<?php
+/**
+ * Methods to play with strings.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ * http://www.gnu.org/copyleft/gpl.html
+ *
+ * @file
+ */
+/**
+ * A collection of static methods to play with strings.
+ */
class StringUtils {
/**
- * Perform an operation equivalent to
+ * Perform an operation equivalent to
*
* preg_replace( "!$startDelim(.*?)$endDelim!", $replace, $subject );
*
* except that it's worst-case O(N) instead of O(N^2)
*
* Compared to delimiterReplace(), this implementation is fast but memory-
- * hungry and inflexible. The memory requirements are such that I don't
+ * hungry and inflexible. The memory requirements are such that I don't
* recommend using it on anything but guaranteed small chunks of text.
+ *
+ * @param $startDelim
+ * @param $endDelim
+ * @param $replace
+ * @param $subject
+ *
+ * @return string
*/
static function hungryDelimiterReplace( $startDelim, $endDelim, $replace, $subject ) {
$segments = explode( $startDelim, $subject );
}
return $output;
}
-
+
/**
- * Perform an operation equivalent to
+ * Perform an operation equivalent to
*
* preg_replace_callback( "!$startDelim(.*)$endDelim!s$flags", $callback, $subject )
*
- * This implementation is slower than staticDelimiterReplace but uses far less
- * memory and allows regular expression delimiters.
+ * This implementation is slower than hungryDelimiterReplace but uses far less
+ * memory. The delimiters are literal strings, not regular expressions.
+ *
+ * If the start delimiter ends with an initial substring of the end delimiter,
+ * e.g. in the case of C-style comments, the behaviour differs from the model
+ * regex. In this implementation, the end must share no characters with the
+ * start, so e.g. /*\/ is not considered to be both the start and end of a
+ * comment. /*\/xy/*\/ is considered to be a single comment with contents /xy/.
*
- * @param string $flags Regular expression flags
+ * @param $startDelim String: start delimiter
+ * @param $endDelim String: end delimiter
+ * @param $callback Callback: function to call on each match
+ * @param $subject String
+ * @param $flags String: regular expression flags
+ * @throws MWException
+ * @return string
*/
static function delimiterReplaceCallback( $startDelim, $endDelim, $callback, $subject, $flags = '' ) {
$inputPos = 0;
$outputPos = 0;
$output = '';
$foundStart = false;
+ $encStart = preg_quote( $startDelim, '!' );
+ $encEnd = preg_quote( $endDelim, '!' );
+ $strcmp = strpos( $flags, 'i' ) === false ? 'strcmp' : 'strcasecmp';
+ $endLength = strlen( $endDelim );
+ $m = array();
- while ( $inputPos < strlen( $subject ) &&
- preg_match( "!($startDelim)|($endDelim)!$flags", $subject, $m, PREG_OFFSET_CAPTURE, $inputPos ) )
+ while ( $inputPos < strlen( $subject ) &&
+ preg_match( "!($encStart)|($encEnd)!S$flags", $subject, $m, PREG_OFFSET_CAPTURE, $inputPos ) )
{
+ $tokenOffset = $m[0][1];
if ( $m[1][0] != '' ) {
- # Found start
- # Write out the non-matching section
- $output .= substr( $subject, $outputPos, $m[1][1] - $outputPos );
- $foundStart = true;
- $inputPos = $contentPos = $m[1][1] + strlen( $m[1][0] );
- $outputPos = $m[1][1];
+ if ( $foundStart &&
+ $strcmp( $endDelim, substr( $subject, $tokenOffset, $endLength ) ) == 0 )
+ {
+ # An end match is present at the same location
+ $tokenType = 'end';
+ $tokenLength = $endLength;
+ } else {
+ $tokenType = 'start';
+ $tokenLength = strlen( $m[0][0] );
+ }
} elseif ( $m[2][0] != '' ) {
- # Found end
+ $tokenType = 'end';
+ $tokenLength = strlen( $m[0][0] );
+ } else {
+ throw new MWException( 'Invalid delimiter given to ' . __METHOD__ );
+ }
+
+ if ( $tokenType == 'start' ) {
+ # Only move the start position if we haven't already found a start
+ # This means that START START END matches outer pair
+ if ( !$foundStart ) {
+ # Found start
+ $inputPos = $tokenOffset + $tokenLength;
+ # Write out the non-matching section
+ $output .= substr( $subject, $outputPos, $tokenOffset - $outputPos );
+ $outputPos = $tokenOffset;
+ $contentPos = $inputPos;
+ $foundStart = true;
+ } else {
+ # Move the input position past the *first character* of START,
+ # to protect against missing END when it overlaps with START
+ $inputPos = $tokenOffset + 1;
+ }
+ } elseif ( $tokenType == 'end' ) {
if ( $foundStart ) {
# Found match
$output .= call_user_func( $callback, array(
- substr( $subject, $outputPos, $m[2][1] + strlen( $m[2][0] ) - $outputPos ),
- substr( $subject, $contentPos, $m[2][1] - $contentPos )
+ substr( $subject, $outputPos, $tokenOffset + $tokenLength - $outputPos ),
+ substr( $subject, $contentPos, $tokenOffset - $contentPos )
));
$foundStart = false;
} else {
# Non-matching end, write it out
- $output .= substr( $subject, $inputPos, $m[2][1] + strlen( $m[2][0] ) - $outputPos );
+ $output .= substr( $subject, $inputPos, $tokenOffset + $tokenLength - $outputPos );
}
- $inputPos = $outputPos = $m[2][1] + strlen( $m[2][0] );
+ $inputPos = $outputPos = $tokenOffset + $tokenLength;
} else {
throw new MWException( 'Invalid delimiter given to ' . __METHOD__ );
}
return $output;
}
- /*
- * Perform an operation equivalent to
+ /**
+ * Perform an operation equivalent to
*
* preg_replace( "!$startDelim(.*)$endDelim!$flags", $replace, $subject )
*
- * @param string $startDelim Start delimiter regular expression
- * @param string $endDelim End delimiter regular expression
- * @param string $replace Replacement string. May contain $1, which will be
- * replaced by the text between the delimiters
- * @param string $subject String to search
- * @return string The string with the matches replaced
+ * @param $startDelim String: start delimiter regular expression
+ * @param $endDelim String: end delimiter regular expression
+ * @param $replace String: replacement string. May contain $1, which will be
+ * replaced by the text between the delimiters
+ * @param $subject String to search
+ * @param $flags String: regular expression flags
+ * @return String: The string with the matches replaced
*/
static function delimiterReplace( $startDelim, $endDelim, $replace, $subject, $flags = '' ) {
$replacer = new RegexlikeReplacer( $replace );
/**
* More or less "markup-safe" explode()
* Ignores any instances of the separator inside <...>
- * @param string $separator
- * @param string $text
+ * @param $separator String
+ * @param $text String
* @return array
*/
static function explodeMarkup( $separator, $text ) {
$placeholder = "\x00";
-
+
// Remove placeholder instances
$text = str_replace( $placeholder, '', $text );
-
+
// Replace instances of the separator inside HTML-like tags with the placeholder
$replacer = new DoubleReplacer( $separator, $placeholder );
$cleaned = StringUtils::delimiterReplaceCallback( '<', '>', $replacer->cb(), $text );
foreach( $items as $i => $str ) {
$items[$i] = str_replace( $placeholder, $separator, $str );
}
-
+
return $items;
}
* Escape a string to make it suitable for inclusion in a preg_replace()
* replacement parameter.
*
- * @param string $string
- * @return string
+ * @param $string String
+ * @return String
*/
static function escapeRegexReplacement( $string ) {
$string = str_replace( '\\', '\\\\', $string );
$string = str_replace( '$', '\\$', $string );
return $string;
}
+
+ /**
+ * Workalike for explode() with limited memory usage.
+ * Returns an Iterator
+ * @param $separator
+ * @param $subject
+ * @return ArrayIterator|\ExplodeIterator
+ */
+ static function explode( $separator, $subject ) {
+ if ( substr_count( $subject, $separator ) > 1000 ) {
+ return new ExplodeIterator( $separator, $subject );
+ } else {
+ return new ArrayIterator( explode( $separator, $subject ) );
+ }
+ }
}
/**
- * Base class for "replacers", objects used in preg_replace_callback() and
+ * Base class for "replacers", objects used in preg_replace_callback() and
* StringUtils::delimiterReplaceCallback()
*/
class Replacer {
+
+ /**
+ * @return array
+ */
function cb() {
return array( &$this, 'replace' );
}
*/
class RegexlikeReplacer extends Replacer {
var $r;
+
+ /**
+ * @param $r string
+ */
function __construct( $r ) {
$this->r = $r;
}
+ /**
+ * @param $matches array
+ * @return string
+ */
function replace( $matches ) {
$pairs = array();
foreach ( $matches as $i => $match ) {
* Class to perform secondary replacement within each replacement string
*/
class DoubleReplacer extends Replacer {
+
+ /**
+ * @param $from
+ * @param $to
+ * @param $index int
+ */
function __construct( $from, $to, $index = 0 ) {
$this->from = $from;
$this->to = $to;
$this->index = $index;
}
-
+
+ /**
+ * @param $matches array
+ * @return mixed
+ */
function replace( $matches ) {
return str_replace( $this->from, $this->to, $matches[$this->index] );
}
class HashtableReplacer extends Replacer {
var $table, $index;
+ /**
+ * @param $table
+ * @param $index int
+ */
function __construct( $table, $index = 0 ) {
$this->table = $table;
$this->index = $index;
}
+ /**
+ * @param $matches array
+ * @return mixed
+ */
function replace( $matches ) {
return $this->table[$matches[$this->index]];
}
/**
* Create an object with the specified replacement array
* The array should have the same form as the replacement array for strtr()
+ * @param array $data
*/
function __construct( $data = array() ) {
$this->data = $data;
}
+ /**
+ * @return array
+ */
function __sleep() {
return array( 'data' );
}
$this->fss = false;
}
+ /**
+ * @return array|bool
+ */
function getArray() {
return $this->data;
}
/**
* Set an element of the replacement array
+ * @param $from string
+ * @param $to stromg
*/
function setPair( $from, $to ) {
$this->data[$from] = $to;
$this->fss = false;
}
+ /**
+ * @param $data array
+ */
function mergeArray( $data ) {
$this->data = array_merge( $this->data, $data );
$this->fss = false;
}
+ /**
+ * @param $other
+ */
function merge( $other ) {
$this->data = array_merge( $this->data, $other->data );
$this->fss = false;
}
+ /**
+ * @param $from string
+ */
+ function removePair( $from ) {
+ unset($this->data[$from]);
+ $this->fss = false;
+ }
+
+ /**
+ * @param $data array
+ */
+ function removeArray( $data ) {
+ foreach( $data as $from => $to ) {
+ $this->removePair( $from );
+ }
+ $this->fss = false;
+ }
+
+ /**
+ * @param $subject string
+ * @return string
+ */
function replace( $subject ) {
if ( function_exists( 'fss_prep_replace' ) ) {
wfProfileIn( __METHOD__.'-fss' );
}
}
-?>
+/**
+ * An iterator which works exactly like:
+ *
+ * foreach ( explode( $delim, $s ) as $element ) {
+ * ...
+ * }
+ *
+ * Except it doesn't use 193 byte per element
+ */
+class ExplodeIterator implements Iterator {
+ // The subject string
+ var $subject, $subjectLength;
+
+ // The delimiter
+ var $delim, $delimLength;
+
+ // The position of the start of the line
+ var $curPos;
+
+ // The position after the end of the next delimiter
+ var $endPos;
+
+ // The current token
+ var $current;
+
+ /**
+ * Construct a DelimIterator
+ * @param $delim string
+ * @param $s string
+ */
+ function __construct( $delim, $s ) {
+ $this->subject = $s;
+ $this->delim = $delim;
+
+ // Micro-optimisation (theoretical)
+ $this->subjectLength = strlen( $s );
+ $this->delimLength = strlen( $delim );
+
+ $this->rewind();
+ }
+
+ function rewind() {
+ $this->curPos = 0;
+ $this->endPos = strpos( $this->subject, $this->delim );
+ $this->refreshCurrent();
+ }
+
+ function refreshCurrent() {
+ if ( $this->curPos === false ) {
+ $this->current = false;
+ } elseif ( $this->curPos >= $this->subjectLength ) {
+ $this->current = '';
+ } elseif ( $this->endPos === false ) {
+ $this->current = substr( $this->subject, $this->curPos );
+ } else {
+ $this->current = substr( $this->subject, $this->curPos, $this->endPos - $this->curPos );
+ }
+ }
+
+ function current() {
+ return $this->current;
+ }
+
+ function key() {
+ return $this->curPos;
+ }
+
+ /**
+ * @return string
+ */
+ function next() {
+ if ( $this->endPos === false ) {
+ $this->curPos = false;
+ } else {
+ $this->curPos = $this->endPos + $this->delimLength;
+ if ( $this->curPos >= $this->subjectLength ) {
+ $this->endPos = false;
+ } else {
+ $this->endPos = strpos( $this->subject, $this->delim, $this->curPos );
+ }
+ }
+ $this->refreshCurrent();
+ return $this->current;
+ }
+
+ /**
+ * @return bool
+ */
+ function valid() {
+ return $this->curPos !== false;
+ }
+}