Skip to content

GH-36433: [C++] Update fast_float version to 3.10.1 #36434

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

Merged
kou merged 18 commits into from
Jul 19, 2023
Merged

GH-36433: [C++] Update fast_float version to 3.10.1 #36434

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
68c7eec
GH-36433 updated fast_float version to 5.2.0
xxlaykxx Jul 2, 2023
92b5adf
Update fast_float version
xxlaykxx Jul 4, 2023
81abdaa
Define FASTFLOAT_ALLOWS_LEADING_PLUS
xxlaykxx Jul 7, 2023
99f4b61
define FASTFLOAT_ALLOWS_LEADING_PLUS
xxlaykxx Jul 7, 2023
75f4fc3
move define to valuer_parsing class, added tests
xxlaykxx Jul 10, 2023
dcc6cfa
fixed test
xxlaykxx Jul 10, 2023
d1fd764
fixed tests
xxlaykxx Jul 10, 2023
ab215ad
fixed tests
xxlaykxx Jul 10, 2023
cfbf691
fixed formating
xxlaykxx Jul 10, 2023
7da736e
fixed formating
xxlaykxx Jul 10, 2023
028b175
Merge branch 'apache:main' into GH-36433
xxlaykxx Jul 11, 2023
3a3b5fc
Moving define
xxlaykxx Jul 14, 2023
d8ca8b9
fix define
xxlaykxx Jul 14, 2023
b0ddd25
fixed formating
xxlaykxx Jul 14, 2023
5045b35
fixed comment
xxlaykxx Jul 17, 2023
dbdaea0
Update cpp/src/arrow/util/value_parsing.cc
xxlaykxx Jul 18, 2023
2a5eef6
Merge branch 'apache:main' into GH-36433
xxlaykxx Jul 18, 2023
89e16f0
Merge branch 'apache:main' into GH-36433
xxlaykxx Jul 18, 2023
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
2 changes: 2 additions & 0 deletions cpp/src/arrow/util/value_parsing.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,8 @@
// specific language governing permissions and limitations
// under the License.

#define FASTFLOAT_ALLOWS_LEADING_PLUS 1

#include "arrow/util/value_parsing.h"

#include <string>
Expand Down
6 changes: 6 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 @@ -119,6 +119,9 @@ TEST(StringConversion, ToFloat) {
AssertConversion<FloatType>("0", 0.0f);
AssertConversion<FloatType>("-0.0", -0.0f);
AssertConversion<FloatType>("-1e20", -1e20f);
AssertConversion<FloatType>("+Infinity", std::numeric_limits<float>::infinity());
AssertConversion<FloatType>("-Infinity", -std::numeric_limits<float>::infinity());
AssertConversion<FloatType>("Infinity", std::numeric_limits<float>::infinity());

AssertConversionFails<FloatType>("");
AssertConversionFails<FloatType>("e");
Expand All @@ -135,6 +138,9 @@ TEST(StringConversion, ToDouble) {
AssertConversion<DoubleType>("0", 0);
AssertConversion<DoubleType>("-0.0", -0.0);
AssertConversion<DoubleType>("-1e100", -1e100);
AssertConversion<DoubleType>("+Infinity", std::numeric_limits<double>::infinity());
AssertConversion<DoubleType>("-Infinity", -std::numeric_limits<double>::infinity());
AssertConversion<DoubleType>("Infinity", std::numeric_limits<double>::infinity());

AssertConversionFails<DoubleType>("");
AssertConversionFails<DoubleType>("e");
Expand Down
4 changes: 2 additions & 2 deletions cpp/src/arrow/vendored/fast_float/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@
# fast_float

The files in this directory are vendored from fast_float
git tag `v3.8.1`.
git tag `v3.10.1`.
Copy link
Member

@kou kou Jul 5, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you really want to use 3.10.1?
It seems that the latest version is 5.2.0: https://github.com/fastfloat/fast_float/releases

Copy link
Contributor Author

@xxlaykxx xxlaykxx Jul 5, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

after 3.10.1 they started using c++ 20 and it cause a lot of build errors. In reality - we just need in parse_number.h

#ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default
  if (*first == UC('+')) {
      ++first;
  }

Copy link
Member

@kou kou Jul 5, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK. Could you describe it in descriptions of the PR and the associated issue?


See https://github.com/fastfloat/fast_float

Expand All @@ -31,7 +31,7 @@ See https://github.com/fastfloat/fast_float
## How to update

You must replace `VERSION` in the command lines with suitable version
such as `3.8.1`.
such as `3.10.1`.

```bash
cpp/src/arrow/vendoered/fast_float/update.sh VERSION
Expand Down
15 changes: 11 additions & 4 deletions cpp/src/arrow/vendored/fast_float/ascii_number.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,9 +13,11 @@ namespace fast_float {

// Next function can be micro-optimized, but compilers are entirely
// able to optimize it well.
fastfloat_really_inline bool is_integer(char c) noexcept { return c >= '0' && c <= '9'; }
fastfloat_really_inline constexpr bool is_integer(char c) noexcept {
return c >= '0' && c <= '9';
}

fastfloat_really_inline uint64_t byteswap(uint64_t val) {
fastfloat_really_inline constexpr uint64_t byteswap(uint64_t val) {
return (val & 0xFF00000000000000) >> 56
| (val & 0x00FF000000000000) >> 40
| (val & 0x0000FF0000000000) >> 24
Expand Down Expand Up @@ -45,7 +47,8 @@ fastfloat_really_inline void write_u64(uint8_t *chars, uint64_t val) {
}

// credit @aqrit
fastfloat_really_inline uint32_t parse_eight_digits_unrolled(uint64_t val) {
fastfloat_really_inline FASTFLOAT_CONSTEXPR14
uint32_t parse_eight_digits_unrolled(uint64_t val) {
const uint64_t mask = 0x000000FF000000FF;
const uint64_t mul1 = 0x000F424000000064; // 100 + (1000000ULL << 32)
const uint64_t mul2 = 0x0000271000000001; // 1 + (10000ULL << 32)
Expand All @@ -60,7 +63,7 @@ fastfloat_really_inline uint32_t parse_eight_digits_unrolled(const char *chars)
}

// credit @aqrit
fastfloat_really_inline bool is_made_of_eight_digits_fast(uint64_t val) noexcept {
fastfloat_really_inline constexpr bool is_made_of_eight_digits_fast(uint64_t val) noexcept {
return !((((val + 0x4646464646464646) | (val - 0x3030303030303030)) &
0x8080808080808080));
}
Expand Down Expand Up @@ -94,7 +97,11 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
answer.valid = false;
answer.too_many_digits = false;
answer.negative = (*p == '-');
#if FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default
if ((*p == '-') || (*p == '+')) {
#else
if (*p == '-') { // C++17 20.19.3.(7.1) explicitly forbids '+' sign here
#endif
++p;
if (p == pend) {
return answer;
Expand Down
Loading