vaev-markup: in table cell html insertion mode

src/web/vaev-markup/html.cpp

@@ -4117,7 +4117,7 @@ static constexpr Array IMPLIED_END_TAGS = {
     Html::DD, Html::DT, Html::LI, Html::OPTION, Html::OPTGROUP, Html::P, Html::RB, Html::RP, Html::RT, Html::RTC
 };
 
-static void generateImpliedEndTags(HtmlParser &b, TagName except) {
+static void generateImpliedEndTags(HtmlParser &b, Opt<TagName> except = NONE) {
     while (contains(IMPLIED_END_TAGS, last(b._openElements)->tagName) and
            last(b._openElements)->tagName != except) {
         b._openElements.popBack();
@@ -5483,6 +5483,116 @@ void HtmlParser::_handleInTableRow(HtmlToken const &t) {
     }
 }
 
+// 13.2.6.4.15 MARK: The "in cell" insertion mode
+// https://html.spec.whatwg.org/multipage/parsing.html#parsing-main-intd
+void HtmlParser::_handleInCell(HtmlToken const &t) {
+    auto _closeTheCell = [&]() {
+        // Generate implied end tags.
+        generateImpliedEndTags(*this);
+
+        // If the current node is not now a td element or a th element, then this is a parse error.
+        if (last(_openElements)->tagName != Html::TD and last(_openElements)->tagName != Html::TR) {
+            _raise();
+        }
+
+        // Pop elements from the stack of open elements until a td element or a th element has been popped from the stack.
+        while (Karm::any(_openElements)) {
+            auto poppedEl = _openElements.popBack();
+            if (poppedEl == Html::TD or poppedEl == Html::TH)
+                break;
+        }
+
+        // TODO: Clear the list of active formatting elements up to the last marker.
+
+        // Switch the insertion mode to "in row".
+        _switchTo(Mode::IN_ROW);
+    };
+
+    // An end tag whose tag name is one of: "td", "th"
+    if (t.type == HtmlToken::END_TAG and (t.name == "td" or t.name == "th")) {
+        // If the stack of open elements does not have an element in table scope that is an HTML element with the same
+        // tag name as that of the token,
+        TagName tokenTagName{TagName::make(t.name, Vaev::HTML)};
+
+        if (not _hasElementInTableScope(tokenTagName)) {
+            // this is a parse error; ignore the token.
+            _raise();
+            return;
+        }
+
+        // Otherwise:
+
+        // Generate implied end tags.
+        generateImpliedEndTags(*this);
+
+        // Now, if the current node is not an HTML element with the same tag name as the token,
+        if (last(_openElements)->tagName != tokenTagName) {
+            // then this is a parse error.
+            _raise();
+        }
+
+        // Pop elements from the stack of open elements until an HTML element with the same tag name as
+        // the token has been popped from the stack.
+        while (Karm::any(_openElements) and _openElements.popBack()->tagName != tokenTagName) {
+            // do nothing
+        }
+
+        // TODO: Clear the list of active formatting elements up to the last marker.
+
+        // Switch the insertion mode to "in row".
+        _switchTo(Mode::IN_ROW);
+    }
+
+    // A start tag whose tag name is one of: "caption", "col", "colgroup", "tbody", "td", "tfoot", "th", "thead", "tr"
+    else if (t.type == HtmlToken::START_TAG and
+             (t.name == "caption" or t.name == "col" or t.name == "colgroup" or
+              t.name == "tbody" or t.name == "td" or t.name == "tfoot" or
+              t.name == "th" or t.name == "thead" or t.name == "tr"
+             )) {
+
+        // Assert: The stack of open elements has a td or th element in table scope.
+        if (not _hasElementInTableScope(Html::TD) and not _hasElementInTableScope(Html::TR)) {
+            _raise();
+            // FIXME: should this be a panic()?
+        }
+
+        // Close the cell (see below) and reprocess the token.
+        _closeTheCell();
+        accept(t);
+    }
+
+    // An end tag whose tag name is one of: "body", "caption", "col", "colgroup", "html"
+    else if (t.type == HtmlToken::END_TAG and
+             (t.name == "body" or t.name == "caption" or t.name == "col" or
+              t.name == "colgroup" or t.name == "html"
+             )) {
+        // Parse error. Ignore the token.
+        _raise();
+    }
+
+    // An end tag whose tag name is one of: "table", "tbody", "tfoot", "thead", "tr"
+    else if (t.type == HtmlToken::END_TAG and
+             (t.name == "table" or t.name == "tbody" or t.name == "tfoot" or t.name == "thead" or t.name == "tr")) {
+
+        // If the stack of open elements does not have an element in table scope that is an HTML element with the same
+        // tag name as the token,
+        if (not _hasElementInTableScope(TagName::make(t.name, Vaev::HTML))) {
+            // this is a parse error; ignore the token.
+            _raise();
+            return;
+        }
+
+        // Otherwise, close the cell (see below) and reprocess the token.
+        _closeTheCell();
+        accept(t);
+    }
+
+    else {
+        // Process the token using the rules for the "in body" insertion mode.
+        _acceptIn(Mode::IN_BODY, t);
+    }
+}
+
 // 3.2.6.4.22 MARK: The "after after body" insertion mode
 // https://html.spec.whatwg.org/multipage/parsing.html#the-after-after-body-insertion-mode
 void HtmlParser::_handleAfterBody(HtmlToken const &t) {
@@ -5579,8 +5689,8 @@ void HtmlParser::_acceptIn(Mode mode, HtmlToken const &t) {
         _handleInTableRow(t);
         break;
 
-    // TODO: https://html.spec.whatwg.org/multipage/parsing.html#parsing-main-incell
     case Mode::IN_CELL:
+        _handleInCell(t);
         break;
 
     // TODO: https://html.spec.whatwg.org/multipage/parsing.html#parsing-main-inselect

src/web/vaev-markup/html.h

@@ -332,6 +332,8 @@ struct HtmlParser : public HtmlSink {
 
     void _handleInTableRow(HtmlToken const &t);
 
+    void _handleInCell(HtmlToken const &t);
+
     void _handleAfterBody(HtmlToken const &t);
 
     void _switchTo(Mode mode);