dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "Simplify translatewiki.net links from Special:AllMessages"
[lhc/web/wiklou.git] / languages / Language.php
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 use CLDRPluralRuleParser\Evaluator;
34
35 /**
36 * Internationalisation code
37 * @ingroup Language
38 */
39 class Language {
40 /**
41 * @var LanguageConverter
42 */
43 public $mConverter;
44
45 public $mVariants, $mCode, $mLoaded = false;
46 public $mMagicExtensions = [], $mMagicHookDone = false;
47 private $mHtmlCode = null, $mParentLanguage = false;
48
49 public $dateFormatStrings = [];
50 public $mExtendedSpecialPageAliases;
51
52 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
53
54 /**
55 * ReplacementArray object caches
56 */
57 public $transformData = [];
58
59 /**
60 * @var LocalisationCache
61 */
62 static public $dataCache;
63
64 static public $mLangObjCache = [];
65
66 static public $mWeekdayMsgs = [
67 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
68 'friday', 'saturday'
69 ];
70
71 static public $mWeekdayAbbrevMsgs = [
72 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
73 ];
74
75 static public $mMonthMsgs = [
76 'january', 'february', 'march', 'april', 'may_long', 'june',
77 'july', 'august', 'september', 'october', 'november',
78 'december'
79 ];
80 static public $mMonthGenMsgs = [
81 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
82 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
83 'december-gen'
84 ];
85 static public $mMonthAbbrevMsgs = [
86 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
87 'sep', 'oct', 'nov', 'dec'
88 ];
89
90 static public $mIranianCalendarMonthMsgs = [
91 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
92 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
93 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
94 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
95 ];
96
97 static public $mHebrewCalendarMonthMsgs = [
98 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
99 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
100 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
101 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
102 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
103 ];
104
105 static public $mHebrewCalendarMonthGenMsgs = [
106 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
107 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
108 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
109 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
110 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
111 ];
112
113 static public $mHijriCalendarMonthMsgs = [
114 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
115 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
116 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
117 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
118 ];
119
120 /**
121 * @since 1.20
122 * @var array
123 */
124 static public $durationIntervals = [
125 'millennia' => 31556952000,
126 'centuries' => 3155695200,
127 'decades' => 315569520,
128 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
129 'weeks' => 604800,
130 'days' => 86400,
131 'hours' => 3600,
132 'minutes' => 60,
133 'seconds' => 1,
134 ];
135
136 /**
137 * Cache for language fallbacks.
138 * @see Language::getFallbacksIncludingSiteLanguage
139 * @since 1.21
140 * @var array
141 */
142 static private $fallbackLanguageCache = [];
143
144 /**
145 * Cache for language names
146 * @var HashBagOStuff|null
147 */
148 static private $languageNameCache;
149
150 /**
151 * Unicode directional formatting characters, for embedBidi()
152 */
153 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
154 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
155 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
156
157 /**
158 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
159 * - in group 1 if it is LTR
160 * - in group 2 if it is RTL
161 * Does not match if there is no strong directionality codepoint.
162 *
163 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
164 *
165 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
166 * https://git.wikimedia.org/summary/unicodejs.git .
167 */
168 // @codingStandardsIgnoreStart
169 // @codeCoverageIgnoreStart
170 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';
171 // @codeCoverageIgnoreEnd
172 // @codingStandardsIgnoreEnd
173
174 /**
175 * Get a cached or new language object for a given language code
176 * @param string $code
177 * @return Language
178 */
179 static function factory( $code ) {
180 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
181
182 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
183 $code = $wgDummyLanguageCodes[$code];
184 }
185
186 // get the language object to process
187 $langObj = isset( self::$mLangObjCache[$code] )
188 ? self::$mLangObjCache[$code]
189 : self::newFromCode( $code );
190
191 // merge the language object in to get it up front in the cache
192 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
193 // get rid of the oldest ones in case we have an overflow
194 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
195
196 return $langObj;
197 }
198
199 /**
200 * Create a language object for a given language code
201 * @param string $code
202 * @throws MWException
203 * @return Language
204 */
205 protected static function newFromCode( $code ) {
206 if ( !Language::isValidCode( $code ) ) {
207 throw new MWException( "Invalid language code \"$code\"" );
208 }
209
210 if ( !Language::isValidBuiltInCode( $code ) ) {
211 // It's not possible to customise this code with class files, so
212 // just return a Language object. This is to support uselang= hacks.
213 $lang = new Language;
214 $lang->setCode( $code );
215 return $lang;
216 }
217
218 // Check if there is a language class for the code
219 $class = self::classFromCode( $code );
220 if ( class_exists( $class ) ) {
221 $lang = new $class;
222 return $lang;
223 }
224
225 // Keep trying the fallback list until we find an existing class
226 $fallbacks = Language::getFallbacksFor( $code );
227 foreach ( $fallbacks as $fallbackCode ) {
228 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
229 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
230 }
231
232 $class = self::classFromCode( $fallbackCode );
233 if ( class_exists( $class ) ) {
234 $lang = new $class;
235 $lang->setCode( $code );
236 return $lang;
237 }
238 }
239
240 throw new MWException( "Invalid fallback sequence for language '$code'" );
241 }
242
243 /**
244 * Checks whether any localisation is available for that language tag
245 * in MediaWiki (MessagesXx.php exists).
246 *
247 * @param string $code Language tag (in lower case)
248 * @return bool Whether language is supported
249 * @since 1.21
250 */
251 public static function isSupportedLanguage( $code ) {
252 if ( !self::isValidBuiltInCode( $code ) ) {
253 return false;
254 }
255
256 if ( $code === 'qqq' ) {
257 return false;
258 }
259
260 return is_readable( self::getMessagesFileName( $code ) ) ||
261 is_readable( self::getJsonMessagesFileName( $code ) );
262 }
263
264 /**
265 * Returns true if a language code string is a well-formed language tag
266 * according to RFC 5646.
267 * This function only checks well-formedness; it doesn't check that
268 * language, script or variant codes actually exist in the repositories.
269 *
270 * Based on regexes by Mark Davis of the Unicode Consortium:
271 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
272 *
273 * @param string $code
274 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
275 *
276 * @return bool
277 * @since 1.21
278 */
279 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
280 $alpha = '[a-z]';
281 $digit = '[0-9]';
282 $alphanum = '[a-z0-9]';
283 $x = 'x'; # private use singleton
284 $singleton = '[a-wy-z]'; # other singleton
285 $s = $lenient ? '[-_]' : '-';
286
287 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
288 $script = "$alpha{4}"; # ISO 15924
289 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
290 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
291 $extension = "$singleton(?:$s$alphanum{2,8})+";
292 $privateUse = "$x(?:$s$alphanum{1,8})+";
293
294 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
295 # Since these are limited, this is safe even later changes to the registry --
296 # the only oddity is that it might change the type of the tag, and thus
297 # the results from the capturing groups.
298 # http://www.iana.org/assignments/language-subtag-registry
299
300 $grandfathered = "en{$s}GB{$s}oed"
301 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
302 . "|no{$s}(?:bok|nyn)"
303 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
304 . "|zh{$s}min{$s}nan";
305
306 $variantList = "$variant(?:$s$variant)*";
307 $extensionList = "$extension(?:$s$extension)*";
308
309 $langtag = "(?:($language)"
310 . "(?:$s$script)?"
311 . "(?:$s$region)?"
312 . "(?:$s$variantList)?"
313 . "(?:$s$extensionList)?"
314 . "(?:$s$privateUse)?)";
315
316 # The final breakdown, with capturing groups for each of these components
317 # The variants, extensions, grandfathered, and private-use may have interior '-'
318
319 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
320
321 return (bool)preg_match( "/$root/", strtolower( $code ) );
322 }
323
324 /**
325 * Returns true if a language code string is of a valid form, whether or
326 * not it exists. This includes codes which are used solely for
327 * customisation via the MediaWiki namespace.
328 *
329 * @param string $code
330 *
331 * @return bool
332 */
333 public static function isValidCode( $code ) {
334 static $cache = [];
335 if ( !isset( $cache[$code] ) ) {
336 // People think language codes are html safe, so enforce it.
337 // Ideally we should only allow a-zA-Z0-9-
338 // but, .+ and other chars are often used for {{int:}} hacks
339 // see bugs 37564, 37587, 36938
340 $cache[$code] =
341 // Protect against path traversal
342 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
343 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
344 }
345 return $cache[$code];
346 }
347
348 /**
349 * Returns true if a language code is of a valid form for the purposes of
350 * internal customisation of MediaWiki, via Messages*.php or *.json.
351 *
352 * @param string $code
353 *
354 * @throws MWException
355 * @since 1.18
356 * @return bool
357 */
358 public static function isValidBuiltInCode( $code ) {
359
360 if ( !is_string( $code ) ) {
361 if ( is_object( $code ) ) {
362 $addmsg = " of class " . get_class( $code );
363 } else {
364 $addmsg = '';
365 }
366 $type = gettype( $code );
367 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
368 }
369
370 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
371 }
372
373 /**
374 * Returns true if a language code is an IETF tag known to MediaWiki.
375 *
376 * @param string $tag
377 *
378 * @since 1.21
379 * @return bool
380 */
381 public static function isKnownLanguageTag( $tag ) {
382 // Quick escape for invalid input to avoid exceptions down the line
383 // when code tries to process tags which are not valid at all.
384 if ( !self::isValidBuiltInCode( $tag ) ) {
385 return false;
386 }
387
388 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
389 || self::fetchLanguageName( $tag, $tag ) !== ''
390 ) {
391 return true;
392 }
393
394 return false;
395 }
396
397 /**
398 * Get the LocalisationCache instance
399 *
400 * @return LocalisationCache
401 */
402 public static function getLocalisationCache() {
403 if ( is_null( self::$dataCache ) ) {
404 global $wgLocalisationCacheConf;
405 $class = $wgLocalisationCacheConf['class'];
406 self::$dataCache = new $class( $wgLocalisationCacheConf );
407 }
408 return self::$dataCache;
409 }
410
411 function __construct() {
412 $this->mConverter = new FakeConverter( $this );
413 // Set the code to the name of the descendant
414 if ( get_class( $this ) == 'Language' ) {
415 $this->mCode = 'en';
416 } else {
417 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
418 }
419 self::getLocalisationCache();
420 }
421
422 /**
423 * Reduce memory usage
424 */
425 function __destruct() {
426 foreach ( $this as $name => $value ) {
427 unset( $this->$name );
428 }
429 }
430
431 /**
432 * Hook which will be called if this is the content language.
433 * Descendants can use this to register hook functions or modify globals
434 */
435 function initContLang() {
436 }
437
438 /**
439 * @return array
440 * @since 1.19
441 */
442 public function getFallbackLanguages() {
443 return self::getFallbacksFor( $this->mCode );
444 }
445
446 /**
447 * Exports $wgBookstoreListEn
448 * @return array
449 */
450 public function getBookstoreList() {
451 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
452 }
453
454 /**
455 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
456 * available in localised form, it will be included in English.
457 *
458 * @return array
459 */
460 public function getNamespaces() {
461 if ( is_null( $this->namespaceNames ) ) {
462 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
463
464 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
465 $validNamespaces = MWNamespace::getCanonicalNamespaces();
466
467 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
468
469 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
470 if ( $wgMetaNamespaceTalk ) {
471 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
472 } else {
473 $talk = $this->namespaceNames[NS_PROJECT_TALK];
474 $this->namespaceNames[NS_PROJECT_TALK] =
475 $this->fixVariableInNamespace( $talk );
476 }
477
478 # Sometimes a language will be localised but not actually exist on this wiki.
479 foreach ( $this->namespaceNames as $key => $text ) {
480 if ( !isset( $validNamespaces[$key] ) ) {
481 unset( $this->namespaceNames[$key] );
482 }
483 }
484
485 # The above mixing may leave namespaces out of canonical order.
486 # Re-order by namespace ID number...
487 ksort( $this->namespaceNames );
488
489 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
490 }
491
492 return $this->namespaceNames;
493 }
494
495 /**
496 * Arbitrarily set all of the namespace names at once. Mainly used for testing
497 * @param array $namespaces Array of namespaces (id => name)
498 */
499 public function setNamespaces( array $namespaces ) {
500 $this->namespaceNames = $namespaces;
501 $this->mNamespaceIds = null;
502 }
503
504 /**
505 * Resets all of the namespace caches. Mainly used for testing
506 */
507 public function resetNamespaces() {
508 $this->namespaceNames = null;
509 $this->mNamespaceIds = null;
510 $this->namespaceAliases = null;
511 }
512
513 /**
514 * A convenience function that returns getNamespaces() with spaces instead of underscores
515 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
516 *
517 * @return array
518 */
519 public function getFormattedNamespaces() {
520 $ns = $this->getNamespaces();
521 foreach ( $ns as $k => $v ) {
522 $ns[$k] = strtr( $v, '_', ' ' );
523 }
524 return $ns;
525 }
526
527 /**
528 * Get a namespace value by key
529 *
530 * <code>
531 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
532 * echo $mw_ns; // prints 'MediaWiki'
533 * </code>
534 *
535 * @param int $index The array key of the namespace to return
536 * @return string|bool String if the namespace value exists, otherwise false
537 */
538 public function getNsText( $index ) {
539 $ns = $this->getNamespaces();
540 return isset( $ns[$index] ) ? $ns[$index] : false;
541 }
542
543 /**
544 * A convenience function that returns the same thing as
545 * getNsText() except with '_' changed to ' ', useful for
546 * producing output.
547 *
548 * <code>
549 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
550 * echo $mw_ns; // prints 'MediaWiki talk'
551 * </code>
552 *
553 * @param int $index The array key of the namespace to return
554 * @return string Namespace name without underscores (empty string if namespace does not exist)
555 */
556 public function getFormattedNsText( $index ) {
557 $ns = $this->getNsText( $index );
558 return strtr( $ns, '_', ' ' );
559 }
560
561 /**
562 * Returns gender-dependent namespace alias if available.
563 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
564 * @param int $index Namespace index
565 * @param string $gender Gender key (male, female... )
566 * @return string
567 * @since 1.18
568 */
569 public function getGenderNsText( $index, $gender ) {
570 global $wgExtraGenderNamespaces;
571
572 $ns = $wgExtraGenderNamespaces +
573 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
574
575 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
576 }
577
578 /**
579 * Whether this language uses gender-dependent namespace aliases.
580 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
581 * @return bool
582 * @since 1.18
583 */
584 public function needsGenderDistinction() {
585 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
586 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
587 // $wgExtraGenderNamespaces overrides everything
588 return true;
589 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
590 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
591 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
592 return false;
593 } else {
594 // Check what is in i18n files
595 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
596 return count( $aliases ) > 0;
597 }
598 }
599
600 /**
601 * Get a namespace key by value, case insensitive.
602 * Only matches namespace names for the current language, not the
603 * canonical ones defined in Namespace.php.
604 *
605 * @param string $text
606 * @return int|bool An integer if $text is a valid value otherwise false
607 */
608 function getLocalNsIndex( $text ) {
609 $lctext = $this->lc( $text );
610 $ids = $this->getNamespaceIds();
611 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
612 }
613
614 /**
615 * @return array
616 */
617 public function getNamespaceAliases() {
618 if ( is_null( $this->namespaceAliases ) ) {
619 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
620 if ( !$aliases ) {
621 $aliases = [];
622 } else {
623 foreach ( $aliases as $name => $index ) {
624 if ( $index === NS_PROJECT_TALK ) {
625 unset( $aliases[$name] );
626 $name = $this->fixVariableInNamespace( $name );
627 $aliases[$name] = $index;
628 }
629 }
630 }
631
632 global $wgExtraGenderNamespaces;
633 $genders = $wgExtraGenderNamespaces +
634 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
635 foreach ( $genders as $index => $forms ) {
636 foreach ( $forms as $alias ) {
637 $aliases[$alias] = $index;
638 }
639 }
640
641 # Also add converted namespace names as aliases, to avoid confusion.
642 $convertedNames = [];
643 foreach ( $this->getVariants() as $variant ) {
644 if ( $variant === $this->mCode ) {
645 continue;
646 }
647 foreach ( $this->getNamespaces() as $ns => $_ ) {
648 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
649 }
650 }
651
652 $this->namespaceAliases = $aliases + $convertedNames;
653 }
654
655 return $this->namespaceAliases;
656 }
657
658 /**
659 * @return array
660 */
661 public function getNamespaceIds() {
662 if ( is_null( $this->mNamespaceIds ) ) {
663 global $wgNamespaceAliases;
664 # Put namespace names and aliases into a hashtable.
665 # If this is too slow, then we should arrange it so that it is done
666 # before caching. The catch is that at pre-cache time, the above
667 # class-specific fixup hasn't been done.
668 $this->mNamespaceIds = [];
669 foreach ( $this->getNamespaces() as $index => $name ) {
670 $this->mNamespaceIds[$this->lc( $name )] = $index;
671 }
672 foreach ( $this->getNamespaceAliases() as $name => $index ) {
673 $this->mNamespaceIds[$this->lc( $name )] = $index;
674 }
675 if ( $wgNamespaceAliases ) {
676 foreach ( $wgNamespaceAliases as $name => $index ) {
677 $this->mNamespaceIds[$this->lc( $name )] = $index;
678 }
679 }
680 }
681 return $this->mNamespaceIds;
682 }
683
684 /**
685 * Get a namespace key by value, case insensitive. Canonical namespace
686 * names override custom ones defined for the current language.
687 *
688 * @param string $text
689 * @return int|bool An integer if $text is a valid value otherwise false
690 */
691 public function getNsIndex( $text ) {
692 $lctext = $this->lc( $text );
693 $ns = MWNamespace::getCanonicalIndex( $lctext );
694 if ( $ns !== null ) {
695 return $ns;
696 }
697 $ids = $this->getNamespaceIds();
698 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
699 }
700
701 /**
702 * short names for language variants used for language conversion links.
703 *
704 * @param string $code
705 * @param bool $usemsg Use the "variantname-xyz" message if it exists
706 * @return string
707 */
708 public function getVariantname( $code, $usemsg = true ) {
709 $msg = "variantname-$code";
710 if ( $usemsg && wfMessage( $msg )->exists() ) {
711 return $this->getMessageFromDB( $msg );
712 }
713 $name = self::fetchLanguageName( $code );
714 if ( $name ) {
715 return $name; # if it's defined as a language name, show that
716 } else {
717 # otherwise, output the language code
718 return $code;
719 }
720 }
721
722 /**
723 * @return array
724 */
725 public function getDatePreferences() {
726 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
727 }
728
729 /**
730 * @return array
731 */
732 function getDateFormats() {
733 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
734 }
735
736 /**
737 * @return array|string
738 */
739 public function getDefaultDateFormat() {
740 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
741 if ( $df === 'dmy or mdy' ) {
742 global $wgAmericanDates;
743 return $wgAmericanDates ? 'mdy' : 'dmy';
744 } else {
745 return $df;
746 }
747 }
748
749 /**
750 * @return array
751 */
752 public function getDatePreferenceMigrationMap() {
753 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
754 }
755
756 /**
757 * @param string $image
758 * @return array|null
759 */
760 function getImageFile( $image ) {
761 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
762 }
763
764 /**
765 * @return array
766 * @since 1.24
767 */
768 public function getImageFiles() {
769 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
770 }
771
772 /**
773 * @return array
774 */
775 public function getExtraUserToggles() {
776 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
777 }
778
779 /**
780 * @param string $tog
781 * @return string
782 */
783 function getUserToggle( $tog ) {
784 return $this->getMessageFromDB( "tog-$tog" );
785 }
786
787 /**
788 * Get an array of language names, indexed by code.
789 * @param null|string $inLanguage Code of language in which to return the names
790 * Use null for autonyms (native names)
791 * @param string $include One of:
792 * 'all' all available languages
793 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
794 * 'mwfile' only if the language is in 'mw' *and* has a message file
795 * @return array Language code => language name
796 * @since 1.20
797 */
798 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
799 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
800 $cacheKey .= ":$include";
801 if ( self::$languageNameCache === null ) {
802 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
803 }
804
805 $ret = self::$languageNameCache->get( $cacheKey );
806 if ( !$ret ) {
807 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
808 self::$languageNameCache->set( $cacheKey, $ret );
809 }
810 return $ret;
811 }
812
813 /**
814 * Uncached helper for fetchLanguageNames
815 * @param null|string $inLanguage Code of language in which to return the names
816 * Use null for autonyms (native names)
817 * @param string $include One of:
818 * 'all' all available languages
819 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
820 * 'mwfile' only if the language is in 'mw' *and* has a message file
821 * @return array Language code => language name
822 */
823 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
824 global $wgExtraLanguageNames;
825
826 // If passed an invalid language code to use, fallback to en
827 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
828 $inLanguage = 'en';
829 }
830
831 $names = [];
832
833 if ( $inLanguage ) {
834 # TODO: also include when $inLanguage is null, when this code is more efficient
835 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
836 }
837
838 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
839 foreach ( $mwNames as $mwCode => $mwName ) {
840 # - Prefer own MediaWiki native name when not using the hook
841 # - For other names just add if not added through the hook
842 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
843 $names[$mwCode] = $mwName;
844 }
845 }
846
847 if ( $include === 'all' ) {
848 ksort( $names );
849 return $names;
850 }
851
852 $returnMw = [];
853 $coreCodes = array_keys( $mwNames );
854 foreach ( $coreCodes as $coreCode ) {
855 $returnMw[$coreCode] = $names[$coreCode];
856 }
857
858 if ( $include === 'mwfile' ) {
859 $namesMwFile = [];
860 # We do this using a foreach over the codes instead of a directory
861 # loop so that messages files in extensions will work correctly.
862 foreach ( $returnMw as $code => $value ) {
863 if ( is_readable( self::getMessagesFileName( $code ) )
864 || is_readable( self::getJsonMessagesFileName( $code ) )
865 ) {
866 $namesMwFile[$code] = $names[$code];
867 }
868 }
869
870 ksort( $namesMwFile );
871 return $namesMwFile;
872 }
873
874 ksort( $returnMw );
875 # 'mw' option; default if it's not one of the other two options (all/mwfile)
876 return $returnMw;
877 }
878
879 /**
880 * @param string $code The code of the language for which to get the name
881 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
882 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
883 * @return string Language name or empty
884 * @since 1.20
885 */
886 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
887 $code = strtolower( $code );
888 $array = self::fetchLanguageNames( $inLanguage, $include );
889 return !array_key_exists( $code, $array ) ? '' : $array[$code];
890 }
891
892 /**
893 * Get a message from the MediaWiki namespace.
894 *
895 * @param string $msg Message name
896 * @return string
897 */
898 public function getMessageFromDB( $msg ) {
899 return $this->msg( $msg )->text();
900 }
901
902 /**
903 * Get message object in this language. Only for use inside this class.
904 *
905 * @param string $msg Message name
906 * @return Message
907 */
908 protected function msg( $msg ) {
909 return wfMessage( $msg )->inLanguage( $this );
910 }
911
912 /**
913 * @param string $key
914 * @return string
915 */
916 public function getMonthName( $key ) {
917 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
918 }
919
920 /**
921 * @return array
922 */
923 public function getMonthNamesArray() {
924 $monthNames = [ '' ];
925 for ( $i = 1; $i < 13; $i++ ) {
926 $monthNames[] = $this->getMonthName( $i );
927 }
928 return $monthNames;
929 }
930
931 /**
932 * @param string $key
933 * @return string
934 */
935 public function getMonthNameGen( $key ) {
936 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
937 }
938
939 /**
940 * @param string $key
941 * @return string
942 */
943 public function getMonthAbbreviation( $key ) {
944 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
945 }
946
947 /**
948 * @return array
949 */
950 public function getMonthAbbreviationsArray() {
951 $monthNames = [ '' ];
952 for ( $i = 1; $i < 13; $i++ ) {
953 $monthNames[] = $this->getMonthAbbreviation( $i );
954 }
955 return $monthNames;
956 }
957
958 /**
959 * @param string $key
960 * @return string
961 */
962 public function getWeekdayName( $key ) {
963 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
964 }
965
966 /**
967 * @param string $key
968 * @return string
969 */
970 function getWeekdayAbbreviation( $key ) {
971 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
972 }
973
974 /**
975 * @param string $key
976 * @return string
977 */
978 function getIranianCalendarMonthName( $key ) {
979 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
980 }
981
982 /**
983 * @param string $key
984 * @return string
985 */
986 function getHebrewCalendarMonthName( $key ) {
987 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
988 }
989
990 /**
991 * @param string $key
992 * @return string
993 */
994 function getHebrewCalendarMonthNameGen( $key ) {
995 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
996 }
997
998 /**
999 * @param string $key
1000 * @return string
1001 */
1002 function getHijriCalendarMonthName( $key ) {
1003 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1004 }
1005
1006 /**
1007 * Pass through result from $dateTimeObj->format()
1008 * @param DateTime|bool|null &$dateTimeObj
1009 * @param string $ts
1010 * @param DateTimeZone|bool|null $zone
1011 * @param string $code
1012 * @return string
1013 */
1014 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1015 if ( !$dateTimeObj ) {
1016 $dateTimeObj = DateTime::createFromFormat(
1017 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1018 );
1019 }
1020 return $dateTimeObj->format( $code );
1021 }
1022
1023 /**
1024 * This is a workalike of PHP's date() function, but with better
1025 * internationalisation, a reduced set of format characters, and a better
1026 * escaping format.
1027 *
1028 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1029 * the PHP manual for definitions. There are a number of extensions, which
1030 * start with "x":
1031 *
1032 * xn Do not translate digits of the next numeric format character
1033 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1034 * xr Use roman numerals for the next numeric format character
1035 * xh Use hebrew numerals for the next numeric format character
1036 * xx Literal x
1037 * xg Genitive month name
1038 *
1039 * xij j (day number) in Iranian calendar
1040 * xiF F (month name) in Iranian calendar
1041 * xin n (month number) in Iranian calendar
1042 * xiy y (two digit year) in Iranian calendar
1043 * xiY Y (full year) in Iranian calendar
1044 *
1045 * xjj j (day number) in Hebrew calendar
1046 * xjF F (month name) in Hebrew calendar
1047 * xjt t (days in month) in Hebrew calendar
1048 * xjx xg (genitive month name) in Hebrew calendar
1049 * xjn n (month number) in Hebrew calendar
1050 * xjY Y (full year) in Hebrew calendar
1051 *
1052 * xmj j (day number) in Hijri calendar
1053 * xmF F (month name) in Hijri calendar
1054 * xmn n (month number) in Hijri calendar
1055 * xmY Y (full year) in Hijri calendar
1056 *
1057 * xkY Y (full year) in Thai solar calendar. Months and days are
1058 * identical to the Gregorian calendar
1059 * xoY Y (full year) in Minguo calendar or Juche year.
1060 * Months and days are identical to the
1061 * Gregorian calendar
1062 * xtY Y (full year) in Japanese nengo. Months and days are
1063 * identical to the Gregorian calendar
1064 *
1065 * Characters enclosed in double quotes will be considered literal (with
1066 * the quotes themselves removed). Unmatched quotes will be considered
1067 * literal quotes. Example:
1068 *
1069 * "The month is" F => The month is January
1070 * i's" => 20'11"
1071 *
1072 * Backslash escaping is also supported.
1073 *
1074 * Input timestamp is assumed to be pre-normalized to the desired local
1075 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1076 * $ts is UTC if $zone is not given.
1077 *
1078 * @param string $format
1079 * @param string $ts 14-character timestamp
1080 * YYYYMMDDHHMMSS
1081 * 01234567890123
1082 * @param DateTimeZone $zone Timezone of $ts
1083 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1084 * Only makes sense if $ts is the current time.
1085 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1086 *
1087 * @throws MWException
1088 * @return string
1089 */
1090 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = null ) {
1091 $s = '';
1092 $raw = false;
1093 $roman = false;
1094 $hebrewNum = false;
1095 $dateTimeObj = false;
1096 $rawToggle = false;
1097 $iranian = false;
1098 $hebrew = false;
1099 $hijri = false;
1100 $thai = false;
1101 $minguo = false;
1102 $tenno = false;
1103
1104 $usedSecond = false;
1105 $usedMinute = false;
1106 $usedHour = false;
1107 $usedAMPM = false;
1108 $usedDay = false;
1109 $usedWeek = false;
1110 $usedMonth = false;
1111 $usedYear = false;
1112 $usedISOYear = false;
1113 $usedIsLeapYear = false;
1114
1115 $usedHebrewMonth = false;
1116 $usedIranianMonth = false;
1117 $usedHijriMonth = false;
1118 $usedHebrewYear = false;
1119 $usedIranianYear = false;
1120 $usedHijriYear = false;
1121 $usedTennoYear = false;
1122
1123 if ( strlen( $ts ) !== 14 ) {
1124 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1125 }
1126
1127 if ( !ctype_digit( $ts ) ) {
1128 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1129 }
1130
1131 $formatLength = strlen( $format );
1132 for ( $p = 0; $p < $formatLength; $p++ ) {
1133 $num = false;
1134 $code = $format[$p];
1135 if ( $code == 'x' && $p < $formatLength - 1 ) {
1136 $code .= $format[++$p];
1137 }
1138
1139 if ( ( $code === 'xi'
1140 || $code === 'xj'
1141 || $code === 'xk'
1142 || $code === 'xm'
1143 || $code === 'xo'
1144 || $code === 'xt' )
1145 && $p < $formatLength - 1 ) {
1146 $code .= $format[++$p];
1147 }
1148
1149 switch ( $code ) {
1150 case 'xx':
1151 $s .= 'x';
1152 break;
1153 case 'xn':
1154 $raw = true;
1155 break;
1156 case 'xN':
1157 $rawToggle = !$rawToggle;
1158 break;
1159 case 'xr':
1160 $roman = true;
1161 break;
1162 case 'xh':
1163 $hebrewNum = true;
1164 break;
1165 case 'xg':
1166 $usedMonth = true;
1167 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1168 break;
1169 case 'xjx':
1170 $usedHebrewMonth = true;
1171 if ( !$hebrew ) {
1172 $hebrew = self::tsToHebrew( $ts );
1173 }
1174 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1175 break;
1176 case 'd':
1177 $usedDay = true;
1178 $num = substr( $ts, 6, 2 );
1179 break;
1180 case 'D':
1181 $usedDay = true;
1182 $s .= $this->getWeekdayAbbreviation(
1183 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1184 );
1185 break;
1186 case 'j':
1187 $usedDay = true;
1188 $num = intval( substr( $ts, 6, 2 ) );
1189 break;
1190 case 'xij':
1191 $usedDay = true;
1192 if ( !$iranian ) {
1193 $iranian = self::tsToIranian( $ts );
1194 }
1195 $num = $iranian[2];
1196 break;
1197 case 'xmj':
1198 $usedDay = true;
1199 if ( !$hijri ) {
1200 $hijri = self::tsToHijri( $ts );
1201 }
1202 $num = $hijri[2];
1203 break;
1204 case 'xjj':
1205 $usedDay = true;
1206 if ( !$hebrew ) {
1207 $hebrew = self::tsToHebrew( $ts );
1208 }
1209 $num = $hebrew[2];
1210 break;
1211 case 'l':
1212 $usedDay = true;
1213 $s .= $this->getWeekdayName(
1214 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1215 );
1216 break;
1217 case 'F':
1218 $usedMonth = true;
1219 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1220 break;
1221 case 'xiF':
1222 $usedIranianMonth = true;
1223 if ( !$iranian ) {
1224 $iranian = self::tsToIranian( $ts );
1225 }
1226 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1227 break;
1228 case 'xmF':
1229 $usedHijriMonth = true;
1230 if ( !$hijri ) {
1231 $hijri = self::tsToHijri( $ts );
1232 }
1233 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1234 break;
1235 case 'xjF':
1236 $usedHebrewMonth = true;
1237 if ( !$hebrew ) {
1238 $hebrew = self::tsToHebrew( $ts );
1239 }
1240 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1241 break;
1242 case 'm':
1243 $usedMonth = true;
1244 $num = substr( $ts, 4, 2 );
1245 break;
1246 case 'M':
1247 $usedMonth = true;
1248 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1249 break;
1250 case 'n':
1251 $usedMonth = true;
1252 $num = intval( substr( $ts, 4, 2 ) );
1253 break;
1254 case 'xin':
1255 $usedIranianMonth = true;
1256 if ( !$iranian ) {
1257 $iranian = self::tsToIranian( $ts );
1258 }
1259 $num = $iranian[1];
1260 break;
1261 case 'xmn':
1262 $usedHijriMonth = true;
1263 if ( !$hijri ) {
1264 $hijri = self::tsToHijri( $ts );
1265 }
1266 $num = $hijri[1];
1267 break;
1268 case 'xjn':
1269 $usedHebrewMonth = true;
1270 if ( !$hebrew ) {
1271 $hebrew = self::tsToHebrew( $ts );
1272 }
1273 $num = $hebrew[1];
1274 break;
1275 case 'xjt':
1276 $usedHebrewMonth = true;
1277 if ( !$hebrew ) {
1278 $hebrew = self::tsToHebrew( $ts );
1279 }
1280 $num = $hebrew[3];
1281 break;
1282 case 'Y':
1283 $usedYear = true;
1284 $num = substr( $ts, 0, 4 );
1285 break;
1286 case 'xiY':
1287 $usedIranianYear = true;
1288 if ( !$iranian ) {
1289 $iranian = self::tsToIranian( $ts );
1290 }
1291 $num = $iranian[0];
1292 break;
1293 case 'xmY':
1294 $usedHijriYear = true;
1295 if ( !$hijri ) {
1296 $hijri = self::tsToHijri( $ts );
1297 }
1298 $num = $hijri[0];
1299 break;
1300 case 'xjY':
1301 $usedHebrewYear = true;
1302 if ( !$hebrew ) {
1303 $hebrew = self::tsToHebrew( $ts );
1304 }
1305 $num = $hebrew[0];
1306 break;
1307 case 'xkY':
1308 $usedYear = true;
1309 if ( !$thai ) {
1310 $thai = self::tsToYear( $ts, 'thai' );
1311 }
1312 $num = $thai[0];
1313 break;
1314 case 'xoY':
1315 $usedYear = true;
1316 if ( !$minguo ) {
1317 $minguo = self::tsToYear( $ts, 'minguo' );
1318 }
1319 $num = $minguo[0];
1320 break;
1321 case 'xtY':
1322 $usedTennoYear = true;
1323 if ( !$tenno ) {
1324 $tenno = self::tsToYear( $ts, 'tenno' );
1325 }
1326 $num = $tenno[0];
1327 break;
1328 case 'y':
1329 $usedYear = true;
1330 $num = substr( $ts, 2, 2 );
1331 break;
1332 case 'xiy':
1333 $usedIranianYear = true;
1334 if ( !$iranian ) {
1335 $iranian = self::tsToIranian( $ts );
1336 }
1337 $num = substr( $iranian[0], -2 );
1338 break;
1339 case 'a':
1340 $usedAMPM = true;
1341 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1342 break;
1343 case 'A':
1344 $usedAMPM = true;
1345 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1346 break;
1347 case 'g':
1348 $usedHour = true;
1349 $h = substr( $ts, 8, 2 );
1350 $num = $h % 12 ? $h % 12 : 12;
1351 break;
1352 case 'G':
1353 $usedHour = true;
1354 $num = intval( substr( $ts, 8, 2 ) );
1355 break;
1356 case 'h':
1357 $usedHour = true;
1358 $h = substr( $ts, 8, 2 );
1359 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1360 break;
1361 case 'H':
1362 $usedHour = true;
1363 $num = substr( $ts, 8, 2 );
1364 break;
1365 case 'i':
1366 $usedMinute = true;
1367 $num = substr( $ts, 10, 2 );
1368 break;
1369 case 's':
1370 $usedSecond = true;
1371 $num = substr( $ts, 12, 2 );
1372 break;
1373 case 'c':
1374 case 'r':
1375 $usedSecond = true;
1376 // fall through
1377 case 'e':
1378 case 'O':
1379 case 'P':
1380 case 'T':
1381 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1382 break;
1383 case 'w':
1384 case 'N':
1385 case 'z':
1386 $usedDay = true;
1387 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1388 break;
1389 case 'W':
1390 $usedWeek = true;
1391 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1392 break;
1393 case 't':
1394 $usedMonth = true;
1395 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1396 break;
1397 case 'L':
1398 $usedIsLeapYear = true;
1399 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1400 break;
1401 case 'o':
1402 $usedISOYear = true;
1403 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1404 break;
1405 case 'U':
1406 $usedSecond = true;
1407 // fall through
1408 case 'I':
1409 case 'Z':
1410 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1411 break;
1412 case '\\':
1413 # Backslash escaping
1414 if ( $p < $formatLength - 1 ) {
1415 $s .= $format[++$p];
1416 } else {
1417 $s .= '\\';
1418 }
1419 break;
1420 case '"':
1421 # Quoted literal
1422 if ( $p < $formatLength - 1 ) {
1423 $endQuote = strpos( $format, '"', $p + 1 );
1424 if ( $endQuote === false ) {
1425 # No terminating quote, assume literal "
1426 $s .= '"';
1427 } else {
1428 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1429 $p = $endQuote;
1430 }
1431 } else {
1432 # Quote at end of string, assume literal "
1433 $s .= '"';
1434 }
1435 break;
1436 default:
1437 $s .= $format[$p];
1438 }
1439 if ( $num !== false ) {
1440 if ( $rawToggle || $raw ) {
1441 $s .= $num;
1442 $raw = false;
1443 } elseif ( $roman ) {
1444 $s .= Language::romanNumeral( $num );
1445 $roman = false;
1446 } elseif ( $hebrewNum ) {
1447 $s .= self::hebrewNumeral( $num );
1448 $hebrewNum = false;
1449 } else {
1450 $s .= $this->formatNum( $num, true );
1451 }
1452 }
1453 }
1454
1455 if ( $usedSecond ) {
1456 $ttl = 1;
1457 } elseif ( $usedMinute ) {
1458 $ttl = 60 - substr( $ts, 12, 2 );
1459 } elseif ( $usedHour ) {
1460 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1461 } elseif ( $usedAMPM ) {
1462 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1463 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1464 } elseif (
1465 $usedDay ||
1466 $usedHebrewMonth ||
1467 $usedIranianMonth ||
1468 $usedHijriMonth ||
1469 $usedHebrewYear ||
1470 $usedIranianYear ||
1471 $usedHijriYear ||
1472 $usedTennoYear
1473 ) {
1474 // @todo Someone who understands the non-Gregorian calendars
1475 // should write proper logic for them so that they don't need purged every day.
1476 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1477 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1478 } else {
1479 $possibleTtls = [];
1480 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1481 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1482 if ( $usedWeek ) {
1483 $possibleTtls[] =
1484 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1485 $timeRemainingInDay;
1486 } elseif ( $usedISOYear ) {
1487 // December 28th falls on the last ISO week of the year, every year.
1488 // The last ISO week of a year can be 52 or 53.
1489 $lastWeekOfISOYear = DateTime::createFromFormat(
1490 'Ymd',
1491 substr( $ts, 0, 4 ) . '1228',
1492 $zone ?: new DateTimeZone( 'UTC' )
1493 )->format( 'W' );
1494 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1495 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1496 $timeRemainingInWeek =
1497 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1498 + $timeRemainingInDay;
1499 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1500 }
1501
1502 if ( $usedMonth ) {
1503 $possibleTtls[] =
1504 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1505 substr( $ts, 6, 2 ) ) * 86400
1506 + $timeRemainingInDay;
1507 } elseif ( $usedYear ) {
1508 $possibleTtls[] =
1509 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1510 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1511 + $timeRemainingInDay;
1512 } elseif ( $usedIsLeapYear ) {
1513 $year = substr( $ts, 0, 4 );
1514 $timeRemainingInYear =
1515 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1516 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1517 + $timeRemainingInDay;
1518 $mod = $year % 4;
1519 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1520 // this isn't a leap year. see when the next one starts
1521 $nextCandidate = $year - $mod + 4;
1522 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1523 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1524 $timeRemainingInYear;
1525 } else {
1526 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1527 $timeRemainingInYear;
1528 }
1529 } else {
1530 // this is a leap year, so the next year isn't
1531 $possibleTtls[] = $timeRemainingInYear;
1532 }
1533 }
1534
1535 if ( $possibleTtls ) {
1536 $ttl = min( $possibleTtls );
1537 }
1538 }
1539
1540 return $s;
1541 }
1542
1543 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1544 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1545
1546 /**
1547 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1548 * Gregorian dates to Iranian dates. Originally written in C, it
1549 * is released under the terms of GNU Lesser General Public
1550 * License. Conversion to PHP was performed by Niklas Laxström.
1551 *
1552 * Link: http://www.farsiweb.info/jalali/jalali.c
1553 *
1554 * @param string $ts
1555 *
1556 * @return int[]
1557 */
1558 private static function tsToIranian( $ts ) {
1559 $gy = substr( $ts, 0, 4 ) -1600;
1560 $gm = substr( $ts, 4, 2 ) -1;
1561 $gd = substr( $ts, 6, 2 ) -1;
1562
1563 # Days passed from the beginning (including leap years)
1564 $gDayNo = 365 * $gy
1565 + floor( ( $gy + 3 ) / 4 )
1566 - floor( ( $gy + 99 ) / 100 )
1567 + floor( ( $gy + 399 ) / 400 );
1568
1569 // Add days of the past months of this year
1570 for ( $i = 0; $i < $gm; $i++ ) {
1571 $gDayNo += self::$GREG_DAYS[$i];
1572 }
1573
1574 // Leap years
1575 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1576 $gDayNo++;
1577 }
1578
1579 // Days passed in current month
1580 $gDayNo += (int)$gd;
1581
1582 $jDayNo = $gDayNo - 79;
1583
1584 $jNp = floor( $jDayNo / 12053 );
1585 $jDayNo %= 12053;
1586
1587 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1588 $jDayNo %= 1461;
1589
1590 if ( $jDayNo >= 366 ) {
1591 $jy += floor( ( $jDayNo - 1 ) / 365 );
1592 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1593 }
1594
1595 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1596 $jDayNo -= self::$IRANIAN_DAYS[$i];
1597 }
1598
1599 $jm = $i + 1;
1600 $jd = $jDayNo + 1;
1601
1602 return [ $jy, $jm, $jd ];
1603 }
1604
1605 /**
1606 * Converting Gregorian dates to Hijri dates.
1607 *
1608 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1609 *
1610 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1611 *
1612 * @param string $ts
1613 *
1614 * @return int[]
1615 */
1616 private static function tsToHijri( $ts ) {
1617 $year = substr( $ts, 0, 4 );
1618 $month = substr( $ts, 4, 2 );
1619 $day = substr( $ts, 6, 2 );
1620
1621 $zyr = $year;
1622 $zd = $day;
1623 $zm = $month;
1624 $zy = $zyr;
1625
1626 if (
1627 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1628 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1629 ) {
1630 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1631 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1632 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1633 $zd - 32075;
1634 } else {
1635 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1636 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1637 }
1638
1639 $zl = $zjd -1948440 + 10632;
1640 $zn = (int)( ( $zl - 1 ) / 10631 );
1641 $zl = $zl - 10631 * $zn + 354;
1642 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1643 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1644 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1645 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1646 $zm = (int)( ( 24 * $zl ) / 709 );
1647 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1648 $zy = 30 * $zn + $zj - 30;
1649
1650 return [ $zy, $zm, $zd ];
1651 }
1652
1653 /**
1654 * Converting Gregorian dates to Hebrew dates.
1655 *
1656 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1657 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1658 * to translate the relevant functions into PHP and release them under
1659 * GNU GPL.
1660 *
1661 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1662 * and Adar II is 14. In a non-leap year, Adar is 6.
1663 *
1664 * @param string $ts
1665 *
1666 * @return int[]
1667 */
1668 private static function tsToHebrew( $ts ) {
1669 # Parse date
1670 $year = substr( $ts, 0, 4 );
1671 $month = substr( $ts, 4, 2 );
1672 $day = substr( $ts, 6, 2 );
1673
1674 # Calculate Hebrew year
1675 $hebrewYear = $year + 3760;
1676
1677 # Month number when September = 1, August = 12
1678 $month += 4;
1679 if ( $month > 12 ) {
1680 # Next year
1681 $month -= 12;
1682 $year++;
1683 $hebrewYear++;
1684 }
1685
1686 # Calculate day of year from 1 September
1687 $dayOfYear = $day;
1688 for ( $i = 1; $i < $month; $i++ ) {
1689 if ( $i == 6 ) {
1690 # February
1691 $dayOfYear += 28;
1692 # Check if the year is leap
1693 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1694 $dayOfYear++;
1695 }
1696 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1697 $dayOfYear += 30;
1698 } else {
1699 $dayOfYear += 31;
1700 }
1701 }
1702
1703 # Calculate the start of the Hebrew year
1704 $start = self::hebrewYearStart( $hebrewYear );
1705
1706 # Calculate next year's start
1707 if ( $dayOfYear <= $start ) {
1708 # Day is before the start of the year - it is the previous year
1709 # Next year's start
1710 $nextStart = $start;
1711 # Previous year
1712 $year--;
1713 $hebrewYear--;
1714 # Add days since previous year's 1 September
1715 $dayOfYear += 365;
1716 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1717 # Leap year
1718 $dayOfYear++;
1719 }
1720 # Start of the new (previous) year
1721 $start = self::hebrewYearStart( $hebrewYear );
1722 } else {
1723 # Next year's start
1724 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1725 }
1726
1727 # Calculate Hebrew day of year
1728 $hebrewDayOfYear = $dayOfYear - $start;
1729
1730 # Difference between year's days
1731 $diff = $nextStart - $start;
1732 # Add 12 (or 13 for leap years) days to ignore the difference between
1733 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1734 # difference is only about the year type
1735 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1736 $diff += 13;
1737 } else {
1738 $diff += 12;
1739 }
1740
1741 # Check the year pattern, and is leap year
1742 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1743 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1744 # and non-leap years
1745 $yearPattern = $diff % 30;
1746 # Check if leap year
1747 $isLeap = $diff >= 30;
1748
1749 # Calculate day in the month from number of day in the Hebrew year
1750 # Don't check Adar - if the day is not in Adar, we will stop before;
1751 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1752 $hebrewDay = $hebrewDayOfYear;
1753 $hebrewMonth = 1;
1754 $days = 0;
1755 while ( $hebrewMonth <= 12 ) {
1756 # Calculate days in this month
1757 if ( $isLeap && $hebrewMonth == 6 ) {
1758 # Adar in a leap year
1759 if ( $isLeap ) {
1760 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1761 $days = 30;
1762 if ( $hebrewDay <= $days ) {
1763 # Day in Adar I
1764 $hebrewMonth = 13;
1765 } else {
1766 # Subtract the days of Adar I
1767 $hebrewDay -= $days;
1768 # Try Adar II
1769 $days = 29;
1770 if ( $hebrewDay <= $days ) {
1771 # Day in Adar II
1772 $hebrewMonth = 14;
1773 }
1774 }
1775 }
1776 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1777 # Cheshvan in a complete year (otherwise as the rule below)
1778 $days = 30;
1779 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1780 # Kislev in an incomplete year (otherwise as the rule below)
1781 $days = 29;
1782 } else {
1783 # Odd months have 30 days, even have 29
1784 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1785 }
1786 if ( $hebrewDay <= $days ) {
1787 # In the current month
1788 break;
1789 } else {
1790 # Subtract the days of the current month
1791 $hebrewDay -= $days;
1792 # Try in the next month
1793 $hebrewMonth++;
1794 }
1795 }
1796
1797 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1798 }
1799
1800 /**
1801 * This calculates the Hebrew year start, as days since 1 September.
1802 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1803 * Used for Hebrew date.
1804 *
1805 * @param int $year
1806 *
1807 * @return string
1808 */
1809 private static function hebrewYearStart( $year ) {
1810 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1811 $b = intval( ( $year - 1 ) % 4 );
1812 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1813 if ( $m < 0 ) {
1814 $m--;
1815 }
1816 $Mar = intval( $m );
1817 if ( $m < 0 ) {
1818 $m++;
1819 }
1820 $m -= $Mar;
1821
1822 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1823 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1824 $Mar++;
1825 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1826 $Mar += 2;
1827 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1828 $Mar++;
1829 }
1830
1831 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1832 return $Mar;
1833 }
1834
1835 /**
1836 * Algorithm to convert Gregorian dates to Thai solar dates,
1837 * Minguo dates or Minguo dates.
1838 *
1839 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1840 * http://en.wikipedia.org/wiki/Minguo_calendar
1841 * http://en.wikipedia.org/wiki/Japanese_era_name
1842 *
1843 * @param string $ts 14-character timestamp
1844 * @param string $cName Calender name
1845 * @return array Converted year, month, day
1846 */
1847 private static function tsToYear( $ts, $cName ) {
1848 $gy = substr( $ts, 0, 4 );
1849 $gm = substr( $ts, 4, 2 );
1850 $gd = substr( $ts, 6, 2 );
1851
1852 if ( !strcmp( $cName, 'thai' ) ) {
1853 # Thai solar dates
1854 # Add 543 years to the Gregorian calendar
1855 # Months and days are identical
1856 $gy_offset = $gy + 543;
1857 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1858 # Minguo dates
1859 # Deduct 1911 years from the Gregorian calendar
1860 # Months and days are identical
1861 $gy_offset = $gy - 1911;
1862 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1863 # Nengō dates up to Meiji period
1864 # Deduct years from the Gregorian calendar
1865 # depending on the nengo periods
1866 # Months and days are identical
1867 if ( ( $gy < 1912 )
1868 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1869 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1870 ) {
1871 # Meiji period
1872 $gy_gannen = $gy - 1868 + 1;
1873 $gy_offset = $gy_gannen;
1874 if ( $gy_gannen == 1 ) {
1875 $gy_offset = '元';
1876 }
1877 $gy_offset = '明治' . $gy_offset;
1878 } elseif (
1879 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1880 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1881 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1882 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1883 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1884 ) {
1885 # Taishō period
1886 $gy_gannen = $gy - 1912 + 1;
1887 $gy_offset = $gy_gannen;
1888 if ( $gy_gannen == 1 ) {
1889 $gy_offset = '元';
1890 }
1891 $gy_offset = '大正' . $gy_offset;
1892 } elseif (
1893 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1894 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1895 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1896 ) {
1897 # Shōwa period
1898 $gy_gannen = $gy - 1926 + 1;
1899 $gy_offset = $gy_gannen;
1900 if ( $gy_gannen == 1 ) {
1901 $gy_offset = '元';
1902 }
1903 $gy_offset = '昭和' . $gy_offset;
1904 } else {
1905 # Heisei period
1906 $gy_gannen = $gy - 1989 + 1;
1907 $gy_offset = $gy_gannen;
1908 if ( $gy_gannen == 1 ) {
1909 $gy_offset = '元';
1910 }
1911 $gy_offset = '平成' . $gy_offset;
1912 }
1913 } else {
1914 $gy_offset = $gy;
1915 }
1916
1917 return [ $gy_offset, $gm, $gd ];
1918 }
1919
1920 /**
1921 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1922 *
1923 * This is the rule the BIDI algorithm uses to determine the directionality of
1924 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1925 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1926 *
1927 * TODO: Does not handle BIDI control characters inside the text.
1928 * TODO: Does not handle unallocated characters.
1929 *
1930 * @param string $text Text to test
1931 * @return null|string Directionality ('ltr' or 'rtl') or null
1932 */
1933 private static function strongDirFromContent( $text = '' ) {
1934 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1935 return null;
1936 }
1937 if ( $matches[1] === '' ) {
1938 return 'rtl';
1939 }
1940 return 'ltr';
1941 }
1942
1943 /**
1944 * Roman number formatting up to 10000
1945 *
1946 * @param int $num
1947 *
1948 * @return string
1949 */
1950 static function romanNumeral( $num ) {
1951 static $table = [
1952 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1953 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1954 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1955 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1956 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1957 ];
1958
1959 $num = intval( $num );
1960 if ( $num > 10000 || $num <= 0 ) {
1961 return $num;
1962 }
1963
1964 $s = '';
1965 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1966 if ( $num >= $pow10 ) {
1967 $s .= $table[$i][(int)floor( $num / $pow10 )];
1968 }
1969 $num = $num % $pow10;
1970 }
1971 return $s;
1972 }
1973
1974 /**
1975 * Hebrew Gematria number formatting up to 9999
1976 *
1977 * @param int $num
1978 *
1979 * @return string
1980 */
1981 static function hebrewNumeral( $num ) {
1982 static $table = [
1983 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
1984 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
1985 [ '',
1986 [ 'ק' ],
1987 [ 'ר' ],
1988 [ 'ש' ],
1989 [ 'ת' ],
1990 [ 'ת', 'ק' ],
1991 [ 'ת', 'ר' ],
1992 [ 'ת', 'ש' ],
1993 [ 'ת', 'ת' ],
1994 [ 'ת', 'ת', 'ק' ],
1995 [ 'ת', 'ת', 'ר' ],
1996 ],
1997 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
1998 ];
1999
2000 $num = intval( $num );
2001 if ( $num > 9999 || $num <= 0 ) {
2002 return $num;
2003 }
2004
2005 // Round thousands have special notations
2006 if ( $num === 1000 ) {
2007 return "א' אלף";
2008 } elseif ( $num % 1000 === 0 ) {
2009 return $table[0][$num / 1000] . "' אלפים";
2010 }
2011
2012 $letters = [];
2013
2014 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2015 if ( $num >= $pow10 ) {
2016 if ( $num === 15 || $num === 16 ) {
2017 $letters[] = $table[0][9];
2018 $letters[] = $table[0][$num - 9];
2019 $num = 0;
2020 } else {
2021 $letters = array_merge(
2022 $letters,
2023 (array)$table[$i][intval( $num / $pow10 )]
2024 );
2025
2026 if ( $pow10 === 1000 ) {
2027 $letters[] = "'";
2028 }
2029 }
2030 }
2031
2032 $num = $num % $pow10;
2033 }
2034
2035 $preTransformLength = count( $letters );
2036 if ( $preTransformLength === 1 ) {
2037 // Add geresh (single quote) to one-letter numbers
2038 $letters[] = "'";
2039 } else {
2040 $lastIndex = $preTransformLength - 1;
2041 $letters[$lastIndex] = str_replace(
2042 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2043 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2044 $letters[$lastIndex]
2045 );
2046
2047 // Add gershayim (double quote) to multiple-letter numbers,
2048 // but exclude numbers with only one letter after the thousands
2049 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2050 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2051 $letters[] = "'";
2052 } else {
2053 array_splice( $letters, -1, 0, '"' );
2054 }
2055 }
2056
2057 return implode( $letters );
2058 }
2059
2060 /**
2061 * Used by date() and time() to adjust the time output.
2062 *
2063 * @param string $ts The time in date('YmdHis') format
2064 * @param mixed $tz Adjust the time by this amount (default false, mean we
2065 * get user timecorrection setting)
2066 * @return int
2067 */
2068 public function userAdjust( $ts, $tz = false ) {
2069 global $wgUser, $wgLocalTZoffset;
2070
2071 if ( $tz === false ) {
2072 $tz = $wgUser->getOption( 'timecorrection' );
2073 }
2074
2075 $data = explode( '|', $tz, 3 );
2076
2077 if ( $data[0] == 'ZoneInfo' ) {
2078 MediaWiki\suppressWarnings();
2079 $userTZ = timezone_open( $data[2] );
2080 MediaWiki\restoreWarnings();
2081 if ( $userTZ !== false ) {
2082 $date = date_create( $ts, timezone_open( 'UTC' ) );
2083 date_timezone_set( $date, $userTZ );
2084 $date = date_format( $date, 'YmdHis' );
2085 return $date;
2086 }
2087 # Unrecognized timezone, default to 'Offset' with the stored offset.
2088 $data[0] = 'Offset';
2089 }
2090
2091 if ( $data[0] == 'System' || $tz == '' ) {
2092 # Global offset in minutes.
2093 $minDiff = $wgLocalTZoffset;
2094 } elseif ( $data[0] == 'Offset' ) {
2095 $minDiff = intval( $data[1] );
2096 } else {
2097 $data = explode( ':', $tz );
2098 if ( count( $data ) == 2 ) {
2099 $data[0] = intval( $data[0] );
2100 $data[1] = intval( $data[1] );
2101 $minDiff = abs( $data[0] ) * 60 + $data[1];
2102 if ( $data[0] < 0 ) {
2103 $minDiff = -$minDiff;
2104 }
2105 } else {
2106 $minDiff = intval( $data[0] ) * 60;
2107 }
2108 }
2109
2110 # No difference ? Return time unchanged
2111 if ( 0 == $minDiff ) {
2112 return $ts;
2113 }
2114
2115 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2116 # Generate an adjusted date; take advantage of the fact that mktime
2117 # will normalize out-of-range values so we don't have to split $minDiff
2118 # into hours and minutes.
2119 $t = mktime( (
2120 (int)substr( $ts, 8, 2 ) ), # Hours
2121 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2122 (int)substr( $ts, 12, 2 ), # Seconds
2123 (int)substr( $ts, 4, 2 ), # Month
2124 (int)substr( $ts, 6, 2 ), # Day
2125 (int)substr( $ts, 0, 4 ) ); # Year
2126
2127 $date = date( 'YmdHis', $t );
2128 MediaWiki\restoreWarnings();
2129
2130 return $date;
2131 }
2132
2133 /**
2134 * This is meant to be used by time(), date(), and timeanddate() to get
2135 * the date preference they're supposed to use, it should be used in
2136 * all children.
2137 *
2138 *<code>
2139 * function timeanddate([...], $format = true) {
2140 * $datePreference = $this->dateFormat($format);
2141 * [...]
2142 * }
2143 *</code>
2144 *
2145 * @param int|string|bool $usePrefs If true, the user's preference is used
2146 * if false, the site/language default is used
2147 * if int/string, assumed to be a format.
2148 * @return string
2149 */
2150 function dateFormat( $usePrefs = true ) {
2151 global $wgUser;
2152
2153 if ( is_bool( $usePrefs ) ) {
2154 if ( $usePrefs ) {
2155 $datePreference = $wgUser->getDatePreference();
2156 } else {
2157 $datePreference = (string)User::getDefaultOption( 'date' );
2158 }
2159 } else {
2160 $datePreference = (string)$usePrefs;
2161 }
2162
2163 // return int
2164 if ( $datePreference == '' ) {
2165 return 'default';
2166 }
2167
2168 return $datePreference;
2169 }
2170
2171 /**
2172 * Get a format string for a given type and preference
2173 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2174 * @param string $pref The format name as it appears in Messages*.php under
2175 * $datePreferences.
2176 *
2177 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2178 *
2179 * @return string
2180 */
2181 function getDateFormatString( $type, $pref ) {
2182 $wasDefault = false;
2183 if ( $pref == 'default' ) {
2184 $wasDefault = true;
2185 $pref = $this->getDefaultDateFormat();
2186 }
2187
2188 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2189 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2190
2191 if ( $type === 'pretty' && $df === null ) {
2192 $df = $this->getDateFormatString( 'date', $pref );
2193 }
2194
2195 if ( !$wasDefault && $df === null ) {
2196 $pref = $this->getDefaultDateFormat();
2197 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2198 }
2199
2200 $this->dateFormatStrings[$type][$pref] = $df;
2201 }
2202 return $this->dateFormatStrings[$type][$pref];
2203 }
2204
2205 /**
2206 * @param string $ts The time format which needs to be turned into a
2207 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2208 * @param bool $adj Whether to adjust the time output according to the
2209 * user configured offset ($timecorrection)
2210 * @param mixed $format True to use user's date format preference
2211 * @param string|bool $timecorrection The time offset as returned by
2212 * validateTimeZone() in Special:Preferences
2213 * @return string
2214 */
2215 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2216 $ts = wfTimestamp( TS_MW, $ts );
2217 if ( $adj ) {
2218 $ts = $this->userAdjust( $ts, $timecorrection );
2219 }
2220 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2221 return $this->sprintfDate( $df, $ts );
2222 }
2223
2224 /**
2225 * @param string $ts The time format which needs to be turned into a
2226 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2227 * @param bool $adj Whether to adjust the time output according to the
2228 * user configured offset ($timecorrection)
2229 * @param mixed $format True to use user's date format preference
2230 * @param string|bool $timecorrection The time offset as returned by
2231 * validateTimeZone() in Special:Preferences
2232 * @return string
2233 */
2234 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2235 $ts = wfTimestamp( TS_MW, $ts );
2236 if ( $adj ) {
2237 $ts = $this->userAdjust( $ts, $timecorrection );
2238 }
2239 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2240 return $this->sprintfDate( $df, $ts );
2241 }
2242
2243 /**
2244 * @param string $ts The time format which needs to be turned into a
2245 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2246 * @param bool $adj Whether to adjust the time output according to the
2247 * user configured offset ($timecorrection)
2248 * @param mixed $format What format to return, if it's false output the
2249 * default one (default true)
2250 * @param string|bool $timecorrection The time offset as returned by
2251 * validateTimeZone() in Special:Preferences
2252 * @return string
2253 */
2254 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2255 $ts = wfTimestamp( TS_MW, $ts );
2256 if ( $adj ) {
2257 $ts = $this->userAdjust( $ts, $timecorrection );
2258 }
2259 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2260 return $this->sprintfDate( $df, $ts );
2261 }
2262
2263 /**
2264 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2265 *
2266 * @since 1.20
2267 *
2268 * @param int $seconds The amount of seconds.
2269 * @param array $chosenIntervals The intervals to enable.
2270 *
2271 * @return string
2272 */
2273 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2274 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2275
2276 $segments = [];
2277
2278 foreach ( $intervals as $intervalName => $intervalValue ) {
2279 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2280 // duration-years, duration-decades, duration-centuries, duration-millennia
2281 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2282 $segments[] = $message->inLanguage( $this )->escaped();
2283 }
2284
2285 return $this->listToText( $segments );
2286 }
2287
2288 /**
2289 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2290 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2291 *
2292 * @since 1.20
2293 *
2294 * @param int $seconds The amount of seconds.
2295 * @param array $chosenIntervals The intervals to enable.
2296 *
2297 * @return array
2298 */
2299 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2300 if ( empty( $chosenIntervals ) ) {
2301 $chosenIntervals = [
2302 'millennia',
2303 'centuries',
2304 'decades',
2305 'years',
2306 'days',
2307 'hours',
2308 'minutes',
2309 'seconds'
2310 ];
2311 }
2312
2313 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2314 $sortedNames = array_keys( $intervals );
2315 $smallestInterval = array_pop( $sortedNames );
2316
2317 $segments = [];
2318
2319 foreach ( $intervals as $name => $length ) {
2320 $value = floor( $seconds / $length );
2321
2322 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2323 $seconds -= $value * $length;
2324 $segments[$name] = $value;
2325 }
2326 }
2327
2328 return $segments;
2329 }
2330
2331 /**
2332 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2333 *
2334 * @param string $type Can be 'date', 'time' or 'both'
2335 * @param string $ts The time format which needs to be turned into a
2336 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2337 * @param User $user User object used to get preferences for timezone and format
2338 * @param array $options Array, can contain the following keys:
2339 * - 'timecorrection': time correction, can have the following values:
2340 * - true: use user's preference
2341 * - false: don't use time correction
2342 * - int: value of time correction in minutes
2343 * - 'format': format to use, can have the following values:
2344 * - true: use user's preference
2345 * - false: use default preference
2346 * - string: format to use
2347 * @since 1.19
2348 * @return string
2349 */
2350 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2351 $ts = wfTimestamp( TS_MW, $ts );
2352 $options += [ 'timecorrection' => true, 'format' => true ];
2353 if ( $options['timecorrection'] !== false ) {
2354 if ( $options['timecorrection'] === true ) {
2355 $offset = $user->getOption( 'timecorrection' );
2356 } else {
2357 $offset = $options['timecorrection'];
2358 }
2359 $ts = $this->userAdjust( $ts, $offset );
2360 }
2361 if ( $options['format'] === true ) {
2362 $format = $user->getDatePreference();
2363 } else {
2364 $format = $options['format'];
2365 }
2366 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2367 return $this->sprintfDate( $df, $ts );
2368 }
2369
2370 /**
2371 * Get the formatted date for the given timestamp and formatted for
2372 * the given user.
2373 *
2374 * @param mixed $ts Mixed: the time format which needs to be turned into a
2375 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2376 * @param User $user User object used to get preferences for timezone and format
2377 * @param array $options Array, can contain the following keys:
2378 * - 'timecorrection': time correction, can have the following values:
2379 * - true: use user's preference
2380 * - false: don't use time correction
2381 * - int: value of time correction in minutes
2382 * - 'format': format to use, can have the following values:
2383 * - true: use user's preference
2384 * - false: use default preference
2385 * - string: format to use
2386 * @since 1.19
2387 * @return string
2388 */
2389 public function userDate( $ts, User $user, array $options = [] ) {
2390 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2391 }
2392
2393 /**
2394 * Get the formatted time for the given timestamp and formatted for
2395 * the given user.
2396 *
2397 * @param mixed $ts The time format which needs to be turned into a
2398 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2399 * @param User $user User object used to get preferences for timezone and format
2400 * @param array $options Array, can contain the following keys:
2401 * - 'timecorrection': time correction, can have the following values:
2402 * - true: use user's preference
2403 * - false: don't use time correction
2404 * - int: value of time correction in minutes
2405 * - 'format': format to use, can have the following values:
2406 * - true: use user's preference
2407 * - false: use default preference
2408 * - string: format to use
2409 * @since 1.19
2410 * @return string
2411 */
2412 public function userTime( $ts, User $user, array $options = [] ) {
2413 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2414 }
2415
2416 /**
2417 * Get the formatted date and time for the given timestamp and formatted for
2418 * the given user.
2419 *
2420 * @param mixed $ts The time format which needs to be turned into a
2421 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2422 * @param User $user User object used to get preferences for timezone and format
2423 * @param array $options Array, can contain the following keys:
2424 * - 'timecorrection': time correction, can have the following values:
2425 * - true: use user's preference
2426 * - false: don't use time correction
2427 * - int: value of time correction in minutes
2428 * - 'format': format to use, can have the following values:
2429 * - true: use user's preference
2430 * - false: use default preference
2431 * - string: format to use
2432 * @since 1.19
2433 * @return string
2434 */
2435 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2436 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2437 }
2438
2439 /**
2440 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2441 *
2442 * Determine the difference between the timestamp and the current time, and
2443 * generate a readable timestamp by returning "<N> <units> ago", where the
2444 * largest possible unit is used.
2445 *
2446 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2447 *
2448 * @param MWTimestamp $time
2449 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2450 * @param User|null $user User the timestamp is being generated for
2451 * (or null to use main context's user)
2452 * @return string Formatted timestamp
2453 */
2454 public function getHumanTimestamp(
2455 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2456 ) {
2457 if ( $relativeTo === null ) {
2458 $relativeTo = new MWTimestamp();
2459 }
2460 if ( $user === null ) {
2461 $user = RequestContext::getMain()->getUser();
2462 }
2463
2464 // Adjust for the user's timezone.
2465 $offsetThis = $time->offsetForUser( $user );
2466 $offsetRel = $relativeTo->offsetForUser( $user );
2467
2468 $ts = '';
2469 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2470 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2471 }
2472
2473 // Reset the timezone on the objects.
2474 $time->timestamp->sub( $offsetThis );
2475 $relativeTo->timestamp->sub( $offsetRel );
2476
2477 return $ts;
2478 }
2479
2480 /**
2481 * Convert an MWTimestamp into a pretty human-readable timestamp using
2482 * the given user preferences and relative base time.
2483 *
2484 * @see Language::getHumanTimestamp
2485 * @param MWTimestamp $ts Timestamp to prettify
2486 * @param MWTimestamp $relativeTo Base timestamp
2487 * @param User $user User preferences to use
2488 * @return string Human timestamp
2489 * @since 1.26
2490 */
2491 private function getHumanTimestampInternal(
2492 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2493 ) {
2494 $diff = $ts->diff( $relativeTo );
2495 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2496 (int)$relativeTo->timestamp->format( 'w' ) );
2497 $days = $diff->days ?: (int)$diffDay;
2498 if ( $diff->invert || $days > 5
2499 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2500 ) {
2501 // Timestamps are in different years: use full timestamp
2502 // Also do full timestamp for future dates
2503 /**
2504 * @todo FIXME: Add better handling of future timestamps.
2505 */
2506 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2507 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2508 } elseif ( $days > 5 ) {
2509 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2510 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2511 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2512 } elseif ( $days > 1 ) {
2513 // Timestamp within the past week: show the day of the week and time
2514 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2515 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2516 // Messages:
2517 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2518 $ts = wfMessage( "$weekday-at" )
2519 ->inLanguage( $this )
2520 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2521 ->text();
2522 } elseif ( $days == 1 ) {
2523 // Timestamp was yesterday: say 'yesterday' and the time.
2524 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2525 $ts = wfMessage( 'yesterday-at' )
2526 ->inLanguage( $this )
2527 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2528 ->text();
2529 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2530 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2531 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2532 $ts = wfMessage( 'today-at' )
2533 ->inLanguage( $this )
2534 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2535 ->text();
2536
2537 // From here on in, the timestamp was soon enough ago so that we can simply say
2538 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2539 } elseif ( $diff->h == 1 ) {
2540 // Less than 90 minutes, but more than an hour ago.
2541 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2542 } elseif ( $diff->i >= 1 ) {
2543 // A few minutes ago.
2544 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2545 } elseif ( $diff->s >= 30 ) {
2546 // Less than a minute, but more than 30 sec ago.
2547 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2548 } else {
2549 // Less than 30 seconds ago.
2550 $ts = wfMessage( 'just-now' )->text();
2551 }
2552
2553 return $ts;
2554 }
2555
2556 /**
2557 * @param string $key
2558 * @return array|null
2559 */
2560 public function getMessage( $key ) {
2561 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2562 }
2563
2564 /**
2565 * @return array
2566 */
2567 function getAllMessages() {
2568 return self::$dataCache->getItem( $this->mCode, 'messages' );
2569 }
2570
2571 /**
2572 * @param string $in
2573 * @param string $out
2574 * @param string $string
2575 * @return string
2576 */
2577 public function iconv( $in, $out, $string ) {
2578 # This is a wrapper for iconv in all languages except esperanto,
2579 # which does some nasty x-conversions beforehand
2580
2581 # Even with //IGNORE iconv can whine about illegal characters in
2582 # *input* string. We just ignore those too.
2583 # REF: http://bugs.php.net/bug.php?id=37166
2584 # REF: https://phabricator.wikimedia.org/T18885
2585 MediaWiki\suppressWarnings();
2586 $text = iconv( $in, $out . '//IGNORE', $string );
2587 MediaWiki\restoreWarnings();
2588 return $text;
2589 }
2590
2591 // callback functions for ucwords(), ucwordbreaks()
2592
2593 /**
2594 * @param array $matches
2595 * @return mixed|string
2596 */
2597 function ucwordbreaksCallbackAscii( $matches ) {
2598 return $this->ucfirst( $matches[1] );
2599 }
2600
2601 /**
2602 * @param array $matches
2603 * @return string
2604 */
2605 function ucwordbreaksCallbackMB( $matches ) {
2606 return mb_strtoupper( $matches[0] );
2607 }
2608
2609 /**
2610 * @param array $matches
2611 * @return string
2612 */
2613 function ucwordsCallbackMB( $matches ) {
2614 return mb_strtoupper( $matches[0] );
2615 }
2616
2617 /**
2618 * Make a string's first character uppercase
2619 *
2620 * @param string $str
2621 *
2622 * @return string
2623 */
2624 public function ucfirst( $str ) {
2625 $o = ord( $str );
2626 if ( $o < 96 ) { // if already uppercase...
2627 return $str;
2628 } elseif ( $o < 128 ) {
2629 return ucfirst( $str ); // use PHP's ucfirst()
2630 } else {
2631 // fall back to more complex logic in case of multibyte strings
2632 return $this->uc( $str, true );
2633 }
2634 }
2635
2636 /**
2637 * Convert a string to uppercase
2638 *
2639 * @param string $str
2640 * @param bool $first
2641 *
2642 * @return string
2643 */
2644 public function uc( $str, $first = false ) {
2645 if ( $first ) {
2646 if ( $this->isMultibyte( $str ) ) {
2647 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2648 } else {
2649 return ucfirst( $str );
2650 }
2651 } else {
2652 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2653 }
2654 }
2655
2656 /**
2657 * @param string $str
2658 * @return mixed|string
2659 */
2660 function lcfirst( $str ) {
2661 $o = ord( $str );
2662 if ( !$o ) {
2663 return strval( $str );
2664 } elseif ( $o >= 128 ) {
2665 return $this->lc( $str, true );
2666 } elseif ( $o > 96 ) {
2667 return $str;
2668 } else {
2669 $str[0] = strtolower( $str[0] );
2670 return $str;
2671 }
2672 }
2673
2674 /**
2675 * @param string $str
2676 * @param bool $first
2677 * @return mixed|string
2678 */
2679 function lc( $str, $first = false ) {
2680 if ( $first ) {
2681 if ( $this->isMultibyte( $str ) ) {
2682 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2683 } else {
2684 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2685 }
2686 } else {
2687 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2688 }
2689 }
2690
2691 /**
2692 * @param string $str
2693 * @return bool
2694 */
2695 function isMultibyte( $str ) {
2696 return strlen( $str ) !== mb_strlen( $str );
2697 }
2698
2699 /**
2700 * @param string $str
2701 * @return mixed|string
2702 */
2703 function ucwords( $str ) {
2704 if ( $this->isMultibyte( $str ) ) {
2705 $str = $this->lc( $str );
2706
2707 // regexp to find first letter in each word (i.e. after each space)
2708 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2709
2710 // function to use to capitalize a single char
2711 return preg_replace_callback(
2712 $replaceRegexp,
2713 [ $this, 'ucwordsCallbackMB' ],
2714 $str
2715 );
2716 } else {
2717 return ucwords( strtolower( $str ) );
2718 }
2719 }
2720
2721 /**
2722 * capitalize words at word breaks
2723 *
2724 * @param string $str
2725 * @return mixed
2726 */
2727 function ucwordbreaks( $str ) {
2728 if ( $this->isMultibyte( $str ) ) {
2729 $str = $this->lc( $str );
2730
2731 // since \b doesn't work for UTF-8, we explicitely define word break chars
2732 $breaks = "[ \-\(\)\}\{\.,\?!]";
2733
2734 // find first letter after word break
2735 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2736 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2737
2738 return preg_replace_callback(
2739 $replaceRegexp,
2740 [ $this, 'ucwordbreaksCallbackMB' ],
2741 $str
2742 );
2743 } else {
2744 return preg_replace_callback(
2745 '/\b([\w\x80-\xff]+)\b/',
2746 [ $this, 'ucwordbreaksCallbackAscii' ],
2747 $str
2748 );
2749 }
2750 }
2751
2752 /**
2753 * Return a case-folded representation of $s
2754 *
2755 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2756 * and $s2 are the same except for the case of their characters. It is not
2757 * necessary for the value returned to make sense when displayed.
2758 *
2759 * Do *not* perform any other normalisation in this function. If a caller
2760 * uses this function when it should be using a more general normalisation
2761 * function, then fix the caller.
2762 *
2763 * @param string $s
2764 *
2765 * @return string
2766 */
2767 function caseFold( $s ) {
2768 return $this->uc( $s );
2769 }
2770
2771 /**
2772 * @param string $s
2773 * @return string
2774 * @throws MWException
2775 */
2776 function checkTitleEncoding( $s ) {
2777 if ( is_array( $s ) ) {
2778 throw new MWException( 'Given array to checkTitleEncoding.' );
2779 }
2780 if ( StringUtils::isUtf8( $s ) ) {
2781 return $s;
2782 }
2783
2784 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2785 }
2786
2787 /**
2788 * @return array
2789 */
2790 function fallback8bitEncoding() {
2791 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2792 }
2793
2794 /**
2795 * Most writing systems use whitespace to break up words.
2796 * Some languages such as Chinese don't conventionally do this,
2797 * which requires special handling when breaking up words for
2798 * searching etc.
2799 *
2800 * @return bool
2801 */
2802 function hasWordBreaks() {
2803 return true;
2804 }
2805
2806 /**
2807 * Some languages such as Chinese require word segmentation,
2808 * Specify such segmentation when overridden in derived class.
2809 *
2810 * @param string $string
2811 * @return string
2812 */
2813 function segmentByWord( $string ) {
2814 return $string;
2815 }
2816
2817 /**
2818 * Some languages have special punctuation need to be normalized.
2819 * Make such changes here.
2820 *
2821 * @param string $string
2822 * @return string
2823 */
2824 function normalizeForSearch( $string ) {
2825 return self::convertDoubleWidth( $string );
2826 }
2827
2828 /**
2829 * convert double-width roman characters to single-width.
2830 * range: ff00-ff5f ~= 0020-007f
2831 *
2832 * @param string $string
2833 *
2834 * @return string
2835 */
2836 protected static function convertDoubleWidth( $string ) {
2837 static $full = null;
2838 static $half = null;
2839
2840 if ( $full === null ) {
2841 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2842 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2843 $full = str_split( $fullWidth, 3 );
2844 $half = str_split( $halfWidth );
2845 }
2846
2847 $string = str_replace( $full, $half, $string );
2848 return $string;
2849 }
2850
2851 /**
2852 * @param string $string
2853 * @param string $pattern
2854 * @return string
2855 */
2856 protected static function insertSpace( $string, $pattern ) {
2857 $string = preg_replace( $pattern, " $1 ", $string );
2858 $string = preg_replace( '/ +/', ' ', $string );
2859 return $string;
2860 }
2861
2862 /**
2863 * @param array $termsArray
2864 * @return array
2865 */
2866 function convertForSearchResult( $termsArray ) {
2867 # some languages, e.g. Chinese, need to do a conversion
2868 # in order for search results to be displayed correctly
2869 return $termsArray;
2870 }
2871
2872 /**
2873 * Get the first character of a string.
2874 *
2875 * @param string $s
2876 * @return string
2877 */
2878 function firstChar( $s ) {
2879 $matches = [];
2880 preg_match(
2881 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2882 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2883 $s,
2884 $matches
2885 );
2886
2887 if ( isset( $matches[1] ) ) {
2888 if ( strlen( $matches[1] ) != 3 ) {
2889 return $matches[1];
2890 }
2891
2892 // Break down Hangul syllables to grab the first jamo
2893 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2894 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2895 return $matches[1];
2896 } elseif ( $code < 0xb098 ) {
2897 return "\xe3\x84\xb1";
2898 } elseif ( $code < 0xb2e4 ) {
2899 return "\xe3\x84\xb4";
2900 } elseif ( $code < 0xb77c ) {
2901 return "\xe3\x84\xb7";
2902 } elseif ( $code < 0xb9c8 ) {
2903 return "\xe3\x84\xb9";
2904 } elseif ( $code < 0xbc14 ) {
2905 return "\xe3\x85\x81";
2906 } elseif ( $code < 0xc0ac ) {
2907 return "\xe3\x85\x82";
2908 } elseif ( $code < 0xc544 ) {
2909 return "\xe3\x85\x85";
2910 } elseif ( $code < 0xc790 ) {
2911 return "\xe3\x85\x87";
2912 } elseif ( $code < 0xcc28 ) {
2913 return "\xe3\x85\x88";
2914 } elseif ( $code < 0xce74 ) {
2915 return "\xe3\x85\x8a";
2916 } elseif ( $code < 0xd0c0 ) {
2917 return "\xe3\x85\x8b";
2918 } elseif ( $code < 0xd30c ) {
2919 return "\xe3\x85\x8c";
2920 } elseif ( $code < 0xd558 ) {
2921 return "\xe3\x85\x8d";
2922 } else {
2923 return "\xe3\x85\x8e";
2924 }
2925 } else {
2926 return '';
2927 }
2928 }
2929
2930 function initEncoding() {
2931 # Some languages may have an alternate char encoding option
2932 # (Esperanto X-coding, Japanese furigana conversion, etc)
2933 # If this language is used as the primary content language,
2934 # an override to the defaults can be set here on startup.
2935 }
2936
2937 /**
2938 * @param string $s
2939 * @return string
2940 */
2941 function recodeForEdit( $s ) {
2942 # For some languages we'll want to explicitly specify
2943 # which characters make it into the edit box raw
2944 # or are converted in some way or another.
2945 global $wgEditEncoding;
2946 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2947 return $s;
2948 } else {
2949 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2950 }
2951 }
2952
2953 /**
2954 * @param string $s
2955 * @return string
2956 */
2957 function recodeInput( $s ) {
2958 # Take the previous into account.
2959 global $wgEditEncoding;
2960 if ( $wgEditEncoding != '' ) {
2961 $enc = $wgEditEncoding;
2962 } else {
2963 $enc = 'UTF-8';
2964 }
2965 if ( $enc == 'UTF-8' ) {
2966 return $s;
2967 } else {
2968 return $this->iconv( $enc, 'UTF-8', $s );
2969 }
2970 }
2971
2972 /**
2973 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2974 * also cleans up certain backwards-compatible sequences, converting them
2975 * to the modern Unicode equivalent.
2976 *
2977 * This is language-specific for performance reasons only.
2978 *
2979 * @param string $s
2980 *
2981 * @return string
2982 */
2983 function normalize( $s ) {
2984 global $wgAllUnicodeFixes;
2985 $s = UtfNormal\Validator::cleanUp( $s );
2986 if ( $wgAllUnicodeFixes ) {
2987 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2988 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2989 }
2990
2991 return $s;
2992 }
2993
2994 /**
2995 * Transform a string using serialized data stored in the given file (which
2996 * must be in the serialized subdirectory of $IP). The file contains pairs
2997 * mapping source characters to destination characters.
2998 *
2999 * The data is cached in process memory. This will go faster if you have the
3000 * FastStringSearch extension.
3001 *
3002 * @param string $file
3003 * @param string $string
3004 *
3005 * @throws MWException
3006 * @return string
3007 */
3008 function transformUsingPairFile( $file, $string ) {
3009 if ( !isset( $this->transformData[$file] ) ) {
3010 $data = wfGetPrecompiledData( $file );
3011 if ( $data === false ) {
3012 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
3013 }
3014 $this->transformData[$file] = new ReplacementArray( $data );
3015 }
3016 return $this->transformData[$file]->replace( $string );
3017 }
3018
3019 /**
3020 * For right-to-left language support
3021 *
3022 * @return bool
3023 */
3024 function isRTL() {
3025 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3026 }
3027
3028 /**
3029 * Return the correct HTML 'dir' attribute value for this language.
3030 * @return string
3031 */
3032 function getDir() {
3033 return $this->isRTL() ? 'rtl' : 'ltr';
3034 }
3035
3036 /**
3037 * Return 'left' or 'right' as appropriate alignment for line-start
3038 * for this language's text direction.
3039 *
3040 * Should be equivalent to CSS3 'start' text-align value....
3041 *
3042 * @return string
3043 */
3044 function alignStart() {
3045 return $this->isRTL() ? 'right' : 'left';
3046 }
3047
3048 /**
3049 * Return 'right' or 'left' as appropriate alignment for line-end
3050 * for this language's text direction.
3051 *
3052 * Should be equivalent to CSS3 'end' text-align value....
3053 *
3054 * @return string
3055 */
3056 function alignEnd() {
3057 return $this->isRTL() ? 'left' : 'right';
3058 }
3059
3060 /**
3061 * A hidden direction mark (LRM or RLM), depending on the language direction.
3062 * Unlike getDirMark(), this function returns the character as an HTML entity.
3063 * This function should be used when the output is guaranteed to be HTML,
3064 * because it makes the output HTML source code more readable. When
3065 * the output is plain text or can be escaped, getDirMark() should be used.
3066 *
3067 * @param bool $opposite Get the direction mark opposite to your language
3068 * @return string
3069 * @since 1.20
3070 */
3071 function getDirMarkEntity( $opposite = false ) {
3072 if ( $opposite ) {
3073 return $this->isRTL() ? '&lrm;' : '&rlm;';
3074 }
3075 return $this->isRTL() ? '&rlm;' : '&lrm;';
3076 }
3077
3078 /**
3079 * A hidden direction mark (LRM or RLM), depending on the language direction.
3080 * This function produces them as invisible Unicode characters and
3081 * the output may be hard to read and debug, so it should only be used
3082 * when the output is plain text or can be escaped. When the output is
3083 * HTML, use getDirMarkEntity() instead.
3084 *
3085 * @param bool $opposite Get the direction mark opposite to your language
3086 * @return string
3087 */
3088 function getDirMark( $opposite = false ) {
3089 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3090 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3091 if ( $opposite ) {
3092 return $this->isRTL() ? $lrm : $rlm;
3093 }
3094 return $this->isRTL() ? $rlm : $lrm;
3095 }
3096
3097 /**
3098 * @return array
3099 */
3100 function capitalizeAllNouns() {
3101 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3102 }
3103
3104 /**
3105 * An arrow, depending on the language direction.
3106 *
3107 * @param string $direction The direction of the arrow: forwards (default),
3108 * backwards, left, right, up, down.
3109 * @return string
3110 */
3111 function getArrow( $direction = 'forwards' ) {
3112 switch ( $direction ) {
3113 case 'forwards':
3114 return $this->isRTL() ? '←' : '→';
3115 case 'backwards':
3116 return $this->isRTL() ? '→' : '←';
3117 case 'left':
3118 return '←';
3119 case 'right':
3120 return '→';
3121 case 'up':
3122 return '↑';
3123 case 'down':
3124 return '↓';
3125 }
3126 }
3127
3128 /**
3129 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3130 *
3131 * @return bool
3132 */
3133 function linkPrefixExtension() {
3134 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3135 }
3136
3137 /**
3138 * Get all magic words from cache.
3139 * @return array
3140 */
3141 function getMagicWords() {
3142 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3143 }
3144
3145 /**
3146 * Run the LanguageGetMagic hook once.
3147 */
3148 protected function doMagicHook() {
3149 if ( $this->mMagicHookDone ) {
3150 return;
3151 }
3152 $this->mMagicHookDone = true;
3153 Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
3154 }
3155
3156 /**
3157 * Fill a MagicWord object with data from here
3158 *
3159 * @param MagicWord $mw
3160 */
3161 function getMagic( $mw ) {
3162 // Saves a function call
3163 if ( !$this->mMagicHookDone ) {
3164 $this->doMagicHook();
3165 }
3166
3167 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3168 $rawEntry = $this->mMagicExtensions[$mw->mId];
3169 } else {
3170 $rawEntry = self::$dataCache->getSubitem(
3171 $this->mCode, 'magicWords', $mw->mId );
3172 }
3173
3174 if ( !is_array( $rawEntry ) ) {
3175 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3176 } else {
3177 $mw->mCaseSensitive = $rawEntry[0];
3178 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3179 }
3180 }
3181
3182 /**
3183 * Add magic words to the extension array
3184 *
3185 * @param array $newWords
3186 */
3187 function addMagicWordsByLang( $newWords ) {
3188 $fallbackChain = $this->getFallbackLanguages();
3189 $fallbackChain = array_reverse( $fallbackChain );
3190 foreach ( $fallbackChain as $code ) {
3191 if ( isset( $newWords[$code] ) ) {
3192 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3193 }
3194 }
3195 }
3196
3197 /**
3198 * Get special page names, as an associative array
3199 * canonical name => array of valid names, including aliases
3200 * @return array
3201 */
3202 function getSpecialPageAliases() {
3203 // Cache aliases because it may be slow to load them
3204 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3205 // Initialise array
3206 $this->mExtendedSpecialPageAliases =
3207 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3208 Hooks::run( 'LanguageGetSpecialPageAliases',
3209 [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
3210 }
3211
3212 return $this->mExtendedSpecialPageAliases;
3213 }
3214
3215 /**
3216 * Italic is unsuitable for some languages
3217 *
3218 * @param string $text The text to be emphasized.
3219 * @return string
3220 */
3221 function emphasize( $text ) {
3222 return "<em>$text</em>";
3223 }
3224
3225 /**
3226 * Normally we output all numbers in plain en_US style, that is
3227 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3228 * point twohundredthirtyfive. However this is not suitable for all
3229 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3230 * Icelandic just want to use commas instead of dots, and dots instead
3231 * of commas like "293.291,235".
3232 *
3233 * An example of this function being called:
3234 * <code>
3235 * wfMessage( 'message' )->numParams( $num )->text()
3236 * </code>
3237 *
3238 * See $separatorTransformTable on MessageIs.php for
3239 * the , => . and . => , implementation.
3240 *
3241 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3242 * @param int|float $number The string to be formatted, should be an integer
3243 * or a floating point number.
3244 * @param bool $nocommafy Set to true for special numbers like dates
3245 * @return string
3246 */
3247 public function formatNum( $number, $nocommafy = false ) {
3248 global $wgTranslateNumerals;
3249 if ( !$nocommafy ) {
3250 $number = $this->commafy( $number );
3251 $s = $this->separatorTransformTable();
3252 if ( $s ) {
3253 $number = strtr( $number, $s );
3254 }
3255 }
3256
3257 if ( $wgTranslateNumerals ) {
3258 $s = $this->digitTransformTable();
3259 if ( $s ) {
3260 $number = strtr( $number, $s );
3261 }
3262 }
3263
3264 return $number;
3265 }
3266
3267 /**
3268 * Front-end for non-commafied formatNum
3269 *
3270 * @param int|float $number The string to be formatted, should be an integer
3271 * or a floating point number.
3272 * @since 1.21
3273 * @return string
3274 */
3275 public function formatNumNoSeparators( $number ) {
3276 return $this->formatNum( $number, true );
3277 }
3278
3279 /**
3280 * @param string $number
3281 * @return string
3282 */
3283 public function parseFormattedNumber( $number ) {
3284 $s = $this->digitTransformTable();
3285 if ( $s ) {
3286 // eliminate empty array values such as ''. (bug 64347)
3287 $s = array_filter( $s );
3288 $number = strtr( $number, array_flip( $s ) );
3289 }
3290
3291 $s = $this->separatorTransformTable();
3292 if ( $s ) {
3293 // eliminate empty array values such as ''. (bug 64347)
3294 $s = array_filter( $s );
3295 $number = strtr( $number, array_flip( $s ) );
3296 }
3297
3298 $number = strtr( $number, [ ',' => '' ] );
3299 return $number;
3300 }
3301
3302 /**
3303 * Adds commas to a given number
3304 * @since 1.19
3305 * @param mixed $number
3306 * @return string
3307 */
3308 function commafy( $number ) {
3309 $digitGroupingPattern = $this->digitGroupingPattern();
3310 if ( $number === null ) {
3311 return '';
3312 }
3313
3314 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3315 // default grouping is at thousands, use the same for ###,###,### pattern too.
3316 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3317 } else {
3318 // Ref: http://cldr.unicode.org/translation/number-patterns
3319 $sign = "";
3320 if ( intval( $number ) < 0 ) {
3321 // For negative numbers apply the algorithm like positive number and add sign.
3322 $sign = "-";
3323 $number = substr( $number, 1 );
3324 }
3325 $integerPart = [];
3326 $decimalPart = [];
3327 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3328 preg_match( "/\d+/", $number, $integerPart );
3329 preg_match( "/\.\d*/", $number, $decimalPart );
3330 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3331 if ( $groupedNumber === $number ) {
3332 // the string does not have any number part. Eg: .12345
3333 return $sign . $groupedNumber;
3334 }
3335 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3336 while ( $start > 0 ) {
3337 $match = $matches[0][$numMatches - 1];
3338 $matchLen = strlen( $match );
3339 $start = $end - $matchLen;
3340 if ( $start < 0 ) {
3341 $start = 0;
3342 }
3343 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3344 $end = $start;
3345 if ( $numMatches > 1 ) {
3346 // use the last pattern for the rest of the number
3347 $numMatches--;
3348 }
3349 if ( $start > 0 ) {
3350 $groupedNumber = "," . $groupedNumber;
3351 }
3352 }
3353 return $sign . $groupedNumber;
3354 }
3355 }
3356
3357 /**
3358 * @return string
3359 */
3360 function digitGroupingPattern() {
3361 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3362 }
3363
3364 /**
3365 * @return array
3366 */
3367 function digitTransformTable() {
3368 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3369 }
3370
3371 /**
3372 * @return array
3373 */
3374 function separatorTransformTable() {
3375 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3376 }
3377
3378 /**
3379 * Take a list of strings and build a locale-friendly comma-separated
3380 * list, using the local comma-separator message.
3381 * The last two strings are chained with an "and".
3382 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3383 *
3384 * @param string[] $l
3385 * @return string
3386 */
3387 function listToText( array $l ) {
3388 $m = count( $l ) - 1;
3389 if ( $m < 0 ) {
3390 return '';
3391 }
3392 if ( $m > 0 ) {
3393 $and = $this->msg( 'and' )->escaped();
3394 $space = $this->msg( 'word-separator' )->escaped();
3395 if ( $m > 1 ) {
3396 $comma = $this->msg( 'comma-separator' )->escaped();
3397 }
3398 }
3399 $s = $l[$m];
3400 for ( $i = $m - 1; $i >= 0; $i-- ) {
3401 if ( $i == $m - 1 ) {
3402 $s = $l[$i] . $and . $space . $s;
3403 } else {
3404 $s = $l[$i] . $comma . $s;
3405 }
3406 }
3407 return $s;
3408 }
3409
3410 /**
3411 * Take a list of strings and build a locale-friendly comma-separated
3412 * list, using the local comma-separator message.
3413 * @param string[] $list Array of strings to put in a comma list
3414 * @return string
3415 */
3416 function commaList( array $list ) {
3417 return implode(
3418 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3419 $list
3420 );
3421 }
3422
3423 /**
3424 * Take a list of strings and build a locale-friendly semicolon-separated
3425 * list, using the local semicolon-separator message.
3426 * @param string[] $list Array of strings to put in a semicolon list
3427 * @return string
3428 */
3429 function semicolonList( array $list ) {
3430 return implode(
3431 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3432 $list
3433 );
3434 }
3435
3436 /**
3437 * Same as commaList, but separate it with the pipe instead.
3438 * @param string[] $list Array of strings to put in a pipe list
3439 * @return string
3440 */
3441 function pipeList( array $list ) {
3442 return implode(
3443 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3444 $list
3445 );
3446 }
3447
3448 /**
3449 * Truncate a string to a specified length in bytes, appending an optional
3450 * string (e.g. for ellipses)
3451 *
3452 * The database offers limited byte lengths for some columns in the database;
3453 * multi-byte character sets mean we need to ensure that only whole characters
3454 * are included, otherwise broken characters can be passed to the user
3455 *
3456 * If $length is negative, the string will be truncated from the beginning
3457 *
3458 * @param string $string String to truncate
3459 * @param int $length Maximum length (including ellipses)
3460 * @param string $ellipsis String to append to the truncated text
3461 * @param bool $adjustLength Subtract length of ellipsis from $length.
3462 * $adjustLength was introduced in 1.18, before that behaved as if false.
3463 * @return string
3464 */
3465 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3466 # Use the localized ellipsis character
3467 if ( $ellipsis == '...' ) {
3468 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3469 }
3470 # Check if there is no need to truncate
3471 if ( $length == 0 ) {
3472 return $ellipsis; // convention
3473 } elseif ( strlen( $string ) <= abs( $length ) ) {
3474 return $string; // no need to truncate
3475 }
3476 $stringOriginal = $string;
3477 # If ellipsis length is >= $length then we can't apply $adjustLength
3478 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3479 $string = $ellipsis; // this can be slightly unexpected
3480 # Otherwise, truncate and add ellipsis...
3481 } else {
3482 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3483 if ( $length > 0 ) {
3484 $length -= $eLength;
3485 $string = substr( $string, 0, $length ); // xyz...
3486 $string = $this->removeBadCharLast( $string );
3487 $string = rtrim( $string );
3488 $string = $string . $ellipsis;
3489 } else {
3490 $length += $eLength;
3491 $string = substr( $string, $length ); // ...xyz
3492 $string = $this->removeBadCharFirst( $string );
3493 $string = ltrim( $string );
3494 $string = $ellipsis . $string;
3495 }
3496 }
3497 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3498 # This check is *not* redundant if $adjustLength, due to the single case where
3499 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3500 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3501 return $string;
3502 } else {
3503 return $stringOriginal;
3504 }
3505 }
3506
3507 /**
3508 * Remove bytes that represent an incomplete Unicode character
3509 * at the end of string (e.g. bytes of the char are missing)
3510 *
3511 * @param string $string
3512 * @return string
3513 */
3514 protected function removeBadCharLast( $string ) {
3515 if ( $string != '' ) {
3516 $char = ord( $string[strlen( $string ) - 1] );
3517 $m = [];
3518 if ( $char >= 0xc0 ) {
3519 # We got the first byte only of a multibyte char; remove it.
3520 $string = substr( $string, 0, -1 );
3521 } elseif ( $char >= 0x80 &&
3522 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3523 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3524 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3525 ) {
3526 # We chopped in the middle of a character; remove it
3527 $string = $m[1];
3528 }
3529 }
3530 return $string;
3531 }
3532
3533 /**
3534 * Remove bytes that represent an incomplete Unicode character
3535 * at the start of string (e.g. bytes of the char are missing)
3536 *
3537 * @param string $string
3538 * @return string
3539 */
3540 protected function removeBadCharFirst( $string ) {
3541 if ( $string != '' ) {
3542 $char = ord( $string[0] );
3543 if ( $char >= 0x80 && $char < 0xc0 ) {
3544 # We chopped in the middle of a character; remove the whole thing
3545 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3546 }
3547 }
3548 return $string;
3549 }
3550
3551 /**
3552 * Truncate a string of valid HTML to a specified length in bytes,
3553 * appending an optional string (e.g. for ellipses), and return valid HTML
3554 *
3555 * This is only intended for styled/linked text, such as HTML with
3556 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3557 * Also, this will not detect things like "display:none" CSS.
3558 *
3559 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3560 *
3561 * @param string $text HTML string to truncate
3562 * @param int $length (zero/positive) Maximum length (including ellipses)
3563 * @param string $ellipsis String to append to the truncated text
3564 * @return string
3565 */
3566 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3567 # Use the localized ellipsis character
3568 if ( $ellipsis == '...' ) {
3569 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3570 }
3571 # Check if there is clearly no need to truncate
3572 if ( $length <= 0 ) {
3573 return $ellipsis; // no text shown, nothing to format (convention)
3574 } elseif ( strlen( $text ) <= $length ) {
3575 return $text; // string short enough even *with* HTML (short-circuit)
3576 }
3577
3578 $dispLen = 0; // innerHTML legth so far
3579 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3580 $tagType = 0; // 0-open, 1-close
3581 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3582 $entityState = 0; // 0-not entity, 1-entity
3583 $tag = $ret = ''; // accumulated tag name, accumulated result string
3584 $openTags = []; // open tag stack
3585 $maybeState = null; // possible truncation state
3586
3587 $textLen = strlen( $text );
3588 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3589 for ( $pos = 0; true; ++$pos ) {
3590 # Consider truncation once the display length has reached the maximim.
3591 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3592 # Check that we're not in the middle of a bracket/entity...
3593 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3594 if ( !$testingEllipsis ) {
3595 $testingEllipsis = true;
3596 # Save where we are; we will truncate here unless there turn out to
3597 # be so few remaining characters that truncation is not necessary.
3598 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3599 $maybeState = [ $ret, $openTags ]; // save state
3600 }
3601 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3602 # String in fact does need truncation, the truncation point was OK.
3603 list( $ret, $openTags ) = $maybeState; // reload state
3604 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3605 $ret .= $ellipsis; // add ellipsis
3606 break;
3607 }
3608 }
3609 if ( $pos >= $textLen ) {
3610 break; // extra iteration just for above checks
3611 }
3612
3613 # Read the next char...
3614 $ch = $text[$pos];
3615 $lastCh = $pos ? $text[$pos - 1] : '';
3616 $ret .= $ch; // add to result string
3617 if ( $ch == '<' ) {
3618 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3619 $entityState = 0; // for bad HTML
3620 $bracketState = 1; // tag started (checking for backslash)
3621 } elseif ( $ch == '>' ) {
3622 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3623 $entityState = 0; // for bad HTML
3624 $bracketState = 0; // out of brackets
3625 } elseif ( $bracketState == 1 ) {
3626 if ( $ch == '/' ) {
3627 $tagType = 1; // close tag (e.g. "</span>")
3628 } else {
3629 $tagType = 0; // open tag (e.g. "<span>")
3630 $tag .= $ch;
3631 }
3632 $bracketState = 2; // building tag name
3633 } elseif ( $bracketState == 2 ) {
3634 if ( $ch != ' ' ) {
3635 $tag .= $ch;
3636 } else {
3637 // Name found (e.g. "<a href=..."), add on tag attributes...
3638 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3639 }
3640 } elseif ( $bracketState == 0 ) {
3641 if ( $entityState ) {
3642 if ( $ch == ';' ) {
3643 $entityState = 0;
3644 $dispLen++; // entity is one displayed char
3645 }
3646 } else {
3647 if ( $neLength == 0 && !$maybeState ) {
3648 // Save state without $ch. We want to *hit* the first
3649 // display char (to get tags) but not *use* it if truncating.
3650 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3651 }
3652 if ( $ch == '&' ) {
3653 $entityState = 1; // entity found, (e.g. "&#160;")
3654 } else {
3655 $dispLen++; // this char is displayed
3656 // Add the next $max display text chars after this in one swoop...
3657 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3658 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3659 $dispLen += $skipped;
3660 $pos += $skipped;
3661 }
3662 }
3663 }
3664 }
3665 // Close the last tag if left unclosed by bad HTML
3666 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3667 while ( count( $openTags ) > 0 ) {
3668 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3669 }
3670 return $ret;
3671 }
3672
3673 /**
3674 * truncateHtml() helper function
3675 * like strcspn() but adds the skipped chars to $ret
3676 *
3677 * @param string $ret
3678 * @param string $text
3679 * @param string $search
3680 * @param int $start
3681 * @param null|int $len
3682 * @return int
3683 */
3684 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3685 if ( $len === null ) {
3686 $len = -1; // -1 means "no limit" for strcspn
3687 } elseif ( $len < 0 ) {
3688 $len = 0; // sanity
3689 }
3690 $skipCount = 0;
3691 if ( $start < strlen( $text ) ) {
3692 $skipCount = strcspn( $text, $search, $start, $len );
3693 $ret .= substr( $text, $start, $skipCount );
3694 }
3695 return $skipCount;
3696 }
3697
3698 /**
3699 * truncateHtml() helper function
3700 * (a) push or pop $tag from $openTags as needed
3701 * (b) clear $tag value
3702 * @param string &$tag Current HTML tag name we are looking at
3703 * @param int $tagType (0-open tag, 1-close tag)
3704 * @param string $lastCh Character before the '>' that ended this tag
3705 * @param array &$openTags Open tag stack (not accounting for $tag)
3706 */
3707 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3708 $tag = ltrim( $tag );
3709 if ( $tag != '' ) {
3710 if ( $tagType == 0 && $lastCh != '/' ) {
3711 $openTags[] = $tag; // tag opened (didn't close itself)
3712 } elseif ( $tagType == 1 ) {
3713 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3714 array_pop( $openTags ); // tag closed
3715 }
3716 }
3717 $tag = '';
3718 }
3719 }
3720
3721 /**
3722 * Grammatical transformations, needed for inflected languages
3723 * Invoked by putting {{grammar:case|word}} in a message
3724 *
3725 * @param string $word
3726 * @param string $case
3727 * @return string
3728 */
3729 function convertGrammar( $word, $case ) {
3730 global $wgGrammarForms;
3731 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3732 return $wgGrammarForms[$this->getCode()][$case][$word];
3733 }
3734
3735 return $word;
3736 }
3737 /**
3738 * Get the grammar forms for the content language
3739 * @return array Array of grammar forms
3740 * @since 1.20
3741 */
3742 function getGrammarForms() {
3743 global $wgGrammarForms;
3744 if ( isset( $wgGrammarForms[$this->getCode()] )
3745 && is_array( $wgGrammarForms[$this->getCode()] )
3746 ) {
3747 return $wgGrammarForms[$this->getCode()];
3748 }
3749
3750 return [];
3751 }
3752 /**
3753 * Provides an alternative text depending on specified gender.
3754 * Usage {{gender:username|masculine|feminine|unknown}}.
3755 * username is optional, in which case the gender of current user is used,
3756 * but only in (some) interface messages; otherwise default gender is used.
3757 *
3758 * If no forms are given, an empty string is returned. If only one form is
3759 * given, it will be returned unconditionally. These details are implied by
3760 * the caller and cannot be overridden in subclasses.
3761 *
3762 * If three forms are given, the default is to use the third (unknown) form.
3763 * If fewer than three forms are given, the default is to use the first (masculine) form.
3764 * These details can be overridden in subclasses.
3765 *
3766 * @param string $gender
3767 * @param array $forms
3768 *
3769 * @return string
3770 */
3771 function gender( $gender, $forms ) {
3772 if ( !count( $forms ) ) {
3773 return '';
3774 }
3775 $forms = $this->preConvertPlural( $forms, 2 );
3776 if ( $gender === 'male' ) {
3777 return $forms[0];
3778 }
3779 if ( $gender === 'female' ) {
3780 return $forms[1];
3781 }
3782 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3783 }
3784
3785 /**
3786 * Plural form transformations, needed for some languages.
3787 * For example, there are 3 form of plural in Russian and Polish,
3788 * depending on "count mod 10". See [[w:Plural]]
3789 * For English it is pretty simple.
3790 *
3791 * Invoked by putting {{plural:count|wordform1|wordform2}}
3792 * or {{plural:count|wordform1|wordform2|wordform3}}
3793 *
3794 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3795 *
3796 * @param int $count Non-localized number
3797 * @param array $forms Different plural forms
3798 * @return string Correct form of plural for $count in this language
3799 */
3800 function convertPlural( $count, $forms ) {
3801 // Handle explicit n=pluralform cases
3802 $forms = $this->handleExplicitPluralForms( $count, $forms );
3803 if ( is_string( $forms ) ) {
3804 return $forms;
3805 }
3806 if ( !count( $forms ) ) {
3807 return '';
3808 }
3809
3810 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3811 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3812 return $forms[$pluralForm];
3813 }
3814
3815 /**
3816 * Handles explicit plural forms for Language::convertPlural()
3817 *
3818 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3819 * If an explicitly defined plural form matches the $count, then
3820 * string value returned, otherwise array returned for further consideration
3821 * by CLDR rules or overridden convertPlural().
3822 *
3823 * @since 1.23
3824 *
3825 * @param int $count Non-localized number
3826 * @param array $forms Different plural forms
3827 *
3828 * @return array|string
3829 */
3830 protected function handleExplicitPluralForms( $count, array $forms ) {
3831 foreach ( $forms as $index => $form ) {
3832 if ( preg_match( '/\d+=/i', $form ) ) {
3833 $pos = strpos( $form, '=' );
3834 if ( substr( $form, 0, $pos ) === (string)$count ) {
3835 return substr( $form, $pos + 1 );
3836 }
3837 unset( $forms[$index] );
3838 }
3839 }
3840 return array_values( $forms );
3841 }
3842
3843 /**
3844 * Checks that convertPlural was given an array and pads it to requested
3845 * amount of forms by copying the last one.
3846 *
3847 * @param array $forms Array of forms given to convertPlural
3848 * @param int $count How many forms should there be at least
3849 * @return array Padded array of forms or an exception if not an array
3850 */
3851 protected function preConvertPlural( /* Array */ $forms, $count ) {
3852 while ( count( $forms ) < $count ) {
3853 $forms[] = $forms[count( $forms ) - 1];
3854 }
3855 return $forms;
3856 }
3857
3858 /**
3859 * Wraps argument with unicode control characters for directionality safety
3860 *
3861 * This solves the problem where directionality-neutral characters at the edge of
3862 * the argument string get interpreted with the wrong directionality from the
3863 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3864 *
3865 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3866 * directionality of the argument string, using the BIDI algorithm's own "First
3867 * strong directional codepoint" rule. Essentially, this works round the fact that
3868 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3869 * direction inference). The latter is cleaner but still not widely supported.
3870 *
3871 * @param string $text Text to wrap
3872 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3873 */
3874 public function embedBidi( $text = '' ) {
3875 $dir = Language::strongDirFromContent( $text );
3876 if ( $dir === 'ltr' ) {
3877 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3878 return self::$lre . $text . self::$pdf;
3879 }
3880 if ( $dir === 'rtl' ) {
3881 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3882 return self::$rle . $text . self::$pdf;
3883 }
3884 // No strong directionality: do not wrap
3885 return $text;
3886 }
3887
3888 /**
3889 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3890 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3891 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3892 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3893 * match up with it.
3894 *
3895 * @param string $str The validated block duration in English
3896 * @param User $user User object to use timezone from or null for $wgUser
3897 * @return string Somehow translated block duration
3898 * @see LanguageFi.php for example implementation
3899 */
3900 function translateBlockExpiry( $str, User $user = null ) {
3901 $duration = SpecialBlock::getSuggestedDurations( $this );
3902 foreach ( $duration as $show => $value ) {
3903 if ( strcmp( $str, $value ) == 0 ) {
3904 return htmlspecialchars( trim( $show ) );
3905 }
3906 }
3907
3908 if ( wfIsInfinity( $str ) ) {
3909 foreach ( $duration as $show => $value ) {
3910 if ( wfIsInfinity( $value ) ) {
3911 return htmlspecialchars( trim( $show ) );
3912 }
3913 }
3914 }
3915
3916 // If all else fails, return a standard duration or timestamp description.
3917 $time = strtotime( $str, 0 );
3918 if ( $time === false ) { // Unknown format. Return it as-is in case.
3919 return $str;
3920 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3921 // $time is relative to 0 so it's a duration length.
3922 return $this->formatDuration( $time );
3923 } else { // It's an absolute timestamp.
3924 if ( $time === 0 ) {
3925 // wfTimestamp() handles 0 as current time instead of epoch.
3926 $time = '19700101000000';
3927 }
3928 if ( $user ) {
3929 return $this->userTimeAndDate( $time, $user );
3930 }
3931 return $this->timeanddate( $time );
3932 }
3933 }
3934
3935 /**
3936 * languages like Chinese need to be segmented in order for the diff
3937 * to be of any use
3938 *
3939 * @param string $text
3940 * @return string
3941 */
3942 public function segmentForDiff( $text ) {
3943 return $text;
3944 }
3945
3946 /**
3947 * and unsegment to show the result
3948 *
3949 * @param string $text
3950 * @return string
3951 */
3952 public function unsegmentForDiff( $text ) {
3953 return $text;
3954 }
3955
3956 /**
3957 * Return the LanguageConverter used in the Language
3958 *
3959 * @since 1.19
3960 * @return LanguageConverter
3961 */
3962 public function getConverter() {
3963 return $this->mConverter;
3964 }
3965
3966 /**
3967 * convert text to all supported variants
3968 *
3969 * @param string $text
3970 * @return array
3971 */
3972 public function autoConvertToAllVariants( $text ) {
3973 return $this->mConverter->autoConvertToAllVariants( $text );
3974 }
3975
3976 /**
3977 * convert text to different variants of a language.
3978 *
3979 * @param string $text
3980 * @return string
3981 */
3982 public function convert( $text ) {
3983 return $this->mConverter->convert( $text );
3984 }
3985
3986 /**
3987 * Convert a Title object to a string in the preferred variant
3988 *
3989 * @param Title $title
3990 * @return string
3991 */
3992 public function convertTitle( $title ) {
3993 return $this->mConverter->convertTitle( $title );
3994 }
3995
3996 /**
3997 * Convert a namespace index to a string in the preferred variant
3998 *
3999 * @param int $ns
4000 * @return string
4001 */
4002 public function convertNamespace( $ns ) {
4003 return $this->mConverter->convertNamespace( $ns );
4004 }
4005
4006 /**
4007 * Check if this is a language with variants
4008 *
4009 * @return bool
4010 */
4011 public function hasVariants() {
4012 return count( $this->getVariants() ) > 1;
4013 }
4014
4015 /**
4016 * Check if the language has the specific variant
4017 *
4018 * @since 1.19
4019 * @param string $variant
4020 * @return bool
4021 */
4022 public function hasVariant( $variant ) {
4023 return (bool)$this->mConverter->validateVariant( $variant );
4024 }
4025
4026 /**
4027 * Perform output conversion on a string, and encode for safe HTML output.
4028 * @param string $text Text to be converted
4029 * @param bool $isTitle Whether this conversion is for the article title
4030 * @return string
4031 * @todo this should get integrated somewhere sane
4032 */
4033 public function convertHtml( $text, $isTitle = false ) {
4034 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4035 }
4036
4037 /**
4038 * @param string $key
4039 * @return string
4040 */
4041 public function convertCategoryKey( $key ) {
4042 return $this->mConverter->convertCategoryKey( $key );
4043 }
4044
4045 /**
4046 * Get the list of variants supported by this language
4047 * see sample implementation in LanguageZh.php
4048 *
4049 * @return array An array of language codes
4050 */
4051 public function getVariants() {
4052 return $this->mConverter->getVariants();
4053 }
4054
4055 /**
4056 * @return string
4057 */
4058 public function getPreferredVariant() {
4059 return $this->mConverter->getPreferredVariant();
4060 }
4061
4062 /**
4063 * @return string
4064 */
4065 public function getDefaultVariant() {
4066 return $this->mConverter->getDefaultVariant();
4067 }
4068
4069 /**
4070 * @return string
4071 */
4072 public function getURLVariant() {
4073 return $this->mConverter->getURLVariant();
4074 }
4075
4076 /**
4077 * If a language supports multiple variants, it is
4078 * possible that non-existing link in one variant
4079 * actually exists in another variant. this function
4080 * tries to find it. See e.g. LanguageZh.php
4081 * The input parameters may be modified upon return
4082 *
4083 * @param string &$link The name of the link
4084 * @param Title &$nt The title object of the link
4085 * @param bool $ignoreOtherCond To disable other conditions when
4086 * we need to transclude a template or update a category's link
4087 */
4088 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4089 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4090 }
4091
4092 /**
4093 * returns language specific options used by User::getPageRenderHash()
4094 * for example, the preferred language variant
4095 *
4096 * @return string
4097 */
4098 function getExtraHashOptions() {
4099 return $this->mConverter->getExtraHashOptions();
4100 }
4101
4102 /**
4103 * For languages that support multiple variants, the title of an
4104 * article may be displayed differently in different variants. this
4105 * function returns the apporiate title defined in the body of the article.
4106 *
4107 * @return string
4108 */
4109 public function getParsedTitle() {
4110 return $this->mConverter->getParsedTitle();
4111 }
4112
4113 /**
4114 * Refresh the cache of conversion tables when
4115 * MediaWiki:Conversiontable* is updated.
4116 *
4117 * @param Title $title The Title of the page being updated
4118 */
4119 public function updateConversionTable( Title $title ) {
4120 $this->mConverter->updateConversionTable( $title );
4121 }
4122
4123 /**
4124 * Prepare external link text for conversion. When the text is
4125 * a URL, it shouldn't be converted, and it'll be wrapped in
4126 * the "raw" tag (-{R| }-) to prevent conversion.
4127 *
4128 * This function is called "markNoConversion" for historical
4129 * reasons.
4130 *
4131 * @param string $text Text to be used for external link
4132 * @param bool $noParse Wrap it without confirming it's a real URL first
4133 * @return string The tagged text
4134 */
4135 public function markNoConversion( $text, $noParse = false ) {
4136 // Excluding protocal-relative URLs may avoid many false positives.
4137 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4138 return $this->mConverter->markNoConversion( $text );
4139 } else {
4140 return $text;
4141 }
4142 }
4143
4144 /**
4145 * A regular expression to match legal word-trailing characters
4146 * which should be merged onto a link of the form [[foo]]bar.
4147 *
4148 * @return string
4149 */
4150 public function linkTrail() {
4151 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4152 }
4153
4154 /**
4155 * A regular expression character set to match legal word-prefixing
4156 * characters which should be merged onto a link of the form foo[[bar]].
4157 *
4158 * @return string
4159 */
4160 public function linkPrefixCharset() {
4161 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4162 }
4163
4164 /**
4165 * Get the "parent" language which has a converter to convert a "compatible" language
4166 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4167 *
4168 * @return Language|null
4169 * @since 1.22
4170 */
4171 public function getParentLanguage() {
4172 if ( $this->mParentLanguage !== false ) {
4173 return $this->mParentLanguage;
4174 }
4175
4176 $code = explode( '-', $this->getCode() )[0];
4177 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4178 $this->mParentLanguage = null;
4179 return null;
4180 }
4181 $lang = Language::factory( $code );
4182 if ( !$lang->hasVariant( $this->getCode() ) ) {
4183 $this->mParentLanguage = null;
4184 return null;
4185 }
4186
4187 $this->mParentLanguage = $lang;
4188 return $lang;
4189 }
4190
4191 /**
4192 * Compare with an other language object
4193 *
4194 * @since 1.28
4195 * @param Language $lang
4196 * @return boolean
4197 */
4198 public function equals( Language $lang ) {
4199 return $lang->getCode() === $this->mCode;
4200 }
4201
4202 /**
4203 * Get the internal language code for this language object
4204 *
4205 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4206 * htmlspecialchars() or similar
4207 *
4208 * @return string
4209 */
4210 public function getCode() {
4211 return $this->mCode;
4212 }
4213
4214 /**
4215 * Get the code in BCP 47 format which we can use
4216 * inside of html lang="" tags.
4217 *
4218 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4219 * htmlspecialchars() or similar.
4220 *
4221 * @since 1.19
4222 * @return string
4223 */
4224 public function getHtmlCode() {
4225 if ( is_null( $this->mHtmlCode ) ) {
4226 $this->mHtmlCode = wfBCP47( $this->getCode() );
4227 }
4228 return $this->mHtmlCode;
4229 }
4230
4231 /**
4232 * @param string $code
4233 */
4234 public function setCode( $code ) {
4235 $this->mCode = $code;
4236 // Ensure we don't leave incorrect cached data lying around
4237 $this->mHtmlCode = null;
4238 $this->mParentLanguage = false;
4239 }
4240
4241 /**
4242 * Get the language code from a file name. Inverse of getFileName()
4243 * @param string $filename $prefix . $languageCode . $suffix
4244 * @param string $prefix Prefix before the language code
4245 * @param string $suffix Suffix after the language code
4246 * @return string Language code, or false if $prefix or $suffix isn't found
4247 */
4248 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4249 $m = null;
4250 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4251 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4252 if ( !count( $m ) ) {
4253 return false;
4254 }
4255 return str_replace( '_', '-', strtolower( $m[1] ) );
4256 }
4257
4258 /**
4259 * @param string $code
4260 * @return string Name of the language class
4261 */
4262 public static function classFromCode( $code ) {
4263 if ( $code == 'en' ) {
4264 return 'Language';
4265 } else {
4266 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4267 }
4268 }
4269
4270 /**
4271 * Get the name of a file for a certain language code
4272 * @param string $prefix Prepend this to the filename
4273 * @param string $code Language code
4274 * @param string $suffix Append this to the filename
4275 * @throws MWException
4276 * @return string $prefix . $mangledCode . $suffix
4277 */
4278 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4279 if ( !self::isValidBuiltInCode( $code ) ) {
4280 throw new MWException( "Invalid language code \"$code\"" );
4281 }
4282
4283 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4284 }
4285
4286 /**
4287 * @param string $code
4288 * @return string
4289 */
4290 public static function getMessagesFileName( $code ) {
4291 global $IP;
4292 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4293 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4294 return $file;
4295 }
4296
4297 /**
4298 * @param string $code
4299 * @return string
4300 * @throws MWException
4301 * @since 1.23
4302 */
4303 public static function getJsonMessagesFileName( $code ) {
4304 global $IP;
4305
4306 if ( !self::isValidBuiltInCode( $code ) ) {
4307 throw new MWException( "Invalid language code \"$code\"" );
4308 }
4309
4310 return "$IP/languages/i18n/$code.json";
4311 }
4312
4313 /**
4314 * Get the first fallback for a given language.
4315 *
4316 * @param string $code
4317 *
4318 * @return bool|string
4319 */
4320 public static function getFallbackFor( $code ) {
4321 $fallbacks = self::getFallbacksFor( $code );
4322 if ( $fallbacks ) {
4323 return $fallbacks[0];
4324 }
4325 return false;
4326 }
4327
4328 /**
4329 * Get the ordered list of fallback languages.
4330 *
4331 * @since 1.19
4332 * @param string $code Language code
4333 * @return array Non-empty array, ending in "en"
4334 */
4335 public static function getFallbacksFor( $code ) {
4336 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4337 return [];
4338 }
4339 // For unknown languages, fallbackSequence returns an empty array,
4340 // hardcode fallback to 'en' in that case.
4341 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4342 }
4343
4344 /**
4345 * Get the ordered list of fallback languages, ending with the fallback
4346 * language chain for the site language.
4347 *
4348 * @since 1.22
4349 * @param string $code Language code
4350 * @return array Array( fallbacks, site fallbacks )
4351 */
4352 public static function getFallbacksIncludingSiteLanguage( $code ) {
4353 global $wgLanguageCode;
4354
4355 // Usually, we will only store a tiny number of fallback chains, so we
4356 // keep them in static memory.
4357 $cacheKey = "{$code}-{$wgLanguageCode}";
4358
4359 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4360 $fallbacks = self::getFallbacksFor( $code );
4361
4362 // Append the site's fallback chain, including the site language itself
4363 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4364 array_unshift( $siteFallbacks, $wgLanguageCode );
4365
4366 // Eliminate any languages already included in the chain
4367 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4368
4369 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4370 }
4371 return self::$fallbackLanguageCache[$cacheKey];
4372 }
4373
4374 /**
4375 * Get all messages for a given language
4376 * WARNING: this may take a long time. If you just need all message *keys*
4377 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4378 *
4379 * @param string $code
4380 *
4381 * @return array
4382 */
4383 public static function getMessagesFor( $code ) {
4384 return self::getLocalisationCache()->getItem( $code, 'messages' );
4385 }
4386
4387 /**
4388 * Get a message for a given language
4389 *
4390 * @param string $key
4391 * @param string $code
4392 *
4393 * @return string
4394 */
4395 public static function getMessageFor( $key, $code ) {
4396 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4397 }
4398
4399 /**
4400 * Get all message keys for a given language. This is a faster alternative to
4401 * array_keys( Language::getMessagesFor( $code ) )
4402 *
4403 * @since 1.19
4404 * @param string $code Language code
4405 * @return array Array of message keys (strings)
4406 */
4407 public static function getMessageKeysFor( $code ) {
4408 return self::getLocalisationCache()->getSubitemList( $code, 'messages' );
4409 }
4410
4411 /**
4412 * @param string $talk
4413 * @return mixed
4414 */
4415 function fixVariableInNamespace( $talk ) {
4416 if ( strpos( $talk, '$1' ) === false ) {
4417 return $talk;
4418 }
4419
4420 global $wgMetaNamespace;
4421 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4422
4423 # Allow grammar transformations
4424 # Allowing full message-style parsing would make simple requests
4425 # such as action=raw much more expensive than they need to be.
4426 # This will hopefully cover most cases.
4427 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4428 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4429 return str_replace( ' ', '_', $talk );
4430 }
4431
4432 /**
4433 * @param string $m
4434 * @return string
4435 */
4436 function replaceGrammarInNamespace( $m ) {
4437 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4438 }
4439
4440 /**
4441 * Decode an expiry (block, protection, etc) which has come from the DB
4442 *
4443 * @param string $expiry Database expiry String
4444 * @param bool|int $format True to process using language functions, or TS_ constant
4445 * to return the expiry in a given timestamp
4446 * @param string $infinity If $format is not true, use this string for infinite expiry
4447 * @return string
4448 * @since 1.18
4449 */
4450 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4451 static $dbInfinity;
4452 if ( $dbInfinity === null ) {
4453 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4454 }
4455
4456 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4457 return $format === true
4458 ? $this->getMessageFromDB( 'infiniteblock' )
4459 : $infinity;
4460 } else {
4461 return $format === true
4462 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4463 : wfTimestamp( $format, $expiry );
4464 }
4465 }
4466
4467 /**
4468 * @todo Document
4469 * @param int|float $seconds
4470 * @param array $format Optional
4471 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4472 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4473 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4474 * and friends.
4475 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4476 * or 'avoidminutes'.
4477 * @return string
4478 */
4479 function formatTimePeriod( $seconds, $format = [] ) {
4480 if ( !is_array( $format ) ) {
4481 $format = [ 'avoid' => $format ]; // For backwards compatibility
4482 }
4483 if ( !isset( $format['avoid'] ) ) {
4484 $format['avoid'] = false;
4485 }
4486 if ( !isset( $format['noabbrevs'] ) ) {
4487 $format['noabbrevs'] = false;
4488 }
4489 $secondsMsg = wfMessage(
4490 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4491 $minutesMsg = wfMessage(
4492 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4493 $hoursMsg = wfMessage(
4494 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4495 $daysMsg = wfMessage(
4496 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4497
4498 if ( round( $seconds * 10 ) < 100 ) {
4499 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4500 $s = $secondsMsg->params( $s )->text();
4501 } elseif ( round( $seconds ) < 60 ) {
4502 $s = $this->formatNum( round( $seconds ) );
4503 $s = $secondsMsg->params( $s )->text();
4504 } elseif ( round( $seconds ) < 3600 ) {
4505 $minutes = floor( $seconds / 60 );
4506 $secondsPart = round( fmod( $seconds, 60 ) );
4507 if ( $secondsPart == 60 ) {
4508 $secondsPart = 0;
4509 $minutes++;
4510 }
4511 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4512 $s .= ' ';
4513 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4514 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4515 $hours = floor( $seconds / 3600 );
4516 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4517 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4518 if ( $secondsPart == 60 ) {
4519 $secondsPart = 0;
4520 $minutes++;
4521 }
4522 if ( $minutes == 60 ) {
4523 $minutes = 0;
4524 $hours++;
4525 }
4526 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4527 $s .= ' ';
4528 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4529 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4530 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4531 }
4532 } else {
4533 $days = floor( $seconds / 86400 );
4534 if ( $format['avoid'] === 'avoidminutes' ) {
4535 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4536 if ( $hours == 24 ) {
4537 $hours = 0;
4538 $days++;
4539 }
4540 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4541 $s .= ' ';
4542 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4543 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4544 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4545 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4546 if ( $minutes == 60 ) {
4547 $minutes = 0;
4548 $hours++;
4549 }
4550 if ( $hours == 24 ) {
4551 $hours = 0;
4552 $days++;
4553 }
4554 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4555 $s .= ' ';
4556 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4557 $s .= ' ';
4558 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4559 } else {
4560 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4561 $s .= ' ';
4562 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4563 }
4564 }
4565 return $s;
4566 }
4567
4568 /**
4569 * Format a bitrate for output, using an appropriate
4570 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4571 * the magnitude in question.
4572 *
4573 * This use base 1000. For base 1024 use formatSize(), for another base
4574 * see formatComputingNumbers().
4575 *
4576 * @param int $bps
4577 * @return string
4578 */
4579 function formatBitrate( $bps ) {
4580 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4581 }
4582
4583 /**
4584 * @param int $size Size of the unit
4585 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4586 * @param string $messageKey Message key to be uesd
4587 * @return string
4588 */
4589 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4590 if ( $size <= 0 ) {
4591 return str_replace( '$1', $this->formatNum( $size ),
4592 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4593 );
4594 }
4595 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4596 $index = 0;
4597
4598 $maxIndex = count( $sizes ) - 1;
4599 while ( $size >= $boundary && $index < $maxIndex ) {
4600 $index++;
4601 $size /= $boundary;
4602 }
4603
4604 // For small sizes no decimal places necessary
4605 $round = 0;
4606 if ( $index > 1 ) {
4607 // For MB and bigger two decimal places are smarter
4608 $round = 2;
4609 }
4610 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4611
4612 $size = round( $size, $round );
4613 $text = $this->getMessageFromDB( $msg );
4614 return str_replace( '$1', $this->formatNum( $size ), $text );
4615 }
4616
4617 /**
4618 * Format a size in bytes for output, using an appropriate
4619 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4620 *
4621 * This method use base 1024. For base 1000 use formatBitrate(), for
4622 * another base see formatComputingNumbers()
4623 *
4624 * @param int $size Size to format
4625 * @return string Plain text (not HTML)
4626 */
4627 function formatSize( $size ) {
4628 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4629 }
4630
4631 /**
4632 * Make a list item, used by various special pages
4633 *
4634 * @param string $page Page link
4635 * @param string $details HTML safe text between brackets
4636 * @param bool $oppositedm Add the direction mark opposite to your
4637 * language, to display text properly
4638 * @return HTML escaped string
4639 */
4640 function specialList( $page, $details, $oppositedm = true ) {
4641 if ( !$details ) {
4642 return $page;
4643 }
4644
4645 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4646 return
4647 $page .
4648 $dirmark .
4649 $this->msg( 'word-separator' )->escaped() .
4650 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4651 }
4652
4653 /**
4654 * Generate (prev x| next x) (20|50|100...) type links for paging
4655 *
4656 * @param Title $title Title object to link
4657 * @param int $offset
4658 * @param int $limit
4659 * @param array $query Optional URL query parameter string
4660 * @param bool $atend Optional param for specified if this is the last page
4661 * @return string
4662 */
4663 public function viewPrevNext( Title $title, $offset, $limit,
4664 array $query = [], $atend = false
4665 ) {
4666 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4667
4668 # Make 'previous' link
4669 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4670 if ( $offset > 0 ) {
4671 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4672 $query, $prev, 'prevn-title', 'mw-prevlink' );
4673 } else {
4674 $plink = htmlspecialchars( $prev );
4675 }
4676
4677 # Make 'next' link
4678 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4679 if ( $atend ) {
4680 $nlink = htmlspecialchars( $next );
4681 } else {
4682 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4683 $query, $next, 'nextn-title', 'mw-nextlink' );
4684 }
4685
4686 # Make links to set number of items per page
4687 $numLinks = [];
4688 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4689 $numLinks[] = $this->numLink( $title, $offset, $num,
4690 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4691 }
4692
4693 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4694 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4695 }
4696
4697 /**
4698 * Helper function for viewPrevNext() that generates links
4699 *
4700 * @param Title $title Title object to link
4701 * @param int $offset
4702 * @param int $limit
4703 * @param array $query Extra query parameters
4704 * @param string $link Text to use for the link; will be escaped
4705 * @param string $tooltipMsg Name of the message to use as tooltip
4706 * @param string $class Value of the "class" attribute of the link
4707 * @return string HTML fragment
4708 */
4709 private function numLink( Title $title, $offset, $limit, array $query, $link,
4710 $tooltipMsg, $class
4711 ) {
4712 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4713 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4714 ->numParams( $limit )->text();
4715
4716 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4717 'title' => $tooltip, 'class' => $class ], $link );
4718 }
4719
4720 /**
4721 * Get the conversion rule title, if any.
4722 *
4723 * @return string
4724 */
4725 public function getConvRuleTitle() {
4726 return $this->mConverter->getConvRuleTitle();
4727 }
4728
4729 /**
4730 * Get the compiled plural rules for the language
4731 * @since 1.20
4732 * @return array Associative array with plural form, and plural rule as key-value pairs
4733 */
4734 public function getCompiledPluralRules() {
4735 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4736 $fallbacks = Language::getFallbacksFor( $this->mCode );
4737 if ( !$pluralRules ) {
4738 foreach ( $fallbacks as $fallbackCode ) {
4739 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4740 if ( $pluralRules ) {
4741 break;
4742 }
4743 }
4744 }
4745 return $pluralRules;
4746 }
4747
4748 /**
4749 * Get the plural rules for the language
4750 * @since 1.20
4751 * @return array Associative array with plural form number and plural rule as key-value pairs
4752 */
4753 public function getPluralRules() {
4754 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4755 $fallbacks = Language::getFallbacksFor( $this->mCode );
4756 if ( !$pluralRules ) {
4757 foreach ( $fallbacks as $fallbackCode ) {
4758 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4759 if ( $pluralRules ) {
4760 break;
4761 }
4762 }
4763 }
4764 return $pluralRules;
4765 }
4766
4767 /**
4768 * Get the plural rule types for the language
4769 * @since 1.22
4770 * @return array Associative array with plural form number and plural rule type as key-value pairs
4771 */
4772 public function getPluralRuleTypes() {
4773 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4774 $fallbacks = Language::getFallbacksFor( $this->mCode );
4775 if ( !$pluralRuleTypes ) {
4776 foreach ( $fallbacks as $fallbackCode ) {
4777 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4778 if ( $pluralRuleTypes ) {
4779 break;
4780 }
4781 }
4782 }
4783 return $pluralRuleTypes;
4784 }
4785
4786 /**
4787 * Find the index number of the plural rule appropriate for the given number
4788 * @param int $number
4789 * @return int The index number of the plural rule
4790 */
4791 public function getPluralRuleIndexNumber( $number ) {
4792 $pluralRules = $this->getCompiledPluralRules();
4793 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4794 return $form;
4795 }
4796
4797 /**
4798 * Find the plural rule type appropriate for the given number
4799 * For example, if the language is set to Arabic, getPluralType(5) should
4800 * return 'few'.
4801 * @since 1.22
4802 * @param int $number
4803 * @return string The name of the plural rule type, e.g. one, two, few, many
4804 */
4805 public function getPluralRuleType( $number ) {
4806 $index = $this->getPluralRuleIndexNumber( $number );
4807 $pluralRuleTypes = $this->getPluralRuleTypes();
4808 if ( isset( $pluralRuleTypes[$index] ) ) {
4809 return $pluralRuleTypes[$index];
4810 } else {
4811 return 'other';
4812 }
4813 }
4814 }
Wiklou, le Wiki du Wiklou