X-Git-Url: https://jasonwoof.com/gitweb/?a=blobdiff_plain;f=parse-html.coffee;h=06c2a1b04e67c55ad0fbf7e5e1b2b78396efa99f;hb=3ff49c30096e8e97599b98755157f9a692937f58;hp=fee4202f92205ab9df44d834ffacb3826534bd56;hpb=66e9aafa4b99a448ce8082a4b9d238f2d1009f2c;p=peach-html5-editor.git diff --git a/parse-html.coffee b/parse-html.coffee index fee4202..06c2a1b 100644 --- a/parse-html.coffee +++ b/parse-html.coffee @@ -54,6 +54,15 @@ unless module?.exports? window.wheic = {} module = exports: window.wheic +from_code_point = (x) -> + if String.fromCodePoint? + return String.fromCodePoint x + else + if x <= 0xffff + return String.fromCharCode x + x -= 0x10000 + return String.fromCharCode((x >> 10) + 0xd800, (x % 0x400) + 0xdc00) + # Each node is an obect of the Node class. Here are the Node types: TYPE_TAG = 0 # name, {attributes}, [children] TYPE_TEXT = 1 # "text" @@ -206,6 +215,36 @@ is_input_hidden_tok = (t) -> # https://en.wikipedia.org/wiki/Whitespace_character#Unicode whitespace_chars = "\u0009\u000a\u000b\u000c\u000d\u0020\u0085\u00a0\u1680\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a\u2028\u2029\u202f\u205f\u3000" +unicode_fixes = {} +unicode_fixes[0x00] = "\uFFFD" +unicode_fixes[0x80] = "\u20AC" +unicode_fixes[0x82] = "\u201A" +unicode_fixes[0x83] = "\u0192" +unicode_fixes[0x84] = "\u201E" +unicode_fixes[0x85] = "\u2026" +unicode_fixes[0x86] = "\u2020" +unicode_fixes[0x87] = "\u2021" +unicode_fixes[0x88] = "\u02C6" +unicode_fixes[0x89] = "\u2030" +unicode_fixes[0x8A] = "\u0160" +unicode_fixes[0x8B] = "\u2039" +unicode_fixes[0x8C] = "\u0152" +unicode_fixes[0x8E] = "\u017D" +unicode_fixes[0x91] = "\u2018" +unicode_fixes[0x92] = "\u2019" +unicode_fixes[0x93] = "\u201C" +unicode_fixes[0x94] = "\u201D" +unicode_fixes[0x95] = "\u2022" +unicode_fixes[0x96] = "\u2013" +unicode_fixes[0x97] = "\u2014" +unicode_fixes[0x98] = "\u02DC" +unicode_fixes[0x99] = "\u2122" +unicode_fixes[0x9A] = "\u0161" +unicode_fixes[0x9B] = "\u203A" +unicode_fixes[0x9C] = "\u0153" +unicode_fixes[0x9E] = "\u017E" +unicode_fixes[0x9F] = "\u0178" + # 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 = { @@ -324,14 +363,15 @@ 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 + return mathml_text_integration[el.name] is 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 + if el.namespace is NS_MATHML + if el.name is 'annotation-xml' + if el.attrs.encoding? + if el.attrs.encoding.toLowerCase() is 'text/html' + return true + if el.attrs.encoding.toLowerCase() is 'application/xhtml+xml' + return true return false if el.namespace is NS_SVG if el.name is 'foreignObject' or el.name is 'desc' or el.name is 'title' @@ -342,28 +382,25 @@ 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 + table: NS_HTML + tbody: NS_HTML + tfoot: NS_HTML + thead: NS_HTML + tr: NS_HTML } -# 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 + dd: NS_HTML + dt: NS_HTML + li: NS_HTML + option: NS_HTML + optgroup: NS_HTML + p: NS_HTML + rb: NS_HTML + rp: NS_HTML + rt: NS_HTML + rtc: NS_HTML } el_is_special = (e) -> @@ -566,7 +603,7 @@ parse_html = (args) -> # 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' + if t.name is 'template' and t.namespace is NS_HTML return true return false is_in_scope_x = (tag_name, scope, namespace) -> @@ -615,11 +652,12 @@ parse_html = (args) -> return false return false # this checks for a particular element, not by name - el_is_in_scope = (el) -> - for t in open_els - if t is el + # this requires a namespace match + el_is_in_scope = (needle) -> + for el in open_els + if el is needle return true - if standard_scopers[t.name] is t.namespace + if standard_scopers[el.name] is el.namespace return false return false @@ -635,15 +673,15 @@ parse_html = (args) -> open_els.shift() return clear_to_table_body_stopers = { - 'tbody': true - 'tfoot': true - 'thead': true - 'template': true - 'html': true + tbody: NS_HTML + tfoot: NS_HTML + thead: NS_HTML + template: NS_HTML + html: NS_HTML } clear_stack_to_table_body_context = -> loop - if clear_to_table_body_stopers[open_els[0].name]? + if clear_to_table_body_stopers[open_els[0].name] is open_els[0].namespace break open_els.shift() return @@ -684,7 +722,7 @@ parse_html = (args) -> # fixfull (fragment case) # 4. If node is a select element, run these substeps: - if node.name is 'select' + if node.name is 'select' and node.namespace is NS_HTML # 1. If last is true, jump to the step below labeled done. unless last # 2. Let ancestor be node. @@ -701,11 +739,11 @@ parse_html = (args) -> ancestor = open_els[ancestor_i] # 5. If ancestor is a template node, jump to the step below # labeled done. - if ancestor.name is 'template' + if ancestor.name is 'template' and ancestor.namespace is NS_HTML break # 6. If ancestor is a table node, switch the insertion mode # to "in select in table" and abort these steps. - if ancestor.name is 'table' + if ancestor.name is 'table' and ancestor.namespace is NS_HTML ins_mode = ins_mode_in_select_in_table return # 7. Jump back to the step labeled loop. @@ -715,61 +753,62 @@ parse_html = (args) -> 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 + if (node.name is 'td' or node.name is 'th') and node.namespace is NS_HTML and last is false ins_mode = ins_mode_in_cell return # 6. If node is a tr element, then switch the insertion mode to "in # row" and abort these steps. - if node.name is 'tr' + if node.name is 'tr' and node.namespace is NS_HTML ins_mode = ins_mode_in_row return # 7. If node is a tbody, thead, or tfoot element, then switch the # insertion mode to "in table body" and abort these steps. - if node.name is 'tbody' or node.name is 'thead' or node.name is 'tfoot' + if (node.name is 'tbody' or node.name is 'thead' or node.name is 'tfoot') and node.namespace is NS_HTML ins_mode = ins_mode_in_table_body return # 8. If node is a caption element, then switch the insertion mode # to "in caption" and abort these steps. - if node.name is 'caption' + if node.name is 'caption' and node.namespace is NS_HTML ins_mode = ins_mode_in_caption return # 9. If node is a colgroup element, then switch the insertion mode # to "in column group" and abort these steps. - if node.name is 'colgroup' + if node.name is 'colgroup' and node.namespace is NS_HTML ins_mode = ins_mode_in_column_group return # 10. If node is a table element, then switch the insertion mode to # "in table" and abort these steps. - if node.name is 'table' + if node.name is 'table' and node.namespace is NS_HTML ins_mode = ins_mode_in_table return # 11. If node is a template element, then switch the insertion mode # to the current template insertion mode and abort these steps. - # fixfull (template insertion mode stack) - + if node.name is 'template' and node.namespace is NS_HTML + ins_mode = template_ins_modes[0] + return # 12. If node is a head element and last is true, then switch the # insertion mode to "in body" ("in body"! not "in head"!) and abort # these steps. (fragment case) - if node.name is 'head' and last + if node.name is 'head' and node.namespace is NS_HTML and last ins_mode = ins_mode_in_body return # 13. If node is a head element and last is false, then switch the # insertion mode to "in head" and abort these steps. - if node.name is 'head' and last is false + if node.name is 'head' and node.namespace is NS_HTML and last is false ins_mode = ins_mode_in_head return # 14. If node is a body element, then switch the insertion mode to # "in body" and abort these steps. - if node.name is 'body' + if node.name is 'body' and node.namespace is NS_HTML ins_mode = ins_mode_in_body return # 15. If node is a frameset element, then switch the insertion mode # to "in frameset" and abort these steps. (fragment case) - if node.name is 'frameset' + if node.name is 'frameset' and node.namespace is NS_HTML ins_mode = ins_mode_in_frameset return # 16. If node is an html element, run these substeps: - if node.name is 'html' + if node.name is 'html' and node.namespace is NS_HTML # 1. If the head element pointer is null, switch the insertion # mode to "before head" and abort these steps. (fragment case) if head_element_pointer is null @@ -833,7 +872,7 @@ parse_html = (args) -> 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 + if open_els[0].name is subject and open_els[0].namespace is NS_HTML el = open_els[0] open_els.shift() # remove it from the list of active formatting elements (if found) @@ -1117,14 +1156,14 @@ parse_html = (args) -> # http://www.w3.org/TR/html5/syntax.html#close-a-p-element close_p_element = -> generate_implied_end_tags 'p' # arg is exception - if open_els[0].name isnt 'p' + unless open_els[0].name is 'p' and open_els[0].namespace is NS_HTML parse_error() while open_els.length > 1 # just in case el = open_els.shift() - if el.name is 'p' + if el.name is 'p' and el.namespace is NS_HTML return close_p_if_in_button_scope = -> - if is_in_button_scope 'p' + if is_in_button_scope 'p', NS_HTML close_p_element() # http://www.w3.org/TR/html5/syntax.html#insert-a-character @@ -1185,7 +1224,7 @@ parse_html = (args) -> # If foster parenting is enabled and target is a table, tbody, tfoot, # thead, or tr element Foster parenting happens when content is # misnested in tables. - if flag_foster_parenting and foster_parenting_targets[target.name] + if flag_foster_parenting and foster_parenting_targets[target.name] is target.namespace loop # once. this is here so we can ``break`` to "abort these substeps" # 1. Let last template be the last template element in the # stack of open elements, if any. @@ -1318,7 +1357,7 @@ parse_html = (args) -> # 8.2.5.3 http://www.w3.org/TR/html5/syntax.html#closing-elements-that-have-implied-end-tags # http://www.w3.org/TR/html5/syntax.html#generate-implied-end-tags generate_implied_end_tags = (except = null) -> - while end_tag_implied[open_els[0].name] and open_els[0].name isnt except + while end_tag_implied[open_els[0].name] is open_els[0].namespace and open_els[0].name isnt except open_els.shift() # 8.2.5.4 The rules for parsing tokens in HTML content @@ -1480,7 +1519,7 @@ parse_html = (args) -> parse_error() loop el = open_els.shift() - if el.name is 'template' + if el.name is 'template' and el.namespace is NS_HTML break clear_afe_to_marker() template_ins_modes.shift() @@ -1579,7 +1618,7 @@ parse_html = (args) -> # 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 + if el.name is name and el.namespace is NS_HTML generate_implied_end_tags name # arg is exception parse_error() unless i is 0 while i >= 0 @@ -1667,7 +1706,7 @@ parse_html = (args) -> stop_parsing() return if t.type is TYPE_END_TAG and t.name is 'body' - unless is_in_scope 'body' + unless is_in_scope 'body', NS_HTML parse_error() return ok_tags = { @@ -1684,7 +1723,7 @@ parse_html = (args) -> ins_mode = ins_mode_after_body return if t.type is TYPE_END_TAG and t.name is 'html' - unless is_in_scope 'body' + unless is_in_scope 'body', NS_HTML parse_error() return ok_tags = { @@ -1707,7 +1746,7 @@ parse_html = (args) -> 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 + if h_tags[open_els[0].name] is open_els[0].namespace parse_error() open_els.shift() insert_html_element t @@ -2110,7 +2149,7 @@ parse_html = (args) -> return if t.type is TYPE_EOF parse_error() - if open_els[0].name is 'script' + if open_els[0].name is 'script' and open_els[0].namespace is NS_HTML open_els[0].flag 'already started', true open_els.shift() ins_mode = original_ins_mode @@ -2138,17 +2177,10 @@ parse_html = (args) -> 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] + if t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr' original_ins_mode = ins_mode ins_mode = ins_mode_in_table_text process_token t @@ -2185,10 +2217,10 @@ parse_html = (args) -> process_token t when 'table' parse_error() - if is_in_table_scope 'table' + if is_in_table_scope 'table', NS_HTML loop el = open_els.shift() - if el.name is 'table' + if el.name is 'table' and el.namespace is NS_HTML break reset_ins_mode() process_token t @@ -2215,14 +2247,14 @@ parse_html = (args) -> when TYPE_END_TAG switch t.name when 'table' - if is_in_table_scope 'table' + if is_in_table_scope 'table', NS_HTML loop el = open_els.shift() - if el.name is 'table' + if el.name is 'table' and el.namespace is NS_HTML break reset_ins_mode() else - parse_error + parse_error() when 'body', 'caption', 'col', 'colgroup', 'html', 'tbody', 'td', 'tfoot', 'th', 'thead', 'tr' parse_error() when 'template' @@ -2263,13 +2295,13 @@ parse_html = (args) -> # 8.2.5.4.11 http://www.w3.org/TR/html5/syntax.html#parsing-main-incaption ins_mode_in_caption = (t) -> if t.type is TYPE_END_TAG and t.name is 'caption' - if is_in_table_scope 'caption' + if is_in_table_scope 'caption', NS_HTML generate_implied_end_tags() if open_els[0].name isnt 'caption' parse_error() loop el = open_els.shift() - if el.name is 'caption' + if el.name is 'caption' and el.namespace is NS_HTML break clear_afe_to_marker() ins_mode = ins_mode_in_table @@ -2279,10 +2311,10 @@ parse_html = (args) -> return if (t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr')) or t.type is TYPE_END_TAG and t.name is 'table' parse_error() - if is_in_table_scope 'caption' + if is_in_table_scope 'caption', NS_HTML loop el = open_els.shift() - if el.name is 'caption' + if el.name is 'caption' and el.namespace is NS_HTML break clear_afe_to_marker() ins_mode = ins_mode_in_table @@ -2315,7 +2347,7 @@ parse_html = (args) -> t.acknowledge_self_closing() return if t.type is TYPE_END_TAG and t.name is 'colgroup' - if open_els[0].name is 'colgroup' + if open_els[0].name is 'colgroup' and open_els.namespace is NS_HTML open_els.shift() ins_mode = ins_mode_in_table else @@ -2354,7 +2386,7 @@ parse_html = (args) -> process_token t return if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead') - unless is_in_table_scope t.name # fixfull check namespace + unless is_in_table_scope t.name, NS_HTML parse_error() return clear_stack_to_table_body_context() @@ -2364,10 +2396,10 @@ parse_html = (args) -> if (t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead')) or (t.type is TYPE_END_TAG and t.name is 'table') has = false for el in open_els - if el.name is 'tbody' or el.name is 'tfoot' or el.name is 'thead' + if el.namespace is NS_HTML and (el.name is 'tbody' or el.name is 'tfoot' or el.name is 'thead') has = true break - if table_scopers[el.name] + if table_scopers[el.name] is el.namespace break if !has parse_error() @@ -2392,7 +2424,7 @@ parse_html = (args) -> afe_push_marker() return if t.type is TYPE_END_TAG and t.name is 'tr' - if is_in_table_scope 'tr' + if is_in_table_scope 'tr', NS_HTML clear_stack_to_table_row_context() open_els.shift() ins_mode = ins_mode_in_table_body @@ -2400,7 +2432,7 @@ parse_html = (args) -> parse_error() return if (t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr')) or t.type is TYPE_END_TAG and t.name is 'table' - if is_in_table_scope 'tr' + if is_in_table_scope 'tr', NS_HTML clear_stack_to_table_row_context() open_els.shift() ins_mode = ins_mode_in_table_body @@ -2409,8 +2441,8 @@ parse_html = (args) -> parse_error() return if t.type is TYPE_END_TAG and (t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead') - if is_in_table_scope t.name # fixfull namespace - if is_in_table_scope 'tr' + if is_in_table_scope t.name, NS_HTML + if is_in_table_scope 'tr', NS_HTML clear_stack_to_table_row_context() open_els.shift() ins_mode = ins_mode_in_table_body @@ -2427,11 +2459,11 @@ parse_html = (args) -> # http://www.w3.org/TR/html5/syntax.html#close-the-cell close_the_cell = -> generate_implied_end_tags() - unless open_els[0].name is 'td' or open_els[0] is 'th' + unless (open_els[0].name is 'td' or open_els[0] is 'th') and open_els[0].namespace is NS_HTML parse_error() loop el = open_els.shift() - if el.name is 'td' or el.name is 'th' + if el.namespace is NS_HTML and (el.name is 'td' or el.name is 'th') break clear_afe_to_marker() ins_mode = ins_mode_in_row @@ -2439,13 +2471,13 @@ parse_html = (args) -> # 8.2.5.4.15 http://www.w3.org/TR/html5/syntax.html#parsing-main-intd ins_mode_in_cell = (t) -> if t.type is TYPE_END_TAG and (t.name is 'td' or t.name is 'th') - if is_in_table_scope t.name + if is_in_table_scope t.name, NS_HTML generate_implied_end_tags() - if open_els[0].name isnt t.name - parse_error + unless (open_els[0].name is t.name) and open_els[0].namespace is NS_HTML + parse_error() loop el = open_els.shift() - if el.name is t.name + if el.name is t.name and el.namespace is NS_HTML break clear_afe_to_marker() ins_mode = ins_mode_in_row @@ -2455,10 +2487,10 @@ parse_html = (args) -> if t.type is TYPE_START_TAG and (t.name is 'caption' or t.name is 'col' or t.name is 'colgroup' or t.name is 'tbody' or t.name is 'td' or t.name is 'tfoot' or t.name is 'th' or t.name is 'thead' or t.name is 'tr') has = false for el in open_els - if el.name is 'td' or el.name is 'th' + if el.namespace is NS_HTML and (el.name is 'td' or el.name is 'th') has = true break - if table_scopers[el.name] + if table_scopers[el.name] is el.namespace break if !has parse_error() @@ -2470,7 +2502,7 @@ parse_html = (args) -> parse_error() return if t.type is TYPE_END_TAG and (t.name is 'table' or t.name is 'tbody' or t.name is 'tfoot' or t.name is 'thead' or t.name is 'tr') - if is_in_table_scope t.name # fixfull namespace + if is_in_table_scope t.name, NS_HTML close_the_cell() process_token t else @@ -2497,36 +2529,37 @@ parse_html = (args) -> ins_mode_in_body t return if t.type is TYPE_START_TAG and t.name is 'option' - if open_els[0].name is 'option' + if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML open_els.shift() insert_html_element t return if t.type is TYPE_START_TAG and t.name is 'optgroup' - if open_els[0].name is 'option' + if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML open_els.shift() - if open_els[0].name is 'optgroup' + if open_els[0].name is 'optgroup' and open_els[0].namespace is NS_HTML open_els.shift() insert_html_element t return if t.type is TYPE_END_TAG and t.name is 'optgroup' - if open_els[0].name is 'option' and open_els[1].name is 'optgroup' - open_els.shift() - if open_els[0].name is 'optgroup' + if open_els[0].name is 'option' and open_els[0].namespace in NS_HTML + if open_els[1].name is 'optgroup' and open_els[0].namespace is NS_HTML + open_els.shift() + if open_els[0].name is 'optgroup' and open_els[0].namespace is NS_HTML open_els.shift() else parse_error() return if t.type is TYPE_END_TAG and t.name is 'option' - if open_els[0].name is 'option' + if open_els[0].name is 'option' and open_els[0].namespace is NS_HTML open_els.shift() else parse_error() return if t.type is TYPE_END_TAG and t.name is 'select' - if is_in_select_scope 'select' + if is_in_select_scope 'select', NS_HTML loop el = open_els.shift() - if el.name is 'select' + if el.name is 'select' and el.namespace is NS_HTML break reset_ins_mode() else @@ -2536,7 +2569,7 @@ parse_html = (args) -> parse_error() loop el = open_els.shift() - if el.name is 'select' + if el.name is 'select' and el.namespace is NS_HTML break reset_ins_mode() # spec says that this is the same as but it doesn't say @@ -2544,11 +2577,11 @@ parse_html = (args) -> 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' + if is_in_select_scope 'select', NS_HTML return loop el = open_els.shift() - if el.name is 'select' + if el.name is 'select' and el.namespace is NS_HTML break reset_ins_mode() process_token t @@ -2569,7 +2602,7 @@ parse_html = (args) -> parse_error() loop el = open_els.shift() - if el.name is 'select' + if el.name is 'select' and el.namespace is NS_HTML break reset_ins_mode() process_token t @@ -2580,7 +2613,7 @@ parse_html = (args) -> return loop el = open_els.shift() - if el.name is 'select' + if el.name is 'select' and el.namespace is NS_HTML break reset_ins_mode() process_token t @@ -2637,7 +2670,7 @@ parse_html = (args) -> parse_error() loop el = open_els.shift() - if el.name is 'template' # fixfull check namespace + if el.name is 'template' and el.namespace is NS_HTML break clear_afe_to_marker() template_ins_modes.shift() @@ -2688,7 +2721,6 @@ parse_html = (args) -> 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 @@ -2705,7 +2737,6 @@ parse_html = (args) -> 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() @@ -3984,12 +4015,12 @@ parse_html = (args) -> return if c is '"' parse_error() - tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text? + tok_cur_tag.public_identifier = '' tok_state = tok_state_doctype_public_identifier_double_quoted return if c is "'" parse_error() - tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text? + tok_cur_tag.public_identifier = '' tok_state = tok_state_doctype_public_identifier_single_quoted return if c is '>' @@ -4016,12 +4047,12 @@ parse_html = (args) -> return if c is '"' parse_error() - tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text? + tok_cur_tag.public_identifier = '' tok_state = tok_state_doctype_public_identifier_double_quoted return if c is "'" parse_error() - tok_cur_tag.public_identifier = '' # FIXME should this go in @attrs or @text? + tok_cur_tag.public_identifier = '' tok_state = tok_state_doctype_public_identifier_single_quoted return if c is '>' @@ -4331,26 +4362,39 @@ parse_html = (args) -> if cur + 1 >= txt.length return '&' if txt.charAt(cur + 1).toLowerCase() is 'x' - prefix = '#x' + base = 16 charset = hex_chars start = cur + 2 else charset = digits start = cur + 1 - prefix = '#' + base = 10 i = 0 while start + i < txt.length and charset.indexOf(txt.charAt(start + i)) > -1 i += 1 if i is 0 return '&' + cur = start + i if txt.charAt(start + i) is ';' - i += 1 - # FIXME This is supposed to generate parse errors for some chars - decoded = decode_named_char_ref(prefix + txt.substr(start, i).toLowerCase()) - if decoded? - cur = start + i - return decoded - return '&' + cur += 1 + else + parse_error() + code_point = txt.substr(start, i) + while code_point.charAt(0) is '0' and code_point.length > 1 + code_point = code_point.substr 1 + code_point = parseInt(code_point, base) + if unicode_fixes[code_point]? + parse_error() + return unicode_fixes[code_point] + else + if (code_point >= 0xd800 and code_point <= 0xdfff) or code_point > 0x10ffff + parse_error() + return "\ufffd" + else + if (code_point >= 0x0001 and code_point <= 0x0008) or (code_point >= 0x000D and code_point <= 0x001F) or (code_point >= 0x007F and code_point <= 0x009F) or (code_point >= 0xFDD0 and code_point <= 0xFDEF) or code_point is 0x000B or code_point is 0xFFFE or code_point is 0xFFFF or code_point is 0x1FFFE or code_point is 0x1FFFF or code_point is 0x2FFFE or code_point is 0x2FFFF or code_point is 0x3FFFE or code_point is 0x3FFFF or code_point is 0x4FFFE or code_point is 0x4FFFF or code_point is 0x5FFFE or code_point is 0x5FFFF or code_point is 0x6FFFE or code_point is 0x6FFFF or code_point is 0x7FFFE or code_point is 0x7FFFF or code_point is 0x8FFFE or code_point is 0x8FFFF or code_point is 0x9FFFE or code_point is 0x9FFFF or code_point is 0xAFFFE or code_point is 0xAFFFF or code_point is 0xBFFFE or code_point is 0xBFFFF or code_point is 0xCFFFE or code_point is 0xCFFFF or code_point is 0xDFFFE or code_point is 0xDFFFF or code_point is 0xEFFFE or code_point is 0xEFFFF or code_point is 0xFFFFE or code_point is 0xFFFFF or code_point is 0x10FFFE or code_point is 0x10FFFF + parse_error() + return from_code_point code_point + return else for i in [0...31] if alnum.indexOf(txt.charAt(cur + i)) is -1 @@ -4414,8 +4458,8 @@ parse_html = (args) -> # tokenizer initialization tok_state = tok_state_data - if args.name is "one_that_breaks #1" - throw "hi" # console.log "hi" + if args.name is "namespace-sensitivity.dat #1" + console.log "hi" # proccess input # http://www.w3.org/TR/html5/syntax.html#tree-construction while flag_parsing @@ -4434,7 +4478,6 @@ serialize_els = (els, shallow, show_ids) -> 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