HTMLDocumentParser.cpp 117 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416141714181419142014211422142314241425142614271428142914301431143214331434143514361437143814391440144114421443144414451446144714481449145014511452145314541455145614571458145914601461146214631464146514661467146814691470147114721473147414751476147714781479148014811482148314841485148614871488148914901491149214931494149514961497149814991500150115021503150415051506150715081509151015111512151315141515151615171518151915201521152215231524152515261527152815291530153115321533153415351536153715381539154015411542154315441545154615471548154915501551155215531554155515561557155815591560156115621563156415651566156715681569157015711572157315741575157615771578157915801581158215831584158515861587158815891590159115921593159415951596159715981599160016011602160316041605160616071608160916101611161216131614161516161617161816191620162116221623162416251626162716281629163016311632163316341635163616371638163916401641164216431644164516461647164816491650165116521653165416551656165716581659166016611662166316641665166616671668166916701671167216731674167516761677167816791680168116821683168416851686168716881689169016911692169316941695169616971698169917001701170217031704170517061707170817091710171117121713171417151716171717181719172017211722172317241725172617271728172917301731173217331734173517361737173817391740174117421743174417451746174717481749175017511752175317541755175617571758175917601761176217631764176517661767176817691770177117721773177417751776177717781779178017811782178317841785178617871788178917901791179217931794179517961797179817991800180118021803180418051806180718081809181018111812181318141815181618171818181918201821182218231824182518261827182818291830183118321833183418351836183718381839184018411842184318441845184618471848184918501851185218531854185518561857185818591860186118621863186418651866186718681869187018711872187318741875187618771878187918801881188218831884188518861887188818891890189118921893189418951896189718981899190019011902190319041905190619071908190919101911191219131914191519161917191819191920192119221923192419251926192719281929193019311932193319341935193619371938193919401941194219431944194519461947194819491950195119521953195419551956195719581959196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019202020212022202320242025202620272028202920302031203220332034203520362037203820392040204120422043204420452046204720482049205020512052205320542055205620572058205920602061206220632064206520662067206820692070207120722073207420752076207720782079208020812082208320842085208620872088208920902091209220932094209520962097209820992100210121022103210421052106210721082109211021112112211321142115211621172118211921202121212221232124212521262127212821292130213121322133213421352136213721382139214021412142214321442145214621472148214921502151215221532154215521562157215821592160216121622163216421652166216721682169217021712172217321742175217621772178217921802181218221832184218521862187218821892190219121922193219421952196219721982199220022012202220322042205220622072208220922102211221222132214221522162217221822192220222122222223222422252226222722282229223022312232223322342235223622372238223922402241224222432244224522462247224822492250225122522253225422552256225722582259226022612262226322642265226622672268226922702271227222732274227522762277227822792280228122822283228422852286228722882289229022912292229322942295229622972298229923002301230223032304230523062307230823092310231123122313231423152316231723182319232023212322232323242325232623272328232923302331233223332334233523362337233823392340234123422343234423452346234723482349235023512352235323542355235623572358235923602361236223632364236523662367236823692370237123722373237423752376237723782379238023812382238323842385238623872388238923902391239223932394239523962397239823992400240124022403240424052406240724082409241024112412241324142415241624172418241924202421242224232424242524262427242824292430243124322433243424352436243724382439244024412442244324442445244624472448244924502451245224532454245524562457245824592460246124622463246424652466246724682469247024712472247324742475247624772478247924802481248224832484248524862487248824892490249124922493249424952496249724982499250025012502250325042505250625072508250925102511251225132514251525162517251825192520252125222523252425252526252725282529253025312532253325342535253625372538253925402541254225432544254525462547254825492550255125522553255425552556255725582559256025612562256325642565256625672568256925702571257225732574257525762577257825792580258125822583258425852586258725882589259025912592259325942595259625972598259926002601260226032604260526062607260826092610261126122613261426152616261726182619262026212622262326242625262626272628262926302631263226332634263526362637263826392640264126422643264426452646264726482649265026512652265326542655265626572658265926602661266226632664266526662667266826692670267126722673267426752676267726782679268026812682268326842685268626872688268926902691269226932694269526962697269826992700270127022703270427052706270727082709271027112712271327142715271627172718271927202721272227232724272527262727272827292730273127322733273427352736273727382739274027412742274327442745274627472748274927502751275227532754275527562757275827592760276127622763276427652766276727682769277027712772277327742775277627772778277927802781278227832784278527862787278827892790279127922793279427952796279727982799280028012802280328042805280628072808280928102811281228132814281528162817281828192820282128222823282428252826282728282829283028312832283328342835283628372838283928402841284228432844284528462847284828492850285128522853285428552856285728582859286028612862286328642865286628672868286928702871287228732874287528762877287828792880288128822883288428852886288728882889289028912892289328942895289628972898289929002901290229032904290529062907290829092910291129122913291429152916291729182919292029212922292329242925292629272928292929302931293229332934293529362937293829392940294129422943294429452946294729482949295029512952295329542955295629572958295929602961296229632964296529662967296829692970297129722973297429752976297729782979298029812982298329842985298629872988298929902991299229932994299529962997299829993000300130023003300430053006300730083009301030113012301330143015301630173018301930203021302230233024302530263027302830293030303130323033303430353036303730383039304030413042304330443045304630473048304930503051
  1. /*
  2. * Copyright (c) 2020, Andreas Kling <kling@serenityos.org>
  3. *
  4. * SPDX-License-Identifier: BSD-2-Clause
  5. */
  6. #include <AK/Debug.h>
  7. #include <AK/SourceLocation.h>
  8. #include <AK/Utf32View.h>
  9. #include <LibTextCodec/Decoder.h>
  10. #include <LibWeb/DOM/Comment.h>
  11. #include <LibWeb/DOM/Document.h>
  12. #include <LibWeb/DOM/DocumentType.h>
  13. #include <LibWeb/DOM/ElementFactory.h>
  14. #include <LibWeb/DOM/Event.h>
  15. #include <LibWeb/DOM/Text.h>
  16. #include <LibWeb/DOM/Window.h>
  17. #include <LibWeb/HTML/EventNames.h>
  18. #include <LibWeb/HTML/HTMLFormElement.h>
  19. #include <LibWeb/HTML/HTMLHeadElement.h>
  20. #include <LibWeb/HTML/HTMLScriptElement.h>
  21. #include <LibWeb/HTML/HTMLTableElement.h>
  22. #include <LibWeb/HTML/HTMLTemplateElement.h>
  23. #include <LibWeb/HTML/Parser/HTMLDocumentParser.h>
  24. #include <LibWeb/HTML/Parser/HTMLEncodingDetection.h>
  25. #include <LibWeb/HTML/Parser/HTMLToken.h>
  26. #include <LibWeb/Namespace.h>
  27. #include <LibWeb/SVG/TagNames.h>
  28. namespace Web::HTML {
  29. static inline void log_parse_error(const SourceLocation& location = SourceLocation::current())
  30. {
  31. dbgln("Parse error! {}", location);
  32. }
  33. static Vector<FlyString> s_quirks_public_ids = {
  34. "+//Silmaril//dtd html Pro v0r11 19970101//",
  35. "-//AS//DTD HTML 3.0 asWedit + extensions//",
  36. "-//AdvaSoft Ltd//DTD HTML 3.0 asWedit + extensions//",
  37. "-//IETF//DTD HTML 2.0 Level 1//",
  38. "-//IETF//DTD HTML 2.0 Level 2//",
  39. "-//IETF//DTD HTML 2.0 Strict Level 1//",
  40. "-//IETF//DTD HTML 2.0 Strict Level 2//",
  41. "-//IETF//DTD HTML 2.0 Strict//",
  42. "-//IETF//DTD HTML 2.0//",
  43. "-//IETF//DTD HTML 2.1E//",
  44. "-//IETF//DTD HTML 3.0//",
  45. "-//IETF//DTD HTML 3.2 Final//",
  46. "-//IETF//DTD HTML 3.2//",
  47. "-//IETF//DTD HTML 3//",
  48. "-//IETF//DTD HTML Level 0//",
  49. "-//IETF//DTD HTML Level 1//",
  50. "-//IETF//DTD HTML Level 2//",
  51. "-//IETF//DTD HTML Level 3//",
  52. "-//IETF//DTD HTML Strict Level 0//",
  53. "-//IETF//DTD HTML Strict Level 1//",
  54. "-//IETF//DTD HTML Strict Level 2//",
  55. "-//IETF//DTD HTML Strict Level 3//",
  56. "-//IETF//DTD HTML Strict//",
  57. "-//IETF//DTD HTML//",
  58. "-//Metrius//DTD Metrius Presentational//",
  59. "-//Microsoft//DTD Internet Explorer 2.0 HTML Strict//",
  60. "-//Microsoft//DTD Internet Explorer 2.0 HTML//",
  61. "-//Microsoft//DTD Internet Explorer 2.0 Tables//",
  62. "-//Microsoft//DTD Internet Explorer 3.0 HTML Strict//",
  63. "-//Microsoft//DTD Internet Explorer 3.0 HTML//",
  64. "-//Microsoft//DTD Internet Explorer 3.0 Tables//",
  65. "-//Netscape Comm. Corp.//DTD HTML//",
  66. "-//Netscape Comm. Corp.//DTD Strict HTML//",
  67. "-//O'Reilly and Associates//DTD HTML 2.0//",
  68. "-//O'Reilly and Associates//DTD HTML Extended 1.0//",
  69. "-//O'Reilly and Associates//DTD HTML Extended Relaxed 1.0//",
  70. "-//SQ//DTD HTML 2.0 HoTMetaL + extensions//",
  71. "-//SoftQuad Software//DTD HoTMetaL PRO 6.0::19990601::extensions to HTML 4.0//",
  72. "-//SoftQuad//DTD HoTMetaL PRO 4.0::19971010::extensions to HTML 4.0//",
  73. "-//Spyglass//DTD HTML 2.0 Extended//",
  74. "-//Sun Microsystems Corp.//DTD HotJava HTML//",
  75. "-//Sun Microsystems Corp.//DTD HotJava Strict HTML//",
  76. "-//W3C//DTD HTML 3 1995-03-24//",
  77. "-//W3C//DTD HTML 3.2 Draft//",
  78. "-//W3C//DTD HTML 3.2 Final//",
  79. "-//W3C//DTD HTML 3.2//",
  80. "-//W3C//DTD HTML 3.2S Draft//",
  81. "-//W3C//DTD HTML 4.0 Frameset//",
  82. "-//W3C//DTD HTML 4.0 Transitional//",
  83. "-//W3C//DTD HTML Experimental 19960712//",
  84. "-//W3C//DTD HTML Experimental 970421//",
  85. "-//W3C//DTD W3 HTML//",
  86. "-//W3O//DTD W3 HTML 3.0//",
  87. "-//WebTechs//DTD Mozilla HTML 2.0//",
  88. "-//WebTechs//DTD Mozilla HTML//"
  89. };
  90. RefPtr<DOM::Document> parse_html_document(const StringView& data, const URL& url, const String& encoding)
  91. {
  92. auto document = DOM::Document::create(url);
  93. HTMLDocumentParser parser(document, data, encoding);
  94. parser.run(url);
  95. return document;
  96. }
  97. HTMLDocumentParser::HTMLDocumentParser(DOM::Document& document, const StringView& input, const String& encoding)
  98. : m_tokenizer(input, encoding)
  99. , m_document(document)
  100. {
  101. m_document->set_should_invalidate_styles_on_attribute_changes(false);
  102. auto standardized_encoding = TextCodec::get_standardized_encoding(encoding);
  103. VERIFY(standardized_encoding.has_value());
  104. m_document->set_encoding(standardized_encoding.value());
  105. }
  106. HTMLDocumentParser::~HTMLDocumentParser()
  107. {
  108. m_document->set_should_invalidate_styles_on_attribute_changes(true);
  109. }
  110. void HTMLDocumentParser::run(const URL& url)
  111. {
  112. m_document->set_url(url);
  113. m_document->set_source(m_tokenizer.source());
  114. for (;;) {
  115. auto optional_token = m_tokenizer.next_token();
  116. if (!optional_token.has_value())
  117. break;
  118. auto& token = optional_token.value();
  119. dbgln_if(PARSER_DEBUG, "[{}] {}", insertion_mode_name(), token.to_string());
  120. // FIXME: If the adjusted current node is a MathML text integration point and the token is a start tag whose tag name is neither "mglyph" nor "malignmark"
  121. // FIXME: If the adjusted current node is a MathML text integration point and the token is a character token
  122. // FIXME: If the adjusted current node is a MathML annotation-xml element and the token is a start tag whose tag name is "svg"
  123. // FIXME: If the adjusted current node is an HTML integration point and the token is a start tag
  124. // FIXME: If the adjusted current node is an HTML integration point and the token is a character token
  125. if (m_stack_of_open_elements.is_empty()
  126. || adjusted_current_node().namespace_() == Namespace::HTML
  127. || token.is_end_of_file()) {
  128. process_using_the_rules_for(m_insertion_mode, token);
  129. } else {
  130. process_using_the_rules_for_foreign_content(token);
  131. }
  132. if (m_stop_parsing) {
  133. dbgln_if(PARSER_DEBUG, "Stop parsing{}! :^)", m_parsing_fragment ? " fragment" : "");
  134. break;
  135. }
  136. }
  137. flush_character_insertions();
  138. // "The end"
  139. m_document->set_ready_state("interactive");
  140. auto scripts_to_execute_when_parsing_has_finished = m_document->take_scripts_to_execute_when_parsing_has_finished({});
  141. for (auto& script : scripts_to_execute_when_parsing_has_finished) {
  142. // FIXME: Spin the event loop until the script is ready to be parser executed and there's no style sheets blocking scripts.
  143. script.execute_script();
  144. }
  145. auto content_loaded_event = DOM::Event::create(HTML::EventNames::DOMContentLoaded);
  146. content_loaded_event->set_bubbles(true);
  147. m_document->dispatch_event(content_loaded_event);
  148. // FIXME: The document parser shouldn't execute these, it should just spin the event loop until the list becomes empty.
  149. // FIXME: Once the set has been added, also spin the event loop until the set becomes empty.
  150. auto scripts_to_execute_as_soon_as_possible = m_document->take_scripts_to_execute_as_soon_as_possible({});
  151. for (auto& script : scripts_to_execute_as_soon_as_possible) {
  152. script.execute_script();
  153. }
  154. // FIXME: Spin the event loop until there is nothing that delays the load event in the Document.
  155. m_document->set_ready_state("complete");
  156. m_document->window().dispatch_event(DOM::Event::create(HTML::EventNames::load));
  157. m_document->set_ready_for_post_load_tasks(true);
  158. m_document->completely_finish_loading();
  159. }
  160. void HTMLDocumentParser::process_using_the_rules_for(InsertionMode mode, HTMLToken& token)
  161. {
  162. switch (mode) {
  163. case InsertionMode::Initial:
  164. handle_initial(token);
  165. break;
  166. case InsertionMode::BeforeHTML:
  167. handle_before_html(token);
  168. break;
  169. case InsertionMode::BeforeHead:
  170. handle_before_head(token);
  171. break;
  172. case InsertionMode::InHead:
  173. handle_in_head(token);
  174. break;
  175. case InsertionMode::InHeadNoscript:
  176. handle_in_head_noscript(token);
  177. break;
  178. case InsertionMode::AfterHead:
  179. handle_after_head(token);
  180. break;
  181. case InsertionMode::InBody:
  182. handle_in_body(token);
  183. break;
  184. case InsertionMode::AfterBody:
  185. handle_after_body(token);
  186. break;
  187. case InsertionMode::AfterAfterBody:
  188. handle_after_after_body(token);
  189. break;
  190. case InsertionMode::Text:
  191. handle_text(token);
  192. break;
  193. case InsertionMode::InTable:
  194. handle_in_table(token);
  195. break;
  196. case InsertionMode::InTableBody:
  197. handle_in_table_body(token);
  198. break;
  199. case InsertionMode::InRow:
  200. handle_in_row(token);
  201. break;
  202. case InsertionMode::InCell:
  203. handle_in_cell(token);
  204. break;
  205. case InsertionMode::InTableText:
  206. handle_in_table_text(token);
  207. break;
  208. case InsertionMode::InSelectInTable:
  209. handle_in_select_in_table(token);
  210. break;
  211. case InsertionMode::InSelect:
  212. handle_in_select(token);
  213. break;
  214. case InsertionMode::InCaption:
  215. handle_in_caption(token);
  216. break;
  217. case InsertionMode::InColumnGroup:
  218. handle_in_column_group(token);
  219. break;
  220. case InsertionMode::InTemplate:
  221. handle_in_template(token);
  222. break;
  223. case InsertionMode::InFrameset:
  224. handle_in_frameset(token);
  225. break;
  226. case InsertionMode::AfterFrameset:
  227. handle_after_frameset(token);
  228. break;
  229. case InsertionMode::AfterAfterFrameset:
  230. handle_after_after_frameset(token);
  231. break;
  232. default:
  233. VERIFY_NOT_REACHED();
  234. }
  235. }
  236. DOM::QuirksMode HTMLDocumentParser::which_quirks_mode(const HTMLToken& doctype_token) const
  237. {
  238. if (doctype_token.doctype_data().force_quirks)
  239. return DOM::QuirksMode::Yes;
  240. // NOTE: The tokenizer puts the name into lower case for us.
  241. if (doctype_token.doctype_data().name != "html")
  242. return DOM::QuirksMode::Yes;
  243. auto const& public_identifier = doctype_token.doctype_data().public_identifier;
  244. auto const& system_identifier = doctype_token.doctype_data().system_identifier;
  245. if (public_identifier.equals_ignoring_case("-//W3O//DTD W3 HTML Strict 3.0//EN//"))
  246. return DOM::QuirksMode::Yes;
  247. if (public_identifier.equals_ignoring_case("-/W3C/DTD HTML 4.0 Transitional/EN"))
  248. return DOM::QuirksMode::Yes;
  249. if (public_identifier.equals_ignoring_case("HTML"))
  250. return DOM::QuirksMode::Yes;
  251. if (system_identifier.equals_ignoring_case("http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd"))
  252. return DOM::QuirksMode::Yes;
  253. for (auto& public_id : s_quirks_public_ids) {
  254. if (public_identifier.starts_with(public_id, CaseSensitivity::CaseInsensitive))
  255. return DOM::QuirksMode::Yes;
  256. }
  257. if (doctype_token.doctype_data().missing_system_identifier) {
  258. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  259. return DOM::QuirksMode::Yes;
  260. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  261. return DOM::QuirksMode::Yes;
  262. }
  263. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Frameset//", CaseSensitivity::CaseInsensitive))
  264. return DOM::QuirksMode::Limited;
  265. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Transitional//", CaseSensitivity::CaseInsensitive))
  266. return DOM::QuirksMode::Limited;
  267. if (!doctype_token.doctype_data().missing_system_identifier) {
  268. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  269. return DOM::QuirksMode::Limited;
  270. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  271. return DOM::QuirksMode::Limited;
  272. }
  273. return DOM::QuirksMode::No;
  274. }
  275. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  276. {
  277. if (token.is_character() && token.is_parser_whitespace()) {
  278. return;
  279. }
  280. if (token.is_comment()) {
  281. auto comment = adopt_ref(*new DOM::Comment(document(), token.comment()));
  282. document().append_child(move(comment));
  283. return;
  284. }
  285. if (token.is_doctype()) {
  286. auto doctype = adopt_ref(*new DOM::DocumentType(document()));
  287. doctype->set_name(token.doctype_data().name);
  288. doctype->set_public_id(token.doctype_data().public_identifier);
  289. doctype->set_system_id(token.doctype_data().system_identifier);
  290. document().append_child(move(doctype));
  291. document().set_quirks_mode(which_quirks_mode(token));
  292. m_insertion_mode = InsertionMode::BeforeHTML;
  293. return;
  294. }
  295. log_parse_error();
  296. document().set_quirks_mode(DOM::QuirksMode::Yes);
  297. m_insertion_mode = InsertionMode::BeforeHTML;
  298. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  299. }
  300. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  301. {
  302. if (token.is_doctype()) {
  303. log_parse_error();
  304. return;
  305. }
  306. if (token.is_comment()) {
  307. auto comment = adopt_ref(*new DOM::Comment(document(), token.comment()));
  308. document().append_child(move(comment));
  309. return;
  310. }
  311. if (token.is_character() && token.is_parser_whitespace()) {
  312. return;
  313. }
  314. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  315. auto element = create_element_for(token, Namespace::HTML);
  316. document().append_child(element);
  317. m_stack_of_open_elements.push(move(element));
  318. m_insertion_mode = InsertionMode::BeforeHead;
  319. return;
  320. }
  321. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  322. goto AnythingElse;
  323. }
  324. if (token.is_end_tag()) {
  325. log_parse_error();
  326. return;
  327. }
  328. AnythingElse:
  329. auto element = create_element(document(), HTML::TagNames::html, Namespace::HTML);
  330. document().append_child(element);
  331. m_stack_of_open_elements.push(element);
  332. // FIXME: If the Document is being loaded as part of navigation of a browsing context, then: run the application cache selection algorithm with no manifest, passing it the Document object.
  333. m_insertion_mode = InsertionMode::BeforeHead;
  334. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  335. return;
  336. }
  337. DOM::Element& HTMLDocumentParser::current_node()
  338. {
  339. return m_stack_of_open_elements.current_node();
  340. }
  341. DOM::Element& HTMLDocumentParser::adjusted_current_node()
  342. {
  343. if (m_parsing_fragment && m_stack_of_open_elements.elements().size() == 1)
  344. return *m_context_element;
  345. return current_node();
  346. }
  347. DOM::Element& HTMLDocumentParser::node_before_current_node()
  348. {
  349. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  350. }
  351. HTMLDocumentParser::AdjustedInsertionLocation HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  352. {
  353. auto& target = current_node();
  354. HTMLDocumentParser::AdjustedInsertionLocation adjusted_insertion_location;
  355. if (m_foster_parenting && target.local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  356. auto last_template = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::template_);
  357. auto last_table = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::table);
  358. if (last_template.element && (!last_table.element || last_template.index > last_table.index)) {
  359. // This returns the template content, so no need to check the parent is a template.
  360. return { verify_cast<HTMLTemplateElement>(last_template.element)->content(), nullptr };
  361. }
  362. if (!last_table.element) {
  363. VERIFY(m_parsing_fragment);
  364. // Guaranteed not to be a template element (it will be the html element),
  365. // so no need to check the parent is a template.
  366. return { m_stack_of_open_elements.elements().first(), nullptr };
  367. }
  368. if (last_table.element->parent_node())
  369. adjusted_insertion_location = { last_table.element->parent_node(), last_table.element };
  370. else
  371. adjusted_insertion_location = { m_stack_of_open_elements.element_before(*last_table.element), nullptr };
  372. } else {
  373. adjusted_insertion_location = { target, nullptr };
  374. }
  375. if (is<HTMLTemplateElement>(*adjusted_insertion_location.parent))
  376. return { verify_cast<HTMLTemplateElement>(*adjusted_insertion_location.parent).content(), nullptr };
  377. return adjusted_insertion_location;
  378. }
  379. NonnullRefPtr<DOM::Element> HTMLDocumentParser::create_element_for(const HTMLToken& token, const FlyString& namespace_)
  380. {
  381. auto element = create_element(document(), token.tag_name(), namespace_);
  382. token.for_each_attribute([&](auto& attribute) {
  383. element->set_attribute(attribute.local_name, attribute.value);
  384. return IterationDecision::Continue;
  385. });
  386. return element;
  387. }
  388. // https://html.spec.whatwg.org/multipage/parsing.html#insert-a-foreign-element
  389. NonnullRefPtr<DOM::Element> HTMLDocumentParser::insert_foreign_element(const HTMLToken& token, const FlyString& namespace_)
  390. {
  391. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  392. // FIXME: Pass in adjusted_insertion_location.parent as the intended parent.
  393. auto element = create_element_for(token, namespace_);
  394. auto pre_insertion_validity = adjusted_insertion_location.parent->ensure_pre_insertion_validity(element, adjusted_insertion_location.insert_before_sibling);
  395. // NOTE: If it's not possible to insert the element at the adjusted insertion location, the element is simply dropped.
  396. if (!pre_insertion_validity.is_exception()) {
  397. if (!m_parsing_fragment) {
  398. // FIXME: push a new element queue onto element's relevant agent's custom element reactions stack.
  399. }
  400. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling);
  401. if (!m_parsing_fragment) {
  402. // FIXME: pop the element queue from element's relevant agent's custom element reactions stack, and invoke custom element reactions in that queue.
  403. }
  404. }
  405. m_stack_of_open_elements.push(element);
  406. return element;
  407. }
  408. NonnullRefPtr<DOM::Element> HTMLDocumentParser::insert_html_element(const HTMLToken& token)
  409. {
  410. return insert_foreign_element(token, Namespace::HTML);
  411. }
  412. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  413. {
  414. if (token.is_character() && token.is_parser_whitespace()) {
  415. return;
  416. }
  417. if (token.is_comment()) {
  418. insert_comment(token);
  419. return;
  420. }
  421. if (token.is_doctype()) {
  422. log_parse_error();
  423. return;
  424. }
  425. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  426. process_using_the_rules_for(InsertionMode::InBody, token);
  427. return;
  428. }
  429. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  430. auto element = insert_html_element(token);
  431. m_head_element = verify_cast<HTMLHeadElement>(*element);
  432. m_insertion_mode = InsertionMode::InHead;
  433. return;
  434. }
  435. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  436. goto AnythingElse;
  437. }
  438. if (token.is_end_tag()) {
  439. log_parse_error();
  440. return;
  441. }
  442. AnythingElse:
  443. m_head_element = verify_cast<HTMLHeadElement>(*insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::head)));
  444. m_insertion_mode = InsertionMode::InHead;
  445. process_using_the_rules_for(InsertionMode::InHead, token);
  446. return;
  447. }
  448. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  449. {
  450. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  451. adjusted_insertion_location.parent->insert_before(adopt_ref(*new DOM::Comment(document(), token.comment())), adjusted_insertion_location.insert_before_sibling);
  452. }
  453. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  454. {
  455. if (token.is_parser_whitespace()) {
  456. insert_character(token.code_point());
  457. return;
  458. }
  459. if (token.is_comment()) {
  460. insert_comment(token);
  461. return;
  462. }
  463. if (token.is_doctype()) {
  464. log_parse_error();
  465. return;
  466. }
  467. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  468. process_using_the_rules_for(InsertionMode::InBody, token);
  469. return;
  470. }
  471. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  472. insert_html_element(token);
  473. m_stack_of_open_elements.pop();
  474. token.acknowledge_self_closing_flag_if_set();
  475. return;
  476. }
  477. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  478. auto element = insert_html_element(token);
  479. m_stack_of_open_elements.pop();
  480. token.acknowledge_self_closing_flag_if_set();
  481. return;
  482. }
  483. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  484. insert_html_element(token);
  485. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  486. m_original_insertion_mode = m_insertion_mode;
  487. m_insertion_mode = InsertionMode::Text;
  488. return;
  489. }
  490. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled) || token.tag_name() == HTML::TagNames::noframes || token.tag_name() == HTML::TagNames::style)) {
  491. parse_generic_raw_text_element(token);
  492. return;
  493. }
  494. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noscript && !m_scripting_enabled) {
  495. insert_html_element(token);
  496. m_insertion_mode = InsertionMode::InHeadNoscript;
  497. return;
  498. }
  499. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  500. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  501. auto element = create_element_for(token, Namespace::HTML);
  502. auto& script_element = verify_cast<HTMLScriptElement>(*element);
  503. script_element.set_parser_document({}, document());
  504. script_element.set_non_blocking({}, false);
  505. if (m_parsing_fragment) {
  506. script_element.set_already_started({}, true);
  507. }
  508. if (m_invoked_via_document_write) {
  509. TODO();
  510. }
  511. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling, false);
  512. m_stack_of_open_elements.push(element);
  513. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  514. m_original_insertion_mode = m_insertion_mode;
  515. m_insertion_mode = InsertionMode::Text;
  516. return;
  517. }
  518. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  519. m_stack_of_open_elements.pop();
  520. m_insertion_mode = InsertionMode::AfterHead;
  521. return;
  522. }
  523. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  524. goto AnythingElse;
  525. }
  526. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  527. insert_html_element(token);
  528. m_list_of_active_formatting_elements.add_marker();
  529. m_frameset_ok = false;
  530. m_insertion_mode = InsertionMode::InTemplate;
  531. m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  532. return;
  533. }
  534. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  535. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  536. log_parse_error();
  537. return;
  538. }
  539. generate_all_implied_end_tags_thoroughly();
  540. if (current_node().local_name() != HTML::TagNames::template_)
  541. log_parse_error();
  542. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  543. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  544. m_stack_of_template_insertion_modes.take_last();
  545. reset_the_insertion_mode_appropriately();
  546. return;
  547. }
  548. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  549. log_parse_error();
  550. return;
  551. }
  552. AnythingElse:
  553. m_stack_of_open_elements.pop();
  554. m_insertion_mode = InsertionMode::AfterHead;
  555. process_using_the_rules_for(m_insertion_mode, token);
  556. }
  557. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken& token)
  558. {
  559. if (token.is_doctype()) {
  560. log_parse_error();
  561. return;
  562. }
  563. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  564. process_using_the_rules_for(InsertionMode::InBody, token);
  565. return;
  566. }
  567. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::noscript) {
  568. m_stack_of_open_elements.pop();
  569. m_insertion_mode = InsertionMode::InHead;
  570. return;
  571. }
  572. if (token.is_parser_whitespace() || token.is_comment() || (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::style))) {
  573. process_using_the_rules_for(InsertionMode::InHead, token);
  574. return;
  575. }
  576. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  577. goto AnythingElse;
  578. }
  579. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::noscript)) {
  580. log_parse_error();
  581. return;
  582. }
  583. AnythingElse:
  584. log_parse_error();
  585. m_stack_of_open_elements.pop();
  586. m_insertion_mode = InsertionMode::InHead;
  587. process_using_the_rules_for(m_insertion_mode, token);
  588. }
  589. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  590. {
  591. insert_html_element(token);
  592. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  593. m_original_insertion_mode = m_insertion_mode;
  594. m_insertion_mode = InsertionMode::Text;
  595. }
  596. DOM::Text* HTMLDocumentParser::find_character_insertion_node()
  597. {
  598. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  599. if (adjusted_insertion_location.insert_before_sibling) {
  600. TODO();
  601. }
  602. if (adjusted_insertion_location.parent->is_document())
  603. return nullptr;
  604. if (adjusted_insertion_location.parent->last_child() && adjusted_insertion_location.parent->last_child()->is_text())
  605. return verify_cast<DOM::Text>(adjusted_insertion_location.parent->last_child());
  606. auto new_text_node = adopt_ref(*new DOM::Text(document(), ""));
  607. adjusted_insertion_location.parent->append_child(new_text_node);
  608. return new_text_node;
  609. }
  610. void HTMLDocumentParser::flush_character_insertions()
  611. {
  612. if (m_character_insertion_builder.is_empty())
  613. return;
  614. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  615. m_character_insertion_node->parent()->children_changed();
  616. m_character_insertion_builder.clear();
  617. }
  618. void HTMLDocumentParser::insert_character(u32 data)
  619. {
  620. auto node = find_character_insertion_node();
  621. if (node == m_character_insertion_node) {
  622. m_character_insertion_builder.append(Utf32View { &data, 1 });
  623. return;
  624. }
  625. if (!m_character_insertion_node) {
  626. m_character_insertion_node = node;
  627. m_character_insertion_builder.append(Utf32View { &data, 1 });
  628. return;
  629. }
  630. flush_character_insertions();
  631. m_character_insertion_node = node;
  632. m_character_insertion_builder.append(Utf32View { &data, 1 });
  633. }
  634. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  635. {
  636. if (token.is_character() && token.is_parser_whitespace()) {
  637. insert_character(token.code_point());
  638. return;
  639. }
  640. if (token.is_comment()) {
  641. insert_comment(token);
  642. return;
  643. }
  644. if (token.is_doctype()) {
  645. log_parse_error();
  646. return;
  647. }
  648. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  649. process_using_the_rules_for(InsertionMode::InBody, token);
  650. return;
  651. }
  652. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  653. insert_html_element(token);
  654. m_frameset_ok = false;
  655. m_insertion_mode = InsertionMode::InBody;
  656. return;
  657. }
  658. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  659. insert_html_element(token);
  660. m_insertion_mode = InsertionMode::InFrameset;
  661. return;
  662. }
  663. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  664. log_parse_error();
  665. m_stack_of_open_elements.push(*m_head_element);
  666. process_using_the_rules_for(InsertionMode::InHead, token);
  667. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  668. return entry.ptr() == m_head_element;
  669. });
  670. return;
  671. }
  672. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  673. process_using_the_rules_for(InsertionMode::InHead, token);
  674. return;
  675. }
  676. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  677. goto AnythingElse;
  678. }
  679. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  680. log_parse_error();
  681. return;
  682. }
  683. AnythingElse:
  684. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::body));
  685. m_insertion_mode = InsertionMode::InBody;
  686. process_using_the_rules_for(m_insertion_mode, token);
  687. }
  688. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  689. {
  690. while (current_node().local_name() != exception && current_node().local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc))
  691. m_stack_of_open_elements.pop();
  692. }
  693. void HTMLDocumentParser::generate_all_implied_end_tags_thoroughly()
  694. {
  695. while (current_node().local_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::colgroup, HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))
  696. m_stack_of_open_elements.pop();
  697. }
  698. void HTMLDocumentParser::close_a_p_element()
  699. {
  700. generate_implied_end_tags(HTML::TagNames::p);
  701. if (current_node().local_name() != HTML::TagNames::p) {
  702. log_parse_error();
  703. }
  704. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  705. }
  706. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  707. {
  708. if (token.is_character() && token.is_parser_whitespace()) {
  709. process_using_the_rules_for(InsertionMode::InBody, token);
  710. return;
  711. }
  712. if (token.is_comment()) {
  713. auto& insertion_location = m_stack_of_open_elements.first();
  714. insertion_location.append_child(adopt_ref(*new DOM::Comment(document(), token.comment())));
  715. return;
  716. }
  717. if (token.is_doctype()) {
  718. log_parse_error();
  719. return;
  720. }
  721. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  722. process_using_the_rules_for(InsertionMode::InBody, token);
  723. return;
  724. }
  725. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  726. if (m_parsing_fragment) {
  727. log_parse_error();
  728. return;
  729. }
  730. m_insertion_mode = InsertionMode::AfterAfterBody;
  731. return;
  732. }
  733. if (token.is_end_of_file()) {
  734. stop_parsing();
  735. return;
  736. }
  737. log_parse_error();
  738. m_insertion_mode = InsertionMode::InBody;
  739. process_using_the_rules_for(InsertionMode::InBody, token);
  740. }
  741. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  742. {
  743. if (token.is_comment()) {
  744. auto comment = adopt_ref(*new DOM::Comment(document(), token.comment()));
  745. document().append_child(move(comment));
  746. return;
  747. }
  748. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  749. process_using_the_rules_for(InsertionMode::InBody, token);
  750. return;
  751. }
  752. if (token.is_end_of_file()) {
  753. stop_parsing();
  754. return;
  755. }
  756. log_parse_error();
  757. m_insertion_mode = InsertionMode::InBody;
  758. process_using_the_rules_for(m_insertion_mode, token);
  759. }
  760. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  761. {
  762. // FIXME: This needs to care about "markers"
  763. if (m_list_of_active_formatting_elements.is_empty())
  764. return;
  765. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  766. return;
  767. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  768. return;
  769. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  770. RefPtr<DOM::Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  771. VERIFY(entry);
  772. Rewind:
  773. if (index == 0) {
  774. goto Create;
  775. }
  776. --index;
  777. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  778. VERIFY(entry);
  779. if (!m_stack_of_open_elements.contains(*entry))
  780. goto Rewind;
  781. Advance:
  782. ++index;
  783. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  784. VERIFY(entry);
  785. Create:
  786. // FIXME: Hold on to the real token!
  787. auto new_element = insert_html_element(HTMLToken::make_start_tag(entry->local_name()));
  788. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  789. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  790. goto Advance;
  791. }
  792. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  793. {
  794. auto subject = token.tag_name();
  795. // If the current node is an HTML element whose tag name is subject,
  796. // and the current node is not in the list of active formatting elements,
  797. // then pop the current node off the stack of open elements, and return.
  798. if (current_node().local_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  799. m_stack_of_open_elements.pop();
  800. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  801. }
  802. size_t outer_loop_counter = 0;
  803. //OuterLoop:
  804. if (outer_loop_counter >= 8)
  805. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  806. ++outer_loop_counter;
  807. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  808. if (!formatting_element)
  809. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  810. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  811. log_parse_error();
  812. m_list_of_active_formatting_elements.remove(*formatting_element);
  813. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  814. }
  815. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  816. log_parse_error();
  817. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  818. }
  819. if (formatting_element != &current_node()) {
  820. log_parse_error();
  821. }
  822. RefPtr<DOM::Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  823. if (!furthest_block) {
  824. while (&current_node() != formatting_element)
  825. m_stack_of_open_elements.pop();
  826. m_stack_of_open_elements.pop();
  827. m_list_of_active_formatting_elements.remove(*formatting_element);
  828. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  829. }
  830. // FIXME: Implement the rest of the AAA :^)
  831. TODO();
  832. }
  833. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name, const FlyString& namespace_)
  834. {
  835. if (namespace_ == Namespace::HTML) {
  836. return tag_name.is_one_of(
  837. HTML::TagNames::address,
  838. HTML::TagNames::applet,
  839. HTML::TagNames::area,
  840. HTML::TagNames::article,
  841. HTML::TagNames::aside,
  842. HTML::TagNames::base,
  843. HTML::TagNames::basefont,
  844. HTML::TagNames::bgsound,
  845. HTML::TagNames::blockquote,
  846. HTML::TagNames::body,
  847. HTML::TagNames::br,
  848. HTML::TagNames::button,
  849. HTML::TagNames::caption,
  850. HTML::TagNames::center,
  851. HTML::TagNames::col,
  852. HTML::TagNames::colgroup,
  853. HTML::TagNames::dd,
  854. HTML::TagNames::details,
  855. HTML::TagNames::dir,
  856. HTML::TagNames::div,
  857. HTML::TagNames::dl,
  858. HTML::TagNames::dt,
  859. HTML::TagNames::embed,
  860. HTML::TagNames::fieldset,
  861. HTML::TagNames::figcaption,
  862. HTML::TagNames::figure,
  863. HTML::TagNames::footer,
  864. HTML::TagNames::form,
  865. HTML::TagNames::frame,
  866. HTML::TagNames::frameset,
  867. HTML::TagNames::h1,
  868. HTML::TagNames::h2,
  869. HTML::TagNames::h3,
  870. HTML::TagNames::h4,
  871. HTML::TagNames::h5,
  872. HTML::TagNames::h6,
  873. HTML::TagNames::head,
  874. HTML::TagNames::header,
  875. HTML::TagNames::hgroup,
  876. HTML::TagNames::hr,
  877. HTML::TagNames::html,
  878. HTML::TagNames::iframe,
  879. HTML::TagNames::img,
  880. HTML::TagNames::input,
  881. HTML::TagNames::keygen,
  882. HTML::TagNames::li,
  883. HTML::TagNames::link,
  884. HTML::TagNames::listing,
  885. HTML::TagNames::main,
  886. HTML::TagNames::marquee,
  887. HTML::TagNames::menu,
  888. HTML::TagNames::meta,
  889. HTML::TagNames::nav,
  890. HTML::TagNames::noembed,
  891. HTML::TagNames::noframes,
  892. HTML::TagNames::noscript,
  893. HTML::TagNames::object,
  894. HTML::TagNames::ol,
  895. HTML::TagNames::p,
  896. HTML::TagNames::param,
  897. HTML::TagNames::plaintext,
  898. HTML::TagNames::pre,
  899. HTML::TagNames::script,
  900. HTML::TagNames::section,
  901. HTML::TagNames::select,
  902. HTML::TagNames::source,
  903. HTML::TagNames::style,
  904. HTML::TagNames::summary,
  905. HTML::TagNames::table,
  906. HTML::TagNames::tbody,
  907. HTML::TagNames::td,
  908. HTML::TagNames::template_,
  909. HTML::TagNames::textarea,
  910. HTML::TagNames::tfoot,
  911. HTML::TagNames::th,
  912. HTML::TagNames::thead,
  913. HTML::TagNames::title,
  914. HTML::TagNames::tr,
  915. HTML::TagNames::track,
  916. HTML::TagNames::ul,
  917. HTML::TagNames::wbr,
  918. HTML::TagNames::xmp);
  919. } else if (namespace_ == Namespace::SVG) {
  920. return tag_name.is_one_of(
  921. SVG::TagNames::desc,
  922. SVG::TagNames::foreignObject,
  923. SVG::TagNames::title);
  924. } else if (namespace_ == Namespace::MathML) {
  925. TODO();
  926. }
  927. return false;
  928. }
  929. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  930. {
  931. if (token.is_character()) {
  932. if (token.code_point() == 0) {
  933. log_parse_error();
  934. return;
  935. }
  936. if (token.is_parser_whitespace()) {
  937. reconstruct_the_active_formatting_elements();
  938. insert_character(token.code_point());
  939. return;
  940. }
  941. reconstruct_the_active_formatting_elements();
  942. insert_character(token.code_point());
  943. m_frameset_ok = false;
  944. return;
  945. }
  946. if (token.is_comment()) {
  947. insert_comment(token);
  948. return;
  949. }
  950. if (token.is_doctype()) {
  951. log_parse_error();
  952. return;
  953. }
  954. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  955. log_parse_error();
  956. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  957. return;
  958. token.for_each_attribute([&](auto& attribute) {
  959. if (!current_node().has_attribute(attribute.local_name))
  960. current_node().set_attribute(attribute.local_name, attribute.value);
  961. return IterationDecision::Continue;
  962. });
  963. return;
  964. }
  965. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  966. process_using_the_rules_for(InsertionMode::InHead, token);
  967. return;
  968. }
  969. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  970. process_using_the_rules_for(InsertionMode::InHead, token);
  971. return;
  972. }
  973. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  974. log_parse_error();
  975. if (m_stack_of_open_elements.elements().size() == 1
  976. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body
  977. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  978. VERIFY(m_parsing_fragment);
  979. return;
  980. }
  981. m_frameset_ok = false;
  982. auto& body_element = m_stack_of_open_elements.elements().at(1);
  983. token.for_each_attribute([&](auto& attribute) {
  984. if (!body_element.has_attribute(attribute.local_name))
  985. body_element.set_attribute(attribute.local_name, attribute.value);
  986. return IterationDecision::Continue;
  987. });
  988. return;
  989. }
  990. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  991. log_parse_error();
  992. if (m_stack_of_open_elements.elements().size() == 1
  993. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body) {
  994. VERIFY(m_parsing_fragment);
  995. return;
  996. }
  997. if (!m_frameset_ok)
  998. return;
  999. TODO();
  1000. }
  1001. if (token.is_end_of_file()) {
  1002. if (!m_stack_of_template_insertion_modes.is_empty()) {
  1003. process_using_the_rules_for(InsertionMode::InTemplate, token);
  1004. return;
  1005. }
  1006. for (auto& node : m_stack_of_open_elements.elements()) {
  1007. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1008. log_parse_error();
  1009. break;
  1010. }
  1011. }
  1012. stop_parsing();
  1013. return;
  1014. }
  1015. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  1016. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1017. log_parse_error();
  1018. return;
  1019. }
  1020. for (auto& node : m_stack_of_open_elements.elements()) {
  1021. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1022. log_parse_error();
  1023. break;
  1024. }
  1025. }
  1026. m_insertion_mode = InsertionMode::AfterBody;
  1027. return;
  1028. }
  1029. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  1030. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1031. log_parse_error();
  1032. return;
  1033. }
  1034. for (auto& node : m_stack_of_open_elements.elements()) {
  1035. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1036. log_parse_error();
  1037. break;
  1038. }
  1039. }
  1040. m_insertion_mode = InsertionMode::AfterBody;
  1041. process_using_the_rules_for(m_insertion_mode, token);
  1042. return;
  1043. }
  1044. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::article, HTML::TagNames::aside, HTML::TagNames::blockquote, HTML::TagNames::center, HTML::TagNames::details, HTML::TagNames::dialog, HTML::TagNames::dir, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::fieldset, HTML::TagNames::figcaption, HTML::TagNames::figure, HTML::TagNames::footer, HTML::TagNames::header, HTML::TagNames::hgroup, HTML::TagNames::main, HTML::TagNames::menu, HTML::TagNames::nav, HTML::TagNames::ol, HTML::TagNames::p, HTML::TagNames::section, HTML::TagNames::summary, HTML::TagNames::ul)) {
  1045. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1046. close_a_p_element();
  1047. insert_html_element(token);
  1048. return;
  1049. }
  1050. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1051. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1052. close_a_p_element();
  1053. if (current_node().local_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1054. log_parse_error();
  1055. m_stack_of_open_elements.pop();
  1056. }
  1057. insert_html_element(token);
  1058. return;
  1059. }
  1060. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  1061. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1062. close_a_p_element();
  1063. insert_html_element(token);
  1064. m_frameset_ok = false;
  1065. // If the next token is a U+000A LINE FEED (LF) character token,
  1066. // then ignore that token and move on to the next one.
  1067. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1068. auto next_token = m_tokenizer.next_token();
  1069. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1070. // Ignore it.
  1071. } else {
  1072. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1073. }
  1074. return;
  1075. }
  1076. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1077. if (m_form_element && !m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1078. log_parse_error();
  1079. return;
  1080. }
  1081. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1082. close_a_p_element();
  1083. auto element = insert_html_element(token);
  1084. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  1085. m_form_element = verify_cast<HTMLFormElement>(*element);
  1086. return;
  1087. }
  1088. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  1089. m_frameset_ok = false;
  1090. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1091. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1092. if (node->local_name() == HTML::TagNames::li) {
  1093. generate_implied_end_tags(HTML::TagNames::li);
  1094. if (current_node().local_name() != HTML::TagNames::li) {
  1095. log_parse_error();
  1096. }
  1097. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1098. break;
  1099. }
  1100. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1101. break;
  1102. }
  1103. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1104. close_a_p_element();
  1105. insert_html_element(token);
  1106. return;
  1107. }
  1108. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1109. m_frameset_ok = false;
  1110. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1111. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1112. if (node->local_name() == HTML::TagNames::dd) {
  1113. generate_implied_end_tags(HTML::TagNames::dd);
  1114. if (current_node().local_name() != HTML::TagNames::dd) {
  1115. log_parse_error();
  1116. }
  1117. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  1118. break;
  1119. }
  1120. if (node->local_name() == HTML::TagNames::dt) {
  1121. generate_implied_end_tags(HTML::TagNames::dt);
  1122. if (current_node().local_name() != HTML::TagNames::dt) {
  1123. log_parse_error();
  1124. }
  1125. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  1126. break;
  1127. }
  1128. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1129. break;
  1130. }
  1131. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1132. close_a_p_element();
  1133. insert_html_element(token);
  1134. return;
  1135. }
  1136. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  1137. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1138. close_a_p_element();
  1139. insert_html_element(token);
  1140. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  1141. return;
  1142. }
  1143. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  1144. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  1145. log_parse_error();
  1146. generate_implied_end_tags();
  1147. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  1148. }
  1149. reconstruct_the_active_formatting_elements();
  1150. insert_html_element(token);
  1151. m_frameset_ok = false;
  1152. return;
  1153. }
  1154. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::article, HTML::TagNames::aside, HTML::TagNames::blockquote, HTML::TagNames::button, HTML::TagNames::center, HTML::TagNames::details, HTML::TagNames::dialog, HTML::TagNames::dir, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::fieldset, HTML::TagNames::figcaption, HTML::TagNames::figure, HTML::TagNames::footer, HTML::TagNames::header, HTML::TagNames::hgroup, HTML::TagNames::listing, HTML::TagNames::main, HTML::TagNames::menu, HTML::TagNames::nav, HTML::TagNames::ol, HTML::TagNames::pre, HTML::TagNames::section, HTML::TagNames::summary, HTML::TagNames::ul)) {
  1155. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1156. log_parse_error();
  1157. return;
  1158. }
  1159. generate_implied_end_tags();
  1160. if (current_node().local_name() != token.tag_name()) {
  1161. log_parse_error();
  1162. }
  1163. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1164. return;
  1165. }
  1166. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  1167. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1168. auto node = m_form_element;
  1169. m_form_element = nullptr;
  1170. if (!node || !m_stack_of_open_elements.has_in_scope(*node)) {
  1171. log_parse_error();
  1172. return;
  1173. }
  1174. generate_implied_end_tags();
  1175. if (&current_node() != node) {
  1176. log_parse_error();
  1177. }
  1178. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  1179. } else {
  1180. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  1181. log_parse_error();
  1182. return;
  1183. }
  1184. generate_implied_end_tags();
  1185. if (current_node().local_name() != HTML::TagNames::form) {
  1186. log_parse_error();
  1187. }
  1188. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  1189. }
  1190. return;
  1191. }
  1192. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  1193. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1194. log_parse_error();
  1195. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::p));
  1196. }
  1197. close_a_p_element();
  1198. return;
  1199. }
  1200. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  1201. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  1202. log_parse_error();
  1203. return;
  1204. }
  1205. generate_implied_end_tags(HTML::TagNames::li);
  1206. if (current_node().local_name() != HTML::TagNames::li) {
  1207. log_parse_error();
  1208. dbgln("Expected <li> current node, but had <{}>", current_node().local_name());
  1209. }
  1210. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1211. return;
  1212. }
  1213. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1214. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1215. log_parse_error();
  1216. return;
  1217. }
  1218. generate_implied_end_tags(token.tag_name());
  1219. if (current_node().local_name() != token.tag_name()) {
  1220. log_parse_error();
  1221. }
  1222. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1223. return;
  1224. }
  1225. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1226. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  1227. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  1228. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  1229. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  1230. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  1231. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  1232. log_parse_error();
  1233. return;
  1234. }
  1235. generate_implied_end_tags();
  1236. if (current_node().local_name() != token.tag_name()) {
  1237. log_parse_error();
  1238. }
  1239. for (;;) {
  1240. auto popped_element = m_stack_of_open_elements.pop();
  1241. if (popped_element->local_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6))
  1242. break;
  1243. }
  1244. return;
  1245. }
  1246. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1247. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1248. log_parse_error();
  1249. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1250. goto AnyOtherEndTag;
  1251. m_list_of_active_formatting_elements.remove(*element);
  1252. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1253. return entry.ptr() == element;
  1254. });
  1255. }
  1256. reconstruct_the_active_formatting_elements();
  1257. auto element = insert_html_element(token);
  1258. m_list_of_active_formatting_elements.add(*element);
  1259. return;
  1260. }
  1261. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::code, HTML::TagNames::em, HTML::TagNames::font, HTML::TagNames::i, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::strike, HTML::TagNames::strong, HTML::TagNames::tt, HTML::TagNames::u)) {
  1262. reconstruct_the_active_formatting_elements();
  1263. auto element = insert_html_element(token);
  1264. m_list_of_active_formatting_elements.add(*element);
  1265. return;
  1266. }
  1267. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1268. reconstruct_the_active_formatting_elements();
  1269. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1270. log_parse_error();
  1271. run_the_adoption_agency_algorithm(token);
  1272. reconstruct_the_active_formatting_elements();
  1273. }
  1274. auto element = insert_html_element(token);
  1275. m_list_of_active_formatting_elements.add(*element);
  1276. return;
  1277. }
  1278. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::a, HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::code, HTML::TagNames::em, HTML::TagNames::font, HTML::TagNames::i, HTML::TagNames::nobr, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::strike, HTML::TagNames::strong, HTML::TagNames::tt, HTML::TagNames::u)) {
  1279. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1280. goto AnyOtherEndTag;
  1281. return;
  1282. }
  1283. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1284. reconstruct_the_active_formatting_elements();
  1285. insert_html_element(token);
  1286. m_list_of_active_formatting_elements.add_marker();
  1287. m_frameset_ok = false;
  1288. return;
  1289. }
  1290. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1291. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1292. log_parse_error();
  1293. return;
  1294. }
  1295. generate_implied_end_tags();
  1296. if (current_node().local_name() != token.tag_name()) {
  1297. log_parse_error();
  1298. }
  1299. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1300. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1301. return;
  1302. }
  1303. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1304. if (!document().in_quirks_mode()) {
  1305. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1306. close_a_p_element();
  1307. }
  1308. insert_html_element(token);
  1309. m_frameset_ok = false;
  1310. m_insertion_mode = InsertionMode::InTable;
  1311. return;
  1312. }
  1313. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1314. token.drop_attributes();
  1315. goto BRStartTag;
  1316. }
  1317. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::area, HTML::TagNames::br, HTML::TagNames::embed, HTML::TagNames::img, HTML::TagNames::keygen, HTML::TagNames::wbr)) {
  1318. BRStartTag:
  1319. reconstruct_the_active_formatting_elements();
  1320. insert_html_element(token);
  1321. m_stack_of_open_elements.pop();
  1322. token.acknowledge_self_closing_flag_if_set();
  1323. m_frameset_ok = false;
  1324. return;
  1325. }
  1326. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1327. reconstruct_the_active_formatting_elements();
  1328. insert_html_element(token);
  1329. m_stack_of_open_elements.pop();
  1330. token.acknowledge_self_closing_flag_if_set();
  1331. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1332. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1333. m_frameset_ok = false;
  1334. }
  1335. return;
  1336. }
  1337. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1338. insert_html_element(token);
  1339. m_stack_of_open_elements.pop();
  1340. token.acknowledge_self_closing_flag_if_set();
  1341. return;
  1342. }
  1343. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1344. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1345. close_a_p_element();
  1346. insert_html_element(token);
  1347. m_stack_of_open_elements.pop();
  1348. token.acknowledge_self_closing_flag_if_set();
  1349. m_frameset_ok = false;
  1350. return;
  1351. }
  1352. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::image) {
  1353. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1354. log_parse_error();
  1355. token.set_tag_name("img");
  1356. process_using_the_rules_for(m_insertion_mode, token);
  1357. return;
  1358. }
  1359. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1360. insert_html_element(token);
  1361. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1362. // If the next token is a U+000A LINE FEED (LF) character token,
  1363. // then ignore that token and move on to the next one.
  1364. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1365. auto next_token = m_tokenizer.next_token();
  1366. m_original_insertion_mode = m_insertion_mode;
  1367. m_frameset_ok = false;
  1368. m_insertion_mode = InsertionMode::Text;
  1369. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1370. // Ignore it.
  1371. } else {
  1372. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1373. }
  1374. return;
  1375. }
  1376. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1377. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1378. close_a_p_element();
  1379. }
  1380. reconstruct_the_active_formatting_elements();
  1381. m_frameset_ok = false;
  1382. parse_generic_raw_text_element(token);
  1383. return;
  1384. }
  1385. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1386. m_frameset_ok = false;
  1387. parse_generic_raw_text_element(token);
  1388. return;
  1389. }
  1390. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1391. parse_generic_raw_text_element(token);
  1392. return;
  1393. }
  1394. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1395. reconstruct_the_active_formatting_elements();
  1396. insert_html_element(token);
  1397. m_frameset_ok = false;
  1398. switch (m_insertion_mode) {
  1399. case InsertionMode::InTable:
  1400. case InsertionMode::InCaption:
  1401. case InsertionMode::InTableBody:
  1402. case InsertionMode::InRow:
  1403. case InsertionMode::InCell:
  1404. m_insertion_mode = InsertionMode::InSelectInTable;
  1405. break;
  1406. default:
  1407. m_insertion_mode = InsertionMode::InSelect;
  1408. break;
  1409. }
  1410. return;
  1411. }
  1412. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1413. if (current_node().local_name() == HTML::TagNames::option)
  1414. m_stack_of_open_elements.pop();
  1415. reconstruct_the_active_formatting_elements();
  1416. insert_html_element(token);
  1417. return;
  1418. }
  1419. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1420. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1421. generate_implied_end_tags();
  1422. if (current_node().local_name() != HTML::TagNames::ruby)
  1423. log_parse_error();
  1424. insert_html_element(token);
  1425. return;
  1426. }
  1427. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1428. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1429. generate_implied_end_tags(HTML::TagNames::rtc);
  1430. if (current_node().local_name() != HTML::TagNames::rtc || current_node().local_name() != HTML::TagNames::ruby)
  1431. log_parse_error();
  1432. insert_html_element(token);
  1433. return;
  1434. }
  1435. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::math) {
  1436. reconstruct_the_active_formatting_elements();
  1437. adjust_mathml_attributes(token);
  1438. adjust_foreign_attributes(token);
  1439. insert_foreign_element(token, Namespace::MathML);
  1440. if (token.is_self_closing()) {
  1441. m_stack_of_open_elements.pop();
  1442. token.acknowledge_self_closing_flag_if_set();
  1443. }
  1444. return;
  1445. }
  1446. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::svg) {
  1447. reconstruct_the_active_formatting_elements();
  1448. adjust_svg_attributes(token);
  1449. adjust_foreign_attributes(token);
  1450. insert_foreign_element(token, Namespace::SVG);
  1451. if (token.is_self_closing()) {
  1452. m_stack_of_open_elements.pop();
  1453. token.acknowledge_self_closing_flag_if_set();
  1454. }
  1455. return;
  1456. }
  1457. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::frame, HTML::TagNames::head, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))) {
  1458. log_parse_error();
  1459. return;
  1460. }
  1461. // Any other start tag
  1462. if (token.is_start_tag()) {
  1463. reconstruct_the_active_formatting_elements();
  1464. insert_html_element(token);
  1465. return;
  1466. }
  1467. if (token.is_end_tag()) {
  1468. AnyOtherEndTag:
  1469. RefPtr<DOM::Element> node;
  1470. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1471. node = m_stack_of_open_elements.elements()[i];
  1472. if (node->local_name() == token.tag_name()) {
  1473. generate_implied_end_tags(token.tag_name());
  1474. if (node != current_node()) {
  1475. log_parse_error();
  1476. }
  1477. while (&current_node() != node) {
  1478. m_stack_of_open_elements.pop();
  1479. }
  1480. m_stack_of_open_elements.pop();
  1481. break;
  1482. }
  1483. if (is_special_tag(node->local_name(), node->namespace_())) {
  1484. log_parse_error();
  1485. return;
  1486. }
  1487. }
  1488. return;
  1489. }
  1490. }
  1491. void HTMLDocumentParser::adjust_mathml_attributes(HTMLToken& token)
  1492. {
  1493. token.adjust_attribute_name("definitionurl", "definitionURL");
  1494. }
  1495. void HTMLDocumentParser::adjust_svg_tag_names(HTMLToken& token)
  1496. {
  1497. token.adjust_tag_name("altglyph", "altGlyph");
  1498. token.adjust_tag_name("altglyphdef", "altGlyphDef");
  1499. token.adjust_tag_name("altglyphitem", "altGlyphItem");
  1500. token.adjust_tag_name("animatecolor", "animateColor");
  1501. token.adjust_tag_name("animatemotion", "animateMotion");
  1502. token.adjust_tag_name("animatetransform", "animateTransform");
  1503. token.adjust_tag_name("clippath", "clipPath");
  1504. token.adjust_tag_name("feblend", "feBlend");
  1505. token.adjust_tag_name("fecolormatrix", "feColorMatrix");
  1506. token.adjust_tag_name("fecomponenttransfer", "feComponentTransfer");
  1507. token.adjust_tag_name("fecomposite", "feComposite");
  1508. token.adjust_tag_name("feconvolvematrix", "feConvolveMatrix");
  1509. token.adjust_tag_name("fediffuselighting", "feDiffuseLighting");
  1510. token.adjust_tag_name("fedisplacementmap", "feDisplacementMap");
  1511. token.adjust_tag_name("fedistantlight", "feDistantLight");
  1512. token.adjust_tag_name("fedropshadow", "feDropShadow");
  1513. token.adjust_tag_name("feflood", "feFlood");
  1514. token.adjust_tag_name("fefunca", "feFuncA");
  1515. token.adjust_tag_name("fefuncb", "feFuncB");
  1516. token.adjust_tag_name("fefuncg", "feFuncG");
  1517. token.adjust_tag_name("fefuncr", "feFuncR");
  1518. token.adjust_tag_name("fegaussianblur", "feGaussianBlur");
  1519. token.adjust_tag_name("feimage", "feImage");
  1520. token.adjust_tag_name("femerge", "feMerge");
  1521. token.adjust_tag_name("femergenode", "feMergeNode");
  1522. token.adjust_tag_name("femorphology", "feMorphology");
  1523. token.adjust_tag_name("feoffset", "feOffset");
  1524. token.adjust_tag_name("fepointlight", "fePointLight");
  1525. token.adjust_tag_name("fespecularlighting", "feSpecularLighting");
  1526. token.adjust_tag_name("fespotlight", "feSpotlight");
  1527. token.adjust_tag_name("glyphref", "glyphRef");
  1528. token.adjust_tag_name("lineargradient", "linearGradient");
  1529. token.adjust_tag_name("radialgradient", "radialGradient");
  1530. token.adjust_tag_name("textpath", "textPath");
  1531. }
  1532. void HTMLDocumentParser::adjust_svg_attributes(HTMLToken& token)
  1533. {
  1534. token.adjust_attribute_name("attributename", "attributeName");
  1535. token.adjust_attribute_name("attributetype", "attributeType");
  1536. token.adjust_attribute_name("basefrequency", "baseFrequency");
  1537. token.adjust_attribute_name("baseprofile", "baseProfile");
  1538. token.adjust_attribute_name("calcmode", "calcMode");
  1539. token.adjust_attribute_name("clippathunits", "clipPathUnits");
  1540. token.adjust_attribute_name("diffuseconstant", "diffuseConstant");
  1541. token.adjust_attribute_name("edgemode", "edgeMode");
  1542. token.adjust_attribute_name("filterunits", "filterUnits");
  1543. token.adjust_attribute_name("glyphref", "glyphRef");
  1544. token.adjust_attribute_name("gradienttransform", "gradientTransform");
  1545. token.adjust_attribute_name("gradientunits", "gradientUnits");
  1546. token.adjust_attribute_name("kernelmatrix", "kernelMatrix");
  1547. token.adjust_attribute_name("kernelunitlength", "kernelUnitLength");
  1548. token.adjust_attribute_name("keypoints", "keyPoints");
  1549. token.adjust_attribute_name("keysplines", "keySplines");
  1550. token.adjust_attribute_name("keytimes", "keyTimes");
  1551. token.adjust_attribute_name("lengthadjust", "lengthAdjust");
  1552. token.adjust_attribute_name("limitingconeangle", "limitingConeAngle");
  1553. token.adjust_attribute_name("markerheight", "markerHeight");
  1554. token.adjust_attribute_name("markerunits", "markerUnits");
  1555. token.adjust_attribute_name("markerwidth", "markerWidth");
  1556. token.adjust_attribute_name("maskcontentunits", "maskContentUnits");
  1557. token.adjust_attribute_name("maskunits", "maskUnits");
  1558. token.adjust_attribute_name("numoctaves", "numOctaves");
  1559. token.adjust_attribute_name("pathlength", "pathLength");
  1560. token.adjust_attribute_name("patterncontentunits", "patternContentUnits");
  1561. token.adjust_attribute_name("patterntransform", "patternTransform");
  1562. token.adjust_attribute_name("patternunits", "patternUnits");
  1563. token.adjust_attribute_name("pointsatx", "pointsAtX");
  1564. token.adjust_attribute_name("pointsaty", "pointsAtY");
  1565. token.adjust_attribute_name("pointsatz", "pointsAtZ");
  1566. token.adjust_attribute_name("preservealpha", "preserveAlpha");
  1567. token.adjust_attribute_name("preserveaspectratio", "preserveAspectRatio");
  1568. token.adjust_attribute_name("primitiveunits", "primitiveUnits");
  1569. token.adjust_attribute_name("refx", "refX");
  1570. token.adjust_attribute_name("refy", "refY");
  1571. token.adjust_attribute_name("repeatcount", "repeatCount");
  1572. token.adjust_attribute_name("repeatdur", "repeatDur");
  1573. token.adjust_attribute_name("requiredextensions", "requiredExtensions");
  1574. token.adjust_attribute_name("requiredfeatures", "requiredFeatures");
  1575. token.adjust_attribute_name("specularconstant", "specularConstant");
  1576. token.adjust_attribute_name("specularexponent", "specularExponent");
  1577. token.adjust_attribute_name("spreadmethod", "spreadMethod");
  1578. token.adjust_attribute_name("startoffset", "startOffset");
  1579. token.adjust_attribute_name("stddeviation", "stdDeviation");
  1580. token.adjust_attribute_name("stitchtiles", "stitchTiles");
  1581. token.adjust_attribute_name("surfacescale", "surfaceScale");
  1582. token.adjust_attribute_name("systemlanguage", "systemLanguage");
  1583. token.adjust_attribute_name("tablevalues", "tableValues");
  1584. token.adjust_attribute_name("targetx", "targetX");
  1585. token.adjust_attribute_name("targety", "targetY");
  1586. token.adjust_attribute_name("textlength", "textLength");
  1587. token.adjust_attribute_name("viewbox", "viewBox");
  1588. token.adjust_attribute_name("viewtarget", "viewTarget");
  1589. token.adjust_attribute_name("xchannelselector", "xChannelSelector");
  1590. token.adjust_attribute_name("ychannelselector", "yChannelSelector");
  1591. token.adjust_attribute_name("zoomandpan", "zoomAndPan");
  1592. }
  1593. void HTMLDocumentParser::adjust_foreign_attributes(HTMLToken& token)
  1594. {
  1595. token.adjust_foreign_attribute("xlink:actuate", "xlink", "actuate", Namespace::XLink);
  1596. token.adjust_foreign_attribute("xlink:arcrole", "xlink", "arcrole", Namespace::XLink);
  1597. token.adjust_foreign_attribute("xlink:href", "xlink", "href", Namespace::XLink);
  1598. token.adjust_foreign_attribute("xlink:role", "xlink", "role", Namespace::XLink);
  1599. token.adjust_foreign_attribute("xlink:show", "xlink", "show", Namespace::XLink);
  1600. token.adjust_foreign_attribute("xlink:title", "xlink", "title", Namespace::XLink);
  1601. token.adjust_foreign_attribute("xlink:type", "xlink", "type", Namespace::XLink);
  1602. token.adjust_foreign_attribute("xml:lang", "xml", "lang", Namespace::XML);
  1603. token.adjust_foreign_attribute("xml:space", "xml", "space", Namespace::XML);
  1604. token.adjust_foreign_attribute("xmlns", "", "xmlns", Namespace::XMLNS);
  1605. token.adjust_foreign_attribute("xmlns:xlink", "xmlns", "xlink", Namespace::XMLNS);
  1606. }
  1607. void HTMLDocumentParser::increment_script_nesting_level()
  1608. {
  1609. ++m_script_nesting_level;
  1610. }
  1611. void HTMLDocumentParser::decrement_script_nesting_level()
  1612. {
  1613. VERIFY(m_script_nesting_level);
  1614. --m_script_nesting_level;
  1615. }
  1616. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1617. {
  1618. if (token.is_character()) {
  1619. insert_character(token.code_point());
  1620. return;
  1621. }
  1622. if (token.is_end_of_file()) {
  1623. log_parse_error();
  1624. if (current_node().local_name() == HTML::TagNames::script)
  1625. verify_cast<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1626. m_stack_of_open_elements.pop();
  1627. m_insertion_mode = m_original_insertion_mode;
  1628. process_using_the_rules_for(m_insertion_mode, token);
  1629. return;
  1630. }
  1631. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1632. // Make sure the <script> element has up-to-date text content before preparing the script.
  1633. flush_character_insertions();
  1634. NonnullRefPtr<HTMLScriptElement> script = verify_cast<HTMLScriptElement>(current_node());
  1635. m_stack_of_open_elements.pop();
  1636. m_insertion_mode = m_original_insertion_mode;
  1637. // FIXME: Handle tokenizer insertion point stuff here.
  1638. increment_script_nesting_level();
  1639. script->prepare_script({});
  1640. decrement_script_nesting_level();
  1641. if (script_nesting_level() == 0)
  1642. m_parser_pause_flag = false;
  1643. // FIXME: Handle tokenizer insertion point stuff here too.
  1644. while (document().pending_parsing_blocking_script()) {
  1645. if (script_nesting_level() != 0) {
  1646. m_parser_pause_flag = true;
  1647. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1648. // yielding control back to the caller. (Tokenization will resume when
  1649. // the caller returns to the "outer" tree construction stage.)
  1650. TODO();
  1651. } else {
  1652. auto the_script = document().take_pending_parsing_blocking_script({});
  1653. m_tokenizer.set_blocked(true);
  1654. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1655. // or the script's "ready to be parser-executed" flag is not set:
  1656. // spin the event loop until the parser's Document has no style sheet
  1657. // that is blocking scripts and the script's "ready to be parser-executed"
  1658. // flag is set.
  1659. if (the_script->failed_to_load())
  1660. return;
  1661. VERIFY(the_script->is_ready_to_be_parser_executed());
  1662. if (m_aborted)
  1663. return;
  1664. m_tokenizer.set_blocked(false);
  1665. // FIXME: Handle tokenizer insertion point stuff here too.
  1666. VERIFY(script_nesting_level() == 0);
  1667. increment_script_nesting_level();
  1668. the_script->execute_script();
  1669. decrement_script_nesting_level();
  1670. VERIFY(script_nesting_level() == 0);
  1671. m_parser_pause_flag = false;
  1672. // FIXME: Handle tokenizer insertion point stuff here too.
  1673. }
  1674. }
  1675. return;
  1676. }
  1677. if (token.is_end_tag()) {
  1678. m_stack_of_open_elements.pop();
  1679. m_insertion_mode = m_original_insertion_mode;
  1680. return;
  1681. }
  1682. TODO();
  1683. }
  1684. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1685. {
  1686. while (!current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1687. m_stack_of_open_elements.pop();
  1688. if (current_node().local_name() == HTML::TagNames::html)
  1689. VERIFY(m_parsing_fragment);
  1690. }
  1691. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1692. {
  1693. while (!current_node().local_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1694. m_stack_of_open_elements.pop();
  1695. if (current_node().local_name() == HTML::TagNames::html)
  1696. VERIFY(m_parsing_fragment);
  1697. }
  1698. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1699. {
  1700. while (!current_node().local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1701. m_stack_of_open_elements.pop();
  1702. if (current_node().local_name() == HTML::TagNames::html)
  1703. VERIFY(m_parsing_fragment);
  1704. }
  1705. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1706. {
  1707. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1708. clear_the_stack_back_to_a_table_row_context();
  1709. insert_html_element(token);
  1710. m_insertion_mode = InsertionMode::InCell;
  1711. m_list_of_active_formatting_elements.add_marker();
  1712. return;
  1713. }
  1714. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1715. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1716. log_parse_error();
  1717. return;
  1718. }
  1719. clear_the_stack_back_to_a_table_row_context();
  1720. m_stack_of_open_elements.pop();
  1721. m_insertion_mode = InsertionMode::InTableBody;
  1722. return;
  1723. }
  1724. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr))
  1725. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1726. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1727. log_parse_error();
  1728. return;
  1729. }
  1730. clear_the_stack_back_to_a_table_row_context();
  1731. m_stack_of_open_elements.pop();
  1732. m_insertion_mode = InsertionMode::InTableBody;
  1733. process_using_the_rules_for(m_insertion_mode, token);
  1734. return;
  1735. }
  1736. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1737. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1738. log_parse_error();
  1739. return;
  1740. }
  1741. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1742. return;
  1743. }
  1744. clear_the_stack_back_to_a_table_row_context();
  1745. m_stack_of_open_elements.pop();
  1746. m_insertion_mode = InsertionMode::InTableBody;
  1747. process_using_the_rules_for(m_insertion_mode, token);
  1748. return;
  1749. }
  1750. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::td, HTML::TagNames::th)) {
  1751. log_parse_error();
  1752. return;
  1753. }
  1754. process_using_the_rules_for(InsertionMode::InTable, token);
  1755. }
  1756. void HTMLDocumentParser::close_the_cell()
  1757. {
  1758. generate_implied_end_tags();
  1759. if (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1760. log_parse_error();
  1761. }
  1762. while (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1763. m_stack_of_open_elements.pop();
  1764. m_stack_of_open_elements.pop();
  1765. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1766. m_insertion_mode = InsertionMode::InRow;
  1767. }
  1768. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1769. {
  1770. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1771. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1772. log_parse_error();
  1773. return;
  1774. }
  1775. generate_implied_end_tags();
  1776. if (current_node().local_name() != token.tag_name()) {
  1777. log_parse_error();
  1778. }
  1779. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1780. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1781. m_insertion_mode = InsertionMode::InRow;
  1782. return;
  1783. }
  1784. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1785. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::td) && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::th)) {
  1786. VERIFY(m_parsing_fragment);
  1787. log_parse_error();
  1788. return;
  1789. }
  1790. close_the_cell();
  1791. process_using_the_rules_for(m_insertion_mode, token);
  1792. return;
  1793. }
  1794. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html)) {
  1795. log_parse_error();
  1796. return;
  1797. }
  1798. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1799. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1800. log_parse_error();
  1801. return;
  1802. }
  1803. close_the_cell();
  1804. // Reprocess the token.
  1805. process_using_the_rules_for(m_insertion_mode, token);
  1806. return;
  1807. }
  1808. process_using_the_rules_for(InsertionMode::InBody, token);
  1809. }
  1810. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1811. {
  1812. if (token.is_character()) {
  1813. if (token.code_point() == 0) {
  1814. log_parse_error();
  1815. return;
  1816. }
  1817. m_pending_table_character_tokens.append(move(token));
  1818. return;
  1819. }
  1820. for (auto& pending_token : m_pending_table_character_tokens) {
  1821. VERIFY(pending_token.is_character());
  1822. if (!pending_token.is_parser_whitespace()) {
  1823. // If any of the tokens in the pending table character tokens list
  1824. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1825. // reprocess the character tokens in the pending table character tokens list using
  1826. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1827. log_parse_error();
  1828. m_foster_parenting = true;
  1829. process_using_the_rules_for(InsertionMode::InBody, token);
  1830. m_foster_parenting = false;
  1831. return;
  1832. }
  1833. }
  1834. for (auto& pending_token : m_pending_table_character_tokens) {
  1835. insert_character(pending_token.code_point());
  1836. }
  1837. m_insertion_mode = m_original_insertion_mode;
  1838. process_using_the_rules_for(m_insertion_mode, token);
  1839. }
  1840. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1841. {
  1842. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1843. clear_the_stack_back_to_a_table_body_context();
  1844. insert_html_element(token);
  1845. m_insertion_mode = InsertionMode::InRow;
  1846. return;
  1847. }
  1848. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1849. log_parse_error();
  1850. clear_the_stack_back_to_a_table_body_context();
  1851. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tr));
  1852. m_insertion_mode = InsertionMode::InRow;
  1853. process_using_the_rules_for(m_insertion_mode, token);
  1854. return;
  1855. }
  1856. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1857. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1858. log_parse_error();
  1859. return;
  1860. }
  1861. clear_the_stack_back_to_a_table_body_context();
  1862. m_stack_of_open_elements.pop();
  1863. m_insertion_mode = InsertionMode::InTable;
  1864. return;
  1865. }
  1866. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead))
  1867. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1868. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tbody)
  1869. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::thead)
  1870. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tfoot)) {
  1871. log_parse_error();
  1872. return;
  1873. }
  1874. clear_the_stack_back_to_a_table_body_context();
  1875. m_stack_of_open_elements.pop();
  1876. m_insertion_mode = InsertionMode::InTable;
  1877. process_using_the_rules_for(InsertionMode::InTable, token);
  1878. return;
  1879. }
  1880. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1881. log_parse_error();
  1882. return;
  1883. }
  1884. process_using_the_rules_for(InsertionMode::InTable, token);
  1885. }
  1886. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1887. {
  1888. if (token.is_character() && current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1889. m_pending_table_character_tokens.clear();
  1890. m_original_insertion_mode = m_insertion_mode;
  1891. m_insertion_mode = InsertionMode::InTableText;
  1892. process_using_the_rules_for(InsertionMode::InTableText, token);
  1893. return;
  1894. }
  1895. if (token.is_comment()) {
  1896. insert_comment(token);
  1897. return;
  1898. }
  1899. if (token.is_doctype()) {
  1900. log_parse_error();
  1901. return;
  1902. }
  1903. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1904. clear_the_stack_back_to_a_table_context();
  1905. m_list_of_active_formatting_elements.add_marker();
  1906. insert_html_element(token);
  1907. m_insertion_mode = InsertionMode::InCaption;
  1908. return;
  1909. }
  1910. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1911. clear_the_stack_back_to_a_table_context();
  1912. insert_html_element(token);
  1913. m_insertion_mode = InsertionMode::InColumnGroup;
  1914. return;
  1915. }
  1916. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1917. clear_the_stack_back_to_a_table_context();
  1918. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::colgroup));
  1919. m_insertion_mode = InsertionMode::InColumnGroup;
  1920. process_using_the_rules_for(m_insertion_mode, token);
  1921. return;
  1922. }
  1923. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1924. clear_the_stack_back_to_a_table_context();
  1925. insert_html_element(token);
  1926. m_insertion_mode = InsertionMode::InTableBody;
  1927. return;
  1928. }
  1929. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1930. clear_the_stack_back_to_a_table_context();
  1931. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tbody));
  1932. m_insertion_mode = InsertionMode::InTableBody;
  1933. process_using_the_rules_for(m_insertion_mode, token);
  1934. return;
  1935. }
  1936. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1937. log_parse_error();
  1938. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table))
  1939. return;
  1940. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1941. reset_the_insertion_mode_appropriately();
  1942. process_using_the_rules_for(m_insertion_mode, token);
  1943. return;
  1944. }
  1945. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::table) {
  1946. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1947. log_parse_error();
  1948. return;
  1949. }
  1950. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1951. reset_the_insertion_mode_appropriately();
  1952. return;
  1953. }
  1954. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1955. log_parse_error();
  1956. return;
  1957. }
  1958. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::style, HTML::TagNames::script, HTML::TagNames::template_))
  1959. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_)) {
  1960. process_using_the_rules_for(InsertionMode::InHead, token);
  1961. return;
  1962. }
  1963. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1964. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1965. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1966. goto AnythingElse;
  1967. }
  1968. log_parse_error();
  1969. insert_html_element(token);
  1970. // FIXME: Is this the correct interpretation of "Pop that input element off the stack of open elements."?
  1971. // Because this wording is the first time it's seen in the spec.
  1972. // Other times it's worded as: "Immediately pop the current node off the stack of open elements."
  1973. m_stack_of_open_elements.pop();
  1974. token.acknowledge_self_closing_flag_if_set();
  1975. return;
  1976. }
  1977. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1978. log_parse_error();
  1979. if (m_form_element || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1980. return;
  1981. }
  1982. m_form_element = verify_cast<HTMLFormElement>(*insert_html_element(token));
  1983. // FIXME: See previous FIXME, as this is the same situation but for form.
  1984. m_stack_of_open_elements.pop();
  1985. return;
  1986. }
  1987. if (token.is_end_of_file()) {
  1988. process_using_the_rules_for(InsertionMode::InBody, token);
  1989. return;
  1990. }
  1991. AnythingElse:
  1992. log_parse_error();
  1993. m_foster_parenting = true;
  1994. process_using_the_rules_for(InsertionMode::InBody, token);
  1995. m_foster_parenting = false;
  1996. }
  1997. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1998. {
  1999. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::td, HTML::TagNames::th)) {
  2000. log_parse_error();
  2001. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2002. reset_the_insertion_mode_appropriately();
  2003. process_using_the_rules_for(m_insertion_mode, token);
  2004. return;
  2005. }
  2006. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::td, HTML::TagNames::th)) {
  2007. log_parse_error();
  2008. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name()))
  2009. return;
  2010. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2011. reset_the_insertion_mode_appropriately();
  2012. process_using_the_rules_for(m_insertion_mode, token);
  2013. return;
  2014. }
  2015. process_using_the_rules_for(InsertionMode::InSelect, token);
  2016. }
  2017. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  2018. {
  2019. if (token.is_character()) {
  2020. if (token.code_point() == 0) {
  2021. log_parse_error();
  2022. return;
  2023. }
  2024. insert_character(token.code_point());
  2025. return;
  2026. }
  2027. if (token.is_comment()) {
  2028. insert_comment(token);
  2029. return;
  2030. }
  2031. if (token.is_doctype()) {
  2032. log_parse_error();
  2033. return;
  2034. }
  2035. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2036. process_using_the_rules_for(InsertionMode::InBody, token);
  2037. return;
  2038. }
  2039. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  2040. if (current_node().local_name() == HTML::TagNames::option) {
  2041. m_stack_of_open_elements.pop();
  2042. }
  2043. insert_html_element(token);
  2044. return;
  2045. }
  2046. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2047. if (current_node().local_name() == HTML::TagNames::option) {
  2048. m_stack_of_open_elements.pop();
  2049. }
  2050. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2051. m_stack_of_open_elements.pop();
  2052. }
  2053. insert_html_element(token);
  2054. return;
  2055. }
  2056. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2057. if (current_node().local_name() == HTML::TagNames::option && node_before_current_node().local_name() == HTML::TagNames::optgroup)
  2058. m_stack_of_open_elements.pop();
  2059. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2060. m_stack_of_open_elements.pop();
  2061. } else {
  2062. log_parse_error();
  2063. return;
  2064. }
  2065. return;
  2066. }
  2067. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  2068. if (current_node().local_name() == HTML::TagNames::option) {
  2069. m_stack_of_open_elements.pop();
  2070. } else {
  2071. log_parse_error();
  2072. return;
  2073. }
  2074. return;
  2075. }
  2076. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  2077. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2078. VERIFY(m_parsing_fragment);
  2079. log_parse_error();
  2080. return;
  2081. }
  2082. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2083. reset_the_insertion_mode_appropriately();
  2084. return;
  2085. }
  2086. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  2087. log_parse_error();
  2088. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2089. VERIFY(m_parsing_fragment);
  2090. return;
  2091. }
  2092. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2093. reset_the_insertion_mode_appropriately();
  2094. return;
  2095. }
  2096. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  2097. log_parse_error();
  2098. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2099. VERIFY(m_parsing_fragment);
  2100. return;
  2101. }
  2102. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2103. reset_the_insertion_mode_appropriately();
  2104. process_using_the_rules_for(m_insertion_mode, token);
  2105. return;
  2106. }
  2107. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  2108. process_using_the_rules_for(InsertionMode::InHead, token);
  2109. return;
  2110. }
  2111. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2112. process_using_the_rules_for(InsertionMode::InHead, token);
  2113. return;
  2114. }
  2115. if (token.is_end_of_file()) {
  2116. process_using_the_rules_for(InsertionMode::InBody, token);
  2117. return;
  2118. }
  2119. log_parse_error();
  2120. }
  2121. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  2122. {
  2123. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  2124. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2125. VERIFY(m_parsing_fragment);
  2126. log_parse_error();
  2127. return;
  2128. }
  2129. generate_implied_end_tags();
  2130. if (current_node().local_name() != HTML::TagNames::caption)
  2131. log_parse_error();
  2132. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2133. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2134. m_insertion_mode = InsertionMode::InTable;
  2135. return;
  2136. }
  2137. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))
  2138. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  2139. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2140. VERIFY(m_parsing_fragment);
  2141. log_parse_error();
  2142. return;
  2143. }
  2144. generate_implied_end_tags();
  2145. if (current_node().local_name() != HTML::TagNames::caption)
  2146. log_parse_error();
  2147. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2148. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2149. m_insertion_mode = InsertionMode::InTable;
  2150. process_using_the_rules_for(m_insertion_mode, token);
  2151. return;
  2152. }
  2153. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  2154. log_parse_error();
  2155. return;
  2156. }
  2157. process_using_the_rules_for(InsertionMode::InBody, token);
  2158. }
  2159. void HTMLDocumentParser::handle_in_column_group(HTMLToken& token)
  2160. {
  2161. if (token.is_character() && token.is_parser_whitespace()) {
  2162. insert_character(token.code_point());
  2163. return;
  2164. }
  2165. if (token.is_comment()) {
  2166. insert_comment(token);
  2167. return;
  2168. }
  2169. if (token.is_doctype()) {
  2170. log_parse_error();
  2171. return;
  2172. }
  2173. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2174. process_using_the_rules_for(InsertionMode::InBody, token);
  2175. return;
  2176. }
  2177. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2178. insert_html_element(token);
  2179. m_stack_of_open_elements.pop();
  2180. token.acknowledge_self_closing_flag_if_set();
  2181. return;
  2182. }
  2183. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  2184. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2185. log_parse_error();
  2186. return;
  2187. }
  2188. m_stack_of_open_elements.pop();
  2189. m_insertion_mode = InsertionMode::InTable;
  2190. return;
  2191. }
  2192. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::col) {
  2193. log_parse_error();
  2194. return;
  2195. }
  2196. if ((token.is_start_tag() || token.is_end_tag()) && token.tag_name() == HTML::TagNames::template_) {
  2197. process_using_the_rules_for(InsertionMode::InHead, token);
  2198. return;
  2199. }
  2200. if (token.is_end_of_file()) {
  2201. process_using_the_rules_for(InsertionMode::InBody, token);
  2202. return;
  2203. }
  2204. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2205. log_parse_error();
  2206. return;
  2207. }
  2208. m_stack_of_open_elements.pop();
  2209. m_insertion_mode = InsertionMode::InTable;
  2210. process_using_the_rules_for(m_insertion_mode, token);
  2211. }
  2212. void HTMLDocumentParser::handle_in_template(HTMLToken& token)
  2213. {
  2214. if (token.is_character() || token.is_comment() || token.is_doctype()) {
  2215. process_using_the_rules_for(InsertionMode::InBody, token);
  2216. return;
  2217. }
  2218. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  2219. process_using_the_rules_for(InsertionMode::InHead, token);
  2220. return;
  2221. }
  2222. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2223. process_using_the_rules_for(InsertionMode::InHead, token);
  2224. return;
  2225. }
  2226. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  2227. m_stack_of_template_insertion_modes.take_last();
  2228. m_stack_of_template_insertion_modes.append(InsertionMode::InTable);
  2229. m_insertion_mode = InsertionMode::InTable;
  2230. process_using_the_rules_for(m_insertion_mode, token);
  2231. return;
  2232. }
  2233. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2234. m_stack_of_template_insertion_modes.take_last();
  2235. m_stack_of_template_insertion_modes.append(InsertionMode::InColumnGroup);
  2236. m_insertion_mode = InsertionMode::InColumnGroup;
  2237. process_using_the_rules_for(m_insertion_mode, token);
  2238. return;
  2239. }
  2240. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  2241. m_stack_of_template_insertion_modes.take_last();
  2242. m_stack_of_template_insertion_modes.append(InsertionMode::InTableBody);
  2243. m_insertion_mode = InsertionMode::InTableBody;
  2244. process_using_the_rules_for(m_insertion_mode, token);
  2245. return;
  2246. }
  2247. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2248. m_stack_of_template_insertion_modes.take_last();
  2249. m_stack_of_template_insertion_modes.append(InsertionMode::InRow);
  2250. m_insertion_mode = InsertionMode::InRow;
  2251. process_using_the_rules_for(m_insertion_mode, token);
  2252. return;
  2253. }
  2254. if (token.is_start_tag()) {
  2255. m_stack_of_template_insertion_modes.take_last();
  2256. m_stack_of_template_insertion_modes.append(InsertionMode::InBody);
  2257. m_insertion_mode = InsertionMode::InBody;
  2258. process_using_the_rules_for(m_insertion_mode, token);
  2259. return;
  2260. }
  2261. if (token.is_end_tag()) {
  2262. log_parse_error();
  2263. return;
  2264. }
  2265. if (token.is_end_of_file()) {
  2266. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  2267. VERIFY(m_parsing_fragment);
  2268. stop_parsing();
  2269. return;
  2270. }
  2271. log_parse_error();
  2272. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  2273. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2274. m_stack_of_template_insertion_modes.take_last();
  2275. reset_the_insertion_mode_appropriately();
  2276. process_using_the_rules_for(m_insertion_mode, token);
  2277. }
  2278. }
  2279. void HTMLDocumentParser::handle_in_frameset(HTMLToken& token)
  2280. {
  2281. if (token.is_character() && token.is_parser_whitespace()) {
  2282. insert_character(token.code_point());
  2283. return;
  2284. }
  2285. if (token.is_comment()) {
  2286. insert_comment(token);
  2287. return;
  2288. }
  2289. if (token.is_doctype()) {
  2290. log_parse_error();
  2291. return;
  2292. }
  2293. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2294. process_using_the_rules_for(InsertionMode::InBody, token);
  2295. return;
  2296. }
  2297. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2298. insert_html_element(token);
  2299. return;
  2300. }
  2301. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2302. // FIXME: If the current node is the root html element, then this is a parse error; ignore the token. (fragment case)
  2303. m_stack_of_open_elements.pop();
  2304. if (!m_parsing_fragment && current_node().local_name() != HTML::TagNames::frameset) {
  2305. m_insertion_mode = InsertionMode::AfterFrameset;
  2306. }
  2307. return;
  2308. }
  2309. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frame) {
  2310. insert_html_element(token);
  2311. m_stack_of_open_elements.pop();
  2312. token.acknowledge_self_closing_flag_if_set();
  2313. return;
  2314. }
  2315. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2316. process_using_the_rules_for(InsertionMode::InHead, token);
  2317. return;
  2318. }
  2319. if (token.is_end_of_file()) {
  2320. //FIXME: If the current node is not the root html element, then this is a parse error.
  2321. stop_parsing();
  2322. return;
  2323. }
  2324. log_parse_error();
  2325. }
  2326. void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
  2327. {
  2328. if (token.is_character() && token.is_parser_whitespace()) {
  2329. insert_character(token.code_point());
  2330. return;
  2331. }
  2332. if (token.is_comment()) {
  2333. insert_comment(token);
  2334. return;
  2335. }
  2336. if (token.is_doctype()) {
  2337. log_parse_error();
  2338. return;
  2339. }
  2340. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2341. process_using_the_rules_for(InsertionMode::InBody, token);
  2342. return;
  2343. }
  2344. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  2345. m_insertion_mode = InsertionMode::AfterAfterFrameset;
  2346. return;
  2347. }
  2348. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2349. process_using_the_rules_for(InsertionMode::InHead, token);
  2350. return;
  2351. }
  2352. if (token.is_end_of_file()) {
  2353. stop_parsing();
  2354. return;
  2355. }
  2356. log_parse_error();
  2357. }
  2358. void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
  2359. {
  2360. if (token.is_comment()) {
  2361. auto comment = adopt_ref(*new DOM::Comment(document(), token.comment()));
  2362. document().append_child(move(comment));
  2363. return;
  2364. }
  2365. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  2366. process_using_the_rules_for(InsertionMode::InBody, token);
  2367. return;
  2368. }
  2369. if (token.is_end_of_file()) {
  2370. stop_parsing();
  2371. return;
  2372. }
  2373. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2374. process_using_the_rules_for(InsertionMode::InHead, token);
  2375. return;
  2376. }
  2377. log_parse_error();
  2378. }
  2379. void HTMLDocumentParser::process_using_the_rules_for_foreign_content(HTMLToken& token)
  2380. {
  2381. if (token.is_character()) {
  2382. if (token.code_point() == 0) {
  2383. log_parse_error();
  2384. insert_character(0xFFFD);
  2385. return;
  2386. }
  2387. if (token.is_parser_whitespace()) {
  2388. insert_character(token.code_point());
  2389. return;
  2390. }
  2391. insert_character(token.code_point());
  2392. m_frameset_ok = false;
  2393. return;
  2394. }
  2395. if (token.is_comment()) {
  2396. insert_comment(token);
  2397. return;
  2398. }
  2399. if (token.is_doctype()) {
  2400. log_parse_error();
  2401. return;
  2402. }
  2403. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::blockquote, HTML::TagNames::body, HTML::TagNames::br, HTML::TagNames::center, HTML::TagNames::code, HTML::TagNames::dd, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::dt, HTML::TagNames::em, HTML::TagNames::embed, HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6, HTML::TagNames::head, HTML::TagNames::hr, HTML::TagNames::i, HTML::TagNames::img, HTML::TagNames::li, HTML::TagNames::listing, HTML::TagNames::menu, HTML::TagNames::meta, HTML::TagNames::nobr, HTML::TagNames::ol, HTML::TagNames::p, HTML::TagNames::pre, HTML::TagNames::ruby, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::span, HTML::TagNames::strong, HTML::TagNames::strike, HTML::TagNames::sub, HTML::TagNames::sup, HTML::TagNames::table, HTML::TagNames::tt, HTML::TagNames::u, HTML::TagNames::ul, HTML::TagNames::var))
  2404. || (token.is_start_tag() && token.tag_name() == HTML::TagNames::font && (token.has_attribute(HTML::AttributeNames::color) || token.has_attribute(HTML::AttributeNames::face) || token.has_attribute(HTML::AttributeNames::size)))) {
  2405. log_parse_error();
  2406. if (m_parsing_fragment) {
  2407. goto AnyOtherStartTag;
  2408. }
  2409. TODO();
  2410. }
  2411. if (token.is_start_tag()) {
  2412. AnyOtherStartTag:
  2413. if (adjusted_current_node().namespace_() == Namespace::MathML) {
  2414. adjust_mathml_attributes(token);
  2415. } else if (adjusted_current_node().namespace_() == Namespace::SVG) {
  2416. adjust_svg_tag_names(token);
  2417. adjust_svg_attributes(token);
  2418. }
  2419. adjust_foreign_attributes(token);
  2420. insert_foreign_element(token, adjusted_current_node().namespace_());
  2421. if (token.is_self_closing()) {
  2422. if (token.tag_name() == SVG::TagNames::script && current_node().namespace_() == Namespace::SVG) {
  2423. token.acknowledge_self_closing_flag_if_set();
  2424. goto ScriptEndTag;
  2425. }
  2426. m_stack_of_open_elements.pop();
  2427. token.acknowledge_self_closing_flag_if_set();
  2428. }
  2429. return;
  2430. }
  2431. if (token.is_end_tag() && current_node().namespace_() == Namespace::SVG && current_node().tag_name() == SVG::TagNames::script) {
  2432. ScriptEndTag:
  2433. m_stack_of_open_elements.pop();
  2434. TODO();
  2435. }
  2436. if (token.is_end_tag()) {
  2437. RefPtr<DOM::Element> node = current_node();
  2438. // FIXME: Not sure if this is the correct to_lowercase, as the specification says "to ASCII lowercase"
  2439. if (node->tag_name().to_lowercase() != token.tag_name())
  2440. log_parse_error();
  2441. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2442. if (node == m_stack_of_open_elements.first()) {
  2443. VERIFY(m_parsing_fragment);
  2444. return;
  2445. }
  2446. // FIXME: See the above FIXME
  2447. if (node->tag_name().to_lowercase() == token.tag_name()) {
  2448. while (current_node() != node)
  2449. m_stack_of_open_elements.pop();
  2450. m_stack_of_open_elements.pop();
  2451. return;
  2452. }
  2453. node = m_stack_of_open_elements.elements().at(i - 1);
  2454. if (node->namespace_() != Namespace::HTML)
  2455. continue;
  2456. process_using_the_rules_for(m_insertion_mode, token);
  2457. return;
  2458. }
  2459. }
  2460. VERIFY_NOT_REACHED();
  2461. }
  2462. // https://html.spec.whatwg.org/multipage/parsing.html#reset-the-insertion-mode-appropriately
  2463. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  2464. {
  2465. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2466. bool last = i == 0;
  2467. // NOTE: When parsing fragments, we substitute the context element for the root of the stack of open elements.
  2468. RefPtr<DOM::Element> node;
  2469. if (last && m_parsing_fragment) {
  2470. node = m_context_element;
  2471. } else {
  2472. node = m_stack_of_open_elements.elements().at(i);
  2473. }
  2474. if (node->local_name() == HTML::TagNames::select) {
  2475. if (!last) {
  2476. for (ssize_t j = i; j > 0; --j) {
  2477. auto& ancestor = m_stack_of_open_elements.elements().at(j - 1);
  2478. if (is<HTMLTemplateElement>(ancestor))
  2479. break;
  2480. if (is<HTMLTableElement>(ancestor)) {
  2481. m_insertion_mode = InsertionMode::InSelectInTable;
  2482. return;
  2483. }
  2484. }
  2485. }
  2486. m_insertion_mode = InsertionMode::InSelect;
  2487. return;
  2488. }
  2489. if (!last && node->local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2490. m_insertion_mode = InsertionMode::InCell;
  2491. return;
  2492. }
  2493. if (node->local_name() == HTML::TagNames::tr) {
  2494. m_insertion_mode = InsertionMode::InRow;
  2495. return;
  2496. }
  2497. if (node->local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  2498. m_insertion_mode = InsertionMode::InTableBody;
  2499. return;
  2500. }
  2501. if (node->local_name() == HTML::TagNames::caption) {
  2502. m_insertion_mode = InsertionMode::InCaption;
  2503. return;
  2504. }
  2505. if (node->local_name() == HTML::TagNames::colgroup) {
  2506. m_insertion_mode = InsertionMode::InColumnGroup;
  2507. return;
  2508. }
  2509. if (node->local_name() == HTML::TagNames::table) {
  2510. m_insertion_mode = InsertionMode::InTable;
  2511. return;
  2512. }
  2513. if (node->local_name() == HTML::TagNames::template_) {
  2514. m_insertion_mode = m_stack_of_template_insertion_modes.last();
  2515. return;
  2516. }
  2517. if (!last && node->local_name() == HTML::TagNames::head) {
  2518. m_insertion_mode = InsertionMode::InHead;
  2519. return;
  2520. }
  2521. if (node->local_name() == HTML::TagNames::body) {
  2522. m_insertion_mode = InsertionMode::InBody;
  2523. return;
  2524. }
  2525. if (node->local_name() == HTML::TagNames::frameset) {
  2526. VERIFY(m_parsing_fragment);
  2527. m_insertion_mode = InsertionMode::InFrameset;
  2528. return;
  2529. }
  2530. if (node->local_name() == HTML::TagNames::html) {
  2531. if (!m_head_element) {
  2532. VERIFY(m_parsing_fragment);
  2533. m_insertion_mode = InsertionMode::BeforeHead;
  2534. return;
  2535. }
  2536. m_insertion_mode = InsertionMode::AfterHead;
  2537. return;
  2538. }
  2539. }
  2540. VERIFY(m_parsing_fragment);
  2541. m_insertion_mode = InsertionMode::InBody;
  2542. }
  2543. const char* HTMLDocumentParser::insertion_mode_name() const
  2544. {
  2545. switch (m_insertion_mode) {
  2546. #define __ENUMERATE_INSERTION_MODE(mode) \
  2547. case InsertionMode::mode: \
  2548. return #mode;
  2549. ENUMERATE_INSERTION_MODES
  2550. #undef __ENUMERATE_INSERTION_MODE
  2551. }
  2552. VERIFY_NOT_REACHED();
  2553. }
  2554. DOM::Document& HTMLDocumentParser::document()
  2555. {
  2556. return *m_document;
  2557. }
  2558. NonnullRefPtrVector<DOM::Node> HTMLDocumentParser::parse_html_fragment(DOM::Element& context_element, const StringView& markup)
  2559. {
  2560. auto temp_document = DOM::Document::create();
  2561. HTMLDocumentParser parser(*temp_document, markup, "utf-8");
  2562. parser.m_context_element = context_element;
  2563. parser.m_parsing_fragment = true;
  2564. parser.document().set_quirks_mode(context_element.document().mode());
  2565. if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
  2566. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  2567. } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
  2568. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2569. } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
  2570. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  2571. } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
  2572. if (context_element.document().is_scripting_enabled())
  2573. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2574. } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
  2575. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  2576. }
  2577. auto root = create_element(context_element.document(), HTML::TagNames::html, Namespace::HTML);
  2578. parser.document().append_child(root);
  2579. parser.m_stack_of_open_elements.push(root);
  2580. if (context_element.local_name() == HTML::TagNames::template_) {
  2581. parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  2582. }
  2583. // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
  2584. parser.reset_the_insertion_mode_appropriately();
  2585. for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
  2586. if (is<HTMLFormElement>(*form_candidate)) {
  2587. parser.m_form_element = verify_cast<HTMLFormElement>(*form_candidate);
  2588. break;
  2589. }
  2590. }
  2591. parser.run(context_element.document().url());
  2592. NonnullRefPtrVector<DOM::Node> children;
  2593. while (RefPtr<DOM::Node> child = root->first_child()) {
  2594. root->remove_child(*child);
  2595. context_element.document().adopt_node(*child);
  2596. children.append(*child);
  2597. }
  2598. return children;
  2599. }
  2600. NonnullOwnPtr<HTMLDocumentParser> HTMLDocumentParser::create_with_uncertain_encoding(DOM::Document& document, const ByteBuffer& input)
  2601. {
  2602. if (document.has_encoding())
  2603. return make<HTMLDocumentParser>(document, input, document.encoding().value());
  2604. auto encoding = run_encoding_sniffing_algorithm(input);
  2605. dbgln("The encoding sniffing algorithm returned encoding '{}'", encoding);
  2606. return make<HTMLDocumentParser>(document, input, encoding);
  2607. }
  2608. }