setters should know nothing

Parsedown.php

@@ -37,11 +37,11 @@ class Parsedown
 	# Setters
 	#
 
-	private $break_marker = "  \n";
+	private $breaks_enabled = false;
 
 	function set_breaks_enabled($breaks_enabled)
 	{
-		$this->break_marker = $breaks_enabled ? "\n" : "  \n";
+		$this->breaks_enabled = $breaks_enabled;
 
 		return $this;
 	}
@@ -51,7 +51,6 @@ class Parsedown
 	#
 
 	private $reference_map = array();
-	private $escape_sequence_map = array();
 
 	#
 	# Public Methods
@@ -66,25 +65,6 @@ class Parsedown
 		# replaces tabs with spaces
 		$text = str_replace("\t", '    ', $text);
 
-		# encodes escape sequences
-
-		if (strpos($text, '\\') !== FALSE)
-		{
-			$escape_sequences = array('\\\\', '\`', '\*', '\_', '\{', '\}', '\[', '\]', '\(', '\)', '\>', '\#', '\+', '\-', '\.', '\!');
-
-			foreach ($escape_sequences as $index => $escape_sequence)
-			{
-				if (strpos($text, $escape_sequence) !== FALSE)
-				{
-					$code = "\x1A".'\\'.$index.';';
-
-					$text = str_replace($escape_sequence, $code, $text);
-
-					$this->escape_sequence_map[$code] = $escape_sequence;
-				}
-			}
-		}
-
 		# ~
 
 		$text = trim($text, "\n");
@@ -93,15 +73,6 @@ class Parsedown
 
 		$text = $this->parse_block_elements($lines);
 
-		# decodes escape sequences
-
-		foreach ($this->escape_sequence_map as $code => $escape_sequence)
-		{
-			$text = str_replace($code, $escape_sequence[1], $text);
-		}
-
-		# ~
-
 		$text = rtrim($text, "\n");
 
 		return $text;
@@ -149,12 +120,12 @@ class Parsedown
 
 					if ( ! isset($element['closed']))
 					{
-						if (preg_match('{<'.$element['root '].'>$}', $line)) # opening tag
+						if (strpos($line, $element['start']) !== false) # opening tag
 						{
 							$element['depth']++;
 						}
 
-						if (preg_match('{</'.$element['root '].'>$}', $line)) # closing tag
+						if (strpos($line, $element['end']) !== false) # closing tag
 						{
 							$element['depth'] > 0
 								? $element['depth']--
@@ -322,8 +293,6 @@ class Parsedown
 
 					if ($element['type'] === 'paragraph' and isset($element['interrupted']) === false)
 					{
-						$is_heading = true;
-
 						$chopped_line = rtrim($line);
 
 						$i = 1;
@@ -332,19 +301,14 @@ class Parsedown
 						{
 							if ($chopped_line[$i] !== $line[0])
 							{
-								$is_heading = false;
-
-								break;
+								break 2;
 							}
 
 							$i++;
 						}
 
-						if ($is_heading)
-						{
-							$element['type'] = 'heading';
-							$element['level'] = $line[0] === '-' ? 2 : 1;
-						}
+						$element['type'] = 'heading';
+						$element['level'] = $line[0] === '-' ? 2 : 1;
 
 						continue 2;
 					}
@@ -358,34 +322,63 @@ class Parsedown
 			{
 				case '<':
 
-					# self-closing tag
+					$position = strpos($deindented_line, '>');
 
-					if (preg_match('{^<.+?/>$}', $deindented_line))
+					if ($position > 1) # tag
 					{
+						$name = substr($deindented_line, 1, $position - 1);
+						$name = rtrim($name);
+
+						if (substr($name, -1) === '/')
+						{
+							$self_closing = true;
+
+							$name = substr($name, 0, -1);
+						}
+
+						$position = strpos($name, ' ');
+
+						if ($position)
+						{
+							$name = substr($name, 0, $position);
+						}
+
+						if ( ! ctype_alpha($name))
+						{
+							break;
+						}
+
+						if (in_array($name, $this->inline_tags))
+						{
+							break;
+						}
+
 						$elements []= $element;
 
-						$element = array(
-							'type' => 'self-closing tag',
-							'text' => $deindented_line,
-						);
+						if (isset($self_closing))
+						{
+							$element = array(
+								'type' => 'self-closing tag',
+								'text' => $deindented_line,
+							);
 
-						continue 2;
-					}
+							unset($self_closing);
 
-					# opening tag
-
-					if (preg_match('{^<(\w+)(?:[ ].*?)?>}', $deindented_line, $matches))
-					{
-						$elements []= $element;
+							continue 2;
+						}
 
 						$element = array(
 							'type' => 'block-level markup',
 							'text' => $deindented_line,
-							'root ' => strtolower($matches[1]),
+							'start' => '<'.$name.'>',
+							'end' => '</'.$name.'>',
 							'depth' => 0,
 						);
 
-						preg_match('{</'.$matches[1].'>\s*$}', $deindented_line) and $element['closed'] = true;
+						if (strpos($deindented_line, $element['end']))
+						{
+							$element['closed'] = true;
+						}
 
 						continue 2;
 					}
@@ -527,6 +520,8 @@ class Parsedown
 				}
 				else
 				{
+					$this->breaks_enabled and $element['text'] .= '  ';
+
 					$element['text'] .= "\n".$line;
 				}
 			}
@@ -586,17 +581,22 @@ class Parsedown
 					break;
 
 				case 'code block':
+
+					$text = htmlspecialchars($element['text'], ENT_NOQUOTES, 'UTF-8');
+
+					$markup .= '<pre><code>'.$text.'</code></pre>'."\n";
+
+					break;
+
 				case 'fenced block':
 
 					$text = htmlspecialchars($element['text'], ENT_NOQUOTES, 'UTF-8');
 
-					strpos($text, "\x1A\\") !== FALSE and $text = strtr($text, $this->escape_sequence_map);
+					$markup .= '<pre><code';
 
-					$markup .= isset($element['language'])
-						? '<pre><code class="language-'.$element['language'].'">'.$text.'</code></pre>'
-						: '<pre><code>'.$text.'</code></pre>';
+					isset($element['language']) and $markup .= ' class="language-'.$element['language'].'"';
 
-					$markup .= "\n";
+					$markup .= '>'.$text.'</code></pre>'."\n";
 
 					break;
 
@@ -638,7 +638,7 @@ class Parsedown
 
 				case 'block-level markup':
 
-					$markup .= $this->parse_span_elements($element['text'])."\n";
+					$markup .= $element['text']."\n";
 
 					break;
 
@@ -651,31 +651,9 @@ class Parsedown
 		return $markup;
 	}
 
-	# ~
-
-	private $strong_regex = array(
-		'*' => '/^[*]{2}([^*]+?)[*]{2}(?![*])/s',
-		'_' => '/^__([^_]+?)__(?!_)/s',
-	);
-
-	private $em_regex = array(
-		'*' => '/^[*]([^*]+?)[*](?![*])/s',
-		'_' => '/^_([^_]+?)[_](?![_])\b/s',
-	);
-
-	private $strong_em_regex = array(
-		'*' => '/^[*]{2}(.*?)[*](.+?)[*](.*?)[*]{2}/s',
-		'_' => '/^__(.*?)_(.+?)_(.*?)__/s',
-	);
-
-	private $em_strong_regex = array(
-		'*' => '/^[*](.*?)[*]{2}(.+?)[*]{2}(.*?)[*]/s',
-		'_' => '/^_(.*?)__(.+?)__(.*?)_/s',
-	);
-
-	private function parse_span_elements($text, $markers = array('![', '&', '*', '<', '[', '_', '`', 'http', '~~'))
+	private function parse_span_elements($text, $markers = array("  \n", '![', '&', '*', '<', '[', '\\', '_', '`', 'http', '~~'))
 	{
-		if (isset($text[2]) === false or $markers === array())
+		if (isset($text[1]) === false or $markers === array())
 		{
 			return $text;
 		}
@@ -711,7 +689,7 @@ class Parsedown
 
 			# ~
 
-			if ($closest_marker === null or isset($text[$closest_marker_position + 2]) === false)
+			if ($closest_marker === null or isset($text[$closest_marker_position + 1]) === false)
 			{
 				$markup .= $text;
 
@@ -732,6 +710,14 @@ class Parsedown
 
 			switch ($closest_marker)
 			{
+				case "  \n":
+
+					$markup .= '<br />'."\n";
+
+					$offset = 3;
+
+					break;
+
 				case '![':
 				case '[':
 
@@ -823,9 +809,18 @@ class Parsedown
 
 				case '&':
 
-					$markup .= '&amp;';
+					if (preg_match('/^&#?\w+;/', $text, $matches))
+					{
+						$markup .= $matches[0];
 
-					$offset = substr($text, 0, 5) === '&amp;' ? 5 : 1;
+						$offset = strlen($matches[0]);
+					}
+					else
+					{
+						$markup .= '&amp;';
+
+						$offset = 1;
+					}
 
 					break;
 
@@ -912,6 +907,23 @@ class Parsedown
 
 					break;
 
+				case '\\':
+
+					if (in_array($text[1], $this->special_characters))
+					{
+						$markup .= $text[1];
+
+						$offset = 2;
+					}
+					else
+					{
+						$markup .= '\\';
+
+						$offset = 1;
+					}
+
+					break;
+
 				case '`':
 
 					if (preg_match('/^`(.+?)`/', $text, $matches))
@@ -919,11 +931,6 @@ class Parsedown
 						$element_text = $matches[1];
 						$element_text = htmlspecialchars($element_text, ENT_NOQUOTES, 'UTF-8');
 
-						if ($this->escape_sequence_map and strpos($element_text, "\x1A") !== false)
-						{
-							$element_text = strtr($element_text, $this->escape_sequence_map);
-						}
-
 						$markup .= '<code>'.$element_text.'</code>';
 
 						$offset = strlen($matches[0]);
@@ -986,8 +993,41 @@ class Parsedown
 			$markers[$closest_marker_index] = $closest_marker;
 		}
 
-		$markup = str_replace($this->break_marker, '<br />'."\n", $markup);
-
 		return $markup;
 	}
+
+	#
+	# Read-only
+	#
+
+	private $inline_tags = array(
+		'a', 'abbr', 'acronym', 'b', 'bdo', 'big', 'br', 'button',
+		'cite', 'code', 'dfn', 'em', 'i', 'img', 'input', 'kbd',
+		'label', 'map', 'object', 'q', 'samp', 'script', 'select', 'small',
+		'span', 'strong', 'sub', 'sup', 'textarea', 'tt', 'var',
+	);
+
+	private $special_characters = array('\\', '`', '*', '_', '{', '}', '[', ']', '(', ')', '>', '#', '+', '-', '.', '!');
+
+	# ~
+
+	private $strong_regex = array(
+		'*' => '/^[*]{2}([^*]+?)[*]{2}(?![*])/s',
+		'_' => '/^__([^_]+?)__(?!_)/s',
+	);
+
+	private $em_regex = array(
+		'*' => '/^[*]([^*]+?)[*](?![*])/s',
+		'_' => '/^_([^_]+?)[_](?![_])\b/s',
+	);
+
+	private $strong_em_regex = array(
+		'*' => '/^[*]{2}(.*?)[*](.+?)[*](.*?)[*]{2}/s',
+		'_' => '/^__(.*?)_(.+?)_(.*?)__/s',
+	);
+
+	private $em_strong_regex = array(
+		'*' => '/^[*](.*?)[*]{2}(.+?)[*]{2}(.*?)[*]/s',
+		'_' => '/^_(.*?)__(.+?)__(.*?)_/s',
+	);
 }

README.md

@@ -14,7 +14,7 @@ Better [Markdown](http://en.wikipedia.org/wiki/Markdown) parser for PHP.
 * [consistent](http://parsedown.org/consistency)
 * [GitHub Flavored](https://help.github.com/articles/github-flavored-markdown)
 * friendly to international input
-* [tested](https://travis-ci.org/erusev/parsedown) in PHP 5.2, 5.3, 5.4 and 5.5 as well as in [HHVM](http://www.hhvm.com/)
+* [tested](https://travis-ci.org/erusev/parsedown) in PHP 5.2, 5.3, 5.4, 5.5 and [hhvm](http://www.hhvm.com/)
 
 ### Installation
 
@@ -23,9 +23,9 @@ Include `Parsedown.php` or install [the composer package](https://packagist.org/
 ### Example
 
 ```php
-$text = 'Hello **Parsedown**!';
+$text = 'Hello _Parsedown_!';
 
 $result = Parsedown::instance()->parse($text);
 
-echo $result; # prints: <p>Hello <strong>Parsedown</strong>!</p>
+echo $result; # prints: <p>Hello <em>Parsedown</em>!</p>
 ```

tests/data/block-level_html.html

@@ -1,9 +1,5 @@
-<div>content</div>
-<hr style="background: #eaa;" />
-<p>nested elements:</p>
+<div>_content_</div>
+<p>sparse:</p>
 <div>
-parent
-<div>
-child
-</div>
+_content_
 </div>

tests/data/block-level_html.md

@@ -1,12 +1,7 @@
-<div>content</div>
+<div>_content_</div>
 
-<hr style="background: #eaa;" />
-
-nested elements:
+sparse:
 
 <div>
-parent
-<div>
-child
-</div>
+_content_
 </div>

tests/data/html_entity.html

@@ -0,0 +1 @@
+<p>&amp; &copy; &#123;</p>

tests/data/html_entity.md

@@ -0,0 +1 @@
+& © {

tests/data/nested_block-level_html.html

@@ -0,0 +1,10 @@
+<div>
+_parent_
+<div>
+_child_
+</div>
+<pre>
+_adopted child_
+</pre>
+</div>
+<p><em>outside</em></p>

tests/data/nested_block-level_html.md

@@ -0,0 +1,11 @@
+<div>
+_parent_
+<div>
+_child_
+</div>
+<pre>
+_adopted child_
+</pre>
+</div>
+
+_outside_

tests/data/paragraph_list.html

@@ -0,0 +1,12 @@
+<p>paragraph</p>
+<ul>
+<li>li</li>
+<li>li</li>
+</ul>
+<p>paragraph</p>
+<ul>
+<li>
+<p>li</p>
+</li>
+<li>li</li>
+</ul>

tests/data/paragraph_list.md

@@ -0,0 +1,9 @@
+paragraph
+- li
+- li
+
+paragraph
+
+   * li
+   
+   * li

tests/data/self-closing_block-level_html.html

@@ -0,0 +1,4 @@
+<hr />
+<p>attributes:</p>
+<hr style="background: #9bd;" />
+<p>...</p>

tests/data/self-closing_block-level_html.md

@@ -0,0 +1,7 @@
+<hr />
+
+attributes:
+
+<hr style="background: #9bd;" />
+
+...

tests/data/span-level_html.html

@@ -1,3 +1,4 @@
 <p>an <b>important</b> <a href=''>link</a></p>
 <p>broken<br/>
-line</p>
+line</p>
+<p><b>inline tag</b> at the beginning</p>

tests/data/span-level_html.md

@@ -1,4 +1,6 @@
 an <b>important</b> <a href=''>link</a>
 
 broken<br/>
-line
+line
+
+<b>inline tag</b> at the beginning

tests/data/special_characters.html

@@ -1,5 +1,4 @@
 <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>

tests/data/special_characters.md

@@ -1,7 +1,5 @@
 AT&T has an ampersand in their name
 
-AT&T is another way to write it
-
 this & that
 
 4 < 5 and 6 > 5