Issue #306 - The multibute string library: Logic upgrades, optimizations for speed, part 2.

main/inc/lib/multibyte_string_functions.lib.php

@@ -66,7 +66,6 @@ function api_byte_count($string) {
 		return mb_strlen($string, '8bit');
 	}
 	return strlen($string);
-
 	// For PHP6 this function probably will contain:
 	//return strlen((binary)$string);
 }
@@ -88,30 +87,32 @@ function api_byte_count($string) {
  * @link http://php.net/manual/en/function.mb-convert-encoding
  */
 function api_convert_encoding($string, $to_encoding, $from_encoding = null) {
+	static $equal_encodings = array();
 	if (empty($from_encoding)) {
 		$from_encoding = api_mb_internal_encoding();
 	}
-	if (api_equal_encodings($to_encoding, $from_encoding)) {
+	if (!isset($equal_encodings[$to_encoding][$from_encoding])) {
-		// When conversion is not needed, the string is returned directly, without validation.
+		$equal_encodings[$to_encoding][$from_encoding] = api_equal_encodings($to_encoding, $from_encoding);
-		return $string;
+	}
+	if ($equal_encodings[$to_encoding][$from_encoding]) {
+		return $string; // When conversion is not needed, the string is returned directly, without validation.
 	}
-	elseif (api_mb_supports($to_encoding) && api_mb_supports($from_encoding)) {
+	if (api_mb_supports($to_encoding) && api_mb_supports($from_encoding)) {
 		return @mb_convert_encoding($string, $to_encoding, $from_encoding);
 	}
-	elseif (api_iconv_supports($to_encoding) && api_iconv_supports($from_encoding)) {
+	if (api_iconv_supports($to_encoding) && api_iconv_supports($from_encoding)) {
 		return @iconv($from_encoding, $to_encoding, $string);
 	}
-	elseif (api_is_utf8($to_encoding) && api_is_latin1($from_encoding, true)) {
+	if (api_is_utf8($to_encoding) && api_is_latin1($from_encoding, true)) {
 		return utf8_encode($string);
 	}
-	elseif (api_is_latin1($to_encoding, true) && api_is_utf8($from_encoding)) {
+	if (api_is_latin1($to_encoding, true) && api_is_utf8($from_encoding)) {
 		return utf8_decode($string);
 	}
-	elseif (_api_convert_encoding_supports($to_encoding) && _api_convert_encoding_supports($from_encoding)) {
+	if (_api_convert_encoding_supports($to_encoding) && _api_convert_encoding_supports($from_encoding)) {
 		return _api_convert_encoding($string, $to_encoding, $from_encoding);
 	}
-	// Here the function gives up.
+	return $string; // Here the function gives up.
-	return $string;
 }
 
 /**
@@ -127,23 +128,21 @@ function api_utf8_encode($string, $from_encoding = null) {
 		$from_encoding = api_mb_internal_encoding();
 	}
 	if (api_is_utf8($from_encoding)) {
-		// When conversion is not needed, the string is returned directly, without validation.
+		return $string; // When conversion is not needed, the string is returned directly, without validation.
-		return $string;
 	}
-	elseif (api_mb_supports($from_encoding)) {
+	if (api_mb_supports($from_encoding)) {
 		return @mb_convert_encoding($string, 'UTF-8', $from_encoding);
 	}
-	elseif (api_iconv_supports($from_encoding)) {
+	if (api_iconv_supports($from_encoding)) {
 		return @iconv($from_encoding, 'UTF-8', $string);
 	}
-	elseif (api_is_latin1($from_encoding, true)) {
+	if (api_is_latin1($from_encoding, true)) {
 		return utf8_encode($string);
 	}
-	elseif (_api_convert_encoding_supports($from_encoding)) {
+	if (_api_convert_encoding_supports($from_encoding)) {
 		return _api_convert_encoding($string, 'UTF-8', $from_encoding);
 	}
-	// Here the function gives up.
+	return $string; // Here the function gives up.
-	return $string;
 }
 
 /**
@@ -159,23 +158,21 @@ function api_utf8_decode($string, $to_encoding = null) {
 		$to_encoding = api_mb_internal_encoding();
 	}
 	if (api_is_utf8($to_encoding)) {
-		// When conversion is not needed, the string is returned directly, without validation.
+		return $string; // When conversion is not needed, the string is returned directly, without validation.
-		return $string;
 	}
-	elseif (api_mb_supports($to_encoding)) {
+	if (api_mb_supports($to_encoding)) {
 		return @mb_convert_encoding($string, $to_encoding, 'UTF-8');
 	}
-	elseif (api_iconv_supports($to_encoding)) {
+	if (api_iconv_supports($to_encoding)) {
 		return @iconv('UTF-8', $to_encoding, $string);
 	}
-	elseif (api_is_latin1($to_encoding, true)) {
+	if (api_is_latin1($to_encoding, true)) {
 		return utf8_decode($string);
 	}
-	elseif (_api_convert_encoding_supports($to_encoding)) {
+	if (_api_convert_encoding_supports($to_encoding)) {
 		return _api_convert_encoding($string, $to_encoding, 'UTF-8');
 	}
-	// Here the function gives up.
+	return $string; // Here the function gives up.
-	return $string;
 }
 
 /**
@@ -288,10 +285,17 @@ function api_html_entity_decode($string, $quote_style = ENT_COMPAT, $encoding =
 		return html_entity_decode($string, $quote_style, $encoding);
 	}
 	if (api_is_encoding_supported($encoding)) {
-		return api_utf8_decode(html_entity_decode(api_convert_encoding($string, 'UTF-8', $encoding), $quote_style, 'UTF-8'), $encoding);
+		if (!api_is_utf8($encoding)) {
+			$string = api_utf8_encode($string, $encoding);
+		}
+		$string = html_entity_decode($string, $quote_style, 'UTF-8');
+		if (!api_is_utf8($encoding)) {
+			return api_utf8_decode($string, $encoding);
 		}
 		return $string;
 	}
+	return $string; // Here the function guves up.
+}
 
 /**
  * This function encodes (conditionally) a given string to UTF-8 if XmlHttp-request has been detected.
@@ -301,7 +305,12 @@ function api_html_entity_decode($string, $quote_style = ENT_COMPAT, $encoding =
  */
 function api_xml_http_response_encode($string, $from_encoding = null) {
 	if (isset($_SERVER['HTTP_X_REQUESTED_WITH']) && strtolower($_SERVER['HTTP_X_REQUESTED_WITH']) == 'xmlhttprequest') {
-		return api_convert_encoding($string, 'UTF-8', $from_encoding);
+		if (empty($from_encoding)) {
+			$from_encoding = api_mb_internal_encoding();
+		}
+		if (!api_is_utf8($from_encoding)) {
+			return api_utf8_encode($string, $from_encoding);
+		}
 	}
 	return $string;
 }
@@ -464,7 +473,11 @@ function api_stripos($haystack, $needle, $offset = 0, $encoding = null) {
 	}
 	elseif (api_is_encoding_supported($encoding)) {
 		if (MBSTRING_INSTALLED) {
-			return @mb_stripos(api_utf8_encode($haystack, $encoding), api_utf8_encode($needle, $encoding), $offset, 'UTF-8');
+			if (!api_is_utf8($encoding)) {
+				$haystack = api_utf8_encode($haystack, $encoding);
+				$needle = api_utf8_encode($needle, $encoding);
+			}
+			return @mb_stripos($haystack, $needle, $offset, 'UTF-8');
 		}
 		return api_strpos(api_strtolower($haystack, $encoding), api_strtolower($needle, $encoding), $offset, $encoding);
 	}
@@ -503,12 +516,19 @@ function api_stristr($haystack, $needle, $before_needle = false, $encoding = nul
 	}
 	elseif (api_is_encoding_supported($encoding)) {
 		if (MBSTRING_INSTALLED) {
-			$result = @mb_stristr(api_utf8_encode($haystack, $encoding), api_utf8_encode($needle, $encoding), $before_needle, 'UTF-8');
+			if (!api_is_utf8($encoding)) {
+				$haystack = api_utf8_encode($haystack, $encoding);
+				$needle = api_utf8_encode($needle, $encoding);
+			}
+			$result = @mb_stristr($haystack, $needle, $before_needle, 'UTF-8');
 			if ($result === false) {
 				return false;
 			}
+			if (!api_is_utf8($encoding)) {
 				return api_utf8_decode($result, $encoding);
 			}
+			return $result;
+		}
 		$result = api_strstr(api_strtolower($haystack, $encoding), api_strtolower($needle, $encoding), $before_needle, $encoding);
 		if ($result === false) {
 			return false;
@@ -545,13 +565,13 @@ function api_strlen($string, $encoding = null) {
 	if (_api_is_single_byte_encoding($encoding)) {
 		return strlen($string);
 	}
-	elseif (api_mb_supports($encoding)) {
+	if (api_mb_supports($encoding)) {
 		return @mb_strlen($string, $encoding);
 	}
-	elseif (api_iconv_supports($encoding)) {
+	if (api_iconv_supports($encoding)) {
 		return @iconv_strlen($string, $encoding);
 	}
-	elseif (api_is_utf8($encoding)) {
+	if (api_is_utf8($encoding)) {
     	return api_byte_count(preg_replace("/[\x80-\xBF]/", '', $string));
 	}
 	return strlen($string);
@@ -580,13 +600,13 @@ function api_strpos($haystack, $needle, $offset = 0, $encoding = null) {
 		return @mb_strpos($haystack, $needle, $offset, $encoding);
 	}
 	elseif (api_is_encoding_supported($encoding)) {
-		if (MBSTRING_INSTALLED) {
-			return @mb_strpos(api_utf8_encode($haystack, $encoding), api_utf8_encode($needle, $encoding), $offset, 'UTF-8');
-		}
 		if (!api_is_utf8($encoding)) {
 			$haystack = api_utf8_encode($haystack, $encoding);
 			$needle = api_utf8_encode($needle, $encoding);
 		}
+		if (MBSTRING_INSTALLED) {
+			return @mb_strpos($haystack, $needle, $offset, 'UTF-8');
+		}
 		if (empty($offset)) {
 			$haystack = explode($needle, $haystack, 2);
 			if (count($haystack) > 1) {
@@ -644,12 +664,19 @@ function api_strrchr($haystack, $needle, $before_needle = false, $encoding = nul
 		return @mb_strrchr($haystack, $needle, $before_needle, $encoding);
 	}
 	elseif (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
-		$result = @mb_strrchr(api_utf8_encode($haystack, $encoding), api_utf8_encode($needle, $encoding), $before_needle, 'UTF-8');
+		if (!api_is_utf8($encoding)) {
+			$haystack = api_utf8_encode($haystack, $encoding);
+			$needle = api_utf8_encode($needle, $encoding);
+		}
+		$result = @mb_strrchr($haystack, $needle, $before_needle, 'UTF-8');
 		if ($result === false) {
 			return false;
 		}
+		if (!api_is_utf8($encoding)) {
 			return api_utf8_decode($result, $encoding);
 		}
+		return $result;
+	}
 	if (!$before_needle) {
 		return strrchr($haystack, $needle);
 	}
@@ -707,14 +734,14 @@ function api_strrpos($haystack, $needle, $offset = 0, $encoding = null) {
 		return @mb_strrpos($haystack, $needle, $offset, $encoding);
 	}
 	elseif (api_is_encoding_supported($encoding)) {
-		if (MBSTRING_INSTALLED) {
-			return @mb_strrpos(api_utf8_encode($haystack, $encoding), api_utf8_encode($needle, $encoding), $offset, 'UTF-8');
-		}
-		// This branch (this fragment of code) is an adaptation from the CakePHP(tm) Project, http://www.cakefoundation.org
 		if (!api_is_utf8($encoding)) {
 			$haystack = api_utf8_encode($haystack, $encoding);
 			$needle = api_utf8_encode($needle, $encoding);
 		}
+		if (MBSTRING_INSTALLED) {
+			return @mb_strrpos($haystack, $needle, $offset, 'UTF-8');
+		}
+		// This branch (this fragment of code) is an adaptation from the CakePHP(tm) Project, http://www.cakefoundation.org
 		$found = false;
 		$haystack = _api_utf8_to_unicode($haystack);
 		$haystack_count = count($haystack);
@@ -792,12 +819,18 @@ function api_strstr($haystack, $needle, $before_needle = false, $encoding = null
 		return @mb_strstr($haystack, $needle, $before_needle, $encoding);
 	}
 	elseif (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
-		$result = @mb_strstr(api_utf8_encode($haystack, $encoding), api_utf8_encode($needle, $encoding), $before_needle, 'UTF-8');
+		if (!api_is_utf8($encoding)) {
+			$haystack = api_utf8_encode($haystack, $encoding);
+			$needle = api_utf8_encode($needle, $encoding);
+		}
+		$result = @mb_strstr($haystack, $needle, $before_needle, 'UTF-8');
 		if ($result !== false) {
+			if (!api_is_utf8($encoding)) {
 				return api_utf8_decode($result, $encoding);
-		} else {
-			return false;
 			}
+			return $result;
+		}
+		return false;
 	}
 	// Adding the missing parameter $before_needle to the original function strstr(), PHP_VERSION < 5.3
 	if (!$before_needle) {
@@ -830,13 +863,13 @@ function api_strtolower($string, $encoding = null) {
 		return @mb_strtolower($string, $encoding);
 	}
 	elseif (api_is_encoding_supported($encoding)) {
-		if (MBSTRING_INSTALLED) {
-			return api_utf8_decode(@mb_strtolower(api_utf8_encode($string, $encoding), 'UTF-8'), $encoding);
-		}
-		// This branch (this fragment of code) is an adaptation from the CakePHP(tm) Project, http://www.cakefoundation.org
 		if (!api_is_utf8($encoding)) {
 			$string = api_utf8_encode($string, $encoding);
 		}
+		if (MBSTRING_INSTALLED) {
+			$string = @mb_strtolower($string, 'UTF-8');
+		} else {
+			// This branch (this fragment of code) is an adaptation from the CakePHP(tm) Project, http://www.cakefoundation.org
 			$codepoints = _api_utf8_to_unicode($string);
 			$length = count($codepoints);
 			$matched = false;
@@ -869,8 +902,9 @@ function api_strtolower($string, $encoding = null) {
 				}
 			}
 			$string = _api_utf8_from_unicode($result);
+		}
 		if (!api_is_utf8($encoding)) {
-			$string = api_utf8_decode($string, $encoding);
+			return api_utf8_decode($string, $encoding);
 		}
 		return $string;
 	}
@@ -894,13 +928,13 @@ function api_strtoupper($string, $encoding = null) {
 		return @mb_strtoupper($string, $encoding);
 	}
 	elseif (api_is_encoding_supported($encoding)) {
-		if (MBSTRING_INSTALLED) {
-			return api_utf8_decode(@mb_strtoupper(api_utf8_encode($string, $encoding), 'UTF-8'), $encoding);
-		}
-		// This branch (this fragment of code) is an adaptation from the CakePHP(tm) Project, http://www.cakefoundation.org
 		if (!api_is_utf8($encoding)) {
 			$string = api_utf8_encode($string, $encoding);
 		}
+		if (MBSTRING_INSTALLED) {
+			$string = @mb_strtoupper($string, 'UTF-8');
+		} else {
+			// This branch (this fragment of code) is an adaptation from the CakePHP(tm) Project, http://www.cakefoundation.org
 			$codepoints = _api_utf8_to_unicode($string);
 			$length = count($codepoints);
 			$matched = false;
@@ -971,8 +1005,9 @@ function api_strtoupper($string, $encoding = null) {
 				}
 			}
 			$string = _api_utf8_from_unicode($result);
+		}
 		if (!api_is_utf8($encoding)) {
-			$string = api_utf8_decode($string, $encoding);
+			return api_utf8_decode($string, $encoding);
 		}
 		return $string;
 	}
@@ -989,6 +1024,7 @@ function api_strtoupper($string, $encoding = null) {
  * This function is aimed at replacing the function strtr() for human-language strings.
  * @link http://php.net/manual/en/function.strtr
  * TODO: To be revised and tested. Probably this function will not be needed.
+ * TODO: This function will be removed. It is not needed. 21-AUG-2009.
  */
 function api_strtr($string, $from, $to = null, $encoding = null) {
 	if (empty($string)) {
@@ -1060,13 +1096,13 @@ function api_substr($string, $start, $length = null, $encoding = null) {
 		return @mb_substr($string, $start, $length, $encoding);
 	}
 	elseif (api_is_encoding_supported($encoding)) {
-		if (MBSTRING_INSTALLED) {
-			return api_utf8_decode(@mb_substr(api_utf8_encode($string, $encoding), $start, $length, 'UTF-8'), $encoding);
-		}
-		// The following branch of code is from the Drupal CMS, see the function drupal_substr().
 		if (!api_is_utf8($encoding)) {
 			$string = api_utf8_encode($string, $encoding);
 		}
+		if (MBSTRING_INSTALLED) {
+			$string = @mb_substr($string, $start, $length, 'UTF-8');
+		} else {
+			// The following branch of code is from the Drupal CMS, see the function drupal_substr().
 			$strlen = api_byte_count($string);
 			// Find the starting byte offset
 			$bytes = 0;
@@ -1128,6 +1164,7 @@ function api_substr($string, $start, $length = null, $encoding = null) {
 			}
 			$iend = $bytes;
 			$string = substr($string, $istart, max(0, $iend - $istart + 1));
+		}
 		if (!api_is_utf8($encoding)) {
 			$string = api_utf8_decode($string, $encoding);
 		}
@@ -1159,7 +1196,12 @@ function api_substr_replace($string, $replacement, $start, $length = null, $enco
 	if (empty($encoding)) {
 		$encoding = api_mb_internal_encoding();
 	}
-	if (api_is_encoding_supported($encoding) && !_api_is_single_byte_encoding($encoding)) {
+	if (_api_is_single_byte_encoding($encoding)) {
+		return substr_replace($string, $replacement, $start, $length);
+	}
+	if (api_is_encoding_supported($encoding)) {
+		// This fragment (branch) of code is adaptation of a published proposition:
+		// http://php.net/manual/en/function.substr-replace.php#90146
 		$string_length = api_strlen($string, $encoding);
 		if ($start < 0) {
 			$start = max(0, $string_length + $start);
@@ -1214,8 +1256,24 @@ function api_ucwords($string, $encoding = null) {
     if (api_mb_supports($encoding)) {
 		return @mb_convert_case($string, MB_CASE_TITLE, $encoding);
 	}
-	elseif (MBSTRING_INSTALLED && api_iconv_supports($encoding)) {
+	if (api_is_encoding_supported($encoding)) {
-		return api_utf8_decode(@mb_convert_case(api_utf8_encode($string, $encoding), MB_CASE_TITLE, 'UTF-8'), $encoding);
+		if (!api_is_utf8($encoding)) {
+			$string = api_utf8_encode($string, $encoding);
+		}
+		if (MBSTRING_INSTALLED) {
+			$string = @mb_convert_case($string, MB_CASE_TITLE, 'UTF-8');
+		} else {
+			// The following fragment (branch) of code is based on the function utf8_ucwords() by Harry Fuecks
+			// See http://dev.splitbrain.org/view/darcs/dokuwiki/inc/utf8.php
+			// Note: [\x0c\x09\x0b\x0a\x0d\x20] matches - form feeds, horizontal tabs, vertical tabs, linefeeds and carriage returns.
+			// This corresponds to the definition of a "word" defined at http://www.php.net/ucwords
+			$pattern = '/(^|([\x0c\x09\x0b\x0a\x0d\x20]+))([^\x0c\x09\x0b\x0a\x0d\x20]{1})[^\x0c\x09\x0b\x0a\x0d\x20]*/u';
+			$string = preg_replace_callback($pattern, '_api_utf8_ucwords_callback', $string);
+		}
+		if (!api_is_utf8($encoding)) {
+			return api_utf8_decode($string, $encoding);
+		}
+		return $string;
 	}
 	return ucwords($string);
 }
@@ -1228,6 +1286,8 @@ function api_ucwords($string, $encoding = null) {
  */
 
 /**
+ * Note: Try to avoid using this function. Use api_preg_match() with Perl-compatible regular expression syntax.
+ *
  * Executes a regular expression match with extended multibyte support.
  * By default this function uses the platform character set.
  * @param string $pattern			The regular expression pattern.
@@ -1244,12 +1304,10 @@ function api_ereg($pattern, $string, & $regs = null) {
 	if (api_mb_supports($encoding)) {
 		if ($count < 3) {
 			return @mb_ereg($pattern, $string);
-		} else {
-			$result = @mb_ereg($pattern, $string, $regs);
-			return $result;
 		}
+		return @mb_ereg($pattern, $string, $regs);
 	}
-	elseif (MBSTRING_INSTALLED && api_iconv_supports($encoding)) {
+	if (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
 		global $_api_encoding;
 		$_api_encoding = $encoding;
 		api_mb_regex_encoding('UTF-8');
@@ -1261,16 +1319,16 @@ function api_ereg($pattern, $string, & $regs = null) {
 		}
 		api_mb_regex_encoding($encoding);
 		return $result;
-	} else {
+	}
 	if ($count < 3) {
 		return ereg($pattern, $string);
-		} else {
-			return ereg($pattern, $string, $regs);
-		}
 	}
+	return ereg($pattern, $string, $regs);
 }
 
 /**
+ * Note: Try to avoid using this function. Use api_preg_replace() with Perl-compatible regular expression syntax.
+ *
  * Scans string for matches to pattern, then replaces the matched text with replacement, with extended multibyte support.
  * By default this function uses the platform character set.
  * @param string $pattern				The regular expression pattern.
@@ -1292,13 +1350,11 @@ function api_ereg_replace($pattern, $replacement, $string, $option = null) {
 	if (api_mb_supports($encoding)) {
 		if (is_null($option)) {
 			return @mb_ereg_replace($pattern, $replacement, $string);
-		} else {
-			return @mb_ereg_replace($pattern, $replacement, $string, $option);
 		}
+		return @mb_ereg_replace($pattern, $replacement, $string, $option);
 	}
-	elseif (MBSTRING_INSTALLED && api_iconv_supports($encoding)) {
+	if (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
 		api_mb_regex_encoding('UTF-8');
-
 		if (is_null($option)) {
 			$result = api_utf8_decode(@mb_ereg_replace(api_utf8_encode($pattern, $encoding), api_utf8_encode($replacement, $encoding), api_utf8_encode($string, $encoding)), $encoding);
 		} else {
@@ -1306,12 +1362,13 @@ function api_ereg_replace($pattern, $replacement, $string, $option = null) {
 		}
 		api_mb_regex_encoding($encoding);
 		return $result;
-	} else {
-		return ereg_replace($pattern, $replacement, $string);
 	}
+	return ereg_replace($pattern, $replacement, $string);
 }
 
 /**
+ * Note: Try to avoid using this function. Use api_preg_match() with Perl-compatible regular expression syntax.
+ *
  * Executes a regular expression match, ignoring case, with extended multibyte support.
  * By default this function uses the platform character set.
  * @param string $pattern			The regular expression pattern.
@@ -1328,11 +1385,10 @@ function api_eregi($pattern, $string, & $regs = null) {
 	if (api_mb_supports($encoding)) {
 		if ($count < 3) {
 			return @mb_eregi($pattern, $string);
-		} else {
-			return @mb_eregi($pattern, $string, $regs);
 		}
+		return @mb_eregi($pattern, $string, $regs);
 	}
-	elseif (MBSTRING_INSTALLED && api_iconv_supports($encoding)) {
+	if (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
 		global $_api_encoding;
 		$_api_encoding = $encoding;
 		api_mb_regex_encoding('UTF-8');
@@ -1344,16 +1400,16 @@ function api_eregi($pattern, $string, & $regs = null) {
 		}
 		api_mb_regex_encoding($encoding);
 		return $result;
-	} else {
+	}
 	if ($count < 3) {
 		return eregi($pattern, $string);
-		} else {
-			return eregi($pattern, $string, $regs);
-		}
 	}
+	return eregi($pattern, $string, $regs);
 }
 
 /**
+ * Note: Try to avoid using this function. Use api_preg_replace() with Perl-compatible regular expression syntax.
+ *
  * Scans string for matches to pattern, then replaces the matched text with replacement, ignoring case, with extended multibyte support.
  * By default this function uses the platform character set.
  * @param string $pattern				The regular expression pattern.
@@ -1375,11 +1431,10 @@ function api_eregi_replace($pattern, $replacement, $string, $option = null) {
 	if (api_mb_supports($encoding)) {
 		if (is_null($option)) {
 			return @mb_eregi_replace($pattern, $replacement, $string);
-		} else {
-			return @mb_eregi_replace($pattern, $replacement, $string, $option);
 		}
+		return @mb_eregi_replace($pattern, $replacement, $string, $option);
 	}
-	elseif (MBSTRING_INSTALLED && api_iconv_supports($encoding)) {
+	if (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
 		api_mb_regex_encoding('UTF-8');
 		if (is_null($option)) {
 			$result = api_utf8_decode(@mb_eregi_replace(api_utf8_encode($pattern, $encoding), api_utf8_encode($replacement, $encoding), api_utf8_encode($string, $encoding)), $encoding);
@@ -1388,9 +1443,8 @@ function api_eregi_replace($pattern, $replacement, $string, $option = null) {
 		}
 		api_mb_regex_encoding($encoding);
 		return $result;
-	} else {
-		return eregi_replace($pattern, $replacement, $string);
 	}
+	return eregi_replace($pattern, $replacement, $string);
 }
 
 /**
@@ -1452,12 +1506,13 @@ function api_preg_replace($pattern, $replacement, $subject, $limit= -1, &$count
 	if (empty($encoding)){
 		$encoding = api_get_system_encoding();
 	}
+	$is_utf8 = api_is_utf8($encoding);
 	if (is_array($pattern)) {
 		foreach ($pattern as &$p) {
-			$p = api_is_utf8($encoding) ? $p.'u' : $p;
+			$p = $is_utf8 ? $p.'u' : $p;
 		}
 	} else {
-		$pattern = api_is_utf8($encoding) ? $pattern.'u' : $pattern;
+		$pattern = $is_utf8 ? $pattern.'u' : $pattern;
 	}
 	return preg_replace($pattern, $replacement, $subject, $limit, $count);
 }
@@ -1509,6 +1564,8 @@ function api_preg_split($pattern, $subject, $limit = -1, $flags = 0, $encoding =
 }
 
 /**
+ * Note: Try to avoid using this function. Use api_preg_split() with Perl-compatible regular expression syntax.
+ *
  * Splits a multibyte string using regular expression pattern and returns the result as an array.
  * By default this function uses the platform character set.
  * @param string $pattern			The regular expression pattern.
@@ -1524,11 +1581,10 @@ function api_split($pattern, $string, $limit = null) {
 	if (api_mb_supports($encoding)) {
 		if (is_null($limit)) {
 			return @mb_split($pattern, $string);
-		} else {
-			return @mb_split($pattern, $string, $limit);
 		}
+		return @mb_split($pattern, $string, $limit);
 	}
-	elseif (MBSTRING_INSTALLED && api_iconv_supports($encoding)) {
+	if (MBSTRING_INSTALLED && api_is_encoding_supported($encoding)) {
 		global $_api_encoding;
 		$_api_encoding = $encoding;
 		api_mb_regex_encoding('UTF-8');
@@ -1540,13 +1596,11 @@ function api_split($pattern, $string, $limit = null) {
 		$result = _api_array_utf8_decode($result);
 		api_mb_regex_encoding($encoding);
 		return $result;
-	} else {
+	}
 	if (is_null($limit)) {
 		return split($pattern, $string);
-		} else {
-			return split($pattern, $string, $limit);
-		}
 	}
+	return split($pattern, $string, $limit);
 }
 
 /**
@@ -2265,12 +2319,10 @@ function api_get_non_utf8_encoding($language = null) {
 	if (is_array($encodings[$language])) {
 		if (!empty($encodings[$language][0])) {
 			return $encodings[$language][0];
-		} else {
-			return 'ISO-8859-15';
 		}
-	} else {
 		return 'ISO-8859-15';
 	}
+	return 'ISO-8859-15';
 }
 
 /**
@@ -2375,11 +2427,10 @@ function api_refine_encoding_id($encoding) {
  * @return bool							Returns TRUE if the encodings are equal, FALSE otherwise.
  */
 function api_equal_encodings($encoding1, $encoding2) {
-	$is_array_encoding1 = is_array($encoding1);
 	$is_array_encoding2 = is_array($encoding2);
 	$encoding1 = api_refine_encoding_id($encoding1);
 	$encoding2 = api_refine_encoding_id($encoding2);
-	if (!$is_array_encoding1 && !$is_array_encoding2) {
+	if (!is_array($encoding1) && !$is_array_encoding2) {
 		return $encoding1 == $encoding2;
 	}
 	if ($is_array_encoding2) {
@@ -2587,13 +2638,10 @@ function api_iconv_set_encoding($type, $encoding = null) {
 				if(@iconv_set_encoding($type, $encoding)) {
 					$iconv_internal_encoding = $encoding;
 					return true;
-				} else {
-					return false;
 				}
-			} else {
 				return false;
 			}
-			break;
+			return false;
 		case 'iconv_input_encoding':
 			if (empty($encoding)) {
 				if (is_null($iconv_input_encoding)) {
@@ -2605,13 +2653,10 @@ function api_iconv_set_encoding($type, $encoding = null) {
 				if(@iconv_set_encoding($type, $encoding)) {
 					$iconv_input_encoding = $encoding;
 					return true;
-				} else {
-					return false;
 				}
-			} else {
 				return false;
 			}
-			break;
+			return false;
 		case 'iconv_output_encoding':
 			if (empty($encoding)) {
 				if (is_null($iconv_output_encoding)) {
@@ -2623,16 +2668,12 @@ function api_iconv_set_encoding($type, $encoding = null) {
 				if(@iconv_set_encoding($type, $encoding)) {
 					$iconv_output_encoding = $encoding;
 					return true;
-				} else {
-					return false;
 				}
-			} else {
 				return false;
 			}
-			break;
-		default:
 			return false;
 	}
+	return false;
 }
 
 /**

main/inc/lib/multibyte_string_functions_internal.lib.php

@@ -8,6 +8,10 @@
  * @author: Ivan Tcholakov, ivantcholakov@gmail.com, 2009
  * @package dokeos.library
  * ==============================================================================
+ * 
+ * Note: All functions and data structures here are not to be used directly.
+ * See the file multibyte_string_functions.lib.php which contains the "public" API.
+ * 
  */
 
 // Global variables used by some callback functions.
@@ -24,7 +28,7 @@ $_api_collator = null;
 // This is a php-implementation of the function api_convert_encoding().
 function _api_convert_encoding($string, $to_encoding, $from_encoding) {
 	static $character_map = array();
-	static $utf8_like = array('UTF-8', 'US-ASCII');
+	static $utf8_compatible = array('UTF-8', 'US-ASCII');
 	if (empty($string)) {
 		return $string;
 	}
@@ -35,7 +39,7 @@ function _api_convert_encoding($string, $to_encoding, $from_encoding) {
 	}
 	$to = _api_get_character_map_name($to_encoding);
 	$from = _api_get_character_map_name($from_encoding);
-	if (empty($to) || empty($from) || $to == $from || (in_array($to, $utf8_like) && in_array($from, $utf8_like))) {
+	if (empty($to) || empty($from) || $to == $from || (in_array($to, $utf8_compatible) && in_array($from, $utf8_compatible))) {
 		return $string;
 	}
 	if (!isset($character_map[$to])) {
@@ -395,6 +399,20 @@ function _api_utf8_get_letter_case_properties($codepoint, $type = 'lower') {
 	return $result;
 }
 
+/**
+ * A callback function for serving the function api_ucwords()
+ * @author Harry Fuecks
+ * @link http://dev.splitbrain.org/view/darcs/dokuwiki/inc/utf8.php
+ * @author Ivan Tcholakov, adaptation for the Dokeos LMS, 2009
+ * @param array $matches	Input array of matches corresponding to a single word
+ * @return string			Returns a with first char of the word in uppercase
+ */
+function _api_utf8_ucwords_callback($matches) {
+	$leadingws = $matches[2];
+	$ucfirst = api_strtoupper($matches[3], 'UTF-8');
+	$ucword = api_substr_replace(ltrim($matches[0]), $ucfirst, 0, 1, 'UTF-8');
+	return $leadingws . $ucword;
+}
 
 /**
  * ----------------------------------------------------------------------------
@@ -571,11 +589,10 @@ if (MBSTRING_INSTALLED && !function_exists('mb_stristr')) {
 		if ($pos === false) {
 			return false;
 		}
-		elseif($part == true) {
+		if($part == true) {
 			return mb_substr($haystack, 0, $pos + 1, $encoding);
-		} else {
-			return mb_substr($haystack, $pos, mb_strlen($haystack, $encoding), $encoding);
 		}
+		return mb_substr($haystack, $pos, mb_strlen($haystack, $encoding), $encoding);
 	}
 }
 
@@ -591,11 +608,11 @@ if (MBSTRING_INSTALLED && !function_exists('mb_strrchr')) {
 		$pos = mb_strrpos($haystack, $needle, mb_strlen($haystack, $encoding) - 1, $encoding);
 		if ($pos === false) {
 			return false;
-		} elseif($part == true) {
+		} 
+		if($part == true) {
 			return mb_substr($haystack, 0, $pos + 1, $encoding);
-		} else {
-			return mb_substr($haystack, $pos, mb_strlen($haystack, $encoding), $encoding);
 		}
+		return mb_substr($haystack, $pos, mb_strlen($haystack, $encoding), $encoding);
 	}
 }
 
@@ -610,10 +627,10 @@ if (MBSTRING_INSTALLED && !function_exists('mb_strstr')) {
 		$pos = mb_strpos($haystack, $needle, 0, $encoding);
 		if ($pos === false) {
 			return false;
-		} elseif($part == true) {
+		}
+		if($part == true) {
 			return mb_substr($haystack, 0, $pos + 1, $encoding);
-		} else {
-			return mb_substr($haystack, $pos, mb_strlen($haystack, $encoding), $encoding);
 		}
+		return mb_substr($haystack, $pos, mb_strlen($haystack, $encoding), $encoding);
 	}
 }