and
tags within them. var $skip_elements = 'p|pre|ol|ul'; // Tags we want the parser to completely ignore when splitting the string. var $ignore_elements = 'a|b|i|em|strong|span|img|li'; /** * Nothing to do here... * */ function CI_Typography() { } /** * Auto Typography * * This function converts text, making it typographically correct: * - Converts double spaces into paragraphs. * - Converts single line breaks into
tags * - Converts single and double quotes into correctly facing curly quote entities. * - Converts three dots into ellipsis. * - Converts double dashes into em-dashes. * - Converts two spaces into entities * * @access public * @param string * @param bool whether to strip javascript event handlers for security * @param bool whether to reduce more then two consecutive newlines to two * @return string */ function auto_typography($str, $strip_js_event_handlers = TRUE, $reduce_linebreaks = FALSE) { if ($str == '') { return ''; } // Standardize Newlines to make matching easier if (strpos($str, "\r") !== FALSE) { $str = str_replace(array("\r\n", "\r"), "\n", $str); } // Reduce line breaks. If there are more than two consecutive linebreaks // we'll compress them down to a maximum of two since there's no benefit to more. if ($reduce_linebreaks === TRUE) { $str = preg_replace("/\n\n+/", "\n\n", $str); } // Do we allow JavaScript event handlers? If not, we strip them from within all tags if ($strip_js_event_handlers === TRUE) { $str = preg_replace("#<([^><]+?)([^a-z_\-]on\w*|xmlns)(\s*=\s*[^><]*)([><]*)#i", "<\\1\\4", $str); } // Convert quotes within tags to temporary markers. We don't want quotes converted // within tags so we'll temporarily convert them to {@DQ} and {@SQ} if (preg_match_all("#\<.+?>#si", $str, $matches)) { for ($i = 0; $i < count($matches['0']); $i++) { $str = str_replace($matches['0'][$i], str_replace(array("'",'"'), array('{@SQ}', '{@DQ}'), $matches['0'][$i]), $str); } } // Convert "ignore" tags to temporary marker. The parser splits out the string at every tag // it encounters. Certain inline tags, like image tags, links, span tags, etc. will be // adversely affected if they are split out so we'll convert the opening bracket < temporarily to: {@TAG} $str = preg_replace("#<(/*)(".$this->ignore_elements.")#i", "{@TAG}\\1\\2", $str); // Split the string at every tag. This expression creates an array with this prototype: // // [array] // { // [0] = // [1] = Content... // [2] = // Etc... // } $chunks = preg_split('/(<(?:[^<>]+(?:"[^"]*"|\'[^\']*\')?)+>)/', $str, -1, PREG_SPLIT_DELIM_CAPTURE|PREG_SPLIT_NO_EMPTY); // Build our finalized string. We cycle through the array, skipping tags, and processing the contained text $str = ''; $process = TRUE; $paragraph = FALSE; foreach ($chunks as $chunk) { // Are we dealing with a tag? If so, we'll skip the processing for this cycle. // Well also set the "process" flag which allows us to skip

 tags and a few other things.
			if (preg_match("#<(/*)(".$this->block_elements.").*?\>#", $chunk, $match))
			{
				if (preg_match("#".$this->skip_elements."#", $match[2]))
				{
					$process =  ($match[1] == '/') ? TRUE : FALSE;					
				}
				
				$str .= $chunk;
				continue;
			}
				
			if ($process == FALSE)
			{
				$str .= $chunk;
				continue;
			}
			
			//  Convert Newlines into 
 and 
 tags
			$str .= $this->_format_newlines($chunk);
		}

		// Convert quotes, elipsis, and em-dashes
		$str = $this->format_characters($str);
	
		// Final clean up
		$table = array(
		
						// If the user submitted their own paragraph tags within the text
						// we will retain them instead of using our tags.
						'/()
/'		=> '$1', // )+#'		=> '
',
						'/(
)+/'		=> '
',
						
						// Clean up stray paragraph tags that appear before block level elements
						'#
<('.$this->block_elements.')#'	=> '<$1',
			
						// Replace the temporary markers we added earlier
						'/\{@TAG\}/'		=> '<',
						'/\{@DQ\}/'			=> '"',
						'/\{@SQ\}/'			=> "'"

						);
	
		// Do we need to reduce empty lines?
		if ($reduce_linebreaks === TRUE)
		{
			$table['#
\n*
#'] = '';
		}
		else
		{
			// If we have empty paragraph tags we add a non-breaking space
			// otherwise most browsers won't treat them as true paragraphs
			$table['#
#'] = '
 
';
		}
	
		return preg_replace(array_keys($table), $table, $str);

	}
	
	// --------------------------------------------------------------------

	/**
	 * Format Characters
	 *
	 * This function mainly converts double and single quotes
	 * to curly entities, but it also converts em-dashes,
	 * double spaces, and ampersands
	 *
	 * @access	public
	 * @param	string
	 * @return	string
	 */
	function format_characters($str)
	{
		static $table;
		
		if ( ! isset($table))
		{
	        $table = array(					
							// nested smart quotes, opening and closing
							// note that rules for grammar (English) allow only for two levels deep
							// and that single quotes are _supposed_ to always be on the outside
							// but we'll accommodate both
							'/(^|\W|\s)\'"/'				=> '$1‘“',
							'/\'"(\s|\W|$)/'				=> '’”$1',
							'/(^|\W|\s)"\'/'				=> '$1“‘',
							'/"\'(\s|\W|$)/'				=> '”’$1',

							// single quote smart quotes
							'/\'(\s|\W|$)/'					=> '’$1',
							'/(^|\W|\s)\'/'					=> '$1‘',

							// double quote smart quotes
							'/"(\s|\W|$)/'					=> '”$1',
							'/(^|\W|\s)"/'					=> '$1“',

							// apostrophes
							"/(\w)'(\w)/"       	    	=> '$1’$2',

							// Em dash and ellipses dots
							'/\s?\-\-\s?/'					=> '—',
							'/\w\.{3}/'						=> '…',

							// double space after sentences
							'/(\W)  /'						=> '$1  ',

							// ampersands, if not a character entity
							'/&(?!#?[a-zA-Z0-9]{2,};)/'		=> '&'
	        			);			
		}	

		return preg_replace(array_keys($table), $table, $str);
	}
	
	// --------------------------------------------------------------------

	/**
	 * Format Newlines
	 *
	 * Converts newline characters into either 
 tags or 

	 *
	 */	
	function _format_newlines($str)
	{
		if ($str == '')
		{
			return $str;
		}

		if (strpos($str, "\n") === FALSE)
		{
			return $str;
		}

		$str = str_replace("\n\n", "
\n\n
", $str);		
		$str = preg_replace("/([^\n])(\n)([^\n])/", "\\1
\\2\\3", $str);
		
		return '
'.$str.'
';
	}
	
	// ------------------------------------------------------------------------
	
	/**
	 * Convert newlines to HTML line breaks except within PRE tags
	 *
	 * @access	public
	 * @param	string
	 * @return	string
	 */		
	function nl2br_except_pre($str)
	{
		$ex = explode("pre>",$str);
		$ct = count($ex);
	
		$newstr = "";
		for ($i = 0; $i < $ct; $i++)
		{
			if (($i % 2) == 0)
			{
				$newstr .= nl2br($ex[$i]);
			}
			else
			{
				$newstr .= $ex[$i];
			}
		
			if ($ct - 1 != $i)
				$newstr .= "pre>";
		}
	
		return $newstr;
	}
	
}
// END Typography Class

/* End of file Typography.php */
/* Location: ./system/libraries/Typography.php */