {
2544 // Less than 90 minutes, but more than an hour ago.
2545 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2546 } elseif ( $diff->i >= 1 ) {
2547 // A few minutes ago.
2548 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2549 } elseif ( $diff->s >= 30 ) {
2550 // Less than a minute, but more than 30 sec ago.
2551 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2552 } else {
2553 // Less than 30 seconds ago.
2554 $ts = wfMessage( 'just-now' )->text();
2555 }
2556
2557 return $ts;
2558 }
2559
2560 /**
2561 * @param string $key
2562 * @return array|null
2563 */
2564 function getMessage( $key ) {
2565 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2566 }
2567
2568 /**
2569 * @return array
2570 */
2571 function getAllMessages() {
2572 return self::$dataCache->getItem( $this->mCode, 'messages' );
2573 }
2574
2575 /**
2576 * @param string $in
2577 * @param string $out
2578 * @param string $string
2579 * @return string
2580 */
2581 function iconv( $in, $out, $string ) {
2582 # This is a wrapper for iconv in all languages except esperanto,
2583 # which does some nasty x-conversions beforehand
2584
2585 # Even with //IGNORE iconv can whine about illegal characters in
2586 # *input* string. We just ignore those too.
2587 # REF: http://bugs.php.net/bug.php?id=37166
2588 # REF: https://phabricator.wikimedia.org/T18885
2589 MediaWiki\suppressWarnings();
2590 $text = iconv( $in, $out . '//IGNORE', $string );
2591 MediaWiki\restoreWarnings();
2592 return $text;
2593 }
2594
2595 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2596
2597 /**
2598 * @param array $matches
2599 * @return mixed|string
2600 */
2601 function ucwordbreaksCallbackAscii( $matches ) {
2602 return $this->ucfirst( $matches[1] );
2603 }
2604
2605 /**
2606 * @param array $matches
2607 * @return string
2608 */
2609 function ucwordbreaksCallbackMB( $matches ) {
2610 return mb_strtoupper( $matches[0] );
2611 }
2612
2613 /**
2614 * @param array $matches
2615 * @return string
2616 */
2617 function ucCallback( $matches ) {
2618 list( $wikiUpperChars ) = self::getCaseMaps();
2619 return strtr( $matches[1], $wikiUpperChars );
2620 }
2621
2622 /**
2623 * @param array $matches
2624 * @return string
2625 */
2626 function lcCallback( $matches ) {
2627 list( , $wikiLowerChars ) = self::getCaseMaps();
2628 return strtr( $matches[1], $wikiLowerChars );
2629 }
2630
2631 /**
2632 * @param array $matches
2633 * @return string
2634 */
2635 function ucwordsCallbackMB( $matches ) {
2636 return mb_strtoupper( $matches[0] );
2637 }
2638
2639 /**
2640 * @param array $matches
2641 * @return string
2642 */
2643 function ucwordsCallbackWiki( $matches ) {
2644 list( $wikiUpperChars ) = self::getCaseMaps();
2645 return strtr( $matches[0], $wikiUpperChars );
2646 }
2647
2648 /**
2649 * Make a string's first character uppercase
2650 *
2651 * @param string $str
2652 *
2653 * @return string
2654 */
2655 function ucfirst( $str ) {
2656 $o = ord( $str );
2657 if ( $o < 96 ) { // if already uppercase...
2658 return $str;
2659 } elseif ( $o < 128 ) {
2660 return ucfirst( $str ); // use PHP's ucfirst()
2661 } else {
2662 // fall back to more complex logic in case of multibyte strings
2663 return $this->uc( $str, true );
2664 }
2665 }
2666
2667 /**
2668 * Convert a string to uppercase
2669 *
2670 * @param string $str
2671 * @param bool $first
2672 *
2673 * @return string
2674 */
2675 function uc( $str, $first = false ) {
2676 if ( function_exists( 'mb_strtoupper' ) ) {
2677 if ( $first ) {
2678 if ( $this->isMultibyte( $str ) ) {
2679 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2680 } else {
2681 return ucfirst( $str );
2682 }
2683 } else {
2684 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2685 }
2686 } else {
2687 if ( $this->isMultibyte( $str ) ) {
2688 $x = $first ? '^' : '';
2689 return preg_replace_callback(
2690 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2691 [ $this, 'ucCallback' ],
2692 $str
2693 );
2694 } else {
2695 return $first ? ucfirst( $str ) : strtoupper( $str );
2696 }
2697 }
2698 }
2699
2700 /**
2701 * @param string $str
2702 * @return mixed|string
2703 */
2704 function lcfirst( $str ) {
2705 $o = ord( $str );
2706 if ( !$o ) {
2707 return strval( $str );
2708 } elseif ( $o >= 128 ) {
2709 return $this->lc( $str, true );
2710 } elseif ( $o > 96 ) {
2711 return $str;
2712 } else {
2713 $str[0] = strtolower( $str[0] );
2714 return $str;
2715 }
2716 }
2717
2718 /**
2719 * @param string $str
2720 * @param bool $first
2721 * @return mixed|string
2722 */
2723 function lc( $str, $first = false ) {
2724 if ( function_exists( 'mb_strtolower' ) ) {
2725 if ( $first ) {
2726 if ( $this->isMultibyte( $str ) ) {
2727 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2728 } else {
2729 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2730 }
2731 } else {
2732 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2733 }
2734 } else {
2735 if ( $this->isMultibyte( $str ) ) {
2736 $x = $first ? '^' : '';
2737 return preg_replace_callback(
2738 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2739 [ $this, 'lcCallback' ],
2740 $str
2741 );
2742 } else {
2743 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2744 }
2745 }
2746 }
2747
2748 /**
2749 * @param string $str
2750 * @return bool
2751 */
2752 function isMultibyte( $str ) {
2753 return strlen( $str ) !== mb_strlen( $str );
2754 }
2755
2756 /**
2757 * @param string $str
2758 * @return mixed|string
2759 */
2760 function ucwords( $str ) {
2761 if ( $this->isMultibyte( $str ) ) {
2762 $str = $this->lc( $str );
2763
2764 // regexp to find first letter in each word (i.e. after each space)
2765 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2766
2767 // function to use to capitalize a single char
2768 if ( function_exists( 'mb_strtoupper' ) ) {
2769 return preg_replace_callback(
2770 $replaceRegexp,
2771 [ $this, 'ucwordsCallbackMB' ],
2772 $str
2773 );
2774 } else {
2775 return preg_replace_callback(
2776 $replaceRegexp,
2777 [ $this, 'ucwordsCallbackWiki' ],
2778 $str
2779 );
2780 }
2781 } else {
2782 return ucwords( strtolower( $str ) );
2783 }
2784 }
2785
2786 /**
2787 * capitalize words at word breaks
2788 *
2789 * @param string $str
2790 * @return mixed
2791 */
2792 function ucwordbreaks( $str ) {
2793 if ( $this->isMultibyte( $str ) ) {
2794 $str = $this->lc( $str );
2795
2796 // since \b doesn't work for UTF-8, we explicitely define word break chars
2797 $breaks = "[ \-\(\)\}\{\.,\?!]";
2798
2799 // find first letter after word break
2800 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2801 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2802
2803 if ( function_exists( 'mb_strtoupper' ) ) {
2804 return preg_replace_callback(
2805 $replaceRegexp,
2806 [ $this, 'ucwordbreaksCallbackMB' ],
2807 $str
2808 );
2809 } else {
2810 return preg_replace_callback(
2811 $replaceRegexp,
2812 [ $this, 'ucwordsCallbackWiki' ],
2813 $str
2814 );
2815 }
2816 } else {
2817 return preg_replace_callback(
2818 '/\b([\w\x80-\xff]+)\b/',
2819 [ $this, 'ucwordbreaksCallbackAscii' ],
2820 $str
2821 );
2822 }
2823 }
2824
2825 /**
2826 * Return a case-folded representation of $s
2827 *
2828 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2829 * and $s2 are the same except for the case of their characters. It is not
2830 * necessary for the value returned to make sense when displayed.
2831 *
2832 * Do *not* perform any other normalisation in this function. If a caller
2833 * uses this function when it should be using a more general normalisation
2834 * function, then fix the caller.
2835 *
2836 * @param string $s
2837 *
2838 * @return string
2839 */
2840 function caseFold( $s ) {
2841 return $this->uc( $s );
2842 }
2843
2844 /**
2845 * @param string $s
2846 * @return string
2847 * @throws MWException
2848 */
2849 function checkTitleEncoding( $s ) {
2850 if ( is_array( $s ) ) {
2851 throw new MWException( 'Given array to checkTitleEncoding.' );
2852 }
2853 if ( StringUtils::isUtf8( $s ) ) {
2854 return $s;
2855 }
2856
2857 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2858 }
2859
2860 /**
2861 * @return array
2862 */
2863 function fallback8bitEncoding() {
2864 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2865 }
2866
2867 /**
2868 * Most writing systems use whitespace to break up words.
2869 * Some languages such as Chinese don't conventionally do this,
2870 * which requires special handling when breaking up words for
2871 * searching etc.
2872 *
2873 * @return bool
2874 */
2875 function hasWordBreaks() {
2876 return true;
2877 }
2878
2879 /**
2880 * Some languages such as Chinese require word segmentation,
2881 * Specify such segmentation when overridden in derived class.
2882 *
2883 * @param string $string
2884 * @return string
2885 */
2886 function segmentByWord( $string ) {
2887 return $string;
2888 }
2889
2890 /**
2891 * Some languages have special punctuation need to be normalized.
2892 * Make such changes here.
2893 *
2894 * @param string $string
2895 * @return string
2896 */
2897 function normalizeForSearch( $string ) {
2898 return self::convertDoubleWidth( $string );
2899 }
2900
2901 /**
2902 * convert double-width roman characters to single-width.
2903 * range: ff00-ff5f ~= 0020-007f
2904 *
2905 * @param string $string
2906 *
2907 * @return string
2908 */
2909 protected static function convertDoubleWidth( $string ) {
2910 static $full = null;
2911 static $half = null;
2912
2913 if ( $full === null ) {
2914 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2915 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2916 $full = str_split( $fullWidth, 3 );
2917 $half = str_split( $halfWidth );
2918 }
2919
2920 $string = str_replace( $full, $half, $string );
2921 return $string;
2922 }
2923
2924 /**
2925 * @param string $string
2926 * @param string $pattern
2927 * @return string
2928 */
2929 protected static function insertSpace( $string, $pattern ) {
2930 $string = preg_replace( $pattern, " $1 ", $string );
2931 $string = preg_replace( '/ +/', ' ', $string );
2932 return $string;
2933 }
2934
2935 /**
2936 * @param array $termsArray
2937 * @return array
2938 */
2939 function convertForSearchResult( $termsArray ) {
2940 # some languages, e.g. Chinese, need to do a conversion
2941 # in order for search results to be displayed correctly
2942 return $termsArray;
2943 }
2944
2945 /**
2946 * Get the first character of a string.
2947 *
2948 * @param string $s
2949 * @return string
2950 */
2951 function firstChar( $s ) {
2952 $matches = [];
2953 preg_match(
2954 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2955 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2956 $s,
2957 $matches
2958 );
2959
2960 if ( isset( $matches[1] ) ) {
2961 if ( strlen( $matches[1] ) != 3 ) {
2962 return $matches[1];
2963 }
2964
2965 // Break down Hangul syllables to grab the first jamo
2966 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2967 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2968 return $matches[1];
2969 } elseif ( $code < 0xb098 ) {
2970 return "\xe3\x84\xb1";
2971 } elseif ( $code < 0xb2e4 ) {
2972 return "\xe3\x84\xb4";
2973 } elseif ( $code < 0xb77c ) {
2974 return "\xe3\x84\xb7";
2975 } elseif ( $code < 0xb9c8 ) {
2976 return "\xe3\x84\xb9";
2977 } elseif ( $code < 0xbc14 ) {
2978 return "\xe3\x85\x81";
2979 } elseif ( $code < 0xc0ac ) {
2980 return "\xe3\x85\x82";
2981 } elseif ( $code < 0xc544 ) {
2982 return "\xe3\x85\x85";
2983 } elseif ( $code < 0xc790 ) {
2984 return "\xe3\x85\x87";
2985 } elseif ( $code < 0xcc28 ) {
2986 return "\xe3\x85\x88";
2987 } elseif ( $code < 0xce74 ) {
2988 return "\xe3\x85\x8a";
2989 } elseif ( $code < 0xd0c0 ) {
2990 return "\xe3\x85\x8b";
2991 } elseif ( $code < 0xd30c ) {
2992 return "\xe3\x85\x8c";
2993 } elseif ( $code < 0xd558 ) {
2994 return "\xe3\x85\x8d";
2995 } else {
2996 return "\xe3\x85\x8e";
2997 }
2998 } else {
2999 return '';
3000 }
3001 }
3002
3003 function initEncoding() {
3004 # Some languages may have an alternate char encoding option
3005 # (Esperanto X-coding, Japanese furigana conversion, etc)
3006 # If this language is used as the primary content language,
3007 # an override to the defaults can be set here on startup.
3008 }
3009
3010 /**
3011 * @param string $s
3012 * @return string
3013 */
3014 function recodeForEdit( $s ) {
3015 # For some languages we'll want to explicitly specify
3016 # which characters make it into the edit box raw
3017 # or are converted in some way or another.
3018 global $wgEditEncoding;
3019 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
3020 return $s;
3021 } else {
3022 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3023 }
3024 }
3025
3026 /**
3027 * @param string $s
3028 * @return string
3029 */
3030 function recodeInput( $s ) {
3031 # Take the previous into account.
3032 global $wgEditEncoding;
3033 if ( $wgEditEncoding != '' ) {
3034 $enc = $wgEditEncoding;
3035 } else {
3036 $enc = 'UTF-8';
3037 }
3038 if ( $enc == 'UTF-8' ) {
3039 return $s;
3040 } else {
3041 return $this->iconv( $enc, 'UTF-8', $s );
3042 }
3043 }
3044
3045 /**
3046 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3047 * also cleans up certain backwards-compatible sequences, converting them
3048 * to the modern Unicode equivalent.
3049 *
3050 * This is language-specific for performance reasons only.
3051 *
3052 * @param string $s
3053 *
3054 * @return string
3055 */
3056 function normalize( $s ) {
3057 global $wgAllUnicodeFixes;
3058 $s = UtfNormal\Validator::cleanUp( $s );
3059 if ( $wgAllUnicodeFixes ) {
3060 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3061 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3062 }
3063
3064 return $s;
3065 }
3066
3067 /**
3068 * Transform a string using serialized data stored in the given file (which
3069 * must be in the serialized subdirectory of $IP). The file contains pairs
3070 * mapping source characters to destination characters.
3071 *
3072 * The data is cached in process memory. This will go faster if you have the
3073 * FastStringSearch extension.
3074 *
3075 * @param string $file
3076 * @param string $string
3077 *
3078 * @throws MWException
3079 * @return string
3080 */
3081 function transformUsingPairFile( $file, $string ) {
3082 if ( !isset( $this->transformData[$file] ) ) {
3083 $data = wfGetPrecompiledData( $file );
3084 if ( $data === false ) {
3085 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
3086 }
3087 $this->transformData[$file] = new ReplacementArray( $data );
3088 }
3089 return $this->transformData[$file]->replace( $string );
3090 }
3091
3092 /**
3093 * For right-to-left language support
3094 *
3095 * @return bool
3096 */
3097 function isRTL() {
3098 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3099 }
3100
3101 /**
3102 * Return the correct HTML 'dir' attribute value for this language.
3103 * @return string
3104 */
3105 function getDir() {
3106 return $this->isRTL() ? 'rtl' : 'ltr';
3107 }
3108
3109 /**
3110 * Return 'left' or 'right' as appropriate alignment for line-start
3111 * for this language's text direction.
3112 *
3113 * Should be equivalent to CSS3 'start' text-align value....
3114 *
3115 * @return string
3116 */
3117 function alignStart() {
3118 return $this->isRTL() ? 'right' : 'left';
3119 }
3120
3121 /**
3122 * Return 'right' or 'left' as appropriate alignment for line-end
3123 * for this language's text direction.
3124 *
3125 * Should be equivalent to CSS3 'end' text-align value....
3126 *
3127 * @return string
3128 */
3129 function alignEnd() {
3130 return $this->isRTL() ? 'left' : 'right';
3131 }
3132
3133 /**
3134 * A hidden direction mark (LRM or RLM), depending on the language direction.
3135 * Unlike getDirMark(), this function returns the character as an HTML entity.
3136 * This function should be used when the output is guaranteed to be HTML,
3137 * because it makes the output HTML source code more readable. When
3138 * the output is plain text or can be escaped, getDirMark() should be used.
3139 *
3140 * @param bool $opposite Get the direction mark opposite to your language
3141 * @return string
3142 * @since 1.20
3143 */
3144 function getDirMarkEntity( $opposite = false ) {
3145 if ( $opposite ) {
3146 return $this->isRTL() ? '&lrm;' : '&rlm;';
3147 }
3148 return $this->isRTL() ? '&rlm;' : '&lrm;';
3149 }
3150
3151 /**
3152 * A hidden direction mark (LRM or RLM), depending on the language direction.
3153 * This function produces them as invisible Unicode characters and
3154 * the output may be hard to read and debug, so it should only be used
3155 * when the output is plain text or can be escaped. When the output is
3156 * HTML, use getDirMarkEntity() instead.
3157 *
3158 * @param bool $opposite Get the direction mark opposite to your language
3159 * @return string
3160 */
3161 function getDirMark( $opposite = false ) {
3162 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3163 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3164 if ( $opposite ) {
3165 return $this->isRTL() ? $lrm : $rlm;
3166 }
3167 return $this->isRTL() ? $rlm : $lrm;
3168 }
3169
3170 /**
3171 * @return array
3172 */
3173 function capitalizeAllNouns() {
3174 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3175 }
3176
3177 /**
3178 * An arrow, depending on the language direction.
3179 *
3180 * @param string $direction The direction of the arrow: forwards (default),
3181 * backwards, left, right, up, down.
3182 * @return string
3183 */
3184 function getArrow( $direction = 'forwards' ) {
3185 switch ( $direction ) {
3186 case 'forwards':
3187 return $this->isRTL() ? '←' : '→';
3188 case 'backwards':
3189 return $this->isRTL() ? '→' : '←';
3190 case 'left':
3191 return '←';
3192 case 'right':
3193 return '→';
3194 case 'up':
3195 return '↑';
3196 case 'down':
3197 return '↓';
3198 }
3199 }
3200
3201 /**
3202 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3203 *
3204 * @return bool
3205 */
3206 function linkPrefixExtension() {
3207 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3208 }
3209
3210 /**
3211 * Get all magic words from cache.
3212 * @return array
3213 */
3214 function getMagicWords() {
3215 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3216 }
3217
3218 /**
3219 * Run the LanguageGetMagic hook once.
3220 */
3221 protected function doMagicHook() {
3222 if ( $this->mMagicHookDone ) {
3223 return;
3224 }
3225 $this->mMagicHookDone = true;
3226 Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
3227 }
3228
3229 /**
3230 * Fill a MagicWord object with data from here
3231 *
3232 * @param MagicWord $mw
3233 */
3234 function getMagic( $mw ) {
3235 // Saves a function call
3236 if ( !$this->mMagicHookDone ) {
3237 $this->doMagicHook();
3238 }
3239
3240 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3241 $rawEntry = $this->mMagicExtensions[$mw->mId];
3242 } else {
3243 $rawEntry = self::$dataCache->getSubitem(
3244 $this->mCode, 'magicWords', $mw->mId );
3245 }
3246
3247 if ( !is_array( $rawEntry ) ) {
3248 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3249 } else {
3250 $mw->mCaseSensitive = $rawEntry[0];
3251 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3252 }
3253 }
3254
3255 /**
3256 * Add magic words to the extension array
3257 *
3258 * @param array $newWords
3259 */
3260 function addMagicWordsByLang( $newWords ) {
3261 $fallbackChain = $this->getFallbackLanguages();
3262 $fallbackChain = array_reverse( $fallbackChain );
3263 foreach ( $fallbackChain as $code ) {
3264 if ( isset( $newWords[$code] ) ) {
3265 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3266 }
3267 }
3268 }
3269
3270 /**
3271 * Get special page names, as an associative array
3272 * canonical name => array of valid names, including aliases
3273 * @return array
3274 */
3275 function getSpecialPageAliases() {
3276 // Cache aliases because it may be slow to load them
3277 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3278 // Initialise array
3279 $this->mExtendedSpecialPageAliases =
3280 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3281 Hooks::run( 'LanguageGetSpecialPageAliases',
3282 [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
3283 }
3284
3285 return $this->mExtendedSpecialPageAliases;
3286 }
3287
3288 /**
3289 * Italic is unsuitable for some languages
3290 *
3291 * @param string $text The text to be emphasized.
3292 * @return string
3293 */
3294 function emphasize( $text ) {
3295 return "<em>$text</em>";
3296 }
3297
3298 /**
3299 * Normally we output all numbers in plain en_US style, that is
3300 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3301 * point twohundredthirtyfive. However this is not suitable for all
3302 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3303 * Icelandic just want to use commas instead of dots, and dots instead
3304 * of commas like "293.291,235".
3305 *
3306 * An example of this function being called:
3307 * <code>
3308 * wfMessage( 'message' )->numParams( $num )->text()
3309 * </code>
3310 *
3311 * See $separatorTransformTable on MessageIs.php for
3312 * the , => . and . => , implementation.
3313 *
3314 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3315 * @param int|float $number The string to be formatted, should be an integer
3316 * or a floating point number.
3317 * @param bool $nocommafy Set to true for special numbers like dates
3318 * @return string
3319 */
3320 public function formatNum( $number, $nocommafy = false ) {
3321 global $wgTranslateNumerals;
3322 if ( !$nocommafy ) {
3323 $number = $this->commafy( $number );
3324 $s = $this->separatorTransformTable();
3325 if ( $s ) {
3326 $number = strtr( $number, $s );
3327 }
3328 }
3329
3330 if ( $wgTranslateNumerals ) {
3331 $s = $this->digitTransformTable();
3332 if ( $s ) {
3333 $number = strtr( $number, $s );
3334 }
3335 }
3336
3337 return $number;
3338 }
3339
3340 /**
3341 * Front-end for non-commafied formatNum
3342 *
3343 * @param int|float $number The string to be formatted, should be an integer
3344 * or a floating point number.
3345 * @since 1.21
3346 * @return string
3347 */
3348 public function formatNumNoSeparators( $number ) {
3349 return $this->formatNum( $number, true );
3350 }
3351
3352 /**
3353 * @param string $number
3354 * @return string
3355 */
3356 public function parseFormattedNumber( $number ) {
3357 $s = $this->digitTransformTable();
3358 if ( $s ) {
3359 // eliminate empty array values such as ''. (bug 64347)
3360 $s = array_filter( $s );
3361 $number = strtr( $number, array_flip( $s ) );
3362 }
3363
3364 $s = $this->separatorTransformTable();
3365 if ( $s ) {
3366 // eliminate empty array values such as ''. (bug 64347)
3367 $s = array_filter( $s );
3368 $number = strtr( $number, array_flip( $s ) );
3369 }
3370
3371 $number = strtr( $number, [ ',' => '' ] );
3372 return $number;
3373 }
3374
3375 /**
3376 * Adds commas to a given number
3377 * @since 1.19
3378 * @param mixed $number
3379 * @return string
3380 */
3381 function commafy( $number ) {
3382 $digitGroupingPattern = $this->digitGroupingPattern();
3383 if ( $number === null ) {
3384 return '';
3385 }
3386
3387 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3388 // default grouping is at thousands, use the same for ###,###,### pattern too.
3389 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3390 } else {
3391 // Ref: http://cldr.unicode.org/translation/number-patterns
3392 $sign = "";
3393 if ( intval( $number ) < 0 ) {
3394 // For negative numbers apply the algorithm like positive number and add sign.
3395 $sign = "-";
3396 $number = substr( $number, 1 );
3397 }
3398 $integerPart = [];
3399 $decimalPart = [];
3400 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3401 preg_match( "/\d+/", $number, $integerPart );
3402 preg_match( "/\.\d*/", $number, $decimalPart );
3403 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3404 if ( $groupedNumber === $number ) {
3405 // the string does not have any number part. Eg: .12345
3406 return $sign . $groupedNumber;
3407 }
3408 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3409 while ( $start > 0 ) {
3410 $match = $matches[0][$numMatches - 1];
3411 $matchLen = strlen( $match );
3412 $start = $end - $matchLen;
3413 if ( $start < 0 ) {
3414 $start = 0;
3415 }
3416 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3417 $end = $start;
3418 if ( $numMatches > 1 ) {
3419 // use the last pattern for the rest of the number
3420 $numMatches--;
3421 }
3422 if ( $start > 0 ) {
3423 $groupedNumber = "," . $groupedNumber;
3424 }
3425 }
3426 return $sign . $groupedNumber;
3427 }
3428 }
3429
3430 /**
3431 * @return string
3432 */
3433 function digitGroupingPattern() {
3434 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3435 }
3436
3437 /**
3438 * @return array
3439 */
3440 function digitTransformTable() {
3441 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3442 }
3443
3444 /**
3445 * @return array
3446 */
3447 function separatorTransformTable() {
3448 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3449 }
3450
3451 /**
3452 * Take a list of strings and build a locale-friendly comma-separated
3453 * list, using the local comma-separator message.
3454 * The last two strings are chained with an "and".
3455 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3456 *
3457 * @param string[] $l
3458 * @return string
3459 */
3460 function listToText( array $l ) {
3461 $m = count( $l ) - 1;
3462 if ( $m < 0 ) {
3463 return '';
3464 }
3465 if ( $m > 0 ) {
3466 $and = $this->msg( 'and' )->escaped();
3467 $space = $this->msg( 'word-separator' )->escaped();
3468 if ( $m > 1 ) {
3469 $comma = $this->msg( 'comma-separator' )->escaped();
3470 }
3471 }
3472 $s = $l[$m];
3473 for ( $i = $m - 1; $i >= 0; $i-- ) {
3474 if ( $i == $m - 1 ) {
3475 $s = $l[$i] . $and . $space . $s;
3476 } else {
3477 $s = $l[$i] . $comma . $s;
3478 }
3479 }
3480 return $s;
3481 }
3482
3483 /**
3484 * Take a list of strings and build a locale-friendly comma-separated
3485 * list, using the local comma-separator message.
3486 * @param string[] $list Array of strings to put in a comma list
3487 * @return string
3488 */
3489 function commaList( array $list ) {
3490 return implode(
3491 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3492 $list
3493 );
3494 }
3495
3496 /**
3497 * Take a list of strings and build a locale-friendly semicolon-separated
3498 * list, using the local semicolon-separator message.
3499 * @param string[] $list Array of strings to put in a semicolon list
3500 * @return string
3501 */
3502 function semicolonList( array $list ) {
3503 return implode(
3504 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3505 $list
3506 );
3507 }
3508
3509 /**
3510 * Same as commaList, but separate it with the pipe instead.
3511 * @param string[] $list Array of strings to put in a pipe list
3512 * @return string
3513 */
3514 function pipeList( array $list ) {
3515 return implode(
3516 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3517 $list
3518 );
3519 }
3520
3521 /**
3522 * Truncate a string to a specified length in bytes, appending an optional
3523 * string (e.g. for ellipses)
3524 *
3525 * The database offers limited byte lengths for some columns in the database;
3526 * multi-byte character sets mean we need to ensure that only whole characters
3527 * are included, otherwise broken characters can be passed to the user
3528 *
3529 * If $length is negative, the string will be truncated from the beginning
3530 *
3531 * @param string $string String to truncate
3532 * @param int $length Maximum length (including ellipses)
3533 * @param string $ellipsis String to append to the truncated text
3534 * @param bool $adjustLength Subtract length of ellipsis from $length.
3535 * $adjustLength was introduced in 1.18, before that behaved as if false.
3536 * @return string
3537 */
3538 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3539 # Use the localized ellipsis character
3540 if ( $ellipsis == '...' ) {
3541 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3542 }
3543 # Check if there is no need to truncate
3544 if ( $length == 0 ) {
3545 return $ellipsis; // convention
3546 } elseif ( strlen( $string ) <= abs( $length ) ) {
3547 return $string; // no need to truncate
3548 }
3549 $stringOriginal = $string;
3550 # If ellipsis length is >= $length then we can't apply $adjustLength
3551 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3552 $string = $ellipsis; // this can be slightly unexpected
3553 # Otherwise, truncate and add ellipsis...
3554 } else {
3555 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3556 if ( $length > 0 ) {
3557 $length -= $eLength;
3558 $string = substr( $string, 0, $length ); // xyz...
3559 $string = $this->removeBadCharLast( $string );
3560 $string = rtrim( $string );
3561 $string = $string . $ellipsis;
3562 } else {
3563 $length += $eLength;
3564 $string = substr( $string, $length ); // ...xyz
3565 $string = $this->removeBadCharFirst( $string );
3566 $string = ltrim( $string );
3567 $string = $ellipsis . $string;
3568 }
3569 }
3570 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3571 # This check is *not* redundant if $adjustLength, due to the single case where
3572 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3573 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3574 return $string;
3575 } else {
3576 return $stringOriginal;
3577 }
3578 }
3579
3580 /**
3581 * Remove bytes that represent an incomplete Unicode character
3582 * at the end of string (e.g. bytes of the char are missing)
3583 *
3584 * @param string $string
3585 * @return string
3586 */
3587 protected function removeBadCharLast( $string ) {
3588 if ( $string != '' ) {
3589 $char = ord( $string[strlen( $string ) - 1] );
3590 $m = [];
3591 if ( $char >= 0xc0 ) {
3592 # We got the first byte only of a multibyte char; remove it.
3593 $string = substr( $string, 0, -1 );
3594 } elseif ( $char >= 0x80 &&
3595 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3596 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3597 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3598 ) {
3599 # We chopped in the middle of a character; remove it
3600 $string = $m[1];
3601 }
3602 }
3603 return $string;
3604 }
3605
3606 /**
3607 * Remove bytes that represent an incomplete Unicode character
3608 * at the start of string (e.g. bytes of the char are missing)
3609 *
3610 * @param string $string
3611 * @return string
3612 */
3613 protected function removeBadCharFirst( $string ) {
3614 if ( $string != '' ) {
3615 $char = ord( $string[0] );
3616 if ( $char >= 0x80 && $char < 0xc0 ) {
3617 # We chopped in the middle of a character; remove the whole thing
3618 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3619 }
3620 }
3621 return $string;
3622 }
3623
3624 /**
3625 * Truncate a string of valid HTML to a specified length in bytes,
3626 * appending an optional string (e.g. for ellipses), and return valid HTML
3627 *
3628 * This is only intended for styled/linked text, such as HTML with
3629 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3630 * Also, this will not detect things like "display:none" CSS.
3631 *
3632 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3633 *
3634 * @param string $text HTML string to truncate
3635 * @param int $length (zero/positive) Maximum length (including ellipses)
3636 * @param string $ellipsis String to append to the truncated text
3637 * @return string
3638 */
3639 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3640 # Use the localized ellipsis character
3641 if ( $ellipsis == '...' ) {
3642 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3643 }
3644 # Check if there is clearly no need to truncate
3645 if ( $length <= 0 ) {
3646 return $ellipsis; // no text shown, nothing to format (convention)
3647 } elseif ( strlen( $text ) <= $length ) {
3648 return $text; // string short enough even *with* HTML (short-circuit)
3649 }
3650
3651 $dispLen = 0; // innerHTML legth so far
3652 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3653 $tagType = 0; // 0-open, 1-close
3654 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3655 $entityState = 0; // 0-not entity, 1-entity
3656 $tag = $ret = ''; // accumulated tag name, accumulated result string
3657 $openTags = []; // open tag stack
3658 $maybeState = null; // possible truncation state
3659
3660 $textLen = strlen( $text );
3661 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3662 for ( $pos = 0; true; ++$pos ) {
3663 # Consider truncation once the display length has reached the maximim.
3664 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3665 # Check that we're not in the middle of a bracket/entity...
3666 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3667 if ( !$testingEllipsis ) {
3668 $testingEllipsis = true;
3669 # Save where we are; we will truncate here unless there turn out to
3670 # be so few remaining characters that truncation is not necessary.
3671 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3672 $maybeState = [ $ret, $openTags ]; // save state
3673 }
3674 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3675 # String in fact does need truncation, the truncation point was OK.
3676 list( $ret, $openTags ) = $maybeState; // reload state
3677 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3678 $ret .= $ellipsis; // add ellipsis
3679 break;
3680 }
3681 }
3682 if ( $pos >= $textLen ) {
3683 break; // extra iteration just for above checks
3684 }
3685
3686 # Read the next char...
3687 $ch = $text[$pos];
3688 $lastCh = $pos ? $text[$pos - 1] : '';
3689 $ret .= $ch; // add to result string
3690 if ( $ch == '<' ) {
3691 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3692 $entityState = 0; // for bad HTML
3693 $bracketState = 1; // tag started (checking for backslash)
3694 } elseif ( $ch == '>' ) {
3695 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3696 $entityState = 0; // for bad HTML
3697 $bracketState = 0; // out of brackets
3698 } elseif ( $bracketState == 1 ) {
3699 if ( $ch == '/' ) {
3700 $tagType = 1; // close tag (e.g. "</span>")
3701 } else {
3702 $tagType = 0; // open tag (e.g. "<span>")
3703 $tag .= $ch;
3704 }
3705 $bracketState = 2; // building tag name
3706 } elseif ( $bracketState == 2 ) {
3707 if ( $ch != ' ' ) {
3708 $tag .= $ch;
3709 } else {
3710 // Name found (e.g. "<a href=..."), add on tag attributes...
3711 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3712 }
3713 } elseif ( $bracketState == 0 ) {
3714 if ( $entityState ) {
3715 if ( $ch == ';' ) {
3716 $entityState = 0;
3717 $dispLen++; // entity is one displayed char
3718 }
3719 } else {
3720 if ( $neLength == 0 && !$maybeState ) {
3721 // Save state without $ch. We want to *hit* the first
3722 // display char (to get tags) but not *use* it if truncating.
3723 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3724 }
3725 if ( $ch == '&' ) {
3726 $entityState = 1; // entity found, (e.g. "&#160;")
3727 } else {
3728 $dispLen++; // this char is displayed
3729 // Add the next $max display text chars after this in one swoop...
3730 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3731 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3732 $dispLen += $skipped;
3733 $pos += $skipped;
3734 }
3735 }
3736 }
3737 }
3738 // Close the last tag if left unclosed by bad HTML
3739 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3740 while ( count( $openTags ) > 0 ) {
3741 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3742 }
3743 return $ret;
3744 }
3745
3746 /**
3747 * truncateHtml() helper function
3748 * like strcspn() but adds the skipped chars to $ret
3749 *
3750 * @param string $ret
3751 * @param string $text
3752 * @param string $search
3753 * @param int $start
3754 * @param null|int $len
3755 * @return int
3756 */
3757 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3758 if ( $len === null ) {
3759 $len = -1; // -1 means "no limit" for strcspn
3760 } elseif ( $len < 0 ) {
3761 $len = 0; // sanity
3762 }
3763 $skipCount = 0;
3764 if ( $start < strlen( $text ) ) {
3765 $skipCount = strcspn( $text, $search, $start, $len );
3766 $ret .= substr( $text, $start, $skipCount );
3767 }
3768 return $skipCount;
3769 }
3770
3771 /**
3772 * truncateHtml() helper function
3773 * (a) push or pop $tag from $openTags as needed
3774 * (b) clear $tag value
3775 * @param string &$tag Current HTML tag name we are looking at
3776 * @param int $tagType (0-open tag, 1-close tag)
3777 * @param string $lastCh Character before the '>' that ended this tag
3778 * @param array &$openTags Open tag stack (not accounting for $tag)
3779 */
3780 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3781 $tag = ltrim( $tag );
3782 if ( $tag != '' ) {
3783 if ( $tagType == 0 && $lastCh != '/' ) {
3784 $openTags[] = $tag; // tag opened (didn't close itself)
3785 } elseif ( $tagType == 1 ) {
3786 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3787 array_pop( $openTags ); // tag closed
3788 }
3789 }
3790 $tag = '';
3791 }
3792 }
3793
3794 /**
3795 * Grammatical transformations, needed for inflected languages
3796 * Invoked by putting {{grammar:case|word}} in a message
3797 *
3798 * @param string $word
3799 * @param string $case
3800 * @return string
3801 */
3802 function convertGrammar( $word, $case ) {
3803 global $wgGrammarForms;
3804 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3805 return $wgGrammarForms[$this->getCode()][$case][$word];
3806 }
3807
3808 return $word;
3809 }
3810 /**
3811 * Get the grammar forms for the content language
3812 * @return array Array of grammar forms
3813 * @since 1.20
3814 */
3815 function getGrammarForms() {
3816 global $wgGrammarForms;
3817 if ( isset( $wgGrammarForms[$this->getCode()] )
3818 && is_array( $wgGrammarForms[$this->getCode()] )
3819 ) {
3820 return $wgGrammarForms[$this->getCode()];
3821 }
3822
3823 return [];
3824 }
3825 /**
3826 * Provides an alternative text depending on specified gender.
3827 * Usage {{gender:username|masculine|feminine|unknown}}.
3828 * username is optional, in which case the gender of current user is used,
3829 * but only in (some) interface messages; otherwise default gender is used.
3830 *
3831 * If no forms are given, an empty string is returned. If only one form is
3832 * given, it will be returned unconditionally. These details are implied by
3833 * the caller and cannot be overridden in subclasses.
3834 *
3835 * If three forms are given, the default is to use the third (unknown) form.
3836 * If fewer than three forms are given, the default is to use the first (masculine) form.
3837 * These details can be overridden in subclasses.
3838 *
3839 * @param string $gender
3840 * @param array $forms
3841 *
3842 * @return string
3843 */
3844 function gender( $gender, $forms ) {
3845 if ( !count( $forms ) ) {
3846 return '';
3847 }
3848 $forms = $this->preConvertPlural( $forms, 2 );
3849 if ( $gender === 'male' ) {
3850 return $forms[0];
3851 }
3852 if ( $gender === 'female' ) {
3853 return $forms[1];
3854 }
3855 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3856 }
3857
3858 /**
3859 * Plural form transformations, needed for some languages.
3860 * For example, there are 3 form of plural in Russian and Polish,
3861 * depending on "count mod 10". See [[w:Plural]]
3862 * For English it is pretty simple.
3863 *
3864 * Invoked by putting {{plural:count|wordform1|wordform2}}
3865 * or {{plural:count|wordform1|wordform2|wordform3}}
3866 *
3867 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3868 *
3869 * @param int $count Non-localized number
3870 * @param array $forms Different plural forms
3871 * @return string Correct form of plural for $count in this language
3872 */
3873 function convertPlural( $count, $forms ) {
3874 // Handle explicit n=pluralform cases
3875 $forms = $this->handleExplicitPluralForms( $count, $forms );
3876 if ( is_string( $forms ) ) {
3877 return $forms;
3878 }
3879 if ( !count( $forms ) ) {
3880 return '';
3881 }
3882
3883 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3884 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3885 return $forms[$pluralForm];
3886 }
3887
3888 /**
3889 * Handles explicit plural forms for Language::convertPlural()
3890 *
3891 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3892 * If an explicitly defined plural form matches the $count, then
3893 * string value returned, otherwise array returned for further consideration
3894 * by CLDR rules or overridden convertPlural().
3895 *
3896 * @since 1.23
3897 *
3898 * @param int $count Non-localized number
3899 * @param array $forms Different plural forms
3900 *
3901 * @return array|string
3902 */
3903 protected function handleExplicitPluralForms( $count, array $forms ) {
3904 foreach ( $forms as $index => $form ) {
3905 if ( preg_match( '/\d+=/i', $form ) ) {
3906 $pos = strpos( $form, '=' );
3907 if ( substr( $form, 0, $pos ) === (string)$count ) {
3908 return substr( $form, $pos + 1 );
3909 }
3910 unset( $forms[$index] );
3911 }
3912 }
3913 return array_values( $forms );
3914 }
3915
3916 /**
3917 * Checks that convertPlural was given an array and pads it to requested
3918 * amount of forms by copying the last one.
3919 *
3920 * @param array $forms Array of forms given to convertPlural
3921 * @param int $count How many forms should there be at least
3922 * @return array Padded array of forms or an exception if not an array
3923 */
3924 protected function preConvertPlural( /* Array */ $forms, $count ) {
3925 while ( count( $forms ) < $count ) {
3926 $forms[] = $forms[count( $forms ) - 1];
3927 }
3928 return $forms;
3929 }
3930
3931 /**
3932 * Wraps argument with unicode control characters for directionality safety
3933 *
3934 * This solves the problem where directionality-neutral characters at the edge of
3935 * the argument string get interpreted with the wrong directionality from the
3936 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3937 *
3938 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3939 * directionality of the argument string, using the BIDI algorithm's own "First
3940 * strong directional codepoint" rule. Essentially, this works round the fact that
3941 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3942 * direction inference). The latter is cleaner but still not widely supported.
3943 *
3944 * @param string $text Text to wrap
3945 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3946 */
3947 public function embedBidi( $text = '' ) {
3948 $dir = Language::strongDirFromContent( $text );
3949 if ( $dir === 'ltr' ) {
3950 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3951 return self::$lre . $text . self::$pdf;
3952 }
3953 if ( $dir === 'rtl' ) {
3954 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3955 return self::$rle . $text . self::$pdf;
3956 }
3957 // No strong directionality: do not wrap
3958 return $text;
3959 }
3960
3961 /**
3962 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3963 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3964 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3965 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3966 * match up with it.
3967 *
3968 * @param string $str The validated block duration in English
3969 * @return string Somehow translated block duration
3970 * @see LanguageFi.php for example implementation
3971 */
3972 function translateBlockExpiry( $str ) {
3973 $duration = SpecialBlock::getSuggestedDurations( $this );
3974 foreach ( $duration as $show => $value ) {
3975 if ( strcmp( $str, $value ) == 0 ) {
3976 return htmlspecialchars( trim( $show ) );
3977 }
3978 }
3979
3980 if ( wfIsInfinity( $str ) ) {
3981 foreach ( $duration as $show => $value ) {
3982 if ( wfIsInfinity( $value ) ) {
3983 return htmlspecialchars( trim( $show ) );
3984 }
3985 }
3986 }
3987
3988 // If all else fails, return a standard duration or timestamp description.
3989 $time = strtotime( $str, 0 );
3990 if ( $time === false ) { // Unknown format. Return it as-is in case.
3991 return $str;
3992 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3993 // $time is relative to 0 so it's a duration length.
3994 return $this->formatDuration( $time );
3995 } else { // It's an absolute timestamp.
3996 if ( $time === 0 ) {
3997 // wfTimestamp() handles 0 as current time instead of epoch.
3998 return $this->timeanddate( '19700101000000' );
3999 } else {
4000 return $this->timeanddate( $time );
4001 }
4002 }
4003 }
4004
4005 /**
4006 * languages like Chinese need to be segmented in order for the diff
4007 * to be of any use
4008 *
4009 * @param string $text
4010 * @return string
4011 */
4012 public function segmentForDiff( $text ) {
4013 return $text;
4014 }
4015
4016 /**
4017 * and unsegment to show the result
4018 *
4019 * @param string $text
4020 * @return string
4021 */
4022 public function unsegmentForDiff( $text ) {
4023 return $text;
4024 }
4025
4026 /**
4027 * Return the LanguageConverter used in the Language
4028 *
4029 * @since 1.19
4030 * @return LanguageConverter
4031 */
4032 public function getConverter() {
4033 return $this->mConverter;
4034 }
4035
4036 /**
4037 * convert text to all supported variants
4038 *
4039 * @param string $text
4040 * @return array
4041 */
4042 public function autoConvertToAllVariants( $text ) {
4043 return $this->mConverter->autoConvertToAllVariants( $text );
4044 }
4045
4046 /**
4047 * convert text to different variants of a language.
4048 *
4049 * @param string $text
4050 * @return string
4051 */
4052 public function convert( $text ) {
4053 return $this->mConverter->convert( $text );
4054 }
4055
4056 /**
4057 * Convert a Title object to a string in the preferred variant
4058 *
4059 * @param Title $title
4060 * @return string
4061 */
4062 public function convertTitle( $title ) {
4063 return $this->mConverter->convertTitle( $title );
4064 }
4065
4066 /**
4067 * Convert a namespace index to a string in the preferred variant
4068 *
4069 * @param int $ns
4070 * @return string
4071 */
4072 public function convertNamespace( $ns ) {
4073 return $this->mConverter->convertNamespace( $ns );
4074 }
4075
4076 /**
4077 * Check if this is a language with variants
4078 *
4079 * @return bool
4080 */
4081 public function hasVariants() {
4082 return count( $this->getVariants() ) > 1;
4083 }
4084
4085 /**
4086 * Check if the language has the specific variant
4087 *
4088 * @since 1.19
4089 * @param string $variant
4090 * @return bool
4091 */
4092 public function hasVariant( $variant ) {
4093 return (bool)$this->mConverter->validateVariant( $variant );
4094 }
4095
4096 /**
4097 * Perform output conversion on a string, and encode for safe HTML output.
4098 * @param string $text Text to be converted
4099 * @param bool $isTitle Whether this conversion is for the article title
4100 * @return string
4101 * @todo this should get integrated somewhere sane
4102 */
4103 public function convertHtml( $text, $isTitle = false ) {
4104 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4105 }
4106
4107 /**
4108 * @param string $key
4109 * @return string
4110 */
4111 public function convertCategoryKey( $key ) {
4112 return $this->mConverter->convertCategoryKey( $key );
4113 }
4114
4115 /**
4116 * Get the list of variants supported by this language
4117 * see sample implementation in LanguageZh.php
4118 *
4119 * @return array An array of language codes
4120 */
4121 public function getVariants() {
4122 return $this->mConverter->getVariants();
4123 }
4124
4125 /**
4126 * @return string
4127 */
4128 public function getPreferredVariant() {
4129 return $this->mConverter->getPreferredVariant();
4130 }
4131
4132 /**
4133 * @return string
4134 */
4135 public function getDefaultVariant() {
4136 return $this->mConverter->getDefaultVariant();
4137 }
4138
4139 /**
4140 * @return string
4141 */
4142 public function getURLVariant() {
4143 return $this->mConverter->getURLVariant();
4144 }
4145
4146 /**
4147 * If a language supports multiple variants, it is
4148 * possible that non-existing link in one variant
4149 * actually exists in another variant. this function
4150 * tries to find it. See e.g. LanguageZh.php
4151 * The input parameters may be modified upon return
4152 *
4153 * @param string &$link The name of the link
4154 * @param Title &$nt The title object of the link
4155 * @param bool $ignoreOtherCond To disable other conditions when
4156 * we need to transclude a template or update a category's link
4157 */
4158 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4159 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4160 }
4161
4162 /**
4163 * returns language specific options used by User::getPageRenderHash()
4164 * for example, the preferred language variant
4165 *
4166 * @return string
4167 */
4168 function getExtraHashOptions() {
4169 return $this->mConverter->getExtraHashOptions();
4170 }
4171
4172 /**
4173 * For languages that support multiple variants, the title of an
4174 * article may be displayed differently in different variants. this
4175 * function returns the apporiate title defined in the body of the article.
4176 *
4177 * @return string
4178 */
4179 public function getParsedTitle() {
4180 return $this->mConverter->getParsedTitle();
4181 }
4182
4183 /**
4184 * Refresh the cache of conversion tables when
4185 * MediaWiki:Conversiontable* is updated.
4186 *
4187 * @param Title $title The Title of the page being updated
4188 */
4189 public function updateConversionTable( Title $title ) {
4190 $this->mConverter->updateConversionTable( $title );
4191 }
4192
4193 /**
4194 * Prepare external link text for conversion. When the text is
4195 * a URL, it shouldn't be converted, and it'll be wrapped in
4196 * the "raw" tag (-{R| }-) to prevent conversion.
4197 *
4198 * This function is called "markNoConversion" for historical
4199 * reasons.
4200 *
4201 * @param string $text Text to be used for external link
4202 * @param bool $noParse Wrap it without confirming it's a real URL first
4203 * @return string The tagged text
4204 */
4205 public function markNoConversion( $text, $noParse = false ) {
4206 // Excluding protocal-relative URLs may avoid many false positives.
4207 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4208 return $this->mConverter->markNoConversion( $text );
4209 } else {
4210 return $text;
4211 }
4212 }
4213
4214 /**
4215 * A regular expression to match legal word-trailing characters
4216 * which should be merged onto a link of the form [[foo]]bar.
4217 *
4218 * @return string
4219 */
4220 public function linkTrail() {
4221 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4222 }
4223
4224 /**
4225 * A regular expression character set to match legal word-prefixing
4226 * characters which should be merged onto a link of the form foo[[bar]].
4227 *
4228 * @return string
4229 */
4230 public function linkPrefixCharset() {
4231 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4232 }
4233
4234 /**
4235 * Get the "parent" language which has a converter to convert a "compatible" language
4236 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4237 *
4238 * @return Language|null
4239 * @since 1.22
4240 */
4241 public function getParentLanguage() {
4242 if ( $this->mParentLanguage !== false ) {
4243 return $this->mParentLanguage;
4244 }
4245
4246 $code = explode( '-', $this->getCode() )[0];
4247 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4248 $this->mParentLanguage = null;
4249 return null;
4250 }
4251 $lang = Language::factory( $code );
4252 if ( !$lang->hasVariant( $this->getCode() ) ) {
4253 $this->mParentLanguage = null;
4254 return null;
4255 }
4256
4257 $this->mParentLanguage = $lang;
4258 return $lang;
4259 }
4260
4261 /**
4262 * Get the internal language code for this language object
4263 *
4264 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4265 * htmlspecialchars() or similar
4266 *
4267 * @return string
4268 */
4269 public function getCode() {
4270 return $this->mCode;
4271 }
4272
4273 /**
4274 * Get the code in BCP 47 format which we can use
4275 * inside of html lang="" tags.
4276 *
4277 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4278 * htmlspecialchars() or similar.
4279 *
4280 * @since 1.19
4281 * @return string
4282 */
4283 public function getHtmlCode() {
4284 if ( is_null( $this->mHtmlCode ) ) {
4285 $this->mHtmlCode = wfBCP47( $this->getCode() );
4286 }
4287 return $this->mHtmlCode;
4288 }
4289
4290 /**
4291 * @param string $code
4292 */
4293 public function setCode( $code ) {
4294 $this->mCode = $code;
4295 // Ensure we don't leave incorrect cached data lying around
4296 $this->mHtmlCode = null;
4297 $this->mParentLanguage = false;
4298 }
4299
4300 /**
4301 * Get the language code from a file name. Inverse of getFileName()
4302 * @param string $filename $prefix . $languageCode . $suffix
4303 * @param string $prefix Prefix before the language code
4304 * @param string $suffix Suffix after the language code
4305 * @return string Language code, or false if $prefix or $suffix isn't found
4306 */
4307 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4308 $m = null;
4309 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4310 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4311 if ( !count( $m ) ) {
4312 return false;
4313 }
4314 return str_replace( '_', '-', strtolower( $m[1] ) );
4315 }
4316
4317 /**
4318 * @param string $code
4319 * @return string Name of the language class
4320 */
4321 public static function classFromCode( $code ) {
4322 if ( $code == 'en' ) {
4323 return 'Language';
4324 } else {
4325 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4326 }
4327 }
4328
4329 /**
4330 * Get the name of a file for a certain language code
4331 * @param string $prefix Prepend this to the filename
4332 * @param string $code Language code
4333 * @param string $suffix Append this to the filename
4334 * @throws MWException
4335 * @return string $prefix . $mangledCode . $suffix
4336 */
4337 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4338 if ( !self::isValidBuiltInCode( $code ) ) {
4339 throw new MWException( "Invalid language code \"$code\"" );
4340 }
4341
4342 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4343 }
4344
4345 /**
4346 * @param string $code
4347 * @return string
4348 */
4349 public static function getMessagesFileName( $code ) {
4350 global $IP;
4351 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4352 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4353 return $file;
4354 }
4355
4356 /**
4357 * @param string $code
4358 * @return string
4359 * @throws MWException
4360 * @since 1.23
4361 */
4362 public static function getJsonMessagesFileName( $code ) {
4363 global $IP;
4364
4365 if ( !self::isValidBuiltInCode( $code ) ) {
4366 throw new MWException( "Invalid language code \"$code\"" );
4367 }
4368
4369 return "$IP/languages/i18n/$code.json";
4370 }
4371
4372 /**
4373 * Get the first fallback for a given language.
4374 *
4375 * @param string $code
4376 *
4377 * @return bool|string
4378 */
4379 public static function getFallbackFor( $code ) {
4380 $fallbacks = self::getFallbacksFor( $code );
4381 if ( $fallbacks ) {
4382 return $fallbacks[0];
4383 }
4384 return false;
4385 }
4386
4387 /**
4388 * Get the ordered list of fallback languages.
4389 *
4390 * @since 1.19
4391 * @param string $code Language code
4392 * @return array Non-empty array, ending in "en"
4393 */
4394 public static function getFallbacksFor( $code ) {
4395 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4396 return [];
4397 }
4398 // For unknown languages, fallbackSequence returns an empty array,
4399 // hardcode fallback to 'en' in that case.
4400 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4401 }
4402
4403 /**
4404 * Get the ordered list of fallback languages, ending with the fallback
4405 * language chain for the site language.
4406 *
4407 * @since 1.22
4408 * @param string $code Language code
4409 * @return array Array( fallbacks, site fallbacks )
4410 */
4411 public static function getFallbacksIncludingSiteLanguage( $code ) {
4412 global $wgLanguageCode;
4413
4414 // Usually, we will only store a tiny number of fallback chains, so we
4415 // keep them in static memory.
4416 $cacheKey = "{$code}-{$wgLanguageCode}";
4417
4418 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4419 $fallbacks = self::getFallbacksFor( $code );
4420
4421 // Append the site's fallback chain, including the site language itself
4422 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4423 array_unshift( $siteFallbacks, $wgLanguageCode );
4424
4425 // Eliminate any languages already included in the chain
4426 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4427
4428 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4429 }
4430 return self::$fallbackLanguageCache[$cacheKey];
4431 }
4432
4433 /**
4434 * Get all messages for a given language
4435 * WARNING: this may take a long time. If you just need all message *keys*
4436 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4437 *
4438 * @param string $code
4439 *
4440 * @return array
4441 */
4442 public static function getMessagesFor( $code ) {
4443 return self::getLocalisationCache()->getItem( $code, 'messages' );
4444 }
4445
4446 /**
4447 * Get a message for a given language
4448 *
4449 * @param string $key
4450 * @param string $code
4451 *
4452 * @return string
4453 */
4454 public static function getMessageFor( $key, $code ) {
4455 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4456 }
4457
4458 /**
4459 * Get all message keys for a given language. This is a faster alternative to
4460 * array_keys( Language::getMessagesFor( $code ) )
4461 *
4462 * @since 1.19
4463 * @param string $code Language code
4464 * @return array Array of message keys (strings)
4465 */
4466 public static function getMessageKeysFor( $code ) {
4467 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4468 }
4469
4470 /**
4471 * @param string $talk
4472 * @return mixed
4473 */
4474 function fixVariableInNamespace( $talk ) {
4475 if ( strpos( $talk, '$1' ) === false ) {
4476 return $talk;
4477 }
4478
4479 global $wgMetaNamespace;
4480 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4481
4482 # Allow grammar transformations
4483 # Allowing full message-style parsing would make simple requests
4484 # such as action=raw much more expensive than they need to be.
4485 # This will hopefully cover most cases.
4486 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4487 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4488 return str_replace( ' ', '_', $talk );
4489 }
4490
4491 /**
4492 * @param string $m
4493 * @return string
4494 */
4495 function replaceGrammarInNamespace( $m ) {
4496 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4497 }
4498
4499 /**
4500 * @throws MWException
4501 * @return array
4502 */
4503 static function getCaseMaps() {
4504 static $wikiUpperChars, $wikiLowerChars;
4505 if ( isset( $wikiUpperChars ) ) {
4506 return [ $wikiUpperChars, $wikiLowerChars ];
4507 }
4508
4509 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4510 if ( $arr === false ) {
4511 throw new MWException(
4512 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4513 }
4514 $wikiUpperChars = $arr['wikiUpperChars'];
4515 $wikiLowerChars = $arr['wikiLowerChars'];
4516 return [ $wikiUpperChars, $wikiLowerChars ];
4517 }
4518
4519 /**
4520 * Decode an expiry (block, protection, etc) which has come from the DB
4521 *
4522 * @param string $expiry Database expiry String
4523 * @param bool|int $format True to process using language functions, or TS_ constant
4524 * to return the expiry in a given timestamp
4525 * @param string $infinity If $format is not true, use this string for infinite expiry
4526 * @return string
4527 * @since 1.18
4528 */
4529 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4530 static $dbInfinity;
4531 if ( $dbInfinity === null ) {
4532 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4533 }
4534
4535 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4536 return $format === true
4537 ? $this->getMessageFromDB( 'infiniteblock' )
4538 : $infinity;
4539 } else {
4540 return $format === true
4541 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4542 : wfTimestamp( $format, $expiry );
4543 }
4544 }
4545
4546 /**
4547 * @todo Document
4548 * @param int|float $seconds
4549 * @param array $format Optional
4550 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4551 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4552 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4553 * and friends.
4554 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4555 * or 'avoidminutes'.
4556 * @return string
4557 */
4558 function formatTimePeriod( $seconds, $format = [] ) {
4559 if ( !is_array( $format ) ) {
4560 $format = [ 'avoid' => $format ]; // For backwards compatibility
4561 }
4562 if ( !isset( $format['avoid'] ) ) {
4563 $format['avoid'] = false;
4564 }
4565 if ( !isset( $format['noabbrevs'] ) ) {
4566 $format['noabbrevs'] = false;
4567 }
4568 $secondsMsg = wfMessage(
4569 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4570 $minutesMsg = wfMessage(
4571 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4572 $hoursMsg = wfMessage(
4573 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4574 $daysMsg = wfMessage(
4575 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4576
4577 if ( round( $seconds * 10 ) < 100 ) {
4578 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4579 $s = $secondsMsg->params( $s )->text();
4580 } elseif ( round( $seconds ) < 60 ) {
4581 $s = $this->formatNum( round( $seconds ) );
4582 $s = $secondsMsg->params( $s )->text();
4583 } elseif ( round( $seconds ) < 3600 ) {
4584 $minutes = floor( $seconds / 60 );
4585 $secondsPart = round( fmod( $seconds, 60 ) );
4586 if ( $secondsPart == 60 ) {
4587 $secondsPart = 0;
4588 $minutes++;
4589 }
4590 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4591 $s .= ' ';
4592 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4593 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4594 $hours = floor( $seconds / 3600 );
4595 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4596 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4597 if ( $secondsPart == 60 ) {
4598 $secondsPart = 0;
4599 $minutes++;
4600 }
4601 if ( $minutes == 60 ) {
4602 $minutes = 0;
4603 $hours++;
4604 }
4605 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4606 $s .= ' ';
4607 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4608 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4609 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4610 }
4611 } else {
4612 $days = floor( $seconds / 86400 );
4613 if ( $format['avoid'] === 'avoidminutes' ) {
4614 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4615 if ( $hours == 24 ) {
4616 $hours = 0;
4617 $days++;
4618 }
4619 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4620 $s .= ' ';
4621 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4622 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4623 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4624 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4625 if ( $minutes == 60 ) {
4626 $minutes = 0;
4627 $hours++;
4628 }
4629 if ( $hours == 24 ) {
4630 $hours = 0;
4631 $days++;
4632 }
4633 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4634 $s .= ' ';
4635 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4636 $s .= ' ';
4637 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4638 } else {
4639 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4640 $s .= ' ';
4641 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4642 }
4643 }
4644 return $s;
4645 }
4646
4647 /**
4648 * Format a bitrate for output, using an appropriate
4649 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4650 * the magnitude in question.
4651 *
4652 * This use base 1000. For base 1024 use formatSize(), for another base
4653 * see formatComputingNumbers().
4654 *
4655 * @param int $bps
4656 * @return string
4657 */
4658 function formatBitrate( $bps ) {
4659 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4660 }
4661
4662 /**
4663 * @param int $size Size of the unit
4664 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4665 * @param string $messageKey Message key to be uesd
4666 * @return string
4667 */
4668 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4669 if ( $size <= 0 ) {
4670 return str_replace( '$1', $this->formatNum( $size ),
4671 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4672 );
4673 }
4674 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4675 $index = 0;
4676
4677 $maxIndex = count( $sizes ) - 1;
4678 while ( $size >= $boundary && $index < $maxIndex ) {
4679 $index++;
4680 $size /= $boundary;
4681 }
4682
4683 // For small sizes no decimal places necessary
4684 $round = 0;
4685 if ( $index > 1 ) {
4686 // For MB and bigger two decimal places are smarter
4687 $round = 2;
4688 }
4689 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4690
4691 $size = round( $size, $round );
4692 $text = $this->getMessageFromDB( $msg );
4693 return str_replace( '$1', $this->formatNum( $size ), $text );
4694 }
4695
4696 /**
4697 * Format a size in bytes for output, using an appropriate
4698 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4699 *
4700 * This method use base 1024. For base 1000 use formatBitrate(), for
4701 * another base see formatComputingNumbers()
4702 *
4703 * @param int $size Size to format
4704 * @return string Plain text (not HTML)
4705 */
4706 function formatSize( $size ) {
4707 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4708 }
4709
4710 /**
4711 * Make a list item, used by various special pages
4712 *
4713 * @param string $page Page link
4714 * @param string $details HTML safe text between brackets
4715 * @param bool $oppositedm Add the direction mark opposite to your
4716 * language, to display text properly
4717 * @return HTML escaped string
4718 */
4719 function specialList( $page, $details, $oppositedm = true ) {
4720 if ( !$details ) {
4721 return $page;
4722 }
4723
4724 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4725 return
4726 $page .
4727 $dirmark .
4728 $this->msg( 'word-separator' )->escaped() .
4729 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4730 }
4731
4732 /**
4733 * Generate (prev x| next x) (20|50|100...) type links for paging
4734 *
4735 * @param Title $title Title object to link
4736 * @param int $offset
4737 * @param int $limit
4738 * @param array $query Optional URL query parameter string
4739 * @param bool $atend Optional param for specified if this is the last page
4740 * @return string
4741 */
4742 public function viewPrevNext( Title $title, $offset, $limit,
4743 array $query = [], $atend = false
4744 ) {
4745 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4746
4747 # Make 'previous' link
4748 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4749 if ( $offset > 0 ) {
4750 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4751 $query, $prev, 'prevn-title', 'mw-prevlink' );
4752 } else {
4753 $plink = htmlspecialchars( $prev );
4754 }
4755
4756 # Make 'next' link
4757 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4758 if ( $atend ) {
4759 $nlink = htmlspecialchars( $next );
4760 } else {
4761 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4762 $query, $next, 'nextn-title', 'mw-nextlink' );
4763 }
4764
4765 # Make links to set number of items per page
4766 $numLinks = [];
4767 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4768 $numLinks[] = $this->numLink( $title, $offset, $num,
4769 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4770 }
4771
4772 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4773 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4774 }
4775
4776 /**
4777 * Helper function for viewPrevNext() that generates links
4778 *
4779 * @param Title $title Title object to link
4780 * @param int $offset
4781 * @param int $limit
4782 * @param array $query Extra query parameters
4783 * @param string $link Text to use for the link; will be escaped
4784 * @param string $tooltipMsg Name of the message to use as tooltip
4785 * @param string $class Value of the "class" attribute of the link
4786 * @return string HTML fragment
4787 */
4788 private function numLink( Title $title, $offset, $limit, array $query, $link,
4789 $tooltipMsg, $class
4790 ) {
4791 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4792 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4793 ->numParams( $limit )->text();
4794
4795 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4796 'title' => $tooltip, 'class' => $class ], $link );
4797 }
4798
4799 /**
4800 * Get the conversion rule title, if any.
4801 *
4802 * @return string
4803 */
4804 public function getConvRuleTitle() {
4805 return $this->mConverter->getConvRuleTitle();
4806 }
4807
4808 /**
4809 * Get the compiled plural rules for the language
4810 * @since 1.20
4811 * @return array Associative array with plural form, and plural rule as key-value pairs
4812 */
4813 public function getCompiledPluralRules() {
4814 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4815 $fallbacks = Language::getFallbacksFor( $this->mCode );
4816 if ( !$pluralRules ) {
4817 foreach ( $fallbacks as $fallbackCode ) {
4818 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4819 if ( $pluralRules ) {
4820 break;
4821 }
4822 }
4823 }
4824 return $pluralRules;
4825 }
4826
4827 /**
4828 * Get the plural rules for the language
4829 * @since 1.20
4830 * @return array Associative array with plural form number and plural rule as key-value pairs
4831 */
4832 public function getPluralRules() {
4833 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4834 $fallbacks = Language::getFallbacksFor( $this->mCode );
4835 if ( !$pluralRules ) {
4836 foreach ( $fallbacks as $fallbackCode ) {
4837 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4838 if ( $pluralRules ) {
4839 break;
4840 }
4841 }
4842 }
4843 return $pluralRules;
4844 }
4845
4846 /**
4847 * Get the plural rule types for the language
4848 * @since 1.22
4849 * @return array Associative array with plural form number and plural rule type as key-value pairs
4850 */
4851 public function getPluralRuleTypes() {
4852 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4853 $fallbacks = Language::getFallbacksFor( $this->mCode );
4854 if ( !$pluralRuleTypes ) {
4855 foreach ( $fallbacks as $fallbackCode ) {
4856 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4857 if ( $pluralRuleTypes ) {
4858 break;
4859 }
4860 }
4861 }
4862 return $pluralRuleTypes;
4863 }
4864
4865 /**
4866 * Find the index number of the plural rule appropriate for the given number
4867 * @param int $number
4868 * @return int The index number of the plural rule
4869 */
4870 public function getPluralRuleIndexNumber( $number ) {
4871 $pluralRules = $this->getCompiledPluralRules();
4872 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4873 return $form;
4874 }
4875
4876 /**
4877 * Find the plural rule type appropriate for the given number
4878 * For example, if the language is set to Arabic, getPluralType(5) should
4879 * return 'few'.
4880 * @since 1.22
4881 * @param int $number
4882 * @return string The name of the plural rule type, e.g. one, two, few, many
4883 */
4884 public function getPluralRuleType( $number ) {
4885 $index = $this->getPluralRuleIndexNumber( $number );
4886 $pluralRuleTypes = $this->getPluralRuleTypes();
4887 if ( isset( $pluralRuleTypes[$index] ) ) {
4888 return $pluralRuleTypes[$index];
4889 } else {
4890 return 'other';
4891 }
4892 }
4893 }
MediaWiki Core