Implement UTF-8 decode without BOM in TextResourceDecoder

This adds no BOM decoding option to TextResourceDecoderOptions which
disabled BOM checking in the decoder.

Bug: 796192
Change-Id: Id0eb00dca451c4898d9ae00f11bf08e7fbb33a1c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1796366Reviewed-by: Yutaka Hirano <yhirano@chromium.org>
Reviewed-by: Hiroshige Hayashizaki <hiroshige@chromium.org>
Commit-Queue: Eero Häkkinen <eero.hakkinen@intel.com>
Cr-Commit-Position: refs/heads/master@{#711196}

third_party/blink/renderer/core/html/parser/text_resource_decoder.cc

@@ -211,6 +211,11 @@ wtf_size_t TextResourceDecoder::CheckForBOM(const char* data, wtf_size_t len) {
   // respectively.
   DCHECK(!checked_for_bom_);
 
+  if (options_.GetNoBOMDecoding()) {
+    checked_for_bom_ = true;
+    return 0;
+  }
+
   wtf_size_t length_of_bom = 0;
   const wtf_size_t max_bom_length = 3;
 

third_party/blink/renderer/core/html/parser/text_resource_decoder_test.cc

@@ -8,6 +8,33 @@
 
 namespace blink {
 
+TEST(TextResourceDecoderTest, AlwaysUseUTF8) {
+  std::unique_ptr<TextResourceDecoder> decoder =
+      std::make_unique<TextResourceDecoder>(
+          TextResourceDecoderOptions::CreateAlwaysUseUTF8ForText());
+  const unsigned char kFooUTF8WithBOM[] = {0xef, 0xbb, 0xbf, 0x66, 0x6f, 0x6f};
+  WTF::String decoded = decoder->Decode(
+      reinterpret_cast<const char*>(kFooUTF8WithBOM), sizeof(kFooUTF8WithBOM));
+  decoded = decoded + decoder->Flush();
+  EXPECT_EQ(WTF::UTF8Encoding(), decoder->Encoding());
+  EXPECT_EQ("foo", decoded);
+}
+
+TEST(TextResourceDecoderTest, AlwaysUseUTF8WithoutBOM) {
+  std::unique_ptr<TextResourceDecoder> decoder =
+      std::make_unique<TextResourceDecoder>(
+          TextResourceDecoderOptions::CreateAlwaysUseUTF8WithoutBOMForText());
+  const unsigned char kFooUTF8WithBOM[] = {0xef, 0xbb, 0xbf, 0x66, 0x6f, 0x6f};
+  WTF::String decoded = decoder->Decode(
+      reinterpret_cast<const char*>(kFooUTF8WithBOM), sizeof(kFooUTF8WithBOM));
+  decoded = decoded + decoder->Flush();
+  EXPECT_EQ(WTF::UTF8Encoding(), decoder->Encoding());
+  EXPECT_EQ(
+      "\xef\xbb\xbf"
+      "foo",
+      decoded.Utf8());
+}
+
 TEST(TextResourceDecoderTest, BasicUTF16) {
   std::unique_ptr<TextResourceDecoder> decoder =
       std::make_unique<TextResourceDecoder>(TextResourceDecoderOptions(

third_party/blink/renderer/platform/loader/fetch/text_resource_decoder_options.cc

@@ -22,6 +22,13 @@ TextResourceDecoderOptions::CreateAlwaysUseUTF8ForText() {
                                     UTF8Encoding(), nullptr, NullURL());
 }
 
+TextResourceDecoderOptions
+TextResourceDecoderOptions::CreateAlwaysUseUTF8WithoutBOMForText() {
+  TextResourceDecoderOptions options = CreateAlwaysUseUTF8ForText();
+  options.no_bom_decoding_ = true;
+  return options;
+}
+
 TextResourceDecoderOptions TextResourceDecoderOptions::CreateWithAutoDetection(
     ContentType content_type,
     const WTF::TextEncoding& default_encoding,
@@ -41,6 +48,7 @@ TextResourceDecoderOptions::TextResourceDecoderOptions(
     : encoding_detection_option_(encoding_detection_option),
       content_type_(content_type),
       default_encoding_(default_encoding),
+      no_bom_decoding_(false),
       use_lenient_xml_decoding_(false),
       hint_encoding_(hint_encoding),
       hint_url_(hint_url) {

third_party/blink/renderer/platform/loader/fetch/text_resource_decoder_options.h

@@ -40,6 +40,10 @@ class PLATFORM_EXPORT TextResourceDecoderOptions final {
   // https://encoding.spec.whatwg.org/#utf-8-decode.
   static TextResourceDecoderOptions CreateAlwaysUseUTF8ForText();
 
+  // Corresponds to utf-8 decode without BOM in Encoding spec:
+  // https://encoding.spec.whatwg.org/#utf-8-decode-without-bom.
+  static TextResourceDecoderOptions CreateAlwaysUseUTF8WithoutBOMForText();
+
   static TextResourceDecoderOptions CreateWithAutoDetection(
       ContentType,
       const WTF::TextEncoding& default_encoding,
@@ -79,6 +83,7 @@ class PLATFORM_EXPORT TextResourceDecoderOptions final {
   }
   ContentType GetContentType() const { return content_type_; }
   const WTF::TextEncoding& DefaultEncoding() const { return default_encoding_; }
+  bool GetNoBOMDecoding() const { return no_bom_decoding_; }
   bool GetUseLenientXMLDecoding() const { return use_lenient_xml_decoding_; }
 
   const char* HintEncoding() const { return hint_encoding_; }
@@ -95,6 +100,7 @@ class PLATFORM_EXPORT TextResourceDecoderOptions final {
   EncodingDetectionOption encoding_detection_option_;
   ContentType content_type_;
   WTF::TextEncoding default_encoding_;
+  bool no_bom_decoding_;
   bool use_lenient_xml_decoding_;  // Don't stop on XML decoding errors.
 
   // Hints for DetectTextEncoding().