dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "Don't fallback from uk to ru"
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
5 *
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
10 *
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
15 *
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
20 *
21 * @file
22 * @ingroup Language
23 */
24
25 /**
26 * @defgroup Language Language
27 */
28
29 use CLDRPluralRuleParser\Evaluator;
30
31 /**
32 * Internationalisation code
33 * @ingroup Language
34 */
35 class Language {
36 /**
37 * @var LanguageConverter
38 */
39 public $mConverter;
40
41 public $mVariants, $mCode, $mLoaded = false;
42 public $mMagicExtensions = [], $mMagicHookDone = false;
43 private $mHtmlCode = null, $mParentLanguage = false;
44
45 public $dateFormatStrings = [];
46 public $mExtendedSpecialPageAliases;
47
48 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
49
50 /**
51 * ReplacementArray object caches
52 */
53 public $transformData = [];
54
55 /**
56 * @var LocalisationCache
57 */
58 static public $dataCache;
59
60 static public $mLangObjCache = [];
61
62 static public $mWeekdayMsgs = [
63 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
64 'friday', 'saturday'
65 ];
66
67 static public $mWeekdayAbbrevMsgs = [
68 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
69 ];
70
71 static public $mMonthMsgs = [
72 'january', 'february', 'march', 'april', 'may_long', 'june',
73 'july', 'august', 'september', 'october', 'november',
74 'december'
75 ];
76 static public $mMonthGenMsgs = [
77 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
78 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
79 'december-gen'
80 ];
81 static public $mMonthAbbrevMsgs = [
82 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
83 'sep', 'oct', 'nov', 'dec'
84 ];
85
86 static public $mIranianCalendarMonthMsgs = [
87 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
88 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
89 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
90 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
91 ];
92
93 static public $mHebrewCalendarMonthMsgs = [
94 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
95 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
96 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
97 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
98 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
99 ];
100
101 static public $mHebrewCalendarMonthGenMsgs = [
102 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
103 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
104 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
105 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
106 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
107 ];
108
109 static public $mHijriCalendarMonthMsgs = [
110 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
111 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
112 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
113 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
114 ];
115
116 /**
117 * @since 1.20
118 * @var array
119 */
120 static public $durationIntervals = [
121 'millennia' => 31556952000,
122 'centuries' => 3155695200,
123 'decades' => 315569520,
124 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
125 'weeks' => 604800,
126 'days' => 86400,
127 'hours' => 3600,
128 'minutes' => 60,
129 'seconds' => 1,
130 ];
131
132 /**
133 * Cache for language fallbacks.
134 * @see Language::getFallbacksIncludingSiteLanguage
135 * @since 1.21
136 * @var array
137 */
138 static private $fallbackLanguageCache = [];
139
140 /**
141 * Cache for grammar rules data
142 * @var MapCacheLRU|null
143 */
144 static private $grammarTransformations;
145
146 /**
147 * Cache for language names
148 * @var HashBagOStuff|null
149 */
150 static private $languageNameCache;
151
152 /**
153 * Unicode directional formatting characters, for embedBidi()
154 */
155 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
156 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
157 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
158
159 /**
160 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
161 * - in group 1 if it is LTR
162 * - in group 2 if it is RTL
163 * Does not match if there is no strong directionality codepoint.
164 *
165 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
166 *
167 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
168 * https://git.wikimedia.org/summary/unicodejs.git .
169 */
170 // @codingStandardsIgnoreStart
171 // @codeCoverageIgnoreStart
172 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';
173 // @codeCoverageIgnoreEnd
174 // @codingStandardsIgnoreEnd
175
176 /**
177 * Get a cached or new language object for a given language code
178 * @param string $code
179 * @return Language
180 */
181 static function factory( $code ) {
182 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
183
184 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
185 $code = $wgDummyLanguageCodes[$code];
186 }
187
188 // get the language object to process
189 $langObj = isset( self::$mLangObjCache[$code] )
190 ? self::$mLangObjCache[$code]
191 : self::newFromCode( $code );
192
193 // merge the language object in to get it up front in the cache
194 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
195 // get rid of the oldest ones in case we have an overflow
196 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
197
198 return $langObj;
199 }
200
201 /**
202 * Create a language object for a given language code
203 * @param string $code
204 * @throws MWException
205 * @return Language
206 */
207 protected static function newFromCode( $code ) {
208 if ( !Language::isValidCode( $code ) ) {
209 throw new MWException( "Invalid language code \"$code\"" );
210 }
211
212 if ( !Language::isValidBuiltInCode( $code ) ) {
213 // It's not possible to customise this code with class files, so
214 // just return a Language object. This is to support uselang= hacks.
215 $lang = new Language;
216 $lang->setCode( $code );
217 return $lang;
218 }
219
220 // Check if there is a language class for the code
221 $class = self::classFromCode( $code );
222 if ( class_exists( $class ) ) {
223 $lang = new $class;
224 return $lang;
225 }
226
227 // Keep trying the fallback list until we find an existing class
228 $fallbacks = Language::getFallbacksFor( $code );
229 foreach ( $fallbacks as $fallbackCode ) {
230 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
231 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
232 }
233
234 $class = self::classFromCode( $fallbackCode );
235 if ( class_exists( $class ) ) {
236 $lang = new $class;
237 $lang->setCode( $code );
238 return $lang;
239 }
240 }
241
242 throw new MWException( "Invalid fallback sequence for language '$code'" );
243 }
244
245 /**
246 * Checks whether any localisation is available for that language tag
247 * in MediaWiki (MessagesXx.php exists).
248 *
249 * @param string $code Language tag (in lower case)
250 * @return bool Whether language is supported
251 * @since 1.21
252 */
253 public static function isSupportedLanguage( $code ) {
254 if ( !self::isValidBuiltInCode( $code ) ) {
255 return false;
256 }
257
258 if ( $code === 'qqq' ) {
259 return false;
260 }
261
262 return is_readable( self::getMessagesFileName( $code ) ) ||
263 is_readable( self::getJsonMessagesFileName( $code ) );
264 }
265
266 /**
267 * Returns true if a language code string is a well-formed language tag
268 * according to RFC 5646.
269 * This function only checks well-formedness; it doesn't check that
270 * language, script or variant codes actually exist in the repositories.
271 *
272 * Based on regexes by Mark Davis of the Unicode Consortium:
273 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
274 *
275 * @param string $code
276 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
277 *
278 * @return bool
279 * @since 1.21
280 */
281 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
282 $alpha = '[a-z]';
283 $digit = '[0-9]';
284 $alphanum = '[a-z0-9]';
285 $x = 'x'; # private use singleton
286 $singleton = '[a-wy-z]'; # other singleton
287 $s = $lenient ? '[-_]' : '-';
288
289 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
290 $script = "$alpha{4}"; # ISO 15924
291 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
292 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
293 $extension = "$singleton(?:$s$alphanum{2,8})+";
294 $privateUse = "$x(?:$s$alphanum{1,8})+";
295
296 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
297 # Since these are limited, this is safe even later changes to the registry --
298 # the only oddity is that it might change the type of the tag, and thus
299 # the results from the capturing groups.
300 # https://www.iana.org/assignments/language-subtag-registry
301
302 $grandfathered = "en{$s}GB{$s}oed"
303 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
304 . "|no{$s}(?:bok|nyn)"
305 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
306 . "|zh{$s}min{$s}nan";
307
308 $variantList = "$variant(?:$s$variant)*";
309 $extensionList = "$extension(?:$s$extension)*";
310
311 $langtag = "(?:($language)"
312 . "(?:$s$script)?"
313 . "(?:$s$region)?"
314 . "(?:$s$variantList)?"
315 . "(?:$s$extensionList)?"
316 . "(?:$s$privateUse)?)";
317
318 # The final breakdown, with capturing groups for each of these components
319 # The variants, extensions, grandfathered, and private-use may have interior '-'
320
321 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
322
323 return (bool)preg_match( "/$root/", strtolower( $code ) );
324 }
325
326 /**
327 * Returns true if a language code string is of a valid form, whether or
328 * not it exists. This includes codes which are used solely for
329 * customisation via the MediaWiki namespace.
330 *
331 * @param string $code
332 *
333 * @return bool
334 */
335 public static function isValidCode( $code ) {
336 static $cache = [];
337 if ( !isset( $cache[$code] ) ) {
338 // People think language codes are html safe, so enforce it.
339 // Ideally we should only allow a-zA-Z0-9-
340 // but, .+ and other chars are often used for {{int:}} hacks
341 // see bugs T39564, T39587, T38938
342 $cache[$code] =
343 // Protect against path traversal
344 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
345 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
346 }
347 return $cache[$code];
348 }
349
350 /**
351 * Returns true if a language code is of a valid form for the purposes of
352 * internal customisation of MediaWiki, via Messages*.php or *.json.
353 *
354 * @param string $code
355 *
356 * @throws MWException
357 * @since 1.18
358 * @return bool
359 */
360 public static function isValidBuiltInCode( $code ) {
361
362 if ( !is_string( $code ) ) {
363 if ( is_object( $code ) ) {
364 $addmsg = " of class " . get_class( $code );
365 } else {
366 $addmsg = '';
367 }
368 $type = gettype( $code );
369 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
370 }
371
372 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
373 }
374
375 /**
376 * Returns true if a language code is an IETF tag known to MediaWiki.
377 *
378 * @param string $tag
379 *
380 * @since 1.21
381 * @return bool
382 */
383 public static function isKnownLanguageTag( $tag ) {
384 // Quick escape for invalid input to avoid exceptions down the line
385 // when code tries to process tags which are not valid at all.
386 if ( !self::isValidBuiltInCode( $tag ) ) {
387 return false;
388 }
389
390 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
391 || self::fetchLanguageName( $tag, $tag ) !== ''
392 ) {
393 return true;
394 }
395
396 return false;
397 }
398
399 /**
400 * Get the LocalisationCache instance
401 *
402 * @return LocalisationCache
403 */
404 public static function getLocalisationCache() {
405 if ( is_null( self::$dataCache ) ) {
406 global $wgLocalisationCacheConf;
407 $class = $wgLocalisationCacheConf['class'];
408 self::$dataCache = new $class( $wgLocalisationCacheConf );
409 }
410 return self::$dataCache;
411 }
412
413 function __construct() {
414 $this->mConverter = new FakeConverter( $this );
415 // Set the code to the name of the descendant
416 if ( get_class( $this ) == 'Language' ) {
417 $this->mCode = 'en';
418 } else {
419 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
420 }
421 self::getLocalisationCache();
422 }
423
424 /**
425 * Reduce memory usage
426 */
427 function __destruct() {
428 foreach ( $this as $name => $value ) {
429 unset( $this->$name );
430 }
431 }
432
433 /**
434 * Hook which will be called if this is the content language.
435 * Descendants can use this to register hook functions or modify globals
436 */
437 function initContLang() {
438 }
439
440 /**
441 * @return array
442 * @since 1.19
443 */
444 public function getFallbackLanguages() {
445 return self::getFallbacksFor( $this->mCode );
446 }
447
448 /**
449 * Exports $wgBookstoreListEn
450 * @return array
451 */
452 public function getBookstoreList() {
453 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
454 }
455
456 /**
457 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
458 * available in localised form, it will be included in English.
459 *
460 * @return array
461 */
462 public function getNamespaces() {
463 if ( is_null( $this->namespaceNames ) ) {
464 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
465
466 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
467 $validNamespaces = MWNamespace::getCanonicalNamespaces();
468
469 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
470
471 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
472 if ( $wgMetaNamespaceTalk ) {
473 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
474 } else {
475 $talk = $this->namespaceNames[NS_PROJECT_TALK];
476 $this->namespaceNames[NS_PROJECT_TALK] =
477 $this->fixVariableInNamespace( $talk );
478 }
479
480 # Sometimes a language will be localised but not actually exist on this wiki.
481 foreach ( $this->namespaceNames as $key => $text ) {
482 if ( !isset( $validNamespaces[$key] ) ) {
483 unset( $this->namespaceNames[$key] );
484 }
485 }
486
487 # The above mixing may leave namespaces out of canonical order.
488 # Re-order by namespace ID number...
489 ksort( $this->namespaceNames );
490
491 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
492 }
493
494 return $this->namespaceNames;
495 }
496
497 /**
498 * Arbitrarily set all of the namespace names at once. Mainly used for testing
499 * @param array $namespaces Array of namespaces (id => name)
500 */
501 public function setNamespaces( array $namespaces ) {
502 $this->namespaceNames = $namespaces;
503 $this->mNamespaceIds = null;
504 }
505
506 /**
507 * Resets all of the namespace caches. Mainly used for testing
508 */
509 public function resetNamespaces() {
510 $this->namespaceNames = null;
511 $this->mNamespaceIds = null;
512 $this->namespaceAliases = null;
513 }
514
515 /**
516 * A convenience function that returns getNamespaces() with spaces instead of underscores
517 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
518 *
519 * @return array
520 */
521 public function getFormattedNamespaces() {
522 $ns = $this->getNamespaces();
523 foreach ( $ns as $k => $v ) {
524 $ns[$k] = strtr( $v, '_', ' ' );
525 }
526 return $ns;
527 }
528
529 /**
530 * Get a namespace value by key
531 *
532 * <code>
533 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
534 * echo $mw_ns; // prints 'MediaWiki'
535 * </code>
536 *
537 * @param int $index The array key of the namespace to return
538 * @return string|bool String if the namespace value exists, otherwise false
539 */
540 public function getNsText( $index ) {
541 $ns = $this->getNamespaces();
542 return isset( $ns[$index] ) ? $ns[$index] : false;
543 }
544
545 /**
546 * A convenience function that returns the same thing as
547 * getNsText() except with '_' changed to ' ', useful for
548 * producing output.
549 *
550 * <code>
551 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
552 * echo $mw_ns; // prints 'MediaWiki talk'
553 * </code>
554 *
555 * @param int $index The array key of the namespace to return
556 * @return string Namespace name without underscores (empty string if namespace does not exist)
557 */
558 public function getFormattedNsText( $index ) {
559 $ns = $this->getNsText( $index );
560 return strtr( $ns, '_', ' ' );
561 }
562
563 /**
564 * Returns gender-dependent namespace alias if available.
565 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
566 * @param int $index Namespace index
567 * @param string $gender Gender key (male, female... )
568 * @return string
569 * @since 1.18
570 */
571 public function getGenderNsText( $index, $gender ) {
572 global $wgExtraGenderNamespaces;
573
574 $ns = $wgExtraGenderNamespaces +
575 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
576
577 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
578 }
579
580 /**
581 * Whether this language uses gender-dependent namespace aliases.
582 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
583 * @return bool
584 * @since 1.18
585 */
586 public function needsGenderDistinction() {
587 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
588 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
589 // $wgExtraGenderNamespaces overrides everything
590 return true;
591 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
592 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
593 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
594 return false;
595 } else {
596 // Check what is in i18n files
597 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
598 return count( $aliases ) > 0;
599 }
600 }
601
602 /**
603 * Get a namespace key by value, case insensitive.
604 * Only matches namespace names for the current language, not the
605 * canonical ones defined in Namespace.php.
606 *
607 * @param string $text
608 * @return int|bool An integer if $text is a valid value otherwise false
609 */
610 function getLocalNsIndex( $text ) {
611 $lctext = $this->lc( $text );
612 $ids = $this->getNamespaceIds();
613 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
614 }
615
616 /**
617 * @return array
618 */
619 public function getNamespaceAliases() {
620 if ( is_null( $this->namespaceAliases ) ) {
621 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
622 if ( !$aliases ) {
623 $aliases = [];
624 } else {
625 foreach ( $aliases as $name => $index ) {
626 if ( $index === NS_PROJECT_TALK ) {
627 unset( $aliases[$name] );
628 $name = $this->fixVariableInNamespace( $name );
629 $aliases[$name] = $index;
630 }
631 }
632 }
633
634 global $wgExtraGenderNamespaces;
635 $genders = $wgExtraGenderNamespaces +
636 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
637 foreach ( $genders as $index => $forms ) {
638 foreach ( $forms as $alias ) {
639 $aliases[$alias] = $index;
640 }
641 }
642
643 # Also add converted namespace names as aliases, to avoid confusion.
644 $convertedNames = [];
645 foreach ( $this->getVariants() as $variant ) {
646 if ( $variant === $this->mCode ) {
647 continue;
648 }
649 foreach ( $this->getNamespaces() as $ns => $_ ) {
650 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
651 }
652 }
653
654 $this->namespaceAliases = $aliases + $convertedNames;
655 }
656
657 return $this->namespaceAliases;
658 }
659
660 /**
661 * @return array
662 */
663 public function getNamespaceIds() {
664 if ( is_null( $this->mNamespaceIds ) ) {
665 global $wgNamespaceAliases;
666 # Put namespace names and aliases into a hashtable.
667 # If this is too slow, then we should arrange it so that it is done
668 # before caching. The catch is that at pre-cache time, the above
669 # class-specific fixup hasn't been done.
670 $this->mNamespaceIds = [];
671 foreach ( $this->getNamespaces() as $index => $name ) {
672 $this->mNamespaceIds[$this->lc( $name )] = $index;
673 }
674 foreach ( $this->getNamespaceAliases() as $name => $index ) {
675 $this->mNamespaceIds[$this->lc( $name )] = $index;
676 }
677 if ( $wgNamespaceAliases ) {
678 foreach ( $wgNamespaceAliases as $name => $index ) {
679 $this->mNamespaceIds[$this->lc( $name )] = $index;
680 }
681 }
682 }
683 return $this->mNamespaceIds;
684 }
685
686 /**
687 * Get a namespace key by value, case insensitive. Canonical namespace
688 * names override custom ones defined for the current language.
689 *
690 * @param string $text
691 * @return int|bool An integer if $text is a valid value otherwise false
692 */
693 public function getNsIndex( $text ) {
694 $lctext = $this->lc( $text );
695 $ns = MWNamespace::getCanonicalIndex( $lctext );
696 if ( $ns !== null ) {
697 return $ns;
698 }
699 $ids = $this->getNamespaceIds();
700 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
701 }
702
703 /**
704 * short names for language variants used for language conversion links.
705 *
706 * @param string $code
707 * @param bool $usemsg Use the "variantname-xyz" message if it exists
708 * @return string
709 */
710 public function getVariantname( $code, $usemsg = true ) {
711 $msg = "variantname-$code";
712 if ( $usemsg && wfMessage( $msg )->exists() ) {
713 return $this->getMessageFromDB( $msg );
714 }
715 $name = self::fetchLanguageName( $code );
716 if ( $name ) {
717 return $name; # if it's defined as a language name, show that
718 } else {
719 # otherwise, output the language code
720 return $code;
721 }
722 }
723
724 /**
725 * @return array
726 */
727 public function getDatePreferences() {
728 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
729 }
730
731 /**
732 * @return array
733 */
734 function getDateFormats() {
735 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
736 }
737
738 /**
739 * @return array|string
740 */
741 public function getDefaultDateFormat() {
742 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
743 if ( $df === 'dmy or mdy' ) {
744 global $wgAmericanDates;
745 return $wgAmericanDates ? 'mdy' : 'dmy';
746 } else {
747 return $df;
748 }
749 }
750
751 /**
752 * @return array
753 */
754 public function getDatePreferenceMigrationMap() {
755 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
756 }
757
758 /**
759 * @param string $image
760 * @return array|null
761 */
762 function getImageFile( $image ) {
763 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
764 }
765
766 /**
767 * @return array
768 * @since 1.24
769 */
770 public function getImageFiles() {
771 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
772 }
773
774 /**
775 * @return array
776 */
777 public function getExtraUserToggles() {
778 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
779 }
780
781 /**
782 * @param string $tog
783 * @return string
784 */
785 function getUserToggle( $tog ) {
786 return $this->getMessageFromDB( "tog-$tog" );
787 }
788
789 /**
790 * Get an array of language names, indexed by code.
791 * @param null|string $inLanguage Code of language in which to return the names
792 * Use null for autonyms (native names)
793 * @param string $include One of:
794 * 'all' all available languages
795 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
796 * 'mwfile' only if the language is in 'mw' *and* has a message file
797 * @return array Language code => language name
798 * @since 1.20
799 */
800 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
801 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
802 $cacheKey .= ":$include";
803 if ( self::$languageNameCache === null ) {
804 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
805 }
806
807 $ret = self::$languageNameCache->get( $cacheKey );
808 if ( !$ret ) {
809 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
810 self::$languageNameCache->set( $cacheKey, $ret );
811 }
812 return $ret;
813 }
814
815 /**
816 * Uncached helper for fetchLanguageNames
817 * @param null|string $inLanguage Code of language in which to return the names
818 * Use null for autonyms (native names)
819 * @param string $include One of:
820 * 'all' all available languages
821 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
822 * 'mwfile' only if the language is in 'mw' *and* has a message file
823 * @return array Language code => language name
824 */
825 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
826 global $wgExtraLanguageNames;
827
828 // If passed an invalid language code to use, fallback to en
829 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
830 $inLanguage = 'en';
831 }
832
833 $names = [];
834
835 if ( $inLanguage ) {
836 # TODO: also include when $inLanguage is null, when this code is more efficient
837 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
838 }
839
840 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
841 foreach ( $mwNames as $mwCode => $mwName ) {
842 # - Prefer own MediaWiki native name when not using the hook
843 # - For other names just add if not added through the hook
844 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
845 $names[$mwCode] = $mwName;
846 }
847 }
848
849 if ( $include === 'all' ) {
850 ksort( $names );
851 return $names;
852 }
853
854 $returnMw = [];
855 $coreCodes = array_keys( $mwNames );
856 foreach ( $coreCodes as $coreCode ) {
857 $returnMw[$coreCode] = $names[$coreCode];
858 }
859
860 if ( $include === 'mwfile' ) {
861 $namesMwFile = [];
862 # We do this using a foreach over the codes instead of a directory
863 # loop so that messages files in extensions will work correctly.
864 foreach ( $returnMw as $code => $value ) {
865 if ( is_readable( self::getMessagesFileName( $code ) )
866 || is_readable( self::getJsonMessagesFileName( $code ) )
867 ) {
868 $namesMwFile[$code] = $names[$code];
869 }
870 }
871
872 ksort( $namesMwFile );
873 return $namesMwFile;
874 }
875
876 ksort( $returnMw );
877 # 'mw' option; default if it's not one of the other two options (all/mwfile)
878 return $returnMw;
879 }
880
881 /**
882 * @param string $code The code of the language for which to get the name
883 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
884 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
885 * @return string Language name or empty
886 * @since 1.20
887 */
888 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
889 $code = strtolower( $code );
890 $array = self::fetchLanguageNames( $inLanguage, $include );
891 return !array_key_exists( $code, $array ) ? '' : $array[$code];
892 }
893
894 /**
895 * Get a message from the MediaWiki namespace.
896 *
897 * @param string $msg Message name
898 * @return string
899 */
900 public function getMessageFromDB( $msg ) {
901 return $this->msg( $msg )->text();
902 }
903
904 /**
905 * Get message object in this language. Only for use inside this class.
906 *
907 * @param string $msg Message name
908 * @return Message
909 */
910 protected function msg( $msg ) {
911 return wfMessage( $msg )->inLanguage( $this );
912 }
913
914 /**
915 * @param string $key
916 * @return string
917 */
918 public function getMonthName( $key ) {
919 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
920 }
921
922 /**
923 * @return array
924 */
925 public function getMonthNamesArray() {
926 $monthNames = [ '' ];
927 for ( $i = 1; $i < 13; $i++ ) {
928 $monthNames[] = $this->getMonthName( $i );
929 }
930 return $monthNames;
931 }
932
933 /**
934 * @param string $key
935 * @return string
936 */
937 public function getMonthNameGen( $key ) {
938 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
939 }
940
941 /**
942 * @param string $key
943 * @return string
944 */
945 public function getMonthAbbreviation( $key ) {
946 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
947 }
948
949 /**
950 * @return array
951 */
952 public function getMonthAbbreviationsArray() {
953 $monthNames = [ '' ];
954 for ( $i = 1; $i < 13; $i++ ) {
955 $monthNames[] = $this->getMonthAbbreviation( $i );
956 }
957 return $monthNames;
958 }
959
960 /**
961 * @param string $key
962 * @return string
963 */
964 public function getWeekdayName( $key ) {
965 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
966 }
967
968 /**
969 * @param string $key
970 * @return string
971 */
972 function getWeekdayAbbreviation( $key ) {
973 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
974 }
975
976 /**
977 * @param string $key
978 * @return string
979 */
980 function getIranianCalendarMonthName( $key ) {
981 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
982 }
983
984 /**
985 * @param string $key
986 * @return string
987 */
988 function getHebrewCalendarMonthName( $key ) {
989 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
990 }
991
992 /**
993 * @param string $key
994 * @return string
995 */
996 function getHebrewCalendarMonthNameGen( $key ) {
997 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
998 }
999
1000 /**
1001 * @param string $key
1002 * @return string
1003 */
1004 function getHijriCalendarMonthName( $key ) {
1005 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1006 }
1007
1008 /**
1009 * Pass through result from $dateTimeObj->format()
1010 * @param DateTime|bool|null &$dateTimeObj
1011 * @param string $ts
1012 * @param DateTimeZone|bool|null $zone
1013 * @param string $code
1014 * @return string
1015 */
1016 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1017 if ( !$dateTimeObj ) {
1018 $dateTimeObj = DateTime::createFromFormat(
1019 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1020 );
1021 }
1022 return $dateTimeObj->format( $code );
1023 }
1024
1025 /**
1026 * This is a workalike of PHP's date() function, but with better
1027 * internationalisation, a reduced set of format characters, and a better
1028 * escaping format.
1029 *
1030 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1031 * the PHP manual for definitions. There are a number of extensions, which
1032 * start with "x":
1033 *
1034 * xn Do not translate digits of the next numeric format character
1035 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1036 * xr Use roman numerals for the next numeric format character
1037 * xh Use hebrew numerals for the next numeric format character
1038 * xx Literal x
1039 * xg Genitive month name
1040 *
1041 * xij j (day number) in Iranian calendar
1042 * xiF F (month name) in Iranian calendar
1043 * xin n (month number) in Iranian calendar
1044 * xiy y (two digit year) in Iranian calendar
1045 * xiY Y (full year) in Iranian calendar
1046 * xit t (days in month) in Iranian calendar
1047 * xiz z (day of the year) in Iranian calendar
1048 *
1049 * xjj j (day number) in Hebrew calendar
1050 * xjF F (month name) in Hebrew calendar
1051 * xjt t (days in month) in Hebrew calendar
1052 * xjx xg (genitive month name) in Hebrew calendar
1053 * xjn n (month number) in Hebrew calendar
1054 * xjY Y (full year) in Hebrew calendar
1055 *
1056 * xmj j (day number) in Hijri calendar
1057 * xmF F (month name) in Hijri calendar
1058 * xmn n (month number) in Hijri calendar
1059 * xmY Y (full year) in Hijri calendar
1060 *
1061 * xkY Y (full year) in Thai solar calendar. Months and days are
1062 * identical to the Gregorian calendar
1063 * xoY Y (full year) in Minguo calendar or Juche year.
1064 * Months and days are identical to the
1065 * Gregorian calendar
1066 * xtY Y (full year) in Japanese nengo. Months and days are
1067 * identical to the Gregorian calendar
1068 *
1069 * Characters enclosed in double quotes will be considered literal (with
1070 * the quotes themselves removed). Unmatched quotes will be considered
1071 * literal quotes. Example:
1072 *
1073 * "The month is" F => The month is January
1074 * i's" => 20'11"
1075 *
1076 * Backslash escaping is also supported.
1077 *
1078 * Input timestamp is assumed to be pre-normalized to the desired local
1079 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1080 * $ts is UTC if $zone is not given.
1081 *
1082 * @param string $format
1083 * @param string $ts 14-character timestamp
1084 * YYYYMMDDHHMMSS
1085 * 01234567890123
1086 * @param DateTimeZone $zone Timezone of $ts
1087 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1088 * Only makes sense if $ts is the current time.
1089 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1090 *
1091 * @throws MWException
1092 * @return string
1093 */
1094 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1095 $s = '';
1096 $raw = false;
1097 $roman = false;
1098 $hebrewNum = false;
1099 $dateTimeObj = false;
1100 $rawToggle = false;
1101 $iranian = false;
1102 $hebrew = false;
1103 $hijri = false;
1104 $thai = false;
1105 $minguo = false;
1106 $tenno = false;
1107
1108 $usedSecond = false;
1109 $usedMinute = false;
1110 $usedHour = false;
1111 $usedAMPM = false;
1112 $usedDay = false;
1113 $usedWeek = false;
1114 $usedMonth = false;
1115 $usedYear = false;
1116 $usedISOYear = false;
1117 $usedIsLeapYear = false;
1118
1119 $usedHebrewMonth = false;
1120 $usedIranianMonth = false;
1121 $usedHijriMonth = false;
1122 $usedHebrewYear = false;
1123 $usedIranianYear = false;
1124 $usedHijriYear = false;
1125 $usedTennoYear = false;
1126
1127 if ( strlen( $ts ) !== 14 ) {
1128 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1129 }
1130
1131 if ( !ctype_digit( $ts ) ) {
1132 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1133 }
1134
1135 $formatLength = strlen( $format );
1136 for ( $p = 0; $p < $formatLength; $p++ ) {
1137 $num = false;
1138 $code = $format[$p];
1139 if ( $code == 'x' && $p < $formatLength - 1 ) {
1140 $code .= $format[++$p];
1141 }
1142
1143 if ( ( $code === 'xi'
1144 || $code === 'xj'
1145 || $code === 'xk'
1146 || $code === 'xm'
1147 || $code === 'xo'
1148 || $code === 'xt' )
1149 && $p < $formatLength - 1 ) {
1150 $code .= $format[++$p];
1151 }
1152
1153 switch ( $code ) {
1154 case 'xx':
1155 $s .= 'x';
1156 break;
1157 case 'xn':
1158 $raw = true;
1159 break;
1160 case 'xN':
1161 $rawToggle = !$rawToggle;
1162 break;
1163 case 'xr':
1164 $roman = true;
1165 break;
1166 case 'xh':
1167 $hebrewNum = true;
1168 break;
1169 case 'xg':
1170 $usedMonth = true;
1171 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1172 break;
1173 case 'xjx':
1174 $usedHebrewMonth = true;
1175 if ( !$hebrew ) {
1176 $hebrew = self::tsToHebrew( $ts );
1177 }
1178 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1179 break;
1180 case 'd':
1181 $usedDay = true;
1182 $num = substr( $ts, 6, 2 );
1183 break;
1184 case 'D':
1185 $usedDay = true;
1186 $s .= $this->getWeekdayAbbreviation(
1187 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1188 );
1189 break;
1190 case 'j':
1191 $usedDay = true;
1192 $num = intval( substr( $ts, 6, 2 ) );
1193 break;
1194 case 'xij':
1195 $usedDay = true;
1196 if ( !$iranian ) {
1197 $iranian = self::tsToIranian( $ts );
1198 }
1199 $num = $iranian[2];
1200 break;
1201 case 'xmj':
1202 $usedDay = true;
1203 if ( !$hijri ) {
1204 $hijri = self::tsToHijri( $ts );
1205 }
1206 $num = $hijri[2];
1207 break;
1208 case 'xjj':
1209 $usedDay = true;
1210 if ( !$hebrew ) {
1211 $hebrew = self::tsToHebrew( $ts );
1212 }
1213 $num = $hebrew[2];
1214 break;
1215 case 'l':
1216 $usedDay = true;
1217 $s .= $this->getWeekdayName(
1218 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1219 );
1220 break;
1221 case 'F':
1222 $usedMonth = true;
1223 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1224 break;
1225 case 'xiF':
1226 $usedIranianMonth = true;
1227 if ( !$iranian ) {
1228 $iranian = self::tsToIranian( $ts );
1229 }
1230 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1231 break;
1232 case 'xmF':
1233 $usedHijriMonth = true;
1234 if ( !$hijri ) {
1235 $hijri = self::tsToHijri( $ts );
1236 }
1237 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1238 break;
1239 case 'xjF':
1240 $usedHebrewMonth = true;
1241 if ( !$hebrew ) {
1242 $hebrew = self::tsToHebrew( $ts );
1243 }
1244 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1245 break;
1246 case 'm':
1247 $usedMonth = true;
1248 $num = substr( $ts, 4, 2 );
1249 break;
1250 case 'M':
1251 $usedMonth = true;
1252 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1253 break;
1254 case 'n':
1255 $usedMonth = true;
1256 $num = intval( substr( $ts, 4, 2 ) );
1257 break;
1258 case 'xin':
1259 $usedIranianMonth = true;
1260 if ( !$iranian ) {
1261 $iranian = self::tsToIranian( $ts );
1262 }
1263 $num = $iranian[1];
1264 break;
1265 case 'xmn':
1266 $usedHijriMonth = true;
1267 if ( !$hijri ) {
1268 $hijri = self::tsToHijri( $ts );
1269 }
1270 $num = $hijri[1];
1271 break;
1272 case 'xjn':
1273 $usedHebrewMonth = true;
1274 if ( !$hebrew ) {
1275 $hebrew = self::tsToHebrew( $ts );
1276 }
1277 $num = $hebrew[1];
1278 break;
1279 case 'xjt':
1280 $usedHebrewMonth = true;
1281 if ( !$hebrew ) {
1282 $hebrew = self::tsToHebrew( $ts );
1283 }
1284 $num = $hebrew[3];
1285 break;
1286 case 'Y':
1287 $usedYear = true;
1288 $num = substr( $ts, 0, 4 );
1289 break;
1290 case 'xiY':
1291 $usedIranianYear = true;
1292 if ( !$iranian ) {
1293 $iranian = self::tsToIranian( $ts );
1294 }
1295 $num = $iranian[0];
1296 break;
1297 case 'xmY':
1298 $usedHijriYear = true;
1299 if ( !$hijri ) {
1300 $hijri = self::tsToHijri( $ts );
1301 }
1302 $num = $hijri[0];
1303 break;
1304 case 'xjY':
1305 $usedHebrewYear = true;
1306 if ( !$hebrew ) {
1307 $hebrew = self::tsToHebrew( $ts );
1308 }
1309 $num = $hebrew[0];
1310 break;
1311 case 'xkY':
1312 $usedYear = true;
1313 if ( !$thai ) {
1314 $thai = self::tsToYear( $ts, 'thai' );
1315 }
1316 $num = $thai[0];
1317 break;
1318 case 'xoY':
1319 $usedYear = true;
1320 if ( !$minguo ) {
1321 $minguo = self::tsToYear( $ts, 'minguo' );
1322 }
1323 $num = $minguo[0];
1324 break;
1325 case 'xtY':
1326 $usedTennoYear = true;
1327 if ( !$tenno ) {
1328 $tenno = self::tsToYear( $ts, 'tenno' );
1329 }
1330 $num = $tenno[0];
1331 break;
1332 case 'y':
1333 $usedYear = true;
1334 $num = substr( $ts, 2, 2 );
1335 break;
1336 case 'xiy':
1337 $usedIranianYear = true;
1338 if ( !$iranian ) {
1339 $iranian = self::tsToIranian( $ts );
1340 }
1341 $num = substr( $iranian[0], -2 );
1342 break;
1343 case 'xit':
1344 $usedIranianYear = true;
1345 if ( !$iranian ) {
1346 $iranian = self::tsToIranian( $ts );
1347 }
1348 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1349 break;
1350 case 'xiz':
1351 $usedIranianYear = true;
1352 if ( !$iranian ) {
1353 $iranian = self::tsToIranian( $ts );
1354 }
1355 $num = $iranian[3];
1356 break;
1357 case 'a':
1358 $usedAMPM = true;
1359 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1360 break;
1361 case 'A':
1362 $usedAMPM = true;
1363 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1364 break;
1365 case 'g':
1366 $usedHour = true;
1367 $h = substr( $ts, 8, 2 );
1368 $num = $h % 12 ? $h % 12 : 12;
1369 break;
1370 case 'G':
1371 $usedHour = true;
1372 $num = intval( substr( $ts, 8, 2 ) );
1373 break;
1374 case 'h':
1375 $usedHour = true;
1376 $h = substr( $ts, 8, 2 );
1377 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1378 break;
1379 case 'H':
1380 $usedHour = true;
1381 $num = substr( $ts, 8, 2 );
1382 break;
1383 case 'i':
1384 $usedMinute = true;
1385 $num = substr( $ts, 10, 2 );
1386 break;
1387 case 's':
1388 $usedSecond = true;
1389 $num = substr( $ts, 12, 2 );
1390 break;
1391 case 'c':
1392 case 'r':
1393 $usedSecond = true;
1394 // fall through
1395 case 'e':
1396 case 'O':
1397 case 'P':
1398 case 'T':
1399 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1400 break;
1401 case 'w':
1402 case 'N':
1403 case 'z':
1404 $usedDay = true;
1405 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1406 break;
1407 case 'W':
1408 $usedWeek = true;
1409 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1410 break;
1411 case 't':
1412 $usedMonth = true;
1413 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1414 break;
1415 case 'L':
1416 $usedIsLeapYear = true;
1417 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1418 break;
1419 case 'o':
1420 $usedISOYear = true;
1421 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1422 break;
1423 case 'U':
1424 $usedSecond = true;
1425 // fall through
1426 case 'I':
1427 case 'Z':
1428 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1429 break;
1430 case '\\':
1431 # Backslash escaping
1432 if ( $p < $formatLength - 1 ) {
1433 $s .= $format[++$p];
1434 } else {
1435 $s .= '\\';
1436 }
1437 break;
1438 case '"':
1439 # Quoted literal
1440 if ( $p < $formatLength - 1 ) {
1441 $endQuote = strpos( $format, '"', $p + 1 );
1442 if ( $endQuote === false ) {
1443 # No terminating quote, assume literal "
1444 $s .= '"';
1445 } else {
1446 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1447 $p = $endQuote;
1448 }
1449 } else {
1450 # Quote at end of string, assume literal "
1451 $s .= '"';
1452 }
1453 break;
1454 default:
1455 $s .= $format[$p];
1456 }
1457 if ( $num !== false ) {
1458 if ( $rawToggle || $raw ) {
1459 $s .= $num;
1460 $raw = false;
1461 } elseif ( $roman ) {
1462 $s .= Language::romanNumeral( $num );
1463 $roman = false;
1464 } elseif ( $hebrewNum ) {
1465 $s .= self::hebrewNumeral( $num );
1466 $hebrewNum = false;
1467 } else {
1468 $s .= $this->formatNum( $num, true );
1469 }
1470 }
1471 }
1472
1473 if ( $ttl === 'unused' ) {
1474 // No need to calculate the TTL, the caller wont use it anyway.
1475 } elseif ( $usedSecond ) {
1476 $ttl = 1;
1477 } elseif ( $usedMinute ) {
1478 $ttl = 60 - substr( $ts, 12, 2 );
1479 } elseif ( $usedHour ) {
1480 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1481 } elseif ( $usedAMPM ) {
1482 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1483 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1484 } elseif (
1485 $usedDay ||
1486 $usedHebrewMonth ||
1487 $usedIranianMonth ||
1488 $usedHijriMonth ||
1489 $usedHebrewYear ||
1490 $usedIranianYear ||
1491 $usedHijriYear ||
1492 $usedTennoYear
1493 ) {
1494 // @todo Someone who understands the non-Gregorian calendars
1495 // should write proper logic for them so that they don't need purged every day.
1496 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1497 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1498 } else {
1499 $possibleTtls = [];
1500 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1501 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1502 if ( $usedWeek ) {
1503 $possibleTtls[] =
1504 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1505 $timeRemainingInDay;
1506 } elseif ( $usedISOYear ) {
1507 // December 28th falls on the last ISO week of the year, every year.
1508 // The last ISO week of a year can be 52 or 53.
1509 $lastWeekOfISOYear = DateTime::createFromFormat(
1510 'Ymd',
1511 substr( $ts, 0, 4 ) . '1228',
1512 $zone ?: new DateTimeZone( 'UTC' )
1513 )->format( 'W' );
1514 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1515 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1516 $timeRemainingInWeek =
1517 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1518 + $timeRemainingInDay;
1519 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1520 }
1521
1522 if ( $usedMonth ) {
1523 $possibleTtls[] =
1524 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1525 substr( $ts, 6, 2 ) ) * 86400
1526 + $timeRemainingInDay;
1527 } elseif ( $usedYear ) {
1528 $possibleTtls[] =
1529 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1530 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1531 + $timeRemainingInDay;
1532 } elseif ( $usedIsLeapYear ) {
1533 $year = substr( $ts, 0, 4 );
1534 $timeRemainingInYear =
1535 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1536 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1537 + $timeRemainingInDay;
1538 $mod = $year % 4;
1539 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1540 // this isn't a leap year. see when the next one starts
1541 $nextCandidate = $year - $mod + 4;
1542 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1543 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1544 $timeRemainingInYear;
1545 } else {
1546 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1547 $timeRemainingInYear;
1548 }
1549 } else {
1550 // this is a leap year, so the next year isn't
1551 $possibleTtls[] = $timeRemainingInYear;
1552 }
1553 }
1554
1555 if ( $possibleTtls ) {
1556 $ttl = min( $possibleTtls );
1557 }
1558 }
1559
1560 return $s;
1561 }
1562
1563 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1564 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1565
1566 /**
1567 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1568 * Gregorian dates to Iranian dates. Originally written in C, it
1569 * is released under the terms of GNU Lesser General Public
1570 * License. Conversion to PHP was performed by Niklas Laxström.
1571 *
1572 * Link: http://www.farsiweb.info/jalali/jalali.c
1573 *
1574 * @param string $ts
1575 *
1576 * @return int[]
1577 */
1578 private static function tsToIranian( $ts ) {
1579 $gy = substr( $ts, 0, 4 ) -1600;
1580 $gm = substr( $ts, 4, 2 ) -1;
1581 $gd = substr( $ts, 6, 2 ) -1;
1582
1583 # Days passed from the beginning (including leap years)
1584 $gDayNo = 365 * $gy
1585 + floor( ( $gy + 3 ) / 4 )
1586 - floor( ( $gy + 99 ) / 100 )
1587 + floor( ( $gy + 399 ) / 400 );
1588
1589 // Add days of the past months of this year
1590 for ( $i = 0; $i < $gm; $i++ ) {
1591 $gDayNo += self::$GREG_DAYS[$i];
1592 }
1593
1594 // Leap years
1595 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1596 $gDayNo++;
1597 }
1598
1599 // Days passed in current month
1600 $gDayNo += (int)$gd;
1601
1602 $jDayNo = $gDayNo - 79;
1603
1604 $jNp = floor( $jDayNo / 12053 );
1605 $jDayNo %= 12053;
1606
1607 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1608 $jDayNo %= 1461;
1609
1610 if ( $jDayNo >= 366 ) {
1611 $jy += floor( ( $jDayNo - 1 ) / 365 );
1612 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1613 }
1614
1615 $jz = $jDayNo;
1616
1617 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1618 $jDayNo -= self::$IRANIAN_DAYS[$i];
1619 }
1620
1621 $jm = $i + 1;
1622 $jd = $jDayNo + 1;
1623
1624 return [ $jy, $jm, $jd, $jz ];
1625 }
1626
1627 /**
1628 * Converting Gregorian dates to Hijri dates.
1629 *
1630 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1631 *
1632 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1633 *
1634 * @param string $ts
1635 *
1636 * @return int[]
1637 */
1638 private static function tsToHijri( $ts ) {
1639 $year = substr( $ts, 0, 4 );
1640 $month = substr( $ts, 4, 2 );
1641 $day = substr( $ts, 6, 2 );
1642
1643 $zyr = $year;
1644 $zd = $day;
1645 $zm = $month;
1646 $zy = $zyr;
1647
1648 if (
1649 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1650 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1651 ) {
1652 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1653 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1654 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1655 $zd - 32075;
1656 } else {
1657 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1658 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1659 }
1660
1661 $zl = $zjd -1948440 + 10632;
1662 $zn = (int)( ( $zl - 1 ) / 10631 );
1663 $zl = $zl - 10631 * $zn + 354;
1664 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1665 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1666 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1667 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1668 $zm = (int)( ( 24 * $zl ) / 709 );
1669 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1670 $zy = 30 * $zn + $zj - 30;
1671
1672 return [ $zy, $zm, $zd ];
1673 }
1674
1675 /**
1676 * Converting Gregorian dates to Hebrew dates.
1677 *
1678 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1679 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1680 * to translate the relevant functions into PHP and release them under
1681 * GNU GPL.
1682 *
1683 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1684 * and Adar II is 14. In a non-leap year, Adar is 6.
1685 *
1686 * @param string $ts
1687 *
1688 * @return int[]
1689 */
1690 private static function tsToHebrew( $ts ) {
1691 # Parse date
1692 $year = substr( $ts, 0, 4 );
1693 $month = substr( $ts, 4, 2 );
1694 $day = substr( $ts, 6, 2 );
1695
1696 # Calculate Hebrew year
1697 $hebrewYear = $year + 3760;
1698
1699 # Month number when September = 1, August = 12
1700 $month += 4;
1701 if ( $month > 12 ) {
1702 # Next year
1703 $month -= 12;
1704 $year++;
1705 $hebrewYear++;
1706 }
1707
1708 # Calculate day of year from 1 September
1709 $dayOfYear = $day;
1710 for ( $i = 1; $i < $month; $i++ ) {
1711 if ( $i == 6 ) {
1712 # February
1713 $dayOfYear += 28;
1714 # Check if the year is leap
1715 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1716 $dayOfYear++;
1717 }
1718 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1719 $dayOfYear += 30;
1720 } else {
1721 $dayOfYear += 31;
1722 }
1723 }
1724
1725 # Calculate the start of the Hebrew year
1726 $start = self::hebrewYearStart( $hebrewYear );
1727
1728 # Calculate next year's start
1729 if ( $dayOfYear <= $start ) {
1730 # Day is before the start of the year - it is the previous year
1731 # Next year's start
1732 $nextStart = $start;
1733 # Previous year
1734 $year--;
1735 $hebrewYear--;
1736 # Add days since previous year's 1 September
1737 $dayOfYear += 365;
1738 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1739 # Leap year
1740 $dayOfYear++;
1741 }
1742 # Start of the new (previous) year
1743 $start = self::hebrewYearStart( $hebrewYear );
1744 } else {
1745 # Next year's start
1746 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1747 }
1748
1749 # Calculate Hebrew day of year
1750 $hebrewDayOfYear = $dayOfYear - $start;
1751
1752 # Difference between year's days
1753 $diff = $nextStart - $start;
1754 # Add 12 (or 13 for leap years) days to ignore the difference between
1755 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1756 # difference is only about the year type
1757 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1758 $diff += 13;
1759 } else {
1760 $diff += 12;
1761 }
1762
1763 # Check the year pattern, and is leap year
1764 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1765 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1766 # and non-leap years
1767 $yearPattern = $diff % 30;
1768 # Check if leap year
1769 $isLeap = $diff >= 30;
1770
1771 # Calculate day in the month from number of day in the Hebrew year
1772 # Don't check Adar - if the day is not in Adar, we will stop before;
1773 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1774 $hebrewDay = $hebrewDayOfYear;
1775 $hebrewMonth = 1;
1776 $days = 0;
1777 while ( $hebrewMonth <= 12 ) {
1778 # Calculate days in this month
1779 if ( $isLeap && $hebrewMonth == 6 ) {
1780 # Adar in a leap year
1781 if ( $isLeap ) {
1782 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1783 $days = 30;
1784 if ( $hebrewDay <= $days ) {
1785 # Day in Adar I
1786 $hebrewMonth = 13;
1787 } else {
1788 # Subtract the days of Adar I
1789 $hebrewDay -= $days;
1790 # Try Adar II
1791 $days = 29;
1792 if ( $hebrewDay <= $days ) {
1793 # Day in Adar II
1794 $hebrewMonth = 14;
1795 }
1796 }
1797 }
1798 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1799 # Cheshvan in a complete year (otherwise as the rule below)
1800 $days = 30;
1801 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1802 # Kislev in an incomplete year (otherwise as the rule below)
1803 $days = 29;
1804 } else {
1805 # Odd months have 30 days, even have 29
1806 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1807 }
1808 if ( $hebrewDay <= $days ) {
1809 # In the current month
1810 break;
1811 } else {
1812 # Subtract the days of the current month
1813 $hebrewDay -= $days;
1814 # Try in the next month
1815 $hebrewMonth++;
1816 }
1817 }
1818
1819 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1820 }
1821
1822 /**
1823 * This calculates the Hebrew year start, as days since 1 September.
1824 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1825 * Used for Hebrew date.
1826 *
1827 * @param int $year
1828 *
1829 * @return string
1830 */
1831 private static function hebrewYearStart( $year ) {
1832 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1833 $b = intval( ( $year - 1 ) % 4 );
1834 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1835 if ( $m < 0 ) {
1836 $m--;
1837 }
1838 $Mar = intval( $m );
1839 if ( $m < 0 ) {
1840 $m++;
1841 }
1842 $m -= $Mar;
1843
1844 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1845 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1846 $Mar++;
1847 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1848 $Mar += 2;
1849 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1850 $Mar++;
1851 }
1852
1853 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1854 return $Mar;
1855 }
1856
1857 /**
1858 * Algorithm to convert Gregorian dates to Thai solar dates,
1859 * Minguo dates or Minguo dates.
1860 *
1861 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1862 * https://en.wikipedia.org/wiki/Minguo_calendar
1863 * https://en.wikipedia.org/wiki/Japanese_era_name
1864 *
1865 * @param string $ts 14-character timestamp
1866 * @param string $cName Calender name
1867 * @return array Converted year, month, day
1868 */
1869 private static function tsToYear( $ts, $cName ) {
1870 $gy = substr( $ts, 0, 4 );
1871 $gm = substr( $ts, 4, 2 );
1872 $gd = substr( $ts, 6, 2 );
1873
1874 if ( !strcmp( $cName, 'thai' ) ) {
1875 # Thai solar dates
1876 # Add 543 years to the Gregorian calendar
1877 # Months and days are identical
1878 $gy_offset = $gy + 543;
1879 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1880 # Minguo dates
1881 # Deduct 1911 years from the Gregorian calendar
1882 # Months and days are identical
1883 $gy_offset = $gy - 1911;
1884 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1885 # Nengō dates up to Meiji period
1886 # Deduct years from the Gregorian calendar
1887 # depending on the nengo periods
1888 # Months and days are identical
1889 if ( ( $gy < 1912 )
1890 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1891 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1892 ) {
1893 # Meiji period
1894 $gy_gannen = $gy - 1868 + 1;
1895 $gy_offset = $gy_gannen;
1896 if ( $gy_gannen == 1 ) {
1897 $gy_offset = '元';
1898 }
1899 $gy_offset = '明治' . $gy_offset;
1900 } elseif (
1901 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1902 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1903 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1904 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1905 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1906 ) {
1907 # Taishō period
1908 $gy_gannen = $gy - 1912 + 1;
1909 $gy_offset = $gy_gannen;
1910 if ( $gy_gannen == 1 ) {
1911 $gy_offset = '元';
1912 }
1913 $gy_offset = '大正' . $gy_offset;
1914 } elseif (
1915 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1916 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1917 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1918 ) {
1919 # Shōwa period
1920 $gy_gannen = $gy - 1926 + 1;
1921 $gy_offset = $gy_gannen;
1922 if ( $gy_gannen == 1 ) {
1923 $gy_offset = '元';
1924 }
1925 $gy_offset = '昭和' . $gy_offset;
1926 } else {
1927 # Heisei period
1928 $gy_gannen = $gy - 1989 + 1;
1929 $gy_offset = $gy_gannen;
1930 if ( $gy_gannen == 1 ) {
1931 $gy_offset = '元';
1932 }
1933 $gy_offset = '平成' . $gy_offset;
1934 }
1935 } else {
1936 $gy_offset = $gy;
1937 }
1938
1939 return [ $gy_offset, $gm, $gd ];
1940 }
1941
1942 /**
1943 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1944 *
1945 * This is the rule the BIDI algorithm uses to determine the directionality of
1946 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1947 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1948 *
1949 * TODO: Does not handle BIDI control characters inside the text.
1950 * TODO: Does not handle unallocated characters.
1951 *
1952 * @param string $text Text to test
1953 * @return null|string Directionality ('ltr' or 'rtl') or null
1954 */
1955 private static function strongDirFromContent( $text = '' ) {
1956 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1957 return null;
1958 }
1959 if ( $matches[1] === '' ) {
1960 return 'rtl';
1961 }
1962 return 'ltr';
1963 }
1964
1965 /**
1966 * Roman number formatting up to 10000
1967 *
1968 * @param int $num
1969 *
1970 * @return string
1971 */
1972 static function romanNumeral( $num ) {
1973 static $table = [
1974 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1975 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1976 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1977 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1978 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1979 ];
1980
1981 $num = intval( $num );
1982 if ( $num > 10000 || $num <= 0 ) {
1983 return $num;
1984 }
1985
1986 $s = '';
1987 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1988 if ( $num >= $pow10 ) {
1989 $s .= $table[$i][(int)floor( $num / $pow10 )];
1990 }
1991 $num = $num % $pow10;
1992 }
1993 return $s;
1994 }
1995
1996 /**
1997 * Hebrew Gematria number formatting up to 9999
1998 *
1999 * @param int $num
2000 *
2001 * @return string
2002 */
2003 static function hebrewNumeral( $num ) {
2004 static $table = [
2005 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2006 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2007 [ '',
2008 [ 'ק' ],
2009 [ 'ר' ],
2010 [ 'ש' ],
2011 [ 'ת' ],
2012 [ 'ת', 'ק' ],
2013 [ 'ת', 'ר' ],
2014 [ 'ת', 'ש' ],
2015 [ 'ת', 'ת' ],
2016 [ 'ת', 'ת', 'ק' ],
2017 [ 'ת', 'ת', 'ר' ],
2018 ],
2019 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2020 ];
2021
2022 $num = intval( $num );
2023 if ( $num > 9999 || $num <= 0 ) {
2024 return $num;
2025 }
2026
2027 // Round thousands have special notations
2028 if ( $num === 1000 ) {
2029 return "א' אלף";
2030 } elseif ( $num % 1000 === 0 ) {
2031 return $table[0][$num / 1000] . "' אלפים";
2032 }
2033
2034 $letters = [];
2035
2036 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2037 if ( $num >= $pow10 ) {
2038 if ( $num === 15 || $num === 16 ) {
2039 $letters[] = $table[0][9];
2040 $letters[] = $table[0][$num - 9];
2041 $num = 0;
2042 } else {
2043 $letters = array_merge(
2044 $letters,
2045 (array)$table[$i][intval( $num / $pow10 )]
2046 );
2047
2048 if ( $pow10 === 1000 ) {
2049 $letters[] = "'";
2050 }
2051 }
2052 }
2053
2054 $num = $num % $pow10;
2055 }
2056
2057 $preTransformLength = count( $letters );
2058 if ( $preTransformLength === 1 ) {
2059 // Add geresh (single quote) to one-letter numbers
2060 $letters[] = "'";
2061 } else {
2062 $lastIndex = $preTransformLength - 1;
2063 $letters[$lastIndex] = str_replace(
2064 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2065 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2066 $letters[$lastIndex]
2067 );
2068
2069 // Add gershayim (double quote) to multiple-letter numbers,
2070 // but exclude numbers with only one letter after the thousands
2071 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2072 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2073 $letters[] = "'";
2074 } else {
2075 array_splice( $letters, -1, 0, '"' );
2076 }
2077 }
2078
2079 return implode( $letters );
2080 }
2081
2082 /**
2083 * Used by date() and time() to adjust the time output.
2084 *
2085 * @param string $ts The time in date('YmdHis') format
2086 * @param mixed $tz Adjust the time by this amount (default false, mean we
2087 * get user timecorrection setting)
2088 * @return int
2089 */
2090 public function userAdjust( $ts, $tz = false ) {
2091 global $wgUser, $wgLocalTZoffset;
2092
2093 if ( $tz === false ) {
2094 $tz = $wgUser->getOption( 'timecorrection' );
2095 }
2096
2097 $data = explode( '|', $tz, 3 );
2098
2099 if ( $data[0] == 'ZoneInfo' ) {
2100 MediaWiki\suppressWarnings();
2101 $userTZ = timezone_open( $data[2] );
2102 MediaWiki\restoreWarnings();
2103 if ( $userTZ !== false ) {
2104 $date = date_create( $ts, timezone_open( 'UTC' ) );
2105 date_timezone_set( $date, $userTZ );
2106 $date = date_format( $date, 'YmdHis' );
2107 return $date;
2108 }
2109 # Unrecognized timezone, default to 'Offset' with the stored offset.
2110 $data[0] = 'Offset';
2111 }
2112
2113 if ( $data[0] == 'System' || $tz == '' ) {
2114 # Global offset in minutes.
2115 $minDiff = $wgLocalTZoffset;
2116 } elseif ( $data[0] == 'Offset' ) {
2117 $minDiff = intval( $data[1] );
2118 } else {
2119 $data = explode( ':', $tz );
2120 if ( count( $data ) == 2 ) {
2121 $data[0] = intval( $data[0] );
2122 $data[1] = intval( $data[1] );
2123 $minDiff = abs( $data[0] ) * 60 + $data[1];
2124 if ( $data[0] < 0 ) {
2125 $minDiff = -$minDiff;
2126 }
2127 } else {
2128 $minDiff = intval( $data[0] ) * 60;
2129 }
2130 }
2131
2132 # No difference ? Return time unchanged
2133 if ( 0 == $minDiff ) {
2134 return $ts;
2135 }
2136
2137 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2138 # Generate an adjusted date; take advantage of the fact that mktime
2139 # will normalize out-of-range values so we don't have to split $minDiff
2140 # into hours and minutes.
2141 $t = mktime( (
2142 (int)substr( $ts, 8, 2 ) ), # Hours
2143 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2144 (int)substr( $ts, 12, 2 ), # Seconds
2145 (int)substr( $ts, 4, 2 ), # Month
2146 (int)substr( $ts, 6, 2 ), # Day
2147 (int)substr( $ts, 0, 4 ) ); # Year
2148
2149 $date = date( 'YmdHis', $t );
2150 MediaWiki\restoreWarnings();
2151
2152 return $date;
2153 }
2154
2155 /**
2156 * This is meant to be used by time(), date(), and timeanddate() to get
2157 * the date preference they're supposed to use, it should be used in
2158 * all children.
2159 *
2160 *<code>
2161 * function timeanddate([...], $format = true) {
2162 * $datePreference = $this->dateFormat($format);
2163 * [...]
2164 * }
2165 *</code>
2166 *
2167 * @param int|string|bool $usePrefs If true, the user's preference is used
2168 * if false, the site/language default is used
2169 * if int/string, assumed to be a format.
2170 * @return string
2171 */
2172 function dateFormat( $usePrefs = true ) {
2173 global $wgUser;
2174
2175 if ( is_bool( $usePrefs ) ) {
2176 if ( $usePrefs ) {
2177 $datePreference = $wgUser->getDatePreference();
2178 } else {
2179 $datePreference = (string)User::getDefaultOption( 'date' );
2180 }
2181 } else {
2182 $datePreference = (string)$usePrefs;
2183 }
2184
2185 // return int
2186 if ( $datePreference == '' ) {
2187 return 'default';
2188 }
2189
2190 return $datePreference;
2191 }
2192
2193 /**
2194 * Get a format string for a given type and preference
2195 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2196 * @param string $pref The format name as it appears in Messages*.php under
2197 * $datePreferences.
2198 *
2199 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2200 *
2201 * @return string
2202 */
2203 function getDateFormatString( $type, $pref ) {
2204 $wasDefault = false;
2205 if ( $pref == 'default' ) {
2206 $wasDefault = true;
2207 $pref = $this->getDefaultDateFormat();
2208 }
2209
2210 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2211 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2212
2213 if ( $type === 'pretty' && $df === null ) {
2214 $df = $this->getDateFormatString( 'date', $pref );
2215 }
2216
2217 if ( !$wasDefault && $df === null ) {
2218 $pref = $this->getDefaultDateFormat();
2219 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2220 }
2221
2222 $this->dateFormatStrings[$type][$pref] = $df;
2223 }
2224 return $this->dateFormatStrings[$type][$pref];
2225 }
2226
2227 /**
2228 * @param string $ts The time format which needs to be turned into a
2229 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2230 * @param bool $adj Whether to adjust the time output according to the
2231 * user configured offset ($timecorrection)
2232 * @param mixed $format True to use user's date format preference
2233 * @param string|bool $timecorrection The time offset as returned by
2234 * validateTimeZone() in Special:Preferences
2235 * @return string
2236 */
2237 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2238 $ts = wfTimestamp( TS_MW, $ts );
2239 if ( $adj ) {
2240 $ts = $this->userAdjust( $ts, $timecorrection );
2241 }
2242 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2243 return $this->sprintfDate( $df, $ts );
2244 }
2245
2246 /**
2247 * @param string $ts The time format which needs to be turned into a
2248 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2249 * @param bool $adj Whether to adjust the time output according to the
2250 * user configured offset ($timecorrection)
2251 * @param mixed $format True to use user's date format preference
2252 * @param string|bool $timecorrection The time offset as returned by
2253 * validateTimeZone() in Special:Preferences
2254 * @return string
2255 */
2256 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2257 $ts = wfTimestamp( TS_MW, $ts );
2258 if ( $adj ) {
2259 $ts = $this->userAdjust( $ts, $timecorrection );
2260 }
2261 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2262 return $this->sprintfDate( $df, $ts );
2263 }
2264
2265 /**
2266 * @param string $ts The time format which needs to be turned into a
2267 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2268 * @param bool $adj Whether to adjust the time output according to the
2269 * user configured offset ($timecorrection)
2270 * @param mixed $format What format to return, if it's false output the
2271 * default one (default true)
2272 * @param string|bool $timecorrection The time offset as returned by
2273 * validateTimeZone() in Special:Preferences
2274 * @return string
2275 */
2276 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2277 $ts = wfTimestamp( TS_MW, $ts );
2278 if ( $adj ) {
2279 $ts = $this->userAdjust( $ts, $timecorrection );
2280 }
2281 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2282 return $this->sprintfDate( $df, $ts );
2283 }
2284
2285 /**
2286 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2287 *
2288 * @since 1.20
2289 *
2290 * @param int $seconds The amount of seconds.
2291 * @param array $chosenIntervals The intervals to enable.
2292 *
2293 * @return string
2294 */
2295 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2296 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2297
2298 $segments = [];
2299
2300 foreach ( $intervals as $intervalName => $intervalValue ) {
2301 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2302 // duration-years, duration-decades, duration-centuries, duration-millennia
2303 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2304 $segments[] = $message->inLanguage( $this )->escaped();
2305 }
2306
2307 return $this->listToText( $segments );
2308 }
2309
2310 /**
2311 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2312 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2313 *
2314 * @since 1.20
2315 *
2316 * @param int $seconds The amount of seconds.
2317 * @param array $chosenIntervals The intervals to enable.
2318 *
2319 * @return array
2320 */
2321 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2322 if ( empty( $chosenIntervals ) ) {
2323 $chosenIntervals = [
2324 'millennia',
2325 'centuries',
2326 'decades',
2327 'years',
2328 'days',
2329 'hours',
2330 'minutes',
2331 'seconds'
2332 ];
2333 }
2334
2335 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2336 $sortedNames = array_keys( $intervals );
2337 $smallestInterval = array_pop( $sortedNames );
2338
2339 $segments = [];
2340
2341 foreach ( $intervals as $name => $length ) {
2342 $value = floor( $seconds / $length );
2343
2344 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2345 $seconds -= $value * $length;
2346 $segments[$name] = $value;
2347 }
2348 }
2349
2350 return $segments;
2351 }
2352
2353 /**
2354 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2355 *
2356 * @param string $type Can be 'date', 'time' or 'both'
2357 * @param string $ts The time format which needs to be turned into a
2358 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2359 * @param User $user User object used to get preferences for timezone and format
2360 * @param array $options Array, can contain the following keys:
2361 * - 'timecorrection': time correction, can have the following values:
2362 * - true: use user's preference
2363 * - false: don't use time correction
2364 * - int: value of time correction in minutes
2365 * - 'format': format to use, can have the following values:
2366 * - true: use user's preference
2367 * - false: use default preference
2368 * - string: format to use
2369 * @since 1.19
2370 * @return string
2371 */
2372 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2373 $ts = wfTimestamp( TS_MW, $ts );
2374 $options += [ 'timecorrection' => true, 'format' => true ];
2375 if ( $options['timecorrection'] !== false ) {
2376 if ( $options['timecorrection'] === true ) {
2377 $offset = $user->getOption( 'timecorrection' );
2378 } else {
2379 $offset = $options['timecorrection'];
2380 }
2381 $ts = $this->userAdjust( $ts, $offset );
2382 }
2383 if ( $options['format'] === true ) {
2384 $format = $user->getDatePreference();
2385 } else {
2386 $format = $options['format'];
2387 }
2388 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2389 return $this->sprintfDate( $df, $ts );
2390 }
2391
2392 /**
2393 * Get the formatted date for the given timestamp and formatted for
2394 * the given user.
2395 *
2396 * @param mixed $ts Mixed: the time format which needs to be turned into a
2397 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2398 * @param User $user User object used to get preferences for timezone and format
2399 * @param array $options Array, can contain the following keys:
2400 * - 'timecorrection': time correction, can have the following values:
2401 * - true: use user's preference
2402 * - false: don't use time correction
2403 * - int: value of time correction in minutes
2404 * - 'format': format to use, can have the following values:
2405 * - true: use user's preference
2406 * - false: use default preference
2407 * - string: format to use
2408 * @since 1.19
2409 * @return string
2410 */
2411 public function userDate( $ts, User $user, array $options = [] ) {
2412 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2413 }
2414
2415 /**
2416 * Get the formatted time for the given timestamp and formatted for
2417 * the given user.
2418 *
2419 * @param mixed $ts The time format which needs to be turned into a
2420 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2421 * @param User $user User object used to get preferences for timezone and format
2422 * @param array $options Array, can contain the following keys:
2423 * - 'timecorrection': time correction, can have the following values:
2424 * - true: use user's preference
2425 * - false: don't use time correction
2426 * - int: value of time correction in minutes
2427 * - 'format': format to use, can have the following values:
2428 * - true: use user's preference
2429 * - false: use default preference
2430 * - string: format to use
2431 * @since 1.19
2432 * @return string
2433 */
2434 public function userTime( $ts, User $user, array $options = [] ) {
2435 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2436 }
2437
2438 /**
2439 * Get the formatted date and time for the given timestamp and formatted for
2440 * the given user.
2441 *
2442 * @param mixed $ts The time format which needs to be turned into a
2443 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2444 * @param User $user User object used to get preferences for timezone and format
2445 * @param array $options Array, can contain the following keys:
2446 * - 'timecorrection': time correction, can have the following values:
2447 * - true: use user's preference
2448 * - false: don't use time correction
2449 * - int: value of time correction in minutes
2450 * - 'format': format to use, can have the following values:
2451 * - true: use user's preference
2452 * - false: use default preference
2453 * - string: format to use
2454 * @since 1.19
2455 * @return string
2456 */
2457 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2458 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2459 }
2460
2461 /**
2462 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2463 *
2464 * Determine the difference between the timestamp and the current time, and
2465 * generate a readable timestamp by returning "<N> <units> ago", where the
2466 * largest possible unit is used.
2467 *
2468 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2469 *
2470 * @param MWTimestamp $time
2471 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2472 * @param User|null $user User the timestamp is being generated for
2473 * (or null to use main context's user)
2474 * @return string Formatted timestamp
2475 */
2476 public function getHumanTimestamp(
2477 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2478 ) {
2479 if ( $relativeTo === null ) {
2480 $relativeTo = new MWTimestamp();
2481 }
2482 if ( $user === null ) {
2483 $user = RequestContext::getMain()->getUser();
2484 }
2485
2486 // Adjust for the user's timezone.
2487 $offsetThis = $time->offsetForUser( $user );
2488 $offsetRel = $relativeTo->offsetForUser( $user );
2489
2490 $ts = '';
2491 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2492 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2493 }
2494
2495 // Reset the timezone on the objects.
2496 $time->timestamp->sub( $offsetThis );
2497 $relativeTo->timestamp->sub( $offsetRel );
2498
2499 return $ts;
2500 }
2501
2502 /**
2503 * Convert an MWTimestamp into a pretty human-readable timestamp using
2504 * the given user preferences and relative base time.
2505 *
2506 * @see Language::getHumanTimestamp
2507 * @param MWTimestamp $ts Timestamp to prettify
2508 * @param MWTimestamp $relativeTo Base timestamp
2509 * @param User $user User preferences to use
2510 * @return string Human timestamp
2511 * @since 1.26
2512 */
2513 private function getHumanTimestampInternal(
2514 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2515 ) {
2516 $diff = $ts->diff( $relativeTo );
2517 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2518 (int)$relativeTo->timestamp->format( 'w' ) );
2519 $days = $diff->days ?: (int)$diffDay;
2520 if ( $diff->invert || $days > 5
2521 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2522 ) {
2523 // Timestamps are in different years: use full timestamp
2524 // Also do full timestamp for future dates
2525 /**
2526 * @todo FIXME: Add better handling of future timestamps.
2527 */
2528 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2529 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2530 } elseif ( $days > 5 ) {
2531 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2532 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2533 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2534 } elseif ( $days > 1 ) {
2535 // Timestamp within the past week: show the day of the week and time
2536 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2537 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2538 // Messages:
2539 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2540 $ts = wfMessage( "$weekday-at" )
2541 ->inLanguage( $this )
2542 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2543 ->text();
2544 } elseif ( $days == 1 ) {
2545 // Timestamp was yesterday: say 'yesterday' and the time.
2546 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2547 $ts = wfMessage( 'yesterday-at' )
2548 ->inLanguage( $this )
2549 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2550 ->text();
2551 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2552 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2553 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2554 $ts = wfMessage( 'today-at' )
2555 ->inLanguage( $this )
2556 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2557 ->text();
2558
2559 // From here on in, the timestamp was soon enough ago so that we can simply say
2560 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2561 } elseif ( $diff->h == 1 ) {
2562 // Less than 90 minutes, but more than an hour ago.
2563 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2564 } elseif ( $diff->i >= 1 ) {
2565 // A few minutes ago.
2566 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2567 } elseif ( $diff->s >= 30 ) {
2568 // Less than a minute, but more than 30 sec ago.
2569 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2570 } else {
2571 // Less than 30 seconds ago.
2572 $ts = wfMessage( 'just-now' )->text();
2573 }
2574
2575 return $ts;
2576 }
2577
2578 /**
2579 * @param string $key
2580 * @return array|null
2581 */
2582 public function getMessage( $key ) {
2583 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2584 }
2585
2586 /**
2587 * @return array
2588 */
2589 function getAllMessages() {
2590 return self::$dataCache->getItem( $this->mCode, 'messages' );
2591 }
2592
2593 /**
2594 * @param string $in
2595 * @param string $out
2596 * @param string $string
2597 * @return string
2598 */
2599 public function iconv( $in, $out, $string ) {
2600 # Even with //IGNORE iconv can whine about illegal characters in
2601 # *input* string. We just ignore those too.
2602 # REF: https://bugs.php.net/bug.php?id=37166
2603 # REF: https://phabricator.wikimedia.org/T18885
2604 MediaWiki\suppressWarnings();
2605 $text = iconv( $in, $out . '//IGNORE', $string );
2606 MediaWiki\restoreWarnings();
2607 return $text;
2608 }
2609
2610 // callback functions for ucwords(), ucwordbreaks()
2611
2612 /**
2613 * @param array $matches
2614 * @return mixed|string
2615 */
2616 function ucwordbreaksCallbackAscii( $matches ) {
2617 return $this->ucfirst( $matches[1] );
2618 }
2619
2620 /**
2621 * @param array $matches
2622 * @return string
2623 */
2624 function ucwordbreaksCallbackMB( $matches ) {
2625 return mb_strtoupper( $matches[0] );
2626 }
2627
2628 /**
2629 * @param array $matches
2630 * @return string
2631 */
2632 function ucwordsCallbackMB( $matches ) {
2633 return mb_strtoupper( $matches[0] );
2634 }
2635
2636 /**
2637 * Make a string's first character uppercase
2638 *
2639 * @param string $str
2640 *
2641 * @return string
2642 */
2643 public function ucfirst( $str ) {
2644 $o = ord( $str );
2645 if ( $o < 96 ) { // if already uppercase...
2646 return $str;
2647 } elseif ( $o < 128 ) {
2648 return ucfirst( $str ); // use PHP's ucfirst()
2649 } else {
2650 // fall back to more complex logic in case of multibyte strings
2651 return $this->uc( $str, true );
2652 }
2653 }
2654
2655 /**
2656 * Convert a string to uppercase
2657 *
2658 * @param string $str
2659 * @param bool $first
2660 *
2661 * @return string
2662 */
2663 public function uc( $str, $first = false ) {
2664 if ( $first ) {
2665 if ( $this->isMultibyte( $str ) ) {
2666 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2667 } else {
2668 return ucfirst( $str );
2669 }
2670 } else {
2671 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2672 }
2673 }
2674
2675 /**
2676 * @param string $str
2677 * @return mixed|string
2678 */
2679 function lcfirst( $str ) {
2680 $o = ord( $str );
2681 if ( !$o ) {
2682 return strval( $str );
2683 } elseif ( $o >= 128 ) {
2684 return $this->lc( $str, true );
2685 } elseif ( $o > 96 ) {
2686 return $str;
2687 } else {
2688 $str[0] = strtolower( $str[0] );
2689 return $str;
2690 }
2691 }
2692
2693 /**
2694 * @param string $str
2695 * @param bool $first
2696 * @return mixed|string
2697 */
2698 function lc( $str, $first = false ) {
2699 if ( $first ) {
2700 if ( $this->isMultibyte( $str ) ) {
2701 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2702 } else {
2703 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2704 }
2705 } else {
2706 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2707 }
2708 }
2709
2710 /**
2711 * @param string $str
2712 * @return bool
2713 */
2714 function isMultibyte( $str ) {
2715 return strlen( $str ) !== mb_strlen( $str );
2716 }
2717
2718 /**
2719 * @param string $str
2720 * @return mixed|string
2721 */
2722 function ucwords( $str ) {
2723 if ( $this->isMultibyte( $str ) ) {
2724 $str = $this->lc( $str );
2725
2726 // regexp to find first letter in each word (i.e. after each space)
2727 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2728
2729 // function to use to capitalize a single char
2730 return preg_replace_callback(
2731 $replaceRegexp,
2732 [ $this, 'ucwordsCallbackMB' ],
2733 $str
2734 );
2735 } else {
2736 return ucwords( strtolower( $str ) );
2737 }
2738 }
2739
2740 /**
2741 * capitalize words at word breaks
2742 *
2743 * @param string $str
2744 * @return mixed
2745 */
2746 function ucwordbreaks( $str ) {
2747 if ( $this->isMultibyte( $str ) ) {
2748 $str = $this->lc( $str );
2749
2750 // since \b doesn't work for UTF-8, we explicitely define word break chars
2751 $breaks = "[ \-\(\)\}\{\.,\?!]";
2752
2753 // find first letter after word break
2754 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2755 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2756
2757 return preg_replace_callback(
2758 $replaceRegexp,
2759 [ $this, 'ucwordbreaksCallbackMB' ],
2760 $str
2761 );
2762 } else {
2763 return preg_replace_callback(
2764 '/\b([\w\x80-\xff]+)\b/',
2765 [ $this, 'ucwordbreaksCallbackAscii' ],
2766 $str
2767 );
2768 }
2769 }
2770
2771 /**
2772 * Return a case-folded representation of $s
2773 *
2774 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2775 * and $s2 are the same except for the case of their characters. It is not
2776 * necessary for the value returned to make sense when displayed.
2777 *
2778 * Do *not* perform any other normalisation in this function. If a caller
2779 * uses this function when it should be using a more general normalisation
2780 * function, then fix the caller.
2781 *
2782 * @param string $s
2783 *
2784 * @return string
2785 */
2786 function caseFold( $s ) {
2787 return $this->uc( $s );
2788 }
2789
2790 /**
2791 * @param string $s
2792 * @return string
2793 * @throws MWException
2794 */
2795 function checkTitleEncoding( $s ) {
2796 if ( is_array( $s ) ) {
2797 throw new MWException( 'Given array to checkTitleEncoding.' );
2798 }
2799 if ( StringUtils::isUtf8( $s ) ) {
2800 return $s;
2801 }
2802
2803 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2804 }
2805
2806 /**
2807 * @return array
2808 */
2809 function fallback8bitEncoding() {
2810 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2811 }
2812
2813 /**
2814 * Most writing systems use whitespace to break up words.
2815 * Some languages such as Chinese don't conventionally do this,
2816 * which requires special handling when breaking up words for
2817 * searching etc.
2818 *
2819 * @return bool
2820 */
2821 function hasWordBreaks() {
2822 return true;
2823 }
2824
2825 /**
2826 * Some languages such as Chinese require word segmentation,
2827 * Specify such segmentation when overridden in derived class.
2828 *
2829 * @param string $string
2830 * @return string
2831 */
2832 function segmentByWord( $string ) {
2833 return $string;
2834 }
2835
2836 /**
2837 * Some languages have special punctuation need to be normalized.
2838 * Make such changes here.
2839 *
2840 * @param string $string
2841 * @return string
2842 */
2843 function normalizeForSearch( $string ) {
2844 return self::convertDoubleWidth( $string );
2845 }
2846
2847 /**
2848 * convert double-width roman characters to single-width.
2849 * range: ff00-ff5f ~= 0020-007f
2850 *
2851 * @param string $string
2852 *
2853 * @return string
2854 */
2855 protected static function convertDoubleWidth( $string ) {
2856 static $full = null;
2857 static $half = null;
2858
2859 if ( $full === null ) {
2860 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2861 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2862 $full = str_split( $fullWidth, 3 );
2863 $half = str_split( $halfWidth );
2864 }
2865
2866 $string = str_replace( $full, $half, $string );
2867 return $string;
2868 }
2869
2870 /**
2871 * @param string $string
2872 * @param string $pattern
2873 * @return string
2874 */
2875 protected static function insertSpace( $string, $pattern ) {
2876 $string = preg_replace( $pattern, " $1 ", $string );
2877 $string = preg_replace( '/ +/', ' ', $string );
2878 return $string;
2879 }
2880
2881 /**
2882 * @param array $termsArray
2883 * @return array
2884 */
2885 function convertForSearchResult( $termsArray ) {
2886 # some languages, e.g. Chinese, need to do a conversion
2887 # in order for search results to be displayed correctly
2888 return $termsArray;
2889 }
2890
2891 /**
2892 * Get the first character of a string.
2893 *
2894 * @param string $s
2895 * @return string
2896 */
2897 function firstChar( $s ) {
2898 $matches = [];
2899 preg_match(
2900 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2901 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2902 $s,
2903 $matches
2904 );
2905
2906 if ( isset( $matches[1] ) ) {
2907 if ( strlen( $matches[1] ) != 3 ) {
2908 return $matches[1];
2909 }
2910
2911 // Break down Hangul syllables to grab the first jamo
2912 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2913 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2914 return $matches[1];
2915 } elseif ( $code < 0xb098 ) {
2916 return "\xe3\x84\xb1";
2917 } elseif ( $code < 0xb2e4 ) {
2918 return "\xe3\x84\xb4";
2919 } elseif ( $code < 0xb77c ) {
2920 return "\xe3\x84\xb7";
2921 } elseif ( $code < 0xb9c8 ) {
2922 return "\xe3\x84\xb9";
2923 } elseif ( $code < 0xbc14 ) {
2924 return "\xe3\x85\x81";
2925 } elseif ( $code < 0xc0ac ) {
2926 return "\xe3\x85\x82";
2927 } elseif ( $code < 0xc544 ) {
2928 return "\xe3\x85\x85";
2929 } elseif ( $code < 0xc790 ) {
2930 return "\xe3\x85\x87";
2931 } elseif ( $code < 0xcc28 ) {
2932 return "\xe3\x85\x88";
2933 } elseif ( $code < 0xce74 ) {
2934 return "\xe3\x85\x8a";
2935 } elseif ( $code < 0xd0c0 ) {
2936 return "\xe3\x85\x8b";
2937 } elseif ( $code < 0xd30c ) {
2938 return "\xe3\x85\x8c";
2939 } elseif ( $code < 0xd558 ) {
2940 return "\xe3\x85\x8d";
2941 } else {
2942 return "\xe3\x85\x8e";
2943 }
2944 } else {
2945 return '';
2946 }
2947 }
2948
2949 /**
2950 * @deprecated No-op since 1.28
2951 */
2952 function initEncoding() {
2953 // No-op.
2954 }
2955
2956 /**
2957 * @param string $s
2958 * @return string
2959 * @deprecated No-op since 1.28
2960 */
2961 function recodeForEdit( $s ) {
2962 return $s;
2963 }
2964
2965 /**
2966 * @param string $s
2967 * @return string
2968 * @deprecated No-op since 1.28
2969 */
2970 function recodeInput( $s ) {
2971 return $s;
2972 }
2973
2974 /**
2975 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2976 * also cleans up certain backwards-compatible sequences, converting them
2977 * to the modern Unicode equivalent.
2978 *
2979 * This is language-specific for performance reasons only.
2980 *
2981 * @param string $s
2982 *
2983 * @return string
2984 */
2985 function normalize( $s ) {
2986 global $wgAllUnicodeFixes;
2987 $s = UtfNormal\Validator::cleanUp( $s );
2988 if ( $wgAllUnicodeFixes ) {
2989 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2990 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2991 }
2992
2993 return $s;
2994 }
2995
2996 /**
2997 * Transform a string using serialized data stored in the given file (which
2998 * must be in the serialized subdirectory of $IP). The file contains pairs
2999 * mapping source characters to destination characters.
3000 *
3001 * The data is cached in process memory. This will go faster if you have the
3002 * FastStringSearch extension.
3003 *
3004 * @param string $file
3005 * @param string $string
3006 *
3007 * @throws MWException
3008 * @return string
3009 */
3010 function transformUsingPairFile( $file, $string ) {
3011 if ( !isset( $this->transformData[$file] ) ) {
3012 $data = wfGetPrecompiledData( $file );
3013 if ( $data === false ) {
3014 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
3015 }
3016 $this->transformData[$file] = new ReplacementArray( $data );
3017 }
3018 return $this->transformData[$file]->replace( $string );
3019 }
3020
3021 /**
3022 * For right-to-left language support
3023 *
3024 * @return bool
3025 */
3026 function isRTL() {
3027 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3028 }
3029
3030 /**
3031 * Return the correct HTML 'dir' attribute value for this language.
3032 * @return string
3033 */
3034 function getDir() {
3035 return $this->isRTL() ? 'rtl' : 'ltr';
3036 }
3037
3038 /**
3039 * Return 'left' or 'right' as appropriate alignment for line-start
3040 * for this language's text direction.
3041 *
3042 * Should be equivalent to CSS3 'start' text-align value....
3043 *
3044 * @return string
3045 */
3046 function alignStart() {
3047 return $this->isRTL() ? 'right' : 'left';
3048 }
3049
3050 /**
3051 * Return 'right' or 'left' as appropriate alignment for line-end
3052 * for this language's text direction.
3053 *
3054 * Should be equivalent to CSS3 'end' text-align value....
3055 *
3056 * @return string
3057 */
3058 function alignEnd() {
3059 return $this->isRTL() ? 'left' : 'right';
3060 }
3061
3062 /**
3063 * A hidden direction mark (LRM or RLM), depending on the language direction.
3064 * Unlike getDirMark(), this function returns the character as an HTML entity.
3065 * This function should be used when the output is guaranteed to be HTML,
3066 * because it makes the output HTML source code more readable. When
3067 * the output is plain text or can be escaped, getDirMark() should be used.
3068 *
3069 * @param bool $opposite Get the direction mark opposite to your language
3070 * @return string
3071 * @since 1.20
3072 */
3073 function getDirMarkEntity( $opposite = false ) {
3074 if ( $opposite ) {
3075 return $this->isRTL() ? '&lrm;' : '&rlm;';
3076 }
3077 return $this->isRTL() ? '&rlm;' : '&lrm;';
3078 }
3079
3080 /**
3081 * A hidden direction mark (LRM or RLM), depending on the language direction.
3082 * This function produces them as invisible Unicode characters and
3083 * the output may be hard to read and debug, so it should only be used
3084 * when the output is plain text or can be escaped. When the output is
3085 * HTML, use getDirMarkEntity() instead.
3086 *
3087 * @param bool $opposite Get the direction mark opposite to your language
3088 * @return string
3089 */
3090 function getDirMark( $opposite = false ) {
3091 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3092 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3093 if ( $opposite ) {
3094 return $this->isRTL() ? $lrm : $rlm;
3095 }
3096 return $this->isRTL() ? $rlm : $lrm;
3097 }
3098
3099 /**
3100 * @return array
3101 */
3102 function capitalizeAllNouns() {
3103 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3104 }
3105
3106 /**
3107 * An arrow, depending on the language direction.
3108 *
3109 * @param string $direction The direction of the arrow: forwards (default),
3110 * backwards, left, right, up, down.
3111 * @return string
3112 */
3113 function getArrow( $direction = 'forwards' ) {
3114 switch ( $direction ) {
3115 case 'forwards':
3116 return $this->isRTL() ? '←' : '→';
3117 case 'backwards':
3118 return $this->isRTL() ? '→' : '←';
3119 case 'left':
3120 return '←';
3121 case 'right':
3122 return '→';
3123 case 'up':
3124 return '↑';
3125 case 'down':
3126 return '↓';
3127 }
3128 }
3129
3130 /**
3131 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3132 *
3133 * @return bool
3134 */
3135 function linkPrefixExtension() {
3136 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3137 }
3138
3139 /**
3140 * Get all magic words from cache.
3141 * @return array
3142 */
3143 function getMagicWords() {
3144 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3145 }
3146
3147 /**
3148 * Run the LanguageGetMagic hook once.
3149 */
3150 protected function doMagicHook() {
3151 if ( $this->mMagicHookDone ) {
3152 return;
3153 }
3154 $this->mMagicHookDone = true;
3155 Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
3156 }
3157
3158 /**
3159 * Fill a MagicWord object with data from here
3160 *
3161 * @param MagicWord $mw
3162 */
3163 function getMagic( $mw ) {
3164 // Saves a function call
3165 if ( !$this->mMagicHookDone ) {
3166 $this->doMagicHook();
3167 }
3168
3169 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3170 $rawEntry = $this->mMagicExtensions[$mw->mId];
3171 } else {
3172 $rawEntry = self::$dataCache->getSubitem(
3173 $this->mCode, 'magicWords', $mw->mId );
3174 }
3175
3176 if ( !is_array( $rawEntry ) ) {
3177 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3178 } else {
3179 $mw->mCaseSensitive = $rawEntry[0];
3180 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3181 }
3182 }
3183
3184 /**
3185 * Add magic words to the extension array
3186 *
3187 * @param array $newWords
3188 */
3189 function addMagicWordsByLang( $newWords ) {
3190 $fallbackChain = $this->getFallbackLanguages();
3191 $fallbackChain = array_reverse( $fallbackChain );
3192 foreach ( $fallbackChain as $code ) {
3193 if ( isset( $newWords[$code] ) ) {
3194 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3195 }
3196 }
3197 }
3198
3199 /**
3200 * Get special page names, as an associative array
3201 * canonical name => array of valid names, including aliases
3202 * @return array
3203 */
3204 function getSpecialPageAliases() {
3205 // Cache aliases because it may be slow to load them
3206 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3207 // Initialise array
3208 $this->mExtendedSpecialPageAliases =
3209 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3210 Hooks::run( 'LanguageGetSpecialPageAliases',
3211 [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
3212 }
3213
3214 return $this->mExtendedSpecialPageAliases;
3215 }
3216
3217 /**
3218 * Italic is unsuitable for some languages
3219 *
3220 * @param string $text The text to be emphasized.
3221 * @return string
3222 */
3223 function emphasize( $text ) {
3224 return "<em>$text</em>";
3225 }
3226
3227 /**
3228 * Normally we output all numbers in plain en_US style, that is
3229 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3230 * point twohundredthirtyfive. However this is not suitable for all
3231 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3232 * Icelandic just want to use commas instead of dots, and dots instead
3233 * of commas like "293.291,235".
3234 *
3235 * An example of this function being called:
3236 * <code>
3237 * wfMessage( 'message' )->numParams( $num )->text()
3238 * </code>
3239 *
3240 * See $separatorTransformTable on MessageIs.php for
3241 * the , => . and . => , implementation.
3242 *
3243 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3244 * @param int|float $number The string to be formatted, should be an integer
3245 * or a floating point number.
3246 * @param bool $nocommafy Set to true for special numbers like dates
3247 * @return string
3248 */
3249 public function formatNum( $number, $nocommafy = false ) {
3250 global $wgTranslateNumerals;
3251 if ( !$nocommafy ) {
3252 $number = $this->commafy( $number );
3253 $s = $this->separatorTransformTable();
3254 if ( $s ) {
3255 $number = strtr( $number, $s );
3256 }
3257 }
3258
3259 if ( $wgTranslateNumerals ) {
3260 $s = $this->digitTransformTable();
3261 if ( $s ) {
3262 $number = strtr( $number, $s );
3263 }
3264 }
3265
3266 return $number;
3267 }
3268
3269 /**
3270 * Front-end for non-commafied formatNum
3271 *
3272 * @param int|float $number The string to be formatted, should be an integer
3273 * or a floating point number.
3274 * @since 1.21
3275 * @return string
3276 */
3277 public function formatNumNoSeparators( $number ) {
3278 return $this->formatNum( $number, true );
3279 }
3280
3281 /**
3282 * @param string $number
3283 * @return string
3284 */
3285 public function parseFormattedNumber( $number ) {
3286 $s = $this->digitTransformTable();
3287 if ( $s ) {
3288 // eliminate empty array values such as ''. (bug 64347)
3289 $s = array_filter( $s );
3290 $number = strtr( $number, array_flip( $s ) );
3291 }
3292
3293 $s = $this->separatorTransformTable();
3294 if ( $s ) {
3295 // eliminate empty array values such as ''. (bug 64347)
3296 $s = array_filter( $s );
3297 $number = strtr( $number, array_flip( $s ) );
3298 }
3299
3300 $number = strtr( $number, [ ',' => '' ] );
3301 return $number;
3302 }
3303
3304 /**
3305 * Adds commas to a given number
3306 * @since 1.19
3307 * @param mixed $number
3308 * @return string
3309 */
3310 function commafy( $number ) {
3311 $digitGroupingPattern = $this->digitGroupingPattern();
3312 if ( $number === null ) {
3313 return '';
3314 }
3315
3316 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3317 // default grouping is at thousands, use the same for ###,###,### pattern too.
3318 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3319 } else {
3320 // Ref: http://cldr.unicode.org/translation/number-patterns
3321 $sign = "";
3322 if ( intval( $number ) < 0 ) {
3323 // For negative numbers apply the algorithm like positive number and add sign.
3324 $sign = "-";
3325 $number = substr( $number, 1 );
3326 }
3327 $integerPart = [];
3328 $decimalPart = [];
3329 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3330 preg_match( "/\d+/", $number, $integerPart );
3331 preg_match( "/\.\d*/", $number, $decimalPart );
3332 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3333 if ( $groupedNumber === $number ) {
3334 // the string does not have any number part. Eg: .12345
3335 return $sign . $groupedNumber;
3336 }
3337 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3338 while ( $start > 0 ) {
3339 $match = $matches[0][$numMatches - 1];
3340 $matchLen = strlen( $match );
3341 $start = $end - $matchLen;
3342 if ( $start < 0 ) {
3343 $start = 0;
3344 }
3345 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3346 $end = $start;
3347 if ( $numMatches > 1 ) {
3348 // use the last pattern for the rest of the number
3349 $numMatches--;
3350 }
3351 if ( $start > 0 ) {
3352 $groupedNumber = "," . $groupedNumber;
3353 }
3354 }
3355 return $sign . $groupedNumber;
3356 }
3357 }
3358
3359 /**
3360 * @return string
3361 */
3362 function digitGroupingPattern() {
3363 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3364 }
3365
3366 /**
3367 * @return array
3368 */
3369 function digitTransformTable() {
3370 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3371 }
3372
3373 /**
3374 * @return array
3375 */
3376 function separatorTransformTable() {
3377 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3378 }
3379
3380 /**
3381 * Take a list of strings and build a locale-friendly comma-separated
3382 * list, using the local comma-separator message.
3383 * The last two strings are chained with an "and".
3384 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3385 *
3386 * @param string[] $l
3387 * @return string
3388 */
3389 function listToText( array $l ) {
3390 $m = count( $l ) - 1;
3391 if ( $m < 0 ) {
3392 return '';
3393 }
3394 if ( $m > 0 ) {
3395 $and = $this->msg( 'and' )->escaped();
3396 $space = $this->msg( 'word-separator' )->escaped();
3397 if ( $m > 1 ) {
3398 $comma = $this->msg( 'comma-separator' )->escaped();
3399 }
3400 }
3401 $s = $l[$m];
3402 for ( $i = $m - 1; $i >= 0; $i-- ) {
3403 if ( $i == $m - 1 ) {
3404 $s = $l[$i] . $and . $space . $s;
3405 } else {
3406 $s = $l[$i] . $comma . $s;
3407 }
3408 }
3409 return $s;
3410 }
3411
3412 /**
3413 * Take a list of strings and build a locale-friendly comma-separated
3414 * list, using the local comma-separator message.
3415 * @param string[] $list Array of strings to put in a comma list
3416 * @return string
3417 */
3418 function commaList( array $list ) {
3419 return implode(
3420 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3421 $list
3422 );
3423 }
3424
3425 /**
3426 * Take a list of strings and build a locale-friendly semicolon-separated
3427 * list, using the local semicolon-separator message.
3428 * @param string[] $list Array of strings to put in a semicolon list
3429 * @return string
3430 */
3431 function semicolonList( array $list ) {
3432 return implode(
3433 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3434 $list
3435 );
3436 }
3437
3438 /**
3439 * Same as commaList, but separate it with the pipe instead.
3440 * @param string[] $list Array of strings to put in a pipe list
3441 * @return string
3442 */
3443 function pipeList( array $list ) {
3444 return implode(
3445 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3446 $list
3447 );
3448 }
3449
3450 /**
3451 * Truncate a string to a specified length in bytes, appending an optional
3452 * string (e.g. for ellipses)
3453 *
3454 * The database offers limited byte lengths for some columns in the database;
3455 * multi-byte character sets mean we need to ensure that only whole characters
3456 * are included, otherwise broken characters can be passed to the user
3457 *
3458 * If $length is negative, the string will be truncated from the beginning
3459 *
3460 * @param string $string String to truncate
3461 * @param int $length Maximum length (including ellipses)
3462 * @param string $ellipsis String to append to the truncated text
3463 * @param bool $adjustLength Subtract length of ellipsis from $length.
3464 * $adjustLength was introduced in 1.18, before that behaved as if false.
3465 * @return string
3466 */
3467 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3468 # Use the localized ellipsis character
3469 if ( $ellipsis == '...' ) {
3470 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3471 }
3472 # Check if there is no need to truncate
3473 if ( $length == 0 ) {
3474 return $ellipsis; // convention
3475 } elseif ( strlen( $string ) <= abs( $length ) ) {
3476 return $string; // no need to truncate
3477 }
3478 $stringOriginal = $string;
3479 # If ellipsis length is >= $length then we can't apply $adjustLength
3480 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3481 $string = $ellipsis; // this can be slightly unexpected
3482 # Otherwise, truncate and add ellipsis...
3483 } else {
3484 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3485 if ( $length > 0 ) {
3486 $length -= $eLength;
3487 $string = substr( $string, 0, $length ); // xyz...
3488 $string = $this->removeBadCharLast( $string );
3489 $string = rtrim( $string );
3490 $string = $string . $ellipsis;
3491 } else {
3492 $length += $eLength;
3493 $string = substr( $string, $length ); // ...xyz
3494 $string = $this->removeBadCharFirst( $string );
3495 $string = ltrim( $string );
3496 $string = $ellipsis . $string;
3497 }
3498 }
3499 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3500 # This check is *not* redundant if $adjustLength, due to the single case where
3501 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3502 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3503 return $string;
3504 } else {
3505 return $stringOriginal;
3506 }
3507 }
3508
3509 /**
3510 * Remove bytes that represent an incomplete Unicode character
3511 * at the end of string (e.g. bytes of the char are missing)
3512 *
3513 * @param string $string
3514 * @return string
3515 */
3516 protected function removeBadCharLast( $string ) {
3517 if ( $string != '' ) {
3518 $char = ord( $string[strlen( $string ) - 1] );
3519 $m = [];
3520 if ( $char >= 0xc0 ) {
3521 # We got the first byte only of a multibyte char; remove it.
3522 $string = substr( $string, 0, -1 );
3523 } elseif ( $char >= 0x80 &&
3524 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3525 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3526 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3527 ) {
3528 # We chopped in the middle of a character; remove it
3529 $string = $m[1];
3530 }
3531 }
3532 return $string;
3533 }
3534
3535 /**
3536 * Remove bytes that represent an incomplete Unicode character
3537 * at the start of string (e.g. bytes of the char are missing)
3538 *
3539 * @param string $string
3540 * @return string
3541 */
3542 protected function removeBadCharFirst( $string ) {
3543 if ( $string != '' ) {
3544 $char = ord( $string[0] );
3545 if ( $char >= 0x80 && $char < 0xc0 ) {
3546 # We chopped in the middle of a character; remove the whole thing
3547 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3548 }
3549 }
3550 return $string;
3551 }
3552
3553 /**
3554 * Truncate a string of valid HTML to a specified length in bytes,
3555 * appending an optional string (e.g. for ellipses), and return valid HTML
3556 *
3557 * This is only intended for styled/linked text, such as HTML with
3558 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3559 * Also, this will not detect things like "display:none" CSS.
3560 *
3561 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3562 *
3563 * @param string $text HTML string to truncate
3564 * @param int $length (zero/positive) Maximum length (including ellipses)
3565 * @param string $ellipsis String to append to the truncated text
3566 * @return string
3567 */
3568 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3569 # Use the localized ellipsis character
3570 if ( $ellipsis == '...' ) {
3571 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3572 }
3573 # Check if there is clearly no need to truncate
3574 if ( $length <= 0 ) {
3575 return $ellipsis; // no text shown, nothing to format (convention)
3576 } elseif ( strlen( $text ) <= $length ) {
3577 return $text; // string short enough even *with* HTML (short-circuit)
3578 }
3579
3580 $dispLen = 0; // innerHTML legth so far
3581 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3582 $tagType = 0; // 0-open, 1-close
3583 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3584 $entityState = 0; // 0-not entity, 1-entity
3585 $tag = $ret = ''; // accumulated tag name, accumulated result string
3586 $openTags = []; // open tag stack
3587 $maybeState = null; // possible truncation state
3588
3589 $textLen = strlen( $text );
3590 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3591 for ( $pos = 0; true; ++$pos ) {
3592 # Consider truncation once the display length has reached the maximim.
3593 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3594 # Check that we're not in the middle of a bracket/entity...
3595 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3596 if ( !$testingEllipsis ) {
3597 $testingEllipsis = true;
3598 # Save where we are; we will truncate here unless there turn out to
3599 # be so few remaining characters that truncation is not necessary.
3600 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3601 $maybeState = [ $ret, $openTags ]; // save state
3602 }
3603 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3604 # String in fact does need truncation, the truncation point was OK.
3605 list( $ret, $openTags ) = $maybeState; // reload state
3606 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3607 $ret .= $ellipsis; // add ellipsis
3608 break;
3609 }
3610 }
3611 if ( $pos >= $textLen ) {
3612 break; // extra iteration just for above checks
3613 }
3614
3615 # Read the next char...
3616 $ch = $text[$pos];
3617 $lastCh = $pos ? $text[$pos - 1] : '';
3618 $ret .= $ch; // add to result string
3619 if ( $ch == '<' ) {
3620 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3621 $entityState = 0; // for bad HTML
3622 $bracketState = 1; // tag started (checking for backslash)
3623 } elseif ( $ch == '>' ) {
3624 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3625 $entityState = 0; // for bad HTML
3626 $bracketState = 0; // out of brackets
3627 } elseif ( $bracketState == 1 ) {
3628 if ( $ch == '/' ) {
3629 $tagType = 1; // close tag (e.g. "</span>")
3630 } else {
3631 $tagType = 0; // open tag (e.g. "<span>")
3632 $tag .= $ch;
3633 }
3634 $bracketState = 2; // building tag name
3635 } elseif ( $bracketState == 2 ) {
3636 if ( $ch != ' ' ) {
3637 $tag .= $ch;
3638 } else {
3639 // Name found (e.g. "<a href=..."), add on tag attributes...
3640 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3641 }
3642 } elseif ( $bracketState == 0 ) {
3643 if ( $entityState ) {
3644 if ( $ch == ';' ) {
3645 $entityState = 0;
3646 $dispLen++; // entity is one displayed char
3647 }
3648 } else {
3649 if ( $neLength == 0 && !$maybeState ) {
3650 // Save state without $ch. We want to *hit* the first
3651 // display char (to get tags) but not *use* it if truncating.
3652 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3653 }
3654 if ( $ch == '&' ) {
3655 $entityState = 1; // entity found, (e.g. "&#160;")
3656 } else {
3657 $dispLen++; // this char is displayed
3658 // Add the next $max display text chars after this in one swoop...
3659 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3660 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3661 $dispLen += $skipped;
3662 $pos += $skipped;
3663 }
3664 }
3665 }
3666 }
3667 // Close the last tag if left unclosed by bad HTML
3668 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3669 while ( count( $openTags ) > 0 ) {
3670 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3671 }
3672 return $ret;
3673 }
3674
3675 /**
3676 * truncateHtml() helper function
3677 * like strcspn() but adds the skipped chars to $ret
3678 *
3679 * @param string $ret
3680 * @param string $text
3681 * @param string $search
3682 * @param int $start
3683 * @param null|int $len
3684 * @return int
3685 */
3686 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3687 if ( $len === null ) {
3688 $len = -1; // -1 means "no limit" for strcspn
3689 } elseif ( $len < 0 ) {
3690 $len = 0; // sanity
3691 }
3692 $skipCount = 0;
3693 if ( $start < strlen( $text ) ) {
3694 $skipCount = strcspn( $text, $search, $start, $len );
3695 $ret .= substr( $text, $start, $skipCount );
3696 }
3697 return $skipCount;
3698 }
3699
3700 /**
3701 * truncateHtml() helper function
3702 * (a) push or pop $tag from $openTags as needed
3703 * (b) clear $tag value
3704 * @param string &$tag Current HTML tag name we are looking at
3705 * @param int $tagType (0-open tag, 1-close tag)
3706 * @param string $lastCh Character before the '>' that ended this tag
3707 * @param array &$openTags Open tag stack (not accounting for $tag)
3708 */
3709 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3710 $tag = ltrim( $tag );
3711 if ( $tag != '' ) {
3712 if ( $tagType == 0 && $lastCh != '/' ) {
3713 $openTags[] = $tag; // tag opened (didn't close itself)
3714 } elseif ( $tagType == 1 ) {
3715 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3716 array_pop( $openTags ); // tag closed
3717 }
3718 }
3719 $tag = '';
3720 }
3721 }
3722
3723 /**
3724 * Grammatical transformations, needed for inflected languages
3725 * Invoked by putting {{grammar:case|word}} in a message
3726 *
3727 * @param string $word
3728 * @param string $case
3729 * @return string
3730 */
3731 function convertGrammar( $word, $case ) {
3732 global $wgGrammarForms;
3733 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3734 return $wgGrammarForms[$this->getCode()][$case][$word];
3735 }
3736
3737 return $word;
3738 }
3739
3740 /**
3741 * Get the grammar forms for the content language
3742 * @return array Array of grammar forms
3743 * @since 1.20
3744 */
3745 function getGrammarForms() {
3746 global $wgGrammarForms;
3747 if ( isset( $wgGrammarForms[$this->getCode()] )
3748 && is_array( $wgGrammarForms[$this->getCode()] )
3749 ) {
3750 return $wgGrammarForms[$this->getCode()];
3751 }
3752
3753 return [];
3754 }
3755
3756 /**
3757 * Get the grammar transformations data for the language.
3758 * Used like grammar forms, with {{GRAMMAR}} and cases,
3759 * but uses pairs of regexes and replacements instead of code.
3760 *
3761 * @return array[] Array of grammar transformations.
3762 * @throws MWException
3763 * @since 1.28
3764 */
3765 public function getGrammarTransformations() {
3766 $languageCode = $this->getCode();
3767
3768 if ( self::$grammarTransformations === null ) {
3769 self::$grammarTransformations = new MapCacheLRU( 10 );
3770 }
3771
3772 if ( self::$grammarTransformations->has( $languageCode ) ) {
3773 return self::$grammarTransformations->get( $languageCode );
3774 }
3775
3776 $data = [];
3777
3778 $grammarDataFile = __DIR__ . "/data/grammarTransformations/$languageCode.json";
3779 if ( is_readable( $grammarDataFile ) ) {
3780 $data = FormatJson::decode(
3781 file_get_contents( $grammarDataFile ),
3782 true
3783 );
3784
3785 if ( $data === null ) {
3786 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3787 }
3788
3789 self::$grammarTransformations->set( $languageCode, $data );
3790 }
3791
3792 return $data;
3793 }
3794
3795 /**
3796 * Provides an alternative text depending on specified gender.
3797 * Usage {{gender:username|masculine|feminine|unknown}}.
3798 * username is optional, in which case the gender of current user is used,
3799 * but only in (some) interface messages; otherwise default gender is used.
3800 *
3801 * If no forms are given, an empty string is returned. If only one form is
3802 * given, it will be returned unconditionally. These details are implied by
3803 * the caller and cannot be overridden in subclasses.
3804 *
3805 * If three forms are given, the default is to use the third (unknown) form.
3806 * If fewer than three forms are given, the default is to use the first (masculine) form.
3807 * These details can be overridden in subclasses.
3808 *
3809 * @param string $gender
3810 * @param array $forms
3811 *
3812 * @return string
3813 */
3814 function gender( $gender, $forms ) {
3815 if ( !count( $forms ) ) {
3816 return '';
3817 }
3818 $forms = $this->preConvertPlural( $forms, 2 );
3819 if ( $gender === 'male' ) {
3820 return $forms[0];
3821 }
3822 if ( $gender === 'female' ) {
3823 return $forms[1];
3824 }
3825 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3826 }
3827
3828 /**
3829 * Plural form transformations, needed for some languages.
3830 * For example, there are 3 form of plural in Russian and Polish,
3831 * depending on "count mod 10". See [[w:Plural]]
3832 * For English it is pretty simple.
3833 *
3834 * Invoked by putting {{plural:count|wordform1|wordform2}}
3835 * or {{plural:count|wordform1|wordform2|wordform3}}
3836 *
3837 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3838 *
3839 * @param int $count Non-localized number
3840 * @param array $forms Different plural forms
3841 * @return string Correct form of plural for $count in this language
3842 */
3843 function convertPlural( $count, $forms ) {
3844 // Handle explicit n=pluralform cases
3845 $forms = $this->handleExplicitPluralForms( $count, $forms );
3846 if ( is_string( $forms ) ) {
3847 return $forms;
3848 }
3849 if ( !count( $forms ) ) {
3850 return '';
3851 }
3852
3853 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3854 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3855 return $forms[$pluralForm];
3856 }
3857
3858 /**
3859 * Handles explicit plural forms for Language::convertPlural()
3860 *
3861 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3862 * If an explicitly defined plural form matches the $count, then
3863 * string value returned, otherwise array returned for further consideration
3864 * by CLDR rules or overridden convertPlural().
3865 *
3866 * @since 1.23
3867 *
3868 * @param int $count Non-localized number
3869 * @param array $forms Different plural forms
3870 *
3871 * @return array|string
3872 */
3873 protected function handleExplicitPluralForms( $count, array $forms ) {
3874 foreach ( $forms as $index => $form ) {
3875 if ( preg_match( '/\d+=/i', $form ) ) {
3876 $pos = strpos( $form, '=' );
3877 if ( substr( $form, 0, $pos ) === (string)$count ) {
3878 return substr( $form, $pos + 1 );
3879 }
3880 unset( $forms[$index] );
3881 }
3882 }
3883 return array_values( $forms );
3884 }
3885
3886 /**
3887 * Checks that convertPlural was given an array and pads it to requested
3888 * amount of forms by copying the last one.
3889 *
3890 * @param array $forms Array of forms given to convertPlural
3891 * @param int $count How many forms should there be at least
3892 * @return array Padded array of forms or an exception if not an array
3893 */
3894 protected function preConvertPlural( /* Array */ $forms, $count ) {
3895 while ( count( $forms ) < $count ) {
3896 $forms[] = $forms[count( $forms ) - 1];
3897 }
3898 return $forms;
3899 }
3900
3901 /**
3902 * Wraps argument with unicode control characters for directionality safety
3903 *
3904 * This solves the problem where directionality-neutral characters at the edge of
3905 * the argument string get interpreted with the wrong directionality from the
3906 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3907 *
3908 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3909 * directionality of the argument string, using the BIDI algorithm's own "First
3910 * strong directional codepoint" rule. Essentially, this works round the fact that
3911 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3912 * direction inference). The latter is cleaner but still not widely supported.
3913 *
3914 * @param string $text Text to wrap
3915 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3916 */
3917 public function embedBidi( $text = '' ) {
3918 $dir = Language::strongDirFromContent( $text );
3919 if ( $dir === 'ltr' ) {
3920 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3921 return self::$lre . $text . self::$pdf;
3922 }
3923 if ( $dir === 'rtl' ) {
3924 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3925 return self::$rle . $text . self::$pdf;
3926 }
3927 // No strong directionality: do not wrap
3928 return $text;
3929 }
3930
3931 /**
3932 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3933 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3934 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3935 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3936 * match up with it.
3937 *
3938 * @param string $str The validated block duration in English
3939 * @param User $user User object to use timezone from or null for $wgUser
3940 * @return string Somehow translated block duration
3941 * @see LanguageFi.php for example implementation
3942 */
3943 function translateBlockExpiry( $str, User $user = null ) {
3944 $duration = SpecialBlock::getSuggestedDurations( $this );
3945 foreach ( $duration as $show => $value ) {
3946 if ( strcmp( $str, $value ) == 0 ) {
3947 return htmlspecialchars( trim( $show ) );
3948 }
3949 }
3950
3951 if ( wfIsInfinity( $str ) ) {
3952 foreach ( $duration as $show => $value ) {
3953 if ( wfIsInfinity( $value ) ) {
3954 return htmlspecialchars( trim( $show ) );
3955 }
3956 }
3957 }
3958
3959 // If all else fails, return a standard duration or timestamp description.
3960 $time = strtotime( $str, 0 );
3961 if ( $time === false ) { // Unknown format. Return it as-is in case.
3962 return $str;
3963 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3964 // $time is relative to 0 so it's a duration length.
3965 return $this->formatDuration( $time );
3966 } else { // It's an absolute timestamp.
3967 if ( $time === 0 ) {
3968 // wfTimestamp() handles 0 as current time instead of epoch.
3969 $time = '19700101000000';
3970 }
3971 if ( $user ) {
3972 return $this->userTimeAndDate( $time, $user );
3973 }
3974 return $this->timeanddate( $time );
3975 }
3976 }
3977
3978 /**
3979 * languages like Chinese need to be segmented in order for the diff
3980 * to be of any use
3981 *
3982 * @param string $text
3983 * @return string
3984 */
3985 public function segmentForDiff( $text ) {
3986 return $text;
3987 }
3988
3989 /**
3990 * and unsegment to show the result
3991 *
3992 * @param string $text
3993 * @return string
3994 */
3995 public function unsegmentForDiff( $text ) {
3996 return $text;
3997 }
3998
3999 /**
4000 * Return the LanguageConverter used in the Language
4001 *
4002 * @since 1.19
4003 * @return LanguageConverter
4004 */
4005 public function getConverter() {
4006 return $this->mConverter;
4007 }
4008
4009 /**
4010 * convert text to all supported variants
4011 *
4012 * @param string $text
4013 * @return array
4014 */
4015 public function autoConvertToAllVariants( $text ) {
4016 return $this->mConverter->autoConvertToAllVariants( $text );
4017 }
4018
4019 /**
4020 * convert text to different variants of a language.
4021 *
4022 * @param string $text
4023 * @return string
4024 */
4025 public function convert( $text ) {
4026 return $this->mConverter->convert( $text );
4027 }
4028
4029 /**
4030 * Convert a Title object to a string in the preferred variant
4031 *
4032 * @param Title $title
4033 * @return string
4034 */
4035 public function convertTitle( $title ) {
4036 return $this->mConverter->convertTitle( $title );
4037 }
4038
4039 /**
4040 * Convert a namespace index to a string in the preferred variant
4041 *
4042 * @param int $ns
4043 * @return string
4044 */
4045 public function convertNamespace( $ns ) {
4046 return $this->mConverter->convertNamespace( $ns );
4047 }
4048
4049 /**
4050 * Check if this is a language with variants
4051 *
4052 * @return bool
4053 */
4054 public function hasVariants() {
4055 return count( $this->getVariants() ) > 1;
4056 }
4057
4058 /**
4059 * Check if the language has the specific variant
4060 *
4061 * @since 1.19
4062 * @param string $variant
4063 * @return bool
4064 */
4065 public function hasVariant( $variant ) {
4066 return (bool)$this->mConverter->validateVariant( $variant );
4067 }
4068
4069 /**
4070 * Perform output conversion on a string, and encode for safe HTML output.
4071 * @param string $text Text to be converted
4072 * @param bool $isTitle Whether this conversion is for the article title
4073 * @return string
4074 * @todo this should get integrated somewhere sane
4075 */
4076 public function convertHtml( $text, $isTitle = false ) {
4077 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4078 }
4079
4080 /**
4081 * @param string $key
4082 * @return string
4083 */
4084 public function convertCategoryKey( $key ) {
4085 return $this->mConverter->convertCategoryKey( $key );
4086 }
4087
4088 /**
4089 * Get the list of variants supported by this language
4090 * see sample implementation in LanguageZh.php
4091 *
4092 * @return array An array of language codes
4093 */
4094 public function getVariants() {
4095 return $this->mConverter->getVariants();
4096 }
4097
4098 /**
4099 * @return string
4100 */
4101 public function getPreferredVariant() {
4102 return $this->mConverter->getPreferredVariant();
4103 }
4104
4105 /**
4106 * @return string
4107 */
4108 public function getDefaultVariant() {
4109 return $this->mConverter->getDefaultVariant();
4110 }
4111
4112 /**
4113 * @return string
4114 */
4115 public function getURLVariant() {
4116 return $this->mConverter->getURLVariant();
4117 }
4118
4119 /**
4120 * If a language supports multiple variants, it is
4121 * possible that non-existing link in one variant
4122 * actually exists in another variant. this function
4123 * tries to find it. See e.g. LanguageZh.php
4124 * The input parameters may be modified upon return
4125 *
4126 * @param string &$link The name of the link
4127 * @param Title &$nt The title object of the link
4128 * @param bool $ignoreOtherCond To disable other conditions when
4129 * we need to transclude a template or update a category's link
4130 */
4131 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4132 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4133 }
4134
4135 /**
4136 * returns language specific options used by User::getPageRenderHash()
4137 * for example, the preferred language variant
4138 *
4139 * @return string
4140 */
4141 function getExtraHashOptions() {
4142 return $this->mConverter->getExtraHashOptions();
4143 }
4144
4145 /**
4146 * For languages that support multiple variants, the title of an
4147 * article may be displayed differently in different variants. this
4148 * function returns the apporiate title defined in the body of the article.
4149 *
4150 * @return string
4151 */
4152 public function getParsedTitle() {
4153 return $this->mConverter->getParsedTitle();
4154 }
4155
4156 /**
4157 * Refresh the cache of conversion tables when
4158 * MediaWiki:Conversiontable* is updated.
4159 *
4160 * @param Title $title The Title of the page being updated
4161 */
4162 public function updateConversionTable( Title $title ) {
4163 $this->mConverter->updateConversionTable( $title );
4164 }
4165
4166 /**
4167 * Prepare external link text for conversion. When the text is
4168 * a URL, it shouldn't be converted, and it'll be wrapped in
4169 * the "raw" tag (-{R| }-) to prevent conversion.
4170 *
4171 * This function is called "markNoConversion" for historical
4172 * reasons.
4173 *
4174 * @param string $text Text to be used for external link
4175 * @param bool $noParse Wrap it without confirming it's a real URL first
4176 * @return string The tagged text
4177 */
4178 public function markNoConversion( $text, $noParse = false ) {
4179 // Excluding protocal-relative URLs may avoid many false positives.
4180 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4181 return $this->mConverter->markNoConversion( $text );
4182 } else {
4183 return $text;
4184 }
4185 }
4186
4187 /**
4188 * A regular expression to match legal word-trailing characters
4189 * which should be merged onto a link of the form [[foo]]bar.
4190 *
4191 * @return string
4192 */
4193 public function linkTrail() {
4194 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4195 }
4196
4197 /**
4198 * A regular expression character set to match legal word-prefixing
4199 * characters which should be merged onto a link of the form foo[[bar]].
4200 *
4201 * @return string
4202 */
4203 public function linkPrefixCharset() {
4204 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4205 }
4206
4207 /**
4208 * Get the "parent" language which has a converter to convert a "compatible" language
4209 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4210 *
4211 * @return Language|null
4212 * @since 1.22
4213 */
4214 public function getParentLanguage() {
4215 if ( $this->mParentLanguage !== false ) {
4216 return $this->mParentLanguage;
4217 }
4218
4219 $code = explode( '-', $this->getCode() )[0];
4220 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4221 $this->mParentLanguage = null;
4222 return null;
4223 }
4224 $lang = Language::factory( $code );
4225 if ( !$lang->hasVariant( $this->getCode() ) ) {
4226 $this->mParentLanguage = null;
4227 return null;
4228 }
4229
4230 $this->mParentLanguage = $lang;
4231 return $lang;
4232 }
4233
4234 /**
4235 * Compare with an other language object
4236 *
4237 * @since 1.28
4238 * @param Language $lang
4239 * @return boolean
4240 */
4241 public function equals( Language $lang ) {
4242 return $lang->getCode() === $this->mCode;
4243 }
4244
4245 /**
4246 * Get the internal language code for this language object
4247 *
4248 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4249 * htmlspecialchars() or similar
4250 *
4251 * @return string
4252 */
4253 public function getCode() {
4254 return $this->mCode;
4255 }
4256
4257 /**
4258 * Get the code in BCP 47 format which we can use
4259 * inside of html lang="" tags.
4260 *
4261 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4262 * htmlspecialchars() or similar.
4263 *
4264 * @since 1.19
4265 * @return string
4266 */
4267 public function getHtmlCode() {
4268 if ( is_null( $this->mHtmlCode ) ) {
4269 $this->mHtmlCode = wfBCP47( $this->getCode() );
4270 }
4271 return $this->mHtmlCode;
4272 }
4273
4274 /**
4275 * @param string $code
4276 */
4277 public function setCode( $code ) {
4278 $this->mCode = $code;
4279 // Ensure we don't leave incorrect cached data lying around
4280 $this->mHtmlCode = null;
4281 $this->mParentLanguage = false;
4282 }
4283
4284 /**
4285 * Get the language code from a file name. Inverse of getFileName()
4286 * @param string $filename $prefix . $languageCode . $suffix
4287 * @param string $prefix Prefix before the language code
4288 * @param string $suffix Suffix after the language code
4289 * @return string Language code, or false if $prefix or $suffix isn't found
4290 */
4291 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4292 $m = null;
4293 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4294 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4295 if ( !count( $m ) ) {
4296 return false;
4297 }
4298 return str_replace( '_', '-', strtolower( $m[1] ) );
4299 }
4300
4301 /**
4302 * @param string $code
4303 * @return string Name of the language class
4304 */
4305 public static function classFromCode( $code ) {
4306 if ( $code == 'en' ) {
4307 return 'Language';
4308 } else {
4309 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4310 }
4311 }
4312
4313 /**
4314 * Get the name of a file for a certain language code
4315 * @param string $prefix Prepend this to the filename
4316 * @param string $code Language code
4317 * @param string $suffix Append this to the filename
4318 * @throws MWException
4319 * @return string $prefix . $mangledCode . $suffix
4320 */
4321 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4322 if ( !self::isValidBuiltInCode( $code ) ) {
4323 throw new MWException( "Invalid language code \"$code\"" );
4324 }
4325
4326 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4327 }
4328
4329 /**
4330 * @param string $code
4331 * @return string
4332 */
4333 public static function getMessagesFileName( $code ) {
4334 global $IP;
4335 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4336 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4337 return $file;
4338 }
4339
4340 /**
4341 * @param string $code
4342 * @return string
4343 * @throws MWException
4344 * @since 1.23
4345 */
4346 public static function getJsonMessagesFileName( $code ) {
4347 global $IP;
4348
4349 if ( !self::isValidBuiltInCode( $code ) ) {
4350 throw new MWException( "Invalid language code \"$code\"" );
4351 }
4352
4353 return "$IP/languages/i18n/$code.json";
4354 }
4355
4356 /**
4357 * Get the first fallback for a given language.
4358 *
4359 * @param string $code
4360 *
4361 * @return bool|string
4362 */
4363 public static function getFallbackFor( $code ) {
4364 $fallbacks = self::getFallbacksFor( $code );
4365 if ( $fallbacks ) {
4366 return $fallbacks[0];
4367 }
4368 return false;
4369 }
4370
4371 /**
4372 * Get the ordered list of fallback languages.
4373 *
4374 * @since 1.19
4375 * @param string $code Language code
4376 * @return array Non-empty array, ending in "en"
4377 */
4378 public static function getFallbacksFor( $code ) {
4379 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4380 return [];
4381 }
4382 // For unknown languages, fallbackSequence returns an empty array,
4383 // hardcode fallback to 'en' in that case.
4384 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4385 }
4386
4387 /**
4388 * Get the ordered list of fallback languages, ending with the fallback
4389 * language chain for the site language.
4390 *
4391 * @since 1.22
4392 * @param string $code Language code
4393 * @return array Array( fallbacks, site fallbacks )
4394 */
4395 public static function getFallbacksIncludingSiteLanguage( $code ) {
4396 global $wgLanguageCode;
4397
4398 // Usually, we will only store a tiny number of fallback chains, so we
4399 // keep them in static memory.
4400 $cacheKey = "{$code}-{$wgLanguageCode}";
4401
4402 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4403 $fallbacks = self::getFallbacksFor( $code );
4404
4405 // Append the site's fallback chain, including the site language itself
4406 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4407 array_unshift( $siteFallbacks, $wgLanguageCode );
4408
4409 // Eliminate any languages already included in the chain
4410 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4411
4412 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4413 }
4414 return self::$fallbackLanguageCache[$cacheKey];
4415 }
4416
4417 /**
4418 * Get all messages for a given language
4419 * WARNING: this may take a long time. If you just need all message *keys*
4420 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4421 *
4422 * @param string $code
4423 *
4424 * @return array
4425 */
4426 public static function getMessagesFor( $code ) {
4427 return self::getLocalisationCache()->getItem( $code, 'messages' );
4428 }
4429
4430 /**
4431 * Get a message for a given language
4432 *
4433 * @param string $key
4434 * @param string $code
4435 *
4436 * @return string
4437 */
4438 public static function getMessageFor( $key, $code ) {
4439 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4440 }
4441
4442 /**
4443 * Get all message keys for a given language. This is a faster alternative to
4444 * array_keys( Language::getMessagesFor( $code ) )
4445 *
4446 * @since 1.19
4447 * @param string $code Language code
4448 * @return array Array of message keys (strings)
4449 */
4450 public static function getMessageKeysFor( $code ) {
4451 return self::getLocalisationCache()->getSubitemList( $code, 'messages' );
4452 }
4453
4454 /**
4455 * @param string $talk
4456 * @return mixed
4457 */
4458 function fixVariableInNamespace( $talk ) {
4459 if ( strpos( $talk, '$1' ) === false ) {
4460 return $talk;
4461 }
4462
4463 global $wgMetaNamespace;
4464 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4465
4466 # Allow grammar transformations
4467 # Allowing full message-style parsing would make simple requests
4468 # such as action=raw much more expensive than they need to be.
4469 # This will hopefully cover most cases.
4470 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4471 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4472 return str_replace( ' ', '_', $talk );
4473 }
4474
4475 /**
4476 * @param string $m
4477 * @return string
4478 */
4479 function replaceGrammarInNamespace( $m ) {
4480 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4481 }
4482
4483 /**
4484 * Decode an expiry (block, protection, etc) which has come from the DB
4485 *
4486 * @param string $expiry Database expiry String
4487 * @param bool|int $format True to process using language functions, or TS_ constant
4488 * to return the expiry in a given timestamp
4489 * @param string $infinity If $format is not true, use this string for infinite expiry
4490 * @return string
4491 * @since 1.18
4492 */
4493 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4494 static $dbInfinity;
4495 if ( $dbInfinity === null ) {
4496 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4497 }
4498
4499 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4500 return $format === true
4501 ? $this->getMessageFromDB( 'infiniteblock' )
4502 : $infinity;
4503 } else {
4504 return $format === true
4505 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4506 : wfTimestamp( $format, $expiry );
4507 }
4508 }
4509
4510 /**
4511 * Formats a time given in seconds into a string representation of that time.
4512 *
4513 * @param int|float $seconds
4514 * @param array $format An optional argument that formats the returned string in different ways:
4515 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4516 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4517 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4518 * and friends.
4519 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4520 * or 'avoidminutes'.
4521 * @return string
4522 */
4523 function formatTimePeriod( $seconds, $format = [] ) {
4524 if ( !is_array( $format ) ) {
4525 $format = [ 'avoid' => $format ]; // For backwards compatibility
4526 }
4527 if ( !isset( $format['avoid'] ) ) {
4528 $format['avoid'] = false;
4529 }
4530 if ( !isset( $format['noabbrevs'] ) ) {
4531 $format['noabbrevs'] = false;
4532 }
4533 $secondsMsg = wfMessage(
4534 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4535 $minutesMsg = wfMessage(
4536 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4537 $hoursMsg = wfMessage(
4538 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4539 $daysMsg = wfMessage(
4540 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4541
4542 if ( round( $seconds * 10 ) < 100 ) {
4543 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4544 $s = $secondsMsg->params( $s )->text();
4545 } elseif ( round( $seconds ) < 60 ) {
4546 $s = $this->formatNum( round( $seconds ) );
4547 $s = $secondsMsg->params( $s )->text();
4548 } elseif ( round( $seconds ) < 3600 ) {
4549 $minutes = floor( $seconds / 60 );
4550 $secondsPart = round( fmod( $seconds, 60 ) );
4551 if ( $secondsPart == 60 ) {
4552 $secondsPart = 0;
4553 $minutes++;
4554 }
4555 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4556 $s .= ' ';
4557 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4558 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4559 $hours = floor( $seconds / 3600 );
4560 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4561 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4562 if ( $secondsPart == 60 ) {
4563 $secondsPart = 0;
4564 $minutes++;
4565 }
4566 if ( $minutes == 60 ) {
4567 $minutes = 0;
4568 $hours++;
4569 }
4570 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4571 $s .= ' ';
4572 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4573 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4574 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4575 }
4576 } else {
4577 $days = floor( $seconds / 86400 );
4578 if ( $format['avoid'] === 'avoidminutes' ) {
4579 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4580 if ( $hours == 24 ) {
4581 $hours = 0;
4582 $days++;
4583 }
4584 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4585 $s .= ' ';
4586 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4587 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4588 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4589 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4590 if ( $minutes == 60 ) {
4591 $minutes = 0;
4592 $hours++;
4593 }
4594 if ( $hours == 24 ) {
4595 $hours = 0;
4596 $days++;
4597 }
4598 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4599 $s .= ' ';
4600 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4601 $s .= ' ';
4602 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4603 } else {
4604 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4605 $s .= ' ';
4606 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4607 }
4608 }
4609 return $s;
4610 }
4611
4612 /**
4613 * Format a bitrate for output, using an appropriate
4614 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4615 * the magnitude in question.
4616 *
4617 * This use base 1000. For base 1024 use formatSize(), for another base
4618 * see formatComputingNumbers().
4619 *
4620 * @param int $bps
4621 * @return string
4622 */
4623 function formatBitrate( $bps ) {
4624 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4625 }
4626
4627 /**
4628 * @param int $size Size of the unit
4629 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4630 * @param string $messageKey Message key to be uesd
4631 * @return string
4632 */
4633 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4634 if ( $size <= 0 ) {
4635 return str_replace( '$1', $this->formatNum( $size ),
4636 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4637 );
4638 }
4639 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4640 $index = 0;
4641
4642 $maxIndex = count( $sizes ) - 1;
4643 while ( $size >= $boundary && $index < $maxIndex ) {
4644 $index++;
4645 $size /= $boundary;
4646 }
4647
4648 // For small sizes no decimal places necessary
4649 $round = 0;
4650 if ( $index > 1 ) {
4651 // For MB and bigger two decimal places are smarter
4652 $round = 2;
4653 }
4654 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4655
4656 $size = round( $size, $round );
4657 $text = $this->getMessageFromDB( $msg );
4658 return str_replace( '$1', $this->formatNum( $size ), $text );
4659 }
4660
4661 /**
4662 * Format a size in bytes for output, using an appropriate
4663 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4664 *
4665 * This method use base 1024. For base 1000 use formatBitrate(), for
4666 * another base see formatComputingNumbers()
4667 *
4668 * @param int $size Size to format
4669 * @return string Plain text (not HTML)
4670 */
4671 function formatSize( $size ) {
4672 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4673 }
4674
4675 /**
4676 * Make a list item, used by various special pages
4677 *
4678 * @param string $page Page link
4679 * @param string $details HTML safe text between brackets
4680 * @param bool $oppositedm Add the direction mark opposite to your
4681 * language, to display text properly
4682 * @return HTML escaped string
4683 */
4684 function specialList( $page, $details, $oppositedm = true ) {
4685 if ( !$details ) {
4686 return $page;
4687 }
4688
4689 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4690 return
4691 $page .
4692 $dirmark .
4693 $this->msg( 'word-separator' )->escaped() .
4694 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4695 }
4696
4697 /**
4698 * Generate (prev x| next x) (20|50|100...) type links for paging
4699 *
4700 * @param Title $title Title object to link
4701 * @param int $offset
4702 * @param int $limit
4703 * @param array $query Optional URL query parameter string
4704 * @param bool $atend Optional param for specified if this is the last page
4705 * @return string
4706 */
4707 public function viewPrevNext( Title $title, $offset, $limit,
4708 array $query = [], $atend = false
4709 ) {
4710 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4711
4712 # Make 'previous' link
4713 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4714 if ( $offset > 0 ) {
4715 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4716 $query, $prev, 'prevn-title', 'mw-prevlink' );
4717 } else {
4718 $plink = htmlspecialchars( $prev );
4719 }
4720
4721 # Make 'next' link
4722 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4723 if ( $atend ) {
4724 $nlink = htmlspecialchars( $next );
4725 } else {
4726 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4727 $query, $next, 'nextn-title', 'mw-nextlink' );
4728 }
4729
4730 # Make links to set number of items per page
4731 $numLinks = [];
4732 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4733 $numLinks[] = $this->numLink( $title, $offset, $num,
4734 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4735 }
4736
4737 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4738 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4739 }
4740
4741 /**
4742 * Helper function for viewPrevNext() that generates links
4743 *
4744 * @param Title $title Title object to link
4745 * @param int $offset
4746 * @param int $limit
4747 * @param array $query Extra query parameters
4748 * @param string $link Text to use for the link; will be escaped
4749 * @param string $tooltipMsg Name of the message to use as tooltip
4750 * @param string $class Value of the "class" attribute of the link
4751 * @return string HTML fragment
4752 */
4753 private function numLink( Title $title, $offset, $limit, array $query, $link,
4754 $tooltipMsg, $class
4755 ) {
4756 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4757 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4758 ->numParams( $limit )->text();
4759
4760 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4761 'title' => $tooltip, 'class' => $class ], $link );
4762 }
4763
4764 /**
4765 * Get the conversion rule title, if any.
4766 *
4767 * @return string
4768 */
4769 public function getConvRuleTitle() {
4770 return $this->mConverter->getConvRuleTitle();
4771 }
4772
4773 /**
4774 * Get the compiled plural rules for the language
4775 * @since 1.20
4776 * @return array Associative array with plural form, and plural rule as key-value pairs
4777 */
4778 public function getCompiledPluralRules() {
4779 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4780 $fallbacks = Language::getFallbacksFor( $this->mCode );
4781 if ( !$pluralRules ) {
4782 foreach ( $fallbacks as $fallbackCode ) {
4783 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4784 if ( $pluralRules ) {
4785 break;
4786 }
4787 }
4788 }
4789 return $pluralRules;
4790 }
4791
4792 /**
4793 * Get the plural rules for the language
4794 * @since 1.20
4795 * @return array Associative array with plural form number and plural rule as key-value pairs
4796 */
4797 public function getPluralRules() {
4798 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4799 $fallbacks = Language::getFallbacksFor( $this->mCode );
4800 if ( !$pluralRules ) {
4801 foreach ( $fallbacks as $fallbackCode ) {
4802 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4803 if ( $pluralRules ) {
4804 break;
4805 }
4806 }
4807 }
4808 return $pluralRules;
4809 }
4810
4811 /**
4812 * Get the plural rule types for the language
4813 * @since 1.22
4814 * @return array Associative array with plural form number and plural rule type as key-value pairs
4815 */
4816 public function getPluralRuleTypes() {
4817 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4818 $fallbacks = Language::getFallbacksFor( $this->mCode );
4819 if ( !$pluralRuleTypes ) {
4820 foreach ( $fallbacks as $fallbackCode ) {
4821 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4822 if ( $pluralRuleTypes ) {
4823 break;
4824 }
4825 }
4826 }
4827 return $pluralRuleTypes;
4828 }
4829
4830 /**
4831 * Find the index number of the plural rule appropriate for the given number
4832 * @param int $number
4833 * @return int The index number of the plural rule
4834 */
4835 public function getPluralRuleIndexNumber( $number ) {
4836 $pluralRules = $this->getCompiledPluralRules();
4837 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4838 return $form;
4839 }
4840
4841 /**
4842 * Find the plural rule type appropriate for the given number
4843 * For example, if the language is set to Arabic, getPluralType(5) should
4844 * return 'few'.
4845 * @since 1.22
4846 * @param int $number
4847 * @return string The name of the plural rule type, e.g. one, two, few, many
4848 */
4849 public function getPluralRuleType( $number ) {
4850 $index = $this->getPluralRuleIndexNumber( $number );
4851 $pluralRuleTypes = $this->getPluralRuleTypes();
4852 if ( isset( $pluralRuleTypes[$index] ) ) {
4853 return $pluralRuleTypes[$index];
4854 } else {
4855 return 'other';
4856 }
4857 }
4858 }
Wiklou, le Wiki du Wiklou