LibWeb: Avoid re-encoding response headers

isomorphic encoding a value that has already been encoded will
result in garbage data. `response_headers` is already encoded in
ISO-8859-1/latin1, we cannot use `from_string_pair`, as it triggers
ISO-8859-1/latin1 encoding.

Follow-up of https://github.com/LadybirdBrowser/ladybird/pull/1893

Libraries/LibWeb/Fetch/Fetching/Fetching.cpp

@@ -2330,7 +2330,7 @@ WebIDL::ExceptionOr<GC::Ref<PendingResponse>> nonstandard_resource_loader_file_o
             }
 
             for (auto const& [name, value] : response_headers.headers()) {
-                auto header = Infrastructure::Header::from_string_pair(name, value);
+                auto header = Infrastructure::Header::from_latin1_pair(name, value);
                 response->header_list()->append(move(header));
             }
 
@@ -2396,7 +2396,7 @@ WebIDL::ExceptionOr<GC::Ref<PendingResponse>> nonstandard_resource_loader_file_o
             response->set_status(status_code.value_or(200));
             response->set_body(move(body));
             for (auto const& [name, value] : response_headers.headers()) {
-                auto header = Infrastructure::Header::from_string_pair(name, value);
+                auto header = Infrastructure::Header::from_latin1_pair(name, value);
                 response->header_list()->append(move(header));
             }
 
@@ -2421,7 +2421,7 @@ WebIDL::ExceptionOr<GC::Ref<PendingResponse>> nonstandard_resource_loader_file_o
                 auto [body, _] = TRY_OR_IGNORE(extract_body(realm, data));
                 response->set_body(move(body));
                 for (auto const& [name, value] : response_headers.headers()) {
-                    auto header = Infrastructure::Header::from_string_pair(name, value);
+                    auto header = Infrastructure::Header::from_latin1_pair(name, value);
                     response->header_list()->append(move(header));
                 }
 

Libraries/LibWeb/Fetch/Infrastructure/HTTP/Headers.cpp

@@ -58,6 +58,14 @@ Header Header::from_string_pair(StringView name, StringView value)
     };
 }
 
+Header Header::from_latin1_pair(StringView name, StringView value)
+{
+    return Header {
+        .name = MUST(ByteBuffer::copy(name.bytes())),
+        .value = MUST(ByteBuffer::copy(value.bytes())),
+    };
+}
+
 GC::Ref<HeaderList> HeaderList::create(JS::VM& vm)
 {
     return vm.heap().allocate<HeaderList>();

Libraries/LibWeb/Fetch/Infrastructure/HTTP/Headers.h

@@ -29,6 +29,7 @@ struct Header {
 
     [[nodiscard]] static Header copy(Header const&);
     [[nodiscard]] static Header from_string_pair(StringView, StringView);
+    [[nodiscard]] static Header from_latin1_pair(StringView, StringView);
 };
 
 // https://fetch.spec.whatwg.org/#concept-header-list

Libraries/LibWeb/Infra/Strings.cpp

@@ -147,11 +147,12 @@ ErrorOr<String> to_ascii_uppercase(StringView string)
 // https://infra.spec.whatwg.org/#isomorphic-encode
 ByteBuffer isomorphic_encode(StringView input)
 {
+    // To isomorphic encode an isomorphic string input: return a byte sequence whose length is equal to input’s code
+    // point length and whose bytes have the same values as the values of input’s code points, in the same order.
+    // NOTE: This is essentially spec-speak for "Encode as ISO-8859-1 / Latin-1".
     ByteBuffer buf = {};
     for (auto code_point : Utf8View { input }) {
-        // VERIFY(code_point <= 0xFF);
-        if (code_point > 0xFF)
-            dbgln("FIXME: Trying to isomorphic encode a string with code points > U+00FF.");
+        VERIFY(code_point <= 0xFF);
         buf.append((u8)code_point);
     }
     return buf;
@@ -160,6 +161,9 @@ ByteBuffer isomorphic_encode(StringView input)
 // https://infra.spec.whatwg.org/#isomorphic-decode
 String isomorphic_decode(ReadonlyBytes input)
 {
+    // To isomorphic decode a byte sequence input, return a string whose code point length is equal
+    // to input’s length and whose code points have the same values as the values of input’s bytes, in the same order.
+    // NOTE: This is essentially spec-speak for "Decode as ISO-8859-1 / Latin-1".
     StringBuilder builder(input.size());
     for (u8 code_point : input) {
         builder.append_code_point(code_point);

Tests/LibWeb/Text/expected/http-non-ascii-content-type.txt

@@ -1 +1,2 @@
+Content-Type:text/html;test=ÿ;charset=gbk
 PASS (didn't crash)

Tests/LibWeb/Text/input/http-non-ascii-content-type.html

@@ -12,7 +12,8 @@
                 },
             });
 
-            const blob = await fetch(url).then((response) => response.blob());
+            const headers = await fetch(url).then((response) => response.headers);
+            println("Content-Type:" + headers.get("Content-Type"));
             println("PASS (didn't crash)");
         } catch (err) {
             println("FAIL - " + err);