Specs: Actually maintain the stack of open nodes.

sky/specs/parsing.md

 Parsing
 =======
 
 Parsing in Sky is a strict pipeline consisting of five stages:
 
 - decoding, which converts incoming bytes into Unicode characters
   using UTF-8.
 
 - normalising, which manipulates the sequence of characters.
 
@@ skipping 761 matching lines @@
      1. If the value of the token contains only U+0020 and U+000A
         characters, and there is no ``t`` element on the _stack of
         open nodes_, then skip the token.
      2. Create a text node _node_ whose character data is the value of
         the token.
      3. Append _node_ to the top node in the _stack of open nodes_.
    - If _token_ is a start tag token,
      1. Create an element _node_ with tag name and attributes given by
         the token.
      2. Append _node_ to the top node in the _stack of open nodes_.
+     3. Push _node_ onto the top of the _stack of open nodes_.
+     4. If _node_ is a ``template`` element, then:
+        1. Let _fragment_ be the ``DocumentFragment`` object that the
+          ``template`` element uses as its template contents container.
+        2. Push _fragment_ onto the top of the _stack of open nodes_.
    - If _token_ is an end tag token:
      1. Let _node_ be the topmost node in the _stack of open nodes_
         whose tag name is the same as the token's tag name, if any. If
         there isn't one, skip this token.
      2. If there's a ``template`` element in the _stack of open
         nodes_ above _node_, then skip this token.
      3. Pop nodes from the _stack of open nodes_ until _node_ has been
         popped.
      4. If _node_'s tag name is ``script``, then yield until there
         are no pending import loads, then execute the script given by
         the element's contents.
 3. Yield until there are no pending import loads.
 3. Fire a ``load`` event at the _parsing context_ object.