# along with this program. If not, see <http://www.gnu.org/licenses/>.
-# This file implements a parser for html snippets, meant to be used by a
-# WYSIWYG editor. Hence it does not attempt to parse doctypes, <html>, <head>
-# or <body> tags, nor does it produce the top level "document" node in the dom
-# tree, nor nodes for html, head or body. Comments containing "fixfull"
-# indicate places where additional code is needed for full HTML document
-# parsing.
+# This file implements a thorough parser for html5, meant to be used by a
+# WYSIWYG editor.
+
+# The implementation is a pretty direct implementation of the parsing algorithm
+# described here:
+#
+# http://www.w3.org/TR/html5/syntax.html
+#
+# except for some places marked "WHATWG" that are implemented as described here:
+#
+# https://html.spec.whatwg.org/multipage/syntax.html
+#
+# This code passes all of the tests in the .dat files at:
+#
+# https://github.com/JasonWoof/html5lib-tests/tree/patch-1/tree-construction
+
+
+##################################
+## how to use this code
+##################################
#
-# Instead, the data structure produced by this parser is an array of Nodes.
+# See README.md for how to pre-compile this file, or compile it in the browser.
+#
+# This file exports a single useful function: parse_tml
+#
+# Once you include this file in a page (see index.html for an example) you'll
+# have window.wheic
+#
+# Call it like this:
+#
+# wheic.parse_html({html: "<p><b>hi</p>"})
+#
+# Or, if you don't want <html><head><body>/etc, do this:
+#
+# wheic.parse_html({fragment: "body", html: "<p><b>hi</p>"})
+#
+# This code can _almost_ run outside the browser (eg under node.js). To get it
+# to run without the browser would require native implementation of
+# decode_named_char_ref(). The current implementation of that function uses the
+# browser's DOM api, to save space (the list of valid named characters is
+# massive.)
+
+# This code is a work in progress, eg try search this file for "fixfull",
+# "TODO" and "FIXME"
-# stacks/lists
+# Notes: stacks/lists
#
-# the spec uses a many different words do indicate which ends of lists/stacks
-# they are talking about (and relative movement within the lists/stacks). This
-# section splains. I'm implementing "lists" (afe and open_els) the same way
-# (both as stacks)
+# Jason was frequently confused by the terminology used to refer to different
+# parts of the stacks and lists in the spec, so he made this chart to help keep
+# his head straight:
#
# stacks grow downward (current element is index=0)
#
# 1: b
# 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
+from_code_point = (x) ->
+ if String.fromCodePoint?
+ return String.fromCodePoint x
+ else
+ if x <= 0xffff
+ return String.fromCharCode x
+ x -= 0x10000
+ return String.fromCharCode((x >> 10) + 0xd800, (x % 0x400) + 0xdc00)
+
# Each node is an obect of the Node class. Here are the Node types:
TYPE_TAG = 0 # name, {attributes}, [children]
TYPE_TEXT = 1 # "text"
NS_MATHML = 2
NS_SVG = 3
+# quirks mode constants
+QUIRKS_NO = 1
+QUIRKS_LIMITED = 2
+QUIRKS_YES = 3
+
+# queue up debug logs, so eg they can be shown only for tests that fail
g_debug_log = []
debug_log_reset = ->
g_debug_log = []
+ return
debug_log = (str) ->
g_debug_log.push str
+ return
debug_log_each = (cb) ->
for str in g_debug_log
cb str
+ return
prev_node_id = 0
class Node
@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, token: @token
acknowledge_self_closing: ->
if @token?
- @token.flag 'did_self_close'
+ @token.flag 'did_self_close', true
else
@flag 'did_self_close', true
- flag: ->
- # fixfull
- serialize: (shallow = false, show_ids = false) -> # for unit tests
- ret = ''
- switch @type
- when TYPE_TAG
- ret += 'tag:'
- ret += JSON.stringify @name
- ret += ','
- if show_ids
- ret += "##{@id},"
- if shallow
- break
- attr_keys = []
- for k of @attrs
- attr_keys.push k
- attr_keys.sort()
- ret += '{'
- sep = ''
- for k in attr_keys
- ret += sep
- sep = ','
- ret += "#{JSON.stringify k}:#{JSON.stringify @attrs[k]}"
- ret += '},['
- sep = ''
- for c in @children
- ret += sep
- sep = ','
- ret += c.serialize shallow, show_ids
- ret += ']'
- when TYPE_TEXT
- ret += 'text:'
- ret += JSON.stringify @text
- when TYPE_COMMENT
- ret += 'comment:'
- ret += JSON.stringify @text
- when TYPE_DOCTYPE
- ret += 'doctype'
- # FIXME
- when TYPE_AFE_MARKER
- ret += 'marker'
- when TYPE_AAA_BOOKMARK
- ret += 'aaa_bookmark'
- else
- ret += 'unknown:'
- console.log "unknown: #{JSON.stringify @}" # backtrace is just as well
- return ret
+ return
+ flag: (key, value = null) ->
+ if value?
+ @flags[key] = value
+ else
+ return @flags[key]
+ return
# helpers: (only take args that are normally known when parser creates nodes)
new_open_tag = (name) ->
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 false unless t.type is TYPE_START_TAG
+ for a in 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"
+unicode_fixes = {}
+unicode_fixes[0x00] = "\uFFFD"
+unicode_fixes[0x80] = "\u20AC"
+unicode_fixes[0x82] = "\u201A"
+unicode_fixes[0x83] = "\u0192"
+unicode_fixes[0x84] = "\u201E"
+unicode_fixes[0x85] = "\u2026"
+unicode_fixes[0x86] = "\u2020"
+unicode_fixes[0x87] = "\u2021"
+unicode_fixes[0x88] = "\u02C6"
+unicode_fixes[0x89] = "\u2030"
+unicode_fixes[0x8A] = "\u0160"
+unicode_fixes[0x8B] = "\u2039"
+unicode_fixes[0x8C] = "\u0152"
+unicode_fixes[0x8E] = "\u017D"
+unicode_fixes[0x91] = "\u2018"
+unicode_fixes[0x92] = "\u2019"
+unicode_fixes[0x93] = "\u201C"
+unicode_fixes[0x94] = "\u201D"
+unicode_fixes[0x95] = "\u2022"
+unicode_fixes[0x96] = "\u2013"
+unicode_fixes[0x97] = "\u2014"
+unicode_fixes[0x98] = "\u02DC"
+unicode_fixes[0x99] = "\u2122"
+unicode_fixes[0x9A] = "\u0161"
+unicode_fixes[0x9B] = "\u203A"
+unicode_fixes[0x9C] = "\u0153"
+unicode_fixes[0x9E] = "\u017E"
+unicode_fixes[0x9F] = "\u0178"
+
+quirks_yes_pi_prefixes = [
+ "+//silmaril//dtd html pro v0r11 19970101//"
+ "-//as//dtd html 3.0 aswedit + extensions//"
+ "-//advasoft ltd//dtd html 3.0 aswedit + extensions//"
+ "-//ietf//dtd html 2.0 level 1//"
+ "-//ietf//dtd html 2.0 level 2//"
+ "-//ietf//dtd html 2.0 strict level 1//"
+ "-//ietf//dtd html 2.0 strict level 2//"
+ "-//ietf//dtd html 2.0 strict//"
+ "-//ietf//dtd html 2.0//"
+ "-//ietf//dtd html 2.1e//"
+ "-//ietf//dtd html 3.0//"
+ "-//ietf//dtd html 3.2 final//"
+ "-//ietf//dtd html 3.2//"
+ "-//ietf//dtd html 3//"
+ "-//ietf//dtd html level 0//"
+ "-//ietf//dtd html level 1//"
+ "-//ietf//dtd html level 2//"
+ "-//ietf//dtd html level 3//"
+ "-//ietf//dtd html strict level 0//"
+ "-//ietf//dtd html strict level 1//"
+ "-//ietf//dtd html strict level 2//"
+ "-//ietf//dtd html strict level 3//"
+ "-//ietf//dtd html strict//"
+ "-//ietf//dtd html//"
+ "-//metrius//dtd metrius presentational//"
+ "-//microsoft//dtd internet explorer 2.0 html strict//"
+ "-//microsoft//dtd internet explorer 2.0 html//"
+ "-//microsoft//dtd internet explorer 2.0 tables//"
+ "-//microsoft//dtd internet explorer 3.0 html strict//"
+ "-//microsoft//dtd internet explorer 3.0 html//"
+ "-//microsoft//dtd internet explorer 3.0 tables//"
+ "-//netscape comm. corp.//dtd html//"
+ "-//netscape comm. corp.//dtd strict html//"
+ "-//o'reilly and associates//dtd html 2.0//"
+ "-//o'reilly and associates//dtd html extended 1.0//"
+ "-//o'reilly and associates//dtd html extended relaxed 1.0//"
+ "-//sq//dtd html 2.0 hotmetal + extensions//"
+ "-//softquad software//dtd hotmetal pro 6.0::19990601::extensions to html 4.0//"
+ "-//softquad//dtd hotmetal pro 4.0::19971010::extensions to html 4.0//"
+ "-//spyglass//dtd html 2.0 extended//"
+ "-//sun microsystems corp.//dtd hotjava html//"
+ "-//sun microsystems corp.//dtd hotjava strict html//"
+ "-//w3c//dtd html 3 1995-03-24//"
+ "-//w3c//dtd html 3.2 draft//"
+ "-//w3c//dtd html 3.2 final//"
+ "-//w3c//dtd html 3.2//"
+ "-//w3c//dtd html 3.2s draft//"
+ "-//w3c//dtd html 4.0 frameset//"
+ "-//w3c//dtd html 4.0 transitional//"
+ "-//w3c//dtd html experimental 19960712//"
+ "-//w3c//dtd html experimental 970421//"
+ "-//w3c//dtd w3 html//"
+ "-//w3o//dtd w3 html 3.0//"
+ "-//webtechs//dtd mozilla html 2.0//"
+ "-//webtechs//dtd mozilla html//"
+]
+
# These are the character references that don't need a terminating semicolon
# min length: 2, max: 6, none are a prefix of any other.
legacy_char_refs = {
h2:NS_HTML, h3:NS_HTML, h4:NS_HTML, h5:NS_HTML, h6:NS_HTML, head:NS_HTML,
header:NS_HTML, hgroup:NS_HTML, hr:NS_HTML, html:NS_HTML, iframe:NS_HTML,
img:NS_HTML, input:NS_HTML, isindex:NS_HTML, li:NS_HTML, link:NS_HTML,
- listing:NS_HTML, main:NS_HTML, marquee:NS_HTML, meta:NS_HTML, nav:NS_HTML,
- noembed:NS_HTML, noframes:NS_HTML, noscript:NS_HTML, object:NS_HTML,
- ol:NS_HTML, p:NS_HTML, param:NS_HTML, plaintext:NS_HTML, pre:NS_HTML,
- script:NS_HTML, section:NS_HTML, select:NS_HTML, source:NS_HTML,
- style:NS_HTML, summary:NS_HTML, table:NS_HTML, tbody:NS_HTML, td:NS_HTML,
- template:NS_HTML, textarea:NS_HTML, tfoot:NS_HTML, th:NS_HTML,
- thead:NS_HTML, title:NS_HTML, tr:NS_HTML, track:NS_HTML, ul:NS_HTML,
- wbr:NS_HTML, xmp:NS_HTML,
+ listing:NS_HTML, main:NS_HTML, marquee:NS_HTML,
+
+ menu:NS_HTML,menuitem:NS_HTML, # WHATWG adds these
+
+ meta:NS_HTML, nav:NS_HTML, noembed:NS_HTML, noframes:NS_HTML,
+ noscript:NS_HTML, object:NS_HTML, ol:NS_HTML, p:NS_HTML, param:NS_HTML,
+ plaintext:NS_HTML, pre:NS_HTML, script:NS_HTML, section:NS_HTML,
+ select:NS_HTML, source:NS_HTML, style:NS_HTML, summary:NS_HTML,
+ table:NS_HTML, tbody:NS_HTML, td:NS_HTML, template:NS_HTML,
+ textarea:NS_HTML, tfoot:NS_HTML, th:NS_HTML, thead:NS_HTML, title:NS_HTML,
+ tr:NS_HTML, track:NS_HTML, ul:NS_HTML, wbr:NS_HTML, xmp:NS_HTML,
# MathML:
mi:NS_MATHML, mo:NS_MATHML, mn:NS_MATHML, ms:NS_MATHML, mtext:NS_MATHML,
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] is el.namespace
+is_html_integration = (el) -> # DON'T PASS A TOKEN
+ if el.namespace is NS_MATHML
+ if 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
+}
+
foster_parenting_targets = {
- table: true
- tbody: true
- tfoot: true
- thead: true
- tr: true
+ table: NS_HTML
+ tbody: NS_HTML
+ tfoot: NS_HTML
+ thead: NS_HTML
+ tr: NS_HTML
}
-# all html I presume
end_tag_implied = {
- dd: true
- dt: true
- li: true
- option: true
- optgroup: true
- p: true
- rb: true
- rp: true
- rt: true
- rtc: true
+ dd: NS_HTML
+ dt: NS_HTML
+ li: NS_HTML
+ option: NS_HTML
+ optgroup: NS_HTML
+ p: NS_HTML
+ rb: NS_HTML
+ rp: NS_HTML
+ rt: NS_HTML
+ rtc: NS_HTML
}
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'
+ # WHATWG removes this: 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'
+}
+foreign_attr_fixes = {
+ 'xlink:actuate': 'xlink actuate'
+ 'xlink:arcrole': 'xlink arcrole'
+ 'xlink:href': 'xlink href'
+ 'xlink:role': 'xlink role'
+ 'xlink:show': 'xlink show'
+ 'xlink:title': 'xlink title'
+ 'xlink:type': 'xlink type'
+ 'xml:base': 'xml base'
+ 'xml:lang': 'xml lang'
+ 'xml:space': 'xml space'
+ 'xmlns': 'xmlns'
+ 'xmlns:xlink': 'xmlns xlink'
+}
+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
+ for a in t.attrs_a
+ if foreign_attr_fixes[a[0]]?
+ a[0] = foreign_attr_fixes[a[0]]
+ return
+
# decode_named_char_ref()
#
# The list of named character references is _huge_ so ask the browser to decode
return null if decoded is txt
return g_dncr.cache[txt] = decoded
-parse_html = (txt, parse_error_cb = null) ->
- cur = 0 # index of next char in txt to be parsed
+parse_html = (args) ->
+ txt = null
+ cur = null # index of next char in txt to be parsed
# declare doc and tokenizer variables so they're in scope below
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
stop_parsing = ->
flag_parsing = false
+ return
parse_error = ->
- if parse_error_cb?
- parse_error_cb cur
+ if args.error_cb?
+ args.error_cb cur
else
console.log "Parse error at character #{cur} of #{txt.length}"
+ return
+ # http://www.w3.org/TR/html5/syntax.html#push-onto-the-list-of-active-formatting-elements
+ # "Noah's Ark clause" but with three
afe_push = (new_el) ->
matches = 0
for el, i in afe
+ if el.type is TYPE_AFE_MARKER
+ break
if el.name is new_el.name and el.namespace is new_el.namespace
+ attrs_match = true
for k, v of el.attrs
- continue unless new_el.attrs[k] is v
- for k, v of new_el.attrs
- continue unless el.attrs[k] is v
- matches += 1
- if matches is 3
- afe.splice i, 1
- break
+ unless new_el.attrs[k] is v
+ attrs_match = false
+ break
+ if attrs_match
+ for k, v of new_el.attrs
+ unless el.attrs[k] is v
+ attrs_match = false
+ break
+ if attrs_match
+ matches += 1
+ if matches is 3
+ afe.splice i, 1
+ break
afe.unshift new_el
+ return
+
afe_push_marker = ->
afe.unshift new_afe_marker()
+ return
# the functions below impliment the Tree Contstruction algorithm
# http://www.w3.org/TR/html5/syntax.html#tree-construction
# But first... the helpers
template_tag_is_open = ->
- for t in open_els
- if t.name is 'template' # maybe should also check: and t.namespace is 'html'
+ for el in open_els
+ if el.name is 'template' and el.namespace is NS_HTML
return true
return false
is_in_scope_x = (tag_name, scope, namespace) ->
- for t in open_els
- if t.name is tag_name and (namespace is null or namespace is t.namespace)
+ for el in open_els
+ if el.name is tag_name and (namespace is null or namespace is el.namespace)
return true
- if scope[t.name] is t.namespace
+ if scope[el.name] is el.namespace
return false
return false
is_in_scope_x_y = (tag_name, scope, scope2, namespace) ->
- for t in open_els
- if t.name is tag_name and (namespace is null or namespace is t.namespace)
+ for el in open_els
+ if el.name is tag_name and (namespace is null or namespace is el.namespace)
return true
- if scope[t.name] is t.namespace
+ if scope[el.name] is el.namespace
return false
- if scope2[t.name] is t.namespace
+ if scope2[el.name] is el.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,
+ template: NS_HTML,
- mo: NS_MATHML, mn: NS_MATHML, ms: NS_MATHML, mtext: NS_MATHML,
- 'annotation-xml': NS_MATHML,
+ mi: NS_MATHML, mo: NS_MATHML, mn: NS_MATHML, ms: NS_MATHML,
+ mtext: NS_MATHML, 'annotation-xml': NS_MATHML,
foreignObject: NS_SVG, desc: NS_SVG, title: NS_SVG
}
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.namespace isnt NS_HTML and t.name isnt 'optgroup' and t.name isnt 'option'
return false
return false
# this checks for a particular element, not by name
- el_is_in_scope = (el) ->
- for t in open_els
- if t is el
+ # this requires a namespace match
+ el_is_in_scope = (needle) ->
+ for el in open_els
+ if el is needle
return true
- if standard_scopers[t.name] is t.namespace
+ if standard_scopers[el.name] is el.namespace
return false
return false
open_els.shift()
return
clear_to_table_body_stopers = {
- 'tbody': true
- 'tfoot': true
- 'thead': true
- 'template': true
- 'html': true
+ tbody: NS_HTML
+ tfoot: NS_HTML
+ thead: NS_HTML
+ template: NS_HTML
+ html: NS_HTML
}
clear_stack_to_table_body_context = ->
loop
- if clear_to_table_body_stopers[open_els[0].name]?
+ if clear_to_table_body_stopers[open_els[0].name] is open_els[0].namespace
break
open_els.shift()
return
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.
loop
if node_i is open_els.length - 1
last = true
- # fixfull (fragment case)
-
+ if flag_fragment_parsing
+ node = context_element
# 4. If node is a select element, run these substeps:
- if node.name is 'select'
+ if node.name is 'select' and node.namespace is NS_HTML
# 1. If last is true, jump to the step below labeled done.
unless last
# 2. Let ancestor be node.
ancestor = open_els[ancestor_i]
# 5. If ancestor is a template node, jump to the step below
# labeled done.
- if ancestor.name is 'template'
+ if ancestor.name is 'template' and ancestor.namespace is NS_HTML
break
# 6. If ancestor is a table node, switch the insertion mode
# to "in select in table" and abort these steps.
- if ancestor.name is 'table'
- insertion_mode = ins_mode_in_select_in_table
+ if ancestor.name is 'table' and ancestor.namespace is NS_HTML
+ ins_mode = ins_mode_in_select_in_table
return
# 7. Jump back to the step labeled loop.
# 8. Done: Switch the insertion mode to "in select" and abort
# these steps.
- 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
+ if (node.name is 'td' or node.name is 'th') and node.namespace is NS_HTML and last is false
+ ins_mode = ins_mode_in_cell
return
# 6. If node is a tr element, then switch the insertion mode to "in
# row" and abort these steps.
- if node.name is 'tr'
- insertion_mode = ins_mode_in_row
+ if node.name is 'tr' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_row
return
# 7. If node is a tbody, thead, or tfoot element, then switch the
# insertion mode to "in table body" and abort these steps.
- if node.name is 'tbody' or node.name is 'thead' or node.name is 'tfoot'
- insertion_mode = ins_mode_in_table_body
+ if (node.name is 'tbody' or node.name is 'thead' or node.name is 'tfoot') and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_table_body
return
# 8. If node is a caption element, then switch the insertion mode
# to "in caption" and abort these steps.
- if node.name is 'caption'
- insertion_mode = ins_mode_in_caption
+ if node.name is 'caption' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_caption
return
# 9. If node is a colgroup element, then switch the insertion mode
# to "in column group" and abort these steps.
- if node.name is 'colgroup'
- insertion_mode = ins_mode_in_column_group
+ if node.name is 'colgroup' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_column_group
return
# 10. If node is a table element, then switch the insertion mode to
# "in table" and abort these steps.
- if node.name is 'table'
- insertion_mode = ins_mode_in_table
+ if node.name is 'table' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_table
return
# 11. If node is a template element, then switch the insertion mode
# to the current template insertion mode and abort these steps.
- # fixfull (template insertion mode stack)
-
+ if node.name is 'template' and node.namespace is NS_HTML
+ ins_mode = template_ins_modes[0]
+ return
# 12. If node is a head element and last is true, then switch the
# insertion mode to "in body" ("in body"! not "in head"!) and abort
# these steps. (fragment case)
- if node.name is 'head' and last
- insertion_mode = ins_mode_in_body
+ if node.name is 'head' and node.namespace is NS_HTML and last
+ ins_mode = ins_mode_in_body
return
# 13. If node is a head element and last is false, then switch the
# insertion mode to "in head" and abort these steps.
- if node.name is 'head' and last is false
- insertion_mode = ins_mode_in_head
+ if node.name is 'head' and node.namespace is NS_HTML and last is false
+ ins_mode = ins_mode_in_head
return
# 14. If node is a body element, then switch the insertion mode to
# "in body" and abort these steps.
- if node.name is 'body'
- insertion_mode = ins_mode_in_body
+ if node.name is 'body' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_body
return
# 15. If node is a frameset element, then switch the insertion mode
# to "in frameset" and abort these steps. (fragment case)
- if node.name is 'frameset'
- insertion_mode = ins_mode_in_frameset
+ if node.name is 'frameset' and node.namespace is NS_HTML
+ ins_mode = ins_mode_in_frameset
return
# 16. If node is an html element, run these substeps:
- if node.name is 'html'
+ if node.name is 'html' and node.namespace is NS_HTML
# 1. If the head element pointer is null, switch the insertion
# mode to "before head" and abort these steps. (fragment case)
- # 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_i += 1
node = open_els[node_i]
# 19. Return to the step labeled loop.
+ return
# 8.2.3.2
# http://www.w3.org/TR/html5/syntax.html#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
+ return
# http://www.w3.org/TR/html5/syntax.html#adoption-agency-algorithm
# adoption agency algorithm
# http://www.w3.org/TR/html5/syntax.html#misnested-tags:-b-p-/b-/p
# http://www.w3.org/TR/html5/syntax.html#unclosed-formatting-elements
adoption_agency = (subject) ->
- debug_log "adoption_agency()"
- debug_log "tree: #{serialize_els doc.children, false, true}"
- debug_log "open_els: #{serialize_els open_els, true, true}"
- debug_log "afe: #{serialize_els afe, true, true}"
- if open_els[0].name is subject
- el = open_els[0]
- open_els.shift()
+# this block implements tha W3C spec
+# # 1. If the current node is an HTML element whose tag name is subject,
+# # then run these substeps:
+# #
+# # 1. Let element be the current node.
+# #
+# # 2. Pop element off the stack of open elements.
+# #
+# # 3. If element is also in the list of active formatting elements,
+# # remove the element from the list.
+# #
+# # 4. Abort the adoption agency algorithm.
+# if open_els[0].name is subject and open_els[0].namespace is NS_HTML
+# el = open_els.shift()
+# # remove it from the list of active formatting elements (if found)
+# for t, i in afe
+# if t is el
+# afe.splice i, 1
+# break
+# return
+# WHATWG: https://html.spec.whatwg.org/multipage/syntax.html#adoption-agency-algorithm
+ # If the current node is an HTML element whose tag name is subject, and
+ # the current node is not in the list of active formatting elements,
+ # then pop the current node off the stack of open elements, and abort
+ # these steps.
+ if open_els[0].name is subject and open_els[0].namespace is NS_HTML
# remove it from the list of active formatting elements (if found)
- for t, i in afe
- if t is el
- afe.splice i, 1
+ in_afe = false
+ for el, i in afe
+ if el is open_els[0]
+ in_afe = true
break
- debug_log "aaa: starting off with subject on top of stack, exiting"
- return
+ unless in_afe
+ open_els.shift()
+ return
+ # fall through
+# END WHATWG
outer = 0
loop
if outer >= 8
# If there is no such element, then abort these steps and instead
# act as described in the "any other end tag" entry above.
if fe is null
- debug_log "aaa: fe not found in afe"
in_body_any_other_end_tag subject
return
# 6. If formatting element is not in the stack of open elements,
in_open_els = true
break
unless in_open_els
- debug_log "aaa: fe not found in open_els"
parse_error()
# "remove it from the list" must mean afe, since it's not in open_els
afe.splice fe_of_afe, 1
# the element is not in scope, then this is a parse error; abort
# these steps.
unless el_is_in_scope fe
- debug_log "aaa: fe not in scope"
parse_error()
return
# 8. If formatting element is not the current node, this is a parse
# formatting element from the list of active formatting elements,
# and finally abort these steps.
if fb is null
- debug_log "aaa: no fb"
loop
t = open_els.shift()
if t is fe
node_next = open_els[i + 1]
break
node = node_next ? node_above
- debug_log "inner loop #{inner}"
- debug_log "tree: #{serialize_els doc.children, false, true}"
- debug_log "open_els: #{serialize_els open_els, true, true}"
- debug_log "afe: #{serialize_els afe, true, true}"
- debug_log "ca: #{ca.name}##{ca.id} children: #{serialize_els ca.children, true, true}"
- debug_log "fe: #{fe.name}##{fe.id} children: #{serialize_els fe.children, true, true}"
- debug_log "fb: #{fb.name}##{fb.id} children: #{serialize_els fb.children, true, true}"
- debug_log "node: #{node.serialize true, true}"
# TODO make sure node_above gets re-set if/when node is removed from open_els
# 4. If node is formatting element, then go to the next step in
# the overall algorithm.
if node is fe
break
- debug_log "the meat"
# 5. If inner loop counter is greater than three and node is in
# the list of active formatting elements, then remove node from
# the list of active formatting elements.
if t is node
if inner > 3
afe.splice i, 1
- debug_log "max out inner"
else
node_in_afe = true
- debug_log "in afe"
break
# 6. If node is not in the list of active formatting elements,
# then remove node from the stack of open elements and then go
# back to the step labeled inner loop.
unless node_in_afe
- debug_log "not in afe"
for t, i in open_els
if t is node
node_above = open_els[i + 1]
open_els.splice i, 1
break
continue
- debug_log "the bones"
# 7. create an element for the token for which the element node
# was created, in the HTML namespace, with common ancestor as
# the intended parent; replace the entry for node in the list
# 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
- debug_log "replaced in afe"
break
for t, i in open_els
if t is node
node_above = open_els[i + 1]
open_els[i] = new_node
- debug_log "replaced in open_els"
break
node = new_node
# 8. If last node is furthest block, then move the
for t, i in afe
if t is bookmark
afe.splice i, 1
- debug_log "removed bookmark"
break
for t, i in afe
if t is node
# "after" means lower
afe.splice i, 0, bookmark # "after as <-
- debug_log "placed bookmark after node"
- debug_log "node: #{node.id} afe: #{serialize_els afe, true, true}"
break
# 9. Insert last node into node, first removing it from its
# previous parent node if any.
if last_node.parent?
- debug_log "last_node has parent"
for c, i in last_node.parent.children
if c is last_node
- debug_log "removing last_node from parent"
last_node.parent.children.splice i, 1
break
node.children.push last_node
last_node.parent = node
# 10. Let last node be node.
last_node = node
- debug_log "at last"
# 11. Return to the step labeled inner loop.
# 14. Insert whatever last node ended up being in the previous step
# at the appropriate place for inserting a node, but using common
# * last_node is fb
# * last_node is still in the tree (not a duplicate)
if last_node.parent?
- debug_log "FEFIRST? last_node has parent"
for c, i in last_node.parent.children
if c is last_node
- debug_log "removing last_node from parent"
last_node.parent.children.splice i, 1
break
-
- debug_log "after aaa inner loop"
- debug_log "ca: #{ca.name}##{ca.id} children: #{serialize_els ca.children, true, true}"
- debug_log "fe: #{fe.name}##{fe.id} children: #{serialize_els fe.children, true, true}"
- debug_log "fb: #{fb.name}##{fb.id} children: #{serialize_els fb.children, true, true}"
- debug_log "last_node: #{last_node.name}##{last_node.id} children: #{serialize_els last_node.children, true, true}"
- debug_log "tree: #{serialize_els doc.children, false, true}"
-
- debug_log "insert"
-
-
# can't use standard insert token thing, because it's already in
# open_els and must stay at it's current position in open_els
dest = adjusted_insertion_location ca
dest[0].children.splice dest[1], 0, last_node
last_node.parent = dest[0]
-
-
- debug_log "ca: #{ca.name}##{ca.id} children: #{serialize_els ca.children, true, true}"
- debug_log "fe: #{fe.name}##{fe.id} children: #{serialize_els fe.children, true, true}"
- debug_log "fb: #{fb.name}##{fb.id} children: #{serialize_els fb.children, true, true}"
- debug_log "last_node: #{last_node.name}##{last_node.id} children: #{serialize_els last_node.children, true, true}"
- debug_log "tree: #{serialize_els doc.children, false, true}"
-
# 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
open_els.splice i, 0, new_element
break
# 20. Jump back to the step labeled outer loop.
- debug_log "done wrapping fb's children. new_element: #{new_element.name}##{new_element.id}"
- debug_log "tree: #{serialize_els doc.children, false, true}"
- debug_log "open_els: #{serialize_els open_els, true, true}"
- debug_log "afe: #{serialize_els afe, true, true}"
- debug_log "AAA DONE"
+ return
# http://www.w3.org/TR/html5/syntax.html#close-a-p-element
close_p_element = ->
generate_implied_end_tags 'p' # arg is exception
- if open_els[0].name isnt 'p'
+ unless open_els[0].name is 'p' and open_els[0].namespace is NS_HTML
parse_error()
while open_els.length > 1 # just in case
el = open_els.shift()
- if el.name is 'p'
+ if el.name is 'p' and el.namespace is NS_HTML
return
+ return
close_p_if_in_button_scope = ->
- if is_in_button_scope 'p'
+ if is_in_button_scope 'p', NS_HTML
close_p_element()
+ return
# http://www.w3.org/TR/html5/syntax.html#insert-a-character
# aka insert_a_character = (t) ->
prev.text += t.text
return
dest[0].children.splice dest[1], 0, t
+ return
+
+ # 8.2.5 http://www.w3.org/TR/html5/syntax.html#tree-construction
+ process_token = (t) ->
+ acn = adjusted_current_node()
+ unless acn?
+ ins_mode t
+ return
+ if acn.namespace is NS_HTML
+ ins_mode t
+ return
+ if is_mathml_text_integration_point(acn)
+ if t.type is TYPE_START_TAG and not (t.name is 'mglyph' or t.name is 'malignmark')
+ ins_mode t
+ return
+ if t.type is TYPE_TEXT
+ ins_mode t
+ return
+ if acn.namespace is NS_MATHML and acn.name is 'annotation-xml' and t.type is TYPE_START_TAG and t.name is 'svg'
+ ins_mode t
+ return
+ if is_html_integration acn
+ if t.type is TYPE_START_TAG or t.type is TYPE_TEXT
+ ins_mode t
+ return
+ if t.type is TYPE_EOF
+ ins_mode t
+ return
+ in_foreign_content t
+ return
# 8.2.5.1
# http://www.w3.org/TR/html5/syntax.html#creating-and-inserting-nodes
# If foster parenting is enabled and target is a table, tbody, tfoot,
# thead, or tr element Foster parenting happens when content is
# misnested in tables.
- if flag_foster_parenting and foster_parenting_targets[target.name]
+ if flag_foster_parenting and foster_parenting_targets[target.name] is target.namespace
loop # once. this is here so we can ``break`` to "abort these substeps"
# 1. Let last template be the last template element in the
# stack of open elements, if any.
last_template = null
last_template_i = null
for el, i in open_els
- if el.name is 'template'
+ 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, token: t
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) ->
+ return insert_foreign_element token, NS_HTML
# http://www.w3.org/TR/html5/syntax.html#insert-a-comment
# position should be [node, index_within_children]
insert_comment = (t, position = null) ->
position ?= adjusted_insertion_location()
position[0].children.splice position[1], 0, t
+ return
# 8.2.5.2
# http://www.w3.org/TR/html5/syntax.html#generic-raw-text-element-parsing-algorithm
parse_generic_raw_text = (t) ->
insert_html_element t
tok_state = tok_state_rawtext
- original_insertion_mode = insertion_mode
- insertion_mode = ins_mode_text
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_text
+ return
parse_generic_rcdata_text = (t) ->
insert_html_element t
tok_state = tok_state_rcdata
- original_insertion_mode = insertion_mode
- insertion_mode = ins_mode_text
+ original_ins_mode = ins_mode
+ ins_mode = ins_mode_text
+ return
# 8.2.5.3 http://www.w3.org/TR/html5/syntax.html#closing-elements-that-have-implied-end-tags
# http://www.w3.org/TR/html5/syntax.html#generate-implied-end-tags
generate_implied_end_tags = (except = null) ->
- while end_tag_implied[open_els[0].name] and open_els[0].name isnt except
+ while end_tag_implied[open_els[0].name] is open_els[0].namespace and open_els[0].name isnt except
open_els.shift()
+ return
# 8.2.5.4 The rules for parsing tokens in HTML content
# http://www.w3.org/TR/html5/syntax.html#parsing-main-inhtml
# 8.2.5.4.1 The "initial" insertion mode
# http://www.w3.org/TR/html5/syntax.html#the-initial-insertion-mode
+ is_quirks_yes_doctype = (t) ->
+ if t.flag 'force-quirks'
+ return true
+ if t.name isnt 'html'
+ return true
+ if t.public_identifier?
+ pi = t.public_identifier.toLowerCase()
+ for p in quirks_yes_pi_prefixes
+ if pi.substr(0, p.length) is p
+ return true
+ if pi is '-//w3o//dtd w3 html strict 3.0//en//' or pi is '-/w3c/dtd html 4.0 transitional/en' or pi is 'html'
+ return true
+ if t.system_identifier?
+ if t.system_identifier.toLowerCase() is 'http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd'
+ return true
+ else if t.public_identifier?
+ # already did this: pi = t.public_identifier.toLowerCase()
+ if pi.substr(0, 32) is '-//w3c//dtd html 4.01 frameset//' or pi.substr(0, 36) is '-//w3c//dtd html 4.01 transitional//'
+ return true
+ return false
+ is_quirks_limited_doctype = (t) ->
+ if t.public_identifier?
+ pi = t.public_identifier.toLowerCase()
+ if pi.substr(0, 32) is '-//w3c//dtd xhtml 1.0 frameset//' or pi.substr(0, 36) is '-//w3c//dtd xhtml 1.0 transitional//'
+ return true
+ if t.system_identifier?
+ if pi.substr(0, 32) is '-//w3c//dtd html 4.01 frameset//' or pi.substr(0, 36) is '-//w3c//dtd html 4.01 transitional//'
+ return true
+ return false
ins_mode_initial = (t) ->
if is_space_tok t
return
if t.type is TYPE_COMMENT
- # fixfull 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
+ # fixfull syntax error from first paragraph and following bullets
+ # fixfull set doc.doctype
+ # fixfull is the "not an iframe srcdoc" thing relevant?
+ if is_quirks_yes_doctype t
+ doc.flag 'quirks mode', QUIRKS_YES
+ else if is_quirks_limited_doctype t
+ doc.flag 'quirks mode', QUIRKS_LIMITED
doc.children.push t
- 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
+ # fixfull not iframe srcdoc?
+ parse_error()
+ doc.flag 'quirks mode', QUIRKS_YES
+ ins_mode = ins_mode_before_html
+ process_token t
return
# 8.2.5.4.2 http://www.w3.org/TR/html5/syntax.html#the-before-html-insertion-mode
return
if t.type is TYPE_START_TAG and t.name is 'html'
el = token_to_element t, NS_HTML, doc
+ doc.children.push el
+ el.document = doc
open_els.unshift(el)
# fixfull (big paragraph in spec about manifest, fragment, urls, etc)
- 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'
parse_error()
return
# Anything else
- html_tok = new_open_tag 'html'
- el = token_to_element html_tok, NS_HTML, doc
+ el = token_to_element new_open_tag('html'), NS_HTML, doc
doc.children.push el
+ el.document = doc
open_els.unshift el
# ?fixfull browsing context
- 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
+ return
if t.type is TYPE_END_TAG
if t.name is 'head' or t.name is 'body' or t.name is 'html' or t.name is 'br'
# fall through to Anything else below
parse_error()
return
# Anything else
- head_tok = new_open_tag 'head'
- el = insert_html_element head_tok
+ el = insert_html_element new_open_tag 'head'
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
+ return
# 8.2.5.4.4 http://www.w3.org/TR/html5/syntax.html#parsing-main-inhead
ins_mode_in_head_else = (t) -> # factored out for same-as-spec flow control
open_els.shift() # spec says this will be a 'head' node
- insertion_mode = ins_mode_after_head
- insertion_mode t
+ ins_mode = ins_mode_after_head
+ process_token t
+ return
ins_mode_in_head = (t) ->
if t.type is TYPE_TEXT and (t.text is "\t" or t.text is "\n" or t.text is "\u000c" or t.text is ' ')
insert_character t
if t.type is TYPE_START_TAG and t.name is 'title'
parse_generic_rcdata_text t
return
- if t.type is TYPE_START_TAG and ((t.name is 'noscript' and flag_scripting) or (t.name is 'noframes' or t.name is 'style'))
+ 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()
parse_error()
loop
el = open_els.shift()
- if el.name is 'template'
+ if el.name is 'template' and el.namespace is NS_HTML
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
+ return
# 8.2.5.4.5 http://www.w3.org/TR/html5/syntax.html#parsing-main-inheadnoscript
+ ins_mode_in_head_noscript_else = (t) ->
+ parse_error()
+ open_els.shift()
+ ins_mode = ins_mode_in_head
+ process_token t
+ return
ins_mode_in_head_noscript = (t) ->
- # 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 and t.name is 'html'
+ ins_mode_in_body t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'noscript'
+ open_els.shift()
+ ins_mode = ins_mode_in_head
+ return
+ if is_space_tok(t) or t.type is TYPE_COMMENT or (t.type is TYPE_START_TAG and (t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'style'))
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'br'
+ ins_mode_in_head_noscript_else t
+ return
+ if (t.type is TYPE_START_TAG and (t.name is 'head' or t.name is 'noscript')) or t.type is TYPE_END_TAG
+ parse_error()
+ return
+ # Anything else
+ ins_mode_in_head_noscript_else t
+ return
# 8.2.5.4.6 http://www.w3.org/TR/html5/syntax.html#the-after-head-insertion-mode
ins_mode_after_head_else = (t) ->
body_tok = new_open_tag 'body'
insert_html_element body_tok
- 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()
open_els.unshift head_element_pointer
ins_mode_in_head t
- for el, i of open_els
+ for el, i in open_els
if el is head_element_pointer
open_els.splice i, 1
return
- console.log "warning: 23904 couldn't find head element in open_els"
return
if t.type is TYPE_END_TAG and t.name is 'template'
ins_mode_in_head t
return
# Anything else
ins_mode_after_head_else t
+ return
# 8.2.5.4.7 http://www.w3.org/TR/html5/syntax.html#parsing-main-inbody
in_body_any_other_end_tag = (name) -> # factored out because adoption agency calls it
- for node, i in open_els
- if node.name is name # FIXME check namespace too
+ node = open_els[0]
+ loop
+ if node.name is name and node.namespace is NS_HTML
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
+ unless node is open_els[0]
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el is node
+ return
+ if special_elements[node.name] is node.namespace
parse_error()
return
- 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
- 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
+ for el, i in open_els
+ if node is el
+ node = open_els[i + 1]
+ break
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 in 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
+ if (t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'script' or t.name is 'style' or t.name is 'template' or t.name is 'title')) or (t.type is TYPE_END_TAG and t.name is 'template')
+ ins_mode_in_head t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'body'
+ parse_error()
+ return if open_els.length < 2
+ second = open_els[open_els.length - 2]
+ return unless second.namespace is NS_HTML
+ return unless second.name is 'body'
+ return if template_tag_is_open()
+ flag_frameset_ok = false
+ for a in t.attrs_a
+ second.attrs[a[0]] = a[1] unless second.attrs[a[0]]?
+ return
+ if t.type is TYPE_START_TAG and t.name is 'frameset'
+ parse_error()
+ return if open_els.length < 2
+ second_i = open_els.length - 2
+ second = open_els[second_i]
+ return unless second.namespace is NS_HTML
+ return unless second.name is 'body'
+ if flag_frameset_ok is false
+ return
+ if second.parent?
+ for el, i in second.parent.children
+ if el is second
+ second.parent.children.splice i, 1
+ break
+ open_els.splice second_i, 1
+ # pop everything except the "root html element"
+ while open_els.length > 1
+ open_els.shift()
+ insert_html_element t
+ ins_mode = ins_mode_in_frameset
+ return
+ if t.type is TYPE_EOF
+ ok_tags = {
+ dd:NS_HTML, dt:NS_HTML, li:NS_HTML, p:NS_HTML, tbody:NS_HTML,
+ td:NS_HTML, tfoot:NS_HTML, th:NS_HTML, thead:NS_HTML,
+ tr:NS_HTML, body:NS_HTML, html:NS_HTML,
+ }
+ for el in open_els
+ unless ok_tags[t.name] is el.namespace
+ parse_error()
+ break
+ if template_ins_modes.length > 0
+ ins_mode_in_template t
+ else
+ stop_parsing()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'body'
+ unless is_in_scope 'body', NS_HTML
parse_error()
- when TYPE_START_TAG
- switch t.name
- when 'caption'
- clear_stack_to_table_context()
- afe_push_marker()
+ return
+ ok_tags = {
+ dd:NS_HTML, dt:NS_HTML, li:NS_HTML, optgroup:NS_HTML,
+ option:NS_HTML, p:NS_HTML, rb:NS_HTML, rp:NS_HTML, rt:NS_HTML,
+ rtc:NS_HTML, tbody:NS_HTML, td:NS_HTML, tfoot:NS_HTML,
+ th:NS_HTML, thead:NS_HTML, tr:NS_HTML, body:NS_HTML,
+ html:NS_HTML
+ }
+ for el in open_els
+ unless ok_tags[t.name] is el.namespace
+ parse_error()
+ break
+ ins_mode = ins_mode_after_body
+ return
+ if t.type is TYPE_END_TAG and t.name is 'html'
+ unless is_in_scope 'body', NS_HTML
+ parse_error()
+ return
+ ok_tags = {
+ dd:NS_HTML, dt:NS_HTML, li:NS_HTML, optgroup:NS_HTML,
+ option:NS_HTML, p:NS_HTML, rb:NS_HTML, rp:NS_HTML, rt:NS_HTML,
+ rtc:NS_HTML, tbody:NS_HTML, td:NS_HTML, tfoot:NS_HTML,
+ th:NS_HTML, thead:NS_HTML, tr:NS_HTML, body:NS_HTML,
+ html:NS_HTML
+ }
+ for el in open_els
+ unless ok_tags[t.name] is el.namespace
+ parse_error()
+ break
+ ins_mode = ins_mode_after_body
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'address' or t.name is 'article' or t.name is 'aside' or t.name is 'blockquote' or t.name is 'center' or t.name is 'details' or t.name is 'dialog' or t.name is 'dir' or t.name is 'div' or t.name is 'dl' or t.name is 'fieldset' or t.name is 'figcaption' or t.name is 'figure' or t.name is 'footer' or t.name is 'header' or t.name is 'hgroup' or t.name is 'main' or t.name is 'nav' or t.name is 'ol' or t.name is 'p' or t.name is 'section' or t.name is 'summary' or t.name is 'ul')
+ close_p_if_in_button_scope()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and h_tags[t.name]?
+ close_p_if_in_button_scope()
+ if h_tags[open_els[0].name] is open_els[0].namespace
+ parse_error()
+ open_els.shift()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'pre' or t.name is 'listing')
+ close_p_if_in_button_scope()
+ insert_html_element t
+ eat_next_token_if_newline()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and t.name is 'form'
+ unless form_element_pointer is null or template_tag_is_open()
+ parse_error()
+ return
+ close_p_if_in_button_scope()
+ el = insert_html_element t
+ unless template_tag_is_open()
+ form_element_pointer = el
+ return
+ if t.type is TYPE_START_TAG and t.name is 'li'
+ flag_frameset_ok = false
+ for node in open_els
+ if node.name is 'li' and node.namespace is NS_HTML
+ generate_implied_end_tags 'li' # arg is exception
+ if open_els[0].name isnt 'li' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'li' and el.namespace is NS_HTML
+ break
+ break
+ if el_is_special_not_adp node
+ break
+ close_p_if_in_button_scope()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'dd' or t.name is 'dt')
+ flag_frameset_ok = false
+ for node in open_els
+ if node.name is 'dd' and node.namespace is NS_HTML
+ generate_implied_end_tags 'dd' # arg is exception
+ if open_els[0].name isnt 'dd' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'dd' and el.namespace is NS_HTML
+ break
+ break
+ if node.name is 'dt' and node.namespace is NS_HTML
+ generate_implied_end_tags 'dt' # arg is exception
+ if open_els[0].name isnt 'dt' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'dt' and el.namespace is NS_HTML
+ break
+ break
+ if el_is_special_not_adp node
+ break
+ close_p_if_in_button_scope()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'plaintext'
+ close_p_if_in_button_scope()
+ insert_html_element t
+ tok_state = tok_state_plaintext
+ return
+ if t.type is TYPE_START_TAG and t.name is 'button'
+ if is_in_scope 'button', NS_HTML
+ parse_error()
+ generate_implied_end_tags()
+ loop
+ el = open_els.shift()
+ if el.name is 'button' and el.namespace is NS_HTML
+ break
+ reconstruct_afe()
+ insert_html_element t
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'address' or t.name is 'article' or t.name is 'aside' or t.name is 'blockquote' or t.name is 'button' or t.name is 'center' or t.name is 'details' or t.name is 'dialog' or t.name is 'dir' or t.name is 'div' or t.name is 'dl' or t.name is 'fieldset' or t.name is 'figcaption' or t.name is 'figure' or t.name is 'footer' or t.name is 'header' or t.name is 'hgroup' or t.name is 'listing' or t.name is 'main' or t.name is 'nav' or t.name is 'ol' or t.name is 'pre' or t.name is 'section' or t.name is 'summary' or t.name is 'ul')
+ unless is_in_scope t.name, NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags()
+ unless open_els[0].name is t.name and open_els[0].namespace is NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is t.name and el.namespace is NS_HTML
+ return
+ return
+ if t.type is TYPE_END_TAG and t.name is 'form'
+ unless template_tag_is_open()
+ node = form_element_pointer
+ form_element_pointer = null
+ if node is null or not el_is_in_scope node
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0] isnt node
+ parse_error()
+ for el, i in open_els
+ if el is node
+ open_els.splice i, 1
+ break
+ else
+ unless is_in_scope 'form', NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0].name isnt 'form' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'form' and el.namespace is NS_HTML
+ break
+ return
+ if t.type is TYPE_END_TAG and t.name is 'p'
+ unless is_in_button_scope 'p', NS_HTML
+ parse_error()
+ insert_html_element new_open_tag 'p'
+ close_p_element()
+ return
+ if t.type is TYPE_END_TAG and t.name is 'li'
+ unless is_in_li_scope 'li', NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags 'li' # arg is exception
+ if open_els[0].name isnt 'li' or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is 'li' and el.namespace is NS_HTML
+ break
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'dd' or t.name is 'dt')
+ unless is_in_scope t.name, NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags t.name # arg is exception
+ if open_els[0].name isnt t.name or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is t.name and el.namespace is NS_HTML
+ break
+ return
+ if t.type is TYPE_END_TAG and h_tags[t.name]?
+ h_in_scope = false
+ for el in open_els
+ if h_tags[el.name] is el.namespace
+ h_in_scope = true
+ break
+ if standard_scopers[el.name] is el.namespace
+ break
+ unless h_in_scope
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0].name isnt t.name or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if h_tags[el.name] is el.namespace
+ break
+ return
+ # deep breath!
+ if t.type is TYPE_START_TAG and t.name is 'a'
+ # If the list of active formatting elements contains an a element
+ # between the end of the list and the last marker on the list (or
+ # the start of the list if there is no marker on the list), then
+ # this is a parse error; run the adoption agency algorithm for the
+ # tag name "a", then remove that element from the list of active
+ # formatting elements and the stack of open elements if the
+ # adoption agency algorithm didn't already remove it (it might not
+ # have if the element is not in table scope).
+ found = false
+ for el in afe
+ if el.type is TYPE_AFE_MARKER
+ break
+ if el.name is 'a' and el.namespace is NS_HTML
+ found = el
+ if found?
+ parse_error()
+ adoption_agency 'a'
+ for el, i in afe
+ if el is found
+ afe.splice i, 1
+ for el, i in open_els
+ if el is found
+ open_els.splice i, 1
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'b' or t.name is 'big' or t.name is 'code' or t.name is 'em' or t.name is 'font' or t.name is 'i' or t.name is 's' or t.name is 'small' or t.name is 'strike' or t.name is 'strong' or t.name is 'tt' or t.name is 'u')
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
+ return
+ if t.type is TYPE_START_TAG and t.name is 'nobr'
+ reconstruct_afe()
+ if is_in_scope 'nobr', NS_HTML
+ parse_error()
+ adoption_agency 'nobr'
+ reconstruct_afe()
+ el = insert_html_element t
+ afe_push el
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'a' or t.name is 'b' or t.name is 'big' or t.name is 'code' or t.name is 'em' or t.name is 'font' or t.name is 'i' or t.name is 'nobr' or t.name is 's' or t.name is 'small' or t.name is 'strike' or t.name is 'strong' or t.name is 'tt' or t.name is 'u')
+ adoption_agency t.name
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'applet' or t.name is 'marquee' or t.name is 'object')
+ reconstruct_afe()
+ insert_html_element t
+ afe_push_marker()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_END_TAG and (t.name is 'applet' or t.name is 'marquee' or t.name is 'object')
+ unless is_in_scope t.name, NS_HTML
+ parse_error()
+ return
+ generate_implied_end_tags()
+ if open_els[0].name isnt t.name or open_els[0].namespace isnt NS_HTML
+ parse_error()
+ loop
+ el = open_els.shift()
+ if el.name is t.name and el.namespace is NS_HTML
+ break
+ clear_afe_to_marker()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'table'
+ unless doc.flag('quirks mode') is QUIRKS_YES
+ close_p_if_in_button_scope() # test
+ insert_html_element t
+ flag_frameset_ok = false
+ ins_mode = ins_mode_in_table
+ return
+ if t.type is TYPE_END_TAG and t.name is 'br'
+ parse_error()
+ # W3C: t.type = TYPE_START_TAG
+ t = new_open_tag 'br' # WHATWG
+ # fall through
+ if t.type is TYPE_START_TAG and (t.name is 'area' or t.name is 'br' or t.name is 'embed' or t.name is 'img' or t.name is 'keygen' or t.name is 'wbr')
+ reconstruct_afe()
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and t.name is 'input'
+ reconstruct_afe()
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ unless is_input_hidden_tok t
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'menuitem' or t.name is 'param' or t.name is 'source' or t.name is 'track')
+ # WHATWG adds 'menuitem' for this block
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'hr'
+ close_p_if_in_button_scope()
+ insert_html_element t
+ open_els.shift()
+ t.acknowledge_self_closing()
+ flag_frameset_ok = false
+ return
+ if t.type is TYPE_START_TAG and t.name is 'image'
+ parse_error()
+ t.name = 'img'
+ process_token t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'isindex'
+ parse_error()
+ if template_tag_is_open() is false and form_element_pointer isnt null
+ return
+ t.acknowledge_self_closing()
+ flag_frameset_ok = false
+ close_p_if_in_button_scope()
+ el = insert_html_element new_open_tag 'form'
+ unless template_tag_is_open()
+ form_element_pointer = el
+ for a in t.attrs_a
+ if a[0] is 'action'
+ el.attrs['action'] = a[1]
+ break
+ insert_html_element new_open_tag 'hr'
+ open_els.shift()
+ reconstruct_afe()
+ insert_html_element new_open_tag 'label'
+ # note: this is a little out-of-spec-order so we only have to scan t.attrs_a once
+ input_el = new_open_tag 'input'
+ prompt = null
+ for a in t.attrs_a
+ if a[0] is 'prompt'
+ prompt = a[1]
+ if a[0] isnt 'name' and a[0] isnt 'action' and a[0] isnt 'prompt'
+ input_el.attrs_a.push [a[0], a[1]]
+ input_el.attrs_a.push ['name', 'isindex']
+ # fixfull this next bit is in english... internationalize?
+ prompt ?= "This is a searchable index. Enter search keywords: "
+ insert_character new_character_token prompt # fixfull split
+ # TODO submit typo "balue" in spec
+ insert_html_element input_el
+ open_els.shift()
+ # insert_character '' # you can put chars here if promt attr missing
+ open_els.shift()
+ insert_html_element new_open_tag 'hr'
+ open_els.shift()
+ open_els.shift()
+ unless template_tag_is_open()
+ form_element_pointer = null
+ return
+ if t.type is TYPE_START_TAG and t.name is 'textarea'
+ insert_html_element t
+ eat_next_token_if_newline()
+ tok_state = tok_state_rcdata
+ original_ins_mode = ins_mode
+ flag_frameset_ok = false
+ ins_mode = ins_mode_text
+ return
+ if t.type is TYPE_START_TAG and t.name is 'xmp'
+ close_p_if_in_button_scope()
+ reconstruct_afe()
+ flag_frameset_ok = false
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'iframe'
+ flag_frameset_ok = false
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'noembed' or (t.name is 'noscript' and flag_scripting))
+ parse_generic_raw_text t
+ return
+ if t.type is TYPE_START_TAG and t.name is 'select'
+ reconstruct_afe()
+ insert_html_element t
+ flag_frameset_ok = false
+ if ins_mode is ins_mode_in_table or ins_mode is ins_mode_in_caption or ins_mode is ins_mode_in_table_body or ins_mode is ins_mode_in_row or ins_mode is ins_mode_in_cell
+ ins_mode = ins_mode_in_select_in_table
+ else
+ ins_mode = ins_mode_in_select
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'optgroup' or t.name is 'option')
+ if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML
+ open_els.shift()
+ reconstruct_afe()
+ insert_html_element t
+ return
+# this comment block implements the W3C spec
+# if t.type is TYPE_START_TAG and (t.name is 'rb' or t.name is 'rp' or t.name is 'rtc')
+# if is_in_scope 'ruby', NS_HTML
+# generate_implied_end_tags()
+# unless open_els[0].name is 'ruby' and open_els[0].namespace is NS_HTML
+# parse_error()
+# insert_html_element t
+# return
+# if t.type is TYPE_START_TAG and t.name is 'rt'
+# if is_in_scope 'ruby', NS_HTML
+# generate_implied_end_tags 'rtc' # arg is exception
+# unless (open_els[0].name is 'ruby' or open_els[0].name is 'rtc') and open_els[0].namespace is NS_HTML
+# parse_error()
+# insert_html_element t
+# return
+# below implements the WHATWG spec https://html.spec.whatwg.org/multipage/syntax.html#parsing-main-inbody
+ if t.type is TYPE_START_TAG and (t.name is 'rb' or t.name is 'rtc')
+ if is_in_scope 'ruby', NS_HTML
+ generate_implied_end_tags()
+ unless open_els[0].name is 'ruby' and open_els[0].namespace is NS_HTML
+ parse_error()
+ insert_html_element t
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'rp' or t.name is 'rt')
+ if is_in_scope 'ruby', NS_HTML
+ generate_implied_end_tags 'rtc'
+ unless (open_els[0].name is 'ruby' or open_els[0].name is 'rtc') and open_els[0].namespace is NS_HTML
+ parse_error()
+ insert_html_element t
+ return
+# end WHATWG chunk
+ if t.type is TYPE_START_TAG and t.name is 'math'
+ reconstruct_afe()
+ adjust_mathml_attributes t
+ adjust_foreign_attributes t
+ insert_foreign_element t, NS_MATHML
+ if t.flag 'self-closing'
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and t.name is 'svg'
+ reconstruct_afe()
+ adjust_svg_attributes t
+ adjust_foreign_attributes t
+ insert_foreign_element t, NS_SVG
+ if t.flag 'self-closing'
+ open_els.shift()
+ t.acknowledge_self_closing()
+ return
+ if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'frame' or t.name is 'head' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')
+ parse_error()
+ return
+ if t.type is TYPE_START_TAG # any other start tag
+ reconstruct_afe()
+ insert_html_element t
+ return
+ if t.type is TYPE_END_TAG # any other end tag
+ in_body_any_other_end_tag t.name
+ return
+ return
+
+ # 8.2.5.4.8 http://www.w3.org/TR/html5/syntax.html#parsing-main-incdata
+ ins_mode_text = (t) ->
+ if t.type is TYPE_TEXT
+ insert_character t
+ return
+ if t.type is TYPE_EOF
+ parse_error()
+ if open_els[0].name is 'script' and open_els[0].namespace is NS_HTML
+ open_els[0].flag 'already started', true
+ open_els.shift()
+ ins_mode = original_ins_mode
+ process_token t
+ return
+ if t.type is TYPE_END_TAG and t.name is 'script'
+ open_els.shift()
+ ins_mode = original_ins_mode
+ # fixfull the spec seems to assume that I'm going to run the script
+ # http://www.w3.org/TR/html5/syntax.html#scriptEndTag
+ return
+ if t.type is TYPE_END_TAG
+ open_els.shift()
+ ins_mode = original_ins_mode
+ return
+ return
+
+ # 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
+ ins_mode_in_table = (t) ->
+ switch t.type
+ when TYPE_TEXT
+ if (open_els[0].name is 'table' or open_els[0].name is 'tbody' or open_els[0].name is 'tfoot' or open_els[0].name is 'thead' or open_els[0].name is 'tr') and open_els[0].namespace is NS_HTML
+ pending_table_character_tokens = []
+ 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
- insertion_mode = ins_mode_in_caption
+ ins_mode = ins_mode_in_caption
when 'colgroup'
clear_stack_to_table_context()
insert_html_element t
- insertion_mode = ins_mode_in_column_group
+ ins_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
+ ins_mode = ins_mode_in_column_group
+ process_token t
when 'tbody', 'tfoot', 'thead'
clear_stack_to_table_context()
insert_html_element t
- insertion_mode = ins_mode_in_table_body
+ ins_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
+ ins_mode = ins_mode_in_table_body
+ process_token t
when 'table'
parse_error()
- if is_in_table_scope 'table'
+ if is_in_table_scope 'table', NS_HTML
loop
el = open_els.shift()
- if el.name is 'table'
+ if el.name is 'table' and el.namespace is NS_HTML
break
- reset_insertion_mode()
- insertion_mode t
+ reset_ins_mode()
+ process_token t
when 'style', 'script', 'template'
ins_mode_in_head t
when 'input'
- if token_is_input_hidden t
+ unless is_input_hidden_tok t
ins_mode_in_table_else t
else
parse_error()
when TYPE_END_TAG
switch t.name
when 'table'
- if is_in_table_scope 'table'
+ if is_in_table_scope 'table', NS_HTML
loop
el = open_els.shift()
- if el.name is 'table'
+ if el.name is 'table' and el.namespace is NS_HTML
break
- reset_insertion_mode()
+ reset_ins_mode()
else
- parse_error
+ parse_error()
when 'body', 'caption', 'col', 'colgroup', 'html', 'tbody', 'td', 'tfoot', 'th', 'thead', 'tr'
parse_error()
when 'template'
ins_mode_in_body t
else
ins_mode_in_table_else t
+ return
# 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
+ # from javascript?
parse_error()
return
if t.type is TYPE_TEXT
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
+ ins_mode_in_table_else old
+ pending_table_character_tokens = []
+ ins_mode = original_ins_mode
+ process_token t
+ return
# 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'
+ if is_in_table_scope 'caption', NS_HTML
generate_implied_end_tags()
if open_els[0].name isnt 'caption'
parse_error()
loop
el = open_els.shift()
- if el.name is 'caption'
+ if el.name is 'caption' and el.namespace is NS_HTML
break
clear_afe_to_marker()
- insertion_mode = in_table
+ ins_mode = ins_mode_in_table
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'
+ if is_in_table_scope 'caption', NS_HTML
loop
el = open_els.shift()
- if el.name is 'caption'
+ if el.name is 'caption' and el.namespace is NS_HTML
break
clear_afe_to_marker()
- insertion_mode = in_table
- insertion_mode t
+ 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')
return
# Anything else
ins_mode_in_body t
+ return
# 8.2.5.4.12 http://www.w3.org/TR/html5/syntax.html#parsing-main-incolgroup
ins_mode_in_column_group = (t) ->
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[0].shift()
- insertion_mode = ins_mode_in_table
+ if open_els[0].name is 'colgroup' and open_els.namespace is NS_HTML
+ open_els.shift()
+ ins_mode = ins_mode_in_table
else
parse_error()
return
parse_error()
return
open_els.shift()
- insertion_mode = ins_mode_in_table
- insertion_mode t
+ 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
if t.type is TYPE_START_TAG and t.name is 'tr'
clear_stack_to_table_body_context()
insert_html_element t
- insertion_mode = ins_mode_in_row
+ 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'
- insertion_mode = ins_mode_in_row
- insertion_mode t
+ 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
+ unless is_in_table_scope t.name, NS_HTML
parse_error()
return
clear_stack_to_table_body_context()
open_els.shift()
- insertion_mode = ins_mode_in_table
+ 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'
+ if el.namespace is NS_HTML and (el.name is 'tbody' or el.name is 'tfoot' or el.name is 'thead')
has = true
break
- if table_scopers[el.name]
+ if table_scopers[el.name] is el.namespace
break
if !has
parse_error()
return
clear_stack_to_table_body_context()
open_els.shift()
- insertion_mode = ins_mode_in_table
- insertion_mode t
+ 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
+ return
# 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
- insertion_mode = ins_mode_in_cell
+ 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'
+ if is_in_table_scope 'tr', NS_HTML
clear_stack_to_table_row_context()
open_els.shift()
- insertion_mode = ins_mode_in_table_body
+ 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'
+ if is_in_table_scope 'tr', NS_HTML
clear_stack_to_table_row_context()
open_els.shift()
- insertion_mode = ins_mode_in_table_body
- insertion_mode t
+ 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'
+ if is_in_table_scope t.name, NS_HTML
+ if is_in_table_scope 'tr', NS_HTML
clear_stack_to_table_row_context()
open_els.shift()
- insertion_mode = ins_mode_in_table_body
- insertion_mode t
+ ins_mode = ins_mode_in_table_body
+ process_token t
else
parse_error()
return
return
# Anything else
ins_mode_in_table t
+ return
# 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'
+ unless (open_els[0].name is 'td' or open_els[0] is 'th') and open_els[0].namespace is NS_HTML
parse_error()
loop
el = open_els.shift()
- if el.name is 'td' or el.name is 'th'
+ if el.namespace is NS_HTML and (el.name is 'td' or el.name is 'th')
break
clear_afe_to_marker()
- insertion_mode = ins_mode_in_row
+ ins_mode = ins_mode_in_row
+ return
# 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 is_in_table_scope t.name, NS_HTML
generate_implied_end_tags()
- if open_els[0].name isnt t.name
- parse_error
+ 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
+ if el.name is t.name and el.namespace is NS_HTML
break
clear_afe_to_marker()
- insertion_mode = ins_mode_in_row
+ 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'
+ if el.namespace is NS_HTML and (el.name is 'td' or el.name is 'th')
has = true
break
- if table_scopers[el.name]
+ if table_scopers[el.name] is el.namespace
break
if !has
parse_error()
return
close_the_cell()
- insertion_mode t
+ 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
+ if is_in_table_scope t.name, NS_HTML
close_the_cell()
- insertion_mode t
+ process_token t
else
parse_error()
return
# Anything Else
ins_mode_in_body t
+ return
# 8.2.5.4.16 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselect
ins_mode_in_select = (t) ->
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'
+ if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML
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'
+ 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'
+ if open_els[0].name is 'optgroup' and open_els[0].namespace is NS_HTML
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'
+ if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML
+ if open_els[1].name is 'optgroup' and open_els[0].namespace is NS_HTML
+ open_els.shift()
+ if open_els[0].name is 'optgroup' and open_els[0].namespace is NS_HTML
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'
+ if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML
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'
+ if is_in_select_scope 'select', NS_HTML
loop
el = open_els.shift()
- if el.name is 'select'
+ if el.name is 'select' and el.namespace is NS_HTML
break
- reset_insertion_mode()
+ reset_ins_mode()
else
parse_error()
return
parse_error()
loop
el = open_els.shift()
- if el.name is 'select'
+ if el.name is 'select' and el.namespace is NS_HTML
break
- reset_insertion_mode()
+ 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'
+ unless is_in_select_scope 'select', NS_HTML
return
loop
el = open_els.shift()
- if el.name is 'select'
+ if el.name is 'select' and el.namespace is NS_HTML
break
- reset_insertion_mode()
- insertion_mode t
+ 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
parse_error()
loop
el = open_els.shift()
- if el.name is 'select'
+ if el.name is 'select' and el.namespace is NS_HTML
break
- reset_insertion_mode()
- insertion_mode t
+ 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()
return
loop
el = open_els.shift()
- if el.name is 'select'
+ if el.name is 'select' and el.namespace is NS_HTML
break
- reset_insertion_mode()
- insertion_mode t
+ reset_ins_mode()
+ process_token t
return
# Anything else
ins_mode_in_select t
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_insertion_modes.shift()
- template_insertion_modes.unshift ins_mode_in_table
- insertion_mode = ins_mode_in_table
- insertion_mode t
+ 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_insertion_modes.shift()
- template_insertion_modes.unshift ins_mode_in_column_group
- insertion_mode = ins_mode_in_column_group
- insertion_mode t
+ 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_insertion_modes.shift()
- template_insertion_modes.unshift ins_mode_in_table_body
- insertion_mode = ins_mode_in_table_body
- insertion_mode t
+ 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_insertion_modes.shift()
- template_insertion_modes.unshift ins_mode_in_row
- insertion_mode = ins_mode_in_row
- insertion_mode t
+ 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_insertion_modes.shift()
- template_insertion_modes.unshift ins_mode_in_body
- insertion_mode = ins_mode_in_body
- insertion_mode t
+ 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 EOF
+ 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
+ if el.name is 'template' and el.namespace is NS_HTML
break
clear_afe_to_marker()
- template_insertion_modes.shift()
- reset_insertion_mode()
- insertion_mode t
+ template_ins_modes.shift()
+ reset_ins_mode()
+ process_token t
+ return
# 8.2.5.4.19 http://www.w3.org/TR/html5/syntax.html#parsing-main-afterbody
ins_mode_after_body = (t) ->
ins_mode_in_body t
return
if t.type is TYPE_COMMENT
- insert_comment t, [open_els[0], open_els[0].children.length]
+ first = open_els[open_els.length - 1]
+ insert_comment t, [first, first.children.length]
return
if t.type is TYPE_DOCTYPE
parse_error()
ins_mode_in_body t
return
if t.type is TYPE_END_TAG and t.name is 'html'
- # fixfull fragment case
- insertion_mode = ins_mode_after_after_body
+ if flag_fragment_parsing
+ parse_error()
+ return
+ ins_mode = ins_mode_after_after_body
return
if t.type is TYPE_EOF
stop_parsing()
return
# Anything ELse
parse_error()
- insertion_mode = ins_mode_in_body
- insertion_mode t
+ ins_mode = ins_mode_in_body
+ process_token t
+ return
# 8.2.5.4.20 http://www.w3.org/TR/html5/syntax.html#parsing-main-inframeset
ins_mode_in_frameset = (t) ->
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'
- insertion_mode = ins_mode_after_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'
+ 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()
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
+ ins_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
# Anything else
parse_error()
- insertion_mode = ins_mode_in_body
+ ins_mode = ins_mode_in_body
+ process_token t
return
# 8.2.5.4.23 http://www.w3.org/TR/html5/syntax.html#the-after-after-frameset-insertion-mode
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()
+ # fixfull
+ 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()
+ if is_mathml_text_integration_point(open_els[0]) or is_html_integration(open_els[0]) or open_els[0].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
+ i = 0
+ node = open_els[i]
+ if node.name.toLowerCase() isnt t.name
+ parse_error()
+ loop
+ 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
+ i += 1
+ node = open_els[i]
+ if node.namespace is NS_HTML
+ break
+ ins_mode t # explicitly call HTML insertion mode
+ return
# 8.2.4.1 http://www.w3.org/TR/html5/syntax.html#data-state
# 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 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 '<'
- return null
+ c = txt.charAt(cur++)
+ if c is '!'
+ tok_state = tok_state_markup_declaration_open
+ return
+ if c is '/'
+ tok_state = tok_state_end_tag_open
+ return
+ if is_uc_alpha(c)
+ tok_cur_tag = new_open_tag c.toLowerCase()
+ tok_state = tok_state_tag_name
+ return
+ if is_lc_alpha(c)
+ tok_cur_tag = new_open_tag c
+ tok_state = tok_state_tag_name
+ return
+ if c is '?'
+ parse_error()
+ tok_cur_tag = new_comment_token '?' # FIXME right?
+ tok_state = tok_state_bogus_comment
+ return
+ # Anything else
+ parse_error()
+ tok_state = tok_state_data
+ cur -= 1 # we didn't parse/handle the char after <
+ return new_text_node '<'
# 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 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_cur_tag = new_comment_token '/'
- tok_state = tok_state_bogus_comment
+ c = txt.charAt(cur++)
+ if is_uc_alpha(c)
+ tok_cur_tag = new_end_tag c.toLowerCase()
+ tok_state = tok_state_tag_name
+ return
+ if is_lc_alpha(c)
+ tok_cur_tag = new_end_tag c
+ tok_state = tok_state_tag_name
+ return
+ if c is '>'
+ parse_error()
+ tok_state = tok_state_data
+ return
+ if c is '' # EOF
+ parse_error()
+ tok_state = tok_state_data
+ return new_text_node '</'
+ # Anything else
+ parse_error()
+ tok_cur_tag = new_comment_token c
+ tok_state = tok_state_bogus_comment
return null
# 8.2.4.10 http://www.w3.org/TR/html5/syntax.html#tag-name-state
parse_error()
tok_state = tok_state_data
else
- if uc_alpha.indexOf(c) > -1
+ if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
else
tok_cur_tag.name += c
# 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 uc_alpha.indexOf(c) > -1
+ 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 lc_alpha.indexOf(c) > -1
+ if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
temporary_buffer += c
tok_state = tok_state_rcdata_end_tag_name
# 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}"
+ # fixfull: this assumes that open_els[0].name is "the tag name of the last
+ # start tag to have been emitted from this tokenizer"
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 = tok_state_data
return tok_cur_tag
# else fall through to "Anything else"
- if uc_alpha.indexOf(c) > -1
+ if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
temporary_buffer += c
return null
- if lc_alpha.indexOf(c) > -1
+ if is_lc_alpha(c)
tok_cur_tag.name += c
temporary_buffer += c
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 = ->
c = txt.charAt(cur++)
- if uc_alpha.indexOf(c) > -1
+ 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 lc_alpha.indexOf(c) > -1
+ if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
temporary_buffer += c
tok_state = tok_state_rawtext_end_tag_name
tok_state = tok_state_data
return tok_cur_tag
# else fall through to "Anything else"
- if uc_alpha.indexOf(c) > -1
+ if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
temporary_buffer += c
return null
- if lc_alpha.indexOf(c) > -1
+ if is_lc_alpha(c)
tok_cur_tag.name += c
temporary_buffer += c
return null
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
+ # 8.2.4.17 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 c is '>'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_data
+ return tok_cur_tag
+ # 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 '<'
+
+ # 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 c is '>'
+ if is_appropriate_end_tag tok_cur_tag
+ tok_state = tok_state_data
+ return tok_cur_tag
+ # 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 = ->
parse_error()
tok_state = tok_state_data
else
- if uc_alpha.indexOf(c) > -1
+ if is_uc_alpha(c)
attr_name = c.toLowerCase()
else
attr_name = c
return tmp
when "\u0000"
parse_error()
- tok_cur_tag.attrs_a[0][0] = "\ufffd"
+ tok_cur_tag.attrs_a[0][0] += "\ufffd"
when '"', "'", '<'
parse_error()
- tok_cur_tag.attrs_a[0][0] = c
+ 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()
+ 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
return
if c is '>'
tok_state = tok_state_data
- return
- if uc_alpha.indexOf(c) > -1
+ return tok_cur_tag
+ if is_uc_alpha(c)
tok_cur_tag.attrs_a.unshift [c.toLowerCase(), '']
tok_state = tok_state_attribute_name
return
# Anything else
tok_cur_tag.attrs_a.unshift [c, '']
tok_state = tok_state_attribute_name
+ return
# 8.2.4.37 http://www.w3.org/TR/html5/syntax.html#before-attribute-value-state
tok_state_before_attribute_value = ->
cur -= 1 # we didn't handle that char
return null
+ # 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', true
+ 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
+
# 8.2.4.44 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 = ->
else
val = txt.substr cur, (next_gt - cur)
cur = next_gt + 1
- val = val.replace "\u0000", "\ufffd"
+ val = val.replace(new RegExp("\u0000", 'g'), "\ufffd")
tok_cur_tag.text += val
tok_state = tok_state_data
return tok_cur_tag
tok_state = tok_state_doctype
return
acn = adjusted_current_node()
- if acn and acn.namespace isnt NS_HTML and text.substr(cur, 7) is '[CDATA['
+ 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_cur_tag = new_comment_token ''
tok_state = tok_state_bogus_comment
return
tok_state = tok_state_comment_start_dash
when "\u0000"
parse_error()
+ tok_state = tok_state_comment
return new_character_token "\ufffd"
when '>'
parse_error()
return tok_cur_tag
else
tok_cur_tag.text += c
+ tok_state = tok_state_comment
return null
# 8.2.4.47 http://www.w3.org/TR/html5/syntax.html#comment-start-dash-state
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
- if uc_alpha.indexOf(c) > -1
+ if is_uc_alpha(c)
tok_cur_tag = new_doctype_token c.toLowerCase()
tok_state = tok_state_doctype_name
return
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
- if uc_alpha.indexOf(c) > -1
+ if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
return
if c is "\u0000"
return
if c is '"'
parse_error()
- tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
+ tok_cur_tag.public_identifier = ''
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_cur_tag.public_identifier = ''
tok_state = tok_state_doctype_public_identifier_single_quoted
return
if c is '>'
return
if c is '"'
parse_error()
- tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
+ tok_cur_tag.public_identifier = ''
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_cur_tag.public_identifier = ''
tok_state = tok_state_doctype_public_identifier_single_quoted
return
if c is '>'
# 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(new RegExp("\u0000", 'g'), "\ufffd")
+ if val.length > 0
+ return new_character_token val # fixfull split
+ return null
# 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
if cur + 1 >= txt.length
return '&'
if txt.charAt(cur + 1).toLowerCase() is 'x'
- prefix = '#x'
+ base = 16
charset = hex_chars
start = cur + 2
else
charset = digits
start = cur + 1
- prefix = '#'
+ base = 10
i = 0
while start + i < txt.length and charset.indexOf(txt.charAt(start + i)) > -1
i += 1
if i is 0
return '&'
+ cur = start + i
if txt.charAt(start + i) is ';'
- i += 1
- # FIXME This is supposed to generate parse errors for some chars
- decoded = decode_named_char_ref(prefix + txt.substr(start, i).toLowerCase())
- if decoded?
- cur = start + i
- return decoded
- return '&'
+ cur += 1
+ else
+ parse_error()
+ code_point = txt.substr(start, i)
+ while code_point.charAt(0) is '0' and code_point.length > 1
+ code_point = code_point.substr 1
+ code_point = parseInt(code_point, base)
+ if unicode_fixes[code_point]?
+ parse_error()
+ return unicode_fixes[code_point]
+ else
+ if (code_point >= 0xd800 and code_point <= 0xdfff) or code_point > 0x10ffff
+ parse_error()
+ return "\ufffd"
+ else
+ if (code_point >= 0x0001 and code_point <= 0x0008) or (code_point >= 0x000D and code_point <= 0x001F) or (code_point >= 0x007F and code_point <= 0x009F) or (code_point >= 0xFDD0 and code_point <= 0xFDEF) or code_point is 0x000B or code_point is 0xFFFE or code_point is 0xFFFF or code_point is 0x1FFFE or code_point is 0x1FFFF or code_point is 0x2FFFE or code_point is 0x2FFFF or code_point is 0x3FFFE or code_point is 0x3FFFF or code_point is 0x4FFFE or code_point is 0x4FFFF or code_point is 0x5FFFE or code_point is 0x5FFFF or code_point is 0x6FFFE or code_point is 0x6FFFF or code_point is 0x7FFFE or code_point is 0x7FFFF or code_point is 0x8FFFE or code_point is 0x8FFFF or code_point is 0x9FFFE or code_point is 0x9FFFF or code_point is 0xAFFFE or code_point is 0xAFFFF or code_point is 0xBFFFE or code_point is 0xBFFFF or code_point is 0xCFFFE or code_point is 0xCFFFF or code_point is 0xDFFFE or code_point is 0xDFFFF or code_point is 0xEFFFE or code_point is 0xEFFFF or code_point is 0xFFFFE or code_point is 0xFFFFF or code_point is 0x10FFFE or code_point is 0x10FFFF
+ parse_error()
+ return from_code_point code_point
+ return
else
for i in [0...31]
if alnum.indexOf(txt.charAt(cur + i)) is -1
return '&'
return # never reached
+ eat_next_token_if_newline = ->
+ old_cur = cur
+ t = null
+ until t?
+ t = tok_state()
+ if t.type is TYPE_TEXT
+ # definition of a newline depends on whether it was a character ref or not
+ if cur - old_cur is 1
+ # not a character reference
+ if t.text is "\u000d" or t.text is "\u000a"
+ return
+ else
+ if t.text is "\u000a"
+ return
+ # not a "newline"
+ cur = old_cur
+ return
+
# 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]
+ txt = args.html
+ cur = 0
+ doc = new Node TYPE_TAG, name: 'document', namespace: NS_HTML
+ doc.flag 'quirks mode', QUIRKS_NO # TODO bugreport spec for not specifying this
+ fragment_root = null # fragment parsing algorithm returns children of this
+ open_els = []
afe = [] # active formatting elements
- template_insertion_modes = []
- insertion_mode = ins_mode_initial
- original_insertion_mode = insertion_mode # TODO check spec
- flag_scripting = true # TODO might need an extra flag to get <noscript> to parse correctly
+ template_ins_modes = []
+ ins_mode = ins_mode_initial
+ original_ins_mode = ins_mode # TODO check spec
+ flag_scripting = args.scripting ? true # TODO might need an extra flag to get <noscript> to parse correctly
flag_frameset_ok = true
flag_parsing = true
flag_foster_parenting = false
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
+ flag_fragment_parsing = false
+ context_element = null
+ prev_node_id = 0 # just for debugging
# tokenizer initialization
tok_state = tok_state_data
- # proccess input
- while flag_parsing
- t = tok_state()
- if t?
- insertion_mode t
- # fixfull parse error if has self-closing flag, but it wasn't acknolwedged
- return doc.children
+ parse_init = ->
+ # fragment parsing (text arg)
+ if args.fragment?
+ # this handles the fragment from the tests in the format described here:
+ # https://github.com/html5lib/html5lib-tests/blob/master/tree-construction/README.md
+ f = args.fragment
+ ns = NS_HTML
+ if f.substr(0, 5) is 'math '
+ f = f.substr 5
+ ns = NS_MATHML
+ else if f.substr(0, 4) is 'svg '
+ f = f.substr 4
+ ns = NS_SVG
+ t = new_open_tag f
+ context_element = token_to_element t, ns
+ context_element.document = new Node TYPE_TAG, name: 'document', namespace: NS_HTML
+ context_element.document.flag 'quirks mode', QUIRKS_NO
+ # fragment parsing (Node arg)
+ if args.context?
+ context_element = args.context
+
+ # http://www.w3.org/TR/html5/syntax.html#parsing-html-fragments
+ # fragment parsing algorithm
+ if context_element?
+ flag_fragment_parsing = true
+ doc = new Node TYPE_TAG, name: 'html', namespace: NS_HTML
+ # search up the tree from context, to try to find it's document,
+ # because this file only puts a "document" property on the root
+ # element.
+ old_doc = null
+ el = context_element
+ loop
+ if el.document?
+ old_doc = el.document
+ break
+ if el.parent
+ el = el.parent
+ else
+ break
+ if old_doc
+ doc.flag 'quirks mode', old_doc.flag 'quirks mode'
+ # set tok_state
+ if context_element.namespace is NS_HTML
+ switch context_element.name
+ when 'title', 'textarea'
+ tok_state = tok_state_rcdata
+ when 'style', 'xmp', 'iframe', 'noembed', 'noframes'
+ tok_state = tok_state_rawtext
+ when 'script'
+ tok_state = tok_state_script_data
+ when 'noscript'
+ if flag_scripting
+ tok_state = tok_state_rawtext
+ when 'plaintext'
+ tok_state = tok_state_plaintext
+ fragment_root = new Node TYPE_TAG, name: 'html', namespace: NS_HTML
+ doc.children.push fragment_root
+ fragment_root.document = doc
+ open_els = [fragment_root]
+ if context_element.name is 'template' and context_element.namespace is NS_HTML
+ template_ins_modes.unshift ins_mode_in_template
+ # fixfull create token for context (it should have it's original one already)
+ reset_ins_mode()
+ # set form_element pointer... in the foreign doc?!
+ el = context_element
+ loop
+ if el.name is 'form' and el.namespace is NS_HTML
+ form_element_pointer = el
+ break
+ if el.parent
+ el = el.parent
+ else
+ break
+
+ # text pre-processing
+ # FIXME check http://www.w3.org/TR/html5/syntax.html#preprocessing-the-input-stream
+ txt = txt.replace(new RegExp("\r\n", 'g'), "\n") # fixfull spec doesn't say this
+ txt = txt.replace(new RegExp("\r", 'g'), "\n") # fixfull spec doesn't say this
+
+ return
-serialize_els = (els, shallow, show_ids) ->
- serialized = ''
- sep = ''
- for t in els
- serialized += sep
- sep = ','
- serialized += t.serialize shallow, show_ids
- return serialized
+ # http://www.w3.org/TR/html5/syntax.html#tree-construction
+ parse_main_loop = ->
+ while flag_parsing
+ t = tok_state()
+ if t?
+ process_token t
+ # fixfull parse error if has self-closing flag, but it wasn't acknolwedged
+ return
+ parse_init()
+ parse_main_loop()
+
+ if flag_fragment_parsing
+ return fragment_root.children
+ return doc.children
-# 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_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
+module.exports.QUIRKS_NO = QUIRKS_NO
+module.exports.QUIRKS_LIMITED = QUIRKS_LIMITED
+module.exports.QUIRKS_YES = QUIRKS_YES
Got questions, comments, patches, etc.? Contact Jason Woofenden