Don't display multiple language links to the same language
[mediawiki.git] / includes / parser / DateFormatter.php
blob2917b4a74073aff28b43e1c1844ddf90b64961a7
1 <?php
2 /**
3 * Date formatter
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
20 * @file
21 * @ingroup Parser
24 /**
25 * Date formatter, recognises dates in plain text and formats them accoding to user preferences.
26 * @todo preferences, OutputPage
27 * @ingroup Parser
29 class DateFormatter {
30 var $mSource, $mTarget;
31 var $monthNames = '', $rxDM, $rxMD, $rxDMY, $rxYDM, $rxMDY, $rxYMD;
33 var $regexes, $pDays, $pMonths, $pYears;
34 var $rules, $xMonths, $preferences;
36 protected $lang;
38 const ALL = -1;
39 const NONE = 0;
40 const MDY = 1;
41 const DMY = 2;
42 const YMD = 3;
43 const ISO1 = 4;
44 const LASTPREF = 4;
45 const ISO2 = 5;
46 const YDM = 6;
47 const DM = 7;
48 const MD = 8;
49 const LAST = 8;
51 /**
52 * @param $lang Language In which language to format the date
54 function __construct( Language $lang ) {
55 $this->lang = $lang;
57 $this->monthNames = $this->getMonthRegex();
58 for ( $i=1; $i<=12; $i++ ) {
59 $this->xMonths[$this->lang->lc( $this->lang->getMonthName( $i ) )] = $i;
60 $this->xMonths[$this->lang->lc( $this->lang->getMonthAbbreviation( $i ) )] = $i;
63 $this->regexTrail = '(?![a-z])/iu';
65 # Partial regular expressions
66 $this->prxDM = '\[\[(\d{1,2})[ _](' . $this->monthNames . ')\]\]';
67 $this->prxMD = '\[\[(' . $this->monthNames . ')[ _](\d{1,2})\]\]';
68 $this->prxY = '\[\[(\d{1,4}([ _]BC|))\]\]';
69 $this->prxISO1 = '\[\[(-?\d{4})]]-\[\[(\d{2})-(\d{2})\]\]';
70 $this->prxISO2 = '\[\[(-?\d{4})-(\d{2})-(\d{2})\]\]';
72 # Real regular expressions
73 $this->regexes[self::DMY] = "/{$this->prxDM}(?: *, *| +){$this->prxY}{$this->regexTrail}";
74 $this->regexes[self::YDM] = "/{$this->prxY}(?: *, *| +){$this->prxDM}{$this->regexTrail}";
75 $this->regexes[self::MDY] = "/{$this->prxMD}(?: *, *| +){$this->prxY}{$this->regexTrail}";
76 $this->regexes[self::YMD] = "/{$this->prxY}(?: *, *| +){$this->prxMD}{$this->regexTrail}";
77 $this->regexes[self::DM] = "/{$this->prxDM}{$this->regexTrail}";
78 $this->regexes[self::MD] = "/{$this->prxMD}{$this->regexTrail}";
79 $this->regexes[self::ISO1] = "/{$this->prxISO1}{$this->regexTrail}";
80 $this->regexes[self::ISO2] = "/{$this->prxISO2}{$this->regexTrail}";
82 # Extraction keys
83 # See the comments in replace() for the meaning of the letters
84 $this->keys[self::DMY] = 'jFY';
85 $this->keys[self::YDM] = 'Y jF';
86 $this->keys[self::MDY] = 'FjY';
87 $this->keys[self::YMD] = 'Y Fj';
88 $this->keys[self::DM] = 'jF';
89 $this->keys[self::MD] = 'Fj';
90 $this->keys[self::ISO1] = 'ymd'; # y means ISO year
91 $this->keys[self::ISO2] = 'ymd';
93 # Target date formats
94 $this->targets[self::DMY] = '[[F j|j F]] [[Y]]';
95 $this->targets[self::YDM] = '[[Y]], [[F j|j F]]';
96 $this->targets[self::MDY] = '[[F j]], [[Y]]';
97 $this->targets[self::YMD] = '[[Y]] [[F j]]';
98 $this->targets[self::DM] = '[[F j|j F]]';
99 $this->targets[self::MD] = '[[F j]]';
100 $this->targets[self::ISO1] = '[[Y|y]]-[[F j|m-d]]';
101 $this->targets[self::ISO2] = '[[y-m-d]]';
103 # Rules
104 # pref source target
105 $this->rules[self::DMY][self::MD] = self::DM;
106 $this->rules[self::ALL][self::MD] = self::MD;
107 $this->rules[self::MDY][self::DM] = self::MD;
108 $this->rules[self::ALL][self::DM] = self::DM;
109 $this->rules[self::NONE][self::ISO2] = self::ISO1;
111 $this->preferences = array(
112 'default' => self::NONE,
113 'dmy' => self::DMY,
114 'mdy' => self::MDY,
115 'ymd' => self::YMD,
116 'ISO 8601' => self::ISO1,
121 * Get a DateFormatter object
123 * @param $lang Language|string|null In which language to format the date
124 * Defaults to the site content language
125 * @return DateFormatter object
127 public static function &getInstance( $lang = null ) {
128 global $wgMemc, $wgContLang;
129 static $dateFormatter = false;
130 $lang = $lang ? wfGetLangObj( $lang ) : $wgContLang;
131 $key = wfMemcKey( 'dateformatter', $lang->getCode() );
132 if ( !$dateFormatter ) {
133 $dateFormatter = $wgMemc->get( $key );
134 if ( !$dateFormatter ) {
135 $dateFormatter = new DateFormatter( $lang );
136 $wgMemc->set( $key, $dateFormatter, 3600 );
139 return $dateFormatter;
143 * @param $preference String: User preference
144 * @param $text String: Text to reformat
145 * @param $options Array: can contain 'linked' and/or 'match-whole'
146 * @return mixed|String
148 function reformat( $preference, $text, $options = array('linked') ) {
149 $linked = in_array( 'linked', $options );
150 $match_whole = in_array( 'match-whole', $options );
152 if ( isset( $this->preferences[$preference] ) ) {
153 $preference = $this->preferences[$preference];
154 } else {
155 $preference = self::NONE;
157 for ( $i=1; $i<=self::LAST; $i++ ) {
158 $this->mSource = $i;
159 if ( isset ( $this->rules[$preference][$i] ) ) {
160 # Specific rules
161 $this->mTarget = $this->rules[$preference][$i];
162 } elseif ( isset ( $this->rules[self::ALL][$i] ) ) {
163 # General rules
164 $this->mTarget = $this->rules[self::ALL][$i];
165 } elseif ( $preference ) {
166 # User preference
167 $this->mTarget = $preference;
168 } else {
169 # Default
170 $this->mTarget = $i;
172 $regex = $this->regexes[$i];
174 // Horrible hack
175 if (!$linked) {
176 $regex = str_replace( array( '\[\[', '\]\]' ), '', $regex );
179 if ($match_whole) {
180 // Let's hope this works
181 $regex = preg_replace( '!^/!', '/^', $regex );
182 $regex = str_replace( $this->regexTrail,
183 '$'.$this->regexTrail, $regex );
186 // Another horrible hack
187 $this->mLinked = $linked;
188 $text = preg_replace_callback( $regex, array( &$this, 'replace' ), $text );
189 unset($this->mLinked);
191 return $text;
195 * @param $matches
196 * @return string
198 function replace( $matches ) {
199 # Extract information from $matches
200 $linked = true;
201 if ( isset( $this->mLinked ) )
202 $linked = $this->mLinked;
204 $bits = array();
205 $key = $this->keys[$this->mSource];
206 for ( $p=0; $p < strlen($key); $p++ ) {
207 if ( $key[$p] != ' ' ) {
208 $bits[$key[$p]] = $matches[$p+1];
212 return $this->formatDate( $bits, $linked );
216 * @param $bits array
217 * @param $link bool
218 * @return string
220 function formatDate( $bits, $link = true ) {
221 $format = $this->targets[$this->mTarget];
223 if (!$link) {
224 // strip piped links
225 $format = preg_replace( '/\[\[[^|]+\|([^\]]+)\]\]/', '$1', $format );
226 // strip remaining links
227 $format = str_replace( array( '[[', ']]' ), '', $format );
230 # Construct new date
231 $text = '';
232 $fail = false;
234 // Pre-generate y/Y stuff because we need the year for the <span> title.
235 if ( !isset( $bits['y'] ) && isset( $bits['Y'] ) )
236 $bits['y'] = $this->makeIsoYear( $bits['Y'] );
237 if ( !isset( $bits['Y'] ) && isset( $bits['y'] ) )
238 $bits['Y'] = $this->makeNormalYear( $bits['y'] );
240 if ( !isset( $bits['m'] ) ) {
241 $m = $this->makeIsoMonth( $bits['F'] );
242 if ( !$m || $m == '00' ) {
243 $fail = true;
244 } else {
245 $bits['m'] = $m;
249 if ( !isset($bits['d']) ) {
250 $bits['d'] = sprintf( '%02d', $bits['j'] );
253 for ( $p=0; $p < strlen( $format ); $p++ ) {
254 $char = $format[$p];
255 switch ( $char ) {
256 case 'd': # ISO day of month
257 $text .= $bits['d'];
258 break;
259 case 'm': # ISO month
260 $text .= $bits['m'];
261 break;
262 case 'y': # ISO year
263 $text .= $bits['y'];
264 break;
265 case 'j': # ordinary day of month
266 if ( !isset($bits['j']) ) {
267 $text .= intval( $bits['d'] );
268 } else {
269 $text .= $bits['j'];
271 break;
272 case 'F': # long month
273 if ( !isset( $bits['F'] ) ) {
274 $m = intval($bits['m']);
275 if ( $m > 12 || $m < 1 ) {
276 $fail = true;
277 } else {
278 $text .= $this->lang->getMonthName( $m );
280 } else {
281 $text .= ucfirst( $bits['F'] );
283 break;
284 case 'Y': # ordinary (optional BC) year
285 $text .= $bits['Y'];
286 break;
287 default:
288 $text .= $char;
291 if ( $fail ) {
292 $text = $matches[0];
295 $isoBits = array();
296 if ( isset($bits['y']) )
297 $isoBits[] = $bits['y'];
298 $isoBits[] = $bits['m'];
299 $isoBits[] = $bits['d'];
300 $isoDate = implode( '-', $isoBits );
302 // Output is not strictly HTML (it's wikitext), but <span> is whitelisted.
303 $text = Html::rawElement( 'span',
304 array( 'class' => 'mw-formatted-date', 'title' => $isoDate ), $text );
306 return $text;
310 * @todo document
311 * @return string
313 function getMonthRegex() {
314 $names = array();
315 for( $i = 1; $i <= 12; $i++ ) {
316 $names[] = $this->lang->getMonthName( $i );
317 $names[] = $this->lang->getMonthAbbreviation( $i );
319 return implode( '|', $names );
323 * Makes an ISO month, e.g. 02, from a month name
324 * @param $monthName String: month name
325 * @return string ISO month name
327 function makeIsoMonth( $monthName ) {
328 $n = $this->xMonths[$this->lang->lc( $monthName )];
329 return sprintf( '%02d', $n );
333 * @todo document
334 * @param $year String: Year name
335 * @return string ISO year name
337 function makeIsoYear( $year ) {
338 # Assumes the year is in a nice format, as enforced by the regex
339 if ( substr( $year, -2 ) == 'BC' ) {
340 $num = intval(substr( $year, 0, -3 )) - 1;
341 # PHP bug note: sprintf( "%04d", -1 ) fails poorly
342 $text = sprintf( '-%04d', $num );
344 } else {
345 $text = sprintf( '%04d', $year );
347 return $text;
351 * @todo document
352 * @return int|string
354 function makeNormalYear( $iso ) {
355 if ( $iso[0] == '-' ) {
356 $text = (intval( substr( $iso, 1 ) ) + 1) . ' BC';
357 } else {
358 $text = intval( $iso );
360 return $text;