LibWeb: Improve support for "a" and "li" during "in body" insertion

We can now parse welcome.html once again, without resorting to hacks
or fallbacks during "in body" :^)

4 files changed, 47 insertions, 7 deletions

diff --git a/Libraries/LibWeb/Parser/HTMLDocumentParser.cpp b/Libraries/LibWeb/Parser/HTMLDocumentParser.cpp
index d504633f73..e4a72302aa 100644
--- a/Libraries/LibWeb/Parser/HTMLDocumentParser.cpp
+++ b/Libraries/LibWeb/Parser/HTMLDocumentParser.cpp
@@ -623,10 +623,7 @@ void HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
         PARSE_ERROR();
     }
 
-    // FIXME: Let furthest block be the topmost node in the stack of open elements
-    //        that is lower in the stack than formatting element, and is an element
-    //        in the special category. There might not be one.
-    RefPtr<Element> furthest_block = nullptr;
+    RefPtr<Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
 
     if (!furthest_block) {
         while (&current_node() != formatting_element)
@@ -642,7 +639,7 @@ void HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
     TODO();
 }
 
-static bool is_special_tag(const FlyString& tag_name)
+bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name)
 {
     return tag_name.is_one_of(
         "address",
@@ -911,7 +908,17 @@ void HTMLDocumentParser::handle_in_body(HTMLToken& token)
     }
 
     if (token.is_end_tag() && token.tag_name() == "li") {
-        TODO();
+        if (!m_stack_of_open_elements.has_in_list_item_scope("li")) {
+            PARSE_ERROR();
+            return;
+        }
+        generate_implied_end_tags("li");
+        if (current_node().tag_name() != "li") {
+            PARSE_ERROR();
+            dbg() << "Expected <li> current node, but had <" << current_node().tag_name() << ">";
+        }
+        m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped("li");
+        return;
     }
 
     if (token.is_end_tag() && token.tag_name().is_one_of("dd", "dt")) {
@@ -943,7 +950,13 @@ void HTMLDocumentParser::handle_in_body(HTMLToken& token)
     }
 
     if (token.is_start_tag() && token.tag_name() == "a") {
-        TODO();
+        if (m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker("a")) {
+            TODO();
+        }
+        reconstruct_the_active_formatting_elements();
+        auto element = insert_html_element(token);
+        m_list_of_active_formatting_elements.add(*element);
+        return;
     }
 
     if (token.is_start_tag() && token.tag_name().is_one_of("b", "big", "code", "em", "font", "i", "s", "small", "strike", "strong", "tt", "u")) {
diff --git a/Libraries/LibWeb/Parser/HTMLDocumentParser.h b/Libraries/LibWeb/Parser/HTMLDocumentParser.h
index df98658e4c..9ac19091fc 100644
--- a/Libraries/LibWeb/Parser/HTMLDocumentParser.h
+++ b/Libraries/LibWeb/Parser/HTMLDocumentParser.h
@@ -76,6 +76,8 @@ public:
 
     InsertionMode insertion_mode() const { return m_insertion_mode; }
 
+    static bool is_special_tag(const FlyString& tag_name);
+
 private:
     const char* insertion_mode_name() const;
 
diff --git a/Libraries/LibWeb/Parser/StackOfOpenElements.cpp b/Libraries/LibWeb/Parser/StackOfOpenElements.cpp
index ff07739702..807fc40c0e 100644
--- a/Libraries/LibWeb/Parser/StackOfOpenElements.cpp
+++ b/Libraries/LibWeb/Parser/StackOfOpenElements.cpp
@@ -25,6 +25,7 @@
  */
 
 #include <LibWeb/DOM/Element.h>
+#include <LibWeb/Parser/HTMLDocumentParser.h>
 #include <LibWeb/Parser/StackOfOpenElements.h>
 
 namespace Web {
@@ -85,6 +86,14 @@ bool StackOfOpenElements::has_in_table_scope(const FlyString& tag_name) const
     return has_in_scope_impl(tag_name, list);
 }
 
+bool StackOfOpenElements::has_in_list_item_scope(const FlyString& tag_name) const
+{
+    auto list = s_base_list;
+    list.append("ol");
+    list.append("ul");
+    return has_in_scope_impl(tag_name, list);
+}
+
 bool StackOfOpenElements::contains(const Element& element) const
 {
     for (auto& element_on_stack : m_elements) {
@@ -101,4 +110,17 @@ void StackOfOpenElements::pop_until_an_element_with_tag_name_has_been_popped(con
     pop();
 }
 
+Element* StackOfOpenElements::topmost_special_node_below(const Element& formatting_element)
+{
+    Element* found_element = nullptr;
+    for (ssize_t i = m_elements.size() - 1; i >= 0; ++i) {
+        auto& element = m_elements[i];
+        if (&element == &formatting_element)
+            break;
+        if (HTMLDocumentParser::is_special_tag(element.tag_name()))
+            found_element = &element;
+    }
+    return found_element;
+}
+
 }
diff --git a/Libraries/LibWeb/Parser/StackOfOpenElements.h b/Libraries/LibWeb/Parser/StackOfOpenElements.h
index 9918158dc9..e7637c5855 100644
--- a/Libraries/LibWeb/Parser/StackOfOpenElements.h
+++ b/Libraries/LibWeb/Parser/StackOfOpenElements.h
@@ -50,6 +50,7 @@ public:
     bool has_in_scope(const FlyString& tag_name) const;
     bool has_in_button_scope(const FlyString& tag_name) const;
     bool has_in_table_scope(const FlyString& tag_name) const;
+    bool has_in_list_item_scope(const FlyString& tag_name) const;
 
     bool has_in_scope(const Element&) const;
 
@@ -59,6 +60,8 @@ public:
 
     void pop_until_an_element_with_tag_name_has_been_popped(const FlyString&);
 
+    Element* topmost_special_node_below(const Element&);
+
 private:
     bool has_in_scope_impl(const FlyString& tag_name, const Vector<FlyString>&) const;
     bool has_in_scope_impl(const Element& target_node, const Vector<FlyString>&) const;