Added <details> and <summary> tags to list of HTML elements with bloc…

…k content model (fixes #155)

Note that the <summary> element was using the span content model
prior to this change but the block content model is better
suited (both are allowed by the HTML5 spec).

lib/kramdown/parser/html.rb

@@ -34,11 +34,11 @@ module Constants
         HTML_ENTITY_RE = /&([\w:][\-\w\.:]*);|&#(\d+);|&\#x([0-9a-fA-F]+);/
 
         HTML_CONTENT_MODEL_BLOCK = %w{address applet article aside button blockquote body
-             dd div dl fieldset figure figcaption footer form header hgroup iframe li map menu nav
-              noscript object section td}
+             dd details div dl fieldset figure figcaption footer form header hgroup iframe li map menu nav
+              noscript object section summary td}
         HTML_CONTENT_MODEL_SPAN  = %w{a abbr acronym b bdo big button cite caption del dfn dt em
              h1 h2 h3 h4 h5 h6 i ins label legend optgroup p q rb rbc
-             rp rt rtc ruby select small span strong sub sup summary th tt}
+             rp rt rtc ruby select small span strong sub sup th tt}
         HTML_CONTENT_MODEL_RAW   = %w{script style math option textarea pre code kbd samp var}
         # The following elements are also parsed as raw since they need child elements that cannot
         # be expressed using kramdown syntax: colgroup table tbody thead tfoot tr ul ol