improve parsing of emphasis

Prior to PHP 5.4.0 the default encoding for `htmlentities()`
and `htmlspecialchars` is "ISO-8859-1". For PHP 5.4+ is "UTF-8".

This ensures always the right encoding is used no matter the PHP version
and the locale settings.

.travis.yml

@@ -4,3 +4,4 @@ php:
   - 5.5
   - 5.4
   - 5.3
+  - 5.2

Parsedown.php

@@ -46,17 +46,17 @@ class Parsedown
 
 	function parse($text)
 	{
-		# Removes UTF-8 BOM and marker characters.
+		# removes UTF-8 BOM and marker characters
 		$text = preg_replace('{^\xEF\xBB\xBF|\x1A}', '', $text);
 
-		# Removes \r characters.
+		# removes \r characters
 		$text = str_replace("\r\n", "\n", $text);
 		$text = str_replace("\r", "\n", $text);
 
-		# Replaces tabs with spaces.
+		# replaces tabs with spaces
 		$text = str_replace("\t", '    ', $text);
 
-		# Encodes escape sequences.
+		# encodes escape sequences
 
 		if (strpos($text, '\\') !== FALSE)
 		{
@@ -66,7 +66,7 @@ class Parsedown
 			{
 				if (strpos($text, $escape_sequence) !== FALSE)
 				{
-					$code = "\x1A".'\\'.$index;
+					$code = "\x1A".'\\'.$index.';';
 
 					$text = str_replace($escape_sequence, $code, $text);
 
@@ -78,13 +78,13 @@ class Parsedown
 		# ~
 
 		$text = preg_replace('/\n\s*\n/', "\n\n", $text);
-		$text = trim($text, "\n ");
+		$text = trim($text, "\n");
 
 		$lines = explode("\n", $text);
 
 		$text = $this->parse_block_elements($lines);
 
-		# Decodes escape sequences (leaves out backslashes).
+		# decodes escape sequences
 
 		foreach ($this->escape_sequence_map as $code => $escape_sequence)
 		{
@@ -110,28 +110,55 @@ class Parsedown
 
 		foreach ($lines as $line)
 		{
-			# Block-Level HTML 
+			# fenced elements
 
-			if ($element['type'] === 'block' and ! isset($element['closed']))
+			switch ($element['type'])
 			{
-				if (preg_match('{<'.$element['subtype'].'>$}', $line)) # <open>
-				{
-					$element['depth']++;
-				}
+				case 'fenced_code_block':
 
-				if (preg_match('{</'.$element['subtype'].'>$}', $line)) # </close>
-				{
-					$element['depth'] > 0 
-						? $element['depth']-- 
-						: $element['closed'] = true;
-				}
+					if ( ! isset($element['closed']))
+					{
+						if (preg_match('/^[ ]*'.$element['fence'][0].'{3,}[ ]*$/', $line))
+						{
+							$element['closed'] = true;
+						}
+						else
+						{
+							$element['text'] !== '' and $element['text'] .= "\n";
 
-				$element['text'] .= "\n".$line;
+							$element['text'] .= $line;
+						}
 
-				continue;
+						continue 2;
+					}
+
+					break;
+
+				case 'markup':
+
+					if ( ! isset($element['closed']))
+					{
+						if (preg_match('{<'.$element['subtype'].'>$}', $line)) # opening tag
+						{
+							$element['depth']++;
+						}
+
+						if (preg_match('{</'.$element['subtype'].'>$}', $line)) # closing tag
+						{
+							$element['depth'] > 0
+								? $element['depth']--
+								: $element['closed'] = true;
+						}
+
+						$element['text'] .= "\n".$line;
+
+						continue 2;
+					}
+
+					break;
 			}
 
-			# Empty 
+			# *
 
 			if ($line === '')
 			{
@@ -140,259 +167,329 @@ class Parsedown
 				continue;
 			}
 
-			# Lazy Blockquote 
+			# composite elements
 
-			if ($element['type'] === 'blockquote' and ! isset($element['interrupted']))
+			switch ($element['type'])
 			{
-				$line = preg_replace('/^[ ]*>[ ]?/', '', $line);
+				case 'blockquote':
 
-				$element['lines'] []= $line;
-				
-				continue;
-			}
-			
-			# Lazy List Item 
-			
-			if ($element['type'] === 'li')
-			{
-				if (preg_match('/^([ ]{0,3})(\d+[.]|[*+-])[ ](.*)/', $line, $matches))
-				{
-					if ($element['indentation'] !== $matches[1]) 
+					if ( ! isset($element['interrupted']))
 					{
+						$line = preg_replace('/^[ ]*>[ ]?/', '', $line);
+
 						$element['lines'] []= $line;
+
+						continue 2;
+					}
+
+					break;
+
+				case 'li':
+
+					if (preg_match('/^([ ]{0,3})(\d+[.]|[*+-])[ ](.*)/', $line, $matches))
+					{
+						if ($element['indentation'] !== $matches[1])
+						{
+							$element['lines'] []= $line;
+						}
+						else
+						{
+							unset($element['last']);
+
+							$elements []= $element;
+
+							$element = array(
+								'type' => 'li',
+								'indentation' => $matches[1],
+								'last' => true,
+								'lines' => array(
+									preg_replace('/^[ ]{0,4}/', '', $matches[3]),
+								),
+							);
+						}
+
+						continue 2;
+					}
+
+					if (isset($element['interrupted']))
+					{
+						if ($line[0] === ' ')
+						{
+							$element['lines'] []= '';
+
+							$line = preg_replace('/^[ ]{0,4}/', '', $line);
+
+							$element['lines'] []= $line;
+
+							unset($element['interrupted']);
+
+							continue 2;
+						}
 					}
 					else
 					{
-						unset($element['last']);
+						$line = preg_replace('/^[ ]{0,4}/', '', $line);
 
+						$element['lines'] []= $line;
+
+						continue 2;
+					}
+
+					break;
+			}
+
+			# indentation sensitive types
+
+			$deindented_line = $line;
+
+			switch ($line[0])
+			{
+				case ' ':
+
+					# ~
+
+					$deindented_line = ltrim($line);
+
+					if ($deindented_line === '')
+					{
+						continue 2;
+					}
+
+					# code block
+
+					if (preg_match('/^[ ]{4}(.*)/', $line, $matches))
+					{
+						if ($element['type'] === 'code_block')
+						{
+							if (isset($element['interrupted']))
+							{
+								$element['text'] .= "\n";
+
+								unset ($element['interrupted']);
+							}
+
+							$element['text'] .= "\n".$matches[1];
+						}
+						else
+						{
+							$elements []= $element;
+
+							$element = array(
+								'type' => 'code_block',
+								'text' => $matches[1],
+							);
+						}
+
+						continue 2;
+					}
+
+					break;
+
+				case '#':
+
+					# atx heading (#)
+
+					if (preg_match('/^(#{1,6})[ ]*(.+?)[ ]*#*$/', $line, $matches))
+					{
+						$elements []= $element;
+
+						$level = strlen($matches[1]);
+
+						$element = array(
+							'type' => 'h.',
+							'text' => $matches[2],
+							'level' => $level,
+						);
+
+						continue 2;
+					}
+
+					break;
+
+				case '-':
+
+					# setext heading (---)
+
+					if ($line[0] === '-' and $element['type'] === 'p' and ! isset($element['interrupted']) and preg_match('/^[-]+[ ]*$/', $line))
+					{
+						$element['type'] = 'h.';
+						$element['level'] = 2;
+
+						continue 2;
+					}
+
+					break;
+
+				case '=':
+
+					# setext heading (===)
+
+					if ($line[0] === '=' and $element['type'] === 'p' and ! isset($element['interrupted']) and preg_match('/^[=]+[ ]*$/', $line))
+					{
+						$element['type'] = 'h.';
+						$element['level'] = 1;
+
+						continue 2;
+					}
+
+					break;
+			}
+
+			# indentation insensitive types
+
+			switch ($deindented_line[0])
+			{
+				case '<':
+
+					# self-closing tag
+
+					if (preg_match('{^<.+?/>$}', $deindented_line))
+					{
+						$elements []= $element;
+
+						$element = array(
+							'type' => '',
+							'text' => $deindented_line,
+						);
+
+						continue 2;
+					}
+
+					# opening tag
+
+					if (preg_match('{^<(\w+)(?:[ ].*?)?>}', $deindented_line, $matches))
+					{
+						$elements []= $element;
+
+						$element = array(
+							'type' => 'markup',
+							'subtype' => strtolower($matches[1]),
+							'text' => $deindented_line,
+							'depth' => 0,
+						);
+
+						preg_match('{</'.$matches[1].'>\s*$}', $deindented_line) and $element['closed'] = true;
+
+						continue 2;
+					}
+
+					break;
+
+				case '>':
+
+					# quote
+
+					if (preg_match('/^>[ ]?(.*)/', $deindented_line, $matches))
+					{
+						$elements []= $element;
+
+						$element = array(
+							'type' => 'blockquote',
+							'lines' => array(
+								$matches[1],
+							),
+						);
+
+						continue 2;
+					}
+
+					break;
+
+				case '[':
+
+					# reference
+
+					if (preg_match('/^\[(.+?)\]:[ ]*([^ ]+)/', $deindented_line, $matches))
+					{
+						$label = strtolower($matches[1]);
+
+						$this->reference_map[$label] = trim($matches[2], '<>');;
+
+						continue 2;
+					}
+
+					break;
+
+				case '`':
+				case '~':
+
+					# fenced code block
+
+					if (preg_match('/^([`]{3,}|[~]{3,})[ ]*(\S+)?[ ]*$/', $deindented_line, $matches))
+					{
+						$elements []= $element;
+
+						$element = array(
+							'type' => 'fenced_code_block',
+							'text' => '',
+							'fence' => $matches[1],
+						);
+
+						isset($matches[2]) and $element['language'] = $matches[2];
+
+						continue 2;
+					}
+
+					break;
+
+				case '*':
+				case '+':
+				case '-':
+				case '_':
+
+					# hr
+
+					if (preg_match('/^([-*_])([ ]{0,2}\1){2,}[ ]*$/', $deindented_line))
+					{
+						$elements []= $element;
+
+						$element = array(
+							'type' => 'hr',
+						);
+
+						continue 2;
+					}
+
+					# li
+
+					if (preg_match('/^([ ]*)[*+-][ ](.*)/', $line, $matches))
+					{
 						$elements []= $element;
 
 						$element = array(
 							'type' => 'li',
+							'ordered' => false,
 							'indentation' => $matches[1],
 							'last' => true,
 							'lines' => array(
-								preg_replace('/^[ ]{0,4}/', '', $matches[3]),
+								preg_replace('/^[ ]{0,4}/', '', $matches[2]),
 							),
 						);
+
+						continue 2;
 					}
-					
-					continue;
-				}
-				
-				if (isset($element['interrupted']))
-				{
-					if ($line[0] === ' ')
-					{
-						$element['lines'] []= '';
-						
-						$line = preg_replace('/^[ ]{0,4}/', '', $line);;
-						
-						$element['lines'] []= $line;
-						
-						continue;
-					}
-				}
-				else
-				{
-					$line = preg_replace('/^[ ]{0,4}/', '', $line);;
-						
-					$element['lines'] []= $line;
-					
-					continue;
-				}
 			}
 
-			# Quick Paragraph 
+			# li
 
-			if ($line[0] >= 'a' or $line[0] >= 'A' and $line[0] <= 'Z')
-			{
-				goto paragraph; # trust me 
-			}
-			
-			# Code 
-			
-			if ($line[0] === ' ' and preg_match('/^[ ]{4}(.*)/', $line, $matches))
-			{
-				if ($element['type'] === 'code')
-				{
-					if (isset($element['interrupted']))
-					{
-						$element['text'] .= "\n";
-						
-						unset ($element['interrupted']);
-					}
-					
-					$element['text'] .= "\n".$matches[1];
-				}
-				else
-				{
-					$elements []= $element;
-					
-					$element = array(
-						'type' => 'code',
-						'text' => $matches[1],
-					);
-				}
-
-				continue;
-			}
-			
-			# Setext Header (---)  
-			
-			if ($line[0] === '-' and $element['type'] === 'p' and ! isset($element['interrupted']) and preg_match('/^[-]+[ ]*$/', $line))
-			{
-				$element['type'] = 'h.';
-				$element['level'] = 2;
-				
-				continue;
-			}
-			
-			# Atx Header (#)
-			
-			if ($line[0] === '#' and preg_match('/^(#{1,6})[ ]*(.+?)[ ]*#*$/', $line, $matches))
-			{
-				$elements []= $element;
-				
-				$level = strlen($matches[1]);
-				
-				$element = array(
-					'type' => 'h.',
-					'text' => $matches[2],
-					'level' => $level,
-				);
-				
-				continue;
-			}
-			
-			# Setext Header (===) 
-			
-			if ($line[0] === '=' and $element['type'] === 'p' and ! isset($element['interrupted']) and preg_match('/^[=]+[ ]*$/', $line))
-			{
-				$element['type'] = 'h.';
-				$element['level'] = 1;
-				
-				continue;
-			}
-			
-			# ~ 
-			
-			$pure_line = $line[0] !== ' ' ? $line : ltrim($line);
-			
-			# Link Reference 
-			
-			if ($pure_line[0] === '[' and preg_match('/^\[(.+?)\]:[ ]*([^ ]+)/', $pure_line, $matches))
-			{
-				$label = $matches[1];
-				$url = trim($matches[2], '<>');
-				
-				$this->reference_map[$label] = $url;
-				
-				continue;
-			}
-			
-			# Blockquote 
-			
-			if ($pure_line[0] === '>' and preg_match('/^>[ ]?(.*)/', $pure_line, $matches))
-			{
-				if ($element['type'] === 'blockquote')
-				{
-					if (isset($element['interrupted']))
-					{
-						$element['lines'] []= '';
-						
-						unset($element['interrupted']);
-					}
-					
-					$element['lines'] []= $matches[1];
-				}
-				else
-				{
-					$elements []= $element;
-					
-					$element = array(
-						'type' => 'blockquote',
-						'lines' => array(
-							$matches[1],
-						),
-					);
-				}
-				
-				continue;
-			}
-			
-			# HTML  
-			
-			if ($pure_line[0] === '<')
-			{
-				# Block-Level HTML <self-closing/>
-
-				if (preg_match('{^<.+?/>$}', $pure_line))
-				{
-					$elements []= $element;
-
-					$element = array(
-						'type' => '',
-						'text' => $pure_line,
-					);
-
-					continue;
-				}
-				
-				# Block-Level HTML <open>
-
-				if (preg_match('{^<(\w+)(?:[ ].*?)?>}', $pure_line, $matches))
-				{
-					$elements []= $element;
-
-					$element = array(
-						'type' => 'block',
-						'subtype' => strtolower($matches[1]),
-						'text' => $pure_line,
-						'depth' => 0,
-					);
-					
-					preg_match('{</'.$matches[1].'>\s*$}', $pure_line) and $element['closed'] = true;
-
-					continue;
-				}
-			}
-			
-			# Horizontal Rule   
-			
-			if (preg_match('/^([-*_])([ ]{0,2}\1){2,}[ ]*$/', $pure_line))
-			{
-				$elements []= $element;
-				
-				$element = array(
-					'type' => 'hr',
-				);
-				
-				continue;
-			}
-			
-			# List Item 
-			
-			if (preg_match('/^([ ]*)(\d+[.]|[*+-])[ ](.*)/', $line, $matches))
+			if ($deindented_line[0] <= '9' and $deindented_line >= '0' and preg_match('/^([ ]*)\d+[.][ ](.*)/', $line, $matches))
 			{
 				$elements []= $element;
 
 				$element = array(
 					'type' => 'li',
-					'ordered' => isset($matches[2][1]),
+					'ordered' => true,
 					'indentation' => $matches[1],
 					'last' => true,
 					'lines' => array(
-						preg_replace('/^[ ]{0,4}/', '', $matches[3]),
+						preg_replace('/^[ ]{0,4}/', '', $matches[2]),
 					),
 				);
 
 				continue;
 			}
 
-			# ~ 
-			
-			paragraph:
+			# paragraph
 
 			if ($element['type'] === 'p')
 			{
@@ -422,7 +519,7 @@ class Parsedown
 
 		$elements []= $element;
 
-		array_shift($elements);
+		unset($elements[0]);
 
 		#
 		# ~
@@ -430,10 +527,71 @@ class Parsedown
 
 		$markup = '';
 
-		foreach ($elements as $index => $element)
+		foreach ($elements as $element)
 		{
 			switch ($element['type'])
 			{
+				case 'p':
+
+					$text = $this->parse_span_elements($element['text']);
+
+					$text = preg_replace('/[ ]{2}\n/', '<br />'."\n", $text);
+
+					if ($context === 'li' and $markup === '')
+					{
+						if (isset($element['interrupted']))
+						{
+							$markup .= "\n".'<p>'.$text.'</p>'."\n";
+						}
+						else
+						{
+							$markup .= $text;
+						}
+					}
+					else
+					{
+						$markup .= '<p>'.$text.'</p>'."\n";
+					}
+
+					break;
+
+				case 'blockquote':
+
+					$text = $this->parse_block_elements($element['lines']);
+
+					$markup .= '<blockquote>'."\n".$text.'</blockquote>'."\n";
+
+					break;
+
+				case 'code_block':
+				case 'fenced_code_block':
+
+					$text = htmlspecialchars($element['text'], ENT_NOQUOTES, 'UTF-8');
+
+					strpos($text, "\x1A\\") !== FALSE and $text = strtr($text, $this->escape_sequence_map);
+
+					$markup .= isset($element['language'])
+						? '<pre><code class="language-'.$element['language'].'">'.$text.'</code></pre>'
+						: '<pre><code>'.$text.'</code></pre>';
+
+					$markup .= "\n";
+
+					break;
+
+				case 'h.':
+
+					$text = $this->parse_span_elements($element['text']);
+
+					$markup .= '<h'.$element['level'].'>'.$text.'</h'.$element['level'].'>'."\n";
+
+					break;
+
+				case 'hr':
+
+					$markup .= '<hr />'."\n";
+
+					break;
+
 				case 'li':
 
 					if (isset($element['ordered'])) # first
@@ -456,62 +614,6 @@ class Parsedown
 
 					break;
 
-				case 'p':
-					
-					$text = $this->parse_inline_elements($element['text']);
-					
-					$text = preg_replace('/[ ]{2}\n/', '<br />'."\n", $text);
-					
-					if ($context === 'li' and $index === 0)
-					{
-						if (isset($element['interrupted']))
-						{
-							$markup .= "\n".'<p>'.$text.'</p>'."\n";
-						}
-						else 
-						{
-							$markup .= $text;
-						}
-					}
-					else 
-					{
-						$markup .= '<p>'.$text.'</p>'."\n";
-					}
-					
-					break;
-				
-				case 'code':
-					
-					$text = htmlentities($element['text'], ENT_NOQUOTES);
-					
-					strpos($text, "\x1A\\") !== FALSE and $text = strtr($text, $this->escape_sequence_map);
-					
-					$markup .= '<pre><code>'.$text.'</code></pre>'."\n";
-					
-					break;
-				
-				case 'blockquote':
-					
-					$text = $this->parse_block_elements($element['lines']);
-					
-					$markup .= '<blockquote>'."\n".$text.'</blockquote>'."\n";
-					
-					break;
-				
-				case 'h.':
-					
-					$text = $this->parse_inline_elements($element['text']);
-					
-					$markup .= '<h'.$element['level'].'>'.$text.'</h'.$element['level'].'>'."\n";
-					
-					break;
-				
-				case 'hr':
-					
-					$markup .= '<hr />'."\n";
-					
-					break;
-
 				default:
 
 					$markup .= $element['text']."\n";
@@ -521,46 +623,15 @@ class Parsedown
 		return $markup;
 	}
 
-	private function parse_inline_elements($text)
+	private function parse_span_elements($text)
 	{
 		$map = array();
 
 		$index = 0;
 
-		# Code Span 
+		# inline link / inline image (recursive)
 
-		if (strpos($text, '`') !== FALSE and preg_match_all('/`(.+?)`/', $text, $matches, PREG_SET_ORDER)) 
-		{
-			foreach ($matches as $matches)
-			{
-				$element_text = $matches[1];
-				$element_text = htmlentities($element_text, ENT_NOQUOTES);
-				
-				# Decodes escape sequences.
-				
-				$this->escape_sequence_map 
-					and strpos($element_text, "\x1A") !== FALSE 
-					and $element_text = strtr($element_text, $this->escape_sequence_map);
-				
-				# Composes element.
-				
-				$element = '<code>'.$element_text.'</code>';
-				
-				# Encodes element.
-				
-				$code = "\x1A".'$'.$index;
-				
-				$text = str_replace($matches[0], $code, $text);
-				
-				$map[$code] = $element;
-				
-				$index ++;
-			}
-		}
-		
-		# Inline Link / Image 
-		
-		if (strpos($text, '](') !== FALSE and preg_match_all('/(!?)(\[((?:[^\[\]]|(?2))*)\])\((.*?)\)/', $text, $matches, PREG_SET_ORDER)) # inline 
+		if (strpos($text, '](') !== FALSE and preg_match_all('/(!?)(\[((?:[^\[\]]|(?2))*)\])\((.*?)\)/', $text, $matches, PREG_SET_ORDER))
 		{
 			foreach ($matches as $matches)
 			{
@@ -572,9 +643,9 @@ class Parsedown
 				{
 					$element = '<img alt="'.$matches[3].'" src="'.$url.'">';
 				}
-				else
+				else # link
 				{
-					$element_text = $this->parse_inline_elements($matches[3]);
+					$element_text = $this->parse_span_elements($matches[3]);
 
 					$element = '<a href="'.$url.'">'.$element_text.'</a>';
 				}
@@ -591,7 +662,7 @@ class Parsedown
 			}
 		}
 
-		# Reference(d) Link / Image 
+		# reference link / reference image (recursive)
 
 		if ($this->reference_map and strpos($text, '[') !== FALSE and preg_match_all('/(!?)\[(.+?)\](?:\n?[ ]?\[(.*?)\])?/ms', $text, $matches, PREG_SET_ORDER))
 		{
@@ -613,9 +684,9 @@ class Parsedown
 					{
 						$element = '<img alt="'.$matches[2].'" src="'.$url.'">';
 					}
-					else # anchor 
+					else # link
 					{
-						$element_text = $this->parse_inline_elements($matches[2]);
+						$element_text = $this->parse_span_elements($matches[2]);
 
 						$element = '<a href="'.$url.'">'.$element_text.'</a>';
 					}
@@ -633,21 +704,26 @@ class Parsedown
 			}
 		}
 
-		# Automatic Links 
+		# code span
 
-		if (strpos($text, '<') !== FALSE and preg_match_all('/<((https?|ftp|dict):[^\^\s]+?)>/i', $text, $matches, PREG_SET_ORDER))
+		if (strpos($text, '`') !== FALSE and preg_match_all('/`(.+?)`/', $text, $matches, PREG_SET_ORDER))
 		{
 			foreach ($matches as $matches)
 			{
-				$url = $matches[1];
+				$element_text = $matches[1];
+				$element_text = htmlspecialchars($element_text, ENT_NOQUOTES, 'UTF-8');
 
-				strpos($url, '&') !== FALSE and $url = preg_replace('/&(?!#?\w+;)/', '&amp;', $url);
+				# decodes escape sequences
 
-				$element = '<a href=":href">:text</a>';
-				$element = str_replace(':text', $url, $element);
-				$element = str_replace(':href', $url, $element);
+				$this->escape_sequence_map
+					and strpos($element_text, "\x1A") !== FALSE
+					and $element_text = strtr($element_text, $this->escape_sequence_map);
 
-				# ~ 
+				# composes element
+
+				$element = '<code>'.$element_text.'</code>';
+
+				# encodes element
 
 				$code = "\x1A".'$'.$index;
 
@@ -659,6 +735,40 @@ class Parsedown
 			}
 		}
 
+		# automatic link
+
+		if (strpos($text, '://') !== FALSE)
+		{
+			switch (TRUE)
+			{
+				case preg_match_all('{<(https?:[/]{2}[^\s]+)>}i', $text, $matches, PREG_SET_ORDER):
+				case preg_match_all('{\b(https?:[/]{2}[^\s]+)\b}i', $text, $matches, PREG_SET_ORDER):
+
+					foreach ($matches as $matches)
+					{
+						$url = $matches[1];
+
+						strpos($url, '&') !== FALSE and $url = preg_replace('/&(?!#?\w+;)/', '&amp;', $url);
+
+						$element = '<a href=":href">:text</a>';
+						$element = str_replace(':text', $url, $element);
+						$element = str_replace(':href', $url, $element);
+
+						# ~
+
+						$code = "\x1A".'$'.$index;
+
+						$text = str_replace($matches[0], $code, $text);
+
+						$map[$code] = $element;
+
+						$index ++;
+					}
+
+					break;
+			}
+		}
+
 		# ~
 
 		strpos($text, '&') !== FALSE and $text = preg_replace('/&(?!#?\w+;)/', '&amp;', $text);
@@ -666,16 +776,23 @@ class Parsedown
 
 		# ~
 
+		if (strpos($text, '~~') !== FALSE)
+		{
+			$text = preg_replace('/~~(?=\S)(.+?)(?<=\S)~~/s', '<del>$1</del>', $text);
+		}
+
 		if (strpos($text, '_') !== FALSE)
 		{
-			$text = preg_replace('/__(?=\S)(.+?)(?<=\S)__/', '<strong>$1</strong>', $text);
-			$text = preg_replace('/_(?=\S)(.+?)(?<=\S)_/', '<em>$1</em>', $text);
+			$text = preg_replace('/__(?=\S)([^_]+?)(?<=\S)__/s', '<strong>$1</strong>', $text, -1, $count);
+			$text = preg_replace('/(\b|_)_(?=\S)([^_]+?)(?<=\S)_(\b|_)/s', '$1<em>$2</em>$3', $text);
+			$text = preg_replace('/__(?=\S)([^_]+?)(?<=\S)__/s', '<strong>$1</strong>', $text, -1, $count);
 		}
 
 		if (strpos($text, '*') !== FALSE)
 		{
-			$text = preg_replace('/\*\*(?=\S)(.+?)(?<=\S)\*\*/', '<strong>$1</strong>', $text);
-			$text = preg_replace('/\*(?=\S)(.+?)(?<=\S)\*/', '<em>$1</em>', $text);
+			$text = preg_replace('/\*\*(?=\S)([^*]+?)(?<=\S)\*\*/s', '<strong>$1</strong>', $text);
+			$text = preg_replace('/\*(?=\S)([^*]+?)(?<=\S)\*/s', '<em>$1</em>', $text);
+			$text = preg_replace('/\*\*(?=\S)([^*]+?)(?<=\S)\*\*/s', '<strong>$1</strong>', $text);
 		}
 
 		$text = strtr($text, $map);

README.md

@@ -1,9 +1,20 @@
-## Parsedown PHP
+## Parsedown
 
-Parsedown is a parser for Markdown. It parses Markdown text the way people do. First, it divides texts into blocks. Then it looks at how these blocks start and how they relate to each other. Finally, it looks for special characters to identify inline elements. As a result, Parsedown is (super) fast, consistent and clean.
+Better [Markdown](http://daringfireball.net/projects/markdown/) parser for PHP.
 
-[Explorer (demo)](http://parsedown.org/explorer/)  
-[Tests](http://parsedown.org/tests/)
+***
+
+[demo](http://parsedown.org/demo) · [tests](http://parsedown.org/tests/)
+
+***
+
+### Features
+
+* [fast](http://parsedown.org/speed)
+* [consistent](http://parsedown.org/consistency)
+* [GitHub Flavored](https://help.github.com/articles/github-flavored-markdown)
+* [tested](https://travis-ci.org/erusev/parsedown) in PHP 5.2, 5.3, 5.4 and 5.5
+* friendly to international input
 
 ### Installation
 

tests/Test.php

@@ -20,20 +20,29 @@ class Test extends PHPUnit_Framework_TestCase
 	{
 		$provider = array();
 
-		$DirectoryIterator = new DirectoryIterator(__DIR__ . '/' . self::provider_dir);
+		$path = dirname(__FILE__).'/';
+
+		$DirectoryIterator = new DirectoryIterator($path . '/' . self::provider_dir);
 
 		foreach ($DirectoryIterator as $Item)
 		{
-			if ($Item->isFile() and $Item->getExtension() === 'md')
+			if ($Item->isFile())
 			{
+				$filename = $Item->getFilename();
+
+				$extension = pathinfo($filename, PATHINFO_EXTENSION);
+
+				if ($extension !== 'md')
+					continue;
+
 				$basename = $Item->getBasename('.md');
 
-				$markdown = file_get_contents(__DIR__ . '/' . self::provider_dir . $basename . '.md');
+				$markdown = file_get_contents($path . '/' . self::provider_dir . $basename . '.md');
 
 				if (!$markdown)
 					continue;
 
-				$expected_markup = file_get_contents(__DIR__ . '/' . self::provider_dir . $basename . '.html');
+				$expected_markup = file_get_contents($path . '/' . self::provider_dir . $basename . '.html');
 				$expected_markup = str_replace("\r\n", "\n", $expected_markup);
 				$expected_markup = str_replace("\r", "\n", $expected_markup);
 
@@ -44,4 +53,3 @@ class Test extends PHPUnit_Framework_TestCase
 		return $provider;
 	}
 }
-

tests/data/atx_heading.html

@@ -1,6 +1,7 @@
-<h1>This is an h1</h1>
-<h2>This is an h2</h2>
-<h3>This is an h3</h3>
-<h4>This is an h4</h4>
-<h5>This is an h5</h5>
-<h6>This is an h6</h6>
+<h1>h1</h1>
+<h2>h2</h2>
+<h3>h3</h3>
+<h4>h4</h4>
+<h5>h5</h5>
+<h6>h6</h6>
+<h1>closed h1</h1>

tests/data/atx_heading.md

@@ -1,11 +1,13 @@
-# This is an h1
+# h1
 
-## This is an h2
+## h2
 
-### This is an h3
+### h3
 
-#### This is an h4
+#### h4
 
-##### This is an h5
+##### h5
 
-###### This is an h6
+###### h6
+
+# closed h1 #

tests/data/block-level_html.html

@@ -0,0 +1,9 @@
+<div>content</div>
+<hr style="background: #eaa;" />
+<p>nested elements:</p>
+<div>
+parent
+<div>
+child
+</div>
+</div>

tests/data/block-level_html.md

@@ -0,0 +1,12 @@
+<div>content</div>
+
+<hr style="background: #eaa;" />
+
+nested elements:
+
+<div>
+parent
+<div>
+child
+</div>
+</div>

tests/data/blockquote.html

@@ -1,13 +0,0 @@
-<p>Here's a regular blockquote:</p>
-<blockquote>
-<p>blockquote</p>
-</blockquote>
-<p>Here's one with no space after the ">":</p>
-<blockquote>
-<p>blockquote</p>
-</blockquote>
-<p>Here's one on multiple lines:</p>
-<blockquote>
-<p>line 1
-line 2</p>
-</blockquote>

tests/data/blockquote.md

@@ -1,12 +0,0 @@
-Here's a regular blockquote:
-
-> blockquote
-
-Here's one with no space after the ">":
-
->blockquote
-
-Here's one on multiple lines:
-
-> line 1
-> line 2

tests/data/closed_atx_heading.html

@@ -1,6 +0,0 @@
-<h1>h1</h1>
-<h2>h2</h2>
-<h3>h3</h3>
-<h4>h4</h4>
-<h5>h5</h5>
-<h6>h6</h6>

tests/data/closed_atx_heading.md

@@ -1,11 +0,0 @@
-# h1 #
-
-## h2 ##
-
-### h3 ###
-
-#### h4 ####
-
-##### h5 #####
-
-###### h6 ######

tests/data/code_block.html

@@ -1,8 +1,8 @@
-<p>Here's a code block:</p>
 <pre><code>&lt;?php
 
 $message = 'Hello World!'; 
 echo $message;</code></pre>
-<p>Here's one that holds a list:</p>
-<pre><code>- list item
-- another list item</code></pre>
+<hr />
+<pre><code>&gt; not a quote 
+- not a list item
+[not a reference]: http://foo.com</code></pre>

tests/data/code_block.md

@@ -1,12 +1,11 @@
-Here's a code block:
-
     <?php
     
     $message = 'Hello World!'; 
     echo $message;
 
-Here's one that holds a list:
+---
 
-    - list item
-    - another list item
+    > not a quote 
+    - not a list item
+    [not a reference]: http://foo.com
     

tests/data/code_span.html

@@ -1 +1 @@
-<p>This is a <code>code span</code>.</p>
+<p>a <code>code span</code></p>

tests/data/code_span.md

@@ -1 +1 @@
-This is a `code span`.
+a `code span`

tests/data/compound_blockquote.html

@@ -1,16 +1,9 @@
-<p>Here's one with multiple paragraphs:</p>
 <blockquote>
-<p>This is line one.</p>
-<p>This is line two.</p>
-</blockquote>
-<p>Here's one with multiple types of blocks:</p>
-<blockquote>
-<p>This is a quoted paragraph.</p>
+<h2>header</h2>
+<p>paragraph</p>
 <ul>
-<li>This is a list item of a quoted list.</li>
-<li>This is another list item.</li>
+<li>li</li>
 </ul>
-<blockquote>
-<p>This is a nested quote block.</p>
-</blockquote>
+<hr />
+<p>paragraph</p>
 </blockquote>

tests/data/compound_blockquote.md

@@ -1,14 +1,10 @@
-Here's one with multiple paragraphs:
-
-> This is line one.
+> header
+> ------
 >
-> This is line two.
-
-Here's one with multiple types of blocks:
-
-> This is a quoted paragraph.
+> paragraph
 >
-> - This is a list item of a quoted list.
-> - This is another list item.
+> - li
 >
-> > This is a nested quote block.
+> ---
+>
+> paragraph

tests/data/compound_list.html

@@ -1,13 +1,12 @@
-<p>Here's a compound list:</p>
 <ul>
 <li>
-<p>This is the first paragraph of the list item.</p>
-<p>This is the second one.</p>
+<p>paragraph</p>
+<p>paragraph</p>
 </li>
 <li>
-<p>This is another list item.</p>
+<p>paragraph</p>
 <blockquote>
-<p>This is a quote block that belongs to it.</p>
+<p>quote</p>
 </blockquote>
 </li>
 </ul>

tests/data/compound_list.md

@@ -1,10 +1,7 @@
-Here's a compound list:
+- paragraph
 
-- This is the first paragraph of the list item.
+  paragraph
 
-  This is the second one.
-
-- This is another list item.
-
-  > This is a quote block that belongs to it.
+- paragraph
 
+  > quote

tests/data/dense_list.html

@@ -1,24 +0,0 @@
-<p>Here's a regular list:</p>
-<ul>
-<li>list item</li>
-<li>another list item</li>
-<li>3rd list item</li>
-</ul>
-<p>Here's one with white space around items:</p>
-<ul>
-<li>list item  </li>
-<li>another list item  </li>
-</ul>
-<p>Here's one with too much space before items:</p>
-<pre><code>- list item
-- another list item</code></pre>
-<p>Here's one with no space after markers:</p>
-<p>-list item
--another list item</p>
-<p>Here's one where items contain line breaks:</p>
-<ul>
-<li>list
-item</li>
-<li>another
-list item</li>
-</ul>

tests/data/dense_list.md

@@ -1,27 +0,0 @@
-Here's a regular list:
-
-- list item
-- another list item
-- 3rd list item
-
-Here's one with white space around items:
-
-  - list item  
-  - another list item  
-
-Here's one with too much space before items:
-
-    - list item
-    - another list item
-
-Here's one with no space after markers:
-
--list item
--another list item
-
-Here's one where items contain line breaks:
-
-- list
-item
-- another
-list item

tests/data/em_strong.html

@@ -0,0 +1,6 @@
+<p><strong><em>em strong</em> strong</strong></p>
+<p><strong>strong <em>em strong</em></strong></p>
+<p><strong>strong <em>em strong</em> strong</strong></p>
+<p><strong><em>em strong</em> strong</strong></p>
+<p><strong>strong <em>em strong</em></strong></p>
+<p><strong>strong <em>em strong</em> strong</strong></p>

tests/data/em_strong.md

@@ -0,0 +1,11 @@
+___em strong_ strong__
+
+__strong _em strong___
+
+__strong _em strong_ strong__
+
+***em strong* strong**
+
+**strong *em strong***
+
+**strong *em strong* strong**

tests/data/emphasis.html

@@ -1,7 +1,7 @@
-<p>Here's <em>an emphasis</em>.</p>
-<p>A short emphasis <em>a</em> <em>b</em> .</p>
-<p>Here's <strong>a strong one</strong>. </p>
-<p>Here's <em>an emphasis that uses underscores</em>. </p>
-<p>Here's <strong>a strong emphasis that uses underscores</strong>.</p>
-<p>This is not _ an emphasis _ neither is * this * neither is _ this_ neither is _this _.</p>
-<p>Empty emphasis ** is not __ an emphasis.</p>
+<p><em>underscore</em>, <em>asterisk</em>, <em>one two</em>, <em>three four</em>, <em>a</em>, <em>b</em></p>
+<p><em>multiline
+emphasis</em></p>
+<p>_ this _ is not an emphasis, neither is _ this_, _this _, or _this*</p>
+<p>this_is_not_an_emphasis</p>
+<p>an empty emphasis __ ** is not an emphasis</p>
+<p>*mixed *<em>double and</em> single asterisk** spans</p>

tests/data/emphasis.md

@@ -1,13 +1,12 @@
-Here's *an emphasis*.
+_underscore_, *asterisk*, _one two_, *three four*, _a_, *b*
 
-A short emphasis _a_ *b* .
+_multiline
+emphasis_
 
-Here's **a strong one**. 
+_ this _ is not an emphasis, neither is _ this_, _this _, or _this*
 
-Here's _an emphasis that uses underscores_. 
+this_is_not_an_emphasis
 
-Here's __a strong emphasis that uses underscores__.
+an empty emphasis __ ** is not an emphasis
 
-This is not _ an emphasis _ neither is * this * neither is _ this_ neither is _this _.
-
-Empty emphasis ** is not __ an emphasis.
+*mixed **double and* single asterisk** spans

tests/data/escaping.html

@@ -1,6 +1,4 @@
-<p>Here's an <em>emphasis</em> and here's an escaped *emphasis*. Here are also an escaped `code span`, escaped [inline link](http://example.com).</p>
-<p>Here's <code>an escaped \*emphasis\* inside of a code span</code>.</p>
-<p>Here's one inside of a code block:</p>
-<pre><code>An escaped \*emphasis\*.</code></pre>
-<p>Finally, an escaped reference:</p>
-<p>[1]: http://example.com</p>
+<p>escaped *emphasis*.</p>
+<p><code>escaped \*emphasis\* in a code span</code></p>
+<pre><code>escaped \*emphasis\* in a code block</code></pre>
+<p>\ ` * _ { } [ ] ( ) > # + - . !</p>

tests/data/escaping.md

@@ -1,11 +1,7 @@
-Here's an *emphasis* and here's an escaped \*emphasis\*. Here are also an escaped \`code span\`, escaped \[inline link](http://example.com).
+escaped \*emphasis\*.
 
-Here's `an escaped \*emphasis\* inside of a code span`.
+`escaped \*emphasis\* in a code span`
 
-Here's one inside of a code block:
+    escaped \*emphasis\* in a code block
 
-    An escaped \*emphasis\*.
-
-Finally, an escaped reference:
-
-\[1]: http://example.com
+\\ \` \* \_ \{ \} \[ \] \( \) \> \# \+ \- \. \!

tests/data/fenced_code_block.html

@@ -0,0 +1,5 @@
+<pre><code>&lt;?php
+
+$message = 'fenced code block';
+echo $message;</code></pre>
+<pre><code>tilde</code></pre>

tests/data/fenced_code_block.md

@@ -0,0 +1,10 @@
+```
+<?php
+
+$message = 'fenced code block';
+echo $message;
+```
+
+~~~
+tilde
+~~~

tests/data/horizontal_rule.html

@@ -1,16 +1,5 @@
-<p>Dashes:</p>
 <hr />
 <hr />
 <hr />
 <hr />
-<pre><code>---</code></pre>
 <hr />
-<hr />
-<hr />
-<hr />
-<pre><code>- - -</code></pre>
-<p>Asterisks:</p>
-<hr />
-<p>Underscores:</p>
-<hr />
-<p>Based on <a href="http://daringfireball.net/projects/downloads/MarkdownTest_1.0.zip">the original</a> test suite.</p>

tests/data/horizontal_rule.md

@@ -1,31 +1,9 @@
-Dashes:
-
 ---
 
- ---
- 
-  ---
-
-   ---
-
-	---
-
 - - -
 
- - - -
- 
-  - - -
-
    - - -
 
-	- - -
-
-Asterisks:
-
 ***
 
-Underscores:
-
 ___
-
-Based on [the original](http://daringfireball.net/projects/downloads/MarkdownTest_1.0.zip) test suite.

tests/data/html.html

@@ -1,15 +0,0 @@
-<p>Self-closing tag:</p>
-<hr/>
-<p>Self-closing tag with attributes:</p>
-<hr style="background: #eaa" />
-<p>Bare element:</p>
-<div>content</div>
-<p>Element with attributes:</p>
-<a href="http://parsedown.org">link</a>
-<p>Nested elements:</p>
-<div>
-parent
-<div>
-child
-</div>
-</div>

tests/data/html.md

@@ -1,24 +0,0 @@
-Self-closing tag:
-
-<hr/>
-
-Self-closing tag with attributes:
-
-<hr style="background: #eaa" />
-
-Bare element:
-
-<div>content</div>
-
-Element with attributes:
-
-<a href="http://parsedown.org">link</a>
-
-Nested elements:
-
-<div>
-parent
-<div>
-child
-</div>
-</div>

tests/data/image_reference.html

@@ -0,0 +1 @@
+<p><img alt="Markdown Logo" src="/md.png"></p>

tests/data/image_reference.md

@@ -0,0 +1,3 @@
+![Markdown Logo][image]
+
+[image]: /md.png

tests/data/implicit_reference.html

@@ -0,0 +1,2 @@
+<p>an <a href="http://example.com">implicit</a> reference link</p>
+<p>an <a href="http://example.com">implicit</a> reference link with an empty link definition</p>

tests/data/implicit_reference.md

@@ -0,0 +1,5 @@
+an [implicit] reference link
+
+[implicit]: http://example.com
+
+an [implicit][] reference link with an empty link definition

tests/data/inline_link.html

@@ -1,2 +1,3 @@
-<p>Here's a <a href="http://parsedown.org">link</a>.</p>
-<p>Here's an image link: <a href="http://daringfireball.net/projects/markdown/"><img alt="MD Logo" src="http://parsedown.org/md.png"></a>.</p>
+<p><a href="http://example.com">link</a></p>
+<p><a href="http://example.com"><code>link</code></a></p>
+<p><a href="http://example.com"><img alt="MD Logo" src="http://parsedown.org/md.png"></a></p>

tests/data/inline_link.md

@@ -1,3 +1,5 @@
-Here's a [link](http://parsedown.org).
+[link](http://example.com)
 
-Here's an image link: [![MD Logo](http://parsedown.org/md.png)](http://daringfireball.net/projects/markdown/).
+[`link`](http://example.com)
+
+[![MD Logo](http://parsedown.org/md.png)](http://example.com)

tests/data/lazy_blockquote.html

@@ -1,4 +1,4 @@
 <blockquote>
-<p>line 1
-line 2</p>
+<p>quote
+the rest of it</p>
 </blockquote>

tests/data/lazy_blockquote.md

@@ -1,2 +1,2 @@
-> line 1
-line 2
+> quote
+the rest of it

tests/data/lazy_list.html

@@ -0,0 +1,4 @@
+<ul>
+<li>li
+the rest of it</li>
+</ul>

tests/data/lazy_list.md

@@ -0,0 +1,2 @@
+- li
+the rest of it

tests/data/lazy_list_item.html

@@ -1,4 +0,0 @@
-<ul>
-<li>li
-more text</li>
-</ul>

tests/data/lazy_list_item.md

@@ -1,2 +0,0 @@
-- li
-more text

tests/data/multiline_list_paragraph.html

@@ -0,0 +1,7 @@
+<ul>
+<li>
+<p>li</p>
+<p>line
+line</p>
+</li>
+</ul>

tests/data/multiline_list_paragraph.md

@@ -0,0 +1,4 @@
+- li
+
+  line
+  line

tests/data/ordered_list.html

@@ -1,16 +1,13 @@
-<p>Here's a regular ordered list:</p>
-<ol>
-<li>one</li>
-<li>two</li>
-<li>three</li>
-</ol>
-<p>Here's one with repeating numbers:</p>
 <ol>
 <li>one</li>
 <li>two</li>
 </ol>
-<p>Here's one with large numbers:</p>
+<p>repeating numbers:</p>
 <ol>
 <li>one</li>
 <li>two</li>
 </ol>
+<p>large numbers:</p>
+<ol>
+<li>one</li>
+</ol>

tests/data/ordered_list.md

@@ -1,16 +1,11 @@
-Here's a regular ordered list:
-
 1. one
 2. two
-3. three
 
-Here's one with repeating numbers:
+repeating numbers:
 
 1. one
 1. two
 
-Here's one with large numbers:
+large numbers:
 
 123. one
-123. two
-

tests/data/paragraph_blockquote.html

@@ -1,4 +0,0 @@
-<p>Here's a paragraph.</p>
-<blockquote>
-<p>a block quote that belongs to it.</p>
-</blockquote>

tests/data/paragraph_blockquote.md

@@ -1,2 +0,0 @@
-Here's a paragraph.
-> a block quote that belongs to it.

tests/data/paragraph_list.html

@@ -1,5 +0,0 @@
-<p>Here's a list that's "inside" a paragraph:</p>
-<ul>
-<li>list item</li>
-<li>another list item</li>
-</ul>

tests/data/paragraph_list.md

@@ -1,4 +0,0 @@
-Here's a list that's "inside" a paragraph:
-- list item
-- another list item
-

tests/data/quote_block.html

@@ -1,20 +0,0 @@
-<p>Here's a regular quote block:</p>
-<blockquote>
-<p>Some quoted text.
-Here goes some more.</p>
-</blockquote>
-<p>Here's one with space before lines:</p>
-<blockquote>
-<p>Some quoted text.
-Here goes some more.</p>
-</blockquote>
-<p>Here's one with no space after >:</p>
-<blockquote>
-<p>Some quoted text.
-Here goes some more.</p>
-</blockquote>
-<p>Here's one with no > on the second line:</p>
-<blockquote>
-<p>Some quoted text.
-Here goes some more.</p>
-</blockquote>

tests/data/quote_block.md

@@ -1,19 +0,0 @@
-Here's a regular quote block:
-
-> Some quoted text.
-> Here goes some more.
-
-Here's one with space before lines:
-
-   > Some quoted text.
-   > Here goes some more.
-
-Here's one with no space after >:
-
->Some quoted text.
->Here goes some more.
-
-Here's one with no > on the second line:
-
-> Some quoted text.
-Here goes some more.

tests/data/reference_link.html

@@ -1,11 +0,0 @@
-<p>Here's a <a href="http://parsedown.org">reference link</a>.</p>
-<p>Here's <a href="http://parsedown.org">one</a> on the next line.</p>
-<p>Here's <a href="http://parsedown.org/tests/">one</a> with a different URL.</p>
-<p>Here's <a href="http://parsedown.org">one</a> with a semantic name.</p>
-<p>Here's <a href="http://parsedown.org">one</a> with definition name on the next line.</p>
-<p>Here's [one][404] with no definition.</p>
-<p>Here's an image: <img alt="Markdown Logo" src="/md.png"></p>
-<p>Here's an <a href="http://google.com">implicit one</a>.</p>
-<p>Here's an <a href="http://google.com">implicit one</a> with an empty link definition.</p>
-<p>Here's a <a href="http://parsedown.org">multiline
-one</a> defined on 2 lines.</p>

tests/data/reference_link.md

@@ -1,32 +0,0 @@
-Here's a [reference link][1].
-
-[1]: http://parsedown.org
-
-Here's [one][2] on the next line.
-[2]: http://parsedown.org
-
-Here's [one][3] with a different URL.
-
-[3]: http://parsedown.org/tests/
-
-Here's [one][website] with a semantic name.
-
-[website]: http://parsedown.org
-
-Here's [one]
-[website] with definition name on the next line.
-
-Here's [one][404] with no definition.
-
-Here's an image: ![Markdown Logo][image]
-
-[image]: /md.png
-
-Here's an [implicit one].
-
-[implicit one]: http://google.com
-
-Here's an [implicit one][] with an empty link definition.
-
-Here's a [multiline
-one][website] defined on 2 lines.

tests/data/setext_heading.html

@@ -1,6 +0,0 @@
-<h1>Heading 1</h1>
-<h2>Heading 2</h2>
-<h2>Block Heading</h2>
-<p>This is the rest of the block.</p>
-<h1>Single "="</h1>
-<h2>Single "-"</h2>

tests/data/setext_heading.md

@@ -1,16 +0,0 @@
-Heading 1
-=========
-
-Heading 2
----------
-
-Block Heading
--------------
-This is the rest of the block.
-
-Single "="
-=
-
-Single "-"
--
-

tests/data/simple_blockquote.html

@@ -0,0 +1,11 @@
+<blockquote>
+<p>quote</p>
+</blockquote>
+<p>indented:</p>
+<blockquote>
+<p>quote</p>
+</blockquote>
+<p>no space after <code>&gt;</code>:</p>
+<blockquote>
+<p>quote</p>
+</blockquote>

tests/data/simple_blockquote.md

@@ -0,0 +1,7 @@
+> quote
+
+indented:
+   > quote
+
+no space after `>`:
+>quote

tests/data/span-level_html.html

@@ -0,0 +1,3 @@
+<p>an <b>important</b> <a href=''>link</a></p>
+<p>broken<br/>
+line</p>

tests/data/span-level_html.md

@@ -0,0 +1,4 @@
+an <b>important</b> <a href=''>link</a>
+
+broken<br/>
+line

tests/data/span_level_html.html

@@ -1 +0,0 @@
-<p>Here's an <b>important</b> <a href=''>link</a>.</p>

tests/data/span_level_html.md

@@ -1 +0,0 @@
-Here's an <b>important</b> <a href=''>link</a>.

tests/data/sparse_list.html

@@ -1,16 +1,15 @@
-<p>Here's a sparse list:</p>
 <ul>
 <li>
-<p>list item</p>
+<p>li</p>
 </li>
-<li>another list item</li>
+<li>li</li>
 </ul>
-<p>Here's one with an indented list item:</p>
+<hr />
 <ul>
 <li>
 <p>li</p>
 <ul>
-<li>li</li>
+<li>indented li</li>
 </ul>
 </li>
 </ul>

tests/data/sparse_list.md

@@ -1,11 +1,9 @@
-Here's a sparse list:
-
-- list item
-
-- another list item
-
-Here's one with an indented list item:
+- li
 
 - li
 
-    - li
+---
+
+- li
+
+    - indented li

tests/data/special_characters.html

@@ -1,9 +1,7 @@
-<p>AT&amp;T has an ampersand in their name.</p>
-<p>AT&amp;T is another way to write it.</p>
-<p>This &amp; that.</p>
-<p>4 &lt; 5 and 6 > 5.</p>
-<p>Here's a <a href="http://example.com/?foo=1&amp;bar=2">link</a> with an ampersand in the URL.</p>
-<p>Here's an inline <a href="/script?foo=1&amp;bar=2">link</a>.</p>
-<p><a href="http://example.com/autolink?a=1&amp;b=2">http://example.com/autolink?a=1&amp;b=2</a></p>
-<hr />
-<p>Based on <a href="http://daringfireball.net/projects/downloads/MarkdownTest_1.0.zip">the original</a> test suite.</p>
+<p>AT&amp;T has an ampersand in their name</p>
+<p>AT&amp;T is another way to write it</p>
+<p>this &amp; that</p>
+<p>4 &lt; 5 and 6 > 5</p>
+<p><a href="http://example.com/autolink?a=1&amp;b=2">http://example.com/autolink?a=1&amp;b=2</a> </p>
+<p><a href="/script?a=1&amp;b=2">inline link</a></p>
+<p><a href="http://example.com/?a=1&amp;b=2">reference link</a></p>

tests/data/special_characters.md

@@ -1,19 +1,15 @@
-AT&T has an ampersand in their name.
+AT&T has an ampersand in their name
 
-AT&T is another way to write it.
+AT&T is another way to write it
 
-This & that.
+this & that
 
-4 < 5 and 6 > 5.
-
-Here's a [link] [1] with an ampersand in the URL.
-
-Here's an inline [link](/script?foo=1&bar=2).
-
-[1]: http://example.com/?foo=1&bar=2
+4 < 5 and 6 > 5
 
 <http://example.com/autolink?a=1&b=2> 
 
----
+[inline link](/script?a=1&b=2)
 
-Based on [the original](http://daringfireball.net/projects/downloads/MarkdownTest_1.0.zip) test suite.
+[reference link][1]
+
+[1]: http://example.com/?a=1&b=2

tests/data/strikethrough.html

@@ -0,0 +1,3 @@
+<p><del>strikethrough</del></p>
+<p>in the <del>middle</del> of a sentence</p>
+<p>in the middle of a w<del>or</del>d</p>

tests/data/strikethrough.md

@@ -0,0 +1,5 @@
+~~strikethrough~~
+
+in the ~~middle~~ of a sentence
+
+in the middle of a w~~or~~d

tests/data/strong_em.html

@@ -0,0 +1,6 @@
+<p><em><strong>strong em</strong></em> </p>
+<p><em>em <strong>strong em</strong></em></p>
+<p><em><strong>strong em</strong> em</em></p>
+<p><em><strong>strong em</strong></em></p>
+<p><em>em <strong>strong em</strong></em></p>
+<p><em><strong>strong em</strong> em</em></p>

tests/data/strong_em.md

@@ -0,0 +1,11 @@
+***strong em*** 
+
+*em **strong em***
+
+***strong em** em*
+
+___strong em___
+
+_em __strong em___
+
+___strong em__ em_

tests/data/text_reference.html

@@ -0,0 +1,7 @@
+<p><a href="http://example.com">reference link</a></p>
+<p><a href="http://example.com">one</a> with a semantic name</p>
+<p>[one][404] with no definition</p>
+<p><a href="http://example.com">multiline
+one</a> defined on 2 lines</p>
+<p><a href="http://example.com">one</a> with an upper case label</p>
+<p><a href="http://example.com"><code>link</code></a></p>

tests/data/text_reference.md

@@ -0,0 +1,18 @@
+[reference link][1]
+
+[1]: http://example.com
+
+[one][website] with a semantic name
+
+[website]: http://example.com
+
+[one][404] with no definition
+
+[multiline
+one][website] defined on 2 lines
+
+[one][label] with an upper case label
+
+[LABEL]: http://example.com
+
+[`link`][website]

tests/data/unordered_list.html

@@ -1,20 +1,10 @@
-<p>Here's a regular unordered list:</p>
 <ul>
-<li>list item</li>
-<li>another list item</li>
-<li>3rd list item</li>
+<li>li</li>
+<li>li</li>
 </ul>
-<p>Here's one with a variety of markers:</p>
+<p>mixed markers:</p>
 <ul>
-<li>hyphen</li>
-<li>plus</li>
-<li>asterisk</li>
+<li>li</li>
+<li>li</li>
+<li>li</li>
 </ul>
-<p>Here's one with white space around items:</p>
-<ul>
-<li>list item  </li>
-<li>another list item  </li>
-</ul>
-<p>Here's one with no space after markers:</p>
-<p>-list item
--another list item</p>

tests/data/unordered_list.md

@@ -1,21 +1,8 @@
-Here's a regular unordered list:
+- li
+- li
 
-- list item
-- another list item
-- 3rd list item
+mixed markers:
 
-Here's one with a variety of markers:
-
-- hyphen
-+ plus
-* asterisk
-
-Here's one with white space around items:
-
-  - list item  
-  - another list item  
-
-Here's one with no space after markers:
-
--list item
--another list item
+* li
++ li
+- li

tests/data/url_autolinking.html

@@ -0,0 +1 @@
+<p>Here's an autolink <a href="http://example.com">http://example.com</a>.</p>

tests/data/url_autolinking.md

@@ -0,0 +1 @@
+Here's an autolink http://example.com.

tests/data/whitespace.html

@@ -0,0 +1 @@
+<pre><code>code</code></pre>

tests/data/whitespace.md

@@ -0,0 +1,5 @@
+    
+
+    code
+
+