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