JasonWoof Got questions, comments, patches, etc.? Contact Jason Woofenden
fix garbage after </
authorJason Woofenden <jason@jasonwoof.com>
Thu, 24 Dec 2015 05:04:45 +0000 (00:04 -0500)
committerJason Woofenden <jason@jasonwoof.com>
Thu, 24 Dec 2015 05:04:45 +0000 (00:04 -0500)
parse-html.coffee

index b25359d..581002b 100644 (file)
@@ -896,9 +896,9 @@ parse_html = (args) ->
                debug_log "tree: #{serialize_els doc.children, false, true}"
                debug_log "open_els: #{serialize_els open_els, true, true}"
                debug_log "afe: #{serialize_els afe, true, true}"
+               # FIXME CONTINUE do WATWG thing here
                if open_els[0].name is subject and open_els[0].namespace is NS_HTML
-                       el = open_els[0]
-                       open_els.shift()
+                       el = open_els.shift()
                        # remove it from the list of active formatting elements (if found)
                        for t, i in afe
                                if t is el
@@ -3048,25 +3048,27 @@ parse_html = (args) ->
 
        # 8.2.4.9 http://www.w3.org/TR/html5/syntax.html#end-tag-open-state
        tok_state_end_tag_open = ->
-               switch c = txt.charAt(cur++)
-                       when '>'
-                               parse_error()
-                               tok_state = tok_state_data
-                       when '' # EOF
-                               parse_error()
-                               tok_state = tok_state_data
-                               return new_text_node '</'
-                       else
-                               if is_uc_alpha(c)
-                                       tok_cur_tag = new_end_tag c.toLowerCase()
-                                       tok_state = tok_state_tag_name
-                               else if is_lc_alpha(c)
-                                       tok_cur_tag = new_end_tag c
-                                       tok_state = tok_state_tag_name
-                               else
-                                       parse_error()
-                                       tok_cur_tag = new_comment_token '/'
-                                       tok_state = tok_state_bogus_comment
+               c = txt.charAt(cur++)
+               if is_uc_alpha(c)
+                       tok_cur_tag = new_end_tag c.toLowerCase()
+                       tok_state = tok_state_tag_name
+                       return
+               if is_lc_alpha(c)
+                       tok_cur_tag = new_end_tag c
+                       tok_state = tok_state_tag_name
+                       return
+               if c is '>'
+                       parse_error()
+                       tok_state = tok_state_data
+                       return
+               if c is '' # EOF
+                       parse_error()
+                       tok_state = tok_state_data
+                       return new_text_node '</'
+               # Anything else
+               parse_error()
+               tok_cur_tag = new_comment_token c
+               tok_state = tok_state_bogus_comment
                return null
 
        # 8.2.4.10 http://www.w3.org/TR/html5/syntax.html#tag-name-state