# HTML parser meant to run in a browser, in support of WYSIWYG editor
# Copyright 2015 Jason Woofenden
#
# This program is free software: you can redistribute it and/or modify it under
# the terms of the GNU Affero General Public License as published by the Free
# Software Foundation, either version 3 of the License, or (at your option) any
# later version.
#
# This program is distributed in the hope that it will be useful, but WITHOUT
# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
# FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more
# details.
#
# You should have received a copy of the GNU Affero General Public License
# along with this program. If not, see .
# This file implements a parser for html snippets, meant to be used by a
# WYSIWYG editor. Hence it does not attempt to parse doctypes, ,
# or 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.
#
# Instead, the data structure produced by this parser is an array of Nodes.
# 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)
#
# stacks grow downward (current element is index=0)
#
# example: open_els = [a, b, c, d, e, f, g]
#
# "grows downwards" means it's visualized like this: (index: el, names)
#
# 6: g "start of the list", "topmost", "first"
# 5: f
# 4: e "previous" (to d), "above", "before"
# 3: d (previous/next are relative to this element)
# 2: c "next", "after", "lower", "below"
# 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
# Each node is an obect of the Node class. Here are the Node types:
TYPE_TAG = 0 # name, {attributes}, [children]
TYPE_TEXT = 1 # "text"
TYPE_COMMENT = 2
TYPE_DOCTYPE = 3
# the following types are emited by the tokenizer, but shouldn't end up in the tree:
TYPE_START_TAG = 4 # name, [attributes ([key,value]...) in reverse order], [children]
TYPE_END_TAG = 5 # name
TYPE_EOF = 6
TYPE_AFE_MARKER = 7 # http://www.w3.org/TR/html5/syntax.html#reconstruct-the-active-formatting-elements
TYPE_AAA_BOOKMARK = 8 # http://www.w3.org/TR/html5/syntax.html#adoption-agency-algorithm
# namespace constants
NS_HTML = 1
NS_MATHML = 2
NS_SVG = 3
g_debug_log = []
debug_log_reset = ->
g_debug_log = []
debug_log = (str) ->
g_debug_log.push str
debug_log_each = (cb) ->
for str in g_debug_log
cb str
prev_node_id = 0
class Node
constructor: (type, args = {}) ->
@type = type # one of the TYPE_* constants above
@name = args.name ? '' # tag name
@text = args.text ? '' # contents for text/comment nodes
@attrs = args.attrs ? {}
@attrs_a = args.attr_k ? [] # attrs in progress, TYPE_START_TAG only
@children = args.children ? []
@namespace = args.namespace ? NS_HTML
@parent = args.parent ? null
@token = args.token ? null
@flags = args.flags ? {}
if args.id?
@id = "#{args.id}+"
else
@id = "#{++prev_node_id}"
acknowledge_self_closing: ->
if @token?
@token.flag 'did_self_close'
else
@flag 'did_self_close', true
flag: (key, value = null) ->
if value?
@flags[key] = value
else
return @flags[key]
serialize: (shallow = false, show_ids = false) -> # for unit tests
ret = ''
switch @type
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:#{@name},#{JSON.stringify(@public_identifier ? '')},#{JSON.stringify(@system_identifier ? '')}"
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
# helpers: (only take args that are normally known when parser creates nodes)
new_open_tag = (name) ->
return new Node TYPE_START_TAG, name: name
new_end_tag = (name) ->
return new Node TYPE_END_TAG, name: name
new_element = (name) ->
return new Node TYPE_TAG, name: name
new_text_node = (txt) ->
return new Node TYPE_TEXT, text: txt
new_character_token = new_text_node
new_comment_token = (txt) ->
return new Node TYPE_COMMENT, text: txt
new_doctype_token = (name) ->
return new Node TYPE_DOCTYPE, name: name
new_eof_token = ->
return new Node TYPE_EOF
new_afe_marker = ->
return new Node TYPE_AFE_MARKER
new_aaa_bookmark = ->
return new Node TYPE_AAA_BOOKMARK
lc_alpha = "abcdefghijklmnopqrstuvwxyz"
uc_alpha = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
digits = "0123456789"
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 + "-"
# http://www.w3.org/TR/html5/infrastructure.html#space-character
space_chars = "\u0009\u000a\u000c\u000d\u0020"
is_space = (txt) ->
return txt.length is 1 and space_chars.indexOf(txt) > -1
is_space_tok = (t) ->
return t.type is TYPE_TEXT && t.text.length is 1 and space_chars.indexOf(t.text) > -1
is_input_hidden_tok = (t) ->
return unless t.type is TYPE_START_TAG
for a of t.attrs_a
if a[0] is 'type'
if a[1].toLowerCase() is 'hidden'
return true
return false
return false
# https://en.wikipedia.org/wiki/Whitespace_character#Unicode
whitespace_chars = "\u0009\u000a\u000b\u000c\u000d\u0020\u0085\u00a0\u1680\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a\u2028\u2029\u202f\u205f\u3000"
# 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 = {
Aacute: 'Á', aacute: 'á', Acirc: 'Â', acirc: 'â', acute: '´', AElig: 'Æ',
aelig: 'æ', Agrave: 'À', agrave: 'à', AMP: '&', amp: '&', Aring: 'Å',
aring: 'å', Atilde: 'Ã', atilde: 'ã', Auml: 'Ä', auml: 'ä', brvbar: '¦',
Ccedil: 'Ç', ccedil: 'ç', cedil: '¸', cent: '¢', COPY: '©', copy: '©',
curren: '¤', deg: '°', divide: '÷', Eacute: 'É', eacute: 'é', Ecirc: 'Ê',
ecirc: 'ê', Egrave: 'È', egrave: 'è', ETH: 'Ð', eth: 'ð', Euml: 'Ë',
euml: 'ë', frac12: '½', frac14: '¼', frac34: '¾', GT: '>', gt: '>',
Iacute: 'Í', iacute: 'í', Icirc: 'Î', icirc: 'î', iexcl: '¡', Igrave: 'Ì',
igrave: 'ì', iquest: '¿', Iuml: 'Ï', iuml: 'ï', laquo: '«', LT: '<',
lt: '<', macr: '¯', micro: 'µ', middot: '·', nbsp: "\u00a0", not: '¬',
Ntilde: 'Ñ', ntilde: 'ñ', Oacute: 'Ó', oacute: 'ó', Ocirc: 'Ô', ocirc: 'ô',
Ograve: 'Ò', ograve: 'ò', ordf: 'ª', ordm: 'º', Oslash: 'Ø', oslash: 'ø',
Otilde: 'Õ', otilde: 'õ', Ouml: 'Ö', ouml: 'ö', para: '¶', plusmn: '±',
pound: '£', QUOT: '"', quot: '"', raquo: '»', REG: '®', reg: '®', sect: '§',
shy: '', sup1: '¹', sup2: '²', sup3: '³', szlig: 'ß', THORN: 'Þ', thorn: 'þ',
times: '×', Uacute: 'Ú', uacute: 'ú', Ucirc: 'Û', ucirc: 'û', Ugrave: 'Ù',
ugrave: 'ù', uml: '¨', Uuml: 'Ü', uuml: 'ü', Yacute: 'Ý', yacute: 'ý',
yen: '¥', yuml: 'ÿ'
}
void_elements = ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'keygen', 'link', 'meta', 'param', 'source', 'track', 'wbr']
raw_text_elements = ['script', 'style']
escapable_raw_text_elements = ['textarea', 'title']
# http://www.w3.org/TR/SVG/ 1.1 (Second Edition)
svg_elements = [
'a', 'altGlyph', 'altGlyphDef', 'altGlyphItem', 'animate', 'animateColor',
'animateMotion', 'animateTransform', 'circle', 'clipPath', 'color-profile',
'cursor', 'defs', 'desc', 'ellipse', 'feBlend', 'feColorMatrix',
'feComponentTransfer', 'feComposite', 'feConvolveMatrix',
'feDiffuseLighting', 'feDisplacementMap', 'feDistantLight', 'feFlood',
'feFuncA', 'feFuncB', 'feFuncG', 'feFuncR', 'feGaussianBlur', 'feImage',
'feMerge', 'feMergeNode', 'feMorphology', 'feOffset', 'fePointLight',
'feSpecularLighting', 'feSpotLight', 'feTile', 'feTurbulence', 'filter',
'font', 'font-face', 'font-face-format', 'font-face-name', 'font-face-src',
'font-face-uri', 'foreignObject', 'g', 'glyph', 'glyphRef', 'hkern',
'image', 'line', 'linearGradient', 'marker', 'mask', 'metadata',
'missing-glyph', 'mpath', 'path', 'pattern', 'polygon', 'polyline',
'radialGradient', 'rect', 'script', 'set', 'stop', 'style', 'svg',
'switch', 'symbol', 'text', 'textPath', 'title', 'tref', 'tspan', 'use',
'view', 'vkern'
]
# http://www.w3.org/TR/MathML/ Version 3.0 2nd Edition
mathml_elements = [
'abs', 'and', 'annotation', 'annotation-xml', 'apply', 'approx', 'arccos',
'arccosh', 'arccot', 'arccoth', 'arccsc', 'arccsch', 'arcsec', 'arcsech',
'arcsin', 'arcsinh', 'arctan', 'arctanh', 'arg', 'bind', 'bvar', 'card',
'cartesianproduct', 'cbytes', 'ceiling', 'cerror', 'ci', 'cn', 'codomain',
'complexes', 'compose', 'condition', 'conjugate', 'cos', 'cosh', 'cot',
'coth', 'cs', 'csc', 'csch', 'csymbol', 'curl', 'declare', 'degree',
'determinant', 'diff', 'divergence', 'divide', 'domain',
'domainofapplication', 'emptyset', 'eq', 'equivalent', 'eulergamma',
'exists', 'exp', 'exponentiale', 'factorial', 'factorof', 'false', 'floor',
'fn', 'forall', 'gcd', 'geq', 'grad', 'gt', 'ident', 'image', 'imaginary',
'imaginaryi', 'implies', 'in', 'infinity', 'int', 'integers', 'intersect',
'interval', 'inverse', 'lambda', 'laplacian', 'lcm', 'leq', 'limit',
'list', 'ln', 'log', 'logbase', 'lowlimit', 'lt', 'maction', 'maligngroup',
'malignmark', 'math', 'matrix', 'matrixrow', 'max', 'mean', 'median',
'menclose', 'merror', 'mfenced', 'mfrac', 'mglyph', 'mi', 'mi', 'min',
'minus', 'mlabeledtr', 'mlongdiv', 'mmultiscripts', 'mn', 'mo', 'mode',
'moment', 'momentabout', 'mover', 'mpadded', 'mphantom', 'mprescripts',
'mroot', 'mrow', 'ms', 'mscarries', 'mscarry', 'msgroup', 'msline',
'mspace', 'msqrt', 'msrow', 'mstack', 'mstyle', 'msub', 'msubsup', 'msup',
'mtable', 'mtd', 'mtext', 'mtr', 'munder', 'munderover', 'naturalnumbers',
'neq', 'none', 'not', 'notanumber', 'notin', 'notprsubset', 'notsubset',
'or', 'otherwise', 'outerproduct', 'partialdiff', 'pi', 'piece',
'piecewise', 'plus', 'power', 'primes', 'product', 'prsubset', 'quotient',
'rationals', 'real', 'reals', 'reln', 'rem', 'root', 'scalarproduct',
'sdev', 'sec', 'sech', 'selector', 'semantics', 'sep', 'set', 'setdiff',
'share', 'sin', 'sinh', 'span', 'subset', 'sum', 'tan', 'tanh', 'tendsto',
'times', 'transpose', 'true', 'union', 'uplimit', 'variance', 'vector',
'vectorproduct', 'xor'
]
# foreign_elements = [svg_elements..., mathml_elements...]
#normal_elements = All other allowed HTML elements are normal elements.
special_elements = {
# HTML:
address:NS_HTML, applet:NS_HTML, area:NS_HTML, article:NS_HTML,
aside:NS_HTML, base:NS_HTML, basefont:NS_HTML, bgsound:NS_HTML,
blockquote:NS_HTML, body:NS_HTML, br:NS_HTML, button:NS_HTML,
caption:NS_HTML, center:NS_HTML, col:NS_HTML, colgroup:NS_HTML, dd:NS_HTML,
details:NS_HTML, dir:NS_HTML, div:NS_HTML, dl:NS_HTML, dt:NS_HTML,
embed:NS_HTML, fieldset:NS_HTML, figcaption:NS_HTML, figure:NS_HTML,
footer:NS_HTML, form:NS_HTML, frame:NS_HTML, frameset:NS_HTML, h1:NS_HTML,
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,
# MathML:
mi:NS_MATHML, mo:NS_MATHML, mn:NS_MATHML, ms:NS_MATHML, mtext:NS_MATHML,
'annotation-xml':NS_MATHML,
# SVG:
foreignObject:NS_SVG, desc:NS_SVG, title:NS_SVG
}
formatting_elements = {
a: true, b: true, big: true, code: true, em: true, font: true, i: true,
nobr: true, s: true, small: true, strike: true, strong: true, tt: true,
u: true
}
h_tags = {
h1:NS_HTML, h2:NS_HTML, h3:NS_HTML, h4:NS_HTML, h5:NS_HTML, h6:NS_HTML
}
# FIXME namespacify
foster_parenting_targets = {
table: true
tbody: true
tfoot: true
thead: true
tr: true
}
# FIXME namespacify
# 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
}
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_attribute_fixes = {
attributename: 'attributeName'
attributetype: 'attributeType'
basefrequency: 'baseFrequency'
baseprofile: 'baseProfile'
calcmode: 'calcMode'
clippathunits: 'clipPathUnits'
contentscripttype: 'contentScriptType'
contentstyletype: 'contentStyleType'
diffuseconstant: 'diffuseConstant'
edgemode: 'edgeMode'
externalresourcesrequired: 'externalResourcesRequired'
filterres: 'filterRes'
filterunits: 'filterUnits'
glyphref: 'glyphRef'
gradienttransform: 'gradientTransform'
gradientunits: 'gradientUnits'
kernelmatrix: 'kernelMatrix'
kernelunitlength: 'kernelUnitLength'
keypoints: 'keyPoints'
keysplines: 'keySplines'
keytimes: 'keyTimes'
lengthadjust: 'lengthAdjust'
limitingconeangle: 'limitingConeAngle'
markerheight: 'markerHeight'
markerunits: 'markerUnits'
markerwidth: 'markerWidth'
maskcontentunits: 'maskContentUnits'
maskunits: 'maskUnits'
numoctaves: 'numOctaves'
pathlength: 'pathLength'
patterncontentunits: 'patternContentUnits'
patterntransform: 'patternTransform'
patternunits: 'patternUnits'
pointsatx: 'pointsAtX'
pointsaty: 'pointsAtY'
pointsatz: 'pointsAtZ'
preservealpha: 'preserveAlpha'
preserveaspectratio: 'preserveAspectRatio'
primitiveunits: 'primitiveUnits'
refx: 'refX'
refy: 'refY'
repeatcount: 'repeatCount'
repeatdur: 'repeatDur'
requiredextensions: 'requiredExtensions'
requiredfeatures: 'requiredFeatures'
specularconstant: 'specularConstant'
specularexponent: 'specularExponent'
spreadmethod: 'spreadMethod'
startoffset: 'startOffset'
stddeviation: 'stdDeviation'
stitchtiles: 'stitchTiles'
surfacescale: 'surfaceScale'
systemlanguage: 'systemLanguage'
tablevalues: 'tableValues'
targetx: 'targetX'
targety: 'targetY'
textlength: 'textLength'
viewbox: 'viewBox'
viewtarget: 'viewTarget'
xchannelselector: 'xChannelSelector'
ychannelselector: 'yChannelSelector'
zoomandpan: 'zoomAndPan'
}
adjust_mathml_attributes = (t) ->
for a in t.attrs_a
if a[0] is 'definitionurl'
a[0] = 'definitionURL'
return
adjust_svg_attributes = (t) ->
for a in t.attrs_a
if svg_attribute_fixes[a[0]]?
a[0] = svg_attribute_fixes[a[0]]
return
adjust_foreign_attributes = (t) ->
# fixfull
return
# decode_named_char_ref()
#
# The list of named character references is _huge_ so ask the browser to decode
# for us instead of wasting bandwidth/space on including the table here.
#
# Pass without the "&" but with the ";" examples:
# for "&" pass "amp;"
# for "′" pass "x2032;"
g_dncr = {
cache: {}
textarea: document.createElement('textarea')
}
# TODO test this in IE8
decode_named_char_ref = (txt) ->
txt = "{txt}"
decoded = g_dncr.cache[txt]
return decoded if decoded?
g_dncr.textarea.innerHTML = txt
decoded = g_dncr.textarea.value
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
# 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_ins_modes = null
ins_mode = null
original_ins_mode = null
tok_state = null
tok_cur_tag = null # partially parsed tag
flag_scripting = null
flag_frameset_ok = null
flag_parsing = null
flag_foster_parenting = null
form_element_pointer = null
temporary_buffer = null
pending_table_character_tokens = null
head_element_pointer = null
flag_fragment_parsing = null
context_element = null
stop_parsing = ->
flag_parsing = false
parse_error = ->
if parse_error_cb?
parse_error_cb cur
else
console.log "Parse error at character #{cur} of #{txt.length}"
afe_push = (new_el) ->
matches = 0
for el, i in afe
if el.name is new_el.name and el.namespace is new_el.namespace
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
afe.unshift new_el
afe_push_marker = ->
afe.unshift new_afe_marker()
# 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'
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)
return true
if scope[t.name] is t.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)
return true
if scope[t.name] is t.namespace
return false
if scope2[t.name] is t.namespace
return false
return false
standard_scopers = {
applet: NS_HTML, caption: NS_HTML, html: NS_HTML, table: NS_HTML,
td: NS_HTML, th: NS_HTML, marquee: NS_HTML, object: NS_HTML,
template: NS_HTML, mi: NS_MATHML,
mo: NS_MATHML, mn: NS_MATHML, ms: NS_MATHML, mtext: NS_MATHML,
'annotation-xml': NS_MATHML,
foreignObject: NS_SVG, desc: NS_SVG, title: NS_SVG
}
button_scopers = button: NS_HTML
li_scopers = ol: NS_HTML, ul: NS_HTML
table_scopers = html: NS_HTML, table: NS_HTML, template: NS_HTML
is_in_scope = (tag_name, namespace = null) ->
return is_in_scope_x tag_name, standard_scopers, namespace
is_in_button_scope = (tag_name, namespace = null) ->
return is_in_scope_x_y tag_name, standard_scopers, button_scopers, namespace
is_in_table_scope = (tag_name, namespace = null) ->
return is_in_scope_x tag_name, table_scopers, namespace
# aka is_in_list_item_scope
is_in_li_scope = (tag_name, namespace = null) ->
return is_in_scope_x_y tag_name, standard_scopers, li_scopers, namespace
is_in_select_scope = (tag_name, namespace = null) ->
for t in open_els
if t.name is tag_name and (namespace is null or namespace is t.namespace)
return true
if t.ns isnt NS_HTML and t.name isnt 'optgroup' and t.name isnt 'option'
return false
return false
# this checks for a particular element, not by name
el_is_in_scope = (el) ->
for t in open_els
if t is el
return true
if standard_scopers[t.name] is t.namespace
return false
return false
clear_to_table_stopers = {
'table': true
'template': true
'html': true
}
clear_stack_to_table_context = ->
loop
if clear_to_table_stopers[open_els[0].name]?
break
open_els.shift()
return
clear_to_table_body_stopers = {
'tbody': true
'tfoot': true
'thead': true
'template': true
'html': true
}
clear_stack_to_table_body_context = ->
loop
if clear_to_table_body_stopers[open_els[0].name]?
break
open_els.shift()
return
clear_to_table_row_stopers = {
'tr': true
'template': true
'html': true
}
clear_stack_to_table_row_context = ->
loop
if clear_to_table_row_stopers[open_els[0].name]?
break
open_els.shift()
return
clear_afe_to_marker = ->
loop
return unless afe.length > 0 # this happens in fragment case, ?spec error
el = afe.shift()
if el.type is TYPE_AFE_MARKER
return
return
# 8.2.3.1 ...
# http://www.w3.org/TR/html5/syntax.html#reset-the-insertion-mode-appropriately
reset_ins_mode = ->
# 1. Let last be false.
last = false
# 2. Let node be the last node in the stack of open elements.
node_i = 0
node = open_els[node_i]
# 3. Loop: If node is the first node in the stack of open elements,
# then set last to true, and, if the parser was originally created as
# part of the HTML fragment parsing algorithm (fragment case) set node
# to the context element.
loop
if node_i is open_els.length - 1
last = true
# fixfull (fragment case)
# 4. If node is a select element, run these substeps:
if node.name is 'select'
# 1. If last is true, jump to the step below labeled done.
unless last
# 2. Let ancestor be node.
ancestor_i = node_i
ancestor = node
# 3. Loop: If ancestor is the first node in the stack of
# open elements, jump to the step below labeled done.
loop
if ancestor_i is open_els.length - 1
break
# 4. Let ancestor be the node before ancestor in the stack
# of open elements.
ancestor_i += 1
ancestor = open_els[ancestor_i]
# 5. If ancestor is a template node, jump to the step below
# labeled done.
if ancestor.name is 'template'
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'
ins_mode = ins_mode_in_select_in_table
return
# 7. Jump back to the step labeled loop.
# 8. Done: Switch the insertion mode to "in select" and abort
# these steps.
ins_mode = ins_mode_in_select
return
# 5. If node is a td or th element and last is false, then switch
# the insertion mode to "in cell" and abort these steps.
if (node.name is 'td' or node.name is 'th') and 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'
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'
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'
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'
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'
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)
# 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
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
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'
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'
ins_mode = ins_mode_in_frameset
return
# 16. If node is an html element, run these substeps:
if node.name is 'html'
# 1. If the head element pointer is null, switch the insertion
# mode to "before head" and abort these steps. (fragment case)
if head_element_pointer is null
ins_mode = ins_mode_before_head
else
# 2. Otherwise, the head element pointer is not null,
# switch the insertion mode to "after head" and abort these
# steps.
ins_mode = ins_mode_after_head
return
# 17. If last is true, then switch the insertion mode to "in body"
# and abort these steps. (fragment case)
if last
ins_mode = ins_mode_in_body
return
# 18. Let node now be the node before node in the stack of open
# elements.
node_i += 1
node = open_els[node_i]
# 19. Return to the step labeled loop.
# 8.2.3.2
# http://www.w3.org/TR/html5/syntax.html#adjusted-current-node
adjusted_current_node = ->
if open_els.length is 1 and flag_fragment_parsing
return context_element
return open_els[0]
# http://www.w3.org/TR/html5/syntax.html#reconstruct-the-active-formatting-elements
# this implementation is structured (mostly) as described at the link above.
# capitalized comments are the "labels" described at the link above.
reconstruct_afe = ->
return if afe.length is 0
if afe[0].type is TYPE_AFE_MARKER or afe[0] in open_els
return
# Rewind
i = 0
loop
if i is afe.length - 1
break
i += 1
if afe[i].type is TYPE_AFE_MARKER or afe[i] in open_els
i -= 1 # Advance
break
# Create
loop
el = insert_html_element afe[i].token
afe[i] = el
break if i is 0
i -= 1 # Advance
# http://www.w3.org/TR/html5/syntax.html#adoption-agency-algorithm
# adoption agency algorithm
# overview here:
# http://www.w3.org/TR/html5/syntax.html#misnested-tags:-b-i-/b-/i
# http://www.w3.org/TR/html5/syntax.html#misnested-tags:-b-p-/b-/p
# http://www.w3.org/TR/html5/syntax.html#unclosed-formatting-elements
adoption_agency = (subject) ->
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()
# 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
debug_log "aaa: starting off with subject on top of stack, exiting"
return
outer = 0
loop
if outer >= 8
return
outer += 1
# 5. Let formatting element be the last element in the list of
# active formatting elements that: is between the end of the list
# and the last scope marker in the list, if any, or the start of
# the list otherwise, and has the tag name subject.
fe = null
for t, fe_of_afe in afe
if t.type is TYPE_AFE_MARKER
break
if t.name is subject
fe = t
break
# If there is no such element, then abort these steps and instead
# act as described in the "any other end tag" entry above.
if fe is null
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,
# then this is a parse error; remove the element from the list, and
# abort these steps.
in_open_els = false
for t, fe_of_open_els in open_els
if t is fe
in_open_els = true
break
unless in_open_els
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
return
# 7. If formatting element is in the stack of open elements, but
# the element is not in scope, then this is a parse error; abort
# these steps.
unless el_is_in_scope fe
debug_log "aaa: fe not in scope"
parse_error()
return
# 8. If formatting element is not the current node, this is a parse
# error. (But do not abort these steps.)
unless open_els[0] is fe
parse_error()
# continue
# 9. Let furthest block be the topmost node in the stack of open
# elements that is lower in the stack than formatting element, and
# is an element in the special category. There might not be one.
fb = null
fb_of_open_els = null
for t, i in open_els
if t is fe
break
if el_is_special t
fb = t
fb_of_open_els = i
# and continue, to see if there's one that's more "topmost"
# 10. If there is no furthest block, then the UA must first pop all
# the nodes from the bottom of the stack of open elements, from the
# current node up to and including formatting element, then remove
# formatting element from the list of active formatting elements,
# and finally abort these steps.
if fb is null
debug_log "aaa: no fb"
loop
t = open_els.shift()
if t is fe
afe.splice fe_of_afe, 1
return
# 11. Let common ancestor be the element immediately above
# formatting element in the stack of open elements.
ca = open_els[fe_of_open_els + 1] # common ancestor
node_above = open_els[fb_of_open_els + 1] # next node if node isn't in open_els anymore
# 12. Let a bookmark note the position of formatting element in the list of active formatting elements relative to the elements on either side of it in the list.
bookmark = new_aaa_bookmark()
for t, i in afe
if t is fe
afe.splice i, 0, bookmark
break
node = last_node = fb
inner = 0
loop
inner += 1
# 3. Let node be the element immediately above node in the
# stack of open elements, or if node is no longer in the stack
# of open elements (e.g. because it got removed by this
# algorithm), the element that was immediately above node in
# the stack of open elements before node was removed.
node_next = null
for t, i in open_els
if t is node
node_next = open_els[i + 1]
break
node = node_next ? node_above
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.
node_in_afe = false
for t, i in afe
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
# of active formatting elements with an entry for the new
# element, replace the entry for node in the stack of open
# elements with an entry for the new element, and let node be
# the new element.
new_node = token_to_element node.token, NS_HTML, ca
for t, i in afe
if t is node
afe[i] = new_node
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
# aforementioned bookmark to be immediately after the new node
# in the list of active formatting elements.
if last_node is fb
for t, i in afe
if t is bookmark
afe.splice i, 1
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
# ancestor as the override target.
# In the case where fe is immediately followed by fb:
# * inner loop exits out early (node==fe)
# * last_node is fb
# * last_node is still in the tree (not a duplicate)
if last_node.parent?
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 = token_to_element fe.token, NS_HTML, fb
# 16. Take all of the child nodes of furthest block and append them
# to the element created in the last step.
while fb.children.length
t = fb.children.shift()
t.parent = new_element
new_element.children.push t
# 17. Append that new element to furthest block.
new_element.parent = fb
fb.children.push new_element
# 18. Remove formatting element from the list of active formatting
# elements, and insert the new element into the list of active
# formatting elements at the position of the aforementioned
# bookmark.
for t, i in afe
if t is fe
afe.splice i, 1
break
for t, i in afe
if t is bookmark
afe[i] = new_element
break
# 19. Remove formatting element from the stack of open elements,
# and insert the new element into the stack of open elements
# immediately below the position of furthest block in that stack.
for t, i in open_els
if t is fe
open_els.splice i, 1
break
for t, i in open_els
if t is fb
open_els.splice i, 0, new_element
break
# 20. Jump back to the step labeled outer loop.
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"
# 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'
parse_error()
while open_els.length > 1 # just in case
el = open_els.shift()
if el.name is 'p'
return
close_p_if_in_button_scope = ->
if is_in_button_scope 'p'
close_p_element()
# http://www.w3.org/TR/html5/syntax.html#insert-a-character
# aka insert_a_character = (t) ->
insert_character = (t) ->
dest = adjusted_insertion_location()
# fixfull check for Document node
if dest[1] > 0
prev = dest[0].children[dest[1] - 1]
if prev.type is TYPE_TEXT
prev.text += t.text
return
dest[0].children.splice dest[1], 0, t
# 8.2.5.1
# http://www.w3.org/TR/html5/syntax.html#creating-and-inserting-nodes
# http://www.w3.org/TR/html5/syntax.html#appropriate-place-for-inserting-a-node
adjusted_insertion_location = (override_target = null) ->
# 1. If there was an override target specified, then let target be the
# override target.
if override_target?
target = override_target
else # Otherwise, let target be the current node.
target = open_els[0]
# 2. Determine the adjusted insertion location using the first matching
# steps from the following list:
#
# If foster parenting is enabled and target is a table, tbody, tfoot,
# thead, or tr element Foster parenting happens when content is
# misnested in tables.
if flag_foster_parenting and foster_parenting_targets[target.name]
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'
last_template = el
last_template_i = i
break
# 2. Let last table be the last table element in the stack of
# open elements, if any.
last_table = null
last_table_i
for el, i in open_els
if el.name is 'table'
last_table = el
last_table_i = i
break
# 3. If there is a last template and either there is no last
# table, or there is one, but last template is lower (more
# recently added) than last table in the stack of open
# elements, then: let adjusted insertion location be inside
# last template's template contents, after its last child (if
# any), and abort these substeps.
if last_template and (last_table is null or last_template_i < last_table_i)
target = last_template # fixfull should be it's contents
target_i = target.children.length
break
# 4. If there is no last table, then let adjusted insertion
# location be inside the first element in the stack of open
# elements (the html element), after its last child (if any),
# and abort these substeps. (fragment case)
if last_table is null
# this is odd
target = open_els[open_els.length - 1]
target_i = target.children.length
# 5. If last table has a parent element, then let adjusted
# insertion location be inside last table's parent element,
# immediately before last table, and abort these substeps.
if last_table.parent?
for c, i in last_table.parent.children
if c is last_table
target = last_table.parent
target_i = i
break
break
# 6. Let previous element be the element immediately above last
# table in the stack of open elements.
#
# huh? how could it not have a parent?
previous_element = open_els[last_table_i + 1]
# 7. Let adjusted insertion location be inside previous
# element, after its last child (if any).
target = previous_element
target_i = target.children.length
# Note: These steps are involved in part because it's possible
# for elements, the table element in this case in particular,
# to have been moved by a script around in the DOM, or indeed
# removed from the DOM entirely, after the element was inserted
# by the parser.
break # don't really loop
else
# Otherwise Let adjusted insertion location be inside target, after
# its last child (if any).
target_i = target.children.length
# 3. If the adjusted insertion location is inside a template element,
# let it instead be inside the template element's template contents,
# after its last child (if any).
# fixfull (template)
# 4. Return the adjusted insertion location.
return [target, target_i]
# http://www.w3.org/TR/html5/syntax.html#create-an-element-for-the-token
# aka create_an_element_for_token
token_to_element = (t, namespace, intended_parent) ->
# convert attributes into a hash
attrs = {}
for a in t.attrs_a
attrs[a[0]] = a[1] # TODO check what to do with dupilcate attrs
el = new Node TYPE_TAG, name: t.name, namespace: namespace, attrs: attrs, token: t
# TODO 2. If the newly created element has an xmlns attribute in the
# XMLNS namespace whose value is not exactly the same as the element's
# namespace, that is a parse error. Similarly, if the newly created
# element has an xmlns:xlink attribute in the XMLNS namespace whose
# value is not the XLink Namespace, that is a parse error.
# fixfull: the spec says stuff about form pointers and ownerDocument
return el
# http://www.w3.org/TR/html5/syntax.html#insert-a-foreign-element
insert_foreign_element = (token, namespace) ->
ail = adjusted_insertion_location()
ail_el = ail[0]
ail_i = ail[1]
el = token_to_element token, namespace, ail_el
# TODO skip this next step if it's broken (eg ail_el is document with child already)
el.parent = ail_el
ail_el.children.splice ail_i, 0, el
open_els.unshift el
return el
# http://www.w3.org/TR/html5/syntax.html#insert-an-html-element
insert_html_element = (token) ->
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
# 8.2.5.2
# http://www.w3.org/TR/html5/syntax.html#generic-raw-text-element-parsing-algorithm
parse_generic_raw_text = (t) ->
insert_html_element t
tok_state = tok_state_rawtext
original_ins_mode = ins_mode
ins_mode = ins_mode_text
parse_generic_rcdata_text = (t) ->
insert_html_element t
tok_state = tok_state_rcdata
original_ins_mode = ins_mode
ins_mode = ins_mode_text
# 8.2.5.3 http://www.w3.org/TR/html5/syntax.html#closing-elements-that-have-implied-end-tags
# http://www.w3.org/TR/html5/syntax.html#generate-implied-end-tags
generate_implied_end_tags = (except = null) ->
while end_tag_implied[open_els[0].name] and open_els[0].name isnt except
open_els.shift()
# 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
ins_mode_initial = (t) ->
if is_space_tok t
return
if t.type is TYPE_COMMENT
# ?fixfull
doc.children.push t
return
if t.type is TYPE_DOCTYPE
# FIXME check identifiers, set quirks, etc
# fixfull
doc.children.push t
ins_mode = ins_mode_before_html
return
# Anything else
#fixfull (iframe, quirks)
ins_mode = ins_mode_before_html
ins_mode t # reprocess the token
return
# 8.2.5.4.2 http://www.w3.org/TR/html5/syntax.html#the-before-html-insertion-mode
ins_mode_before_html = (t) ->
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_COMMENT
doc.children.push t
return
if is_space_tok t
return
if t.type is TYPE_START_TAG and t.name is 'html'
el = token_to_element t, NS_HTML, doc
doc.children.push el
open_els.unshift(el)
# fixfull (big paragraph in spec about manifest, fragment, urls, etc)
ins_mode = ins_mode_before_head
return
if t.type is TYPE_END_TAG
if t.name is 'head' or t.name is 'body' or t.name is 'html' or t.name is 'br'
# fall through to "anything else"
else
parse_error()
return
# Anything else
html_tok = new_open_tag 'html'
el = token_to_element html_tok, NS_HTML, doc
doc.children.push el
open_els.unshift el
# ?fixfull browsing context
ins_mode = ins_mode_before_head
ins_mode t
return
# 8.2.5.4.3 http://www.w3.org/TR/html5/syntax.html#the-before-head-insertion-mode
ins_mode_before_head = (t) ->
if is_space_tok t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_START_TAG and t.name is 'head'
el = insert_html_element t
head_element_pointer = el
ins_mode = ins_mode_in_head
if t.type is TYPE_END_TAG
if t.name is 'head' or t.name is 'body' or t.name is 'html' or t.name is 'br'
# fall through to Anything else below
else
parse_error()
return
# Anything else
head_tok = new_open_tag 'head'
el = insert_html_element head_tok
head_element_pointer = el
ins_mode = ins_mode_in_head
ins_mode t # reprocess current token
# 8.2.5.4.4 http://www.w3.org/TR/html5/syntax.html#parsing-main-inhead
ins_mode_in_head_else = (t) -> # factored out for same-as-spec flow control
open_els.shift() # spec says this will be a 'head' node
ins_mode = ins_mode_after_head
ins_mode t
ins_mode_in_head = (t) ->
if t.type is TYPE_TEXT and (t.text is "\t" or t.text is "\n" or t.text is "\u000c" or t.text is ' ')
insert_character t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link')
el = insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
return
if t.type is TYPE_START_TAG and t.name is 'meta'
el = insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
# fixfull encoding stuff
return
if t.type is TYPE_START_TAG and t.name is 'title'
parse_generic_rcdata_text t
return
if t.type is TYPE_START_TAG and ((t.name is 'noscript' and flag_scripting) or (t.name is 'noframes' or t.name is 'style'))
parse_generic_raw_text t
return
if t.type is TYPE_START_TAG and t.name is 'noscript' and flag_scripting is false
insert_html_element t
ins_mode = ins_mode_in_head_noscript
return
if t.type is TYPE_START_TAG and t.name is 'script'
ail = adjusted_insertion_location()
el = token_to_element t, NS_HTML, ail
el.flag 'parser-inserted', true
# fixfull frament case
ail[0].children.splice ail[1], 0, el
open_els.unshift el
tok_state = tok_state_script_data
original_ins_mode = ins_mode # make sure orig... is defined
ins_mode = ins_mode_text
return
if t.type is TYPE_END_TAG and t.name is 'head'
open_els.shift() # will be a head element... spec says so
ins_mode = ins_mode_after_head
return
if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'html' or t.name is 'br')
ins_mode_in_head_else t
return
if t.type is TYPE_START_TAG and t.name is 'template'
insert_html_element t
afe_push_marker()
flag_frameset_ok = false
ins_mode = ins_mode_in_template
template_ins_modes.unshift ins_mode_in_template
return
if t.type is TYPE_END_TAG and t.name is 'template'
if template_tag_is_open()
generate_implied_end_tags
if open_els[0].name isnt 'template'
parse_error()
loop
el = open_els.shift()
if el.name is 'template'
break
clear_afe_to_marker()
template_ins_modes.shift()
reset_ins_mode()
else
parse_error()
return
if (t.type is TYPE_START_TAG and t.name is 'head') or t.type is TYPE_END_TAG
parse_error()
return
ins_mode_in_head_else t
# 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
ins_mode t
ins_mode_in_head_noscript = (t) ->
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG
ins_mode_in_body t
return
if t.type is TYPE_END_TAG and t.name is 'noscript'
open_els.shift()
ins_mode = ins_mode_in_head
return
if (t.type is TYPE_TEXT and (t.text is "\t" or t.text is "\u000a" or t.text is "\u000c" or t.text is "\u000d" or t.text is ' ')) or t.type is TYPE_COMMENT or (t.type is TYPE_START_TAG and (t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'style'))
ins_mode_in_head t
return
if t.type is TYPE_END_TAG and t.name is 'br'
ins_mode_in_head_noscript_else t
return
if (t.type is TYPE_START_TAG and (t.name is 'head' or t.name is 'noscript')) or t.type is TYPE_END_TAG
parse_error()
return
# Anything else
ins_mode_in_head_noscript_else t
return
# 8.2.5.4.6 http://www.w3.org/TR/html5/syntax.html#the-after-head-insertion-mode
ins_mode_after_head_else = (t) ->
body_tok = new_open_tag 'body'
insert_html_element body_tok
ins_mode = ins_mode_in_body
ins_mode t # reprocess token
return
ins_mode_after_head = (t) ->
if is_space_tok t
insert_character t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_START_TAG and t.name is 'body'
insert_html_element t
flag_frameset_ok = false
ins_mode = ins_mode_in_body
return
if t.type is TYPE_START_TAG and t.name is 'frameset'
insert_html_element t
ins_mode = ins_mode_in_frameset
return
if t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'script' or t.name is 'style' or t.name is 'template' or t.name is 'title')
parse_error()
open_els.unshift head_element_pointer
ins_mode_in_head t
for el, i of 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
if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'html' or t.name is 'br')
ins_mode_after_head_else t
return
if (t.type is TYPE_START_TAG and t.name is 'head') or t.type is TYPE_END_TAG
parse_error()
return
# Anything else
ins_mode_after_head_else t
# 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 el, i in open_els
if el.namespace is NS_HTML and el.name is name
generate_implied_end_tags name # arg is exception
parse_error() unless i is 0
while i >= 0
open_els.shift()
i -= 1
return
if special_elements[el.name] is el.namespace
parse_error()
return
return
ins_mode_in_body = (t) ->
if t.type is TYPE_TEXT and t.text is "\u0000"
parse_error()
return
if is_space_tok t
reconstruct_afe()
insert_character t
return
if t.type is TYPE_TEXT
reconstruct_afe()
insert_character t
flag_frameset_ok = false
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
parse_error()
return if template_tag_is_open()
root_attrs = open_els[open_els.length - 1].attrs
for a of t.attrs_a
root_attrs[a[0]] = a[1] unless root_attrs[a[0]]?
return
if (t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'script' or t.name is 'style' or t.name is 'template' or t.name is 'title')) or (t.type is TYPE_END_TAG and t.name is 'template')
ins_mode_in_head t
return
if t.type is TYPE_START_TAG and t.name is 'body'
parse_error()
return if open_els.length < 2
second = open_els[open_els.length - 2]
return unless second.ns is NS_HTML
return unless second.name is 'body'
return if template_tag_is_open()
frameset_ok_flag = false
for a of t.attrs_a
second.attrs[a[0]] = a[1] unless second.attrs[a[0]]?
return
if t.type is TYPE_START_TAG and t.name is 'frameset'
parse_error()
return if open_els.length < 2
second_i = open_els.length - 2
second = open_els[second_i]
return unless second.ns is NS_HTML
return unless second.name is 'body'
flag_frameset_ok = false
if second.parent?
for el, i in second.parent.children
if el is second
second.parent.children.splice i, 1
break
open_els.splice second_i, 1
# pop everything except the "root html element"
while open_els.length > 1
open_els.shift()
insert_html_element t
ins_mode = ins_mode_in_frameset
return
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'
parse_error()
return
ok_tags = {
dd:NS_HTML, dt:NS_HTML, li:NS_HTML, optgroup:NS_HTML,
option:NS_HTML, p:NS_HTML, rb:NS_HTML, rp:NS_HTML, rt:NS_HTML,
rtc:NS_HTML, tbody:NS_HTML, td:NS_HTML, tfoot:NS_HTML,
th:NS_HTML, thead:NS_HTML, tr:NS_HTML, body:NS_HTML,
html:NS_HTML
}
for el in open_els
unless ok_tags[t.name] is el.namespace
parse_error()
break
ins_mode = ins_mode_after_body
return
if t.type is TYPE_END_TAG and t.name is 'html'
unless is_in_scope 'body'
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
ins_mode 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]] is NS_HTML
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
# spec: If the next token is a "LF" (U+000A) character token, then
# ignore that token and move on to the next one. (Newlines at the
# start of pre blocks are ignored as an authoring convenience.)
if txt.charAt(cur) is "\u000a" # FIXME check for crlf?
cur += 1
flag_frameset_ok = false
return
if t.type is TYPE_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()
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'
close_p_if_in_button_scope() # fixfull quirksmode thing
insert_html_element t
flag_frameset_ok = false
ins_mode = ins_mode_in_table
return
if t.type is TYPE_END_TAG and t.name is 'br'
parse_error()
t.type is TYPE_START_TAG
# fall through
if t.type is TYPE_START_TAG and (t.name is 'area' or t.name is 'br' or t.name is 'embed' or t.name is 'img' or t.name is 'keygen' or t.name is 'wbr')
reconstruct_afe()
insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
flag_frameset_ok = false
return
if t.type is TYPE_START_TAG and t.name is 'input'
reconstruct_afe()
insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
unless is_input_hidden_tok t
flag_frameset_ok = false
return
if t.type is TYPE_START_TAG and (t.name is 'param' or t.name is 'source' or t.name is 'track')
insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
return
if t.type is TYPE_START_TAG and t.name is 'hr'
close_p_if_in_button_scope()
insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
flag_frameset_ok = false
return
if t.type is TYPE_START_TAG and t.name is 'image'
parse_error()
t.name = 'img'
ins_mode 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 prompt # fixfull split
# TODO submit typo "balue" in spec
insert_html_element input_el
open_els.shift()
# insert_character '' # you can put chars here if promt attr missing
open_els.shift()
insert_html_element new_open_tag 'hr'
open_els.shift()
open_els.shift()
unless template_tag_is_open()
form_element_pointer = null
return
if t.type is TYPE_START_TAG and t.name is 'textarea'
insert_html_element t
if txt.charAt(cur) is "\u000a" # FIXME check for crlf?
cur += 1
tok_state = tok_state_rcdata
original_ins_mode = ins_mode
flag_frameset_ok = false
ins_mode = ins_mode_text
return
if t.type is TYPE_START_TAG and t.name is 'xmp'
close_p_if_in_button_scope()
reconstruct_afe()
flag_frameset_ok = false
parse_generic_raw_text t
return
if t.type is TYPE_START_TAG and t.name is 'iframe'
flag_frameset_ok = false
parse_generic_raw_text t
return
if t.type is TYPE_START_TAG and (t.name is 'noembed' or (t.name is 'noscript' and flag_scripting))
parse_generic_raw_text t
return
if t.type is TYPE_START_TAG and t.name is 'select'
reconstruct_afe()
insert_html_element t
flag_frameset_ok = false
if ins_mode is ins_mode_in_table or ins_mode is ins_mode_in_caption or ins_mode is ins_mode_in_table_body or ins_mode is ins_mode_in_row or ins_mode is ins_mode_in_cell
ins_mode = ins_mode_in_select_in_table
else
ins_mode = ins_mode_in_select
return
if t.type is TYPE_START_TAG and (t.name is 'optgroup' or t.name is 'option')
if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML
open_els.shift()
reconstruct_afe()
insert_html_element t
return
if t.type is TYPE_START_TAG and (t.name is 'rb' or t.name is 'rp' or t.name is 'rtc')
if is_in_scope 'ruby', NS_HTML
generate_implied_end_tags()
unless open_els[0].name is 'ruby' and open_els[0].namespace is NS_HTML
parse_error()
insert_html_element t
return
if t.type is TYPE_START_TAG and t.name is 'rt'
if is_in_scope 'ruby', NS_HTML
generate_implied_end_tags 'rtc' # arg is exception
unless (open_els[0].name is 'ruby' or open_els[0].name is 'rtc') and open_els[0].namespace is NS_HTML
parse_error()
insert_html_element t
return
if t.type is TYPE_START_TAG and t.name is 'math'
reconstruct_afe()
adjust_mathml_attributes t
adjust_foreign_attributes t
insert_foreign_element t, NS_MATHML
if t.flag 'self-closing'
open_els.shift()
t.acknowledge_self_closing()
return
if t.type is TYPE_START_TAG and t.name is 'svg'
reconstruct_afe()
adjust_svg_attributes t
adjust_foreign_attributes t
insert_foreign_element t, NS_SVG
if t.flag 'self-closing'
open_els.shift()
t.acknowledge_self_closing()
return
if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'frame' or t.name is 'head' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')
parse_error()
return
if t.type is TYPE_START_TAG # any other start tag
reconstruct_afe()
insert_html_element t
return
if t.type is TYPE_END_TAG # any other end tag
in_body_any_other_end_tag t.name
return
return
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) ->
if t.type is TYPE_TEXT
insert_character t
return
if t.type is TYPE_EOF
parse_error()
if open_els[0].name is 'script'
open_els[0].flag 'already started', true
open_els.shift()
ins_mode = original_ins_mode
ins_mode t
return
if t.type is TYPE_END_TAG and t.name is 'script'
open_els.shift()
ins_mode = original_ins_mode
# fixfull the spec seems to assume that I'm going to run the script
# http://www.w3.org/TR/html5/syntax.html#scriptEndTag
return
if t.type is TYPE_END_TAG
open_els.shift()
ins_mode = original_ins_mode
return
console.log 'warning: end of ins_mode_text reached'
# the functions below implement the tokenizer stats described here:
# http://www.w3.org/TR/html5/syntax.html#tokenization
# 8.2.5.4.9 http://www.w3.org/TR/html5/syntax.html#parsing-main-intable
ins_mode_in_table = (t) ->
switch t.type
when TYPE_TEXT
if can_in_table[t.name]
original_ins_mode = ins_mode
ins_mode = ins_mode_in_table_text
ins_mode t
else
ins_mode_in_table_else t
when TYPE_COMMENT
insert_comment t
when TYPE_DOCTYPE
parse_error()
when TYPE_START_TAG
switch t.name
when 'caption'
clear_stack_to_table_context()
afe_push_marker()
insert_html_element t
ins_mode = ins_mode_in_caption
when 'colgroup'
clear_stack_to_table_context()
insert_html_element t
ins_mode = ins_mode_in_column_group
when 'col'
clear_stack_to_table_context()
insert_html_element new_open_tag 'colgroup'
ins_mode = ins_mode_in_column_group
ins_mode t
when 'tbody', 'tfoot', 'thead'
clear_stack_to_table_context()
insert_html_element t
ins_mode = ins_mode_in_table_body
when 'td', 'th', 'tr'
clear_stack_to_table_context()
insert_html_element new_open_tag 'tbody'
ins_mode = ins_mode_in_table_body
ins_mode t
when 'table'
parse_error()
if is_in_table_scope 'table'
loop
el = open_els.shift()
if el.name is 'table'
break
reset_ins_mode()
ins_mode t
when 'style', 'script', 'template'
ins_mode_in_head t
when 'input'
if is_input_hidden_tok t
ins_mode_in_table_else t
else
parse_error()
el = insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
when 'form'
parse_error()
if form_element_pointer?
return
if template_tag_is_open()
return
form_element_pointer = insert_html_element t
open_els.shift()
else
ins_mode_in_table_else t
when TYPE_END_TAG
switch t.name
when 'table'
if is_in_table_scope 'table'
loop
el = open_els.shift()
if el.name is 'table'
break
reset_ins_mode()
else
parse_error
when 'body', 'caption', 'col', 'colgroup', 'html', 'tbody', 'td', 'tfoot', 'th', 'thead', 'tr'
parse_error()
when 'template'
ins_mode_in_head t
else
ins_mode_in_table_else t
when TYPE_EOF
ins_mode_in_body t
else
ins_mode_in_table_else t
# 8.2.5.4.10 http://www.w3.org/TR/html5/syntax.html#parsing-main-intabletext
ins_mode_in_table_text = (t) ->
if t.type is TYPE_TEXT and t.text is "\u0000"
# huh? I thought the tokenizer didn't emit these
parse_error()
return
if t.type is TYPE_TEXT
pending_table_character_tokens.push t
return
# Anything else
all_space = true
for old in pending_table_character_tokens
unless is_space_tok old
all_space = false
break
if all_space
for old in pending_table_character_tokens
insert_character old
else
for old in pending_table_character_tokens
ins_mode_table_else old
pending_table_character_tokens = [] # FIXME test (spec doesn't say this)
ins_mode = original_ins_mode
ins_mode t
# 8.2.5.4.11 http://www.w3.org/TR/html5/syntax.html#parsing-main-incaption
ins_mode_in_caption = (t) ->
if t.type is TYPE_END_TAG and t.name is 'caption'
if is_in_table_scope 'caption'
generate_implied_end_tags()
if open_els[0].name isnt 'caption'
parse_error()
loop
el = open_els.shift()
if el.name is 'caption'
break
clear_afe_to_marker()
ins_mode = ins_mode_in_table
else
parse_error()
# fragment case
return
if (t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')) or t.type is TYPE_END_TAG and t.name is 'table'
parse_error()
if is_in_table_scope 'caption'
loop
el = open_els.shift()
if el.name is 'caption'
break
clear_afe_to_marker()
ins_mode = ins_mode_in_table
ins_mode t
# else fragment case
return
if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')
parse_error()
return
# Anything else
ins_mode_in_body t
# 8.2.5.4.12 http://www.w3.org/TR/html5/syntax.html#parsing-main-incolgroup
ins_mode_in_column_group = (t) ->
if is_space_tok t
insert_character t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_START_TAG and t.name is 'col'
el = insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
return
if t.type is TYPE_END_TAG and t.name is 'colgroup'
if open_els[0].name is 'colgroup'
open_els.shift()
ins_mode = ins_mode_in_table
else
parse_error()
return
if t.type is TYPE_END_TAG and t.name is 'col'
parse_error()
return
if (t.type is TYPE_START_TAG or t.type is TYPE_END_TAG) and t.name is 'template'
ins_mode_in_head t
return
if t.type is TYPE_EOF
ins_mode_in_body t
return
# Anything else
if open_els[0].name isnt 'colgroup'
parse_error()
return
open_els.shift()
ins_mode = ins_mode_in_table
ins_mode t
return
# 8.2.5.4.13 http://www.w3.org/TR/html5/syntax.html#parsing-main-intbody
ins_mode_in_table_body = (t) ->
if t.type is TYPE_START_TAG and t.name is 'tr'
clear_stack_to_table_body_context()
insert_html_element t
ins_mode = ins_mode_in_row
return
if t.type is TYPE_START_TAG and (t.name is 'th' or t.name is 'td')
parse_error()
clear_stack_to_table_body_context()
insert_html_element new_open_tag 'tr'
ins_mode = ins_mode_in_row
ins_mode t
return
if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
unless is_in_table_scope t.name # fixfull check namespace
parse_error()
return
clear_stack_to_table_body_context()
open_els.shift()
ins_mode = ins_mode_in_table
return
if (t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')) or (t.type is TYPE_END_TAG and t.name is 'table')
has = false
for el in open_els
if el.name is 'tbody' or el.name is 'tfoot' or el.name is 'thead'
has = true
break
if table_scopers[el.name]
break
if !has
parse_error()
return
clear_stack_to_table_body_context()
open_els.shift()
ins_mode = ins_mode_in_table
ins_mode t
return
if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th' or t.name is 'tr')
parse_error()
return
# Anything else
ins_mode_in_table t
# 8.2.5.4.14 http://www.w3.org/TR/html5/syntax.html#parsing-main-intr
ins_mode_in_row = (t) ->
if t.type is TYPE_START_TAG and (t.name is 'th' or t.name is 'td')
clear_stack_to_table_row_context()
insert_html_element t
ins_mode = ins_mode_in_cell
afe_push_marker()
return
if t.type is TYPE_END_TAG and t.name is 'tr'
if is_in_table_scope 'tr'
clear_stack_to_table_row_context()
open_els.shift()
ins_mode = ins_mode_in_table_body
else
parse_error()
return
if (t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')) or t.type is TYPE_END_TAG and t.name is 'table'
if is_in_table_scope 'tr'
clear_stack_to_table_row_context()
open_els.shift()
ins_mode = ins_mode_in_table_body
ins_mode t
else
parse_error()
return
if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
if is_in_table_scope t.name # fixfull namespace
if is_in_table_scope 'tr'
clear_stack_to_table_row_context()
open_els.shift()
ins_mode = ins_mode_in_table_body
ins_mode t
else
parse_error()
return
if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th')
parse_error()
return
# Anything else
ins_mode_in_table t
# http://www.w3.org/TR/html5/syntax.html#close-the-cell
close_the_cell = ->
generate_implied_end_tags()
unless open_els[0].name is 'td' or open_els[0] is 'th'
parse_error()
loop
el = open_els.shift()
if el.name is 'td' or el.name is 'th'
break
clear_afe_to_marker()
ins_mode = ins_mode_in_row
# 8.2.5.4.15 http://www.w3.org/TR/html5/syntax.html#parsing-main-intd
ins_mode_in_cell = (t) ->
if t.type is TYPE_END_TAG and (t.name is 'td' or t.name is 'th')
if is_in_table_scope t.name
generate_implied_end_tags()
if open_els[0].name isnt t.name
parse_error
loop
el = open_els.shift()
if el.name is t.name
break
clear_afe_to_marker()
ins_mode = ins_mode_in_row
else
parse_error()
return
if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')
has = false
for el in open_els
if el.name is 'td' or el.name is 'th'
has = true
break
if table_scopers[el.name]
break
if !has
parse_error()
return
close_the_cell()
ins_mode t
return
if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html')
parse_error()
return
if t.type is TYPE_END_TAG and (t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')
if is_in_table_scope t.name # fixfull namespace
close_the_cell()
ins_mode t
else
parse_error()
return
# Anything Else
ins_mode_in_body t
# 8.2.5.4.16 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselect
ins_mode_in_select = (t) ->
if t.type is TYPE_TEXT and t.text is "\u0000"
parse_error()
return
if t.type is TYPE_TEXT
insert_character t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_START_TAG and t.name is 'option'
if open_els[0].name is 'option'
open_els.shift()
insert_html_element t
return
if t.type is TYPE_START_TAG and t.name is 'optgroup'
if open_els[0].name is 'option'
open_els.shift()
if open_els[0].name is 'optgroup'
open_els.shift()
insert_html_element t
return
if t.type is TYPE_END_TAG and t.name is 'optgroup'
if open_els[0].name is 'option' and open_els[1].name is 'optgroup'
open_els.shift()
if open_els[0].name is 'optgroup'
open_els.shift()
else
parse_error()
return
if t.type is TYPE_END_TAG and t.name is 'option'
if open_els[0].name is 'option'
open_els.shift()
else
parse_error()
return
if t.type is TYPE_END_TAG and t.name is 'select'
if is_in_select_scope 'select'
loop
el = open_els.shift()
if el.name is 'select'
break
reset_ins_mode()
else
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'select'
parse_error()
loop
el = open_els.shift()
if el.name is 'select'
break
reset_ins_mode()
# spec says that this is the same as but it doesn't say
# to check scope first
return
if t.type is TYPE_START_TAG and (t.name is 'input' or t.name is 'keygen' or t.name is 'textarea')
parse_error()
if is_in_select_scope 'select'
return
loop
el = open_els.shift()
if el.name is 'select'
break
reset_ins_mode()
ins_mode t
return
if t.type is TYPE_START_TAG and (t.name is 'script' or t.name is 'template')
ins_mode_in_head t
return
if t.type is TYPE_EOF
ins_mode_in_body t
return
# Anything else
parse_error()
return
# 8.2.5.4.17 http://www.w3.org/TR/html5/syntax.html#parsing-main-inselectintable
ins_mode_in_select_in_table = (t) ->
if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th')
parse_error()
loop
el = open_els.shift()
if el.name is 'select'
break
reset_ins_mode()
ins_mode t
return
if t.type is TYPE_END_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th')
parse_error()
unless is_in_table_scope t.name, NS_HTML
return
loop
el = open_els.shift()
if el.name is 'select'
break
reset_ins_mode()
ins_mode t
return
# Anything else
ins_mode_in_select t
return
# 8.2.5.4.18 http://www.w3.org/TR/html5/syntax.html#parsing-main-intemplate
ins_mode_in_template = (t) ->
if t.type is TYPE_TEXT or t.type is TYPE_COMMENT or t.type is TYPE_DOCTYPE
ins_mode_in_body t
return
if (t.type is TYPE_START_TAG and (t.name is 'base' or t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'script' or t.name is 'style' or t.name is 'template' or t.name is 'title')) or (t.type is TYPE_END_TAG and t.name is 'template')
ins_mode_in_head t
return
if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')
template_ins_modes.shift()
template_ins_modes.unshift ins_mode_in_table
ins_mode = ins_mode_in_table
ins_mode t
return
if t.type is TYPE_START_TAG and t.name is 'col'
template_ins_modes.shift()
template_ins_modes.unshift ins_mode_in_column_group
ins_mode = ins_mode_in_column_group
ins_mode t
return
if t.type is TYPE_START_TAG and t.name is 'tr'
template_ins_modes.shift()
template_ins_modes.unshift ins_mode_in_table_body
ins_mode = ins_mode_in_table_body
ins_mode t
return
if t.type is TYPE_START_TAG and (t.name is 'td' or t.name is 'th')
template_ins_modes.shift()
template_ins_modes.unshift ins_mode_in_row
ins_mode = ins_mode_in_row
ins_mode t
return
if t.type is TYPE_START_TAG
template_ins_modes.shift()
template_ins_modes.unshift ins_mode_in_body
ins_mode = ins_mode_in_body
ins_mode t
return
if t.type is TYPE_END_TAG
parse_error()
return
if t.type is TYPE_EOF
unless template_tag_is_open()
stop_parsing()
return
parse_error()
loop
el = open_els.shift()
if el.name is 'template' # fixfull check namespace
break
clear_afe_to_marker()
template_ins_modes.shift()
reset_ins_mode()
ins_mode t
# 8.2.5.4.19 http://www.w3.org/TR/html5/syntax.html#parsing-main-afterbody
ins_mode_after_body = (t) ->
if is_space_tok t
ins_mode_in_body t
return
if t.type is TYPE_COMMENT
insert_comment t, [open_els[0], open_els[0].children.length]
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_END_TAG and t.name is 'html'
# fixfull fragment case
ins_mode = ins_mode_after_after_body
return
if t.type is TYPE_EOF
stop_parsing()
return
# Anything ELse
parse_error()
ins_mode = ins_mode_in_body
ins_mode t
# 8.2.5.4.20 http://www.w3.org/TR/html5/syntax.html#parsing-main-inframeset
ins_mode_in_frameset = (t) ->
if is_space_tok t
insert_character t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_START_TAG and t.name is 'frameset'
insert_html_element t
return
if t.type is TYPE_END_TAG and t.name is 'frameset'
# TODO ?correct for: "if the current node is the root html element"
if open_els.length is 1
parse_error()
return # fragment case
open_els.shift()
if flag_fragment_parsing is false and open_els[0].name isnt 'frameset'
ins_mode = ins_mode_after_frameset
return
if t.type is TYPE_START_TAG and t.name is 'frame'
insert_html_element t
open_els.shift()
t.acknowledge_self_closing()
return
if t.type is TYPE_START_TAG and t.name is 'noframes'
ins_mode_in_head t
return
if t.type is TYPE_EOF
# TODO ?correct for: "if the current node is not the root html element"
if open_els.length isnt 1
parse_error()
stop_parsing()
return
# Anything else
parse_error()
return
# 8.2.5.4.21 http://www.w3.org/TR/html5/syntax.html#parsing-main-afterframeset
ins_mode_after_frameset = (t) ->
if is_space_tok t
insert_character t
return
if t.type is TYPE_COMMENT
insert_comment t
return
if t.type is TYPE_DOCTYPE
parse_error()
return
if t.type is TYPE_START_TAG and t.name is 'html'
ins_mode_in_body t
return
if t.type is TYPE_END_TAG and t.name is 'html'
insert_mode = ins_mode_after_after_frameset
return
if t.type is TYPE_START_TAG and t.name is 'noframes'
ins_mode_in_head t
return
if t.type is TYPE_EOF
stop_parsing()
return
# Anything else
parse_error()
return
# 8.2.5.4.22 http://www.w3.org/TR/html5/syntax.html#the-after-after-body-insertion-mode
ins_mode_after_after_body = (t) ->
if t.type is TYPE_COMMENT
insert_comment t, [doc, doc.children.length]
return
if t.type is TYPE_DOCTYPE or is_space_tok(t) or (t.type is TYPE_START_TAG and t.name is 'html')
ins_mode_in_body t
return
if t.type is TYPE_EOF
stop_parsing()
return
# Anything else
parse_error()
ins_mode = ins_mode_in_body
return
# 8.2.5.4.23 http://www.w3.org/TR/html5/syntax.html#the-after-after-frameset-insertion-mode
ins_mode_after_after_frameset = (t) ->
if t.type is TYPE_COMMENT
insert_comment t, [doc, doc.children.length]
return
if t.type is TYPE_DOCTYPE or is_space_tok(t) or (t.type is TYPE_START_TAG and t.name is 'html')
ins_mode_in_body t
return
if t.type is TYPE_EOF
stop_parsing()
return
if t.type is TYPE_START_TAG and t.name is 'noframes'
ins_mode_in_head t
return
# Anything else
parse_error()
return
# 8.2.4.1 http://www.w3.org/TR/html5/syntax.html#data-state
tok_state_data = ->
switch c = txt.charAt(cur++)
when '&'
return new_text_node parse_character_reference()
when '<'
tok_state = tok_state_tag_open
when "\u0000"
parse_error()
return new_text_node c
when '' # EOF
return new_eof_token()
else
return new_text_node c
return null
# 8.2.4.2 http://www.w3.org/TR/html5/syntax.html#character-reference-in-data-state
# not needed: tok_state_character_reference_in_data = ->
# just call parse_character_reference()
# 8.2.4.3 http://www.w3.org/TR/html5/syntax.html#rcdata-state
tok_state_rcdata = ->
switch c = txt.charAt(cur++)
when '&'
return new_text_node parse_character_reference()
when '<'
tok_state = tok_state_rcdata_less_than_sign
when "\u0000"
parse_error()
return new_character_token "\ufffd"
when '' # EOF
return new_eof_token()
else
return new_character_token c
return null
# 8.2.4.4 http://www.w3.org/TR/html5/syntax.html#character-reference-in-rcdata-state
# not needed: tok_state_character_reference_in_rcdata = ->
# just call parse_character_reference()
# 8.2.4.5 http://www.w3.org/TR/html5/syntax.html#rawtext-state
tok_state_rawtext = ->
switch c = txt.charAt(cur++)
when '<'
tok_state = tok_state_rawtext_less_than_sign
when "\u0000"
parse_error()
return new_character_token "\ufffd"
when '' # EOF
return new_eof_token()
else
return new_character_token c
return null
# 8.2.4.6 http://www.w3.org/TR/html5/syntax.html#script-data-state
tok_state_script_data = ->
switch c = txt.charAt(cur++)
when '<'
tok_state = tok_state_script_data_less_than_sign
when "\u0000"
parse_error()
return new_character_token "\ufffd"
when '' # EOF
return new_eof_token()
else
return new_character_token c
return null
# 8.2.4.7 http://www.w3.org/TR/html5/syntax.html#plaintext-state
tok_state_plaintext = ->
switch c = txt.charAt(cur++)
when "\u0000"
parse_error()
return new_character_token "\ufffd"
when '' # EOF
return new_eof_token()
else
return new_character_token c
return null
# 8.2.4.8 http://www.w3.org/TR/html5/syntax.html#tag-open-state
tok_state_tag_open = ->
switch c = txt.charAt(cur++)
when '!'
tok_state = tok_state_markup_declaration_open
when '/'
tok_state = tok_state_end_tag_open
when '?'
parse_error()
tok_cur_tag = new_comment_token '?'
tok_state = tok_state_bogus_comment
else
if is_lc_alpha(c)
tok_cur_tag = new_open_tag c
tok_state = tok_state_tag_name
else if is_uc_alpha(c)
tok_cur_tag = new_open_tag c.toLowerCase()
tok_state = tok_state_tag_name
else
parse_error()
tok_state = tok_state_data
cur -= 1 # we didn't parse/handle the char after <
return new_text_node '<'
return null
# 8.2.4.9 http://www.w3.org/TR/html5/syntax.html#end-tag-open-state
tok_state_end_tag_open = ->
switch c = txt.charAt(cur++)
when '>'
parse_error()
tok_state = tok_state_data
when '' # EOF
parse_error()
tok_state = tok_state_data
return new_text_node ''
else
if is_uc_alpha(c)
tok_cur_tag = new_end_tag c.toLowerCase()
tok_state = tok_state_tag_name
else if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
tok_state = tok_state_tag_name
else
parse_error()
tok_cur_tag = new_comment_token '/'
tok_state = tok_state_bogus_comment
return null
# 8.2.4.10 http://www.w3.org/TR/html5/syntax.html#tag-name-state
tok_state_tag_name = ->
switch c = txt.charAt(cur++)
when "\t", "\n", "\u000c", ' '
tok_state = tok_state_before_attribute_name
when '/'
tok_state = tok_state_self_closing_start_tag
when '>'
tok_state = tok_state_data
tmp = tok_cur_tag
tok_cur_tag = null
return tmp
when "\u0000"
parse_error()
tok_cur_tag.name += "\ufffd"
when '' # EOF
parse_error()
tok_state = tok_state_data
else
if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
else
tok_cur_tag.name += c
return null
# 8.2.4.11 http://www.w3.org/TR/html5/syntax.html#rcdata-less-than-sign-state
tok_state_rcdata_less_than_sign = ->
c = txt.charAt(cur++)
if c is '/'
temporary_buffer = ''
tok_state = tok_state_rcdata_end_tag_open
return null
# Anything else
tok_state = tok_state_rcdata
cur -= 1 # reconsume the input character
return new_character_token '<'
# 8.2.4.12 http://www.w3.org/TR/html5/syntax.html#rcdata-end-tag-open-state
tok_state_rcdata_end_tag_open = ->
c = txt.charAt(cur++)
if is_uc_alpha(c)
tok_cur_tag = new_end_tag c.toLowerCase()
temporary_buffer += c
tok_state = tok_state_rcdata_end_tag_name
return null
if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
temporary_buffer += c
tok_state = tok_state_rcdata_end_tag_name
return null
# Anything else
tok_state = tok_state_rcdata
cur -= 1 # reconsume the input character
return new_character_token "" # fixfull separate these
# http://www.w3.org/TR/html5/syntax.html#appropriate-end-tag-token
is_appropriate_end_tag = (t) ->
# spec says to check against "the tag name of the last start tag to
# have been emitted from this tokenizer", but this is only called from
# the various "raw" states, which I'm pretty sure all push the start
# token onto open_els. TODO: verify this after the script data states
# are implemented
debug_log "#{t.type}, #{t.name} open_els: #{serialize_els open_els, true, true}"
return t.type is TYPE_END_TAG and t.name is open_els[0].name
# 8.2.4.13 http://www.w3.org/TR/html5/syntax.html#rcdata-end-tag-name-state
tok_state_rcdata_end_tag_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_before_attribute_name
return
# else fall through to "Anything else"
if c is '/'
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_self_closing_start_tag # FIXME spec typo?
return
# else fall through to "Anything else"
if c is '>'
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_data
return tok_cur_tag
# else fall through to "Anything else"
if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
temporary_buffer += c
return null
if is_lc_alpha(c)
tok_cur_tag.name += c
temporary_buffer += c
return null
# Anything else
tok_state = tok_state_rcdata
cur -= 1 # reconsume the input character
return new_character_token '' + temporary_buffer # fixfull separate these
# 8.2.4.14 http://www.w3.org/TR/html5/syntax.html#rawtext-less-than-sign-state
tok_state_rawtext_less_than_sign = ->
c = txt.charAt(cur++)
if c is '/'
temporary_buffer = ''
tok_state = tok_state_rawtext_end_tag_open
return null
# Anything else
tok_state = tok_state_rawtext
cur -= 1 # reconsume the input character
return new_character_token '<'
# 8.2.4.15 http://www.w3.org/TR/html5/syntax.html#rawtext-end-tag-open-state
tok_state_rawtext_end_tag_open = ->
c = txt.charAt(cur++)
if is_uc_alpha(c)
tok_cur_tag = new_end_tag c.toLowerCase()
temporary_buffer += c
tok_state = tok_state_rawtext_end_tag_name
return null
if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
temporary_buffer += c
tok_state = tok_state_rawtext_end_tag_name
return null
# Anything else
tok_state = tok_state_rawtext
cur -= 1 # reconsume the input character
return new_character_token "" # fixfull separate these
# 8.2.4.16 http://www.w3.org/TR/html5/syntax.html#rawtext-end-tag-name-state
tok_state_rawtext_end_tag_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_before_attribute_name
return
# else fall through to "Anything else"
if c is '/'
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_self_closing_start_tag
return
# else fall through to "Anything else"
if c is '>'
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_data
return tok_cur_tag
# else fall through to "Anything else"
if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
temporary_buffer += c
return null
if is_lc_alpha(c)
tok_cur_tag.name += c
temporary_buffer += c
return null
# Anything else
tok_state = tok_state_rawtext
cur -= 1 # reconsume the input character
return new_character_token '' + temporary_buffer # fixfull separate these
# 8.2.4.17 http://www.w3.org/TR/html5/syntax.html#script-data-less-than-sign-state
tok_state_script_data_less_than_sign = ->
c = txt.charAt(cur++)
if c is '/'
temporary_buffer = ''
tok_state = tok_state_script_data_end_tag_open
return
if c is '!'
tok_state = tok_state_script_data_escape_start
return new_character_token '
c = txt.charAt(cur++)
if is_uc_alpha(c)
tok_cur_tag = new_end_tag c.toLowerCase()
temporary_buffer += c
tok_state = tok_state_script_data_end_tag_name
return
if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
temporary_buffer += c
tok_state = tok_state_script_data_end_tag_name
return
# Anything else
tok_state = tok_state_script_data
cur -= 1 # Reconsume
return new_character_token ''
# 8.2.4.19 http://www.w3.org/TR/html5/syntax.html#script-data-end-tag-open-state
tok_state_script_data_end_tag_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_before_attribute_name
return
# fall through
if c is '/'
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_self_closing_start_tag
return
# fall through
if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
temporary_buffer += c
return
if is_lc_alpha(c)
tok_cur_tag.name += c
temporary_buffer += c
return
# Anything else
tok_state = tok_state_script_data
cur -= 1 # Reconsume
return new_character_token "#{temporary_buffer}" # fixfull split
# 8.2.4.20 http://www.w3.org/TR/html5/syntax.html#script-data-escape-start-state
tok_state_script_data_escape_start = ->
c = txt.charAt(cur++)
if c is '-'
tok_state = tok_state_script_data_escape_start_dash
return new_character_token '-'
# Anything else
tok_state = tok_state_script_data
cur -= 1 # Reconsume
return
# 8.2.4.21 http://www.w3.org/TR/html5/syntax.html#script-data-escape-start-dash-state
tok_state_script_data_escape_start_dash = ->
c = txt.charAt(cur++)
if c is '-'
tok_state = tok_state_script_data_escaped_dash_dash
return new_character_token '-'
# Anything else
tok_state = tok_state_script_data
cur -= 1 # Reconsume
return
# 8.2.4.22 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-state
tok_state_script_data_escaped = ->
c = txt.charAt(cur++)
if c is '-'
tok_state = tok_state_script_data_escaped_dash
return new_character_token '-'
if c is '<'
tok_state = tok_state_script_data_escaped_less_than_sign
return
if c is "\u0000"
parse_error()
return new_character_token "\ufffd"
if c is '' # EOF
tok_state = tok_state_data
parse_error()
cur -= 1 # Reconsume
return
# Anything else
return new_character_token c
# 8.2.4.23 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-dash-state
tok_state_script_data_escaped_dash = ->
c = txt.charAt(cur++)
if c is '-'
tok_state = tok_state_script_data_escaped_dash_dash
return new_character_token '-'
if c is '<'
tok_state = tok_state_script_data_escaped_less_than_sign
return
if c is "\u0000"
parse_error()
tok_state = tok_state_script_data_escaped
return new_character_token "\ufffd"
if c is '' # EOF
tok_state = tok_state_data
parse_error()
cur -= 1 # Reconsume
return
# Anything else
tok_state = tok_state_script_data_escaped
return new_character_token c
# 8.2.4.24 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-dash-dash-state
tok_state_script_data_escaped_dash_dash = ->
c = txt.charAt(cur++)
if c is '-'
return new_character_token '-'
if c is '<'
tok_state = tok_state_script_data_escaped_less_than_sign
return
if c is '>'
tok_state = tok_state_script_data
return new_character_token '>'
if c is "\u0000"
parse_error()
tok_state = tok_state_script_data_escaped
return new_character_token "\ufffd"
if c is '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return
# Anything else
tok_state = tok_state_script_data_escaped
return new_character_token c
# 8.2.4.25 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-less-than-sign-state
tok_state_script_data_escaped_less_than_sign = ->
c = txt.charAt(cur++)
if c is '/'
temporary_buffer = ''
tok_state = tok_state_script_data_escaped_end_tag_open
return
if is_uc_alpha(c)
temporary_buffer = c.toLowerCase() # yes, really
tok_state = tok_state_script_data_double_escape_start
return new_character_token "<#{c}" # fixfull split
if is_lc_alpha(c)
temporary_buffer = c
tok_state = tok_state_script_data_double_escape_start
return new_character_token "<#{c}" # fixfull split
# Anything else
tok_state = tok_state_script_data_escaped
cur -= 1 # Reconsume
return new_character_token c
# 8.2.4.26 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-end-tag-open-state
tok_state_script_data_escaped_end_tag_open = ->
c = txt.charAt(cur++)
if is_uc_alpha(c)
tok_cur_tag = new_end_tag c.toLowerCase()
temporary_buffer += c
tok_state = tok_state_script_data_escaped_end_tag_name
return
if is_lc_alpha(c)
tok_cur_tag = new_end_tag c
temporary_buffer += c
tok_state = tok_state_script_data_escaped_end_tag_name
return
# Anything else
tok_state = tok_state_script_data_escaped
cur -= 1 # Reconsume
return new_character_token '' # fixfull split
# 8.2.4.27 http://www.w3.org/TR/html5/syntax.html#script-data-escaped-end-tag-name-state
tok_state_script_data_escaped_end_tag_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_before_attribute_name
return
# fall through
if c is '/'
if is_appropriate_end_tag tok_cur_tag
tok_state = tok_state_self_closing_start_tag
return
# fall through
if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
temporary_buffer += c.toLowerCase()
return
if is_lc_alpha(c)
tok_cur_tag.name += c
temporary_buffer += c.toLowerCase()
return
# Anything else
tok_state = tok_state_script_data_escaped
cur -= 1 # Reconsume
return new_character_token "#{temporary_buffer}" # fixfull split
# 8.2.4.28 http://www.w3.org/TR/html5/syntax.html#script-data-double-escape-start-state
tok_state_script_data_double_escape_start = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' ' or c is '/' or c is '>'
if temporary_buffer is 'script'
tok_state = tok_state_script_data_double_escaped
else
tok_state = tok_state_script_data_escaped
return new_character_token c
if is_uc_alpha(c)
temporary_buffer += c.toLowerCase() # yes, really lowercase
return new_character_token c
if is_lc_alpha(c)
temporary_buffer += c
return new_character_token c
# Anything else
tok_state = tok_state_script_data_escaped
cur -= 1 # Reconsume
return
# 8.2.4.29 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-state
tok_state_script_data_double_escaped = ->
c = txt.charAt(cur++)
if c is '-'
tok_state = tok_state_script_data_double_escaped_dash
return new_character_token '-'
if c is '<'
tok_state = tok_state_script_data_double_escaped_less_than_sign
return new_character_token '<'
if c is "\u0000"
parse_error()
return new_character_token "\ufffd"
if c is '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return
# Anything else
return new_character_token c
# 8.2.4.30 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-dash-state
tok_state_script_data_double_escaped_dash = ->
c = txt.charAt(cur++)
if c is '-'
tok_state = tok_state_script_data_double_escaped_dash_dash
return new_character_token '-'
if c is '<'
tok_state = tok_state_script_data_double_escaped_less_than_sign
return new_character_token '<'
if c is "\u0000"
parse_error()
tok_state = tok_state_script_data_double_escaped
return new_character_token "\ufffd"
if c is '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return
# Anything else
tok_state = tok_state_script_data_double_escaped
return new_character_token c
# 8.2.4.31 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-dash-dash-state
tok_state_script_data_double_escaped_dash_dash = ->
c = txt.charAt(cur++)
if c is '-'
return new_character_token '-'
if c is '<'
tok_state = tok_state_script_data_double_escaped_less_than_sign
return new_character_token '<'
if c is '>'
tok_state = tok_state_script_data
return new_character_token '>'
if c is "\u0000"
parse_error()
tok_state = tok_state_script_data_double_escaped
return new_character_token "\ufffd"
if c is '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return
# Anything else
tok_state = tok_state_script_data_double_escaped
return new_character_token c
# 8.2.4.32 http://www.w3.org/TR/html5/syntax.html#script-data-double-escaped-less-than-sign-state
tok_state_script_data_double_escaped_less_than_sign = ->
c = txt.charAt(cur++)
if c is '/'
temporary_buffer = ''
tok_state = tok_state_script_data_double_escape_end
return new_character_token '/'
# Anything else
tok_state = tok_state_script_data_double_escaped
cur -= 1 # Reconsume
return
# 8.2.4.33 http://www.w3.org/TR/html5/syntax.html#script-data-double-escape-end-state
tok_state_script_data_double_escape_end = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' ' or c is '/' or c is '>'
if temporary_buffer is 'script'
tok_state = tok_state_script_data_escaped
else
tok_state = tok_state_script_data_double_escaped
return new_character_token c
if is_uc_alpha(c)
temporary_buffer += c.toLowerCase() # yes, really lowercase
return new_character_token c
if is_lc_alpha(c)
temporary_buffer += c
return new_character_token c
# Anything else
tok_state = tok_state_script_data_double_escaped
cur -= 1 # Reconsume
return
# 8.2.4.34 http://www.w3.org/TR/html5/syntax.html#before-attribute-name-state
tok_state_before_attribute_name = ->
attr_name = null
switch c = txt.charAt(cur++)
when "\t", "\n", "\u000c", ' '
return null
when '/'
tok_state = tok_state_self_closing_start_tag
return null
when '>'
tok_state = tok_state_data
tmp = tok_cur_tag
tok_cur_tag = null
return tmp
when "\u0000"
parse_error()
attr_name = "\ufffd"
when '"', "'", '<', '='
parse_error()
attr_name = c
when '' # EOF
parse_error()
tok_state = tok_state_data
else
if is_uc_alpha(c)
attr_name = c.toLowerCase()
else
attr_name = c
if attr_name?
tok_cur_tag.attrs_a.unshift [attr_name, '']
tok_state = tok_state_attribute_name
return null
# 8.2.4.35 http://www.w3.org/TR/html5/syntax.html#attribute-name-state
tok_state_attribute_name = ->
switch c = txt.charAt(cur++)
when "\t", "\n", "\u000c", ' '
tok_state = tok_state_after_attribute_name
when '/'
tok_state = tok_state_self_closing_start_tag
when '='
tok_state = tok_state_before_attribute_value
when '>'
tok_state = tok_state_data
tmp = tok_cur_tag
tok_cur_tag = null
return tmp
when "\u0000"
parse_error()
tok_cur_tag.attrs_a[0][0] = "\ufffd"
when '"', "'", '<'
parse_error()
tok_cur_tag.attrs_a[0][0] = c
when '' # EOF
parse_error()
tok_state = tok_state_data
else
if is_uc_alpha(c)
tok_cur_tag.attrs_a[0][0] = c.toLowerCase()
else
tok_cur_tag.attrs_a[0][0] += c
return null
# 8.2.4.36 http://www.w3.org/TR/html5/syntax.html#after-attribute-name-state
tok_state_after_attribute_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\n" or c is "\u000c" or c is ' '
return
if c is '/'
tok_state = tok_state_self_closing_start_tag
return
if c is '='
tok_state = tok_state_before_attribute_value
return
if c is '>'
tok_state = tok_state_data
return
if is_uc_alpha(c)
tok_cur_tag.attrs_a.unshift [c.toLowerCase(), '']
tok_state = tok_state_attribute_name
return
if c is "\u0000"
parse_error()
tok_cur_tag.attrs_a.unshift ["\ufffd", '']
tok_state = tok_state_attribute_name
return
if c is '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # reconsume
return
if c is '"' or c is "'" or c is '<'
parse_error()
# fall through to Anything else
# Anything else
tok_cur_tag.attrs_a.unshift [c, '']
tok_state = tok_state_attribute_name
# 8.2.4.37 http://www.w3.org/TR/html5/syntax.html#before-attribute-value-state
tok_state_before_attribute_value = ->
switch c = txt.charAt(cur++)
when "\t", "\n", "\u000c", ' '
return null
when '"'
tok_state = tok_state_attribute_value_double_quoted
when '&'
tok_state = tok_state_attribute_value_unquoted
cur -= 1
when "'"
tok_state = tok_state_attribute_value_single_quoted
when "\u0000"
# Parse error
tok_cur_tag.attrs_a[0][1] += "\ufffd"
tok_state = tok_state_attribute_value_unquoted
when '>'
# Parse error
tok_state = tok_state_data
tmp = tok_cur_tag
tok_cur_tag = null
return tmp
when '' # EOF
parse_error()
tok_state = tok_state_data
else
tok_cur_tag.attrs_a[0][1] += c
tok_state = tok_state_attribute_value_unquoted
return null
# 8.2.4.38 http://www.w3.org/TR/html5/syntax.html#attribute-value-(double-quoted)-state
tok_state_attribute_value_double_quoted = ->
switch c = txt.charAt(cur++)
when '"'
tok_state = tok_state_after_attribute_value_quoted
when '&'
tok_cur_tag.attrs_a[0][1] += parse_character_reference '"', true
when "\u0000"
# Parse error
tok_cur_tag.attrs_a[0][1] += "\ufffd"
when '' # EOF
parse_error()
tok_state = tok_state_data
else
tok_cur_tag.attrs_a[0][1] += c
return null
# 8.2.4.39 http://www.w3.org/TR/html5/syntax.html#attribute-value-(single-quoted)-state
tok_state_attribute_value_single_quoted = ->
switch c = txt.charAt(cur++)
when "'"
tok_state = tok_state_after_attribute_value_quoted
when '&'
tok_cur_tag.attrs_a[0][1] += parse_character_reference "'", true
when "\u0000"
# Parse error
tok_cur_tag.attrs_a[0][1] += "\ufffd"
when '' # EOF
parse_error()
tok_state = tok_state_data
else
tok_cur_tag.attrs_a[0][1] += c
return null
# 8.2.4.40 http://www.w3.org/TR/html5/syntax.html#attribute-value-(unquoted)-state
tok_state_attribute_value_unquoted = ->
switch c = txt.charAt(cur++)
when "\t", "\n", "\u000c", ' '
tok_state = tok_state_before_attribute_name
when '&'
tok_cur_tag.attrs_a[0][1] += parse_character_reference '>', true
when '>'
tok_state = tok_state_data
tmp = tok_cur_tag
tok_cur_tag = null
return tmp
when "\u0000"
tok_cur_tag.attrs_a[0][1] += "\ufffd"
when '' # EOF
parse_error()
tok_state = tok_state_data
else
# Parse Error if ', <, = or ` (backtick)
tok_cur_tag.attrs_a[0][1] += c
return null
# 8.2.4.42 http://www.w3.org/TR/html5/syntax.html#after-attribute-value-(quoted)-state
tok_state_after_attribute_value_quoted = ->
switch c = txt.charAt(cur++)
when "\t", "\n", "\u000c", ' '
tok_state = tok_state_before_attribute_name
when '/'
tok_state = tok_state_self_closing_start_tag
when '>'
tok_state = tok_state_data
tmp = tok_cur_tag
tok_cur_tag = null
return tmp
when '' # EOF
parse_error()
tok_state = tok_state_data
else
# Parse Error
tok_state = tok_state_before_attribute_name
cur -= 1 # we didn't handle that char
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'
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 = ->
next_gt = txt.indexOf '>', cur
if next_gt is -1
val = txt.substr cur
cur = txt.length
else
val = txt.substr cur, (next_gt - cur)
cur = next_gt + 1
val = val.replace "\u0000", "\ufffd"
tok_cur_tag.text += val
tok_state = tok_state_data
return tok_cur_tag
# 8.2.4.45 http://www.w3.org/TR/html5/syntax.html#markup-declaration-open-state
tok_state_markup_declaration_open = ->
if txt.substr(cur, 2) is '--'
cur += 2
tok_cur_tag = new_comment_token ''
tok_state = tok_state_comment_start
return
if txt.substr(cur, 7).toLowerCase() is 'doctype'
cur += 7
tok_state = tok_state_doctype
return
acn = adjusted_current_node()
if acn and acn.namespace isnt NS_HTML and txt.substr(cur, 7) is '[CDATA['
cur += 7
tok_state = tok_state_cdata_section
return
# Otherwise
parse_error()
tok_cur_tag = new_comment_token '!' # TODO test ("!" right?)
tok_state = tok_state_bogus_comment
return
# 8.2.4.46 http://www.w3.org/TR/html5/syntax.html#comment-start-state
tok_state_comment_start = ->
switch c = txt.charAt(cur++)
when '-'
tok_state = tok_state_comment_start_dash
when "\u0000"
parse_error()
return new_character_token "\ufffd"
when '>'
parse_error()
tok_state = tok_state_data
return tok_cur_tag
when '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
else
tok_cur_tag.text += c
return null
# 8.2.4.47 http://www.w3.org/TR/html5/syntax.html#comment-start-dash-state
tok_state_comment_start_dash = ->
switch c = txt.charAt(cur++)
when '-'
tok_state = tok_state_comment_end
when "\u0000"
parse_error()
tok_cur_tag.text += "-\ufffd"
tok_state = tok_state_comment
when '>'
parse_error()
tok_state = tok_state_data
return tok_cur_tag
when '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
else
tok_cur_tag.text += "-#{c}"
tok_state = tok_state_comment
return null
# 8.2.4.48 http://www.w3.org/TR/html5/syntax.html#comment-state
tok_state_comment = ->
switch c = txt.charAt(cur++)
when '-'
tok_state = tok_state_comment_end_dash
when "\u0000"
parse_error()
tok_cur_tag.text += "\ufffd"
when '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
else
tok_cur_tag.text += c
return null
# 8.2.4.49 http://www.w3.org/TR/html5/syntax.html#comment-end-dash-state
tok_state_comment_end_dash = ->
switch c = txt.charAt(cur++)
when '-'
tok_state = tok_state_comment_end
when "\u0000"
parse_error()
tok_cur_tag.text += "-\ufffd"
tok_state = tok_state_comment
when '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
else
tok_cur_tag.text += "-#{c}"
tok_state = tok_state_comment
return null
# 8.2.4.50 http://www.w3.org/TR/html5/syntax.html#comment-end-state
tok_state_comment_end = ->
switch c = txt.charAt(cur++)
when '>'
tok_state = tok_state_data
return tok_cur_tag
when "\u0000"
parse_error()
tok_cur_tag.text += "--\ufffd"
tok_state = tok_state_comment
when '!'
parse_error()
tok_state = tok_state_comment_end_bang
when '-'
parse_error()
tok_cur_tag.text += '-'
when '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
else
parse_error()
tok_cur_tag.text += "--#{c}"
tok_state = tok_state_comment
return null
# 8.2.4.51 http://www.w3.org/TR/html5/syntax.html#comment-end-bang-state
tok_state_comment_end_bang = ->
switch c = txt.charAt(cur++)
when '-'
tok_cur_tag.text += "--!#{c}"
tok_state = tok_state_comment_end_dash
when '>'
tok_state = tok_state_data
return tok_cur_tag
when "\u0000"
parse_error()
tok_cur_tag.text += "--!\ufffd"
tok_state = tok_state_comment
when '' # EOF
parse_error()
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
else
tok_cur_tag.text += "--!#{c}"
tok_state = tok_state_comment
return null
# 8.2.4.52 http://www.w3.org/TR/html5/syntax.html#doctype-state
tok_state_doctype = ->
switch c = txt.charAt(cur++)
when "\t", "\u000a", "\u000c", ' '
tok_state = tok_state_before_doctype_name
when '' # EOF
parse_error()
tok_state = tok_state_data
el = new_doctype_token ''
el.flag 'force-quirks', true
cur -= 1 # Reconsume
return el
else
parse_error()
tok_state = tok_state_before_doctype_name
cur -= 1 # Reconsume
return null
# 8.2.4.52 http://www.w3.org/TR/html5/syntax.html#doctype-state
tok_state_before_doctype_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
if is_uc_alpha(c)
tok_cur_tag = new_doctype_token c.toLowerCase()
tok_state = tok_state_doctype_name
return
if c is "\u0000"
parse_error()
tok_cur_tag = new_doctype_token "\ufffd"
tok_state = tok_state_doctype_name
return
if c is '>'
parse_error()
el = new_doctype_token ''
el.flag 'force-quirks', true
tok_state = tok_state_data
return el
if c is '' # EOF
parse_error()
tok_state = tok_state_data
el = new_doctype_token ''
el.flag 'force-quirks', true
cur -= 1 # Reconsume
return el
# Anything else
tok_cur_tag = new_doctype_token c
tok_state = tok_state_doctype_name
return null
# 8.2.4.54 http://www.w3.org/TR/html5/syntax.html#doctype-name-state
tok_state_doctype_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
tok_state = tok_state_after_doctype_name
return
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
if is_uc_alpha(c)
tok_cur_tag.name += c.toLowerCase()
return
if c is "\u0000"
parse_error()
tok_cur_tag.name += "\ufffd"
return
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
tok_cur_tag.name += c
return null
# 8.2.4.55 http://www.w3.org/TR/html5/syntax.html#after-doctype-name-state
tok_state_after_doctype_name = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
if txt.substr(cur - 1, 6).toLowerCase() is 'public'
cur += 5
tok_state = tok_state_after_doctype_public_keyword
return
if txt.substr(cur - 1, 6).toLowerCase() is 'system'
cur += 5
tok_state = tok_state_after_doctype_system_keyword
return
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.56 http://www.w3.org/TR/html5/syntax.html#after-doctype-public-keyword-state
tok_state_after_doctype_public_keyword = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
tok_state = tok_state_before_doctype_public_identifier
return
if c is '"'
parse_error()
tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
tok_state = tok_state_doctype_public_identifier_double_quoted
return
if c is "'"
parse_error()
tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
tok_state = tok_state_doctype_public_identifier_single_quoted
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.57 http://www.w3.org/TR/html5/syntax.html#before-doctype-public-identifier-state
tok_state_before_doctype_public_identifier = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
if c is '"'
parse_error()
tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
tok_state = tok_state_doctype_public_identifier_double_quoted
return
if c is "'"
parse_error()
tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text?
tok_state = tok_state_doctype_public_identifier_single_quoted
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.58 http://www.w3.org/TR/html5/syntax.html#doctype-public-identifier-(double-quoted)-state
tok_state_doctype_public_identifier_double_quoted = ->
c = txt.charAt(cur++)
if c is '"'
tok_state = tok_state_after_doctype_public_identifier
return
if c is "\u0000"
parse_error()
tok_cur_tag.public_identifier += "\ufffd"
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
tok_cur_tag.public_identifier += c
return null
# 8.2.4.59 http://www.w3.org/TR/html5/syntax.html#doctype-public-identifier-(single-quoted)-state
tok_state_doctype_public_identifier_single_quoted = ->
c = txt.charAt(cur++)
if c is "'"
tok_state = tok_state_after_doctype_public_identifier
return
if c is "\u0000"
parse_error()
tok_cur_tag.public_identifier += "\ufffd"
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
tok_cur_tag.public_identifier += c
return null
# 8.2.4.60 http://www.w3.org/TR/html5/syntax.html#after-doctype-public-identifier-state
tok_state_after_doctype_public_identifier = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
tok_state = tok_state_between_doctype_public_and_system_identifiers
return
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
if c is '"'
parse_error()
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_double_quoted
return
if c is "'"
parse_error()
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_single_quoted
return
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.61 http://www.w3.org/TR/html5/syntax.html#between-doctype-public-and-system-identifiers-state
tok_state_between_doctype_public_and_system_identifiers = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
if c is '"'
parse_error()
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_double_quoted
return
if c is "'"
parse_error()
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_single_quoted
return
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.62 http://www.w3.org/TR/html5/syntax.html#after-doctype-system-keyword-state
tok_state_after_doctype_system_keyword = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
tok_state = tok_state_before_doctype_system_identifier
return
if c is '"'
parse_error()
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_double_quoted
return
if c is "'"
parse_error()
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_single_quoted
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.63 http://www.w3.org/TR/html5/syntax.html#before-doctype-system-identifier-state
tok_state_before_doctype_system_identifier = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
if c is '"'
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_double_quoted
return
if c is "'"
tok_cur_tag.system_identifier = ''
tok_state = tok_state_doctype_system_identifier_single_quoted
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.64 http://www.w3.org/TR/html5/syntax.html#doctype-system-identifier-(double-quoted)-state
tok_state_doctype_system_identifier_double_quoted = ->
c = txt.charAt(cur++)
if c is '"'
tok_state = tok_state_after_doctype_system_identifier
return
if c is "\u0000"
parse_error()
tok_cur_tag.system_identifier += "\ufffd"
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
tok_cur_tag.system_identifier += c
return null
# 8.2.4.65 http://www.w3.org/TR/html5/syntax.html#doctype-system-identifier-(single-quoted)-state
tok_state_doctype_system_identifier_single_quoted = ->
c = txt.charAt(cur++)
if c is "'"
tok_state = tok_state_after_doctype_system_identifier
return
if c is "\u0000"
parse_error()
tok_cur_tag.system_identifier += "\ufffd"
return
if c is '>'
parse_error()
tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
tok_cur_tag.system_identifier += c
return null
# 8.2.4.66 http://www.w3.org/TR/html5/syntax.html#after-doctype-system-identifier-state
tok_state_after_doctype_system_identifier = ->
c = txt.charAt(cur++)
if c is "\t" or c is "\u000a" or c is "\u000c" or c is ' '
return
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
parse_error()
tok_state = tok_state_data
tok_cur_tag.flag 'force-quirks', true
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
parse_error()
# do _not_ tok_cur_tag.flag 'force-quirks', true
tok_state = tok_state_bogus_doctype
return null
# 8.2.4.67 http://www.w3.org/TR/html5/syntax.html#bogus-doctype-state
tok_state_bogus_doctype = ->
c = txt.charAt(cur++)
if c is '>'
tok_state = tok_state_data
return tok_cur_tag
if c is '' # EOF
tok_state = tok_state_data
cur -= 1 # Reconsume
return tok_cur_tag
# Anything else
return null
# 8.2.4.68 http://www.w3.org/TR/html5/syntax.html#cdata-section-state
tok_state_cdata_section = ->
tok_state = tok_state_data
next_gt = txt.indexOf ']]>', cur
if next_gt is -1
val = txt.substr cur
cur = txt.length
else
val = txt.substr cur, (next_gt - cur)
cur = next_gt + 3
val = val.replace "\u0000", "\ufffd" # fixfull spec doesn't say this
return new_character_token val # fixfull split
# 8.2.4.69 http://www.w3.org/TR/html5/syntax.html#consume-a-character-reference
# Don't set this as a state, just call it
# returns a string (NOT a text node)
parse_character_reference = (allowed_char = null, in_attr = false) ->
if cur >= txt.length
return '&'
switch c = txt.charAt(cur)
when "\t", "\n", "\u000c", ' ', '<', '&', '', allowed_char
# explicitly not a parse error
return '&'
when ';'
# there has to be "one or more" alnums between & and ; to be a parse error
return '&'
when '#'
if cur + 1 >= txt.length
return '&'
if txt.charAt(cur + 1).toLowerCase() is 'x'
prefix = '#x'
charset = hex_chars
start = cur + 2
else
charset = digits
start = cur + 1
prefix = '#'
i = 0
while start + i < txt.length and charset.indexOf(txt.charAt(start + i)) > -1
i += 1
if i is 0
return '&'
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 '&'
else
for i in [0...31]
if alnum.indexOf(txt.charAt(cur + i)) is -1
break
if i is 0
# exit early, because parse_error() below needs at least one alnum
return '&'
if txt.charAt(cur + i) is ';'
i += 1 # include ';' terminator in value
decoded = decode_named_char_ref txt.substr(cur, i)
if decoded?
cur += i
return decoded
parse_error()
return '&'
else
# no ';' terminator (only legacy char refs)
max = i
for i in [2..max] # no prefix matches, so ok to check shortest first
c = legacy_char_refs[txt.substr(cur, i)]
if c?
if in_attr
if txt.charAt(cur + i) is '='
# "because some legacy user agents will
# misinterpret the markup in those cases"
parse_error()
return '&'
if alnum.indexOf(txt.charAt(cur + i)) > -1
# this makes attributes forgiving about url args
return '&'
# ok, and besides the weird exceptions for attributes...
# return the matching char
cur += i # consume entity chars
parse_error() # because no terminating ";"
return c
parse_error()
return '&'
return # never reached
# 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 = []
afe = [] # active formatting elements
template_ins_modes = []
ins_mode = ins_mode_initial
original_ins_mode = ins_mode # TODO check spec
flag_scripting = true # TODO might need an extra flag to get to parse correctly
flag_frameset_ok = true
flag_parsing = true
flag_foster_parenting = false
form_element_pointer = null
temporary_buffer = null
pending_table_character_tokens = []
head_element_pointer = null
flag_fragment_parsing = false # parser originally created as part of the html fragment parsing algorithm (fragment case)
context_element = null # FIXME initialize from args.fragment http://www.w3.org/TR/html5/syntax.html#parsing-html-fragments
# tokenizer initialization
tok_state = tok_state_data
# proccess input
while flag_parsing
t = tok_state()
if t?
ins_mode t
# fixfull parse error if has self-closing flag, but it wasn't acknolwedged
return doc.children
serialize_els = (els, shallow, show_ids) ->
serialized = ''
sep = ''
for t in els
serialized += sep
sep = ','
serialized += t.serialize shallow, show_ids
return serialized
# TODO export TYPE_*
module.exports.parse_html = parse_html
module.exports.debug_log_reset = debug_log_reset
module.exports.debug_log_each = debug_log_each
module.exports.TYPE_TAG = TYPE_TAG
module.exports.TYPE_TEXT = TYPE_TEXT
module.exports.TYPE_COMMENT = TYPE_COMMENT
module.exports.TYPE_DOCTYPE = TYPE_DOCTYPE