dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "Remove redundant ignore_user_abort() call in ApiStashEdit"
[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 MediaWiki\MediaWikiServices;
31 use Wikimedia\Assert\Assert;
32
33 /**
34 * Internationalisation code
35 * @ingroup Language
36 */
37 class Language {
38 /**
39 * Return autonyms in fetchLanguageName(s).
40 * @since 1.32
41 */
42 const AS_AUTONYMS = null;
43
44 /**
45 * Return all known languages in fetchLanguageName(s).
46 * @since 1.32
47 */
48 const ALL = 'all';
49
50 /**
51 * Return in fetchLanguageName(s) only the languages for which we have at
52 * least some localisation.
53 * @since 1.32
54 */
55 const SUPPORTED = 'mwfile';
56
57 /**
58 * @var LanguageConverter
59 */
60 public $mConverter;
61
62 public $mVariants, $mCode, $mLoaded = false;
63 public $mMagicExtensions = [];
64 private $mHtmlCode = null, $mParentLanguage = false;
65
66 public $dateFormatStrings = [];
67 public $mExtendedSpecialPageAliases;
68
69 /** @var array|null */
70 protected $namespaceNames;
71 protected $mNamespaceIds, $namespaceAliases;
72
73 /**
74 * ReplacementArray object caches
75 */
76 public $transformData = [];
77
78 /**
79 * @var LocalisationCache
80 */
81 public static $dataCache;
82
83 public static $mLangObjCache = [];
84
85 /**
86 * Return a fallback chain for messages in getFallbacksFor
87 * @since 1.32
88 */
89 const MESSAGES_FALLBACKS = 0;
90
91 /**
92 * Return a strict fallback chain in getFallbacksFor
93 * @since 1.32
94 */
95 const STRICT_FALLBACKS = 1;
96
97 public static $mWeekdayMsgs = [
98 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
99 'friday', 'saturday'
100 ];
101
102 public static $mWeekdayAbbrevMsgs = [
103 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
104 ];
105
106 public static $mMonthMsgs = [
107 'january', 'february', 'march', 'april', 'may_long', 'june',
108 'july', 'august', 'september', 'october', 'november',
109 'december'
110 ];
111 public static $mMonthGenMsgs = [
112 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
113 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
114 'december-gen'
115 ];
116 public static $mMonthAbbrevMsgs = [
117 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
118 'sep', 'oct', 'nov', 'dec'
119 ];
120
121 public static $mIranianCalendarMonthMsgs = [
122 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
123 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
124 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
125 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
126 ];
127
128 public static $mHebrewCalendarMonthMsgs = [
129 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
130 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
131 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
132 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
133 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
134 ];
135
136 public static $mHebrewCalendarMonthGenMsgs = [
137 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
138 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
139 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
140 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
141 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
142 ];
143
144 public static $mHijriCalendarMonthMsgs = [
145 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
146 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
147 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
148 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
149 ];
150
151 /**
152 * @since 1.20
153 * @var array
154 */
155 public static $durationIntervals = [
156 'millennia' => 31556952000,
157 'centuries' => 3155695200,
158 'decades' => 315569520,
159 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
160 'weeks' => 604800,
161 'days' => 86400,
162 'hours' => 3600,
163 'minutes' => 60,
164 'seconds' => 1,
165 ];
166
167 /**
168 * Cache for language fallbacks.
169 * @see Language::getFallbacksIncludingSiteLanguage
170 * @since 1.21
171 * @var array
172 */
173 private static $fallbackLanguageCache = [];
174
175 /**
176 * Cache for grammar rules data
177 * @var MapCacheLRU|null
178 */
179 private static $grammarTransformations;
180
181 /**
182 * Cache for language names
183 * @var HashBagOStuff|null
184 */
185 private static $languageNameCache;
186
187 /**
188 * Unicode directional formatting characters, for embedBidi()
189 */
190 private static $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
191 private static $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
192 private static $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
193
194 /**
195 * Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
196 * - in group 1 if it is LTR
197 * - in group 2 if it is RTL
198 * Does not match if there is no strong directionality codepoint.
199 *
200 * The form is '/(?:([strong ltr codepoint])|([strong rtl codepoint]))/u' .
201 *
202 * Generated by UnicodeJS (see tools/strongDir) from the UCD; see
203 * https://phabricator.wikimedia.org/diffusion/GUJS/ .
204 */
205 // @codeCoverageIgnoreStart
206 // phpcs:ignore Generic.Files.LineLength
207 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';
208 // @codeCoverageIgnoreEnd
209
210 /**
211 * Get a cached or new language object for a given language code
212 * @param string $code
213 * @throws MWException
214 * @return Language
215 */
216 static function factory( $code ) {
217 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
218
219 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
220 $code = $wgDummyLanguageCodes[$code];
221 }
222
223 // get the language object to process
224 $langObj = self::$mLangObjCache[$code] ?? self::newFromCode( $code );
225
226 // merge the language object in to get it up front in the cache
227 self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
228 // get rid of the oldest ones in case we have an overflow
229 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
230
231 return $langObj;
232 }
233
234 /**
235 * Create a language object for a given language code
236 * @param string $code
237 * @param bool $fallback Whether we're going through language fallback chain
238 * @throws MWException
239 * @return Language
240 */
241 protected static function newFromCode( $code, $fallback = false ) {
242 if ( !self::isValidCode( $code ) ) {
243 throw new MWException( "Invalid language code \"$code\"" );
244 }
245
246 if ( !self::isValidBuiltInCode( $code ) ) {
247 // It's not possible to customise this code with class files, so
248 // just return a Language object. This is to support uselang= hacks.
249 $lang = new Language;
250 $lang->mCode = $code;
251 return $lang;
252 }
253
254 // Check if there is a language class for the code
255 $class = self::classFromCode( $code, $fallback );
256 // LanguageCode does not inherit Language
257 if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
258 $lang = new $class;
259 return $lang;
260 }
261
262 // Keep trying the fallback list until we find an existing class
263 $fallbacks = self::getFallbacksFor( $code );
264 foreach ( $fallbacks as $fallbackCode ) {
265 if ( !self::isValidBuiltInCode( $fallbackCode ) ) {
266 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
267 }
268
269 $class = self::classFromCode( $fallbackCode );
270 if ( class_exists( $class ) ) {
271 $lang = new $class;
272 $lang->mCode = $code;
273 return $lang;
274 }
275 }
276
277 throw new MWException( "Invalid fallback sequence for language '$code'" );
278 }
279
280 /**
281 * Intended for tests that may change configuration in a way that invalidates caches.
282 *
283 * @since 1.32
284 */
285 public static function clearCaches() {
286 if ( !defined( 'MW_PHPUNIT_TEST' ) ) {
287 throw new MWException( __METHOD__ . ' must not be used outside tests' );
288 }
289 self::$dataCache = null;
290 // Reinitialize $dataCache, since it's expected to always be available
291 self::getLocalisationCache();
292 self::$mLangObjCache = [];
293 self::$fallbackLanguageCache = [];
294 self::$grammarTransformations = null;
295 self::$languageNameCache = null;
296 }
297
298 /**
299 * Checks whether any localisation is available for that language tag
300 * in MediaWiki (MessagesXx.php exists).
301 *
302 * @param string $code Language tag (in lower case)
303 * @return bool Whether language is supported
304 * @since 1.21
305 */
306 public static function isSupportedLanguage( $code ) {
307 if ( !self::isValidBuiltInCode( $code ) ) {
308 return false;
309 }
310
311 if ( $code === 'qqq' ) {
312 return false;
313 }
314
315 return is_readable( self::getMessagesFileName( $code ) ) ||
316 is_readable( self::getJsonMessagesFileName( $code ) );
317 }
318
319 /**
320 * Returns true if a language code string is a well-formed language tag
321 * according to RFC 5646.
322 * This function only checks well-formedness; it doesn't check that
323 * language, script or variant codes actually exist in the repositories.
324 *
325 * Based on regexes by Mark Davis of the Unicode Consortium:
326 * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
327 *
328 * @param string $code
329 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
330 *
331 * @return bool
332 * @since 1.21
333 */
334 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
335 $alpha = '[a-z]';
336 $digit = '[0-9]';
337 $alphanum = '[a-z0-9]';
338 $x = 'x'; # private use singleton
339 $singleton = '[a-wy-z]'; # other singleton
340 $s = $lenient ? '[-_]' : '-';
341
342 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
343 $script = "$alpha{4}"; # ISO 15924
344 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
345 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
346 $extension = "$singleton(?:$s$alphanum{2,8})+";
347 $privateUse = "$x(?:$s$alphanum{1,8})+";
348
349 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
350 # Since these are limited, this is safe even later changes to the registry --
351 # the only oddity is that it might change the type of the tag, and thus
352 # the results from the capturing groups.
353 # https://www.iana.org/assignments/language-subtag-registry
354
355 $grandfathered = "en{$s}GB{$s}oed"
356 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
357 . "|no{$s}(?:bok|nyn)"
358 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
359 . "|zh{$s}min{$s}nan";
360
361 $variantList = "$variant(?:$s$variant)*";
362 $extensionList = "$extension(?:$s$extension)*";
363
364 $langtag = "(?:($language)"
365 . "(?:$s$script)?"
366 . "(?:$s$region)?"
367 . "(?:$s$variantList)?"
368 . "(?:$s$extensionList)?"
369 . "(?:$s$privateUse)?)";
370
371 # The final breakdown, with capturing groups for each of these components
372 # The variants, extensions, grandfathered, and private-use may have interior '-'
373
374 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
375
376 return (bool)preg_match( "/$root/", strtolower( $code ) );
377 }
378
379 /**
380 * Returns true if a language code string is of a valid form, whether or
381 * not it exists. This includes codes which are used solely for
382 * customisation via the MediaWiki namespace.
383 *
384 * @param string $code
385 *
386 * @return bool
387 */
388 public static function isValidCode( $code ) {
389 static $cache = [];
390 Assert::parameterType( 'string', $code, '$code' );
391 if ( !isset( $cache[$code] ) ) {
392 // People think language codes are html safe, so enforce it.
393 // Ideally we should only allow a-zA-Z0-9-
394 // but, .+ and other chars are often used for {{int:}} hacks
395 // see bugs T39564, T39587, T38938
396 $cache[$code] =
397 // Protect against path traversal
398 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
399 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
400 }
401 return $cache[$code];
402 }
403
404 /**
405 * Returns true if a language code is of a valid form for the purposes of
406 * internal customisation of MediaWiki, via Messages*.php or *.json.
407 *
408 * @param string $code
409 *
410 * @since 1.18
411 * @return bool
412 */
413 public static function isValidBuiltInCode( $code ) {
414 Assert::parameterType( 'string', $code, '$code' );
415
416 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
417 }
418
419 /**
420 * Returns true if a language code is an IETF tag known to MediaWiki.
421 *
422 * @param string $tag
423 *
424 * @since 1.21
425 * @return bool
426 */
427 public static function isKnownLanguageTag( $tag ) {
428 // Quick escape for invalid input to avoid exceptions down the line
429 // when code tries to process tags which are not valid at all.
430 if ( !self::isValidBuiltInCode( $tag ) ) {
431 return false;
432 }
433
434 if ( isset( MediaWiki\Languages\Data\Names::$names[$tag] )
435 || self::fetchLanguageName( $tag, $tag ) !== ''
436 ) {
437 return true;
438 }
439
440 return false;
441 }
442
443 /**
444 * Get the LocalisationCache instance
445 *
446 * @return LocalisationCache
447 */
448 public static function getLocalisationCache() {
449 if ( is_null( self::$dataCache ) ) {
450 global $wgLocalisationCacheConf;
451 $class = $wgLocalisationCacheConf['class'];
452 self::$dataCache = new $class( $wgLocalisationCacheConf );
453 }
454 return self::$dataCache;
455 }
456
457 function __construct() {
458 $this->mConverter = new FakeConverter( $this );
459 // Set the code to the name of the descendant
460 if ( static::class === 'Language' ) {
461 $this->mCode = 'en';
462 } else {
463 $this->mCode = str_replace( '_', '-', strtolower( substr( static::class, 8 ) ) );
464 }
465 self::getLocalisationCache();
466 }
467
468 /**
469 * Reduce memory usage
470 * @suppress PhanTypeSuspiciousNonTraversableForeach
471 */
472 function __destruct() {
473 foreach ( $this as $name => $value ) {
474 unset( $this->$name );
475 }
476 }
477
478 /**
479 * Hook which will be called if this is the content language.
480 * Descendants can use this to register hook functions or modify globals
481 */
482 function initContLang() {
483 }
484
485 /**
486 * @return array
487 * @since 1.19
488 */
489 public function getFallbackLanguages() {
490 return self::getFallbacksFor( $this->mCode );
491 }
492
493 /**
494 * Exports $wgBookstoreListEn
495 * @return array
496 */
497 public function getBookstoreList() {
498 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
499 }
500
501 /**
502 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
503 * available in localised form, it will be included in English.
504 *
505 * @return string[] List of localized namespace names, indexed by numeric namespace ID.
506 */
507 public function getNamespaces() {
508 if ( is_null( $this->namespaceNames ) ) {
509 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
510
511 $validNamespaces = MediaWikiServices::getInstance()->getNamespaceInfo()->
512 getCanonicalNamespaces();
513
514 $this->namespaceNames = $wgExtraNamespaces +
515 self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
516 $this->namespaceNames += $validNamespaces;
517
518 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
519 if ( $wgMetaNamespaceTalk ) {
520 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
521 } else {
522 $talk = $this->namespaceNames[NS_PROJECT_TALK];
523 $this->namespaceNames[NS_PROJECT_TALK] =
524 $this->fixVariableInNamespace( $talk );
525 }
526
527 # Sometimes a language will be localised but not actually exist on this wiki.
528 // @phan-suppress-next-line PhanTypeMismatchForeach
529 foreach ( $this->namespaceNames as $key => $text ) {
530 if ( !isset( $validNamespaces[$key] ) ) {
531 unset( $this->namespaceNames[$key] );
532 }
533 }
534
535 # The above mixing may leave namespaces out of canonical order.
536 # Re-order by namespace ID number...
537 // @phan-suppress-next-line PhanTypeMismatchArgumentInternal
538 ksort( $this->namespaceNames );
539
540 Hooks::run( 'LanguageGetNamespaces', [ &$this->namespaceNames ] );
541 }
542
543 return $this->namespaceNames;
544 }
545
546 /**
547 * Arbitrarily set all of the namespace names at once. Mainly used for testing
548 * @param array $namespaces Array of namespaces (id => name)
549 */
550 public function setNamespaces( array $namespaces ) {
551 $this->namespaceNames = $namespaces;
552 $this->mNamespaceIds = null;
553 }
554
555 /**
556 * Resets all of the namespace caches. Mainly used for testing
557 */
558 public function resetNamespaces() {
559 $this->namespaceNames = null;
560 $this->mNamespaceIds = null;
561 $this->namespaceAliases = null;
562 }
563
564 /**
565 * A convenience function that returns getNamespaces() with spaces instead of underscores
566 * in values. Useful for producing output to be displayed e.g. in `<select>` forms.
567 *
568 * @return array
569 */
570 public function getFormattedNamespaces() {
571 $ns = $this->getNamespaces();
572 foreach ( $ns as $k => $v ) {
573 $ns[$k] = strtr( $v, '_', ' ' );
574 }
575 return $ns;
576 }
577
578 /**
579 * Get a namespace value by key
580 *
581 * <code>
582 * $mw_ns = $lang->getNsText( NS_MEDIAWIKI );
583 * echo $mw_ns; // prints 'MediaWiki'
584 * </code>
585 *
586 * @param int $index The array key of the namespace to return
587 * @return string|bool String if the namespace value exists, otherwise false
588 */
589 public function getNsText( $index ) {
590 $ns = $this->getNamespaces();
591 return $ns[$index] ?? false;
592 }
593
594 /**
595 * A convenience function that returns the same thing as
596 * getNsText() except with '_' changed to ' ', useful for
597 * producing output.
598 *
599 * <code>
600 * $mw_ns = $lang->getFormattedNsText( NS_MEDIAWIKI_TALK );
601 * echo $mw_ns; // prints 'MediaWiki talk'
602 * </code>
603 *
604 * @param int $index The array key of the namespace to return
605 * @return string Namespace name without underscores (empty string if namespace does not exist)
606 */
607 public function getFormattedNsText( $index ) {
608 $ns = $this->getNsText( $index );
609 return strtr( $ns, '_', ' ' );
610 }
611
612 /**
613 * Returns gender-dependent namespace alias if available.
614 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
615 * @param int $index Namespace index
616 * @param string $gender Gender key (male, female... )
617 * @return string
618 * @since 1.18
619 */
620 public function getGenderNsText( $index, $gender ) {
621 global $wgExtraGenderNamespaces;
622
623 $ns = $wgExtraGenderNamespaces +
624 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
625
626 return $ns[$index][$gender] ?? $this->getNsText( $index );
627 }
628
629 /**
630 * Whether this language uses gender-dependent namespace aliases.
631 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
632 * @return bool
633 * @since 1.18
634 */
635 public function needsGenderDistinction() {
636 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
637 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
638 // $wgExtraGenderNamespaces overrides everything
639 return true;
640 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
641 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
642 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
643 return false;
644 } else {
645 // Check what is in i18n files
646 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
647 return count( $aliases ) > 0;
648 }
649 }
650
651 /**
652 * Get a namespace key by value, case insensitive.
653 * Only matches namespace names for the current language, not the
654 * canonical ones defined in Namespace.php.
655 *
656 * @param string $text
657 * @return int|bool An integer if $text is a valid value otherwise false
658 */
659 function getLocalNsIndex( $text ) {
660 $lctext = $this->lc( $text );
661 $ids = $this->getNamespaceIds();
662 return $ids[$lctext] ?? false;
663 }
664
665 /**
666 * @return array
667 */
668 public function getNamespaceAliases() {
669 if ( is_null( $this->namespaceAliases ) ) {
670 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
671 if ( !$aliases ) {
672 $aliases = [];
673 } else {
674 foreach ( $aliases as $name => $index ) {
675 if ( $index === NS_PROJECT_TALK ) {
676 unset( $aliases[$name] );
677 $name = $this->fixVariableInNamespace( $name );
678 $aliases[$name] = $index;
679 }
680 }
681 }
682
683 global $wgExtraGenderNamespaces;
684 $genders = $wgExtraGenderNamespaces +
685 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
686 foreach ( $genders as $index => $forms ) {
687 foreach ( $forms as $alias ) {
688 $aliases[$alias] = $index;
689 }
690 }
691
692 # Also add converted namespace names as aliases, to avoid confusion.
693 $convertedNames = [];
694 foreach ( $this->getVariants() as $variant ) {
695 if ( $variant === $this->mCode ) {
696 continue;
697 }
698 foreach ( $this->getNamespaces() as $ns => $_ ) {
699 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
700 }
701 }
702
703 $this->namespaceAliases = $aliases + $convertedNames;
704
705 # Filter out aliases to namespaces that don't exist, e.g. from extensions
706 # that aren't loaded here but are included in the l10n cache.
707 # (array_intersect preserves keys from its first argument)
708 $this->namespaceAliases = array_intersect(
709 $this->namespaceAliases,
710 array_keys( $this->getNamespaces() )
711 );
712 }
713
714 return $this->namespaceAliases;
715 }
716
717 /**
718 * @return array
719 */
720 public function getNamespaceIds() {
721 if ( is_null( $this->mNamespaceIds ) ) {
722 global $wgNamespaceAliases;
723 # Put namespace names and aliases into a hashtable.
724 # If this is too slow, then we should arrange it so that it is done
725 # before caching. The catch is that at pre-cache time, the above
726 # class-specific fixup hasn't been done.
727 $this->mNamespaceIds = [];
728 foreach ( $this->getNamespaces() as $index => $name ) {
729 $this->mNamespaceIds[$this->lc( $name )] = $index;
730 }
731 foreach ( $this->getNamespaceAliases() as $name => $index ) {
732 $this->mNamespaceIds[$this->lc( $name )] = $index;
733 }
734 if ( $wgNamespaceAliases ) {
735 foreach ( $wgNamespaceAliases as $name => $index ) {
736 $this->mNamespaceIds[$this->lc( $name )] = $index;
737 }
738 }
739 }
740 return $this->mNamespaceIds;
741 }
742
743 /**
744 * Get a namespace key by value, case insensitive. Canonical namespace
745 * names override custom ones defined for the current language.
746 *
747 * @param string $text
748 * @return int|bool An integer if $text is a valid value otherwise false
749 */
750 public function getNsIndex( $text ) {
751 $lctext = $this->lc( $text );
752 $ns = MediaWikiServices::getInstance()->getNamespaceInfo()->
753 getCanonicalIndex( $lctext );
754 if ( $ns !== null ) {
755 return $ns;
756 }
757 $ids = $this->getNamespaceIds();
758 return $ids[$lctext] ?? false;
759 }
760
761 /**
762 * short names for language variants used for language conversion links.
763 *
764 * @param string $code
765 * @param bool $usemsg Use the "variantname-xyz" message if it exists
766 * @return string
767 */
768 public function getVariantname( $code, $usemsg = true ) {
769 $msg = "variantname-$code";
770 if ( $usemsg && wfMessage( $msg )->exists() ) {
771 return $this->getMessageFromDB( $msg );
772 }
773 $name = self::fetchLanguageName( $code );
774 if ( $name ) {
775 return $name; # if it's defined as a language name, show that
776 } else {
777 # otherwise, output the language code
778 return $code;
779 }
780 }
781
782 /**
783 * @return string[]|bool List of date format preference keys, or false if disabled.
784 */
785 public function getDatePreferences() {
786 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
787 }
788
789 /**
790 * @return array
791 */
792 function getDateFormats() {
793 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
794 }
795
796 /**
797 * @return array|string
798 */
799 public function getDefaultDateFormat() {
800 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
801 if ( $df === 'dmy or mdy' ) {
802 global $wgAmericanDates;
803 return $wgAmericanDates ? 'mdy' : 'dmy';
804 } else {
805 return $df;
806 }
807 }
808
809 /**
810 * @return array
811 */
812 public function getDatePreferenceMigrationMap() {
813 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
814 }
815
816 /**
817 * @return array
818 */
819 public function getExtraUserToggles() {
820 wfDeprecated( __METHOD__, '1.34' );
821 return [];
822 }
823
824 /**
825 * @param string $tog
826 * @return string
827 */
828 function getUserToggle( $tog ) {
829 return $this->getMessageFromDB( "tog-$tog" );
830 }
831
832 /**
833 * Get an array of language names, indexed by code.
834 * @param null|string $inLanguage Code of language in which to return the names
835 * Use self::AS_AUTONYMS for autonyms (native names)
836 * @param string $include One of:
837 * self::ALL all available languages
838 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
839 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
840 * @return array Language code => language name (sorted by key)
841 * @since 1.20
842 */
843 public static function fetchLanguageNames( $inLanguage = self::AS_AUTONYMS, $include = 'mw' ) {
844 $cacheKey = $inLanguage === self::AS_AUTONYMS ? 'null' : $inLanguage;
845 $cacheKey .= ":$include";
846 if ( self::$languageNameCache === null ) {
847 self::$languageNameCache = new HashBagOStuff( [ 'maxKeys' => 20 ] );
848 }
849
850 $ret = self::$languageNameCache->get( $cacheKey );
851 if ( !$ret ) {
852 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
853 self::$languageNameCache->set( $cacheKey, $ret );
854 }
855 return $ret;
856 }
857
858 /**
859 * Uncached helper for fetchLanguageNames
860 * @param null|string $inLanguage Code of language in which to return the names
861 * Use self::AS_AUTONYMS for autonyms (native names)
862 * @param string $include One of:
863 * self::ALL all available languages
864 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
865 * self::SUPPORTED only if the language is in 'mw' *and* has a message file
866 * @return array Language code => language name (sorted by key)
867 */
868 private static function fetchLanguageNamesUncached(
869 $inLanguage = self::AS_AUTONYMS,
870 $include = 'mw'
871 ) {
872 global $wgExtraLanguageNames, $wgUsePigLatinVariant;
873
874 // If passed an invalid language code to use, fallback to en
875 if ( $inLanguage !== self::AS_AUTONYMS && !self::isValidCode( $inLanguage ) ) {
876 $inLanguage = 'en';
877 }
878
879 $names = [];
880
881 if ( $inLanguage ) {
882 # TODO: also include when $inLanguage is null, when this code is more efficient
883 Hooks::run( 'LanguageGetTranslatedLanguageNames', [ &$names, $inLanguage ] );
884 }
885
886 $mwNames = $wgExtraLanguageNames + MediaWiki\Languages\Data\Names::$names;
887 if ( $wgUsePigLatinVariant ) {
888 // Pig Latin (for variant development)
889 $mwNames['en-x-piglatin'] = 'Igpay Atinlay';
890 }
891
892 foreach ( $mwNames as $mwCode => $mwName ) {
893 # - Prefer own MediaWiki native name when not using the hook
894 # - For other names just add if not added through the hook
895 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
896 $names[$mwCode] = $mwName;
897 }
898 }
899
900 if ( $include === self::ALL ) {
901 ksort( $names );
902 return $names;
903 }
904
905 $returnMw = [];
906 $coreCodes = array_keys( $mwNames );
907 foreach ( $coreCodes as $coreCode ) {
908 $returnMw[$coreCode] = $names[$coreCode];
909 }
910
911 if ( $include === self::SUPPORTED ) {
912 $namesMwFile = [];
913 # We do this using a foreach over the codes instead of a directory
914 # loop so that messages files in extensions will work correctly.
915 foreach ( $returnMw as $code => $value ) {
916 if ( is_readable( self::getMessagesFileName( $code ) )
917 || is_readable( self::getJsonMessagesFileName( $code ) )
918 ) {
919 $namesMwFile[$code] = $names[$code];
920 }
921 }
922
923 ksort( $namesMwFile );
924 return $namesMwFile;
925 }
926
927 ksort( $returnMw );
928 # 'mw' option; default if it's not one of the other two options (all/mwfile)
929 return $returnMw;
930 }
931
932 /**
933 * @param string $code The code of the language for which to get the name
934 * @param null|string $inLanguage Code of language in which to return the name
935 * (SELF::AS_AUTONYMS for autonyms)
936 * @param string $include See fetchLanguageNames()
937 * @return string Language name or empty
938 * @since 1.20
939 */
940 public static function fetchLanguageName(
941 $code,
942 $inLanguage = self::AS_AUTONYMS,
943 $include = self::ALL
944 ) {
945 $code = strtolower( $code );
946 $array = self::fetchLanguageNames( $inLanguage, $include );
947 return !array_key_exists( $code, $array ) ? '' : $array[$code];
948 }
949
950 /**
951 * Get a message from the MediaWiki namespace.
952 *
953 * @param string $msg Message name
954 * @return string
955 */
956 public function getMessageFromDB( $msg ) {
957 return $this->msg( $msg )->text();
958 }
959
960 /**
961 * Get message object in this language. Only for use inside this class.
962 *
963 * @param string $msg Message name
964 * @return Message
965 */
966 protected function msg( $msg ) {
967 return wfMessage( $msg )->inLanguage( $this );
968 }
969
970 /**
971 * @param string $key
972 * @return string
973 */
974 public function getMonthName( $key ) {
975 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
976 }
977
978 /**
979 * @return array
980 */
981 public function getMonthNamesArray() {
982 $monthNames = [ '' ];
983 for ( $i = 1; $i < 13; $i++ ) {
984 $monthNames[] = $this->getMonthName( $i );
985 }
986 return $monthNames;
987 }
988
989 /**
990 * @param string $key
991 * @return string
992 */
993 public function getMonthNameGen( $key ) {
994 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
995 }
996
997 /**
998 * @param string $key
999 * @return string
1000 */
1001 public function getMonthAbbreviation( $key ) {
1002 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
1003 }
1004
1005 /**
1006 * @return array
1007 */
1008 public function getMonthAbbreviationsArray() {
1009 $monthNames = [ '' ];
1010 for ( $i = 1; $i < 13; $i++ ) {
1011 $monthNames[] = $this->getMonthAbbreviation( $i );
1012 }
1013 return $monthNames;
1014 }
1015
1016 /**
1017 * @param string $key
1018 * @return string
1019 */
1020 public function getWeekdayName( $key ) {
1021 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
1022 }
1023
1024 /**
1025 * @param string $key
1026 * @return string
1027 */
1028 function getWeekdayAbbreviation( $key ) {
1029 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1030 }
1031
1032 /**
1033 * @param string $key
1034 * @return string
1035 */
1036 function getIranianCalendarMonthName( $key ) {
1037 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1038 }
1039
1040 /**
1041 * @param string $key
1042 * @return string
1043 */
1044 function getHebrewCalendarMonthName( $key ) {
1045 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1046 }
1047
1048 /**
1049 * @param string $key
1050 * @return string
1051 */
1052 function getHebrewCalendarMonthNameGen( $key ) {
1053 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1054 }
1055
1056 /**
1057 * @param string $key
1058 * @return string
1059 */
1060 function getHijriCalendarMonthName( $key ) {
1061 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1062 }
1063
1064 /**
1065 * Pass through result from $dateTimeObj->format()
1066 * @param DateTime|bool|null &$dateTimeObj
1067 * @param string $ts
1068 * @param DateTimeZone|bool|null $zone
1069 * @param string $code
1070 * @return string
1071 */
1072 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1073 if ( !$dateTimeObj ) {
1074 $dateTimeObj = DateTime::createFromFormat(
1075 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1076 );
1077 }
1078 return $dateTimeObj->format( $code );
1079 }
1080
1081 /**
1082 * This is a workalike of PHP's date() function, but with better
1083 * internationalisation, a reduced set of format characters, and a better
1084 * escaping format.
1085 *
1086 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1087 * the PHP manual for definitions. There are a number of extensions, which
1088 * start with "x":
1089 *
1090 * xn Do not translate digits of the next numeric format character
1091 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1092 * xr Use roman numerals for the next numeric format character
1093 * xh Use hebrew numerals for the next numeric format character
1094 * xx Literal x
1095 * xg Genitive month name
1096 *
1097 * xij j (day number) in Iranian calendar
1098 * xiF F (month name) in Iranian calendar
1099 * xin n (month number) in Iranian calendar
1100 * xiy y (two digit year) in Iranian calendar
1101 * xiY Y (full year) in Iranian calendar
1102 * xit t (days in month) in Iranian calendar
1103 * xiz z (day of the year) in Iranian calendar
1104 *
1105 * xjj j (day number) in Hebrew calendar
1106 * xjF F (month name) in Hebrew calendar
1107 * xjt t (days in month) in Hebrew calendar
1108 * xjx xg (genitive month name) in Hebrew calendar
1109 * xjn n (month number) in Hebrew calendar
1110 * xjY Y (full year) in Hebrew calendar
1111 *
1112 * xmj j (day number) in Hijri calendar
1113 * xmF F (month name) in Hijri calendar
1114 * xmn n (month number) in Hijri calendar
1115 * xmY Y (full year) in Hijri calendar
1116 *
1117 * xkY Y (full year) in Thai solar calendar. Months and days are
1118 * identical to the Gregorian calendar
1119 * xoY Y (full year) in Minguo calendar or Juche year.
1120 * Months and days are identical to the
1121 * Gregorian calendar
1122 * xtY Y (full year) in Japanese nengo. Months and days are
1123 * identical to the Gregorian calendar
1124 *
1125 * Characters enclosed in double quotes will be considered literal (with
1126 * the quotes themselves removed). Unmatched quotes will be considered
1127 * literal quotes. Example:
1128 *
1129 * "The month is" F => The month is January
1130 * i's" => 20'11"
1131 *
1132 * Backslash escaping is also supported.
1133 *
1134 * Input timestamp is assumed to be pre-normalized to the desired local
1135 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1136 * $ts is UTC if $zone is not given.
1137 *
1138 * @param string $format
1139 * @param string $ts 14-character timestamp
1140 * YYYYMMDDHHMMSS
1141 * 01234567890123
1142 * @param DateTimeZone|null $zone Timezone of $ts
1143 * @param int &$ttl The amount of time (in seconds) the output may be cached for.
1144 * Only makes sense if $ts is the current time.
1145 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1146 *
1147 * @throws MWException
1148 * @return string
1149 */
1150 public function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = 'unused' ) {
1151 $s = '';
1152 $raw = false;
1153 $roman = false;
1154 $hebrewNum = false;
1155 $dateTimeObj = false;
1156 $rawToggle = false;
1157 $iranian = false;
1158 $hebrew = false;
1159 $hijri = false;
1160 $thai = false;
1161 $minguo = false;
1162 $tenno = false;
1163
1164 $usedSecond = false;
1165 $usedMinute = false;
1166 $usedHour = false;
1167 $usedAMPM = false;
1168 $usedDay = false;
1169 $usedWeek = false;
1170 $usedMonth = false;
1171 $usedYear = false;
1172 $usedISOYear = false;
1173 $usedIsLeapYear = false;
1174
1175 $usedHebrewMonth = false;
1176 $usedIranianMonth = false;
1177 $usedHijriMonth = false;
1178 $usedHebrewYear = false;
1179 $usedIranianYear = false;
1180 $usedHijriYear = false;
1181 $usedTennoYear = false;
1182
1183 if ( strlen( $ts ) !== 14 ) {
1184 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1185 }
1186
1187 if ( !ctype_digit( $ts ) ) {
1188 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1189 }
1190
1191 $formatLength = strlen( $format );
1192 for ( $p = 0; $p < $formatLength; $p++ ) {
1193 $num = false;
1194 $code = $format[$p];
1195 if ( $code == 'x' && $p < $formatLength - 1 ) {
1196 $code .= $format[++$p];
1197 }
1198
1199 if ( ( $code === 'xi'
1200 || $code === 'xj'
1201 || $code === 'xk'
1202 || $code === 'xm'
1203 || $code === 'xo'
1204 || $code === 'xt' )
1205 && $p < $formatLength - 1 ) {
1206 $code .= $format[++$p];
1207 }
1208
1209 switch ( $code ) {
1210 case 'xx':
1211 $s .= 'x';
1212 break;
1213 case 'xn':
1214 $raw = true;
1215 break;
1216 case 'xN':
1217 $rawToggle = !$rawToggle;
1218 break;
1219 case 'xr':
1220 $roman = true;
1221 break;
1222 case 'xh':
1223 $hebrewNum = true;
1224 break;
1225 case 'xg':
1226 $usedMonth = true;
1227 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1228 break;
1229 case 'xjx':
1230 $usedHebrewMonth = true;
1231 if ( !$hebrew ) {
1232 $hebrew = self::tsToHebrew( $ts );
1233 }
1234 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1235 break;
1236 case 'd':
1237 $usedDay = true;
1238 $num = substr( $ts, 6, 2 );
1239 break;
1240 case 'D':
1241 $usedDay = true;
1242 $s .= $this->getWeekdayAbbreviation(
1243 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1244 );
1245 break;
1246 case 'j':
1247 $usedDay = true;
1248 $num = intval( substr( $ts, 6, 2 ) );
1249 break;
1250 case 'xij':
1251 $usedDay = true;
1252 if ( !$iranian ) {
1253 $iranian = self::tsToIranian( $ts );
1254 }
1255 $num = $iranian[2];
1256 break;
1257 case 'xmj':
1258 $usedDay = true;
1259 if ( !$hijri ) {
1260 $hijri = self::tsToHijri( $ts );
1261 }
1262 $num = $hijri[2];
1263 break;
1264 case 'xjj':
1265 $usedDay = true;
1266 if ( !$hebrew ) {
1267 $hebrew = self::tsToHebrew( $ts );
1268 }
1269 $num = $hebrew[2];
1270 break;
1271 case 'l':
1272 $usedDay = true;
1273 $s .= $this->getWeekdayName(
1274 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1275 );
1276 break;
1277 case 'F':
1278 $usedMonth = true;
1279 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1280 break;
1281 case 'xiF':
1282 $usedIranianMonth = true;
1283 if ( !$iranian ) {
1284 $iranian = self::tsToIranian( $ts );
1285 }
1286 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1287 break;
1288 case 'xmF':
1289 $usedHijriMonth = true;
1290 if ( !$hijri ) {
1291 $hijri = self::tsToHijri( $ts );
1292 }
1293 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1294 break;
1295 case 'xjF':
1296 $usedHebrewMonth = true;
1297 if ( !$hebrew ) {
1298 $hebrew = self::tsToHebrew( $ts );
1299 }
1300 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1301 break;
1302 case 'm':
1303 $usedMonth = true;
1304 $num = substr( $ts, 4, 2 );
1305 break;
1306 case 'M':
1307 $usedMonth = true;
1308 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1309 break;
1310 case 'n':
1311 $usedMonth = true;
1312 $num = intval( substr( $ts, 4, 2 ) );
1313 break;
1314 case 'xin':
1315 $usedIranianMonth = true;
1316 if ( !$iranian ) {
1317 $iranian = self::tsToIranian( $ts );
1318 }
1319 $num = $iranian[1];
1320 break;
1321 case 'xmn':
1322 $usedHijriMonth = true;
1323 if ( !$hijri ) {
1324 $hijri = self::tsToHijri( $ts );
1325 }
1326 $num = $hijri[1];
1327 break;
1328 case 'xjn':
1329 $usedHebrewMonth = true;
1330 if ( !$hebrew ) {
1331 $hebrew = self::tsToHebrew( $ts );
1332 }
1333 $num = $hebrew[1];
1334 break;
1335 case 'xjt':
1336 $usedHebrewMonth = true;
1337 if ( !$hebrew ) {
1338 $hebrew = self::tsToHebrew( $ts );
1339 }
1340 $num = $hebrew[3];
1341 break;
1342 case 'Y':
1343 $usedYear = true;
1344 $num = substr( $ts, 0, 4 );
1345 break;
1346 case 'xiY':
1347 $usedIranianYear = true;
1348 if ( !$iranian ) {
1349 $iranian = self::tsToIranian( $ts );
1350 }
1351 $num = $iranian[0];
1352 break;
1353 case 'xmY':
1354 $usedHijriYear = true;
1355 if ( !$hijri ) {
1356 $hijri = self::tsToHijri( $ts );
1357 }
1358 $num = $hijri[0];
1359 break;
1360 case 'xjY':
1361 $usedHebrewYear = true;
1362 if ( !$hebrew ) {
1363 $hebrew = self::tsToHebrew( $ts );
1364 }
1365 $num = $hebrew[0];
1366 break;
1367 case 'xkY':
1368 $usedYear = true;
1369 if ( !$thai ) {
1370 $thai = self::tsToYear( $ts, 'thai' );
1371 }
1372 $num = $thai[0];
1373 break;
1374 case 'xoY':
1375 $usedYear = true;
1376 if ( !$minguo ) {
1377 $minguo = self::tsToYear( $ts, 'minguo' );
1378 }
1379 $num = $minguo[0];
1380 break;
1381 case 'xtY':
1382 $usedTennoYear = true;
1383 if ( !$tenno ) {
1384 $tenno = self::tsToYear( $ts, 'tenno' );
1385 }
1386 $num = $tenno[0];
1387 break;
1388 case 'y':
1389 $usedYear = true;
1390 $num = substr( $ts, 2, 2 );
1391 break;
1392 case 'xiy':
1393 $usedIranianYear = true;
1394 if ( !$iranian ) {
1395 $iranian = self::tsToIranian( $ts );
1396 }
1397 $num = substr( $iranian[0], -2 );
1398 break;
1399 case 'xit':
1400 $usedIranianYear = true;
1401 if ( !$iranian ) {
1402 $iranian = self::tsToIranian( $ts );
1403 }
1404 $num = self::$IRANIAN_DAYS[$iranian[1] - 1];
1405 break;
1406 case 'xiz':
1407 $usedIranianYear = true;
1408 if ( !$iranian ) {
1409 $iranian = self::tsToIranian( $ts );
1410 }
1411 $num = $iranian[3];
1412 break;
1413 case 'a':
1414 $usedAMPM = true;
1415 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1416 break;
1417 case 'A':
1418 $usedAMPM = true;
1419 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1420 break;
1421 case 'g':
1422 $usedHour = true;
1423 $h = substr( $ts, 8, 2 );
1424 $num = $h % 12 ?: 12;
1425 break;
1426 case 'G':
1427 $usedHour = true;
1428 $num = intval( substr( $ts, 8, 2 ) );
1429 break;
1430 case 'h':
1431 $usedHour = true;
1432 $h = substr( $ts, 8, 2 );
1433 $num = sprintf( '%02d', $h % 12 ?: 12 );
1434 break;
1435 case 'H':
1436 $usedHour = true;
1437 $num = substr( $ts, 8, 2 );
1438 break;
1439 case 'i':
1440 $usedMinute = true;
1441 $num = substr( $ts, 10, 2 );
1442 break;
1443 case 's':
1444 $usedSecond = true;
1445 $num = substr( $ts, 12, 2 );
1446 break;
1447 case 'c':
1448 case 'r':
1449 $usedSecond = true;
1450 // fall through
1451 case 'e':
1452 case 'O':
1453 case 'P':
1454 case 'T':
1455 $s .= self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1456 break;
1457 case 'w':
1458 case 'N':
1459 case 'z':
1460 $usedDay = true;
1461 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1462 break;
1463 case 'W':
1464 $usedWeek = true;
1465 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1466 break;
1467 case 't':
1468 $usedMonth = true;
1469 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1470 break;
1471 case 'L':
1472 $usedIsLeapYear = true;
1473 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1474 break;
1475 case 'o':
1476 $usedISOYear = true;
1477 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1478 break;
1479 case 'U':
1480 $usedSecond = true;
1481 // fall through
1482 case 'I':
1483 case 'Z':
1484 $num = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1485 break;
1486 case '\\':
1487 # Backslash escaping
1488 if ( $p < $formatLength - 1 ) {
1489 $s .= $format[++$p];
1490 } else {
1491 $s .= '\\';
1492 }
1493 break;
1494 case '"':
1495 # Quoted literal
1496 if ( $p < $formatLength - 1 ) {
1497 $endQuote = strpos( $format, '"', $p + 1 );
1498 if ( $endQuote === false ) {
1499 # No terminating quote, assume literal "
1500 $s .= '"';
1501 } else {
1502 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1503 $p = $endQuote;
1504 }
1505 } else {
1506 # Quote at end of string, assume literal "
1507 $s .= '"';
1508 }
1509 break;
1510 default:
1511 $s .= $format[$p];
1512 }
1513 if ( $num !== false ) {
1514 if ( $rawToggle || $raw ) {
1515 $s .= $num;
1516 $raw = false;
1517 } elseif ( $roman ) {
1518 $s .= self::romanNumeral( $num );
1519 $roman = false;
1520 } elseif ( $hebrewNum ) {
1521 $s .= self::hebrewNumeral( $num );
1522 $hebrewNum = false;
1523 } else {
1524 $s .= $this->formatNum( $num, true );
1525 }
1526 }
1527 }
1528
1529 if ( $ttl === 'unused' ) {
1530 // No need to calculate the TTL, the caller wont use it anyway.
1531 } elseif ( $usedSecond ) {
1532 $ttl = 1;
1533 } elseif ( $usedMinute ) {
1534 $ttl = 60 - substr( $ts, 12, 2 );
1535 } elseif ( $usedHour ) {
1536 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1537 } elseif ( $usedAMPM ) {
1538 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1539 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1540 } elseif (
1541 $usedDay ||
1542 $usedHebrewMonth ||
1543 $usedIranianMonth ||
1544 $usedHijriMonth ||
1545 $usedHebrewYear ||
1546 $usedIranianYear ||
1547 $usedHijriYear ||
1548 $usedTennoYear
1549 ) {
1550 // @todo Someone who understands the non-Gregorian calendars
1551 // should write proper logic for them so that they don't need purged every day.
1552 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1553 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1554 } else {
1555 $possibleTtls = [];
1556 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1557 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1558 if ( $usedWeek ) {
1559 $possibleTtls[] =
1560 ( 7 - self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1561 $timeRemainingInDay;
1562 } elseif ( $usedISOYear ) {
1563 // December 28th falls on the last ISO week of the year, every year.
1564 // The last ISO week of a year can be 52 or 53.
1565 $lastWeekOfISOYear = DateTime::createFromFormat(
1566 'Ymd',
1567 substr( $ts, 0, 4 ) . '1228',
1568 $zone ?: new DateTimeZone( 'UTC' )
1569 )->format( 'W' );
1570 $currentISOWeek = self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1571 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1572 $timeRemainingInWeek =
1573 ( 7 - self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1574 + $timeRemainingInDay;
1575 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1576 }
1577
1578 if ( $usedMonth ) {
1579 $possibleTtls[] =
1580 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1581 substr( $ts, 6, 2 ) ) * 86400
1582 + $timeRemainingInDay;
1583 } elseif ( $usedYear ) {
1584 $possibleTtls[] =
1585 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1586 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1587 + $timeRemainingInDay;
1588 } elseif ( $usedIsLeapYear ) {
1589 $year = substr( $ts, 0, 4 );
1590 $timeRemainingInYear =
1591 ( self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1592 self::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1593 + $timeRemainingInDay;
1594 $mod = $year % 4;
1595 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1596 // this isn't a leap year. see when the next one starts
1597 $nextCandidate = $year - $mod + 4;
1598 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1599 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1600 $timeRemainingInYear;
1601 } else {
1602 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1603 $timeRemainingInYear;
1604 }
1605 } else {
1606 // this is a leap year, so the next year isn't
1607 $possibleTtls[] = $timeRemainingInYear;
1608 }
1609 }
1610
1611 if ( $possibleTtls ) {
1612 $ttl = min( $possibleTtls );
1613 }
1614 }
1615
1616 return $s;
1617 }
1618
1619 private static $GREG_DAYS = [ 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ];
1620 private static $IRANIAN_DAYS = [ 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 ];
1621
1622 /**
1623 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1624 * Gregorian dates to Iranian dates. Originally written in C, it
1625 * is released under the terms of GNU Lesser General Public
1626 * License. Conversion to PHP was performed by Niklas Laxström.
1627 *
1628 * Link: http://www.farsiweb.info/jalali/jalali.c
1629 *
1630 * @param string $ts
1631 *
1632 * @return int[]
1633 */
1634 private static function tsToIranian( $ts ) {
1635 $gy = substr( $ts, 0, 4 ) - 1600;
1636 $gm = substr( $ts, 4, 2 ) - 1;
1637 $gd = substr( $ts, 6, 2 ) - 1;
1638
1639 # Days passed from the beginning (including leap years)
1640 $gDayNo = 365 * $gy
1641 + floor( ( $gy + 3 ) / 4 )
1642 - floor( ( $gy + 99 ) / 100 )
1643 + floor( ( $gy + 399 ) / 400 );
1644
1645 // Add days of the past months of this year
1646 for ( $i = 0; $i < $gm; $i++ ) {
1647 $gDayNo += self::$GREG_DAYS[$i];
1648 }
1649
1650 // Leap years
1651 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1652 $gDayNo++;
1653 }
1654
1655 // Days passed in current month
1656 $gDayNo += (int)$gd;
1657
1658 $jDayNo = $gDayNo - 79;
1659
1660 $jNp = floor( $jDayNo / 12053 );
1661 $jDayNo %= 12053;
1662
1663 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1664 $jDayNo %= 1461;
1665
1666 if ( $jDayNo >= 366 ) {
1667 $jy += floor( ( $jDayNo - 1 ) / 365 );
1668 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1669 }
1670
1671 $jz = $jDayNo;
1672
1673 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1674 $jDayNo -= self::$IRANIAN_DAYS[$i];
1675 }
1676
1677 $jm = $i + 1;
1678 $jd = $jDayNo + 1;
1679
1680 return [ $jy, $jm, $jd, $jz ];
1681 }
1682
1683 /**
1684 * Converting Gregorian dates to Hijri dates.
1685 *
1686 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1687 *
1688 * @see https://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1689 *
1690 * @param string $ts
1691 *
1692 * @return int[]
1693 */
1694 private static function tsToHijri( $ts ) {
1695 $year = substr( $ts, 0, 4 );
1696 $month = substr( $ts, 4, 2 );
1697 $day = substr( $ts, 6, 2 );
1698
1699 $zyr = $year;
1700 $zd = $day;
1701 $zm = $month;
1702 $zy = $zyr;
1703
1704 if (
1705 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1706 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1707 ) {
1708 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1709 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1710 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1711 $zd - 32075;
1712 } else {
1713 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1714 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1715 }
1716
1717 $zl = $zjd - 1948440 + 10632;
1718 $zn = (int)( ( $zl - 1 ) / 10631 );
1719 $zl = $zl - 10631 * $zn + 354;
1720 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1721 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1722 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1723 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1724 $zm = (int)( ( 24 * $zl ) / 709 );
1725 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1726 $zy = 30 * $zn + $zj - 30;
1727
1728 return [ $zy, $zm, $zd ];
1729 }
1730
1731 /**
1732 * Converting Gregorian dates to Hebrew dates.
1733 *
1734 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1735 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1736 * to translate the relevant functions into PHP and release them under
1737 * GNU GPL.
1738 *
1739 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1740 * and Adar II is 14. In a non-leap year, Adar is 6.
1741 *
1742 * @param string $ts
1743 *
1744 * @return int[]
1745 */
1746 private static function tsToHebrew( $ts ) {
1747 # Parse date
1748 $year = substr( $ts, 0, 4 );
1749 $month = substr( $ts, 4, 2 );
1750 $day = substr( $ts, 6, 2 );
1751
1752 # Calculate Hebrew year
1753 $hebrewYear = $year + 3760;
1754
1755 # Month number when September = 1, August = 12
1756 $month += 4;
1757 if ( $month > 12 ) {
1758 # Next year
1759 $month -= 12;
1760 $year++;
1761 $hebrewYear++;
1762 }
1763
1764 # Calculate day of year from 1 September
1765 $dayOfYear = $day;
1766 for ( $i = 1; $i < $month; $i++ ) {
1767 if ( $i == 6 ) {
1768 # February
1769 $dayOfYear += 28;
1770 # Check if the year is leap
1771 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1772 $dayOfYear++;
1773 }
1774 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1775 $dayOfYear += 30;
1776 } else {
1777 $dayOfYear += 31;
1778 }
1779 }
1780
1781 # Calculate the start of the Hebrew year
1782 $start = self::hebrewYearStart( $hebrewYear );
1783
1784 # Calculate next year's start
1785 if ( $dayOfYear <= $start ) {
1786 # Day is before the start of the year - it is the previous year
1787 # Next year's start
1788 $nextStart = $start;
1789 # Previous year
1790 $year--;
1791 $hebrewYear--;
1792 # Add days since previous year's 1 September
1793 $dayOfYear += 365;
1794 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1795 # Leap year
1796 $dayOfYear++;
1797 }
1798 # Start of the new (previous) year
1799 $start = self::hebrewYearStart( $hebrewYear );
1800 } else {
1801 # Next year's start
1802 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1803 }
1804
1805 # Calculate Hebrew day of year
1806 $hebrewDayOfYear = $dayOfYear - $start;
1807
1808 # Difference between year's days
1809 $diff = $nextStart - $start;
1810 # Add 12 (or 13 for leap years) days to ignore the difference between
1811 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1812 # difference is only about the year type
1813 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1814 $diff += 13;
1815 } else {
1816 $diff += 12;
1817 }
1818
1819 # Check the year pattern, and is leap year
1820 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1821 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1822 # and non-leap years
1823 $yearPattern = $diff % 30;
1824 # Check if leap year
1825 $isLeap = $diff >= 30;
1826
1827 # Calculate day in the month from number of day in the Hebrew year
1828 # Don't check Adar - if the day is not in Adar, we will stop before;
1829 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1830 $hebrewDay = $hebrewDayOfYear;
1831 $hebrewMonth = 1;
1832 $days = 0;
1833 while ( $hebrewMonth <= 12 ) {
1834 # Calculate days in this month
1835 if ( $isLeap && $hebrewMonth == 6 ) {
1836 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1837 $days = 30;
1838 if ( $hebrewDay <= $days ) {
1839 # Day in Adar I
1840 $hebrewMonth = 13;
1841 } else {
1842 # Subtract the days of Adar I
1843 $hebrewDay -= $days;
1844 # Try Adar II
1845 $days = 29;
1846 if ( $hebrewDay <= $days ) {
1847 # Day in Adar II
1848 $hebrewMonth = 14;
1849 }
1850 }
1851 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1852 # Cheshvan in a complete year (otherwise as the rule below)
1853 $days = 30;
1854 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1855 # Kislev in an incomplete year (otherwise as the rule below)
1856 $days = 29;
1857 } else {
1858 # Odd months have 30 days, even have 29
1859 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1860 }
1861 if ( $hebrewDay <= $days ) {
1862 # In the current month
1863 break;
1864 } else {
1865 # Subtract the days of the current month
1866 $hebrewDay -= $days;
1867 # Try in the next month
1868 $hebrewMonth++;
1869 }
1870 }
1871
1872 return [ $hebrewYear, $hebrewMonth, $hebrewDay, $days ];
1873 }
1874
1875 /**
1876 * This calculates the Hebrew year start, as days since 1 September.
1877 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1878 * Used for Hebrew date.
1879 *
1880 * @param int $year
1881 *
1882 * @return string
1883 */
1884 private static function hebrewYearStart( $year ) {
1885 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1886 $b = intval( ( $year - 1 ) % 4 );
1887 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1888 if ( $m < 0 ) {
1889 $m--;
1890 }
1891 $Mar = intval( $m );
1892 if ( $m < 0 ) {
1893 $m++;
1894 }
1895 $m -= $Mar;
1896
1897 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1898 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1899 $Mar++;
1900 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1901 $Mar += 2;
1902 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1903 $Mar++;
1904 }
1905
1906 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1907 return $Mar;
1908 }
1909
1910 /**
1911 * Algorithm to convert Gregorian dates to Thai solar dates,
1912 * Minguo dates or Minguo dates.
1913 *
1914 * Link: https://en.wikipedia.org/wiki/Thai_solar_calendar
1915 * https://en.wikipedia.org/wiki/Minguo_calendar
1916 * https://en.wikipedia.org/wiki/Japanese_era_name
1917 *
1918 * @param string $ts 14-character timestamp
1919 * @param string $cName Calender name
1920 * @return array Converted year, month, day
1921 */
1922 private static function tsToYear( $ts, $cName ) {
1923 $gy = substr( $ts, 0, 4 );
1924 $gm = substr( $ts, 4, 2 );
1925 $gd = substr( $ts, 6, 2 );
1926
1927 if ( !strcmp( $cName, 'thai' ) ) {
1928 # Thai solar dates
1929 # Add 543 years to the Gregorian calendar
1930 # Months and days are identical
1931 $gy_offset = $gy + 543;
1932 # fix for dates between 1912 and 1941
1933 # https://en.wikipedia.org/?oldid=836596673#New_year
1934 if ( $gy >= 1912 && $gy <= 1940 ) {
1935 if ( $gm <= 3 ) {
1936 $gy_offset--;
1937 }
1938 $gm = ( $gm - 3 ) % 12;
1939 }
1940 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1941 # Minguo dates
1942 # Deduct 1911 years from the Gregorian calendar
1943 # Months and days are identical
1944 $gy_offset = $gy - 1911;
1945 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1946 # Nengō dates up to Meiji period
1947 # Deduct years from the Gregorian calendar
1948 # depending on the nengo periods
1949 # Months and days are identical
1950 if ( ( $gy < 1912 )
1951 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1952 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1953 ) {
1954 # Meiji period
1955 $gy_gannen = $gy - 1868 + 1;
1956 $gy_offset = $gy_gannen;
1957 if ( $gy_gannen == 1 ) {
1958 $gy_offset = '元';
1959 }
1960 $gy_offset = '明治' . $gy_offset;
1961 } elseif (
1962 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1963 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1964 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1965 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1966 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1967 ) {
1968 # Taishō period
1969 $gy_gannen = $gy - 1912 + 1;
1970 $gy_offset = $gy_gannen;
1971 if ( $gy_gannen == 1 ) {
1972 $gy_offset = '元';
1973 }
1974 $gy_offset = '大正' . $gy_offset;
1975 } elseif (
1976 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1977 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1978 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1979 ) {
1980 # Shōwa period
1981 $gy_gannen = $gy - 1926 + 1;
1982 $gy_offset = $gy_gannen;
1983 if ( $gy_gannen == 1 ) {
1984 $gy_offset = '元';
1985 }
1986 $gy_offset = '昭和' . $gy_offset;
1987 } elseif (
1988 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd >= 8 ) ) ||
1989 ( ( $gy > 1989 ) && ( $gy < 2019 ) ) ||
1990 ( ( $gy == 2019 ) && ( $gm < 5 ) )
1991 ) {
1992 # Heisei period
1993 $gy_gannen = $gy - 1989 + 1;
1994 $gy_offset = $gy_gannen;
1995 if ( $gy_gannen == 1 ) {
1996 $gy_offset = '元';
1997 }
1998 $gy_offset = '平成' . $gy_offset;
1999 } else {
2000 # Reiwa period
2001 $gy_gannen = $gy - 2019 + 1;
2002 $gy_offset = $gy_gannen;
2003 if ( $gy_gannen == 1 ) {
2004 $gy_offset = '元';
2005 }
2006 $gy_offset = '令和' . $gy_offset;
2007 }
2008 } else {
2009 $gy_offset = $gy;
2010 }
2011
2012 return [ $gy_offset, $gm, $gd ];
2013 }
2014
2015 /**
2016 * Gets directionality of the first strongly directional codepoint, for embedBidi()
2017 *
2018 * This is the rule the BIDI algorithm uses to determine the directionality of
2019 * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
2020 * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
2021 *
2022 * TODO: Does not handle BIDI control characters inside the text.
2023 * TODO: Does not handle unallocated characters.
2024 *
2025 * @param string $text Text to test
2026 * @return null|string Directionality ('ltr' or 'rtl') or null
2027 */
2028 private static function strongDirFromContent( $text = '' ) {
2029 if ( !preg_match( self::$strongDirRegex, $text, $matches ) ) {
2030 return null;
2031 }
2032 if ( $matches[1] === '' ) {
2033 return 'rtl';
2034 }
2035 return 'ltr';
2036 }
2037
2038 /**
2039 * Roman number formatting up to 10000
2040 *
2041 * @param int $num
2042 *
2043 * @return string
2044 */
2045 static function romanNumeral( $num ) {
2046 static $table = [
2047 [ '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ],
2048 [ '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ],
2049 [ '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ],
2050 [ '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2051 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' ]
2052 ];
2053
2054 $num = intval( $num );
2055 if ( $num > 10000 || $num <= 0 ) {
2056 return $num;
2057 }
2058
2059 $s = '';
2060 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2061 if ( $num >= $pow10 ) {
2062 $s .= $table[$i][(int)floor( $num / $pow10 )];
2063 }
2064 $num = $num % $pow10;
2065 }
2066 return $s;
2067 }
2068
2069 /**
2070 * Hebrew Gematria number formatting up to 9999
2071 *
2072 * @param int $num
2073 *
2074 * @return string
2075 */
2076 static function hebrewNumeral( $num ) {
2077 static $table = [
2078 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ],
2079 [ '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ],
2080 [ '',
2081 [ 'ק' ],
2082 [ 'ר' ],
2083 [ 'ש' ],
2084 [ 'ת' ],
2085 [ 'ת', 'ק' ],
2086 [ 'ת', 'ר' ],
2087 [ 'ת', 'ש' ],
2088 [ 'ת', 'ת' ],
2089 [ 'ת', 'ת', 'ק' ],
2090 [ 'ת', 'ת', 'ר' ],
2091 ],
2092 [ '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ]
2093 ];
2094
2095 $num = intval( $num );
2096 if ( $num > 9999 || $num <= 0 ) {
2097 return $num;
2098 }
2099
2100 // Round thousands have special notations
2101 if ( $num === 1000 ) {
2102 return "א' אלף";
2103 } elseif ( $num % 1000 === 0 ) {
2104 return $table[0][$num / 1000] . "' אלפים";
2105 }
2106
2107 $letters = [];
2108
2109 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2110 if ( $num >= $pow10 ) {
2111 if ( $num === 15 || $num === 16 ) {
2112 $letters[] = $table[0][9];
2113 $letters[] = $table[0][$num - 9];
2114 $num = 0;
2115 } else {
2116 $letters = array_merge(
2117 $letters,
2118 (array)$table[$i][intval( $num / $pow10 )]
2119 );
2120
2121 if ( $pow10 === 1000 ) {
2122 $letters[] = "'";
2123 }
2124 }
2125 }
2126
2127 $num = $num % $pow10;
2128 }
2129
2130 $preTransformLength = count( $letters );
2131 if ( $preTransformLength === 1 ) {
2132 // Add geresh (single quote) to one-letter numbers
2133 $letters[] = "'";
2134 } else {
2135 $lastIndex = $preTransformLength - 1;
2136 $letters[$lastIndex] = str_replace(
2137 [ 'כ', 'מ', 'נ', 'פ', 'צ' ],
2138 [ 'ך', 'ם', 'ן', 'ף', 'ץ' ],
2139 $letters[$lastIndex]
2140 );
2141
2142 // Add gershayim (double quote) to multiple-letter numbers,
2143 // but exclude numbers with only one letter after the thousands
2144 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2145 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2146 $letters[] = "'";
2147 } else {
2148 array_splice( $letters, -1, 0, '"' );
2149 }
2150 }
2151
2152 return implode( $letters );
2153 }
2154
2155 /**
2156 * Used by date() and time() to adjust the time output.
2157 *
2158 * @param string $ts The time in date('YmdHis') format
2159 * @param mixed $tz Adjust the time by this amount (default false, mean we
2160 * get user timecorrection setting)
2161 * @return int
2162 */
2163 public function userAdjust( $ts, $tz = false ) {
2164 global $wgUser, $wgLocalTZoffset;
2165
2166 if ( $tz === false ) {
2167 $tz = $wgUser->getOption( 'timecorrection' );
2168 }
2169
2170 $data = explode( '|', $tz, 3 );
2171
2172 if ( $data[0] == 'ZoneInfo' ) {
2173 try {
2174 $userTZ = new DateTimeZone( $data[2] );
2175 $date = new DateTime( $ts, new DateTimeZone( 'UTC' ) );
2176 $date->setTimezone( $userTZ );
2177 return $date->format( 'YmdHis' );
2178 } catch ( Exception $e ) {
2179 // Unrecognized timezone, default to 'Offset' with the stored offset.
2180 $data[0] = 'Offset';
2181 }
2182 }
2183
2184 if ( $data[0] == 'System' || $tz == '' ) {
2185 # Global offset in minutes.
2186 $minDiff = $wgLocalTZoffset;
2187 } elseif ( $data[0] == 'Offset' ) {
2188 $minDiff = intval( $data[1] );
2189 } else {
2190 $data = explode( ':', $tz );
2191 if ( count( $data ) == 2 ) {
2192 $data[0] = intval( $data[0] );
2193 $data[1] = intval( $data[1] );
2194 $minDiff = abs( $data[0] ) * 60 + $data[1];
2195 if ( $data[0] < 0 ) {
2196 $minDiff = -$minDiff;
2197 }
2198 } else {
2199 $minDiff = intval( $data[0] ) * 60;
2200 }
2201 }
2202
2203 # No difference ? Return time unchanged
2204 if ( $minDiff == 0 ) {
2205 return $ts;
2206 }
2207
2208 Wikimedia\suppressWarnings(); // E_STRICT system time bitching
2209 # Generate an adjusted date; take advantage of the fact that mktime
2210 # will normalize out-of-range values so we don't have to split $minDiff
2211 # into hours and minutes.
2212 $t = mktime( (
2213 (int)substr( $ts, 8, 2 ) ), # Hours
2214 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2215 (int)substr( $ts, 12, 2 ), # Seconds
2216 (int)substr( $ts, 4, 2 ), # Month
2217 (int)substr( $ts, 6, 2 ), # Day
2218 (int)substr( $ts, 0, 4 ) ); # Year
2219
2220 $date = date( 'YmdHis', $t );
2221 Wikimedia\restoreWarnings();
2222
2223 return $date;
2224 }
2225
2226 /**
2227 * This is meant to be used by time(), date(), and timeanddate() to get
2228 * the date preference they're supposed to use, it should be used in
2229 * all children.
2230 *
2231 * function timeanddate([...], $format = true) {
2232 * $datePreference = $this->dateFormat($format);
2233 * [...]
2234 * }
2235 *
2236 * @param int|string|bool $usePrefs If true, the user's preference is used
2237 * if false, the site/language default is used
2238 * if int/string, assumed to be a format.
2239 * @return string
2240 */
2241 function dateFormat( $usePrefs = true ) {
2242 global $wgUser;
2243
2244 if ( is_bool( $usePrefs ) ) {
2245 if ( $usePrefs ) {
2246 $datePreference = $wgUser->getDatePreference();
2247 } else {
2248 $datePreference = (string)User::getDefaultOption( 'date' );
2249 }
2250 } else {
2251 $datePreference = (string)$usePrefs;
2252 }
2253
2254 // return int
2255 if ( $datePreference == '' ) {
2256 return 'default';
2257 }
2258
2259 return $datePreference;
2260 }
2261
2262 /**
2263 * Get a format string for a given type and preference
2264 * @param string $type May be 'date', 'time', 'both', or 'pretty'.
2265 * @param string $pref The format name as it appears in Messages*.php under
2266 * $datePreferences.
2267 *
2268 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2269 *
2270 * @return string
2271 */
2272 function getDateFormatString( $type, $pref ) {
2273 $wasDefault = false;
2274 if ( $pref == 'default' ) {
2275 $wasDefault = true;
2276 $pref = $this->getDefaultDateFormat();
2277 }
2278
2279 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2280 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2281
2282 if ( $type === 'pretty' && $df === null ) {
2283 $df = $this->getDateFormatString( 'date', $pref );
2284 }
2285
2286 if ( !$wasDefault && $df === null ) {
2287 $pref = $this->getDefaultDateFormat();
2288 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2289 }
2290
2291 $this->dateFormatStrings[$type][$pref] = $df;
2292 }
2293 return $this->dateFormatStrings[$type][$pref];
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 date( $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( 'date', $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 True to use user's date format preference
2321 * @param string|bool $timecorrection The time offset as returned by
2322 * validateTimeZone() in Special:Preferences
2323 * @return string
2324 */
2325 public function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2326 $ts = wfTimestamp( TS_MW, $ts );
2327 if ( $adj ) {
2328 $ts = $this->userAdjust( $ts, $timecorrection );
2329 }
2330 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2331 return $this->sprintfDate( $df, $ts );
2332 }
2333
2334 /**
2335 * @param string $ts The time format which needs to be turned into a
2336 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2337 * @param bool $adj Whether to adjust the time output according to the
2338 * user configured offset ($timecorrection)
2339 * @param mixed $format What format to return, if it's false output the
2340 * default one (default true)
2341 * @param string|bool $timecorrection The time offset as returned by
2342 * validateTimeZone() in Special:Preferences
2343 * @return string
2344 */
2345 public function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2346 $ts = wfTimestamp( TS_MW, $ts );
2347 if ( $adj ) {
2348 $ts = $this->userAdjust( $ts, $timecorrection );
2349 }
2350 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2351 return $this->sprintfDate( $df, $ts );
2352 }
2353
2354 /**
2355 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2356 *
2357 * @since 1.20
2358 *
2359 * @param int $seconds The amount of seconds.
2360 * @param array $chosenIntervals The intervals to enable.
2361 *
2362 * @return string
2363 */
2364 public function formatDuration( $seconds, array $chosenIntervals = [] ) {
2365 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2366
2367 $segments = [];
2368
2369 foreach ( $intervals as $intervalName => $intervalValue ) {
2370 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2371 // duration-years, duration-decades, duration-centuries, duration-millennia
2372 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2373 $segments[] = $message->inLanguage( $this )->escaped();
2374 }
2375
2376 return $this->listToText( $segments );
2377 }
2378
2379 /**
2380 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2381 * For example 65 will be turned into [ minutes => 1, seconds => 5 ].
2382 *
2383 * @since 1.20
2384 *
2385 * @param int $seconds The amount of seconds.
2386 * @param array $chosenIntervals The intervals to enable.
2387 *
2388 * @return array
2389 */
2390 public function getDurationIntervals( $seconds, array $chosenIntervals = [] ) {
2391 if ( empty( $chosenIntervals ) ) {
2392 $chosenIntervals = [
2393 'millennia',
2394 'centuries',
2395 'decades',
2396 'years',
2397 'days',
2398 'hours',
2399 'minutes',
2400 'seconds'
2401 ];
2402 }
2403
2404 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2405 $sortedNames = array_keys( $intervals );
2406 $smallestInterval = array_pop( $sortedNames );
2407
2408 $segments = [];
2409
2410 foreach ( $intervals as $name => $length ) {
2411 $value = floor( $seconds / $length );
2412
2413 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2414 $seconds -= $value * $length;
2415 $segments[$name] = $value;
2416 }
2417 }
2418
2419 return $segments;
2420 }
2421
2422 /**
2423 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2424 *
2425 * @param string $type Can be 'date', 'time' or 'both'
2426 * @param string $ts The time format which needs to be turned into a
2427 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2428 * @param User $user User object used to get preferences for timezone and format
2429 * @param array $options Array, can contain the following keys:
2430 * - 'timecorrection': time correction, can have the following values:
2431 * - true: use user's preference
2432 * - false: don't use time correction
2433 * - int: value of time correction in minutes
2434 * - 'format': format to use, can have the following values:
2435 * - true: use user's preference
2436 * - false: use default preference
2437 * - string: format to use
2438 * @since 1.19
2439 * @return string
2440 */
2441 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2442 $ts = wfTimestamp( TS_MW, $ts );
2443 $options += [ 'timecorrection' => true, 'format' => true ];
2444 if ( $options['timecorrection'] !== false ) {
2445 if ( $options['timecorrection'] === true ) {
2446 $offset = $user->getOption( 'timecorrection' );
2447 } else {
2448 $offset = $options['timecorrection'];
2449 }
2450 $ts = $this->userAdjust( $ts, $offset );
2451 }
2452 if ( $options['format'] === true ) {
2453 $format = $user->getDatePreference();
2454 } else {
2455 $format = $options['format'];
2456 }
2457 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2458 return $this->sprintfDate( $df, $ts );
2459 }
2460
2461 /**
2462 * Get the formatted date for the given timestamp and formatted for
2463 * the given user.
2464 *
2465 * @param mixed $ts Mixed: the time format which needs to be turned into a
2466 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2467 * @param User $user User object used to get preferences for timezone and format
2468 * @param array $options Array, can contain the following keys:
2469 * - 'timecorrection': time correction, can have the following values:
2470 * - true: use user's preference
2471 * - false: don't use time correction
2472 * - int: value of time correction in minutes
2473 * - 'format': format to use, can have the following values:
2474 * - true: use user's preference
2475 * - false: use default preference
2476 * - string: format to use
2477 * @since 1.19
2478 * @return string
2479 */
2480 public function userDate( $ts, User $user, array $options = [] ) {
2481 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2482 }
2483
2484 /**
2485 * Get the formatted time for the given timestamp and formatted for
2486 * the given user.
2487 *
2488 * @param mixed $ts The time format which needs to be turned into a
2489 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2490 * @param User $user User object used to get preferences for timezone and format
2491 * @param array $options Array, can contain the following keys:
2492 * - 'timecorrection': time correction, can have the following values:
2493 * - true: use user's preference
2494 * - false: don't use time correction
2495 * - int: value of time correction in minutes
2496 * - 'format': format to use, can have the following values:
2497 * - true: use user's preference
2498 * - false: use default preference
2499 * - string: format to use
2500 * @since 1.19
2501 * @return string
2502 */
2503 public function userTime( $ts, User $user, array $options = [] ) {
2504 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2505 }
2506
2507 /**
2508 * Get the formatted date and time for the given timestamp and formatted for
2509 * the given user.
2510 *
2511 * @param mixed $ts The time format which needs to be turned into a
2512 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2513 * @param User $user User object used to get preferences for timezone and format
2514 * @param array $options Array, can contain the following keys:
2515 * - 'timecorrection': time correction, can have the following values:
2516 * - true: use user's preference
2517 * - false: don't use time correction
2518 * - int: value of time correction in minutes
2519 * - 'format': format to use, can have the following values:
2520 * - true: use user's preference
2521 * - false: use default preference
2522 * - string: format to use
2523 * @since 1.19
2524 * @return string
2525 */
2526 public function userTimeAndDate( $ts, User $user, array $options = [] ) {
2527 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2528 }
2529
2530 /**
2531 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2532 *
2533 * Determine the difference between the timestamp and the current time, and
2534 * generate a readable timestamp by returning "<N> <units> ago", where the
2535 * largest possible unit is used.
2536 *
2537 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2538 *
2539 * @param MWTimestamp $time
2540 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2541 * @param User|null $user User the timestamp is being generated for
2542 * (or null to use main context's user)
2543 * @return string Formatted timestamp
2544 */
2545 public function getHumanTimestamp(
2546 MWTimestamp $time, MWTimestamp $relativeTo = null, User $user = null
2547 ) {
2548 if ( $relativeTo === null ) {
2549 $relativeTo = new MWTimestamp();
2550 }
2551 if ( $user === null ) {
2552 $user = RequestContext::getMain()->getUser();
2553 }
2554
2555 // Adjust for the user's timezone.
2556 $offsetThis = $time->offsetForUser( $user );
2557 $offsetRel = $relativeTo->offsetForUser( $user );
2558
2559 $ts = '';
2560 if ( Hooks::run( 'GetHumanTimestamp', [ &$ts, $time, $relativeTo, $user, $this ] ) ) {
2561 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2562 }
2563
2564 // Reset the timezone on the objects.
2565 $time->timestamp->sub( $offsetThis );
2566 $relativeTo->timestamp->sub( $offsetRel );
2567
2568 return $ts;
2569 }
2570
2571 /**
2572 * Convert an MWTimestamp into a pretty human-readable timestamp using
2573 * the given user preferences and relative base time.
2574 *
2575 * @see Language::getHumanTimestamp
2576 * @param MWTimestamp $ts Timestamp to prettify
2577 * @param MWTimestamp $relativeTo Base timestamp
2578 * @param User $user User preferences to use
2579 * @return string Human timestamp
2580 * @since 1.26
2581 */
2582 private function getHumanTimestampInternal(
2583 MWTimestamp $ts, MWTimestamp $relativeTo, User $user
2584 ) {
2585 $diff = $ts->diff( $relativeTo );
2586 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2587 (int)$relativeTo->timestamp->format( 'w' ) );
2588 $days = $diff->days ?: (int)$diffDay;
2589 if ( $diff->invert || $days > 5
2590 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2591 ) {
2592 // Timestamps are in different years: use full timestamp
2593 // Also do full timestamp for future dates
2594 /**
2595 * @todo FIXME: Add better handling of future timestamps.
2596 */
2597 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2598 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2599 } elseif ( $days > 5 ) {
2600 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2601 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2602 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2603 } elseif ( $days > 1 ) {
2604 // Timestamp within the past week: show the day of the week and time
2605 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2606 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2607 // Messages:
2608 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2609 $ts = wfMessage( "$weekday-at" )
2610 ->inLanguage( $this )
2611 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2612 ->text();
2613 } elseif ( $days == 1 ) {
2614 // Timestamp was yesterday: say 'yesterday' and the time.
2615 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2616 $ts = wfMessage( 'yesterday-at' )
2617 ->inLanguage( $this )
2618 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2619 ->text();
2620 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2621 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2622 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2623 $ts = wfMessage( 'today-at' )
2624 ->inLanguage( $this )
2625 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2626 ->text();
2627
2628 // From here on in, the timestamp was soon enough ago so that we can simply say
2629 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2630 } elseif ( $diff->h == 1 ) {
2631 // Less than 90 minutes, but more than an hour ago.
2632 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2633 } elseif ( $diff->i >= 1 ) {
2634 // A few minutes ago.
2635 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2636 } elseif ( $diff->s >= 30 ) {
2637 // Less than a minute, but more than 30 sec ago.
2638 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2639 } else {
2640 // Less than 30 seconds ago.
2641 $ts = wfMessage( 'just-now' )->text();
2642 }
2643
2644 return $ts;
2645 }
2646
2647 /**
2648 * @param string $key
2649 * @return string|null
2650 */
2651 public function getMessage( $key ) {
2652 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2653 }
2654
2655 /**
2656 * @return array
2657 */
2658 function getAllMessages() {
2659 return self::$dataCache->getItem( $this->mCode, 'messages' );
2660 }
2661
2662 /**
2663 * @param string $in
2664 * @param string $out
2665 * @param string $string
2666 * @return string
2667 */
2668 public function iconv( $in, $out, $string ) {
2669 # Even with //IGNORE iconv can whine about illegal characters in
2670 # *input* string. We just ignore those too.
2671 # REF: https://bugs.php.net/bug.php?id=37166
2672 # REF: https://phabricator.wikimedia.org/T18885
2673 Wikimedia\suppressWarnings();
2674 $text = iconv( $in, $out . '//IGNORE', $string );
2675 Wikimedia\restoreWarnings();
2676 return $text;
2677 }
2678
2679 // callback functions for ucwords(), ucwordbreaks()
2680
2681 /**
2682 * @param array $matches
2683 * @return mixed|string
2684 */
2685 function ucwordbreaksCallbackAscii( $matches ) {
2686 return $this->ucfirst( $matches[1] );
2687 }
2688
2689 /**
2690 * @param array $matches
2691 * @return string
2692 */
2693 function ucwordbreaksCallbackMB( $matches ) {
2694 return mb_strtoupper( $matches[0] );
2695 }
2696
2697 /**
2698 * @param array $matches
2699 * @return string
2700 */
2701 function ucwordsCallbackMB( $matches ) {
2702 return mb_strtoupper( $matches[0] );
2703 }
2704
2705 /**
2706 * Make a string's first character uppercase
2707 *
2708 * @param string $str
2709 *
2710 * @return string
2711 */
2712 public function ucfirst( $str ) {
2713 $o = ord( $str );
2714 if ( $o < 96 ) { // if already uppercase...
2715 return $str;
2716 } elseif ( $o < 128 ) {
2717 return ucfirst( $str ); // use PHP's ucfirst()
2718 } else {
2719 // fall back to more complex logic in case of multibyte strings
2720 return $this->uc( $str, true );
2721 }
2722 }
2723
2724 /**
2725 * Convert a string to uppercase
2726 *
2727 * @param string $str
2728 * @param bool $first
2729 *
2730 * @return string
2731 */
2732 public function uc( $str, $first = false ) {
2733 if ( $first ) {
2734 if ( $this->isMultibyte( $str ) ) {
2735 return $this->mbUpperChar( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2736 } else {
2737 return ucfirst( $str );
2738 }
2739 } else {
2740 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2741 }
2742 }
2743
2744 /**
2745 * Convert character to uppercase, allowing overrides of the default mb_upper
2746 * behaviour, which is buggy in many ways. Having a conversion table can be
2747 * useful during transitions between PHP versions where unicode changes happen.
2748 * This can make some resources unreachable on-wiki, see discussion at T219279.
2749 * Providing such a conversion table can allow to manage the transition period.
2750 *
2751 * @since 1.34
2752 *
2753 * @param string $char
2754 *
2755 * @return string
2756 */
2757 protected function mbUpperChar( $char ) {
2758 global $wgOverrideUcfirstCharacters;
2759 if ( array_key_exists( $char, $wgOverrideUcfirstCharacters ) ) {
2760 return $wgOverrideUcfirstCharacters[$char];
2761 } else {
2762 return mb_strtoupper( $char );
2763 }
2764 }
2765
2766 /**
2767 * @param string $str
2768 * @return mixed|string
2769 */
2770 function lcfirst( $str ) {
2771 $o = ord( $str );
2772 if ( !$o ) {
2773 return strval( $str );
2774 } elseif ( $o >= 128 ) {
2775 return $this->lc( $str, true );
2776 } elseif ( $o > 96 ) {
2777 return $str;
2778 } else {
2779 $str[0] = strtolower( $str[0] );
2780 return $str;
2781 }
2782 }
2783
2784 /**
2785 * @param string $str
2786 * @param bool $first
2787 * @return mixed|string
2788 */
2789 function lc( $str, $first = false ) {
2790 if ( $first ) {
2791 if ( $this->isMultibyte( $str ) ) {
2792 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2793 } else {
2794 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2795 }
2796 } else {
2797 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2798 }
2799 }
2800
2801 /**
2802 * @param string $str
2803 * @return bool
2804 */
2805 function isMultibyte( $str ) {
2806 return strlen( $str ) !== mb_strlen( $str );
2807 }
2808
2809 /**
2810 * @param string $str
2811 * @return mixed|string
2812 */
2813 function ucwords( $str ) {
2814 if ( $this->isMultibyte( $str ) ) {
2815 $str = $this->lc( $str );
2816
2817 // regexp to find first letter in each word (i.e. after each space)
2818 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2819
2820 // function to use to capitalize a single char
2821 return preg_replace_callback(
2822 $replaceRegexp,
2823 [ $this, 'ucwordsCallbackMB' ],
2824 $str
2825 );
2826 } else {
2827 return ucwords( strtolower( $str ) );
2828 }
2829 }
2830
2831 /**
2832 * capitalize words at word breaks
2833 *
2834 * @param string $str
2835 * @return mixed
2836 */
2837 function ucwordbreaks( $str ) {
2838 if ( $this->isMultibyte( $str ) ) {
2839 $str = $this->lc( $str );
2840
2841 // since \b doesn't work for UTF-8, we explicitely define word break chars
2842 $breaks = "[ \-\(\)\}\{\.,\?!]";
2843
2844 // find first letter after word break
2845 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2846 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2847
2848 return preg_replace_callback(
2849 $replaceRegexp,
2850 [ $this, 'ucwordbreaksCallbackMB' ],
2851 $str
2852 );
2853 } else {
2854 return preg_replace_callback(
2855 '/\b([\w\x80-\xff]+)\b/',
2856 [ $this, 'ucwordbreaksCallbackAscii' ],
2857 $str
2858 );
2859 }
2860 }
2861
2862 /**
2863 * Return a case-folded representation of $s
2864 *
2865 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2866 * and $s2 are the same except for the case of their characters. It is not
2867 * necessary for the value returned to make sense when displayed.
2868 *
2869 * Do *not* perform any other normalisation in this function. If a caller
2870 * uses this function when it should be using a more general normalisation
2871 * function, then fix the caller.
2872 *
2873 * @param string $s
2874 *
2875 * @return string
2876 */
2877 function caseFold( $s ) {
2878 return $this->uc( $s );
2879 }
2880
2881 /**
2882 * TODO: $s is not always a string per T218883
2883 * @param string $s
2884 * @return string
2885 */
2886 function checkTitleEncoding( $s ) {
2887 if ( is_array( $s ) ) {
2888 throw new MWException( 'Given array to checkTitleEncoding.' );
2889 }
2890 if ( StringUtils::isUtf8( $s ) ) {
2891 return $s;
2892 }
2893
2894 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2895 }
2896
2897 /**
2898 * @return string
2899 */
2900 function fallback8bitEncoding() {
2901 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2902 }
2903
2904 /**
2905 * Most writing systems use whitespace to break up words.
2906 * Some languages such as Chinese don't conventionally do this,
2907 * which requires special handling when breaking up words for
2908 * searching etc.
2909 *
2910 * @return bool
2911 */
2912 function hasWordBreaks() {
2913 return true;
2914 }
2915
2916 /**
2917 * Some languages such as Chinese require word segmentation,
2918 * Specify such segmentation when overridden in derived class.
2919 *
2920 * @param string $string
2921 * @return string
2922 */
2923 function segmentByWord( $string ) {
2924 return $string;
2925 }
2926
2927 /**
2928 * Some languages have special punctuation need to be normalized.
2929 * Make such changes here.
2930 *
2931 * @param string $string
2932 * @return string
2933 */
2934 function normalizeForSearch( $string ) {
2935 return self::convertDoubleWidth( $string );
2936 }
2937
2938 /**
2939 * convert double-width roman characters to single-width.
2940 * range: ff00-ff5f ~= 0020-007f
2941 *
2942 * @param string $string
2943 *
2944 * @return string
2945 */
2946 protected static function convertDoubleWidth( $string ) {
2947 static $full = null;
2948 static $half = null;
2949
2950 if ( $full === null ) {
2951 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2952 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2953 $full = str_split( $fullWidth, 3 );
2954 $half = str_split( $halfWidth );
2955 }
2956
2957 $string = str_replace( $full, $half, $string );
2958 return $string;
2959 }
2960
2961 /**
2962 * @param string $string
2963 * @param string $pattern
2964 * @return string
2965 */
2966 protected static function insertSpace( $string, $pattern ) {
2967 $string = preg_replace( $pattern, " $1 ", $string );
2968 $string = preg_replace( '/ +/', ' ', $string );
2969 return $string;
2970 }
2971
2972 /**
2973 * @param string[] $termsArray
2974 * @return string[]
2975 */
2976 function convertForSearchResult( $termsArray ) {
2977 # some languages, e.g. Chinese, need to do a conversion
2978 # in order for search results to be displayed correctly
2979 return $termsArray;
2980 }
2981
2982 /**
2983 * Get the first character of a string.
2984 *
2985 * @param string $s
2986 * @return string
2987 */
2988 function firstChar( $s ) {
2989 $matches = [];
2990 preg_match(
2991 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2992 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2993 $s,
2994 $matches
2995 );
2996
2997 if ( isset( $matches[1] ) ) {
2998 if ( strlen( $matches[1] ) != 3 ) {
2999 return $matches[1];
3000 }
3001
3002 // Break down Hangul syllables to grab the first jamo
3003 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
3004 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
3005 return $matches[1];
3006 } elseif ( $code < 0xb098 ) {
3007 return "\u{3131}";
3008 } elseif ( $code < 0xb2e4 ) {
3009 return "\u{3134}";
3010 } elseif ( $code < 0xb77c ) {
3011 return "\u{3137}";
3012 } elseif ( $code < 0xb9c8 ) {
3013 return "\u{3139}";
3014 } elseif ( $code < 0xbc14 ) {
3015 return "\u{3141}";
3016 } elseif ( $code < 0xc0ac ) {
3017 return "\u{3142}";
3018 } elseif ( $code < 0xc544 ) {
3019 return "\u{3145}";
3020 } elseif ( $code < 0xc790 ) {
3021 return "\u{3147}";
3022 } elseif ( $code < 0xcc28 ) {
3023 return "\u{3148}";
3024 } elseif ( $code < 0xce74 ) {
3025 return "\u{314A}";
3026 } elseif ( $code < 0xd0c0 ) {
3027 return "\u{314B}";
3028 } elseif ( $code < 0xd30c ) {
3029 return "\u{314C}";
3030 } elseif ( $code < 0xd558 ) {
3031 return "\u{314D}";
3032 } else {
3033 return "\u{314E}";
3034 }
3035 } else {
3036 return '';
3037 }
3038 }
3039
3040 /**
3041 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3042 * also cleans up certain backwards-compatible sequences, converting them
3043 * to the modern Unicode equivalent.
3044 *
3045 * This is language-specific for performance reasons only.
3046 *
3047 * @param string $s
3048 *
3049 * @return string
3050 */
3051 public function normalize( $s ) {
3052 global $wgAllUnicodeFixes;
3053 $s = UtfNormal\Validator::cleanUp( $s );
3054 if ( $wgAllUnicodeFixes ) {
3055 $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
3056 $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
3057 }
3058
3059 return $s;
3060 }
3061
3062 /**
3063 * Transform a string using serialized data stored in the given file (which
3064 * must be in the serialized subdirectory of $IP). The file contains pairs
3065 * mapping source characters to destination characters.
3066 *
3067 * The data is cached in process memory. This will go faster if you have the
3068 * FastStringSearch extension.
3069 *
3070 * @param string $file
3071 * @param string $string
3072 *
3073 * @throws MWException
3074 * @return string
3075 */
3076 protected function transformUsingPairFile( $file, $string ) {
3077 if ( !isset( $this->transformData[$file] ) ) {
3078 global $IP;
3079 $data = require "$IP/languages/data/{$file}";
3080 $this->transformData[$file] = new ReplacementArray( $data );
3081 }
3082 return $this->transformData[$file]->replace( $string );
3083 }
3084
3085 /**
3086 * For right-to-left language support
3087 *
3088 * @return bool
3089 */
3090 function isRTL() {
3091 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3092 }
3093
3094 /**
3095 * Return the correct HTML 'dir' attribute value for this language.
3096 * @return string
3097 */
3098 function getDir() {
3099 return $this->isRTL() ? 'rtl' : 'ltr';
3100 }
3101
3102 /**
3103 * Return 'left' or 'right' as appropriate alignment for line-start
3104 * for this language's text direction.
3105 *
3106 * Should be equivalent to CSS3 'start' text-align value....
3107 *
3108 * @return string
3109 */
3110 function alignStart() {
3111 return $this->isRTL() ? 'right' : 'left';
3112 }
3113
3114 /**
3115 * Return 'right' or 'left' as appropriate alignment for line-end
3116 * for this language's text direction.
3117 *
3118 * Should be equivalent to CSS3 'end' text-align value....
3119 *
3120 * @return string
3121 */
3122 function alignEnd() {
3123 return $this->isRTL() ? 'left' : 'right';
3124 }
3125
3126 /**
3127 * A hidden direction mark (LRM or RLM), depending on the language direction.
3128 * Unlike getDirMark(), this function returns the character as an HTML entity.
3129 * This function should be used when the output is guaranteed to be HTML,
3130 * because it makes the output HTML source code more readable. When
3131 * the output is plain text or can be escaped, getDirMark() should be used.
3132 *
3133 * @param bool $opposite Get the direction mark opposite to your language
3134 * @return string
3135 * @since 1.20
3136 */
3137 function getDirMarkEntity( $opposite = false ) {
3138 if ( $opposite ) {
3139 return $this->isRTL() ? '&lrm;' : '&rlm;';
3140 }
3141 return $this->isRTL() ? '&rlm;' : '&lrm;';
3142 }
3143
3144 /**
3145 * A hidden direction mark (LRM or RLM), depending on the language direction.
3146 * This function produces them as invisible Unicode characters and
3147 * the output may be hard to read and debug, so it should only be used
3148 * when the output is plain text or can be escaped. When the output is
3149 * HTML, use getDirMarkEntity() instead.
3150 *
3151 * @param bool $opposite Get the direction mark opposite to your language
3152 * @return string
3153 */
3154 function getDirMark( $opposite = false ) {
3155 $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3156 $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3157 if ( $opposite ) {
3158 return $this->isRTL() ? $lrm : $rlm;
3159 }
3160 return $this->isRTL() ? $rlm : $lrm;
3161 }
3162
3163 /**
3164 * @return array
3165 */
3166 function capitalizeAllNouns() {
3167 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3168 }
3169
3170 /**
3171 * An arrow, depending on the language direction.
3172 *
3173 * @param string $direction The direction of the arrow: forwards (default),
3174 * backwards, left, right, up, down.
3175 * @return string
3176 */
3177 function getArrow( $direction = 'forwards' ) {
3178 switch ( $direction ) {
3179 case 'forwards':
3180 return $this->isRTL() ? '←' : '→';
3181 case 'backwards':
3182 return $this->isRTL() ? '→' : '←';
3183 case 'left':
3184 return '←';
3185 case 'right':
3186 return '→';
3187 case 'up':
3188 return '↑';
3189 case 'down':
3190 return '↓';
3191 }
3192 }
3193
3194 /**
3195 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3196 *
3197 * @return bool
3198 */
3199 function linkPrefixExtension() {
3200 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3201 }
3202
3203 /**
3204 * Get all magic words from cache.
3205 * @return array
3206 */
3207 function getMagicWords() {
3208 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3209 }
3210
3211 /**
3212 * Fill a MagicWord object with data from here
3213 *
3214 * @param MagicWord $mw
3215 */
3216 function getMagic( $mw ) {
3217 $rawEntry = $this->mMagicExtensions[$mw->mId] ??
3218 self::$dataCache->getSubitem( $this->mCode, 'magicWords', $mw->mId );
3219
3220 if ( !is_array( $rawEntry ) ) {
3221 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3222 } else {
3223 $mw->mCaseSensitive = $rawEntry[0];
3224 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3225 }
3226 }
3227
3228 /**
3229 * Add magic words to the extension array
3230 *
3231 * @param array $newWords
3232 */
3233 function addMagicWordsByLang( $newWords ) {
3234 $fallbackChain = $this->getFallbackLanguages();
3235 $fallbackChain = array_reverse( $fallbackChain );
3236 foreach ( $fallbackChain as $code ) {
3237 if ( isset( $newWords[$code] ) ) {
3238 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3239 }
3240 }
3241 }
3242
3243 /**
3244 * Get special page names, as an associative array
3245 * canonical name => array of valid names, including aliases
3246 * @return array
3247 */
3248 function getSpecialPageAliases() {
3249 // Cache aliases because it may be slow to load them
3250 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3251 // Initialise array
3252 $this->mExtendedSpecialPageAliases =
3253 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3254 }
3255
3256 return $this->mExtendedSpecialPageAliases;
3257 }
3258
3259 /**
3260 * Italic is unsuitable for some languages
3261 *
3262 * @param string $text The text to be emphasized.
3263 * @return string
3264 */
3265 function emphasize( $text ) {
3266 return "<em>$text</em>";
3267 }
3268
3269 /**
3270 * Normally we output all numbers in plain en_US style, that is
3271 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3272 * point twohundredthirtyfive. However this is not suitable for all
3273 * languages, some such as Bengali (bn) want ২,৯৩,২৯১.২৩৫ and others such as
3274 * Icelandic just want to use commas instead of dots, and dots instead
3275 * of commas like "293.291,235".
3276 *
3277 * An example of this function being called:
3278 * <code>
3279 * wfMessage( 'message' )->numParams( $num )->text()
3280 * </code>
3281 *
3282 * See $separatorTransformTable on MessageIs.php for
3283 * the , => . and . => , implementation.
3284 *
3285 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3286 * @param int|float $number The string to be formatted, should be an integer
3287 * or a floating point number.
3288 * @param bool $nocommafy Set to true for special numbers like dates
3289 * @return string
3290 */
3291 public function formatNum( $number, $nocommafy = false ) {
3292 global $wgTranslateNumerals;
3293 if ( !$nocommafy ) {
3294 $number = $this->commafy( $number );
3295 $s = $this->separatorTransformTable();
3296 if ( $s ) {
3297 $number = strtr( $number, $s );
3298 }
3299 }
3300
3301 if ( $wgTranslateNumerals ) {
3302 $s = $this->digitTransformTable();
3303 if ( $s ) {
3304 $number = strtr( $number, $s );
3305 }
3306 }
3307
3308 return (string)$number;
3309 }
3310
3311 /**
3312 * Front-end for non-commafied formatNum
3313 *
3314 * @param int|float $number The string to be formatted, should be an integer
3315 * or a floating point number.
3316 * @since 1.21
3317 * @return string
3318 */
3319 public function formatNumNoSeparators( $number ) {
3320 return $this->formatNum( $number, true );
3321 }
3322
3323 /**
3324 * @param string $number
3325 * @return string
3326 */
3327 public function parseFormattedNumber( $number ) {
3328 $s = $this->digitTransformTable();
3329 if ( $s ) {
3330 // eliminate empty array values such as ''. (T66347)
3331 $s = array_filter( $s );
3332 $number = strtr( $number, array_flip( $s ) );
3333 }
3334
3335 $s = $this->separatorTransformTable();
3336 if ( $s ) {
3337 // eliminate empty array values such as ''. (T66347)
3338 $s = array_filter( $s );
3339 $number = strtr( $number, array_flip( $s ) );
3340 }
3341
3342 $number = strtr( $number, [ ',' => '' ] );
3343 return $number;
3344 }
3345
3346 /**
3347 * Adds commas to a given number
3348 * @since 1.19
3349 * @param mixed $number
3350 * @return string
3351 */
3352 function commafy( $number ) {
3353 $digitGroupingPattern = $this->digitGroupingPattern();
3354 $minimumGroupingDigits = $this->minimumGroupingDigits();
3355 if ( $number === null ) {
3356 return '';
3357 }
3358
3359 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3360 // Default grouping is at thousands, use the same for ###,###,### pattern too.
3361 // In some languages it's conventional not to insert a thousands separator
3362 // in numbers that are four digits long (1000-9999).
3363 if ( $minimumGroupingDigits ) {
3364 // Number of '#' characters after last comma in the grouping pattern.
3365 // The pattern is hardcoded here, but this would vary for different patterns.
3366 $primaryGroupingSize = 3;
3367 // Maximum length of a number to suppress digit grouping for.
3368 $maximumLength = $minimumGroupingDigits + $primaryGroupingSize - 1;
3369 if ( preg_match( '/^\-?\d{1,' . $maximumLength . '}(\.\d+)?$/', $number ) ) {
3370 return $number;
3371 }
3372 }
3373 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3374 } else {
3375 // Ref: http://cldr.unicode.org/translation/number-patterns
3376 $sign = "";
3377 if ( intval( $number ) < 0 ) {
3378 // For negative numbers apply the algorithm like positive number and add sign.
3379 $sign = "-";
3380 $number = substr( $number, 1 );
3381 }
3382 $integerPart = [];
3383 $decimalPart = [];
3384 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3385 preg_match( "/\d+/", $number, $integerPart );
3386 preg_match( "/\.\d*/", $number, $decimalPart );
3387 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3388 if ( $groupedNumber === $number ) {
3389 // the string does not have any number part. Eg: .12345
3390 return $sign . $groupedNumber;
3391 }
3392 $start = $end = ( $integerPart ) ? strlen( $integerPart[0] ) : 0;
3393 while ( $start > 0 ) {
3394 $match = $matches[0][$numMatches - 1];
3395 $matchLen = strlen( $match );
3396 $start = $end - $matchLen;
3397 if ( $start < 0 ) {
3398 $start = 0;
3399 }
3400 $groupedNumber = substr( $number, $start, $end - $start ) . $groupedNumber;
3401 $end = $start;
3402 if ( $numMatches > 1 ) {
3403 // use the last pattern for the rest of the number
3404 $numMatches--;
3405 }
3406 if ( $start > 0 ) {
3407 $groupedNumber = "," . $groupedNumber;
3408 }
3409 }
3410 return $sign . $groupedNumber;
3411 }
3412 }
3413
3414 /**
3415 * @return string
3416 */
3417 function digitGroupingPattern() {
3418 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3419 }
3420
3421 /**
3422 * @return array
3423 */
3424 function digitTransformTable() {
3425 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3426 }
3427
3428 /**
3429 * @return array
3430 */
3431 function separatorTransformTable() {
3432 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3433 }
3434
3435 /**
3436 * @return int|null
3437 */
3438 function minimumGroupingDigits() {
3439 return self::$dataCache->getItem( $this->mCode, 'minimumGroupingDigits' );
3440 }
3441
3442 /**
3443 * Take a list of strings and build a locale-friendly comma-separated
3444 * list, using the local comma-separator message.
3445 * The last two strings are chained with an "and".
3446 *
3447 * @param string[] $list
3448 * @return string
3449 */
3450 public function listToText( array $list ) {
3451 $itemCount = count( $list );
3452 if ( $itemCount < 1 ) {
3453 return '';
3454 }
3455 $text = array_pop( $list );
3456 if ( $itemCount > 1 ) {
3457 $and = $this->msg( 'and' )->escaped();
3458 $space = $this->msg( 'word-separator' )->escaped();
3459 $comma = '';
3460 if ( $itemCount > 2 ) {
3461 $comma = $this->msg( 'comma-separator' )->escaped();
3462 }
3463 $text = implode( $comma, $list ) . $and . $space . $text;
3464 }
3465 return $text;
3466 }
3467
3468 /**
3469 * Take a list of strings and build a locale-friendly comma-separated
3470 * list, using the local comma-separator message.
3471 * @param string[] $list Array of strings to put in a comma list
3472 * @return string
3473 */
3474 function commaList( array $list ) {
3475 return implode(
3476 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3477 $list
3478 );
3479 }
3480
3481 /**
3482 * Take a list of strings and build a locale-friendly semicolon-separated
3483 * list, using the local semicolon-separator message.
3484 * @param string[] $list Array of strings to put in a semicolon list
3485 * @return string
3486 */
3487 function semicolonList( array $list ) {
3488 return implode(
3489 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3490 $list
3491 );
3492 }
3493
3494 /**
3495 * Same as commaList, but separate it with the pipe instead.
3496 * @param string[] $list Array of strings to put in a pipe list
3497 * @return string
3498 */
3499 function pipeList( array $list ) {
3500 return implode(
3501 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3502 $list
3503 );
3504 }
3505
3506 /**
3507 * Truncate a string to a specified length in bytes, appending an optional
3508 * string (e.g. for ellipsis)
3509 *
3510 * If $length is negative, the string will be truncated from the beginning
3511 *
3512 * @since 1.31
3513 *
3514 * @param string $string String to truncate
3515 * @param int $length Maximum length in bytes
3516 * @param string $ellipsis String to append to the end of truncated text
3517 * @param bool $adjustLength Subtract length of ellipsis from $length
3518 *
3519 * @return string
3520 */
3521 function truncateForDatabase( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3522 return $this->truncateInternal(
3523 $string, $length, $ellipsis, $adjustLength, 'strlen', 'substr'
3524 );
3525 }
3526
3527 /**
3528 * Truncate a string to a specified number of characters, appending an optional
3529 * string (e.g. for ellipsis).
3530 *
3531 * This provides multibyte version of truncateForDatabase() method of this class,
3532 * suitable for truncation based on number of characters, instead of number of bytes.
3533 *
3534 * If $length is negative, the string will be truncated from the beginning.
3535 *
3536 * @since 1.31
3537 *
3538 * @param string $string String to truncate
3539 * @param int $length Maximum number of characters
3540 * @param string $ellipsis String to append to the end of truncated text
3541 * @param bool $adjustLength Subtract length of ellipsis from $length
3542 *
3543 * @return string
3544 */
3545 function truncateForVisual( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3546 // Passing encoding to mb_strlen and mb_substr is optional.
3547 // Encoding defaults to mb_internal_encoding(), which is set to UTF-8 in Setup.php, so
3548 // explicit specification of encoding is skipped.
3549 // Note: Both multibyte methods are callables invoked in truncateInternal.
3550 return $this->truncateInternal(
3551 $string, $length, $ellipsis, $adjustLength, 'mb_strlen', 'mb_substr'
3552 );
3553 }
3554
3555 /**
3556 * Internal method used for truncation. This method abstracts text truncation into
3557 * one common method, allowing users to provide length measurement function and
3558 * function for finding substring.
3559 *
3560 * For usages, see truncateForDatabase and truncateForVisual.
3561 *
3562 * @param string $string String to truncate
3563 * @param int $length Maximum length of final text
3564 * @param string $ellipsis String to append to the end of truncated text
3565 * @param bool $adjustLength Subtract length of ellipsis from $length
3566 * @param callable $measureLength Callable function used for determining the length of text
3567 * @param callable $getSubstring Callable function used for getting the substrings
3568 *
3569 * @return string
3570 */
3571 private function truncateInternal(
3572 $string, $length, $ellipsis, $adjustLength, callable $measureLength, callable $getSubstring
3573 ) {
3574 # Check if there is no need to truncate
3575 if ( $measureLength( $string ) <= abs( $length ) ) {
3576 return $string; // no need to truncate
3577 }
3578
3579 # Use the localized ellipsis character
3580 if ( $ellipsis == '...' ) {
3581 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3582 }
3583 if ( $length == 0 ) {
3584 return $ellipsis; // convention
3585 }
3586
3587 $stringOriginal = $string;
3588 # If ellipsis length is >= $length then we can't apply $adjustLength
3589 if ( $adjustLength && $measureLength( $ellipsis ) >= abs( $length ) ) {
3590 $string = $ellipsis; // this can be slightly unexpected
3591 # Otherwise, truncate and add ellipsis...
3592 } else {
3593 $ellipsisLength = $adjustLength ? $measureLength( $ellipsis ) : 0;
3594 if ( $length > 0 ) {
3595 $length -= $ellipsisLength;
3596 $string = $getSubstring( $string, 0, $length ); // xyz...
3597 $string = $this->removeBadCharLast( $string );
3598 $string = rtrim( $string ) . $ellipsis;
3599 } else {
3600 $length += $ellipsisLength;
3601 $string = $getSubstring( $string, $length ); // ...xyz
3602 $string = $this->removeBadCharFirst( $string );
3603 $string = $ellipsis . ltrim( $string );
3604 }
3605 }
3606
3607 # Do not truncate if the ellipsis makes the string longer/equal (T24181).
3608 # This check is *not* redundant if $adjustLength, due to the single case where
3609 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3610 if ( $measureLength( $string ) < $measureLength( $stringOriginal ) ) {
3611 return $string;
3612 } else {
3613 return $stringOriginal;
3614 }
3615 }
3616
3617 /**
3618 * Remove bytes that represent an incomplete Unicode character
3619 * at the end of string (e.g. bytes of the char are missing)
3620 *
3621 * @param string $string
3622 * @return string
3623 */
3624 protected function removeBadCharLast( $string ) {
3625 if ( $string != '' ) {
3626 $char = ord( $string[strlen( $string ) - 1] );
3627 $m = [];
3628 if ( $char >= 0xc0 ) {
3629 # We got the first byte only of a multibyte char; remove it.
3630 $string = substr( $string, 0, -1 );
3631 } elseif ( $char >= 0x80 &&
3632 // Use the /s modifier (PCRE_DOTALL) so (.*) also matches newlines
3633 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3634 '[\xf0-\xf7][\x80-\xbf]{1,2})$/s', $string, $m )
3635 ) {
3636 # We chopped in the middle of a character; remove it
3637 $string = $m[1];
3638 }
3639 }
3640 return $string;
3641 }
3642
3643 /**
3644 * Remove bytes that represent an incomplete Unicode character
3645 * at the start of string (e.g. bytes of the char are missing)
3646 *
3647 * @param string $string
3648 * @return string
3649 */
3650 protected function removeBadCharFirst( $string ) {
3651 if ( $string != '' ) {
3652 $char = ord( $string[0] );
3653 if ( $char >= 0x80 && $char < 0xc0 ) {
3654 # We chopped in the middle of a character; remove the whole thing
3655 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3656 }
3657 }
3658 return $string;
3659 }
3660
3661 /**
3662 * Truncate a string of valid HTML to a specified length in bytes,
3663 * appending an optional string (e.g. for ellipses), and return valid HTML
3664 *
3665 * This is only intended for styled/linked text, such as HTML with
3666 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3667 * Also, this will not detect things like "display:none" CSS.
3668 *
3669 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3670 *
3671 * @param string $text HTML string to truncate
3672 * @param int $length (zero/positive) Maximum length (including ellipses)
3673 * @param string $ellipsis String to append to the truncated text
3674 * @return string
3675 */
3676 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3677 # Use the localized ellipsis character
3678 if ( $ellipsis == '...' ) {
3679 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3680 }
3681 # Check if there is clearly no need to truncate
3682 if ( $length <= 0 ) {
3683 return $ellipsis; // no text shown, nothing to format (convention)
3684 } elseif ( strlen( $text ) <= $length ) {
3685 return $text; // string short enough even *with* HTML (short-circuit)
3686 }
3687
3688 $dispLen = 0; // innerHTML legth so far
3689 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3690 $tagType = 0; // 0-open, 1-close
3691 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3692 $entityState = 0; // 0-not entity, 1-entity
3693 $tag = $ret = ''; // accumulated tag name, accumulated result string
3694 $openTags = []; // open tag stack
3695 $maybeState = null; // possible truncation state
3696
3697 $textLen = strlen( $text );
3698 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3699 for ( $pos = 0; true; ++$pos ) {
3700 # Consider truncation once the display length has reached the maximim.
3701 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3702 # Check that we're not in the middle of a bracket/entity...
3703 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3704 if ( !$testingEllipsis ) {
3705 $testingEllipsis = true;
3706 # Save where we are; we will truncate here unless there turn out to
3707 # be so few remaining characters that truncation is not necessary.
3708 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3709 $maybeState = [ $ret, $openTags ]; // save state
3710 }
3711 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3712 # String in fact does need truncation, the truncation point was OK.
3713 // @phan-suppress-next-line PhanTypeInvalidExpressionArrayDestructuring
3714 list( $ret, $openTags ) = $maybeState; // reload state
3715 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3716 $ret .= $ellipsis; // add ellipsis
3717 break;
3718 }
3719 }
3720 if ( $pos >= $textLen ) {
3721 break; // extra iteration just for above checks
3722 }
3723
3724 # Read the next char...
3725 $ch = $text[$pos];
3726 $lastCh = $pos ? $text[$pos - 1] : '';
3727 $ret .= $ch; // add to result string
3728 if ( $ch == '<' ) {
3729 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3730 $entityState = 0; // for bad HTML
3731 $bracketState = 1; // tag started (checking for backslash)
3732 } elseif ( $ch == '>' ) {
3733 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3734 $entityState = 0; // for bad HTML
3735 $bracketState = 0; // out of brackets
3736 } elseif ( $bracketState == 1 ) {
3737 if ( $ch == '/' ) {
3738 $tagType = 1; // close tag (e.g. "</span>")
3739 } else {
3740 $tagType = 0; // open tag (e.g. "<span>")
3741 $tag .= $ch;
3742 }
3743 $bracketState = 2; // building tag name
3744 } elseif ( $bracketState == 2 ) {
3745 if ( $ch != ' ' ) {
3746 $tag .= $ch;
3747 } else {
3748 // Name found (e.g. "<a href=..."), add on tag attributes...
3749 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3750 }
3751 } elseif ( $bracketState == 0 ) {
3752 if ( $entityState ) {
3753 if ( $ch == ';' ) {
3754 $entityState = 0;
3755 $dispLen++; // entity is one displayed char
3756 }
3757 } else {
3758 if ( $neLength == 0 && !$maybeState ) {
3759 // Save state without $ch. We want to *hit* the first
3760 // display char (to get tags) but not *use* it if truncating.
3761 $maybeState = [ substr( $ret, 0, -1 ), $openTags ];
3762 }
3763 if ( $ch == '&' ) {
3764 $entityState = 1; // entity found, (e.g. "&#160;")
3765 } else {
3766 $dispLen++; // this char is displayed
3767 // Add the next $max display text chars after this in one swoop...
3768 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3769 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3770 $dispLen += $skipped;
3771 $pos += $skipped;
3772 }
3773 }
3774 }
3775 }
3776 // Close the last tag if left unclosed by bad HTML
3777 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3778 while ( count( $openTags ) > 0 ) {
3779 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3780 }
3781 return $ret;
3782 }
3783
3784 /**
3785 * truncateHtml() helper function
3786 * like strcspn() but adds the skipped chars to $ret
3787 *
3788 * @param string $ret
3789 * @param string $text
3790 * @param string $search
3791 * @param int $start
3792 * @param null|int $len
3793 * @return int
3794 */
3795 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3796 if ( $len === null ) {
3797 $len = -1; // -1 means "no limit" for strcspn
3798 } elseif ( $len < 0 ) {
3799 $len = 0; // sanity
3800 }
3801 $skipCount = 0;
3802 if ( $start < strlen( $text ) ) {
3803 $skipCount = strcspn( $text, $search, $start, $len );
3804 $ret .= substr( $text, $start, $skipCount );
3805 }
3806 return $skipCount;
3807 }
3808
3809 /**
3810 * truncateHtml() helper function
3811 * (a) push or pop $tag from $openTags as needed
3812 * (b) clear $tag value
3813 * @param string &$tag Current HTML tag name we are looking at
3814 * @param int $tagType (0-open tag, 1-close tag)
3815 * @param string $lastCh Character before the '>' that ended this tag
3816 * @param array &$openTags Open tag stack (not accounting for $tag)
3817 */
3818 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3819 $tag = ltrim( $tag );
3820 if ( $tag != '' ) {
3821 if ( $tagType == 0 && $lastCh != '/' ) {
3822 $openTags[] = $tag; // tag opened (didn't close itself)
3823 } elseif ( $tagType == 1 ) {
3824 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3825 array_pop( $openTags ); // tag closed
3826 }
3827 }
3828 $tag = '';
3829 }
3830 }
3831
3832 /**
3833 * Grammatical transformations, needed for inflected languages
3834 * Invoked by putting {{grammar:case|word}} in a message
3835 *
3836 * @param string $word
3837 * @param string $case
3838 * @return string
3839 */
3840 function convertGrammar( $word, $case ) {
3841 global $wgGrammarForms;
3842 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3843 return $wgGrammarForms[$this->getCode()][$case][$word];
3844 }
3845
3846 $grammarTransformations = $this->getGrammarTransformations();
3847
3848 if ( isset( $grammarTransformations[$case] ) ) {
3849 $forms = $grammarTransformations[$case];
3850
3851 // Some names of grammar rules are aliases for other rules.
3852 // In such cases the value is a string rather than object,
3853 // so load the actual rules.
3854 if ( is_string( $forms ) ) {
3855 $forms = $grammarTransformations[$forms];
3856 }
3857
3858 foreach ( array_values( $forms ) as $rule ) {
3859 $form = $rule[0];
3860
3861 if ( $form === '@metadata' ) {
3862 continue;
3863 }
3864
3865 $replacement = $rule[1];
3866
3867 $regex = '/' . addcslashes( $form, '/' ) . '/u';
3868 $patternMatches = preg_match( $regex, $word );
3869
3870 if ( $patternMatches === false ) {
3871 wfLogWarning(
3872 'An error occurred while processing grammar. ' .
3873 "Word: '$word'. Regex: /$form/."
3874 );
3875 } elseif ( $patternMatches === 1 ) {
3876 $word = preg_replace( $regex, $replacement, $word );
3877
3878 break;
3879 }
3880 }
3881 }
3882
3883 return $word;
3884 }
3885
3886 /**
3887 * Get the grammar forms for the content language
3888 * @return array Array of grammar forms
3889 * @since 1.20
3890 */
3891 function getGrammarForms() {
3892 global $wgGrammarForms;
3893 if ( isset( $wgGrammarForms[$this->getCode()] )
3894 && is_array( $wgGrammarForms[$this->getCode()] )
3895 ) {
3896 return $wgGrammarForms[$this->getCode()];
3897 }
3898
3899 return [];
3900 }
3901
3902 /**
3903 * Get the grammar transformations data for the language.
3904 * Used like grammar forms, with {{GRAMMAR}} and cases,
3905 * but uses pairs of regexes and replacements instead of code.
3906 *
3907 * @return array[] Array of grammar transformations.
3908 * @throws MWException
3909 * @since 1.28
3910 */
3911 public function getGrammarTransformations() {
3912 $languageCode = $this->getCode();
3913
3914 if ( self::$grammarTransformations === null ) {
3915 self::$grammarTransformations = new MapCacheLRU( 10 );
3916 }
3917
3918 if ( self::$grammarTransformations->has( $languageCode ) ) {
3919 return self::$grammarTransformations->get( $languageCode );
3920 }
3921
3922 $data = [];
3923
3924 $grammarDataFile = __DIR__ . "/data/grammarTransformations/$languageCode.json";
3925 if ( is_readable( $grammarDataFile ) ) {
3926 $data = FormatJson::decode(
3927 file_get_contents( $grammarDataFile ),
3928 true
3929 );
3930
3931 if ( $data === null ) {
3932 throw new MWException( "Invalid grammar data for \"$languageCode\"." );
3933 }
3934
3935 self::$grammarTransformations->set( $languageCode, $data );
3936 }
3937
3938 return $data;
3939 }
3940
3941 /**
3942 * Provides an alternative text depending on specified gender.
3943 * Usage {{gender:username|masculine|feminine|unknown}}.
3944 * username is optional, in which case the gender of current user is used,
3945 * but only in (some) interface messages; otherwise default gender is used.
3946 *
3947 * If no forms are given, an empty string is returned. If only one form is
3948 * given, it will be returned unconditionally. These details are implied by
3949 * the caller and cannot be overridden in subclasses.
3950 *
3951 * If three forms are given, the default is to use the third (unknown) form.
3952 * If fewer than three forms are given, the default is to use the first (masculine) form.
3953 * These details can be overridden in subclasses.
3954 *
3955 * @param string $gender
3956 * @param array $forms
3957 *
3958 * @return string
3959 */
3960 function gender( $gender, $forms ) {
3961 if ( !count( $forms ) ) {
3962 return '';
3963 }
3964 $forms = $this->preConvertPlural( $forms, 2 );
3965 if ( $gender === 'male' ) {
3966 return $forms[0];
3967 }
3968 if ( $gender === 'female' ) {
3969 return $forms[1];
3970 }
3971 return $forms[2] ?? $forms[0];
3972 }
3973
3974 /**
3975 * Plural form transformations, needed for some languages.
3976 * For example, there are 3 form of plural in Russian and Polish,
3977 * depending on "count mod 10". See [[w:Plural]]
3978 * For English it is pretty simple.
3979 *
3980 * Invoked by putting {{plural:count|wordform1|wordform2}}
3981 * or {{plural:count|wordform1|wordform2|wordform3}}
3982 *
3983 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3984 *
3985 * @param int $count Non-localized number
3986 * @param array $forms Different plural forms
3987 * @return string Correct form of plural for $count in this language
3988 */
3989 function convertPlural( $count, $forms ) {
3990 // Handle explicit n=pluralform cases
3991 $forms = $this->handleExplicitPluralForms( $count, $forms );
3992 if ( is_string( $forms ) ) {
3993 return $forms;
3994 }
3995 if ( !count( $forms ) ) {
3996 return '';
3997 }
3998
3999 $pluralForm = $this->getPluralRuleIndexNumber( $count );
4000 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
4001 return $forms[$pluralForm];
4002 }
4003
4004 /**
4005 * Handles explicit plural forms for Language::convertPlural()
4006 *
4007 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
4008 * If an explicitly defined plural form matches the $count, then
4009 * string value returned, otherwise array returned for further consideration
4010 * by CLDR rules or overridden convertPlural().
4011 *
4012 * @since 1.23
4013 *
4014 * @param int $count Non-localized number
4015 * @param array $forms Different plural forms
4016 *
4017 * @return array|string
4018 */
4019 protected function handleExplicitPluralForms( $count, array $forms ) {
4020 foreach ( $forms as $index => $form ) {
4021 if ( preg_match( '/\d+=/i', $form ) ) {
4022 $pos = strpos( $form, '=' );
4023 if ( substr( $form, 0, $pos ) === (string)$count ) {
4024 return substr( $form, $pos + 1 );
4025 }
4026 unset( $forms[$index] );
4027 }
4028 }
4029 return array_values( $forms );
4030 }
4031
4032 /**
4033 * Checks that convertPlural was given an array and pads it to requested
4034 * amount of forms by copying the last one.
4035 *
4036 * @param array $forms
4037 * @param int $count How many forms should there be at least
4038 * @return array Padded array of forms
4039 */
4040 protected function preConvertPlural( /* Array */ $forms, $count ) {
4041 return array_pad( $forms, $count, end( $forms ) );
4042 }
4043
4044 /**
4045 * Wraps argument with unicode control characters for directionality safety
4046 *
4047 * This solves the problem where directionality-neutral characters at the edge of
4048 * the argument string get interpreted with the wrong directionality from the
4049 * enclosing context, giving renderings that look corrupted like "(Ben_(WMF".
4050 *
4051 * The wrapping is LRE...PDF or RLE...PDF, depending on the detected
4052 * directionality of the argument string, using the BIDI algorithm's own "First
4053 * strong directional codepoint" rule. Essentially, this works round the fact that
4054 * there is no embedding equivalent of U+2068 FSI (isolation with heuristic
4055 * direction inference). The latter is cleaner but still not widely supported.
4056 *
4057 * @param string $text Text to wrap
4058 * @return string Text, wrapped in LRE...PDF or RLE...PDF or nothing
4059 */
4060 public function embedBidi( $text = '' ) {
4061 $dir = self::strongDirFromContent( $text );
4062 if ( $dir === 'ltr' ) {
4063 // Wrap in LEFT-TO-RIGHT EMBEDDING ... POP DIRECTIONAL FORMATTING
4064 return self::$lre . $text . self::$pdf;
4065 }
4066 if ( $dir === 'rtl' ) {
4067 // Wrap in RIGHT-TO-LEFT EMBEDDING ... POP DIRECTIONAL FORMATTING
4068 return self::$rle . $text . self::$pdf;
4069 }
4070 // No strong directionality: do not wrap
4071 return $text;
4072 }
4073
4074 /**
4075 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
4076 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
4077 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
4078 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
4079 * match up with it.
4080 *
4081 * @param string $str The validated block duration in English
4082 * @param User|null $user User object to use timezone from or null for $wgUser
4083 * @param int $now Current timestamp, for formatting relative block durations
4084 * @return string Somehow translated block duration
4085 * @see LanguageFi.php for example implementation
4086 */
4087 function translateBlockExpiry( $str, User $user = null, $now = 0 ) {
4088 $duration = SpecialBlock::getSuggestedDurations( $this );
4089 foreach ( $duration as $show => $value ) {
4090 if ( strcmp( $str, $value ) == 0 ) {
4091 return htmlspecialchars( trim( $show ) );
4092 }
4093 }
4094
4095 if ( wfIsInfinity( $str ) ) {
4096 foreach ( $duration as $show => $value ) {
4097 if ( wfIsInfinity( $value ) ) {
4098 return htmlspecialchars( trim( $show ) );
4099 }
4100 }
4101 }
4102
4103 // If all else fails, return a standard duration or timestamp description.
4104 $time = strtotime( $str, $now );
4105 if ( $time === false ) { // Unknown format. Return it as-is in case.
4106 return $str;
4107 } elseif ( $time !== strtotime( $str, $now + 1 ) ) { // It's a relative timestamp.
4108 // The result differs based on current time, so the difference
4109 // is a fixed duration length.
4110 return $this->formatDuration( $time - $now );
4111 } else { // It's an absolute timestamp.
4112 if ( $time === 0 ) {
4113 // wfTimestamp() handles 0 as current time instead of epoch.
4114 $time = '19700101000000';
4115 }
4116 if ( $user ) {
4117 return $this->userTimeAndDate( $time, $user );
4118 }
4119 return $this->timeanddate( $time );
4120 }
4121 }
4122
4123 /**
4124 * languages like Chinese need to be segmented in order for the diff
4125 * to be of any use
4126 *
4127 * @param string $text
4128 * @return string
4129 */
4130 public function segmentForDiff( $text ) {
4131 return $text;
4132 }
4133
4134 /**
4135 * and unsegment to show the result
4136 *
4137 * @param string $text
4138 * @return string
4139 */
4140 public function unsegmentForDiff( $text ) {
4141 return $text;
4142 }
4143
4144 /**
4145 * Return the LanguageConverter used in the Language
4146 *
4147 * @since 1.19
4148 * @return LanguageConverter
4149 */
4150 public function getConverter() {
4151 return $this->mConverter;
4152 }
4153
4154 /**
4155 * convert text to a variant
4156 *
4157 * @param string $text text to convert
4158 * @param string|bool $variant variant to convert to, or false to use the user's preferred
4159 * variant (if logged in), or the project default variant
4160 * @return string the converted string
4161 */
4162 public function autoConvert( $text, $variant = false ) {
4163 return $this->mConverter->autoConvert( $text, $variant );
4164 }
4165
4166 /**
4167 * convert text to all supported variants
4168 *
4169 * @param string $text
4170 * @return array
4171 */
4172 public function autoConvertToAllVariants( $text ) {
4173 return $this->mConverter->autoConvertToAllVariants( $text );
4174 }
4175
4176 /**
4177 * convert text to different variants of a language.
4178 *
4179 * @warning Glossary state is maintained between calls. This means
4180 * if you pass unescaped text to this method it can cause an XSS
4181 * in later calls to this method, even if the later calls have properly
4182 * escaped the input. Never feed this method user controlled text that
4183 * is not properly escaped!
4184 * @param string $text Content that has been already escaped for use in HTML
4185 * @return string HTML
4186 */
4187 public function convert( $text ) {
4188 return $this->mConverter->convert( $text );
4189 }
4190
4191 /**
4192 * Convert a Title object to a string in the preferred variant
4193 *
4194 * @param Title $title
4195 * @return string
4196 */
4197 public function convertTitle( $title ) {
4198 return $this->mConverter->convertTitle( $title );
4199 }
4200
4201 /**
4202 * Convert a namespace index to a string in the preferred variant
4203 *
4204 * @param int $ns namespace index (https://www.mediawiki.org/wiki/Manual:Namespace)
4205 * @param string|null $variant variant to convert to, or null to use the user's preferred
4206 * variant (if logged in), or the project default variant
4207 * @return string a string representation of the namespace
4208 */
4209 public function convertNamespace( $ns, $variant = null ) {
4210 return $this->mConverter->convertNamespace( $ns, $variant );
4211 }
4212
4213 /**
4214 * Check if this is a language with variants
4215 *
4216 * @return bool
4217 */
4218 public function hasVariants() {
4219 return count( $this->getVariants() ) > 1;
4220 }
4221
4222 /**
4223 * Strict check if the language has the specific variant.
4224 *
4225 * Compare to LanguageConverter::validateVariant() which does a more
4226 * lenient check and attempts to coerce the given code to a valid one.
4227 *
4228 * @since 1.19
4229 * @param string $variant
4230 * @return bool
4231 */
4232 public function hasVariant( $variant ) {
4233 return $variant && ( $variant === $this->mConverter->validateVariant( $variant ) );
4234 }
4235
4236 /**
4237 * Perform output conversion on a string, and encode for safe HTML output.
4238 * @param string $text Text to be converted
4239 * @return string
4240 * @todo this should get integrated somewhere sane
4241 */
4242 public function convertHtml( $text ) {
4243 return htmlspecialchars( $this->convert( $text ) );
4244 }
4245
4246 /**
4247 * @param string $key
4248 * @return string
4249 */
4250 public function convertCategoryKey( $key ) {
4251 return $this->mConverter->convertCategoryKey( $key );
4252 }
4253
4254 /**
4255 * Get the list of variants supported by this language
4256 * see sample implementation in LanguageZh.php
4257 *
4258 * @return string[] An array of language codes
4259 */
4260 public function getVariants() {
4261 return $this->mConverter->getVariants();
4262 }
4263
4264 /**
4265 * @return string
4266 */
4267 public function getPreferredVariant() {
4268 return $this->mConverter->getPreferredVariant();
4269 }
4270
4271 /**
4272 * @return string
4273 */
4274 public function getDefaultVariant() {
4275 return $this->mConverter->getDefaultVariant();
4276 }
4277
4278 /**
4279 * @return string
4280 */
4281 public function getURLVariant() {
4282 return $this->mConverter->getURLVariant();
4283 }
4284
4285 /**
4286 * If a language supports multiple variants, it is
4287 * possible that non-existing link in one variant
4288 * actually exists in another variant. this function
4289 * tries to find it. See e.g. LanguageZh.php
4290 * The input parameters may be modified upon return
4291 *
4292 * @param string &$link The name of the link
4293 * @param Title &$nt The title object of the link
4294 * @param bool $ignoreOtherCond To disable other conditions when
4295 * we need to transclude a template or update a category's link
4296 */
4297 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4298 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4299 }
4300
4301 /**
4302 * returns language specific options used by User::getPageRenderHash()
4303 * for example, the preferred language variant
4304 *
4305 * @return string
4306 */
4307 function getExtraHashOptions() {
4308 return $this->mConverter->getExtraHashOptions();
4309 }
4310
4311 /**
4312 * For languages that support multiple variants, the title of an
4313 * article may be displayed differently in different variants. this
4314 * function returns the apporiate title defined in the body of the article.
4315 *
4316 * @return string
4317 */
4318 public function getParsedTitle() {
4319 return $this->mConverter->getParsedTitle();
4320 }
4321
4322 /**
4323 * Refresh the cache of conversion tables when
4324 * MediaWiki:Conversiontable* is updated.
4325 *
4326 * @param Title $title The Title of the page being updated
4327 */
4328 public function updateConversionTable( Title $title ) {
4329 $this->mConverter->updateConversionTable( $title );
4330 }
4331
4332 /**
4333 * A regular expression to match legal word-trailing characters
4334 * which should be merged onto a link of the form [[foo]]bar.
4335 *
4336 * @return string
4337 */
4338 public function linkTrail() {
4339 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4340 }
4341
4342 /**
4343 * A regular expression character set to match legal word-prefixing
4344 * characters which should be merged onto a link of the form foo[[bar]].
4345 *
4346 * @return string
4347 */
4348 public function linkPrefixCharset() {
4349 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4350 }
4351
4352 /**
4353 * Get the "parent" language which has a converter to convert a "compatible" language
4354 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4355 *
4356 * @return Language|null
4357 * @since 1.22
4358 */
4359 public function getParentLanguage() {
4360 if ( $this->mParentLanguage !== false ) {
4361 return $this->mParentLanguage;
4362 }
4363
4364 $code = explode( '-', $this->getCode() )[0];
4365 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4366 $this->mParentLanguage = null;
4367 return null;
4368 }
4369 $lang = self::factory( $code );
4370 if ( !$lang->hasVariant( $this->getCode() ) ) {
4371 $this->mParentLanguage = null;
4372 return null;
4373 }
4374
4375 $this->mParentLanguage = $lang;
4376 return $lang;
4377 }
4378
4379 /**
4380 * Compare with an other language object
4381 *
4382 * @since 1.28
4383 * @param Language $lang
4384 * @return bool
4385 */
4386 public function equals( Language $lang ) {
4387 return $lang === $this || $lang->getCode() === $this->mCode;
4388 }
4389
4390 /**
4391 * Get the internal language code for this language object
4392 *
4393 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4394 * htmlspecialchars() or similar
4395 *
4396 * @return string
4397 */
4398 public function getCode() {
4399 return $this->mCode;
4400 }
4401
4402 /**
4403 * Get the code in BCP 47 format which we can use
4404 * inside of html lang="" tags.
4405 *
4406 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4407 * htmlspecialchars() or similar.
4408 *
4409 * @since 1.19
4410 * @return string
4411 */
4412 public function getHtmlCode() {
4413 if ( is_null( $this->mHtmlCode ) ) {
4414 $this->mHtmlCode = LanguageCode::bcp47( $this->getCode() );
4415 }
4416 return $this->mHtmlCode;
4417 }
4418
4419 /**
4420 * Get the language code from a file name. Inverse of getFileName()
4421 * @param string $filename $prefix . $languageCode . $suffix
4422 * @param string $prefix Prefix before the language code
4423 * @param string $suffix Suffix after the language code
4424 * @return string Language code, or false if $prefix or $suffix isn't found
4425 */
4426 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4427 $m = null;
4428 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4429 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4430 if ( !count( $m ) ) {
4431 return false;
4432 }
4433 return str_replace( '_', '-', strtolower( $m[1] ) );
4434 }
4435
4436 /**
4437 * @param string $code
4438 * @param bool $fallback Whether we're going through language fallback chain
4439 * @return string Name of the language class
4440 */
4441 public static function classFromCode( $code, $fallback = true ) {
4442 if ( $fallback && $code == 'en' ) {
4443 return 'Language';
4444 } else {
4445 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
4446 }
4447 }
4448
4449 /**
4450 * Get the name of a file for a certain language code
4451 * @param string $prefix Prepend this to the filename
4452 * @param string $code Language code
4453 * @param string $suffix Append this to the filename
4454 * @throws MWException
4455 * @return string $prefix . $mangledCode . $suffix
4456 */
4457 public static function getFileName( $prefix, $code, $suffix = '.php' ) {
4458 if ( !self::isValidBuiltInCode( $code ) ) {
4459 throw new MWException( "Invalid language code \"$code\"" );
4460 }
4461
4462 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4463 }
4464
4465 /**
4466 * @param string $code
4467 * @return string
4468 */
4469 public static function getMessagesFileName( $code ) {
4470 global $IP;
4471 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4472 Hooks::run( 'Language::getMessagesFileName', [ $code, &$file ] );
4473 return $file;
4474 }
4475
4476 /**
4477 * @param string $code
4478 * @return string
4479 * @throws MWException
4480 * @since 1.23
4481 */
4482 public static function getJsonMessagesFileName( $code ) {
4483 global $IP;
4484
4485 if ( !self::isValidBuiltInCode( $code ) ) {
4486 throw new MWException( "Invalid language code \"$code\"" );
4487 }
4488
4489 return "$IP/languages/i18n/$code.json";
4490 }
4491
4492 /**
4493 * Get the first fallback for a given language.
4494 *
4495 * @param string $code
4496 *
4497 * @return bool|string
4498 */
4499 public static function getFallbackFor( $code ) {
4500 $fallbacks = self::getFallbacksFor( $code );
4501 if ( $fallbacks ) {
4502 return $fallbacks[0];
4503 }
4504 return false;
4505 }
4506
4507 /**
4508 * Get the ordered list of fallback languages.
4509 *
4510 * @since 1.19
4511 * @param string $code Language code
4512 * @param int $mode Fallback mode, either MESSAGES_FALLBACKS (which always falls back to 'en'),
4513 * or STRICT_FALLBACKS (whic honly falls back to 'en' when explicitly defined)
4514 * @throws MWException
4515 * @return array List of language codes
4516 */
4517 public static function getFallbacksFor( $code, $mode = self::MESSAGES_FALLBACKS ) {
4518 if ( $code === 'en' || !self::isValidBuiltInCode( $code ) ) {
4519 return [];
4520 }
4521 switch ( $mode ) {
4522 case self::MESSAGES_FALLBACKS:
4523 // For unknown languages, fallbackSequence returns an empty array,
4524 // hardcode fallback to 'en' in that case as English messages are
4525 // always defined.
4526 return self::getLocalisationCache()->getItem( $code, 'fallbackSequence' ) ?: [ 'en' ];
4527 case self::STRICT_FALLBACKS:
4528 // Use this mode when you don't want to fallback to English unless
4529 // explicitly defined, for example when you have language-variant icons
4530 // and an international language-independent fallback.
4531 return self::getLocalisationCache()->getItem( $code, 'originalFallbackSequence' );
4532 default:
4533 throw new MWException( "Invalid fallback mode \"$mode\"" );
4534 }
4535 }
4536
4537 /**
4538 * Get the ordered list of fallback languages, ending with the fallback
4539 * language chain for the site language.
4540 *
4541 * @since 1.22
4542 * @param string $code Language code
4543 * @return array [ fallbacks, site fallbacks ]
4544 */
4545 public static function getFallbacksIncludingSiteLanguage( $code ) {
4546 global $wgLanguageCode;
4547
4548 // Usually, we will only store a tiny number of fallback chains, so we
4549 // keep them in static memory.
4550 $cacheKey = "{$code}-{$wgLanguageCode}";
4551
4552 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4553 $fallbacks = self::getFallbacksFor( $code );
4554
4555 // Append the site's fallback chain, including the site language itself
4556 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4557 array_unshift( $siteFallbacks, $wgLanguageCode );
4558
4559 // Eliminate any languages already included in the chain
4560 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4561
4562 self::$fallbackLanguageCache[$cacheKey] = [ $fallbacks, $siteFallbacks ];
4563 }
4564 return self::$fallbackLanguageCache[$cacheKey];
4565 }
4566
4567 /**
4568 * Get all messages for a given language
4569 * WARNING: this may take a long time. If you just need all message *keys*
4570 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4571 *
4572 * @param string $code
4573 *
4574 * @return array
4575 */
4576 public static function getMessagesFor( $code ) {
4577 return self::getLocalisationCache()->getItem( $code, 'messages' );
4578 }
4579
4580 /**
4581 * Get a message for a given language
4582 *
4583 * @param string $key
4584 * @param string $code
4585 *
4586 * @return string
4587 */
4588 public static function getMessageFor( $key, $code ) {
4589 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4590 }
4591
4592 /**
4593 * Get all message keys for a given language. This is a faster alternative to
4594 * array_keys( Language::getMessagesFor( $code ) )
4595 *
4596 * @since 1.19
4597 * @param string $code Language code
4598 * @return array Array of message keys (strings)
4599 */
4600 public static function getMessageKeysFor( $code ) {
4601 return self::getLocalisationCache()->getSubitemList( $code, 'messages' );
4602 }
4603
4604 /**
4605 * @param string $talk
4606 * @return string
4607 */
4608 function fixVariableInNamespace( $talk ) {
4609 if ( strpos( $talk, '$1' ) === false ) {
4610 return $talk;
4611 }
4612
4613 global $wgMetaNamespace;
4614 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4615
4616 # Allow grammar transformations
4617 # Allowing full message-style parsing would make simple requests
4618 # such as action=raw much more expensive than they need to be.
4619 # This will hopefully cover most cases.
4620 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4621 [ $this, 'replaceGrammarInNamespace' ], $talk );
4622 return str_replace( ' ', '_', $talk );
4623 }
4624
4625 /**
4626 * @param string $m
4627 * @return string
4628 */
4629 function replaceGrammarInNamespace( $m ) {
4630 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4631 }
4632
4633 /**
4634 * Decode an expiry (block, protection, etc) which has come from the DB
4635 *
4636 * @param string $expiry Database expiry String
4637 * @param bool|int $format True to process using language functions, or TS_ constant
4638 * to return the expiry in a given timestamp
4639 * @param string $infinity If $format is not true, use this string for infinite expiry
4640 * @return string
4641 * @since 1.18
4642 */
4643 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4644 static $dbInfinity;
4645 if ( $dbInfinity === null ) {
4646 $dbInfinity = wfGetDB( DB_REPLICA )->getInfinity();
4647 }
4648
4649 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4650 return $format === true
4651 ? $this->getMessageFromDB( 'infiniteblock' )
4652 : $infinity;
4653 } else {
4654 return $format === true
4655 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4656 : wfTimestamp( $format, $expiry );
4657 }
4658 }
4659
4660 /**
4661 * Formats a time given in seconds into a string representation of that time.
4662 *
4663 * @param int|float $seconds
4664 * @param array $format An optional argument that formats the returned string in different ways:
4665 * If $format['avoid'] === 'avoidhours': don't show hours, just show days
4666 * If $format['avoid'] === 'avoidseconds': don't show seconds if $seconds >= 1 hour,
4667 * If $format['avoid'] === 'avoidminutes': don't show seconds/minutes if $seconds > 48 hours,
4668 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4669 * and friends.
4670 * @note For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4671 * or 'avoidminutes'.
4672 * @return string
4673 */
4674 function formatTimePeriod( $seconds, $format = [] ) {
4675 if ( !is_array( $format ) ) {
4676 $format = [ 'avoid' => $format ]; // For backwards compatibility
4677 }
4678 if ( !isset( $format['avoid'] ) ) {
4679 $format['avoid'] = false;
4680 }
4681 if ( !isset( $format['noabbrevs'] ) ) {
4682 $format['noabbrevs'] = false;
4683 }
4684 $secondsMsg = wfMessage(
4685 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4686 $minutesMsg = wfMessage(
4687 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4688 $hoursMsg = wfMessage(
4689 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4690 $daysMsg = wfMessage(
4691 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4692
4693 if ( round( $seconds * 10 ) < 100 ) {
4694 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4695 $s = $secondsMsg->params( $s )->text();
4696 } elseif ( round( $seconds ) < 60 ) {
4697 $s = $this->formatNum( round( $seconds ) );
4698 $s = $secondsMsg->params( $s )->text();
4699 } elseif ( round( $seconds ) < 3600 ) {
4700 $minutes = floor( $seconds / 60 );
4701 $secondsPart = round( fmod( $seconds, 60 ) );
4702 if ( $secondsPart == 60 ) {
4703 $secondsPart = 0;
4704 $minutes++;
4705 }
4706 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4707 $s .= ' ';
4708 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4709 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4710 $hours = floor( $seconds / 3600 );
4711 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4712 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4713 if ( $secondsPart == 60 ) {
4714 $secondsPart = 0;
4715 $minutes++;
4716 }
4717 if ( $minutes == 60 ) {
4718 $minutes = 0;
4719 $hours++;
4720 }
4721 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4722 $s .= ' ';
4723 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4724 if ( !in_array( $format['avoid'], [ 'avoidseconds', 'avoidminutes', 'avoidhours' ] ) ) {
4725 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4726 }
4727 } else {
4728 $days = floor( $seconds / 86400 );
4729 if ( $format['avoid'] === 'avoidhours' ) {
4730 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4731 if ( $hours == 24 ) {
4732 $hours = 0;
4733 $days++;
4734 }
4735 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4736 } elseif ( $format['avoid'] === 'avoidminutes' ) {
4737 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4738 if ( $hours == 24 ) {
4739 $hours = 0;
4740 $days++;
4741 }
4742 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4743 $s .= ' ';
4744 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4745 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4746 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4747 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4748 if ( $minutes == 60 ) {
4749 $minutes = 0;
4750 $hours++;
4751 }
4752 if ( $hours == 24 ) {
4753 $hours = 0;
4754 $days++;
4755 }
4756 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4757 $s .= ' ';
4758 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4759 $s .= ' ';
4760 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4761 } else {
4762 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4763 $s .= ' ';
4764 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4765 }
4766 }
4767 return $s;
4768 }
4769
4770 /**
4771 * Format a bitrate for output, using an appropriate
4772 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4773 * the magnitude in question.
4774 *
4775 * This use base 1000. For base 1024 use formatSize(), for another base
4776 * see formatComputingNumbers().
4777 *
4778 * @param int $bps
4779 * @return string
4780 */
4781 function formatBitrate( $bps ) {
4782 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4783 }
4784
4785 /**
4786 * @param int $size Size of the unit
4787 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4788 * @param string $messageKey Message key to be uesd
4789 * @return string
4790 */
4791 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4792 if ( $size <= 0 ) {
4793 return str_replace( '$1', $this->formatNum( $size ),
4794 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4795 );
4796 }
4797 $sizes = [ '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' ];
4798 $index = 0;
4799
4800 $maxIndex = count( $sizes ) - 1;
4801 while ( $size >= $boundary && $index < $maxIndex ) {
4802 $index++;
4803 $size /= $boundary;
4804 }
4805
4806 // For small sizes no decimal places necessary
4807 $round = 0;
4808 if ( $index > 1 ) {
4809 // For MB and bigger two decimal places are smarter
4810 $round = 2;
4811 }
4812 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4813
4814 $size = round( $size, $round );
4815 $text = $this->getMessageFromDB( $msg );
4816 return str_replace( '$1', $this->formatNum( $size ), $text );
4817 }
4818
4819 /**
4820 * Format a size in bytes for output, using an appropriate
4821 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4822 *
4823 * This method use base 1024. For base 1000 use formatBitrate(), for
4824 * another base see formatComputingNumbers()
4825 *
4826 * @param int $size Size to format
4827 * @return string Plain text (not HTML)
4828 */
4829 function formatSize( $size ) {
4830 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4831 }
4832
4833 /**
4834 * Make a list item, used by various special pages
4835 *
4836 * @param string $page Page link
4837 * @param string $details HTML safe text between brackets
4838 * @param bool $oppositedm Add the direction mark opposite to your
4839 * language, to display text properly
4840 * @return string HTML escaped
4841 */
4842 function specialList( $page, $details, $oppositedm = true ) {
4843 if ( !$details ) {
4844 return $page;
4845 }
4846
4847 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4848 return $page .
4849 $dirmark .
4850 $this->msg( 'word-separator' )->escaped() .
4851 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4852 }
4853
4854 /**
4855 * Generate (prev x| next x) (20|50|100...) type links for paging
4856 *
4857 * @param Title $title Title object to link
4858 * @param int $offset
4859 * @param int $limit
4860 * @param array $query Optional URL query parameter string
4861 * @param bool $atend Optional param for specified if this is the last page
4862 * @return string
4863 * @deprecated since 1.34, use PrevNextNavigationRenderer::buildPrevNextNavigation()
4864 * instead.
4865 */
4866 public function viewPrevNext( Title $title, $offset, $limit,
4867 array $query = [], $atend = false
4868 ) {
4869 wfDeprecated( __METHOD__, '1.34' );
4870 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4871
4872 # Make 'previous' link
4873 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4874 if ( $offset > 0 ) {
4875 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4876 $query, $prev, 'prevn-title', 'mw-prevlink' );
4877 } else {
4878 $plink = htmlspecialchars( $prev );
4879 }
4880
4881 # Make 'next' link
4882 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4883 if ( $atend ) {
4884 $nlink = htmlspecialchars( $next );
4885 } else {
4886 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4887 $query, $next, 'nextn-title', 'mw-nextlink' );
4888 }
4889
4890 # Make links to set number of items per page
4891 $numLinks = [];
4892 foreach ( [ 20, 50, 100, 250, 500 ] as $num ) {
4893 $numLinks[] = $this->numLink( $title, $offset, $num,
4894 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4895 }
4896
4897 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4898 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4899 }
4900
4901 /**
4902 * Helper function for viewPrevNext() that generates links
4903 *
4904 * @param Title $title Title object to link
4905 * @param int $offset
4906 * @param int $limit
4907 * @param array $query Extra query parameters
4908 * @param string $link Text to use for the link; will be escaped
4909 * @param string $tooltipMsg Name of the message to use as tooltip
4910 * @param string $class Value of the "class" attribute of the link
4911 * @return string HTML fragment
4912 */
4913 private function numLink( Title $title, $offset, $limit, array $query, $link,
4914 $tooltipMsg, $class
4915 ) {
4916 $query = [ 'limit' => $limit, 'offset' => $offset ] + $query;
4917 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4918 ->numParams( $limit )->text();
4919
4920 return Html::element( 'a', [ 'href' => $title->getLocalURL( $query ),
4921 'title' => $tooltip, 'class' => $class ], $link );
4922 }
4923
4924 /**
4925 * Get the conversion rule title, if any.
4926 *
4927 * @return string
4928 */
4929 public function getConvRuleTitle() {
4930 return $this->mConverter->getConvRuleTitle();
4931 }
4932
4933 /**
4934 * Get the compiled plural rules for the language
4935 * @since 1.20
4936 * @return array Associative array with plural form, and plural rule as key-value pairs
4937 */
4938 public function getCompiledPluralRules() {
4939 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4940 $fallbacks = self::getFallbacksFor( $this->mCode );
4941 if ( !$pluralRules ) {
4942 foreach ( $fallbacks as $fallbackCode ) {
4943 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4944 if ( $pluralRules ) {
4945 break;
4946 }
4947 }
4948 }
4949 return $pluralRules;
4950 }
4951
4952 /**
4953 * Get the plural rules for the language
4954 * @since 1.20
4955 * @return array Associative array with plural form number and plural rule as key-value pairs
4956 */
4957 public function getPluralRules() {
4958 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4959 $fallbacks = self::getFallbacksFor( $this->mCode );
4960 if ( !$pluralRules ) {
4961 foreach ( $fallbacks as $fallbackCode ) {
4962 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4963 if ( $pluralRules ) {
4964 break;
4965 }
4966 }
4967 }
4968 return $pluralRules;
4969 }
4970
4971 /**
4972 * Get the plural rule types for the language
4973 * @since 1.22
4974 * @return array Associative array with plural form number and plural rule type as key-value pairs
4975 */
4976 public function getPluralRuleTypes() {
4977 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4978 $fallbacks = self::getFallbacksFor( $this->mCode );
4979 if ( !$pluralRuleTypes ) {
4980 foreach ( $fallbacks as $fallbackCode ) {
4981 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4982 if ( $pluralRuleTypes ) {
4983 break;
4984 }
4985 }
4986 }
4987 return $pluralRuleTypes;
4988 }
4989
4990 /**
4991 * Find the index number of the plural rule appropriate for the given number
4992 * @param int $number
4993 * @return int The index number of the plural rule
4994 */
4995 public function getPluralRuleIndexNumber( $number ) {
4996 $pluralRules = $this->getCompiledPluralRules();
4997 $form = Evaluator::evaluateCompiled( $number, $pluralRules );
4998 return $form;
4999 }
5000
5001 /**
5002 * Find the plural rule type appropriate for the given number
5003 * For example, if the language is set to Arabic, getPluralType(5) should
5004 * return 'few'.
5005 * @since 1.22
5006 * @param int $number
5007 * @return string The name of the plural rule type, e.g. one, two, few, many
5008 */
5009 public function getPluralRuleType( $number ) {
5010 $index = $this->getPluralRuleIndexNumber( $number );
5011 $pluralRuleTypes = $this->getPluralRuleTypes();
5012 return $pluralRuleTypes[$index] ?? 'other';
5013 }
5014 }
Wiklou, le Wiki du Wiklou