MediaWiki
REL1_22
|
00001 <?php 00026 class StringUtils { 00027 00052 static function isUtf8( $value, $disableMbstring = false ) { 00053 $value = (string)$value; 00054 00055 // If the mbstring extension is loaded, use it. However, before PHP 5.4, values above 00056 // U+10FFFF are incorrectly allowed, so we have to check for them separately. 00057 if ( !$disableMbstring && function_exists( 'mb_check_encoding' ) ) { 00058 static $newPHP; 00059 if ( $newPHP === null ) { 00060 $newPHP = !mb_check_encoding( "\xf4\x90\x80\x80", 'UTF-8' ); 00061 } 00062 00063 return mb_check_encoding( $value, 'UTF-8' ) && 00064 ( $newPHP || preg_match( "/\xf4[\x90-\xbf]|[\xf5-\xff]/S", $value ) === 0 ); 00065 } 00066 00067 if ( preg_match( "/[\x80-\xff]/S", $value ) === 0 ) { 00068 // String contains only ASCII characters, has to be valid 00069 return true; 00070 } 00071 00072 // PCRE implements repetition using recursion; to avoid a stack overflow (and segfault) 00073 // for large input, we check for invalid sequences (<= 5 bytes) rather than valid 00074 // sequences, which can be as long as the input string is. Multiple short regexes are 00075 // used rather than a single long regex for performance. 00076 static $regexes; 00077 if ( $regexes === null ) { 00078 $cont = "[\x80-\xbf]"; 00079 $after = "(?!$cont)"; // "(?:[^\x80-\xbf]|$)" would work here 00080 $regexes = array( 00081 // Continuation byte at the start 00082 "/^$cont/", 00083 00084 // ASCII byte followed by a continuation byte 00085 "/[\\x00-\x7f]$cont/S", 00086 00087 // Illegal byte 00088 "/[\xc0\xc1\xf5-\xff]/S", 00089 00090 // Invalid 2-byte sequence, or valid one then an extra continuation byte 00091 "/[\xc2-\xdf](?!$cont$after)/S", 00092 00093 // Invalid 3-byte sequence, or valid one then an extra continuation byte 00094 "/\xe0(?![\xa0-\xbf]$cont$after)/", 00095 "/[\xe1-\xec\xee\xef](?!$cont{2}$after)/S", 00096 "/\xed(?![\x80-\x9f]$cont$after)/", 00097 00098 // Invalid 4-byte sequence, or valid one then an extra continuation byte 00099 "/\xf0(?![\x90-\xbf]$cont{2}$after)/", 00100 "/[\xf1-\xf3](?!$cont{3}$after)/S", 00101 "/\xf4(?![\x80-\x8f]$cont{2}$after)/", 00102 ); 00103 } 00104 00105 foreach ( $regexes as $regex ) { 00106 if ( preg_match( $regex, $value ) !== 0 ) { 00107 return false; 00108 } 00109 } 00110 return true; 00111 } 00112 00131 static function hungryDelimiterReplace( $startDelim, $endDelim, $replace, $subject ) { 00132 $segments = explode( $startDelim, $subject ); 00133 $output = array_shift( $segments ); 00134 foreach ( $segments as $s ) { 00135 $endDelimPos = strpos( $s, $endDelim ); 00136 if ( $endDelimPos === false ) { 00137 $output .= $startDelim . $s; 00138 } else { 00139 $output .= $replace . substr( $s, $endDelimPos + strlen( $endDelim ) ); 00140 } 00141 } 00142 return $output; 00143 } 00144 00167 static function delimiterReplaceCallback( $startDelim, $endDelim, $callback, $subject, $flags = '' ) { 00168 $inputPos = 0; 00169 $outputPos = 0; 00170 $output = ''; 00171 $foundStart = false; 00172 $encStart = preg_quote( $startDelim, '!' ); 00173 $encEnd = preg_quote( $endDelim, '!' ); 00174 $strcmp = strpos( $flags, 'i' ) === false ? 'strcmp' : 'strcasecmp'; 00175 $endLength = strlen( $endDelim ); 00176 $m = array(); 00177 00178 while ( $inputPos < strlen( $subject ) && 00179 preg_match( "!($encStart)|($encEnd)!S$flags", $subject, $m, PREG_OFFSET_CAPTURE, $inputPos ) ) 00180 { 00181 $tokenOffset = $m[0][1]; 00182 if ( $m[1][0] != '' ) { 00183 if ( $foundStart && 00184 $strcmp( $endDelim, substr( $subject, $tokenOffset, $endLength ) ) == 0 ) 00185 { 00186 # An end match is present at the same location 00187 $tokenType = 'end'; 00188 $tokenLength = $endLength; 00189 } else { 00190 $tokenType = 'start'; 00191 $tokenLength = strlen( $m[0][0] ); 00192 } 00193 } elseif ( $m[2][0] != '' ) { 00194 $tokenType = 'end'; 00195 $tokenLength = strlen( $m[0][0] ); 00196 } else { 00197 throw new MWException( 'Invalid delimiter given to ' . __METHOD__ ); 00198 } 00199 00200 if ( $tokenType == 'start' ) { 00201 # Only move the start position if we haven't already found a start 00202 # This means that START START END matches outer pair 00203 if ( !$foundStart ) { 00204 # Found start 00205 $inputPos = $tokenOffset + $tokenLength; 00206 # Write out the non-matching section 00207 $output .= substr( $subject, $outputPos, $tokenOffset - $outputPos ); 00208 $outputPos = $tokenOffset; 00209 $contentPos = $inputPos; 00210 $foundStart = true; 00211 } else { 00212 # Move the input position past the *first character* of START, 00213 # to protect against missing END when it overlaps with START 00214 $inputPos = $tokenOffset + 1; 00215 } 00216 } elseif ( $tokenType == 'end' ) { 00217 if ( $foundStart ) { 00218 # Found match 00219 $output .= call_user_func( $callback, array( 00220 substr( $subject, $outputPos, $tokenOffset + $tokenLength - $outputPos ), 00221 substr( $subject, $contentPos, $tokenOffset - $contentPos ) 00222 )); 00223 $foundStart = false; 00224 } else { 00225 # Non-matching end, write it out 00226 $output .= substr( $subject, $inputPos, $tokenOffset + $tokenLength - $outputPos ); 00227 } 00228 $inputPos = $outputPos = $tokenOffset + $tokenLength; 00229 } else { 00230 throw new MWException( 'Invalid delimiter given to ' . __METHOD__ ); 00231 } 00232 } 00233 if ( $outputPos < strlen( $subject ) ) { 00234 $output .= substr( $subject, $outputPos ); 00235 } 00236 return $output; 00237 } 00238 00252 static function delimiterReplace( $startDelim, $endDelim, $replace, $subject, $flags = '' ) { 00253 $replacer = new RegexlikeReplacer( $replace ); 00254 return self::delimiterReplaceCallback( $startDelim, $endDelim, 00255 $replacer->cb(), $subject, $flags ); 00256 } 00257 00265 static function explodeMarkup( $separator, $text ) { 00266 $placeholder = "\x00"; 00267 00268 // Remove placeholder instances 00269 $text = str_replace( $placeholder, '', $text ); 00270 00271 // Replace instances of the separator inside HTML-like tags with the placeholder 00272 $replacer = new DoubleReplacer( $separator, $placeholder ); 00273 $cleaned = StringUtils::delimiterReplaceCallback( '<', '>', $replacer->cb(), $text ); 00274 00275 // Explode, then put the replaced separators back in 00276 $items = explode( $separator, $cleaned ); 00277 foreach ( $items as $i => $str ) { 00278 $items[$i] = str_replace( $placeholder, $separator, $str ); 00279 } 00280 00281 return $items; 00282 } 00283 00291 static function escapeRegexReplacement( $string ) { 00292 $string = str_replace( '\\', '\\\\', $string ); 00293 $string = str_replace( '$', '\\$', $string ); 00294 return $string; 00295 } 00296 00304 static function explode( $separator, $subject ) { 00305 if ( substr_count( $subject, $separator ) > 1000 ) { 00306 return new ExplodeIterator( $separator, $subject ); 00307 } else { 00308 return new ArrayIterator( explode( $separator, $subject ) ); 00309 } 00310 } 00311 } 00312 00317 class Replacer { 00318 00322 function cb() { 00323 return array( &$this, 'replace' ); 00324 } 00325 } 00326 00330 class RegexlikeReplacer extends Replacer { 00331 var $r; 00332 00336 function __construct( $r ) { 00337 $this->r = $r; 00338 } 00339 00344 function replace( $matches ) { 00345 $pairs = array(); 00346 foreach ( $matches as $i => $match ) { 00347 $pairs["\$$i"] = $match; 00348 } 00349 return strtr( $this->r, $pairs ); 00350 } 00351 00352 } 00353 00357 class DoubleReplacer extends Replacer { 00358 00364 function __construct( $from, $to, $index = 0 ) { 00365 $this->from = $from; 00366 $this->to = $to; 00367 $this->index = $index; 00368 } 00369 00374 function replace( $matches ) { 00375 return str_replace( $this->from, $this->to, $matches[$this->index] ); 00376 } 00377 } 00378 00382 class HashtableReplacer extends Replacer { 00383 var $table, $index; 00384 00389 function __construct( $table, $index = 0 ) { 00390 $this->table = $table; 00391 $this->index = $index; 00392 } 00393 00398 function replace( $matches ) { 00399 return $this->table[$matches[$this->index]]; 00400 } 00401 } 00402 00407 class ReplacementArray { 00408 /*mostly private*/ var $data = false; 00409 /*mostly private*/ var $fss = false; 00410 00416 function __construct( $data = array() ) { 00417 $this->data = $data; 00418 } 00419 00423 function __sleep() { 00424 return array( 'data' ); 00425 } 00426 00427 function __wakeup() { 00428 $this->fss = false; 00429 } 00430 00435 function setArray( $data ) { 00436 $this->data = $data; 00437 $this->fss = false; 00438 } 00439 00443 function getArray() { 00444 return $this->data; 00445 } 00446 00452 function setPair( $from, $to ) { 00453 $this->data[$from] = $to; 00454 $this->fss = false; 00455 } 00456 00460 function mergeArray( $data ) { 00461 $this->data = array_merge( $this->data, $data ); 00462 $this->fss = false; 00463 } 00464 00468 function merge( $other ) { 00469 $this->data = array_merge( $this->data, $other->data ); 00470 $this->fss = false; 00471 } 00472 00476 function removePair( $from ) { 00477 unset( $this->data[$from] ); 00478 $this->fss = false; 00479 } 00480 00484 function removeArray( $data ) { 00485 foreach ( $data as $from => $to ) { 00486 $this->removePair( $from ); 00487 } 00488 $this->fss = false; 00489 } 00490 00495 function replace( $subject ) { 00496 if ( function_exists( 'fss_prep_replace' ) ) { 00497 wfProfileIn( __METHOD__ . '-fss' ); 00498 if ( $this->fss === false ) { 00499 $this->fss = fss_prep_replace( $this->data ); 00500 } 00501 $result = fss_exec_replace( $this->fss, $subject ); 00502 wfProfileOut( __METHOD__ . '-fss' ); 00503 } else { 00504 wfProfileIn( __METHOD__ . '-strtr' ); 00505 $result = strtr( $subject, $this->data ); 00506 wfProfileOut( __METHOD__ . '-strtr' ); 00507 } 00508 return $result; 00509 } 00510 } 00511 00521 class ExplodeIterator implements Iterator { 00522 // The subject string 00523 var $subject, $subjectLength; 00524 00525 // The delimiter 00526 var $delim, $delimLength; 00527 00528 // The position of the start of the line 00529 var $curPos; 00530 00531 // The position after the end of the next delimiter 00532 var $endPos; 00533 00534 // The current token 00535 var $current; 00536 00542 function __construct( $delim, $subject ) { 00543 $this->subject = $subject; 00544 $this->delim = $delim; 00545 00546 // Micro-optimisation (theoretical) 00547 $this->subjectLength = strlen( $subject ); 00548 $this->delimLength = strlen( $delim ); 00549 00550 $this->rewind(); 00551 } 00552 00553 function rewind() { 00554 $this->curPos = 0; 00555 $this->endPos = strpos( $this->subject, $this->delim ); 00556 $this->refreshCurrent(); 00557 } 00558 00559 function refreshCurrent() { 00560 if ( $this->curPos === false ) { 00561 $this->current = false; 00562 } elseif ( $this->curPos >= $this->subjectLength ) { 00563 $this->current = ''; 00564 } elseif ( $this->endPos === false ) { 00565 $this->current = substr( $this->subject, $this->curPos ); 00566 } else { 00567 $this->current = substr( $this->subject, $this->curPos, $this->endPos - $this->curPos ); 00568 } 00569 } 00570 00571 function current() { 00572 return $this->current; 00573 } 00574 00578 function key() { 00579 return $this->curPos; 00580 } 00581 00585 function next() { 00586 if ( $this->endPos === false ) { 00587 $this->curPos = false; 00588 } else { 00589 $this->curPos = $this->endPos + $this->delimLength; 00590 if ( $this->curPos >= $this->subjectLength ) { 00591 $this->endPos = false; 00592 } else { 00593 $this->endPos = strpos( $this->subject, $this->delim, $this->curPos ); 00594 } 00595 } 00596 $this->refreshCurrent(); 00597 return $this->current; 00598 } 00599 00603 function valid() { 00604 return $this->curPos !== false; 00605 } 00606 }