| // Copyright 2018 The Chromium Authors |
| // Use of this source code is governed by a BSD-style license that can be |
| // found in the LICENSE file. |
| |
| #include "base/strings/string_util.h" |
| |
| #include <cinttypes> |
| |
| #include "base/features.h" |
| #include "base/strings/utf_string_conversions.h" |
| #include "base/test/scoped_feature_list.h" |
| #include "base/time/time.h" |
| #include "build/build_config.h" |
| #include "testing/gtest/include/gtest/gtest.h" |
| |
| namespace base { |
| |
| template <typename String> |
| void MeasureIsStringASCII(size_t str_length, size_t non_ascii_pos) { |
| String str(str_length, 'A'); |
| if (non_ascii_pos < str_length) { |
| str[non_ascii_pos] = '\xAF'; |
| } |
| |
| TimeTicks t0 = TimeTicks::Now(); |
| for (size_t i = 0; i < 10000000; ++i) { |
| IsStringASCII(str); |
| } |
| TimeDelta time = TimeTicks::Now() - t0; |
| printf( |
| "char-size:\t%zu\tlength:\t%zu\tnon-ascii-pos:\t%zu\ttime-ms:\t%" PRIu64 |
| "\n", |
| sizeof(typename String::value_type), str_length, non_ascii_pos, |
| time.InMilliseconds()); |
| } |
| |
| TEST(StringUtilTest, DISABLED_IsStringASCIIPerf) { |
| for (size_t str_length = 4; str_length <= 1024; str_length *= 2) { |
| for (size_t non_ascii_loc = 0; non_ascii_loc < 3; ++non_ascii_loc) { |
| size_t non_ascii_pos = str_length * non_ascii_loc / 2 + 2; |
| MeasureIsStringASCII<std::string>(str_length, non_ascii_pos); |
| MeasureIsStringASCII<std::u16string>(str_length, non_ascii_pos); |
| #if defined(WCHAR_T_IS_32_BIT) |
| MeasureIsStringASCII<std::basic_string<wchar_t>>(str_length, |
| non_ascii_pos); |
| #endif |
| } |
| } |
| } |
| |
| void MeasureUTF8ToUTF16(const std::string& name, |
| std::string_view input, |
| bool enable_feature, |
| size_t iterations = 1000000) { |
| test::ScopedFeatureList feature_list; |
| if (enable_feature) { |
| feature_list.InitAndEnableFeature(features::kUtfConversionAsciiFastPath); |
| } else { |
| feature_list.InitAndDisableFeature(features::kUtfConversionAsciiFastPath); |
| } |
| strings_internal::InitializeUtfStringConversionsFeatures(); |
| |
| std::u16string output; |
| TimeTicks t0 = TimeTicks::Now(); |
| for (size_t i = 0; i < iterations; ++i) { |
| UTF8ToUTF16(input.data(), input.length(), &output); |
| } |
| TimeDelta time = TimeTicks::Now() - t0; |
| uint64_t normalized_ms = |
| iterations == 1000000 |
| ? static_cast<uint64_t>(time.InMilliseconds()) |
| : static_cast<uint64_t>(time.InMicroseconds() * 1000000 / iterations / |
| 1000); |
| printf("UTF8ToUTF16 | mode: %-8s | dataset: %-42s | time-ms: %" PRIu64 "\n", |
| enable_feature ? "FAST" : "LEGACY", name.c_str(), normalized_ms); |
| } |
| |
| void MeasureUTF16ToUTF8(const std::string& name, |
| std::u16string_view input, |
| bool enable_feature, |
| size_t iterations = 1000000) { |
| test::ScopedFeatureList feature_list; |
| if (enable_feature) { |
| feature_list.InitAndEnableFeature(features::kUtfConversionAsciiFastPath); |
| } else { |
| feature_list.InitAndDisableFeature(features::kUtfConversionAsciiFastPath); |
| } |
| strings_internal::InitializeUtfStringConversionsFeatures(); |
| |
| std::string output; |
| TimeTicks t0 = TimeTicks::Now(); |
| for (size_t i = 0; i < iterations; ++i) { |
| UTF16ToUTF8(input.data(), input.length(), &output); |
| } |
| TimeDelta time = TimeTicks::Now() - t0; |
| uint64_t normalized_ms = |
| iterations == 1000000 |
| ? static_cast<uint64_t>(time.InMilliseconds()) |
| : static_cast<uint64_t>(time.InMicroseconds() * 1000000 / iterations / |
| 1000); |
| printf("UTF16ToUTF8 | mode: %-8s | dataset: %-42s | time-ms: %" PRIu64 "\n", |
| enable_feature ? "FAST" : "LEGACY", name.c_str(), normalized_ms); |
| } |
| |
| void MeasureUTF8ToWide(const std::string& name, |
| std::string_view input, |
| bool enable_feature, |
| size_t iterations = 1000000) { |
| test::ScopedFeatureList feature_list; |
| if (enable_feature) { |
| feature_list.InitAndEnableFeature(features::kUtfConversionAsciiFastPath); |
| } else { |
| feature_list.InitAndDisableFeature(features::kUtfConversionAsciiFastPath); |
| } |
| strings_internal::InitializeUtfStringConversionsFeatures(); |
| |
| std::wstring output; |
| TimeTicks t0 = TimeTicks::Now(); |
| for (size_t i = 0; i < iterations; ++i) { |
| UTF8ToWide(input.data(), input.length(), &output); |
| } |
| TimeDelta time = TimeTicks::Now() - t0; |
| uint64_t normalized_ms = |
| iterations == 1000000 |
| ? static_cast<uint64_t>(time.InMilliseconds()) |
| : static_cast<uint64_t>(time.InMicroseconds() * 1000000 / iterations / |
| 1000); |
| printf("UTF8ToWide | mode: %-8s | dataset: %-42s | time-ms: %" PRIu64 "\n", |
| enable_feature ? "FAST" : "LEGACY", name.c_str(), normalized_ms); |
| } |
| |
| TEST(StringUtilTest, DISABLED_UTFConversionPerf) { |
| std::string large_json_10k = "{\"results\": ["; |
| while (large_json_10k.length() < 10000) { |
| large_json_10k += |
| "{\"id\": 1024, \"title\": \"Google 検索\", \"desc\": \"Bonjour le " |
| "monde!\"},"; |
| } |
| large_json_10k += "]}"; |
| |
| std::string large_non_ascii_10k; |
| while (large_non_ascii_10k.length() < 10000) { |
| large_non_ascii_10k += "日本語の文章テキストデータです。"; |
| } |
| |
| struct Dataset { |
| std::string name; |
| std::string text; |
| size_t iterations = 1000000; |
| } datasets[] = { |
| // 1. Very short strings (< 32 bytes) |
| {"Short Pure ASCII (HTML ID)", "main-container", 1000000}, |
| {"Short Mixed ASCII (Token)", "id_café", 1000000}, |
| {"Short Pure Non-ASCII (CJK)", "検索", 1000000}, |
| |
| // 2. Standard real-world mixed datasets |
| {"Pure ASCII (URL)", |
| "https://www.google.com/search?q=chromium+base+strings+performance", |
| 1000000}, |
| {"Mixed ASCII+Japanese (HTML)", |
| "<div class=\"title\">Google 検索と AI 機能</div>", 1000000}, |
| {"Mixed ASCII+Spanish (JSON)", |
| "{\"status\": \"success\", \"message\": \"Bonjour le monde! ¿Cómo " |
| "estás?\"}", |
| 1000000}, |
| {"Mixed ASCII+Cyrillic (Path)", |
| "/usr/share/doc/chromium/документация/README.md", 1000000}, |
| {"Mixed ASCII+Arabic (JSON)", |
| "{\"lang\": \"ar\", \"label\": \"مرحبا بالعالم\", \"code\": 200}", |
| 1000000}, |
| {"Mixed ASCII+Hindi (Header)", |
| "HTTP/1.1 200 OK\r\nX-Custom-Header: नमस्ते दुनिया\r\n\r\n", 1000000}, |
| {"Pure Non-ASCII (Japanese)", "日本語の文章テキストデータです。", |
| 1000000}, |
| |
| // 3. Interspersed non-ASCII & Early non-ASCII in long strings |
| {"Early Non-ASCII in 1KB JSON", |
| "{\"name\": \"José\", \"payload\": \"" + std::string(1000, 'x') + "\"}", |
| 1000000}, |
| {"Interspersed Emoji/Accents", |
| "Hello 👋 World 🌍 Chromium 🚀 UTF8 💡 Strings 🔍 Unicode 🌐 Speed 🏎️", |
| 1000000}, |
| |
| // 4. Synthetic batch boundary tests |
| {"Worst-Case (End of 1st Batch)", |
| std::string(127, 'A') + "\xC2\xA2" + std::string(100, 'A'), 1000000}, |
| {"Worst-Case (End of 2nd Batch)", |
| std::string(255, 'A') + "\xC2\xA2" + std::string(100, 'A'), 1000000}, |
| {"Worst-Case (End of 8th Batch)", |
| std::string(1023, 'A') + "\xC2\xA2" + std::string(100, 'A'), 1000000}, |
| |
| // 5. Large multi-kilobyte payloads (10KB - 50KB) |
| {"Large Payload - 10KB JSON (Mixed)", large_json_10k, 100000}, |
| {"Large Payload - 10KB Non-ASCII (CJK)", large_non_ascii_10k, 100000}, |
| {"Large Payload - 50KB ASCII Document", std::string(50000, 'A'), 20000}, |
| {"Large Payload - 50KB (Worst-Case Suffix)", |
| std::string(50000, 'A') + "\xC2\xA2" + std::string(100, 'A'), 20000}, |
| }; |
| |
| for (const auto& ds : datasets) { |
| MeasureUTF8ToUTF16(ds.name, ds.text, false, ds.iterations); |
| MeasureUTF8ToUTF16(ds.name, ds.text, true, ds.iterations); |
| } |
| |
| std::u16string u16_path = u"/usr/share/doc/chromium/документация/README.md"; |
| MeasureUTF16ToUTF8("Mixed ASCII+Cyrillic (Path)", u16_path, false); |
| MeasureUTF16ToUTF8("Mixed ASCII+Cyrillic (Path)", u16_path, true); |
| |
| MeasureUTF8ToWide("Mixed ASCII+Cyrillic (Path)", |
| "/usr/share/doc/chromium/документация/README.md", false); |
| MeasureUTF8ToWide("Mixed ASCII+Cyrillic (Path)", |
| "/usr/share/doc/chromium/документация/README.md", true); |
| } |
| |
| } // namespace base |
