a9bbc20c1d490b667bd6958e7cc349600906cd89
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code.
4 * See https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation for more information.
5 *
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
10 *
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
15 *
16 * You should have received a copy of the GNU General Public License along
17 * with this program; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
19 * http://www.gnu.org/copyleft/gpl.html
20 *
21 * @file
22 * @ingroup Language
23 */
24
25 /**
26 * @defgroup Language Language
27 */
28
29 use CLDRPluralRuleParser\Evaluator;
30 use Wikimedia\Assert\Assert;
31
32 /**
33 * Internationalisation code
34 * @ingroup Language
35 */
36 class Language {
37 /**
38 * Return autonyms in fetchLanguageName(s).
39 * @since 1.32
40 */
41 const AS_AUTONYMS = null;
42
43 /**
44 * Return all known languages in fetchLanguageName(s).
45 * @since 1.32
46 */
47 const ALL = 'all';
48
49 /**
50 * Return in fetchLanguageName(s) only the languages for which we have at
51 * least some localisation.
52 * @since 1.32
53 */
54 const SUPPORTED = 'mwfile';
55
56 /**
57 * @var LanguageConverter
58 */
59 public $mConverter;
60
61 public $mVariants, $mCode, $mLoaded = false;
62 public $mMagicExtensions = [];
63 private $mHtmlCode = null, $mParentLanguage = false;
64
65 public $dateFormatStrings = [];
66 public $mExtendedSpecialPageAliases;
67
68 /** @var array|null */
69 protected $namespaceNames;
70 protected $mNamespaceIds, $namespaceAliases;
71
72 /**
73 * ReplacementArray object caches
74 */
75 public $transformData = [];
76
77 /**
78 * @var LocalisationCache
79 */
80 public static $dataCache;
81
82 public static $mLangObjCache = [];
83
84 /**
85 * Return a fallback chain for messages in getFallbacksFor
86 * @since 1.32
87 */
88 const MESSAGES_FALLBACKS = 0;
89
90 /**
91 * Return a strict fallback chain in getFallbacksFor
92 * @since 1.32
93 */
94 const STRICT_FALLBACKS = 1;
95
96 public static $mWeekdayMsgs = [
97 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
98 'friday', 'saturday'
99 ];
100
101 public static $mWeekdayAbbrevMsgs = [
102 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
103 ];
104
105 public static $mMonthMsgs = [
106 'january', 'february', 'march', 'april', 'may_long', 'june',
107 'july', 'august', 'september', 'october', 'november',
108 'december'
109 ];
110 public static $mMonthGenMsgs = [
111 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
112 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
113 'december-gen'
114 ];
115 public static $mMonthAbbrevMsgs = [
116 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
117 'sep', 'oct', 'nov', 'dec'
118 ];
119
120 public static $mIranianCalendarMonthMsgs = [
121 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
122 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
123 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
124 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
125 ];
126
127 public static $mHebrewCalendarMonthMsgs = [
128 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
129 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
130 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
131 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
132 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
133 ];
134
135 public static $mHebrewCalendarMonthGenMsgs = [
136 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
137 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
138 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
139 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
140 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
141 ];
142
143 public static $mHijriCalendarMonthMsgs = [
144 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
145 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
146 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
147 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
148 ];
149
150 /**
151 * @since 1.20
152 * @var array
153 */
154 public static $durationIntervals = [
155 'millennia' => 31556952000,
156 'centuries' => 3155695200,
157 'decades' => 315569520,
158 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
159 'weeks' => 604800,
160 'days' => 86400,
161 'hours' => 3600,
162 'minutes' => 60,
163 'seconds' => 1,
164 ];
165
166 /**
167 * Cache for language fallbacks.
168 * @see Language::getFallbacksIncludingSiteLanguage
169 * @since 1.21
170 * @var array
171 */
172 private static $fallbackLanguageCache = [];
173
174 /**
175 * Cache for grammar rules data
176 * @var MapCacheLRU|null
177 */
178 private static $grammarTransformations;
179
180 /**
181 * Cache for language names
182 * @var HashBagOStuff|null
183 */
184 private static $languageNameCache;
185
186 /**
187 * Unicode directional formatting characters, for embedBidi()
188 */
189 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
190 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
191 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
192
193 /**
194 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
195 * - in group 1 if it is LTR
196 * - in group 2 if it is RTL
197 * Does not match if there is no strong directionality codepoint.
198 *
199 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
200 *
201 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
202 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
203 */
204 // @codeCoverageIgnoreStart
205 // phpcs:ignore Generic.Files.LineLength
206 private static $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';
207 // @codeCoverageIgnoreEnd
208
209 /**
210 * Get a cached or new language object for a given language code
211 * @param string $code
212 * @throws MWException
213 * @return Language
214 */
215 static function factory( $code ) {
216 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
217
218 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
219 $code = $wgDummyLanguageCodes[$code];
220 }
221
222 // get the language object to process
223 $langObj = self::$mLangObjCache[$code] ?? self::newFromCode( $code );
224
225 // merge the language object in to get it up front in the cache
226 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
227 // get rid of the oldest ones in case we have an overflow
228 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
229
230 return $langObj;
231 }
232
233 /**
234 * Create a language object for a given language code
235 * @param string $code
236 * @param bool $fallback Whether we're going through language fallback chain
237 * @throws MWException
238 * @return Language
239 */
240 protected static function newFromCode( $code, $fallback = false ) {
241 if ( !self::isValidCode( $code ) ) {
242 throw new MWException( "Invalid language code \"$code\"" );
243 }
244
245 if ( !self::isValidBuiltInCode( $code ) ) {
246 // It's not possible to customise this code with class files, so
247 // just return a Language object. This is to support uselang= hacks.
248 $lang = new Language;
249 $lang->mCode = $code;
250 return $lang;
251 }
252
253 // Check if there is a language class for the code
254 $class = self::classFromCode( $code, $fallback );
255 // LanguageCode does not inherit Language
256 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
257 $lang = new $class;
258 return $lang;
259 }
260
261 // Keep trying the fallback list until we find an existing class
262 $fallbacks = self::getFallbacksFor( $code );
263 foreach ( $fallbacks as $fallbackCode ) {
264 if ( !self::isValidBuiltInCode( $fallbackCode ) ) {
265 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
266 }
267
268 $class = self::classFromCode( $fallbackCode );
269 if ( class_exists( $class ) ) {
270 $lang = new $class;
271 $lang->mCode = $code;
272 return $lang;
273 }
274 }
275
276 throw new MWException( "Invalid fallback sequence for language '$code'" );
277 }
278
279 /**
280 * Intended for tests that may change configuration in a way that invalidates caches.
281 *
282 * @since 1.32
283 */
284 public static function clearCaches() {
285 if ( !defined( 'MW_PHPUNIT_TEST' ) ) {
286 throw new MWException( __METHOD__ . ' must not be used outside tests' );
287 }
288 self::$dataCache = null;
289 // Reinitialize $dataCache, since it's expected to always be available
290 self::getLocalisationCache();
291 self::$mLangObjCache = [];
292 self::$fallbackLanguageCache = [];
293 self::$grammarTransformations = null;
294 self::$languageNameCache = null;
295 }
296
297 /**
298 * Checks whether any localisation is available for that language tag
299 * in MediaWiki (MessagesXx.php exists).
300 *
301 * @param string $code Language tag (in lower case)
302 * @return bool Whether language is supported
303 * @since 1.21
304 */
305 public static function isSupportedLanguage( $code ) {
306 if ( !self::isValidBuiltInCode( $code ) ) {
307 return false;
308 }
309
310 if ( $code === 'qqq' ) {
311 return false;
312 }
313
314 return is_readable( self::getMessagesFileName( $code ) ) ||
315 is_readable( self::getJsonMessagesFileName( $code ) );
316 }
317
318 /**
319 * Returns true if a language code string is a well-formed language tag
320 * according to RFC 5646.
321 * This function only checks well-formedness; it doesn't check that
322 * language, script or variant codes actually exist in the repositories.
323 *
324 * Based on regexes by Mark Davis of the Unicode Consortium:
325 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
326 *
327 * @param string $code
328 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
329 *
330 * @return bool
331 * @since 1.21
332 */
333 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
334 $alpha = '[a-z]';
335 $digit = '[0-9]';
336 $alphanum = '[a-z0-9]';
337 $x = 'x'; # private use singleton
338 $singleton = '[a-wy-z]'; # other singleton
339 $s = $lenient ? '[-_]' : '-';
340
341 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
342 $script = "$alpha{4}"; # ISO 15924
343 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
344 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
345 $extension = "$singleton(?:$s$alphanum{2,8})+";
346 $privateUse = "$x(?:$s$alphanum{1,8})+";
347
348 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
349 # Since these are limited, this is safe even later changes to the registry --
350 # the only oddity is that it might change the type of the tag, and thus
351 # the results from the capturing groups.
352 # https://www.iana.org/assignments/language-subtag-registry
353
354 $grandfathered = "en{$s}GB{$s}oed"
355 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
356 . "|no{$s}(?:bok|nyn)"
357 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
358 . "|zh{$s}min{$s}nan";
359
360 $variantList = "$variant(?:$s$variant)*";
361 $extensionList = "$extension(?:$s$extension)*";
362
363 $langtag = "(?:($language)"
364 . "(?:$s$script)?"
365 . "(?:$s$region)?"
366 . "(?:$s$variantList)?"
367 . "(?:$s$extensionList)?"
368 . "(?:$s$privateUse)?)";
369
370 # The final breakdown, with capturing groups for each of these components
371 # The variants, extensions, grandfathered, and private-use may have interior '-'
372
373 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
374
375 return (bool)preg_match( "/$root/", strtolower( $code ) );
376 }
377
378 /**
379 * Returns true if a language code string is of a valid form, whether or
380 * not it exists. This includes codes which are used solely for
381 * customisation via the MediaWiki namespace.
382 *
383 * @param string $code
384 *
385 * @return bool
386 */
387 public static function isValidCode( $code ) {
388 static $cache = [];
389 Assert::parameterType( 'string', $code, '$code' );
390 if ( !isset( $cache[$code] ) ) {
391 // People think language codes are html safe, so enforce it.
392 // Ideally we should only allow a-zA-Z0-9-
393 // but, .+ and other chars are often used for {{int:}} hacks
394 // see bugs T39564, T39587, T38938
395 $cache[$code] =
396 // Protect against path traversal
397 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
398 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
399 }
400 return $cache[$code];
401 }
402
403 /**
404 * Returns true if a language code is of a valid form for the purposes of
405 * internal customisation of MediaWiki, via Messages*.php or *.json.
406 *
407 * @param string $code
408 *
409 * @since 1.18
410 * @return bool
411 */
412 public static function isValidBuiltInCode( $code ) {
413 Assert::parameterType( 'string', $code, '$code' );
414
415 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
416 }
417
418 /**
419 * Returns true if a language code is an IETF tag known to MediaWiki.
420 *
421 * @param string $tag
422 *
423 * @since 1.21
424 * @return bool
425 */
426 public static function isKnownLanguageTag( $tag ) {
427 // Quick escape for invalid input to avoid exceptions down the line
428 // when code tries to process tags which are not valid at all.
429 if ( !self::isValidBuiltInCode( $tag ) ) {
430 return false;
431 }
432
433 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
434 || self::fetchLanguageName( $tag, $tag ) !== ''
435 ) {
436 return true;
437 }
438
439 return false;
440 }
441
442 /**
443 * Get the LocalisationCache instance
444 *
445 * @return LocalisationCache
446 */
447 public static function getLocalisationCache() {
448 if ( is_null( self::$dataCache ) ) {
449 global $wgLocalisationCacheConf;
450 $class = $wgLocalisationCacheConf['class'];
451 self::$dataCache = new $class( $wgLocalisationCacheConf );
452 }
453 return self::$dataCache;
454 }
455
456 function __construct() {
457 $this->mConverter = new FakeConverter( $this );
458 // Set the code to the name of the descendant
459 if ( static::class === 'Language' ) {
460 $this->mCode = 'en';
461 } else {
462 $this->mCode = str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
463 }
464 self::getLocalisationCache();
465 }
466
467 /**
468 * Reduce memory usage
469 */
470 function __destruct() {
471 foreach ( $this as $name => $value ) {
472 unset( $this->$name );
473 }
474 }
475
476 /**
477 * Hook which will be called if this is the content language.
478 * Descendants can use this to register hook functions or modify globals
479 */
480 function initContLang() {
481 }
482
483 /**
484 * @return array
485 * @since 1.19
486 */
487 public function getFallbackLanguages() {
488 return self::getFallbacksFor( $this->mCode );
489 }
490
491 /**
492 * Exports $wgBookstoreListEn
493 * @return array
494 */
495 public function getBookstoreList() {
496 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
497 }
498
499 /**
500 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
501 * available in localised form, it will be included in English.
502 *
503 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
504 */
505 public function getNamespaces() {
506 if ( is_null( $this->namespaceNames ) ) {
507 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
508
509 $validNamespaces = MWNamespace::getCanonicalNamespaces();
510
511 $this->namespaceNames = $wgExtraNamespaces +
512 self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
513 $this->namespaceNames += $validNamespaces;
514
515 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
516 if ( $wgMetaNamespaceTalk ) {
517 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
518 } else {
519 $talk = $this->namespaceNames[NS_PROJECT_TALK];
520 $this->namespaceNames[NS_PROJECT_TALK] =
521 $this->fixVariableInNamespace( $talk );
522 }
523
524 # Sometimes a language will be localised but not actually exist on this wiki.
525 foreach ( $this->namespaceNames as $key => $text ) {
526 if ( !isset( $validNamespaces[$key] ) ) {
527 unset( $this->namespaceNames[$key] );
528 }
529 }
530
531 # The above mixing may leave namespaces out of canonical order.
532 # Re-order by namespace ID number...
533 ksort( $this->namespaceNames );
534
535 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
536 }
537
538 return $this->namespaceNames;
539 }
540
541 /**
542 * Arbitrarily set all of the namespace names at once. Mainly used for testing
543 * @param array $namespaces Array of namespaces (id => name)
544 */
545 public function setNamespaces( array $namespaces ) {
546 $this->namespaceNames = $namespaces;
547 $this->mNamespaceIds = null;
548 }
549
550 /**
551 * Resets all of the namespace caches. Mainly used for testing
552 */
553 public function resetNamespaces() {
554 $this->namespaceNames = null;
555 $this->mNamespaceIds = null;
556 $this->namespaceAliases = null;
557 }
558
559 /**
560 * A convenience function that returns getNamespaces() with spaces instead of underscores
561 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
562 *
563 * @return array
564 */
565 public function getFormattedNamespaces() {
566 $ns = $this->getNamespaces();
567 foreach ( $ns as $k => $v ) {
568 $ns[$k] = strtr( $v, '_', ' ' );
569 }
570 return $ns;
571 }
572
573 /**
574 * Get a namespace value by key
575 *
576 * <code>
577 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
578 * echo $mw_ns; // prints 'MediaWiki'
579 * </code>
580 *
581 * @param int $index The array key of the namespace to return
582 * @return string|bool String if the namespace value exists, otherwise false
583 */
584 public function getNsText( $index ) {
585 $ns = $this->getNamespaces();
586 return $ns[$index] ?? false;
587 }
588
589 /**
590 * A convenience function that returns the same thing as
591 * getNsText() except with '_' changed to ' ', useful for
592 * producing output.
593 *
594 * <code>
595 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
596 * echo $mw_ns; // prints 'MediaWiki talk'
597 * </code>
598 *
599 * @param int $index The array key of the namespace to return
600 * @return string Namespace name without underscores (empty string if namespace does not exist)
601 */
602 public function getFormattedNsText( $index ) {
603 $ns = $this->getNsText( $index );
604 return strtr( $ns, '_', ' ' );
605 }
606
607 /**
608 * Returns gender-dependent namespace alias if available.
609 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
610 * @param int $index Namespace index
611 * @param string $gender Gender key (male, female... )
612 * @return string
613 * @since 1.18
614 */
615 public function getGenderNsText( $index, $gender ) {
616 global $wgExtraGenderNamespaces;
617
618 $ns = $wgExtraGenderNamespaces +
619 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
620
621 return $ns[$index][$gender] ?? $this->getNsText( $index );
622 }
623
624 /**
625 * Whether this language uses gender-dependent namespace aliases.
626 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
627 * @return bool
628 * @since 1.18
629 */
630 public function needsGenderDistinction() {
631 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
632 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
633 // $wgExtraGenderNamespaces overrides everything
634 return true;
635 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
636 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
637 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
638 return false;
639 } else {
640 // Check what is in i18n files
641 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
642 return count( $aliases ) > 0;
643 }
644 }
645
646 /**
647 * Get a namespace key by value, case insensitive.
648 * Only matches namespace names for the current language, not the
649 * canonical ones defined in Namespace.php.
650 *
651 * @param string $text
652 * @return int|bool An integer if $text is a valid value otherwise false
653 */
654 function getLocalNsIndex( $text ) {
655 $lctext = $this->lc( $text );
656 $ids = $this->getNamespaceIds();
657 return $ids[$lctext] ?? false;
658 }
659
660 /**
661 * @return array
662 */
663 public function getNamespaceAliases() {
664 if ( is_null( $this->namespaceAliases ) ) {
665 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
666 if ( !$aliases ) {
667 $aliases = [];
668 } else {
669 foreach ( $aliases as $name => $index ) {
670 if ( $index === NS_PROJECT_TALK ) {
671 unset( $aliases[$name] );
672 $name = $this->fixVariableInNamespace( $name );
673 $aliases[$name] = $index;
674 }
675 }
676 }
677
678 global $wgExtraGenderNamespaces;
679 $genders = $wgExtraGenderNamespaces +
680 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
681 foreach ( $genders as $index => $forms ) {
682 foreach ( $forms as $alias ) {
683 $aliases[$alias] = $index;
684 }
685 }
686
687 # Also add converted namespace names as aliases, to avoid confusion.
688 $convertedNames = [];
689 foreach ( $this->getVariants() as $variant ) {
690 if ( $variant === $this->mCode ) {
691 continue;
692 }
693 foreach ( $this->getNamespaces() as $ns => $_ ) {
694 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
695 }
696 }
697
698 $this->namespaceAliases = $aliases + $convertedNames;
699
700 # Filter out aliases to namespaces that don't exist, e.g. from extensions
701 # that aren't loaded here but are included in the l10n cache.
702 # (array_intersect preserves keys from its first argument)
703 $this->namespaceAliases = array_intersect(
704 $this->namespaceAliases,
705 array_keys( $this->getNamespaces() )
706 );
707 }
708
709 return $this->namespaceAliases;
710 }
711
712 /**
713 * @return array
714 */
715 public function getNamespaceIds() {
716 if ( is_null( $this->mNamespaceIds ) ) {
717 global $wgNamespaceAliases;
718 # Put namespace names and aliases into a hashtable.
719 # If this is too slow, then we should arrange it so that it is done
720 # before caching. The catch is that at pre-cache time, the above
721 # class-specific fixup hasn't been done.
722 $this->mNamespaceIds = [];
723 foreach ( $this->getNamespaces() as $index => $name ) {
724 $this->mNamespaceIds[$this->lc( $name )] = $index;
725 }
726 foreach ( $this->getNamespaceAliases() as $name => $index ) {
727 $this->mNamespaceIds[$this->lc( $name )] = $index;
728 }
729 if ( $wgNamespaceAliases ) {
730 foreach ( $wgNamespaceAliases as $name => $index ) {
731 $this->mNamespaceIds[$this->lc( $name )] = $index;
732 }
733 }
734 }
735 return $this->mNamespaceIds;
736 }
737
738 /**
739 * Get a namespace key by value, case insensitive. Canonical namespace
740 * names override custom ones defined for the current language.
741 *
742 * @param string $text
743 * @return int|bool An integer if $text is a valid value otherwise false
744 */
745 public function getNsIndex( $text ) {
746 $lctext = $this->lc( $text );
747 $ns = MWNamespace::getCanonicalIndex( $lctext );
748 if ( $ns !== null ) {
749 return $ns;
750 }
751 $ids = $this->getNamespaceIds();
752 return $ids[$lctext] ?? false;
753 }
754
755 /**
756 * short names for language variants used for language conversion links.
757 *
758 * @param string $code
759 * @param bool $usemsg Use the "variantname-xyz" message if it exists
760 * @return string
761 */
762 public function getVariantname( $code, $usemsg = true ) {
763 $msg = "variantname-$code";
764 if ( $usemsg && wfMessage( $msg )->exists() ) {
765 return $this->getMessageFromDB( $msg );
766 }
767 $name = self::fetchLanguageName( $code );
768 if ( $name ) {
769 return $name; # if it's defined as a language name, show that
770 } else {
771 # otherwise, output the language code
772 return $code;
773 }
774 }
775
776 /**
777 * @return string[]|bool List of date format preference keys, or false if disabled.
778 */
779 public function getDatePreferences() {
780 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
781 }
782
783 /**
784 * @return array
785 */
786 function getDateFormats() {
787 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
788 }
789
790 /**
791 * @return array|string
792 */
793 public function getDefaultDateFormat() {
794 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
795 if ( $df === 'dmy or mdy' ) {
796 global $wgAmericanDates;
797 return $wgAmericanDates ? 'mdy' : 'dmy';
798 } else {
799 return $df;
800 }
801 }
802
803 /**
804 * @return array
805 */
806 public function getDatePreferenceMigrationMap() {
807 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
808 }
809
810 /**
811 * @return array
812 */
813 public function getExtraUserToggles() {
814 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
815 }
816
817 /**
818 * @param string $tog
819 * @return string
820 */
821 function getUserToggle( $tog ) {
822 return $this->getMessageFromDB( "tog-$tog" );
823 }
824
825 /**
826 * Get an array of language names, indexed by code.
827 * @param null|string $inLanguage Code of language in which to return the names
828 * Use self::AS_AUTONYMS for autonyms (native names)
829 * @param string $include One of:
830 * self::ALL all available languages
831 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
832 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
833 * @return array Language code => language name (sorted by key)
834 * @since 1.20
835 */
836 public static function fetchLanguageNames( $inLanguage = self::AS_AUTONYMS, $include = 'mw' ) {
837 $cacheKey = $inLanguage === self::AS_AUTONYMS ? 'null' : $inLanguage;
838 $cacheKey .= ":$include";
839 if ( self::$languageNameCache === null ) {
840 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
841 }
842
843 $ret = self::$languageNameCache->get( $cacheKey );
844 if ( !$ret ) {
845 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
846 self::$languageNameCache->set( $cacheKey, $ret );
847 }
848 return $ret;
849 }
850
851 /**
852 * Uncached helper for fetchLanguageNames
853 * @param null|string $inLanguage Code of language in which to return the names
854 * Use self::AS_AUTONYMS for autonyms (native names)
855 * @param string $include One of:
856 * self::ALL all available languages
857 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
858 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
859 * @return array Language code => language name (sorted by key)
860 */
861 private static function fetchLanguageNamesUncached(
862 $inLanguage = self::AS_AUTONYMS,
863 $include = 'mw'
864 ) {
865 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
866
867 // If passed an invalid language code to use, fallback to en
868 if ( $inLanguage !== self::AS_AUTONYMS && !self::isValidCode( $inLanguage ) ) {
869 $inLanguage = 'en';
870 }
871
872 $names = [];
873
874 if ( $inLanguage ) {
875 # TODO: also include when $inLanguage is null, when this code is more efficient
876 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
877 }
878
879 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
880 if ( $wgUsePigLatinVariant ) {
881 // Pig Latin (for variant development)
882 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
883 }
884
885 foreach ( $mwNames as $mwCode => $mwName ) {
886 # - Prefer own MediaWiki native name when not using the hook
887 # - For other names just add if not added through the hook
888 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
889 $names[$mwCode] = $mwName;
890 }
891 }
892
893 if ( $include === self::ALL ) {
894 ksort( $names );
895 return $names;
896 }
897
898 $returnMw = [];
899 $coreCodes = array_keys( $mwNames );
900 foreach ( $coreCodes as $coreCode ) {
901 $returnMw[$coreCode] = $names[$coreCode];
902 }
903
904 if ( $include === self::SUPPORTED ) {
905 $namesMwFile = [];
906 # We do this using a foreach over the codes instead of a directory
907 # loop so that messages files in extensions will work correctly.
908 foreach ( $returnMw as $code => $value ) {
909 if ( is_readable( self::getMessagesFileName( $code ) )
910 || is_readable( self::getJsonMessagesFileName( $code ) )
911 ) {
912 $namesMwFile[$code] = $names[$code];
913 }
914 }
915
916 ksort( $namesMwFile );
917 return $namesMwFile;
918 }
919
920 ksort( $returnMw );
921 # 'mw' option; default if it's not one of the other two options (all/mwfile)
922 return $returnMw;
923 }
924
925 /**
926 * @param string $code The code of the language for which to get the name
927 * @param null|string $inLanguage Code of language in which to return the name
928 * (SELF::AS_AUTONYMS for autonyms)
929 * @param string $include See fetchLanguageNames()
930 * @return string Language name or empty
931 * @since 1.20
932 */
933 public static function fetchLanguageName(
934 $code,
935 $inLanguage = self::AS_AUTONYMS,
936 $include = self::ALL
937 ) {
938 $code = strtolower( $code );
939 $array = self::fetchLanguageNames( $inLanguage, $include );
940 return !array_key_exists( $code, $array ) ? '' : $array[$code];
941 }
942
943 /**
944 * Get a message from the MediaWiki namespace.
945 *
946 * @param string $msg Message name
947 * @return string
948 */
949 public function getMessageFromDB( $msg ) {
950 return $this->msg( $msg )->text();
951 }
952
953 /**
954 * Get message object in this language. Only for use inside this class.
955 *
956 * @param string $msg Message name
957 * @return Message
958 */
959 protected function msg( $msg ) {
960 return wfMessage( $msg )->inLanguage( $this );
961 }
962
963 /**
964 * @param string $key
965 * @return string
966 */
967 public function getMonthName( $key ) {
968 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
969 }
970
971 /**
972 * @return array
973 */
974 public function getMonthNamesArray() {
975 $monthNames = [ '' ];
976 for ( $i = 1; $i < 13; $i++ ) {
977 $monthNames[] = $this->getMonthName( $i );
978 }
979 return $monthNames;
980 }
981
982 /**
983 * @param string $key
984 * @return string
985 */
986 public function getMonthNameGen( $key ) {
987 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
988 }
989
990 /**
991 * @param string $key
992 * @return string
993 */
994 public function getMonthAbbreviation( $key ) {
995 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
996 }
997
998 /**
999 * @return array
1000 */
1001 public function getMonthAbbreviationsArray() {
1002 $monthNames = [ '' ];
1003 for ( $i = 1; $i < 13; $i++ ) {
1004 $monthNames[] = $this->getMonthAbbreviation( $i );
1005 }
1006 return $monthNames;
1007 }
1008
1009 /**
1010 * @param string $key
1011 * @return string
1012 */
1013 public function getWeekdayName( $key ) {
1014 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
1015 }
1016
1017 /**
1018 * @param string $key
1019 * @return string
1020 */
1021 function getWeekdayAbbreviation( $key ) {
1022 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1023 }
1024
1025 /**
1026 * @param string $key
1027 * @return string
1028 */
1029 function getIranianCalendarMonthName( $key ) {
1030 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1031 }
1032
1033 /**
1034 * @param string $key
1035 * @return string
1036 */
1037 function getHebrewCalendarMonthName( $key ) {
1038 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1039 }
1040
1041 /**
1042 * @param string $key
1043 * @return string
1044 */
1045 function getHebrewCalendarMonthNameGen( $key ) {
1046 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1047 }
1048
1049 /**
1050 * @param string $key
1051 * @return string
1052 */
1053 function getHijriCalendarMonthName( $key ) {
1054 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1055 }
1056
1057 /**
1058 * Pass through result from $dateTimeObj->format()
1059 * @param DateTime|bool|null &$dateTimeObj
1060 * @param string $ts
1061 * @param DateTimeZone|bool|null $zone
1062 * @param string $code
1063 * @return string
1064 */
1065 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1066 if ( !$dateTimeObj ) {
1067 $dateTimeObj = DateTime::createFromFormat(
1068 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1069 );
1070 }
1071 return $dateTimeObj->format( $code );
1072 }
1073
1074 /**
1075 * This is a workalike of PHP's date() function, but with better
1076 * internationalisation, a reduced set of format characters, and a better
1077 * escaping format.
1078 *
1079 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1080 * the PHP manual for definitions. There are a number of extensions, which
1081 * start with "x":
1082 *
1083 * xn Do not translate digits of the next numeric format character
1084 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1085 * xr Use roman numerals for the next numeric format character
1086 * xh Use hebrew numerals for the next numeric format character
1087 * xx Literal x
1088 * xg Genitive month name
1089 *
1090 * xij j (day number) in Iranian calendar
1091 * xiF F (month name) in Iranian calendar
1092 * xin n (month number) in Iranian calendar
1093 * xiy y (two digit year) in Iranian calendar
1094 * xiY Y (full year) in Iranian calendar
1095 * xit t (days in month) in Iranian calendar
1096 * xiz z (day of the year) in Iranian calendar
1097 *
1098 * xjj j (day number) in Hebrew calendar
1099 * xjF F (month name) in Hebrew calendar
1100 * xjt t (days in month) in Hebrew calendar
1101 * xjx xg (genitive month name) in Hebrew calendar
1102 * xjn n (month number) in Hebrew calendar
1103 * xjY Y (full year) in Hebrew calendar
1104 *
1105 * xmj j (day number) in Hijri calendar
1106 * xmF F (month name) in Hijri calendar
1107 * xmn n (month number) in Hijri calendar
1108 * xmY Y (full year) in Hijri calendar
1109 *
1110 * xkY Y (full year) in Thai solar calendar. Months and days are
1111 * identical to the Gregorian calendar
1112 * xoY Y (full year) in Minguo calendar or Juche year.
1113 * Months and days are identical to the
1114 * Gregorian calendar
1115 * xtY Y (full year) in Japanese nengo. Months and days are
1116 * identical to the Gregorian calendar
1117 *
1118 * Characters enclosed in double quotes will be considered literal (with
1119 * the quotes themselves removed). Unmatched quotes will be considered
1120 * literal quotes. Example:
1121 *
1122 * "The month is" F => The month is January
1123 * i's" => 20'11"
1124 *
1125 * Backslash escaping is also supported.
1126 *
1127 * Input timestamp is assumed to be pre-normalized to the desired local
1128 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1129 * $ts is UTC if $zone is not given.
1130 *
1131 * @param string $format
1132 * @param string $ts 14-character timestamp
1133 * YYYYMMDDHHMMSS
1134 * 01234567890123
1135 * @param DateTimeZone|null $zone Timezone of $ts
1136 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1137 * Only makes sense if $ts is the current time.
1138 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1139 *
1140 * @throws MWException
1141 * @return string
1142 */
1143 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1144 $s = '';
1145 $raw = false;
1146 $roman = false;
1147 $hebrewNum = false;
1148 $dateTimeObj = false;
1149 $rawToggle = false;
1150 $iranian = false;
1151 $hebrew = false;
1152 $hijri = false;
1153 $thai = false;
1154 $minguo = false;
1155 $tenno = false;
1156
1157 $usedSecond = false;
1158 $usedMinute = false;
1159 $usedHour = false;
1160 $usedAMPM = false;
1161 $usedDay = false;
1162 $usedWeek = false;
1163 $usedMonth = false;
1164 $usedYear = false;
1165 $usedISOYear = false;
1166 $usedIsLeapYear = false;
1167
1168 $usedHebrewMonth = false;
1169 $usedIranianMonth = false;
1170 $usedHijriMonth = false;
1171 $usedHebrewYear = false;
1172 $usedIranianYear = false;
1173 $usedHijriYear = false;
1174 $usedTennoYear = false;
1175
1176 if ( strlen( $ts ) !== 14 ) {
1177 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1178 }
1179
1180 if ( !ctype_digit( $ts ) ) {
1181 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1182 }
1183
1184 $formatLength = strlen( $format );
1185 for ( $p = 0; $p < $formatLength; $p++ ) {
1186 $num = false;
1187 $code = $format[$p];
1188 if ( $code == 'x' && $p < $formatLength - 1 ) {
1189 $code .= $format[++$p];
1190 }
1191
1192 if ( ( $code === 'xi'
1193 || $code === 'xj'
1194 || $code === 'xk'
1195 || $code === 'xm'
1196 || $code === 'xo'
1197 || $code === 'xt' )
1198 && $p < $formatLength - 1 ) {
1199 $code .= $format[++$p];
1200 }
1201
1202 switch ( $code ) {
1203 case 'xx':
1204 $s .= 'x';
1205 break;
1206 case 'xn':
1207 $raw = true;
1208 break;
1209 case 'xN':
1210 $rawToggle = !$rawToggle;
1211 break;
1212 case 'xr':
1213 $roman = true;
1214 break;
1215 case 'xh':
1216 $hebrewNum = true;
1217 break;
1218 case 'xg':
1219 $usedMonth = true;
1220 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1221 break;
1222 case 'xjx':
1223 $usedHebrewMonth = true;
1224 if ( !$hebrew ) {
1225 $hebrew = self::tsToHebrew( $ts );
1226 }
1227 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1228 break;
1229 case 'd':
1230 $usedDay = true;
1231 $num = substr( $ts, 6, 2 );
1232 break;
1233 case 'D':
1234 $usedDay = true;
1235 $s .= $this->getWeekdayAbbreviation(
1236 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1237 );
1238 break;
1239 case 'j':
1240 $usedDay = true;
1241 $num = intval( substr( $ts, 6, 2 ) );
1242 break;
1243 case 'xij':
1244 $usedDay = true;
1245 if ( !$iranian ) {
1246 $iranian = self::tsToIranian( $ts );
1247 }
1248 $num = $iranian[2];
1249 break;
1250 case 'xmj':
1251 $usedDay = true;
1252 if ( !$hijri ) {
1253 $hijri = self::tsToHijri( $ts );
1254 }
1255 $num = $hijri[2];
1256 break;
1257 case 'xjj':
1258 $usedDay = true;
1259 if ( !$hebrew ) {
1260 $hebrew = self::tsToHebrew( $ts );
1261 }
1262 $num = $hebrew[2];
1263 break;
1264 case 'l':
1265 $usedDay = true;
1266 $s .= $this->getWeekdayName(
1267 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1268 );
1269 break;
1270 case 'F':
1271 $usedMonth = true;
1272 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1273 break;
1274 case 'xiF':
1275 $usedIranianMonth = true;
1276 if ( !$iranian ) {
1277 $iranian = self::tsToIranian( $ts );
1278 }
1279 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1280 break;
1281 case 'xmF':
1282 $usedHijriMonth = true;
1283 if ( !$hijri ) {
1284 $hijri = self::tsToHijri( $ts );
1285 }
1286 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1287 break;
1288 case 'xjF':
1289 $usedHebrewMonth = true;
1290 if ( !$hebrew ) {
1291 $hebrew = self::tsToHebrew( $ts );
1292 }
1293 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1294 break;
1295 case 'm':
1296 $usedMonth = true;
1297 $num = substr( $ts, 4, 2 );
1298 break;
1299 case 'M':
1300 $usedMonth = true;
1301 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1302 break;
1303 case 'n':
1304 $usedMonth = true;
1305 $num = intval( substr( $ts, 4, 2 ) );
1306 break;
1307 case 'xin':
1308 $usedIranianMonth = true;
1309 if ( !$iranian ) {
1310 $iranian = self::tsToIranian( $ts );
1311 }
1312 $num = $iranian[1];
1313 break;
1314 case 'xmn':
1315 $usedHijriMonth = true;
1316 if ( !$hijri ) {
1317 $hijri = self::tsToHijri( $ts );
1318 }
1319 $num = $hijri[1];
1320 break;
1321 case 'xjn':
1322 $usedHebrewMonth = true;
1323 if ( !$hebrew ) {
1324 $hebrew = self::tsToHebrew( $ts );
1325 }
1326 $num = $hebrew[1];
1327 break;
1328 case 'xjt':
1329 $usedHebrewMonth = true;
1330 if ( !$hebrew ) {
1331 $hebrew = self::tsToHebrew( $ts );
1332 }
1333 $num = $hebrew[3];
1334 break;
1335 case 'Y':
1336 $usedYear = true;
1337 $num = substr( $ts, 0, 4 );
1338 break;
1339 case 'xiY':
1340 $usedIranianYear = true;
1341 if ( !$iranian ) {
1342 $iranian = self::tsToIranian( $ts );
1343 }
1344 $num = $iranian[0];
1345 break;
1346 case 'xmY':
1347 $usedHijriYear = true;
1348 if ( !$hijri ) {
1349 $hijri = self::tsToHijri( $ts );
1350 }
1351 $num = $hijri[0];
1352 break;
1353 case 'xjY':
1354 $usedHebrewYear = true;
1355 if ( !$hebrew ) {
1356 $hebrew = self::tsToHebrew( $ts );
1357 }
1358 $num = $hebrew[0];
1359 break;
1360 case 'xkY':
1361 $usedYear = true;
1362 if ( !$thai ) {
1363 $thai = self::tsToYear( $ts, 'thai' );
1364 }
1365 $num = $thai[0];
1366 break;
1367 case 'xoY':
1368 $usedYear = true;
1369 if ( !$minguo ) {
1370 $minguo = self::tsToYear( $ts, 'minguo' );
1371 }
1372 $num = $minguo[0];
1373 break;
1374 case 'xtY':
1375 $usedTennoYear = true;
1376 if ( !$tenno ) {
1377 $tenno = self::tsToYear( $ts, 'tenno' );
1378 }
1379 $num = $tenno[0];
1380 break;
1381 case 'y':
1382 $usedYear = true;
1383 $num = substr( $ts, 2, 2 );
1384 break;
1385 case 'xiy':
1386 $usedIranianYear = true;
1387 if ( !$iranian ) {
1388 $iranian = self::tsToIranian( $ts );
1389 }
1390 $num = substr( $iranian[0], -2 );
1391 break;
1392 case 'xit':
1393 $usedIranianYear = true;
1394 if ( !$iranian ) {
1395 $iranian = self::tsToIranian( $ts );
1396 }
1397 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1398 break;
1399 case 'xiz':
1400 $usedIranianYear = true;
1401 if ( !$iranian ) {
1402 $iranian = self::tsToIranian( $ts );
1403 }
1404 $num = $iranian[3];
1405 break;
1406 case 'a':
1407 $usedAMPM = true;
1408 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1409 break;
1410 case 'A':
1411 $usedAMPM = true;
1412 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1413 break;
1414 case 'g':
1415 $usedHour = true;
1416 $h = substr( $ts, 8, 2 );
1417 $num = $h % 12 ?: 12;
1418 break;
1419 case 'G':
1420 $usedHour = true;
1421 $num = intval( substr( $ts, 8, 2 ) );
1422 break;
1423 case 'h':
1424 $usedHour = true;
1425 $h = substr( $ts, 8, 2 );
1426 $num = sprintf( '%02d', $h % 12 ?: 12 );
1427 break;
1428 case 'H':
1429 $usedHour = true;
1430 $num = substr( $ts, 8, 2 );
1431 break;
1432 case 'i':
1433 $usedMinute = true;
1434 $num = substr( $ts, 10, 2 );
1435 break;
1436 case 's':
1437 $usedSecond = true;
1438 $num = substr( $ts, 12, 2 );
1439 break;
1440 case 'c':
1441 case 'r':
1442 $usedSecond = true;
1443 // fall through
1444 case 'e':
1445 case 'O':
1446 case 'P':
1447 case 'T':
1448 $s .= self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1449 break;
1450 case 'w':
1451 case 'N':
1452 case 'z':
1453 $usedDay = true;
1454 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1455 break;
1456 case 'W':
1457 $usedWeek = true;
1458 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1459 break;
1460 case 't':
1461 $usedMonth = true;
1462 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1463 break;
1464 case 'L':
1465 $usedIsLeapYear = true;
1466 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1467 break;
1468 case 'o':
1469 $usedISOYear = true;
1470 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1471 break;
1472 case 'U':
1473 $usedSecond = true;
1474 // fall through
1475 case 'I':
1476 case 'Z':
1477 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1478 break;
1479 case '\\':
1480 # Backslash escaping
1481 if ( $p < $formatLength - 1 ) {
1482 $s .= $format[++$p];
1483 } else {
1484 $s .= '\\';
1485 }
1486 break;
1487 case '"':
1488 # Quoted literal
1489 if ( $p < $formatLength - 1 ) {
1490 $endQuote = strpos( $format, '"', $p + 1 );
1491 if ( $endQuote === false ) {
1492 # No terminating quote, assume literal "
1493 $s .= '"';
1494 } else {
1495 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1496 $p = $endQuote;
1497 }
1498 } else {
1499 # Quote at end of string, assume literal "
1500 $s .= '"';
1501 }
1502 break;
1503 default:
1504 $s .= $format[$p];
1505 }
1506 if ( $num !== false ) {
1507 if ( $rawToggle || $raw ) {
1508 $s .= $num;
1509 $raw = false;
1510 } elseif ( $roman ) {
1511 $s .= self::romanNumeral( $num );
1512 $roman = false;
1513 } elseif ( $hebrewNum ) {
1514 $s .= self::hebrewNumeral( $num );
1515 $hebrewNum = false;
1516 } else {
1517 $s .= $this->formatNum( $num, true );
1518 }
1519 }
1520 }
1521
1522 if ( $ttl === 'unused' ) {
1523 // No need to calculate the TTL, the caller wont use it anyway.
1524 } elseif ( $usedSecond ) {
1525 $ttl = 1;
1526 } elseif ( $usedMinute ) {
1527 $ttl = 60 - substr( $ts, 12, 2 );
1528 } elseif ( $usedHour ) {
1529 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1530 } elseif ( $usedAMPM ) {
1531 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1532 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1533 } elseif (
1534 $usedDay ||
1535 $usedHebrewMonth ||
1536 $usedIranianMonth ||
1537 $usedHijriMonth ||
1538 $usedHebrewYear ||
1539 $usedIranianYear ||
1540 $usedHijriYear ||
1541 $usedTennoYear
1542 ) {
1543 // @todo Someone who understands the non-Gregorian calendars
1544 // should write proper logic for them so that they don't need purged every day.
1545 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1546 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1547 } else {
1548 $possibleTtls = [];
1549 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1550 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1551 if ( $usedWeek ) {
1552 $possibleTtls[] =
1553 ( 7 - self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1554 $timeRemainingInDay;
1555 } elseif ( $usedISOYear ) {
1556 // December 28th falls on the last ISO week of the year, every year.
1557 // The last ISO week of a year can be 52 or 53.
1558 $lastWeekOfISOYear = DateTime::createFromFormat(
1559 'Ymd',
1560 substr( $ts, 0, 4 ) . '1228',
1561 $zone ?: new DateTimeZone( 'UTC' )
1562 )->format( 'W' );
1563 $currentISOWeek = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1564 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1565 $timeRemainingInWeek =
1566 ( 7 - self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1567 + $timeRemainingInDay;
1568 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1569 }
1570
1571 if ( $usedMonth ) {
1572 $possibleTtls[] =
1573 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1574 substr( $ts, 6, 2 ) ) * 86400
1575 + $timeRemainingInDay;
1576 } elseif ( $usedYear ) {
1577 $possibleTtls[] =
1578 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1579 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1580 + $timeRemainingInDay;
1581 } elseif ( $usedIsLeapYear ) {
1582 $year = substr( $ts, 0, 4 );
1583 $timeRemainingInYear =
1584 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1585 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1586 + $timeRemainingInDay;
1587 $mod = $year % 4;
1588 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1589 // this isn't a leap year. see when the next one starts
1590 $nextCandidate = $year - $mod + 4;
1591 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1592 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1593 $timeRemainingInYear;
1594 } else {
1595 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1596 $timeRemainingInYear;
1597 }
1598 } else {
1599 // this is a leap year, so the next year isn't
1600 $possibleTtls[] = $timeRemainingInYear;
1601 }
1602 }
1603
1604 if ( $possibleTtls ) {
1605 $ttl = min( $possibleTtls );
1606 }
1607 }
1608
1609 return $s;
1610 }
1611
1612 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1613 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1614
1615 /**
1616 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1617 * Gregorian dates to Iranian dates. Originally written in C, it
1618 * is released under the terms of GNU Lesser General Public
1619 * License. Conversion to PHP was performed by Niklas Laxström.
1620 *
1621 * Link: http://www.farsiweb.info/jalali/jalali.c
1622 *
1623 * @param string $ts
1624 *
1625 * @return int[]
1626 */
1627 private static function tsToIranian( $ts ) {
1628 $gy = substr( $ts, 0, 4 ) - 1600;
1629 $gm = substr( $ts, 4, 2 ) - 1;
1630 $gd = substr( $ts, 6, 2 ) - 1;
1631
1632 # Days passed from the beginning (including leap years)
1633 $gDayNo = 365 * $gy
1634 + floor( ( $gy + 3 ) / 4 )
1635 - floor( ( $gy + 99 ) / 100 )
1636 + floor( ( $gy + 399 ) / 400 );
1637
1638 // Add days of the past months of this year
1639 for ( $i = 0; $i < $gm; $i++ ) {
1640 $gDayNo += self::$GREG_DAYS[$i];
1641 }
1642
1643 // Leap years
1644 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1645 $gDayNo++;
1646 }
1647
1648 // Days passed in current month
1649 $gDayNo += (int)$gd;
1650
1651 $jDayNo = $gDayNo - 79;
1652
1653 $jNp = floor( $jDayNo / 12053 );
1654 $jDayNo %= 12053;
1655
1656 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1657 $jDayNo %= 1461;
1658
1659 if ( $jDayNo >= 366 ) {
1660 $jy += floor( ( $jDayNo - 1 ) / 365 );
1661 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1662 }
1663
1664 $jz = $jDayNo;
1665
1666 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1667 $jDayNo -= self::$IRANIAN_DAYS[$i];
1668 }
1669
1670 $jm = $i + 1;
1671 $jd = $jDayNo + 1;
1672
1673 return [ $jy, $jm, $jd, $jz ];
1674 }
1675
1676 /**
1677 * Converting Gregorian dates to Hijri dates.
1678 *
1679 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1680 *
1681 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1682 *
1683 * @param string $ts
1684 *
1685 * @return int[]
1686 */
1687 private static function tsToHijri( $ts ) {
1688 $year = substr( $ts, 0, 4 );
1689 $month = substr( $ts, 4, 2 );
1690 $day = substr( $ts, 6, 2 );
1691
1692 $zyr = $year;
1693 $zd = $day;
1694 $zm = $month;
1695 $zy = $zyr;
1696
1697 if (
1698 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1699 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1700 ) {
1701 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1702 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1703 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1704 $zd - 32075;
1705 } else {
1706 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1707 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1708 }
1709
1710 $zl = $zjd - 1948440 + 10632;
1711 $zn = (int)( ( $zl - 1 ) / 10631 );
1712 $zl = $zl - 10631 * $zn + 354;
1713 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1714 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1715 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1716 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1717 $zm = (int)( ( 24 * $zl ) / 709 );
1718 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1719 $zy = 30 * $zn + $zj - 30;
1720
1721 return [ $zy, $zm, $zd ];
1722 }
1723
1724 /**
1725 * Converting Gregorian dates to Hebrew dates.
1726 *
1727 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1728 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1729 * to translate the relevant functions into PHP and release them under
1730 * GNU GPL.
1731 *
1732 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1733 * and Adar II is 14. In a non-leap year, Adar is 6.
1734 *
1735 * @param string $ts
1736 *
1737 * @return int[]
1738 */
1739 private static function tsToHebrew( $ts ) {
1740 # Parse date
1741 $year = substr( $ts, 0, 4 );
1742 $month = substr( $ts, 4, 2 );
1743 $day = substr( $ts, 6, 2 );
1744
1745 # Calculate Hebrew year
1746 $hebrewYear = $year + 3760;
1747
1748 # Month number when September = 1, August = 12
1749 $month += 4;
1750 if ( $month > 12 ) {
1751 # Next year
1752 $month -= 12;
1753 $year++;
1754 $hebrewYear++;
1755 }
1756
1757 # Calculate day of year from 1 September
1758 $dayOfYear = $day;
1759 for ( $i = 1; $i < $month; $i++ ) {
1760 if ( $i == 6 ) {
1761 # February
1762 $dayOfYear += 28;
1763 # Check if the year is leap
1764 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1765 $dayOfYear++;
1766 }
1767 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1768 $dayOfYear += 30;
1769 } else {
1770 $dayOfYear += 31;
1771 }
1772 }
1773
1774 # Calculate the start of the Hebrew year
1775 $start = self::hebrewYearStart( $hebrewYear );
1776
1777 # Calculate next year's start
1778 if ( $dayOfYear <= $start ) {
1779 # Day is before the start of the year - it is the previous year
1780 # Next year's start
1781 $nextStart = $start;
1782 # Previous year
1783 $year--;
1784 $hebrewYear--;
1785 # Add days since previous year's 1 September
1786 $dayOfYear += 365;
1787 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1788 # Leap year
1789 $dayOfYear++;
1790 }
1791 # Start of the new (previous) year
1792 $start = self::hebrewYearStart( $hebrewYear );
1793 } else {
1794 # Next year's start
1795 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1796 }
1797
1798 # Calculate Hebrew day of year
1799 $hebrewDayOfYear = $dayOfYear - $start;
1800
1801 # Difference between year's days
1802 $diff = $nextStart - $start;
1803 # Add 12 (or 13 for leap years) days to ignore the difference between
1804 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1805 # difference is only about the year type
1806 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1807 $diff += 13;
1808 } else {
1809 $diff += 12;
1810 }
1811
1812 # Check the year pattern, and is leap year
1813 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1814 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1815 # and non-leap years
1816 $yearPattern = $diff % 30;
1817 # Check if leap year
1818 $isLeap = $diff >= 30;
1819
1820 # Calculate day in the month from number of day in the Hebrew year
1821 # Don't check Adar - if the day is not in Adar, we will stop before;
1822 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1823 $hebrewDay = $hebrewDayOfYear;
1824 $hebrewMonth = 1;
1825 $days = 0;
1826 while ( $hebrewMonth <= 12 ) {
1827 # Calculate days in this month
1828 if ( $isLeap && $hebrewMonth == 6 ) {
1829 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1830 $days = 30;
1831 if ( $hebrewDay <= $days ) {
1832 # Day in Adar I
1833 $hebrewMonth = 13;
1834 } else {
1835 # Subtract the days of Adar I
1836 $hebrewDay -= $days;
1837 # Try Adar II
1838 $days = 29;
1839 if ( $hebrewDay <= $days ) {
1840 # Day in Adar II
1841 $hebrewMonth = 14;
1842 }
1843 }
1844 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1845 # Cheshvan in a complete year (otherwise as the rule below)
1846 $days = 30;
1847 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1848 # Kislev in an incomplete year (otherwise as the rule below)
1849 $days = 29;
1850 } else {
1851 # Odd months have 30 days, even have 29
1852 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1853 }
1854 if ( $hebrewDay <= $days ) {
1855 # In the current month
1856 break;
1857 } else {
1858 # Subtract the days of the current month
1859 $hebrewDay -= $days;
1860 # Try in the next month
1861 $hebrewMonth++;
1862 }
1863 }
1864
1865 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1866 }
1867
1868 /**
1869 * This calculates the Hebrew year start, as days since 1 September.
1870 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1871 * Used for Hebrew date.
1872 *
1873 * @param int $year
1874 *
1875 * @return string
1876 */
1877 private static function hebrewYearStart( $year ) {
1878 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1879 $b = intval( ( $year - 1 ) % 4 );
1880 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1881 if ( $m < 0 ) {
1882 $m--;
1883 }
1884 $Mar = intval( $m );
1885 if ( $m < 0 ) {
1886 $m++;
1887 }
1888 $m -= $Mar;
1889
1890 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1891 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1892 $Mar++;
1893 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1894 $Mar += 2;
1895 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1896 $Mar++;
1897 }
1898
1899 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1900 return $Mar;
1901 }
1902
1903 /**
1904 * Algorithm to convert Gregorian dates to Thai solar dates,
1905 * Minguo dates or Minguo dates.
1906 *
1907 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1908 * https://en.wikipedia.org/wiki/Minguo_calendar
1909 * https://en.wikipedia.org/wiki/Japanese_era_name
1910 *
1911 * @param string $ts 14-character timestamp
1912 * @param string $cName Calender name
1913 * @return array Converted year, month, day
1914 */
1915 private static function tsToYear( $ts, $cName ) {
1916 $gy = substr( $ts, 0, 4 );
1917 $gm = substr( $ts, 4, 2 );
1918 $gd = substr( $ts, 6, 2 );
1919
1920 if ( !strcmp( $cName, 'thai' ) ) {
1921 # Thai solar dates
1922 # Add 543 years to the Gregorian calendar
1923 # Months and days are identical
1924 $gy_offset = $gy + 543;
1925 # fix for dates between 1912 and 1941
1926 # https://en.wikipedia.org/?oldid=836596673#New_year
1927 if ( $gy >= 1912 && $gy <= 1940 ) {
1928 if ( $gm <= 3 ) {
1929 $gy_offset--;
1930 }
1931 $gm = ( $gm - 3 ) % 12;
1932 }
1933 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1934 # Minguo dates
1935 # Deduct 1911 years from the Gregorian calendar
1936 # Months and days are identical
1937 $gy_offset = $gy - 1911;
1938 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1939 # Nengō dates up to Meiji period
1940 # Deduct years from the Gregorian calendar
1941 # depending on the nengo periods
1942 # Months and days are identical
1943 if ( ( $gy < 1912 )
1944 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1945 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1946 ) {
1947 # Meiji period
1948 $gy_gannen = $gy - 1868 + 1;
1949 $gy_offset = $gy_gannen;
1950 if ( $gy_gannen == 1 ) {
1951 $gy_offset = '元';
1952 }
1953 $gy_offset = '明治' . $gy_offset;
1954 } elseif (
1955 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1956 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1957 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1958 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1959 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1960 ) {
1961 # Taishō period
1962 $gy_gannen = $gy - 1912 + 1;
1963 $gy_offset = $gy_gannen;
1964 if ( $gy_gannen == 1 ) {
1965 $gy_offset = '元';
1966 }
1967 $gy_offset = '大正' . $gy_offset;
1968 } elseif (
1969 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1970 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1971 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1972 ) {
1973 # Shōwa period
1974 $gy_gannen = $gy - 1926 + 1;
1975 $gy_offset = $gy_gannen;
1976 if ( $gy_gannen == 1 ) {
1977 $gy_offset = '元';
1978 }
1979 $gy_offset = '昭和' . $gy_offset;
1980 } else {
1981 # Heisei period
1982 $gy_gannen = $gy - 1989 + 1;
1983 $gy_offset = $gy_gannen;
1984 if ( $gy_gannen == 1 ) {
1985 $gy_offset = '元';
1986 }
1987 $gy_offset = '平成' . $gy_offset;
1988 }
1989 } else {
1990 $gy_offset = $gy;
1991 }
1992
1993 return [ $gy_offset, $gm, $gd ];
1994 }
1995
1996 /**
1997 * Gets directionality of the first strongly directional codepoint, for embedBidi()
1998 *
1999 * This is the rule the BIDI algorithm uses to determine the directionality of
2000 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
2001 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2002 *
2003 * TODO: Does not handle BIDI control characters inside the text.
2004 * TODO: Does not handle unallocated characters.
2005 *
2006 * @param string $text Text to test
2007 * @return null|string Directionality ('ltr' or 'rtl') or null
2008 */
2009 private static function strongDirFromContent( $text = '' ) {
2010 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
2011 return null;
2012 }
2013 if ( $matches[1] === '' ) {
2014 return 'rtl';
2015 }
2016 return 'ltr';
2017 }
2018
2019 /**
2020 * Roman number formatting up to 10000
2021 *
2022 * @param int $num
2023 *
2024 * @return string
2025 */
2026 static function romanNumeral( $num ) {
2027 static $table = [
2028 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
2029 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
2030 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
2031 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2032 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2033 ];
2034
2035 $num = intval( $num );
2036 if ( $num > 10000 || $num <= 0 ) {
2037 return $num;
2038 }
2039
2040 $s = '';
2041 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2042 if ( $num >= $pow10 ) {
2043 $s .= $table[$i][(int)floor( $num / $pow10 )];
2044 }
2045 $num = $num % $pow10;
2046 }
2047 return $s;
2048 }
2049
2050 /**
2051 * Hebrew Gematria number formatting up to 9999
2052 *
2053 * @param int $num
2054 *
2055 * @return string
2056 */
2057 static function hebrewNumeral( $num ) {
2058 static $table = [
2059 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2060 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2061 [ '',
2062 [ 'ק' ],
2063 [ 'ר' ],
2064 [ 'ש' ],
2065 [ 'ת' ],
2066 [ 'ת', 'ק' ],
2067 [ 'ת', 'ר' ],
2068 [ 'ת', 'ש' ],
2069 [ 'ת', 'ת' ],
2070 [ 'ת', 'ת', 'ק' ],
2071 [ 'ת', 'ת', 'ר' ],
2072 ],
2073 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2074 ];
2075
2076 $num = intval( $num );
2077 if ( $num > 9999 || $num <= 0 ) {
2078 return $num;
2079 }
2080
2081 // Round thousands have special notations
2082 if ( $num === 1000 ) {
2083 return "א' אלף";
2084 } elseif ( $num % 1000 === 0 ) {
2085 return $table[0][$num / 1000] . "' אלפים";
2086 }
2087
2088 $letters = [];
2089
2090 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2091 if ( $num >= $pow10 ) {
2092 if ( $num === 15 || $num === 16 ) {
2093 $letters[] = $table[0][9];
2094 $letters[] = $table[0][$num - 9];
2095 $num = 0;
2096 } else {
2097 $letters = array_merge(
2098 $letters,
2099 (array)$table[$i][intval( $num / $pow10 )]
2100 );
2101
2102 if ( $pow10 === 1000 ) {
2103 $letters[] = "'";
2104 }
2105 }
2106 }
2107
2108 $num = $num % $pow10;
2109 }
2110
2111 $preTransformLength = count( $letters );
2112 if ( $preTransformLength === 1 ) {
2113 // Add geresh (single quote) to one-letter numbers
2114 $letters[] = "'";
2115 } else {
2116 $lastIndex = $preTransformLength - 1;
2117 $letters[$lastIndex] = str_replace(
2118 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2119 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2120 $letters[$lastIndex]
2121 );
2122
2123 // Add gershayim (double quote) to multiple-letter numbers,
2124 // but exclude numbers with only one letter after the thousands
2125 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2126 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2127 $letters[] = "'";
2128 } else {
2129 array_splice( $letters, -1, 0, '"' );
2130 }
2131 }
2132
2133 return implode( $letters );
2134 }
2135
2136 /**
2137 * Used by date() and time() to adjust the time output.
2138 *
2139 * @param string $ts The time in date('YmdHis') format
2140 * @param mixed $tz Adjust the time by this amount (default false, mean we
2141 * get user timecorrection setting)
2142 * @return int
2143 */
2144 public function userAdjust( $ts, $tz = false ) {
2145 global $wgUser, $wgLocalTZoffset;
2146
2147 if ( $tz === false ) {
2148 $tz = $wgUser->getOption( 'timecorrection' );
2149 }
2150
2151 $data = explode( '|', $tz, 3 );
2152
2153 if ( $data[0] == 'ZoneInfo' ) {
2154 try {
2155 $userTZ = new DateTimeZone( $data[2] );
2156 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2157 $date->setTimezone( $userTZ );
2158 return $date->format( 'YmdHis' );
2159 } catch ( Exception $e ) {
2160 // Unrecognized timezone, default to 'Offset' with the stored offset.
2161 $data[0] = 'Offset';
2162 }
2163 }
2164
2165 if ( $data[0] == 'System' || $tz == '' ) {
2166 # Global offset in minutes.
2167 $minDiff = $wgLocalTZoffset;
2168 } elseif ( $data[0] == 'Offset' ) {
2169 $minDiff = intval( $data[1] );
2170 } else {
2171 $data = explode( ':', $tz );
2172 if ( count( $data ) == 2 ) {
2173 $data[0] = intval( $data[0] );
2174 $data[1] = intval( $data[1] );
2175 $minDiff = abs( $data[0] ) * 60 + $data[1];
2176 if ( $data[0] < 0 ) {
2177 $minDiff = -$minDiff;
2178 }
2179 } else {
2180 $minDiff = intval( $data[0] ) * 60;
2181 }
2182 }
2183
2184 # No difference ? Return time unchanged
2185 if ( $minDiff == 0 ) {
2186 return $ts;
2187 }
2188
2189 Wikimedia\suppressWarnings(); // E_STRICT system time bitching
2190 # Generate an adjusted date; take advantage of the fact that mktime
2191 # will normalize out-of-range values so we don't have to split $minDiff
2192 # into hours and minutes.
2193 $t = mktime( (
2194 (int)substr( $ts, 8, 2 ) ), # Hours
2195 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2196 (int)substr( $ts, 12, 2 ), # Seconds
2197 (int)substr( $ts, 4, 2 ), # Month
2198 (int)substr( $ts, 6, 2 ), # Day
2199 (int)substr( $ts, 0, 4 ) ); # Year
2200
2201 $date = date( 'YmdHis', $t );
2202 Wikimedia\restoreWarnings();
2203
2204 return $date;
2205 }
2206
2207 /**
2208 * This is meant to be used by time(), date(), and timeanddate() to get
2209 * the date preference they're supposed to use, it should be used in
2210 * all children.
2211 *
2212 * function timeanddate([...], $format = true) {
2213 * $datePreference = $this->dateFormat($format);
2214 * [...]
2215 * }
2216 *
2217 * @param int|string|bool $usePrefs If true, the user's preference is used
2218 * if false, the site/language default is used
2219 * if int/string, assumed to be a format.
2220 * @return string
2221 */
2222 function dateFormat( $usePrefs = true ) {
2223 global $wgUser;
2224
2225 if ( is_bool( $usePrefs ) ) {
2226 if ( $usePrefs ) {
2227 $datePreference = $wgUser->getDatePreference();
2228 } else {
2229 $datePreference = (string)User::getDefaultOption( 'date' );
2230 }
2231 } else {
2232 $datePreference = (string)$usePrefs;
2233 }
2234
2235 // return int
2236 if ( $datePreference == '' ) {
2237 return 'default';
2238 }
2239
2240 return $datePreference;
2241 }
2242
2243 /**
2244 * Get a format string for a given type and preference
2245 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2246 * @param string $pref The format name as it appears in Messages*.php under
2247 * $datePreferences.
2248 *
2249 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2250 *
2251 * @return string
2252 */
2253 function getDateFormatString( $type, $pref ) {
2254 $wasDefault = false;
2255 if ( $pref == 'default' ) {
2256 $wasDefault = true;
2257 $pref = $this->getDefaultDateFormat();
2258 }
2259
2260 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2261 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2262
2263 if ( $type === 'pretty' && $df === null ) {
2264 $df = $this->getDateFormatString( 'date', $pref );
2265 }
2266
2267 if ( !$wasDefault && $df === null ) {
2268 $pref = $this->getDefaultDateFormat();
2269 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2270 }
2271
2272 $this->dateFormatStrings[$type][$pref] = $df;
2273 }
2274 return $this->dateFormatStrings[$type][$pref];
2275 }
2276
2277 /**
2278 * @param string $ts The time format which needs to be turned into a
2279 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2280 * @param bool $adj Whether to adjust the time output according to the
2281 * user configured offset ($timecorrection)
2282 * @param mixed $format True to use user's date format preference
2283 * @param string|bool $timecorrection The time offset as returned by
2284 * validateTimeZone() in Special:Preferences
2285 * @return string
2286 */
2287 public function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2288 $ts = wfTimestamp( TS_MW, $ts );
2289 if ( $adj ) {
2290 $ts = $this->userAdjust( $ts, $timecorrection );
2291 }
2292 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2293 return $this->sprintfDate( $df, $ts );
2294 }
2295
2296 /**
2297 * @param string $ts The time format which needs to be turned into a
2298 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2299 * @param bool $adj Whether to adjust the time output according to the
2300 * user configured offset ($timecorrection)
2301 * @param mixed $format True to use user's date format preference
2302 * @param string|bool $timecorrection The time offset as returned by
2303 * validateTimeZone() in Special:Preferences
2304 * @return string
2305 */
2306 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2307 $ts = wfTimestamp( TS_MW, $ts );
2308 if ( $adj ) {
2309 $ts = $this->userAdjust( $ts, $timecorrection );
2310 }
2311 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2312 return $this->sprintfDate( $df, $ts );
2313 }
2314
2315 /**
2316 * @param string $ts The time format which needs to be turned into a
2317 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2318 * @param bool $adj Whether to adjust the time output according to the
2319 * user configured offset ($timecorrection)
2320 * @param mixed $format What format to return, if it's false output the
2321 * default one (default true)
2322 * @param string|bool $timecorrection The time offset as returned by
2323 * validateTimeZone() in Special:Preferences
2324 * @return string
2325 */
2326 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2327 $ts = wfTimestamp( TS_MW, $ts );
2328 if ( $adj ) {
2329 $ts = $this->userAdjust( $ts, $timecorrection );
2330 }
2331 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2332 return $this->sprintfDate( $df, $ts );
2333 }
2334
2335 /**
2336 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2337 *
2338 * @since 1.20
2339 *
2340 * @param int $seconds The amount of seconds.
2341 * @param array $chosenIntervals The intervals to enable.
2342 *
2343 * @return string
2344 */
2345 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2346 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2347
2348 $segments = [];
2349
2350 foreach ( $intervals as $intervalName => $intervalValue ) {
2351 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2352 // duration-years, duration-decades, duration-centuries, duration-millennia
2353 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2354 $segments[] = $message->inLanguage( $this )->escaped();
2355 }
2356
2357 return $this->listToText( $segments );
2358 }
2359
2360 /**
2361 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2362 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2363 *
2364 * @since 1.20
2365 *
2366 * @param int $seconds The amount of seconds.
2367 * @param array $chosenIntervals The intervals to enable.
2368 *
2369 * @return array
2370 */
2371 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2372 if ( empty( $chosenIntervals ) ) {
2373 $chosenIntervals = [
2374 'millennia',
2375 'centuries',
2376 'decades',
2377 'years',
2378 'days',
2379 'hours',
2380 'minutes',
2381 'seconds'
2382 ];
2383 }
2384
2385 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2386 $sortedNames = array_keys( $intervals );
2387 $smallestInterval = array_pop( $sortedNames );
2388
2389 $segments = [];
2390
2391 foreach ( $intervals as $name => $length ) {
2392 $value = floor( $seconds / $length );
2393
2394 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2395 $seconds -= $value * $length;
2396 $segments[$name] = $value;
2397 }
2398 }
2399
2400 return $segments;
2401 }
2402
2403 /**
2404 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2405 *
2406 * @param string $type Can be 'date', 'time' or 'both'
2407 * @param string $ts The time format which needs to be turned into a
2408 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2409 * @param User $user User object used to get preferences for timezone and format
2410 * @param array $options Array, can contain the following keys:
2411 * - 'timecorrection': time correction, can have the following values:
2412 * - true: use user's preference
2413 * - false: don't use time correction
2414 * - int: value of time correction in minutes
2415 * - 'format': format to use, can have the following values:
2416 * - true: use user's preference
2417 * - false: use default preference
2418 * - string: format to use
2419 * @since 1.19
2420 * @return string
2421 */
2422 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2423 $ts = wfTimestamp( TS_MW, $ts );
2424 $options += [ 'timecorrection' => true, 'format' => true ];
2425 if ( $options['timecorrection'] !== false ) {
2426 if ( $options['timecorrection'] === true ) {
2427 $offset = $user->getOption( 'timecorrection' );
2428 } else {
2429 $offset = $options['timecorrection'];
2430 }
2431 $ts = $this->userAdjust( $ts, $offset );
2432 }
2433 if ( $options['format'] === true ) {
2434 $format = $user->getDatePreference();
2435 } else {
2436 $format = $options['format'];
2437 }
2438 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2439 return $this->sprintfDate( $df, $ts );
2440 }
2441
2442 /**
2443 * Get the formatted date for the given timestamp and formatted for
2444 * the given user.
2445 *
2446 * @param mixed $ts Mixed: the time format which needs to be turned into a
2447 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2448 * @param User $user User object used to get preferences for timezone and format
2449 * @param array $options Array, can contain the following keys:
2450 * - 'timecorrection': time correction, can have the following values:
2451 * - true: use user's preference
2452 * - false: don't use time correction
2453 * - int: value of time correction in minutes
2454 * - 'format': format to use, can have the following values:
2455 * - true: use user's preference
2456 * - false: use default preference
2457 * - string: format to use
2458 * @since 1.19
2459 * @return string
2460 */
2461 public function userDate( $ts, User $user, array $options = [] ) {
2462 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2463 }
2464
2465 /**
2466 * Get the formatted time for the given timestamp and formatted for
2467 * the given user.
2468 *
2469 * @param mixed $ts The time format which needs to be turned into a
2470 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2471 * @param User $user User object used to get preferences for timezone and format
2472 * @param array $options Array, can contain the following keys:
2473 * - 'timecorrection': time correction, can have the following values:
2474 * - true: use user's preference
2475 * - false: don't use time correction
2476 * - int: value of time correction in minutes
2477 * - 'format': format to use, can have the following values:
2478 * - true: use user's preference
2479 * - false: use default preference
2480 * - string: format to use
2481 * @since 1.19
2482 * @return string
2483 */
2484 public function userTime( $ts, User $user, array $options = [] ) {
2485 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2486 }
2487
2488 /**
2489 * Get the formatted date and time for the given timestamp and formatted for
2490 * the given user.
2491 *
2492 * @param mixed $ts The time format which needs to be turned into a
2493 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2494 * @param User $user User object used to get preferences for timezone and format
2495 * @param array $options Array, can contain the following keys:
2496 * - 'timecorrection': time correction, can have the following values:
2497 * - true: use user's preference
2498 * - false: don't use time correction
2499 * - int: value of time correction in minutes
2500 * - 'format': format to use, can have the following values:
2501 * - true: use user's preference
2502 * - false: use default preference
2503 * - string: format to use
2504 * @since 1.19
2505 * @return string
2506 */
2507 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2508 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2509 }
2510
2511 /**
2512 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2513 *
2514 * Determine the difference between the timestamp and the current time, and
2515 * generate a readable timestamp by returning "<N> <units> ago", where the
2516 * largest possible unit is used.
2517 *
2518 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2519 *
2520 * @param MWTimestamp $time
2521 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2522 * @param User|null $user User the timestamp is being generated for
2523 * (or null to use main context's user)
2524 * @return string Formatted timestamp
2525 */
2526 public function getHumanTimestamp(
2527 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2528 ) {
2529 if ( $relativeTo === null ) {
2530 $relativeTo = new MWTimestamp();
2531 }
2532 if ( $user === null ) {
2533 $user = RequestContext::getMain()->getUser();
2534 }
2535
2536 // Adjust for the user's timezone.
2537 $offsetThis = $time->offsetForUser( $user );
2538 $offsetRel = $relativeTo->offsetForUser( $user );
2539
2540 $ts = '';
2541 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2542 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2543 }
2544
2545 // Reset the timezone on the objects.
2546 $time->timestamp->sub( $offsetThis );
2547 $relativeTo->timestamp->sub( $offsetRel );
2548
2549 return $ts;
2550 }
2551
2552 /**
2553 * Convert an MWTimestamp into a pretty human-readable timestamp using
2554 * the given user preferences and relative base time.
2555 *
2556 * @see Language::getHumanTimestamp
2557 * @param MWTimestamp $ts Timestamp to prettify
2558 * @param MWTimestamp $relativeTo Base timestamp
2559 * @param User $user User preferences to use
2560 * @return string Human timestamp
2561 * @since 1.26
2562 */
2563 private function getHumanTimestampInternal(
2564 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2565 ) {
2566 $diff = $ts->diff( $relativeTo );
2567 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2568 (int)$relativeTo->timestamp->format( 'w' ) );
2569 $days = $diff->days ?: (int)$diffDay;
2570 if ( $diff->invert || $days > 5
2571 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2572 ) {
2573 // Timestamps are in different years: use full timestamp
2574 // Also do full timestamp for future dates
2575 /**
2576 * @todo FIXME: Add better handling of future timestamps.
2577 */
2578 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2579 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2580 } elseif ( $days > 5 ) {
2581 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2582 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2583 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2584 } elseif ( $days > 1 ) {
2585 // Timestamp within the past week: show the day of the week and time
2586 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2587 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2588 // Messages:
2589 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2590 $ts = wfMessage( "$weekday-at" )
2591 ->inLanguage( $this )
2592 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2593 ->text();
2594 } elseif ( $days == 1 ) {
2595 // Timestamp was yesterday: say 'yesterday' and the time.
2596 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2597 $ts = wfMessage( 'yesterday-at' )
2598 ->inLanguage( $this )
2599 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2600 ->text();
2601 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2602 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2603 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2604 $ts = wfMessage( 'today-at' )
2605 ->inLanguage( $this )
2606 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2607 ->text();
2608
2609 // From here on in, the timestamp was soon enough ago so that we can simply say
2610 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2611 } elseif ( $diff->h == 1 ) {
2612 // Less than 90 minutes, but more than an hour ago.
2613 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2614 } elseif ( $diff->i >= 1 ) {
2615 // A few minutes ago.
2616 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2617 } elseif ( $diff->s >= 30 ) {
2618 // Less than a minute, but more than 30 sec ago.
2619 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2620 } else {
2621 // Less than 30 seconds ago.
2622 $ts = wfMessage( 'just-now' )->text();
2623 }
2624
2625 return $ts;
2626 }
2627
2628 /**
2629 * @param string $key
2630 * @return string|null
2631 */
2632 public function getMessage( $key ) {
2633 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2634 }
2635
2636 /**
2637 * @return array
2638 */
2639 function getAllMessages() {
2640 return self::$dataCache->getItem( $this->mCode, 'messages' );
2641 }
2642
2643 /**
2644 * @param string $in
2645 * @param string $out
2646 * @param string $string
2647 * @return string
2648 */
2649 public function iconv( $in, $out, $string ) {
2650 # Even with //IGNORE iconv can whine about illegal characters in
2651 # *input* string. We just ignore those too.
2652 # REF: https://bugs.php.net/bug.php?id=37166
2653 # REF: https://phabricator.wikimedia.org/T18885
2654 Wikimedia\suppressWarnings();
2655 $text = iconv( $in, $out . '//IGNORE', $string );
2656 Wikimedia\restoreWarnings();
2657 return $text;
2658 }
2659
2660 // callback functions for ucwords(), ucwordbreaks()
2661
2662 /**
2663 * @param array $matches
2664 * @return mixed|string
2665 */
2666 function ucwordbreaksCallbackAscii( $matches ) {
2667 return $this->ucfirst( $matches[1] );
2668 }
2669
2670 /**
2671 * @param array $matches
2672 * @return string
2673 */
2674 function ucwordbreaksCallbackMB( $matches ) {
2675 return mb_strtoupper( $matches[0] );
2676 }
2677
2678 /**
2679 * @param array $matches
2680 * @return string
2681 */
2682 function ucwordsCallbackMB( $matches ) {
2683 return mb_strtoupper( $matches[0] );
2684 }
2685
2686 /**
2687 * Make a string's first character uppercase
2688 *
2689 * @param string $str
2690 *
2691 * @return string
2692 */
2693 public function ucfirst( $str ) {
2694 $o = ord( $str );
2695 if ( $o < 96 ) { // if already uppercase...
2696 return $str;
2697 } elseif ( $o < 128 ) {
2698 return ucfirst( $str ); // use PHP's ucfirst()
2699 } else {
2700 // fall back to more complex logic in case of multibyte strings
2701 return $this->uc( $str, true );
2702 }
2703 }
2704
2705 /**
2706 * Convert a string to uppercase
2707 *
2708 * @param string $str
2709 * @param bool $first
2710 *
2711 * @return string
2712 */
2713 public function uc( $str, $first = false ) {
2714 if ( $first ) {
2715 if ( $this->isMultibyte( $str ) ) {
2716 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2717 } else {
2718 return ucfirst( $str );
2719 }
2720 } else {
2721 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2722 }
2723 }
2724
2725 /**
2726 * @param string $str
2727 * @return mixed|string
2728 */
2729 function lcfirst( $str ) {
2730 $o = ord( $str );
2731 if ( !$o ) {
2732 return strval( $str );
2733 } elseif ( $o >= 128 ) {
2734 return $this->lc( $str, true );
2735 } elseif ( $o > 96 ) {
2736 return $str;
2737 } else {
2738 $str[0] = strtolower( $str[0] );
2739 return $str;
2740 }
2741 }
2742
2743 /**
2744 * @param string $str
2745 * @param bool $first
2746 * @return mixed|string
2747 */
2748 function lc( $str, $first = false ) {
2749 if ( $first ) {
2750 if ( $this->isMultibyte( $str ) ) {
2751 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2752 } else {
2753 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2754 }
2755 } else {
2756 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2757 }
2758 }
2759
2760 /**
2761 * @param string $str
2762 * @return bool
2763 */
2764 function isMultibyte( $str ) {
2765 return strlen( $str ) !== mb_strlen( $str );
2766 }
2767
2768 /**
2769 * @param string $str
2770 * @return mixed|string
2771 */
2772 function ucwords( $str ) {
2773 if ( $this->isMultibyte( $str ) ) {
2774 $str = $this->lc( $str );
2775
2776 // regexp to find first letter in each word (i.e. after each space)
2777 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2778
2779 // function to use to capitalize a single char
2780 return preg_replace_callback(
2781 $replaceRegexp,
2782 [ $this, 'ucwordsCallbackMB' ],
2783 $str
2784 );
2785 } else {
2786 return ucwords( strtolower( $str ) );
2787 }
2788 }
2789
2790 /**
2791 * capitalize words at word breaks
2792 *
2793 * @param string $str
2794 * @return mixed
2795 */
2796 function ucwordbreaks( $str ) {
2797 if ( $this->isMultibyte( $str ) ) {
2798 $str = $this->lc( $str );
2799
2800 // since \b doesn't work for UTF-8, we explicitely define word break chars
2801 $breaks = "[ \-\(\)\}\{\.,\?!]";
2802
2803 // find first letter after word break
2804 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2805 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2806
2807 return preg_replace_callback(
2808 $replaceRegexp,
2809 [ $this, 'ucwordbreaksCallbackMB' ],
2810 $str
2811 );
2812 } else {
2813 return preg_replace_callback(
2814 '/\b([\w\x80-\xff]+)\b/',
2815 [ $this, 'ucwordbreaksCallbackAscii' ],
2816 $str
2817 );
2818 }
2819 }
2820
2821 /**
2822 * Return a case-folded representation of $s
2823 *
2824 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2825 * and $s2 are the same except for the case of their characters. It is not
2826 * necessary for the value returned to make sense when displayed.
2827 *
2828 * Do *not* perform any other normalisation in this function. If a caller
2829 * uses this function when it should be using a more general normalisation
2830 * function, then fix the caller.
2831 *
2832 * @param string $s
2833 *
2834 * @return string
2835 */
2836 function caseFold( $s ) {
2837 return $this->uc( $s );
2838 }
2839
2840 /**
2841 * TODO: $s is not always a string per T218883
2842 * @param string $s
2843 * @return string
2844 */
2845 function checkTitleEncoding( $s ) {
2846 if ( is_array( $s ) ) {
2847 throw new MWException( 'Given array to checkTitleEncoding.' );
2848 }
2849 if ( StringUtils::isUtf8( $s ) ) {
2850 return $s;
2851 }
2852
2853 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2854 }
2855
2856 /**
2857 * @return string
2858 */
2859 function fallback8bitEncoding() {
2860 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2861 }
2862
2863 /**
2864 * Most writing systems use whitespace to break up words.
2865 * Some languages such as Chinese don't conventionally do this,
2866 * which requires special handling when breaking up words for
2867 * searching etc.
2868 *
2869 * @return bool
2870 */
2871 function hasWordBreaks() {
2872 return true;
2873 }
2874
2875 /**
2876 * Some languages such as Chinese require word segmentation,
2877 * Specify such segmentation when overridden in derived class.
2878 *
2879 * @param string $string
2880 * @return string
2881 */
2882 function segmentByWord( $string ) {
2883 return $string;
2884 }
2885
2886 /**
2887 * Some languages have special punctuation need to be normalized.
2888 * Make such changes here.
2889 *
2890 * @param string $string
2891 * @return string
2892 */
2893 function normalizeForSearch( $string ) {
2894 return self::convertDoubleWidth( $string );
2895 }
2896
2897 /**
2898 * convert double-width roman characters to single-width.
2899 * range: ff00-ff5f ~= 0020-007f
2900 *
2901 * @param string $string
2902 *
2903 * @return string
2904 */
2905 protected static function convertDoubleWidth( $string ) {
2906 static $full = null;
2907 static $half = null;
2908
2909 if ( $full === null ) {
2910 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2911 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2912 $full = str_split( $fullWidth, 3 );
2913 $half = str_split( $halfWidth );
2914 }
2915
2916 $string = str_replace( $full, $half, $string );
2917 return $string;
2918 }
2919
2920 /**
2921 * @param string $string
2922 * @param string $pattern
2923 * @return string
2924 */
2925 protected static function insertSpace( $string, $pattern ) {
2926 $string = preg_replace( $pattern, " $1 ", $string );
2927 $string = preg_replace( '/ +/', ' ', $string );
2928 return $string;
2929 }
2930
2931 /**
2932 * @param array $termsArray
2933 * @return array
2934 */
2935 function convertForSearchResult( $termsArray ) {
2936 # some languages, e.g. Chinese, need to do a conversion
2937 # in order for search results to be displayed correctly
2938 return $termsArray;
2939 }
2940
2941 /**
2942 * Get the first character of a string.
2943 *
2944 * @param string $s
2945 * @return string
2946 */
2947 function firstChar( $s ) {
2948 $matches = [];
2949 preg_match(
2950 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2951 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2952 $s,
2953 $matches
2954 );
2955
2956 if ( isset( $matches[1] ) ) {
2957 if ( strlen( $matches[1] ) != 3 ) {
2958 return $matches[1];
2959 }
2960
2961 // Break down Hangul syllables to grab the first jamo
2962 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2963 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2964 return $matches[1];
2965 } elseif ( $code < 0xb098 ) {
2966 return "\u{3131}";
2967 } elseif ( $code < 0xb2e4 ) {
2968 return "\u{3134}";
2969 } elseif ( $code < 0xb77c ) {
2970 return "\u{3137}";
2971 } elseif ( $code < 0xb9c8 ) {
2972 return "\u{3139}";
2973 } elseif ( $code < 0xbc14 ) {
2974 return "\u{3141}";
2975 } elseif ( $code < 0xc0ac ) {
2976 return "\u{3142}";
2977 } elseif ( $code < 0xc544 ) {
2978 return "\u{3145}";
2979 } elseif ( $code < 0xc790 ) {
2980 return "\u{3147}";
2981 } elseif ( $code < 0xcc28 ) {
2982 return "\u{3148}";
2983 } elseif ( $code < 0xce74 ) {
2984 return "\u{314A}";
2985 } elseif ( $code < 0xd0c0 ) {
2986 return "\u{314B}";
2987 } elseif ( $code < 0xd30c ) {
2988 return "\u{314C}";
2989 } elseif ( $code < 0xd558 ) {
2990 return "\u{314D}";
2991 } else {
2992 return "\u{314E}";
2993 }
2994 } else {
2995 return '';
2996 }
2997 }
2998
2999 /**
3000 * @deprecated No-op since 1.28
3001 */
3002 function initEncoding() {
3003 wfDeprecated( __METHOD__, '1.28' );
3004 // No-op.
3005 }
3006
3007 /**
3008 * @param string $s
3009 * @return string
3010 * @deprecated No-op since 1.28
3011 */
3012 function recodeForEdit( $s ) {
3013 wfDeprecated( __METHOD__, '1.28' );
3014 return $s;
3015 }
3016
3017 /**
3018 * @param string $s
3019 * @return string
3020 * @deprecated No-op since 1.28
3021 */
3022 function recodeInput( $s ) {
3023 wfDeprecated( __METHOD__, '1.28' );
3024 return $s;
3025 }
3026
3027 /**
3028 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3029 * also cleans up certain backwards-compatible sequences, converting them
3030 * to the modern Unicode equivalent.
3031 *
3032 * This is language-specific for performance reasons only.
3033 *
3034 * @param string $s
3035 *
3036 * @return string
3037 */
3038 public function normalize( $s ) {
3039 global $wgAllUnicodeFixes;
3040 $s = UtfNormal\Validator::cleanUp( $s );
3041 if ( $wgAllUnicodeFixes ) {
3042 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
3043 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3044 }
3045
3046 return $s;
3047 }
3048
3049 /**
3050 * Transform a string using serialized data stored in the given file (which
3051 * must be in the serialized subdirectory of $IP). The file contains pairs
3052 * mapping source characters to destination characters.
3053 *
3054 * The data is cached in process memory. This will go faster if you have the
3055 * FastStringSearch extension.
3056 *
3057 * @param string $file
3058 * @param string $string
3059 *
3060 * @throws MWException
3061 * @return string
3062 */
3063 protected function transformUsingPairFile( $file, $string ) {
3064 if ( !isset( $this->transformData[$file] ) ) {
3065 global $IP;
3066 $data = require "$IP/languages/data/{$file}";
3067 $this->transformData[$file] = new ReplacementArray( $data );
3068 }
3069 return $this->transformData[$file]->replace( $string );
3070 }
3071
3072 /**
3073 * For right-to-left language support
3074 *
3075 * @return bool
3076 */
3077 function isRTL() {
3078 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3079 }
3080
3081 /**
3082 * Return the correct HTML 'dir' attribute value for this language.
3083 * @return string
3084 */
3085 function getDir() {
3086 return $this->isRTL() ? 'rtl' : 'ltr';
3087 }
3088
3089 /**
3090 * Return 'left' or 'right' as appropriate alignment for line-start
3091 * for this language's text direction.
3092 *
3093 * Should be equivalent to CSS3 'start' text-align value....
3094 *
3095 * @return string
3096 */
3097 function alignStart() {
3098 return $this->isRTL() ? 'right' : 'left';
3099 }
3100
3101 /**
3102 * Return 'right' or 'left' as appropriate alignment for line-end
3103 * for this language's text direction.
3104 *
3105 * Should be equivalent to CSS3 'end' text-align value....
3106 *
3107 * @return string
3108 */
3109 function alignEnd() {
3110 return $this->isRTL() ? 'left' : 'right';
3111 }
3112
3113 /**
3114 * A hidden direction mark (LRM or RLM), depending on the language direction.
3115 * Unlike getDirMark(), this function returns the character as an HTML entity.
3116 * This function should be used when the output is guaranteed to be HTML,
3117 * because it makes the output HTML source code more readable. When
3118 * the output is plain text or can be escaped, getDirMark() should be used.
3119 *
3120 * @param bool $opposite Get the direction mark opposite to your language
3121 * @return string
3122 * @since 1.20
3123 */
3124 function getDirMarkEntity( $opposite = false ) {
3125 if ( $opposite ) {
3126 return $this->isRTL() ? '&lrm;' : '&rlm;';
3127 }
3128 return $this->isRTL() ? '&rlm;' : '&lrm;';
3129 }
3130
3131 /**
3132 * A hidden direction mark (LRM or RLM), depending on the language direction.
3133 * This function produces them as invisible Unicode characters and
3134 * the output may be hard to read and debug, so it should only be used
3135 * when the output is plain text or can be escaped. When the output is
3136 * HTML, use getDirMarkEntity() instead.
3137 *
3138 * @param bool $opposite Get the direction mark opposite to your language
3139 * @return string
3140 */
3141 function getDirMark( $opposite = false ) {
3142 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3143 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3144 if ( $opposite ) {
3145 return $this->isRTL() ? $lrm : $rlm;
3146 }
3147 return $this->isRTL() ? $rlm : $lrm;
3148 }
3149
3150 /**
3151 * @return array
3152 */
3153 function capitalizeAllNouns() {
3154 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3155 }
3156
3157 /**
3158 * An arrow, depending on the language direction.
3159 *
3160 * @param string $direction The direction of the arrow: forwards (default),
3161 * backwards, left, right, up, down.
3162 * @return string
3163 */
3164 function getArrow( $direction = 'forwards' ) {
3165 switch ( $direction ) {
3166 case 'forwards':
3167 return $this->isRTL() ? '←' : '→';
3168 case 'backwards':
3169 return $this->isRTL() ? '→' : '←';
3170 case 'left':
3171 return '←';
3172 case 'right':
3173 return '→';
3174 case 'up':
3175 return '↑';
3176 case 'down':
3177 return '↓';
3178 }
3179 }
3180
3181 /**
3182 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3183 *
3184 * @return bool
3185 */
3186 function linkPrefixExtension() {
3187 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3188 }
3189
3190 /**
3191 * Get all magic words from cache.
3192 * @return array
3193 */
3194 function getMagicWords() {
3195 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3196 }
3197
3198 /**
3199 * Fill a MagicWord object with data from here
3200 *
3201 * @param MagicWord $mw
3202 */
3203 function getMagic( $mw ) {
3204 $rawEntry = $this->mMagicExtensions[$mw->mId] ??
3205 self::$dataCache->getSubitem( $this->mCode, 'magicWords', $mw->mId );
3206
3207 if ( !is_array( $rawEntry ) ) {
3208 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3209 } else {
3210 $mw->mCaseSensitive = $rawEntry[0];
3211 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3212 }
3213 }
3214
3215 /**
3216 * Add magic words to the extension array
3217 *
3218 * @param array $newWords
3219 */
3220 function addMagicWordsByLang( $newWords ) {
3221 $fallbackChain = $this->getFallbackLanguages();
3222 $fallbackChain = array_reverse( $fallbackChain );
3223 foreach ( $fallbackChain as $code ) {
3224 if ( isset( $newWords[$code] ) ) {
3225 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3226 }
3227 }
3228 }
3229
3230 /**
3231 * Get special page names, as an associative array
3232 * canonical name => array of valid names, including aliases
3233 * @return array
3234 */
3235 function getSpecialPageAliases() {
3236 // Cache aliases because it may be slow to load them
3237 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3238 // Initialise array
3239 $this->mExtendedSpecialPageAliases =
3240 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3241 }
3242
3243 return $this->mExtendedSpecialPageAliases;
3244 }
3245
3246 /**
3247 * Italic is unsuitable for some languages
3248 *
3249 * @param string $text The text to be emphasized.
3250 * @return string
3251 */
3252 function emphasize( $text ) {
3253 return "<em>$text</em>";
3254 }
3255
3256 /**
3257 * Normally we output all numbers in plain en_US style, that is
3258 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3259 * point twohundredthirtyfive. However this is not suitable for all
3260 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3261 * Icelandic just want to use commas instead of dots, and dots instead
3262 * of commas like "293.291,235".
3263 *
3264 * An example of this function being called:
3265 * <code>
3266 * wfMessage( 'message' )->numParams( $num )->text()
3267 * </code>
3268 *
3269 * See $separatorTransformTable on MessageIs.php for
3270 * the , => . and . => , implementation.
3271 *
3272 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3273 * @param int|float $number The string to be formatted, should be an integer
3274 * or a floating point number.
3275 * @param bool $nocommafy Set to true for special numbers like dates
3276 * @return string
3277 */
3278 public function formatNum( $number, $nocommafy = false ) {
3279 global $wgTranslateNumerals;
3280 if ( !$nocommafy ) {
3281 $number = $this->commafy( $number );
3282 $s = $this->separatorTransformTable();
3283 if ( $s ) {
3284 $number = strtr( $number, $s );
3285 }
3286 }
3287
3288 if ( $wgTranslateNumerals ) {
3289 $s = $this->digitTransformTable();
3290 if ( $s ) {
3291 $number = strtr( $number, $s );
3292 }
3293 }
3294
3295 return (string)$number;
3296 }
3297
3298 /**
3299 * Front-end for non-commafied formatNum
3300 *
3301 * @param int|float $number The string to be formatted, should be an integer
3302 * or a floating point number.
3303 * @since 1.21
3304 * @return string
3305 */
3306 public function formatNumNoSeparators( $number ) {
3307 return $this->formatNum( $number, true );
3308 }
3309
3310 /**
3311 * @param string $number
3312 * @return string
3313 */
3314 public function parseFormattedNumber( $number ) {
3315 $s = $this->digitTransformTable();
3316 if ( $s ) {
3317 // eliminate empty array values such as ''. (T66347)
3318 $s = array_filter( $s );
3319 $number = strtr( $number, array_flip( $s ) );
3320 }
3321
3322 $s = $this->separatorTransformTable();
3323 if ( $s ) {
3324 // eliminate empty array values such as ''. (T66347)
3325 $s = array_filter( $s );
3326 $number = strtr( $number, array_flip( $s ) );
3327 }
3328
3329 $number = strtr( $number, [ ',' => '' ] );
3330 return $number;
3331 }
3332
3333 /**