From 241d4f0db7416f51ea1db3bc6b313b31cfb4e967 Mon Sep 17 00:00:00 2001
From: VAN BOSSUYT Nicolas <nivb@odoo.com>
Date: Tue, 19 Mar 2024 13:33:59 +0100
Subject: [PATCH] web-html: More spec comment.

---
 src/web/web-html/builder.cpp | 37 +++++++++++++++++++++++++++++-------
 src/web/web-html/builder.h   | 33 ++++++++++++++++++++++++++++++--
 2 files changed, 61 insertions(+), 9 deletions(-)

diff --git a/src/web/web-html/builder.cpp b/src/web/web-html/builder.cpp
index 77f6b814631..2c16f491d54 100644
--- a/src/web/web-html/builder.cpp
+++ b/src/web/web-html/builder.cpp
@@ -5,18 +5,18 @@
 
 namespace Web::Html {
 
-void Builder::_switchTo(Mode mode) {
-    _mode = mode;
-}
+// 13.2.6 Tree construction
+// https://html.spec.whatwg.org/multipage/parsing.html#tree-construction
+
+// 13.2.2 Parse errors
+// https://html.spec.whatwg.org/multipage/parsing.html#parse-errors
 
 void Builder::_raise(Str msg) {
     logError("{}: {}", toStr(_mode), msg);
 }
 
-// https://html.spec.whatwg.org/multipage/parsing.html#the-initial-insertion-mode
-Dom::QuirkMode Builder::_whichQuirkMode(Token const &) {
-    // NOSPEC: We assume no quirk mode
-    return Dom::QuirkMode::NO;
+void Builder::_switchTo(Mode mode) {
+    _mode = mode;
 }
 
 // https://html.spec.whatwg.org/multipage/parsing.html#create-an-element-for-the-token
@@ -26,7 +26,16 @@ Strong<Dom::Element> Builder::_createElementFor(Token const &t) {
     return el;
 }
 
+// 13.2.6.4 The rules for parsing tokens in HTML content
+
+// 13.2.6.4.1 The "initial" insertion mode
 // https://html.spec.whatwg.org/multipage/parsing.html#the-initial-insertion-mode
+
+Dom::QuirkMode Builder::_whichQuirkMode(Token const &) {
+    // NOSPEC: We assume no quirk mode
+    return Dom::QuirkMode::NO;
+}
+
 void Builder::_handleInitialMode(Token const &t) {
     if (t.type == Token::CHARACTER and
         (t.rune == '\t' or
@@ -87,7 +96,21 @@ void Builder::_handleBeforeHead(Token const &t) {
          t.rune == '\n' or
          t.rune == '\f' or
          t.rune == ' ')) {
+        // Ignore the token.
+    } else if (t.type == Token::COMMENT) {
+        // Insert a comment.
+    } else if (t.type == Token::DOCTYPE) {
+        // Parse error. Ignore the token.
+        _raise();
+    } else if (t.type == Token::START_TAG and t.name == "html") {
+        // Process the token using the rules for the "in body" insertion mode.
+        _acceptIn(Mode::IN_BODY, t);
+    } else if (t.type == Token::START_TAG and t.name == "head") {
+
+    } else if (t.type == Token::END_TAG and not(t.name == "head" or t.name == "body" or t.name == "html" or t.name == "br")) {
         // ignore
+        _raise();
+    } else {
     }
 }
 
diff --git a/src/web/web-html/builder.h b/src/web/web-html/builder.h
index bee23c5ac33..fa8e2374488 100644
--- a/src/web/web-html/builder.h
+++ b/src/web/web-html/builder.h
@@ -34,6 +34,9 @@ namespace Web::Html {
     MODE(AFTER_AFTER_FRAMESET)
 
 struct Builder {
+    // 13.2.6 Tree construction
+    // https://html.spec.whatwg.org/multipage/parsing.html#tree-construction
+
     enum struct Mode {
 #define ITER(NAME) NAME,
         FOREACH_INSERTION_MODE(ITER)
@@ -44,19 +47,45 @@ struct Builder {
     Lexer _lexer;
     Strong<Dom::Document> _document;
     Vec<Strong<Dom::Element>> _openElements;
+    Opt<Strong<Dom::Element>> _headElement;
+    Opt<Strong<Dom::Element>> _formElement;
 
     Builder(Strong<Dom::Document> document)
         : _document(document) {
     }
 
-    void _switchTo(Mode mode);
+    // 13.2.2 Parse errors
+    // https://html.spec.whatwg.org/multipage/parsing.html#parse-errors
 
     void _raise(Str msg = "parse-error");
 
-    Dom::QuirkMode _whichQuirkMode(Token const &);
+    // 13.2.6.1 Creating and inserting nodes
+    // https://html.spec.whatwg.org/multipage/parsing.html#creating-and-inserting-nodes
+
+    void _apropriatePlaceForInsertingANode();
 
     Strong<Dom::Element> _createElementFor(Token const &t);
 
+    void _insertAnElementAtTheAdjustedInsertionLocation();
+
+    void _insertAForeignElement(Token const &t);
+
+    void _insertAnHtmlElement();
+
+    void _insertACharacter();
+
+    void _insertAComment();
+
+    //////////////////////////////////////////
+
+    void _switchTo(Mode mode);
+
+    // 13.2.6.4 The rules for parsing tokens in HTML content
+
+    // 13.2.6.4.1 The "initial" insertion mode
+    // https://html.spec.whatwg.org/multipage/parsing.html#the-initial-insertion-mode
+    Dom::QuirkMode _whichQuirkMode(Token const &);
+
     void _handleInitialMode(Token const &t);
 
     void _handleBeforeHtml(Token const &t);