UTF-8 encoding library to support Rust Punycode decoding.

PiperOrigin-RevId: 641983507
Change-Id: Iad7933884aef6bfd90d159c049a1d698d19456c6

3 files changed, 292 insertions, 0 deletions

diff --git a/absl/debugging/internal/utf8_for_code_point.cc b/absl/debugging/internal/utf8_for_code_point.cc
new file mode 100644
index 00000000..658a3b51
--- /dev/null
+++ b/absl/debugging/internal/utf8_for_code_point.cc
@@ -0,0 +1,70 @@
+// Copyright 2024 The Abseil Authors
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     https://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "absl/debugging/internal/utf8_for_code_point.h"
+
+#include <cstdint>
+
+#include "absl/base/config.h"
+
+namespace absl {
+ABSL_NAMESPACE_BEGIN
+namespace debugging_internal {
+namespace {
+
+// UTF-8 encoding bounds.
+constexpr uint32_t kMinSurrogate = 0xd800, kMaxSurrogate = 0xdfff;
+constexpr uint32_t kMax1ByteCodePoint = 0x7f;
+constexpr uint32_t kMax2ByteCodePoint = 0x7ff;
+constexpr uint32_t kMax3ByteCodePoint = 0xffff;
+constexpr uint32_t kMaxCodePoint = 0x10ffff;
+
+}  // namespace
+
+Utf8ForCodePoint::Utf8ForCodePoint(uint64_t code_point) {
+  if (code_point <= kMax1ByteCodePoint) {
+    length = 1;
+    bytes[0] = static_cast<char>(code_point);
+    return;
+  }
+
+  if (code_point <= kMax2ByteCodePoint) {
+    length = 2;
+    bytes[0] = static_cast<char>(0xc0 | (code_point >> 6));
+    bytes[1] = static_cast<char>(0x80 | (code_point & 0x3f));
+    return;
+  }
+
+  if (kMinSurrogate <= code_point && code_point <= kMaxSurrogate) return;
+
+  if (code_point <= kMax3ByteCodePoint) {
+    length = 3;
+    bytes[0] = static_cast<char>(0xe0 | (code_point >> 12));
+    bytes[1] = static_cast<char>(0x80 | ((code_point >> 6) & 0x3f));
+    bytes[2] = static_cast<char>(0x80 | (code_point & 0x3f));
+    return;
+  }
+
+  if (code_point > kMaxCodePoint) return;
+
+  length = 4;
+  bytes[0] = static_cast<char>(0xf0 | (code_point >> 18));
+  bytes[1] = static_cast<char>(0x80 | ((code_point >> 12) & 0x3f));
+  bytes[2] = static_cast<char>(0x80 | ((code_point >> 6) & 0x3f));
+  bytes[3] = static_cast<char>(0x80 | (code_point & 0x3f));
+}
+
+}  // namespace debugging_internal
+ABSL_NAMESPACE_END
+}  // namespace absl
diff --git a/absl/debugging/internal/utf8_for_code_point.h b/absl/debugging/internal/utf8_for_code_point.h
new file mode 100644
index 00000000..f23cde6d
--- /dev/null
+++ b/absl/debugging/internal/utf8_for_code_point.h
@@ -0,0 +1,47 @@
+// Copyright 2024 The Abseil Authors
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     https://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#ifndef ABSL_DEBUGGING_INTERNAL_UTF8_FOR_CODE_POINT_H_
+#define ABSL_DEBUGGING_INTERNAL_UTF8_FOR_CODE_POINT_H_
+
+#include <cstdint>
+
+#include "absl/base/config.h"
+
+namespace absl {
+ABSL_NAMESPACE_BEGIN
+namespace debugging_internal {
+
+struct Utf8ForCodePoint {
+  // Converts a Unicode code point to the corresponding UTF-8 byte sequence.
+  // Async-signal-safe to support use in symbolizing stack traces from a signal
+  // handler.
+  explicit Utf8ForCodePoint(uint64_t code_point);
+
+  // Returns true if the constructor's code_point argument was valid.
+  bool ok() const { return length != 0; }
+
+  // If code_point was in range, then 1 <= length <= 4, and the UTF-8 encoding
+  // is found in bytes[0 .. (length - 1)].  If code_point was invalid, then
+  // length == 0.  In either case, the contents of bytes[length .. 3] are
+  // unspecified.
+  char bytes[4] = {};
+  uint32_t length = 0;
+};
+
+}  // namespace debugging_internal
+ABSL_NAMESPACE_END
+}  // namespace absl
+
+#endif  // ABSL_DEBUGGING_INTERNAL_UTF8_FOR_CODE_POINT_H_
diff --git a/absl/debugging/internal/utf8_for_code_point_test.cc b/absl/debugging/internal/utf8_for_code_point_test.cc
new file mode 100644
index 00000000..d6817c77
--- /dev/null
+++ b/absl/debugging/internal/utf8_for_code_point_test.cc
@@ -0,0 +1,175 @@
+// Copyright 2024 The Abseil Authors
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     https://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "absl/debugging/internal/utf8_for_code_point.h"
+
+#include <cstdint>
+
+#include "gtest/gtest.h"
+#include "absl/base/config.h"
+
+namespace absl {
+ABSL_NAMESPACE_BEGIN
+namespace debugging_internal {
+namespace {
+
+TEST(Utf8ForCodePointTest, RecognizesTheSmallestCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0});
+  ASSERT_EQ(utf8.length, 1);
+  EXPECT_EQ(utf8.bytes[0], '\0');
+}
+
+TEST(Utf8ForCodePointTest, RecognizesAsciiSmallA) {
+  Utf8ForCodePoint utf8(std::uint64_t{'a'});
+  ASSERT_EQ(utf8.length, 1);
+  EXPECT_EQ(utf8.bytes[0], 'a');
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheLargestOneByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x7f});
+  ASSERT_EQ(utf8.length, 1);
+  EXPECT_EQ(utf8.bytes[0], '\x7f');
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheSmallestTwoByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x80});
+  ASSERT_EQ(utf8.length, 2);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xc2));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0x80));
+}
+
+TEST(Utf8ForCodePointTest, RecognizesSmallNWithTilde) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xf1});
+  ASSERT_EQ(utf8.length, 2);
+  const char* want = "ñ";
+  EXPECT_EQ(utf8.bytes[0], want[0]);
+  EXPECT_EQ(utf8.bytes[1], want[1]);
+}
+
+TEST(Utf8ForCodePointTest, RecognizesCapitalPi) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x3a0});
+  ASSERT_EQ(utf8.length, 2);
+  const char* want = "Π";
+  EXPECT_EQ(utf8.bytes[0], want[0]);
+  EXPECT_EQ(utf8.bytes[1], want[1]);
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheLargestTwoByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x7ff});
+  ASSERT_EQ(utf8.length, 2);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xdf));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0xbf));
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheSmallestThreeByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x800});
+  ASSERT_EQ(utf8.length, 3);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xe0));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0xa0));
+  EXPECT_EQ(utf8.bytes[2], static_cast<char>(0x80));
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheChineseCharacterZhong1AsInZhong1Wen2) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x4e2d});
+  ASSERT_EQ(utf8.length, 3);
+  const char* want = "中";
+  EXPECT_EQ(utf8.bytes[0], want[0]);
+  EXPECT_EQ(utf8.bytes[1], want[1]);
+  EXPECT_EQ(utf8.bytes[2], want[2]);
+}
+
+TEST(Utf8ForCodePointTest, RecognizesOneBeforeTheSmallestSurrogate) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xd7ff});
+  ASSERT_EQ(utf8.length, 3);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xed));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0x9f));
+  EXPECT_EQ(utf8.bytes[2], static_cast<char>(0xbf));
+}
+
+TEST(Utf8ForCodePointTest, RejectsTheSmallestSurrogate) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xd800});
+  EXPECT_EQ(utf8.length, 0);
+}
+
+TEST(Utf8ForCodePointTest, RejectsTheLargestSurrogate) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xdfff});
+  EXPECT_EQ(utf8.length, 0);
+}
+
+TEST(Utf8ForCodePointTest, RecognizesOnePastTheLargestSurrogate) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xe000});
+  ASSERT_EQ(utf8.length, 3);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xee));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0x80));
+  EXPECT_EQ(utf8.bytes[2], static_cast<char>(0x80));
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheLargestThreeByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xffff});
+  ASSERT_EQ(utf8.length, 3);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xef));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0xbf));
+  EXPECT_EQ(utf8.bytes[2], static_cast<char>(0xbf));
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheSmallestFourByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x10000});
+  ASSERT_EQ(utf8.length, 4);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xf0));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0x90));
+  EXPECT_EQ(utf8.bytes[2], static_cast<char>(0x80));
+  EXPECT_EQ(utf8.bytes[3], static_cast<char>(0x80));
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheJackOfHearts) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x1f0bb});
+  ASSERT_EQ(utf8.length, 4);
+  const char* want = "🂻";
+  EXPECT_EQ(utf8.bytes[0], want[0]);
+  EXPECT_EQ(utf8.bytes[1], want[1]);
+  EXPECT_EQ(utf8.bytes[2], want[2]);
+  EXPECT_EQ(utf8.bytes[3], want[3]);
+}
+
+TEST(Utf8ForCodePointTest, RecognizesTheLargestFourByteCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x10ffff});
+  ASSERT_EQ(utf8.length, 4);
+  EXPECT_EQ(utf8.bytes[0], static_cast<char>(0xf4));
+  EXPECT_EQ(utf8.bytes[1], static_cast<char>(0x8f));
+  EXPECT_EQ(utf8.bytes[2], static_cast<char>(0xbf));
+  EXPECT_EQ(utf8.bytes[3], static_cast<char>(0xbf));
+}
+
+TEST(Utf8ForCodePointTest, RejectsTheSmallestOverlargeCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0x110000});
+  EXPECT_EQ(utf8.length, 0);
+}
+
+TEST(Utf8ForCodePointTest, RejectsAThroughlyOverlargeCodePoint) {
+  Utf8ForCodePoint utf8(std::uint64_t{0xffffffff00000000});
+  EXPECT_EQ(utf8.length, 0);
+}
+
+TEST(Utf8ForCodePointTest, OkReturnsTrueForAValidCodePoint) {
+  EXPECT_TRUE(Utf8ForCodePoint(std::uint64_t{0}).ok());
+}
+
+TEST(Utf8ForCodePointTest, OkReturnsFalseForAnInvalidCodePoint) {
+  EXPECT_FALSE(Utf8ForCodePoint(std::uint64_t{0xffffffff00000000}).ok());
+}
+
+}  // namespace
+}  // namespace debugging_internal
+ABSL_NAMESPACE_END
+}  // namespace absl