X-Git-Url: http://git.onelab.eu/?p=www-register-wizard.git;a=blobdiff_plain;f=libraries%2FTypography.php;h=242ca3cb618d7f66d922227409d5e826706e13f0;hp=0ee6a24d08d35ca50a339a7637d0626d7a9e9efb;hb=47598daa8c32dbbd72db83dc33f2ce91b3f6f7b0;hpb=4afb2fe256f094a1caf6bff14f51c6a88938cc9f

diff --git a/libraries/Typography.php b/libraries/Typography.php
index 0ee6a24..242ca3c 100644
--- a/libraries/Typography.php
+++ b/libraries/Typography.php
@@ -6,7 +6,7 @@
  *
  * @package		CodeIgniter
  * @author		ExpressionEngine Dev Team
- * @copyright	Copyright (c) 2008, EllisLab, Inc.
+ * @copyright	Copyright (c) 2008 - 2009, EllisLab, Inc.
  * @license		http://codeigniter.com/user_guide/license.html
  * @link		http://codeigniter.com
  * @since		Version 1.0
@@ -30,11 +30,17 @@ class CI_Typography {
 	var $block_elements = 'address|blockquote|div|dl|fieldset|form|h\d|hr|noscript|object|ol|p|pre|script|table|ul';
 	
 	// Elements that should not have <p> and <br /> tags within them.
-	var $skip_elements	= 'p|pre|ol|ul|dl|object|table';
+	var $skip_elements	= 'p|pre|ol|ul|dl|object|table|h\d';
 	
 	// Tags we want the parser to completely ignore when splitting the string.
-	var $inline_elements = 'a|abbr|acronym|b|bdo|br|button|cite|code|del|dfn|em|i|img|ins|input|label|map|kbd|samp|select|span|strong|sub|sup|textarea|var';
-
+	var $inline_elements = 'a|abbr|acronym|b|bdo|big|br|button|cite|code|del|dfn|em|i|img|ins|input|label|map|kbd|q|samp|select|small|span|strong|sub|sup|textarea|tt|var';
+	
+	// array of block level elements that require inner content to be within another block level element
+	var $inner_block_required = array('blockquote');
+	
+	// the last block element parsed
+	var $last_block_element = '';
+	
 	// whether or not to protect quotes within { curly braces }
 	var $protect_braced_quotes = FALSE;
 	
@@ -59,11 +65,10 @@ class CI_Typography {
 	 *
 	 * @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, $reduce_linebreaks = FALSE)
+	function auto_typography($str, $reduce_linebreaks = FALSE)
 	{
 		if ($str == '')
 		{
@@ -81,39 +86,38 @@ class CI_Typography {
 		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))
+		// HTML comment tags don't conform to patterns of normal tags, so pull them out separately, only if needed
+		$html_comments = array();
+		if (strpos($str, '<!--') !== FALSE)
 		{
-			for ($i = 0; $i < count($matches['0']); $i++)
+			if (preg_match_all("#(<!\-\-.*?\-\->)#s", $str, $matches))
 			{
-				$str = str_replace($matches['0'][$i],
-									str_replace(array("'",'"'), array('{@SQ}', '{@DQ}'), $matches['0'][$i]),
-									$str);
+				for ($i = 0, $total = count($matches[0]); $i < $total; $i++)
+				{
+					$html_comments[] = $matches[0][$i];
+					$str = str_replace($matches[0][$i], '{@HC'.$i.'}', $str);
+				}
 			}
 		}
+		
+		// match and yank <pre> tags if they exist.  It's cheaper to do this separately since most content will
+		// not contain <pre> tags, and it keeps the PCRE patterns below simpler and faster
+		if (strpos($str, '<pre') !== FALSE)
+		{
+			$str = preg_replace_callback("#<pre.*?>.*?</pre>#si", array($this, '_protect_characters'), $str);
+		}
+		
+		// Convert quotes within tags to temporary markers.
+		$str = preg_replace_callback("#<.+?>#si", array($this, '_protect_characters'), $str);
 
+		// Do the same with braces if necessary
 		if ($this->protect_braced_quotes === TRUE)
 		{
-			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);
-				}
-			}			
+			$str = preg_replace_callback("#\{.+?\}#si", array($this, '_protect_characters'), $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}
@@ -134,61 +138,91 @@ class CI_Typography {
 		$str = '';
 		$process = TRUE;
 		$paragraph = FALSE;
+		$current_chunk = 0;
+		$total_chunks = count($chunks);
+		
 		foreach ($chunks as $chunk)
-		{
+		{ 
+			$current_chunk++;
+			
 			// Are we dealing with a tag? If so, we'll skip the processing for this cycle.
 			// Well also set the "process" flag which allows us to skip <pre> tags and a few other things.
-			if (preg_match("#<(/*)(".$this->block_elements.").*?\>#", $chunk, $match))
+			if (preg_match("#<(/*)(".$this->block_elements.").*?>#", $chunk, $match))
 			{
 				if (preg_match("#".$this->skip_elements."#", $match[2]))
 				{
 					$process =  ($match[1] == '/') ? TRUE : FALSE;
 				}
 				
+				if ($match[1] == '')
+				{
+					$this->last_block_element = $match[2];
+				}
+
 				$str .= $chunk;
 				continue;
 			}
-
+			
 			if ($process == FALSE)
 			{
 				$str .= $chunk;
 				continue;
 			}
 			
+			//  Force a newline to make sure end tags get processed by _format_newlines()
+			if ($current_chunk == $total_chunks)
+			{
+				$chunk .= "\n";  
+			}
+			
 			//  Convert Newlines into <p> and <br /> tags
 			$str .= $this->_format_newlines($chunk);
 		}
-
-		// is the whole of the content inside a block level element?
-		if ( ! preg_match("/^<(?:".$this->block_elements.")/i", $str, $match))
+		
+		// No opening block level tag?  Add it if needed.
+		if ( ! preg_match("/^\s*<(?:".$this->block_elements.")/i", $str))
 		{
-			$str = "<p>{$str}</p>";
+			$str = preg_replace("/^(.*?)<(".$this->block_elements.")/i", '<p>$1</p><$2', $str);
 		}
-
-		// Convert quotes, elipsis, and em-dashes
+		
+		// Convert quotes, elipsis, em-dashes, non-breaking spaces, and ampersands
 		$str = $this->format_characters($str);
-	
+		
+		// restore HTML comments
+		for ($i = 0, $total = count($html_comments); $i < $total; $i++)
+		{
+			// remove surrounding paragraph tags, but only if there's an opening paragraph tag
+			// otherwise HTML comments at the ends of paragraphs will have the closing tag removed
+			// if '<p>{@HC1}' then replace <p>{@HC1}</p> with the comment, else replace only {@HC1} with the comment
+			$str = preg_replace('#(?(?=<p>\{@HC'.$i.'\})<p>\{@HC'.$i.'\}(\s*</p>)|\{@HC'.$i.'\})#s', $html_comments[$i], $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.
-						'/(<p.*?>)<p>/'		=> '$1', // <?php BBEdit syntax coloring bug fix
+						'/(<p[^>*?]>)<p>/'		=> '$1', // <?php BBEdit syntax coloring bug fix
 						
 						// Reduce multiple instances of opening/closing paragraph tags to a single one
 						'#(</p>)+#'			=> '</p>',
-						'/(<p><p>)+/'		=> '<p>',
+						'/(<p>\W*<p>)+/'	=> '<p>',
 						
 						// Clean up stray paragraph tags that appear before block level elements
 						'#<p></p><('.$this->block_elements.')#'	=> '<$1',
-			
+
+						// Clean up stray non-breaking spaces preceeding block elements
+						'#(&nbsp;\s*)+<('.$this->block_elements.')#'	=> '  <$2',
+
 						// Replace the temporary markers we added earlier
 						'/\{@TAG\}/'		=> '<',
 						'/\{@DQ\}/'			=> '"',
-						'/\{@SQ\}/'			=> "'"
+						'/\{@SQ\}/'			=> "'",
+						'/\{@DD\}/'			=> '--',
+						'/\{@NBS\}/'		=> '  '
 
 						);
-	
+		
 		// Do we need to reduce empty lines?
 		if ($reduce_linebreaks === TRUE)
 		{
@@ -200,13 +234,13 @@ class CI_Typography {
 			// otherwise most browsers won't treat them as true paragraphs
 			$table['#<p></p>#'] = '<p>&nbsp;</p>';
 		}
-	
+		
 		return preg_replace(array_keys($table), $table, $str);
 
 	}
 	
 	// --------------------------------------------------------------------
-
+	
 	/**
 	 * Format Characters
 	 *
@@ -224,26 +258,37 @@ class CI_Typography {
 		
 		if ( ! isset($table))
 		{
-	        $table = array(					
+			$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)\'"/'				=> '$1&#8216;&#8220;',
-							'/\'"(\s|\W|$)/'				=> '&#8217;&#8221;$1',
-							'/(^|\W|\s)"\'/'				=> '$1&#8220;&#8216;',
-							'/"\'(\s|\W|$)/'				=> '&#8221;&#8217;$1',
+							// Note that in all cases, whitespace is the primary determining factor
+							// on which direction to curl, with non-word characters like punctuation
+							// being a secondary factor only after whitespace is addressed.
+							'/\'"(\s|$)/'					=> '&#8217;&#8221;$1',
+							'/(^|\s|<p>)\'"/'				=> '$1&#8216;&#8220;',
+							'/\'"(\W)/'						=> '&#8217;&#8221;$1',
+							'/(\W)\'"/'						=> '$1&#8216;&#8220;',
+							'/"\'(\s|$)/'					=> '&#8221;&#8217;$1',
+							'/(^|\s|<p>)"\'/'				=> '$1&#8220;&#8216;',
+							'/"\'(\W)/'						=> '&#8221;&#8217;$1',
+							'/(\W)"\'/'						=> '$1&#8220;&#8216;',
 
 							// single quote smart quotes
-							'/\'(\s|\W|$)/'					=> '&#8217;$1',
-							'/(^|\W|\s)\'/'					=> '$1&#8216;',
+							'/\'(\s|$)/'					=> '&#8217;$1',
+							'/(^|\s|<p>)\'/'				=> '$1&#8216;',
+							'/\'(\W)/'						=> '&#8217;$1',
+							'/(\W)\'/'						=> '$1&#8216;',
 
 							// double quote smart quotes
-							'/"(\s|\W|$)/'					=> '&#8221;$1',
-							'/(^|\W|\s)"/'					=> '$1&#8220;',
+							'/"(\s|$)/'						=> '&#8221;$1',
+							'/(^|\s|<p>)"/'					=> '$1&#8220;',
+							'/"(\W)/'						=> '&#8221;$1',
+							'/(\W)"/'						=> '$1&#8220;',
 
 							// apostrophes
-							"/(\w)'(\w)/"       	    	=> '$1&#8217;$2',
+							"/(\w)'(\w)/"					=> '$1&#8217;$2',
 
 							// Em dash and ellipses dots
 							'/\s?\-\-\s?/'					=> '&#8212;',
@@ -254,8 +299,8 @@ class CI_Typography {
 
 							// ampersands, if not a character entity
 							'/&(?!#?[a-zA-Z0-9]{2,};)/'		=> '&amp;'
-	        			);			
-		}	
+						);
+		}
 
 		return preg_replace(array_keys($table), $table, $str);
 	}
@@ -277,8 +322,8 @@ class CI_Typography {
 		{
 			return $str;
 		}
-
-		if (strpos($str, "\n") === FALSE)
+		
+		if (strpos($str, "\n") === FALSE  && ! in_array($this->last_block_element, $this->inner_block_required))
 		{
 			return $str;
 		}
@@ -304,6 +349,25 @@ class CI_Typography {
 	
 	// ------------------------------------------------------------------------
 	
+	/**
+	 * Protect Characters
+	 *
+	 * Protects special characters from being formatted later
+	 * We don't want quotes converted within tags so we'll temporarily convert them to {@DQ} and {@SQ}
+ 	 * and we don't want double dashes converted to emdash entities, so they are marked with {@DD}
+ 	 * likewise double spaces are converted to {@NBS} to prevent entity conversion
+	 *
+	 * @access	public
+	 * @param	array
+	 * @return	string
+	 */
+	function _protect_characters($match)
+	{
+		return str_replace(array("'",'"','--','  '), array('{@SQ}', '{@DQ}', '{@DD}', '{@NBS}'), $match[0]);
+	}
+
+	// --------------------------------------------------------------------
+	
 	/**
 	 * Convert newlines to HTML line breaks except within PRE tags
 	 *