From b49b1b35e4d19c4171004b7eddb6eeac91f00a97 Mon Sep 17 00:00:00 2001
From: Shannon Booth <shannon@serenityos.org>
Date: Sat, 28 Jun 2025 22:19:33 +1200
Subject: [PATCH] LibURL: Correct logic for domains not matched by PSL in
 public_suffix

For the AO defined in the URL specification, in the case the
domain does not match against the PSL, we should be returning
the TLD. This fixes a crash for a bunch of WPT tests using the
Document.domain setter when the test is being served by WPT
locally.

We should be doing similar logic in registrable_domain, but that
unfortunately runs into some other issues, so just leave a FIXME
for now.
---
 Libraries/LibURL/Host.cpp | 19 ++++++++++------
 Tests/LibURL/TestURL.cpp  | 48 +++++++++++++++++++++++++++++++++++++++
 2 files changed, 60 insertions(+), 7 deletions(-)

diff --git a/Libraries/LibURL/Host.cpp b/Libraries/LibURL/Host.cpp
index d0b8954b8d1..2f6c1b2d833 100644
--- a/Libraries/LibURL/Host.cpp
+++ b/Libraries/LibURL/Host.cpp
@@ -192,13 +192,19 @@ Optional<String> Host::public_suffix() const
     auto trailing_dot = host_string.ends_with('.') ? "."sv : ""sv;
 
     // 3. Let publicSuffix be the public suffix determined by running the Public Suffix List algorithm with host as domain. [PSL]
-    // NOTE: The spec algorithm for the public suffix returns "*" by default, but get_public_suffix() returns an empty Optional.
-    //       Remove the `value_or()` if and when we update it.
-    auto public_suffix = PublicSuffixData::the()->get_public_suffix(host_string).value_or("*"_string);
+    // FIXME: Unify this logic with registrable domain.
+    auto public_suffix = PublicSuffixData::the()->get_public_suffix(host_string);
+    if (!public_suffix.has_value()) {
+        auto last_dot = host_string.bytes_as_string_view().find_last('.');
+        if (last_dot.has_value())
+            public_suffix = MUST(host_string.substring_from_byte_offset(last_dot.value() + 1));
+        else
+            public_suffix = host_string;
+    }
 
     // 4. Assert: publicSuffix is an ASCII string that does not end with ".".
-    VERIFY(public_suffix.is_ascii());
-    VERIFY(!public_suffix.ends_with('.'));
+    VERIFY(public_suffix->is_ascii());
+    VERIFY(!public_suffix->ends_with('.'));
 
     // 5. Return publicSuffix and trailingDot concatenated.
     return MUST(String::formatted("{}{}", public_suffix, trailing_dot));
@@ -219,8 +225,7 @@ Optional<String> Host::registrable_domain() const
     auto trailing_dot = host_string.ends_with('.') ? "."sv : ""sv;
 
     // 3. Let registrableDomain be the registrable domain determined by running the Public Suffix List algorithm with host as domain. [PSL]
-    // NOTE: The spec algorithm for the public suffix returns "*" by default, but get_public_suffix() returns an empty Optional.
-    //       Remove the `value_or()` if and when we update it.
+    // FIXME: This is not correct, we should be doing the same as public_suffix() above.
     auto registrable_domain = get_registrable_domain(host_string).value_or("*"_string);
 
     // 4. Assert: registrableDomain is an ASCII string that does not end with ".".
diff --git a/Tests/LibURL/TestURL.cpp b/Tests/LibURL/TestURL.cpp
index 96583f13afa..7cf1982468f 100644
--- a/Tests/LibURL/TestURL.cpp
+++ b/Tests/LibURL/TestURL.cpp
@@ -641,3 +641,51 @@ TEST_CASE(get_registrable_domain)
         EXPECT_EQ(*domain, "ladybird.github.io"sv);
     }
 }
+
+TEST_CASE(public_suffix)
+{
+    {
+        auto domain = URL::Parser::parse_host("com"sv);
+        EXPECT_EQ(domain->public_suffix(), "com"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("example.com"sv);
+        EXPECT_EQ(domain->public_suffix(), "com"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("www.example.com"sv);
+        EXPECT_EQ(domain->public_suffix(), "com"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("EXAMPLE.COM"sv);
+        EXPECT_EQ(domain->public_suffix(), "com"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("www.example.com."sv);
+        EXPECT_EQ(domain->public_suffix(), "com."sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("github.io"sv);
+        EXPECT_EQ(domain->public_suffix(), "github.io"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("whatwg.github.io"sv);
+        EXPECT_EQ(domain->public_suffix(), "github.io"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("إختبار"sv);
+        EXPECT_EQ(domain->public_suffix(), "xn--kgbechtv"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("example.إختبار"sv);
+        EXPECT_EQ(domain->public_suffix(), "xn--kgbechtv"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("sub.example.إختبار"sv);
+        EXPECT_EQ(domain->public_suffix(), "xn--kgbechtv"sv);
+    }
+    {
+        auto domain = URL::Parser::parse_host("[2001:0db8:85a3:0000:0000:8a2e:0370:7334]"sv);
+        EXPECT_EQ(domain->public_suffix(), OptionalNone {});
+    }
+}