Skip to content

ARROW-12657: [C++] Adding String hex to numeric conversion #11160

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wmalpica wants to merge 1 commit into from
Closed

ARROW-12657: [C++] Adding String hex to numeric conversion #11160

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 24 additions & 19 deletions cpp/src/arrow/compute/kernels/scalar_cast_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -315,42 +315,43 @@ class TestCast : public TestBase {
CastOptions options;
auto src_type = TypeTraits<SourceType>::type_singleton();

std::vector<bool> is_valid = {true, false, true, true, true};
std::vector<bool> is_valid = {true, false, true, true, true, true, true, true};

// string to int
std::vector<std::string> v_int = {"0", "1", "127", "-1", "0"};
std::vector<int8_t> e_int8 = {0, 1, 127, -1, 0};
std::vector<int16_t> e_int16 = {0, 1, 127, -1, 0};
std::vector<int32_t> e_int32 = {0, 1, 127, -1, 0};
std::vector<int64_t> e_int64 = {0, 1, 127, -1, 0};
std::vector<std::string> v_int = {"0", "1", "127", "-1", "0", "0x0", "0x7f", "0XFF"};
std::vector<int8_t> e_int8 = {0, 1, 127, -1, 0, 0, 127, -1};
std::vector<int16_t> e_int16 = {0, 1, 127, -1, 0, 0, 127, -1};
std::vector<int32_t> e_int32 = {0, 1, 127, -1, 0, 0, 127, -1};
std::vector<int64_t> e_int64 = {0, 1, 127, -1, 0, 0, 127, -1};
CheckCase<SourceType, Int8Type>(v_int, is_valid, e_int8, options);
CheckCase<SourceType, Int16Type>(v_int, is_valid, e_int16, options);
CheckCase<SourceType, Int32Type>(v_int, is_valid, e_int32, options);
CheckCase<SourceType, Int64Type>(v_int, is_valid, e_int64, options);

v_int = {"2147483647", "0", "-2147483648", "0", "0"};
e_int32 = {2147483647, 0, -2147483648LL, 0, 0};
v_int = {"2147483647", "0", "-2147483648", "0", "0", "0x7fff", "0X0", "0xF000"};
e_int32 = {2147483647, 0, -2147483648LL, 0, 0, 2147483647, 0, -2147483648LL};
CheckCase<SourceType, Int32Type>(v_int, is_valid, e_int32, options);
v_int = {"9223372036854775807", "0", "-9223372036854775808", "0", "0"};
e_int64 = {9223372036854775807LL, 0, (-9223372036854775807LL - 1), 0, 0};
v_int = {"9223372036854775807", "0", "-9223372036854775808", "0", "0", "0x7FFFFFFF", "0x0", "0Xffffffff"};
e_int64 = {9223372036854775807LL, 0, (-9223372036854775807LL - 1), 0, 0,
9223372036854775807LL, 0, (-9223372036854775807LL - 1)};
CheckCase<SourceType, Int64Type>(v_int, is_valid, e_int64, options);

// string to uint
std::vector<std::string> v_uint = {"0", "1", "127", "255", "0"};
std::vector<uint8_t> e_uint8 = {0, 1, 127, 255, 0};
std::vector<uint16_t> e_uint16 = {0, 1, 127, 255, 0};
std::vector<uint32_t> e_uint32 = {0, 1, 127, 255, 0};
std::vector<uint64_t> e_uint64 = {0, 1, 127, 255, 0};
std::vector<std::string> v_uint = {"0", "1", "127", "255", "0", "0x0", "0x7f", "0XFF"};
std::vector<uint8_t> e_uint8 = {0, 1, 127, 255, 0, 0, 127, 255};
std::vector<uint16_t> e_uint16 = {0, 1, 127, 255, 0, 0, 127, 255};
std::vector<uint32_t> e_uint32 = {0, 1, 127, 255, 0, 0, 127, 255};
std::vector<uint64_t> e_uint64 = {0, 1, 127, 255, 0, 0, 127, 255};
CheckCase<SourceType, UInt8Type>(v_uint, is_valid, e_uint8, options);
CheckCase<SourceType, UInt16Type>(v_uint, is_valid, e_uint16, options);
CheckCase<SourceType, UInt32Type>(v_uint, is_valid, e_uint32, options);
CheckCase<SourceType, UInt64Type>(v_uint, is_valid, e_uint64, options);

v_uint = {"4294967295", "0", "0", "0", "0"};
e_uint32 = {4294967295, 0, 0, 0, 0};
v_uint = {"4294967295", "0", "0", "0", "0", "0xFFFF", "0x0", "0x1"};
e_uint32 = {4294967295, 0, 0, 0, 0, 4294967295, 0, 1};
CheckCase<SourceType, UInt32Type>(v_uint, is_valid, e_uint32, options);
v_uint = {"18446744073709551615", "0", "0", "0", "0"};
e_uint64 = {18446744073709551615ULL, 0, 0, 0, 0};
v_uint = {"18446744073709551615", "0", "0", "0", "0", "0xffffFFFF", "0x0", "0x1"};
e_uint64 = {18446744073709551615ULL, 0, 0, 0, 0, 18446744073709551615ULL, 0, 1};
CheckCase<SourceType, UInt64Type>(v_uint, is_valid, e_uint64, options);

// string to float
Expand Down Expand Up @@ -1554,6 +1555,10 @@ TEST_F(TestCast, StringToNumberErrors) {
CheckFails<StringType>({"128"}, is_valid, int8(), options);
CheckFails<StringType>({"-129"}, is_valid, int8(), options);
CheckFails<StringType>({"0.5"}, is_valid, int8(), options);
CheckFails<StringType>({"0x"}, is_valid, int8(), options);
CheckFails<StringType>({"0x12r"}, is_valid, int32(), options);
CheckFails<StringType>({"0x1ffff"}, is_valid, int32(), options);
CheckFails<StringType>({"-0x123"}, is_valid, int32(), options);

CheckFails<StringType>({"256"}, is_valid, uint8(), options);
CheckFails<StringType>({"-1"}, is_valid, uint8(), options);
Expand Down
50 changes: 50 additions & 0 deletions cpp/src/arrow/util/value_parsing_benchmark.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,6 +56,27 @@ static std::vector<std::string> MakeIntStrings(int32_t num_items) {
return strings;
}

template <typename c_int>
static std::vector<std::string> MakeHexStrings(int32_t num_items) {
int32_t num_bytes = sizeof(c_int);
const char* kAsciiTable = "0123456789ABCDEF";
std::vector<char> large_hex_chars(num_bytes * 2 + 2);
large_hex_chars[0] = '0';
large_hex_chars[1] = 'x';
for (int32_t i = 0; i < num_bytes * 2; ++i) {
large_hex_chars[i + 2] = kAsciiTable[i];
}
std::string large_hex(&large_hex_chars[0], large_hex_chars.size());

std::vector<std::string> base_strings = {"0x0", "0xA5", "0x5E", large_hex};
std::vector<std::string> strings;
for (int32_t i = 0; i < num_items; ++i) {
strings.push_back(base_strings[i % base_strings.size()]);
}
return strings;
}


static std::vector<std::string> MakeFloatStrings(int32_t num_items) {
std::vector<std::string> base_strings = {"0.0", "5", "-12.3",
"98765430000", "3456.789", "0.0012345",
Expand Down Expand Up @@ -123,6 +144,26 @@ static void IntegerParsing(benchmark::State& state) { // NOLINT non-const refer
state.SetItemsProcessed(state.iterations() * strings.size());
}

template <typename ARROW_TYPE, typename C_TYPE = typename ARROW_TYPE::c_type>
static void HexParsing(benchmark::State& state) { // NOLINT non-const reference
auto strings = MakeHexStrings<C_TYPE>(1000);

while (state.KeepRunning()) {
C_TYPE total = 0;
for (const auto& s : strings) {
C_TYPE value;
if (!ParseValue<ARROW_TYPE>(s.data(), s.length(), &value)) {
std::cerr << "Conversion failed for '" << s << "'";
std::abort();
}
total = static_cast<C_TYPE>(total + value);
}
benchmark::DoNotOptimize(total);
}
state.SetItemsProcessed(state.iterations() * strings.size());
}


template <typename ARROW_TYPE, typename C_TYPE = typename ARROW_TYPE::c_type>
static void FloatParsing(benchmark::State& state) { // NOLINT non-const reference
auto strings = MakeFloatStrings(1000);
Expand Down Expand Up @@ -230,6 +271,15 @@ BENCHMARK_TEMPLATE(IntegerParsing, UInt16Type);
BENCHMARK_TEMPLATE(IntegerParsing, UInt32Type);
BENCHMARK_TEMPLATE(IntegerParsing, UInt64Type);

BENCHMARK_TEMPLATE(HexParsing, Int8Type);
BENCHMARK_TEMPLATE(HexParsing, Int16Type);
BENCHMARK_TEMPLATE(HexParsing, Int32Type);
BENCHMARK_TEMPLATE(HexParsing, Int64Type);
BENCHMARK_TEMPLATE(HexParsing, UInt8Type);
BENCHMARK_TEMPLATE(HexParsing, UInt16Type);
BENCHMARK_TEMPLATE(HexParsing, UInt32Type);
BENCHMARK_TEMPLATE(HexParsing, UInt64Type);

BENCHMARK_TEMPLATE(FloatParsing, FloatType);
BENCHMARK_TEMPLATE(FloatParsing, DoubleType);

Expand Down
77 changes: 77 additions & 0 deletions cpp/src/arrow/util/value_parsing_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -120,6 +120,17 @@ TEST(StringConversion, ToInt8) {
AssertConversionFails<Int8Type>("-");
AssertConversionFails<Int8Type>("0.0");
AssertConversionFails<Int8Type>("e");

// Hex
AssertConversion<Int8Type>("0x0", 0);
AssertConversion<Int8Type>("0X1A", 26);
AssertConversion<Int8Type>("0xb", 11);
AssertConversion<Int8Type>("0x7F", 127);
AssertConversion<Int8Type>("0xFF", -1);
AssertConversionFails<Int8Type>("0x");
AssertConversionFails<Int8Type>("0x100");
AssertConversionFails<Int8Type>("0x1g");

}

TEST(StringConversion, ToUInt8) {
Expand Down Expand Up @@ -155,6 +166,16 @@ TEST(StringConversion, ToInt16) {
AssertConversionFails<Int16Type>("-");
AssertConversionFails<Int16Type>("0.0");
AssertConversionFails<Int16Type>("e");

// Hex
AssertConversion<Int16Type>("0x0", 0);
AssertConversion<Int16Type>("0X1aA", 426);
AssertConversion<Int16Type>("0xb", 11);
AssertConversion<Int16Type>("0x7ffF", 32767);
AssertConversion<Int16Type>("0XfffF", -1);
AssertConversionFails<Int16Type>("0x");
AssertConversionFails<Int16Type>("0x10000");
AssertConversionFails<Int16Type>("0x1g");
}

TEST(StringConversion, ToUInt16) {
Expand All @@ -172,6 +193,16 @@ TEST(StringConversion, ToUInt16) {
AssertConversionFails<UInt16Type>("-");
AssertConversionFails<UInt16Type>("0.0");
AssertConversionFails<UInt16Type>("e");

// Hex
AssertConversion<UInt16Type>("0x0", 0);
AssertConversion<UInt16Type>("0x1aA", 426);
AssertConversion<UInt16Type>("0xb", 11);
AssertConversion<UInt16Type>("0x7ffF", 32767);
AssertConversion<UInt16Type>("0xFffF", 65535);
AssertConversionFails<UInt16Type>("0x");
AssertConversionFails<UInt16Type>("0x10000");
AssertConversionFails<UInt16Type>("0x1g");
}

TEST(StringConversion, ToInt32) {
Expand All @@ -189,6 +220,18 @@ TEST(StringConversion, ToInt32) {
AssertConversionFails<Int32Type>("-");
AssertConversionFails<Int32Type>("0.0");
AssertConversionFails<Int32Type>("e");

// Hex
AssertConversion<Int32Type>("0x0", 0);
AssertConversion<Int32Type>("0x123ABC", 1194684);
AssertConversion<Int32Type>("0xA4B35", 674613);
AssertConversion<Int32Type>("0x7FFFFFFF", 2147483647);
AssertConversion<Int32Type>("0x123abc", 1194684);
AssertConversion<Int32Type>("0xA4b35", 674613);
AssertConversion<Int32Type>("0x7FFFfFfF", 2147483647);
AssertConversion<Int32Type>("0XFFFFfFfF", -1);
AssertConversionFails<Int32Type>("0X");
AssertConversionFails<Int32Type>("0x23512ak");
}

TEST(StringConversion, ToUInt32) {
Expand All @@ -206,6 +249,18 @@ TEST(StringConversion, ToUInt32) {
AssertConversionFails<UInt32Type>("-");
AssertConversionFails<UInt32Type>("0.0");
AssertConversionFails<UInt32Type>("e");

// Hex
AssertConversion<UInt32Type>("0x0", 0);
AssertConversion<UInt32Type>("0x123ABC", 1194684);
AssertConversion<UInt32Type>("0xA4B35", 674613);
AssertConversion<UInt32Type>("0x7FFFFFFF", 2147483647);
AssertConversion<UInt32Type>("0x123abc", 1194684);
AssertConversion<UInt32Type>("0xA4b35", 674613);
AssertConversion<UInt32Type>("0x7FFFfFfF", 2147483647);
AssertConversion<UInt32Type>("0XFFFFfFfF", 4294967295);
AssertConversionFails<UInt32Type>("0X");
AssertConversionFails<UInt32Type>("0x23512ak");
}

TEST(StringConversion, ToInt64) {
Expand All @@ -223,6 +278,17 @@ TEST(StringConversion, ToInt64) {
AssertConversionFails<Int64Type>("-");
AssertConversionFails<Int64Type>("0.0");
AssertConversionFails<Int64Type>("e");

// Hex
AssertConversion<Int64Type>("0x0", 0);
AssertConversion<Int64Type>("0x5415a123ABC123cb", 6058926048274359243);
AssertConversion<Int64Type>("0xA4B35", 674613);
AssertConversion<Int64Type>("0x7FFFFFFFFFFFFFFf", 9223372036854775807);
AssertConversion<Int64Type>("0XF000000000000001", -1152921504606846975);
AssertConversion<Int64Type>("0xfFFFFFFFFFFFFFFf", -1);
AssertConversionFails<Int64Type>("0X");
AssertConversionFails<Int64Type>("0x12345678901234567");
AssertConversionFails<Int64Type>("0x23512ak");
}

TEST(StringConversion, ToUInt64) {
Expand All @@ -237,6 +303,17 @@ TEST(StringConversion, ToUInt64) {
AssertConversionFails<UInt64Type>("-");
AssertConversionFails<UInt64Type>("0.0");
AssertConversionFails<UInt64Type>("e");

// Hex
AssertConversion<UInt64Type>("0x0", 0);
AssertConversion<UInt64Type>("0x5415a123ABC123cb", 6058926048274359243);
AssertConversion<UInt64Type>("0xA4B35", 674613);
AssertConversion<UInt64Type>("0x7FFFFFFFFFFFFFFf", 9223372036854775807);
AssertConversion<UInt64Type>("0XF000000000000001", 17293822569102704641ULL);
AssertConversion<UInt64Type>("0xfFFFFFFFFFFFFFFf", 18446744073709551615ULL);
AssertConversionFails<UInt64Type>("0x");
AssertConversionFails<UInt64Type>("0x12345678901234567");
AssertConversionFails<UInt64Type>("0x23512ak");
}

TEST(StringConversion, ToTimestampDate_ISO8601) {
Expand Down