X-Git-Url: https://jasonwoof.com/gitweb/?a=blobdiff_plain;f=parse-html.coffee;h=fee4202f92205ab9df44d834ffacb3826534bd56;hb=66e9aafa4b99a448ce8082a4b9d238f2d1009f2c;hp=a6894cd4c2dbaf917b2ab2a6ec47a540ed8289bd;hpb=45a9823ee371376bec0c8295996ccf07c7d76580;p=peach-html5-editor.git diff --git a/parse-html.coffee b/parse-html.coffee index a6894cd..fee4202 100644 --- a/parse-html.coffee +++ b/parse-html.coffee @@ -195,8 +195,8 @@ 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 + return false unless t.type is TYPE_START_TAG + for a in t.attrs_a if a[0] is 'type' if a[1].toLowerCase() is 'hidden' return true @@ -320,6 +320,24 @@ formatting_elements = { u: true } +mathml_text_integration = { + mi: NS_MATHML, mo: NS_MATHML, mn: NS_MATHML, ms: NS_MATHML, mtext: NS_MATHML +} +is_mathml_text_integration_point = (el) -> + return mathml_text_integration[el.name] = el.namespace +is_html_integration = (el) -> # DON'T PASS A TOKEN + if el.namespace is NS_MATHML and el.name is 'annotation-xml' + if el.attrs.encoding? + if el.attrs.encoding.toLowerCase() is 'text/html' + return true + if el.attrs.encoding.toLowerCase() is 'application/xhtml+xml' + return true + return false + if el.namespace is NS_SVG + if el.name is 'foreignObject' or el.name is 'desc' or el.name is 'title' + return true + return false + h_tags = { h1:NS_HTML, h2:NS_HTML, h3:NS_HTML, h4:NS_HTML, h5:NS_HTML, h6:NS_HTML } @@ -355,6 +373,45 @@ adp_els = { address: NS_HTML, div: NS_HTML, p: NS_HTML } el_is_special_not_adp = (el) -> return special_elements[el.name] is el.namespace and adp_els[el.name] isnt el.namespace +svg_name_fixes = { + altglyph: 'altGlyph' + altglyphdef: 'altGlyphDef' + altglyphitem: 'altGlyphItem' + animatecolor: 'animateColor' + animatemotion: 'animateMotion' + animatetransform: 'animateTransform' + clippath: 'clipPath' + feblend: 'feBlend' + fecolormatrix: 'feColorMatrix' + fecomponenttransfer: 'feComponentTransfer' + fecomposite: 'feComposite' + feconvolvematrix: 'feConvolveMatrix' + fediffuselighting: 'feDiffuseLighting' + fedisplacementmap: 'feDisplacementMap' + fedistantlight: 'feDistantLight' + fedropshadow: 'feDropShadow' + feflood: 'feFlood' + fefunca: 'feFuncA' + fefuncb: 'feFuncB' + fefuncg: 'feFuncG' + fefuncr: 'feFuncR' + fegaussianblur: 'feGaussianBlur' + feimage: 'feImage' + femerge: 'feMerge' + femergenode: 'feMergeNode' + femorphology: 'feMorphology' + feoffset: 'feOffset' + fepointlight: 'fePointLight' + fespecularlighting: 'feSpecularLighting' + fespotlight: 'feSpotLight' + fetile: 'feTile' + feturbulence: 'feTurbulence' + foreignobject: 'foreignObject' + glyphref: 'glyphRef' + lineargradient: 'linearGradient' + radialgradient: 'radialGradient' + textpath: 'textPath' +} svg_attribute_fixes = { attributename: 'attributeName' attributetype: 'attributeType' @@ -455,8 +512,9 @@ decode_named_char_ref = (txt) -> return null if decoded is txt return g_dncr.cache[txt] = decoded -parse_html = (txt, parse_error_cb = null) -> - cur = 0 # index of next char in txt to be parsed +parse_html = (args) -> + txt = null + cur = null # index of next char in txt to be parsed # declare doc and tokenizer variables so they're in scope below doc = null open_els = null # stack of open elements @@ -481,8 +539,8 @@ parse_html = (txt, parse_error_cb = null) -> flag_parsing = false parse_error = -> - if parse_error_cb? - parse_error_cb cur + if args.error_cb? + args.error_cb cur else console.log "Parse error at character #{cur} of #{txt.length}" @@ -1081,6 +1139,36 @@ parse_html = (txt, parse_error_cb = null) -> return dest[0].children.splice dest[1], 0, t + + # 8.2.5 http://www.w3.org/TR/html5/syntax.html#tree-construction + process_token = (t) -> + acn = adjusted_current_node() + unless acn? + ins_mode t + return + if acn.namespace is NS_HTML + ins_mode t + return + if is_mathml_text_integration_point(acn) + if t.type is TYPE_START_TAG and (t.name is 'mglyph' or t.name is 'malignmark') + ins_mode t + return + if t.type is TYPE_TEXT + ins_mode t + return + if acn.namespace is NS_MATHML and acn.name is 'annotation-xml' and t.type is TYPE_START_TAG and t.name is 'svg' + ins_mode t + return + if is_html_integration acn + if t.type is TYPE_START_TAG or t.type is TYPE_TEXT + ins_mode t + return + if t.type is TYPE_EOF + ins_mode t + return + in_foreign_content t + return + # 8.2.5.1 # http://www.w3.org/TR/html5/syntax.html#creating-and-inserting-nodes # http://www.w3.org/TR/html5/syntax.html#appropriate-place-for-inserting-a-node @@ -1104,7 +1192,7 @@ parse_html = (txt, parse_error_cb = null) -> last_template = null last_template_i = null for el, i in open_els - if el.name is 'template' + if el.name is 'template' and el.namespace is NS_HTML last_template = el last_template_i = i break @@ -1113,7 +1201,7 @@ parse_html = (txt, parse_error_cb = null) -> last_table = null last_table_i for el, i in open_els - if el.name is 'table' + if el.name is 'table' and el.namespace is NS_HTML last_table = el last_table_i = i break @@ -1135,6 +1223,7 @@ parse_html = (txt, parse_error_cb = null) -> # this is odd target = open_els[open_els.length - 1] target_i = target.children.length + break # 5. If last table has a parent element, then let adjusted # insertion location be inside last table's parent element, # immediately before last table, and abort these substeps. @@ -1253,7 +1342,7 @@ parse_html = (txt, parse_error_cb = null) -> # Anything else #fixfull (iframe, quirks) ins_mode = ins_mode_before_html - ins_mode t # reprocess the token + process_token t return # 8.2.5.4.2 http://www.w3.org/TR/html5/syntax.html#the-before-html-insertion-mode @@ -1286,7 +1375,7 @@ parse_html = (txt, parse_error_cb = null) -> open_els.unshift el # ?fixfull browsing context ins_mode = ins_mode_before_head - ins_mode t + process_token t return # 8.2.5.4.3 http://www.w3.org/TR/html5/syntax.html#the-before-head-insertion-mode @@ -1306,6 +1395,7 @@ parse_html = (txt, parse_error_cb = null) -> el = insert_html_element t head_element_pointer = el ins_mode = ins_mode_in_head + return if t.type is TYPE_END_TAG if t.name is 'head' or t.name is 'body' or t.name is 'html' or t.name is 'br' # fall through to Anything else below @@ -1317,13 +1407,13 @@ parse_html = (txt, parse_error_cb = null) -> el = insert_html_element head_tok head_element_pointer = el ins_mode = ins_mode_in_head - ins_mode t # reprocess current token + process_token t # 8.2.5.4.4 http://www.w3.org/TR/html5/syntax.html#parsing-main-inhead ins_mode_in_head_else = (t) -> # factored out for same-as-spec flow control open_els.shift() # spec says this will be a 'head' node ins_mode = ins_mode_after_head - ins_mode t + process_token t ins_mode_in_head = (t) -> if t.type is TYPE_TEXT and (t.text is "\t" or t.text is "\n" or t.text is "\u000c" or t.text is ' ') insert_character t @@ -1351,7 +1441,7 @@ parse_html = (txt, parse_error_cb = null) -> if t.type is TYPE_START_TAG and t.name is 'title' parse_generic_rcdata_text t return - if t.type is TYPE_START_TAG and ((t.name is 'noscript' and flag_scripting) or (t.name is 'noframes' or t.name is 'style')) + if t.type is TYPE_START_TAG and ((t.name is 'noscript' and flag_scripting) or t.name is 'noframes' or t.name is 'style') parse_generic_raw_text t return if t.type is TYPE_START_TAG and t.name is 'noscript' and flag_scripting is false @@ -1408,19 +1498,19 @@ parse_html = (txt, parse_error_cb = null) -> parse_error() open_els.shift() ins_mode = ins_mode_in_head - ins_mode t + process_token t ins_mode_in_head_noscript = (t) -> if t.type is TYPE_DOCTYPE parse_error() return - if t.type is TYPE_START_TAG + if t.type is TYPE_START_TAG and t.name is 'html' ins_mode_in_body t return if t.type is TYPE_END_TAG and t.name is 'noscript' open_els.shift() ins_mode = ins_mode_in_head return - if (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')) + if is_space_tok(t) or t.type is TYPE_COMMENT or (t.type is TYPE_START_TAG and (t.name is 'basefont' or t.name is 'bgsound' or t.name is 'link' or t.name is 'meta' or t.name is 'noframes' or t.name is 'style')) ins_mode_in_head t return if t.type is TYPE_END_TAG and t.name is 'br' @@ -1440,7 +1530,7 @@ parse_html = (txt, parse_error_cb = null) -> body_tok = new_open_tag 'body' insert_html_element body_tok ins_mode = ins_mode_in_body - ins_mode t # reprocess token + process_token t return ins_mode_after_head = (t) -> if is_space_tok t @@ -1609,7 +1699,7 @@ parse_html = (txt, parse_error_cb = null) -> parse_error() break ins_mode = ins_mode_after_body - ins_mode t + process_token t return if t.type is TYPE_START_TAG and (t.name is 'address' or t.name is 'article' or t.name is 'aside' or t.name is 'blockquote' or t.name is 'center' or t.name is 'details' or t.name is 'dialog' or t.name is 'dir' or t.name is 'div' or t.name is 'dl' or t.name is 'fieldset' or t.name is 'figcaption' or t.name is 'figure' or t.name is 'footer' or t.name is 'header' or t.name is 'hgroup' or t.name is 'main' or t.name is 'nav' or t.name is 'ol' or t.name is 'p' or t.name is 'section' or t.name is 'summary' or t.name is 'ul') close_p_if_in_button_scope() @@ -1889,7 +1979,7 @@ parse_html = (txt, parse_error_cb = null) -> if t.type is TYPE_START_TAG and t.name is 'image' parse_error() t.name = 'img' - ins_mode t + process_token t return if t.type is TYPE_START_TAG and t.name is 'isindex' parse_error() @@ -1920,7 +2010,7 @@ parse_html = (txt, parse_error_cb = null) -> 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 + insert_character new_character_token prompt # fixfull split # TODO submit typo "balue" in spec insert_html_element input_el open_els.shift() @@ -2013,19 +2103,6 @@ parse_html = (txt, parse_error_cb = null) -> 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 @@ -2037,7 +2114,7 @@ parse_html = (txt, parse_error_cb = null) -> open_els[0].flag 'already started', true open_els.shift() ins_mode = original_ins_mode - ins_mode t + process_token t return if t.type is TYPE_END_TAG and t.name is 'script' open_els.shift() @@ -2055,13 +2132,26 @@ parse_html = (txt, parse_error_cb = null) -> # http://www.w3.org/TR/html5/syntax.html#tokenization # 8.2.5.4.9 http://www.w3.org/TR/html5/syntax.html#parsing-main-intable + ins_mode_in_table_else = (t) -> + parse_error() + flag_foster_parenting = true + ins_mode_in_body t + flag_foster_parenting = false + return + can_in_table = { # FIXME do this inline like everywhere else + 'table': true + 'tbody': true + 'tfoot': true + 'thead': true + 'tr': true + } ins_mode_in_table = (t) -> switch t.type when TYPE_TEXT if can_in_table[t.name] original_ins_mode = ins_mode ins_mode = ins_mode_in_table_text - ins_mode t + process_token t else ins_mode_in_table_else t when TYPE_COMMENT @@ -2083,7 +2173,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_stack_to_table_context() insert_html_element new_open_tag 'colgroup' ins_mode = ins_mode_in_column_group - ins_mode t + process_token t when 'tbody', 'tfoot', 'thead' clear_stack_to_table_context() insert_html_element t @@ -2092,7 +2182,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_stack_to_table_context() insert_html_element new_open_tag 'tbody' ins_mode = ins_mode_in_table_body - ins_mode t + process_token t when 'table' parse_error() if is_in_table_scope 'table' @@ -2101,11 +2191,11 @@ parse_html = (txt, parse_error_cb = null) -> if el.name is 'table' break reset_ins_mode() - ins_mode t + process_token t when 'style', 'script', 'template' ins_mode_in_head t when 'input' - if is_input_hidden_tok t + unless is_input_hidden_tok t ins_mode_in_table_else t else parse_error() @@ -2168,7 +2258,7 @@ parse_html = (txt, parse_error_cb = null) -> ins_mode_table_else old pending_table_character_tokens = [] # FIXME test (spec doesn't say this) ins_mode = original_ins_mode - ins_mode t + process_token t # 8.2.5.4.11 http://www.w3.org/TR/html5/syntax.html#parsing-main-incaption ins_mode_in_caption = (t) -> @@ -2196,7 +2286,7 @@ parse_html = (txt, parse_error_cb = null) -> break clear_afe_to_marker() ins_mode = ins_mode_in_table - ins_mode t + process_token t # else fragment case return if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr') @@ -2246,7 +2336,7 @@ parse_html = (txt, parse_error_cb = null) -> return open_els.shift() ins_mode = ins_mode_in_table - ins_mode t + process_token t return # 8.2.5.4.13 http://www.w3.org/TR/html5/syntax.html#parsing-main-intbody @@ -2261,7 +2351,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_stack_to_table_body_context() insert_html_element new_open_tag 'tr' ins_mode = ins_mode_in_row - ins_mode t + process_token t return if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead') unless is_in_table_scope t.name # fixfull check namespace @@ -2285,7 +2375,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_stack_to_table_body_context() open_els.shift() ins_mode = ins_mode_in_table - ins_mode t + process_token t return if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html' or t.name is 'td' or t.name is 'th' or t.name is 'tr') parse_error() @@ -2314,7 +2404,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_stack_to_table_row_context() open_els.shift() ins_mode = ins_mode_in_table_body - ins_mode t + process_token t else parse_error() return @@ -2324,7 +2414,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_stack_to_table_row_context() open_els.shift() ins_mode = ins_mode_in_table_body - ins_mode t + process_token t else parse_error() return @@ -2374,7 +2464,7 @@ parse_html = (txt, parse_error_cb = null) -> parse_error() return close_the_cell() - ins_mode t + process_token t return if t.type is TYPE_END_TAG and (t.name is 'body' or t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'html') parse_error() @@ -2382,7 +2472,7 @@ parse_html = (txt, parse_error_cb = null) -> 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 + process_token t else parse_error() return @@ -2461,7 +2551,7 @@ parse_html = (txt, parse_error_cb = null) -> if el.name is 'select' break reset_ins_mode() - ins_mode t + process_token t return if t.type is TYPE_START_TAG and (t.name is 'script' or t.name is 'template') ins_mode_in_head t @@ -2482,7 +2572,7 @@ parse_html = (txt, parse_error_cb = null) -> if el.name is 'select' break reset_ins_mode() - ins_mode t + process_token t return if t.type is TYPE_END_TAG and (t.name is 'caption' or t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' or t.name is 'td' or t.name is 'th') parse_error() @@ -2493,7 +2583,7 @@ parse_html = (txt, parse_error_cb = null) -> if el.name is 'select' break reset_ins_mode() - ins_mode t + process_token t return # Anything else ins_mode_in_select t @@ -2511,31 +2601,31 @@ parse_html = (txt, parse_error_cb = null) -> template_ins_modes.shift() template_ins_modes.unshift ins_mode_in_table ins_mode = ins_mode_in_table - ins_mode t + process_token t return if t.type is TYPE_START_TAG and t.name is 'col' template_ins_modes.shift() template_ins_modes.unshift ins_mode_in_column_group ins_mode = ins_mode_in_column_group - ins_mode t + process_token t return if t.type is TYPE_START_TAG and t.name is 'tr' template_ins_modes.shift() template_ins_modes.unshift ins_mode_in_table_body ins_mode = ins_mode_in_table_body - ins_mode t + process_token t return if t.type is TYPE_START_TAG and (t.name is 'td' or t.name is 'th') template_ins_modes.shift() template_ins_modes.unshift ins_mode_in_row ins_mode = ins_mode_in_row - ins_mode t + process_token t return if t.type is TYPE_START_TAG template_ins_modes.shift() template_ins_modes.unshift ins_mode_in_body ins_mode = ins_mode_in_body - ins_mode t + process_token t return if t.type is TYPE_END_TAG parse_error() @@ -2552,7 +2642,7 @@ parse_html = (txt, parse_error_cb = null) -> clear_afe_to_marker() template_ins_modes.shift() reset_ins_mode() - ins_mode t + process_token t # 8.2.5.4.19 http://www.w3.org/TR/html5/syntax.html#parsing-main-afterbody ins_mode_after_body = (t) -> @@ -2578,7 +2668,7 @@ parse_html = (txt, parse_error_cb = null) -> # Anything ELse parse_error() ins_mode = ins_mode_in_body - ins_mode t + process_token t # 8.2.5.4.20 http://www.w3.org/TR/html5/syntax.html#parsing-main-inframeset ins_mode_in_frameset = (t) -> @@ -2685,8 +2775,84 @@ parse_html = (txt, parse_error_cb = null) -> parse_error() return - - + # 8.2.5.5 http://www.w3.org/TR/html5/syntax.html#parsing-main-inforeign + has_color_face_or_size = (t) -> + for a in t.attrs_a + if a[0] is 'color' or a[0] is 'face' or a[0] is 'size' + return true + return false + in_foreign_content_end_script = -> + open_els.shift() + # fixfull + return + in_foreign_content_other_start = (t) -> + acn = adjusted_current_node() + if acn.namespace is NS_MATHML + adjust_mathml_attributes t + if acn.namespace is NS_SVG and svg_name_fixes[t.name]? + t.name = svg_name_fixes[t.name] + if acn.namespace is NS_SVG + adjust_svg_attributes t + adjust_foreign_attributes t + insert_foreign_element t, acn.namespace + if t.flag 'self-closing' + if t.name is 'script' + t.acknowledge_self_closing() + in_foreign_content_end_script() + else + open_els.shift() + t.acknowledge_self_closing() + return + in_foreign_content = (t) -> + if t.type is TYPE_TEXT and t.text is "\u0000" + parse_error() + insert_character new_character_token "\ufffd" + return + if is_space_tok t + insert_character t + return + if t.type is TYPE_TEXT + flag_frameset_ok = false + insert_character t + return + if t.type is TYPE_COMMENT + insert_comment t + return + if t.type is TYPE_DOCTYPE + parse_error() + return + if t.type is TYPE_START_TAG and (t.name is 'b' or t.name is 'big' or t.name is 'blockquote' or t.name is 'body' or t.name is 'br' or t.name is 'center' or t.name is 'code' or t.name is 'dd' or t.name is 'div' or t.name is 'dl' or t.name is 'dt' or t.name is 'em' or t.name is 'embed' or t.name is 'h1' or t.name is 'h2' or t.name is 'h3' or t.name is 'h4' or t.name is 'h5' or t.name is 'h6' or t.name is 'head' or t.name is 'hr' or t.name is 'i' or t.name is 'img' or t.name is 'li' or t.name is 'listing' or t.name is 'main' or t.name is 'meta' or t.name is 'nobr' or t.name is 'ol' or t.name is 'p' or t.name is 'pre' or t.name is 'ruby' or t.name is 's' or t.name is 'small' or t.name is 'span' or t.name is 'strong' or t.name is 'strike' or t.name is 'sub' or t.name is 'sup' or t.name is 'table' or t.name is 'tt' or t.name is 'u' or t.name is 'ul' or t.name is 'var' or (t.name is 'font' and has_color_face_or_size(t))) + parse_error() + if flag_fragment_parsing + in_foreign_content_other_start t + return + loop # is this safe? + open_els.shift() + cn = open_els[0] + if is_mathml_text_integration_point(cn) or is_html_integration(cn) or cn.namespace is NS_HTML + break + process_token t + return + if t.type is TYPE_START_TAG + in_foreign_content_other_start t + return + if t.type is TYPE_END_TAG and t.name is 'script' and open_els[0].name is 'script' and open_els[0].namespace is NS_SVG + in_foreign_content_end_script() + return + if t.type is TYPE_END_TAG + if open_els[0].name.toLowerCase() isnt t.name + parse_error() + for node in open_els + if node is open_els[open_els.length - 1] + return + if node.name.toLowerCase() is t.name + loop + el = open_els.shift() + if el is node + return + if node.namespace is NS_HTML + break + ins_mode t # explicitly call HTML insertion mode # 8.2.4.1 http://www.w3.org/TR/html5/syntax.html#data-state @@ -2877,9 +3043,9 @@ parse_html = (txt, parse_error_cb = null) -> 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 + # the various "raw" states, so it's hopefully ok to assume that + # open_els[0].name will work instead 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 @@ -3021,6 +3187,11 @@ parse_html = (txt, parse_error_cb = null) -> tok_state = tok_state_self_closing_start_tag return # fall through + if c is '>' + if is_appropriate_end_tag tok_cur_tag + tok_state = tok_state_data + return tok_cur_tag + # fall through if is_uc_alpha(c) tok_cur_tag.name += c.toLowerCase() temporary_buffer += c @@ -3173,6 +3344,11 @@ parse_html = (txt, parse_error_cb = null) -> tok_state = tok_state_self_closing_start_tag return # fall through + if c is '>' + if is_appropriate_end_tag tok_cur_tag + tok_state = tok_state_data + return tok_cur_tag + # fall through if is_uc_alpha(c) tok_cur_tag.name += c.toLowerCase() temporary_buffer += c.toLowerCase() @@ -3353,16 +3529,16 @@ parse_html = (txt, parse_error_cb = null) -> return tmp when "\u0000" parse_error() - tok_cur_tag.attrs_a[0][0] = "\ufffd" + tok_cur_tag.attrs_a[0][0] += "\ufffd" when '"', "'", '<' parse_error() - tok_cur_tag.attrs_a[0][0] = c + tok_cur_tag.attrs_a[0][0] += c when '' # EOF parse_error() tok_state = tok_state_data else if is_uc_alpha(c) - tok_cur_tag.attrs_a[0][0] = c.toLowerCase() + tok_cur_tag.attrs_a[0][0] += c.toLowerCase() else tok_cur_tag.attrs_a[0][0] += c return null @@ -3560,7 +3736,7 @@ parse_html = (txt, parse_error_cb = null) -> return # Otherwise parse_error() - tok_cur_tag = new_comment_token '!' # TODO test ("!" right?) + tok_cur_tag = new_comment_token '' tok_state = tok_state_bogus_comment return @@ -3571,6 +3747,7 @@ parse_html = (txt, parse_error_cb = null) -> tok_state = tok_state_comment_start_dash when "\u0000" parse_error() + tok_state = tok_state_comment return new_character_token "\ufffd" when '>' parse_error() @@ -3583,6 +3760,7 @@ parse_html = (txt, parse_error_cb = null) -> return tok_cur_tag else tok_cur_tag.text += c + tok_state = tok_state_comment return null # 8.2.4.47 http://www.w3.org/TR/html5/syntax.html#comment-start-dash-state @@ -4131,7 +4309,9 @@ parse_html = (txt, parse_error_cb = null) -> else val = txt.substr cur, (next_gt - cur) cur = next_gt + 3 - val = val.replace "\u0000", "\ufffd" # fixfull spec doesn't say this + val = val.replace(new RegExp("\u0000", 'g'), "\ufffd") # fixfull spec doesn't say this + val = val.replace(new RegExp("\r\n", 'g'), "\n") # fixfull spec doesn't say this + val = val.replace(new RegExp("\r", 'g'), "\n") # 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 @@ -4212,13 +4392,15 @@ parse_html = (txt, parse_error_cb = null) -> # tree constructor initialization # see comments on TYPE_TAG/etc for the structure of this data + txt = args.html + cur = 0 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