?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Merge "Update docs/hooks.txt for ShowSearchHitTitle"
[mediawiki.git] / languages / Language.php
blob5bce76bfcc64b437e70ffc083073f39eebccf7d6
1 <?php
2 /**
3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
5 *
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
10 *
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
15 *
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
20 *
21 * @file
22 * @ingroup Language
23 */
24
25 /**
26 * @defgroup Language Language
27 */
28
29 use CLDRPluralRuleParser\Evaluator;
30
31 /**
32 * Internationalisation code
33 * @ingroup Language
34 */
35 class Language {
36 /**
37 * @var LanguageConverter
38 */
39 public $mConverter;
40
41 public $mVariants, $mCode, $mLoaded = false;
42 public $mMagicExtensions = [], $mMagicHookDone = false;
43 private $mHtmlCode = null, $mParentLanguage = false;
44
45 public $dateFormatStrings = [];
46 public $mExtendedSpecialPageAliases;
47
48 /** @var array|null */
49 protected $namespaceNames;
50 protected $mNamespaceIds, $namespaceAliases;
51
52 /**
53 * ReplacementArray object caches
54 */
55 public $transformData = [];
56
57 /**
58 * @var LocalisationCache
59 */
60 static public $dataCache;
61
62 static public $mLangObjCache = [];
63
64 static public $mWeekdayMsgs = [
65 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
66 'friday', 'saturday'
67 ];
68
69 static public $mWeekdayAbbrevMsgs = [
70 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
71 ];
72
73 static public $mMonthMsgs = [
74 'january', 'february', 'march', 'april', 'may_long', 'june',
75 'july', 'august', 'september', 'october', 'november',
76 'december'
77 ];
78 static public $mMonthGenMsgs = [
79 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
80 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
81 'december-gen'
82 ];
83 static public $mMonthAbbrevMsgs = [
84 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
85 'sep', 'oct', 'nov', 'dec'
86 ];
87
88 static public $mIranianCalendarMonthMsgs = [
89 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
90 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
91 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
92 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
93 ];
94
95 static public $mHebrewCalendarMonthMsgs = [
96 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
97 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
98 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
99 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
100 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
101 ];
102
103 static public $mHebrewCalendarMonthGenMsgs = [
104 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
105 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
106 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
107 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
108 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
109 ];
110
111 static public $mHijriCalendarMonthMsgs = [
112 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
113 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
114 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
115 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
116 ];
117
118 /**
119 * @since 1.20
120 * @var array
121 */
122 static public $durationIntervals = [
123 'millennia' => 31556952000,
124 'centuries' => 3155695200,
125 'decades' => 315569520,
126 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
127 'weeks' => 604800,
128 'days' => 86400,
129 'hours' => 3600,
130 'minutes' => 60,
131 'seconds' => 1,
132 ];
133
134 /**
135 * Cache for language fallbacks.
136 * @see Language::getFallbacksIncludingSiteLanguage
137 * @since 1.21
138 * @var array
139 */
140 static private $fallbackLanguageCache = [];
141
142 /**
143 * Cache for grammar rules data
144 * @var MapCacheLRU|null
145 */
146 static private $grammarTransformations;
147
148 /**
149 * Cache for language names
150 * @var HashBagOStuff|null
151 */
152 static private $languageNameCache;
153
154 /**
155 * Unicode directional formatting characters, for embedBidi()
156 */
157 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
158 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
159 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
160
161 /**
162 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
163 * - in group 1 if it is LTR
164 * - in group 2 if it is RTL
165 * Does not match if there is no strong directionality codepoint.
166 *
167 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
168 *
169 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
170 * https://git.wikimedia.org/summary/unicodejs.git .
171 */
172 // @codingStandardsIgnoreStart
173 // @codeCoverageIgnoreStart
174 static private $strongDirRegex = '/(?:([\x{41}-\x{5a}\x{61}-\x{7a}\x{aa}\x{b5}\x{ba}\x{c0}-\x{d6}\x{d8}-\x{f6}\x{f8}-\x{2b8}\x{2bb}-\x{2c1}\x{2d0}\x{2d1}\x{2e0}-\x{2e4}\x{2ee}\x{370}-\x{373}\x{376}\x{377}\x{37a}-\x{37d}\x{37f}\x{386}\x{388}-\x{38a}\x{38c}\x{38e}-\x{3a1}\x{3a3}-\x{3f5}\x{3f7}-\x{482}\x{48a}-\x{52f}\x{531}-\x{556}\x{559}-\x{55f}\x{561}-\x{587}\x{589}\x{903}-\x{939}\x{93b}\x{93d}-\x{940}\x{949}-\x{94c}\x{94e}-\x{950}\x{958}-\x{961}\x{964}-\x{980}\x{982}\x{983}\x{985}-\x{98c}\x{98f}\x{990}\x{993}-\x{9a8}\x{9aa}-\x{9b0}\x{9b2}\x{9b6}-\x{9b9}\x{9bd}-\x{9c0}\x{9c7}\x{9c8}\x{9cb}\x{9cc}\x{9ce}\x{9d7}\x{9dc}\x{9dd}\x{9df}-\x{9e1}\x{9e6}-\x{9f1}\x{9f4}-\x{9fa}\x{a03}\x{a05}-\x{a0a}\x{a0f}\x{a10}\x{a13}-\x{a28}\x{a2a}-\x{a30}\x{a32}\x{a33}\x{a35}\x{a36}\x{a38}\x{a39}\x{a3e}-\x{a40}\x{a59}-\x{a5c}\x{a5e}\x{a66}-\x{a6f}\x{a72}-\x{a74}\x{a83}\x{a85}-\x{a8d}\x{a8f}-\x{a91}\x{a93}-\x{aa8}\x{aaa}-\x{ab0}\x{ab2}\x{ab3}\x{ab5}-\x{ab9}\x{abd}-\x{ac0}\x{ac9}\x{acb}\x{acc}\x{ad0}\x{ae0}\x{ae1}\x{ae6}-\x{af0}\x{af9}\x{b02}\x{b03}\x{b05}-\x{b0c}\x{b0f}\x{b10}\x{b13}-\x{b28}\x{b2a}-\x{b30}\x{b32}\x{b33}\x{b35}-\x{b39}\x{b3d}\x{b3e}\x{b40}\x{b47}\x{b48}\x{b4b}\x{b4c}\x{b57}\x{b5c}\x{b5d}\x{b5f}-\x{b61}\x{b66}-\x{b77}\x{b83}\x{b85}-\x{b8a}\x{b8e}-\x{b90}\x{b92}-\x{b95}\x{b99}\x{b9a}\x{b9c}\x{b9e}\x{b9f}\x{ba3}\x{ba4}\x{ba8}-\x{baa}\x{bae}-\x{bb9}\x{bbe}\x{bbf}\x{bc1}\x{bc2}\x{bc6}-\x{bc8}\x{bca}-\x{bcc}\x{bd0}\x{bd7}\x{be6}-\x{bf2}\x{c01}-\x{c03}\x{c05}-\x{c0c}\x{c0e}-\x{c10}\x{c12}-\x{c28}\x{c2a}-\x{c39}\x{c3d}\x{c41}-\x{c44}\x{c58}-\x{c5a}\x{c60}\x{c61}\x{c66}-\x{c6f}\x{c7f}\x{c82}\x{c83}\x{c85}-\x{c8c}\x{c8e}-\x{c90}\x{c92}-\x{ca8}\x{caa}-\x{cb3}\x{cb5}-\x{cb9}\x{cbd}-\x{cc4}\x{cc6}-\x{cc8}\x{cca}\x{ccb}\x{cd5}\x{cd6}\x{cde}\x{ce0}\x{ce1}\x{ce6}-\x{cef}\x{cf1}\x{cf2}\x{d02}\x{d03}\x{d05}-\x{d0c}\x{d0e}-\x{d10}\x{d12}-\x{d3a}\x{d3d}-\x{d40}\x{d46}-\x{d48}\x{d4a}-\x{d4c}\x{d4e}\x{d57}\x{d5f}-\x{d61}\x{d66}-\x{d75}\x{d79}-\x{d7f}\x{d82}\x{d83}\x{d85}-\x{d96}\x{d9a}-\x{db1}\x{db3}-\x{dbb}\x{dbd}\x{dc0}-\x{dc6}\x{dcf}-\x{dd1}\x{dd8}-\x{ddf}\x{de6}-\x{def}\x{df2}-\x{df4}\x{e01}-\x{e30}\x{e32}\x{e33}\x{e40}-\x{e46}\x{e4f}-\x{e5b}\x{e81}\x{e82}\x{e84}\x{e87}\x{e88}\x{e8a}\x{e8d}\x{e94}-\x{e97}\x{e99}-\x{e9f}\x{ea1}-\x{ea3}\x{ea5}\x{ea7}\x{eaa}\x{eab}\x{ead}-\x{eb0}\x{eb2}\x{eb3}\x{ebd}\x{ec0}-\x{ec4}\x{ec6}\x{ed0}-\x{ed9}\x{edc}-\x{edf}\x{f00}-\x{f17}\x{f1a}-\x{f34}\x{f36}\x{f38}\x{f3e}-\x{f47}\x{f49}-\x{f6c}\x{f7f}\x{f85}\x{f88}-\x{f8c}\x{fbe}-\x{fc5}\x{fc7}-\x{fcc}\x{fce}-\x{fda}\x{1000}-\x{102c}\x{1031}\x{1038}\x{103b}\x{103c}\x{103f}-\x{1057}\x{105a}-\x{105d}\x{1061}-\x{1070}\x{1075}-\x{1081}\x{1083}\x{1084}\x{1087}-\x{108c}\x{108e}-\x{109c}\x{109e}-\x{10c5}\x{10c7}\x{10cd}\x{10d0}-\x{1248}\x{124a}-\x{124d}\x{1250}-\x{1256}\x{1258}\x{125a}-\x{125d}\x{1260}-\x{1288}\x{128a}-\x{128d}\x{1290}-\x{12b0}\x{12b2}-\x{12b5}\x{12b8}-\x{12be}\x{12c0}\x{12c2}-\x{12c5}\x{12c8}-\x{12d6}\x{12d8}-\x{1310}\x{1312}-\x{1315}\x{1318}-\x{135a}\x{1360}-\x{137c}\x{1380}-\x{138f}\x{13a0}-\x{13f5}\x{13f8}-\x{13fd}\x{1401}-\x{167f}\x{1681}-\x{169a}\x{16a0}-\x{16f8}\x{1700}-\x{170c}\x{170e}-\x{1711}\x{1720}-\x{1731}\x{1735}\x{1736}\x{1740}-\x{1751}\x{1760}-\x{176c}\x{176e}-\x{1770}\x{1780}-\x{17b3}\x{17b6}\x{17be}-\x{17c5}\x{17c7}\x{17c8}\x{17d4}-\x{17da}\x{17dc}\x{17e0}-\x{17e9}\x{1810}-\x{1819}\x{1820}-\x{1877}\x{1880}-\x{18a8}\x{18aa}\x{18b0}-\x{18f5}\x{1900}-\x{191e}\x{1923}-\x{1926}\x{1929}-\x{192b}\x{1930}\x{1931}\x{1933}-\x{1938}\x{1946}-\x{196d}\x{1970}-\x{1974}\x{1980}-\x{19ab}\x{19b0}-\x{19c9}\x{19d0}-\x{19da}\x{1a00}-\x{1a16}\x{1a19}\x{1a1a}\x{1a1e}-\x{1a55}\x{1a57}\x{1a61}\x{1a63}\x{1a64}\x{1a6d}-\x{1a72}\x{1a80}-\x{1a89}\x{1a90}-\x{1a99}\x{1aa0}-\x{1aad}\x{1b04}-\x{1b33}\x{1b35}\x{1b3b}\x{1b3d}-\x{1b41}\x{1b43}-\x{1b4b}\x{1b50}-\x{1b6a}\x{1b74}-\x{1b7c}\x{1b82}-\x{1ba1}\x{1ba6}\x{1ba7}\x{1baa}\x{1bae}-\x{1be5}\x{1be7}\x{1bea}-\x{1bec}\x{1bee}\x{1bf2}\x{1bf3}\x{1bfc}-\x{1c2b}\x{1c34}\x{1c35}\x{1c3b}-\x{1c49}\x{1c4d}-\x{1c7f}\x{1cc0}-\x{1cc7}\x{1cd3}\x{1ce1}\x{1ce9}-\x{1cec}\x{1cee}-\x{1cf3}\x{1cf5}\x{1cf6}\x{1d00}-\x{1dbf}\x{1e00}-\x{1f15}\x{1f18}-\x{1f1d}\x{1f20}-\x{1f45}\x{1f48}-\x{1f4d}\x{1f50}-\x{1f57}\x{1f59}\x{1f5b}\x{1f5d}\x{1f5f}-\x{1f7d}\x{1f80}-\x{1fb4}\x{1fb6}-\x{1fbc}\x{1fbe}\x{1fc2}-\x{1fc4}\x{1fc6}-\x{1fcc}\x{1fd0}-\x{1fd3}\x{1fd6}-\x{1fdb}\x{1fe0}-\x{1fec}\x{1ff2}-\x{1ff4}\x{1ff6}-\x{1ffc}\x{200e}\x{2071}\x{207f}\x{2090}-\x{209c}\x{2102}\x{2107}\x{210a}-\x{2113}\x{2115}\x{2119}-\x{211d}\x{2124}\x{2126}\x{2128}\x{212a}-\x{212d}\x{212f}-\x{2139}\x{213c}-\x{213f}\x{2145}-\x{2149}\x{214e}\x{214f}\x{2160}-\x{2188}\x{2336}-\x{237a}\x{2395}\x{249c}-\x{24e9}\x{26ac}\x{2800}-\x{28ff}\x{2c00}-\x{2c2e}\x{2c30}-\x{2c5e}\x{2c60}-\x{2ce4}\x{2ceb}-\x{2cee}\x{2cf2}\x{2cf3}\x{2d00}-\x{2d25}\x{2d27}\x{2d2d}\x{2d30}-\x{2d67}\x{2d6f}\x{2d70}\x{2d80}-\x{2d96}\x{2da0}-\x{2da6}\x{2da8}-\x{2dae}\x{2db0}-\x{2db6}\x{2db8}-\x{2dbe}\x{2dc0}-\x{2dc6}\x{2dc8}-\x{2dce}\x{2dd0}-\x{2dd6}\x{2dd8}-\x{2dde}\x{3005}-\x{3007}\x{3021}-\x{3029}\x{302e}\x{302f}\x{3031}-\x{3035}\x{3038}-\x{303c}\x{3041}-\x{3096}\x{309d}-\x{309f}\x{30a1}-\x{30fa}\x{30fc}-\x{30ff}\x{3105}-\x{312d}\x{3131}-\x{318e}\x{3190}-\x{31ba}\x{31f0}-\x{321c}\x{3220}-\x{324f}\x{3260}-\x{327b}\x{327f}-\x{32b0}\x{32c0}-\x{32cb}\x{32d0}-\x{32fe}\x{3300}-\x{3376}\x{337b}-\x{33dd}\x{33e0}-\x{33fe}\x{3400}-\x{4db5}\x{4e00}-\x{9fd5}\x{a000}-\x{a48c}\x{a4d0}-\x{a60c}\x{a610}-\x{a62b}\x{a640}-\x{a66e}\x{a680}-\x{a69d}\x{a6a0}-\x{a6ef}\x{a6f2}-\x{a6f7}\x{a722}-\x{a787}\x{a789}-\x{a7ad}\x{a7b0}-\x{a7b7}\x{a7f7}-\x{a801}\x{a803}-\x{a805}\x{a807}-\x{a80a}\x{a80c}-\x{a824}\x{a827}\x{a830}-\x{a837}\x{a840}-\x{a873}\x{a880}-\x{a8c3}\x{a8ce}-\x{a8d9}\x{a8f2}-\x{a8fd}\x{a900}-\x{a925}\x{a92e}-\x{a946}\x{a952}\x{a953}\x{a95f}-\x{a97c}\x{a983}-\x{a9b2}\x{a9b4}\x{a9b5}\x{a9ba}\x{a9bb}\x{a9bd}-\x{a9cd}\x{a9cf}-\x{a9d9}\x{a9de}-\x{a9e4}\x{a9e6}-\x{a9fe}\x{aa00}-\x{aa28}\x{aa2f}\x{aa30}\x{aa33}\x{aa34}\x{aa40}-\x{aa42}\x{aa44}-\x{aa4b}\x{aa4d}\x{aa50}-\x{aa59}\x{aa5c}-\x{aa7b}\x{aa7d}-\x{aaaf}\x{aab1}\x{aab5}\x{aab6}\x{aab9}-\x{aabd}\x{aac0}\x{aac2}\x{aadb}-\x{aaeb}\x{aaee}-\x{aaf5}\x{ab01}-\x{ab06}\x{ab09}-\x{ab0e}\x{ab11}-\x{ab16}\x{ab20}-\x{ab26}\x{ab28}-\x{ab2e}\x{ab30}-\x{ab65}\x{ab70}-\x{abe4}\x{abe6}\x{abe7}\x{abe9}-\x{abec}\x{abf0}-\x{abf9}\x{ac00}-\x{d7a3}\x{d7b0}-\x{d7c6}\x{d7cb}-\x{d7fb}\x{e000}-\x{fa6d}\x{fa70}-\x{fad9}\x{fb00}-\x{fb06}\x{fb13}-\x{fb17}\x{ff21}-\x{ff3a}\x{ff41}-\x{ff5a}\x{ff66}-\x{ffbe}\x{ffc2}-\x{ffc7}\x{ffca}-\x{ffcf}\x{ffd2}-\x{ffd7}\x{ffda}-\x{ffdc}\x{10000}-\x{1000b}\x{1000d}-\x{10026}\x{10028}-\x{1003a}\x{1003c}\x{1003d}\x{1003f}-\x{1004d}\x{10050}-\x{1005d}\x{10080}-\x{100fa}\x{10100}\x{10102}\x{10107}-\x{10133}\x{10137}-\x{1013f}\x{101d0}-\x{101fc}\x{10280}-\x{1029c}\x{102a0}-\x{102d0}\x{10300}-\x{10323}\x{10330}-\x{1034a}\x{10350}-\x{10375}\x{10380}-\x{1039d}\x{1039f}-\x{103c3}\x{103c8}-\x{103d5}\x{10400}-\x{1049d}\x{104a0}-\x{104a9}\x{10500}-\x{10527}\x{10530}-\x{10563}\x{1056f}\x{10600}-\x{10736}\x{10740}-\x{10755}\x{10760}-\x{10767}\x{11000}\x{11002}-\x{11037}\x{11047}-\x{1104d}\x{11066}-\x{1106f}\x{11082}-\x{110b2}\x{110b7}\x{110b8}\x{110bb}-\x{110c1}\x{110d0}-\x{110e8}\x{110f0}-\x{110f9}\x{11103}-\x{11126}\x{1112c}\x{11136}-\x{11143}\x{11150}-\x{11172}\x{11174}-\x{11176}\x{11182}-\x{111b5}\x{111bf}-\x{111c9}\x{111cd}\x{111d0}-\x{111df}\x{111e1}-\x{111f4}\x{11200}-\x{11211}\x{11213}-\x{1122e}\x{11232}\x{11233}\x{11235}\x{11238}-\x{1123d}\x{11280}-\x{11286}\x{11288}\x{1128a}-\x{1128d}\x{1128f}-\x{1129d}\x{1129f}-\x{112a9}\x{112b0}-\x{112de}\x{112e0}-\x{112e2}\x{112f0}-\x{112f9}\x{11302}\x{11303}\x{11305}-\x{1130c}\x{1130f}\x{11310}\x{11313}-\x{11328}\x{1132a}-\x{11330}\x{11332}\x{11333}\x{11335}-\x{11339}\x{1133d}-\x{1133f}\x{11341}-\x{11344}\x{11347}\x{11348}\x{1134b}-\x{1134d}\x{11350}\x{11357}\x{1135d}-\x{11363}\x{11480}-\x{114b2}\x{114b9}\x{114bb}-\x{114be}\x{114c1}\x{114c4}-\x{114c7}\x{114d0}-\x{114d9}\x{11580}-\x{115b1}\x{115b8}-\x{115bb}\x{115be}\x{115c1}-\x{115db}\x{11600}-\x{11632}\x{1163b}\x{1163c}\x{1163e}\x{11641}-\x{11644}\x{11650}-\x{11659}\x{11680}-\x{116aa}\x{116ac}\x{116ae}\x{116af}\x{116b6}\x{116c0}-\x{116c9}\x{11700}-\x{11719}\x{11720}\x{11721}\x{11726}\x{11730}-\x{1173f}\x{118a0}-\x{118f2}\x{118ff}\x{11ac0}-\x{11af8}\x{12000}-\x{12399}\x{12400}-\x{1246e}\x{12470}-\x{12474}\x{12480}-\x{12543}\x{13000}-\x{1342e}\x{14400}-\x{14646}\x{16800}-\x{16a38}\x{16a40}-\x{16a5e}\x{16a60}-\x{16a69}\x{16a6e}\x{16a6f}\x{16ad0}-\x{16aed}\x{16af5}\x{16b00}-\x{16b2f}\x{16b37}-\x{16b45}\x{16b50}-\x{16b59}\x{16b5b}-\x{16b61}\x{16b63}-\x{16b77}\x{16b7d}-\x{16b8f}\x{16f00}-\x{16f44}\x{16f50}-\x{16f7e}\x{16f93}-\x{16f9f}\x{1b000}\x{1b001}\x{1bc00}-\x{1bc6a}\x{1bc70}-\x{1bc7c}\x{1bc80}-\x{1bc88}\x{1bc90}-\x{1bc99}\x{1bc9c}\x{1bc9f}\x{1d000}-\x{1d0f5}\x{1d100}-\x{1d126}\x{1d129}-\x{1d166}\x{1d16a}-\x{1d172}\x{1d183}\x{1d184}\x{1d18c}-\x{1d1a9}\x{1d1ae}-\x{1d1e8}\x{1d360}-\x{1d371}\x{1d400}-\x{1d454}\x{1d456}-\x{1d49c}\x{1d49e}\x{1d49f}\x{1d4a2}\x{1d4a5}\x{1d4a6}\x{1d4a9}-\x{1d4ac}\x{1d4ae}-\x{1d4b9}\x{1d4bb}\x{1d4bd}-\x{1d4c3}\x{1d4c5}-\x{1d505}\x{1d507}-\x{1d50a}\x{1d50d}-\x{1d514}\x{1d516}-\x{1d51c}\x{1d51e}-\x{1d539}\x{1d53b}-\x{1d53e}\x{1d540}-\x{1d544}\x{1d546}\x{1d54a}-\x{1d550}\x{1d552}-\x{1d6a5}\x{1d6a8}-\x{1d6da}\x{1d6dc}-\x{1d714}\x{1d716}-\x{1d74e}\x{1d750}-\x{1d788}\x{1d78a}-\x{1d7c2}\x{1d7c4}-\x{1d7cb}\x{1d800}-\x{1d9ff}\x{1da37}-\x{1da3a}\x{1da6d}-\x{1da74}\x{1da76}-\x{1da83}\x{1da85}-\x{1da8b}\x{1f110}-\x{1f12e}\x{1f130}-\x{1f169}\x{1f170}-\x{1f19a}\x{1f1e6}-\x{1f202}\x{1f210}-\x{1f23a}\x{1f240}-\x{1f248}\x{1f250}\x{1f251}\x{20000}-\x{2a6d6}\x{2a700}-\x{2b734}\x{2b740}-\x{2b81d}\x{2b820}-\x{2cea1}\x{2f800}-\x{2fa1d}\x{f0000}-\x{ffffd}\x{100000}-\x{10fffd}])|([\x{590}\x{5be}\x{5c0}\x{5c3}\x{5c6}\x{5c8}-\x{5ff}\x{7c0}-\x{7ea}\x{7f4}\x{7f5}\x{7fa}-\x{815}\x{81a}\x{824}\x{828}\x{82e}-\x{858}\x{85c}-\x{89f}\x{200f}\x{fb1d}\x{fb1f}-\x{fb28}\x{fb2a}-\x{fb4f}\x{10800}-\x{1091e}\x{10920}-\x{10a00}\x{10a04}\x{10a07}-\x{10a0b}\x{10a10}-\x{10a37}\x{10a3b}-\x{10a3e}\x{10a40}-\x{10ae4}\x{10ae7}-\x{10b38}\x{10b40}-\x{10e5f}\x{10e7f}-\x{10fff}\x{1e800}-\x{1e8cf}\x{1e8d7}-\x{1edff}\x{1ef00}-\x{1efff}\x{608}\x{60b}\x{60d}\x{61b}-\x{64a}\x{66d}-\x{66f}\x{671}-\x{6d5}\x{6e5}\x{6e6}\x{6ee}\x{6ef}\x{6fa}-\x{710}\x{712}-\x{72f}\x{74b}-\x{7a5}\x{7b1}-\x{7bf}\x{8a0}-\x{8e2}\x{fb50}-\x{fd3d}\x{fd40}-\x{fdcf}\x{fdf0}-\x{fdfc}\x{fdfe}\x{fdff}\x{fe70}-\x{fefe}\x{1ee00}-\x{1eeef}\x{1eef2}-\x{1eeff}]))/u';
175 // @codeCoverageIgnoreEnd
176 // @codingStandardsIgnoreEnd
177
178 /**
179 * Get a cached or new language object for a given language code
180 * @param string $code
181 * @return Language
182 */
183 static function factory( $code ) {
184 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
185
186 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
187 $code = $wgDummyLanguageCodes[$code];
188 }
189
190 // get the language object to process
191 $langObj = isset( self::$mLangObjCache[$code] )
192 ? self::$mLangObjCache[$code]
193 : self::newFromCode( $code );
194
195 // merge the language object in to get it up front in the cache
196 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
197 // get rid of the oldest ones in case we have an overflow
198 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
199
200 return $langObj;
201 }
202
203 /**
204 * Create a language object for a given language code
205 * @param string $code
206 * @throws MWException
207 * @return Language
208 */
209 protected static function newFromCode( $code ) {
210 if ( !Language::isValidCode( $code ) ) {
211 throw new MWException( "Invalid language code \"$code\"" );
212 }
213
214 if ( !Language::isValidBuiltInCode( $code ) ) {
215 // It's not possible to customise this code with class files, so
216 // just return a Language object. This is to support uselang= hacks.
217 $lang = new Language;
218 $lang->setCode( $code );
219 return $lang;
220 }
221
222 // Check if there is a language class for the code
223 $class = self::classFromCode( $code );
224 if ( class_exists( $class ) ) {
225 $lang = new $class;
226 return $lang;
227 }
228
229 // Keep trying the fallback list until we find an existing class
230 $fallbacks = Language::getFallbacksFor( $code );
231 foreach ( $fallbacks as $fallbackCode ) {
232 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
233 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
234 }
235
236 $class = self::classFromCode( $fallbackCode );
237 if ( class_exists( $class ) ) {
238 $lang = new $class;
239 $lang->setCode( $code );
240 return $lang;
241 }
242 }
243
244 throw new MWException( "Invalid fallback sequence for language '$code'" );
245 }
246
247 /**
248 * Checks whether any localisation is available for that language tag
249 * in MediaWiki (MessagesXx.php exists).
250 *
251 * @param string $code Language tag (in lower case)
252 * @return bool Whether language is supported
253 * @since 1.21
254 */
255 public static function isSupportedLanguage( $code ) {
256 if ( !self::isValidBuiltInCode( $code ) ) {
257 return false;
258 }
259
260 if ( $code === 'qqq' ) {
261 return false;
262 }
263
264 return is_readable( self::getMessagesFileName( $code ) ) ||
265 is_readable( self::getJsonMessagesFileName( $code ) );
266 }
267
268 /**
269 * Returns true if a language code string is a well-formed language tag
270 * according to RFC 5646.
271 * This function only checks well-formedness; it doesn't check that
272 * language, script or variant codes actually exist in the repositories.
273 *
274 * Based on regexes by Mark Davis of the Unicode Consortium:
275 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
276 *
277 * @param string $code
278 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
279 *
280 * @return bool
281 * @since 1.21
282 */
283 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
284 $alpha = '[a-z]';
285 $digit = '[0-9]';
286 $alphanum = '[a-z0-9]';
287 $x = 'x'; # private use singleton
288 $singleton = '[a-wy-z]'; # other singleton
289 $s = $lenient ? '[-_]' : '-';
290
291 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
292 $script = "$alpha{4}"; # ISO 15924
293 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
294 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
295 $extension = "$singleton(?:$s$alphanum{2,8})+";
296 $privateUse = "$x(?:$s$alphanum{1,8})+";
297
298 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
299 # Since these are limited, this is safe even later changes to the registry --
300 # the only oddity is that it might change the type of the tag, and thus
301 # the results from the capturing groups.
302 # https://www.iana.org/assignments/language-subtag-registry
303
304 $grandfathered = "en{$s}GB{$s}oed"
305 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
306 . "|no{$s}(?:bok|nyn)"
307 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
308 . "|zh{$s}min{$s}nan";
309
310 $variantList = "$variant(?:$s$variant)*";
311 $extensionList = "$extension(?:$s$extension)*";
312
313 $langtag = "(?:($language)"
314 . "(?:$s$script)?"
315 . "(?:$s$region)?"
316 . "(?:$s$variantList)?"
317 . "(?:$s$extensionList)?"
318 . "(?:$s$privateUse)?)";
319
320 # The final breakdown, with capturing groups for each of these components
321 # The variants, extensions, grandfathered, and private-use may have interior '-'
322
323 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
324
325 return (bool)preg_match( "/$root/", strtolower( $code ) );
326 }
327
328 /**
329 * Returns true if a language code string is of a valid form, whether or
330 * not it exists. This includes codes which are used solely for
331 * customisation via the MediaWiki namespace.
332 *
333 * @param string $code
334 *
335 * @return bool
336 */
337 public static function isValidCode( $code ) {
338 static $cache = [];
339 if ( !isset( $cache[$code] ) ) {
340 // People think language codes are html safe, so enforce it.
341 // Ideally we should only allow a-zA-Z0-9-
342 // but, .+ and other chars are often used for {{int:}} hacks
343 // see bugs T39564, T39587, T38938
344 $cache[$code] =
345 // Protect against path traversal
346 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
347 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
348 }
349 return $cache[$code];
350 }
351
352 /**
353 * Returns true if a language code is of a valid form for the purposes of
354 * internal customisation of MediaWiki, via Messages*.php or *.json.
355 *
356 * @param string $code
357 *
358 * @throws MWException
359 * @since 1.18
360 * @return bool
361 */
362 public static function isValidBuiltInCode( $code ) {
363
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
367 } else {
368 $addmsg = '';
369 }
370 $type = gettype( $code );
371 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
372 }
373
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
375 }
376
377 /**
378 * Returns true if a language code is an IETF tag known to MediaWiki.
379 *
380 * @param string $tag
381 *
382 * @since 1.21
383 * @return bool
384 */
385 public static function isKnownLanguageTag( $tag ) {
386 // Quick escape for invalid input to avoid exceptions down the line
387 // when code tries to process tags which are not valid at all.
388 if ( !self::isValidBuiltInCode( $tag ) ) {
389 return false;
390 }
391
392 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
393 || self::fetchLanguageName( $tag, $tag ) !== ''
394 ) {
395 return true;
396 }
397
398 return false;
399 }
400
401 /**
402 * Get the LocalisationCache instance
403 *
404 * @return LocalisationCache
405 */
406 public static function getLocalisationCache() {
407 if ( is_null( self::$dataCache ) ) {
408 global $wgLocalisationCacheConf;
409 $class = $wgLocalisationCacheConf['class'];
410 self::$dataCache = new $class( $wgLocalisationCacheConf );
411 }
412 return self::$dataCache;
413 }
414
415 function __construct() {
416 $this->mConverter = new FakeConverter( $this );
417 // Set the code to the name of the descendant
418 if ( get_class( $this ) == 'Language' ) {
419 $this->mCode = 'en';
420 } else {
421 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
422 }
423 self::getLocalisationCache();
424 }
425
426 /**
427 * Reduce memory usage
428 */
429 function __destruct() {
430 foreach ( $this as $name => $value ) {
431 unset( $this->$name );
432 }
433 }
434
435 /**
436 * Hook which will be called if this is the content language.
437 * Descendants can use this to register hook functions or modify globals
438 */
439 function initContLang() {
440 }
441
442 /**
443 * @return array
444 * @since 1.19
445 */
446 public function getFallbackLanguages() {
447 return self::getFallbacksFor( $this->mCode );
448 }
449
450 /**
451 * Exports $wgBookstoreListEn
452 * @return array
453 */
454 public function getBookstoreList() {
455 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
456 }
457
458 /**
459 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
460 * available in localised form, it will be included in English.
461 *
462 * @return array
463 */
464 public function getNamespaces() {
465 if ( is_null( $this->namespaceNames ) ) {
466 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
467
468 $validNamespaces = MWNamespace::getCanonicalNamespaces();
469
470 $this->namespaceNames = $wgExtraNamespaces +
471 self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
472 $this->namespaceNames += $validNamespaces;
473
474 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
475 if ( $wgMetaNamespaceTalk ) {
476 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
477 } else {
478 $talk = $this->namespaceNames[NS_PROJECT_TALK];
479 $this->namespaceNames[NS_PROJECT_TALK] =
480 $this->fixVariableInNamespace( $talk );
481 }
482
483 # Sometimes a language will be localised but not actually exist on this wiki.
484 foreach ( $this->namespaceNames as $key => $text ) {
485 if ( !isset( $validNamespaces[$key] ) ) {
486 unset( $this->namespaceNames[$key] );
487 }
488 }
489
490 # The above mixing may leave namespaces out of canonical order.
491 # Re-order by namespace ID number...
492 ksort( $this->namespaceNames );
493
494 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
495 }
496
497 return $this->namespaceNames;
498 }
499
500 /**
501 * Arbitrarily set all of the namespace names at once. Mainly used for testing
502 * @param array $namespaces Array of namespaces (id => name)
503 */
504 public function setNamespaces( array $namespaces ) {
505 $this->namespaceNames = $namespaces;
506 $this->mNamespaceIds = null;
507 }
508
509 /**
510 * Resets all of the namespace caches. Mainly used for testing
511 */
512 public function resetNamespaces() {
513 $this->namespaceNames = null;
514 $this->mNamespaceIds = null;
515 $this->namespaceAliases = null;
516 }
517
518 /**
519 * A convenience function that returns getNamespaces() with spaces instead of underscores
520 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
521 *
522 * @return array
523 */
524 public function getFormattedNamespaces() {
525 $ns = $this->getNamespaces();
526 foreach ( $ns as $k => $v ) {
527 $ns[$k] = strtr( $v, '_', ' ' );
528 }
529 return $ns;
530 }
531
532 /**
533 * Get a namespace value by key
534 *
535 * <code>
536 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
537 * echo $mw_ns; // prints 'MediaWiki'
538 * </code>
539 *
540 * @param int $index The array key of the namespace to return
541 * @return string|bool String if the namespace value exists, otherwise false
542 */
543 public function getNsText( $index ) {
544 $ns = $this->getNamespaces();
545 return isset( $ns[$index] ) ? $ns[$index] : false;
546 }
547
548 /**
549 * A convenience function that returns the same thing as
550 * getNsText() except with '_' changed to ' ', useful for
551 * producing output.
552 *
553 * <code>
554 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
555 * echo $mw_ns; // prints 'MediaWiki talk'
556 * </code>
557 *
558 * @param int $index The array key of the namespace to return
559 * @return string Namespace name without underscores (empty string if namespace does not exist)
560 */
561 public function getFormattedNsText( $index ) {
562 $ns = $this->getNsText( $index );
563 return strtr( $ns, '_', ' ' );
564 }
565
566 /**
567 * Returns gender-dependent namespace alias if available.
568 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
569 * @param int $index Namespace index
570 * @param string $gender Gender key (male, female... )
571 * @return string
572 * @since 1.18
573 */
574 public function getGenderNsText( $index, $gender ) {
575 global $wgExtraGenderNamespaces;
576
577 $ns = $wgExtraGenderNamespaces +
578 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
579
580 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
581 }
582
583 /**
584 * Whether this language uses gender-dependent namespace aliases.
585 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
586 * @return bool
587 * @since 1.18
588 */
589 public function needsGenderDistinction() {
590 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
591 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
592 // $wgExtraGenderNamespaces overrides everything
593 return true;
594 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
595 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
596 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
597 return false;
598 } else {
599 // Check what is in i18n files
600 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
601 return count( $aliases ) > 0;
602 }
603 }
604
605 /**
606 * Get a namespace key by value, case insensitive.
607 * Only matches namespace names for the current language, not the
608 * canonical ones defined in Namespace.php.
609 *
610 * @param string $text
611 * @return int|bool An integer if $text is a valid value otherwise false
612 */
613 function getLocalNsIndex( $text ) {
614 $lctext = $this->lc( $text );
615 $ids = $this->getNamespaceIds();
616 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
617 }
618
619 /**
620 * @return array
621 */
622 public function getNamespaceAliases() {
623 if ( is_null( $this->namespaceAliases ) ) {
624 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
625 if ( !$aliases ) {
626 $aliases = [];
627 } else {
628 foreach ( $aliases as $name => $index ) {
629 if ( $index === NS_PROJECT_TALK ) {
630 unset( $aliases[$name] );
631 $name = $this->fixVariableInNamespace( $name );
632 $aliases[$name] = $index;
633 }
634 }
635 }
636
637 global $wgExtraGenderNamespaces;
638 $genders = $wgExtraGenderNamespaces +
639 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
640 foreach ( $genders as $index => $forms ) {
641 foreach ( $forms as $alias ) {
642 $aliases[$alias] = $index;
643 }
644 }
645
646 # Also add converted namespace names as aliases, to avoid confusion.
647 $convertedNames = [];
648 foreach ( $this->getVariants() as $variant ) {
649 if ( $variant === $this->mCode ) {
650 continue;
651 }
652 foreach ( $this->getNamespaces() as $ns => $_ ) {
653 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
654 }
655 }
656
657 $this->namespaceAliases = $aliases + $convertedNames;
658 }
659
660 return $this->namespaceAliases;
661 }
662
663 /**
664 * @return array
665 */
666 public function getNamespaceIds() {
667 if ( is_null( $this->mNamespaceIds ) ) {
668 global $wgNamespaceAliases;
669 # Put namespace names and aliases into a hashtable.
670 # If this is too slow, then we should arrange it so that it is done
671 # before caching. The catch is that at pre-cache time, the above
672 # class-specific fixup hasn't been done.
673 $this->mNamespaceIds = [];
674 foreach ( $this->getNamespaces() as $index => $name ) {
675 $this->mNamespaceIds[$this->lc( $name )] = $index;
676 }
677 foreach ( $this->getNamespaceAliases() as $name => $index ) {
678 $this->mNamespaceIds[$this->lc( $name )] = $index;
679 }
680 if ( $wgNamespaceAliases ) {
681 foreach ( $wgNamespaceAliases as $name => $index ) {
682 $this->mNamespaceIds[$this->lc( $name )] = $index;
683 }
684 }
685 }
686 return $this->mNamespaceIds;
687 }
688
689 /**
690 * Get a namespace key by value, case insensitive. Canonical namespace
691 * names override custom ones defined for the current language.
692 *
693 * @param string $text
694 * @return int|bool An integer if $text is a valid value otherwise false
695 */
696 public function getNsIndex( $text ) {
697 $lctext = $this->lc( $text );
698 $ns = MWNamespace::getCanonicalIndex( $lctext );
699 if ( $ns !== null ) {
700 return $ns;
701 }
702 $ids = $this->getNamespaceIds();
703 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
704 }
705
706 /**
707 * short names for language variants used for language conversion links.
708 *
709 * @param string $code
710 * @param bool $usemsg Use the "variantname-xyz" message if it exists
711 * @return string
712 */
713 public function getVariantname( $code, $usemsg = true ) {
714 $msg = "variantname-$code";
715 if ( $usemsg && wfMessage( $msg )->exists() ) {
716 return $this->getMessageFromDB( $msg );
717 }
718 $name = self::fetchLanguageName( $code );
719 if ( $name ) {
720 return $name; # if it's defined as a language name, show that
721 } else {
722 # otherwise, output the language code
723 return $code;
724 }
725 }
726
727 /**
728 * @return array
729 */
730 public function getDatePreferences() {
731 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
732 }
733
734 /**
735 * @return array
736 */
737 function getDateFormats() {
738 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
739 }
740
741 /**
742 * @return array|string
743 */
744 public function getDefaultDateFormat() {
745 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
746 if ( $df === 'dmy or mdy' ) {
747 global $wgAmericanDates;
748 return $wgAmericanDates ? 'mdy' : 'dmy';
749 } else {
750 return $df;
751 }
752 }
753
754 /**
755 * @return array
756 */
757 public function getDatePreferenceMigrationMap() {
758 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
759 }
760
761 /**
762 * @param string $image
763 * @return array|null
764 */
765 function getImageFile( $image ) {
766 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
767 }
768
769 /**
770 * @return array
771 * @since 1.24
772 */
773 public function getImageFiles() {
774 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
775 }
776
777 /**
778 * @return array
779 */
780 public function getExtraUserToggles() {
781 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
782 }
783
784 /**
785 * @param string $tog
786 * @return string
787 */
788 function getUserToggle( $tog ) {
789 return $this->getMessageFromDB( "tog-$tog" );
790 }
791
792 /**
793 * Get an array of language names, indexed by code.
794 * @param null|string $inLanguage Code of language in which to return the names
795 * Use null for autonyms (native names)
796 * @param string $include One of:
797 * 'all' all available languages
798 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
799 * 'mwfile' only if the language is in 'mw' *and* has a message file
800 * @return array Language code => language name
801 * @since 1.20
802 */
803 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
804 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
805 $cacheKey .= ":$include";
806 if ( self::$languageNameCache === null ) {
807 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
808 }
809
810 $ret = self::$languageNameCache->get( $cacheKey );
811 if ( !$ret ) {
812 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
813 self::$languageNameCache->set( $cacheKey, $ret );
814 }
815 return $ret;
816 }
817
818 /**
819 * Uncached helper for fetchLanguageNames
820 * @param null|string $inLanguage Code of language in which to return the names
821 * Use null for autonyms (native names)
822 * @param string $include One of:
823 * 'all' all available languages
824 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
825 * 'mwfile' only if the language is in 'mw' *and* has a message file
826 * @return array Language code => language name
827 */
828 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
829 global $wgExtraLanguageNames;
830
831 // If passed an invalid language code to use, fallback to en
832 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
833 $inLanguage = 'en';
834 }
835
836 $names = [];
837
838 if ( $inLanguage ) {
839 # TODO: also include when $inLanguage is null, when this code is more efficient
840 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
841 }
842
843 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
844 foreach ( $mwNames as $mwCode => $mwName ) {
845 # - Prefer own MediaWiki native name when not using the hook
846 # - For other names just add if not added through the hook
847 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
848 $names[$mwCode] = $mwName;
849 }
850 }
851
852 if ( $include === 'all' ) {
853 ksort( $names );
854 return $names;
855 }
856
857 $returnMw = [];
858 $coreCodes = array_keys( $mwNames );
859 foreach ( $coreCodes as $coreCode ) {
860 $returnMw[$coreCode] = $names[$coreCode];
861 }
862
863 if ( $include === 'mwfile' ) {
864 $namesMwFile = [];
865 # We do this using a foreach over the codes instead of a directory
866 # loop so that messages files in extensions will work correctly.
867 foreach ( $returnMw as $code => $value ) {
868 if ( is_readable( self::getMessagesFileName( $code ) )
869 || is_readable( self::getJsonMessagesFileName( $code ) )
870 ) {
871 $namesMwFile[$code] = $names[$code];
872 }
873 }
874
875 ksort( $namesMwFile );
876 return $namesMwFile;
877 }
878
879 ksort( $returnMw );
880 # 'mw' option; default if it's not one of the other two options (all/mwfile)
881 return $returnMw;
882 }
883
884 /**
885 * @param string $code The code of the language for which to get the name
886 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
887 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
888 * @return string Language name or empty
889 * @since 1.20
890 */
891 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
892 $code = strtolower( $code );
893 $array = self::fetchLanguageNames( $inLanguage, $include );
894 return !array_key_exists( $code, $array ) ? '' : $array[$code];
895 }
896
897 /**
898 * Get a message from the MediaWiki namespace.
899 *
900 * @param string $msg Message name
901 * @return string
902 */
903 public function getMessageFromDB( $msg ) {
904 return $this->msg( $msg )->text();
905 }
906
907 /**
908 * Get message object in this language. Only for use inside this class.
909 *
910 * @param string $msg Message name
911 * @return Message
912 */
913 protected function msg( $msg ) {
914 return wfMessage( $msg )->inLanguage( $this );
915 }
916
917 /**
918 * @param string $key
919 * @return string
920 */
921 public function getMonthName( $key ) {
922 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
923 }
924
925 /**
926 * @return array
927 */
928 public function getMonthNamesArray() {
929 $monthNames = [ '' ];
930 for ( $i = 1; $i < 13; $i++ ) {
931 $monthNames[] = $this->getMonthName( $i );
932 }
933 return $monthNames;
934 }
935
936 /**
937 * @param string $key
938 * @return string
939 */
940 public function getMonthNameGen( $key ) {
941 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
942 }
943
944 /**
945 * @param string $key
946 * @return string
947 */
948 public function getMonthAbbreviation( $key ) {
949 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
950 }
951
952 /**
953 * @return array
954 */
955 public function getMonthAbbreviationsArray() {
956 $monthNames = [ '' ];
957 for ( $i = 1; $i < 13; $i++ ) {
958 $monthNames[] = $this->getMonthAbbreviation( $i );
959 }
960 return $monthNames;
961 }
962
963 /**
964 * @param string $key
965 * @return string
966 */
967 public function getWeekdayName( $key ) {
968 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
969 }
970
971 /**
972 * @param string $key
973 * @return string
974 */
975 function getWeekdayAbbreviation( $key ) {
976 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
977 }
978
979 /**
980 * @param string $key
981 * @return string
982 */
983 function getIranianCalendarMonthName( $key ) {
984 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
985 }
986
987 /**
988 * @param string $key
989 * @return string
990 */
991 function getHebrewCalendarMonthName( $key ) {
992 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
993 }
994
995 /**
996 * @param string $key
997 * @return string
998 */
999 function getHebrewCalendarMonthNameGen( $key ) {
1000 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1001 }
1002
1003 /**
1004 * @param string $key
1005 * @return string
1006 */
1007 function getHijriCalendarMonthName( $key ) {
1008 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1009 }
1010
1011 /**
1012 * Pass through result from $dateTimeObj->format()
1013 * @param DateTime|bool|null &$dateTimeObj
1014 * @param string $ts
1015 * @param DateTimeZone|bool|null $zone
1016 * @param string $code
1017 * @return string
1018 */
1019 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1020 if ( !$dateTimeObj ) {
1021 $dateTimeObj = DateTime::createFromFormat(
1022 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1023 );
1024 }
1025 return $dateTimeObj->format( $code );
1026 }
1027
1028 /**
1029 * This is a workalike of PHP's date() function, but with better
1030 * internationalisation, a reduced set of format characters, and a better
1031 * escaping format.
1032 *
1033 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1034 * the PHP manual for definitions. There are a number of extensions, which
1035 * start with "x":
1036 *
1037 * xn Do not translate digits of the next numeric format character
1038 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1039 * xr Use roman numerals for the next numeric format character
1040 * xh Use hebrew numerals for the next numeric format character
1041 * xx Literal x
1042 * xg Genitive month name
1043 *
1044 * xij j (day number) in Iranian calendar
1045 * xiF F (month name) in Iranian calendar
1046 * xin n (month number) in Iranian calendar
1047 * xiy y (two digit year) in Iranian calendar
1048 * xiY Y (full year) in Iranian calendar
1049 * xit t (days in month) in Iranian calendar
1050 * xiz z (day of the year) in Iranian calendar
1051 *
1052 * xjj j (day number) in Hebrew calendar
1053 * xjF F (month name) in Hebrew calendar
1054 * xjt t (days in month) in Hebrew calendar
1055 * xjx xg (genitive month name) in Hebrew calendar
1056 * xjn n (month number) in Hebrew calendar
1057 * xjY Y (full year) in Hebrew calendar
1058 *
1059 * xmj j (day number) in Hijri calendar
1060 * xmF F (month name) in Hijri calendar
1061 * xmn n (month number) in Hijri calendar
1062 * xmY Y (full year) in Hijri calendar
1063 *
1064 * xkY Y (full year) in Thai solar calendar. Months and days are
1065 * identical to the Gregorian calendar
1066 * xoY Y (full year) in Minguo calendar or Juche year.
1067 * Months and days are identical to the
1068 * Gregorian calendar
1069 * xtY Y (full year) in Japanese nengo. Months and days are
1070 * identical to the Gregorian calendar
1071 *
1072 * Characters enclosed in double quotes will be considered literal (with
1073 * the quotes themselves removed). Unmatched quotes will be considered
1074 * literal quotes. Example:
1075 *
1076 * "The month is" F => The month is January
1077 * i's" => 20'11"
1078 *
1079 * Backslash escaping is also supported.
1080 *
1081 * Input timestamp is assumed to be pre-normalized to the desired local
1082 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1083 * $ts is UTC if $zone is not given.
1084 *
1085 * @param string $format
1086 * @param string $ts 14-character timestamp
1087 * YYYYMMDDHHMMSS
1088 * 01234567890123
1089 * @param DateTimeZone $zone Timezone of $ts
1090 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1091 * Only makes sense if $ts is the current time.
1092 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1093 *
1094 * @throws MWException
1095 * @return string
1096 */
1097 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1098 $s = '';
1099 $raw = false;
1100 $roman = false;
1101 $hebrewNum = false;
1102 $dateTimeObj = false;
1103 $rawToggle = false;
1104 $iranian = false;
1105 $hebrew = false;
1106 $hijri = false;
1107 $thai = false;
1108 $minguo = false;
1109 $tenno = false;
1110
1111 $usedSecond = false;
1112 $usedMinute = false;
1113 $usedHour = false;
1114 $usedAMPM = false;
1115 $usedDay = false;
1116 $usedWeek = false;
1117 $usedMonth = false;
1118 $usedYear = false;
1119 $usedISOYear = false;
1120 $usedIsLeapYear = false;
1121
1122 $usedHebrewMonth = false;
1123 $usedIranianMonth = false;
1124 $usedHijriMonth = false;
1125 $usedHebrewYear = false;
1126 $usedIranianYear = false;
1127 $usedHijriYear = false;
1128 $usedTennoYear = false;
1129
1130 if ( strlen( $ts ) !== 14 ) {
1131 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1132 }
1133
1134 if ( !ctype_digit( $ts ) ) {
1135 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1136 }
1137
1138 $formatLength = strlen( $format );
1139 for ( $p = 0; $p < $formatLength; $p++ ) {
1140 $num = false;
1141 $code = $format[$p];
1142 if ( $code == 'x' && $p < $formatLength - 1 ) {
1143 $code .= $format[++$p];
1144 }
1145
1146 if ( ( $code === 'xi'
1147 || $code === 'xj'
1148 || $code === 'xk'
1149 || $code === 'xm'
1150 || $code === 'xo'
1151 || $code === 'xt' )
1152 && $p < $formatLength - 1 ) {
1153 $code .= $format[++$p];
1154 }
1155
1156 switch ( $code ) {
1157 case 'xx':
1158 $s .= 'x';
1159 break;
1160 case 'xn':
1161 $raw = true;
1162 break;
1163 case 'xN':
1164 $rawToggle = !$rawToggle;
1165 break;
1166 case 'xr':
1167 $roman = true;
1168 break;
1169 case 'xh':
1170 $hebrewNum = true;
1171 break;
1172 case 'xg':
1173 $usedMonth = true;
1174 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1175 break;
1176 case 'xjx':
1177 $usedHebrewMonth = true;
1178 if ( !$hebrew ) {
1179 $hebrew = self::tsToHebrew( $ts );
1180 }
1181 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1182 break;
1183 case 'd':
1184 $usedDay = true;
1185 $num = substr( $ts, 6, 2 );
1186 break;
1187 case 'D':
1188 $usedDay = true;
1189 $s .= $this->getWeekdayAbbreviation(
1190 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1191 );
1192 break;
1193 case 'j':
1194 $usedDay = true;
1195 $num = intval( substr( $ts, 6, 2 ) );
1196 break;
1197 case 'xij':
1198 $usedDay = true;
1199 if ( !$iranian ) {
1200 $iranian = self::tsToIranian( $ts );
1201 }
1202 $num = $iranian[2];
1203 break;
1204 case 'xmj':
1205 $usedDay = true;
1206 if ( !$hijri ) {
1207 $hijri = self::tsToHijri( $ts );
1208 }
1209 $num = $hijri[2];
1210 break;
1211 case 'xjj':
1212 $usedDay = true;
1213 if ( !$hebrew ) {
1214 $hebrew = self::tsToHebrew( $ts );
1215 }
1216 $num = $hebrew[2];
1217 break;
1218 case 'l':
1219 $usedDay = true;
1220 $s .= $this->getWeekdayName(
1221 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1222 );
1223 break;
1224 case 'F':
1225 $usedMonth = true;
1226 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1227 break;
1228 case 'xiF':
1229 $usedIranianMonth = true;
1230 if ( !$iranian ) {
1231 $iranian = self::tsToIranian( $ts );
1232 }
1233 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1234 break;
1235 case 'xmF':
1236 $usedHijriMonth = true;
1237 if ( !$hijri ) {
1238 $hijri = self::tsToHijri( $ts );
1239 }
1240 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1241 break;
1242 case 'xjF':
1243 $usedHebrewMonth = true;
1244 if ( !$hebrew ) {
1245 $hebrew = self::tsToHebrew( $ts );
1246 }
1247 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1248 break;
1249 case 'm':
1250 $usedMonth = true;
1251 $num = substr( $ts, 4, 2 );
1252 break;
1253 case 'M':
1254 $usedMonth = true;
1255 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1256 break;
1257 case 'n':
1258 $usedMonth = true;
1259 $num = intval( substr( $ts, 4, 2 ) );
1260 break;
1261 case 'xin':
1262 $usedIranianMonth = true;
1263 if ( !$iranian ) {
1264 $iranian = self::tsToIranian( $ts );
1265 }
1266 $num = $iranian[1];
1267 break;
1268 case 'xmn':
1269 $usedHijriMonth = true;
1270 if ( !$hijri ) {
1271 $hijri = self::tsToHijri( $ts );
1272 }
1273 $num = $hijri[1];
1274 break;
1275 case 'xjn':
1276 $usedHebrewMonth = true;
1277 if ( !$hebrew ) {
1278 $hebrew = self::tsToHebrew( $ts );
1279 }
1280 $num = $hebrew[1];
1281 break;
1282 case 'xjt':
1283 $usedHebrewMonth = true;
1284 if ( !$hebrew ) {
1285 $hebrew = self::tsToHebrew( $ts );
1286 }
1287 $num = $hebrew[3];
1288 break;
1289 case 'Y':
1290 $usedYear = true;
1291 $num = substr( $ts, 0, 4 );
1292 break;
1293 case 'xiY':
1294 $usedIranianYear = true;
1295 if ( !$iranian ) {
1296 $iranian = self::tsToIranian( $ts );
1297 }
1298 $num = $iranian[0];
1299 break;
1300 case 'xmY':
1301 $usedHijriYear = true;
1302 if ( !$hijri ) {
1303 $hijri = self::tsToHijri( $ts );
1304 }
1305 $num = $hijri[0];
1306 break;
1307 case 'xjY':
1308 $usedHebrewYear = true;
1309 if ( !$hebrew ) {
1310 $hebrew = self::tsToHebrew( $ts );
1311 }
1312 $num = $hebrew[0];
1313 break;
1314 case 'xkY':
1315 $usedYear = true;
1316 if ( !$thai ) {
1317 $thai = self::tsToYear( $ts, 'thai' );
1318 }
1319 $num = $thai[0];
1320 break;
1321 case 'xoY':
1322 $usedYear = true;
1323 if ( !$minguo ) {
1324 $minguo = self::tsToYear( $ts, 'minguo' );
1325 }
1326 $num = $minguo[0];
1327 break;
1328 case 'xtY':
1329 $usedTennoYear = true;
1330 if ( !$tenno ) {
1331 $tenno = self::tsToYear( $ts, 'tenno' );
1332 }
1333 $num = $tenno[0];
1334 break;
1335 case 'y':
1336 $usedYear = true;
1337 $num = substr( $ts, 2, 2 );
1338 break;
1339 case 'xiy':
1340 $usedIranianYear = true;
1341 if ( !$iranian ) {
1342 $iranian = self::tsToIranian( $ts );
1343 }
1344 $num = substr( $iranian[0], -2 );
1345 break;
1346 case 'xit':
1347 $usedIranianYear = true;
1348 if ( !$iranian ) {
1349 $iranian = self::tsToIranian( $ts );
1350 }
1351 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1352 break;
1353 case 'xiz':
1354 $usedIranianYear = true;
1355 if ( !$iranian ) {
1356 $iranian = self::tsToIranian( $ts );
1357 }
1358 $num = $iranian[3];
1359 break;
1360 case 'a':
1361 $usedAMPM = true;
1362 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1363 break;
1364 case 'A':
1365 $usedAMPM = true;
1366 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1367 break;
1368 case 'g':
1369 $usedHour = true;
1370 $h = substr( $ts, 8, 2 );
1371 $num = $h % 12 ? $h % 12 : 12;
1372 break;
1373 case 'G':
1374 $usedHour = true;
1375 $num = intval( substr( $ts, 8, 2 ) );
1376 break;
1377 case 'h':
1378 $usedHour = true;
1379 $h = substr( $ts, 8, 2 );
1380 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1381 break;
1382 case 'H':
1383 $usedHour = true;
1384 $num = substr( $ts, 8, 2 );
1385 break;
1386 case 'i':
1387 $usedMinute = true;
1388 $num = substr( $ts, 10, 2 );
1389 break;
1390 case 's':
1391 $usedSecond = true;
1392 $num = substr( $ts, 12, 2 );
1393 break;
1394 case 'c':
1395 case 'r':
1396 $usedSecond = true;
1397 // fall through
1398 case 'e':
1399 case 'O':
1400 case 'P':
1401 case 'T':
1402 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1403 break;
1404 case 'w':
1405 case 'N':
1406 case 'z':
1407 $usedDay = true;
1408 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1409 break;
1410 case 'W':
1411 $usedWeek = true;
1412 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1413 break;
1414 case 't':
1415 $usedMonth = true;
1416 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1417 break;
1418 case 'L':
1419 $usedIsLeapYear = true;
1420 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1421 break;
1422 case 'o':
1423 $usedISOYear = true;
1424 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1425 break;
1426 case 'U':
1427 $usedSecond = true;
1428 // fall through
1429 case 'I':
1430 case 'Z':
1431 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1432 break;
1433 case '\\':
1434 # Backslash escaping
1435 if ( $p < $formatLength - 1 ) {
1436 $s .= $format[++$p];
1437 } else {
1438 $s .= '\\';
1439 }
1440 break;
1441 case '"':
1442 # Quoted literal
1443 if ( $p < $formatLength - 1 ) {
1444 $endQuote = strpos( $format, '"', $p + 1 );
1445 if ( $endQuote === false ) {
1446 # No terminating quote, assume literal "
1447 $s .= '"';
1448 } else {
1449 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1450 $p = $endQuote;
1451 }
1452 } else {
1453 # Quote at end of string, assume literal "
1454 $s .= '"';
1455 }
1456 break;
1457 default:
1458 $s .= $format[$p];
1459 }
1460 if ( $num !== false ) {
1461 if ( $rawToggle || $raw ) {
1462 $s .= $num;
1463 $raw = false;
1464 } elseif ( $roman ) {
1465 $s .= Language::romanNumeral( $num );
1466 $roman = false;
1467 } elseif ( $hebrewNum ) {
1468 $s .= self::hebrewNumeral( $num );
1469 $hebrewNum = false;
1470 } else {
1471 $s .= $this->formatNum( $num, true );
1472 }
1473 }
1474 }
1475
1476 if ( $ttl === 'unused' ) {
1477 // No need to calculate the TTL, the caller wont use it anyway.
1478 } elseif ( $usedSecond ) {
1479 $ttl = 1;
1480 } elseif ( $usedMinute ) {
1481 $ttl = 60 - substr( $ts, 12, 2 );
1482 } elseif ( $usedHour ) {
1483 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1484 } elseif ( $usedAMPM ) {
1485 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1486 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1487 } elseif (
1488 $usedDay ||
1489 $usedHebrewMonth ||
1490 $usedIranianMonth ||
1491 $usedHijriMonth ||
1492 $usedHebrewYear ||
1493 $usedIranianYear ||
1494 $usedHijriYear ||
1495 $usedTennoYear
1496 ) {
1497 // @todo Someone who understands the non-Gregorian calendars
1498 // should write proper logic for them so that they don't need purged every day.
1499 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1500 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1501 } else {
1502 $possibleTtls = [];
1503 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1504 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1505 if ( $usedWeek ) {
1506 $possibleTtls[] =
1507 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1508 $timeRemainingInDay;
1509 } elseif ( $usedISOYear ) {
1510 // December 28th falls on the last ISO week of the year, every year.
1511 // The last ISO week of a year can be 52 or 53.
1512 $lastWeekOfISOYear = DateTime::createFromFormat(
1513 'Ymd',
1514 substr( $ts, 0, 4 ) . '1228',
1515 $zone ?: new DateTimeZone( 'UTC' )
1516 )->format( 'W' );
1517 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1518 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1519 $timeRemainingInWeek =
1520 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1521 + $timeRemainingInDay;
1522 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1523 }
1524
1525 if ( $usedMonth ) {
1526 $possibleTtls[] =
1527 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1528 substr( $ts, 6, 2 ) ) * 86400
1529 + $timeRemainingInDay;
1530 } elseif ( $usedYear ) {
1531 $possibleTtls[] =
1532 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1533 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1534 + $timeRemainingInDay;
1535 } elseif ( $usedIsLeapYear ) {
1536 $year = substr( $ts, 0, 4 );
1537 $timeRemainingInYear =
1538 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1539 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1540 + $timeRemainingInDay;
1541 $mod = $year % 4;
1542 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1543 // this isn't a leap year. see when the next one starts
1544 $nextCandidate = $year - $mod + 4;
1545 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1546 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1547 $timeRemainingInYear;
1548 } else {
1549 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1550 $timeRemainingInYear;
1551 }
1552 } else {
1553 // this is a leap year, so the next year isn't
1554 $possibleTtls[] = $timeRemainingInYear;
1555 }
1556 }
1557
1558 if ( $possibleTtls ) {
1559 $ttl = min( $possibleTtls );
1560 }
1561 }
1562
1563 return $s;
1564 }
1565
1566 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1567 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1568
1569 /**
1570 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1571 * Gregorian dates to Iranian dates. Originally written in C, it
1572 * is released under the terms of GNU Lesser General Public
1573 * License. Conversion to PHP was performed by Niklas Laxström.
1574 *
1575 * Link: http://www.farsiweb.info/jalali/jalali.c
1576 *
1577 * @param string $ts
1578 *
1579 * @return int[]
1580 */
1581 private static function tsToIranian( $ts ) {
1582 $gy = substr( $ts, 0, 4 ) -1600;
1583 $gm = substr( $ts, 4, 2 ) -1;
1584 $gd = substr( $ts, 6, 2 ) -1;
1585
1586 # Days passed from the beginning (including leap years)
1587 $gDayNo = 365 * $gy
1588 + floor( ( $gy + 3 ) / 4 )
1589 - floor( ( $gy + 99 ) / 100 )
1590 + floor( ( $gy + 399 ) / 400 );
1591
1592 // Add days of the past months of this year
1593 for ( $i = 0; $i < $gm; $i++ ) {
1594 $gDayNo += self::$GREG_DAYS[$i];
1595 }
1596
1597 // Leap years
1598 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1599 $gDayNo++;
1600 }
1601
1602 // Days passed in current month
1603 $gDayNo += (int)$gd;
1604
1605 $jDayNo = $gDayNo - 79;
1606
1607 $jNp = floor( $jDayNo / 12053 );
1608 $jDayNo %= 12053;
1609
1610 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1611 $jDayNo %= 1461;
1612
1613 if ( $jDayNo >= 366 ) {
1614 $jy += floor( ( $jDayNo - 1 ) / 365 );
1615 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1616 }
1617
1618 $jz = $jDayNo;
1619
1620 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1621 $jDayNo -= self::$IRANIAN_DAYS[$i];
1622 }
1623
1624 $jm = $i + 1;
1625 $jd = $jDayNo + 1;
1626
1627 return [ $jy, $jm, $jd, $jz ];
1628 }
1629
1630 /**
1631 * Converting Gregorian dates to Hijri dates.
1632 *
1633 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1634 *
1635 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1636 *
1637 * @param string $ts
1638 *
1639 * @return int[]
1640 */
1641 private static function tsToHijri( $ts ) {
1642 $year = substr( $ts, 0, 4 );
1643 $month = substr( $ts, 4, 2 );
1644 $day = substr( $ts, 6, 2 );
1645
1646 $zyr = $year;
1647 $zd = $day;
1648 $zm = $month;
1649 $zy = $zyr;
1650
1651 if (
1652 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1653 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1654 ) {
1655 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1656 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1657 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1658 $zd - 32075;
1659 } else {
1660 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1661 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1662 }
1663
1664 $zl = $zjd -1948440 + 10632;
1665 $zn = (int)( ( $zl - 1 ) / 10631 );
1666 $zl = $zl - 10631 * $zn + 354;
1667 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1668 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1669 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1670 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1671 $zm = (int)( ( 24 * $zl ) / 709 );
1672 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1673 $zy = 30 * $zn + $zj - 30;
1674
1675 return [ $zy, $zm, $zd ];
1676 }
1677
1678 /**
1679 * Converting Gregorian dates to Hebrew dates.
1680 *
1681 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1682 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1683 * to translate the relevant functions into PHP and release them under
1684 * GNU GPL.
1685 *
1686 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1687 * and Adar II is 14. In a non-leap year, Adar is 6.
1688 *
1689 * @param string $ts
1690 *
1691 * @return int[]
1692 */
1693 private static function tsToHebrew( $ts ) {
1694 # Parse date
1695 $year = substr( $ts, 0, 4 );
1696 $month = substr( $ts, 4, 2 );
1697 $day = substr( $ts, 6, 2 );
1698
1699 # Calculate Hebrew year
1700 $hebrewYear = $year + 3760;
1701
1702 # Month number when September = 1, August = 12
1703 $month += 4;
1704 if ( $month > 12 ) {
1705 # Next year
1706 $month -= 12;
1707 $year++;
1708 $hebrewYear++;
1709 }
1710
1711 # Calculate day of year from 1 September
1712 $dayOfYear = $day;
1713 for ( $i = 1; $i < $month; $i++ ) {
1714 if ( $i == 6 ) {
1715 # February
1716 $dayOfYear += 28;
1717 # Check if the year is leap
1718 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1719 $dayOfYear++;
1720 }
1721 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1722 $dayOfYear += 30;
1723 } else {
1724 $dayOfYear += 31;
1725 }
1726 }
1727
1728 # Calculate the start of the Hebrew year
1729 $start = self::hebrewYearStart( $hebrewYear );
1730
1731 # Calculate next year's start
1732 if ( $dayOfYear <= $start ) {
1733 # Day is before the start of the year - it is the previous year
1734 # Next year's start
1735 $nextStart = $start;
1736 # Previous year
1737 $year--;
1738 $hebrewYear--;
1739 # Add days since previous year's 1 September
1740 $dayOfYear += 365;
1741 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1742 # Leap year
1743 $dayOfYear++;
1744 }
1745 # Start of the new (previous) year
1746 $start = self::hebrewYearStart( $hebrewYear );
1747 } else {
1748 # Next year's start
1749 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1750 }
1751
1752 # Calculate Hebrew day of year
1753 $hebrewDayOfYear = $dayOfYear - $start;
1754
1755 # Difference between year's days
1756 $diff = $nextStart - $start;
1757 # Add 12 (or 13 for leap years) days to ignore the difference between
1758 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1759 # difference is only about the year type
1760 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1761 $diff += 13;
1762 } else {
1763 $diff += 12;
1764 }
1765
1766 # Check the year pattern, and is leap year
1767 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1768 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1769 # and non-leap years
1770 $yearPattern = $diff % 30;
1771 # Check if leap year
1772 $isLeap = $diff >= 30;
1773
1774 # Calculate day in the month from number of day in the Hebrew year
1775 # Don't check Adar - if the day is not in Adar, we will stop before;
1776 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1777 $hebrewDay = $hebrewDayOfYear;
1778 $hebrewMonth = 1;
1779 $days = 0;
1780 while ( $hebrewMonth <= 12 ) {
1781 # Calculate days in this month
1782 if ( $isLeap && $hebrewMonth == 6 ) {
1783 # Adar in a leap year
1784 if ( $isLeap ) {
1785 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1786 $days = 30;
1787 if ( $hebrewDay <= $days ) {
1788 # Day in Adar I
1789 $hebrewMonth = 13;
1790 } else {
1791 # Subtract the days of Adar I
1792 $hebrewDay -= $days;
1793 # Try Adar II
1794 $days = 29;
1795 if ( $hebrewDay <= $days ) {
1796 # Day in Adar II
1797 $hebrewMonth = 14;
1798 }
1799 }
1800 }
1801 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1802 # Cheshvan in a complete year (otherwise as the rule below)
1803 $days = 30;
1804 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1805 # Kislev in an incomplete year (otherwise as the rule below)
1806 $days = 29;
1807 } else {
1808 # Odd months have 30 days, even have 29
1809 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1810 }
1811 if ( $hebrewDay <= $days ) {
1812 # In the current month
1813 break;
1814 } else {
1815 # Subtract the days of the current month
1816 $hebrewDay -= $days;
1817 # Try in the next month
1818 $hebrewMonth++;
1819 }
1820 }
1821
1822 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1823 }
1824
1825 /**
1826 * This calculates the Hebrew year start, as days since 1 September.
1827 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1828 * Used for Hebrew date.
1829 *
1830 * @param int $year
1831 *
1832 * @return string
1833 */
1834 private static function hebrewYearStart( $year ) {
1835 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1836 $b = intval( ( $year - 1 ) % 4 );
1837 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1838 if ( $m < 0 ) {
1839 $m--;
1840 }
1841 $Mar = intval( $m );
1842 if ( $m < 0 ) {
1843 $m++;
1844 }
1845 $m -= $Mar;
1846
1847 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1848 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1849 $Mar++;
1850 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1851 $Mar += 2;
1852 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1853 $Mar++;
1854 }
1855
1856 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1857 return $Mar;
1858 }
1859
1860 /**
1861 * Algorithm to convert Gregorian dates to Thai solar dates,
1862 * Minguo dates or Minguo dates.
1863 *
1864 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1865 * https://en.wikipedia.org/wiki/Minguo_calendar
1866 * https://en.wikipedia.org/wiki/Japanese_era_name
1867 *
1868 * @param string $ts 14-character timestamp
1869 * @param string $cName Calender name
1870 * @return array Converted year, month, day
1871 */
1872 private static function tsToYear( $ts, $cName ) {
1873 $gy = substr( $ts, 0, 4 );
1874 $gm = substr( $ts, 4, 2 );
1875 $gd = substr( $ts, 6, 2 );
1876
1877 if ( !strcmp( $cName, 'thai' ) ) {
1878 # Thai solar dates
1879 # Add 543 years to the Gregorian calendar
1880 # Months and days are identical
1881 $gy_offset = $gy + 543;
1882 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1883 # Minguo dates
1884 # Deduct 1911 years from the Gregorian calendar
1885 # Months and days are identical
1886 $gy_offset = $gy - 1911;
1887 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1888 # Nengō dates up to Meiji period
1889 # Deduct years from the Gregorian calendar
1890 # depending on the nengo periods
1891 # Months and days are identical
1892 if ( ( $gy < 1912 )
1893 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1894 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1895 ) {
1896 # Meiji period
1897 $gy_gannen = $gy - 1868 + 1;
1898 $gy_offset = $gy_gannen;
1899 if ( $gy_gannen == 1 ) {
1900 $gy_offset = '元';
1901 }
1902 $gy_offset = '明治' . $gy_offset;
1903 } elseif (
1904 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1905 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1906 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1907 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1908 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1909 ) {
1910 # Taishō period
1911 $gy_gannen = $gy - 1912 + 1;
1912 $gy_offset = $gy_gannen;
1913 if ( $gy_gannen == 1 ) {
1914 $gy_offset = '元';
1915 }
1916 $gy_offset = '大正' . $gy_offset;
1917 } elseif (
1918 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1919 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1920 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1921 ) {
1922 # Shōwa period
1923 $gy_gannen = $gy - 1926 + 1;
1924 $gy_offset = $gy_gannen;
1925 if ( $gy_gannen == 1 ) {
1926 $gy_offset = '元';
1927 }
1928 $gy_offset = '昭和' . $gy_offset;
1929 } else {
1930 # Heisei period
1931 $gy_gannen = $gy - 1989 + 1;
1932 $gy_offset = $gy_gannen;
1933 if ( $gy_gannen == 1 ) {
1934 $gy_offset = '元';
1935 }
1936 $gy_offset = '平成' . $gy_offset;
1937 }
1938 } else {
1939 $gy_offset = $gy;
1940 }
1941
1942 return [ $gy_offset, $gm, $gd ];
1943 }
1944
1945 /**
1946 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1947 *
1948 * This is the rule the BIDI algorithm uses to determine the directionality of
1949 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1950 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1951 *
1952 * TODO: Does not handle BIDI control characters inside the text.
1953 * TODO: Does not handle unallocated characters.
1954 *
1955 * @param string $text Text to test
1956 * @return null|string Directionality ('ltr' or 'rtl') or null
1957 */
1958 private static function strongDirFromContent( $text = '' ) {
1959 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1960 return null;
1961 }
1962 if ( $matches[1] === '' ) {
1963 return 'rtl';
1964 }
1965 return 'ltr';
1966 }
1967
1968 /**
1969 * Roman number formatting up to 10000
1970 *
1971 * @param int $num
1972 *
1973 * @return string
1974 */
1975 static function romanNumeral( $num ) {
1976 static $table = [
1977 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1978 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1979 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1980 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1981 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1982 ];
1983
1984 $num = intval( $num );
1985 if ( $num > 10000 || $num <= 0 ) {
1986 return $num;
1987 }
1988
1989 $s = '';
1990 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1991 if ( $num >= $pow10 ) {
1992 $s .= $table[$i][(int)floor( $num / $pow10 )];
1993 }
1994 $num = $num % $pow10;
1995 }
1996 return $s;
1997 }
1998
1999 /**
2000 * Hebrew Gematria number formatting up to 9999
2001 *
2002 * @param int $num
2003 *
2004 * @return string
2005 */
2006 static function hebrewNumeral( $num ) {
2007 static $table = [
2008 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2009 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2010 [ '',
2011 [ 'ק' ],
2012 [ 'ר' ],
2013 [ 'ש' ],
2014 [ 'ת' ],
2015 [ 'ת', 'ק' ],
2016 [ 'ת', 'ר' ],
2017 [ 'ת', 'ש' ],
2018 [ 'ת', 'ת' ],
2019 [ 'ת', 'ת', 'ק' ],
2020 [ 'ת', 'ת', 'ר' ],
2021 ],
2022 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2023 ];
2024
2025 $num = intval( $num );
2026 if ( $num > 9999 || $num <= 0 ) {
2027 return $num;
2028 }
2029
2030 // Round thousands have special notations
2031 if ( $num === 1000 ) {
2032 return "א' אלף";
2033 } elseif ( $num % 1000 === 0 ) {
2034 return $table[0][$num / 1000] . "' אלפים";
2035 }
2036
2037 $letters = [];
2038
2039 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2040 if ( $num >= $pow10 ) {
2041 if ( $num === 15 || $num === 16 ) {
2042 $letters[] = $table[0][9];
2043 $letters[] = $table[0][$num - 9];
2044 $num = 0;
2045 } else {
2046 $letters = array_merge(
2047 $letters,
2048 (array)$table[$i][intval( $num / $pow10 )]
2049 );
2050
2051 if ( $pow10 === 1000 ) {
2052 $letters[] = "'";
2053 }
2054 }
2055 }
2056
2057 $num = $num % $pow10;
2058 }
2059
2060 $preTransformLength = count( $letters );
2061 if ( $preTransformLength === 1 ) {
2062 // Add geresh (single quote) to one-letter numbers
2063 $letters[] = "'";
2064 } else {
2065 $lastIndex = $preTransformLength - 1;
2066 $letters[$lastIndex] = str_replace(
2067 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2068 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2069 $letters[$lastIndex]
2070 );
2071
2072 // Add gershayim (double quote) to multiple-letter numbers,
2073 // but exclude numbers with only one letter after the thousands
2074 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2075 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2076 $letters[] = "'";
2077 } else {
2078 array_splice( $letters, -1, 0, '"' );
2079 }
2080 }
2081
2082 return implode( $letters );
2083 }
2084
2085 /**
2086 * Used by date() and time() to adjust the time output.
2087 *
2088 * @param string $ts The time in date('YmdHis') format
2089 * @param mixed $tz Adjust the time by this amount (default false, mean we
2090 * get user timecorrection setting)
2091 * @return int
2092 */
2093 public function userAdjust( $ts, $tz = false ) {
2094 global $wgUser, $wgLocalTZoffset;
2095
2096 if ( $tz === false ) {
2097 $tz = $wgUser->getOption( 'timecorrection' );
2098 }
2099
2100 $data = explode( '|', $tz, 3 );
2101
2102 if ( $data[0] == 'ZoneInfo' ) {
2103 try {
2104 $userTZ = new DateTimeZone( $data[2] );
2105 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2106 $date->setTimezone( $userTZ );
2107 return $date->format( 'YmdHis' );
2108 } catch ( Exception $e ) {
2109 // Unrecognized timezone, default to 'Offset' with the stored offset.
2110 $data[0] = 'Offset';
2111 }
2112 }
2113
2114 if ( $data[0] == 'System' || $tz == '' ) {
2115 # Global offset in minutes.
2116 $minDiff = $wgLocalTZoffset;
2117 } elseif ( $data[0] == 'Offset' ) {
2118 $minDiff = intval( $data[1] );
2119 } else {
2120 $data = explode( ':', $tz );
2121 if ( count( $data ) == 2 ) {
2122 $data[0] = intval( $data[0] );
2123 $data[1] = intval( $data[1] );
2124 $minDiff = abs( $data[0] ) * 60 + $data[1];
2125 if ( $data[0] < 0 ) {
2126 $minDiff = -$minDiff;
2127 }
2128 } else {
2129 $minDiff = intval( $data[0] ) * 60;
2130 }
2131 }
2132
2133 # No difference ? Return time unchanged
2134 if ( 0 == $minDiff ) {
2135 return $ts;
2136 }
2137
2138 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2139 # Generate an adjusted date; take advantage of the fact that mktime
2140 # will normalize out-of-range values so we don't have to split $minDiff
2141 # into hours and minutes.
2142 $t = mktime( (
2143 (int)substr( $ts, 8, 2 ) ), # Hours
2144 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2145 (int)substr( $ts, 12, 2 ), # Seconds
2146 (int)substr( $ts, 4, 2 ), # Month
2147 (int)substr( $ts, 6, 2 ), # Day
2148 (int)substr( $ts, 0, 4 ) ); # Year
2149
2150 $date = date( 'YmdHis', $t );
2151 MediaWiki\restoreWarnings();
2152
2153 return $date;
2154 }
2155
2156 /**
2157 * This is meant to be used by time(), date(), and timeanddate() to get
2158 * the date preference they're supposed to use, it should be used in
2159 * all children.
2160 *
2161 *<code>
2162 * function timeanddate([...], $format = true) {
2163 * $datePreference = $this->dateFormat($format);
2164 * [...]
2165 * }
2166 *</code>
2167 *
2168 * @param int|string|bool $usePrefs If true, the user's preference is used
2169 * if false, the site/language default is used
2170 * if int/string, assumed to be a format.
2171 * @return string
2172 */
2173 function dateFormat( $usePrefs = true ) {
2174 global $wgUser;
2175
2176 if ( is_bool( $usePrefs ) ) {
2177 if ( $usePrefs ) {
2178 $datePreference = $wgUser->getDatePreference();
2179 } else {
2180 $datePreference = (string)User::getDefaultOption( 'date' );
2181 }
2182 } else {
2183 $datePreference = (string)$usePrefs;
2184 }
2185
2186 // return int
2187 if ( $datePreference == '' ) {
2188 return 'default';
2189 }
2190
2191 return $datePreference;
2192 }
2193
2194 /**
2195 * Get a format string for a given type and preference
2196 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2197 * @param string $pref The format name as it appears in Messages*.php under
2198 * $datePreferences.
2199 *
2200 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2201 *
2202 * @return string
2203 */
2204 function getDateFormatString( $type, $pref ) {
2205 $wasDefault = false;
2206 if ( $pref == 'default' ) {
2207 $wasDefault = true;
2208 $pref = $this->getDefaultDateFormat();
2209 }
2210
2211 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2212 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2213
2214 if ( $type === 'pretty' && $df === null ) {
2215 $df = $this->getDateFormatString( 'date', $pref );
2216 }
2217
2218 if ( !$wasDefault && $df === null ) {
2219 $pref = $this->getDefaultDateFormat();
2220 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2221 }
2222
2223 $this->dateFormatStrings[$type][$pref] = $df;
2224 }
2225 return $this->dateFormatStrings[$type][$pref];
2226 }
2227
2228 /**
2229 * @param string $ts The time format which needs to be turned into a
2230 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2231 * @param bool $adj Whether to adjust the time output according to the
2232 * user configured offset ($timecorrection)
2233 * @param mixed $format True to use user's date format preference
2234 * @param string|bool $timecorrection The time offset as returned by
2235 * validateTimeZone() in Special:Preferences
2236 * @return string
2237 */
2238 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2239 $ts = wfTimestamp( TS_MW, $ts );
2240 if ( $adj ) {
2241 $ts = $this->userAdjust( $ts, $timecorrection );
2242 }
2243 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2244 return $this->sprintfDate( $df, $ts );
2245 }
2246
2247 /**
2248 * @param string $ts The time format which needs to be turned into a
2249 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2250 * @param bool $adj Whether to adjust the time output according to the
2251 * user configured offset ($timecorrection)
2252 * @param mixed $format True to use user's date format preference
2253 * @param string|bool $timecorrection The time offset as returned by
2254 * validateTimeZone() in Special:Preferences
2255 * @return string
2256 */
2257 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2258 $ts = wfTimestamp( TS_MW, $ts );
2259 if ( $adj ) {
2260 $ts = $this->userAdjust( $ts, $timecorrection );
2261 }
2262 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2263 return $this->sprintfDate( $df, $ts );
2264 }
2265
2266 /**
2267 * @param string $ts The time format which needs to be turned into a
2268 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2269 * @param bool $adj Whether to adjust the time output according to the
2270 * user configured offset ($timecorrection)
2271 * @param mixed $format What format to return, if it's false output the
2272 * default one (default true)
2273 * @param string|bool $timecorrection The time offset as returned by
2274 * validateTimeZone() in Special:Preferences
2275 * @return string
2276 */
2277 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2278 $ts = wfTimestamp( TS_MW, $ts );
2279 if ( $adj ) {
2280 $ts = $this->userAdjust( $ts, $timecorrection );
2281 }
2282 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2283 return $this->sprintfDate( $df, $ts );
2284 }
2285
2286 /**
2287 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2288 *
2289 * @since 1.20
2290 *
2291 * @param int $seconds The amount of seconds.
2292 * @param array $chosenIntervals The intervals to enable.
2293 *
2294 * @return string
2295 */
2296 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2297 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2298
2299 $segments = [];
2300
2301 foreach ( $intervals as $intervalName => $intervalValue ) {
2302 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2303 // duration-years, duration-decades, duration-centuries, duration-millennia
2304 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2305 $segments[] = $message->inLanguage( $this )->escaped();
2306 }
2307
2308 return $this->listToText( $segments );
2309 }
2310
2311 /**
2312 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2313 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2314 *
2315 * @since 1.20
2316 *
2317 * @param int $seconds The amount of seconds.
2318 * @param array $chosenIntervals The intervals to enable.
2319 *
2320 * @return array
2321 */
2322 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2323 if ( empty( $chosenIntervals ) ) {
2324 $chosenIntervals = [
2325 'millennia',
2326 'centuries',
2327 'decades',
2328 'years',
2329 'days',
2330 'hours',
2331 'minutes',
2332 'seconds'
2333 ];
2334 }
2335
2336 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2337 $sortedNames = array_keys( $intervals );
2338 $smallestInterval = array_pop( $sortedNames );
2339
2340 $segments = [];
2341
2342 foreach ( $intervals as $name => $length ) {
2343 $value = floor( $seconds / $length );
2344
2345 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2346 $seconds -= $value * $length;
2347 $segments[$name] = $value;
2348 }
2349 }
2350
2351 return $segments;
2352 }
2353
2354 /**
2355 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2356 *
2357 * @param string $type Can be 'date', 'time' or 'both'
2358 * @param string $ts The time format which needs to be turned into a
2359 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2360 * @param User $user User object used to get preferences for timezone and format
2361 * @param array $options Array, can contain the following keys:
2362 * - 'timecorrection': time correction, can have the following values:
2363 * - true: use user's preference
2364 * - false: don't use time correction
2365 * - int: value of time correction in minutes
2366 * - 'format': format to use, can have the following values:
2367 * - true: use user's preference
2368 * - false: use default preference
2369 * - string: format to use
2370 * @since 1.19
2371 * @return string
2372 */
2373 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2374 $ts = wfTimestamp( TS_MW, $ts );
2375 $options += [ 'timecorrection' => true, 'format' => true ];
2376 if ( $options['timecorrection'] !== false ) {
2377 if ( $options['timecorrection'] === true ) {
2378 $offset = $user->getOption( 'timecorrection' );
2379 } else {
2380 $offset = $options['timecorrection'];
2381 }
2382 $ts = $this->userAdjust( $ts, $offset );
2383 }
2384 if ( $options['format'] === true ) {
2385 $format = $user->getDatePreference();
2386 } else {
2387 $format = $options['format'];
2388 }
2389 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2390 return $this->sprintfDate( $df, $ts );
2391 }
2392
2393 /**
2394 * Get the formatted date for the given timestamp and formatted for
2395 * the given user.
2396 *
2397 * @param mixed $ts Mixed: the time format which needs to be turned into a
2398 * date('YmdHis') format with wfTimestamp(TS_MW,$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 userDate( $ts, User $user, array $options = [] ) {
2413 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2414 }
2415
2416 /**
2417 * Get the formatted 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 userTime( $ts, User $user, array $options = [] ) {
2436 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2437 }
2438
2439 /**
2440 * Get the formatted date and time for the given timestamp and formatted for
2441 * the given user.
2442 *
2443 * @param mixed $ts The time format which needs to be turned into a
2444 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2445 * @param User $user User object used to get preferences for timezone and format
2446 * @param array $options Array, can contain the following keys:
2447 * - 'timecorrection': time correction, can have the following values:
2448 * - true: use user's preference
2449 * - false: don't use time correction
2450 * - int: value of time correction in minutes
2451 * - 'format': format to use, can have the following values:
2452 * - true: use user's preference
2453 * - false: use default preference
2454 * - string: format to use
2455 * @since 1.19
2456 * @return string
2457 */
2458 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2459 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2460 }
2461
2462 /**
2463 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2464 *
2465 * Determine the difference between the timestamp and the current time, and
2466 * generate a readable timestamp by returning "<N> <units> ago", where the
2467 * largest possible unit is used.
2468 *
2469 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2470 *
2471 * @param MWTimestamp $time
2472 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2473 * @param User|null $user User the timestamp is being generated for
2474 * (or null to use main context's user)
2475 * @return string Formatted timestamp
2476 */
2477 public function getHumanTimestamp(
2478 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2479 ) {
2480 if ( $relativeTo === null ) {
2481 $relativeTo = new MWTimestamp();
2482 }
2483 if ( $user === null ) {
2484 $user = RequestContext::getMain()->getUser();
2485 }
2486
2487 // Adjust for the user's timezone.
2488 $offsetThis = $time->offsetForUser( $user );
2489 $offsetRel = $relativeTo->offsetForUser( $user );
2490
2491 $ts = '';
2492 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2493 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2494 }
2495
2496 // Reset the timezone on the objects.
2497 $time->timestamp->sub( $offsetThis );
2498 $relativeTo->timestamp->sub( $offsetRel );
2499
2500 return $ts;
2501 }
2502
2503 /**
2504 * Convert an MWTimestamp into a pretty human-readable timestamp using
2505 * the given user preferences and relative base time.
2506 *
2507 * @see Language::getHumanTimestamp
2508 * @param MWTimestamp $ts Timestamp to prettify
2509 * @param MWTimestamp $relativeTo Base timestamp
2510 * @param User $user User preferences to use
2511 * @return string Human timestamp
2512 * @since 1.26
2513 */
2514 private function getHumanTimestampInternal(
2515 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2516 ) {
2517 $diff = $ts->diff( $relativeTo );
2518 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2519 (int)$relativeTo->timestamp->format( 'w' ) );
2520 $days = $diff->days ?: (int)$diffDay;
2521 if ( $diff->invert || $days > 5
2522 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2523 ) {
2524 // Timestamps are in different years: use full timestamp
2525 // Also do full timestamp for future dates
2526 /**
2527 * @todo FIXME: Add better handling of future timestamps.
2528 */
2529 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2530 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2531 } elseif ( $days > 5 ) {
2532 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2533 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2534 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2535 } elseif ( $days > 1 ) {
2536 // Timestamp within the past week: show the day of the week and time
2537 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2538 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2539 // Messages:
2540 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2541 $ts = wfMessage( "$weekday-at" )
2542 ->inLanguage( $this )
2543 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2544 ->text();
2545 } elseif ( $days == 1 ) {
2546 // Timestamp was yesterday: say 'yesterday' and the time.
2547 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2548 $ts = wfMessage( 'yesterday-at' )
2549 ->inLanguage( $this )
2550 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2551 ->text();
2552 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2553 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2554 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2555 $ts = wfMessage( 'today-at' )
2556 ->inLanguage( $this )
2557 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2558 ->text();
2559
2560 // From here on in, the timestamp was soon enough ago so that we can simply say
2561 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2562 } elseif ( $diff->h == 1 ) {
2563 // Less than 90 minutes, but more than an hour ago.
2564 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2565 } elseif ( $diff->i >= 1 ) {
2566 // A few minutes ago.
2567 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2568 } elseif ( $diff->s >= 30 ) {
2569 // Less than a minute, but more than 30 sec ago.
2570 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2571 } else {
2572 // Less than 30 seconds ago.
2573 $ts = wfMessage( 'just-now' )->text();
2574 }
2575
2576 return $ts;
2577 }
2578
2579 /**
2580 * @param string $key
2581 * @return string|null
2582 */
2583 public function getMessage( $key ) {
2584 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2585 }
2586
2587 /**
2588 * @return array
2589 */
2590 function getAllMessages() {
2591 return self::$dataCache->getItem( $this->mCode, 'messages' );
2592 }
2593
2594 /**
2595 * @param string $in
2596 * @param string $out
2597 * @param string $string
2598 * @return string
2599 */
2600 public function iconv( $in, $out, $string ) {
2601 # Even with //IGNORE iconv can whine about illegal characters in
2602 # *input* string. We just ignore those too.
2603 # REF: https://bugs.php.net/bug.php?id=37166
2604 # REF: https://phabricator.wikimedia.org/T18885
2605 MediaWiki\suppressWarnings();
2606 $text = iconv( $in, $out . '//IGNORE', $string );
2607 MediaWiki\restoreWarnings();
2608 return $text;
2609 }
2610
2611 // callback functions for ucwords(), ucwordbreaks()
2612
2613 /**
2614 * @param array $matches
2615 * @return mixed|string
2616 */
2617 function ucwordbreaksCallbackAscii( $matches ) {
2618 return $this->ucfirst( $matches[1] );
2619 }
2620
2621 /**
2622 * @param array $matches
2623 * @return string
2624 */
2625 function ucwordbreaksCallbackMB( $matches ) {
2626 return mb_strtoupper( $matches[0] );
2627 }
2628
2629 /**
2630 * @param array $matches
2631 * @return string
2632 */
2633 function ucwordsCallbackMB( $matches ) {
2634 return mb_strtoupper( $matches[0] );
2635 }
2636
2637 /**
2638 * Make a string's first character uppercase
2639 *
2640 * @param string $str
2641 *
2642 * @return string
2643 */
2644 public function ucfirst( $str ) {
2645 $o = ord( $str );
2646 if ( $o < 96 ) { // if already uppercase...
2647 return $str;
2648 } elseif ( $o < 128 ) {
2649 return ucfirst( $str ); // use PHP's ucfirst()
2650 } else {
2651 // fall back to more complex logic in case of multibyte strings
2652 return $this->uc( $str, true );
2653 }
2654 }
2655
2656 /**
2657 * Convert a string to uppercase
2658 *
2659 * @param string $str
2660 * @param bool $first
2661 *
2662 * @return string
2663 */
2664 public function uc( $str, $first = false ) {
2665 if ( $first ) {
2666 if ( $this->isMultibyte( $str ) ) {
2667 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2668 } else {
2669 return ucfirst( $str );
2670 }
2671 } else {
2672 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2673 }
2674 }
2675
2676 /**
2677 * @param string $str
2678 * @return mixed|string
2679 */
2680 function lcfirst( $str ) {
2681 $o = ord( $str );
2682 if ( !$o ) {
2683 return strval( $str );
2684 } elseif ( $o >= 128 ) {
2685 return $this->lc( $str, true );
2686 } elseif ( $o > 96 ) {
2687 return $str;
2688 } else {
2689 $str[0] = strtolower( $str[0] );
2690 return $str;
2691 }
2692 }
2693
2694 /**
2695 * @param string $str
2696 * @param bool $first
2697 * @return mixed|string
2698 */
2699 function lc( $str, $first = false ) {
2700 if ( $first ) {
2701 if ( $this->isMultibyte( $str ) ) {
2702 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2703 } else {
2704 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2705 }
2706 } else {
2707 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2708 }
2709 }
2710
2711 /**
2712 * @param string $str
2713 * @return bool
2714 */
2715 function isMultibyte( $str ) {
2716 return strlen( $str ) !== mb_strlen( $str );
2717 }
2718
2719 /**
2720 * @param string $str
2721 * @return mixed|string
2722 */
2723 function ucwords( $str ) {
2724 if ( $this->isMultibyte( $str ) ) {
2725 $str = $this->lc( $str );
2726
2727 // regexp to find first letter in each word (i.e. after each space)
2728 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2729
2730 // function to use to capitalize a single char
2731 return preg_replace_callback(
2732 $replaceRegexp,
2733 [ $this, 'ucwordsCallbackMB' ],
2734 $str
2735 );
2736 } else {
2737 return ucwords( strtolower( $str ) );
2738 }
2739 }
2740
2741 /**
2742 * capitalize words at word breaks
2743 *
2744 * @param string $str
2745 * @return mixed
2746 */
2747 function ucwordbreaks( $str ) {
2748 if ( $this->isMultibyte( $str ) ) {
2749 $str = $this->lc( $str );
2750
2751 // since \b doesn't work for UTF-8, we explicitely define word break chars
2752 $breaks = "[ \-\(\)\}\{\.,\?!]";
2753
2754 // find first letter after word break
2755 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2756 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2757
2758 return preg_replace_callback(
2759 $replaceRegexp,
2760 [ $this, 'ucwordbreaksCallbackMB' ],
2761 $str
2762 );
2763 } else {
2764 return preg_replace_callback(
2765 '/\b([\w\x80-\xff]+)\b/',
2766 [ $this, 'ucwordbreaksCallbackAscii' ],
2767 $str
2768 );
2769 }
2770 }
2771
2772 /**
2773 * Return a case-folded representation of $s
2774 *
2775 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2776 * and $s2 are the same except for the case of their characters. It is not
2777 * necessary for the value returned to make sense when displayed.
2778 *
2779 * Do *not* perform any other normalisation in this function. If a caller
2780 * uses this function when it should be using a more general normalisation
2781 * function, then fix the caller.
2782 *
2783 * @param string $s
2784 *
2785 * @return string
2786 */
2787 function caseFold( $s ) {
2788 return $this->uc( $s );
2789 }
2790
2791 /**
2792 * @param string $s
2793 * @return string
2794 * @throws MWException
2795 */
2796 function checkTitleEncoding( $s ) {
2797 if ( is_array( $s ) ) {
2798 throw new MWException( 'Given array to checkTitleEncoding.' );
2799 }
2800 if ( StringUtils::isUtf8( $s ) ) {
2801 return $s;
2802 }
2803
2804 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2805 }
2806
2807 /**
2808 * @return array
2809 */
2810 function fallback8bitEncoding() {
2811 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2812 }
2813
2814 /**
2815 * Most writing systems use whitespace to break up words.
2816 * Some languages such as Chinese don't conventionally do this,
2817 * which requires special handling when breaking up words for
2818 * searching etc.
2819 *
2820 * @return bool
2821 */
2822 function hasWordBreaks() {
2823 return true;
2824 }
2825
2826 /**
2827 * Some languages such as Chinese require word segmentation,
2828 * Specify such segmentation when overridden in derived class.
2829 *
2830 * @param string $string
2831 * @return string
2832 */
2833 function segmentByWord( $string ) {
2834 return $string;
2835 }
2836
2837 /**
2838 * Some languages have special punctuation need to be normalized.
2839 * Make such changes here.
2840 *
2841 * @param string $string
2842 * @return string
2843 */
2844 function normalizeForSearch( $string ) {
2845 return self::convertDoubleWidth( $string );
2846 }
2847
2848 /**
2849 * convert double-width roman characters to single-width.
2850 * range: ff00-ff5f ~= 0020-007f
2851 *
2852 * @param string $string
2853 *
2854 * @return string
2855 */
2856 protected static function convertDoubleWidth( $string ) {
2857 static $full = null;
2858 static $half = null;
2859
2860 if ( $full === null ) {
2861 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2862 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2863 $full = str_split( $fullWidth, 3 );
2864 $half = str_split( $halfWidth );
2865 }
2866
2867 $string = str_replace( $full, $half, $string );
2868 return $string;
2869 }
2870
2871 /**
2872 * @param string $string
2873 * @param string $pattern
2874 * @return string
2875 */
2876 protected static function insertSpace( $string, $pattern ) {
2877 $string = preg_replace( $pattern, " $1 ", $string );
2878 $string = preg_replace( '/ +/', ' ', $string );
2879 return $string;
2880 }
2881
2882 /**
2883 * @param array $termsArray
2884 * @return array
2885 */
2886 function convertForSearchResult( $termsArray ) {
2887 # some languages, e.g. Chinese, need to do a conversion
2888 # in order for search results to be displayed correctly
2889 return $termsArray;
2890 }
2891
2892 /**
2893 * Get the first character of a string.
2894 *
2895 * @param string $s
2896 * @return string
2897 */
2898 function firstChar( $s ) {
2899 $matches = [];
2900 preg_match(
2901 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2902 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2903 $s,
2904 $matches
2905 );
2906
2907 if ( isset( $matches[1] ) ) {
2908 if ( strlen( $matches[1] ) != 3 ) {
2909 return $matches[1];
2910 }
2911
2912 // Break down Hangul syllables to grab the first jamo
2913 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2914 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2915 return $matches[1];
2916 } elseif ( $code < 0xb098 ) {
2917 return "\xe3\x84\xb1";
2918 } elseif ( $code < 0xb2e4 ) {
2919 return "\xe3\x84\xb4";
2920 } elseif ( $code < 0xb77c ) {
2921 return "\xe3\x84\xb7";
2922 } elseif ( $code < 0xb9c8 ) {
2923 return "\xe3\x84\xb9";
2924 } elseif ( $code < 0xbc14 ) {
2925 return "\xe3\x85\x81";
2926 } elseif ( $code < 0xc0ac ) {
2927 return "\xe3\x85\x82";
2928 } elseif ( $code < 0xc544 ) {
2929 return "\xe3\x85\x85";
2930 } elseif ( $code < 0xc790 ) {
2931 return "\xe3\x85\x87";
2932 } elseif ( $code < 0xcc28 ) {
2933 return "\xe3\x85\x88";
2934 } elseif ( $code < 0xce74 ) {
2935 return "\xe3\x85\x8a";
2936 } elseif ( $code < 0xd0c0 ) {
2937 return "\xe3\x85\x8b";
2938 } elseif ( $code < 0xd30c ) {
2939 return "\xe3\x85\x8c";
2940 } elseif ( $code < 0xd558 ) {
2941 return "\xe3\x85\x8d";
2942 } else {
2943 return "\xe3\x85\x8e";
2944 }
2945 } else {
2946 return '';
2947 }
2948 }
2949
2950 /**
2951 * @deprecated No-op since 1.28
2952 */
2953 function initEncoding() {
2954 // No-op.
2955 }
2956
2957 /**
2958 * @param string $s
2959 * @return string
2960 * @deprecated No-op since 1.28
2961 */
2962 function recodeForEdit( $s ) {
2963 return $s;
2964 }
2965
2966 /**
2967 * @param string $s
2968 * @return string
2969 * @deprecated No-op since 1.28
2970 */
2971 function recodeInput( $s ) {
2972 return $s;
2973 }
2974
2975 /**
2976 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2977 * also cleans up certain backwards-compatible sequences, converting them
2978 * to the modern Unicode equivalent.
2979 *
2980 * This is language-specific for performance reasons only.
2981 *
2982 * @param string $s
2983 *
2984 * @return string
2985 */
2986 function normalize( $s ) {
2987 global $wgAllUnicodeFixes;
2988 $s = UtfNormal\Validator::cleanUp( $s );
2989 if ( $wgAllUnicodeFixes ) {
2990 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2991 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2992 }
2993
2994 return $s;
2995 }
2996
2997 /**
2998 * Transform a string using serialized data stored in the given file (which
2999 * must be in the serialized subdirectory of $IP). The file contains pairs
3000 * mapping source characters to destination characters.
3001 *
3002 * The data is cached in process memory. This will go faster if you have the
3003 * FastStringSearch extension.
3004 *
3005 * @param string $file
3006 * @param string $string
3007 *
3008 * @throws MWException
3009 * @return string
3010 */
3011 function transformUsingPairFile( $file, $string ) {
3012 if ( !isset( $this->transformData[$file] ) ) {
3013 $data = wfGetPrecompiledData( $file );
3014 if ( $data === false ) {
3015 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
3016 }
3017 $this->transformData[$file] = new ReplacementArray( $data );
3018 }
3019 return $this->transformData[$file]->replace( $string );
3020 }
3021
3022 /**
3023 * For right-to-left language support
3024 *
3025 * @return bool
3026 */
3027 function isRTL() {
3028 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3029 }
3030
3031 /**
3032 * Return the correct HTML 'dir' attribute value for this language.
3033 * @return string
3034 */
3035 function getDir() {
3036 return $this->isRTL() ? 'rtl' : 'ltr';
3037 }
3038
3039 /**
3040 * Return 'left' or 'right' as appropriate alignment for line-start
3041 * for this language's text direction.
3042 *
3043 * Should be equivalent to CSS3 'start' text-align value....
3044 *
3045 * @return string
3046 */
3047 function alignStart() {
3048 return $this->isRTL() ? 'right' : 'left';
3049 }
3050
3051 /**
3052 * Return 'right' or 'left' as appropriate alignment for line-end
3053 * for this language's text direction.
3054 *
3055 * Should be equivalent to CSS3 'end' text-align value....
3056 *
3057 * @return string
3058 */
3059 function alignEnd() {
3060 return $this->isRTL() ? 'left' : 'right';
3061 }
3062
3063 /**
3064 * A hidden direction mark (LRM or RLM), depending on the language direction.
3065 * Unlike getDirMark(), this function returns the character as an HTML entity.
3066 * This function should be used when the output is guaranteed to be HTML,
3067 * because it makes the output HTML source code more readable. When
3068 * the output is plain text or can be escaped, getDirMark() should be used.
3069 *
3070 * @param bool $opposite Get the direction mark opposite to your language
3071 * @return string
3072 * @since 1.20
3073 */
3074 function getDirMarkEntity( $opposite = false ) {
3075 if ( $opposite ) {
3076 return $this->isRTL() ? '&lrm;' : '&rlm;';
3077 }
3078 return $this->isRTL() ? '&rlm;' : '&lrm;';
3079 }
3080
3081 /**
3082 * A hidden direction mark (LRM or RLM), depending on the language direction.
3083 * This function produces them as invisible Unicode characters and
3084 * the output may be hard to read and debug, so it should only be used
3085 * when the output is plain text or can be escaped. When the output is
3086 * HTML, use getDirMarkEntity() instead.
3087 *
3088 * @param bool $opposite Get the direction mark opposite to your language
3089 * @return string
3090 */
3091 function getDirMark( $opposite = false ) {
3092 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3093 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3094 if ( $opposite ) {
3095 return $this->isRTL() ? $lrm : $rlm;
3096 }
3097 return $this->isRTL() ? $rlm : $lrm;
3098 }
3099
3100 /**
3101 * @return array
3102 */
3103 function capitalizeAllNouns() {
3104 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3105 }
3106
3107 /**
3108 * An arrow, depending on the language direction.
3109 *
3110 * @param string $direction The direction of the arrow: forwards (default),
3111 * backwards, left, right, up, down.
3112 * @return string
3113 */
3114 function getArrow( $direction = 'forwards' ) {
3115 switch ( $direction ) {
3116 case 'forwards':
3117 return $this->isRTL() ? '←' : '→';
3118 case 'backwards':
3119 return $this->isRTL() ? '→' : '←';
3120 case 'left':
3121 return '←';
3122 case 'right':
3123 return '→';
3124 case 'up':
3125 return '↑';
3126 case 'down':
3127 return '↓';
3128 }
3129 }
3130
3131 /**
3132 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3133 *
3134 * @return bool
3135 */
3136 function linkPrefixExtension() {
3137 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3138 }
3139
3140 /**
3141 * Get all magic words from cache.
3142 * @return array
3143 */
3144 function getMagicWords() {
3145 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3146 }
3147
3148 /**
3149 * Run the LanguageGetMagic hook once.
3150 */
3151 protected function doMagicHook() {
3152 if ( $this->mMagicHookDone ) {
3153 return;
3154 }
3155 $this->mMagicHookDone = true;
3156 Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
3157 }
3158
3159 /**
3160 * Fill a MagicWord object with data from here
3161 *
3162 * @param MagicWord $mw
3163 */
3164 function getMagic( $mw ) {
3165 // Saves a function call
3166 if ( !$this->mMagicHookDone ) {
3167 $this->doMagicHook();
3168 }
3169
3170 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3171 $rawEntry = $this->mMagicExtensions[$mw->mId];
3172 } else {
3173 $rawEntry = self::$dataCache->getSubitem(
3174 $this->mCode, 'magicWords', $mw->mId );
3175 }
3176
3177 if ( !is_array( $rawEntry ) ) {
3178 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3179 } else {
3180 $mw->mCaseSensitive = $rawEntry[0];
3181 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3182 }
3183 }
3184
3185 /**
3186 * Add magic words to the extension array
3187 *
3188 * @param array $newWords
3189 */
3190 function addMagicWordsByLang( $newWords ) {
3191 $fallbackChain = $this->getFallbackLanguages();
3192 $fallbackChain = array_reverse( $fallbackChain );
3193 foreach ( $fallbackChain as $code ) {
3194 if ( isset( $newWords[$code] ) ) {
3195 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3196 }
3197 }
3198 }
3199
3200 /**
3201 * Get special page names, as an associative array
3202 * canonical name => array of valid names, including aliases
3203 * @return array
3204 */
3205 function getSpecialPageAliases() {
3206 // Cache aliases because it may be slow to load them
3207 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3208 // Initialise array
3209 $this->mExtendedSpecialPageAliases =
3210 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3211 Hooks::run( 'LanguageGetSpecialPageAliases',
3212 [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
3213 }
3214
3215 return $this->mExtendedSpecialPageAliases;
3216 }
3217
3218 /**
3219 * Italic is unsuitable for some languages
3220 *
3221 * @param string $text The text to be emphasized.
3222 * @return string
3223 */
3224 function emphasize( $text ) {
3225 return "<em>$text</em>";
3226 }
3227
3228 /**
3229 * Normally we output all numbers in plain en_US style, that is
3230 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3231 * point twohundredthirtyfive. However this is not suitable for all
3232 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3233 * Icelandic just want to use commas instead of dots, and dots instead
3234 * of commas like "293.291,235".
3235 *
3236 * An example of this function being called:
3237 * <code>
3238 * wfMessage( 'message' )->numParams( $num )->text()
3239 * </code>
3240 *
3241 * See $separatorTransformTable on MessageIs.php for
3242 * the , => . and . => , implementation.
3243 *
3244 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3245 * @param int|float $number The string to be formatted, should be an integer
3246 * or a floating point number.
3247 * @param bool $nocommafy Set to true for special numbers like dates
3248 * @return string
3249 */
3250 public function formatNum( $number, $nocommafy = false ) {
3251 global $wgTranslateNumerals;
3252 if ( !$nocommafy ) {
3253 $number = $this->commafy( $number );
3254 $s = $this->separatorTransformTable();
3255 if ( $s ) {
3256 $number = strtr( $number, $s );
3257 }
3258 }
3259
3260 if ( $wgTranslateNumerals ) {
3261 $s = $this->digitTransformTable();
3262 if ( $s ) {
3263 $number = strtr( $number, $s );
3264 }
3265 }
3266
3267 return $number;
3268 }
3269
3270 /**
3271 * Front-end for non-commafied formatNum
3272 *
3273 * @param int|float $number The string to be formatted, should be an integer
3274 * or a floating point number.
3275 * @since 1.21
3276 * @return string
3277 */
3278 public function formatNumNoSeparators( $number ) {
3279 return $this->formatNum( $number, true );
3280 }
3281
3282 /**
3283 * @param string $number
3284 * @return string
3285 */
3286 public function parseFormattedNumber( $number ) {
3287 $s = $this->digitTransformTable();
3288 if ( $s ) {
3289 // eliminate empty array values such as ''. (bug 64347)
3290 $s = array_filter( $s );
3291 $number = strtr( $number, array_flip( $s ) );
3292 }
3293
3294 $s = $this->separatorTransformTable();
3295 if ( $s ) {
3296 // eliminate empty array values such as ''. (bug 64347)
3297 $s = array_filter( $s );
3298 $number = strtr( $number, array_flip( $s ) );
3299 }
3300
3301 $number = strtr( $number, [ ',' => '' ] );
3302 return $number;
3303 }
3304
3305 /**
3306 * Adds commas to a given number
3307 * @since 1.19
3308 * @param mixed $number
3309 * @return string
3310 */
3311 function commafy( $number ) {
3312 $digitGroupingPattern = $this->digitGroupingPattern();
3313 if ( $number === null ) {
3314 return '';
3315 }
3316
3317 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3318 // default grouping is at thousands, use the same for ###,###,### pattern too.
3319 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3320 } else {
3321 // Ref: http://cldr.unicode.org/translation/number-patterns
3322 $sign = "";
3323 if ( intval( $number ) < 0 ) {
3324 // For negative numbers apply the algorithm like positive number and add sign.
3325 $sign = "-";
3326 $number = substr( $number, 1 );
3327 }
3328 $integerPart = [];
3329 $decimalPart = [];
3330 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3331 preg_match( "/\d+/", $number, $integerPart );
3332 preg_match( "/\.\d*/", $number, $decimalPart );
3333 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3334 if ( $groupedNumber === $number ) {
3335 // the string does not have any number part. Eg: .12345
3336 return $sign . $groupedNumber;
3337 }
3338 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3339 while ( $start > 0 ) {
3340 $match = $matches[0][$numMatches - 1];
3341 $matchLen = strlen( $match );
3342 $start = $end - $matchLen;
3343 if ( $start < 0 ) {
3344 $start = 0;
3345 }
3346 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3347 $end = $start;
3348 if ( $numMatches > 1 ) {
3349 // use the last pattern for the rest of the number
3350 $numMatches--;
3351 }
3352 if ( $start > 0 ) {
3353 $groupedNumber = "," . $groupedNumber;
3354 }
3355 }
3356 return $sign . $groupedNumber;
3357 }
3358 }
3359
3360 /**
3361 * @return string
3362 */
3363 function digitGroupingPattern() {
3364 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3365 }
3366
3367 /**
3368 * @return array
3369 */
3370 function digitTransformTable() {
3371 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3372 }
3373
3374 /**
3375 * @return array
3376 */
3377 function separatorTransformTable() {
3378 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3379 }
3380
3381 /**
3382 * Take a list of strings and build a locale-friendly comma-separated
3383 * list, using the local comma-separator message.
3384 * The last two strings are chained with an "and".
3385 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3386 *
3387 * @param string[] $l
3388 * @return string
3389 */
3390 function listToText( array $l ) {
3391 $m = count( $l ) - 1;
3392 if ( $m < 0 ) {
3393 return '';
3394 }
3395 if ( $m > 0 ) {
3396 $and = $this->msg( 'and' )->escaped();
3397 $space = $this->msg( 'word-separator' )->escaped();
3398 if ( $m > 1 ) {
3399 $comma = $this->msg( 'comma-separator' )->escaped();
3400 }
3401 }
3402 $s = $l[$m];
3403 for ( $i = $m - 1; $i >= 0; $i-- ) {
3404 if ( $i == $m - 1 ) {
3405 $s = $l[$i] . $and . $space . $s;
3406 } else {
3407 $s = $l[$i] . $comma . $s;
3408 }
3409 }
3410 return $s;
3411 }
3412
3413 /**
3414 * Take a list of strings and build a locale-friendly comma-separated
3415 * list, using the local comma-separator message.
3416 * @param string[] $list Array of strings to put in a comma list
3417 * @return string
3418 */
3419 function commaList( array $list ) {
3420 return implode(
3421 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3422 $list
3423 );
3424 }
3425
3426 /**
3427 * Take a list of strings and build a locale-friendly semicolon-separated
3428 * list, using the local semicolon-separator message.
3429 * @param string[] $list Array of strings to put in a semicolon list
3430 * @return string
3431 */
3432 function semicolonList( array $list ) {
3433 return implode(
3434 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3435 $list
3436 );
3437 }
3438
3439 /**
3440 * Same as commaList, but separate it with the pipe instead.
3441 * @param string[] $list Array of strings to put in a pipe list
3442 * @return string
3443 */
3444 function pipeList( array $list ) {
3445 return implode(
3446 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3447 $list
3448 );
3449 }
3450
3451 /**
3452 * Truncate a string to a specified length in bytes, appending an optional
3453 * string (e.g. for ellipses)
3454 *
3455 * The database offers limited byte lengths for some columns in the database;
3456 * multi-byte character sets mean we need to ensure that only whole characters
3457 * are included, otherwise broken characters can be passed to the user
3458 *
3459 * If $length is negative, the string will be truncated from the beginning
3460 *
3461 * @param string $string String to truncate
3462 * @param int $length Maximum length (including ellipses)
3463 * @param string $ellipsis String to append to the truncated text
3464 * @param bool $adjustLength Subtract length of ellipsis from $length.
3465 * $adjustLength was introduced in 1.18, before that behaved as if false.
3466 * @return string
3467 */
3468 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3469 # Use the localized ellipsis character
3470 if ( $ellipsis == '...' ) {
3471 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3472 }
3473 # Check if there is no need to truncate
3474 if ( $length == 0 ) {
3475 return $ellipsis; // convention
3476 } elseif ( strlen( $string ) <= abs( $length ) ) {
3477 return $string; // no need to truncate
3478 }
3479 $stringOriginal = $string;
3480 # If ellipsis length is >= $length then we can't apply $adjustLength
3481 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3482 $string = $ellipsis; // this can be slightly unexpected
3483 # Otherwise, truncate and add ellipsis...
3484 } else {
3485 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3486 if ( $length > 0 ) {
3487 $length -= $eLength;
3488 $string = substr( $string, 0, $length ); // xyz...
3489 $string = $this->removeBadCharLast( $string );
3490 $string = rtrim( $string );
3491 $string = $string . $ellipsis;
3492 } else {
3493 $length += $eLength;
3494 $string = substr( $string, $length ); // ...xyz
3495 $string = $this->removeBadCharFirst( $string );
3496 $string = ltrim( $string );
3497 $string = $ellipsis . $string;
3498 }
3499 }
3500 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3501 # This check is *not* redundant if $adjustLength, due to the single case where
3502 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3503 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3504 return $string;
3505 } else {
3506 return $stringOriginal;
3507 }
3508 }
3509
3510 /**
3511 * Remove bytes that represent an incomplete Unicode character
3512 * at the end of string (e.g. bytes of the char are missing)
3513 *
3514 * @param string $string
3515 * @return string
3516 */
3517 protected function removeBadCharLast( $string ) {
3518 if ( $string != '' ) {
3519 $char = ord( $string[strlen( $string ) - 1] );
3520 $m = [];
3521 if ( $char >= 0xc0 ) {
3522 # We got the first byte only of a multibyte char; remove it.
3523 $string = substr( $string, 0, -1 );
3524 } elseif ( $char >= 0x80 &&
3525 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3526 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3527 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3528 ) {
3529 # We chopped in the middle of a character; remove it
3530 $string = $m[1];
3531 }
3532 }
3533 return $string;
3534 }
3535
3536 /**
3537 * Remove bytes that represent an incomplete Unicode character
3538 * at the start of string (e.g. bytes of the char are missing)
3539 *
3540 * @param string $string
3541 * @return string
3542 */
3543 protected function removeBadCharFirst( $string ) {
3544 if ( $string != '' ) {
3545 $char = ord( $string[0] );
3546 if ( $char >= 0x80 && $char < 0xc0 ) {
3547 # We chopped in the middle of a character; remove the whole thing
3548 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3549 }
3550 }
3551 return $string;
3552 }
3553
3554 /**
3555 * Truncate a string of valid HTML to a specified length in bytes,
3556 * appending an optional string (e.g. for ellipses), and return valid HTML
3557 *
3558 * This is only intended for styled/linked text, such as HTML with
3559 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3560 * Also, this will not detect things like "display:none" CSS.
3561 *
3562 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3563 *
3564 * @param string $text HTML string to truncate
3565 * @param int $length (zero/positive) Maximum length (including ellipses)
3566 * @param string $ellipsis String to append to the truncated text
3567 * @return string
3568 */
3569 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3570 # Use the localized ellipsis character
3571 if ( $ellipsis == '...' ) {
3572 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3573 }
3574 # Check if there is clearly no need to truncate
3575 if ( $length <= 0 ) {
3576 return $ellipsis; // no text shown, nothing to format (convention)
3577 } elseif ( strlen( $text ) <= $length ) {
3578 return $text; // string short enough even *with* HTML (short-circuit)
3579 }
3580
3581 $dispLen = 0; // innerHTML legth so far
3582 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3583 $tagType = 0; // 0-open, 1-close
3584 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3585 $entityState = 0; // 0-not entity, 1-entity
3586 $tag = $ret = ''; // accumulated tag name, accumulated result string
3587 $openTags = []; // open tag stack
3588 $maybeState = null; // possible truncation state
3589
3590 $textLen = strlen( $text );
3591 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3592 for ( $pos = 0; true; ++$pos ) {
3593 # Consider truncation once the display length has reached the maximim.
3594 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3595 # Check that we're not in the middle of a bracket/entity...
3596 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3597 if ( !$testingEllipsis ) {
3598 $testingEllipsis = true;
3599 # Save where we are; we will truncate here unless there turn out to
3600 # be so few remaining characters that truncation is not necessary.
3601 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3602 $maybeState = [ $ret, $openTags ]; // save state
3603 }
3604 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3605 # String in fact does need truncation, the truncation point was OK.
3606 list( $ret, $openTags ) = $maybeState; // reload state
3607 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3608 $ret .= $ellipsis; // add ellipsis
3609 break;
3610 }
3611 }
3612 if ( $pos >= $textLen ) {
3613 break; // extra iteration just for above checks
3614 }
3615
3616 # Read the next char...
3617 $ch = $text[$pos];
3618 $lastCh = $pos ? $text[$pos - 1] : '';
3619 $ret .= $ch; // add to result string
3620 if ( $ch == '<' ) {
3621 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3622 $entityState = 0; // for bad HTML
3623 $bracketState = 1; // tag started (checking for backslash)
3624 } elseif ( $ch == '>' ) {
3625 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3626 $entityState = 0; // for bad HTML
3627 $bracketState = 0; // out of brackets
3628 } elseif ( $bracketState == 1 ) {
3629 if ( $ch == '/' ) {
3630 $tagType = 1; // close tag (e.g. "</span>")
3631 } else {
3632 $tagType = 0; // open tag (e.g. "<span>")
3633 $tag .= $ch;
3634 }
3635 $bracketState = 2; // building tag name
3636 } elseif ( $bracketState == 2 ) {
3637 if ( $ch != ' ' ) {
3638 $tag .= $ch;
3639 } else {
3640 // Name found (e.g. "<a href=..."), add on tag attributes...
3641 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3642 }
3643 } elseif ( $bracketState == 0 ) {
3644 if ( $entityState ) {
3645 if ( $ch == ';' ) {
3646 $entityState = 0;
3647 $dispLen++; // entity is one displayed char
3648 }
3649 } else {
3650 if ( $neLength == 0 && !$maybeState ) {
3651 // Save state without $ch. We want to *hit* the first
3652 // display char (to get tags) but not *use* it if truncating.
3653 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3654 }
3655 if ( $ch == '&' ) {
3656 $entityState = 1; // entity found, (e.g. "&#160;")
3657 } else {
3658 $dispLen++; // this char is displayed
3659 // Add the next $max display text chars after this in one swoop...
3660 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3661 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3662 $dispLen += $skipped;
3663 $pos += $skipped;
3664 }
3665 }
3666 }
3667 }
3668 // Close the last tag if left unclosed by bad HTML
3669 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3670 while ( count( $openTags ) > 0 ) {
3671 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3672 }
3673 return $ret;
3674 }
3675
3676 /**
3677 * truncateHtml() helper function
3678 * like strcspn() but adds the skipped chars to $ret
3679 *
3680 * @param string $ret
3681 * @param string $text
3682 * @param string $search
3683 * @param int $start
3684 * @param null|int $len
3685 * @return int
3686 */
3687 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3688 if ( $len === null ) {
3689 $len = -1; // -1 means "no limit" for strcspn
3690 } elseif ( $len < 0 ) {
3691 $len = 0; // sanity
3692 }
3693 $skipCount = 0;
3694 if ( $start < strlen( $text ) ) {
3695 $skipCount = strcspn( $text, $search, $start, $len );
3696 $ret .= substr( $text, $start, $skipCount );
3697 }
3698 return $skipCount;
3699 }
3700
3701 /**
3702 * truncateHtml() helper function
3703 * (a) push or pop $tag from $openTags as needed
3704 * (b) clear $tag value
3705 * @param string &$tag Current HTML tag name we are looking at
3706 * @param int $tagType (0-open tag, 1-close tag)
3707 * @param string $lastCh Character before the '>' that ended this tag
3708 * @param array &$openTags Open tag stack (not accounting for $tag)
3709 */
3710 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3711 $tag = ltrim( $tag );
3712 if ( $tag != '' ) {
3713 if ( $tagType == 0 && $lastCh != '/' ) {
3714 $openTags[] = $tag; // tag opened (didn't close itself)
3715 } elseif ( $tagType == 1 ) {
3716 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3717 array_pop( $openTags ); // tag closed
3718 }
3719 }
3720 $tag = '';
3721 }
3722 }
3723
3724 /**
3725 * Grammatical transformations, needed for inflected languages
3726 * Invoked by putting {{grammar:case|word}} in a message
3727 *
3728 * @param string $word
3729 * @param string $case
3730 * @return string
3731 */
3732 function convertGrammar( $word, $case ) {
3733 global $wgGrammarForms;
3734 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3735 return $wgGrammarForms[$this->getCode()][$case][$word];
3736 }
3737
3738 $grammarTransformations = $this->getGrammarTransformations();
3739
3740 if ( isset( $grammarTransformations[$case] ) ) {
3741 $forms = $grammarTransformations[$case];
3742
3743 // Some names of grammar rules are aliases for other rules.
3744 // In such cases the value is a string rather than object,
3745 // so load the actual rules.
3746 if ( is_string( $forms ) ) {
3747 $forms = $grammarTransformations[$forms];
3748 }
3749
3750 foreach ( array_values( $forms ) as $rule ) {
3751 $form = $rule[0];
3752
3753 if ( $form === '@metadata' ) {
3754 continue;
3755 }
3756
3757 $replacement = $rule[1];
3758
3759 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3760 $patternMatches = preg_match( $regex, $word );
3761
3762 if ( $patternMatches === false ) {
3763 wfLogWarning(
3764 'An error occurred while processing grammar. ' .
3765 "Word: '$word'. Regex: /$form/."
3766 );
3767 } elseif ( $patternMatches === 1 ) {
3768 $word = preg_replace( $regex, $replacement, $word );
3769
3770 break;
3771 }
3772 }
3773 }
3774
3775 return $word;
3776 }
3777
3778 /**
3779 * Get the grammar forms for the content language
3780 * @return array Array of grammar forms
3781 * @since 1.20
3782 */
3783 function getGrammarForms() {
3784 global $wgGrammarForms;
3785 if ( isset( $wgGrammarForms[$this->getCode()] )
3786 && is_array( $wgGrammarForms[$this->getCode()] )
3787 ) {
3788 return $wgGrammarForms[$this->getCode()];
3789 }
3790
3791 return [];
3792 }
3793
3794 /**
3795 * Get the grammar transformations data for the language.
3796 * Used like grammar forms, with {{GRAMMAR}} and cases,
3797 * but uses pairs of regexes and replacements instead of code.
3798 *
3799 * @return array[] Array of grammar transformations.
3800 * @throws MWException
3801 * @since 1.28
3802 */
3803 public function getGrammarTransformations() {
3804 $languageCode = $this->getCode();
3805
3806 if ( self::$grammarTransformations === null ) {
3807 self::$grammarTransformations = new MapCacheLRU( 10 );
3808 }
3809
3810 if ( self::$grammarTransformations->has( $languageCode ) ) {
3811 return self::$grammarTransformations->get( $languageCode );
3812 }
3813
3814 $data = [];
3815
3816 $grammarDataFile = __DIR__ . "/data/grammarTransformations/$languageCode.json";
3817 if ( is_readable( $grammarDataFile ) ) {
3818 $data = FormatJson::decode(
3819 file_get_contents( $grammarDataFile ),
3820 true
3821 );
3822
3823 if ( $data === null ) {
3824 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3825 }
3826
3827 self::$grammarTransformations->set( $languageCode, $data );
3828 }
3829
3830 return $data;
3831 }
3832
3833 /**
3834 * Provides an alternative text depending on specified gender.
3835 * Usage {{gender:username|masculine|feminine|unknown}}.
3836 * username is optional, in which case the gender of current user is used,
3837 * but only in (some) interface messages; otherwise default gender is used.
3838 *
3839 * If no forms are given, an empty string is returned. If only one form is
3840 * given, it will be returned unconditionally. These details are implied by
3841 * the caller and cannot be overridden in subclasses.
3842 *
3843 * If three forms are given, the default is to use the third (unknown) form.
3844 * If fewer than three forms are given, the default is to use the first (masculine) form.
3845 * These details can be overridden in subclasses.
3846 *
3847 * @param string $gender
3848 * @param array $forms
3849 *
3850 * @return string
3851 */
3852 function gender( $gender, $forms ) {
3853 if ( !count( $forms ) ) {
3854 return '';
3855 }
3856 $forms = $this->preConvertPlural( $forms, 2 );
3857 if ( $gender === 'male' ) {
3858 return $forms[0];
3859 }
3860 if ( $gender === 'female' ) {
3861 return $forms[1];
3862 }
3863 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3864 }
3865
3866 /**
3867 * Plural form transformations, needed for some languages.
3868 * For example, there are 3 form of plural in Russian and Polish,
3869 * depending on "count mod 10". See [[w:Plural]]
3870 * For English it is pretty simple.
3871 *
3872 * Invoked by putting {{plural:count|wordform1|wordform2}}
3873 * or {{plural:count|wordform1|wordform2|wordform3}}
3874 *
3875 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3876 *
3877 * @param int $count Non-localized number
3878 * @param array $forms Different plural forms
3879 * @return string Correct form of plural for $count in this language
3880 */
3881 function convertPlural( $count, $forms ) {
3882 // Handle explicit n=pluralform cases
3883 $forms = $this->handleExplicitPluralForms( $count, $forms );
3884 if ( is_string( $forms ) ) {
3885 return $forms;
3886 }
3887 if ( !count( $forms ) ) {
3888 return '';
3889 }
3890
3891 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3892 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3893 return $forms[$pluralForm];
3894 }
3895
3896 /**
3897 * Handles explicit plural forms for Language::convertPlural()
3898 *
3899 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3900 * If an explicitly defined plural form matches the $count, then
3901 * string value returned, otherwise array returned for further consideration
3902 * by CLDR rules or overridden convertPlural().
3903 *
3904 * @since 1.23
3905 *
3906 * @param int $count Non-localized number
3907 * @param array $forms Different plural forms
3908 *
3909 * @return array|string
3910 */
3911 protected function handleExplicitPluralForms( $count, array $forms ) {
3912 foreach ( $forms as $index => $form ) {
3913 if ( preg_match( '/\d+=/i', $form ) ) {
3914 $pos = strpos( $form, '=' );
3915 if ( substr( $form, 0, $pos ) === (string)$count ) {
3916 return substr( $form, $pos + 1 );
3917 }
3918 unset( $forms[$index] );
3919 }
3920 }
3921 return array_values( $forms );
3922 }
3923
3924 /**
3925 * Checks that convertPlural was given an array and pads it to requested
3926 * amount of forms by copying the last one.
3927 *
3928 * @param array $forms Array of forms given to convertPlural
3929 * @param int $count How many forms should there be at least
3930 * @return array Padded array of forms or an exception if not an array
3931 */
3932 protected function preConvertPlural( /* Array */ $forms, $count ) {
3933 while ( count( $forms ) < $count ) {
3934 $forms[] = $forms[count( $forms ) - 1];
3935 }
3936 return $forms;
3937 }
3938
3939 /**
3940 * Wraps argument with unicode control characters for directionality safety
3941 *
3942 * This solves the problem where directionality-neutral characters at the edge of
3943 * the argument string get interpreted with the wrong directionality from the
3944 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3945 *
3946 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3947 * directionality of the argument string, using the BIDI algorithm's own "First
3948 * strong directional codepoint" rule. Essentially, this works round the fact that
3949 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3950 * direction inference). The latter is cleaner but still not widely supported.
3951 *
3952 * @param string $text Text to wrap
3953 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3954 */
3955 public function embedBidi( $text = '' ) {
3956 $dir = Language::strongDirFromContent( $text );
3957 if ( $dir === 'ltr' ) {
3958 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3959 return self::$lre . $text . self::$pdf;
3960 }
3961 if ( $dir === 'rtl' ) {
3962 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3963 return self::$rle . $text . self::$pdf;
3964 }
3965 // No strong directionality: do not wrap
3966 return $text;
3967 }
3968
3969 /**
3970 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3971 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3972 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3973 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3974 * match up with it.
3975 *
3976 * @param string $str The validated block duration in English
3977 * @param User $user User object to use timezone from or null for $wgUser
3978 * @return string Somehow translated block duration
3979 * @see LanguageFi.php for example implementation
3980 */
3981 function translateBlockExpiry( $str, User $user = null ) {
3982 $duration = SpecialBlock::getSuggestedDurations( $this );
3983 foreach ( $duration as $show => $value ) {
3984 if ( strcmp( $str, $value ) == 0 ) {
3985 return htmlspecialchars( trim( $show ) );
3986 }
3987 }
3988
3989 if ( wfIsInfinity( $str ) ) {
3990 foreach ( $duration as $show => $value ) {
3991 if ( wfIsInfinity( $value ) ) {
3992 return htmlspecialchars( trim( $show ) );
3993 }
3994 }
3995 }
3996
3997 // If all else fails, return a standard duration or timestamp description.
3998 $time = strtotime( $str, 0 );
3999 if ( $time === false ) { // Unknown format. Return it as-is in case.
4000 return $str;
4001 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4002 // $time is relative to 0 so it's a duration length.
4003 return $this->formatDuration( $time );
4004 } else { // It's an absolute timestamp.
4005 if ( $time === 0 ) {
4006 // wfTimestamp() handles 0 as current time instead of epoch.
4007 $time = '19700101000000';
4008 }
4009 if ( $user ) {
4010 return $this->userTimeAndDate( $time, $user );
4011 }
4012 return $this->timeanddate( $time );
4013 }
4014 }
4015
4016 /**
4017 * languages like Chinese need to be segmented in order for the diff
4018 * to be of any use
4019 *
4020 * @param string $text
4021 * @return string
4022 */
4023 public function segmentForDiff( $text ) {
4024 return $text;
4025 }
4026
4027 /**
4028 * and unsegment to show the result
4029 *
4030 * @param string $text
4031 * @return string
4032 */
4033 public function unsegmentForDiff( $text ) {
4034 return $text;
4035 }
4036
4037 /**
4038 * Return the LanguageConverter used in the Language
4039 *
4040 * @since 1.19
4041 * @return LanguageConverter
4042 */
4043 public function getConverter() {
4044 return $this->mConverter;
4045 }
4046
4047 /**
4048 * convert text to all supported variants
4049 *
4050 * @param string $text
4051 * @return array
4052 */
4053 public function autoConvertToAllVariants( $text ) {
4054 return $this->mConverter->autoConvertToAllVariants( $text );
4055 }
4056
4057 /**
4058 * convert text to different variants of a language.
4059 *
4060 * @param string $text
4061 * @return string
4062 */
4063 public function convert( $text ) {
4064 return $this->mConverter->convert( $text );
4065 }
4066
4067 /**
4068 * Convert a Title object to a string in the preferred variant
4069 *
4070 * @param Title $title
4071 * @return string
4072 */
4073 public function convertTitle( $title ) {
4074 return $this->mConverter->convertTitle( $title );
4075 }
4076
4077 /**
4078 * Convert a namespace index to a string in the preferred variant
4079 *
4080 * @param int $ns
4081 * @return string
4082 */
4083 public function convertNamespace( $ns ) {
4084 return $this->mConverter->convertNamespace( $ns );
4085 }
4086
4087 /**
4088 * Check if this is a language with variants
4089 *
4090 * @return bool
4091 */
4092 public function hasVariants() {
4093 return count( $this->getVariants() ) > 1;
4094 }
4095
4096 /**
4097 * Check if the language has the specific variant
4098 *
4099 * @since 1.19
4100 * @param string $variant
4101 * @return bool
4102 */
4103 public function hasVariant( $variant ) {
4104 return (bool)$this->mConverter->validateVariant( $variant );
4105 }
4106
4107 /**
4108 * Perform output conversion on a string, and encode for safe HTML output.
4109 * @param string $text Text to be converted
4110 * @param bool $isTitle Whether this conversion is for the article title
4111 * @return string
4112 * @todo this should get integrated somewhere sane
4113 */
4114 public function convertHtml( $text, $isTitle = false ) {
4115 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4116 }
4117
4118 /**
4119 * @param string $key
4120 * @return string
4121 */
4122 public function convertCategoryKey( $key ) {
4123 return $this->mConverter->convertCategoryKey( $key );
4124 }
4125
4126 /**
4127 * Get the list of variants supported by this language
4128 * see sample implementation in LanguageZh.php
4129 *
4130 * @return array An array of language codes
4131 */
4132 public function getVariants() {
4133 return $this->mConverter->getVariants();
4134 }
4135
4136 /**
4137 * @return string
4138 */
4139 public function getPreferredVariant() {
4140 return $this->mConverter->getPreferredVariant();
4141 }
4142
4143 /**
4144 * @return string
4145 */
4146 public function getDefaultVariant() {
4147 return $this->mConverter->getDefaultVariant();
4148 }
4149
4150 /**
4151 * @return string
4152 */
4153 public function getURLVariant() {
4154 return $this->mConverter->getURLVariant();
4155 }
4156
4157 /**
4158 * If a language supports multiple variants, it is
4159 * possible that non-existing link in one variant
4160 * actually exists in another variant. this function
4161 * tries to find it. See e.g. LanguageZh.php
4162 * The input parameters may be modified upon return
4163 *
4164 * @param string &$link The name of the link
4165 * @param Title &$nt The title object of the link
4166 * @param bool $ignoreOtherCond To disable other conditions when
4167 * we need to transclude a template or update a category's link
4168 */
4169 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4170 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4171 }
4172
4173 /**
4174 * returns language specific options used by User::getPageRenderHash()
4175 * for example, the preferred language variant
4176 *
4177 * @return string
4178 */
4179 function getExtraHashOptions() {
4180 return $this->mConverter->getExtraHashOptions();
4181 }
4182
4183 /**
4184 * For languages that support multiple variants, the title of an
4185 * article may be displayed differently in different variants. this
4186 * function returns the apporiate title defined in the body of the article.
4187 *
4188 * @return string
4189 */
4190 public function getParsedTitle() {
4191 return $this->mConverter->getParsedTitle();
4192 }
4193
4194 /**
4195 * Refresh the cache of conversion tables when
4196 * MediaWiki:Conversiontable* is updated.
4197 *
4198 * @param Title $title The Title of the page being updated
4199 */
4200 public function updateConversionTable( Title $title ) {
4201 $this->mConverter->updateConversionTable( $title );
4202 }
4203
4204 /**
4205 * Prepare external link text for conversion. When the text is
4206 * a URL, it shouldn't be converted, and it'll be wrapped in
4207 * the "raw" tag (-{R| }-) to prevent conversion.
4208 *
4209 * This function is called "markNoConversion" for historical
4210 * reasons.
4211 *
4212 * @param string $text Text to be used for external link
4213 * @param bool $noParse Wrap it without confirming it's a real URL first
4214 * @return string The tagged text
4215 */
4216 public function markNoConversion( $text, $noParse = false ) {
4217 // Excluding protocal-relative URLs may avoid many false positives.
4218 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4219 return $this->mConverter->markNoConversion( $text );
4220 } else {
4221 return $text;
4222 }
4223 }
4224
4225 /**
4226 * A regular expression to match legal word-trailing characters
4227 * which should be merged onto a link of the form [[foo]]bar.
4228 *
4229 * @return string
4230 */
4231 public function linkTrail() {
4232 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4233 }
4234
4235 /**
4236 * A regular expression character set to match legal word-prefixing
4237 * characters which should be merged onto a link of the form foo[[bar]].
4238 *
4239 * @return string
4240 */
4241 public function linkPrefixCharset() {
4242 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4243 }
4244
4245 /**
4246 * Get the "parent" language which has a converter to convert a "compatible" language
4247 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4248 *
4249 * @return Language|null
4250 * @since 1.22
4251 */
4252 public function getParentLanguage() {
4253 if ( $this->mParentLanguage !== false ) {
4254 return $this->mParentLanguage;
4255 }
4256
4257 $code = explode( '-', $this->getCode() )[0];
4258 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4259 $this->mParentLanguage = null;
4260 return null;
4261 }
4262 $lang = Language::factory( $code );
4263 if ( !$lang->hasVariant( $this->getCode() ) ) {
4264 $this->mParentLanguage = null;
4265 return null;
4266 }
4267
4268 $this->mParentLanguage = $lang;
4269 return $lang;
4270 }
4271
4272 /**
4273 * Compare with an other language object
4274 *
4275 * @since 1.28
4276 * @param Language $lang
4277 * @return boolean
4278 */
4279 public function equals( Language $lang ) {
4280 return $lang->getCode() === $this->mCode;
4281 }
4282
4283 /**
4284 * Get the internal language code for this language object
4285 *
4286 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4287 * htmlspecialchars() or similar
4288 *
4289 * @return string
4290 */
4291 public function getCode() {
4292 return $this->mCode;
4293 }
4294
4295 /**
4296 * Get the code in BCP 47 format which we can use
4297 * inside of html lang="" tags.
4298 *
4299 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4300 * htmlspecialchars() or similar.
4301 *
4302 * @since 1.19
4303 * @return string
4304 */
4305 public function getHtmlCode() {
4306 if ( is_null( $this->mHtmlCode ) ) {
4307 $this->mHtmlCode = wfBCP47( $this->getCode() );
4308 }
4309 return $this->mHtmlCode;
4310 }
4311
4312 /**
4313 * @param string $code
4314 */
4315 public function setCode( $code ) {
4316 $this->mCode = $code;
4317 // Ensure we don't leave incorrect cached data lying around
4318 $this->mHtmlCode = null;
4319 $this->mParentLanguage = false;
4320 }
4321
4322 /**
4323 * Get the language code from a file name. Inverse of getFileName()
4324 * @param string $filename $prefix . $languageCode . $suffix
4325 * @param string $prefix Prefix before the language code
4326 * @param string $suffix Suffix after the language code
4327 * @return string Language code, or false if $prefix or $suffix isn't found
4328 */
4329 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4330 $m = null;
4331 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4332 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4333 if ( !count( $m ) ) {
4334 return false;
4335 }
4336 return str_replace( '_', '-', strtolower( $m[1] ) );
4337 }
4338
4339 /**
4340 * @param string $code
4341 * @return string Name of the language class
4342 */
4343 public static function classFromCode( $code ) {
4344 if ( $code == 'en' ) {
4345 return 'Language';
4346 } else {
4347 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4348 }
4349 }
4350
4351 /**
4352 * Get the name of a file for a certain language code
4353 * @param string $prefix Prepend this to the filename
4354 * @param string $code Language code
4355 * @param string $suffix Append this to the filename
4356 * @throws MWException
4357 * @return string $prefix . $mangledCode . $suffix
4358 */
4359 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4360 if ( !self::isValidBuiltInCode( $code ) ) {
4361 throw new MWException( "Invalid language code \"$code\"" );
4362 }
4363
4364 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4365 }
4366
4367 /**
4368 * @param string $code
4369 * @return string
4370 */
4371 public static function getMessagesFileName( $code ) {
4372 global $IP;
4373 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4374 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4375 return $file;
4376 }
4377
4378 /**
4379 * @param string $code
4380 * @return string
4381 * @throws MWException
4382 * @since 1.23
4383 */
4384 public static function getJsonMessagesFileName( $code ) {
4385 global $IP;
4386
4387 if ( !self::isValidBuiltInCode( $code ) ) {
4388 throw new MWException( "Invalid language code \"$code\"" );
4389 }
4390
4391 return "$IP/languages/i18n/$code.json";
4392 }
4393
4394 /**
4395 * Get the first fallback for a given language.
4396 *
4397 * @param string $code
4398 *
4399 * @return bool|string
4400 */
4401 public static function getFallbackFor( $code ) {
4402 $fallbacks = self::getFallbacksFor( $code );
4403 if ( $fallbacks ) {
4404 return $fallbacks[0];
4405 }
4406 return false;
4407 }
4408
4409 /**
4410 * Get the ordered list of fallback languages.
4411 *
4412 * @since 1.19
4413 * @param string $code Language code
4414 * @return array Non-empty array, ending in "en"
4415 */
4416 public static function getFallbacksFor( $code ) {
4417 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4418 return [];
4419 }
4420 // For unknown languages, fallbackSequence returns an empty array,
4421 // hardcode fallback to 'en' in that case.
4422 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4423 }
4424
4425 /**
4426 * Get the ordered list of fallback languages, ending with the fallback
4427 * language chain for the site language.
4428 *
4429 * @since 1.22
4430 * @param string $code Language code
4431 * @return array Array( fallbacks, site fallbacks )
4432 */
4433 public static function getFallbacksIncludingSiteLanguage( $code ) {
4434 global $wgLanguageCode;
4435
4436 // Usually, we will only store a tiny number of fallback chains, so we
4437 // keep them in static memory.
4438 $cacheKey = "{$code}-{$wgLanguageCode}";
4439
4440 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4441 $fallbacks = self::getFallbacksFor( $code );
4442
4443 // Append the site's fallback chain, including the site language itself
4444 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4445 array_unshift( $siteFallbacks, $wgLanguageCode );
4446
4447 // Eliminate any languages already included in the chain
4448 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4449
4450 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4451 }
4452 return self::$fallbackLanguageCache[$cacheKey];
4453 }
4454
4455 /**
4456 * Get all messages for a given language
4457 * WARNING: this may take a long time. If you just need all message *keys*
4458 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4459 *
4460 * @param string $code
4461 *
4462 * @return array
4463 */
4464 public static function getMessagesFor( $code ) {
4465 return self::getLocalisationCache()->getItem( $code, 'messages' );
4466 }
4467
4468 /**
4469 * Get a message for a given language
4470 *
4471 * @param string $key
4472 * @param string $code
4473 *
4474 * @return string
4475 */
4476 public static function getMessageFor( $key, $code ) {
4477 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4478 }
4479
4480 /**
4481 * Get all message keys for a given language. This is a faster alternative to
4482 * array_keys( Language::getMessagesFor( $code ) )
4483 *
4484 * @since 1.19
4485 * @param string $code Language code
4486 * @return array Array of message keys (strings)
4487 */
4488 public static function getMessageKeysFor( $code ) {
4489 return self::getLocalisationCache()->getSubitemList( $code, 'messages' );
4490 }
4491
4492 /**
4493 * @param string $talk
4494 * @return mixed
4495 */
4496 function fixVariableInNamespace( $talk ) {
4497 if ( strpos( $talk, '$1' ) === false ) {
4498 return $talk;
4499 }
4500
4501 global $wgMetaNamespace;
4502 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4503
4504 # Allow grammar transformations
4505 # Allowing full message-style parsing would make simple requests
4506 # such as action=raw much more expensive than they need to be.
4507 # This will hopefully cover most cases.
4508 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4509 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4510 return str_replace( ' ', '_', $talk );
4511 }
4512
4513 /**
4514 * @param string $m
4515 * @return string
4516 */
4517 function replaceGrammarInNamespace( $m ) {
4518 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4519 }
4520
4521 /**
4522 * Decode an expiry (block, protection, etc) which has come from the DB
4523 *
4524 * @param string $expiry Database expiry String
4525 * @param bool|int $format True to process using language functions, or TS_ constant
4526 * to return the expiry in a given timestamp
4527 * @param string $infinity If $format is not true, use this string for infinite expiry
4528 * @return string
4529 * @since 1.18
4530 */
4531 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4532 static $dbInfinity;
4533 if ( $dbInfinity === null ) {
4534 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4535 }
4536
4537 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4538 return $format === true
4539 ? $this->getMessageFromDB( 'infiniteblock' )
4540 : $infinity;
4541 } else {
4542 return $format === true
4543 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4544 : wfTimestamp( $format, $expiry );
4545 }
4546 }
4547
4548 /**
4549 * Formats a time given in seconds into a string representation of that time.
4550 *
4551 * @param int|float $seconds
4552 * @param array $format An optional argument that formats the returned string in different ways:
4553 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4554 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4555 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4556 * and friends.
4557 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4558 * or 'avoidminutes'.
4559 * @return string
4560 */
4561 function formatTimePeriod( $seconds, $format = [] ) {
4562 if ( !is_array( $format ) ) {
4563 $format = [ 'avoid' => $format ]; // For backwards compatibility
4564 }
4565 if ( !isset( $format['avoid'] ) ) {
4566 $format['avoid'] = false;
4567 }
4568 if ( !isset( $format['noabbrevs'] ) ) {
4569 $format['noabbrevs'] = false;
4570 }
4571 $secondsMsg = wfMessage(
4572 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4573 $minutesMsg = wfMessage(
4574 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4575 $hoursMsg = wfMessage(
4576 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4577 $daysMsg = wfMessage(
4578 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4579
4580 if ( round( $seconds * 10 ) < 100 ) {
4581 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4582 $s = $secondsMsg->params( $s )->text();
4583 } elseif ( round( $seconds ) < 60 ) {
4584 $s = $this->formatNum( round( $seconds ) );
4585 $s = $secondsMsg->params( $s )->text();
4586 } elseif ( round( $seconds ) < 3600 ) {
4587 $minutes = floor( $seconds / 60 );
4588 $secondsPart = round( fmod( $seconds, 60 ) );
4589 if ( $secondsPart == 60 ) {
4590 $secondsPart = 0;
4591 $minutes++;
4592 }
4593 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4594 $s .= ' ';
4595 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4596 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4597 $hours = floor( $seconds / 3600 );
4598 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4599 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4600 if ( $secondsPart == 60 ) {
4601 $secondsPart = 0;
4602 $minutes++;
4603 }
4604 if ( $minutes == 60 ) {
4605 $minutes = 0;
4606 $hours++;
4607 }
4608 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4609 $s .= ' ';
4610 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4611 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4612 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4613 }
4614 } else {
4615 $days = floor( $seconds / 86400 );
4616 if ( $format['avoid'] === 'avoidminutes' ) {
4617 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4618 if ( $hours == 24 ) {
4619 $hours = 0;
4620 $days++;
4621 }
4622 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4623 $s .= ' ';
4624 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4625 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4626 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4627 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4628 if ( $minutes == 60 ) {
4629 $minutes = 0;
4630 $hours++;
4631 }
4632 if ( $hours == 24 ) {
4633 $hours = 0;
4634 $days++;
4635 }
4636 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4637 $s .= ' ';
4638 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4639 $s .= ' ';
4640 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4641 } else {
4642 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4643 $s .= ' ';
4644 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4645 }
4646 }
4647 return $s;
4648 }
4649
4650 /**
4651 * Format a bitrate for output, using an appropriate
4652 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4653 * the magnitude in question.
4654 *
4655 * This use base 1000. For base 1024 use formatSize(), for another base
4656 * see formatComputingNumbers().
4657 *
4658 * @param int $bps
4659 * @return string
4660 */
4661 function formatBitrate( $bps ) {
4662 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4663 }
4664
4665 /**
4666 * @param int $size Size of the unit
4667 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4668 * @param string $messageKey Message key to be uesd
4669 * @return string
4670 */
4671 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4672 if ( $size <= 0 ) {
4673 return str_replace( '$1', $this->formatNum( $size ),
4674 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4675 );
4676 }
4677 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4678 $index = 0;
4679
4680 $maxIndex = count( $sizes ) - 1;
4681 while ( $size >= $boundary && $index < $maxIndex ) {
4682 $index++;
4683 $size /= $boundary;
4684 }
4685
4686 // For small sizes no decimal places necessary
4687 $round = 0;
4688 if ( $index > 1 ) {
4689 // For MB and bigger two decimal places are smarter
4690 $round = 2;
4691 }
4692 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4693
4694 $size = round( $size, $round );
4695 $text = $this->getMessageFromDB( $msg );
4696 return str_replace( '$1', $this->formatNum( $size ), $text );
4697 }
4698
4699 /**
4700 * Format a size in bytes for output, using an appropriate
4701 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4702 *
4703 * This method use base 1024. For base 1000 use formatBitrate(), for
4704 * another base see formatComputingNumbers()
4705 *
4706 * @param int $size Size to format
4707 * @return string Plain text (not HTML)
4708 */
4709 function formatSize( $size ) {
4710 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4711 }
4712
4713 /**
4714 * Make a list item, used by various special pages
4715 *
4716 * @param string $page Page link
4717 * @param string $details HTML safe text between brackets
4718 * @param bool $oppositedm Add the direction mark opposite to your
4719 * language, to display text properly
4720 * @return HTML escaped string
4721 */
4722 function specialList( $page, $details, $oppositedm = true ) {
4723 if ( !$details ) {
4724 return $page;
4725 }
4726
4727 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4728 return
4729 $page .
4730 $dirmark .
4731 $this->msg( 'word-separator' )->escaped() .
4732 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4733 }
4734
4735 /**
4736 * Generate (prev x| next x) (20|50|100...) type links for paging
4737 *
4738 * @param Title $title Title object to link
4739 * @param int $offset
4740 * @param int $limit
4741 * @param array $query Optional URL query parameter string
4742 * @param bool $atend Optional param for specified if this is the last page
4743 * @return string
4744 */
4745 public function viewPrevNext( Title $title, $offset, $limit,
4746 array $query = [], $atend = false
4747 ) {
4748 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4749
4750 # Make 'previous' link
4751 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4752 if ( $offset > 0 ) {
4753 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4754 $query, $prev, 'prevn-title', 'mw-prevlink' );
4755 } else {
4756 $plink = htmlspecialchars( $prev );
4757 }
4758
4759 # Make 'next' link
4760 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4761 if ( $atend ) {
4762 $nlink = htmlspecialchars( $next );
4763 } else {
4764 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4765 $query, $next, 'nextn-title', 'mw-nextlink' );
4766 }
4767
4768 # Make links to set number of items per page
4769 $numLinks = [];
4770 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4771 $numLinks[] = $this->numLink( $title, $offset, $num,
4772 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4773 }
4774
4775 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4776 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4777 }
4778
4779 /**
4780 * Helper function for viewPrevNext() that generates links
4781 *
4782 * @param Title $title Title object to link
4783 * @param int $offset
4784 * @param int $limit
4785 * @param array $query Extra query parameters
4786 * @param string $link Text to use for the link; will be escaped
4787 * @param string $tooltipMsg Name of the message to use as tooltip
4788 * @param string $class Value of the "class" attribute of the link
4789 * @return string HTML fragment
4790 */
4791 private function numLink( Title $title, $offset, $limit, array $query, $link,
4792 $tooltipMsg, $class
4793 ) {
4794 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4795 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4796 ->numParams( $limit )->text();
4797
4798 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4799 'title' => $tooltip, 'class' => $class ], $link );
4800 }
4801
4802 /**
4803 * Get the conversion rule title, if any.
4804 *
4805 * @return string
4806 */
4807 public function getConvRuleTitle() {
4808 return $this->mConverter->getConvRuleTitle();
4809 }
4810
4811 /**
4812 * Get the compiled plural rules for the language
4813 * @since 1.20
4814 * @return array Associative array with plural form, and plural rule as key-value pairs
4815 */
4816 public function getCompiledPluralRules() {
4817 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4818 $fallbacks = Language::getFallbacksFor( $this->mCode );
4819 if ( !$pluralRules ) {
4820 foreach ( $fallbacks as $fallbackCode ) {
4821 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4822 if ( $pluralRules ) {
4823 break;
4824 }
4825 }
4826 }
4827 return $pluralRules;
4828 }
4829
4830 /**
4831 * Get the plural rules for the language
4832 * @since 1.20
4833 * @return array Associative array with plural form number and plural rule as key-value pairs
4834 */
4835 public function getPluralRules() {
4836 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4837 $fallbacks = Language::getFallbacksFor( $this->mCode );
4838 if ( !$pluralRules ) {
4839 foreach ( $fallbacks as $fallbackCode ) {
4840 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4841 if ( $pluralRules ) {
4842 break;
4843 }
4844 }
4845 }
4846 return $pluralRules;
4847 }
4848
4849 /**
4850 * Get the plural rule types for the language
4851 * @since 1.22
4852 * @return array Associative array with plural form number and plural rule type as key-value pairs
4853 */
4854 public function getPluralRuleTypes() {
4855 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4856 $fallbacks = Language::getFallbacksFor( $this->mCode );
4857 if ( !$pluralRuleTypes ) {
4858 foreach ( $fallbacks as $fallbackCode ) {
4859 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4860 if ( $pluralRuleTypes ) {
4861 break;
4862 }
4863 }
4864 }
4865 return $pluralRuleTypes;
4866 }
4867
4868 /**
4869 * Find the index number of the plural rule appropriate for the given number
4870 * @param int $number
4871 * @return int The index number of the plural rule
4872 */
4873 public function getPluralRuleIndexNumber( $number ) {
4874 $pluralRules = $this->getCompiledPluralRules();
4875 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4876 return $form;
4877 }
4878
4879 /**
4880 * Find the plural rule type appropriate for the given number
4881 * For example, if the language is set to Arabic, getPluralType(5) should
4882 * return 'few'.
4883 * @since 1.22
4884 * @param int $number
4885 * @return string The name of the plural rule type, e.g. one, two, few, many
4886 */
4887 public function getPluralRuleType( $number ) {
4888 $index = $this->getPluralRuleIndexNumber( $number );
4889 $pluralRuleTypes = $this->getPluralRuleTypes();
4890 if ( isset( $pluralRuleTypes[$index] ) ) {
4891 return $pluralRuleTypes[$index];
4892 } else {
4893 return 'other';
4894 }
4895 }
4896 }
MediaWiki Core