+ if args.error_cb?
+ args.error_cb cur
+ else
+ console.log "Parse error at character #{cur} of #{txt.length}"
+ return
+
+ # http://www.w3.org/TR/html5/syntax.html#push-onto-the-list-of-active-formatting-elements
+ # "Noah's Ark clause" but with three
+ afe_push = (new_el) ->
+ matches = 0
+ for el, i in afe
+ if el.type is TYPE_AFE_MARKER
+ break
+ if el.name is new_el.name and el.namespace is new_el.namespace
+ attrs_match = true
+ for k, v of el.attrs
+ unless new_el.attrs[k] is v
+ attrs_match = false
+ break
+ if attrs_match
+ for k, v of new_el.attrs
+ unless el.attrs[k] is v
+ attrs_match = false
+ break
+ if attrs_match
+ matches += 1
+ if matches is 3
+ afe.splice i, 1
+ break
+ afe.unshift new_el
+ return
+
+ afe_push_marker = ->
+ afe.unshift new_afe_marker()
+ return
+
+ # the functions below impliment the Tree Contstruction algorithm
+ # http://www.w3.org/TR/html5/syntax.html#tree-construction
+
+ # But first... the helpers
+ template_tag_is_open = ->
+ for el in open_els
+ if el.name is 'template' and el.namespace is NS_HTML
+ return true
+ return false
+ is_in_scope_x = (tag_name, scope, namespace) ->
+ for el in open_els
+ if el.name is tag_name and (namespace is null or namespace is el.namespace)
+ return true
+ if scope[el.name] is el.namespace
+ return false
+ return false
+ is_in_scope_x_y = (tag_name, scope, scope2, namespace) ->
+ for el in open_els
+ if el.name is tag_name and (namespace is null or namespace is el.namespace)
+ return true
+ if scope[el.name] is el.namespace
+ return false
+ if scope2[el.name] is el.namespace
+ return false
+ return false
+ standard_scopers = {
+ applet: NS_HTML, caption: NS_HTML, html: NS_HTML, table: NS_HTML,
+ td: NS_HTML, th: NS_HTML, marquee: NS_HTML, object: NS_HTML,
+ template: NS_HTML,
+
+ mi: NS_MATHML, mo: NS_MATHML, mn: NS_MATHML, ms: NS_MATHML,
+ mtext: NS_MATHML, 'annotation-xml': NS_MATHML,
+
+ foreignObject: NS_SVG, desc: NS_SVG, title: NS_SVG
+ }
+ button_scopers = button: NS_HTML
+ li_scopers = ol: NS_HTML, ul: NS_HTML
+ table_scopers = html: NS_HTML, table: NS_HTML, template: NS_HTML
+ is_in_scope = (tag_name, namespace = null) ->
+ return is_in_scope_x tag_name, standard_scopers, namespace
+ is_in_button_scope = (tag_name, namespace = null) ->
+ return is_in_scope_x_y tag_name, standard_scopers, button_scopers, namespace
+ is_in_table_scope = (tag_name, namespace = null) ->
+ return is_in_scope_x tag_name, table_scopers, namespace
+ # aka is_in_list_item_scope
+ is_in_li_scope = (tag_name, namespace = null) ->
+ return is_in_scope_x_y tag_name, standard_scopers, li_scopers, namespace
+ is_in_select_scope = (tag_name, namespace = null) ->
+ for t in open_els
+ if t.name is tag_name and (namespace is null or namespace is t.namespace)
+ return true
+ if t.namespace isnt NS_HTML and t.name isnt 'optgroup' and t.name isnt 'option'
+ return false
+ return false
+ # this checks for a particular element, not by name
+ # this requires a namespace match
+ el_is_in_scope = (needle) ->
+ for el in open_els
+ if el is needle
+ return true
+ if standard_scopers[el.name] is el.namespace
+ return false
+ return false
+
+ clear_to_table_stopers = {
+ 'table': true
+ 'template': true
+ 'html': true
+ }
+ clear_stack_to_table_context = ->
+ loop
+ if clear_to_table_stopers[open_els[0].name]?
+ break
+ open_els.shift()
+ return
+ clear_to_table_body_stopers = {
+ tbody: NS_HTML
+ tfoot: NS_HTML
+ thead: NS_HTML
+ template: NS_HTML
+ html: NS_HTML
+ }
+ clear_stack_to_table_body_context = ->
+ loop
+ if clear_to_table_body_stopers[open_els[0].name] is open_els[0].namespace
+ break
+ open_els.shift()
+ return
+ clear_to_table_row_stopers = {
+ 'tr': true
+ 'template': true
+ 'html': true
+ }
+ clear_stack_to_table_row_context = ->
+ loop
+ if clear_to_table_row_stopers[open_els[0].name]?
+ break
+ open_els.shift()
+ return
+ clear_afe_to_marker = ->
+ loop
+ return unless afe.length > 0 # this happens in fragment case, ?spec error
+ el = afe.shift()
+ if el.type is TYPE_AFE_MARKER
+ return
+ return
+
+ # 8.2.3.1 ...
+ # http://www.w3.org/TR/html5/syntax.html#reset-the-insertion-mode-appropriately
+ reset_ins_mode = ->
+ # 1. Let last be false.
+ last = false
+ # 2. Let node be the last node in the stack of open elements.
+ node_i = 0
+ node = open_els[node_i]
+ # 3. Loop: If node is the first node in the stack of open elements,
+ # then set last to true, and, if the parser was originally created as
+ # part of the HTML fragment parsing algorithm (fragment case) set node
+ # to the context element.
+ loop
+ if node_i is open_els.length - 1
+ last = true
+ if flag_fragment_parsing
+ node = context_element
+ # 4. If node is a select element, run these substeps:
+ if node.name is 'select' and node.namespace is NS_HTML
+ # 1. If last is true, jump to the step below labeled done.
+ unless last
+ # 2. Let ancestor be node.
+ ancestor_i = node_i
+ ancestor = node
+ # 3. Loop: If ancestor is the first node in the stack of
+ # open elements, jump to the step below labeled done.
+ loop
+ if ancestor_i is open_els.length - 1
+ break
+ # 4. Let ancestor be the node before ancestor in the stack
+ # of open elements.
+ ancestor_i += 1
+ ancestor = open_els[ancestor_i]
+ # 5. If ancestor is a template node, jump to the step below
+ # labeled done.
+ if ancestor.name is 'template' and ancestor.namespace is NS_HTML
+ break
+ # 6. If ancestor is a table node, switch the insertion mode
+ # to "in select in table" and abort these steps.
+ if ancestor.name is 'table' and ancestor.namespace is NS_HTML
+ ins_mode = ins_mode_in_select_in_table
+ return
+ # 7. Jump back to the step labeled loop.
+ # 8. Done: Switch the insertion mode to "in select" and abort
+ # these steps.
+ ins_mode = ins_mode_in_select
+ return
+ # 5. If node is a td or th element and last is false, then switch
+ # the insertion mode to "in cell" and abort these steps.
+ if (node.name is 'td' or node.name is 'th') and node.namespace is NS_HTML and last is false
+ ins_mode = ins_mode_in_cell
+ return
+ # 6. If node is a tr element, then switch the insertion mode to "in
+ # row" and abort these steps.
+ if node.name is 'tr' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_row
+ return
+ # 7. If node is a tbody, thead, or tfoot element, then switch the
+ # insertion mode to "in table body" and abort these steps.
+ if (node.name is 'tbody' or node.name is 'thead' or node.name is 'tfoot') and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_table_body
+ return
+ # 8. If node is a caption element, then switch the insertion mode
+ # to "in caption" and abort these steps.
+ if node.name is 'caption' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_caption
+ return
+ # 9. If node is a colgroup element, then switch the insertion mode
+ # to "in column group" and abort these steps.
+ if node.name is 'colgroup' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_column_group
+ return
+ # 10. If node is a table element, then switch the insertion mode to
+ # "in table" and abort these steps.
+ if node.name is 'table' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_table
+ return
+ # 11. If node is a template element, then switch the insertion mode
+ # to the current template insertion mode and abort these steps.
+ if node.name is 'template' and node.namespace is NS_HTML
+ ins_mode = template_ins_modes[0]
+ return
+ # 12. If node is a head element and last is true, then switch the
+ # insertion mode to "in body" ("in body"! not "in head"!) and abort
+ # these steps. (fragment case)
+ if node.name is 'head' and node.namespace is NS_HTML and last
+ ins_mode = ins_mode_in_body
+ return
+ # 13. If node is a head element and last is false, then switch the
+ # insertion mode to "in head" and abort these steps.
+ if node.name is 'head' and node.namespace is NS_HTML and last is false
+ ins_mode = ins_mode_in_head
+ return
+ # 14. If node is a body element, then switch the insertion mode to
+ # "in body" and abort these steps.
+ if node.name is 'body' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_body
+ return
+ # 15. If node is a frameset element, then switch the insertion mode
+ # to "in frameset" and abort these steps. (fragment case)
+ if node.name is 'frameset' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_frameset
+ return
+ # 16. If node is an html element, run these substeps:
+ if node.name is 'html' and node.namespace is NS_HTML
+ # 1. If the head element pointer is null, switch the insertion
+ # mode to "before head" and abort these steps. (fragment case)
+ if head_element_pointer is null
+ ins_mode = ins_mode_before_head
+ else
+ # 2. Otherwise, the head element pointer is not null,
+ # switch the insertion mode to "after head" and abort these
+ # steps.
+ ins_mode = ins_mode_after_head
+ return
+ # 17. If last is true, then switch the insertion mode to "in body"
+ # and abort these steps. (fragment case)
+ if last
+ ins_mode = ins_mode_in_body
+ return
+ # 18. Let node now be the node before node in the stack of open
+ # elements.
+ node_i += 1
+ node = open_els[node_i]
+ # 19. Return to the step labeled loop.
+ return
+
+ # 8.2.3.2
+
+ # http://www.w3.org/TR/html5/syntax.html#adjusted-current-node
+ adjusted_current_node = ->
+ if open_els.length is 1 and flag_fragment_parsing
+ return context_element
+ return open_els[0]
+
+ # http://www.w3.org/TR/html5/syntax.html#reconstruct-the-active-formatting-elements
+ # this implementation is structured (mostly) as described at the link above.
+ # capitalized comments are the "labels" described at the link above.
+ reconstruct_afe = ->
+ return if afe.length is 0
+ if afe[0].type is TYPE_AFE_MARKER or afe[0] in open_els
+ return
+ # Rewind
+ i = 0
+ loop
+ if i is afe.length - 1
+ break
+ i += 1
+ if afe[i].type is TYPE_AFE_MARKER or afe[i] in open_els
+ i -= 1 # Advance
+ break
+ # Create
+ loop
+ el = insert_html_element afe[i].token
+ afe[i] = el
+ break if i is 0
+ i -= 1 # Advance
+ return
+
+ # http://www.w3.org/TR/html5/syntax.html#adoption-agency-algorithm
+ # adoption agency algorithm
+ # overview here:
+ # http://www.w3.org/TR/html5/syntax.html#misnested-tags:-b-i-/b-/i
+ # http://www.w3.org/TR/html5/syntax.html#misnested-tags:-b-p-/b-/p
+ # http://www.w3.org/TR/html5/syntax.html#unclosed-formatting-elements
+ adoption_agency = (subject) ->
+# this block implements tha W3C spec
+# # 1. If the current node is an HTML element whose tag name is subject,
+# # then run these substeps:
+# #
+# # 1. Let element be the current node.
+# #
+# # 2. Pop element off the stack of open elements.
+# #
+# # 3. If element is also in the list of active formatting elements,
+# # remove the element from the list.
+# #
+# # 4. Abort the adoption agency algorithm.
+# if open_els[0].name is subject and open_els[0].namespace is NS_HTML
+# el = open_els.shift()
+# # remove it from the list of active formatting elements (if found)
+# for t, i in afe
+# if t is el
+# afe.splice i, 1
+# break
+# return
+# WHATWG: https://html.spec.whatwg.org/multipage/syntax.html#adoption-agency-algorithm
+ # If the current node is an HTML element whose tag name is subject, and
+ # the current node is not in the list of active formatting elements,
+ # then pop the current node off the stack of open elements, and abort
+ # these steps.
+ if open_els[0].name is subject and open_els[0].namespace is NS_HTML
+ # remove it from the list of active formatting elements (if found)
+ in_afe = false
+ for el, i in afe
+ if el is open_els[0]
+ in_afe = true
+ break
+ unless in_afe
+ open_els.shift()
+ return
+ # fall through
+# END WHATWG
+ outer = 0
+ loop
+ if outer >= 8
+ return
+ outer += 1
+ # 5. Let formatting element be the last element in the list of
+ # active formatting elements that: is between the end of the list
+ # and the last scope marker in the list, if any, or the start of
+ # the list otherwise, and has the tag name subject.
+ fe = null
+ for t, fe_of_afe in afe
+ if t.type is TYPE_AFE_MARKER
+ break
+ if t.name is subject
+ fe = t
+ break
+ # If there is no such element, then abort these steps and instead
+ # act as described in the "any other end tag" entry above.
+ if fe is null
+ in_body_any_other_end_tag subject
+ return
+ # 6. If formatting element is not in the stack of open elements,
+ # then this is a parse error; remove the element from the list, and
+ # abort these steps.
+ in_open_els = false
+ for t, fe_of_open_els in open_els
+ if t is fe
+ in_open_els = true
+ break
+ unless in_open_els
+ parse_error()
+ # "remove it from the list" must mean afe, since it's not in open_els
+ afe.splice fe_of_afe, 1
+ return
+ # 7. If formatting element is in the stack of open elements, but
+ # the element is not in scope, then this is a parse error; abort
+ # these steps.
+ unless el_is_in_scope fe
+ parse_error()
+ return
+ # 8. If formatting element is not the current node, this is a parse
+ # error. (But do not abort these steps.)
+ unless open_els[0] is fe
+ parse_error()
+ # continue
+ # 9. Let furthest block be the topmost node in the stack of open
+ # elements that is lower in the stack than formatting element, and
+ # is an element in the special category. There might not be one.
+ fb = null
+ fb_of_open_els = null
+ for t, i in open_els
+ if t is fe
+ break
+ if el_is_special t
+ fb = t
+ fb_of_open_els = i
+ # and continue, to see if there's one that's more "topmost"
+ # 10. If there is no furthest block, then the UA must first pop all
+ # the nodes from the bottom of the stack of open elements, from the
+ # current node up to and including formatting element, then remove
+ # formatting element from the list of active formatting elements,
+ # and finally abort these steps.
+ if fb is null
+ loop
+ t = open_els.shift()
+ if t is fe
+ afe.splice fe_of_afe, 1
+ return
+ # 11. Let common ancestor be the element immediately above
+ # formatting element in the stack of open elements.
+ ca = open_els[fe_of_open_els + 1] # common ancestor
+
+ node_above = open_els[fb_of_open_els + 1] # next node if node isn't in open_els anymore
+ # 12. Let a bookmark note the position of formatting element in the list of active formatting elements relative to the elements on either side of it in the list.
+ bookmark = new_aaa_bookmark()
+ for t, i in afe
+ if t is fe
+ afe.splice i, 0, bookmark
+ break
+ node = last_node = fb
+ inner = 0
+ loop
+ inner += 1
+ # 3. Let node be the element immediately above node in the
+ # stack of open elements, or if node is no longer in the stack
+ # of open elements (e.g. because it got removed by this
+ # algorithm), the element that was immediately above node in
+ # the stack of open elements before node was removed.
+ node_next = null
+ for t, i in open_els
+ if t is node
+ node_next = open_els[i + 1]
+ break
+ node = node_next ? node_above
+ # TODO make sure node_above gets re-set if/when node is removed from open_els
+
+ # 4. If node is formatting element, then go to the next step in
+ # the overall algorithm.
+ if node is fe
+ break
+ # 5. If inner loop counter is greater than three and node is in
+ # the list of active formatting elements, then remove node from
+ # the list of active formatting elements.
+ node_in_afe = false
+ for t, i in afe
+ if t is node
+ if inner > 3
+ afe.splice i, 1
+ else
+ node_in_afe = true
+ break
+ # 6. If node is not in the list of active formatting elements,
+ # then remove node from the stack of open elements and then go
+ # back to the step labeled inner loop.
+ unless node_in_afe
+ for t, i in open_els
+ if t is node
+ node_above = open_els[i + 1]
+ open_els.splice i, 1
+ break
+ continue
+ # 7. create an element for the token for which the element node
+ # was created, in the HTML namespace, with common ancestor as
+ # the intended parent; replace the entry for node in the list
+ # of active formatting elements with an entry for the new
+ # element, replace the entry for node in the stack of open
+ # elements with an entry for the new element, and let node be
+ # the new element.
+ new_node = token_to_element node.token, NS_HTML, ca
+ for t, i in afe
+ if t is node
+ afe[i] = new_node
+ break
+ for t, i in open_els
+ if t is node
+ node_above = open_els[i + 1]
+ open_els[i] = new_node
+ break
+ node = new_node
+ # 8. If last node is furthest block, then move the
+ # aforementioned bookmark to be immediately after the new node
+ # in the list of active formatting elements.
+ if last_node is fb
+ for t, i in afe
+ if t is bookmark
+ afe.splice i, 1
+ break
+ for t, i in afe
+ if t is node
+ # "after" means lower
+ afe.splice i, 0, bookmark # "after as <-
+ break
+ # 9. Insert last node into node, first removing it from its
+ # previous parent node if any.
+ if last_node.parent?
+ for c, i in last_node.parent.children
+ if c is last_node
+ last_node.parent.children.splice i, 1
+ break
+ node.children.push last_node
+ last_node.parent = node
+ # 10. Let last node be node.
+ last_node = node
+ # 11. Return to the step labeled inner loop.
+ # 14. Insert whatever last node ended up being in the previous step
+ # at the appropriate place for inserting a node, but using common
+ # ancestor as the override target.
+
+ # In the case where fe is immediately followed by fb:
+ # * inner loop exits out early (node==fe)
+ # * last_node is fb
+ # * last_node is still in the tree (not a duplicate)
+ if last_node.parent?
+ for c, i in last_node.parent.children
+ if c is last_node
+ last_node.parent.children.splice i, 1
+ break
+ # can't use standard insert token thing, because it's already in
+ # open_els and must stay at it's current position in open_els
+ dest = adjusted_insertion_location ca
+ dest[0].children.splice dest[1], 0, last_node
+ last_node.parent = dest[0]
+ # 15. Create an element for the token for which formatting element
+ # was created, in the HTML namespace, with furthest block as the
+ # intended parent.
+ new_element = token_to_element fe.token, NS_HTML, fb
+ # 16. Take all of the child nodes of furthest block and append them
+ # to the element created in the last step.
+ while fb.children.length
+ t = fb.children.shift()
+ t.parent = new_element
+ new_element.children.push t
+ # 17. Append that new element to furthest block.
+ new_element.parent = fb
+ fb.children.push new_element
+ # 18. Remove formatting element from the list of active formatting
+ # elements, and insert the new element into the list of active
+ # formatting elements at the position of the aforementioned
+ # bookmark.
+ for t, i in afe
+ if t is fe
+ afe.splice i, 1
+ break
+ for t, i in afe
+ if t is bookmark
+ afe[i] = new_element
+ break
+ # 19. Remove formatting element from the stack of open elements,
+ # and insert the new element into the stack of open elements
+ # immediately below the position of furthest block in that stack.
+ for t, i in open_els
+ if t is fe
+ open_els.splice i, 1
+ break
+ for t, i in open_els
+ if t is fb
+ open_els.splice i, 0, new_element
+ break
+ # 20. Jump back to the step labeled outer loop.
+ return
+
+ # http://www.w3.org/TR/html5/syntax.html#close-a-p-element
+ close_p_element = ->
+ generate_implied_end_tags 'p' # arg is exception
+ unless open_els[0].name is 'p' and open_els[0].namespace is NS_HTML
+ parse_error()
+ while open_els.length > 1 # just in case
+ el = open_els.shift()
+ if el.name is 'p' and el.namespace is NS_HTML
+ return
+ return
+ close_p_if_in_button_scope = ->
+ if is_in_button_scope 'p', NS_HTML
+ close_p_element()
+ return
+
+ # http://www.w3.org/TR/html5/syntax.html#insert-a-character
+ # aka insert_a_character = (t) ->
+ insert_character = (t) ->
+ dest = adjusted_insertion_location()
+ # fixfull check for Document node
+ if dest[1] > 0
+ prev = dest[0].children[dest[1] - 1]
+ if prev.type is TYPE_TEXT
+ prev.text += t.text
+ return
+ dest[0].children.splice dest[1], 0, t
+ return
+
+ # 8.2.5 http://www.w3.org/TR/html5/syntax.html#tree-construction
+ process_token = (t) ->
+ acn = adjusted_current_node()
+ unless acn?
+ ins_mode t
+ return
+ if acn.namespace is NS_HTML
+ ins_mode t
+ return
+ if is_mathml_text_integration_point(acn)
+ if t.type is TYPE_START_TAG and not (t.name is 'mglyph' or t.name is 'malignmark')
+ ins_mode t
+ return
+ if t.type is TYPE_TEXT
+ ins_mode t
+ return
+ if acn.namespace is NS_MATHML and acn.name is 'annotation-xml' and t.type is TYPE_START_TAG and t.name is 'svg'
+ ins_mode t
+ return
+ if is_html_integration acn
+ if t.type is TYPE_START_TAG or t.type is TYPE_TEXT
+ ins_mode t
+ return
+ if t.type is TYPE_EOF
+ ins_mode t
+ return
+ in_foreign_content t
+ return
+
+ # 8.2.5.1
+ # http://www.w3.org/TR/html5/syntax.html#creating-and-inserting-nodes
+ # http://www.w3.org/TR/html5/syntax.html#appropriate-place-for-inserting-a-node
+ adjusted_insertion_location = (override_target = null) ->
+ # 1. If there was an override target specified, then let target be the
+ # override target.
+ if override_target?
+ target = override_target
+ else # Otherwise, let target be the current node.
+ target = open_els[0]
+ # 2. Determine the adjusted insertion location using the first matching
+ # steps from the following list:
+ #
+ # If foster parenting is enabled and target is a table, tbody, tfoot,
+ # thead, or tr element Foster parenting happens when content is
+ # misnested in tables.
+ if flag_foster_parenting and foster_parenting_targets[target.name] is target.namespace
+ loop # once. this is here so we can ``break`` to "abort these substeps"
+ # 1. Let last template be the last template element in the
+ # stack of open elements, if any.
+ last_template = null
+ last_template_i = null
+ for el, i in open_els
+ if el.name is 'template' and el.namespace is NS_HTML
+ last_template = el
+ last_template_i = i
+ break
+ # 2. Let last table be the last table element in the stack of
+ # open elements, if any.
+ last_table = null
+ last_table_i
+ for el, i in open_els
+ if el.name is 'table' and el.namespace is NS_HTML
+ last_table = el
+ last_table_i = i
+ break
+ # 3. If there is a last template and either there is no last
+ # table, or there is one, but last template is lower (more
+ # recently added) than last table in the stack of open
+ # elements, then: let adjusted insertion location be inside
+ # last template's template contents, after its last child (if
+ # any), and abort these substeps.
+ if last_template and (last_table is null or last_template_i < last_table_i)
+ target = last_template # fixfull should be it's contents
+ target_i = target.children.length
+ break
+ # 4. If there is no last table, then let adjusted insertion
+ # location be inside the first element in the stack of open
+ # elements (the html element), after its last child (if any),
+ # and abort these substeps. (fragment case)
+ if last_table is null
+ # this is odd
+ target = open_els[open_els.length - 1]
+ target_i = target.children.length
+ break
+ # 5. If last table has a parent element, then let adjusted
+ # insertion location be inside last table's parent element,
+ # immediately before last table, and abort these substeps.
+ if last_table.parent?
+ for c, i in last_table.parent.children
+ if c is last_table
+ target = last_table.parent
+ target_i = i
+ break
+ break
+ # 6. Let previous element be the element immediately above last
+ # table in the stack of open elements.
+ #
+ # huh? how could it not have a parent?
+ previous_element = open_els[last_table_i + 1]
+ # 7. Let adjusted insertion location be inside previous
+ # element, after its last child (if any).
+ target = previous_element
+ target_i = target.children.length
+ # Note: These steps are involved in part because it's possible
+ # for elements, the table element in this case in particular,
+ # to have been moved by a script around in the DOM, or indeed
+ # removed from the DOM entirely, after the element was inserted
+ # by the parser.
+ break # don't really loop
+ else
+ # Otherwise Let adjusted insertion location be inside target, after
+ # its last child (if any).
+ target_i = target.children.length
+
+ # 3. If the adjusted insertion location is inside a template element,
+ # let it instead be inside the template element's template contents,
+ # after its last child (if any).
+ # fixfull (template)
+
+ # 4. Return the adjusted insertion location.
+ return [target, target_i]
+
+ # http://www.w3.org/TR/html5/syntax.html#create-an-element-for-the-token
+ # aka create_an_element_for_token
+ token_to_element = (t, namespace, intended_parent) ->
+ # convert attributes into a hash
+ attrs = {}
+ for a in t.attrs_a
+ attrs[a[0]] = a[1] # TODO check what to do with dupilcate attrs
+ el = new Node TYPE_TAG, name: t.name, namespace: namespace, attrs: attrs, token: t
+
+ # TODO 2. If the newly created element has an xmlns attribute in the
+ # XMLNS namespace whose value is not exactly the same as the element's
+ # namespace, that is a parse error. Similarly, if the newly created
+ # element has an xmlns:xlink attribute in the XMLNS namespace whose
+ # value is not the XLink Namespace, that is a parse error.
+
+ # fixfull: the spec says stuff about form pointers and ownerDocument
+
+ return el
+
+ # http://www.w3.org/TR/html5/syntax.html#insert-a-foreign-element
+ insert_foreign_element = (token, namespace) ->
+ ail = adjusted_insertion_location()
+ ail_el = ail[0]
+ ail_i = ail[1]
+ el = token_to_element token, namespace, ail_el
+ # TODO skip this next step if it's broken (eg ail_el is document with child already)
+ el.parent = ail_el
+ ail_el.children.splice ail_i, 0, el
+ open_els.unshift el
+ return el
+ # http://www.w3.org/TR/html5/syntax.html#insert-an-html-element
+ insert_html_element = (token) ->
+ return insert_foreign_element token, NS_HTML
+
+ # http://www.w3.org/TR/html5/syntax.html#insert-a-comment
+ # position should be [node, index_within_children]
+ insert_comment = (t, position = null) ->
+ position ?= adjusted_insertion_location()
+ position[0].children.splice position[1], 0, t
+ return
+
+ # 8.2.5.2
+ # http://www.w3.org/TR/html5/syntax.html#generic-raw-text-element-parsing-algorithm
+ parse_generic_raw_text = (t) ->
+ insert_html_element t
+ tok_state = tok_state_rawtext
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_text
+ return
+ parse_generic_rcdata_text = (t) ->
+ insert_html_element t
+ tok_state = tok_state_rcdata
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_text
+ return
+
+ # 8.2.5.3 http://www.w3.org/TR/html5/syntax.html#closing-elements-that-have-implied-end-tags
+ # http://www.w3.org/TR/html5/syntax.html#generate-implied-end-tags
+ generate_implied_end_tags = (except = null) ->
+ while end_tag_implied[open_els[0].name] is open_els[0].namespace and open_els[0].name isnt except
+ open_els.shift()
+ return
+
+ # 8.2.5.4 The rules for parsing tokens in HTML content
+ # http://www.w3.org/TR/html5/syntax.html#parsing-main-inhtml
+
+ # 8.2.5.4.1 The "initial" insertion mode
+ # http://www.w3.org/TR/html5/syntax.html#the-initial-insertion-mode
+ is_quirks_yes_doctype = (t) ->
+ if t.flag 'force-quirks'
+ return true
+ if t.name isnt 'html'
+ return true
+ if t.public_identifier?
+ pi = t.public_identifier.toLowerCase()
+ for p in quirks_yes_pi_prefixes
+ if pi.substr(0, p.length) is p
+ return true
+ if pi is '-//w3o//dtd w3 html strict 3.0//en//' or pi is '-/w3c/dtd html 4.0 transitional/en' or pi is 'html'
+ return true
+ if t.system_identifier?
+ if t.system_identifier.toLowerCase() is 'http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd'
+ return true
+ else if t.public_identifier?
+ # already did this: pi = t.public_identifier.toLowerCase()
+ if pi.substr(0, 32) is '-//w3c//dtd html 4.01 frameset//' or pi.substr(0, 36) is '-//w3c//dtd html 4.01 transitional//'
+ return true
+ return false
+ is_quirks_limited_doctype = (t) ->
+ if t.public_identifier?
+ pi = t.public_identifier.toLowerCase()
+ if pi.substr(0, 32) is '-//w3c//dtd xhtml 1.0 frameset//' or pi.substr(0, 36) is '-//w3c//dtd xhtml 1.0 transitional//'
+ return true
+ if t.system_identifier?
+ if pi.substr(0, 32) is '-//w3c//dtd html 4.01 frameset//' or pi.substr(0, 36) is '-//w3c//dtd html 4.01 transitional//'
+ return true
+ return false
+ ins_mode_initial = (t) ->
+ if is_space_tok t
+ return
+ if t.type is TYPE_COMMENT
+ # ?fixfull
+ doc.children.push t
+ return
+ if t.type is TYPE_DOCTYPE
+ # fixfull syntax error from first paragraph and following bullets
+ # fixfull set doc.doctype
+ # fixfull is the "not an iframe srcdoc" thing relevant?
+ if is_quirks_yes_doctype t
+ doc.flag 'quirks mode', QUIRKS_YES
+ else if is_quirks_limited_doctype t
+ doc.flag 'quirks mode', QUIRKS_LIMITED
+ doc.children.push t
+ ins_mode = ins_mode_before_html
+ return
+ # Anything else
+ # fixfull not iframe srcdoc?
+ parse_error()
+ doc.flag 'quirks mode', QUIRKS_YES
+ ins_mode = ins_mode_before_html
+ process_token t
+ return
+
+ # 8.2.5.4.2 http://www.w3.org/TR/html5/syntax.html#the-before-html-insertion-mode
+ ins_mode_before_html = (t) ->
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_COMMENT
+ doc.children.push t
+ return
+ if is_space_tok t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'html'
+ el = token_to_element t, NS_HTML, doc
+ doc.children.push el
+ el.document = doc
+ open_els.unshift(el)
+ # fixfull (big paragraph in spec about manifest, fragment, urls, etc)
+ ins_mode = ins_mode_before_head
+ return
+ if t.type is TYPE_END_TAG
+ if t.name is 'head' or t.name is 'body' or t.name is 'html' or t.name is 'br'
+ # fall through to "anything else"
+ else
+ parse_error()
+ return
+ # Anything else
+ el = token_to_element new_open_tag('html'), NS_HTML, doc
+ doc.children.push el
+ el.document = doc
+ open_els.unshift el
+ # ?fixfull browsing context
+ ins_mode = ins_mode_before_head
+ process_token t
+ return
+
+ # 8.2.5.4.3 http://www.w3.org/TR/html5/syntax.html#the-before-head-insertion-mode
+ ins_mode_before_head = (t) ->
+ if is_space_tok t
+ return
+ if t.type is TYPE_COMMENT
+ insert_comment t
+ return
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'html'
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'head'
+ el = insert_html_element t
+ head_element_pointer = el
+ ins_mode = ins_mode_in_head
+ return
+ if t.type is TYPE_END_TAG
+ if t.name is 'head' or t.name is 'body' or t.name is 'html' or t.name is 'br'
+ # fall through to Anything else below
+ else
+ parse_error()
+ return
+ # Anything else
+ el = insert_html_element new_open_tag 'head'
+ head_element_pointer = el
+ ins_mode = ins_mode_in_head
+ process_token t
+ return
+
+ # 8.2.5.4.4 http://www.w3.org/TR/html5/syntax.html#parsing-main-inhead
+ ins_mode_in_head_else = (t) -> # factored out for same-as-spec flow control
+ open_els.shift() # spec says this will be a 'head' node
+ ins_mode = ins_mode_after_head
+ process_token t
+ return
+ ins_mode_in_head = (t) ->
+ if t.type is TYPE_TEXT and (t.text is "\t" or t.text is "\n" or t.text is "\u000c" or t.text is ' ')
+ insert_character t
+ return
+ if t.type is TYPE_COMMENT
+ insert_comment t
+ return
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'html'
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link')
+ el = insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'meta'
+ el = insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ # fixfull encoding stuff
+ return
+ if t.type is TYPE_START_TAG and t.name is 'title'
+ parse_generic_rcdata_text t
+ return
+ if t.type is TYPE_START_TAG and ((t.name is 'noscript' and flag_scripting) or t.name is 'noframes' or t.name is 'style')
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'noscript' and flag_scripting is false
+ insert_html_element t
+ ins_mode = ins_mode_in_head_noscript
+ return
+ if t.type is TYPE_START_TAG and t.name is 'script'
+ ail = adjusted_insertion_location()
+ el = token_to_element t, NS_HTML, ail
+ el.flag 'parser-inserted', true
+ # fixfull frament case
+ ail[0].children.splice ail[1], 0, el
+ open_els.unshift el
+ tok_state = tok_state_script_data
+ original_ins_mode = ins_mode # make sure orig... is defined
+ ins_mode = ins_mode_text
+ return
+ if t.type is TYPE_END_TAG and t.name is 'head'
+ open_els.shift() # will be a head element... spec says so
+ ins_mode = ins_mode_after_head
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'html' or t.name is 'br')
+ ins_mode_in_head_else t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'template'
+ insert_html_element t
+ afe_push_marker()
+ flag_frameset_ok = false
+ ins_mode = ins_mode_in_template
+ template_ins_modes.unshift ins_mode_in_template
+ return
+ if t.type is TYPE_END_TAG and t.name is 'template'
+ if template_tag_is_open()
+ generate_implied_end_tags
+ if open_els[0].name isnt 'template'
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'template' and el.namespace is NS_HTML
+ break
+ clear_afe_to_marker()
+ template_ins_modes.shift()
+ reset_ins_mode()
+ else
+ parse_error()
+ return
+ if (t.type is TYPE_START_TAG and t.name is 'head') or t.type is TYPE_END_TAG
+ parse_error()
+ return
+ ins_mode_in_head_else t
+ return
+
+ # 8.2.5.4.5 http://www.w3.org/TR/html5/syntax.html#parsing-main-inheadnoscript
+ ins_mode_in_head_noscript_else = (t) ->
+ parse_error()
+ open_els.shift()
+ ins_mode = ins_mode_in_head
+ process_token t
+ return
+ ins_mode_in_head_noscript = (t) ->
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'html'
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'noscript'
+ open_els.shift()
+ ins_mode = ins_mode_in_head
+ return
+ if is_space_tok(t) or t.type is TYPE_COMMENT or (t.type is TYPE_START_TAG and (t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'style'))
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'br'
+ ins_mode_in_head_noscript_else t
+ return
+ if (t.type is TYPE_START_TAG and (t.name is 'head' or t.name is 'noscript')) or t.type is TYPE_END_TAG
+ parse_error()
+ return
+ # Anything else
+ ins_mode_in_head_noscript_else t
+ return
+
+ # 8.2.5.4.6 http://www.w3.org/TR/html5/syntax.html#the-after-head-insertion-mode
+ ins_mode_after_head_else = (t) ->
+ body_tok = new_open_tag 'body'
+ insert_html_element body_tok
+ ins_mode = ins_mode_in_body
+ process_token t
+ return
+ ins_mode_after_head = (t) ->
+ if is_space_tok t
+ insert_character t
+ return
+ if t.type is TYPE_COMMENT
+ insert_comment t
+ return
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'html'
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'body'
+ insert_html_element t
+ flag_frameset_ok = false
+ ins_mode = ins_mode_in_body
+ return
+ if t.type is TYPE_START_TAG and t.name is 'frameset'
+ insert_html_element t
+ ins_mode = ins_mode_in_frameset
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'script' or t.name is 'style' or t.name is 'template' or t.name is 'title')
+ parse_error()
+ open_els.unshift head_element_pointer
+ ins_mode_in_head t
+ for el, i in open_els
+ if el is head_element_pointer
+ open_els.splice i, 1
+ return
+ return
+ if t.type is TYPE_END_TAG and t.name is 'template'
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'html' or t.name is 'br')
+ ins_mode_after_head_else t
+ return
+ if (t.type is TYPE_START_TAG and t.name is 'head') or t.type is TYPE_END_TAG
+ parse_error()
+ return
+ # Anything else
+ ins_mode_after_head_else t
+ return
+
+ # 8.2.5.4.7 http://www.w3.org/TR/html5/syntax.html#parsing-main-inbody
+ in_body_any_other_end_tag = (name) -> # factored out because adoption agency calls it
+ node = open_els[0]
+ loop
+ if node.name is name and node.namespace is NS_HTML
+ generate_implied_end_tags name # arg is exception
+ unless node is open_els[0]
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el is node
+ return
+ if special_elements[node.name] is node.namespace
+ parse_error()
+ return
+ for el, i in open_els
+ if node is el
+ node = open_els[i + 1]
+ break
+ return
+ ins_mode_in_body = (t) ->
+ if t.type is TYPE_TEXT and t.text is "\u0000"
+ parse_error()
+ return
+ if is_space_tok t
+ reconstruct_afe()
+ insert_character t
+ return
+ if t.type is TYPE_TEXT
+ reconstruct_afe()
+ insert_character t
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_COMMENT
+ insert_comment t
+ return
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'html'
+ parse_error()
+ return if template_tag_is_open()
+ root_attrs = open_els[open_els.length - 1].attrs
+ for a in t.attrs_a
+ root_attrs[a[0]] = a[1] unless root_attrs[a[0]]?
+ return
+
+ if (t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'script' or t.name is 'style' or t.name is 'template' or t.name is 'title')) or (t.type is TYPE_END_TAG and t.name is 'template')
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'body'
+ parse_error()
+ return if open_els.length < 2
+ second = open_els[open_els.length - 2]
+ return unless second.namespace is NS_HTML
+ return unless second.name is 'body'
+ return if template_tag_is_open()
+ flag_frameset_ok = false
+ for a in t.attrs_a
+ second.attrs[a[0]] = a[1] unless second.attrs[a[0]]?
+ return
+ if t.type is TYPE_START_TAG and t.name is 'frameset'
+ parse_error()
+ return if open_els.length < 2
+ second_i = open_els.length - 2
+ second = open_els[second_i]
+ return unless second.namespace is NS_HTML
+ return unless second.name is 'body'
+ if flag_frameset_ok is false
+ return
+ if second.parent?
+ for el, i in second.parent.children
+ if el is second
+ second.parent.children.splice i, 1
+ break
+ open_els.splice second_i, 1
+ # pop everything except the "root html element"
+ while open_els.length > 1
+ open_els.shift()
+ insert_html_element t
+ ins_mode = ins_mode_in_frameset
+ return
+ if t.type is TYPE_EOF
+ ok_tags = {
+ dd:NS_HTML, dt:NS_HTML, li:NS_HTML, p:NS_HTML, tbody:NS_HTML,
+ td:NS_HTML, tfoot:NS_HTML, th:NS_HTML, thead:NS_HTML,
+ tr:NS_HTML, body:NS_HTML, html:NS_HTML,
+ }
+ for el in open_els
+ unless ok_tags[t.name] is el.namespace
+ parse_error()
+ break
+ if template_ins_modes.length > 0
+ ins_mode_in_template t
+ else
+ stop_parsing()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'body'
+ unless is_in_scope 'body', NS_HTML
+ parse_error()
+ return
+ ok_tags = {
+ dd:NS_HTML, dt:NS_HTML, li:NS_HTML, optgroup:NS_HTML,
+ option:NS_HTML, p:NS_HTML, rb:NS_HTML, rp:NS_HTML, rt:NS_HTML,
+ rtc:NS_HTML, tbody:NS_HTML, td:NS_HTML, tfoot:NS_HTML,
+ th:NS_HTML, thead:NS_HTML, tr:NS_HTML, body:NS_HTML,
+ html:NS_HTML
+ }
+ for el in open_els
+ unless ok_tags[t.name] is el.namespace
+ parse_error()
+ break
+ ins_mode = ins_mode_after_body
+ return
+ if t.type is TYPE_END_TAG and t.name is 'html'
+ unless is_in_scope 'body', NS_HTML
+ parse_error()
+ return
+ ok_tags = {
+ dd:NS_HTML, dt:NS_HTML, li:NS_HTML, optgroup:NS_HTML,
+ option:NS_HTML, p:NS_HTML, rb:NS_HTML, rp:NS_HTML, rt:NS_HTML,
+ rtc:NS_HTML, tbody:NS_HTML, td:NS_HTML, tfoot:NS_HTML,
+ th:NS_HTML, thead:NS_HTML, tr:NS_HTML, body:NS_HTML,
+ html:NS_HTML
+ }
+ for el in open_els
+ unless ok_tags[t.name] is el.namespace
+ parse_error()
+ break
+ ins_mode = ins_mode_after_body
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'address' or t.name is 'article' or t.name is 'aside' or t.name is 'blockquote' or t.name is 'center' or t.name is 'details' or t.name is 'dialog' or t.name is 'dir' or t.name is 'div' or t.name is 'dl' or t.name is 'fieldset' or t.name is 'figcaption' or t.name is 'figure' or t.name is 'footer' or t.name is 'header' or t.name is 'hgroup' or t.name is 'main' or t.name is 'nav' or t.name is 'ol' or t.name is 'p' or t.name is 'section' or t.name is 'summary' or t.name is 'ul')
+ close_p_if_in_button_scope()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and h_tags[t.name]?
+ close_p_if_in_button_scope()
+ if h_tags[open_els[0].name] is open_els[0].namespace
+ parse_error()
+ open_els.shift()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'pre' or t.name is 'listing')
+ close_p_if_in_button_scope()
+ insert_html_element t
+ eat_next_token_if_newline()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and t.name is 'form'
+ unless form_element_pointer is null or template_tag_is_open()
+ parse_error()
+ return
+ close_p_if_in_button_scope()
+ el = insert_html_element t
+ unless template_tag_is_open()
+ form_element_pointer = el
+ return
+ if t.type is TYPE_START_TAG and t.name is 'li'
+ flag_frameset_ok = false
+ for node in open_els
+ if node.name is 'li' and node.namespace is NS_HTML
+ generate_implied_end_tags 'li' # arg is exception
+ if open_els[0].name isnt 'li' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'li' and el.namespace is NS_HTML
+ break
+ break
+ if el_is_special_not_adp node
+ break
+ close_p_if_in_button_scope()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'dd' or t.name is 'dt')
+ flag_frameset_ok = false
+ for node in open_els
+ if node.name is 'dd' and node.namespace is NS_HTML
+ generate_implied_end_tags 'dd' # arg is exception
+ if open_els[0].name isnt 'dd' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'dd' and el.namespace is NS_HTML
+ break
+ break
+ if node.name is 'dt' and node.namespace is NS_HTML
+ generate_implied_end_tags 'dt' # arg is exception
+ if open_els[0].name isnt 'dt' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'dt' and el.namespace is NS_HTML
+ break
+ break
+ if el_is_special_not_adp node
+ break
+ close_p_if_in_button_scope()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'plaintext'
+ close_p_if_in_button_scope()
+ insert_html_element t
+ tok_state = tok_state_plaintext
+ return
+ if t.type is TYPE_START_TAG and t.name is 'button'
+ if is_in_scope 'button', NS_HTML
+ parse_error()
+ generate_implied_end_tags()
+ loop
+ el = open_els.shift()
+ if el.name is 'button' and el.namespace is NS_HTML
+ break
+ reconstruct_afe()
+ insert_html_element t
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'address' or t.name is 'article' or t.name is 'aside' or t.name is 'blockquote' or t.name is 'button' or t.name is 'center' or t.name is 'details' or t.name is 'dialog' or t.name is 'dir' or t.name is 'div' or t.name is 'dl' or t.name is 'fieldset' or t.name is 'figcaption' or t.name is 'figure' or t.name is 'footer' or t.name is 'header' or t.name is 'hgroup' or t.name is 'listing' or t.name is 'main' or t.name is 'nav' or t.name is 'ol' or t.name is 'pre' or t.name is 'section' or t.name is 'summary' or t.name is 'ul')
+ unless is_in_scope t.name, NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags()
+ unless open_els[0].name is t.name and open_els[0].namespace is NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is t.name and el.namespace is NS_HTML
+ return
+ return
+ if t.type is TYPE_END_TAG and t.name is 'form'
+ unless template_tag_is_open()
+ node = form_element_pointer
+ form_element_pointer = null
+ if node is null or not el_is_in_scope node
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0] isnt node
+ parse_error()
+ for el, i in open_els
+ if el is node
+ open_els.splice i, 1
+ break
+ else
+ unless is_in_scope 'form', NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0].name isnt 'form' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'form' and el.namespace is NS_HTML
+ break
+ return
+ if t.type is TYPE_END_TAG and t.name is 'p'
+ unless is_in_button_scope 'p', NS_HTML
+ parse_error()
+ insert_html_element new_open_tag 'p'
+ close_p_element()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'li'
+ unless is_in_li_scope 'li', NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags 'li' # arg is exception
+ if open_els[0].name isnt 'li' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'li' and el.namespace is NS_HTML
+ break
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'dd' or t.name is 'dt')
+ unless is_in_scope t.name, NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags t.name # arg is exception
+ if open_els[0].name isnt t.name or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is t.name and el.namespace is NS_HTML
+ break
+ return
+ if t.type is TYPE_END_TAG and h_tags[t.name]?
+ h_in_scope = false
+ for el in open_els
+ if h_tags[el.name] is el.namespace
+ h_in_scope = true
+ break
+ if standard_scopers[el.name] is el.namespace
+ break
+ unless h_in_scope
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0].name isnt t.name or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if h_tags[el.name] is el.namespace
+ break
+ return
+ # deep breath!
+ if t.type is TYPE_START_TAG and t.name is 'a'
+ # If the list of active formatting elements contains an a element
+ # between the end of the list and the last marker on the list (or
+ # the start of the list if there is no marker on the list), then
+ # this is a parse error; run the adoption agency algorithm for the
+ # tag name "a", then remove that element from the list of active
+ # formatting elements and the stack of open elements if the
+ # adoption agency algorithm didn't already remove it (it might not
+ # have if the element is not in table scope).
+ found = false
+ for el in afe
+ if el.type is TYPE_AFE_MARKER
+ break
+ if el.name is 'a' and el.namespace is NS_HTML
+ found = el
+ if found?
+ parse_error()
+ adoption_agency 'a'
+ for el, i in afe
+ if el is found
+ afe.splice i, 1
+ for el, i in open_els
+ if el is found
+ open_els.splice i, 1
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'b' or t.name is 'big' or t.name is 'code' or t.name is 'em' or t.name is 'font' or t.name is 'i' or t.name is 's' or t.name is 'small' or t.name is 'strike' or t.name is 'strong' or t.name is 'tt' or t.name is 'u')
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
+ return
+ if t.type is TYPE_START_TAG and t.name is 'nobr'
+ reconstruct_afe()
+ if is_in_scope 'nobr', NS_HTML
+ parse_error()
+ adoption_agency 'nobr'
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'a' or t.name is 'b' or t.name is 'big' or t.name is 'code' or t.name is 'em' or t.name is 'font' or t.name is 'i' or t.name is 'nobr' or t.name is 's' or t.name is 'small' or t.name is 'strike' or t.name is 'strong' or t.name is 'tt' or t.name is 'u')
+ adoption_agency t.name
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'applet' or t.name is 'marquee' or t.name is 'object')
+ reconstruct_afe()
+ insert_html_element t
+ afe_push_marker()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'applet' or t.name is 'marquee' or t.name is 'object')
+ unless is_in_scope t.name, NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0].name isnt t.name or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is t.name and el.namespace is NS_HTML
+ break
+ clear_afe_to_marker()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'table'
+ unless doc.flag('quirks mode') is QUIRKS_YES
+ close_p_if_in_button_scope() # test
+ insert_html_element t
+ flag_frameset_ok = false
+ ins_mode = ins_mode_in_table
+ return
+ if t.type is TYPE_END_TAG and t.name is 'br'
+ parse_error()
+ # W3C: t.type = TYPE_START_TAG
+ t = new_open_tag 'br' # WHATWG
+ # fall through
+ if t.type is TYPE_START_TAG and (t.name is 'area' or t.name is 'br' or t.name is 'embed' or t.name is 'img' or t.name is 'keygen' or t.name is 'wbr')
+ reconstruct_afe()
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and t.name is 'input'
+ reconstruct_afe()
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ unless is_input_hidden_tok t
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'menuitem' or t.name is 'param' or t.name is 'source' or t.name is 'track')
+ # WHATWG adds 'menuitem' for this block
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'hr'
+ close_p_if_in_button_scope()
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and t.name is 'image'
+ parse_error()
+ t.name = 'img'
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'isindex'
+ parse_error()
+ if template_tag_is_open() is false and form_element_pointer isnt null
+ return
+ t.acknowledge_self_closing()
+ flag_frameset_ok = false
+ close_p_if_in_button_scope()
+ el = insert_html_element new_open_tag 'form'
+ unless template_tag_is_open()
+ form_element_pointer = el
+ for a in t.attrs_a
+ if a[0] is 'action'
+ el.attrs['action'] = a[1]
+ break
+ insert_html_element new_open_tag 'hr'
+ open_els.shift()
+ reconstruct_afe()
+ insert_html_element new_open_tag 'label'
+ # note: this is a little out-of-spec-order so we only have to scan t.attrs_a once
+ input_el = new_open_tag 'input'
+ prompt = null
+ for a in t.attrs_a
+ if a[0] is 'prompt'
+ prompt = a[1]
+ if a[0] isnt 'name' and a[0] isnt 'action' and a[0] isnt 'prompt'
+ input_el.attrs_a.push [a[0], a[1]]
+ input_el.attrs_a.push ['name', 'isindex']
+ # fixfull this next bit is in english... internationalize?
+ prompt ?= "This is a searchable index. Enter search keywords: "
+ insert_character new_character_token prompt # fixfull split
+ # TODO submit typo "balue" in spec
+ insert_html_element input_el
+ open_els.shift()
+ # insert_character '' # you can put chars here if promt attr missing
+ open_els.shift()
+ insert_html_element new_open_tag 'hr'
+ open_els.shift()
+ open_els.shift()
+ unless template_tag_is_open()
+ form_element_pointer = null
+ return
+ if t.type is TYPE_START_TAG and t.name is 'textarea'
+ insert_html_element t
+ eat_next_token_if_newline()
+ tok_state = tok_state_rcdata
+ original_ins_mode = ins_mode
+ flag_frameset_ok = false
+ ins_mode = ins_mode_text
+ return
+ if t.type is TYPE_START_TAG and t.name is 'xmp'
+ close_p_if_in_button_scope()
+ reconstruct_afe()
+ flag_frameset_ok = false
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'iframe'
+ flag_frameset_ok = false
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'noembed' or (t.name is 'noscript' and flag_scripting))
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'select'
+ reconstruct_afe()
+ insert_html_element t
+ flag_frameset_ok = false
+ if ins_mode is ins_mode_in_table or ins_mode is ins_mode_in_caption or ins_mode is ins_mode_in_table_body or ins_mode is ins_mode_in_row or ins_mode is ins_mode_in_cell
+ ins_mode = ins_mode_in_select_in_table
+ else
+ ins_mode = ins_mode_in_select
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'optgroup' or t.name is 'option')
+ if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML
+ open_els.shift()
+ reconstruct_afe()
+ insert_html_element t
+ return
+# this comment block implements the W3C spec
+# if t.type is TYPE_START_TAG and (t.name is 'rb' or t.name is 'rp' or t.name is 'rtc')
+# if is_in_scope 'ruby', NS_HTML
+# generate_implied_end_tags()
+# unless open_els[0].name is 'ruby' and open_els[0].namespace is NS_HTML
+# parse_error()
+# insert_html_element t
+# return
+# if t.type is TYPE_START_TAG and t.name is 'rt'
+# if is_in_scope 'ruby', NS_HTML
+# generate_implied_end_tags 'rtc' # arg is exception
+# unless (open_els[0].name is 'ruby' or open_els[0].name is 'rtc') and open_els[0].namespace is NS_HTML
+# parse_error()
+# insert_html_element t
+# return
+# below implements the WHATWG spec https://html.spec.whatwg.org/multipage/syntax.html#parsing-main-inbody
+ if t.type is TYPE_START_TAG and (t.name is 'rb' or t.name is 'rtc')
+ if is_in_scope 'ruby', NS_HTML
+ generate_implied_end_tags()
+ unless open_els[0].name is 'ruby' and open_els[0].namespace is NS_HTML
+ parse_error()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'rp' or t.name is 'rt')
+ if is_in_scope 'ruby', NS_HTML
+ generate_implied_end_tags 'rtc'
+ unless (open_els[0].name is 'ruby' or open_els[0].name is 'rtc') and open_els[0].namespace is NS_HTML
+ parse_error()
+ insert_html_element t
+ return
+# end WHATWG chunk
+ if t.type is TYPE_START_TAG and t.name is 'math'
+ reconstruct_afe()
+ adjust_mathml_attributes t
+ adjust_foreign_attributes t
+ insert_foreign_element t, NS_MATHML
+ if t.flag 'self-closing'
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'svg'
+ reconstruct_afe()
+ adjust_svg_attributes t
+ adjust_foreign_attributes t
+ insert_foreign_element t, NS_SVG
+ if t.flag 'self-closing'
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'frame' or t.name is 'head' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG # any other start tag
+ reconstruct_afe()
+ insert_html_element t
+ return
+ if t.type is TYPE_END_TAG # any other end tag
+ in_body_any_other_end_tag t.name
+ return
+ return
+
+ # 8.2.5.4.8 http://www.w3.org/TR/html5/syntax.html#parsing-main-incdata
+ ins_mode_text = (t) ->
+ if t.type is TYPE_TEXT
+ insert_character t
+ return
+ if t.type is TYPE_EOF
+ parse_error()
+ if open_els[0].name is 'script' and open_els[0].namespace is NS_HTML
+ open_els[0].flag 'already started', true
+ open_els.shift()
+ ins_mode = original_ins_mode
+ process_token t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'script'
+ open_els.shift()
+ ins_mode = original_ins_mode
+ # fixfull the spec seems to assume that I'm going to run the script
+ # http://www.w3.org/TR/html5/syntax.html#scriptEndTag
+ return
+ if t.type is TYPE_END_TAG
+ open_els.shift()
+ ins_mode = original_ins_mode
+ return
+ return