HTMLDocumentParser.cpp 107 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416141714181419142014211422142314241425142614271428142914301431143214331434143514361437143814391440144114421443144414451446144714481449145014511452145314541455145614571458145914601461146214631464146514661467146814691470147114721473147414751476147714781479148014811482148314841485148614871488148914901491149214931494149514961497149814991500150115021503150415051506150715081509151015111512151315141515151615171518151915201521152215231524152515261527152815291530153115321533153415351536153715381539154015411542154315441545154615471548154915501551155215531554155515561557155815591560156115621563156415651566156715681569157015711572157315741575157615771578157915801581158215831584158515861587158815891590159115921593159415951596159715981599160016011602160316041605160616071608160916101611161216131614161516161617161816191620162116221623162416251626162716281629163016311632163316341635163616371638163916401641164216431644164516461647164816491650165116521653165416551656165716581659166016611662166316641665166616671668166916701671167216731674167516761677167816791680168116821683168416851686168716881689169016911692169316941695169616971698169917001701170217031704170517061707170817091710171117121713171417151716171717181719172017211722172317241725172617271728172917301731173217331734173517361737173817391740174117421743174417451746174717481749175017511752175317541755175617571758175917601761176217631764176517661767176817691770177117721773177417751776177717781779178017811782178317841785178617871788178917901791179217931794179517961797179817991800180118021803180418051806180718081809181018111812181318141815181618171818181918201821182218231824182518261827182818291830183118321833183418351836183718381839184018411842184318441845184618471848184918501851185218531854185518561857185818591860186118621863186418651866186718681869187018711872187318741875187618771878187918801881188218831884188518861887188818891890189118921893189418951896189718981899190019011902190319041905190619071908190919101911191219131914191519161917191819191920192119221923192419251926192719281929193019311932193319341935193619371938193919401941194219431944194519461947194819491950195119521953195419551956195719581959196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019202020212022202320242025202620272028202920302031203220332034203520362037203820392040204120422043204420452046204720482049205020512052205320542055205620572058205920602061206220632064206520662067206820692070207120722073207420752076207720782079208020812082208320842085208620872088208920902091209220932094209520962097209820992100210121022103210421052106210721082109211021112112211321142115211621172118211921202121212221232124212521262127212821292130213121322133213421352136213721382139214021412142214321442145214621472148214921502151215221532154215521562157215821592160216121622163216421652166216721682169217021712172217321742175217621772178217921802181218221832184218521862187218821892190219121922193219421952196219721982199220022012202220322042205220622072208220922102211221222132214221522162217221822192220222122222223222422252226222722282229223022312232223322342235223622372238223922402241224222432244224522462247224822492250225122522253225422552256225722582259226022612262226322642265226622672268226922702271227222732274227522762277227822792280228122822283228422852286228722882289229022912292229322942295229622972298229923002301230223032304230523062307230823092310231123122313231423152316231723182319232023212322232323242325232623272328232923302331233223332334233523362337233823392340234123422343234423452346234723482349235023512352235323542355235623572358235923602361236223632364236523662367236823692370237123722373237423752376237723782379238023812382238323842385238623872388238923902391239223932394239523962397239823992400240124022403240424052406240724082409241024112412241324142415241624172418241924202421242224232424242524262427242824292430243124322433243424352436243724382439244024412442244324442445244624472448244924502451245224532454245524562457245824592460246124622463246424652466246724682469247024712472247324742475247624772478247924802481248224832484248524862487248824892490249124922493249424952496249724982499250025012502250325042505250625072508250925102511251225132514251525162517251825192520252125222523252425252526252725282529253025312532253325342535253625372538253925402541254225432544254525462547254825492550255125522553255425552556255725582559256025612562256325642565256625672568256925702571257225732574257525762577257825792580258125822583258425852586258725882589259025912592259325942595259625972598259926002601260226032604260526062607260826092610261126122613261426152616261726182619262026212622262326242625262626272628262926302631263226332634263526362637263826392640264126422643264426452646264726482649265026512652265326542655265626572658265926602661266226632664266526662667266826692670267126722673267426752676267726782679268026812682268326842685268626872688268926902691269226932694269526962697269826992700270127022703270427052706270727082709271027112712271327142715271627172718271927202721272227232724272527262727272827292730273127322733273427352736273727382739274027412742274327442745274627472748274927502751275227532754275527562757275827592760276127622763276427652766276727682769277027712772277327742775277627772778277927802781278227832784278527862787278827892790279127922793279427952796279727982799280028012802280328042805280628072808280928102811281228132814281528162817281828192820282128222823282428252826282728282829283028312832283328342835283628372838283928402841284228432844284528462847284828492850
  1. /*
  2. * Copyright (c) 2020, Andreas Kling <kling@serenityos.org>
  3. * All rights reserved.
  4. *
  5. * Redistribution and use in source and binary forms, with or without
  6. * modification, are permitted provided that the following conditions are met:
  7. *
  8. * 1. Redistributions of source code must retain the above copyright notice, this
  9. * list of conditions and the following disclaimer.
  10. *
  11. * 2. Redistributions in binary form must reproduce the above copyright notice,
  12. * this list of conditions and the following disclaimer in the documentation
  13. * and/or other materials provided with the distribution.
  14. *
  15. * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
  16. * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  17. * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
  18. * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
  19. * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  20. * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
  21. * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  22. * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
  23. * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  24. * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  25. */
  26. //#define PARSER_DEBUG
  27. #include <AK/Utf32View.h>
  28. #include <LibWeb/DOM/Comment.h>
  29. #include <LibWeb/DOM/Document.h>
  30. #include <LibWeb/DOM/DocumentType.h>
  31. #include <LibWeb/DOM/ElementFactory.h>
  32. #include <LibWeb/DOM/Event.h>
  33. #include <LibWeb/DOM/Text.h>
  34. #include <LibWeb/HTML/HTMLFormElement.h>
  35. #include <LibWeb/HTML/HTMLHeadElement.h>
  36. #include <LibWeb/HTML/HTMLScriptElement.h>
  37. #include <LibWeb/HTML/HTMLTemplateElement.h>
  38. #include <LibWeb/HTML/Parser/HTMLDocumentParser.h>
  39. #include <LibWeb/HTML/Parser/HTMLToken.h>
  40. namespace Web::HTML {
  41. #define PARSE_ERROR() \
  42. do { \
  43. dbg() << "Parse error! " << __PRETTY_FUNCTION__ << " @ " << __LINE__; \
  44. } while (0)
  45. static Vector<FlyString> s_quirks_public_ids = {
  46. "+//Silmaril//dtd html Pro v0r11 19970101//",
  47. "-//AS//DTD HTML 3.0 asWedit + extensions//",
  48. "-//AdvaSoft Ltd//DTD HTML 3.0 asWedit + extensions//",
  49. "-//IETF//DTD HTML 2.0 Level 1//",
  50. "-//IETF//DTD HTML 2.0 Level 2//",
  51. "-//IETF//DTD HTML 2.0 Strict Level 1//",
  52. "-//IETF//DTD HTML 2.0 Strict Level 2//",
  53. "-//IETF//DTD HTML 2.0 Strict//",
  54. "-//IETF//DTD HTML 2.0//",
  55. "-//IETF//DTD HTML 2.1E//",
  56. "-//IETF//DTD HTML 3.0//",
  57. "-//IETF//DTD HTML 3.2 Final//",
  58. "-//IETF//DTD HTML 3.2//",
  59. "-//IETF//DTD HTML 3//",
  60. "-//IETF//DTD HTML Level 0//",
  61. "-//IETF//DTD HTML Level 1//",
  62. "-//IETF//DTD HTML Level 2//",
  63. "-//IETF//DTD HTML Level 3//",
  64. "-//IETF//DTD HTML Strict Level 0//",
  65. "-//IETF//DTD HTML Strict Level 1//",
  66. "-//IETF//DTD HTML Strict Level 2//",
  67. "-//IETF//DTD HTML Strict Level 3//",
  68. "-//IETF//DTD HTML Strict//",
  69. "-//IETF//DTD HTML//",
  70. "-//Metrius//DTD Metrius Presentational//",
  71. "-//Microsoft//DTD Internet Explorer 2.0 HTML Strict//",
  72. "-//Microsoft//DTD Internet Explorer 2.0 HTML//",
  73. "-//Microsoft//DTD Internet Explorer 2.0 Tables//",
  74. "-//Microsoft//DTD Internet Explorer 3.0 HTML Strict//",
  75. "-//Microsoft//DTD Internet Explorer 3.0 HTML//",
  76. "-//Microsoft//DTD Internet Explorer 3.0 Tables//",
  77. "-//Netscape Comm. Corp.//DTD HTML//",
  78. "-//Netscape Comm. Corp.//DTD Strict HTML//",
  79. "-//O'Reilly and Associates//DTD HTML 2.0//",
  80. "-//O'Reilly and Associates//DTD HTML Extended 1.0//",
  81. "-//O'Reilly and Associates//DTD HTML Extended Relaxed 1.0//",
  82. "-//SQ//DTD HTML 2.0 HoTMetaL + extensions//",
  83. "-//SoftQuad Software//DTD HoTMetaL PRO 6.0::19990601::extensions to HTML 4.0//",
  84. "-//SoftQuad//DTD HoTMetaL PRO 4.0::19971010::extensions to HTML 4.0//",
  85. "-//Spyglass//DTD HTML 2.0 Extended//",
  86. "-//Sun Microsystems Corp.//DTD HotJava HTML//",
  87. "-//Sun Microsystems Corp.//DTD HotJava Strict HTML//",
  88. "-//W3C//DTD HTML 3 1995-03-24//",
  89. "-//W3C//DTD HTML 3.2 Draft//",
  90. "-//W3C//DTD HTML 3.2 Final//",
  91. "-//W3C//DTD HTML 3.2//",
  92. "-//W3C//DTD HTML 3.2S Draft//",
  93. "-//W3C//DTD HTML 4.0 Frameset//",
  94. "-//W3C//DTD HTML 4.0 Transitional//",
  95. "-//W3C//DTD HTML Experimental 19960712//",
  96. "-//W3C//DTD HTML Experimental 970421//",
  97. "-//W3C//DTD W3 HTML//",
  98. "-//W3O//DTD W3 HTML 3.0//",
  99. "-//WebTechs//DTD Mozilla HTML 2.0//",
  100. "-//WebTechs//DTD Mozilla HTML//"
  101. };
  102. RefPtr<DOM::Document> parse_html_document(const StringView& data, const URL& url, const String& encoding)
  103. {
  104. HTMLDocumentParser parser(data, encoding);
  105. parser.run(url);
  106. return parser.document();
  107. }
  108. HTMLDocumentParser::HTMLDocumentParser(const StringView& input, const String& encoding)
  109. : m_tokenizer(input, encoding)
  110. {
  111. m_document = adopt(*new DOM::Document);
  112. }
  113. HTMLDocumentParser::HTMLDocumentParser(const StringView& input, const String& encoding, DOM::Document& existing_document)
  114. : m_tokenizer(input, encoding)
  115. , m_document(existing_document)
  116. {
  117. }
  118. HTMLDocumentParser::~HTMLDocumentParser()
  119. {
  120. }
  121. void HTMLDocumentParser::run(const URL& url)
  122. {
  123. m_document->set_url(url);
  124. m_document->set_source(m_tokenizer.source());
  125. for (;;) {
  126. auto optional_token = m_tokenizer.next_token();
  127. if (!optional_token.has_value())
  128. break;
  129. auto& token = optional_token.value();
  130. #ifdef PARSER_DEBUG
  131. dbg() << "[" << insertion_mode_name() << "] " << token.to_string();
  132. #endif
  133. process_using_the_rules_for(m_insertion_mode, token);
  134. if (m_stop_parsing) {
  135. #ifdef PARSER_DEBUG
  136. dbg() << "Stop parsing" << (m_parsing_fragment ? " fragment" : "") << "! :^)";
  137. #endif
  138. break;
  139. }
  140. }
  141. flush_character_insertions();
  142. // "The end"
  143. m_document->set_ready_state("interactive");
  144. auto scripts_to_execute_when_parsing_has_finished = m_document->take_scripts_to_execute_when_parsing_has_finished({});
  145. for (auto& script : scripts_to_execute_when_parsing_has_finished) {
  146. script.execute_script();
  147. }
  148. m_document->dispatch_event(DOM::Event::create("DOMContentLoaded"));
  149. auto scripts_to_execute_as_soon_as_possible = m_document->take_scripts_to_execute_as_soon_as_possible({});
  150. for (auto& script : scripts_to_execute_as_soon_as_possible) {
  151. script.execute_script();
  152. }
  153. m_document->set_ready_state("complete");
  154. }
  155. void HTMLDocumentParser::process_using_the_rules_for(InsertionMode mode, HTMLToken& token)
  156. {
  157. switch (mode) {
  158. case InsertionMode::Initial:
  159. handle_initial(token);
  160. break;
  161. case InsertionMode::BeforeHTML:
  162. handle_before_html(token);
  163. break;
  164. case InsertionMode::BeforeHead:
  165. handle_before_head(token);
  166. break;
  167. case InsertionMode::InHead:
  168. handle_in_head(token);
  169. break;
  170. case InsertionMode::InHeadNoscript:
  171. handle_in_head_noscript(token);
  172. break;
  173. case InsertionMode::AfterHead:
  174. handle_after_head(token);
  175. break;
  176. case InsertionMode::InBody:
  177. handle_in_body(token);
  178. break;
  179. case InsertionMode::AfterBody:
  180. handle_after_body(token);
  181. break;
  182. case InsertionMode::AfterAfterBody:
  183. handle_after_after_body(token);
  184. break;
  185. case InsertionMode::Text:
  186. handle_text(token);
  187. break;
  188. case InsertionMode::InTable:
  189. handle_in_table(token);
  190. break;
  191. case InsertionMode::InTableBody:
  192. handle_in_table_body(token);
  193. break;
  194. case InsertionMode::InRow:
  195. handle_in_row(token);
  196. break;
  197. case InsertionMode::InCell:
  198. handle_in_cell(token);
  199. break;
  200. case InsertionMode::InTableText:
  201. handle_in_table_text(token);
  202. break;
  203. case InsertionMode::InSelectInTable:
  204. handle_in_select_in_table(token);
  205. break;
  206. case InsertionMode::InSelect:
  207. handle_in_select(token);
  208. break;
  209. case InsertionMode::InCaption:
  210. handle_in_caption(token);
  211. break;
  212. case InsertionMode::InColumnGroup:
  213. handle_in_column_group(token);
  214. break;
  215. case InsertionMode::InTemplate:
  216. handle_in_template(token);
  217. break;
  218. case InsertionMode::InFrameset:
  219. handle_in_frameset(token);
  220. break;
  221. case InsertionMode::AfterFrameset:
  222. handle_after_frameset(token);
  223. break;
  224. case InsertionMode::AfterAfterFrameset:
  225. handle_after_after_frameset(token);
  226. break;
  227. default:
  228. ASSERT_NOT_REACHED();
  229. }
  230. }
  231. DOM::QuirksMode HTMLDocumentParser::which_quirks_mode(const HTMLToken& doctype_token) const
  232. {
  233. if (doctype_token.m_doctype.force_quirks)
  234. return DOM::QuirksMode::Yes;
  235. // NOTE: The tokenizer puts the name into lower case for us.
  236. if (doctype_token.m_doctype.name.to_string() != "html")
  237. return DOM::QuirksMode::Yes;
  238. auto public_identifier = doctype_token.m_doctype.public_identifier.to_string();
  239. auto system_identifier = doctype_token.m_doctype.system_identifier.to_string();
  240. if (public_identifier.equals_ignoring_case("-//W3O//DTD W3 HTML Strict 3.0//EN//"))
  241. return DOM::QuirksMode::Yes;
  242. if (public_identifier.equals_ignoring_case("-/W3C/DTD HTML 4.0 Transitional/EN"))
  243. return DOM::QuirksMode::Yes;
  244. if (public_identifier.equals_ignoring_case("HTML"))
  245. return DOM::QuirksMode::Yes;
  246. if (system_identifier.equals_ignoring_case("http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd"))
  247. return DOM::QuirksMode::Yes;
  248. for (auto& public_id : s_quirks_public_ids) {
  249. if (public_identifier.starts_with(public_id, CaseSensitivity::CaseInsensitive))
  250. return DOM::QuirksMode::Yes;
  251. }
  252. if (doctype_token.m_doctype.missing_system_identifier) {
  253. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  254. return DOM::QuirksMode::Yes;
  255. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  256. return DOM::QuirksMode::Yes;
  257. }
  258. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Frameset//", CaseSensitivity::CaseInsensitive))
  259. return DOM::QuirksMode::Limited;
  260. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Transitional//", CaseSensitivity::CaseInsensitive))
  261. return DOM::QuirksMode::Limited;
  262. if (!doctype_token.m_doctype.missing_system_identifier) {
  263. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  264. return DOM::QuirksMode::Limited;
  265. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  266. return DOM::QuirksMode::Limited;
  267. }
  268. return DOM::QuirksMode::No;
  269. }
  270. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  271. {
  272. if (token.is_character() && token.is_parser_whitespace()) {
  273. return;
  274. }
  275. if (token.is_comment()) {
  276. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  277. document().append_child(move(comment));
  278. return;
  279. }
  280. if (token.is_doctype()) {
  281. auto doctype = adopt(*new DOM::DocumentType(document()));
  282. doctype->set_name(token.m_doctype.name.to_string());
  283. doctype->set_public_id(token.m_doctype.public_identifier.to_string());
  284. doctype->set_system_id(token.m_doctype.system_identifier.to_string());
  285. document().append_child(move(doctype));
  286. document().set_quirks_mode(which_quirks_mode(token));
  287. m_insertion_mode = InsertionMode::BeforeHTML;
  288. return;
  289. }
  290. PARSE_ERROR();
  291. document().set_quirks_mode(DOM::QuirksMode::Yes);
  292. m_insertion_mode = InsertionMode::BeforeHTML;
  293. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  294. }
  295. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  296. {
  297. if (token.is_doctype()) {
  298. PARSE_ERROR();
  299. return;
  300. }
  301. if (token.is_comment()) {
  302. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  303. document().append_child(move(comment));
  304. return;
  305. }
  306. if (token.is_character() && token.is_parser_whitespace()) {
  307. return;
  308. }
  309. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  310. auto element = create_element_for(token);
  311. document().append_child(element);
  312. m_stack_of_open_elements.push(move(element));
  313. m_insertion_mode = InsertionMode::BeforeHead;
  314. return;
  315. }
  316. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  317. goto AnythingElse;
  318. }
  319. if (token.is_end_tag()) {
  320. PARSE_ERROR();
  321. return;
  322. }
  323. AnythingElse:
  324. auto element = create_element(document(), HTML::TagNames::html);
  325. document().append_child(element);
  326. m_stack_of_open_elements.push(element);
  327. // 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.
  328. m_insertion_mode = InsertionMode::BeforeHead;
  329. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  330. return;
  331. }
  332. DOM::Element& HTMLDocumentParser::current_node()
  333. {
  334. return m_stack_of_open_elements.current_node();
  335. }
  336. DOM::Element& HTMLDocumentParser::node_before_current_node()
  337. {
  338. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  339. }
  340. HTMLDocumentParser::AdjustedInsertionLocation HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  341. {
  342. auto& target = current_node();
  343. HTMLDocumentParser::AdjustedInsertionLocation adjusted_insertion_location;
  344. 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)) {
  345. auto last_template = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::template_);
  346. auto last_table = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::table);
  347. if (last_template.element && (!last_table.element || last_template.index > last_table.index)) {
  348. // This returns the template content, so no need to check the parent is a template.
  349. return { downcast<HTMLTemplateElement>(last_template.element)->content(), nullptr };
  350. }
  351. if (!last_table.element) {
  352. ASSERT(m_parsing_fragment);
  353. // Guaranteed not to be a template element (it will be the html element),
  354. // so no need to check the parent is a template.
  355. return { m_stack_of_open_elements.elements().first(), nullptr };
  356. }
  357. if (last_table.element->parent_node())
  358. adjusted_insertion_location = { last_table.element->parent_node(), last_table.element };
  359. else
  360. adjusted_insertion_location = { m_stack_of_open_elements.element_before(*last_table.element), nullptr };
  361. } else {
  362. adjusted_insertion_location = { target, nullptr };
  363. }
  364. if (is<HTMLTemplateElement>(*adjusted_insertion_location.parent))
  365. return { downcast<HTMLTemplateElement>(*adjusted_insertion_location.parent).content(), nullptr };
  366. return adjusted_insertion_location;
  367. }
  368. NonnullRefPtr<DOM::Element> HTMLDocumentParser::create_element_for(const HTMLToken& token)
  369. {
  370. auto element = create_element(document(), token.tag_name());
  371. for (auto& attribute : token.m_tag.attributes) {
  372. element->set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  373. }
  374. return element;
  375. }
  376. RefPtr<DOM::Element> HTMLDocumentParser::insert_html_element(const HTMLToken& token)
  377. {
  378. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  379. auto element = create_element_for(token);
  380. // FIXME: Check if it's possible to insert `element` at `adjusted_insertion_location`
  381. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling);
  382. m_stack_of_open_elements.push(element);
  383. return element;
  384. }
  385. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  386. {
  387. if (token.is_character() && token.is_parser_whitespace()) {
  388. return;
  389. }
  390. if (token.is_comment()) {
  391. insert_comment(token);
  392. return;
  393. }
  394. if (token.is_doctype()) {
  395. PARSE_ERROR();
  396. return;
  397. }
  398. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  399. process_using_the_rules_for(InsertionMode::InBody, token);
  400. return;
  401. }
  402. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  403. auto element = insert_html_element(token);
  404. m_head_element = downcast<HTMLHeadElement>(*element);
  405. m_insertion_mode = InsertionMode::InHead;
  406. return;
  407. }
  408. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  409. goto AnythingElse;
  410. }
  411. if (token.is_end_tag()) {
  412. PARSE_ERROR();
  413. return;
  414. }
  415. AnythingElse:
  416. m_head_element = downcast<HTMLHeadElement>(*insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::head)));
  417. m_insertion_mode = InsertionMode::InHead;
  418. process_using_the_rules_for(InsertionMode::InHead, token);
  419. return;
  420. }
  421. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  422. {
  423. auto data = token.m_comment_or_character.data.to_string();
  424. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  425. adjusted_insertion_location.parent->insert_before(adopt(*new DOM::Comment(document(), data)), adjusted_insertion_location.insert_before_sibling);
  426. }
  427. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  428. {
  429. if (token.is_parser_whitespace()) {
  430. insert_character(token.code_point());
  431. return;
  432. }
  433. if (token.is_comment()) {
  434. insert_comment(token);
  435. return;
  436. }
  437. if (token.is_doctype()) {
  438. PARSE_ERROR();
  439. return;
  440. }
  441. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  442. process_using_the_rules_for(InsertionMode::InBody, token);
  443. return;
  444. }
  445. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  446. insert_html_element(token);
  447. m_stack_of_open_elements.pop();
  448. token.acknowledge_self_closing_flag_if_set();
  449. return;
  450. }
  451. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  452. auto element = insert_html_element(token);
  453. m_stack_of_open_elements.pop();
  454. token.acknowledge_self_closing_flag_if_set();
  455. return;
  456. }
  457. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  458. insert_html_element(token);
  459. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  460. m_original_insertion_mode = m_insertion_mode;
  461. m_insertion_mode = InsertionMode::Text;
  462. return;
  463. }
  464. 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)) {
  465. parse_generic_raw_text_element(token);
  466. return;
  467. }
  468. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noscript && !m_scripting_enabled) {
  469. insert_html_element(token);
  470. m_insertion_mode = InsertionMode::InHeadNoscript;
  471. return;
  472. }
  473. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  474. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  475. auto element = create_element_for(token);
  476. auto& script_element = downcast<HTMLScriptElement>(*element);
  477. script_element.set_parser_document({}, document());
  478. script_element.set_non_blocking({}, false);
  479. if (m_parsing_fragment) {
  480. TODO();
  481. }
  482. if (m_invoked_via_document_write) {
  483. TODO();
  484. }
  485. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling, false);
  486. m_stack_of_open_elements.push(element);
  487. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  488. m_original_insertion_mode = m_insertion_mode;
  489. m_insertion_mode = InsertionMode::Text;
  490. return;
  491. }
  492. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  493. m_stack_of_open_elements.pop();
  494. m_insertion_mode = InsertionMode::AfterHead;
  495. return;
  496. }
  497. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  498. goto AnythingElse;
  499. }
  500. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  501. insert_html_element(token);
  502. m_list_of_active_formatting_elements.add_marker();
  503. m_frameset_ok = false;
  504. m_insertion_mode = InsertionMode::InTemplate;
  505. m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  506. return;
  507. }
  508. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  509. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  510. PARSE_ERROR();
  511. return;
  512. }
  513. generate_all_implied_end_tags_thoroughly();
  514. if (current_node().local_name() != HTML::TagNames::template_)
  515. PARSE_ERROR();
  516. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  517. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  518. m_stack_of_template_insertion_modes.take_last();
  519. reset_the_insertion_mode_appropriately();
  520. return;
  521. }
  522. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  523. PARSE_ERROR();
  524. return;
  525. }
  526. AnythingElse:
  527. m_stack_of_open_elements.pop();
  528. m_insertion_mode = InsertionMode::AfterHead;
  529. process_using_the_rules_for(m_insertion_mode, token);
  530. }
  531. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken& token)
  532. {
  533. if (token.is_doctype()) {
  534. PARSE_ERROR();
  535. return;
  536. }
  537. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  538. process_using_the_rules_for(InsertionMode::InBody, token);
  539. return;
  540. }
  541. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::noscript) {
  542. m_stack_of_open_elements.pop();
  543. m_insertion_mode = InsertionMode::InHead;
  544. return;
  545. }
  546. 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))) {
  547. process_using_the_rules_for(InsertionMode::InHead, token);
  548. return;
  549. }
  550. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  551. goto AnythingElse;
  552. }
  553. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::noscript)) {
  554. PARSE_ERROR();
  555. return;
  556. }
  557. AnythingElse:
  558. PARSE_ERROR();
  559. m_stack_of_open_elements.pop();
  560. m_insertion_mode = InsertionMode::InHead;
  561. process_using_the_rules_for(m_insertion_mode, token);
  562. }
  563. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  564. {
  565. insert_html_element(token);
  566. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  567. m_original_insertion_mode = m_insertion_mode;
  568. m_insertion_mode = InsertionMode::Text;
  569. }
  570. DOM::Text* HTMLDocumentParser::find_character_insertion_node()
  571. {
  572. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  573. if (adjusted_insertion_location.insert_before_sibling) {
  574. TODO();
  575. }
  576. if (adjusted_insertion_location.parent->is_document())
  577. return nullptr;
  578. if (adjusted_insertion_location.parent->last_child() && adjusted_insertion_location.parent->last_child()->is_text())
  579. return downcast<DOM::Text>(adjusted_insertion_location.parent->last_child());
  580. auto new_text_node = adopt(*new DOM::Text(document(), ""));
  581. adjusted_insertion_location.parent->append_child(new_text_node);
  582. return new_text_node;
  583. }
  584. void HTMLDocumentParser::flush_character_insertions()
  585. {
  586. if (m_character_insertion_builder.is_empty())
  587. return;
  588. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  589. m_character_insertion_node->parent()->children_changed();
  590. m_character_insertion_builder.clear();
  591. }
  592. void HTMLDocumentParser::insert_character(u32 data)
  593. {
  594. auto node = find_character_insertion_node();
  595. if (node == m_character_insertion_node) {
  596. m_character_insertion_builder.append(Utf32View { &data, 1 });
  597. return;
  598. }
  599. if (!m_character_insertion_node) {
  600. m_character_insertion_node = node;
  601. m_character_insertion_builder.append(Utf32View { &data, 1 });
  602. return;
  603. }
  604. flush_character_insertions();
  605. m_character_insertion_node = node;
  606. m_character_insertion_builder.append(Utf32View { &data, 1 });
  607. }
  608. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  609. {
  610. if (token.is_character() && token.is_parser_whitespace()) {
  611. insert_character(token.code_point());
  612. return;
  613. }
  614. if (token.is_comment()) {
  615. insert_comment(token);
  616. return;
  617. }
  618. if (token.is_doctype()) {
  619. PARSE_ERROR();
  620. return;
  621. }
  622. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  623. process_using_the_rules_for(InsertionMode::InBody, token);
  624. return;
  625. }
  626. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  627. insert_html_element(token);
  628. m_frameset_ok = false;
  629. m_insertion_mode = InsertionMode::InBody;
  630. return;
  631. }
  632. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  633. insert_html_element(token);
  634. m_insertion_mode = InsertionMode::InFrameset;
  635. return;
  636. }
  637. 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)) {
  638. PARSE_ERROR();
  639. m_stack_of_open_elements.push(*m_head_element);
  640. process_using_the_rules_for(InsertionMode::InHead, token);
  641. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  642. return entry.ptr() == m_head_element;
  643. });
  644. return;
  645. }
  646. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  647. process_using_the_rules_for(InsertionMode::InHead, token);
  648. return;
  649. }
  650. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  651. goto AnythingElse;
  652. }
  653. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  654. PARSE_ERROR();
  655. return;
  656. }
  657. AnythingElse:
  658. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::body));
  659. m_insertion_mode = InsertionMode::InBody;
  660. process_using_the_rules_for(m_insertion_mode, token);
  661. }
  662. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  663. {
  664. 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))
  665. m_stack_of_open_elements.pop();
  666. }
  667. void HTMLDocumentParser::generate_all_implied_end_tags_thoroughly()
  668. {
  669. 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))
  670. m_stack_of_open_elements.pop();
  671. }
  672. void HTMLDocumentParser::close_a_p_element()
  673. {
  674. generate_implied_end_tags(HTML::TagNames::p);
  675. if (current_node().local_name() != HTML::TagNames::p) {
  676. PARSE_ERROR();
  677. }
  678. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  679. }
  680. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  681. {
  682. if (token.is_character() && token.is_parser_whitespace()) {
  683. process_using_the_rules_for(InsertionMode::InBody, token);
  684. return;
  685. }
  686. if (token.is_comment()) {
  687. auto data = token.m_comment_or_character.data.to_string();
  688. auto& insertion_location = m_stack_of_open_elements.first();
  689. insertion_location.append_child(adopt(*new DOM::Comment(document(), data)));
  690. return;
  691. }
  692. if (token.is_doctype()) {
  693. PARSE_ERROR();
  694. return;
  695. }
  696. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  697. process_using_the_rules_for(InsertionMode::InBody, token);
  698. return;
  699. }
  700. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  701. if (m_parsing_fragment) {
  702. PARSE_ERROR();
  703. return;
  704. }
  705. m_insertion_mode = InsertionMode::AfterAfterBody;
  706. return;
  707. }
  708. if (token.is_end_of_file()) {
  709. stop_parsing();
  710. return;
  711. }
  712. PARSE_ERROR();
  713. m_insertion_mode = InsertionMode::InBody;
  714. process_using_the_rules_for(InsertionMode::InBody, token);
  715. }
  716. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  717. {
  718. if (token.is_comment()) {
  719. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  720. document().append_child(move(comment));
  721. return;
  722. }
  723. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  724. process_using_the_rules_for(InsertionMode::InBody, token);
  725. return;
  726. }
  727. if (token.is_end_of_file()) {
  728. stop_parsing();
  729. return;
  730. }
  731. PARSE_ERROR();
  732. m_insertion_mode = InsertionMode::InBody;
  733. process_using_the_rules_for(m_insertion_mode, token);
  734. }
  735. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  736. {
  737. // FIXME: This needs to care about "markers"
  738. if (m_list_of_active_formatting_elements.is_empty())
  739. return;
  740. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  741. return;
  742. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  743. return;
  744. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  745. RefPtr<DOM::Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  746. ASSERT(entry);
  747. Rewind:
  748. if (index == 0) {
  749. goto Create;
  750. }
  751. --index;
  752. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  753. ASSERT(entry);
  754. if (!m_stack_of_open_elements.contains(*entry))
  755. goto Rewind;
  756. Advance:
  757. ++index;
  758. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  759. ASSERT(entry);
  760. Create:
  761. // FIXME: Hold on to the real token!
  762. auto new_element = insert_html_element(HTMLToken::make_start_tag(entry->local_name()));
  763. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  764. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  765. goto Advance;
  766. }
  767. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  768. {
  769. auto subject = token.tag_name();
  770. // If the current node is an HTML element whose tag name is subject,
  771. // and the current node is not in the list of active formatting elements,
  772. // then pop the current node off the stack of open elements, and return.
  773. if (current_node().local_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  774. m_stack_of_open_elements.pop();
  775. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  776. }
  777. size_t outer_loop_counter = 0;
  778. //OuterLoop:
  779. if (outer_loop_counter >= 8)
  780. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  781. ++outer_loop_counter;
  782. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  783. if (!formatting_element)
  784. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  785. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  786. PARSE_ERROR();
  787. // FIXME: If formatting element is not in the stack of open elements,
  788. // then this is a parse error; remove the element from the list, and return.
  789. TODO();
  790. }
  791. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  792. PARSE_ERROR();
  793. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  794. }
  795. if (formatting_element != &current_node()) {
  796. PARSE_ERROR();
  797. }
  798. RefPtr<DOM::Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  799. if (!furthest_block) {
  800. while (&current_node() != formatting_element)
  801. m_stack_of_open_elements.pop();
  802. m_stack_of_open_elements.pop();
  803. m_list_of_active_formatting_elements.remove(*formatting_element);
  804. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  805. }
  806. // FIXME: Implement the rest of the AAA :^)
  807. TODO();
  808. }
  809. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name)
  810. {
  811. return tag_name.is_one_of(
  812. HTML::TagNames::address,
  813. HTML::TagNames::applet,
  814. HTML::TagNames::area,
  815. HTML::TagNames::article,
  816. HTML::TagNames::aside,
  817. HTML::TagNames::base,
  818. HTML::TagNames::basefont,
  819. HTML::TagNames::bgsound,
  820. HTML::TagNames::blockquote,
  821. HTML::TagNames::body,
  822. HTML::TagNames::br,
  823. HTML::TagNames::button,
  824. HTML::TagNames::caption,
  825. HTML::TagNames::center,
  826. HTML::TagNames::col,
  827. HTML::TagNames::colgroup,
  828. HTML::TagNames::dd,
  829. HTML::TagNames::details,
  830. HTML::TagNames::dir,
  831. HTML::TagNames::div,
  832. HTML::TagNames::dl,
  833. HTML::TagNames::dt,
  834. HTML::TagNames::embed,
  835. HTML::TagNames::fieldset,
  836. HTML::TagNames::figcaption,
  837. HTML::TagNames::figure,
  838. HTML::TagNames::footer,
  839. HTML::TagNames::form,
  840. HTML::TagNames::frame,
  841. HTML::TagNames::frameset,
  842. HTML::TagNames::h1,
  843. HTML::TagNames::h2,
  844. HTML::TagNames::h3,
  845. HTML::TagNames::h4,
  846. HTML::TagNames::h5,
  847. HTML::TagNames::h6,
  848. HTML::TagNames::head,
  849. HTML::TagNames::header,
  850. HTML::TagNames::hgroup,
  851. HTML::TagNames::hr,
  852. HTML::TagNames::html,
  853. HTML::TagNames::iframe,
  854. HTML::TagNames::img,
  855. HTML::TagNames::input,
  856. HTML::TagNames::keygen,
  857. HTML::TagNames::li,
  858. HTML::TagNames::link,
  859. HTML::TagNames::listing,
  860. HTML::TagNames::main,
  861. HTML::TagNames::marquee,
  862. HTML::TagNames::menu,
  863. HTML::TagNames::meta,
  864. HTML::TagNames::nav,
  865. HTML::TagNames::noembed,
  866. HTML::TagNames::noframes,
  867. HTML::TagNames::noscript,
  868. HTML::TagNames::object,
  869. HTML::TagNames::ol,
  870. HTML::TagNames::p,
  871. HTML::TagNames::param,
  872. HTML::TagNames::plaintext,
  873. HTML::TagNames::pre,
  874. HTML::TagNames::script,
  875. HTML::TagNames::section,
  876. HTML::TagNames::select,
  877. HTML::TagNames::source,
  878. HTML::TagNames::style,
  879. HTML::TagNames::summary,
  880. HTML::TagNames::table,
  881. HTML::TagNames::tbody,
  882. HTML::TagNames::td,
  883. HTML::TagNames::template_,
  884. HTML::TagNames::textarea,
  885. HTML::TagNames::tfoot,
  886. HTML::TagNames::th,
  887. HTML::TagNames::thead,
  888. HTML::TagNames::title,
  889. HTML::TagNames::tr,
  890. HTML::TagNames::track,
  891. HTML::TagNames::ul,
  892. HTML::TagNames::wbr,
  893. HTML::TagNames::xmp);
  894. }
  895. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  896. {
  897. if (token.is_character()) {
  898. if (token.code_point() == 0) {
  899. PARSE_ERROR();
  900. return;
  901. }
  902. if (token.is_parser_whitespace()) {
  903. reconstruct_the_active_formatting_elements();
  904. insert_character(token.code_point());
  905. return;
  906. }
  907. reconstruct_the_active_formatting_elements();
  908. insert_character(token.code_point());
  909. m_frameset_ok = false;
  910. return;
  911. }
  912. if (token.is_comment()) {
  913. insert_comment(token);
  914. return;
  915. }
  916. if (token.is_doctype()) {
  917. PARSE_ERROR();
  918. return;
  919. }
  920. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  921. PARSE_ERROR();
  922. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  923. return;
  924. for (auto& attribute : token.m_tag.attributes) {
  925. if (current_node().has_attribute(attribute.local_name_builder.string_view()))
  926. continue;
  927. current_node().set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  928. }
  929. return;
  930. }
  931. 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)) {
  932. process_using_the_rules_for(InsertionMode::InHead, token);
  933. return;
  934. }
  935. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  936. process_using_the_rules_for(InsertionMode::InHead, token);
  937. return;
  938. }
  939. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  940. PARSE_ERROR();
  941. if (m_stack_of_open_elements.elements().size() == 1
  942. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body
  943. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  944. ASSERT(m_parsing_fragment);
  945. return;
  946. }
  947. m_frameset_ok = false;
  948. auto& body_element = m_stack_of_open_elements.elements().at(1);
  949. for (auto& attribute : token.m_tag.attributes) {
  950. if (body_element.has_attribute(attribute.local_name_builder.string_view()))
  951. continue;
  952. body_element.set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  953. }
  954. return;
  955. }
  956. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  957. PARSE_ERROR();
  958. if (m_stack_of_open_elements.elements().size() == 1
  959. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body) {
  960. ASSERT(m_parsing_fragment);
  961. return;
  962. }
  963. if (!m_frameset_ok)
  964. return;
  965. TODO();
  966. }
  967. if (token.is_end_of_file()) {
  968. if (!m_stack_of_template_insertion_modes.is_empty()) {
  969. process_using_the_rules_for(InsertionMode::InTemplate, token);
  970. return;
  971. }
  972. for (auto& node : m_stack_of_open_elements.elements()) {
  973. 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)) {
  974. PARSE_ERROR();
  975. break;
  976. }
  977. }
  978. stop_parsing();
  979. return;
  980. }
  981. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  982. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  983. PARSE_ERROR();
  984. return;
  985. }
  986. for (auto& node : m_stack_of_open_elements.elements()) {
  987. 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)) {
  988. PARSE_ERROR();
  989. break;
  990. }
  991. }
  992. m_insertion_mode = InsertionMode::AfterBody;
  993. return;
  994. }
  995. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  996. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  997. PARSE_ERROR();
  998. return;
  999. }
  1000. for (auto& node : m_stack_of_open_elements.elements()) {
  1001. 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)) {
  1002. PARSE_ERROR();
  1003. break;
  1004. }
  1005. }
  1006. m_insertion_mode = InsertionMode::AfterBody;
  1007. process_using_the_rules_for(m_insertion_mode, token);
  1008. return;
  1009. }
  1010. 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)) {
  1011. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1012. close_a_p_element();
  1013. insert_html_element(token);
  1014. return;
  1015. }
  1016. 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)) {
  1017. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1018. close_a_p_element();
  1019. 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)) {
  1020. PARSE_ERROR();
  1021. m_stack_of_open_elements.pop();
  1022. }
  1023. insert_html_element(token);
  1024. return;
  1025. }
  1026. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  1027. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1028. close_a_p_element();
  1029. insert_html_element(token);
  1030. m_frameset_ok = false;
  1031. // If the next token is a U+000A LINE FEED (LF) character token,
  1032. // then ignore that token and move on to the next one.
  1033. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1034. auto next_token = m_tokenizer.next_token();
  1035. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1036. // Ignore it.
  1037. } else {
  1038. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1039. }
  1040. return;
  1041. }
  1042. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1043. if (m_form_element && !m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1044. PARSE_ERROR();
  1045. return;
  1046. }
  1047. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1048. close_a_p_element();
  1049. auto element = insert_html_element(token);
  1050. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  1051. m_form_element = downcast<HTMLFormElement>(*element);
  1052. return;
  1053. }
  1054. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  1055. m_frameset_ok = false;
  1056. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1057. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1058. if (node->local_name() == HTML::TagNames::li) {
  1059. generate_implied_end_tags(HTML::TagNames::li);
  1060. if (current_node().local_name() != HTML::TagNames::li) {
  1061. PARSE_ERROR();
  1062. }
  1063. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1064. break;
  1065. }
  1066. if (is_special_tag(node->local_name()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1067. break;
  1068. }
  1069. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1070. close_a_p_element();
  1071. insert_html_element(token);
  1072. return;
  1073. }
  1074. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1075. m_frameset_ok = false;
  1076. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1077. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1078. if (node->local_name() == HTML::TagNames::dd) {
  1079. generate_implied_end_tags(HTML::TagNames::dd);
  1080. if (current_node().local_name() != HTML::TagNames::dd) {
  1081. PARSE_ERROR();
  1082. }
  1083. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  1084. break;
  1085. }
  1086. if (node->local_name() == HTML::TagNames::dt) {
  1087. generate_implied_end_tags(HTML::TagNames::dt);
  1088. if (current_node().local_name() != HTML::TagNames::dt) {
  1089. PARSE_ERROR();
  1090. }
  1091. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  1092. break;
  1093. }
  1094. if (is_special_tag(node->local_name()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1095. break;
  1096. }
  1097. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1098. close_a_p_element();
  1099. insert_html_element(token);
  1100. return;
  1101. }
  1102. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  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. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  1107. return;
  1108. }
  1109. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  1110. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  1111. PARSE_ERROR();
  1112. generate_implied_end_tags();
  1113. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  1114. }
  1115. reconstruct_the_active_formatting_elements();
  1116. insert_html_element(token);
  1117. m_frameset_ok = false;
  1118. return;
  1119. }
  1120. 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)) {
  1121. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1122. PARSE_ERROR();
  1123. return;
  1124. }
  1125. generate_implied_end_tags();
  1126. if (current_node().local_name() != token.tag_name()) {
  1127. PARSE_ERROR();
  1128. }
  1129. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1130. return;
  1131. }
  1132. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  1133. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1134. auto node = m_form_element;
  1135. m_form_element = nullptr;
  1136. if (!node || !m_stack_of_open_elements.has_in_scope(*node)) {
  1137. PARSE_ERROR();
  1138. return;
  1139. }
  1140. generate_implied_end_tags();
  1141. if (&current_node() != node) {
  1142. PARSE_ERROR();
  1143. }
  1144. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  1145. } else {
  1146. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  1147. PARSE_ERROR();
  1148. return;
  1149. }
  1150. generate_implied_end_tags();
  1151. if (current_node().local_name() != HTML::TagNames::form) {
  1152. PARSE_ERROR();
  1153. }
  1154. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  1155. }
  1156. return;
  1157. }
  1158. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  1159. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1160. PARSE_ERROR();
  1161. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::p));
  1162. }
  1163. close_a_p_element();
  1164. return;
  1165. }
  1166. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  1167. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  1168. PARSE_ERROR();
  1169. return;
  1170. }
  1171. generate_implied_end_tags(HTML::TagNames::li);
  1172. if (current_node().local_name() != HTML::TagNames::li) {
  1173. PARSE_ERROR();
  1174. dbg() << "Expected <li> current node, but had <" << current_node().local_name() << ">";
  1175. }
  1176. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1177. return;
  1178. }
  1179. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1180. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1181. PARSE_ERROR();
  1182. return;
  1183. }
  1184. generate_implied_end_tags(token.tag_name());
  1185. if (current_node().local_name() != token.tag_name()) {
  1186. PARSE_ERROR();
  1187. }
  1188. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1189. return;
  1190. }
  1191. 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)) {
  1192. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  1193. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  1194. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  1195. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  1196. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  1197. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  1198. PARSE_ERROR();
  1199. return;
  1200. }
  1201. generate_implied_end_tags();
  1202. if (current_node().local_name() != token.tag_name()) {
  1203. PARSE_ERROR();
  1204. }
  1205. for (;;) {
  1206. auto popped_element = m_stack_of_open_elements.pop();
  1207. 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))
  1208. break;
  1209. }
  1210. return;
  1211. }
  1212. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1213. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1214. PARSE_ERROR();
  1215. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1216. goto AnyOtherEndTag;
  1217. m_list_of_active_formatting_elements.remove(*element);
  1218. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1219. return entry.ptr() == element;
  1220. });
  1221. }
  1222. reconstruct_the_active_formatting_elements();
  1223. auto element = insert_html_element(token);
  1224. m_list_of_active_formatting_elements.add(*element);
  1225. return;
  1226. }
  1227. 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)) {
  1228. reconstruct_the_active_formatting_elements();
  1229. auto element = insert_html_element(token);
  1230. m_list_of_active_formatting_elements.add(*element);
  1231. return;
  1232. }
  1233. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1234. reconstruct_the_active_formatting_elements();
  1235. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1236. PARSE_ERROR();
  1237. run_the_adoption_agency_algorithm(token);
  1238. reconstruct_the_active_formatting_elements();
  1239. }
  1240. auto element = insert_html_element(token);
  1241. m_list_of_active_formatting_elements.add(*element);
  1242. return;
  1243. }
  1244. 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)) {
  1245. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1246. goto AnyOtherEndTag;
  1247. return;
  1248. }
  1249. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1250. reconstruct_the_active_formatting_elements();
  1251. insert_html_element(token);
  1252. m_list_of_active_formatting_elements.add_marker();
  1253. m_frameset_ok = false;
  1254. return;
  1255. }
  1256. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1257. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1258. PARSE_ERROR();
  1259. return;
  1260. }
  1261. generate_implied_end_tags();
  1262. if (current_node().local_name() != token.tag_name()) {
  1263. PARSE_ERROR();
  1264. }
  1265. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1266. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1267. return;
  1268. }
  1269. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1270. if (!document().in_quirks_mode()) {
  1271. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1272. close_a_p_element();
  1273. }
  1274. insert_html_element(token);
  1275. m_frameset_ok = false;
  1276. m_insertion_mode = InsertionMode::InTable;
  1277. return;
  1278. }
  1279. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1280. token.drop_attributes();
  1281. goto BRStartTag;
  1282. }
  1283. 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)) {
  1284. BRStartTag:
  1285. reconstruct_the_active_formatting_elements();
  1286. insert_html_element(token);
  1287. m_stack_of_open_elements.pop();
  1288. token.acknowledge_self_closing_flag_if_set();
  1289. m_frameset_ok = false;
  1290. return;
  1291. }
  1292. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1293. reconstruct_the_active_formatting_elements();
  1294. insert_html_element(token);
  1295. m_stack_of_open_elements.pop();
  1296. token.acknowledge_self_closing_flag_if_set();
  1297. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1298. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1299. m_frameset_ok = false;
  1300. }
  1301. return;
  1302. }
  1303. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1304. insert_html_element(token);
  1305. m_stack_of_open_elements.pop();
  1306. token.acknowledge_self_closing_flag_if_set();
  1307. return;
  1308. }
  1309. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1310. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1311. close_a_p_element();
  1312. insert_html_element(token);
  1313. m_stack_of_open_elements.pop();
  1314. token.acknowledge_self_closing_flag_if_set();
  1315. m_frameset_ok = false;
  1316. return;
  1317. }
  1318. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::image) {
  1319. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1320. PARSE_ERROR();
  1321. token.m_tag.tag_name.clear();
  1322. token.m_tag.tag_name.append(HTML::TagNames::img);
  1323. process_using_the_rules_for(m_insertion_mode, token);
  1324. return;
  1325. }
  1326. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1327. insert_html_element(token);
  1328. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1329. // If the next token is a U+000A LINE FEED (LF) character token,
  1330. // then ignore that token and move on to the next one.
  1331. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1332. auto next_token = m_tokenizer.next_token();
  1333. m_original_insertion_mode = m_insertion_mode;
  1334. m_frameset_ok = false;
  1335. m_insertion_mode = InsertionMode::Text;
  1336. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1337. // Ignore it.
  1338. } else {
  1339. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1340. }
  1341. return;
  1342. }
  1343. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1344. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1345. close_a_p_element();
  1346. }
  1347. reconstruct_the_active_formatting_elements();
  1348. m_frameset_ok = false;
  1349. parse_generic_raw_text_element(token);
  1350. return;
  1351. }
  1352. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1353. m_frameset_ok = false;
  1354. parse_generic_raw_text_element(token);
  1355. return;
  1356. }
  1357. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1358. parse_generic_raw_text_element(token);
  1359. return;
  1360. }
  1361. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1362. reconstruct_the_active_formatting_elements();
  1363. insert_html_element(token);
  1364. m_frameset_ok = false;
  1365. switch (m_insertion_mode) {
  1366. case InsertionMode::InTable:
  1367. case InsertionMode::InCaption:
  1368. case InsertionMode::InTableBody:
  1369. case InsertionMode::InRow:
  1370. case InsertionMode::InCell:
  1371. m_insertion_mode = InsertionMode::InSelectInTable;
  1372. break;
  1373. default:
  1374. m_insertion_mode = InsertionMode::InSelect;
  1375. break;
  1376. }
  1377. return;
  1378. }
  1379. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1380. if (current_node().local_name() == HTML::TagNames::option)
  1381. m_stack_of_open_elements.pop();
  1382. reconstruct_the_active_formatting_elements();
  1383. insert_html_element(token);
  1384. return;
  1385. }
  1386. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1387. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1388. generate_implied_end_tags();
  1389. if (current_node().local_name() != HTML::TagNames::ruby)
  1390. PARSE_ERROR();
  1391. insert_html_element(token);
  1392. return;
  1393. }
  1394. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1395. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1396. generate_implied_end_tags(HTML::TagNames::rtc);
  1397. if (current_node().local_name() != HTML::TagNames::rtc || current_node().local_name() != HTML::TagNames::ruby)
  1398. PARSE_ERROR();
  1399. insert_html_element(token);
  1400. return;
  1401. }
  1402. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::math) {
  1403. dbg() << "<math> element encountered.";
  1404. reconstruct_the_active_formatting_elements();
  1405. adjust_mathml_attributes(token);
  1406. adjust_foreign_attributes(token);
  1407. // FIXME: this should insert a foreign element, but lets just insert it normally for now :^)
  1408. insert_html_element(token);
  1409. if (token.is_self_closing()) {
  1410. m_stack_of_open_elements.pop();
  1411. token.acknowledge_self_closing_flag_if_set();
  1412. }
  1413. return;
  1414. }
  1415. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::svg) {
  1416. dbg() << "<svg> element encountered.";
  1417. reconstruct_the_active_formatting_elements();
  1418. adjust_svg_attributes(token);
  1419. adjust_foreign_attributes(token);
  1420. // FIXME: this should insert a foreign element, but lets just insert it normally for now :^)
  1421. insert_html_element(token);
  1422. if (token.is_self_closing()) {
  1423. m_stack_of_open_elements.pop();
  1424. token.acknowledge_self_closing_flag_if_set();
  1425. }
  1426. return;
  1427. }
  1428. 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))) {
  1429. PARSE_ERROR();
  1430. return;
  1431. }
  1432. // Any other start tag
  1433. if (token.is_start_tag()) {
  1434. reconstruct_the_active_formatting_elements();
  1435. insert_html_element(token);
  1436. return;
  1437. }
  1438. if (token.is_end_tag()) {
  1439. AnyOtherEndTag:
  1440. RefPtr<DOM::Element> node;
  1441. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1442. node = m_stack_of_open_elements.elements()[i];
  1443. if (node->local_name() == token.tag_name()) {
  1444. generate_implied_end_tags(token.tag_name());
  1445. if (node != current_node()) {
  1446. PARSE_ERROR();
  1447. }
  1448. while (&current_node() != node) {
  1449. m_stack_of_open_elements.pop();
  1450. }
  1451. m_stack_of_open_elements.pop();
  1452. break;
  1453. }
  1454. if (is_special_tag(node->local_name())) {
  1455. PARSE_ERROR();
  1456. return;
  1457. }
  1458. }
  1459. return;
  1460. }
  1461. }
  1462. void HTMLDocumentParser::adjust_mathml_attributes(HTMLToken& token)
  1463. {
  1464. token.adjust_attribute_name("definitionurl", "definitionURL");
  1465. }
  1466. void HTMLDocumentParser::adjust_svg_attributes(HTMLToken& token)
  1467. {
  1468. token.adjust_attribute_name("attributename", "attributeName");
  1469. token.adjust_attribute_name("attributetype", "attributeType");
  1470. token.adjust_attribute_name("basefrequency", "baseFrequency");
  1471. token.adjust_attribute_name("baseprofile", "baseProfile");
  1472. token.adjust_attribute_name("calcmode", "calcMode");
  1473. token.adjust_attribute_name("clippathunits", "clipPathUnits");
  1474. token.adjust_attribute_name("diffuseconstant", "diffuseConstant");
  1475. token.adjust_attribute_name("edgemode", "edgeMode");
  1476. token.adjust_attribute_name("filterunits", "filterUnits");
  1477. token.adjust_attribute_name("glyphref", "glyphRef");
  1478. token.adjust_attribute_name("gradienttransform", "gradientTransform");
  1479. token.adjust_attribute_name("gradientunits", "gradientUnits");
  1480. token.adjust_attribute_name("kernelmatrix", "kernelMatrix");
  1481. token.adjust_attribute_name("kernelunitlength", "kernelUnitLength");
  1482. token.adjust_attribute_name("keypoints", "keyPoints");
  1483. token.adjust_attribute_name("keysplines", "keySplines");
  1484. token.adjust_attribute_name("keytimes", "keyTimes");
  1485. token.adjust_attribute_name("lengthadjust", "lengthAdjust");
  1486. token.adjust_attribute_name("limitingconeangle", "limitingConeAngle");
  1487. token.adjust_attribute_name("markerheight", "markerHeight");
  1488. token.adjust_attribute_name("markerunits", "markerUnits");
  1489. token.adjust_attribute_name("markerwidth", "markerWidth");
  1490. token.adjust_attribute_name("maskcontentunits", "maskContentUnits");
  1491. token.adjust_attribute_name("maskunits", "maskUnits");
  1492. token.adjust_attribute_name("numoctaves", "numOctaves");
  1493. token.adjust_attribute_name("pathlength", "pathLength");
  1494. token.adjust_attribute_name("patterncontentunits", "patternContentUnits");
  1495. token.adjust_attribute_name("patterntransform", "patternTransform");
  1496. token.adjust_attribute_name("patternunits", "patternUnits");
  1497. token.adjust_attribute_name("pointsatx", "pointsAtX");
  1498. token.adjust_attribute_name("pointsaty", "pointsAtY");
  1499. token.adjust_attribute_name("pointsatz", "pointsAtZ");
  1500. token.adjust_attribute_name("preservealpha", "preserveAlpha");
  1501. token.adjust_attribute_name("preserveaspectratio", "preserveAspectRatio");
  1502. token.adjust_attribute_name("primitiveunits", "primitiveUnits");
  1503. token.adjust_attribute_name("refx", "refX");
  1504. token.adjust_attribute_name("refy", "refY");
  1505. token.adjust_attribute_name("repeatcount", "repeatCount");
  1506. token.adjust_attribute_name("repeatdur", "repeatDur");
  1507. token.adjust_attribute_name("requiredextensions", "requiredExtensions");
  1508. token.adjust_attribute_name("requiredfeatures", "requiredFeatures");
  1509. token.adjust_attribute_name("specularconstant", "specularConstant");
  1510. token.adjust_attribute_name("specularexponent", "specularExponent");
  1511. token.adjust_attribute_name("spreadmethod", "spreadMethod");
  1512. token.adjust_attribute_name("startoffset", "startOffset");
  1513. token.adjust_attribute_name("stddeviation", "stdDeviation");
  1514. token.adjust_attribute_name("stitchtiles", "stitchTiles");
  1515. token.adjust_attribute_name("surfacescale", "surfaceScale");
  1516. token.adjust_attribute_name("systemlanguage", "systemLanguage");
  1517. token.adjust_attribute_name("tablevalues", "tableValues");
  1518. token.adjust_attribute_name("targetx", "targetX");
  1519. token.adjust_attribute_name("targety", "targetY");
  1520. token.adjust_attribute_name("textlength", "textLength");
  1521. token.adjust_attribute_name("viewbox", "viewBox");
  1522. token.adjust_attribute_name("viewtarget", "viewTarget");
  1523. token.adjust_attribute_name("xchannelselector", "xChannelSelector");
  1524. token.adjust_attribute_name("ychannelselector", "yChannelSelector");
  1525. token.adjust_attribute_name("zoomandpan", "zoomAndPan");
  1526. }
  1527. void HTMLDocumentParser::adjust_foreign_attributes(HTMLToken& token)
  1528. {
  1529. auto xlink_namespace = "http://www.w3.org/1999/xlink";
  1530. token.adjust_foreign_attribute("xlink:actuate", "xlink", "actuate", xlink_namespace);
  1531. token.adjust_foreign_attribute("xlink:arcrole", "xlink", "arcrole", xlink_namespace);
  1532. token.adjust_foreign_attribute("xlink:href", "xlink", "href", xlink_namespace);
  1533. token.adjust_foreign_attribute("xlink:role", "xlink", "role", xlink_namespace);
  1534. token.adjust_foreign_attribute("xlink:show", "xlink", "show", xlink_namespace);
  1535. token.adjust_foreign_attribute("xlink:title", "xlink", "title", xlink_namespace);
  1536. token.adjust_foreign_attribute("xlink:type", "xlink", "type", xlink_namespace);
  1537. auto xml_namespace = "http://www.w3.org/XML/1998/namespace";
  1538. token.adjust_foreign_attribute("xml:lang", "xml", "lang", xml_namespace);
  1539. token.adjust_foreign_attribute("xml:space", "xml", "space", xml_namespace);
  1540. auto xmlns_namespace = "http://www.w3.org/2000/xmlns/";
  1541. token.adjust_foreign_attribute("xmlns", "", "xmlns", xmlns_namespace);
  1542. token.adjust_foreign_attribute("xmlns:xlink", "xmlns", "xlink", xmlns_namespace);
  1543. }
  1544. void HTMLDocumentParser::increment_script_nesting_level()
  1545. {
  1546. ++m_script_nesting_level;
  1547. }
  1548. void HTMLDocumentParser::decrement_script_nesting_level()
  1549. {
  1550. ASSERT(m_script_nesting_level);
  1551. --m_script_nesting_level;
  1552. }
  1553. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1554. {
  1555. if (token.is_character()) {
  1556. insert_character(token.code_point());
  1557. return;
  1558. }
  1559. if (token.is_end_of_file()) {
  1560. PARSE_ERROR();
  1561. if (current_node().local_name() == HTML::TagNames::script)
  1562. downcast<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1563. m_stack_of_open_elements.pop();
  1564. m_insertion_mode = m_original_insertion_mode;
  1565. process_using_the_rules_for(m_insertion_mode, token);
  1566. return;
  1567. }
  1568. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1569. // Make sure the <script> element has up-to-date text content before preparing the script.
  1570. flush_character_insertions();
  1571. NonnullRefPtr<HTMLScriptElement> script = downcast<HTMLScriptElement>(current_node());
  1572. m_stack_of_open_elements.pop();
  1573. m_insertion_mode = m_original_insertion_mode;
  1574. // FIXME: Handle tokenizer insertion point stuff here.
  1575. increment_script_nesting_level();
  1576. script->prepare_script({});
  1577. decrement_script_nesting_level();
  1578. if (script_nesting_level() == 0)
  1579. m_parser_pause_flag = false;
  1580. // FIXME: Handle tokenizer insertion point stuff here too.
  1581. while (document().pending_parsing_blocking_script()) {
  1582. if (script_nesting_level() != 0) {
  1583. m_parser_pause_flag = true;
  1584. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1585. // yielding control back to the caller. (Tokenization will resume when
  1586. // the caller returns to the "outer" tree construction stage.)
  1587. TODO();
  1588. } else {
  1589. auto the_script = document().take_pending_parsing_blocking_script({});
  1590. m_tokenizer.set_blocked(true);
  1591. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1592. // or the script's "ready to be parser-executed" flag is not set:
  1593. // spin the event loop until the parser's Document has no style sheet
  1594. // that is blocking scripts and the script's "ready to be parser-executed"
  1595. // flag is set.
  1596. if (the_script->failed_to_load())
  1597. return;
  1598. ASSERT(the_script->is_ready_to_be_parser_executed());
  1599. if (m_aborted)
  1600. return;
  1601. m_tokenizer.set_blocked(false);
  1602. // FIXME: Handle tokenizer insertion point stuff here too.
  1603. ASSERT(script_nesting_level() == 0);
  1604. increment_script_nesting_level();
  1605. the_script->execute_script();
  1606. decrement_script_nesting_level();
  1607. ASSERT(script_nesting_level() == 0);
  1608. m_parser_pause_flag = false;
  1609. // FIXME: Handle tokenizer insertion point stuff here too.
  1610. }
  1611. }
  1612. return;
  1613. }
  1614. if (token.is_end_tag()) {
  1615. m_stack_of_open_elements.pop();
  1616. m_insertion_mode = m_original_insertion_mode;
  1617. return;
  1618. }
  1619. TODO();
  1620. }
  1621. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1622. {
  1623. while (!current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1624. m_stack_of_open_elements.pop();
  1625. if (current_node().local_name() == HTML::TagNames::html)
  1626. ASSERT(m_parsing_fragment);
  1627. }
  1628. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1629. {
  1630. while (!current_node().local_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1631. m_stack_of_open_elements.pop();
  1632. if (current_node().local_name() == HTML::TagNames::html)
  1633. ASSERT(m_parsing_fragment);
  1634. }
  1635. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1636. {
  1637. while (!current_node().local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1638. m_stack_of_open_elements.pop();
  1639. if (current_node().local_name() == HTML::TagNames::html)
  1640. ASSERT(m_parsing_fragment);
  1641. }
  1642. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1643. {
  1644. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1645. clear_the_stack_back_to_a_table_row_context();
  1646. insert_html_element(token);
  1647. m_insertion_mode = InsertionMode::InCell;
  1648. m_list_of_active_formatting_elements.add_marker();
  1649. return;
  1650. }
  1651. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1652. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1653. PARSE_ERROR();
  1654. return;
  1655. }
  1656. clear_the_stack_back_to_a_table_row_context();
  1657. m_stack_of_open_elements.pop();
  1658. m_insertion_mode = InsertionMode::InTableBody;
  1659. return;
  1660. }
  1661. 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))
  1662. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1663. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1664. PARSE_ERROR();
  1665. return;
  1666. }
  1667. clear_the_stack_back_to_a_table_row_context();
  1668. m_stack_of_open_elements.pop();
  1669. m_insertion_mode = InsertionMode::InTableBody;
  1670. process_using_the_rules_for(m_insertion_mode, token);
  1671. return;
  1672. }
  1673. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1674. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1675. PARSE_ERROR();
  1676. return;
  1677. }
  1678. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1679. return;
  1680. }
  1681. clear_the_stack_back_to_a_table_row_context();
  1682. m_stack_of_open_elements.pop();
  1683. m_insertion_mode = InsertionMode::InTableBody;
  1684. process_using_the_rules_for(m_insertion_mode, token);
  1685. return;
  1686. }
  1687. 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)) {
  1688. PARSE_ERROR();
  1689. return;
  1690. }
  1691. process_using_the_rules_for(InsertionMode::InTable, token);
  1692. }
  1693. void HTMLDocumentParser::close_the_cell()
  1694. {
  1695. generate_implied_end_tags();
  1696. if (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1697. PARSE_ERROR();
  1698. }
  1699. while (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1700. m_stack_of_open_elements.pop();
  1701. m_stack_of_open_elements.pop();
  1702. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1703. m_insertion_mode = InsertionMode::InRow;
  1704. }
  1705. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1706. {
  1707. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1708. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1709. PARSE_ERROR();
  1710. return;
  1711. }
  1712. generate_implied_end_tags();
  1713. if (current_node().local_name() != token.tag_name()) {
  1714. PARSE_ERROR();
  1715. }
  1716. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1717. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1718. m_insertion_mode = InsertionMode::InRow;
  1719. return;
  1720. }
  1721. 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)) {
  1722. 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)) {
  1723. ASSERT(m_parsing_fragment);
  1724. PARSE_ERROR();
  1725. return;
  1726. }
  1727. close_the_cell();
  1728. process_using_the_rules_for(m_insertion_mode, token);
  1729. return;
  1730. }
  1731. 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)) {
  1732. PARSE_ERROR();
  1733. return;
  1734. }
  1735. 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)) {
  1736. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1737. PARSE_ERROR();
  1738. return;
  1739. }
  1740. close_the_cell();
  1741. // Reprocess the token.
  1742. process_using_the_rules_for(m_insertion_mode, token);
  1743. return;
  1744. }
  1745. process_using_the_rules_for(InsertionMode::InBody, token);
  1746. }
  1747. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1748. {
  1749. if (token.is_character()) {
  1750. if (token.code_point() == 0) {
  1751. PARSE_ERROR();
  1752. return;
  1753. }
  1754. m_pending_table_character_tokens.append(token);
  1755. return;
  1756. }
  1757. for (auto& pending_token : m_pending_table_character_tokens) {
  1758. ASSERT(pending_token.is_character());
  1759. if (!pending_token.is_parser_whitespace()) {
  1760. // If any of the tokens in the pending table character tokens list
  1761. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1762. // reprocess the character tokens in the pending table character tokens list using
  1763. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1764. PARSE_ERROR();
  1765. m_foster_parenting = true;
  1766. process_using_the_rules_for(InsertionMode::InBody, token);
  1767. m_foster_parenting = false;
  1768. return;
  1769. }
  1770. }
  1771. for (auto& pending_token : m_pending_table_character_tokens) {
  1772. insert_character(pending_token.code_point());
  1773. }
  1774. m_insertion_mode = m_original_insertion_mode;
  1775. process_using_the_rules_for(m_insertion_mode, token);
  1776. }
  1777. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1778. {
  1779. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1780. clear_the_stack_back_to_a_table_body_context();
  1781. insert_html_element(token);
  1782. m_insertion_mode = InsertionMode::InRow;
  1783. return;
  1784. }
  1785. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1786. PARSE_ERROR();
  1787. clear_the_stack_back_to_a_table_body_context();
  1788. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tr));
  1789. m_insertion_mode = InsertionMode::InRow;
  1790. process_using_the_rules_for(m_insertion_mode, token);
  1791. return;
  1792. }
  1793. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1794. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1795. PARSE_ERROR();
  1796. return;
  1797. }
  1798. clear_the_stack_back_to_a_table_body_context();
  1799. m_stack_of_open_elements.pop();
  1800. m_insertion_mode = InsertionMode::InTable;
  1801. return;
  1802. }
  1803. 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))
  1804. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1805. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tbody)
  1806. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::thead)
  1807. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tfoot)) {
  1808. PARSE_ERROR();
  1809. return;
  1810. }
  1811. clear_the_stack_back_to_a_table_body_context();
  1812. m_stack_of_open_elements.pop();
  1813. m_insertion_mode = InsertionMode::InTable;
  1814. process_using_the_rules_for(InsertionMode::InTable, token);
  1815. return;
  1816. }
  1817. 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)) {
  1818. PARSE_ERROR();
  1819. return;
  1820. }
  1821. process_using_the_rules_for(InsertionMode::InTable, token);
  1822. }
  1823. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1824. {
  1825. 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)) {
  1826. m_pending_table_character_tokens.clear();
  1827. m_original_insertion_mode = m_insertion_mode;
  1828. m_insertion_mode = InsertionMode::InTableText;
  1829. process_using_the_rules_for(InsertionMode::InTableText, token);
  1830. return;
  1831. }
  1832. if (token.is_comment()) {
  1833. insert_comment(token);
  1834. return;
  1835. }
  1836. if (token.is_doctype()) {
  1837. PARSE_ERROR();
  1838. return;
  1839. }
  1840. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1841. clear_the_stack_back_to_a_table_context();
  1842. m_list_of_active_formatting_elements.add_marker();
  1843. insert_html_element(token);
  1844. m_insertion_mode = InsertionMode::InCaption;
  1845. return;
  1846. }
  1847. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1848. clear_the_stack_back_to_a_table_context();
  1849. insert_html_element(token);
  1850. m_insertion_mode = InsertionMode::InColumnGroup;
  1851. return;
  1852. }
  1853. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1854. clear_the_stack_back_to_a_table_context();
  1855. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::colgroup));
  1856. m_insertion_mode = InsertionMode::InColumnGroup;
  1857. process_using_the_rules_for(m_insertion_mode, token);
  1858. return;
  1859. }
  1860. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1861. clear_the_stack_back_to_a_table_context();
  1862. insert_html_element(token);
  1863. m_insertion_mode = InsertionMode::InTableBody;
  1864. return;
  1865. }
  1866. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1867. clear_the_stack_back_to_a_table_context();
  1868. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tbody));
  1869. m_insertion_mode = InsertionMode::InTableBody;
  1870. process_using_the_rules_for(m_insertion_mode, token);
  1871. return;
  1872. }
  1873. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1874. PARSE_ERROR();
  1875. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table))
  1876. return;
  1877. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1878. reset_the_insertion_mode_appropriately();
  1879. process_using_the_rules_for(m_insertion_mode, token);
  1880. return;
  1881. }
  1882. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::table) {
  1883. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1884. PARSE_ERROR();
  1885. return;
  1886. }
  1887. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1888. reset_the_insertion_mode_appropriately();
  1889. return;
  1890. }
  1891. 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)) {
  1892. PARSE_ERROR();
  1893. return;
  1894. }
  1895. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::style, HTML::TagNames::script, HTML::TagNames::template_))
  1896. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_)) {
  1897. process_using_the_rules_for(InsertionMode::InHead, token);
  1898. return;
  1899. }
  1900. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1901. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1902. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1903. goto AnythingElse;
  1904. }
  1905. PARSE_ERROR();
  1906. insert_html_element(token);
  1907. // FIXME: Is this the correct interpretation of "Pop that input element off the stack of open elements."?
  1908. // Because this wording is the first time it's seen in the spec.
  1909. // Other times it's worded as: "Immediately pop the current node off the stack of open elements."
  1910. m_stack_of_open_elements.pop();
  1911. token.acknowledge_self_closing_flag_if_set();
  1912. return;
  1913. }
  1914. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1915. PARSE_ERROR();
  1916. if (m_form_element || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1917. return;
  1918. }
  1919. m_form_element = downcast<HTMLFormElement>(*insert_html_element(token));
  1920. // FIXME: See previous FIXME, as this is the same situation but for form.
  1921. m_stack_of_open_elements.pop();
  1922. return;
  1923. }
  1924. if (token.is_end_of_file()) {
  1925. process_using_the_rules_for(InsertionMode::InBody, token);
  1926. return;
  1927. }
  1928. AnythingElse:
  1929. PARSE_ERROR();
  1930. m_foster_parenting = true;
  1931. process_using_the_rules_for(InsertionMode::InBody, token);
  1932. m_foster_parenting = false;
  1933. }
  1934. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1935. {
  1936. 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)) {
  1937. PARSE_ERROR();
  1938. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1939. reset_the_insertion_mode_appropriately();
  1940. process_using_the_rules_for(m_insertion_mode, token);
  1941. return;
  1942. }
  1943. 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)) {
  1944. PARSE_ERROR();
  1945. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name()))
  1946. return;
  1947. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1948. reset_the_insertion_mode_appropriately();
  1949. process_using_the_rules_for(m_insertion_mode, token);
  1950. return;
  1951. }
  1952. process_using_the_rules_for(InsertionMode::InSelect, token);
  1953. }
  1954. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  1955. {
  1956. if (token.is_character()) {
  1957. if (token.code_point() == 0) {
  1958. PARSE_ERROR();
  1959. return;
  1960. }
  1961. insert_character(token.code_point());
  1962. return;
  1963. }
  1964. if (token.is_comment()) {
  1965. insert_comment(token);
  1966. return;
  1967. }
  1968. if (token.is_doctype()) {
  1969. PARSE_ERROR();
  1970. return;
  1971. }
  1972. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  1973. process_using_the_rules_for(InsertionMode::InBody, token);
  1974. return;
  1975. }
  1976. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  1977. if (current_node().local_name() == HTML::TagNames::option) {
  1978. m_stack_of_open_elements.pop();
  1979. }
  1980. insert_html_element(token);
  1981. return;
  1982. }
  1983. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1984. if (current_node().local_name() == HTML::TagNames::option) {
  1985. m_stack_of_open_elements.pop();
  1986. }
  1987. if (current_node().local_name() == HTML::TagNames::optgroup) {
  1988. m_stack_of_open_elements.pop();
  1989. }
  1990. insert_html_element(token);
  1991. return;
  1992. }
  1993. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1994. if (current_node().local_name() == HTML::TagNames::option && node_before_current_node().local_name() == HTML::TagNames::optgroup)
  1995. m_stack_of_open_elements.pop();
  1996. if (current_node().local_name() == HTML::TagNames::optgroup) {
  1997. m_stack_of_open_elements.pop();
  1998. } else {
  1999. PARSE_ERROR();
  2000. return;
  2001. }
  2002. return;
  2003. }
  2004. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  2005. if (current_node().local_name() == HTML::TagNames::option) {
  2006. m_stack_of_open_elements.pop();
  2007. } else {
  2008. PARSE_ERROR();
  2009. return;
  2010. }
  2011. return;
  2012. }
  2013. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  2014. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2015. ASSERT(m_parsing_fragment);
  2016. PARSE_ERROR();
  2017. return;
  2018. }
  2019. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2020. reset_the_insertion_mode_appropriately();
  2021. return;
  2022. }
  2023. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  2024. PARSE_ERROR();
  2025. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2026. ASSERT(m_parsing_fragment);
  2027. return;
  2028. }
  2029. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2030. reset_the_insertion_mode_appropriately();
  2031. return;
  2032. }
  2033. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  2034. PARSE_ERROR();
  2035. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2036. ASSERT(m_parsing_fragment);
  2037. return;
  2038. }
  2039. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2040. reset_the_insertion_mode_appropriately();
  2041. process_using_the_rules_for(m_insertion_mode, token);
  2042. return;
  2043. }
  2044. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  2045. process_using_the_rules_for(InsertionMode::InHead, token);
  2046. return;
  2047. }
  2048. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2049. process_using_the_rules_for(InsertionMode::InHead, token);
  2050. return;
  2051. }
  2052. if (token.is_end_of_file()) {
  2053. process_using_the_rules_for(InsertionMode::InBody, token);
  2054. return;
  2055. }
  2056. PARSE_ERROR();
  2057. }
  2058. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  2059. {
  2060. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  2061. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2062. ASSERT(m_parsing_fragment);
  2063. PARSE_ERROR();
  2064. return;
  2065. }
  2066. generate_implied_end_tags();
  2067. if (current_node().local_name() != HTML::TagNames::caption)
  2068. PARSE_ERROR();
  2069. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2070. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2071. m_insertion_mode = InsertionMode::InTable;
  2072. return;
  2073. }
  2074. 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))
  2075. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  2076. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2077. ASSERT(m_parsing_fragment);
  2078. PARSE_ERROR();
  2079. return;
  2080. }
  2081. generate_implied_end_tags();
  2082. if (current_node().local_name() != HTML::TagNames::caption)
  2083. PARSE_ERROR();
  2084. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2085. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2086. m_insertion_mode = InsertionMode::InTable;
  2087. process_using_the_rules_for(m_insertion_mode, token);
  2088. return;
  2089. }
  2090. 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)) {
  2091. PARSE_ERROR();
  2092. return;
  2093. }
  2094. process_using_the_rules_for(InsertionMode::InBody, token);
  2095. }
  2096. void HTMLDocumentParser::handle_in_column_group(HTMLToken& token)
  2097. {
  2098. if (token.is_character() && token.is_parser_whitespace()) {
  2099. insert_character(token.code_point());
  2100. return;
  2101. }
  2102. if (token.is_comment()) {
  2103. insert_comment(token);
  2104. return;
  2105. }
  2106. if (token.is_doctype()) {
  2107. PARSE_ERROR();
  2108. return;
  2109. }
  2110. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2111. process_using_the_rules_for(InsertionMode::InBody, token);
  2112. return;
  2113. }
  2114. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2115. insert_html_element(token);
  2116. m_stack_of_open_elements.pop();
  2117. token.acknowledge_self_closing_flag_if_set();
  2118. return;
  2119. }
  2120. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  2121. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2122. PARSE_ERROR();
  2123. return;
  2124. }
  2125. m_stack_of_open_elements.pop();
  2126. m_insertion_mode = InsertionMode::InTable;
  2127. return;
  2128. }
  2129. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::col) {
  2130. PARSE_ERROR();
  2131. return;
  2132. }
  2133. if ((token.is_start_tag() || token.is_end_tag()) && token.tag_name() == HTML::TagNames::template_) {
  2134. process_using_the_rules_for(InsertionMode::InHead, token);
  2135. return;
  2136. }
  2137. if (token.is_end_of_file()) {
  2138. process_using_the_rules_for(InsertionMode::InBody, token);
  2139. return;
  2140. }
  2141. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2142. PARSE_ERROR();
  2143. return;
  2144. }
  2145. m_stack_of_open_elements.pop();
  2146. m_insertion_mode = InsertionMode::InTable;
  2147. process_using_the_rules_for(m_insertion_mode, token);
  2148. }
  2149. void HTMLDocumentParser::handle_in_template(HTMLToken& token)
  2150. {
  2151. if (token.is_character() || token.is_comment() || token.is_doctype()) {
  2152. process_using_the_rules_for(InsertionMode::InBody, token);
  2153. return;
  2154. }
  2155. 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)) {
  2156. process_using_the_rules_for(InsertionMode::InHead, token);
  2157. return;
  2158. }
  2159. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2160. process_using_the_rules_for(InsertionMode::InHead, token);
  2161. return;
  2162. }
  2163. 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)) {
  2164. m_stack_of_template_insertion_modes.take_last();
  2165. m_stack_of_template_insertion_modes.append(InsertionMode::InTable);
  2166. m_insertion_mode = InsertionMode::InTable;
  2167. process_using_the_rules_for(m_insertion_mode, token);
  2168. return;
  2169. }
  2170. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2171. m_stack_of_template_insertion_modes.take_last();
  2172. m_stack_of_template_insertion_modes.append(InsertionMode::InColumnGroup);
  2173. m_insertion_mode = InsertionMode::InColumnGroup;
  2174. process_using_the_rules_for(m_insertion_mode, token);
  2175. return;
  2176. }
  2177. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  2178. m_stack_of_template_insertion_modes.take_last();
  2179. m_stack_of_template_insertion_modes.append(InsertionMode::InTableBody);
  2180. m_insertion_mode = InsertionMode::InTableBody;
  2181. process_using_the_rules_for(m_insertion_mode, token);
  2182. return;
  2183. }
  2184. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2185. m_stack_of_template_insertion_modes.take_last();
  2186. m_stack_of_template_insertion_modes.append(InsertionMode::InRow);
  2187. m_insertion_mode = InsertionMode::InRow;
  2188. process_using_the_rules_for(m_insertion_mode, token);
  2189. return;
  2190. }
  2191. if (token.is_start_tag()) {
  2192. m_stack_of_template_insertion_modes.take_last();
  2193. m_stack_of_template_insertion_modes.append(InsertionMode::InBody);
  2194. m_insertion_mode = InsertionMode::InBody;
  2195. process_using_the_rules_for(m_insertion_mode, token);
  2196. return;
  2197. }
  2198. if (token.is_end_tag()) {
  2199. PARSE_ERROR();
  2200. return;
  2201. }
  2202. if (token.is_end_of_file()) {
  2203. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  2204. ASSERT(m_parsing_fragment);
  2205. stop_parsing();
  2206. return;
  2207. }
  2208. PARSE_ERROR();
  2209. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  2210. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2211. m_stack_of_template_insertion_modes.take_last();
  2212. reset_the_insertion_mode_appropriately();
  2213. process_using_the_rules_for(m_insertion_mode, token);
  2214. }
  2215. }
  2216. void HTMLDocumentParser::handle_in_frameset(HTMLToken& token)
  2217. {
  2218. if (token.is_character() && token.is_parser_whitespace()) {
  2219. insert_character(token.code_point());
  2220. return;
  2221. }
  2222. if (token.is_comment()) {
  2223. insert_comment(token);
  2224. return;
  2225. }
  2226. if (token.is_doctype()) {
  2227. PARSE_ERROR();
  2228. return;
  2229. }
  2230. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2231. process_using_the_rules_for(InsertionMode::InBody, token);
  2232. return;
  2233. }
  2234. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2235. insert_html_element(token);
  2236. return;
  2237. }
  2238. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2239. // FIXME: If the current node is the root html element, then this is a parse error; ignore the token. (fragment case)
  2240. m_stack_of_open_elements.pop();
  2241. if (!m_parsing_fragment && current_node().local_name() != HTML::TagNames::frameset) {
  2242. m_insertion_mode = InsertionMode::AfterFrameset;
  2243. }
  2244. return;
  2245. }
  2246. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frame) {
  2247. insert_html_element(token);
  2248. m_stack_of_open_elements.pop();
  2249. token.acknowledge_self_closing_flag_if_set();
  2250. return;
  2251. }
  2252. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2253. process_using_the_rules_for(InsertionMode::InHead, token);
  2254. return;
  2255. }
  2256. if (token.is_end_of_file()) {
  2257. //FIXME: If the current node is not the root html element, then this is a parse error.
  2258. stop_parsing();
  2259. return;
  2260. }
  2261. PARSE_ERROR();
  2262. }
  2263. void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
  2264. {
  2265. if (token.is_character() && token.is_parser_whitespace()) {
  2266. insert_character(token.code_point());
  2267. return;
  2268. }
  2269. if (token.is_comment()) {
  2270. insert_comment(token);
  2271. return;
  2272. }
  2273. if (token.is_doctype()) {
  2274. PARSE_ERROR();
  2275. return;
  2276. }
  2277. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2278. process_using_the_rules_for(InsertionMode::InBody, token);
  2279. return;
  2280. }
  2281. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  2282. m_insertion_mode = InsertionMode::AfterAfterFrameset;
  2283. return;
  2284. }
  2285. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2286. process_using_the_rules_for(InsertionMode::InHead, token);
  2287. return;
  2288. }
  2289. if (token.is_end_of_file()) {
  2290. stop_parsing();
  2291. return;
  2292. }
  2293. PARSE_ERROR();
  2294. }
  2295. void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
  2296. {
  2297. if (token.is_comment()) {
  2298. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  2299. document().append_child(move(comment));
  2300. return;
  2301. }
  2302. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  2303. process_using_the_rules_for(InsertionMode::InBody, token);
  2304. return;
  2305. }
  2306. if (token.is_end_of_file()) {
  2307. stop_parsing();
  2308. return;
  2309. }
  2310. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2311. process_using_the_rules_for(InsertionMode::InHead, token);
  2312. return;
  2313. }
  2314. PARSE_ERROR();
  2315. }
  2316. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  2317. {
  2318. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2319. bool last = i == 0;
  2320. // NOTE: When parsing fragments, we substitute the context element for the root of the stack of open elements.
  2321. RefPtr<DOM::Element> node;
  2322. if (last && m_parsing_fragment) {
  2323. node = m_context_element;
  2324. } else {
  2325. node = m_stack_of_open_elements.elements().at(i);
  2326. }
  2327. if (node->local_name() == HTML::TagNames::select) {
  2328. TODO();
  2329. }
  2330. if (!last && node->local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2331. m_insertion_mode = InsertionMode::InCell;
  2332. return;
  2333. }
  2334. if (node->local_name() == HTML::TagNames::tr) {
  2335. m_insertion_mode = InsertionMode::InRow;
  2336. return;
  2337. }
  2338. if (node->local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  2339. m_insertion_mode = InsertionMode::InTableBody;
  2340. return;
  2341. }
  2342. if (node->local_name() == HTML::TagNames::caption) {
  2343. m_insertion_mode = InsertionMode::InCaption;
  2344. return;
  2345. }
  2346. if (node->local_name() == HTML::TagNames::colgroup) {
  2347. m_insertion_mode = InsertionMode::InColumnGroup;
  2348. return;
  2349. }
  2350. if (node->local_name() == HTML::TagNames::table) {
  2351. m_insertion_mode = InsertionMode::InTable;
  2352. return;
  2353. }
  2354. if (node->local_name() == HTML::TagNames::template_) {
  2355. m_insertion_mode = m_stack_of_template_insertion_modes.last();
  2356. return;
  2357. }
  2358. if (!last && node->local_name() == HTML::TagNames::head) {
  2359. m_insertion_mode = InsertionMode::InHead;
  2360. return;
  2361. }
  2362. if (node->local_name() == HTML::TagNames::body) {
  2363. m_insertion_mode = InsertionMode::InBody;
  2364. return;
  2365. }
  2366. if (node->local_name() == HTML::TagNames::frameset) {
  2367. ASSERT(m_parsing_fragment);
  2368. m_insertion_mode = InsertionMode::InFrameset;
  2369. return;
  2370. }
  2371. if (node->local_name() == HTML::TagNames::html) {
  2372. if (!m_head_element) {
  2373. ASSERT(m_parsing_fragment);
  2374. m_insertion_mode = InsertionMode::BeforeHead;
  2375. return;
  2376. }
  2377. m_insertion_mode = InsertionMode::AfterHead;
  2378. return;
  2379. }
  2380. }
  2381. ASSERT(m_parsing_fragment);
  2382. m_insertion_mode = InsertionMode::InBody;
  2383. }
  2384. const char* HTMLDocumentParser::insertion_mode_name() const
  2385. {
  2386. switch (m_insertion_mode) {
  2387. #define __ENUMERATE_INSERTION_MODE(mode) \
  2388. case InsertionMode::mode: \
  2389. return #mode;
  2390. ENUMERATE_INSERTION_MODES
  2391. #undef __ENUMERATE_INSERTION_MODE
  2392. }
  2393. ASSERT_NOT_REACHED();
  2394. }
  2395. DOM::Document& HTMLDocumentParser::document()
  2396. {
  2397. return *m_document;
  2398. }
  2399. NonnullRefPtrVector<DOM::Node> HTMLDocumentParser::parse_html_fragment(DOM::Element& context_element, const StringView& markup)
  2400. {
  2401. HTMLDocumentParser parser(markup, "utf-8");
  2402. parser.m_context_element = context_element;
  2403. parser.m_parsing_fragment = true;
  2404. parser.document().set_quirks_mode(context_element.document().mode());
  2405. if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
  2406. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  2407. } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
  2408. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2409. } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
  2410. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  2411. } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
  2412. if (context_element.document().is_scripting_enabled())
  2413. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2414. } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
  2415. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  2416. }
  2417. auto root = create_element(context_element.document(), HTML::TagNames::html);
  2418. parser.document().append_child(root);
  2419. parser.m_stack_of_open_elements.push(root);
  2420. if (context_element.local_name() == HTML::TagNames::template_) {
  2421. parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  2422. }
  2423. // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
  2424. parser.reset_the_insertion_mode_appropriately();
  2425. for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
  2426. if (is<HTMLFormElement>(*form_candidate)) {
  2427. parser.m_form_element = downcast<HTMLFormElement>(*form_candidate);
  2428. break;
  2429. }
  2430. }
  2431. parser.run(context_element.document().url());
  2432. NonnullRefPtrVector<DOM::Node> children;
  2433. while (RefPtr<DOM::Node> child = root->first_child()) {
  2434. root->remove_child(*child);
  2435. context_element.document().adopt_node(*child);
  2436. children.append(*child);
  2437. }
  2438. return children;
  2439. }
  2440. }