# 0: a "end of the list", "current node", "bottommost", "last"
+# browser
+# note: to get this to run outside a browser, you'll have to write a native
+# implementation of decode_named_char_ref()
+unless module?.exports?
+ window.wheic = {}
+ module = exports: window.wheic
# Each node is an obect of the Node class. Here are the Node types:
TYPE_TAG = 0 # name, {attributes}, [children]
@children = args.children ? []
@namespace = args.namespace ? NS_HTML
@parent = args.parent ? null
+ @token = args.token ? null
+ @flags = args.flags ? {}
if args.id?
@id = "#{args.id}+"
else
@id = "#{++prev_node_id}"
- shallow_clone: -> # return a new node that's the same except without the children or parent
- # WARNING this doesn't work right on open tags that are still being parsed
- attrs = {}
- attrs[k] = v for k, v of @attrs
- return new Node @type, name: @name, text: @text, attrs: attrs, namespace: @namespace, id: @id
acknowledge_self_closing: ->
- @flag 'did_self_close', true
- flag: ->
- # fixfull
+ if @token?
+ @token.flag 'did_self_close'
+ else
+ @flag 'did_self_close', true
+ flag: (key, value = null) ->
+ if value?
+ @flags[key] = value
+ else
+ return @flags[key]
serialize: (shallow = false, show_ids = false) -> # for unit tests
ret = ''
switch @type
ret += 'comment:'
ret += JSON.stringify @text
when TYPE_DOCTYPE
- ret += 'doctype'
- # FIXME
+ ret += "doctype:#{@name},#{JSON.stringify(@public_identifier ? '')},#{JSON.stringify(@system_identifier ? '')}"
when TYPE_AFE_MARKER
ret += 'marker'
when TYPE_AAA_BOOKMARK
new_text_node = (txt) ->
return new Node TYPE_TEXT, text: txt
new_character_token = new_text_node
-new_comment_node = (txt) ->
+new_comment_token = (txt) ->
return new Node TYPE_COMMENT, text: txt
+new_doctype_token = (name) ->
+ return new Node TYPE_DOCTYPE, name: name
new_eof_token = ->
return new Node TYPE_EOF
new_afe_marker = ->
alnum = lc_alpha + uc_alpha + digits
hex_chars = digits + "abcdefABCDEF"
+is_uc_alpha = (str) ->
+ return str.length is 1 and uc_alpha.indexOf(str) > -1
+is_lc_alpha = (str) ->
+ return str.length is 1 and lc_alpha.indexOf(str) > -1
+
# some SVG elements have dashes in them
tag_name_chars = alnum + "-"
is_space_tok = (t) ->
return t.type is TYPE_TEXT && t.text.length is 1 and space_chars.indexOf(t.text) > -1
+is_input_hidden_tok = (t) ->
+ return unless t.type is TYPE_START_TAG
+ for a of t.attrs_a
+ if a[0] is 'type'
+ if a[1].toLowerCase() is 'hidden'
+ return true
+ return false
+ return false
+
# https://en.wikipedia.org/wiki/Whitespace_character#Unicode
whitespace_chars = "\u0009\u000a\u000b\u000c\u000d\u0020\u0085\u00a0\u1680\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a\u2028\u2029\u202f\u205f\u3000"
u: true
}
+mathml_text_integration = {
+ mi: NS_MATHML, mo: NS_MATHML, mn: NS_MATHML, ms: NS_MATHML, mtext: NS_MATHML
+}
+is_mathml_text_integration_point = (el) ->
+ return mathml_text_integration[el.name] = el.namespace
+is_html_integration = (el) -> # DON'T PASS A TOKEN
+ if el.namespace is NS_MATHML and el.name is 'annotation-xml'
+ if el.attrs.encoding?
+ if el.attrs.encoding.toLowerCase() is 'text/html'
+ return true
+ if el.attrs.encoding.toLowerCase() is 'application/xhtml+xml'
+ return true
+ return false
+ if el.namespace is NS_SVG
+ if el.name is 'foreignObject' or el.name is 'desc' or el.name is 'title'
+ return true
+ return false
+
+h_tags = {
+ h1:NS_HTML, h2:NS_HTML, h3:NS_HTML, h4:NS_HTML, h5:NS_HTML, h6:NS_HTML
+}
+
+# FIXME namespacify
foster_parenting_targets = {
table: true
tbody: true
tr: true
}
+# FIXME namespacify
# all html I presume
end_tag_implied = {
dd: true
el_is_special = (e) ->
return special_elements[e.name] is e.namespace
+adp_els = { address: NS_HTML, div: NS_HTML, p: NS_HTML }
+el_is_special_not_adp = (el) ->
+ return special_elements[el.name] is el.namespace and adp_els[el.name] isnt el.namespace
+
+svg_name_fixes = {
+ altglyph: 'altGlyph'
+ altglyphdef: 'altGlyphDef'
+ altglyphitem: 'altGlyphItem'
+ animatecolor: 'animateColor'
+ animatemotion: 'animateMotion'
+ animatetransform: 'animateTransform'
+ clippath: 'clipPath'
+ feblend: 'feBlend'
+ fecolormatrix: 'feColorMatrix'
+ fecomponenttransfer: 'feComponentTransfer'
+ fecomposite: 'feComposite'
+ feconvolvematrix: 'feConvolveMatrix'
+ fediffuselighting: 'feDiffuseLighting'
+ fedisplacementmap: 'feDisplacementMap'
+ fedistantlight: 'feDistantLight'
+ fedropshadow: 'feDropShadow'
+ feflood: 'feFlood'
+ fefunca: 'feFuncA'
+ fefuncb: 'feFuncB'
+ fefuncg: 'feFuncG'
+ fefuncr: 'feFuncR'
+ fegaussianblur: 'feGaussianBlur'
+ feimage: 'feImage'
+ femerge: 'feMerge'
+ femergenode: 'feMergeNode'
+ femorphology: 'feMorphology'
+ feoffset: 'feOffset'
+ fepointlight: 'fePointLight'
+ fespecularlighting: 'feSpecularLighting'
+ fespotlight: 'feSpotLight'
+ fetile: 'feTile'
+ feturbulence: 'feTurbulence'
+ foreignobject: 'foreignObject'
+ glyphref: 'glyphRef'
+ lineargradient: 'linearGradient'
+ radialgradient: 'radialGradient'
+ textpath: 'textPath'
+}
+svg_attribute_fixes = {
+ attributename: 'attributeName'
+ attributetype: 'attributeType'
+ basefrequency: 'baseFrequency'
+ baseprofile: 'baseProfile'
+ calcmode: 'calcMode'
+ clippathunits: 'clipPathUnits'
+ contentscripttype: 'contentScriptType'
+ contentstyletype: 'contentStyleType'
+ diffuseconstant: 'diffuseConstant'
+ edgemode: 'edgeMode'
+ externalresourcesrequired: 'externalResourcesRequired'
+ filterres: 'filterRes'
+ filterunits: 'filterUnits'
+ glyphref: 'glyphRef'
+ gradienttransform: 'gradientTransform'
+ gradientunits: 'gradientUnits'
+ kernelmatrix: 'kernelMatrix'
+ kernelunitlength: 'kernelUnitLength'
+ keypoints: 'keyPoints'
+ keysplines: 'keySplines'
+ keytimes: 'keyTimes'
+ lengthadjust: 'lengthAdjust'
+ limitingconeangle: 'limitingConeAngle'
+ markerheight: 'markerHeight'
+ markerunits: 'markerUnits'
+ markerwidth: 'markerWidth'
+ maskcontentunits: 'maskContentUnits'
+ maskunits: 'maskUnits'
+ numoctaves: 'numOctaves'
+ pathlength: 'pathLength'
+ patterncontentunits: 'patternContentUnits'
+ patterntransform: 'patternTransform'
+ patternunits: 'patternUnits'
+ pointsatx: 'pointsAtX'
+ pointsaty: 'pointsAtY'
+ pointsatz: 'pointsAtZ'
+ preservealpha: 'preserveAlpha'
+ preserveaspectratio: 'preserveAspectRatio'
+ primitiveunits: 'primitiveUnits'
+ refx: 'refX'
+ refy: 'refY'
+ repeatcount: 'repeatCount'
+ repeatdur: 'repeatDur'
+ requiredextensions: 'requiredExtensions'
+ requiredfeatures: 'requiredFeatures'
+ specularconstant: 'specularConstant'
+ specularexponent: 'specularExponent'
+ spreadmethod: 'spreadMethod'
+ startoffset: 'startOffset'
+ stddeviation: 'stdDeviation'
+ stitchtiles: 'stitchTiles'
+ surfacescale: 'surfaceScale'
+ systemlanguage: 'systemLanguage'
+ tablevalues: 'tableValues'
+ targetx: 'targetX'
+ targety: 'targetY'
+ textlength: 'textLength'
+ viewbox: 'viewBox'
+ viewtarget: 'viewTarget'
+ xchannelselector: 'xChannelSelector'
+ ychannelselector: 'yChannelSelector'
+ zoomandpan: 'zoomAndPan'
+}
+adjust_mathml_attributes = (t) ->
+ for a in t.attrs_a
+ if a[0] is 'definitionurl'
+ a[0] = 'definitionURL'
+ return
+adjust_svg_attributes = (t) ->
+ for a in t.attrs_a
+ if svg_attribute_fixes[a[0]]?
+ a[0] = svg_attribute_fixes[a[0]]
+ return
+adjust_foreign_attributes = (t) ->
+ # fixfull
+ return
+
# decode_named_char_ref()
#
# The list of named character references is _huge_ so ask the browser to decode
doc = null
open_els = null # stack of open elements
afe = null # active formatting elements
- template_insertion_modes = null
- insertion_mode = null
- original_insertion_mode = null
+ template_ins_modes = null
+ ins_mode = null
+ original_ins_mode = null
tok_state = null
tok_cur_tag = null # partially parsed tag
flag_scripting = null
temporary_buffer = null
pending_table_character_tokens = null
head_element_pointer = null
+ flag_fragment_parsing = null
+ context_element = null
+
+ stop_parsing = ->
+ flag_parsing = false
parse_error = ->
if parse_error_cb?
if scope2[t.name] is t.namespace
return false
return false
- standard_scopers = { # FIXME these are supposed to be namespace specific
+ 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,
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.ns isnt NS_HTML t.name isnt 'optgroup' and t.name isnt 'option'
+ if t.ns 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
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_insertion_mode = ->
+ reset_ins_mode = ->
# 1. Let last be false.
last = false
# 2. Let node be the last node in the stack of open elements.
# 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'
- insertion_mode = ins_mode_in_select_in_table
+ 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.
- insertion_mode = ins_mode_in_select
+ 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 last is false
- insertion_mode = ins_mode_in_cell
+ 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'
- insertion_mode = ins_mode_in_row
+ 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'
- insertion_mode = ins_mode_in_table_body
+ 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'
- insertion_mode = ins_mode_in_caption
+ 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'
- insertion_mode = ins_mode_in_column_group
+ 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'
- insertion_mode = ins_mode_in_table
+ 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.
# insertion mode to "in body" ("in body"! not "in head"!) and abort
# these steps. (fragment case)
if node.name is 'head' and last
- insertion_mode = ins_mode_in_body
+ 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 last is false
- insertion_mode = ins_mode_in_head
+ 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'
- insertion_mode = ins_mode_in_body
+ 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'
- insertion_mode = ins_mode_in_frameset
+ ins_mode = ins_mode_in_frameset
return
# 16. If node is an html element, run these substeps:
if node.name is 'html'
# 1. If the head element pointer is null, switch the insertion
# mode to "before head" and abort these steps. (fragment case)
- # fixfull (fragment case)
-
- # 2. Otherwise, the head element pointer is not null, switch
- # the insertion mode to "after head" and abort these steps.
- insertion_mode = ins_mode_in_body # FIXME fixfull
+ 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
- insertion_mode = ins_mode_in_body
+ ins_mode = ins_mode_in_body
return
# 18. Let node now be the node before node in the stack of open
# elements.
node = open_els[node_i]
# 19. Return to the step labeled loop.
+ # 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_active_formatting_elements = ->
+ reconstruct_afe = ->
return if afe.length is 0
if afe[0].type is TYPE_AFE_MARKER or afe[0] in open_els
return
break
# Create
loop
- el = afe[i].shallow_clone()
- tree_insert_element el
+ el = insert_html_element afe[i].token
afe[i] = el
break if i is 0
i -= 1 # Advance
# 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 = node.shallow_clone()
+ new_node = token_to_element node.token, NS_HTML, ca
for t, i in afe
if t is node
afe[i] = new_node
# 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 = fe.shallow_clone() # FIXME intended parent thing
+ 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
return
dest[0].children.splice dest[1], 0, t
+
+ # 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 (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
last_template = null
last_template_i = null
for el, i in open_els
- if el.name is 'template'
+ if el.name is 'template' and el.namespace is NS_HTML
last_template = el
last_template_i = i
break
last_table = null
last_table_i
for el, i in open_els
- if el.name is 'table'
+ if el.name is 'table' and el.namespace is NS_HTML
last_table = el
last_table_i = i
break
# 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 = template # fixfull should be it's contents
+ 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
# 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.
# 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) ->
- t.type = TYPE_TAG # not TYPE_START_TAG
# convert attributes into a hash
attrs = {}
- while t.attrs_a.length
- a = t.attrs_a.pop()
+ 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
+ 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
open_els.unshift el
return el
# http://www.w3.org/TR/html5/syntax.html#insert-an-html-element
- insert_html_element = insert_foreign_element # (token, namespace) ->
-
- # FIXME read implement "foster parenting" part
- # FIXME read spec, do this right
- # FIXME implement the override target thing
- # note: this assumes it's an open tag
- # FIXME what part of the spec is this?
- # TODO look through all callers of this, and see what they should really be doing.
- # eg probably insert_html_element for tokens
- tree_insert_element = (el, override_target = null, namespace = null) ->
- if namespace?
- el.namespace = namespace
- dest = adjusted_insertion_location override_target
- if el.type is TYPE_START_TAG # means it's a "token"
- el = token_to_element el, namespace, dest[0]
- unless el.namespace?
- namespace = dest.namespace
- # fixfull: Document nodes sometimes can't accept more chidren
- dest[0].children.splice dest[1], 0, el
- el.parent = dest[0]
- open_els.unshift el
- return el
+ insert_html_element = (token) ->
+ insert_foreign_element token, NS_HTML
# http://www.w3.org/TR/html5/syntax.html#insert-a-comment
# position should be [node, index_within_children]
parse_generic_raw_text = (t) ->
insert_html_element t
tok_state = tok_state_rawtext
- original_insertion_mode = insertion_mode
- insertion_mode = ins_mode_text
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_text
parse_generic_rcdata_text = (t) ->
insert_html_element t
tok_state = tok_state_rcdata
- original_insertion_mode = insertion_mode
- insertion_mode = ins_mode_text
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_text
# 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
if is_space_tok t
return
if t.type is TYPE_COMMENT
- # fixfull this is supposed to be "the last child of the document object"
+ # ?fixfull
doc.children.push t
return
if t.type is TYPE_DOCTYPE
+ # FIXME check identifiers, set quirks, etc
# fixfull
- t.name = 'html'
doc.children.push t
- insertion_mode = ins_mode_before_html
+ ins_mode = ins_mode_before_html
return
# Anything else
#fixfull (iframe, quirks)
- insertion_mode = ins_mode_before_html
- insertion_mode t # reprocess the token
+ 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
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
open_els.unshift(el)
# fixfull (big paragraph in spec about manifest, fragment, urls, etc)
- insertion_mode = ins_mode_before_head
+ 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'
doc.children.push el
open_els.unshift el
# ?fixfull browsing context
- insertion_mode = ins_mode_before_head
- insertion_mode t
+ 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
if t.type is TYPE_START_TAG and t.name is 'head'
el = insert_html_element t
head_element_pointer = el
- insertion_mode = ins_mode_in_head
+ ins_mode = ins_mode_in_head
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
head_tok = new_open_tag 'head'
el = insert_html_element head_tok
head_element_pointer = el
- insertion_mode = ins_mode_in_head
- insertion_mode t # reprocess current token
+ ins_mode = ins_mode_in_head
+ process_token t
# 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
- insertion_mode = ins_mode_after_head
- insertion_mode t
+ ins_mode = ins_mode_after_head
+ process_token t
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
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()
- el.acknowledge_self_closing()
+ 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()
- el.acknowledge_self_closing()
+ t.acknowledge_self_closing()
# fixfull encoding stuff
return
if t.type is TYPE_START_TAG and t.name is 'title'
- parse_generic_rcdata_element t
+ 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
- insertion_mode = in_head_noscript # FIXME implement
+ 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 # FIXME implement
+ 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_insertion_mode = insertion_mode # make sure orig... is defined
- insertion_mode = ins_mode_text # FIXME implement
+ 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
- insertion_mode = ins_mode_after_head
+ 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
insert_html_element t
afe_push_marker()
flag_frameset_ok = false
- insertion_mode = ins_mode_in_template
- template_insertion_modes.unshift ins_mode_in_template # FIXME implement
+ 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()
if el.name is 'template'
break
clear_afe_to_marker()
- template_insertion_modes.shift()
- reset_insertion_mode()
+ template_ins_modes.shift()
+ reset_ins_mode()
else
parse_error()
return
parse_error()
return
ins_mode_in_head_else t
-
+
# 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
ins_mode_in_head_noscript = (t) ->
- # FIXME ?fixfull
- console.log "ins_mode_in_head_noscript unimplemented"
-
+ if t.type is TYPE_DOCTYPE
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG
+ 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 (t.type is TYPE_TEXT and (t.text is "\t" or t.text is "\u000a" or t.text is "\u000c" or t.text is "\u000d" or t.text is ' ')) 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
- insertion_mode = ins_mode_in_body
- insertion_mode t # reprocess token
+ ins_mode = ins_mode_in_body
+ process_token t
return
ins_mode_after_head = (t) ->
if is_space_tok t
if t.type is TYPE_START_TAG and t.name is 'body'
insert_html_element t
flag_frameset_ok = false
- insertion_mode = ins_mode_in_body
+ ins_mode = ins_mode_in_body
return
if t.type is TYPE_START_TAG and t.name is 'frameset'
insert_html_element t
- insertion_mode = ins_mode_in_frameset
+ 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()
# 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
- for node, i in open_els
- if node.name is name # FIXME check namespace too
+ for el, i in open_els
+ if el.namespace is NS_HTML and el.name is name
generate_implied_end_tags name # arg is exception
parse_error() unless i is 0
while i >= 0
open_els.shift()
i -= 1
return
- if special_elements[node.name]? # FIXME check namespac too
+ if special_elements[el.name] is el.namespace
parse_error()
return
- ins_mode_in_body = (t) ->
- switch t.type
- when TYPE_TEXT
- switch t.text
- when "\u0000"
- parse_error()
- when "\t", "\u000a", "\u000c", "\u000d", ' '
- reconstruct_active_formatting_elements()
- insert_character t
- else
- reconstruct_active_formatting_elements()
- insert_character t
- flag_frameset_ok = false
- when TYPE_COMMENT
- insert_comment t
- when TYPE_DOCTYPE
- parse_error()
- when TYPE_START_TAG
- switch t.name
- when 'html'
- parse_error()
- return if template_tag_is_open()
- root_attrs = open_els[open_els.length - 1].attrs
- for k, v of t.attrs
- root_attrs[k] = v unless root_attrs[k]?
- when 'base', 'basefont', 'bgsound', 'link', 'meta', 'noframes', 'script', 'style', 'template', 'title'
- # FIXME also do this for </template> (end tag)
- return ins_mode_in_head t
- when 'body'
- parse_error()
- # TODO
- when 'frameset'
- parse_error()
- # TODO
- when 'address', 'article', 'aside', 'blockquote', 'center', 'details', 'dialog', 'dir', 'div', 'dl', 'fieldset', 'figcaption', 'figure', 'footer', 'header', 'hgroup', 'main', 'nav', 'ol', 'p', 'section', 'summary', 'ul'
- close_p_if_in_button_scope()
- insert_html_element t
- when 'h1', 'h2', 'h3', 'h4', 'h5', 'h6'
- close_p_if_in_button_scope()
- if open_els[0].name in ['h1', 'h2', 'h3', 'h4', 'h5', 'h6']
- parse_error()
- open_els.shift()
- insert_html_element t
- # TODO lots more to implement here
- when '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'
- 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_active_formatting_elements()
- el = insert_html_element t
- afe_push el
- when 'b', 'big', 'code', 'em', 'font', 'i', 's', 'small', 'strike', 'strong', 'tt', 'u'
- reconstruct_active_formatting_elements()
- el = insert_html_element t
- afe_push el
- when 'table'
- # fixfull quirksmode thing
- close_p_if_in_button_scope()
- insert_html_element t
- insertion_mode = ins_mode_in_table
- # TODO lots more to implement here
- else # any other start tag
- reconstruct_active_formatting_elements()
- insert_html_element t
- when TYPE_EOF
- ok_tags = {
- dd: true, dt: true, li: true, p: true, tbody: true, td: true,
- tfoot: true, th: true, thead: true, tr: true, body: true, html: true,
- }
- for t in open_els
- unless ok_tags[t.name]?
- parse_error()
- break
- # TODO stack of template insertion modes thing
- flag_parsing = false # stop parsing
- when TYPE_END_TAG
- switch t.name
- when 'body'
- unless is_in_scope 'body'
- parse_error()
- return
- # TODO implement parse error and move to tree_after_body
- when 'html'
- unless is_in_scope 'body' # weird, but it's what the spec says
- parse_error()
- return
- # TODO implement parse error and move to tree_after_body, reprocess
- when 'address', 'article', 'aside', 'blockquote', 'button', 'center', 'details', 'dialog', 'dir', 'div', 'dl', 'fieldset', 'figcaption', 'figure', 'footer', 'header', 'hgroup', 'listing', 'main', 'nav', 'ol', 'pre', 'section', 'summary', '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
- # TODO lots more close tags to implement here
- when 'p'
- unless is_in_button_scope 'p'
- parse_error()
- insert_html_element new_open_tag 'p'
- close_p_element()
- # TODO lots more close tags to implement here
- when 'a', 'b', 'big', 'code', 'em', 'font', 'i', 'nobr', 's', 'small', 'strike', 'strong', 'tt', 'u'
- adoption_agency t.name
- # TODO lots more close tags to implement here
- else
- in_body_any_other_end_tag t.name
return
-
- ins_mode_in_table_else = (t) ->
- parse_error()
- flag_foster_parenting = true # FIXME
- ins_mode_in_body t
- flag_foster_parenting = false
- can_in_table = { # FIXME do this inline like everywhere else
- 'table': true
- 'tbody': true
- 'tfoot': true
- 'thead': true
- 'tr': true
- }
-
- # 8.2.5.4.8 http://www.w3.org/TR/html5/syntax.html#parsing-main-incdata
- ins_mode_text = (t) ->
+ 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_EOF
- parse_error()
- if open_els[0].name is 'script'
- open_els[0].flag 'already started', true
- open_els.shift()
- insertion_mode = original_insertion_mode
- insertion_mode t
+ if t.type is TYPE_COMMENT
+ insert_comment t
return
- if t.type is TYPE_END_TAG and t.name is 'script'
- open_els.shift()
- insertion_mode = original_insertion_mode
- # fixfull the spec seems to assume that I'm going to run the script
- # http://www.w3.org/TR/html5/syntax.html#scriptEndTag
+ if t.type is TYPE_DOCTYPE
+ parse_error()
return
- if t.type is TYPE_END_TAG
- open_els.shift()
- insertion_mode = original_insertion_mode
+ 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 of t.attrs_a
+ root_attrs[a[0]] = a[1] unless root_attrs[a[0]]?
return
- console.log 'warning: end of ins_mode_text reached'
-
- # the functions below implement the tokenizer stats described here:
- # http://www.w3.org/TR/html5/syntax.html#tokenization
-
- # 8.2.5.4.9 http://www.w3.org/TR/html5/syntax.html#parsing-main-intable
- ins_mode_in_table = (t) ->
- switch t.type
- when TYPE_TEXT
- if can_in_table[t.name]
- original_insertion_mode = insertion_mode
- insertion_mode = ins_mode_in_table_text
- insertion_mode t
- else
- ins_mode_in_table_else t
- when TYPE_COMMENT
- insert_comment t
- when TYPE_DOCTYPE
- parse_error()
- when TYPE_START_TAG
- switch t.name
- when 'caption'
- clear_stack_to_table_context()
- afe_push_marker()
- insert_html_element t
- insertion_mode = ins_mode_in_caption
- when 'colgroup'
- clear_stack_to_table_context()
- insert_html_element t
- insertion_mode = ins_mode_in_column_group
- when 'col'
- clear_stack_to_table_context()
- insert_html_element new_open_tag 'colgroup'
- insertion_mode = ins_mode_in_column_group
- insertion_mode t
- when 'tbody', 'tfoot', 'thead'
- clear_stack_to_table_context()
- insert_html_element t
- insertion_mode = ins_mode_in_table_body
- when 'td', 'th', 'tr'
- clear_stack_to_table_context()
- insert_html_element new_open_tag 'tbody'
- insertion_mode = ins_mode_in_table_body
- insertion_mode t
- when 'table'
- parse_error()
- if is_in_table_scope 'table'
- loop
- el = open_els.shift()
- if el.name is 'table'
- break
- reset_insertion_mode()
- insertion_mode t
- when 'style', 'script', 'template'
- ins_mode_in_head t
- when 'input'
- if token_is_input_hidden t
- ins_mode_in_table_else t
- else
- parse_error()
- el = insert_html_element t
- open_els.shift()
- el.acknowledge_self_closing()
- when 'form'
- parse_error()
- if form_element_pointer?
- return
- if template_tag_is_open()
- return
- form_element_pointer = insert_html_element t
- open_els.shift()
- else
- ins_mode_in_table_else t
- when TYPE_END_TAG
- switch t.name
- when 'table'
- if is_in_table_scope 'table'
- loop
- el = open_els.shift()
- if el.name is 'table'
- break
- reset_insertion_mode()
- else
- parse_error
- when 'body', 'caption', 'col', 'colgroup', 'html', 'tbody', 'td', 'tfoot', 'th', 'thead', 'tr'
- parse_error()
- when 'template'
- ins_mode_in_head t
- else
- ins_mode_in_table_else t
- when TYPE_EOF
- ins_mode_in_body t
- else
- ins_mode_in_table_else t
-
- # 8.2.5.4.10 http://www.w3.org/TR/html5/syntax.html#parsing-main-intabletext
- ins_mode_in_table_text = (t) ->
- if t.type is TYPE_TEXT and t.text is "\u0000"
- # huh? I thought the tokenizer didn't emit these
+ 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.ns is NS_HTML
+ return unless second.name is 'body'
+ return if template_tag_is_open()
+ frameset_ok_flag = false
+ for a of t.attrs_a
+ second.attrs[a[0]] = a[1] unless second.attrs[a[0]]?
return
- if t.type is TYPE_TEXT
- pending_table_character_tokens.push t
+ 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.ns is NS_HTML
+ return unless second.name is 'body'
+ flag_frameset_ok = false
+ 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
- # Anything else
- all_space = true
- for old in pending_table_character_tokens
- unless is_space_tok old
- all_space = false
- break
- if all_space
- for old in pending_table_character_tokens
- insert_character old
- else
- for old in pending_table_character_tokens
- ins_mode_table_else old
- pending_table_character_tokens = [] # FIXME test (spec doesn't say this)
- insertion_mode = original_insertion_mode
- insertion_mode t
-
- # 8.2.5.4.11 http://www.w3.org/TR/html5/syntax.html#parsing-main-incaption
- ins_mode_in_caption = (t) ->
- if t.type is TYPE_END_TAG and t.name is 'caption'
- if is_in_table_scope 'caption'
- generate_implied_end_tags()
- if open_els[0].name isnt 'caption'
+ 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()
- loop
- el = open_els.shift()
- if el.name is 'caption'
- break
- clear_afe_to_marker()
- insertion_mode = in_table
+ break
+ if template_ins_modes.length > 0
+ ins_mode_in_template t
else
- parse_error()
- # fragment case
- 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 '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')) or t.type is TYPE_END_TAG and t.name is 'table'
- parse_error()
- if is_in_table_scope 'caption'
- loop
- el = open_els.shift()
- if el.name is 'caption'
- break
- clear_afe_to_marker()
- insertion_mode = in_table
- insertion_mode t
- # else fragment case
- return
- if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' 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
- # Anything else
- ins_mode_in_body t
-
- # 8.2.5.4.12 http://www.w3.org/TR/html5/syntax.html#parsing-main-incolgroup
- ins_mode_in_column_group = (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 'col'
- el = insert_html_element t
- open_els.shift()
- el.acknowledge_self_closing()
+ stop_parsing()
return
- if t.type is TYPE_END_TAG and t.name is 'colgroup'
- if open_els[0].name is 'colgroup'
- open_els[0].shift()
- insertion_mode = ins_mode_in_table
- else
+ if t.type is TYPE_END_TAG and t.name is 'body'
+ unless is_in_scope 'body'
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 'col'
- parse_error()
- return
- if (t.type is TYPE_START_TAG or t.type is TYPE_END_TAG) and t.name is 'template'
- ins_mode_in_head t
- return
- if t.type is TYPE_EOF
- ins_mode_in_body t
+ if t.type is TYPE_END_TAG and t.name is 'html'
+ unless is_in_scope 'body'
+ 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
- # Anything else
- if open_els[0].name isnt 'colgroup'
- parse_error()
+ 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
- open_els.shift()
- insertion_mode = ins_mode_in_table
- insertion_mode t
- return
-
- # 8.2.5.4.13 http://www.w3.org/TR/html5/syntax.html#parsing-main-intbody
- ins_mode_in_table_body = (t) ->
- if t.type is TYPE_START_TAG and t.name is 'tr'
- clear_stack_to_table_body_context()
+ if t.type is TYPE_START_TAG and h_tags[t.name]?
+ close_p_if_in_button_scope()
+ if h_tags[open_els[0]] is NS_HTML
+ parse_error()
+ open_els.shift()
insert_html_element t
- insertion_mode = ins_mode_in_row
return
- if t.type is TYPE_START_TAG and (t.name is 'th' or t.name is 'td')
- parse_error()
- clear_stack_to_table_body_context()
- insert_html_element new_open_tag 'tr'
- insertion_mode = ins_mode_in_row
- insertion_mode t
+ 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
+ # spec: If the next token is a "LF" (U+000A) character token, then
+ # ignore that token and move on to the next one. (Newlines at the
+ # start of pre blocks are ignored as an authoring convenience.)
+ if txt.charAt(cur) is "\u000a" # FIXME check for crlf?
+ cur += 1
+ flag_frameset_ok = false
return
- if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
- unless is_in_table_scope t.name # fixfull check namespace
+ 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
- clear_stack_to_table_body_context()
- open_els.shift()
- insertion_mode = ins_mode_in_table
+ 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 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')) or (t.type is TYPE_END_TAG and t.name is 'table')
- has = false
- for el in open_els
- if el.name is 'tbody' or el.name is 'tfoot' or el.name is 'thead'
- has = true
- break
- if table_scopers[el.name]
+ 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 !has
- parse_error()
- return
- clear_stack_to_table_body_context()
- open_els.shift()
- insertion_mode = ins_mode_in_table
- insertion_mode t
+ if el_is_special_not_adp node
+ break
+ close_p_if_in_button_scope()
+ insert_html_element t
return
- if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th' or t.name is 'tr')
- parse_error()
+ 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
- # Anything else
- ins_mode_in_table t
-
- # 8.2.5.4.14 http://www.w3.org/TR/html5/syntax.html#parsing-main-intr
- ins_mode_in_row = (t) ->
- if t.type is TYPE_START_TAG and (t.name is 'th' or t.name is 'td')
- clear_stack_to_table_row_context()
+ if t.type is TYPE_START_TAG and t.name is 'plaintext'
+ close_p_if_in_button_scope()
insert_html_element t
- insertion_mode = ins_mode_in_cell
- afe_push_marker()
+ tok_state = tok_state_plaintext
return
- if t.type is TYPE_END_TAG and t.name is 'tr'
- if is_in_table_scope 'tr'
- clear_stack_to_table_row_context()
- open_els.shift()
- insertion_mode = ins_mode_in_table_body
- else
+ 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_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')) or t.type is TYPE_END_TAG and t.name is 'table'
- if is_in_table_scope 'tr'
- clear_stack_to_table_row_context()
- open_els.shift()
- insertion_mode = ins_mode_in_table_body
- insertion_mode t
- else
+ 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
- if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
- if is_in_table_scope t.name # fixfull namespace
- if is_in_table_scope 'tr'
- clear_stack_to_table_row_context()
- open_els.shift()
- insertion_mode = ins_mode_in_table_body
- insertion_mode t
- else
+ 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 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th')
- parse_error()
- return
- # Anything else
- ins_mode_in_table t
-
- # http://www.w3.org/TR/html5/syntax.html#close-the-cell
- close_the_cell = ->
- generate_implied_end_tags()
- unless open_els[0].name is 'td' or open_els[0] is 'th'
- parse_error()
- loop
- el = open_els.shift()
- if el.name is 'td' or el.name is 'th'
- break
- clear_afe_to_marker()
- insertion_mode = ins_mode_in_row
-
- # 8.2.5.4.15 http://www.w3.org/TR/html5/syntax.html#parsing-main-intd
- ins_mode_in_cell = (t) ->
- if t.type is TYPE_END_TAG and (t.name is 'td' or t.name is 'th')
- if is_in_table_scope t.name
+ 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].name isnt t.name
- parse_error
+ 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 t.name
+ if el.name is 'form' and el.namespace is NS_HTML
break
- clear_afe_to_marker()
- insertion_mode = ins_mode_in_row
- else
+ 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_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' 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')
- has = false
+ 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 el.name is 'td' or el.name is 'th'
- has = true
+ if h_tags[el.name] is el.namespace
+ h_in_scope = true
break
- if table_scopers[el.name]
+ if standard_scopers[el.name] is el.namespace
break
- if !has
+ unless h_in_scope
parse_error()
return
- close_the_cell()
- insertion_mode t
- return
- if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html')
- parse_error()
+ 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
- if t.type is TYPE_END_TAG and (t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')
- if is_in_table_scope t.name # fixfull namespace
- close_the_cell()
- insertion_mode t
- else
+ # 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
- # Anything Else
- ins_mode_in_body t
-
- # 8.2.5.4.16 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselect
- ins_mode_in_select = (t) ->
- if t.type is TYPE_TEXT and t.text is "\u0000"
- parse_error()
+ 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_TEXT
- insert_character t
+ if t.type is TYPE_START_TAG and t.name is 'nobr'
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
return
- if t.type is TYPE_COMMENT
- insert_comment t
+ 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_DOCTYPE
- parse_error()
+ 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_START_TAG and t.name is 'html'
- ins_mode_in_body t
+ 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 'option'
- if open_els[0].name is 'option'
- open_els.shift()
+ if t.type is TYPE_START_TAG and t.name is 'table'
+ close_p_if_in_button_scope() # fixfull quirksmode thing
insert_html_element t
+ flag_frameset_ok = false
+ ins_mode = ins_mode_in_table
return
- if t.type is TYPE_START_TAG and t.name is 'optgroup'
- if open_els[0].name is 'option'
- open_els.shift()
- if open_els[0].name is 'optgroup'
- open_els.shift()
+ if t.type is TYPE_END_TAG and t.name is 'br'
+ parse_error()
+ t.type is TYPE_START_TAG
+ # 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_END_TAG and t.name is 'optgroup'
- if open_els[0].name is 'option' and open_els[1].name is 'optgroup'
+ 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 'param' or t.name is 'source' or t.name is 'track')
+ 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
+ if txt.charAt(cur) is "\u000a" # FIXME check for crlf?
+ cur += 1
+ 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()
- if open_els[0].name is 'optgroup'
+ reconstruct_afe()
+ insert_html_element t
+ return
+ 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
+ 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'
+ 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
+ console.log 'warning: end of ins_mode_text reached'
+
+ # the functions below implement the tokenizer stats described here:
+ # http://www.w3.org/TR/html5/syntax.html#tokenization
+
+ # 8.2.5.4.9 http://www.w3.org/TR/html5/syntax.html#parsing-main-intable
+ ins_mode_in_table_else = (t) ->
+ parse_error()
+ flag_foster_parenting = true
+ ins_mode_in_body t
+ flag_foster_parenting = false
+ return
+ can_in_table = { # FIXME do this inline like everywhere else
+ 'table': true
+ 'tbody': true
+ 'tfoot': true
+ 'thead': true
+ 'tr': true
+ }
+ ins_mode_in_table = (t) ->
+ switch t.type
+ when TYPE_TEXT
+ if can_in_table[t.name]
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_in_table_text
+ process_token t
+ else
+ ins_mode_in_table_else t
+ when TYPE_COMMENT
+ insert_comment t
+ when TYPE_DOCTYPE
+ parse_error()
+ when TYPE_START_TAG
+ switch t.name
+ when 'caption'
+ clear_stack_to_table_context()
+ afe_push_marker()
+ insert_html_element t
+ ins_mode = ins_mode_in_caption
+ when 'colgroup'
+ clear_stack_to_table_context()
+ insert_html_element t
+ ins_mode = ins_mode_in_column_group
+ when 'col'
+ clear_stack_to_table_context()
+ insert_html_element new_open_tag 'colgroup'
+ ins_mode = ins_mode_in_column_group
+ process_token t
+ when 'tbody', 'tfoot', 'thead'
+ clear_stack_to_table_context()
+ insert_html_element t
+ ins_mode = ins_mode_in_table_body
+ when 'td', 'th', 'tr'
+ clear_stack_to_table_context()
+ insert_html_element new_open_tag 'tbody'
+ ins_mode = ins_mode_in_table_body
+ process_token t
+ when 'table'
+ parse_error()
+ if is_in_table_scope 'table'
+ loop
+ el = open_els.shift()
+ if el.name is 'table'
+ break
+ reset_ins_mode()
+ process_token t
+ when 'style', 'script', 'template'
+ ins_mode_in_head t
+ when 'input'
+ unless is_input_hidden_tok t
+ ins_mode_in_table_else t
+ else
+ parse_error()
+ el = insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ when 'form'
+ parse_error()
+ if form_element_pointer?
+ return
+ if template_tag_is_open()
+ return
+ form_element_pointer = insert_html_element t
+ open_els.shift()
+ else
+ ins_mode_in_table_else t
+ when TYPE_END_TAG
+ switch t.name
+ when 'table'
+ if is_in_table_scope 'table'
+ loop
+ el = open_els.shift()
+ if el.name is 'table'
+ break
+ reset_ins_mode()
+ else
+ parse_error
+ when 'body', 'caption', 'col', 'colgroup', 'html', 'tbody', 'td', 'tfoot', 'th', 'thead', 'tr'
+ parse_error()
+ when 'template'
+ ins_mode_in_head t
+ else
+ ins_mode_in_table_else t
+ when TYPE_EOF
+ ins_mode_in_body t
+ else
+ ins_mode_in_table_else t
+
+
+ # 8.2.5.4.10 http://www.w3.org/TR/html5/syntax.html#parsing-main-intabletext
+ ins_mode_in_table_text = (t) ->
+ if t.type is TYPE_TEXT and t.text is "\u0000"
+ # huh? I thought the tokenizer didn't emit these
+ parse_error()
+ return
+ if t.type is TYPE_TEXT
+ pending_table_character_tokens.push t
+ return
+ # Anything else
+ all_space = true
+ for old in pending_table_character_tokens
+ unless is_space_tok old
+ all_space = false
+ break
+ if all_space
+ for old in pending_table_character_tokens
+ insert_character old
+ else
+ for old in pending_table_character_tokens
+ ins_mode_table_else old
+ pending_table_character_tokens = [] # FIXME test (spec doesn't say this)
+ ins_mode = original_ins_mode
+ process_token t
+
+ # 8.2.5.4.11 http://www.w3.org/TR/html5/syntax.html#parsing-main-incaption
+ ins_mode_in_caption = (t) ->
+ if t.type is TYPE_END_TAG and t.name is 'caption'
+ if is_in_table_scope 'caption'
+ generate_implied_end_tags()
+ if open_els[0].name isnt 'caption'
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'caption'
+ break
+ clear_afe_to_marker()
+ ins_mode = ins_mode_in_table
else
parse_error()
+ # fragment case
return
- if t.type is TYPE_END_TAG and t.name is 'option'
- if open_els[0].name is 'option'
+ 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 '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')) or t.type is TYPE_END_TAG and t.name is 'table'
+ parse_error()
+ if is_in_table_scope 'caption'
+ loop
+ el = open_els.shift()
+ if el.name is 'caption'
+ break
+ clear_afe_to_marker()
+ ins_mode = ins_mode_in_table
+ process_token t
+ # else fragment case
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' 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
+ # Anything else
+ ins_mode_in_body t
+
+ # 8.2.5.4.12 http://www.w3.org/TR/html5/syntax.html#parsing-main-incolgroup
+ ins_mode_in_column_group = (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 'col'
+ el = insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'colgroup'
+ if open_els[0].name is 'colgroup'
open_els.shift()
+ ins_mode = ins_mode_in_table
else
parse_error()
- return
- if t.type is TYPE_END_TAG and t.name is 'select'
- if is_in_select_scope 'select'
- loop
- el = open_els.shift()
- if el.name is 'select'
- break
- reset_insertion_mode()
- else
+ return
+ if t.type is TYPE_END_TAG and t.name is 'col'
+ parse_error()
+ return
+ if (t.type is TYPE_START_TAG or t.type is TYPE_END_TAG) and t.name is 'template'
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_EOF
+ ins_mode_in_body t
+ return
+ # Anything else
+ if open_els[0].name isnt 'colgroup'
+ parse_error()
+ return
+ open_els.shift()
+ ins_mode = ins_mode_in_table
+ process_token t
+ return
+
+ # 8.2.5.4.13 http://www.w3.org/TR/html5/syntax.html#parsing-main-intbody
+ ins_mode_in_table_body = (t) ->
+ if t.type is TYPE_START_TAG and t.name is 'tr'
+ clear_stack_to_table_body_context()
+ insert_html_element t
+ ins_mode = ins_mode_in_row
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'th' or t.name is 'td')
+ parse_error()
+ clear_stack_to_table_body_context()
+ insert_html_element new_open_tag 'tr'
+ ins_mode = ins_mode_in_row
+ process_token t
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
+ unless is_in_table_scope t.name # fixfull check namespace
+ parse_error()
+ return
+ clear_stack_to_table_body_context()
+ open_els.shift()
+ ins_mode = ins_mode_in_table
+ 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 'tbody' or t.name is 'tfoot' or t.name is 'thead')) or (t.type is TYPE_END_TAG and t.name is 'table')
+ has = false
+ for el in open_els
+ if el.name is 'tbody' or el.name is 'tfoot' or el.name is 'thead'
+ has = true
+ break
+ if table_scopers[el.name]
+ break
+ if !has
+ parse_error()
+ return
+ clear_stack_to_table_body_context()
+ open_els.shift()
+ ins_mode = ins_mode_in_table
+ process_token t
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th' or t.name is 'tr')
+ parse_error()
+ return
+ # Anything else
+ ins_mode_in_table t
+
+ # 8.2.5.4.14 http://www.w3.org/TR/html5/syntax.html#parsing-main-intr
+ ins_mode_in_row = (t) ->
+ if t.type is TYPE_START_TAG and (t.name is 'th' or t.name is 'td')
+ clear_stack_to_table_row_context()
+ insert_html_element t
+ ins_mode = ins_mode_in_cell
+ afe_push_marker()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'tr'
+ if is_in_table_scope 'tr'
+ clear_stack_to_table_row_context()
+ open_els.shift()
+ ins_mode = ins_mode_in_table_body
+ else
+ parse_error()
+ 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 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')) or t.type is TYPE_END_TAG and t.name is 'table'
+ if is_in_table_scope 'tr'
+ clear_stack_to_table_row_context()
+ open_els.shift()
+ ins_mode = ins_mode_in_table_body
+ process_token t
+ else
+ parse_error()
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
+ if is_in_table_scope t.name # fixfull namespace
+ if is_in_table_scope 'tr'
+ clear_stack_to_table_row_context()
+ open_els.shift()
+ ins_mode = ins_mode_in_table_body
+ process_token t
+ else
+ parse_error()
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th')
+ parse_error()
+ return
+ # Anything else
+ ins_mode_in_table t
+
+ # http://www.w3.org/TR/html5/syntax.html#close-the-cell
+ close_the_cell = ->
+ generate_implied_end_tags()
+ unless open_els[0].name is 'td' or open_els[0] is 'th'
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'td' or el.name is 'th'
+ break
+ clear_afe_to_marker()
+ ins_mode = ins_mode_in_row
+
+ # 8.2.5.4.15 http://www.w3.org/TR/html5/syntax.html#parsing-main-intd
+ ins_mode_in_cell = (t) ->
+ if t.type is TYPE_END_TAG and (t.name is 'td' or t.name is 'th')
+ if is_in_table_scope t.name
+ generate_implied_end_tags()
+ if open_els[0].name isnt t.name
+ parse_error
+ loop
+ el = open_els.shift()
+ if el.name is t.name
+ break
+ clear_afe_to_marker()
+ ins_mode = ins_mode_in_row
+ else
+ parse_error()
+ 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 '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')
+ has = false
+ for el in open_els
+ if el.name is 'td' or el.name is 'th'
+ has = true
+ break
+ if table_scopers[el.name]
+ break
+ if !has
+ parse_error()
+ return
+ close_the_cell()
+ process_token t
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html')
+ parse_error()
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')
+ if is_in_table_scope t.name # fixfull namespace
+ close_the_cell()
+ process_token t
+ else
+ parse_error()
+ return
+ # Anything Else
+ ins_mode_in_body t
+
+ # 8.2.5.4.16 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselect
+ ins_mode_in_select = (t) ->
+ if t.type is TYPE_TEXT and t.text is "\u0000"
+ parse_error()
+ return
+ if t.type is TYPE_TEXT
+ 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 'option'
+ if open_els[0].name is 'option'
+ open_els.shift()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'optgroup'
+ if open_els[0].name is 'option'
+ open_els.shift()
+ if open_els[0].name is 'optgroup'
+ open_els.shift()
+ insert_html_element t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'optgroup'
+ if open_els[0].name is 'option' and open_els[1].name is 'optgroup'
+ open_els.shift()
+ if open_els[0].name is 'optgroup'
+ open_els.shift()
+ else
+ parse_error()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'option'
+ if open_els[0].name is 'option'
+ open_els.shift()
+ else
+ parse_error()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'select'
+ if is_in_select_scope 'select'
+ loop
+ el = open_els.shift()
+ if el.name is 'select'
+ break
+ reset_ins_mode()
+ else
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'select'
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'select'
+ break
+ reset_ins_mode()
+ # spec says that this is the same as </select> but it doesn't say
+ # to check scope first
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'input' or t.name is 'keygen' or t.name is 'textarea')
+ parse_error()
+ if is_in_select_scope 'select'
+ return
+ loop
+ el = open_els.shift()
+ if el.name is 'select'
+ break
+ reset_ins_mode()
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'script' or t.name is 'template')
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_EOF
+ ins_mode_in_body t
+ return
+ # Anything else
+ parse_error()
+ return
+
+ # 8.2.5.4.17 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselectintable
+ ins_mode_in_select_in_table = (t) ->
+ if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th')
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'select'
+ break
+ reset_ins_mode()
+ process_token t
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th')
+ parse_error()
+ unless is_in_table_scope t.name, NS_HTML
+ return
+ loop
+ el = open_els.shift()
+ if el.name is 'select'
+ break
+ reset_ins_mode()
+ process_token t
+ return
+ # Anything else
+ ins_mode_in_select t
+ return
+
+ # 8.2.5.4.18 http://www.w3.org/TR/html5/syntax.html#parsing-main-intemplate
+ ins_mode_in_template = (t) ->
+ if t.type is TYPE_TEXT or t.type is TYPE_COMMENT or t.type is TYPE_DOCTYPE
+ 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' 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 'caption' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
+ template_ins_modes.shift()
+ template_ins_modes.unshift ins_mode_in_table
+ ins_mode = ins_mode_in_table
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'col'
+ template_ins_modes.shift()
+ template_ins_modes.unshift ins_mode_in_column_group
+ ins_mode = ins_mode_in_column_group
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'tr'
+ template_ins_modes.shift()
+ template_ins_modes.unshift ins_mode_in_table_body
+ ins_mode = ins_mode_in_table_body
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'td' or t.name is 'th')
+ template_ins_modes.shift()
+ template_ins_modes.unshift ins_mode_in_row
+ ins_mode = ins_mode_in_row
+ process_token t
+ return
+ if t.type is TYPE_START_TAG
+ template_ins_modes.shift()
+ template_ins_modes.unshift ins_mode_in_body
+ ins_mode = ins_mode_in_body
+ process_token t
+ return
+ if t.type is TYPE_END_TAG
+ parse_error()
+ return
+ if t.type is TYPE_EOF
+ unless template_tag_is_open()
+ stop_parsing()
+ return
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'template' # fixfull check namespace
+ break
+ clear_afe_to_marker()
+ template_ins_modes.shift()
+ reset_ins_mode()
+ process_token t
+
+ # 8.2.5.4.19 http://www.w3.org/TR/html5/syntax.html#parsing-main-afterbody
+ ins_mode_after_body = (t) ->
+ if is_space_tok t
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_COMMENT
+ insert_comment t, [open_els[0], open_els[0].children.length]
+ 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_END_TAG and t.name is 'html'
+ # fixfull fragment case
+ ins_mode = ins_mode_after_after_body
+ return
+ if t.type is TYPE_EOF
+ stop_parsing()
+ return
+ # Anything ELse
+ parse_error()
+ ins_mode = ins_mode_in_body
+ process_token t
+
+ # 8.2.5.4.20 http://www.w3.org/TR/html5/syntax.html#parsing-main-inframeset
+ ins_mode_in_frameset = (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 'frameset'
+ insert_html_element t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'frameset'
+ # TODO ?correct for: "if the current node is the root html element"
+ if open_els.length is 1
+ parse_error()
+ return # fragment case
+ open_els.shift()
+ if flag_fragment_parsing is false and open_els[0].name isnt 'frameset'
+ ins_mode = ins_mode_after_frameset
+ return
+ if t.type is TYPE_START_TAG and t.name is 'frame'
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'noframes'
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_EOF
+ # TODO ?correct for: "if the current node is not the root html element"
+ if open_els.length isnt 1
+ parse_error()
+ stop_parsing()
+ return
+ # Anything else
+ parse_error()
+ return
+
+ # 8.2.5.4.21 http://www.w3.org/TR/html5/syntax.html#parsing-main-afterframeset
+ ins_mode_after_frameset = (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_END_TAG and t.name is 'html'
+ insert_mode = ins_mode_after_after_frameset
+ return
+ if t.type is TYPE_START_TAG and t.name is 'noframes'
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_EOF
+ stop_parsing()
+ return
+ # Anything else
+ parse_error()
+ return
+
+ # 8.2.5.4.22 http://www.w3.org/TR/html5/syntax.html#the-after-after-body-insertion-mode
+ ins_mode_after_after_body = (t) ->
+ if t.type is TYPE_COMMENT
+ insert_comment t, [doc, doc.children.length]
+ return
+ if t.type is TYPE_DOCTYPE or is_space_tok(t) or (t.type is TYPE_START_TAG and t.name is 'html')
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_EOF
+ stop_parsing()
+ return
+ # Anything else
+ parse_error()
+ ins_mode = ins_mode_in_body
+ return
+
+ # 8.2.5.4.23 http://www.w3.org/TR/html5/syntax.html#the-after-after-frameset-insertion-mode
+ ins_mode_after_after_frameset = (t) ->
+ if t.type is TYPE_COMMENT
+ insert_comment t, [doc, doc.children.length]
+ return
+ if t.type is TYPE_DOCTYPE or is_space_tok(t) or (t.type is TYPE_START_TAG and t.name is 'html')
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_EOF
+ stop_parsing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'noframes'
+ ins_mode_in_head t
+ return
+ # Anything else
+ parse_error()
+ return
+
+ # 8.2.5.5 http://www.w3.org/TR/html5/syntax.html#parsing-main-inforeign
+ has_color_face_or_size = (t) ->
+ for a in t.attrs_a
+ if a[0] is 'color' or a[0] is 'face' or a[0] is 'size'
+ return true
+ return false
+ in_foreign_content_end_script = ->
+ open_els.shift()
+ # fixfull
+ return
+ in_foreign_content_other_start = (t) ->
+ acn = adjusted_current_node()
+ if acn.namespace is NS_MATHML
+ adjust_mathml_attributes t
+ if acn.namespace is NS_SVG and svg_name_fixes[t.name]?
+ t.name = svg_name_fixes[t.name]
+ if acn.namespace is NS_SVG
+ adjust_svg_attributes t
+ adjust_foreign_attributes t
+ insert_foreign_element t, acn.namespace
+ if t.flag 'self-closing'
+ if t.name is 'script'
+ t.acknowledge_self_closing()
+ in_foreign_content_end_script()
+ else
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ in_foreign_content = (t) ->
+ if t.type is TYPE_TEXT and t.text is "\u0000"
+ parse_error()
+ insert_character new_character_token "\ufffd"
+ return
+ if is_space_tok t
+ insert_character t
+ return
+ if t.type is TYPE_TEXT
+ flag_frameset_ok = false
+ 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 'b' or t.name is 'big' or t.name is 'blockquote' or t.name is 'body' or t.name is 'br' or t.name is 'center' or t.name is 'code' or t.name is 'dd' or t.name is 'div' or t.name is 'dl' or t.name is 'dt' or t.name is 'em' or t.name is 'embed' or t.name is 'h1' or t.name is 'h2' or t.name is 'h3' or t.name is 'h4' or t.name is 'h5' or t.name is 'h6' or t.name is 'head' or t.name is 'hr' or t.name is 'i' or t.name is 'img' or t.name is 'li' or t.name is 'listing' or t.name is 'main' or t.name is 'meta' or t.name is 'nobr' or t.name is 'ol' or t.name is 'p' or t.name is 'pre' or t.name is 'ruby' or t.name is 's' or t.name is 'small' or t.name is 'span' or t.name is 'strong' or t.name is 'strike' or t.name is 'sub' or t.name is 'sup' or t.name is 'table' or t.name is 'tt' or t.name is 'u' or t.name is 'ul' or t.name is 'var' or (t.name is 'font' and has_color_face_or_size(t)))
+ parse_error()
+ if flag_fragment_parsing
+ in_foreign_content_other_start t
+ return
+ loop # is this safe?
+ open_els.shift()
+ cn = open_els[0]
+ if is_mathml_text_integration_point(cn) or is_html_integration(cn) or cn.namespace is NS_HTML
+ break
+ process_token t
+ return
+ if t.type is TYPE_START_TAG
+ in_foreign_content_other_start t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'script' and open_els[0].name is 'script' and open_els[0].namespace is NS_SVG
+ in_foreign_content_end_script()
+ return
+ if t.type is TYPE_END_TAG
+ if open_els[0].name.toLowerCase() isnt t.name
+ parse_error()
+ for node in open_els
+ if node is open_els[open_els.length - 1]
+ return
+ if node.name.toLowerCase() is t.name
+ loop
+ el = open_els.shift()
+ if el is node
+ return
+ if node.namespace is NS_HTML
+ break
+ ins_mode t # explicitly call HTML insertion mode
+
+
+ # 8.2.4.1 http://www.w3.org/TR/html5/syntax.html#data-state
+ tok_state_data = ->
+ switch c = txt.charAt(cur++)
+ when '&'
+ return new_text_node parse_character_reference()
+ when '<'
+ tok_state = tok_state_tag_open
+ when "\u0000"
+ parse_error()
+ return new_text_node c
+ when '' # EOF
+ return new_eof_token()
+ else
+ return new_text_node c
+ return null
+
+ # 8.2.4.2 http://www.w3.org/TR/html5/syntax.html#character-reference-in-data-state
+ # not needed: tok_state_character_reference_in_data = ->
+ # just call parse_character_reference()
+
+ # 8.2.4.3 http://www.w3.org/TR/html5/syntax.html#rcdata-state
+ tok_state_rcdata = ->
+ switch c = txt.charAt(cur++)
+ when '&'
+ return new_text_node parse_character_reference()
+ when '<'
+ tok_state = tok_state_rcdata_less_than_sign
+ when "\u0000"
+ parse_error()
+ return new_character_token "\ufffd"
+ when '' # EOF
+ return new_eof_token()
+ else
+ return new_character_token c
+ return null
+
+ # 8.2.4.4 http://www.w3.org/TR/html5/syntax.html#character-reference-in-rcdata-state
+ # not needed: tok_state_character_reference_in_rcdata = ->
+ # just call parse_character_reference()
+
+ # 8.2.4.5 http://www.w3.org/TR/html5/syntax.html#rawtext-state
+ tok_state_rawtext = ->
+ switch c = txt.charAt(cur++)
+ when '<'
+ tok_state = tok_state_rawtext_less_than_sign
+ when "\u0000"
+ parse_error()
+ return new_character_token "\ufffd"
+ when '' # EOF
+ return new_eof_token()
+ else
+ return new_character_token c
+ return null
+
+ # 8.2.4.6 http://www.w3.org/TR/html5/syntax.html#script-data-state
+ tok_state_script_data = ->
+ switch c = txt.charAt(cur++)
+ when '<'
+ tok_state = tok_state_script_data_less_than_sign
+ when "\u0000"
+ parse_error()
+ return new_character_token "\ufffd"
+ when '' # EOF
+ return new_eof_token()
+ else
+ return new_character_token c
+ return null
+
+ # 8.2.4.7 http://www.w3.org/TR/html5/syntax.html#plaintext-state
+ tok_state_plaintext = ->
+ switch c = txt.charAt(cur++)
+ when "\u0000"
+ parse_error()
+ return new_character_token "\ufffd"
+ when '' # EOF
+ return new_eof_token()
+ else
+ return new_character_token c
+ return null
+
+
+ # 8.2.4.8 http://www.w3.org/TR/html5/syntax.html#tag-open-state
+ tok_state_tag_open = ->
+ switch c = txt.charAt(cur++)
+ when '!'
+ tok_state = tok_state_markup_declaration_open
+ when '/'
+ tok_state = tok_state_end_tag_open
+ when '?'
+ parse_error()
+ tok_cur_tag = new_comment_token '?'
+ tok_state = tok_state_bogus_comment
+ else
+ if is_lc_alpha(c)
+ tok_cur_tag = new_open_tag c
+ tok_state = tok_state_tag_name
+ else if is_uc_alpha(c)
+ tok_cur_tag = new_open_tag c.toLowerCase()
+ tok_state = tok_state_tag_name
+ else
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # we didn't parse/handle the char after <
+ return new_text_node '<'
+ return null
+
+ # 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
+ return null
+
+ # 8.2.4.10 http://www.w3.org/TR/html5/syntax.html#tag-name-state
+ tok_state_tag_name = ->
+ switch c = txt.charAt(cur++)
+ when "\t", "\n", "\u000c", ' '
+ tok_state = tok_state_before_attribute_name
+ when '/'
+ tok_state = tok_state_self_closing_start_tag
+ when '>'
+ tok_state = tok_state_data
+ tmp = tok_cur_tag
+ tok_cur_tag = null
+ return tmp
+ when "\u0000"
+ parse_error()
+ tok_cur_tag.name += "\ufffd"
+ when '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ else
+ if is_uc_alpha(c)
+ tok_cur_tag.name += c.toLowerCase()
+ else
+ tok_cur_tag.name += c
+ return null
+
+ # 8.2.4.11 http://www.w3.org/TR/html5/syntax.html#rcdata-less-than-sign-state
+ tok_state_rcdata_less_than_sign = ->
+ c = txt.charAt(cur++)
+ if c is '/'
+ temporary_buffer = ''
+ tok_state = tok_state_rcdata_end_tag_open
+ return null
+ # Anything else
+ tok_state = tok_state_rcdata
+ cur -= 1 # reconsume the input character
+ return new_character_token '<'
+
+ # 8.2.4.12 http://www.w3.org/TR/html5/syntax.html#rcdata-end-tag-open-state
+ tok_state_rcdata_end_tag_open = ->
+ c = txt.charAt(cur++)
+ if is_uc_alpha(c)
+ tok_cur_tag = new_end_tag c.toLowerCase()
+ temporary_buffer += c
+ tok_state = tok_state_rcdata_end_tag_name
+ return null
+ if is_lc_alpha(c)
+ tok_cur_tag = new_end_tag c
+ temporary_buffer += c
+ tok_state = tok_state_rcdata_end_tag_name
+ return null
+ # Anything else
+ tok_state = tok_state_rcdata
+ cur -= 1 # reconsume the input character
+ return new_character_token "</" # fixfull separate these
+
+ # http://www.w3.org/TR/html5/syntax.html#appropriate-end-tag-token
+ is_appropriate_end_tag = (t) ->
+ # spec says to check against "the tag name of the last start tag to
+ # have been emitted from this tokenizer", but this is only called from
+ # the various "raw" states, which I'm pretty sure all push the start
+ # token onto open_els. TODO: verify this after the script data states
+ # are implemented
+ debug_log "#{t.type}, #{t.name} open_els: #{serialize_els open_els, true, true}"
+ return t.type is TYPE_END_TAG and t.name is open_els[0].name
+
+ # 8.2.4.13 http://www.w3.org/TR/html5/syntax.html#rcdata-end-tag-name-state
+ tok_state_rcdata_end_tag_name = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_before_attribute_name
+ return
+ # else fall through to "Anything else"
+ if c is '/'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_self_closing_start_tag # FIXME spec typo?
+ return
+ # else fall through to "Anything else"
+ if c is '>'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_data
+ return tok_cur_tag
+ # else fall through to "Anything else"
+ if is_uc_alpha(c)
+ tok_cur_tag.name += c.toLowerCase()
+ temporary_buffer += c
+ return null
+ if is_lc_alpha(c)
+ tok_cur_tag.name += c
+ temporary_buffer += c
+ return null
+ # Anything else
+ tok_state = tok_state_rcdata
+ cur -= 1 # reconsume the input character
+ return new_character_token '</' + temporary_buffer # fixfull separate these
+
+ # 8.2.4.14 http://www.w3.org/TR/html5/syntax.html#rawtext-less-than-sign-state
+ tok_state_rawtext_less_than_sign = ->
+ c = txt.charAt(cur++)
+ if c is '/'
+ temporary_buffer = ''
+ tok_state = tok_state_rawtext_end_tag_open
+ return null
+ # Anything else
+ tok_state = tok_state_rawtext
+ cur -= 1 # reconsume the input character
+ return new_character_token '<'
+
+ # 8.2.4.15 http://www.w3.org/TR/html5/syntax.html#rawtext-end-tag-open-state
+ tok_state_rawtext_end_tag_open = ->
+ c = txt.charAt(cur++)
+ if is_uc_alpha(c)
+ tok_cur_tag = new_end_tag c.toLowerCase()
+ temporary_buffer += c
+ tok_state = tok_state_rawtext_end_tag_name
+ return null
+ if is_lc_alpha(c)
+ tok_cur_tag = new_end_tag c
+ temporary_buffer += c
+ tok_state = tok_state_rawtext_end_tag_name
+ return null
+ # Anything else
+ tok_state = tok_state_rawtext
+ cur -= 1 # reconsume the input character
+ return new_character_token "</" # fixfull separate these
+
+ # 8.2.4.16 http://www.w3.org/TR/html5/syntax.html#rawtext-end-tag-name-state
+ tok_state_rawtext_end_tag_name = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_before_attribute_name
+ return
+ # else fall through to "Anything else"
+ if c is '/'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_self_closing_start_tag
+ return
+ # else fall through to "Anything else"
+ if c is '>'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_data
+ return tok_cur_tag
+ # else fall through to "Anything else"
+ if is_uc_alpha(c)
+ tok_cur_tag.name += c.toLowerCase()
+ temporary_buffer += c
+ return null
+ if is_lc_alpha(c)
+ tok_cur_tag.name += c
+ temporary_buffer += c
+ return null
+ # Anything else
+ tok_state = tok_state_rawtext
+ cur -= 1 # reconsume the input character
+ return new_character_token '</' + temporary_buffer # fixfull separate these
+
+ # 8.2.4.17 http://www.w3.org/TR/html5/syntax.html#script-data-less-than-sign-state
+ tok_state_script_data_less_than_sign = ->
+ c = txt.charAt(cur++)
+ if c is '/'
+ temporary_buffer = ''
+ tok_state = tok_state_script_data_end_tag_open
+ return
+ if c is '!'
+ tok_state = tok_state_script_data_escape_start
+ return new_character_token '<!' # fixfull split
+ # Anything else
+ tok_state = tok_state_script_data
+ cur -= 1 # Reconsume
+ return new_character_token '<'
+
+ # 8.2.4.18 http://www.w3.org/TR/html5/syntax.html#script-data-end-tag-open-state
+ tok_state_script_data_end_tag_open = ->
+ c = txt.charAt(cur++)
+ if is_uc_alpha(c)
+ tok_cur_tag = new_end_tag c.toLowerCase()
+ temporary_buffer += c
+ tok_state = tok_state_script_data_end_tag_name
+ return
+ if is_lc_alpha(c)
+ tok_cur_tag = new_end_tag c
+ temporary_buffer += c
+ tok_state = tok_state_script_data_end_tag_name
+ return
+ # Anything else
+ tok_state = tok_state_script_data
+ cur -= 1 # Reconsume
+ return new_character_token '</'
+
+ # 8.2.4.19 http://www.w3.org/TR/html5/syntax.html#script-data-end-tag-open-state
+ tok_state_script_data_end_tag_name = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_before_attribute_name
+ return
+ # fall through
+ if c is '/'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_self_closing_start_tag
+ return
+ # fall through
+ if is_uc_alpha(c)
+ tok_cur_tag.name += c.toLowerCase()
+ temporary_buffer += c
+ return
+ if is_lc_alpha(c)
+ tok_cur_tag.name += c
+ temporary_buffer += c
+ return
+ # Anything else
+ tok_state = tok_state_script_data
+ cur -= 1 # Reconsume
+ return new_character_token "</#{temporary_buffer}" # fixfull split
+
+ # 8.2.4.20 http://www.w3.org/TR/html5/syntax.html#script-data-escape-start-state
+ tok_state_script_data_escape_start = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ tok_state = tok_state_script_data_escape_start_dash
+ return new_character_token '-'
+ # Anything else
+ tok_state = tok_state_script_data
+ cur -= 1 # Reconsume
+ return
+
+ # 8.2.4.21 http://www.w3.org/TR/html5/syntax.html#script-data-escape-start-dash-state
+ tok_state_script_data_escape_start_dash = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ tok_state = tok_state_script_data_escaped_dash_dash
+ return new_character_token '-'
+ # Anything else
+ tok_state = tok_state_script_data
+ cur -= 1 # Reconsume
+ return
+
+ # 8.2.4.22 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-state
+ tok_state_script_data_escaped = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ tok_state = tok_state_script_data_escaped_dash
+ return new_character_token '-'
+ if c is '<'
+ tok_state = tok_state_script_data_escaped_less_than_sign
+ return
+ if c is "\u0000"
+ parse_error()
+ return new_character_token "\ufffd"
+ if c is '' # EOF
+ tok_state = tok_state_data
+ parse_error()
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ return new_character_token c
+
+ # 8.2.4.23 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-dash-state
+ tok_state_script_data_escaped_dash = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ tok_state = tok_state_script_data_escaped_dash_dash
+ return new_character_token '-'
+ if c is '<'
+ tok_state = tok_state_script_data_escaped_less_than_sign
+ return
+ if c is "\u0000"
+ parse_error()
+ tok_state = tok_state_script_data_escaped
+ return new_character_token "\ufffd"
+ if c is '' # EOF
+ tok_state = tok_state_data
+ parse_error()
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ tok_state = tok_state_script_data_escaped
+ return new_character_token c
+
+ # 8.2.4.24 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-dash-dash-state
+ tok_state_script_data_escaped_dash_dash = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ return new_character_token '-'
+ if c is '<'
+ tok_state = tok_state_script_data_escaped_less_than_sign
+ return
+ if c is '>'
+ tok_state = tok_state_script_data
+ return new_character_token '>'
+ if c is "\u0000"
+ parse_error()
+ tok_state = tok_state_script_data_escaped
+ return new_character_token "\ufffd"
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ tok_state = tok_state_script_data_escaped
+ return new_character_token c
+
+ # 8.2.4.25 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-less-than-sign-state
+ tok_state_script_data_escaped_less_than_sign = ->
+ c = txt.charAt(cur++)
+ if c is '/'
+ temporary_buffer = ''
+ tok_state = tok_state_script_data_escaped_end_tag_open
+ return
+ if is_uc_alpha(c)
+ temporary_buffer = c.toLowerCase() # yes, really
+ tok_state = tok_state_script_data_double_escape_start
+ return new_character_token "<#{c}" # fixfull split
+ if is_lc_alpha(c)
+ temporary_buffer = c
+ tok_state = tok_state_script_data_double_escape_start
+ return new_character_token "<#{c}" # fixfull split
+ # Anything else
+ tok_state = tok_state_script_data_escaped
+ cur -= 1 # Reconsume
+ return new_character_token c
+
+ # 8.2.4.26 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-end-tag-open-state
+ tok_state_script_data_escaped_end_tag_open = ->
+ c = txt.charAt(cur++)
+ if is_uc_alpha(c)
+ tok_cur_tag = new_end_tag c.toLowerCase()
+ temporary_buffer += c
+ tok_state = tok_state_script_data_escaped_end_tag_name
+ return
+ if is_lc_alpha(c)
+ tok_cur_tag = new_end_tag c
+ temporary_buffer += c
+ tok_state = tok_state_script_data_escaped_end_tag_name
+ return
+ # Anything else
+ tok_state = tok_state_script_data_escaped
+ cur -= 1 # Reconsume
+ return new_character_token '</' # fixfull split
+
+ # 8.2.4.27 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-end-tag-name-state
+ tok_state_script_data_escaped_end_tag_name = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_before_attribute_name
+ return
+ # fall through
+ if c is '/'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_self_closing_start_tag
+ return
+ # fall through
+ if is_uc_alpha(c)
+ tok_cur_tag.name += c.toLowerCase()
+ temporary_buffer += c.toLowerCase()
+ return
+ if is_lc_alpha(c)
+ tok_cur_tag.name += c
+ temporary_buffer += c.toLowerCase()
+ return
+ # Anything else
+ tok_state = tok_state_script_data_escaped
+ cur -= 1 # Reconsume
+ return new_character_token "</#{temporary_buffer}" # fixfull split
+
+ # 8.2.4.28 http://www.w3.org/TR/html5/syntax.html#script-data-double-escape-start-state
+ tok_state_script_data_double_escape_start = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' ' or c is '/' or c is '>'
+ if temporary_buffer is 'script'
+ tok_state = tok_state_script_data_double_escaped
+ else
+ tok_state = tok_state_script_data_escaped
+ return new_character_token c
+ if is_uc_alpha(c)
+ temporary_buffer += c.toLowerCase() # yes, really lowercase
+ return new_character_token c
+ if is_lc_alpha(c)
+ temporary_buffer += c
+ return new_character_token c
+ # Anything else
+ tok_state = tok_state_script_data_escaped
+ cur -= 1 # Reconsume
+ return
+
+ # 8.2.4.29 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-state
+ tok_state_script_data_double_escaped = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ tok_state = tok_state_script_data_double_escaped_dash
+ return new_character_token '-'
+ if c is '<'
+ tok_state = tok_state_script_data_double_escaped_less_than_sign
+ return new_character_token '<'
+ if c is "\u0000"
+ parse_error()
+ return new_character_token "\ufffd"
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ return new_character_token c
+
+ # 8.2.4.30 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-dash-state
+ tok_state_script_data_double_escaped_dash = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ tok_state = tok_state_script_data_double_escaped_dash_dash
+ return new_character_token '-'
+ if c is '<'
+ tok_state = tok_state_script_data_double_escaped_less_than_sign
+ return new_character_token '<'
+ if c is "\u0000"
+ parse_error()
+ tok_state = tok_state_script_data_double_escaped
+ return new_character_token "\ufffd"
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ tok_state = tok_state_script_data_double_escaped
+ return new_character_token c
+
+ # 8.2.4.31 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-dash-dash-state
+ tok_state_script_data_double_escaped_dash_dash = ->
+ c = txt.charAt(cur++)
+ if c is '-'
+ return new_character_token '-'
+ if c is '<'
+ tok_state = tok_state_script_data_double_escaped_less_than_sign
+ return new_character_token '<'
+ if c is '>'
+ tok_state = tok_state_script_data
+ return new_character_token '>'
+ if c is "\u0000"
+ parse_error()
+ tok_state = tok_state_script_data_double_escaped
+ return new_character_token "\ufffd"
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ tok_state = tok_state_script_data_double_escaped
+ return new_character_token c
+
+ # 8.2.4.32 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-less-than-sign-state
+ tok_state_script_data_double_escaped_less_than_sign = ->
+ c = txt.charAt(cur++)
+ if c is '/'
+ temporary_buffer = ''
+ tok_state = tok_state_script_data_double_escape_end
+ return new_character_token '/'
+ # Anything else
+ tok_state = tok_state_script_data_double_escaped
+ cur -= 1 # Reconsume
+ return
+
+ # 8.2.4.33 http://www.w3.org/TR/html5/syntax.html#script-data-double-escape-end-state
+ tok_state_script_data_double_escape_end = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' ' or c is '/' or c is '>'
+ if temporary_buffer is 'script'
+ tok_state = tok_state_script_data_escaped
+ else
+ tok_state = tok_state_script_data_double_escaped
+ return new_character_token c
+ if is_uc_alpha(c)
+ temporary_buffer += c.toLowerCase() # yes, really lowercase
+ return new_character_token c
+ if is_lc_alpha(c)
+ temporary_buffer += c
+ return new_character_token c
+ # Anything else
+ tok_state = tok_state_script_data_double_escaped
+ cur -= 1 # Reconsume
+ return
+
+ # 8.2.4.34 http://www.w3.org/TR/html5/syntax.html#before-attribute-name-state
+ tok_state_before_attribute_name = ->
+ attr_name = null
+ switch c = txt.charAt(cur++)
+ when "\t", "\n", "\u000c", ' '
+ return null
+ when '/'
+ tok_state = tok_state_self_closing_start_tag
+ return null
+ when '>'
+ tok_state = tok_state_data
+ tmp = tok_cur_tag
+ tok_cur_tag = null
+ return tmp
+ when "\u0000"
+ parse_error()
+ attr_name = "\ufffd"
+ when '"', "'", '<', '='
+ parse_error()
+ attr_name = c
+ when '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ else
+ if is_uc_alpha(c)
+ attr_name = c.toLowerCase()
+ else
+ attr_name = c
+ if attr_name?
+ tok_cur_tag.attrs_a.unshift [attr_name, '']
+ tok_state = tok_state_attribute_name
+ return null
+
+ # 8.2.4.35 http://www.w3.org/TR/html5/syntax.html#attribute-name-state
+ tok_state_attribute_name = ->
+ switch c = txt.charAt(cur++)
+ when "\t", "\n", "\u000c", ' '
+ tok_state = tok_state_after_attribute_name
+ when '/'
+ tok_state = tok_state_self_closing_start_tag
+ when '='
+ tok_state = tok_state_before_attribute_value
+ when '>'
+ tok_state = tok_state_data
+ tmp = tok_cur_tag
+ tok_cur_tag = null
+ return tmp
+ when "\u0000"
+ parse_error()
+ tok_cur_tag.attrs_a[0][0] = "\ufffd"
+ when '"', "'", '<'
+ parse_error()
+ tok_cur_tag.attrs_a[0][0] = c
+ when '' # EOF
parse_error()
+ tok_state = tok_state_data
+ else
+ if is_uc_alpha(c)
+ tok_cur_tag.attrs_a[0][0] = c.toLowerCase()
+ else
+ tok_cur_tag.attrs_a[0][0] += c
+ return null
+
+ # 8.2.4.36 http://www.w3.org/TR/html5/syntax.html#after-attribute-name-state
+ tok_state_after_attribute_name = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
return
- if t.type is TYPE_START_TAG and t.name is 'select'
- parse_error()
- loop
- el = open_els.shift()
- if el.name is 'select'
- break
- reset_insertion_mode()
- # spec says that this is the same as </select> but it doesn't say
- # to check scope first
+ if c is '/'
+ tok_state = tok_state_self_closing_start_tag
return
- if t.type is TYPE_START_TAG and (t.name is 'input' or t.name is 'keygen' or t.name is 'textarea')
- parse_error()
- if is_in_select_scope 'select'
- return
- loop
- el = open_els.shift()
- if el.name is 'select'
- break
- reset_insertion_mode()
- insertion_mode t
+ if c is '='
+ tok_state = tok_state_before_attribute_value
return
- if t.type is TYPE_START_TAG and (t.name is 'script' or t.name is 'template')
- ins_mode_in_head t
+ if c is '>'
+ tok_state = tok_state_data
return
- if t.type is TYPE_EOF
- ins_mode_in_body t
+ if is_uc_alpha(c)
+ tok_cur_tag.attrs_a.unshift [c.toLowerCase(), '']
+ tok_state = tok_state_attribute_name
return
- # Anything else
- parse_error()
- return
-
- # 8.2.5.4.17 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselectintable
- ins_mode_in_select_in_table = (t) ->
- if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th')
+ if c is "\u0000"
parse_error()
- loop
- el = open_els.shift()
- if el.name is 'select'
- break
- reset_insertion_mode()
- insertion_mode t
+ tok_cur_tag.attrs_a.unshift ["\ufffd", '']
+ tok_state = tok_state_attribute_name
return
- if t.type is TYPE_END_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th')
+ if c is '' # EOF
parse_error()
- unless is_in_table_scope t.name, NS_HTML
- return
- loop
- el = open_els.shift()
- if el.name is 'select'
- break
- reset_insertion_mode()
- insertion_mode t
+ tok_state = tok_state_data
+ cur -= 1 # reconsume
return
+ if c is '"' or c is "'" or c is '<'
+ parse_error()
+ # fall through to Anything else
# Anything else
- ins_mode_in_select t
- return
-
- # CONTINUE more insertion modes!
-
-
-
-
-
-
-
-
-
-
-
+ tok_cur_tag.attrs_a.unshift [c, '']
+ tok_state = tok_state_attribute_name
- # 8.2.4.1 http://www.w3.org/TR/html5/syntax.html#data-state
- tok_state_data = ->
+ # 8.2.4.37 http://www.w3.org/TR/html5/syntax.html#before-attribute-value-state
+ tok_state_before_attribute_value = ->
switch c = txt.charAt(cur++)
+ when "\t", "\n", "\u000c", ' '
+ return null
+ when '"'
+ tok_state = tok_state_attribute_value_double_quoted
when '&'
- return new_text_node parse_character_reference()
- when '<'
- tok_state = tok_state_tag_open
+ tok_state = tok_state_attribute_value_unquoted
+ cur -= 1
+ when "'"
+ tok_state = tok_state_attribute_value_single_quoted
when "\u0000"
+ # Parse error
+ tok_cur_tag.attrs_a[0][1] += "\ufffd"
+ tok_state = tok_state_attribute_value_unquoted
+ when '>'
+ # Parse error
+ tok_state = tok_state_data
+ tmp = tok_cur_tag
+ tok_cur_tag = null
+ return tmp
+ when '' # EOF
parse_error()
- return new_text_node c
+ tok_state = tok_state_data
+ else
+ tok_cur_tag.attrs_a[0][1] += c
+ tok_state = tok_state_attribute_value_unquoted
+ return null
+
+ # 8.2.4.38 http://www.w3.org/TR/html5/syntax.html#attribute-value-(double-quoted)-state
+ tok_state_attribute_value_double_quoted = ->
+ switch c = txt.charAt(cur++)
+ when '"'
+ tok_state = tok_state_after_attribute_value_quoted
+ when '&'
+ tok_cur_tag.attrs_a[0][1] += parse_character_reference '"', true
+ when "\u0000"
+ # Parse error
+ tok_cur_tag.attrs_a[0][1] += "\ufffd"
when '' # EOF
- return new_eof_token()
+ parse_error()
+ tok_state = tok_state_data
else
- return new_text_node c
+ tok_cur_tag.attrs_a[0][1] += c
return null
- # 8.2.4.2 http://www.w3.org/TR/html5/syntax.html#character-reference-in-data-state
- # not needed: tok_state_character_reference_in_data = ->
- # just call parse_character_reference()
+ # 8.2.4.39 http://www.w3.org/TR/html5/syntax.html#attribute-value-(single-quoted)-state
+ tok_state_attribute_value_single_quoted = ->
+ switch c = txt.charAt(cur++)
+ when "'"
+ tok_state = tok_state_after_attribute_value_quoted
+ when '&'
+ tok_cur_tag.attrs_a[0][1] += parse_character_reference "'", true
+ when "\u0000"
+ # Parse error
+ tok_cur_tag.attrs_a[0][1] += "\ufffd"
+ when '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ else
+ tok_cur_tag.attrs_a[0][1] += c
+ return null
- # 8.2.4.3 http://www.w3.org/TR/html5/syntax.html#rcdata-state
- tok_state_rcdata = ->
+ # 8.2.4.40 http://www.w3.org/TR/html5/syntax.html#attribute-value-(unquoted)-state
+ tok_state_attribute_value_unquoted = ->
switch c = txt.charAt(cur++)
+ when "\t", "\n", "\u000c", ' '
+ tok_state = tok_state_before_attribute_name
when '&'
- return new_text_node parse_character_reference()
- when '<'
- tok_state = tok_state_rcdata_less_than_sign
+ tok_cur_tag.attrs_a[0][1] += parse_character_reference '>', true
+ when '>'
+ tok_state = tok_state_data
+ tmp = tok_cur_tag
+ tok_cur_tag = null
+ return tmp
when "\u0000"
+ tok_cur_tag.attrs_a[0][1] += "\ufffd"
+ when '' # EOF
parse_error()
- return new_character_token "\ufffd"
+ tok_state = tok_state_data
+ else
+ # Parse Error if ', <, = or ` (backtick)
+ tok_cur_tag.attrs_a[0][1] += c
+ return null
+
+ # 8.2.4.42 http://www.w3.org/TR/html5/syntax.html#after-attribute-value-(quoted)-state
+ tok_state_after_attribute_value_quoted = ->
+ switch c = txt.charAt(cur++)
+ when "\t", "\n", "\u000c", ' '
+ tok_state = tok_state_before_attribute_name
+ when '/'
+ tok_state = tok_state_self_closing_start_tag
+ when '>'
+ tok_state = tok_state_data
+ tmp = tok_cur_tag
+ tok_cur_tag = null
+ return tmp
when '' # EOF
- return new_eof_token()
+ parse_error()
+ tok_state = tok_state_data
else
- return new_character_token c
+ # Parse Error
+ tok_state = tok_state_before_attribute_name
+ cur -= 1 # we didn't handle that char
return null
- # 8.2.4.4 http://www.w3.org/TR/html5/syntax.html#character-reference-in-rcdata-state
- # not needed: tok_state_character_reference_in_rcdata = ->
- # just call parse_character_reference()
+ # 8.2.4.43 http://www.w3.org/TR/html5/syntax.html#self-closing-start-tag-state
+ tok_state_self_closing_start_tag = ->
+ c = txt.charAt(cur++)
+ if c is '>'
+ tok_cur_tag.flag 'self-closing'
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is ''
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return
+ # Anything else
+ parse_error()
+ tok_state = tok_state_before_attribute_name
+ cur -= 1 # Reconsume
+ return
- # 8.2.4.5 http://www.w3.org/TR/html5/syntax.html#rawtext-state
- tok_state_rawtext = ->
+ # 8.2.4.44 http://www.w3.org/TR/html5/syntax.html#bogus-comment-state
+ # WARNING: put a comment token in tok_cur_tag before setting this state
+ tok_state_bogus_comment = ->
+ next_gt = txt.indexOf '>', cur
+ if next_gt is -1
+ val = txt.substr cur
+ cur = txt.length
+ else
+ val = txt.substr cur, (next_gt - cur)
+ cur = next_gt + 1
+ val = val.replace "\u0000", "\ufffd"
+ tok_cur_tag.text += val
+ tok_state = tok_state_data
+ return tok_cur_tag
+
+ # 8.2.4.45 http://www.w3.org/TR/html5/syntax.html#markup-declaration-open-state
+ tok_state_markup_declaration_open = ->
+ if txt.substr(cur, 2) is '--'
+ cur += 2
+ tok_cur_tag = new_comment_token ''
+ tok_state = tok_state_comment_start
+ return
+ if txt.substr(cur, 7).toLowerCase() is 'doctype'
+ cur += 7
+ tok_state = tok_state_doctype
+ return
+ acn = adjusted_current_node()
+ if acn and acn.namespace isnt NS_HTML and txt.substr(cur, 7) is '[CDATA['
+ cur += 7
+ tok_state = tok_state_cdata_section
+ return
+ # Otherwise
+ parse_error()
+ tok_cur_tag = new_comment_token '!' # TODO test ("!" right?)
+ tok_state = tok_state_bogus_comment
+ return
+
+ # 8.2.4.46 http://www.w3.org/TR/html5/syntax.html#comment-start-state
+ tok_state_comment_start = ->
switch c = txt.charAt(cur++)
- when '<'
- tok_state = tok_state_rawtext_less_than_sign
+ when '-'
+ tok_state = tok_state_comment_start_dash
when "\u0000"
parse_error()
return new_character_token "\ufffd"
+ when '>'
+ parse_error()
+ tok_state = tok_state_data
+ return tok_cur_tag
when '' # EOF
- return new_eof_token()
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return tok_cur_tag
else
- return new_character_token c
+ tok_cur_tag.text += c
return null
- # 8.2.4.6 http://www.w3.org/TR/html5/syntax.html#script-data-state
- tok_state_script_data = ->
+ # 8.2.4.47 http://www.w3.org/TR/html5/syntax.html#comment-start-dash-state
+ tok_state_comment_start_dash = ->
switch c = txt.charAt(cur++)
- when '<'
- tok_state = tok_state_script_data_less_than_sign
+ when '-'
+ tok_state = tok_state_comment_end
when "\u0000"
parse_error()
- return new_character_token "\ufffd"
+ tok_cur_tag.text += "-\ufffd"
+ tok_state = tok_state_comment
+ when '>'
+ parse_error()
+ tok_state = tok_state_data
+ return tok_cur_tag
when '' # EOF
- return new_eof_token()
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return tok_cur_tag
else
- return new_character_token c
+ tok_cur_tag.text += "-#{c}"
+ tok_state = tok_state_comment
return null
- # 8.2.4.7 http://www.w3.org/TR/html5/syntax.html#plaintext-state
- tok_state_plaintext = ->
+ # 8.2.4.48 http://www.w3.org/TR/html5/syntax.html#comment-state
+ tok_state_comment = ->
switch c = txt.charAt(cur++)
+ when '-'
+ tok_state = tok_state_comment_end_dash
when "\u0000"
parse_error()
- return new_character_token "\ufffd"
+ tok_cur_tag.text += "\ufffd"
when '' # EOF
- return new_eof_token()
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return tok_cur_tag
else
- return new_character_token c
+ tok_cur_tag.text += c
return null
-
- # 8.2.4.8 http://www.w3.org/TR/html5/syntax.html#tag-open-state
- tok_state_tag_open = ->
+ # 8.2.4.49 http://www.w3.org/TR/html5/syntax.html#comment-end-dash-state
+ tok_state_comment_end_dash = ->
switch c = txt.charAt(cur++)
- when '!'
- tok_state = tok_state_markup_declaration_open
- when '/'
- tok_state = tok_state_end_tag_open
- when '?'
+ when '-'
+ tok_state = tok_state_comment_end
+ when "\u0000"
parse_error()
- tok_state = tok_state_bogus_comment
+ tok_cur_tag.text += "-\ufffd"
+ tok_state = tok_state_comment
+ when '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return tok_cur_tag
else
- if lc_alpha.indexOf(c) > -1
- tok_cur_tag = new_open_tag c
- tok_state = tok_state_tag_name
- else if uc_alpha.indexOf(c) > -1
- tok_cur_tag = new_open_tag c.toLowerCase()
- tok_state = tok_state_tag_name
- else
- parse_error()
- tok_state = tok_state_data
- cur -= 1 # we didn't parse/handle the char after <
- return new_text_node '<'
+ tok_cur_tag.text += "-#{c}"
+ tok_state = tok_state_comment
return null
- # 8.2.4.9 http://www.w3.org/TR/html5/syntax.html#end-tag-open-state
- tok_state_end_tag_open = ->
+ # 8.2.4.50 http://www.w3.org/TR/html5/syntax.html#comment-end-state
+ tok_state_comment_end = ->
switch c = txt.charAt(cur++)
when '>'
- parse_error()
tok_state = tok_state_data
+ return tok_cur_tag
+ when "\u0000"
+ parse_error()
+ tok_cur_tag.text += "--\ufffd"
+ tok_state = tok_state_comment
+ when '!'
+ parse_error()
+ tok_state = tok_state_comment_end_bang
+ when '-'
+ parse_error()
+ tok_cur_tag.text += '-'
when '' # EOF
parse_error()
tok_state = tok_state_data
- return new_text_node '</'
+ cur -= 1 # Reconsume
+ return tok_cur_tag
else
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag = new_end_tag c.toLowerCase()
- tok_state = tok_state_tag_name
- else if lc_alpha.indexOf(c) > -1
- tok_cur_tag = new_end_tag c
- tok_state = tok_state_tag_name
- else
- parse_error()
- tok_state = tok_state_bogus_comment
+ parse_error()
+ tok_cur_tag.text += "--#{c}"
+ tok_state = tok_state_comment
return null
- # 8.2.4.10 http://www.w3.org/TR/html5/syntax.html#tag-name-state
- tok_state_tag_name = ->
+ # 8.2.4.51 http://www.w3.org/TR/html5/syntax.html#comment-end-bang-state
+ tok_state_comment_end_bang = ->
switch c = txt.charAt(cur++)
- when "\t", "\n", "\u000c", ' '
- tok_state = tok_state_before_attribute_name
- when '/'
- tok_state = tok_state_self_closing_start_tag
+ when '-'
+ tok_cur_tag.text += "--!#{c}"
+ tok_state = tok_state_comment_end_dash
when '>'
tok_state = tok_state_data
- tmp = tok_cur_tag
- tok_cur_tag = null
- return tmp
+ return tok_cur_tag
when "\u0000"
parse_error()
- tok_cur_tag.name += "\ufffd"
+ tok_cur_tag.text += "--!\ufffd"
+ tok_state = tok_state_comment
when '' # EOF
parse_error()
tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return tok_cur_tag
else
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag.name += c.toLowerCase()
- else
- tok_cur_tag.name += c
+ tok_cur_tag.text += "--!#{c}"
+ tok_state = tok_state_comment
return null
- # 8.2.4.11 http://www.w3.org/TR/html5/syntax.html#rcdata-less-than-sign-state
- tok_state_rcdata_less_than_sign = ->
- c = txt.charAt(cur++)
- if c is '/'
- temporary_buffer = ''
- tok_state = tok_state_rcdata_end_tag_open
- return null
- # Anything else
- tok_state = tok_state_rcdata
- cur -= 1 # reconsume the input character
- return new_character_token '<'
+ # 8.2.4.52 http://www.w3.org/TR/html5/syntax.html#doctype-state
+ tok_state_doctype = ->
+ switch c = txt.charAt(cur++)
+ when "\t", "\u000a", "\u000c", ' '
+ tok_state = tok_state_before_doctype_name
+ when '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ el = new_doctype_token ''
+ el.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return el
+ else
+ parse_error()
+ tok_state = tok_state_before_doctype_name
+ cur -= 1 # Reconsume
+ return null
- # 8.2.4.12 http://www.w3.org/TR/html5/syntax.html#rcdata-end-tag-open-state
- tok_state_rcdata_end_tag_open = ->
+ # 8.2.4.52 http://www.w3.org/TR/html5/syntax.html#doctype-state
+ tok_state_before_doctype_name = ->
c = txt.charAt(cur++)
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag = new_end_tag c.toLowerCase()
- temporary_buffer += c
- tok_state = tok_state_rcdata_end_tag_name
- return null
- if lc_alpha.indexOf(c) > -1
- tok_cur_tag = new_end_tag c
- temporary_buffer += c
- tok_state = tok_state_rcdata_end_tag_name
- return null
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ return
+ if is_uc_alpha(c)
+ tok_cur_tag = new_doctype_token c.toLowerCase()
+ tok_state = tok_state_doctype_name
+ return
+ if c is "\u0000"
+ parse_error()
+ tok_cur_tag = new_doctype_token "\ufffd"
+ tok_state = tok_state_doctype_name
+ return
+ if c is '>'
+ parse_error()
+ el = new_doctype_token ''
+ el.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return el
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ el = new_doctype_token ''
+ el.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return el
# Anything else
- tok_state = tok_state_rcdata
- cur -= 1 # reconsume the input character
- return new_character_token "</" # fixfull separate these
-
- # http://www.w3.org/TR/html5/syntax.html#appropriate-end-tag-token
- is_appropriate_end_tag = (t) ->
- # spec says to check against "the tag name of the last start tag to
- # have been emitted from this tokenizer", but this is only called from
- # the various "raw" states, which I'm pretty sure all push the start
- # token onto open_els. TODO: verify this after the script data states
- # are implemented
- debug_log "#{t.type}, #{t.name} open_els: #{serialize_els open_els, true, true}"
- return t.type is TYPE_END_TAG and t.name is open_els[0].name
+ tok_cur_tag = new_doctype_token c
+ tok_state = tok_state_doctype_name
+ return null
- # 8.2.4.13 http://www.w3.org/TR/html5/syntax.html#rcdata-end-tag-name-state
- tok_state_rcdata_end_tag_name = ->
+ # 8.2.4.54 http://www.w3.org/TR/html5/syntax.html#doctype-name-state
+ tok_state_doctype_name = ->
c = txt.charAt(cur++)
- if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
- if is_appropriate_end_tag tok_cur_tag
- tok_state = tok_state_before_attribute_name
- return
- # else fall through to "Anything else"
- if c is '/'
- if is_appropriate_end_tag tok_cur_tag
- tok_state = tok_state_self_closing_start_tag # FIXME spec typo?
- return
- # else fall through to "Anything else"
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ tok_state = tok_state_after_doctype_name
+ return
if c is '>'
- if is_appropriate_end_tag tok_cur_tag
- tok_state = tok_state_data
- return tok_cur_tag
- # else fall through to "Anything else"
- if uc_alpha.indexOf(c) > -1
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
- temporary_buffer += c
- return null
- if lc_alpha.indexOf(c) > -1
- tok_cur_tag.name += c
- temporary_buffer += c
- return null
+ return
+ if c is "\u0000"
+ parse_error()
+ tok_cur_tag.name += "\ufffd"
+ return
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
# Anything else
- tok_state = tok_state_rcdata
- cur -= 1 # reconsume the input character
- return new_character_token '</' + temporary_buffer # fixfull separate these
+ tok_cur_tag.name += c
+ return null
- # 8.2.4.14 http://www.w3.org/TR/html5/syntax.html#rawtext-less-than-sign-state
- tok_state_rawtext_less_than_sign = ->
+ # 8.2.4.55 http://www.w3.org/TR/html5/syntax.html#after-doctype-name-state
+ tok_state_after_doctype_name = ->
c = txt.charAt(cur++)
- if c is '/'
- temporary_buffer = ''
- tok_state = tok_state_rawtext_end_tag_open
- return null
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ return
+ if c is '>'
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
# Anything else
- tok_state = tok_state_rawtext
- cur -= 1 # reconsume the input character
- return new_character_token '<'
+ if txt.substr(cur - 1, 6).toLowerCase() is 'public'
+ cur += 5
+ tok_state = tok_state_after_doctype_public_keyword
+ return
+ if txt.substr(cur - 1, 6).toLowerCase() is 'system'
+ cur += 5
+ tok_state = tok_state_after_doctype_system_keyword
+ return
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
+ return null
- # 8.2.4.15 http://www.w3.org/TR/html5/syntax.html#rawtext-end-tag-open-state
- tok_state_rawtext_end_tag_open = ->
+ # 8.2.4.56 http://www.w3.org/TR/html5/syntax.html#after-doctype-public-keyword-state
+ tok_state_after_doctype_public_keyword = ->
c = txt.charAt(cur++)
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag = new_end_tag c.toLowerCase()
- temporary_buffer += c
- tok_state = tok_state_rawtext_end_tag_name
- return null
- if lc_alpha.indexOf(c) > -1
- tok_cur_tag = new_end_tag c
- temporary_buffer += c
- tok_state = tok_state_rawtext_end_tag_name
- return null
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ tok_state = tok_state_before_doctype_public_identifier
+ return
+ if c is '"'
+ parse_error()
+ tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
+ tok_state = tok_state_doctype_public_identifier_double_quoted
+ return
+ if c is "'"
+ parse_error()
+ tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
+ tok_state = tok_state_doctype_public_identifier_single_quoted
+ return
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
# Anything else
- tok_state = tok_state_rawtext
- cur -= 1 # reconsume the input character
- return new_character_token "</" # fixfull separate these
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
+ return null
- # 8.2.4.16 http://www.w3.org/TR/html5/syntax.html#rawtext-end-tag-name-state
- tok_state_rawtext_end_tag_name = ->
+ # 8.2.4.57 http://www.w3.org/TR/html5/syntax.html#before-doctype-public-identifier-state
+ tok_state_before_doctype_public_identifier = ->
c = txt.charAt(cur++)
- if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
- if is_appropriate_end_tag tok_cur_tag
- tok_state = tok_state_before_attribute_name
- return
- # else fall through to "Anything else"
- if c is '/'
- if is_appropriate_end_tag tok_cur_tag
- tok_state = tok_state_self_closing_start_tag
- return
- # else fall through to "Anything else"
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ return
+ if c is '"'
+ parse_error()
+ tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
+ tok_state = tok_state_doctype_public_identifier_double_quoted
+ return
+ if c is "'"
+ parse_error()
+ tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
+ tok_state = tok_state_doctype_public_identifier_single_quoted
+ return
if c is '>'
- if is_appropriate_end_tag tok_cur_tag
- tok_state = tok_state_data
- return tok_cur_tag
- # else fall through to "Anything else"
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag.name += c.toLowerCase()
- temporary_buffer += c
- return null
- if lc_alpha.indexOf(c) > -1
- tok_cur_tag.name += c
- temporary_buffer += c
- return null
- # Anything else
- tok_state = tok_state_rawtext
- cur -= 1 # reconsume the input character
- return new_character_token '</' + temporary_buffer # fixfull separate these
-
- # TODO _all_ of the missing states here (17-33) are for parsing script tags
-
- # 8.2.4.34 http://www.w3.org/TR/html5/syntax.html#before-attribute-name-state
- tok_state_before_attribute_name = ->
- attr_name = null
- switch c = txt.charAt(cur++)
- when "\t", "\n", "\u000c", ' '
- return null
- when '/'
- tok_state = tok_state_self_closing_start_tag
- return null
- when '>'
- tok_state = tok_state_data
- tmp = tok_cur_tag
- tok_cur_tag = null
- return tmp
- when "\u0000"
- parse_error()
- attr_name = "\ufffd"
- when '"', "'", '<', '='
- parse_error()
- attr_name = c
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- if uc_alpha.indexOf(c) > -1
- attr_name = c.toLowerCase()
- else
- attr_name = c
- if attr_name?
- tok_cur_tag.attrs_a.unshift [attr_name, '']
- tok_state = tok_state_attribute_name
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
return null
- # 8.2.4.35 http://www.w3.org/TR/html5/syntax.html#attribute-name-state
- tok_state_attribute_name = ->
- switch c = txt.charAt(cur++)
- when "\t", "\n", "\u000c", ' '
- tok_state = tok_state_after_attribute_name
- when '/'
- tok_state = tok_state_self_closing_start_tag
- when '='
- tok_state = tok_state_before_attribute_value
- when '>'
- tok_state = tok_state_data
- tmp = tok_cur_tag
- tok_cur_tag = null
- return tmp
- when "\u0000"
- parse_error()
- tok_cur_tag.attrs_a[0][0] = "\ufffd"
- when '"', "'", '<'
- parse_error()
- tok_cur_tag.attrs_a[0][0] = c
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag.attrs_a[0][0] = c.toLowerCase()
- else
- tok_cur_tag.attrs_a[0][0] += c
+
+ # 8.2.4.58 http://www.w3.org/TR/html5/syntax.html#doctype-public-identifier-(double-quoted)-state
+ tok_state_doctype_public_identifier_double_quoted = ->
+ c = txt.charAt(cur++)
+ if c is '"'
+ tok_state = tok_state_after_doctype_public_identifier
+ return
+ if c is "\u0000"
+ parse_error()
+ tok_cur_tag.public_identifier += "\ufffd"
+ return
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ tok_cur_tag.public_identifier += c
return null
- # 8.2.4.36 http://www.w3.org/TR/html5/syntax.html#after-attribute-name-state
- tok_state_after_attribute_name = ->
+ # 8.2.4.59 http://www.w3.org/TR/html5/syntax.html#doctype-public-identifier-(single-quoted)-state
+ tok_state_doctype_public_identifier_single_quoted = ->
c = txt.charAt(cur++)
- if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
+ if c is "'"
+ tok_state = tok_state_after_doctype_public_identifier
return
- if c is '/'
- tok_state = tok_state_self_closing_start_tag
+ if c is "\u0000"
+ parse_error()
+ tok_cur_tag.public_identifier += "\ufffd"
return
- if c is '='
- tok_state = tok_state_before_attribute_value
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ tok_cur_tag.public_identifier += c
+ return null
+
+ # 8.2.4.60 http://www.w3.org/TR/html5/syntax.html#after-doctype-public-identifier-state
+ tok_state_after_doctype_public_identifier = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ tok_state = tok_state_between_doctype_public_and_system_identifiers
return
if c is '>'
tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '"'
+ parse_error()
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_double_quoted
return
- if uc_alpha.indexOf(c) > -1
- tok_cur_tag.attrs_a.unshift [c.toLowerCase(), '']
- tok_state = tok_state_attribute_name
+ if c is "'"
+ parse_error()
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_single_quoted
return
- if c is "\u0000"
+ if c is '' # EOF
parse_error()
- tok_cur_tag.attrs_a.unshift ["\ufffd", '']
- tok_state = tok_state_attribute_name
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
+ return null
+
+ # 8.2.4.61 http://www.w3.org/TR/html5/syntax.html#between-doctype-public-and-system-identifiers-state
+ tok_state_between_doctype_public_and_system_identifiers = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ return
+ if c is '>'
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '"'
+ parse_error()
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_double_quoted
+ return
+ if c is "'"
+ parse_error()
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_single_quoted
return
if c is '' # EOF
parse_error()
tok_state = tok_state_data
- cur -= 1 # reconsume
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
+ return null
+
+ # 8.2.4.62 http://www.w3.org/TR/html5/syntax.html#after-doctype-system-keyword-state
+ tok_state_after_doctype_system_keyword = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ tok_state = tok_state_before_doctype_system_identifier
return
- if c is '"' or c is "'" or c is '<'
+ if c is '"'
parse_error()
- # fall through to Anything else
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_double_quoted
+ return
+ if c is "'"
+ parse_error()
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_single_quoted
+ return
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
# Anything else
- tok_cur_tag.attrs_a.unshift [c, '']
- tok_state = tok_state_attribute_name
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
+ return null
- # 8.2.4.37 http://www.w3.org/TR/html5/syntax.html#before-attribute-value-state
- tok_state_before_attribute_value = ->
- switch c = txt.charAt(cur++)
- when "\t", "\n", "\u000c", ' '
- return null
- when '"'
- tok_state = tok_state_attribute_value_double_quoted
- when '&'
- tok_state = tok_state_attribute_value_unquoted
- cur -= 1
- when "'"
- tok_state = tok_state_attribute_value_single_quoted
- when "\u0000"
- # Parse error
- tok_cur_tag.attrs_a[0][1] += "\ufffd"
- tok_state = tok_state_attribute_value_unquoted
- when '>'
- # Parse error
- tok_state = tok_state_data
- tmp = tok_cur_tag
- tok_cur_tag = null
- return tmp
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- tok_cur_tag.attrs_a[0][1] += c
- tok_state = tok_state_attribute_value_unquoted
+ # 8.2.4.63 http://www.w3.org/TR/html5/syntax.html#before-doctype-system-identifier-state
+ tok_state_before_doctype_system_identifier = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ return
+ if c is '"'
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_double_quoted
+ return
+ if c is "'"
+ tok_cur_tag.system_identifier = ''
+ tok_state = tok_state_doctype_system_identifier_single_quoted
+ return
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
return null
- # 8.2.4.38 http://www.w3.org/TR/html5/syntax.html#attribute-value-(double-quoted)-state
- tok_state_attribute_value_double_quoted = ->
- switch c = txt.charAt(cur++)
- when '"'
- tok_state = tok_state_after_attribute_value_quoted
- when '&'
- tok_cur_tag.attrs_a[0][1] += parse_character_reference '"', true
- when "\u0000"
- # Parse error
- tok_cur_tag.attrs_a[0][1] += "\ufffd"
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- tok_cur_tag.attrs_a[0][1] += c
+ # 8.2.4.64 http://www.w3.org/TR/html5/syntax.html#doctype-system-identifier-(double-quoted)-state
+ tok_state_doctype_system_identifier_double_quoted = ->
+ c = txt.charAt(cur++)
+ if c is '"'
+ tok_state = tok_state_after_doctype_system_identifier
+ return
+ if c is "\u0000"
+ parse_error()
+ tok_cur_tag.system_identifier += "\ufffd"
+ return
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ tok_cur_tag.system_identifier += c
return null
- # 8.2.4.39 http://www.w3.org/TR/html5/syntax.html#attribute-value-(single-quoted)-state
- tok_state_attribute_value_single_quoted = ->
- switch c = txt.charAt(cur++)
- when "'"
- tok_state = tok_state_after_attribute_value_quoted
- when '&'
- tok_cur_tag.attrs_a[0][1] += parse_character_reference "'", true
- when "\u0000"
- # Parse error
- tok_cur_tag.attrs_a[0][1] += "\ufffd"
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- tok_cur_tag.attrs_a[0][1] += c
+ # 8.2.4.65 http://www.w3.org/TR/html5/syntax.html#doctype-system-identifier-(single-quoted)-state
+ tok_state_doctype_system_identifier_single_quoted = ->
+ c = txt.charAt(cur++)
+ if c is "'"
+ tok_state = tok_state_after_doctype_system_identifier
+ return
+ if c is "\u0000"
+ parse_error()
+ tok_cur_tag.system_identifier += "\ufffd"
+ return
+ if c is '>'
+ parse_error()
+ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ tok_cur_tag.system_identifier += c
return null
- # 8.2.4.40 http://www.w3.org/TR/html5/syntax.html#attribute-value-(unquoted)-state
- tok_state_attribute_value_unquoted = ->
- switch c = txt.charAt(cur++)
- when "\t", "\n", "\u000c", ' '
- tok_state = tok_state_before_attribute_name
- when '&'
- tok_cur_tag.attrs_a[0][1] += parse_character_reference '>', true
- when '>'
- tok_state = tok_state_data
- tmp = tok_cur_tag
- tok_cur_tag = null
- return tmp
- when "\u0000"
- tok_cur_tag.attrs_a[0][1] += "\ufffd"
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- # Parse Error if ', <, = or ` (backtick)
- tok_cur_tag.attrs_a[0][1] += c
+ # 8.2.4.66 http://www.w3.org/TR/html5/syntax.html#after-doctype-system-identifier-state
+ tok_state_after_doctype_system_identifier = ->
+ c = txt.charAt(cur++)
+ if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
+ return
+ if c is '>'
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ tok_cur_tag.flag 'force-quirks', true
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
+ parse_error()
+ # do _not_ tok_cur_tag.flag 'force-quirks', true
+ tok_state = tok_state_bogus_doctype
return null
- # 8.2.4.42 http://www.w3.org/TR/html5/syntax.html#after-attribute-value-(quoted)-state
- tok_state_after_attribute_value_quoted = ->
- switch c = txt.charAt(cur++)
- when "\t", "\n", "\u000c", ' '
- tok_state = tok_state_before_attribute_name
- when '/'
- tok_state = tok_state_self_closing_start_tag
- when '>'
- tok_state = tok_state_data
- tmp = tok_cur_tag
- tok_cur_tag = null
- return tmp
- when '' # EOF
- parse_error()
- tok_state = tok_state_data
- else
- # Parse Error
- tok_state = tok_state_before_attribute_name
- cur -= 1 # we didn't handle that char
+ # 8.2.4.67 http://www.w3.org/TR/html5/syntax.html#bogus-doctype-state
+ tok_state_bogus_doctype = ->
+ c = txt.charAt(cur++)
+ if c is '>'
+ tok_state = tok_state_data
+ return tok_cur_tag
+ if c is '' # EOF
+ tok_state = tok_state_data
+ cur -= 1 # Reconsume
+ return tok_cur_tag
+ # Anything else
return null
+ # 8.2.4.68 http://www.w3.org/TR/html5/syntax.html#cdata-section-state
+ tok_state_cdata_section = ->
+ tok_state = tok_state_data
+ next_gt = txt.indexOf ']]>', cur
+ if next_gt is -1
+ val = txt.substr cur
+ cur = txt.length
+ else
+ val = txt.substr cur, (next_gt - cur)
+ cur = next_gt + 3
+ val = val.replace "\u0000", "\ufffd" # fixfull spec doesn't say this
+ return new_character_token val # fixfull split
+
# 8.2.4.69 http://www.w3.org/TR/html5/syntax.html#consume-a-character-reference
# Don't set this as a state, just call it
# returns a string (NOT a text node)
# tree constructor initialization
# see comments on TYPE_TAG/etc for the structure of this data
doc = new Node TYPE_TAG, name: 'html', namespace: NS_HTML
- open_els = [doc]
+ open_els = []
afe = [] # active formatting elements
- template_insertion_modes = []
- insertion_mode = ins_mode_initial
- original_insertion_mode = insertion_mode # TODO check spec
+ template_ins_modes = []
+ ins_mode = ins_mode_initial
+ original_ins_mode = ins_mode # TODO check spec
flag_scripting = true # TODO might need an extra flag to get <noscript> to parse correctly
flag_frameset_ok = true
flag_parsing = true
temporary_buffer = null
pending_table_character_tokens = []
head_element_pointer = null
+ flag_fragment_parsing = false # parser originally created as part of the html fragment parsing algorithm (fragment case)
+ context_element = null # FIXME initialize from args.fragment http://www.w3.org/TR/html5/syntax.html#parsing-html-fragments
# tokenizer initialization
tok_state = tok_state_data
# proccess input
+ # http://www.w3.org/TR/html5/syntax.html#tree-construction
while flag_parsing
t = tok_state()
if t?
- insertion_mode t
+ process_token t
+ # fixfull parse error if has self-closing flag, but it wasn't acknolwedged
return doc.children
-# everything below is tests on the above
-test_equals = (description, output, expected_output) ->
- if output is expected_output
- console.log "passed." # don't say name, so smart consoles can merge all of these
- else
- console.log "FAILED: \"#{description}\""
- console.log " Expected: #{expected_output}"
- console.log " Actual: #{output}"
serialize_els = (els, shallow, show_ids) ->
serialized = ''
sep = ''
sep = ','
serialized += t.serialize shallow, show_ids
return serialized
-test_parser = (args) ->
- debug_log_reset()
- parse_errors = []
- errors_cb = (i) ->
- parse_errors.push i
- prev_node_id = 0 # reset counter
- parsed = parse_html args.html, errors_cb
- serialized = serialize_els parsed, false, false
- expected = 'tag:"html",{},[tag:"head",{},[],tag:"body",{},[' + args.expected + ']]'
- if serialized isnt expected
- debug_log_each (str) ->
- console.log str
- console.log "FAILED: \"#{args.name}\""
- console.log " Input: #{args.html}"
- console.log " Correct: #{expected}"
- console.log " Output: #{serialized}"
- if parse_errors.length > 0
- console.log " parse errs: #{JSON.stringify parse_errors}"
- else
- console.log " No parse errors"
- else
- console.log "passed \"#{args.name}\""
-
-test_parser name: "empty", \
- html: "",
- expected: ''
-test_parser name: "just text", \
- html: "abc",
- expected: 'text:"abc"'
-test_parser name: "named entity", \
- html: "a&1234",
- expected: 'text:"a&1234"'
-test_parser name: "broken named character references", \
- html: "1&2&&3&aabbcc;",
- expected: 'text:"1&2&&3&aabbcc;"'
-test_parser name: "numbered entity overrides", \
- html: "1€€ ƒ",
- expected: 'text:"1€€ ƒ"'
-test_parser name: "open tag", \
- html: "foo<span>bar",
- expected: 'text:"foo",tag:"span",{},[text:"bar"]'
-test_parser name: "open tag with attributes", \
- html: "foo<span style=\"foo: bar\" title=\"hi\">bar",
- expected: 'text:"foo",tag:"span",{"style":"foo: bar","title":"hi"},[text:"bar"]'
-test_parser name: "open tag with attributes of various quotings", \
- html: "foo<span abc=\"def\" g=hij klm='nopqrstuv\"' autofocus>bar",
- expected: 'text:"foo",tag:"span",{"abc":"def","autofocus":"","g":"hij","klm":"nopqrstuv\\""},[text:"bar"]'
-test_parser name: "attribute entity exceptions dq", \
- html: "foo<a href=\"foo?t=1&=2&o=3&lt=foo\">bar",
- expected: 'text:"foo",tag:"a",{"href":"foo?t=1&=2&o=3<=foo"},[text:"bar"]'
-test_parser name: "attribute entity exceptions sq", \
- html: "foo<a href='foo?t=1&=2&o=3&lt=foo'>bar",
- expected: 'text:"foo",tag:"a",{"href":"foo?t=1&=2&o=3<=foo"},[text:"bar"]'
-test_parser name: "attribute entity exceptions uq", \
- html: "foo<a href=foo?t=1&=2&o=3&lt=foo>bar",
- expected: 'text:"foo",tag:"a",{"href":"foo?t=1&=2&o=3<=foo"},[text:"bar"]'
-test_parser name: "matching closing tags", \
- html: "foo<a href=\"hi\">hi</a><div>1<div>foo</div>2</div>bar",
- expected: 'text:"foo",tag:"a",{"href":"hi"},[text:"hi"],tag:"div",{},[text:"1",tag:"div",{},[text:"foo"],text:"2"],text:"bar"'
-test_parser name: "missing closing tag inside", \
- html: "foo<div>bar<span>baz</div>qux",
- expected: 'text:"foo",tag:"div",{},[text:"bar",tag:"span",{},[text:"baz"]],text:"qux"'
-test_parser name: "mis-matched closing tags", \
- html: "<span>12<div>34</span>56</div>78",
- expected: 'tag:"span",{},[text:"12",tag:"div",{},[text:"3456"],text:"78"]'
-test_parser name: "mis-matched formatting elements", \
- html: "12<b>34<i>56</b>78</i>90",
- expected: 'text:"12",tag:"b",{},[text:"34",tag:"i",{},[text:"56"]],tag:"i",{},[text:"78"],text:"90"'
-test_parser name: "8.2.8.1 Misnested tags: <b><i></b></i>", \
- html: '<p>1<b>2<i>3</b>4</i>5</p>',
- expected: 'tag:"p",{},[text:"1",tag:"b",{},[text:"2",tag:"i",{},[text:"3"]],tag:"i",{},[text:"4"],text:"5"]'
-test_parser name: "8.2.8.2 Misnested tags: <b><p></b></p>", \
- html: '<b>1<p>2</b>3</p>',
- expected: 'tag:"b",{},[text:"1"],tag:"p",{},[tag:"b",{},[text:"2"],text:"3"]'
-test_parser name: "crazy formatting elements test", \
- html: "<b><i><a><s><tt><div></b>first</b></div></tt></s></a>second</i>",
- # chrome does this: expected: 'tag:"b",{},[tag:"i",{},[tag:"a",{},[tag:"s",{},[tag:"tt",{},[]]],text:"second"]],tag:"a",{},[tag:"s",{},[tag:"tt",{},[tag:"div",{},[tag:"b",{},[],text:"first"]]]]'
- # firefox does this:
- expected: 'tag:"b",{},[tag:"i",{},[tag:"a",{},[tag:"s",{},[tag:"tt",{},[]]]]],tag:"a",{},[tag:"s",{},[tag:"tt",{},[tag:"div",{},[tag:"b",{},[],text:"first"]]]],text:"second"'
-# tests from https://github.com/html5lib/html5lib-tests/blob/master/tree-construction/adoption01.dat
-test_parser name: "html5lib aaa 1", \
- html: '<a><p></a></p>',
- expected: 'tag:"a",{},[],tag:"p",{},[tag:"a",{},[]]'
-test_parser name: "html5lib aaa 2", \
- html: '<a>1<p>2</a>3</p>',
- expected: 'tag:"a",{},[text:"1"],tag:"p",{},[tag:"a",{},[text:"2"],text:"3"]'
-test_parser name: "html5lib aaa 3", \
- html: '<a>1<button>2</a>3</button>',
- expected: 'tag:"a",{},[text:"1"],tag:"button",{},[tag:"a",{},[text:"2"],text:"3"]'
-test_parser name: "html5lib aaa 4", \
- html: '<a>1<b>2</a>3</b>',
- expected: 'tag:"a",{},[text:"1",tag:"b",{},[text:"2"]],tag:"b",{},[text:"3"]'
-test_parser name: "html5lib aaa 5 (two divs deep)", \
- html: '<a>1<div>2<div>3</a>4</div>5</div>',
- expected: 'tag:"a",{},[text:"1"],tag:"div",{},[tag:"a",{},[text:"2"],tag:"div",{},[tag:"a",{},[text:"3"],text:"4"],text:"5"]'
-test_parser name: "html5lib aaa 6 (foster parenting)", \
- html: '<table><a>1<p>2</a>3</p>',
- expected: 'tag:"a",{},[text:"1"],tag:"p",{},[tag:"a",{},[text:"2"],text:"3"],tag:"table",{},[]'
-test_parser name: "html5lib aaa 7 (aaa, eof) 1", \
- html: '<b><b><a><p></a>',
- expected: 'tag:"b",{},[tag:"b",{},[tag:"a",{},[],tag:"p",{},[tag:"a",{},[]]]]'
-test_parser name: "html5lib aaa 8 (aaa, eof) 2", \
- html: '<b><a><b><p></a>',
- expected: 'tag:"b",{},[tag:"a",{},[tag:"b",{},[]],tag:"b",{},[tag:"p",{},[tag:"a",{},[]]]]'
-test_parser name: "html5lib aaa 9 (aaa, eof) 3", \
- html: '<a><b><b><p></a>',
- expected: 'tag:"a",{},[tag:"b",{},[tag:"b",{},[]]],tag:"b",{},[tag:"b",{},[tag:"p",{},[tag:"a",{},[]]]]'
-test_parser name: "html5lib aaa 10 (formatting, nesting, attrs, aaa)", \
- html: '<p>1<s id="A">2<b id="B">3</p>4</s>5</b>',
- expected: 'tag:"p",{},[text:"1",tag:"s",{"id":"A"},[text:"2",tag:"b",{"id":"B"},[text:"3"]]],tag:"s",{"id":"A"},[tag:"b",{"id":"B"},[text:"4"]],tag:"b",{"id":"B"},[text:"5"]'
-test_parser name: "html5lib aaa 11 (table with foster parenting, formatting el and td)", \
- html: '<table><a>1<td>2</td>3</table>',
- expected: 'tag:"a",{},[text:"1"],tag:"a",{},[text:"3"],tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[text:"2"]]]]'
-test_parser name: "html5lib aaa 12 (table with foster parenting, split text)", \
- html: '<table>A<td>B</td>C</table>',
- expected: 'text:"AC",tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[text:"B"]]]]'
-# TODO implement svg and namespacing
-#test_parser name: "html5lib aaa 13 (svg tr input)", \
-# html: '<a><svg><tr><input></a>',
-# expected: 'tag:"a",{},[svg:"svg",{},[svg:"tr",{},[svg:"input"]]]'
-test_parser name: "html5lib aaa 14 (deep ?outer aaa)", \
- html: '<div><a><b><div><div><div><div><div><div><div><div><div><div></a>',
- expected: 'tag:"div",{},[tag:"a",{},[tag:"b",{},[]],tag:"b",{},[tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[],tag:"div",{},[tag:"a",{},[tag:"div",{},[tag:"div",{},[]]]]]]]]]]]]]'
-test_parser name: "html5lib aaa 15 (deep ?inner aaa)", \
- html: '<div><a><b><u><i><code><div></a>',
- expected: 'tag:"div",{},[tag:"a",{},[tag:"b",{},[tag:"u",{},[tag:"i",{},[tag:"code",{},[]]]]],tag:"u",{},[tag:"i",{},[tag:"code",{},[tag:"div",{},[tag:"a",{},[]]]]]]'
-test_parser name: "html5lib aaa 16 (correctly nested 4b)", \
- html: '<b><b><b><b>x</b></b></b></b>y',
- expected: 'tag:"b",{},[tag:"b",{},[tag:"b",{},[tag:"b",{},[text:"x"]]]],text:"y"'
-test_parser name: "html5lib aaa 17 (formatting, implied /p, noah's ark)", \
- html: '<p><b><b><b><b><p>x',
- expected: 'tag:"p",{},[tag:"b",{},[tag:"b",{},[tag:"b",{},[tag:"b",{},[]]]]],tag:"p",{},[tag:"b",{},[tag:"b",{},[tag:"b",{},[text:"x"]]]]'
-test_parser name: "variation on html5lib aaa 17 (with attributes in various orders)", \
- html: '<p><b c="d" e="f"><b e="f" c="d"><b e="f" c="d"><b c="d" e="f"><p>x',
- expected: 'tag:"p",{},[tag:"b",{"c":"d","e":"f"},[tag:"b",{"c":"d","e":"f"},[tag:"b",{"c":"d","e":"f"},[tag:"b",{"c":"d","e":"f"},[]]]]],tag:"p",{},[tag:"b",{"c":"d","e":"f"},[tag:"b",{"c":"d","e":"f"},[tag:"b",{"c":"d","e":"f"},[text:"x"]]]]'
-test_parser name: "junk after attribute close-quote", \
- html: '<p><b c="d", e="f">foo<p>x',
- expected: 'tag:"p",{},[tag:"b",{",":"","c":"d","e":"f"},[text:"foo"]],tag:"p",{},[tag:"b",{",":"","c":"d","e":"f"},[text:"x"]]'
-test_parser name: "html5lib aaa02 1", \
- html: '<b>1<i>2<p>3</b>4',
- expected: 'tag:"b",{},[text:"1",tag:"i",{},[text:"2"]],tag:"i",{},[tag:"p",{},[tag:"b",{},[text:"3"],text:"4"]]'
-test_parser name: "html5lib aaa02 2", \
- html: '<a><div><style></style><address><a>',
- expected: 'tag:"a",{},[],tag:"div",{},[tag:"a",{},[tag:"style",{},[]],tag:"address",{},[tag:"a",{},[],tag:"a",{},[]]]'
-test_parser name: "html5lib tables 1", \
- html: '<table><th>',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"th",{},[]]]]'
-test_parser name: "html5lib tables 2", \
- html: '<table><td>',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[]]]]'
-test_parser name: "html5lib tables 3", \
- html: "<table><col foo='bar'>",
- expected: 'tag:"table",{},[tag:"colgroup",{},[tag:"col",{"foo":"bar"},[]]]'
-test_parser name: "html5lib tables 4", \
- html: '<table><colgroup></html>foo',
- expected: 'text:"foo",tag:"table",{},[tag:"colgroup",{},[]]'
-test_parser name: "html5lib tables 5", \
- html: '<table></table><p>foo',
- expected: 'tag:"table",{},[],tag:"p",{},[text:"foo"]'
-test_parser name: "html5lib tables 6", \
- html: '<table></body></caption></col></colgroup></html></tbody></td></tfoot></th></thead></tr><td>',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[]]]]'
-test_parser name: "html5lib tables 7", \
- html: '<table><select><option>3</select></table>',
- expected: 'tag:"select",{},[tag:"option",{},[text:"3"]],tag:"table",{},[]'
-test_parser name: "html5lib tables 8", \
- html: '<table><select><table></table></select></table>',
- expected: 'tag:"select",{},[],tag:"table",{},[],tag:"table",{},[]'
-test_parser name: "html5lib tables 9", \
- html: '<table><select></table>',
- expected: 'tag:"select",{},[],tag:"table",{},[]'
-test_parser name: "html5lib tables 10", \
- html: '<table><select><option>A<tr><td>B</td></tr></table>',
- expected: 'tag:"select",{},[tag:"option",{},[text:"A"]],tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[text:"B"]]]]'
-test_parser name: "html5lib tables 11", \
- html: '<table><td></body></caption></col></colgroup></html>foo',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[text:"foo"]]]]'
-test_parser name: "html5lib tables 12", \
- html: '<table><td>A</table>B',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[text:"A"]]]],text:"B"'
-test_parser name: "html5lib tables 13", \
- html: '<table><tr><caption>',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[]],tag:"caption",{},[]]'
-test_parser name: "html5lib tables 14", \
- html: '<table><tr></body></caption></col></colgroup></html></td></th><td>foo',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[text:"foo"]]]]'
-test_parser name: "html5lib tables 15", \
- html: '<table><td><tr>',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[]],tag:"tr",{},[]]]'
-test_parser name: "html5lib tables 16", \
- html: '<table><td><button><td>',
- expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[tag:"button",{},[]],tag:"td",{},[]]]]'
-# TODO implement svg parsing
-#test_parser name: "html5lib tables 17", \
-# html: '<table><tr><td><svg><desc><td>',
-# expected: 'tag:"table",{},[tag:"tbody",{},[tag:"tr",{},[tag:"td",{},[svg:"svg",{},[svg:"desc",{},[]]],tag:"td",{},[]]]]'
+
+# TODO export TYPE_*
+module.exports.parse_html = parse_html
+module.exports.debug_log_reset = debug_log_reset
+module.exports.debug_log_each = debug_log_each
+module.exports.TYPE_TAG = TYPE_TAG
+module.exports.TYPE_TEXT = TYPE_TEXT
+module.exports.TYPE_COMMENT = TYPE_COMMENT
+module.exports.TYPE_DOCTYPE = TYPE_DOCTYPE
+module.exports.NS_HTML = NS_HTML
+module.exports.NS_MATHML = NS_MATHML
+module.exports.NS_SVG = NS_SVG