simplify test for em strong

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.

Parsedown.php

@@ -568,7 +568,7 @@ class Parsedown
 				case 'code_block':
 				case 'fenced_code_block':
 
-					$text = htmlentities($element['text'], ENT_NOQUOTES);
+					$text = htmlspecialchars($element['text'], ENT_NOQUOTES, 'UTF-8');
 
 					strpos($text, "\x1A\\") !== FALSE and $text = strtr($text, $this->escape_sequence_map);
 
@@ -634,7 +634,7 @@ class Parsedown
 			foreach ($matches as $matches)
 			{
 				$element_text = $matches[1];
-				$element_text = htmlentities($element_text, ENT_NOQUOTES);
+				$element_text = htmlspecialchars($element_text, ENT_NOQUOTES, 'UTF-8');
 
 				# decodes escape sequences
 
@@ -733,29 +733,35 @@ class Parsedown
 			}
 		}
 
-		# automatic link
-
-		if (strpos($text, '<') !== FALSE and preg_match_all('/<((https?|ftp|dict):[^\^\s]+?)>/i', $text, $matches, PREG_SET_ORDER))
+		if (strpos($text, '://') !== FALSE)
 		{
-			foreach ($matches as $matches)
+			switch (TRUE)
 			{
-				$url = $matches[1];
+				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):
 
-				strpos($url, '&') !== FALSE and $url = preg_replace('/&(?!#?\w+;)/', '&amp;', $url);
+					foreach ($matches as $matches)
+					{
+						$url = $matches[1];
 
-				$element = '<a href=":href">:text</a>';
-				$element = str_replace(':text', $url, $element);
-				$element = str_replace(':href', $url, $element);
+						strpos($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);
+						$code = "\x1A".'$'.$index;
 
-				$map[$code] = $element;
+						$text = str_replace($matches[0], $code, $text);
 
-				$index ++;
+						$map[$code] = $element;
+
+						$index ++;
+					}
+
+					break;
 			}
 		}
 
@@ -766,20 +772,30 @@ 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)__(?!_)/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, -1, $count);
+			$count or $text = preg_replace('/__(?=\S)(.+?)(?<=\S)__(?!_)/s', '<strong>$1</strong>', $text);
+
+			$text = preg_replace('/\b_(?=\S)(.+?)(?<=\S)_\b/s', '<em>$1</em>', $text);
 		}
 
 		if (strpos($text, '*') !== FALSE)
 		{
-			$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, -1, $count);
+			$count or $text = preg_replace('/\*\*(?=\S)(.+?)(?<=\S)\*\*(?!\*)/s', '<strong>$1</strong>', $text);
+
+			$text = preg_replace('/\*(?=\S)([^*]+?)(?<=\S)\*/s', '<em>$1</em>', $text, -1, $count);
+			$count or $text = preg_replace('/\*(?=\S)(.+?)(?<=\S)\*(?!\*)/s', '<em>$1</em>', $text);
 		}
 
 		$text = strtr($text, $map);
 
 		return $text;
 	}
-}
+}

README.md

@@ -1,8 +1,8 @@
 ## Parsedown
 
-Parsedown is a Markdown parser for PHP. It is fast, consistent and easy to use.
+Fast and consistent [Markdown][1] parser for PHP.
 
-[Home](http://parsedown.org)  · [Demo](http://parsedown.org/explorer/) · [Tests](http://parsedown.org/tests/)
+[Home](http://parsedown.org) · [Demo](http://parsedown.org/explorer/) · [Tests](http://parsedown.org/tests/)
 
 ### Installation
 
@@ -17,3 +17,5 @@ $result = Parsedown::instance()->parse($text);
 
 echo $result; # prints: <p>Hello <strong>Parsedown</strong>!</p>
 ```
+
+[1]: http://daringfireball.net/projects/markdown/

tests/data/em_strong.html

@@ -1,5 +1,4 @@
-<p><strong><em>em strong</em></strong></p>
-<p><strong><em>one</em> at the start</strong></p>
-<p><strong>one at the <em>end</em></strong></p>
-<p><strong>one <em>in the</em> middle</strong></p>
-<p><strong>one with <em>asterisks</em></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>
+<p><strong>strong <em>em strong</em></strong></p>

tests/data/em_strong.md

@@ -1,9 +1,7 @@
-___em strong___
+___em strong_ strong__
 
-___one_ at the start__
+__strong _em strong___
 
-__one at the _end___
+__strong _em strong_ strong__
 
-__one _in the_ middle__
-
-**one with *asterisks***
+**strong *em strong***

tests/data/emphasis.html

@@ -2,4 +2,5 @@
 <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>

tests/data/emphasis.md

@@ -5,4 +5,6 @@ emphasis_
 
 _ this _ is not an emphasis, neither is _ this_, _this _, or _this*
 
+this_is_not_an_emphasis
+
 an empty emphasis __ ** is not an emphasis

tests/data/special_characters.html

@@ -1,4 +1,5 @@
 <p>AT&amp;T has an ampersand in their name</p>
+<pre><code>Let's play some cards ♠ ♣ ♥ ♦</code></pre>
 <p>AT&amp;T is another way to write it</p>
 <p>this &amp; that</p>
 <p>4 &lt; 5 and 6 > 5</p>

tests/data/special_characters.md

@@ -1,5 +1,7 @@
 AT&T has an ampersand in their name
 
+    Let's play some cards ♠ ♣ ♥ ♦
+
 AT&T is another way to write it
 
 this & that

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/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.