59 lines
2.3 KiB
C++
59 lines
2.3 KiB
C++
|
/*
|
||
|
|
* Copyright (c) Meta Platforms, Inc. and affiliates.
|
||
|
|
*
|
||
|
|
* This source code is licensed under the MIT license found in the
|
||
|
|
* LICENSE file in the root directory of this source tree.
|
||
|
|
*/
|
||
|
|
|
||
|
|
#include <jsinspector-modern/Utf8.h>
|
||
|
|
|
||
|
|
#include <gmock/gmock.h>
|
||
|
|
#include <gtest/gtest.h>
|
||
|
|
|
||
|
|
using namespace ::testing;
|
||
|
|
|
||
|
|
namespace facebook::react::jsinspector_modern {
|
||
|
|
|
||
|
|
TEST(Utf8Test, TruncateToValidUtf8) {
|
||
|
|
auto buffer = std::vector<char>();
|
||
|
|
std::vector<std::pair<std::string, size_t>> expectedStringsUpToSizes;
|
||
|
|
// Construct a buffer with a concatenation of all code points, and a vector
|
||
|
|
// or "expectedStringsUpToSizes", pairs of valid UTF8 prefix strings and sizes
|
||
|
|
// "n" such that the string would be the expected truncation of the first "n"
|
||
|
|
// bytes of the buffer.
|
||
|
|
for (const std::string& codePoint : {
|
||
|
|
"a", // 1 byte
|
||
|
|
"é", // 2 bytes
|
||
|
|
"✓", // 3 bytes
|
||
|
|
"😀" // 4 bytes
|
||
|
|
}) {
|
||
|
|
auto partial = std::string(buffer.data(), buffer.size());
|
||
|
|
buffer.insert(buffer.end(), codePoint.begin(), codePoint.end());
|
||
|
|
expectedStringsUpToSizes.push_back(std::pair(partial, buffer.size()));
|
||
|
|
}
|
||
|
|
// The constructed buffer is 10 bytes long, comprised of 4 code points of
|
||
|
|
// varied size. Range over naive slices of length 0-9 ensuring that the
|
||
|
|
// truncated result matches the valid UTF8 substring of length <= n.
|
||
|
|
size_t n = 0;
|
||
|
|
for (const auto& expectedStringUpToSize : expectedStringsUpToSizes) {
|
||
|
|
auto nextSize = expectedStringUpToSize.second;
|
||
|
|
auto expectedString = expectedStringUpToSize.first;
|
||
|
|
for (; n < nextSize; ++n) {
|
||
|
|
// Take the first n bytes of the whole buffer, which may be slicing
|
||
|
|
// through the middle of a code point.
|
||
|
|
std::vector<char> slice(buffer.begin(), buffer.begin() + n);
|
||
|
|
truncateToValidUTF8(slice);
|
||
|
|
// Expect the final code point fragment has been discarded and that the
|
||
|
|
// contents are equal to expectedString, which is valid UTF8.
|
||
|
|
EXPECT_EQ(std::string(slice.begin(), slice.end()), expectedString);
|
||
|
|
}
|
||
|
|
}
|
||
|
|
// Finally verify that truncating the whole buffer, which is already valid
|
||
|
|
// UTF8, is a no-op.
|
||
|
|
auto wholeString = std::string(buffer.begin(), buffer.end());
|
||
|
|
truncateToValidUTF8(buffer);
|
||
|
|
EXPECT_EQ(std::string(buffer.begin(), buffer.end()), wholeString);
|
||
|
|
}
|
||
|
|
|
||
|
|
} // namespace facebook::react::jsinspector_modern
|