?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Remove commit() calls from JobQueueDB
[mediawiki.git] / languages / Language.php
blobcb3b4b8832e86e786507d7db588b1441ccb8d387
1 <?php
2 /**
3 * Internationalisation code.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Language
22 */
23
24 /**
25 * @defgroup Language Language
26 */
27
28 use CLDRPluralRuleParser\Evaluator;
29
30 /**
31 * Internationalisation code
32 * @ingroup Language
33 */
34 class Language {
35 /**
36 * @var LanguageConverter
37 */
38 public $mConverter;
39
40 public $mVariants, $mCode, $mLoaded = false;
41 public $mMagicExtensions = [], $mMagicHookDone = false;
42 private $mHtmlCode = null, $mParentLanguage = false;
43
44 public $dateFormatStrings = [];
45 public $mExtendedSpecialPageAliases;
46
47 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
48
49 /**
50 * ReplacementArray object caches
51 */
52 public $transformData = [];
53
54 /**
55 * @var LocalisationCache
56 */
57 static public $dataCache;
58
59 static public $mLangObjCache = [];
60
61 static public $mWeekdayMsgs = [
62 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
63 'friday', 'saturday'
64 ];
65
66 static public $mWeekdayAbbrevMsgs = [
67 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
68 ];
69
70 static public $mMonthMsgs = [
71 'january', 'february', 'march', 'april', 'may_long', 'june',
72 'july', 'august', 'september', 'october', 'november',
73 'december'
74 ];
75 static public $mMonthGenMsgs = [
76 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
77 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
78 'december-gen'
79 ];
80 static public $mMonthAbbrevMsgs = [
81 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
82 'sep', 'oct', 'nov', 'dec'
83 ];
84
85 static public $mIranianCalendarMonthMsgs = [
86 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
87 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
88 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
89 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
90 ];
91
92 static public $mHebrewCalendarMonthMsgs = [
93 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
94 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
95 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
96 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
97 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
98 ];
99
100 static public $mHebrewCalendarMonthGenMsgs = [
101 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
102 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
103 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
104 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
105 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
106 ];
107
108 static public $mHijriCalendarMonthMsgs = [
109 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
110 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
111 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
112 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
113 ];
114
115 /**
116 * @since 1.20
117 * @var array
118 */
119 static public $durationIntervals = [
120 'millennia' => 31556952000,
121 'centuries' => 3155695200,
122 'decades' => 315569520,
123 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
124 'weeks' => 604800,
125 'days' => 86400,
126 'hours' => 3600,
127 'minutes' => 60,
128 'seconds' => 1,
129 ];
130
131 /**
132 * Cache for language fallbacks.
133 * @see Language::getFallbacksIncludingSiteLanguage
134 * @since 1.21
135 * @var array
136 */
137 static private $fallbackLanguageCache = [];
138
139 /**
140 * Cache for language names
141 * @var HashBagOStuff|null
142 */
143 static private $languageNameCache;
144
145 /**
146 * Unicode directional formatting characters, for embedBidi()
147 */
148 static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
149 static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
150 static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
151
152 /**
153 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
154 * - in group 1 if it is LTR
155 * - in group 2 if it is RTL
156 * Does not match if there is no strong directionality codepoint.
157 *
158 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
159 *
160 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
161 * https://git.wikimedia.org/summary/unicodejs.git .
162 */
163 // @codingStandardsIgnoreStart
164 // @codeCoverageIgnoreStart
165 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';
166 // @codeCoverageIgnoreEnd
167 // @codingStandardsIgnoreEnd
168
169 /**
170 * Get a cached or new language object for a given language code
171 * @param string $code
172 * @return Language
173 */
174 static function factory( $code ) {
175 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
176
177 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
178 $code = $wgDummyLanguageCodes[$code];
179 }
180
181 // get the language object to process
182 $langObj = isset( self::$mLangObjCache[$code] )
183 ? self::$mLangObjCache[$code]
184 : self::newFromCode( $code );
185
186 // merge the language object in to get it up front in the cache
187 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
188 // get rid of the oldest ones in case we have an overflow
189 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
190
191 return $langObj;
192 }
193
194 /**
195 * Create a language object for a given language code
196 * @param string $code
197 * @throws MWException
198 * @return Language
199 */
200 protected static function newFromCode( $code ) {
201 if ( !Language::isValidCode( $code ) ) {
202 throw new MWException( "Invalid language code \"$code\"" );
203 }
204
205 if ( !Language::isValidBuiltInCode( $code ) ) {
206 // It's not possible to customise this code with class files, so
207 // just return a Language object. This is to support uselang= hacks.
208 $lang = new Language;
209 $lang->setCode( $code );
210 return $lang;
211 }
212
213 // Check if there is a language class for the code
214 $class = self::classFromCode( $code );
215 if ( class_exists( $class ) ) {
216 $lang = new $class;
217 return $lang;
218 }
219
220 // Keep trying the fallback list until we find an existing class
221 $fallbacks = Language::getFallbacksFor( $code );
222 foreach ( $fallbacks as $fallbackCode ) {
223 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
224 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
225 }
226
227 $class = self::classFromCode( $fallbackCode );
228 if ( class_exists( $class ) ) {
229 $lang = new $class;
230 $lang->setCode( $code );
231 return $lang;
232 }
233 }
234
235 throw new MWException( "Invalid fallback sequence for language '$code'" );
236 }
237
238 /**
239 * Checks whether any localisation is available for that language tag
240 * in MediaWiki (MessagesXx.php exists).
241 *
242 * @param string $code Language tag (in lower case)
243 * @return bool Whether language is supported
244 * @since 1.21
245 */
246 public static function isSupportedLanguage( $code ) {
247 if ( !self::isValidBuiltInCode( $code ) ) {
248 return false;
249 }
250
251 if ( $code === 'qqq' ) {
252 return false;
253 }
254
255 return is_readable( self::getMessagesFileName( $code ) ) ||
256 is_readable( self::getJsonMessagesFileName( $code ) );
257 }
258
259 /**
260 * Returns true if a language code string is a well-formed language tag
261 * according to RFC 5646.
262 * This function only checks well-formedness; it doesn't check that
263 * language, script or variant codes actually exist in the repositories.
264 *
265 * Based on regexes by Mark Davis of the Unicode Consortium:
266 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
267 *
268 * @param string $code
269 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
270 *
271 * @return bool
272 * @since 1.21
273 */
274 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
275 $alpha = '[a-z]';
276 $digit = '[0-9]';
277 $alphanum = '[a-z0-9]';
278 $x = 'x'; # private use singleton
279 $singleton = '[a-wy-z]'; # other singleton
280 $s = $lenient ? '[-_]' : '-';
281
282 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
283 $script = "$alpha{4}"; # ISO 15924
284 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
285 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
286 $extension = "$singleton(?:$s$alphanum{2,8})+";
287 $privateUse = "$x(?:$s$alphanum{1,8})+";
288
289 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
290 # Since these are limited, this is safe even later changes to the registry --
291 # the only oddity is that it might change the type of the tag, and thus
292 # the results from the capturing groups.
293 # http://www.iana.org/assignments/language-subtag-registry
294
295 $grandfathered = "en{$s}GB{$s}oed"
296 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
297 . "|no{$s}(?:bok|nyn)"
298 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
299 . "|zh{$s}min{$s}nan";
300
301 $variantList = "$variant(?:$s$variant)*";
302 $extensionList = "$extension(?:$s$extension)*";
303
304 $langtag = "(?:($language)"
305 . "(?:$s$script)?"
306 . "(?:$s$region)?"
307 . "(?:$s$variantList)?"
308 . "(?:$s$extensionList)?"
309 . "(?:$s$privateUse)?)";
310
311 # The final breakdown, with capturing groups for each of these components
312 # The variants, extensions, grandfathered, and private-use may have interior '-'
313
314 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
315
316 return (bool)preg_match( "/$root/", strtolower( $code ) );
317 }
318
319 /**
320 * Returns true if a language code string is of a valid form, whether or
321 * not it exists. This includes codes which are used solely for
322 * customisation via the MediaWiki namespace.
323 *
324 * @param string $code
325 *
326 * @return bool
327 */
328 public static function isValidCode( $code ) {
329 static $cache = [];
330 if ( !isset( $cache[$code] ) ) {
331 // People think language codes are html safe, so enforce it.
332 // Ideally we should only allow a-zA-Z0-9-
333 // but, .+ and other chars are often used for {{int:}} hacks
334 // see bugs 37564, 37587, 36938
335 $cache[$code] =
336 // Protect against path traversal
337 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
338 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
339 }
340 return $cache[$code];
341 }
342
343 /**
344 * Returns true if a language code is of a valid form for the purposes of
345 * internal customisation of MediaWiki, via Messages*.php or *.json.
346 *
347 * @param string $code
348 *
349 * @throws MWException
350 * @since 1.18
351 * @return bool
352 */
353 public static function isValidBuiltInCode( $code ) {
354
355 if ( !is_string( $code ) ) {
356 if ( is_object( $code ) ) {
357 $addmsg = " of class " . get_class( $code );
358 } else {
359 $addmsg = '';
360 }
361 $type = gettype( $code );
362 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
363 }
364
365 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
366 }
367
368 /**
369 * Returns true if a language code is an IETF tag known to MediaWiki.
370 *
371 * @param string $tag
372 *
373 * @since 1.21
374 * @return bool
375 */
376 public static function isKnownLanguageTag( $tag ) {
377 // Quick escape for invalid input to avoid exceptions down the line
378 // when code tries to process tags which are not valid at all.
379 if ( !self::isValidBuiltInCode( $tag ) ) {
380 return false;
381 }
382
383 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
384 || self::fetchLanguageName( $tag, $tag ) !== ''
385 ) {
386 return true;
387 }
388
389 return false;
390 }
391
392 /**
393 * Get the LocalisationCache instance
394 *
395 * @return LocalisationCache
396 */
397 public static function getLocalisationCache() {
398 if ( is_null( self::$dataCache ) ) {
399 global $wgLocalisationCacheConf;
400 $class = $wgLocalisationCacheConf['class'];
401 self::$dataCache = new $class( $wgLocalisationCacheConf );
402 }
403 return self::$dataCache;
404 }
405
406 function __construct() {
407 $this->mConverter = new FakeConverter( $this );
408 // Set the code to the name of the descendant
409 if ( get_class( $this ) == 'Language' ) {
410 $this->mCode = 'en';
411 } else {
412 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
413 }
414 self::getLocalisationCache();
415 }
416
417 /**
418 * Reduce memory usage
419 */
420 function __destruct() {
421 foreach ( $this as $name => $value ) {
422 unset( $this->$name );
423 }
424 }
425
426 /**
427 * Hook which will be called if this is the content language.
428 * Descendants can use this to register hook functions or modify globals
429 */
430 function initContLang() {
431 }
432
433 /**
434 * @return array
435 * @since 1.19
436 */
437 public function getFallbackLanguages() {
438 return self::getFallbacksFor( $this->mCode );
439 }
440
441 /**
442 * Exports $wgBookstoreListEn
443 * @return array
444 */
445 public function getBookstoreList() {
446 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
447 }
448
449 /**
450 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
451 * available in localised form, it will be included in English.
452 *
453 * @return array
454 */
455 public function getNamespaces() {
456 if ( is_null( $this->namespaceNames ) ) {
457 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
458
459 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
460 $validNamespaces = MWNamespace::getCanonicalNamespaces();
461
462 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
463
464 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
465 if ( $wgMetaNamespaceTalk ) {
466 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
467 } else {
468 $talk = $this->namespaceNames[NS_PROJECT_TALK];
469 $this->namespaceNames[NS_PROJECT_TALK] =
470 $this->fixVariableInNamespace( $talk );
471 }
472
473 # Sometimes a language will be localised but not actually exist on this wiki.
474 foreach ( $this->namespaceNames as $key => $text ) {
475 if ( !isset( $validNamespaces[$key] ) ) {
476 unset( $this->namespaceNames[$key] );
477 }
478 }
479
480 # The above mixing may leave namespaces out of canonical order.
481 # Re-order by namespace ID number...
482 ksort( $this->namespaceNames );
483
484 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
485 }
486
487 return $this->namespaceNames;
488 }
489
490 /**
491 * Arbitrarily set all of the namespace names at once. Mainly used for testing
492 * @param array $namespaces Array of namespaces (id => name)
493 */
494 public function setNamespaces( array $namespaces ) {
495 $this->namespaceNames = $namespaces;
496 $this->mNamespaceIds = null;
497 }
498
499 /**
500 * Resets all of the namespace caches. Mainly used for testing
501 */
502 public function resetNamespaces() {
503 $this->namespaceNames = null;
504 $this->mNamespaceIds = null;
505 $this->namespaceAliases = null;
506 }
507
508 /**
509 * A convenience function that returns getNamespaces() with spaces instead of underscores
510 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
511 *
512 * @return array
513 */
514 public function getFormattedNamespaces() {
515 $ns = $this->getNamespaces();
516 foreach ( $ns as $k => $v ) {
517 $ns[$k] = strtr( $v, '_', ' ' );
518 }
519 return $ns;
520 }
521
522 /**
523 * Get a namespace value by key
524 *
525 * <code>
526 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
527 * echo $mw_ns; // prints 'MediaWiki'
528 * </code>
529 *
530 * @param int $index The array key of the namespace to return
531 * @return string|bool String if the namespace value exists, otherwise false
532 */
533 public function getNsText( $index ) {
534 $ns = $this->getNamespaces();
535 return isset( $ns[$index] ) ? $ns[$index] : false;
536 }
537
538 /**
539 * A convenience function that returns the same thing as
540 * getNsText() except with '_' changed to ' ', useful for
541 * producing output.
542 *
543 * <code>
544 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
545 * echo $mw_ns; // prints 'MediaWiki talk'
546 * </code>
547 *
548 * @param int $index The array key of the namespace to return
549 * @return string Namespace name without underscores (empty string if namespace does not exist)
550 */
551 public function getFormattedNsText( $index ) {
552 $ns = $this->getNsText( $index );
553 return strtr( $ns, '_', ' ' );
554 }
555
556 /**
557 * Returns gender-dependent namespace alias if available.
558 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
559 * @param int $index Namespace index
560 * @param string $gender Gender key (male, female... )
561 * @return string
562 * @since 1.18
563 */
564 public function getGenderNsText( $index, $gender ) {
565 global $wgExtraGenderNamespaces;
566
567 $ns = $wgExtraGenderNamespaces +
568 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
569
570 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
571 }
572
573 /**
574 * Whether this language uses gender-dependent namespace aliases.
575 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
576 * @return bool
577 * @since 1.18
578 */
579 public function needsGenderDistinction() {
580 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
581 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
582 // $wgExtraGenderNamespaces overrides everything
583 return true;
584 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
585 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
586 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
587 return false;
588 } else {
589 // Check what is in i18n files
590 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
591 return count( $aliases ) > 0;
592 }
593 }
594
595 /**
596 * Get a namespace key by value, case insensitive.
597 * Only matches namespace names for the current language, not the
598 * canonical ones defined in Namespace.php.
599 *
600 * @param string $text
601 * @return int|bool An integer if $text is a valid value otherwise false
602 */
603 function getLocalNsIndex( $text ) {
604 $lctext = $this->lc( $text );
605 $ids = $this->getNamespaceIds();
606 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
607 }
608
609 /**
610 * @return array
611 */
612 public function getNamespaceAliases() {
613 if ( is_null( $this->namespaceAliases ) ) {
614 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
615 if ( !$aliases ) {
616 $aliases = [];
617 } else {
618 foreach ( $aliases as $name => $index ) {
619 if ( $index === NS_PROJECT_TALK ) {
620 unset( $aliases[$name] );
621 $name = $this->fixVariableInNamespace( $name );
622 $aliases[$name] = $index;
623 }
624 }
625 }
626
627 global $wgExtraGenderNamespaces;
628 $genders = $wgExtraGenderNamespaces +
629 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
630 foreach ( $genders as $index => $forms ) {
631 foreach ( $forms as $alias ) {
632 $aliases[$alias] = $index;
633 }
634 }
635
636 # Also add converted namespace names as aliases, to avoid confusion.
637 $convertedNames = [];
638 foreach ( $this->getVariants() as $variant ) {
639 if ( $variant === $this->mCode ) {
640 continue;
641 }
642 foreach ( $this->getNamespaces() as $ns => $_ ) {
643 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
644 }
645 }
646
647 $this->namespaceAliases = $aliases + $convertedNames;
648 }
649
650 return $this->namespaceAliases;
651 }
652
653 /**
654 * @return array
655 */
656 public function getNamespaceIds() {
657 if ( is_null( $this->mNamespaceIds ) ) {
658 global $wgNamespaceAliases;
659 # Put namespace names and aliases into a hashtable.
660 # If this is too slow, then we should arrange it so that it is done
661 # before caching. The catch is that at pre-cache time, the above
662 # class-specific fixup hasn't been done.
663 $this->mNamespaceIds = [];
664 foreach ( $this->getNamespaces() as $index => $name ) {
665 $this->mNamespaceIds[$this->lc( $name )] = $index;
666 }
667 foreach ( $this->getNamespaceAliases() as $name => $index ) {
668 $this->mNamespaceIds[$this->lc( $name )] = $index;
669 }
670 if ( $wgNamespaceAliases ) {
671 foreach ( $wgNamespaceAliases as $name => $index ) {
672 $this->mNamespaceIds[$this->lc( $name )] = $index;
673 }
674 }
675 }
676 return $this->mNamespaceIds;
677 }
678
679 /**
680 * Get a namespace key by value, case insensitive. Canonical namespace
681 * names override custom ones defined for the current language.
682 *
683 * @param string $text
684 * @return int|bool An integer if $text is a valid value otherwise false
685 */
686 public function getNsIndex( $text ) {
687 $lctext = $this->lc( $text );
688 $ns = MWNamespace::getCanonicalIndex( $lctext );
689 if ( $ns !== null ) {
690 return $ns;
691 }
692 $ids = $this->getNamespaceIds();
693 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
694 }
695
696 /**
697 * short names for language variants used for language conversion links.
698 *
699 * @param string $code
700 * @param bool $usemsg Use the "variantname-xyz" message if it exists
701 * @return string
702 */
703 public function getVariantname( $code, $usemsg = true ) {
704 $msg = "variantname-$code";
705 if ( $usemsg && wfMessage( $msg )->exists() ) {
706 return $this->getMessageFromDB( $msg );
707 }
708 $name = self::fetchLanguageName( $code );
709 if ( $name ) {
710 return $name; # if it's defined as a language name, show that
711 } else {
712 # otherwise, output the language code
713 return $code;
714 }
715 }
716
717 /**
718 * @return array
719 */
720 public function getDatePreferences() {
721 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
722 }
723
724 /**
725 * @return array
726 */
727 function getDateFormats() {
728 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
729 }
730
731 /**
732 * @return array|string
733 */
734 public function getDefaultDateFormat() {
735 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
736 if ( $df === 'dmy or mdy' ) {
737 global $wgAmericanDates;
738 return $wgAmericanDates ? 'mdy' : 'dmy';
739 } else {
740 return $df;
741 }
742 }
743
744 /**
745 * @return array
746 */
747 public function getDatePreferenceMigrationMap() {
748 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
749 }
750
751 /**
752 * @param string $image
753 * @return array|null
754 */
755 function getImageFile( $image ) {
756 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
757 }
758
759 /**
760 * @return array
761 * @since 1.24
762 */
763 public function getImageFiles() {
764 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
765 }
766
767 /**
768 * @return array
769 */
770 public function getExtraUserToggles() {
771 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
772 }
773
774 /**
775 * @param string $tog
776 * @return string
777 */
778 function getUserToggle( $tog ) {
779 return $this->getMessageFromDB( "tog-$tog" );
780 }
781
782 /**
783 * Get an array of language names, indexed by code.
784 * @param null|string $inLanguage Code of language in which to return the names
785 * Use null for autonyms (native names)
786 * @param string $include One of:
787 * 'all' all available languages
788 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
789 * 'mwfile' only if the language is in 'mw' *and* has a message file
790 * @return array Language code => language name
791 * @since 1.20
792 */
793 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
794 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
795 $cacheKey .= ":$include";
796 if ( self::$languageNameCache === null ) {
797 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
798 }
799
800 $ret = self::$languageNameCache->get( $cacheKey );
801 if ( !$ret ) {
802 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
803 self::$languageNameCache->set( $cacheKey, $ret );
804 }
805 return $ret;
806 }
807
808 /**
809 * Uncached helper for fetchLanguageNames
810 * @param null|string $inLanguage Code of language in which to return the names
811 * Use null for autonyms (native names)
812 * @param string $include One of:
813 * 'all' all available languages
814 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
815 * 'mwfile' only if the language is in 'mw' *and* has a message file
816 * @return array Language code => language name
817 */
818 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
819 global $wgExtraLanguageNames;
820
821 // If passed an invalid language code to use, fallback to en
822 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
823 $inLanguage = 'en';
824 }
825
826 $names = [];
827
828 if ( $inLanguage ) {
829 # TODO: also include when $inLanguage is null, when this code is more efficient
830 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
831 }
832
833 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
834 foreach ( $mwNames as $mwCode => $mwName ) {
835 # - Prefer own MediaWiki native name when not using the hook
836 # - For other names just add if not added through the hook
837 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
838 $names[$mwCode] = $mwName;
839 }
840 }
841
842 if ( $include === 'all' ) {
843 ksort( $names );
844 return $names;
845 }
846
847 $returnMw = [];
848 $coreCodes = array_keys( $mwNames );
849 foreach ( $coreCodes as $coreCode ) {
850 $returnMw[$coreCode] = $names[$coreCode];
851 }
852
853 if ( $include === 'mwfile' ) {
854 $namesMwFile = [];
855 # We do this using a foreach over the codes instead of a directory
856 # loop so that messages files in extensions will work correctly.
857 foreach ( $returnMw as $code => $value ) {
858 if ( is_readable( self::getMessagesFileName( $code ) )
859 || is_readable( self::getJsonMessagesFileName( $code ) )
860 ) {
861 $namesMwFile[$code] = $names[$code];
862 }
863 }
864
865 ksort( $namesMwFile );
866 return $namesMwFile;
867 }
868
869 ksort( $returnMw );
870 # 'mw' option; default if it's not one of the other two options (all/mwfile)
871 return $returnMw;
872 }
873
874 /**
875 * @param string $code The code of the language for which to get the name
876 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
877 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
878 * @return string Language name or empty
879 * @since 1.20
880 */
881 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
882 $code = strtolower( $code );
883 $array = self::fetchLanguageNames( $inLanguage, $include );
884 return !array_key_exists( $code, $array ) ? '' : $array[$code];
885 }
886
887 /**
888 * Get a message from the MediaWiki namespace.
889 *
890 * @param string $msg Message name
891 * @return string
892 */
893 public function getMessageFromDB( $msg ) {
894 return $this->msg( $msg )->text();
895 }
896
897 /**
898 * Get message object in this language. Only for use inside this class.
899 *
900 * @param string $msg Message name
901 * @return Message
902 */
903 protected function msg( $msg ) {
904 return wfMessage( $msg )->inLanguage( $this );
905 }
906
907 /**
908 * @param string $key
909 * @return string
910 */
911 public function getMonthName( $key ) {
912 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
913 }
914
915 /**
916 * @return array
917 */
918 public function getMonthNamesArray() {
919 $monthNames = [ '' ];
920 for ( $i = 1; $i < 13; $i++ ) {
921 $monthNames[] = $this->getMonthName( $i );
922 }
923 return $monthNames;
924 }
925
926 /**
927 * @param string $key
928 * @return string
929 */
930 public function getMonthNameGen( $key ) {
931 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
932 }
933
934 /**
935 * @param string $key
936 * @return string
937 */
938 public function getMonthAbbreviation( $key ) {
939 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
940 }
941
942 /**
943 * @return array
944 */
945 public function getMonthAbbreviationsArray() {
946 $monthNames = [ '' ];
947 for ( $i = 1; $i < 13; $i++ ) {
948 $monthNames[] = $this->getMonthAbbreviation( $i );
949 }
950 return $monthNames;
951 }
952
953 /**
954 * @param string $key
955 * @return string
956 */
957 public function getWeekdayName( $key ) {
958 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
959 }
960
961 /**
962 * @param string $key
963 * @return string
964 */
965 function getWeekdayAbbreviation( $key ) {
966 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
967 }
968
969 /**
970 * @param string $key
971 * @return string
972 */
973 function getIranianCalendarMonthName( $key ) {
974 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
975 }
976
977 /**
978 * @param string $key
979 * @return string
980 */
981 function getHebrewCalendarMonthName( $key ) {
982 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
983 }
984
985 /**
986 * @param string $key
987 * @return string
988 */
989 function getHebrewCalendarMonthNameGen( $key ) {
990 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
991 }
992
993 /**
994 * @param string $key
995 * @return string
996 */
997 function getHijriCalendarMonthName( $key ) {
998 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
999 }
1000
1001 /**
1002 * Pass through result from $dateTimeObj->format()
1003 * @param DateTime|bool|null &$dateTimeObj
1004 * @param string $ts
1005 * @param DateTimeZone|bool|null $zone
1006 * @param string $code
1007 * @return string
1008 */
1009 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1010 if ( !$dateTimeObj ) {
1011 $dateTimeObj = DateTime::createFromFormat(
1012 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1013 );
1014 }
1015 return $dateTimeObj->format( $code );
1016 }
1017
1018 /**
1019 * This is a workalike of PHP's date() function, but with better
1020 * internationalisation, a reduced set of format characters, and a better
1021 * escaping format.
1022 *
1023 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1024 * the PHP manual for definitions. There are a number of extensions, which
1025 * start with "x":
1026 *
1027 * xn Do not translate digits of the next numeric format character
1028 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1029 * xr Use roman numerals for the next numeric format character
1030 * xh Use hebrew numerals for the next numeric format character
1031 * xx Literal x
1032 * xg Genitive month name
1033 *
1034 * xij j (day number) in Iranian calendar
1035 * xiF F (month name) in Iranian calendar
1036 * xin n (month number) in Iranian calendar
1037 * xiy y (two digit year) in Iranian calendar
1038 * xiY Y (full year) in Iranian calendar
1039 * xit t (days in month) in Iranian calendar
1040 *
1041 * xjj j (day number) in Hebrew calendar
1042 * xjF F (month name) in Hebrew calendar
1043 * xjt t (days in month) in Hebrew calendar
1044 * xjx xg (genitive month name) in Hebrew calendar
1045 * xjn n (month number) in Hebrew calendar
1046 * xjY Y (full year) in Hebrew calendar
1047 *
1048 * xmj j (day number) in Hijri calendar
1049 * xmF F (month name) in Hijri calendar
1050 * xmn n (month number) in Hijri calendar
1051 * xmY Y (full year) in Hijri calendar
1052 *
1053 * xkY Y (full year) in Thai solar calendar. Months and days are
1054 * identical to the Gregorian calendar
1055 * xoY Y (full year) in Minguo calendar or Juche year.
1056 * Months and days are identical to the
1057 * Gregorian calendar
1058 * xtY Y (full year) in Japanese nengo. Months and days are
1059 * identical to the Gregorian calendar
1060 *
1061 * Characters enclosed in double quotes will be considered literal (with
1062 * the quotes themselves removed). Unmatched quotes will be considered
1063 * literal quotes. Example:
1064 *
1065 * "The month is" F => The month is January
1066 * i's" => 20'11"
1067 *
1068 * Backslash escaping is also supported.
1069 *
1070 * Input timestamp is assumed to be pre-normalized to the desired local
1071 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1072 * $ts is UTC if $zone is not given.
1073 *
1074 * @param string $format
1075 * @param string $ts 14-character timestamp
1076 * YYYYMMDDHHMMSS
1077 * 01234567890123
1078 * @param DateTimeZone $zone Timezone of $ts
1079 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1080 * Only makes sense if $ts is the current time.
1081 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1082 *
1083 * @throws MWException
1084 * @return string
1085 */
1086 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1087 $s = '';
1088 $raw = false;
1089 $roman = false;
1090 $hebrewNum = false;
1091 $dateTimeObj = false;
1092 $rawToggle = false;
1093 $iranian = false;
1094 $hebrew = false;
1095 $hijri = false;
1096 $thai = false;
1097 $minguo = false;
1098 $tenno = false;
1099
1100 $usedSecond = false;
1101 $usedMinute = false;
1102 $usedHour = false;
1103 $usedAMPM = false;
1104 $usedDay = false;
1105 $usedWeek = false;
1106 $usedMonth = false;
1107 $usedYear = false;
1108 $usedISOYear = false;
1109 $usedIsLeapYear = false;
1110
1111 $usedHebrewMonth = false;
1112 $usedIranianMonth = false;
1113 $usedHijriMonth = false;
1114 $usedHebrewYear = false;
1115 $usedIranianYear = false;
1116 $usedHijriYear = false;
1117 $usedTennoYear = false;
1118
1119 if ( strlen( $ts ) !== 14 ) {
1120 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1121 }
1122
1123 if ( !ctype_digit( $ts ) ) {
1124 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1125 }
1126
1127 $formatLength = strlen( $format );
1128 for ( $p = 0; $p < $formatLength; $p++ ) {
1129 $num = false;
1130 $code = $format[$p];
1131 if ( $code == 'x' && $p < $formatLength - 1 ) {
1132 $code .= $format[++$p];
1133 }
1134
1135 if ( ( $code === 'xi'
1136 || $code === 'xj'
1137 || $code === 'xk'
1138 || $code === 'xm'
1139 || $code === 'xo'
1140 || $code === 'xt' )
1141 && $p < $formatLength - 1 ) {
1142 $code .= $format[++$p];
1143 }
1144
1145 switch ( $code ) {
1146 case 'xx':
1147 $s .= 'x';
1148 break;
1149 case 'xn':
1150 $raw = true;
1151 break;
1152 case 'xN':
1153 $rawToggle = !$rawToggle;
1154 break;
1155 case 'xr':
1156 $roman = true;
1157 break;
1158 case 'xh':
1159 $hebrewNum = true;
1160 break;
1161 case 'xg':
1162 $usedMonth = true;
1163 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1164 break;
1165 case 'xjx':
1166 $usedHebrewMonth = true;
1167 if ( !$hebrew ) {
1168 $hebrew = self::tsToHebrew( $ts );
1169 }
1170 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1171 break;
1172 case 'd':
1173 $usedDay = true;
1174 $num = substr( $ts, 6, 2 );
1175 break;
1176 case 'D':
1177 $usedDay = true;
1178 $s .= $this->getWeekdayAbbreviation(
1179 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1180 );
1181 break;
1182 case 'j':
1183 $usedDay = true;
1184 $num = intval( substr( $ts, 6, 2 ) );
1185 break;
1186 case 'xij':
1187 $usedDay = true;
1188 if ( !$iranian ) {
1189 $iranian = self::tsToIranian( $ts );
1190 }
1191 $num = $iranian[2];
1192 break;
1193 case 'xmj':
1194 $usedDay = true;
1195 if ( !$hijri ) {
1196 $hijri = self::tsToHijri( $ts );
1197 }
1198 $num = $hijri[2];
1199 break;
1200 case 'xjj':
1201 $usedDay = true;
1202 if ( !$hebrew ) {
1203 $hebrew = self::tsToHebrew( $ts );
1204 }
1205 $num = $hebrew[2];
1206 break;
1207 case 'l':
1208 $usedDay = true;
1209 $s .= $this->getWeekdayName(
1210 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1211 );
1212 break;
1213 case 'F':
1214 $usedMonth = true;
1215 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1216 break;
1217 case 'xiF':
1218 $usedIranianMonth = true;
1219 if ( !$iranian ) {
1220 $iranian = self::tsToIranian( $ts );
1221 }
1222 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1223 break;
1224 case 'xmF':
1225 $usedHijriMonth = true;
1226 if ( !$hijri ) {
1227 $hijri = self::tsToHijri( $ts );
1228 }
1229 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1230 break;
1231 case 'xjF':
1232 $usedHebrewMonth = true;
1233 if ( !$hebrew ) {
1234 $hebrew = self::tsToHebrew( $ts );
1235 }
1236 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1237 break;
1238 case 'm':
1239 $usedMonth = true;
1240 $num = substr( $ts, 4, 2 );
1241 break;
1242 case 'M':
1243 $usedMonth = true;
1244 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1245 break;
1246 case 'n':
1247 $usedMonth = true;
1248 $num = intval( substr( $ts, 4, 2 ) );
1249 break;
1250 case 'xin':
1251 $usedIranianMonth = true;
1252 if ( !$iranian ) {
1253 $iranian = self::tsToIranian( $ts );
1254 }
1255 $num = $iranian[1];
1256 break;
1257 case 'xmn':
1258 $usedHijriMonth = true;
1259 if ( !$hijri ) {
1260 $hijri = self::tsToHijri( $ts );
1261 }
1262 $num = $hijri[1];
1263 break;
1264 case 'xjn':
1265 $usedHebrewMonth = true;
1266 if ( !$hebrew ) {
1267 $hebrew = self::tsToHebrew( $ts );
1268 }
1269 $num = $hebrew[1];
1270 break;
1271 case 'xjt':
1272 $usedHebrewMonth = true;
1273 if ( !$hebrew ) {
1274 $hebrew = self::tsToHebrew( $ts );
1275 }
1276 $num = $hebrew[3];
1277 break;
1278 case 'Y':
1279 $usedYear = true;
1280 $num = substr( $ts, 0, 4 );
1281 break;
1282 case 'xiY':
1283 $usedIranianYear = true;
1284 if ( !$iranian ) {
1285 $iranian = self::tsToIranian( $ts );
1286 }
1287 $num = $iranian[0];
1288 break;
1289 case 'xmY':
1290 $usedHijriYear = true;
1291 if ( !$hijri ) {
1292 $hijri = self::tsToHijri( $ts );
1293 }
1294 $num = $hijri[0];
1295 break;
1296 case 'xjY':
1297 $usedHebrewYear = true;
1298 if ( !$hebrew ) {
1299 $hebrew = self::tsToHebrew( $ts );
1300 }
1301 $num = $hebrew[0];
1302 break;
1303 case 'xkY':
1304 $usedYear = true;
1305 if ( !$thai ) {
1306 $thai = self::tsToYear( $ts, 'thai' );
1307 }
1308 $num = $thai[0];
1309 break;
1310 case 'xoY':
1311 $usedYear = true;
1312 if ( !$minguo ) {
1313 $minguo = self::tsToYear( $ts, 'minguo' );
1314 }
1315 $num = $minguo[0];
1316 break;
1317 case 'xtY':
1318 $usedTennoYear = true;
1319 if ( !$tenno ) {
1320 $tenno = self::tsToYear( $ts, 'tenno' );
1321 }
1322 $num = $tenno[0];
1323 break;
1324 case 'y':
1325 $usedYear = true;
1326 $num = substr( $ts, 2, 2 );
1327 break;
1328 case 'xiy':
1329 $usedIranianYear = true;
1330 if ( !$iranian ) {
1331 $iranian = self::tsToIranian( $ts );
1332 }
1333 $num = substr( $iranian[0], -2 );
1334 break;
1335 case 'xit':
1336 $usedIranianYear = true;
1337 if ( !$iranian ) {
1338 $iranian = self::tsToIranian( $ts );
1339 }
1340 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1341 break;
1342 case 'a':
1343 $usedAMPM = true;
1344 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1345 break;
1346 case 'A':
1347 $usedAMPM = true;
1348 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1349 break;
1350 case 'g':
1351 $usedHour = true;
1352 $h = substr( $ts, 8, 2 );
1353 $num = $h % 12 ? $h % 12 : 12;
1354 break;
1355 case 'G':
1356 $usedHour = true;
1357 $num = intval( substr( $ts, 8, 2 ) );
1358 break;
1359 case 'h':
1360 $usedHour = true;
1361 $h = substr( $ts, 8, 2 );
1362 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1363 break;
1364 case 'H':
1365 $usedHour = true;
1366 $num = substr( $ts, 8, 2 );
1367 break;
1368 case 'i':
1369 $usedMinute = true;
1370 $num = substr( $ts, 10, 2 );
1371 break;
1372 case 's':
1373 $usedSecond = true;
1374 $num = substr( $ts, 12, 2 );
1375 break;
1376 case 'c':
1377 case 'r':
1378 $usedSecond = true;
1379 // fall through
1380 case 'e':
1381 case 'O':
1382 case 'P':
1383 case 'T':
1384 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1385 break;
1386 case 'w':
1387 case 'N':
1388 case 'z':
1389 $usedDay = true;
1390 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1391 break;
1392 case 'W':
1393 $usedWeek = true;
1394 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1395 break;
1396 case 't':
1397 $usedMonth = true;
1398 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1399 break;
1400 case 'L':
1401 $usedIsLeapYear = true;
1402 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1403 break;
1404 case 'o':
1405 $usedISOYear = true;
1406 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1407 break;
1408 case 'U':
1409 $usedSecond = true;
1410 // fall through
1411 case 'I':
1412 case 'Z':
1413 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1414 break;
1415 case '\\':
1416 # Backslash escaping
1417 if ( $p < $formatLength - 1 ) {
1418 $s .= $format[++$p];
1419 } else {
1420 $s .= '\\';
1421 }
1422 break;
1423 case '"':
1424 # Quoted literal
1425 if ( $p < $formatLength - 1 ) {
1426 $endQuote = strpos( $format, '"', $p + 1 );
1427 if ( $endQuote === false ) {
1428 # No terminating quote, assume literal "
1429 $s .= '"';
1430 } else {
1431 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1432 $p = $endQuote;
1433 }
1434 } else {
1435 # Quote at end of string, assume literal "
1436 $s .= '"';
1437 }
1438 break;
1439 default:
1440 $s .= $format[$p];
1441 }
1442 if ( $num !== false ) {
1443 if ( $rawToggle || $raw ) {
1444 $s .= $num;
1445 $raw = false;
1446 } elseif ( $roman ) {
1447 $s .= Language::romanNumeral( $num );
1448 $roman = false;
1449 } elseif ( $hebrewNum ) {
1450 $s .= self::hebrewNumeral( $num );
1451 $hebrewNum = false;
1452 } else {
1453 $s .= $this->formatNum( $num, true );
1454 }
1455 }
1456 }
1457
1458 if ( $ttl === 'unused' ) {
1459 // No need to calculate the TTL, the caller wont use it anyway.
1460 } elseif ( $usedSecond ) {
1461 $ttl = 1;
1462 } elseif ( $usedMinute ) {
1463 $ttl = 60 - substr( $ts, 12, 2 );
1464 } elseif ( $usedHour ) {
1465 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1466 } elseif ( $usedAMPM ) {
1467 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1468 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1469 } elseif (
1470 $usedDay ||
1471 $usedHebrewMonth ||
1472 $usedIranianMonth ||
1473 $usedHijriMonth ||
1474 $usedHebrewYear ||
1475 $usedIranianYear ||
1476 $usedHijriYear ||
1477 $usedTennoYear
1478 ) {
1479 // @todo Someone who understands the non-Gregorian calendars
1480 // should write proper logic for them so that they don't need purged every day.
1481 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1482 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1483 } else {
1484 $possibleTtls = [];
1485 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1486 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1487 if ( $usedWeek ) {
1488 $possibleTtls[] =
1489 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1490 $timeRemainingInDay;
1491 } elseif ( $usedISOYear ) {
1492 // December 28th falls on the last ISO week of the year, every year.
1493 // The last ISO week of a year can be 52 or 53.
1494 $lastWeekOfISOYear = DateTime::createFromFormat(
1495 'Ymd',
1496 substr( $ts, 0, 4 ) . '1228',
1497 $zone ?: new DateTimeZone( 'UTC' )
1498 )->format( 'W' );
1499 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1500 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1501 $timeRemainingInWeek =
1502 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1503 + $timeRemainingInDay;
1504 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1505 }
1506
1507 if ( $usedMonth ) {
1508 $possibleTtls[] =
1509 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1510 substr( $ts, 6, 2 ) ) * 86400
1511 + $timeRemainingInDay;
1512 } elseif ( $usedYear ) {
1513 $possibleTtls[] =
1514 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1515 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1516 + $timeRemainingInDay;
1517 } elseif ( $usedIsLeapYear ) {
1518 $year = substr( $ts, 0, 4 );
1519 $timeRemainingInYear =
1520 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1521 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1522 + $timeRemainingInDay;
1523 $mod = $year % 4;
1524 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1525 // this isn't a leap year. see when the next one starts
1526 $nextCandidate = $year - $mod + 4;
1527 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1528 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1529 $timeRemainingInYear;
1530 } else {
1531 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1532 $timeRemainingInYear;
1533 }
1534 } else {
1535 // this is a leap year, so the next year isn't
1536 $possibleTtls[] = $timeRemainingInYear;
1537 }
1538 }
1539
1540 if ( $possibleTtls ) {
1541 $ttl = min( $possibleTtls );
1542 }
1543 }
1544
1545 return $s;
1546 }
1547
1548 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1549 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1550
1551 /**
1552 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1553 * Gregorian dates to Iranian dates. Originally written in C, it
1554 * is released under the terms of GNU Lesser General Public
1555 * License. Conversion to PHP was performed by Niklas Laxström.
1556 *
1557 * Link: http://www.farsiweb.info/jalali/jalali.c
1558 *
1559 * @param string $ts
1560 *
1561 * @return int[]
1562 */
1563 private static function tsToIranian( $ts ) {
1564 $gy = substr( $ts, 0, 4 ) -1600;
1565 $gm = substr( $ts, 4, 2 ) -1;
1566 $gd = substr( $ts, 6, 2 ) -1;
1567
1568 # Days passed from the beginning (including leap years)
1569 $gDayNo = 365 * $gy
1570 + floor( ( $gy + 3 ) / 4 )
1571 - floor( ( $gy + 99 ) / 100 )
1572 + floor( ( $gy + 399 ) / 400 );
1573
1574 // Add days of the past months of this year
1575 for ( $i = 0; $i < $gm; $i++ ) {
1576 $gDayNo += self::$GREG_DAYS[$i];
1577 }
1578
1579 // Leap years
1580 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1581 $gDayNo++;
1582 }
1583
1584 // Days passed in current month
1585 $gDayNo += (int)$gd;
1586
1587 $jDayNo = $gDayNo - 79;
1588
1589 $jNp = floor( $jDayNo / 12053 );
1590 $jDayNo %= 12053;
1591
1592 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1593 $jDayNo %= 1461;
1594
1595 if ( $jDayNo >= 366 ) {
1596 $jy += floor( ( $jDayNo - 1 ) / 365 );
1597 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1598 }
1599
1600 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1601 $jDayNo -= self::$IRANIAN_DAYS[$i];
1602 }
1603
1604 $jm = $i + 1;
1605 $jd = $jDayNo + 1;
1606
1607 return [ $jy, $jm, $jd ];
1608 }
1609
1610 /**
1611 * Converting Gregorian dates to Hijri dates.
1612 *
1613 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1614 *
1615 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1616 *
1617 * @param string $ts
1618 *
1619 * @return int[]
1620 */
1621 private static function tsToHijri( $ts ) {
1622 $year = substr( $ts, 0, 4 );
1623 $month = substr( $ts, 4, 2 );
1624 $day = substr( $ts, 6, 2 );
1625
1626 $zyr = $year;
1627 $zd = $day;
1628 $zm = $month;
1629 $zy = $zyr;
1630
1631 if (
1632 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1633 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1634 ) {
1635 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1636 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1637 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1638 $zd - 32075;
1639 } else {
1640 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1641 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1642 }
1643
1644 $zl = $zjd -1948440 + 10632;
1645 $zn = (int)( ( $zl - 1 ) / 10631 );
1646 $zl = $zl - 10631 * $zn + 354;
1647 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1648 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1649 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1650 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1651 $zm = (int)( ( 24 * $zl ) / 709 );
1652 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1653 $zy = 30 * $zn + $zj - 30;
1654
1655 return [ $zy, $zm, $zd ];
1656 }
1657
1658 /**
1659 * Converting Gregorian dates to Hebrew dates.
1660 *
1661 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1662 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1663 * to translate the relevant functions into PHP and release them under
1664 * GNU GPL.
1665 *
1666 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1667 * and Adar II is 14. In a non-leap year, Adar is 6.
1668 *
1669 * @param string $ts
1670 *
1671 * @return int[]
1672 */
1673 private static function tsToHebrew( $ts ) {
1674 # Parse date
1675 $year = substr( $ts, 0, 4 );
1676 $month = substr( $ts, 4, 2 );
1677 $day = substr( $ts, 6, 2 );
1678
1679 # Calculate Hebrew year
1680 $hebrewYear = $year + 3760;
1681
1682 # Month number when September = 1, August = 12
1683 $month += 4;
1684 if ( $month > 12 ) {
1685 # Next year
1686 $month -= 12;
1687 $year++;
1688 $hebrewYear++;
1689 }
1690
1691 # Calculate day of year from 1 September
1692 $dayOfYear = $day;
1693 for ( $i = 1; $i < $month; $i++ ) {
1694 if ( $i == 6 ) {
1695 # February
1696 $dayOfYear += 28;
1697 # Check if the year is leap
1698 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1699 $dayOfYear++;
1700 }
1701 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1702 $dayOfYear += 30;
1703 } else {
1704 $dayOfYear += 31;
1705 }
1706 }
1707
1708 # Calculate the start of the Hebrew year
1709 $start = self::hebrewYearStart( $hebrewYear );
1710
1711 # Calculate next year's start
1712 if ( $dayOfYear <= $start ) {
1713 # Day is before the start of the year - it is the previous year
1714 # Next year's start
1715 $nextStart = $start;
1716 # Previous year
1717 $year--;
1718 $hebrewYear--;
1719 # Add days since previous year's 1 September
1720 $dayOfYear += 365;
1721 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1722 # Leap year
1723 $dayOfYear++;
1724 }
1725 # Start of the new (previous) year
1726 $start = self::hebrewYearStart( $hebrewYear );
1727 } else {
1728 # Next year's start
1729 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1730 }
1731
1732 # Calculate Hebrew day of year
1733 $hebrewDayOfYear = $dayOfYear - $start;
1734
1735 # Difference between year's days
1736 $diff = $nextStart - $start;
1737 # Add 12 (or 13 for leap years) days to ignore the difference between
1738 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1739 # difference is only about the year type
1740 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1741 $diff += 13;
1742 } else {
1743 $diff += 12;
1744 }
1745
1746 # Check the year pattern, and is leap year
1747 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1748 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1749 # and non-leap years
1750 $yearPattern = $diff % 30;
1751 # Check if leap year
1752 $isLeap = $diff >= 30;
1753
1754 # Calculate day in the month from number of day in the Hebrew year
1755 # Don't check Adar - if the day is not in Adar, we will stop before;
1756 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1757 $hebrewDay = $hebrewDayOfYear;
1758 $hebrewMonth = 1;
1759 $days = 0;
1760 while ( $hebrewMonth <= 12 ) {
1761 # Calculate days in this month
1762 if ( $isLeap && $hebrewMonth == 6 ) {
1763 # Adar in a leap year
1764 if ( $isLeap ) {
1765 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1766 $days = 30;
1767 if ( $hebrewDay <= $days ) {
1768 # Day in Adar I
1769 $hebrewMonth = 13;
1770 } else {
1771 # Subtract the days of Adar I
1772 $hebrewDay -= $days;
1773 # Try Adar II
1774 $days = 29;
1775 if ( $hebrewDay <= $days ) {
1776 # Day in Adar II
1777 $hebrewMonth = 14;
1778 }
1779 }
1780 }
1781 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1782 # Cheshvan in a complete year (otherwise as the rule below)
1783 $days = 30;
1784 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1785 # Kislev in an incomplete year (otherwise as the rule below)
1786 $days = 29;
1787 } else {
1788 # Odd months have 30 days, even have 29
1789 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1790 }
1791 if ( $hebrewDay <= $days ) {
1792 # In the current month
1793 break;
1794 } else {
1795 # Subtract the days of the current month
1796 $hebrewDay -= $days;
1797 # Try in the next month
1798 $hebrewMonth++;
1799 }
1800 }
1801
1802 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1803 }
1804
1805 /**
1806 * This calculates the Hebrew year start, as days since 1 September.
1807 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1808 * Used for Hebrew date.
1809 *
1810 * @param int $year
1811 *
1812 * @return string
1813 */
1814 private static function hebrewYearStart( $year ) {
1815 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1816 $b = intval( ( $year - 1 ) % 4 );
1817 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1818 if ( $m < 0 ) {
1819 $m--;
1820 }
1821 $Mar = intval( $m );
1822 if ( $m < 0 ) {
1823 $m++;
1824 }
1825 $m -= $Mar;
1826
1827 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1828 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1829 $Mar++;
1830 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1831 $Mar += 2;
1832 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1833 $Mar++;
1834 }
1835
1836 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1837 return $Mar;
1838 }
1839
1840 /**
1841 * Algorithm to convert Gregorian dates to Thai solar dates,
1842 * Minguo dates or Minguo dates.
1843 *
1844 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1845 * http://en.wikipedia.org/wiki/Minguo_calendar
1846 * http://en.wikipedia.org/wiki/Japanese_era_name
1847 *
1848 * @param string $ts 14-character timestamp
1849 * @param string $cName Calender name
1850 * @return array Converted year, month, day
1851 */
1852 private static function tsToYear( $ts, $cName ) {
1853 $gy = substr( $ts, 0, 4 );
1854 $gm = substr( $ts, 4, 2 );
1855 $gd = substr( $ts, 6, 2 );
1856
1857 if ( !strcmp( $cName, 'thai' ) ) {
1858 # Thai solar dates
1859 # Add 543 years to the Gregorian calendar
1860 # Months and days are identical
1861 $gy_offset = $gy + 543;
1862 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1863 # Minguo dates
1864 # Deduct 1911 years from the Gregorian calendar
1865 # Months and days are identical
1866 $gy_offset = $gy - 1911;
1867 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1868 # Nengō dates up to Meiji period
1869 # Deduct years from the Gregorian calendar
1870 # depending on the nengo periods
1871 # Months and days are identical
1872 if ( ( $gy < 1912 )
1873 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1874 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1875 ) {
1876 # Meiji period
1877 $gy_gannen = $gy - 1868 + 1;
1878 $gy_offset = $gy_gannen;
1879 if ( $gy_gannen == 1 ) {
1880 $gy_offset = '元';
1881 }
1882 $gy_offset = '明治' . $gy_offset;
1883 } elseif (
1884 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1885 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1886 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1887 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1888 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1889 ) {
1890 # Taishō period
1891 $gy_gannen = $gy - 1912 + 1;
1892 $gy_offset = $gy_gannen;
1893 if ( $gy_gannen == 1 ) {
1894 $gy_offset = '元';
1895 }
1896 $gy_offset = '大正' . $gy_offset;
1897 } elseif (
1898 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1899 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1900 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1901 ) {
1902 # Shōwa period
1903 $gy_gannen = $gy - 1926 + 1;
1904 $gy_offset = $gy_gannen;
1905 if ( $gy_gannen == 1 ) {
1906 $gy_offset = '元';
1907 }
1908 $gy_offset = '昭和' . $gy_offset;
1909 } else {
1910 # Heisei period
1911 $gy_gannen = $gy - 1989 + 1;
1912 $gy_offset = $gy_gannen;
1913 if ( $gy_gannen == 1 ) {
1914 $gy_offset = '元';
1915 }
1916 $gy_offset = '平成' . $gy_offset;
1917 }
1918 } else {
1919 $gy_offset = $gy;
1920 }
1921
1922 return [ $gy_offset, $gm, $gd ];
1923 }
1924
1925 /**
1926 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1927 *
1928 * This is the rule the BIDI algorithm uses to determine the directionality of
1929 * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
1930 * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
1931 *
1932 * TODO: Does not handle BIDI control characters inside the text.
1933 * TODO: Does not handle unallocated characters.
1934 *
1935 * @param string $text Text to test
1936 * @return null|string Directionality ('ltr' or 'rtl') or null
1937 */
1938 private static function strongDirFromContent( $text = '' ) {
1939 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
1940 return null;
1941 }
1942 if ( $matches[1] === '' ) {
1943 return 'rtl';
1944 }
1945 return 'ltr';
1946 }
1947
1948 /**
1949 * Roman number formatting up to 10000
1950 *
1951 * @param int $num
1952 *
1953 * @return string
1954 */
1955 static function romanNumeral( $num ) {
1956 static $table = [
1957 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
1958 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
1959 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
1960 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1961 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
1962 ];
1963
1964 $num = intval( $num );
1965 if ( $num > 10000 || $num <= 0 ) {
1966 return $num;
1967 }
1968
1969 $s = '';
1970 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1971 if ( $num >= $pow10 ) {
1972 $s .= $table[$i][(int)floor( $num / $pow10 )];
1973 }
1974 $num = $num % $pow10;
1975 }
1976 return $s;
1977 }
1978
1979 /**
1980 * Hebrew Gematria number formatting up to 9999
1981 *
1982 * @param int $num
1983 *
1984 * @return string
1985 */
1986 static function hebrewNumeral( $num ) {
1987 static $table = [
1988 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
1989 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
1990 [ '',
1991 [ 'ק' ],
1992 [ 'ר' ],
1993 [ 'ש' ],
1994 [ 'ת' ],
1995 [ 'ת', 'ק' ],
1996 [ 'ת', 'ר' ],
1997 [ 'ת', 'ש' ],
1998 [ 'ת', 'ת' ],
1999 [ 'ת', 'ת', 'ק' ],
2000 [ 'ת', 'ת', 'ר' ],
2001 ],
2002 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2003 ];
2004
2005 $num = intval( $num );
2006 if ( $num > 9999 || $num <= 0 ) {
2007 return $num;
2008 }
2009
2010 // Round thousands have special notations
2011 if ( $num === 1000 ) {
2012 return "א' אלף";
2013 } elseif ( $num % 1000 === 0 ) {
2014 return $table[0][$num / 1000] . "' אלפים";
2015 }
2016
2017 $letters = [];
2018
2019 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2020 if ( $num >= $pow10 ) {
2021 if ( $num === 15 || $num === 16 ) {
2022 $letters[] = $table[0][9];
2023 $letters[] = $table[0][$num - 9];
2024 $num = 0;
2025 } else {
2026 $letters = array_merge(
2027 $letters,
2028 (array)$table[$i][intval( $num / $pow10 )]
2029 );
2030
2031 if ( $pow10 === 1000 ) {
2032 $letters[] = "'";
2033 }
2034 }
2035 }
2036
2037 $num = $num % $pow10;
2038 }
2039
2040 $preTransformLength = count( $letters );
2041 if ( $preTransformLength === 1 ) {
2042 // Add geresh (single quote) to one-letter numbers
2043 $letters[] = "'";
2044 } else {
2045 $lastIndex = $preTransformLength - 1;
2046 $letters[$lastIndex] = str_replace(
2047 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2048 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2049 $letters[$lastIndex]
2050 );
2051
2052 // Add gershayim (double quote) to multiple-letter numbers,
2053 // but exclude numbers with only one letter after the thousands
2054 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2055 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2056 $letters[] = "'";
2057 } else {
2058 array_splice( $letters, -1, 0, '"' );
2059 }
2060 }
2061
2062 return implode( $letters );
2063 }
2064
2065 /**
2066 * Used by date() and time() to adjust the time output.
2067 *
2068 * @param string $ts The time in date('YmdHis') format
2069 * @param mixed $tz Adjust the time by this amount (default false, mean we
2070 * get user timecorrection setting)
2071 * @return int
2072 */
2073 public function userAdjust( $ts, $tz = false ) {
2074 global $wgUser, $wgLocalTZoffset;
2075
2076 if ( $tz === false ) {
2077 $tz = $wgUser->getOption( 'timecorrection' );
2078 }
2079
2080 $data = explode( '|', $tz, 3 );
2081
2082 if ( $data[0] == 'ZoneInfo' ) {
2083 MediaWiki\suppressWarnings();
2084 $userTZ = timezone_open( $data[2] );
2085 MediaWiki\restoreWarnings();
2086 if ( $userTZ !== false ) {
2087 $date = date_create( $ts, timezone_open( 'UTC' ) );
2088 date_timezone_set( $date, $userTZ );
2089 $date = date_format( $date, 'YmdHis' );
2090 return $date;
2091 }
2092 # Unrecognized timezone, default to 'Offset' with the stored offset.
2093 $data[0] = 'Offset';
2094 }
2095
2096 if ( $data[0] == 'System' || $tz == '' ) {
2097 # Global offset in minutes.
2098 $minDiff = $wgLocalTZoffset;
2099 } elseif ( $data[0] == 'Offset' ) {
2100 $minDiff = intval( $data[1] );
2101 } else {
2102 $data = explode( ':', $tz );
2103 if ( count( $data ) == 2 ) {
2104 $data[0] = intval( $data[0] );
2105 $data[1] = intval( $data[1] );
2106 $minDiff = abs( $data[0] ) * 60 + $data[1];
2107 if ( $data[0] < 0 ) {
2108 $minDiff = -$minDiff;
2109 }
2110 } else {
2111 $minDiff = intval( $data[0] ) * 60;
2112 }
2113 }
2114
2115 # No difference ? Return time unchanged
2116 if ( 0 == $minDiff ) {
2117 return $ts;
2118 }
2119
2120 MediaWiki\suppressWarnings(); // E_STRICT system time bitching
2121 # Generate an adjusted date; take advantage of the fact that mktime
2122 # will normalize out-of-range values so we don't have to split $minDiff
2123 # into hours and minutes.
2124 $t = mktime( (
2125 (int)substr( $ts, 8, 2 ) ), # Hours
2126 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2127 (int)substr( $ts, 12, 2 ), # Seconds
2128 (int)substr( $ts, 4, 2 ), # Month
2129 (int)substr( $ts, 6, 2 ), # Day
2130 (int)substr( $ts, 0, 4 ) ); # Year
2131
2132 $date = date( 'YmdHis', $t );
2133 MediaWiki\restoreWarnings();
2134
2135 return $date;
2136 }
2137
2138 /**
2139 * This is meant to be used by time(), date(), and timeanddate() to get
2140 * the date preference they're supposed to use, it should be used in
2141 * all children.
2142 *
2143 *<code>
2144 * function timeanddate([...], $format = true) {
2145 * $datePreference = $this->dateFormat($format);
2146 * [...]
2147 * }
2148 *</code>
2149 *
2150 * @param int|string|bool $usePrefs If true, the user's preference is used
2151 * if false, the site/language default is used
2152 * if int/string, assumed to be a format.
2153 * @return string
2154 */
2155 function dateFormat( $usePrefs = true ) {
2156 global $wgUser;
2157
2158 if ( is_bool( $usePrefs ) ) {
2159 if ( $usePrefs ) {
2160 $datePreference = $wgUser->getDatePreference();
2161 } else {
2162 $datePreference = (string)User::getDefaultOption( 'date' );
2163 }
2164 } else {
2165 $datePreference = (string)$usePrefs;
2166 }
2167
2168 // return int
2169 if ( $datePreference == '' ) {
2170 return 'default';
2171 }
2172
2173 return $datePreference;
2174 }
2175
2176 /**
2177 * Get a format string for a given type and preference
2178 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2179 * @param string $pref The format name as it appears in Messages*.php under
2180 * $datePreferences.
2181 *
2182 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2183 *
2184 * @return string
2185 */
2186 function getDateFormatString( $type, $pref ) {
2187 $wasDefault = false;
2188 if ( $pref == 'default' ) {
2189 $wasDefault = true;
2190 $pref = $this->getDefaultDateFormat();
2191 }
2192
2193 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2194 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2195
2196 if ( $type === 'pretty' && $df === null ) {
2197 $df = $this->getDateFormatString( 'date', $pref );
2198 }
2199
2200 if ( !$wasDefault && $df === null ) {
2201 $pref = $this->getDefaultDateFormat();
2202 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2203 }
2204
2205 $this->dateFormatStrings[$type][$pref] = $df;
2206 }
2207 return $this->dateFormatStrings[$type][$pref];
2208 }
2209
2210 /**
2211 * @param string $ts The time format which needs to be turned into a
2212 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2213 * @param bool $adj Whether to adjust the time output according to the
2214 * user configured offset ($timecorrection)
2215 * @param mixed $format True to use user's date format preference
2216 * @param string|bool $timecorrection The time offset as returned by
2217 * validateTimeZone() in Special:Preferences
2218 * @return string
2219 */
2220 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2221 $ts = wfTimestamp( TS_MW, $ts );
2222 if ( $adj ) {
2223 $ts = $this->userAdjust( $ts, $timecorrection );
2224 }
2225 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2226 return $this->sprintfDate( $df, $ts );
2227 }
2228
2229 /**
2230 * @param string $ts The time format which needs to be turned into a
2231 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2232 * @param bool $adj Whether to adjust the time output according to the
2233 * user configured offset ($timecorrection)
2234 * @param mixed $format True to use user's date format preference
2235 * @param string|bool $timecorrection The time offset as returned by
2236 * validateTimeZone() in Special:Preferences
2237 * @return string
2238 */
2239 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2240 $ts = wfTimestamp( TS_MW, $ts );
2241 if ( $adj ) {
2242 $ts = $this->userAdjust( $ts, $timecorrection );
2243 }
2244 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2245 return $this->sprintfDate( $df, $ts );
2246 }
2247
2248 /**
2249 * @param string $ts The time format which needs to be turned into a
2250 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2251 * @param bool $adj Whether to adjust the time output according to the
2252 * user configured offset ($timecorrection)
2253 * @param mixed $format What format to return, if it's false output the
2254 * default one (default true)
2255 * @param string|bool $timecorrection The time offset as returned by
2256 * validateTimeZone() in Special:Preferences
2257 * @return string
2258 */
2259 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2260 $ts = wfTimestamp( TS_MW, $ts );
2261 if ( $adj ) {
2262 $ts = $this->userAdjust( $ts, $timecorrection );
2263 }
2264 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2265 return $this->sprintfDate( $df, $ts );
2266 }
2267
2268 /**
2269 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2270 *
2271 * @since 1.20
2272 *
2273 * @param int $seconds The amount of seconds.
2274 * @param array $chosenIntervals The intervals to enable.
2275 *
2276 * @return string
2277 */
2278 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2279 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2280
2281 $segments = [];
2282
2283 foreach ( $intervals as $intervalName => $intervalValue ) {
2284 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2285 // duration-years, duration-decades, duration-centuries, duration-millennia
2286 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2287 $segments[] = $message->inLanguage( $this )->escaped();
2288 }
2289
2290 return $this->listToText( $segments );
2291 }
2292
2293 /**
2294 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2295 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2296 *
2297 * @since 1.20
2298 *
2299 * @param int $seconds The amount of seconds.
2300 * @param array $chosenIntervals The intervals to enable.
2301 *
2302 * @return array
2303 */
2304 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2305 if ( empty( $chosenIntervals ) ) {
2306 $chosenIntervals = [
2307 'millennia',
2308 'centuries',
2309 'decades',
2310 'years',
2311 'days',
2312 'hours',
2313 'minutes',
2314 'seconds'
2315 ];
2316 }
2317
2318 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2319 $sortedNames = array_keys( $intervals );
2320 $smallestInterval = array_pop( $sortedNames );
2321
2322 $segments = [];
2323
2324 foreach ( $intervals as $name => $length ) {
2325 $value = floor( $seconds / $length );
2326
2327 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2328 $seconds -= $value * $length;
2329 $segments[$name] = $value;
2330 }
2331 }
2332
2333 return $segments;
2334 }
2335
2336 /**
2337 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2338 *
2339 * @param string $type Can be 'date', 'time' or 'both'
2340 * @param string $ts The time format which needs to be turned into a
2341 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2342 * @param User $user User object used to get preferences for timezone and format
2343 * @param array $options Array, can contain the following keys:
2344 * - 'timecorrection': time correction, can have the following values:
2345 * - true: use user's preference
2346 * - false: don't use time correction
2347 * - int: value of time correction in minutes
2348 * - 'format': format to use, can have the following values:
2349 * - true: use user's preference
2350 * - false: use default preference
2351 * - string: format to use
2352 * @since 1.19
2353 * @return string
2354 */
2355 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2356 $ts = wfTimestamp( TS_MW, $ts );
2357 $options += [ 'timecorrection' => true, 'format' => true ];
2358 if ( $options['timecorrection'] !== false ) {
2359 if ( $options['timecorrection'] === true ) {
2360 $offset = $user->getOption( 'timecorrection' );
2361 } else {
2362 $offset = $options['timecorrection'];
2363 }
2364 $ts = $this->userAdjust( $ts, $offset );
2365 }
2366 if ( $options['format'] === true ) {
2367 $format = $user->getDatePreference();
2368 } else {
2369 $format = $options['format'];
2370 }
2371 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2372 return $this->sprintfDate( $df, $ts );
2373 }
2374
2375 /**
2376 * Get the formatted date for the given timestamp and formatted for
2377 * the given user.
2378 *
2379 * @param mixed $ts Mixed: the time format which needs to be turned into a
2380 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2381 * @param User $user User object used to get preferences for timezone and format
2382 * @param array $options Array, can contain the following keys:
2383 * - 'timecorrection': time correction, can have the following values:
2384 * - true: use user's preference
2385 * - false: don't use time correction
2386 * - int: value of time correction in minutes
2387 * - 'format': format to use, can have the following values:
2388 * - true: use user's preference
2389 * - false: use default preference
2390 * - string: format to use
2391 * @since 1.19
2392 * @return string
2393 */
2394 public function userDate( $ts, User $user, array $options = [] ) {
2395 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2396 }
2397
2398 /**
2399 * Get the formatted time for the given timestamp and formatted for
2400 * the given user.
2401 *
2402 * @param mixed $ts The time format which needs to be turned into a
2403 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2404 * @param User $user User object used to get preferences for timezone and format
2405 * @param array $options Array, can contain the following keys:
2406 * - 'timecorrection': time correction, can have the following values:
2407 * - true: use user's preference
2408 * - false: don't use time correction
2409 * - int: value of time correction in minutes
2410 * - 'format': format to use, can have the following values:
2411 * - true: use user's preference
2412 * - false: use default preference
2413 * - string: format to use
2414 * @since 1.19
2415 * @return string
2416 */
2417 public function userTime( $ts, User $user, array $options = [] ) {
2418 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2419 }
2420
2421 /**
2422 * Get the formatted date and time for the given timestamp and formatted for
2423 * the given user.
2424 *
2425 * @param mixed $ts The time format which needs to be turned into a
2426 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2427 * @param User $user User object used to get preferences for timezone and format
2428 * @param array $options Array, can contain the following keys:
2429 * - 'timecorrection': time correction, can have the following values:
2430 * - true: use user's preference
2431 * - false: don't use time correction
2432 * - int: value of time correction in minutes
2433 * - 'format': format to use, can have the following values:
2434 * - true: use user's preference
2435 * - false: use default preference
2436 * - string: format to use
2437 * @since 1.19
2438 * @return string
2439 */
2440 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2441 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2442 }
2443
2444 /**
2445 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2446 *
2447 * Determine the difference between the timestamp and the current time, and
2448 * generate a readable timestamp by returning "<N> <units> ago", where the
2449 * largest possible unit is used.
2450 *
2451 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2452 *
2453 * @param MWTimestamp $time
2454 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2455 * @param User|null $user User the timestamp is being generated for
2456 * (or null to use main context's user)
2457 * @return string Formatted timestamp
2458 */
2459 public function getHumanTimestamp(
2460 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2461 ) {
2462 if ( $relativeTo === null ) {
2463 $relativeTo = new MWTimestamp();
2464 }
2465 if ( $user === null ) {
2466 $user = RequestContext::getMain()->getUser();
2467 }
2468
2469 // Adjust for the user's timezone.
2470 $offsetThis = $time->offsetForUser( $user );
2471 $offsetRel = $relativeTo->offsetForUser( $user );
2472
2473 $ts = '';
2474 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2475 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2476 }
2477
2478 // Reset the timezone on the objects.
2479 $time->timestamp->sub( $offsetThis );
2480 $relativeTo->timestamp->sub( $offsetRel );
2481
2482 return $ts;
2483 }
2484
2485 /**
2486 * Convert an MWTimestamp into a pretty human-readable timestamp using
2487 * the given user preferences and relative base time.
2488 *
2489 * @see Language::getHumanTimestamp
2490 * @param MWTimestamp $ts Timestamp to prettify
2491 * @param MWTimestamp $relativeTo Base timestamp
2492 * @param User $user User preferences to use
2493 * @return string Human timestamp
2494 * @since 1.26
2495 */
2496 private function getHumanTimestampInternal(
2497 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2498 ) {
2499 $diff = $ts->diff( $relativeTo );
2500 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2501 (int)$relativeTo->timestamp->format( 'w' ) );
2502 $days = $diff->days ?: (int)$diffDay;
2503 if ( $diff->invert || $days > 5
2504 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2505 ) {
2506 // Timestamps are in different years: use full timestamp
2507 // Also do full timestamp for future dates
2508 /**
2509 * @todo FIXME: Add better handling of future timestamps.
2510 */
2511 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2512 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2513 } elseif ( $days > 5 ) {
2514 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2515 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2516 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2517 } elseif ( $days > 1 ) {
2518 // Timestamp within the past week: show the day of the week and time
2519 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2520 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2521 // Messages:
2522 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2523 $ts = wfMessage( "$weekday-at" )
2524 ->inLanguage( $this )
2525 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2526 ->text();
2527 } elseif ( $days == 1 ) {
2528 // Timestamp was yesterday: say 'yesterday' and the time.
2529 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2530 $ts = wfMessage( 'yesterday-at' )
2531 ->inLanguage( $this )
2532 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2533 ->text();
2534 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2535 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2536 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2537 $ts = wfMessage( 'today-at' )
2538 ->inLanguage( $this )
2539 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2540 ->text();
2541
2542 // From here on in, the timestamp was soon enough ago so that we can simply say
2543 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2544 } elseif ( $diff->h == 1 ) {
2545 // Less than 90 minutes, but more than an hour ago.
2546 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2547 } elseif ( $diff->i >= 1 ) {
2548 // A few minutes ago.
2549 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2550 } elseif ( $diff->s >= 30 ) {
2551 // Less than a minute, but more than 30 sec ago.
2552 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2553 } else {
2554 // Less than 30 seconds ago.
2555 $ts = wfMessage( 'just-now' )->text();
2556 }
2557
2558 return $ts;
2559 }
2560
2561 /**
2562 * @param string $key
2563 * @return array|null
2564 */
2565 public function getMessage( $key ) {
2566 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2567 }
2568
2569 /**
2570 * @return array
2571 */
2572 function getAllMessages() {
2573 return self::$dataCache->getItem( $this->mCode, 'messages' );
2574 }
2575
2576 /**
2577 * @param string $in
2578 * @param string $out
2579 * @param string $string
2580 * @return string
2581 */
2582 public function iconv( $in, $out, $string ) {
2583 # Even with //IGNORE iconv can whine about illegal characters in
2584 # *input* string. We just ignore those too.
2585 # REF: http://bugs.php.net/bug.php?id=37166
2586 # REF: https://phabricator.wikimedia.org/T18885
2587 MediaWiki\suppressWarnings();
2588 $text = iconv( $in, $out . '//IGNORE', $string );
2589 MediaWiki\restoreWarnings();
2590 return $text;
2591 }
2592
2593 // callback functions for ucwords(), ucwordbreaks()
2594
2595 /**
2596 * @param array $matches
2597 * @return mixed|string
2598 */
2599 function ucwordbreaksCallbackAscii( $matches ) {
2600 return $this->ucfirst( $matches[1] );
2601 }
2602
2603 /**
2604 * @param array $matches
2605 * @return string
2606 */
2607 function ucwordbreaksCallbackMB( $matches ) {
2608 return mb_strtoupper( $matches[0] );
2609 }
2610
2611 /**
2612 * @param array $matches
2613 * @return string
2614 */
2615 function ucwordsCallbackMB( $matches ) {
2616 return mb_strtoupper( $matches[0] );
2617 }
2618
2619 /**
2620 * Make a string's first character uppercase
2621 *
2622 * @param string $str
2623 *
2624 * @return string
2625 */
2626 public function ucfirst( $str ) {
2627 $o = ord( $str );
2628 if ( $o < 96 ) { // if already uppercase...
2629 return $str;
2630 } elseif ( $o < 128 ) {
2631 return ucfirst( $str ); // use PHP's ucfirst()
2632 } else {
2633 // fall back to more complex logic in case of multibyte strings
2634 return $this->uc( $str, true );
2635 }
2636 }
2637
2638 /**
2639 * Convert a string to uppercase
2640 *
2641 * @param string $str
2642 * @param bool $first
2643 *
2644 * @return string
2645 */
2646 public function uc( $str, $first = false ) {
2647 if ( $first ) {
2648 if ( $this->isMultibyte( $str ) ) {
2649 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2650 } else {
2651 return ucfirst( $str );
2652 }
2653 } else {
2654 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2655 }
2656 }
2657
2658 /**
2659 * @param string $str
2660 * @return mixed|string
2661 */
2662 function lcfirst( $str ) {
2663 $o = ord( $str );
2664 if ( !$o ) {
2665 return strval( $str );
2666 } elseif ( $o >= 128 ) {
2667 return $this->lc( $str, true );
2668 } elseif ( $o > 96 ) {
2669 return $str;
2670 } else {
2671 $str[0] = strtolower( $str[0] );
2672 return $str;
2673 }
2674 }
2675
2676 /**
2677 * @param string $str
2678 * @param bool $first
2679 * @return mixed|string
2680 */
2681 function lc( $str, $first = false ) {
2682 if ( $first ) {
2683 if ( $this->isMultibyte( $str ) ) {
2684 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2685 } else {
2686 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2687 }
2688 } else {
2689 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2690 }
2691 }
2692
2693 /**
2694 * @param string $str
2695 * @return bool
2696 */
2697 function isMultibyte( $str ) {
2698 return strlen( $str ) !== mb_strlen( $str );
2699 }
2700
2701 /**
2702 * @param string $str
2703 * @return mixed|string
2704 */
2705 function ucwords( $str ) {
2706 if ( $this->isMultibyte( $str ) ) {
2707 $str = $this->lc( $str );
2708
2709 // regexp to find first letter in each word (i.e. after each space)
2710 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2711
2712 // function to use to capitalize a single char
2713 return preg_replace_callback(
2714 $replaceRegexp,
2715 [ $this, 'ucwordsCallbackMB' ],
2716 $str
2717 );
2718 } else {
2719 return ucwords( strtolower( $str ) );
2720 }
2721 }
2722
2723 /**
2724 * capitalize words at word breaks
2725 *
2726 * @param string $str
2727 * @return mixed
2728 */
2729 function ucwordbreaks( $str ) {
2730 if ( $this->isMultibyte( $str ) ) {
2731 $str = $this->lc( $str );
2732
2733 // since \b doesn't work for UTF-8, we explicitely define word break chars
2734 $breaks = "[ \-\(\)\}\{\.,\?!]";
2735
2736 // find first letter after word break
2737 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2738 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2739
2740 return preg_replace_callback(
2741 $replaceRegexp,
2742 [ $this, 'ucwordbreaksCallbackMB' ],
2743 $str
2744 );
2745 } else {
2746 return preg_replace_callback(
2747 '/\b([\w\x80-\xff]+)\b/',
2748 [ $this, 'ucwordbreaksCallbackAscii' ],
2749 $str
2750 );
2751 }
2752 }
2753
2754 /**
2755 * Return a case-folded representation of $s
2756 *
2757 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2758 * and $s2 are the same except for the case of their characters. It is not
2759 * necessary for the value returned to make sense when displayed.
2760 *
2761 * Do *not* perform any other normalisation in this function. If a caller
2762 * uses this function when it should be using a more general normalisation
2763 * function, then fix the caller.
2764 *
2765 * @param string $s
2766 *
2767 * @return string
2768 */
2769 function caseFold( $s ) {
2770 return $this->uc( $s );
2771 }
2772
2773 /**
2774 * @param string $s
2775 * @return string
2776 * @throws MWException
2777 */
2778 function checkTitleEncoding( $s ) {
2779 if ( is_array( $s ) ) {
2780 throw new MWException( 'Given array to checkTitleEncoding.' );
2781 }
2782 if ( StringUtils::isUtf8( $s ) ) {
2783 return $s;
2784 }
2785
2786 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2787 }
2788
2789 /**
2790 * @return array
2791 */
2792 function fallback8bitEncoding() {
2793 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2794 }
2795
2796 /**
2797 * Most writing systems use whitespace to break up words.
2798 * Some languages such as Chinese don't conventionally do this,
2799 * which requires special handling when breaking up words for
2800 * searching etc.
2801 *
2802 * @return bool
2803 */
2804 function hasWordBreaks() {
2805 return true;
2806 }
2807
2808 /**
2809 * Some languages such as Chinese require word segmentation,
2810 * Specify such segmentation when overridden in derived class.
2811 *
2812 * @param string $string
2813 * @return string
2814 */
2815 function segmentByWord( $string ) {
2816 return $string;
2817 }
2818
2819 /**
2820 * Some languages have special punctuation need to be normalized.
2821 * Make such changes here.
2822 *
2823 * @param string $string
2824 * @return string
2825 */
2826 function normalizeForSearch( $string ) {
2827 return self::convertDoubleWidth( $string );
2828 }
2829
2830 /**
2831 * convert double-width roman characters to single-width.
2832 * range: ff00-ff5f ~= 0020-007f
2833 *
2834 * @param string $string
2835 *
2836 * @return string
2837 */
2838 protected static function convertDoubleWidth( $string ) {
2839 static $full = null;
2840 static $half = null;
2841
2842 if ( $full === null ) {
2843 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2844 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2845 $full = str_split( $fullWidth, 3 );
2846 $half = str_split( $halfWidth );
2847 }
2848
2849 $string = str_replace( $full, $half, $string );
2850 return $string;
2851 }
2852
2853 /**
2854 * @param string $string
2855 * @param string $pattern
2856 * @return string
2857 */
2858 protected static function insertSpace( $string, $pattern ) {
2859 $string = preg_replace( $pattern, " $1 ", $string );
2860 $string = preg_replace( '/ +/', ' ', $string );
2861 return $string;
2862 }
2863
2864 /**
2865 * @param array $termsArray
2866 * @return array
2867 */
2868 function convertForSearchResult( $termsArray ) {
2869 # some languages, e.g. Chinese, need to do a conversion
2870 # in order for search results to be displayed correctly
2871 return $termsArray;
2872 }
2873
2874 /**
2875 * Get the first character of a string.
2876 *
2877 * @param string $s
2878 * @return string
2879 */
2880 function firstChar( $s ) {
2881 $matches = [];
2882 preg_match(
2883 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2884 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2885 $s,
2886 $matches
2887 );
2888
2889 if ( isset( $matches[1] ) ) {
2890 if ( strlen( $matches[1] ) != 3 ) {
2891 return $matches[1];
2892 }
2893
2894 // Break down Hangul syllables to grab the first jamo
2895 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2896 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2897 return $matches[1];
2898 } elseif ( $code < 0xb098 ) {
2899 return "\xe3\x84\xb1";
2900 } elseif ( $code < 0xb2e4 ) {
2901 return "\xe3\x84\xb4";
2902 } elseif ( $code < 0xb77c ) {
2903 return "\xe3\x84\xb7";
2904 } elseif ( $code < 0xb9c8 ) {
2905 return "\xe3\x84\xb9";
2906 } elseif ( $code < 0xbc14 ) {
2907 return "\xe3\x85\x81";
2908 } elseif ( $code < 0xc0ac ) {
2909 return "\xe3\x85\x82";
2910 } elseif ( $code < 0xc544 ) {
2911 return "\xe3\x85\x85";
2912 } elseif ( $code < 0xc790 ) {
2913 return "\xe3\x85\x87";
2914 } elseif ( $code < 0xcc28 ) {
2915 return "\xe3\x85\x88";
2916 } elseif ( $code < 0xce74 ) {
2917 return "\xe3\x85\x8a";
2918 } elseif ( $code < 0xd0c0 ) {
2919 return "\xe3\x85\x8b";
2920 } elseif ( $code < 0xd30c ) {
2921 return "\xe3\x85\x8c";
2922 } elseif ( $code < 0xd558 ) {
2923 return "\xe3\x85\x8d";
2924 } else {
2925 return "\xe3\x85\x8e";
2926 }
2927 } else {
2928 return '';
2929 }
2930 }
2931
2932 /**
2933 * @deprecated No-op since 1.28
2934 */
2935 function initEncoding() {
2936 // No-op.
2937 }
2938
2939 /**
2940 * @param string $s
2941 * @return string
2942 * @deprecated No-op since 1.28
2943 */
2944 function recodeForEdit( $s ) {
2945 return $s;
2946 }
2947
2948 /**
2949 * @param string $s
2950 * @return string
2951 * @deprecated No-op since 1.28
2952 */
2953 function recodeInput( $s ) {
2954 return $s;
2955 }
2956
2957 /**
2958 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2959 * also cleans up certain backwards-compatible sequences, converting them
2960 * to the modern Unicode equivalent.
2961 *
2962 * This is language-specific for performance reasons only.
2963 *
2964 * @param string $s
2965 *
2966 * @return string
2967 */
2968 function normalize( $s ) {
2969 global $wgAllUnicodeFixes;
2970 $s = UtfNormal\Validator::cleanUp( $s );
2971 if ( $wgAllUnicodeFixes ) {
2972 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2973 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2974 }
2975
2976 return $s;
2977 }
2978
2979 /**
2980 * Transform a string using serialized data stored in the given file (which
2981 * must be in the serialized subdirectory of $IP). The file contains pairs
2982 * mapping source characters to destination characters.
2983 *
2984 * The data is cached in process memory. This will go faster if you have the
2985 * FastStringSearch extension.
2986 *
2987 * @param string $file
2988 * @param string $string
2989 *
2990 * @throws MWException
2991 * @return string
2992 */
2993 function transformUsingPairFile( $file, $string ) {
2994 if ( !isset( $this->transformData[$file] ) ) {
2995 $data = wfGetPrecompiledData( $file );
2996 if ( $data === false ) {
2997 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2998 }
2999 $this->transformData[$file] = new ReplacementArray( $data );
3000 }
3001 return $this->transformData[$file]->replace( $string );
3002 }
3003
3004 /**
3005 * For right-to-left language support
3006 *
3007 * @return bool
3008 */
3009 function isRTL() {
3010 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3011 }
3012
3013 /**
3014 * Return the correct HTML 'dir' attribute value for this language.
3015 * @return string
3016 */
3017 function getDir() {
3018 return $this->isRTL() ? 'rtl' : 'ltr';
3019 }
3020
3021 /**
3022 * Return 'left' or 'right' as appropriate alignment for line-start
3023 * for this language's text direction.
3024 *
3025 * Should be equivalent to CSS3 'start' text-align value....
3026 *
3027 * @return string
3028 */
3029 function alignStart() {
3030 return $this->isRTL() ? 'right' : 'left';
3031 }
3032
3033 /**
3034 * Return 'right' or 'left' as appropriate alignment for line-end
3035 * for this language's text direction.
3036 *
3037 * Should be equivalent to CSS3 'end' text-align value....
3038 *
3039 * @return string
3040 */
3041 function alignEnd() {
3042 return $this->isRTL() ? 'left' : 'right';
3043 }
3044
3045 /**
3046 * A hidden direction mark (LRM or RLM), depending on the language direction.
3047 * Unlike getDirMark(), this function returns the character as an HTML entity.
3048 * This function should be used when the output is guaranteed to be HTML,
3049 * because it makes the output HTML source code more readable. When
3050 * the output is plain text or can be escaped, getDirMark() should be used.
3051 *
3052 * @param bool $opposite Get the direction mark opposite to your language
3053 * @return string
3054 * @since 1.20
3055 */
3056 function getDirMarkEntity( $opposite = false ) {
3057 if ( $opposite ) {
3058 return $this->isRTL() ? '&lrm;' : '&rlm;';
3059 }
3060 return $this->isRTL() ? '&rlm;' : '&lrm;';
3061 }
3062
3063 /**
3064 * A hidden direction mark (LRM or RLM), depending on the language direction.
3065 * This function produces them as invisible Unicode characters and
3066 * the output may be hard to read and debug, so it should only be used
3067 * when the output is plain text or can be escaped. When the output is
3068 * HTML, use getDirMarkEntity() instead.
3069 *
3070 * @param bool $opposite Get the direction mark opposite to your language
3071 * @return string
3072 */
3073 function getDirMark( $opposite = false ) {
3074 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3075 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3076 if ( $opposite ) {
3077 return $this->isRTL() ? $lrm : $rlm;
3078 }
3079 return $this->isRTL() ? $rlm : $lrm;
3080 }
3081
3082 /**
3083 * @return array
3084 */
3085 function capitalizeAllNouns() {
3086 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3087 }
3088
3089 /**
3090 * An arrow, depending on the language direction.
3091 *
3092 * @param string $direction The direction of the arrow: forwards (default),
3093 * backwards, left, right, up, down.
3094 * @return string
3095 */
3096 function getArrow( $direction = 'forwards' ) {
3097 switch ( $direction ) {
3098 case 'forwards':
3099 return $this->isRTL() ? '←' : '→';
3100 case 'backwards':
3101 return $this->isRTL() ? '→' : '←';
3102 case 'left':
3103 return '←';
3104 case 'right':
3105 return '→';
3106 case 'up':
3107 return '↑';
3108 case 'down':
3109 return '↓';
3110 }
3111 }
3112
3113 /**
3114 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3115 *
3116 * @return bool
3117 */
3118 function linkPrefixExtension() {
3119 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3120 }
3121
3122 /**
3123 * Get all magic words from cache.
3124 * @return array
3125 */
3126 function getMagicWords() {
3127 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3128 }
3129
3130 /**
3131 * Run the LanguageGetMagic hook once.
3132 */
3133 protected function doMagicHook() {
3134 if ( $this->mMagicHookDone ) {
3135 return;
3136 }
3137 $this->mMagicHookDone = true;
3138 Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
3139 }
3140
3141 /**
3142 * Fill a MagicWord object with data from here
3143 *
3144 * @param MagicWord $mw
3145 */
3146 function getMagic( $mw ) {
3147 // Saves a function call
3148 if ( !$this->mMagicHookDone ) {
3149 $this->doMagicHook();
3150 }
3151
3152 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3153 $rawEntry = $this->mMagicExtensions[$mw->mId];
3154 } else {
3155 $rawEntry = self::$dataCache->getSubitem(
3156 $this->mCode, 'magicWords', $mw->mId );
3157 }
3158
3159 if ( !is_array( $rawEntry ) ) {
3160 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3161 } else {
3162 $mw->mCaseSensitive = $rawEntry[0];
3163 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3164 }
3165 }
3166
3167 /**
3168 * Add magic words to the extension array
3169 *
3170 * @param array $newWords
3171 */
3172 function addMagicWordsByLang( $newWords ) {
3173 $fallbackChain = $this->getFallbackLanguages();
3174 $fallbackChain = array_reverse( $fallbackChain );
3175 foreach ( $fallbackChain as $code ) {
3176 if ( isset( $newWords[$code] ) ) {
3177 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3178 }
3179 }
3180 }
3181
3182 /**
3183 * Get special page names, as an associative array
3184 * canonical name => array of valid names, including aliases
3185 * @return array
3186 */
3187 function getSpecialPageAliases() {
3188 // Cache aliases because it may be slow to load them
3189 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3190 // Initialise array
3191 $this->mExtendedSpecialPageAliases =
3192 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3193 Hooks::run( 'LanguageGetSpecialPageAliases',
3194 [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
3195 }
3196
3197 return $this->mExtendedSpecialPageAliases;
3198 }
3199
3200 /**
3201 * Italic is unsuitable for some languages
3202 *
3203 * @param string $text The text to be emphasized.
3204 * @return string
3205 */
3206 function emphasize( $text ) {
3207 return "<em>$text</em>";
3208 }
3209
3210 /**
3211 * Normally we output all numbers in plain en_US style, that is
3212 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3213 * point twohundredthirtyfive. However this is not suitable for all
3214 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3215 * Icelandic just want to use commas instead of dots, and dots instead
3216 * of commas like "293.291,235".
3217 *
3218 * An example of this function being called:
3219 * <code>
3220 * wfMessage( 'message' )->numParams( $num )->text()
3221 * </code>
3222 *
3223 * See $separatorTransformTable on MessageIs.php for
3224 * the , => . and . => , implementation.
3225 *
3226 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3227 * @param int|float $number The string to be formatted, should be an integer
3228 * or a floating point number.
3229 * @param bool $nocommafy Set to true for special numbers like dates
3230 * @return string
3231 */
3232 public function formatNum( $number, $nocommafy = false ) {
3233 global $wgTranslateNumerals;
3234 if ( !$nocommafy ) {
3235 $number = $this->commafy( $number );
3236 $s = $this->separatorTransformTable();
3237 if ( $s ) {
3238 $number = strtr( $number, $s );
3239 }
3240 }
3241
3242 if ( $wgTranslateNumerals ) {
3243 $s = $this->digitTransformTable();
3244 if ( $s ) {
3245 $number = strtr( $number, $s );
3246 }
3247 }
3248
3249 return $number;
3250 }
3251
3252 /**
3253 * Front-end for non-commafied formatNum
3254 *
3255 * @param int|float $number The string to be formatted, should be an integer
3256 * or a floating point number.
3257 * @since 1.21
3258 * @return string
3259 */
3260 public function formatNumNoSeparators( $number ) {
3261 return $this->formatNum( $number, true );
3262 }
3263
3264 /**
3265 * @param string $number
3266 * @return string
3267 */
3268 public function parseFormattedNumber( $number ) {
3269 $s = $this->digitTransformTable();
3270 if ( $s ) {
3271 // eliminate empty array values such as ''. (bug 64347)
3272 $s = array_filter( $s );
3273 $number = strtr( $number, array_flip( $s ) );
3274 }
3275
3276 $s = $this->separatorTransformTable();
3277 if ( $s ) {
3278 // eliminate empty array values such as ''. (bug 64347)
3279 $s = array_filter( $s );
3280 $number = strtr( $number, array_flip( $s ) );
3281 }
3282
3283 $number = strtr( $number, [ ',' => '' ] );
3284 return $number;
3285 }
3286
3287 /**
3288 * Adds commas to a given number
3289 * @since 1.19
3290 * @param mixed $number
3291 * @return string
3292 */
3293 function commafy( $number ) {
3294 $digitGroupingPattern = $this->digitGroupingPattern();
3295 if ( $number === null ) {
3296 return '';
3297 }
3298
3299 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3300 // default grouping is at thousands, use the same for ###,###,### pattern too.
3301 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3302 } else {
3303 // Ref: http://cldr.unicode.org/translation/number-patterns
3304 $sign = "";
3305 if ( intval( $number ) < 0 ) {
3306 // For negative numbers apply the algorithm like positive number and add sign.
3307 $sign = "-";
3308 $number = substr( $number, 1 );
3309 }
3310 $integerPart = [];
3311 $decimalPart = [];
3312 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3313 preg_match( "/\d+/", $number, $integerPart );
3314 preg_match( "/\.\d*/", $number, $decimalPart );
3315 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3316 if ( $groupedNumber === $number ) {
3317 // the string does not have any number part. Eg: .12345
3318 return $sign . $groupedNumber;
3319 }
3320 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3321 while ( $start > 0 ) {
3322 $match = $matches[0][$numMatches - 1];
3323 $matchLen = strlen( $match );
3324 $start = $end - $matchLen;
3325 if ( $start < 0 ) {
3326 $start = 0;
3327 }
3328 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3329 $end = $start;
3330 if ( $numMatches > 1 ) {
3331 // use the last pattern for the rest of the number
3332 $numMatches--;
3333 }
3334 if ( $start > 0 ) {
3335 $groupedNumber = "," . $groupedNumber;
3336 }
3337 }
3338 return $sign . $groupedNumber;
3339 }
3340 }
3341
3342 /**
3343 * @return string
3344 */
3345 function digitGroupingPattern() {
3346 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3347 }
3348
3349 /**
3350 * @return array
3351 */
3352 function digitTransformTable() {
3353 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3354 }
3355
3356 /**
3357 * @return array
3358 */
3359 function separatorTransformTable() {
3360 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3361 }
3362
3363 /**
3364 * Take a list of strings and build a locale-friendly comma-separated
3365 * list, using the local comma-separator message.
3366 * The last two strings are chained with an "and".
3367 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3368 *
3369 * @param string[] $l
3370 * @return string
3371 */
3372 function listToText( array $l ) {
3373 $m = count( $l ) - 1;
3374 if ( $m < 0 ) {
3375 return '';
3376 }
3377 if ( $m > 0 ) {
3378 $and = $this->msg( 'and' )->escaped();
3379 $space = $this->msg( 'word-separator' )->escaped();
3380 if ( $m > 1 ) {
3381 $comma = $this->msg( 'comma-separator' )->escaped();
3382 }
3383 }
3384 $s = $l[$m];
3385 for ( $i = $m - 1; $i >= 0; $i-- ) {
3386 if ( $i == $m - 1 ) {
3387 $s = $l[$i] . $and . $space . $s;
3388 } else {
3389 $s = $l[$i] . $comma . $s;
3390 }
3391 }
3392 return $s;
3393 }
3394
3395 /**
3396 * Take a list of strings and build a locale-friendly comma-separated
3397 * list, using the local comma-separator message.
3398 * @param string[] $list Array of strings to put in a comma list
3399 * @return string
3400 */
3401 function commaList( array $list ) {
3402 return implode(
3403 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3404 $list
3405 );
3406 }
3407
3408 /**
3409 * Take a list of strings and build a locale-friendly semicolon-separated
3410 * list, using the local semicolon-separator message.
3411 * @param string[] $list Array of strings to put in a semicolon list
3412 * @return string
3413 */
3414 function semicolonList( array $list ) {
3415 return implode(
3416 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3417 $list
3418 );
3419 }
3420
3421 /**
3422 * Same as commaList, but separate it with the pipe instead.
3423 * @param string[] $list Array of strings to put in a pipe list
3424 * @return string
3425 */
3426 function pipeList( array $list ) {
3427 return implode(
3428 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3429 $list
3430 );
3431 }
3432
3433 /**
3434 * Truncate a string to a specified length in bytes, appending an optional
3435 * string (e.g. for ellipses)
3436 *
3437 * The database offers limited byte lengths for some columns in the database;
3438 * multi-byte character sets mean we need to ensure that only whole characters
3439 * are included, otherwise broken characters can be passed to the user
3440 *
3441 * If $length is negative, the string will be truncated from the beginning
3442 *
3443 * @param string $string String to truncate
3444 * @param int $length Maximum length (including ellipses)
3445 * @param string $ellipsis String to append to the truncated text
3446 * @param bool $adjustLength Subtract length of ellipsis from $length.
3447 * $adjustLength was introduced in 1.18, before that behaved as if false.
3448 * @return string
3449 */
3450 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3451 # Use the localized ellipsis character
3452 if ( $ellipsis == '...' ) {
3453 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3454 }
3455 # Check if there is no need to truncate
3456 if ( $length == 0 ) {
3457 return $ellipsis; // convention
3458 } elseif ( strlen( $string ) <= abs( $length ) ) {
3459 return $string; // no need to truncate
3460 }
3461 $stringOriginal = $string;
3462 # If ellipsis length is >= $length then we can't apply $adjustLength
3463 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3464 $string = $ellipsis; // this can be slightly unexpected
3465 # Otherwise, truncate and add ellipsis...
3466 } else {
3467 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3468 if ( $length > 0 ) {
3469 $length -= $eLength;
3470 $string = substr( $string, 0, $length ); // xyz...
3471 $string = $this->removeBadCharLast( $string );
3472 $string = rtrim( $string );
3473 $string = $string . $ellipsis;
3474 } else {
3475 $length += $eLength;
3476 $string = substr( $string, $length ); // ...xyz
3477 $string = $this->removeBadCharFirst( $string );
3478 $string = ltrim( $string );
3479 $string = $ellipsis . $string;
3480 }
3481 }
3482 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3483 # This check is *not* redundant if $adjustLength, due to the single case where
3484 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3485 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3486 return $string;
3487 } else {
3488 return $stringOriginal;
3489 }
3490 }
3491
3492 /**
3493 * Remove bytes that represent an incomplete Unicode character
3494 * at the end of string (e.g. bytes of the char are missing)
3495 *
3496 * @param string $string
3497 * @return string
3498 */
3499 protected function removeBadCharLast( $string ) {
3500 if ( $string != '' ) {
3501 $char = ord( $string[strlen( $string ) - 1] );
3502 $m = [];
3503 if ( $char >= 0xc0 ) {
3504 # We got the first byte only of a multibyte char; remove it.
3505 $string = substr( $string, 0, -1 );
3506 } elseif ( $char >= 0x80 &&
3507 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3508 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3509 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3510 ) {
3511 # We chopped in the middle of a character; remove it
3512 $string = $m[1];
3513 }
3514 }
3515 return $string;
3516 }
3517
3518 /**
3519 * Remove bytes that represent an incomplete Unicode character
3520 * at the start of string (e.g. bytes of the char are missing)
3521 *
3522 * @param string $string
3523 * @return string
3524 */
3525 protected function removeBadCharFirst( $string ) {
3526 if ( $string != '' ) {
3527 $char = ord( $string[0] );
3528 if ( $char >= 0x80 && $char < 0xc0 ) {
3529 # We chopped in the middle of a character; remove the whole thing
3530 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3531 }
3532 }
3533 return $string;
3534 }
3535
3536 /**
3537 * Truncate a string of valid HTML to a specified length in bytes,
3538 * appending an optional string (e.g. for ellipses), and return valid HTML
3539 *
3540 * This is only intended for styled/linked text, such as HTML with
3541 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3542 * Also, this will not detect things like "display:none" CSS.
3543 *
3544 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3545 *
3546 * @param string $text HTML string to truncate
3547 * @param int $length (zero/positive) Maximum length (including ellipses)
3548 * @param string $ellipsis String to append to the truncated text
3549 * @return string
3550 */
3551 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3552 # Use the localized ellipsis character
3553 if ( $ellipsis == '...' ) {
3554 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3555 }
3556 # Check if there is clearly no need to truncate
3557 if ( $length <= 0 ) {
3558 return $ellipsis; // no text shown, nothing to format (convention)
3559 } elseif ( strlen( $text ) <= $length ) {
3560 return $text; // string short enough even *with* HTML (short-circuit)
3561 }
3562
3563 $dispLen = 0; // innerHTML legth so far
3564 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3565 $tagType = 0; // 0-open, 1-close
3566 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3567 $entityState = 0; // 0-not entity, 1-entity
3568 $tag = $ret = ''; // accumulated tag name, accumulated result string
3569 $openTags = []; // open tag stack
3570 $maybeState = null; // possible truncation state
3571
3572 $textLen = strlen( $text );
3573 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3574 for ( $pos = 0; true; ++$pos ) {
3575 # Consider truncation once the display length has reached the maximim.
3576 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3577 # Check that we're not in the middle of a bracket/entity...
3578 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3579 if ( !$testingEllipsis ) {
3580 $testingEllipsis = true;
3581 # Save where we are; we will truncate here unless there turn out to
3582 # be so few remaining characters that truncation is not necessary.
3583 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3584 $maybeState = [ $ret, $openTags ]; // save state
3585 }
3586 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3587 # String in fact does need truncation, the truncation point was OK.
3588 list( $ret, $openTags ) = $maybeState; // reload state
3589 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3590 $ret .= $ellipsis; // add ellipsis
3591 break;
3592 }
3593 }
3594 if ( $pos >= $textLen ) {
3595 break; // extra iteration just for above checks
3596 }
3597
3598 # Read the next char...
3599 $ch = $text[$pos];
3600 $lastCh = $pos ? $text[$pos - 1] : '';
3601 $ret .= $ch; // add to result string
3602 if ( $ch == '<' ) {
3603 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3604 $entityState = 0; // for bad HTML
3605 $bracketState = 1; // tag started (checking for backslash)
3606 } elseif ( $ch == '>' ) {
3607 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3608 $entityState = 0; // for bad HTML
3609 $bracketState = 0; // out of brackets
3610 } elseif ( $bracketState == 1 ) {
3611 if ( $ch == '/' ) {
3612 $tagType = 1; // close tag (e.g. "</span>")
3613 } else {
3614 $tagType = 0; // open tag (e.g. "<span>")
3615 $tag .= $ch;
3616 }
3617 $bracketState = 2; // building tag name
3618 } elseif ( $bracketState == 2 ) {
3619 if ( $ch != ' ' ) {
3620 $tag .= $ch;
3621 } else {
3622 // Name found (e.g. "<a href=..."), add on tag attributes...
3623 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3624 }
3625 } elseif ( $bracketState == 0 ) {
3626 if ( $entityState ) {
3627 if ( $ch == ';' ) {
3628 $entityState = 0;
3629 $dispLen++; // entity is one displayed char
3630 }
3631 } else {
3632 if ( $neLength == 0 && !$maybeState ) {
3633 // Save state without $ch. We want to *hit* the first
3634 // display char (to get tags) but not *use* it if truncating.
3635 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3636 }
3637 if ( $ch == '&' ) {
3638 $entityState = 1; // entity found, (e.g. "&#160;")
3639 } else {
3640 $dispLen++; // this char is displayed
3641 // Add the next $max display text chars after this in one swoop...
3642 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3643 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3644 $dispLen += $skipped;
3645 $pos += $skipped;
3646 }
3647 }
3648 }
3649 }
3650 // Close the last tag if left unclosed by bad HTML
3651 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3652 while ( count( $openTags ) > 0 ) {
3653 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3654 }
3655 return $ret;
3656 }
3657
3658 /**
3659 * truncateHtml() helper function
3660 * like strcspn() but adds the skipped chars to $ret
3661 *
3662 * @param string $ret
3663 * @param string $text
3664 * @param string $search
3665 * @param int $start
3666 * @param null|int $len
3667 * @return int
3668 */
3669 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3670 if ( $len === null ) {
3671 $len = -1; // -1 means "no limit" for strcspn
3672 } elseif ( $len < 0 ) {
3673 $len = 0; // sanity
3674 }
3675 $skipCount = 0;
3676 if ( $start < strlen( $text ) ) {
3677 $skipCount = strcspn( $text, $search, $start, $len );
3678 $ret .= substr( $text, $start, $skipCount );
3679 }
3680 return $skipCount;
3681 }
3682
3683 /**
3684 * truncateHtml() helper function
3685 * (a) push or pop $tag from $openTags as needed
3686 * (b) clear $tag value
3687 * @param string &$tag Current HTML tag name we are looking at
3688 * @param int $tagType (0-open tag, 1-close tag)
3689 * @param string $lastCh Character before the '>' that ended this tag
3690 * @param array &$openTags Open tag stack (not accounting for $tag)
3691 */
3692 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3693 $tag = ltrim( $tag );
3694 if ( $tag != '' ) {
3695 if ( $tagType == 0 && $lastCh != '/' ) {
3696 $openTags[] = $tag; // tag opened (didn't close itself)
3697 } elseif ( $tagType == 1 ) {
3698 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3699 array_pop( $openTags ); // tag closed
3700 }
3701 }
3702 $tag = '';
3703 }
3704 }
3705
3706 /**
3707 * Grammatical transformations, needed for inflected languages
3708 * Invoked by putting {{grammar:case|word}} in a message
3709 *
3710 * @param string $word
3711 * @param string $case
3712 * @return string
3713 */
3714 function convertGrammar( $word, $case ) {
3715 global $wgGrammarForms;
3716 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3717 return $wgGrammarForms[$this->getCode()][$case][$word];
3718 }
3719
3720 return $word;
3721 }
3722 /**
3723 * Get the grammar forms for the content language
3724 * @return array Array of grammar forms
3725 * @since 1.20
3726 */
3727 function getGrammarForms() {
3728 global $wgGrammarForms;
3729 if ( isset( $wgGrammarForms[$this->getCode()] )
3730 && is_array( $wgGrammarForms[$this->getCode()] )
3731 ) {
3732 return $wgGrammarForms[$this->getCode()];
3733 }
3734
3735 return [];
3736 }
3737 /**
3738 * Provides an alternative text depending on specified gender.
3739 * Usage {{gender:username|masculine|feminine|unknown}}.
3740 * username is optional, in which case the gender of current user is used,
3741 * but only in (some) interface messages; otherwise default gender is used.
3742 *
3743 * If no forms are given, an empty string is returned. If only one form is
3744 * given, it will be returned unconditionally. These details are implied by
3745 * the caller and cannot be overridden in subclasses.
3746 *
3747 * If three forms are given, the default is to use the third (unknown) form.
3748 * If fewer than three forms are given, the default is to use the first (masculine) form.
3749 * These details can be overridden in subclasses.
3750 *
3751 * @param string $gender
3752 * @param array $forms
3753 *
3754 * @return string
3755 */
3756 function gender( $gender, $forms ) {
3757 if ( !count( $forms ) ) {
3758 return '';
3759 }
3760 $forms = $this->preConvertPlural( $forms, 2 );
3761 if ( $gender === 'male' ) {
3762 return $forms[0];
3763 }
3764 if ( $gender === 'female' ) {
3765 return $forms[1];
3766 }
3767 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3768 }
3769
3770 /**
3771 * Plural form transformations, needed for some languages.
3772 * For example, there are 3 form of plural in Russian and Polish,
3773 * depending on "count mod 10". See [[w:Plural]]
3774 * For English it is pretty simple.
3775 *
3776 * Invoked by putting {{plural:count|wordform1|wordform2}}
3777 * or {{plural:count|wordform1|wordform2|wordform3}}
3778 *
3779 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3780 *
3781 * @param int $count Non-localized number
3782 * @param array $forms Different plural forms
3783 * @return string Correct form of plural for $count in this language
3784 */
3785 function convertPlural( $count, $forms ) {
3786 // Handle explicit n=pluralform cases
3787 $forms = $this->handleExplicitPluralForms( $count, $forms );
3788 if ( is_string( $forms ) ) {
3789 return $forms;
3790 }
3791 if ( !count( $forms ) ) {
3792 return '';
3793 }
3794
3795 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3796 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3797 return $forms[$pluralForm];
3798 }
3799
3800 /**
3801 * Handles explicit plural forms for Language::convertPlural()
3802 *
3803 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3804 * If an explicitly defined plural form matches the $count, then
3805 * string value returned, otherwise array returned for further consideration
3806 * by CLDR rules or overridden convertPlural().
3807 *
3808 * @since 1.23
3809 *
3810 * @param int $count Non-localized number
3811 * @param array $forms Different plural forms
3812 *
3813 * @return array|string
3814 */
3815 protected function handleExplicitPluralForms( $count, array $forms ) {
3816 foreach ( $forms as $index => $form ) {
3817 if ( preg_match( '/\d+=/i', $form ) ) {
3818 $pos = strpos( $form, '=' );
3819 if ( substr( $form, 0, $pos ) === (string)$count ) {
3820 return substr( $form, $pos + 1 );
3821 }
3822 unset( $forms[$index] );
3823 }
3824 }
3825 return array_values( $forms );
3826 }
3827
3828 /**
3829 * Checks that convertPlural was given an array and pads it to requested
3830 * amount of forms by copying the last one.
3831 *
3832 * @param array $forms Array of forms given to convertPlural
3833 * @param int $count How many forms should there be at least
3834 * @return array Padded array of forms or an exception if not an array
3835 */
3836 protected function preConvertPlural( /* Array */ $forms, $count ) {
3837 while ( count( $forms ) < $count ) {
3838 $forms[] = $forms[count( $forms ) - 1];
3839 }
3840 return $forms;
3841 }
3842
3843 /**
3844 * Wraps argument with unicode control characters for directionality safety
3845 *
3846 * This solves the problem where directionality-neutral characters at the edge of
3847 * the argument string get interpreted with the wrong directionality from the
3848 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
3849 *
3850 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
3851 * directionality of the argument string, using the BIDI algorithm's own "First
3852 * strong directional codepoint" rule. Essentially, this works round the fact that
3853 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
3854 * direction inference). The latter is cleaner but still not widely supported.
3855 *
3856 * @param string $text Text to wrap
3857 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
3858 */
3859 public function embedBidi( $text = '' ) {
3860 $dir = Language::strongDirFromContent( $text );
3861 if ( $dir === 'ltr' ) {
3862 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
3863 return self::$lre . $text . self::$pdf;
3864 }
3865 if ( $dir === 'rtl' ) {
3866 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
3867 return self::$rle . $text . self::$pdf;
3868 }
3869 // No strong directionality: do not wrap
3870 return $text;
3871 }
3872
3873 /**
3874 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3875 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3876 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3877 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3878 * match up with it.
3879 *
3880 * @param string $str The validated block duration in English
3881 * @param User $user User object to use timezone from or null for $wgUser
3882 * @return string Somehow translated block duration
3883 * @see LanguageFi.php for example implementation
3884 */
3885 function translateBlockExpiry( $str, User $user = null ) {
3886 $duration = SpecialBlock::getSuggestedDurations( $this );
3887 foreach ( $duration as $show => $value ) {
3888 if ( strcmp( $str, $value ) == 0 ) {
3889 return htmlspecialchars( trim( $show ) );
3890 }
3891 }
3892
3893 if ( wfIsInfinity( $str ) ) {
3894 foreach ( $duration as $show => $value ) {
3895 if ( wfIsInfinity( $value ) ) {
3896 return htmlspecialchars( trim( $show ) );
3897 }
3898 }
3899 }
3900
3901 // If all else fails, return a standard duration or timestamp description.
3902 $time = strtotime( $str, 0 );
3903 if ( $time === false ) { // Unknown format. Return it as-is in case.
3904 return $str;
3905 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3906 // $time is relative to 0 so it's a duration length.
3907 return $this->formatDuration( $time );
3908 } else { // It's an absolute timestamp.
3909 if ( $time === 0 ) {
3910 // wfTimestamp() handles 0 as current time instead of epoch.
3911 $time = '19700101000000';
3912 }
3913 if ( $user ) {
3914 return $this->userTimeAndDate( $time, $user );
3915 }
3916 return $this->timeanddate( $time );
3917 }
3918 }
3919
3920 /**
3921 * languages like Chinese need to be segmented in order for the diff
3922 * to be of any use
3923 *
3924 * @param string $text
3925 * @return string
3926 */
3927 public function segmentForDiff( $text ) {
3928 return $text;
3929 }
3930
3931 /**
3932 * and unsegment to show the result
3933 *
3934 * @param string $text
3935 * @return string
3936 */
3937 public function unsegmentForDiff( $text ) {
3938 return $text;
3939 }
3940
3941 /**
3942 * Return the LanguageConverter used in the Language
3943 *
3944 * @since 1.19
3945 * @return LanguageConverter
3946 */
3947 public function getConverter() {
3948 return $this->mConverter;
3949 }
3950
3951 /**
3952 * convert text to all supported variants
3953 *
3954 * @param string $text
3955 * @return array
3956 */
3957 public function autoConvertToAllVariants( $text ) {
3958 return $this->mConverter->autoConvertToAllVariants( $text );
3959 }
3960
3961 /**
3962 * convert text to different variants of a language.
3963 *
3964 * @param string $text
3965 * @return string
3966 */
3967 public function convert( $text ) {
3968 return $this->mConverter->convert( $text );
3969 }
3970
3971 /**
3972 * Convert a Title object to a string in the preferred variant
3973 *
3974 * @param Title $title
3975 * @return string
3976 */
3977 public function convertTitle( $title ) {
3978 return $this->mConverter->convertTitle( $title );
3979 }
3980
3981 /**
3982 * Convert a namespace index to a string in the preferred variant
3983 *
3984 * @param int $ns
3985 * @return string
3986 */
3987 public function convertNamespace( $ns ) {
3988 return $this->mConverter->convertNamespace( $ns );
3989 }
3990
3991 /**
3992 * Check if this is a language with variants
3993 *
3994 * @return bool
3995 */
3996 public function hasVariants() {
3997 return count( $this->getVariants() ) > 1;
3998 }
3999
4000 /**
4001 * Check if the language has the specific variant
4002 *
4003 * @since 1.19
4004 * @param string $variant
4005 * @return bool
4006 */
4007 public function hasVariant( $variant ) {
4008 return (bool)$this->mConverter->validateVariant( $variant );
4009 }
4010
4011 /**
4012 * Perform output conversion on a string, and encode for safe HTML output.
4013 * @param string $text Text to be converted
4014 * @param bool $isTitle Whether this conversion is for the article title
4015 * @return string
4016 * @todo this should get integrated somewhere sane
4017 */
4018 public function convertHtml( $text, $isTitle = false ) {
4019 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4020 }
4021
4022 /**
4023 * @param string $key
4024 * @return string
4025 */
4026 public function convertCategoryKey( $key ) {
4027 return $this->mConverter->convertCategoryKey( $key );
4028 }
4029
4030 /**
4031 * Get the list of variants supported by this language
4032 * see sample implementation in LanguageZh.php
4033 *
4034 * @return array An array of language codes
4035 */
4036 public function getVariants() {
4037 return $this->mConverter->getVariants();
4038 }
4039
4040 /**
4041 * @return string
4042 */
4043 public function getPreferredVariant() {
4044 return $this->mConverter->getPreferredVariant();
4045 }
4046
4047 /**
4048 * @return string
4049 */
4050 public function getDefaultVariant() {
4051 return $this->mConverter->getDefaultVariant();
4052 }
4053
4054 /**
4055 * @return string
4056 */
4057 public function getURLVariant() {
4058 return $this->mConverter->getURLVariant();
4059 }
4060
4061 /**
4062 * If a language supports multiple variants, it is
4063 * possible that non-existing link in one variant
4064 * actually exists in another variant. this function
4065 * tries to find it. See e.g. LanguageZh.php
4066 * The input parameters may be modified upon return
4067 *
4068 * @param string &$link The name of the link
4069 * @param Title &$nt The title object of the link
4070 * @param bool $ignoreOtherCond To disable other conditions when
4071 * we need to transclude a template or update a category's link
4072 */
4073 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4074 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4075 }
4076
4077 /**
4078 * returns language specific options used by User::getPageRenderHash()
4079 * for example, the preferred language variant
4080 *
4081 * @return string
4082 */
4083 function getExtraHashOptions() {
4084 return $this->mConverter->getExtraHashOptions();
4085 }
4086
4087 /**
4088 * For languages that support multiple variants, the title of an
4089 * article may be displayed differently in different variants. this
4090 * function returns the apporiate title defined in the body of the article.
4091 *
4092 * @return string
4093 */
4094 public function getParsedTitle() {
4095 return $this->mConverter->getParsedTitle();
4096 }
4097
4098 /**
4099 * Refresh the cache of conversion tables when
4100 * MediaWiki:Conversiontable* is updated.
4101 *
4102 * @param Title $title The Title of the page being updated
4103 */
4104 public function updateConversionTable( Title $title ) {
4105 $this->mConverter->updateConversionTable( $title );
4106 }
4107
4108 /**
4109 * Prepare external link text for conversion. When the text is
4110 * a URL, it shouldn't be converted, and it'll be wrapped in
4111 * the "raw" tag (-{R| }-) to prevent conversion.
4112 *
4113 * This function is called "markNoConversion" for historical
4114 * reasons.
4115 *
4116 * @param string $text Text to be used for external link
4117 * @param bool $noParse Wrap it without confirming it's a real URL first
4118 * @return string The tagged text
4119 */
4120 public function markNoConversion( $text, $noParse = false ) {
4121 // Excluding protocal-relative URLs may avoid many false positives.
4122 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4123 return $this->mConverter->markNoConversion( $text );
4124 } else {
4125 return $text;
4126 }
4127 }
4128
4129 /**
4130 * A regular expression to match legal word-trailing characters
4131 * which should be merged onto a link of the form [[foo]]bar.
4132 *
4133 * @return string
4134 */
4135 public function linkTrail() {
4136 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4137 }
4138
4139 /**
4140 * A regular expression character set to match legal word-prefixing
4141 * characters which should be merged onto a link of the form foo[[bar]].
4142 *
4143 * @return string
4144 */
4145 public function linkPrefixCharset() {
4146 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4147 }
4148
4149 /**
4150 * Get the "parent" language which has a converter to convert a "compatible" language
4151 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4152 *
4153 * @return Language|null
4154 * @since 1.22
4155 */
4156 public function getParentLanguage() {
4157 if ( $this->mParentLanguage !== false ) {
4158 return $this->mParentLanguage;
4159 }
4160
4161 $code = explode( '-', $this->getCode() )[0];
4162 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4163 $this->mParentLanguage = null;
4164 return null;
4165 }
4166 $lang = Language::factory( $code );
4167 if ( !$lang->hasVariant( $this->getCode() ) ) {
4168 $this->mParentLanguage = null;
4169 return null;
4170 }
4171
4172 $this->mParentLanguage = $lang;
4173 return $lang;
4174 }
4175
4176 /**
4177 * Compare with an other language object
4178 *
4179 * @since 1.28
4180 * @param Language $lang
4181 * @return boolean
4182 */
4183 public function equals( Language $lang ) {
4184 return $lang->getCode() === $this->mCode;
4185 }
4186
4187 /**
4188 * Get the internal language code for this language object
4189 *
4190 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4191 * htmlspecialchars() or similar
4192 *
4193 * @return string
4194 */
4195 public function getCode() {
4196 return $this->mCode;
4197 }
4198
4199 /**
4200 * Get the code in BCP 47 format which we can use
4201 * inside of html lang="" tags.
4202 *
4203 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4204 * htmlspecialchars() or similar.
4205 *
4206 * @since 1.19
4207 * @return string
4208 */
4209 public function getHtmlCode() {
4210 if ( is_null( $this->mHtmlCode ) ) {
4211 $this->mHtmlCode = wfBCP47( $this->getCode() );
4212 }
4213 return $this->mHtmlCode;
4214 }
4215
4216 /**
4217 * @param string $code
4218 */
4219 public function setCode( $code ) {
4220 $this->mCode = $code;
4221 // Ensure we don't leave incorrect cached data lying around
4222 $this->mHtmlCode = null;
4223 $this->mParentLanguage = false;
4224 }
4225
4226 /**
4227 * Get the language code from a file name. Inverse of getFileName()
4228 * @param string $filename $prefix . $languageCode . $suffix
4229 * @param string $prefix Prefix before the language code
4230 * @param string $suffix Suffix after the language code
4231 * @return string Language code, or false if $prefix or $suffix isn't found
4232 */
4233 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4234 $m = null;
4235 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4236 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4237 if ( !count( $m ) ) {
4238 return false;
4239 }
4240 return str_replace( '_', '-', strtolower( $m[1] ) );
4241 }
4242
4243 /**
4244 * @param string $code
4245 * @return string Name of the language class
4246 */
4247 public static function classFromCode( $code ) {
4248 if ( $code == 'en' ) {
4249 return 'Language';
4250 } else {
4251 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4252 }
4253 }
4254
4255 /**
4256 * Get the name of a file for a certain language code
4257 * @param string $prefix Prepend this to the filename
4258 * @param string $code Language code
4259 * @param string $suffix Append this to the filename
4260 * @throws MWException
4261 * @return string $prefix . $mangledCode . $suffix
4262 */
4263 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4264 if ( !self::isValidBuiltInCode( $code ) ) {
4265 throw new MWException( "Invalid language code \"$code\"" );
4266 }
4267
4268 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4269 }
4270
4271 /**
4272 * @param string $code
4273 * @return string
4274 */
4275 public static function getMessagesFileName( $code ) {
4276 global $IP;
4277 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4278 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4279 return $file;
4280 }
4281
4282 /**
4283 * @param string $code
4284 * @return string
4285 * @throws MWException
4286 * @since 1.23
4287 */
4288 public static function getJsonMessagesFileName( $code ) {
4289 global $IP;
4290
4291 if ( !self::isValidBuiltInCode( $code ) ) {
4292 throw new MWException( "Invalid language code \"$code\"" );
4293 }
4294
4295 return "$IP/languages/i18n/$code.json";
4296 }
4297
4298 /**
4299 * Get the first fallback for a given language.
4300 *
4301 * @param string $code
4302 *
4303 * @return bool|string
4304 */
4305 public static function getFallbackFor( $code ) {
4306 $fallbacks = self::getFallbacksFor( $code );
4307 if ( $fallbacks ) {
4308 return $fallbacks[0];
4309 }
4310 return false;
4311 }
4312
4313 /**
4314 * Get the ordered list of fallback languages.
4315 *
4316 * @since 1.19
4317 * @param string $code Language code
4318 * @return array Non-empty array, ending in "en"
4319 */
4320 public static function getFallbacksFor( $code ) {
4321 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4322 return [];
4323 }
4324 // For unknown languages, fallbackSequence returns an empty array,
4325 // hardcode fallback to 'en' in that case.
4326 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4327 }
4328
4329 /**
4330 * Get the ordered list of fallback languages, ending with the fallback
4331 * language chain for the site language.
4332 *
4333 * @since 1.22
4334 * @param string $code Language code
4335 * @return array Array( fallbacks, site fallbacks )
4336 */
4337 public static function getFallbacksIncludingSiteLanguage( $code ) {
4338 global $wgLanguageCode;
4339
4340 // Usually, we will only store a tiny number of fallback chains, so we
4341 // keep them in static memory.
4342 $cacheKey = "{$code}-{$wgLanguageCode}";
4343
4344 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4345 $fallbacks = self::getFallbacksFor( $code );
4346
4347 // Append the site's fallback chain, including the site language itself
4348 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4349 array_unshift( $siteFallbacks, $wgLanguageCode );
4350
4351 // Eliminate any languages already included in the chain
4352 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4353
4354 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4355 }
4356 return self::$fallbackLanguageCache[$cacheKey];
4357 }
4358
4359 /**
4360 * Get all messages for a given language
4361 * WARNING: this may take a long time. If you just need all message *keys*
4362 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4363 *
4364 * @param string $code
4365 *
4366 * @return array
4367 */
4368 public static function getMessagesFor( $code ) {
4369 return self::getLocalisationCache()->getItem( $code, 'messages' );
4370 }
4371
4372 /**
4373 * Get a message for a given language
4374 *
4375 * @param string $key
4376 * @param string $code
4377 *
4378 * @return string
4379 */
4380 public static function getMessageFor( $key, $code ) {
4381 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4382 }
4383
4384 /**
4385 * Get all message keys for a given language. This is a faster alternative to
4386 * array_keys( Language::getMessagesFor( $code ) )
4387 *
4388 * @since 1.19
4389 * @param string $code Language code
4390 * @return array Array of message keys (strings)
4391 */
4392 public static function getMessageKeysFor( $code ) {
4393 return self::getLocalisationCache()->getSubitemList( $code, 'messages' );
4394 }
4395
4396 /**
4397 * @param string $talk
4398 * @return mixed
4399 */
4400 function fixVariableInNamespace( $talk ) {
4401 if ( strpos( $talk, '$1' ) === false ) {
4402 return $talk;
4403 }
4404
4405 global $wgMetaNamespace;
4406 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4407
4408 # Allow grammar transformations
4409 # Allowing full message-style parsing would make simple requests
4410 # such as action=raw much more expensive than they need to be.
4411 # This will hopefully cover most cases.
4412 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4413 [ &$this, 'replaceGrammarInNamespace' ], $talk );
4414 return str_replace( ' ', '_', $talk );
4415 }
4416
4417 /**
4418 * @param string $m
4419 * @return string
4420 */
4421 function replaceGrammarInNamespace( $m ) {
4422 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4423 }
4424
4425 /**
4426 * Decode an expiry (block, protection, etc) which has come from the DB
4427 *
4428 * @param string $expiry Database expiry String
4429 * @param bool|int $format True to process using language functions, or TS_ constant
4430 * to return the expiry in a given timestamp
4431 * @param string $infinity If $format is not true, use this string for infinite expiry
4432 * @return string
4433 * @since 1.18
4434 */
4435 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4436 static $dbInfinity;
4437 if ( $dbInfinity === null ) {
4438 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4439 }
4440
4441 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4442 return $format === true
4443 ? $this->getMessageFromDB( 'infiniteblock' )
4444 : $infinity;
4445 } else {
4446 return $format === true
4447 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4448 : wfTimestamp( $format, $expiry );
4449 }
4450 }
4451
4452 /**
4453 * @todo Document
4454 * @param int|float $seconds
4455 * @param array $format Optional
4456 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4457 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4458 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4459 * and friends.
4460 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4461 * or 'avoidminutes'.
4462 * @return string
4463 */
4464 function formatTimePeriod( $seconds, $format = [] ) {
4465 if ( !is_array( $format ) ) {
4466 $format = [ 'avoid' => $format ]; // For backwards compatibility
4467 }
4468 if ( !isset( $format['avoid'] ) ) {
4469 $format['avoid'] = false;
4470 }
4471 if ( !isset( $format['noabbrevs'] ) ) {
4472 $format['noabbrevs'] = false;
4473 }
4474 $secondsMsg = wfMessage(
4475 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4476 $minutesMsg = wfMessage(
4477 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4478 $hoursMsg = wfMessage(
4479 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4480 $daysMsg = wfMessage(
4481 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4482
4483 if ( round( $seconds * 10 ) < 100 ) {
4484 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4485 $s = $secondsMsg->params( $s )->text();
4486 } elseif ( round( $seconds ) < 60 ) {
4487 $s = $this->formatNum( round( $seconds ) );
4488 $s = $secondsMsg->params( $s )->text();
4489 } elseif ( round( $seconds ) < 3600 ) {
4490 $minutes = floor( $seconds / 60 );
4491 $secondsPart = round( fmod( $seconds, 60 ) );
4492 if ( $secondsPart == 60 ) {
4493 $secondsPart = 0;
4494 $minutes++;
4495 }
4496 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4497 $s .= ' ';
4498 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4499 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4500 $hours = floor( $seconds / 3600 );
4501 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4502 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4503 if ( $secondsPart == 60 ) {
4504 $secondsPart = 0;
4505 $minutes++;
4506 }
4507 if ( $minutes == 60 ) {
4508 $minutes = 0;
4509 $hours++;
4510 }
4511 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4512 $s .= ' ';
4513 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4514 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes' ] ) ) {
4515 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4516 }
4517 } else {
4518 $days = floor( $seconds / 86400 );
4519 if ( $format['avoid'] === 'avoidminutes' ) {
4520 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4521 if ( $hours == 24 ) {
4522 $hours = 0;
4523 $days++;
4524 }
4525 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4526 $s .= ' ';
4527 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4528 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4529 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4530 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4531 if ( $minutes == 60 ) {
4532 $minutes = 0;
4533 $hours++;
4534 }
4535 if ( $hours == 24 ) {
4536 $hours = 0;
4537 $days++;
4538 }
4539 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4540 $s .= ' ';
4541 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4542 $s .= ' ';
4543 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4544 } else {
4545 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4546 $s .= ' ';
4547 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4548 }
4549 }
4550 return $s;
4551 }
4552
4553 /**
4554 * Format a bitrate for output, using an appropriate
4555 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4556 * the magnitude in question.
4557 *
4558 * This use base 1000. For base 1024 use formatSize(), for another base
4559 * see formatComputingNumbers().
4560 *
4561 * @param int $bps
4562 * @return string
4563 */
4564 function formatBitrate( $bps ) {
4565 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4566 }
4567
4568 /**
4569 * @param int $size Size of the unit
4570 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4571 * @param string $messageKey Message key to be uesd
4572 * @return string
4573 */
4574 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4575 if ( $size <= 0 ) {
4576 return str_replace( '$1', $this->formatNum( $size ),
4577 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4578 );
4579 }
4580 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4581 $index = 0;
4582
4583 $maxIndex = count( $sizes ) - 1;
4584 while ( $size >= $boundary && $index < $maxIndex ) {
4585 $index++;
4586 $size /= $boundary;
4587 }
4588
4589 // For small sizes no decimal places necessary
4590 $round = 0;
4591 if ( $index > 1 ) {
4592 // For MB and bigger two decimal places are smarter
4593 $round = 2;
4594 }
4595 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4596
4597 $size = round( $size, $round );
4598 $text = $this->getMessageFromDB( $msg );
4599 return str_replace( '$1', $this->formatNum( $size ), $text );
4600 }
4601
4602 /**
4603 * Format a size in bytes for output, using an appropriate
4604 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4605 *
4606 * This method use base 1024. For base 1000 use formatBitrate(), for
4607 * another base see formatComputingNumbers()
4608 *
4609 * @param int $size Size to format
4610 * @return string Plain text (not HTML)
4611 */
4612 function formatSize( $size ) {
4613 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4614 }
4615
4616 /**
4617 * Make a list item, used by various special pages
4618 *
4619 * @param string $page Page link
4620 * @param string $details HTML safe text between brackets
4621 * @param bool $oppositedm Add the direction mark opposite to your
4622 * language, to display text properly
4623 * @return HTML escaped string
4624 */
4625 function specialList( $page, $details, $oppositedm = true ) {
4626 if ( !$details ) {
4627 return $page;
4628 }
4629
4630 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4631 return
4632 $page .
4633 $dirmark .
4634 $this->msg( 'word-separator' )->escaped() .
4635 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4636 }
4637
4638 /**
4639 * Generate (prev x| next x) (20|50|100...) type links for paging
4640 *
4641 * @param Title $title Title object to link
4642 * @param int $offset
4643 * @param int $limit
4644 * @param array $query Optional URL query parameter string
4645 * @param bool $atend Optional param for specified if this is the last page
4646 * @return string
4647 */
4648 public function viewPrevNext( Title $title, $offset, $limit,
4649 array $query = [], $atend = false
4650 ) {
4651 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4652
4653 # Make 'previous' link
4654 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4655 if ( $offset > 0 ) {
4656 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4657 $query, $prev, 'prevn-title', 'mw-prevlink' );
4658 } else {
4659 $plink = htmlspecialchars( $prev );
4660 }
4661
4662 # Make 'next' link
4663 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4664 if ( $atend ) {
4665 $nlink = htmlspecialchars( $next );
4666 } else {
4667 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4668 $query, $next, 'nextn-title', 'mw-nextlink' );
4669 }
4670
4671 # Make links to set number of items per page
4672 $numLinks = [];
4673 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4674 $numLinks[] = $this->numLink( $title, $offset, $num,
4675 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4676 }
4677
4678 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4679 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4680 }
4681
4682 /**
4683 * Helper function for viewPrevNext() that generates links
4684 *
4685 * @param Title $title Title object to link
4686 * @param int $offset
4687 * @param int $limit
4688 * @param array $query Extra query parameters
4689 * @param string $link Text to use for the link; will be escaped
4690 * @param string $tooltipMsg Name of the message to use as tooltip
4691 * @param string $class Value of the "class" attribute of the link
4692 * @return string HTML fragment
4693 */
4694 private function numLink( Title $title, $offset, $limit, array $query, $link,
4695 $tooltipMsg, $class
4696 ) {
4697 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4698 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4699 ->numParams( $limit )->text();
4700
4701 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4702 'title' => $tooltip, 'class' => $class ], $link );
4703 }
4704
4705 /**
4706 * Get the conversion rule title, if any.
4707 *
4708 * @return string
4709 */
4710 public function getConvRuleTitle() {
4711 return $this->mConverter->getConvRuleTitle();
4712 }
4713
4714 /**
4715 * Get the compiled plural rules for the language
4716 * @since 1.20
4717 * @return array Associative array with plural form, and plural rule as key-value pairs
4718 */
4719 public function getCompiledPluralRules() {
4720 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4721 $fallbacks = Language::getFallbacksFor( $this->mCode );
4722 if ( !$pluralRules ) {
4723 foreach ( $fallbacks as $fallbackCode ) {
4724 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4725 if ( $pluralRules ) {
4726 break;
4727 }
4728 }
4729 }
4730 return $pluralRules;
4731 }
4732
4733 /**
4734 * Get the plural rules for the language
4735 * @since 1.20
4736 * @return array Associative array with plural form number and plural rule as key-value pairs
4737 */
4738 public function getPluralRules() {
4739 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4740 $fallbacks = Language::getFallbacksFor( $this->mCode );
4741 if ( !$pluralRules ) {
4742 foreach ( $fallbacks as $fallbackCode ) {
4743 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4744 if ( $pluralRules ) {
4745 break;
4746 }
4747 }
4748 }
4749 return $pluralRules;
4750 }
4751
4752 /**
4753 * Get the plural rule types for the language
4754 * @since 1.22
4755 * @return array Associative array with plural form number and plural rule type as key-value pairs
4756 */
4757 public function getPluralRuleTypes() {
4758 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4759 $fallbacks = Language::getFallbacksFor( $this->mCode );
4760 if ( !$pluralRuleTypes ) {
4761 foreach ( $fallbacks as $fallbackCode ) {
4762 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4763 if ( $pluralRuleTypes ) {
4764 break;
4765 }
4766 }
4767 }
4768 return $pluralRuleTypes;
4769 }
4770
4771 /**
4772 * Find the index number of the plural rule appropriate for the given number
4773 * @param int $number
4774 * @return int The index number of the plural rule
4775 */
4776 public function getPluralRuleIndexNumber( $number ) {
4777 $pluralRules = $this->getCompiledPluralRules();
4778 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4779 return $form;
4780 }
4781
4782 /**
4783 * Find the plural rule type appropriate for the given number
4784 * For example, if the language is set to Arabic, getPluralType(5) should
4785 * return 'few'.
4786 * @since 1.22
4787 * @param int $number
4788 * @return string The name of the plural rule type, e.g. one, two, few, many
4789 */
4790 public function getPluralRuleType( $number ) {
4791 $index = $this->getPluralRuleIndexNumber( $number );
4792 $pluralRuleTypes = $this->getPluralRuleTypes();
4793 if ( isset( $pluralRuleTypes[$index] ) ) {
4794 return $pluralRuleTypes[$index];
4795 } else {
4796 return 'other';
4797 }
4798 }
4799 }
MediaWiki Core