utf_tools.php

这些都是我以前学习是用到的源码

* Make a string's first character uppercase
* 
* @author Harry Fuecks
* @param string
* @return string with first character as upper case (if applicable)
*/
function utf8_ucfirst($str)
{
	switch (utf8_strlen($str))
	{
		case 0:
			return '';
		break;

		case 1:
			return utf8_strtoupper($str);
		break;

		default:
			preg_match('/^(.{1})(.*)$/us', $str, $matches);
			return utf8_strtoupper($matches[1]) . $matches[2];
		break;
	}
}

/**
* Recode a string to UTF-8
*
* If the encoding is not supported, the string is returned as-is
*
* @param	string	$string		Original string
* @param	string	$encoding	Original encoding (lowered)
* @return	string				The string, encoded in UTF-8
*/
function utf8_recode($string, $encoding)
{
	$encoding = strtolower($encoding);

	if ($encoding == 'utf-8' || !is_string($string) || !isset($string[0]))
	{
		return $string;
	}

	// start with something simple
	if ($encoding == 'iso-8859-1')
	{
		return utf8_encode($string);
	}

	// First, try iconv()
	if (function_exists('iconv'))
	{
		$ret = @iconv($encoding, 'utf-8', $string);

		if (isset($ret[0]))
		{
			return $ret;
		}
	}

	// Try the mb_string extension
	if (function_exists('mb_convert_encoding'))
	{
		$ret = @mb_convert_encoding($string, 'utf-8', $encoding);

		if (isset($ret[0]))
		{
			return $ret;
		}
	}

	// Try the recode extension
	if (function_exists('recode_string'))
	{
		$ret = @recode_string($encoding . '..utf-8', $string);

		if (isset($ret[0]))
		{
			return $ret;
		}
	}

	// If nothing works, check if we have a custom transcoder available
	if (!preg_match('#^[a-z0-9\\-]+$#', $encoding))
	{
		// Make sure the encoding name is alphanumeric, we don't want it to be abused into loading arbitrary files
		trigger_error('Unknown encoding: ' . $encoding, E_USER_ERROR);
	}

	global $phpbb_root_path;

	if (!file_exists($phpbb_root_path . 'includes/utf/data/'))
	{
		return $string;
	}

	die('Finish me!! ' . basename(__FILE__) . ' at line ' . __LINE__);
}

/**
* Replace all UTF-8 chars that are not in ASCII with their NCR
*
* @param	string	$text		UTF-8 string in NFC
* @return	string				ASCII string using NCRs for non-ASCII chars
*/
function utf8_encode_ncr($text)
{
	return preg_replace_callback('#[\\xC2-\\xF4][\\x80-\\xBF]{1,3}#', 'utf8_encode_ncr_callback', $text);
}

/**
* Callback used in encode_ncr()
*
* Takes a UTF-8 char and replaces it with its NCR. Attention, $m is an array
*
* @param	array	$m			0-based numerically indexed array passed by preg_replace_callback()
* @return	string				A HTML NCR if the character is valid, or the original string otherwise
*/
function utf8_encode_ncr_callback($m)
{
	return '&#' . utf8_ord($m[0]) . ';';
}

/**
* Enter description here...
*
* @param string $chr UTF-8 char
* @return integer UNICODE code point
*/
function utf8_ord($chr)
{
	switch (strlen($chr))
	{
		case 1:
			return ord($chr);
		break;

		case 2:
			return ((ord($chr[0]) & 0x1F) << 6) | (ord($chr[1]) & 0x3F);
		break;

		case 3:
			return ((ord($chr[0]) & 0x0F) << 12) | ((ord($chr[1]) & 0x3F) << 6) | (ord($chr[2]) & 0x3F);
		break;

		case 4:
			return ((ord($chr[0]) & 0x07) << 18) | ((ord($chr[1]) & 0x3F) << 12) | ((ord($chr[2]) & 0x3F) << 6) | (ord($chr[3]) & 0x3F);
		break;

		default:
			return $chr;
	}
}

/**
* Converts an NCR to a UTF-8 char
*
* @param	int		$cp	UNICODE code point
* @return	string		UTF-8 char
*/
function utf8_chr($cp)
{
	if ($cp > 0xFFFF)
	{
		return chr(0xF0 | ($cp >> 18)) . chr(0x80 | (($cp >> 12) & 0x3F)) . chr(0x80 | (($cp >> 6) & 0x3F)) . chr(0x80 | ($cp & 0x3F));
	}
	else if ($cp > 0x7FF)
	{
		return chr(0xE0 | ($cp >> 12)) . chr(0x80 | (($cp >> 6) & 0x3F)) . chr(0x80 | ($cp & 0x3F));
	}
	else if ($cp > 0x7F)
	{
		return chr(0xC0 | ($cp >> 6)) . chr(0x80 | ($cp & 0x3F));
	}
	else
	{
		return chr($cp);
	}
}

/**
* Convert Numeric Character References to UTF-8 chars
*
* Notes:
*	- we do not convert NCRs recursively, if you pass &#38;#38; it will return &#38;
*	- we DO NOT check for the existence of the Unicode characters, therefore an entity may be converted to an inexistent codepoint
*
* @param	string	$text		String to convert, encoded in UTF-8 (no normal form required)
* @return	string				UTF-8 string where NCRs have been replaced with the actual chars
*/
function utf8_decode_ncr($text)
{
	return preg_replace_callback('/&#([0-9]{1,6}|x[0-9A-F]{1,5});/i', 'utf8_decode_ncr_callback', $text);
}

/**
* Callback used in decode_ncr()
*
* Takes a NCR (in decimal or hexadecimal) and returns a UTF-8 char. Attention, $m is an array.
* It will ignore most of invalid NCRs, but not all!
*
* @param	array	$m			0-based numerically indexed array passed by preg_replace_callback()
* @return	string				UTF-8 char
*/
function utf8_decode_ncr_callback($m)
{
	$cp = (strncasecmp($m[1], 'x', 1)) ? $m[1] : hexdec(substr($m[1], 1));

	return utf8_chr($cp);
}

/**
* Takes an array of ints representing the Unicode characters and returns
* a UTF-8 string.
*
* @param	string	$text	text to be case folded
* @param	string	$option	determines how we will fold the cases
* @return	string			case folded text
*/
function utf8_case_fold($text, $option = 'full')
{
	static $uniarray = array();
	global $phpbb_root_path, $phpEx;

	// common is always set
	if (!isset($uniarray['c']))
	{
		$uniarray['c'] = include($phpbb_root_path . 'includes/utf/data/case_fold_c.' . $phpEx);
	}

	// only set full if we need to
	if ($option === 'full' && !isset($uniarray['f']))
	{
		$uniarray['f'] = include($phpbb_root_path . 'includes/utf/data/case_fold_f.' . $phpEx);
	}

	// only set simple if we need to
	if ($option !== 'full' && !isset($uniarray['s']))
	{
		$uniarray['s'] = include($phpbb_root_path . 'includes/utf/data/case_fold_s.' . $phpEx);
	}

	$text = strtr($text, $uniarray['c']);
	if ($option === 'full')
	{
		$text = strtr($text, $uniarray['f']);
	}
	else
	{
		$text = strtr($text, $uniarray['s']);
	}
	return $text;
}

/**
* A wrapper function for the normalizer which takes care of including the class if required and modifies the passed strings
* to be in NFC (Normalization Form Composition).
*
* @param	mixed	$strings	a string or an array of strings to normalize
* @return	mixed				the normalized content, preserving array keys if array given.
*/
function utf8_normalize_nfc($strings)
{
	if (empty($strings))
	{
		return $strings;
	}

	if (!class_exists('utf_normalizer'))
	{
		global $phpbb_root_path, $phpEx;
		include($phpbb_root_path . 'includes/utf/utf_normalizer.' . $phpEx);
	}

	if (!is_array($strings))
	{
		utf_normalizer::nfc($strings);
	}
	else if (is_array($strings))
	{
		foreach ($strings as $key => $string)
		{
			utf_normalizer::nfc($strings[$key]);
		}
	}

	return $strings;
}

/**
* This function is used to generate a "clean" version of a string.
* Clean means that it is a case insensitive form (case folding) and that it is normalized (NFC).
* Additionally a homographs of one character are transformed into one specific character (preferably ASCII
* if it is an ASCII character).
*
* Please be aware that if you change something within this function or within
* functions used here you need to rebuild/update the username_clean column in the users table. And all other
* columns that store a clean string otherwise you will break this functionality.
*
* @param	string	$text	An unclean string, mabye user input (has to be valid UTF-8!)
* @return	string			Cleaned up version of the input string
*/
function utf8_clean_string($text)
{
	$text = utf8_case_fold($text);
	
	if (!class_exists('utf_normalizer'))
	{
		global $phpbb_root_path, $phpEx;
		include($phpbb_root_path . 'includes/utf/utf_normalizer.' . $phpEx);
	}

	utf_normalizer::nfc($text);

	static $homographs = array(
		// cyrllic
		"\xD0\xB0" => "\x61",
		"\xD0\xB5" => "\x65",
		"\xD0\xBE" => "\x6F",
		"\xD1\x80" => "\x70",
		"\xD1\x81" => "\x63",
		"\xD1\x83" => "\x79",
		"\xD1\x85" => "\x78",
		"\xD1\x95" => "\x73",
		"\xD1\x96" => "\x69",
		"\xD1\x98" => "\x6A",
		"\xD2\xBB" => "\x68",
		// greek
		"\xCE\xB1" => "\x61",
		"\xCE\xBF" => "\x6F",
		// other
		"\xC2\xA1" => "\x69",
	);

	$text = strtr($text, $homographs);

	return $text;
}

?>