AK: Decode data URLs to separate class (and parse like every other URL)

Parsing 'data:' URLs took it's own route. It never set standard URL
fields like path, query or fragment (except for scheme) and instead
gave us separate methods called `data_payload()`, `data_mime_type()`,
and `data_payload_is_base64()`.

Because parsing 'data:' didn't use standard fields, running the
following JS code:

    new URL('#a', 'data:text/plain,hello').toString()

not only cleared the path as URLParser doesn't check for data from
data_payload() function (making the result be 'data:#a'), but it also
crashes the program because we forbid having an empty MIME type when we
serialize to string.

With this change, 'data:' URLs will be parsed like every other URLs.
To decode the 'data:' URL contents, one needs to call process_data_url()
on a URL, which will return a struct containing MIME type with already
decoded data! :^)

AK/URL.cpp

@@ -5,6 +5,7 @@
  * SPDX-License-Identifier: BSD-2-Clause
  */
 
+#include <AK/Base64.h>
 #include <AK/CharacterTypes.h>
 #include <AK/Debug.h>
 #include <AK/LexicalPath.h>
@@ -173,18 +174,7 @@ bool URL::compute_validity() const
     if (m_scheme.is_empty())
         return false;
 
-    if (m_scheme == "data") {
-        if (m_data_mime_type.is_empty())
-            return false;
-        if (m_data_payload_is_base64) {
-            if (m_data_payload.length() % 4 != 0)
-                return false;
-            for (auto character : m_data_payload) {
-                if (!is_ascii_alphanumeric(character) || character == '+' || character == '/' || character == '=')
-                    return false;
-            }
-        }
-    } else if (m_cannot_be_a_base_url) {
+    if (m_cannot_be_a_base_url) {
         if (m_paths.size() != 1)
             return false;
         if (m_paths[0].is_empty())
@@ -275,6 +265,22 @@ URL URL::create_with_url_or_path(DeprecatedString const& url_or_path)
     return URL::create_with_file_scheme(path);
 }
 
+URL URL::create_with_data(StringView mime_type, StringView payload, bool is_base64)
+{
+    URL url;
+    url.set_cannot_be_a_base_url(true);
+    url.set_scheme("data"sv);
+
+    StringBuilder builder;
+    builder.append(mime_type);
+    if (is_base64)
+        builder.append(";base64"sv);
+    builder.append(',');
+    builder.append(payload);
+    url.set_paths({ builder.to_deprecated_string() });
+    return url;
+}
+
 // https://url.spec.whatwg.org/#special-scheme
 bool URL::is_special_scheme(StringView scheme)
 {
@@ -293,30 +299,9 @@ DeprecatedString URL::serialize_path(ApplyPercentDecoding apply_percent_decoding
     return builder.to_deprecated_string();
 }
 
-DeprecatedString URL::serialize_data_url() const
-{
-    VERIFY(m_scheme == "data");
-    VERIFY(!m_data_mime_type.is_null());
-    VERIFY(!m_data_payload.is_null());
-    StringBuilder builder;
-    builder.append(m_scheme);
-    builder.append(':');
-    builder.append(m_data_mime_type);
-    if (m_data_payload_is_base64)
-        builder.append(";base64"sv);
-    builder.append(',');
-    // NOTE: The specification does not say anything about encoding this, but we should encode at least control and non-ASCII
-    //       characters (since this is also a valid representation of the same data URL).
-    builder.append(URL::percent_encode(m_data_payload, PercentEncodeSet::C0Control));
-    return builder.to_deprecated_string();
-}
-
 // https://url.spec.whatwg.org/#concept-url-serializer
 DeprecatedString URL::serialize(ExcludeFragment exclude_fragment) const
 {
-    if (m_scheme == "data")
-        return serialize_data_url();
-
     // 1. Let output be url’s scheme and U+003A (:) concatenated.
     StringBuilder output;
     output.append(m_scheme);
@@ -387,8 +372,7 @@ DeprecatedString URL::serialize(ExcludeFragment exclude_fragment) const
 DeprecatedString URL::serialize_for_display() const
 {
     VERIFY(m_valid);
-    if (m_scheme == "data")
-        return serialize_data_url();
+
     StringBuilder builder;
     builder.append(m_scheme);
     builder.append(':');
@@ -466,6 +450,80 @@ bool URL::equals(URL const& other, ExcludeFragment exclude_fragments) const
     return serialize(exclude_fragments) == other.serialize(exclude_fragments);
 }
 
+// https://fetch.spec.whatwg.org/#data-url-processor
+ErrorOr<URL::DataURL> URL::process_data_url() const
+{
+    // 1. Assert: dataURL’s scheme is "data".
+    VERIFY(scheme() == "data");
+
+    // 2. Let input be the result of running the URL serializer on dataURL with exclude fragment set to true.
+    auto input = serialize(URL::ExcludeFragment::Yes);
+
+    // 3. Remove the leading "data:" from input.
+    input = input.substring("data:"sv.length());
+
+    // 4. Let position point at the start of input.
+
+    // 5. Let mimeType be the result of collecting a sequence of code points that are not equal to U+002C (,), given position.
+    auto position = input.find(',');
+    auto mime_type = input.substring_view(0, position.value_or(input.length()));
+
+    // 6. Strip leading and trailing ASCII whitespace from mimeType.
+    mime_type = mime_type.trim_whitespace(TrimMode::Both);
+
+    // 7. If position is past the end of input, then return failure.
+    if (!position.has_value())
+        return Error::from_string_literal("Missing a comma character");
+
+    // 8. Advance position by 1.
+    position = position.value() + 1;
+
+    // 9. Let encodedBody be the remainder of input.
+    auto encoded_body = input.substring_view(position.value());
+
+    // 10. Let body be the percent-decoding of encodedBody.
+    auto body = URL::percent_decode(encoded_body).to_byte_buffer();
+
+    // 11. If mimeType ends with U+003B (;), followed by zero or more U+0020 SPACE, followed by an ASCII case-insensitive match for "base64", then:
+    if (mime_type.ends_with("base64"sv, CaseSensitivity::CaseInsensitive)) {
+        auto trimmed_substring_view = mime_type.substring_view(0, mime_type.length() - 6);
+        trimmed_substring_view = trimmed_substring_view.trim(" "sv, TrimMode::Right);
+        if (trimmed_substring_view.ends_with(';')) {
+            // 1. Let stringBody be the isomorphic decode of body.
+            auto string_body = StringView(body);
+
+            // 2. Set body to the forgiving-base64 decode of stringBody.
+            //    FIXME: Check if it's really forgiving.
+            // 3. If body is failure, then return failure.
+            body = TRY(decode_base64(string_body));
+
+            // 4. Remove the last 6 code points from mimeType.
+            // 5. Remove trailing U+0020 SPACE code points from mimeType, if any.
+            // 6. Remove the last U+003B (;) from mimeType.
+            mime_type = trimmed_substring_view.substring_view(0, trimmed_substring_view.length() - 1);
+        }
+    }
+
+    // 12. If mimeType starts with ";", then prepend "text/plain" to mimeType.
+    StringBuilder builder;
+    if (mime_type.starts_with(';')) {
+        builder.append("text/plain"sv);
+        builder.append(mime_type);
+        mime_type = builder.string_view();
+    }
+
+    // FIXME: Parse the MIME type's components according to https://mimesniff.spec.whatwg.org/#parse-a-mime-type
+    // FIXME: 13. Let mimeTypeRecord be the result of parsing mimeType.
+    auto mime_type_record = mime_type.trim("\n\r\t "sv, TrimMode::Both);
+
+    // 14. If mimeTypeRecord is failure, then set mimeTypeRecord to text/plain;charset=US-ASCII.
+    if (mime_type_record.is_empty())
+        mime_type_record = "text/plain;charset=US-ASCII"sv;
+
+    // 15. Return a new data: URL struct whose MIME type is mimeTypeRecord and body is body.
+    return URL::DataURL { TRY(String::from_utf8(mime_type_record)), body };
+}
+
 void URL::append_percent_encoded(StringBuilder& builder, u32 code_point)
 {
     if (code_point <= 0x7f)

AK/URL.h

@@ -128,14 +128,16 @@ public:
 
     URL complete_url(StringView) const;
 
-    bool data_payload_is_base64() const { return m_data_payload_is_base64; }
-    DeprecatedString const& data_mime_type() const { return m_data_mime_type; }
-    DeprecatedString const& data_payload() const { return m_data_payload; }
+    struct DataURL {
+        String mime_type;
+        ByteBuffer body;
+    };
+    ErrorOr<DataURL> process_data_url() const;
 
     static URL create_with_url_or_path(DeprecatedString const&);
     static URL create_with_file_scheme(DeprecatedString const& path, DeprecatedString const& fragment = {}, DeprecatedString const& hostname = {});
     static URL create_with_help_scheme(DeprecatedString const& path, DeprecatedString const& fragment = {}, DeprecatedString const& hostname = {});
-    static URL create_with_data(DeprecatedString mime_type, DeprecatedString payload, bool is_base64 = false) { return URL(move(mime_type), move(payload), is_base64); }
+    static URL create_with_data(StringView mime_type, StringView payload, bool is_base64 = false);
 
     static u16 default_port_for_scheme(StringView);
     static bool is_special_scheme(StringView);
@@ -152,17 +154,7 @@ public:
     static bool code_point_is_in_percent_encode_set(u32 code_point, URL::PercentEncodeSet);
 
 private:
-    URL(DeprecatedString&& data_mime_type, DeprecatedString&& data_payload, bool payload_is_base64)
-        : m_valid(true)
-        , m_scheme("data")
-        , m_data_payload_is_base64(payload_is_base64)
-        , m_data_mime_type(move(data_mime_type))
-        , m_data_payload(move(data_payload))
-    {
-    }
-
     bool compute_validity() const;
-    DeprecatedString serialize_data_url() const;
 
     static void append_percent_encoded_if_necessary(StringBuilder&, u32 code_point, PercentEncodeSet set = PercentEncodeSet::Userinfo);
     static void append_percent_encoded(StringBuilder&, u32 code_point);
@@ -196,10 +188,6 @@ private:
     DeprecatedString m_fragment;
 
     bool m_cannot_be_a_base_url { false };
-
-    bool m_data_payload_is_base64 { false };
-    DeprecatedString m_data_mime_type;
-    DeprecatedString m_data_payload;
 };
 
 template<>

AK/URLParser.cpp

@@ -697,42 +697,6 @@ DeprecatedString URLParser::percent_encode_after_encoding(StringView input, URL:
     return output.to_deprecated_string();
 }
 
-// https://fetch.spec.whatwg.org/#data-urls
-// FIXME: This only loosely follows the spec, as we use the same class for "regular" and data URLs, unlike the spec.
-Optional<URL> URLParser::parse_data_url(StringView raw_input)
-{
-    dbgln_if(URL_PARSER_DEBUG, "URLParser::parse_data_url: Parsing '{}'.", raw_input);
-    VERIFY(raw_input.starts_with("data:"sv));
-    auto input = raw_input.substring_view(5);
-    auto comma_offset = input.find(',');
-    if (!comma_offset.has_value())
-        return {};
-    auto mime_type = StringUtils::trim(input.substring_view(0, comma_offset.value()), "\t\n\f\r "sv, TrimMode::Both);
-    auto encoded_body = input.substring_view(comma_offset.value() + 1);
-    auto body = URL::percent_decode(encoded_body);
-    bool is_base64_encoded = false;
-    if (mime_type.ends_with("base64"sv, CaseSensitivity::CaseInsensitive)) {
-        auto substring_view = mime_type.substring_view(0, mime_type.length() - 6);
-        auto trimmed_substring_view = StringUtils::trim(substring_view, " "sv, TrimMode::Right);
-        if (trimmed_substring_view.ends_with(';')) {
-            is_base64_encoded = true;
-            mime_type = trimmed_substring_view.substring_view(0, trimmed_substring_view.length() - 1);
-        }
-    }
-
-    StringBuilder builder;
-    if (mime_type.starts_with(";"sv) || mime_type.is_empty()) {
-        builder.append("text/plain"sv);
-        builder.append(mime_type);
-        mime_type = builder.string_view();
-    }
-
-    // FIXME: Parse the MIME type's components according to https://mimesniff.spec.whatwg.org/#parse-a-mime-type
-    URL url { StringUtils::trim(mime_type, "\n\r\t "sv, TrimMode::Both), move(body), is_base64_encoded };
-    dbgln_if(URL_PARSER_DEBUG, "URLParser::parse_data_url: Parsed data URL to be '{}'.", url.serialize());
-    return url;
-}
-
 // https://url.spec.whatwg.org/#concept-basic-url-parser
 // NOTE: This parser assumes a UTF-8 encoding.
 // NOTE: Refrain from using the URL classes setters inside this algorithm. Rather, set the values directly. This bypasses the setters' built-in
@@ -746,13 +710,6 @@ URL URLParser::basic_parse(StringView raw_input, Optional<URL> const& base_url,
     if (raw_input.is_empty())
         return base_url.has_value() ? *base_url : URL {};
 
-    if (raw_input.starts_with("data:"sv)) {
-        auto maybe_url = parse_data_url(raw_input);
-        if (!maybe_url.has_value())
-            return {};
-        return maybe_url.release_value();
-    }
-
     size_t start_index = 0;
     size_t end_index = raw_input.length();
 

AK/URLParser.h

@@ -63,9 +63,6 @@ public:
 
     // https://url.spec.whatwg.org/#concept-host-serializer
     static ErrorOr<String> serialize_host(URL::Host const&);
-
-private:
-    static Optional<URL> parse_data_url(StringView raw_input);
 };
 
 #undef ENUMERATE_STATES

Tests/AK/TestURL.cpp

@@ -7,7 +7,6 @@
 
 #include <LibTest/TestCase.h>
 
-#include <AK/Base64.h>
 #include <AK/URL.h>
 #include <AK/URLParser.h>
 
@@ -111,7 +110,6 @@ TEST_CASE(some_bad_urls)
     EXPECT_EQ(URL("http://serenityos.org:abc"sv).is_valid(), false);
     EXPECT_EQ(URL("http://serenityos.org:abc:80"sv).is_valid(), false);
     EXPECT_EQ(URL("http://serenityos.org:abc:80/"sv).is_valid(), false);
-    EXPECT_EQ(URL("data:"sv).is_valid(), false);
 }
 
 TEST_CASE(serialization)
@@ -246,10 +244,11 @@ TEST_CASE(data_url)
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/html");
-    EXPECT_EQ(url.data_payload(), "test");
-    EXPECT(!url.data_payload_is_base64());
     EXPECT_EQ(url.serialize(), "data:text/html,test");
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/html");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "test"sv);
 }
 
 TEST_CASE(data_url_default_mime_type)
@@ -258,10 +257,11 @@ TEST_CASE(data_url_default_mime_type)
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/plain");
-    EXPECT_EQ(url.data_payload(), "test");
-    EXPECT(!url.data_payload_is_base64());
-    EXPECT_EQ(url.serialize(), "data:text/plain,test");
+    EXPECT_EQ(url.serialize(), "data:,test");
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/plain;charset=US-ASCII");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "test"sv);
 }
 
 TEST_CASE(data_url_encoded)
@@ -270,46 +270,50 @@ TEST_CASE(data_url_encoded)
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/html");
-    EXPECT_EQ(url.data_payload(), "Hello friends,%0X%X0");
-    EXPECT(!url.data_payload_is_base64());
-    EXPECT_EQ(url.serialize(), "data:text/html,Hello friends,%0X%X0");
+    EXPECT_EQ(url.serialize(), "data:text/html,Hello%20friends%2C%0X%X0");
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/html");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "Hello friends,%0X%X0"sv);
 }
 
 TEST_CASE(data_url_base64_encoded)
 {
-    URL url("data:text/html;base64,test"sv);
+    URL url("data:text/html;base64,dGVzdA=="sv);
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/html");
-    EXPECT_EQ(url.data_payload(), "test");
-    EXPECT(url.data_payload_is_base64());
-    EXPECT_EQ(url.serialize(), "data:text/html;base64,test");
+    EXPECT_EQ(url.serialize(), "data:text/html;base64,dGVzdA==");
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/html");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "test"sv);
 }
 
 TEST_CASE(data_url_base64_encoded_default_mime_type)
 {
-    URL url("data:;base64,test"sv);
+    URL url("data:;base64,dGVzdA=="sv);
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/plain");
-    EXPECT_EQ(url.data_payload(), "test");
-    EXPECT(url.data_payload_is_base64());
-    EXPECT_EQ(url.serialize(), "data:text/plain;base64,test");
+    EXPECT_EQ(url.serialize(), "data:;base64,dGVzdA==");
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/plain;charset=US-ASCII");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "test"sv);
 }
 
 TEST_CASE(data_url_base64_encoded_with_whitespace)
 {
-    URL url("data: text/html ;     bAsE64 , test with whitespace "sv);
+    URL url("data: text/html ;     bAsE64 , dGVz dA== "sv);
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/html");
-    EXPECT_EQ(url.data_payload(), " test with whitespace ");
-    EXPECT(url.data_payload_is_base64());
-    EXPECT_EQ(url.serialize(), "data:text/html;base64, test with whitespace ");
+    EXPECT_EQ(url.serialize(), "data: text/html ;     bAsE64 , dGVz dA==");
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/html");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "test");
 }
 
 TEST_CASE(data_url_base64_encoded_with_inline_whitespace)
@@ -318,12 +322,23 @@ TEST_CASE(data_url_base64_encoded_with_inline_whitespace)
     EXPECT(url.is_valid());
     EXPECT_EQ(url.scheme(), "data");
     EXPECT(url.host().has<Empty>());
-    EXPECT_EQ(url.data_mime_type(), "text/javascript");
-    EXPECT(url.data_payload_is_base64());
-    EXPECT_EQ(url.data_payload(), " ZD Qg\r\nPS An Zm91cic\r\n 7 "sv);
-    auto decode_result = decode_base64(url.data_payload());
-    EXPECT_EQ(decode_result.is_error(), false);
-    EXPECT_EQ(StringView(decode_result.value()), "d4 = 'four';"sv);
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/javascript");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "d4 = 'four';"sv);
+}
+
+TEST_CASE(data_url_completed_with_fragment)
+{
+    auto url = URL("data:text/plain,test"sv).complete_url("#a"sv);
+    EXPECT(url.is_valid());
+    EXPECT_EQ(url.scheme(), "data");
+    EXPECT_EQ(url.fragment(), "a");
+    EXPECT(url.host().has<Empty>());
+
+    auto data_url = TRY_OR_FAIL(url.process_data_url());
+    EXPECT_EQ(data_url.mime_type, "text/plain");
+    EXPECT_EQ(StringView(data_url.body.bytes()), "test"sv);
 }
 
 TEST_CASE(trailing_slash_with_complete_url)

Userland/Applications/Spreadsheet/HelpWindow.cpp

@@ -124,7 +124,7 @@ HelpWindow::HelpWindow(GUI::Window* parent)
             window->show();
         } else if (url.host() == String::from_utf8_short_string("doc"sv)) {
             auto entry = LexicalPath::basename(url.serialize_path());
-            m_webview->load(URL::create_with_data("text/html", render(entry)));
+            m_webview->load(URL::create_with_data("text/html"sv, render(entry)));
         } else {
             dbgln("Invalid spreadsheet action domain '{}'", url.serialized_host().release_value_but_fixme_should_propagate_errors());
         }
@@ -135,7 +135,7 @@ HelpWindow::HelpWindow(GUI::Window* parent)
             return;
 
         auto key = static_cast<HelpListModel*>(m_listview->model())->key(index);
-        m_webview->load(URL::create_with_data("text/html", render(key)));
+        m_webview->load(URL::create_with_data("text/html"sv, render(key)));
     };
 }
 

Userland/Libraries/LibWeb/Fetch/Fetching/Fetching.cpp

@@ -714,17 +714,15 @@ WebIDL::ExceptionOr<JS::NonnullGCPtr<PendingResponse>> scheme_fetch(JS::Realm& r
     // -> "data"
     else if (request->current_url().scheme() == "data"sv) {
         // 1. Let dataURLStruct be the result of running the data: URL processor on request’s current URL.
-        auto const& url = request->current_url();
-        auto data_or_error = url.data_payload_is_base64()
-            ? decode_base64(url.data_payload())
-            : TRY_OR_THROW_OOM(vm, ByteBuffer::copy(url.data_payload().bytes()));
+        auto data_url_struct = request->current_url().process_data_url();
 
         // 2. If dataURLStruct is failure, then return a network error.
-        if (data_or_error.is_error())
-            return PendingResponse::create(vm, request, Infrastructure::Response::network_error(vm, "Request has invalid base64 'data:' URL"sv));
+        if (data_url_struct.is_error())
+            return PendingResponse::create(vm, request, Infrastructure::Response::network_error(vm, "Failed to process 'data:' URL"sv));
 
         // 3. Let mimeType be dataURLStruct’s MIME type, serialized.
-        auto const& mime_type = url.data_mime_type();
+        //    FIXME: Serialize MIME type.
+        auto const& mime_type = data_url_struct.value().mime_type;
 
         // 4. Return a new response whose status message is `OK`, header list is « (`Content-Type`, mimeType) », and
         //    body is dataURLStruct’s body as a body.
@@ -732,7 +730,7 @@ WebIDL::ExceptionOr<JS::NonnullGCPtr<PendingResponse>> scheme_fetch(JS::Realm& r
         response->set_status_message(MUST(ByteBuffer::copy("OK"sv.bytes())));
         auto header = TRY_OR_THROW_OOM(vm, Infrastructure::Header::from_string_pair("Content-Type"sv, mime_type));
         TRY_OR_THROW_OOM(vm, response->header_list()->append(move(header)));
-        response->set_body(TRY(Infrastructure::byte_sequence_as_body(realm, data_or_error.value().span())));
+        response->set_body(TRY(Infrastructure::byte_sequence_as_body(realm, data_url_struct.value().body)));
         return PendingResponse::create(vm, request, response);
     }
     // -> "file"

Userland/Libraries/LibWeb/HTML/HTMLCanvasElement.cpp

@@ -195,7 +195,7 @@ DeprecatedString HTMLCanvasElement::to_data_url(DeprecatedString const& type, [[
         // FIXME: propagate error
         return {};
     }
-    return AK::URL::create_with_data(type, base64_encoded_or_error.release_value().to_deprecated_string(), true).to_deprecated_string();
+    return AK::URL::create_with_data(type, base64_encoded_or_error.release_value(), true).to_deprecated_string();
 }
 
 void HTMLCanvasElement::present()

Userland/Libraries/LibWeb/Loader/ResourceLoader.cpp

@@ -5,7 +5,6 @@
  * SPDX-License-Identifier: BSD-2-Clause
  */
 
-#include <AK/Base64.h>
 #include <AK/Debug.h>
 #include <AK/JsonObject.h>
 #include <LibCore/ElapsedTimer.h>
@@ -123,7 +122,7 @@ RefPtr<Resource> ResourceLoader::load_resource(Resource::Type type, LoadRequest&
 static DeprecatedString sanitized_url_for_logging(AK::URL const& url)
 {
     if (url.scheme() == "data"sv)
-        return DeprecatedString::formatted("[data URL, mime-type={}, size={}]", url.data_mime_type(), url.data_payload().length());
+        return "[data URL]"sv;
     return url.to_deprecated_string();
 }
 
@@ -204,30 +203,25 @@ void ResourceLoader::load(LoadRequest& request, Function<void(ReadonlyBytes, Has
     }
 
     if (url.scheme() == "data") {
-        dbgln_if(SPAM_DEBUG, "ResourceLoader loading a data URL with mime-type: '{}', base64={}, payload='{}'",
-            url.data_mime_type(),
-            url.data_payload_is_base64(),
-            url.data_payload());
-
-        ByteBuffer data;
-        if (url.data_payload_is_base64()) {
-            auto data_maybe = decode_base64(url.data_payload());
-            if (data_maybe.is_error()) {
-                auto error_message = data_maybe.error().string_literal();
-                log_failure(request, error_message);
-                error_callback(error_message, {});
-                return;
-            }
-            data = data_maybe.value();
-        } else {
-            data = url.data_payload().to_byte_buffer();
+        auto data_url_or_error = url.process_data_url();
+        if (data_url_or_error.is_error()) {
+            auto error_message = data_url_or_error.error().string_literal();
+            log_failure(request, error_message);
+            error_callback(error_message, {});
+            return;
         }
+        auto data_url = data_url_or_error.release_value();
+
+        dbgln_if(SPAM_DEBUG, "ResourceLoader loading a data URL with mime-type: '{}', payload='{}'",
+            data_url.mime_type,
+            StringView(data_url.body.bytes()));
 
         HashMap<DeprecatedString, DeprecatedString, CaseInsensitiveStringTraits> response_headers;
-        response_headers.set("Content-Type", url.data_mime_type());
+        response_headers.set("Content-Type", data_url.mime_type.to_deprecated_string());
 
         log_success(request);
-        Platform::EventLoopPlugin::the().deferred_invoke([data = move(data), response_headers = move(response_headers), success_callback = move(success_callback)] {
+
+        Platform::EventLoopPlugin::the().deferred_invoke([data = move(data_url.body), response_headers = move(response_headers), success_callback = move(success_callback)] {
             success_callback(data, response_headers, {});
         });
         return;