Chiebot-Mirror
/
abseil-cpp
8
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Change kPower10Table bounds to be half-open

Browse Source 
Half-open intervals are recommended by Dijkstra:
https://www.cs.utexas.edu/users/EWD/transcriptions/EWD08xx/EWD831.html

This also simplifies a static_assert by removing a +1 term.

Also fix a "smaller exponent" copy/pasto.

PiperOrigin-RevId: 477848998
Change-Id: I67af73d37ac08d36a1117ba33313a02932bd5814
pull/1283/head
Abseil Team 3 years ago committed by Copybara-Service
parent cfa5dc7576
commit b39aa365e1
1 changed files with 41 additions and 37 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 78
      absl/strings/charconv.cc

78
absl/strings/charconv.cc
Unescape View File

@ -111,11 +111,11 @@ struct FloatTraits<double> {
// //
// 9999999999999999999, with 19 nines but no decimal point, is the largest // 9999999999999999999, with 19 nines but no decimal point, is the largest
// "repeated nines" integer that fits in a uint64_t. // "repeated nines" integer that fits in a uint64_t.
static constexpr int kEiselLemireMinInclExp10 = -324 - 18; static constexpr int kEiselLemireMinInclusiveExp10 = -324 - 18;
// The smallest positive integer N such that parsing 1eN produces infinity. // The smallest positive integer N such that parsing 1eN produces infinity.
// Parsing a smaller N will produce something finite. // Parsing a smaller N will produce something finite.
static constexpr int kEiselLemireMaxExclExp10 = 309; static constexpr int kEiselLemireMaxExclusiveExp10 = 309;
static double MakeNan(const char* tagp) { static double MakeNan(const char* tagp) {
// Support nan no matter which namespace it's in. Some platforms // Support nan no matter which namespace it's in. Some platforms
@ -180,8 +180,8 @@ struct FloatTraits<float> {
static constexpr int kExponentBias = 127; static constexpr int kExponentBias = 127;
static constexpr int kEiselLemireShift = 38; static constexpr int kEiselLemireShift = 38;
static constexpr uint64_t kEiselLemireMask = uint64_t{0x3FFFFFFFFF}; static constexpr uint64_t kEiselLemireMask = uint64_t{0x3FFFFFFFFF};
static constexpr int kEiselLemireMinInclExp10 = -46 - 18; static constexpr int kEiselLemireMinInclusiveExp10 = -46 - 18;
static constexpr int kEiselLemireMaxExclExp10 = 39; static constexpr int kEiselLemireMaxExclusiveExp10 = 39;
static float MakeNan(const char* tagp) { static float MakeNan(const char* tagp) {
// Support nanf no matter which namespace it's in. Some platforms // Support nanf no matter which namespace it's in. Some platforms
@ -232,8 +232,8 @@ struct FloatTraits<float> {
// Lookups into the power-of-10 table must first check the Power10Overflow() and // Lookups into the power-of-10 table must first check the Power10Overflow() and
// Power10Underflow() functions, to avoid out-of-bounds table access. // Power10Underflow() functions, to avoid out-of-bounds table access.
// //
// Indexes into these tables are biased by -kPower10TableMin, and the table has // Indexes into these tables are biased by -kPower10TableMinInclusive. Valid
// values in the range [kPower10TableMin, kPower10TableMax]. // indexes range from kPower10TableMinInclusive to kPower10TableMaxExclusive.
extern const uint64_t kPower10MantissaHighTable[]; // High 64 of 128 bits. extern const uint64_t kPower10MantissaHighTable[]; // High 64 of 128 bits.
extern const uint64_t kPower10MantissaLowTable[]; // Low 64 of 128 bits. extern const uint64_t kPower10MantissaLowTable[]; // Low 64 of 128 bits.
extern const int16_t kPower10ExponentTable[]; extern const int16_t kPower10ExponentTable[];
@ -241,28 +241,28 @@ extern const int16_t kPower10ExponentTable[];
// The smallest (inclusive) allowed value for use with the Power10Mantissa() // The smallest (inclusive) allowed value for use with the Power10Mantissa()
// and Power10Exponent() functions below. (If a smaller exponent is needed in // and Power10Exponent() functions below. (If a smaller exponent is needed in
// calculations, the end result is guaranteed to underflow.) // calculations, the end result is guaranteed to underflow.)
constexpr int kPower10TableMin = -342; constexpr int kPower10TableMinInclusive = -342;
// The largest (inclusive) allowed value for use with the Power10Mantissa() and // The largest (exclusive) allowed value for use with the Power10Mantissa() and
// Power10Exponent() functions below. (If a smaller exponent is needed in // Power10Exponent() functions below. (If a larger-or-equal exponent is needed
// calculations, the end result is guaranteed to overflow.) // in calculations, the end result is guaranteed to overflow.)
constexpr int kPower10TableMax = 308; constexpr int kPower10TableMaxExclusive = 309;
uint64_t Power10Mantissa(int n) { uint64_t Power10Mantissa(int n) {
return kPower10MantissaHighTable[n - kPower10TableMin]; return kPower10MantissaHighTable[n - kPower10TableMinInclusive];
} }
int Power10Exponent(int n) { int Power10Exponent(int n) {
return kPower10ExponentTable[n - kPower10TableMin]; return kPower10ExponentTable[n - kPower10TableMinInclusive];
} }
// Returns true if n is large enough that 10**n always results in an IEEE // Returns true if n is large enough that 10**n always results in an IEEE
// overflow. // overflow.
bool Power10Overflow(int n) { return n > kPower10TableMax; } bool Power10Overflow(int n) { return n >= kPower10TableMaxExclusive; }
// Returns true if n is small enough that 10**n times a ParsedFloat mantissa // Returns true if n is small enough that 10**n times a ParsedFloat mantissa
// always results in an IEEE underflow. // always results in an IEEE underflow.
bool Power10Underflow(int n) { return n < kPower10TableMin; } bool Power10Underflow(int n) { return n < kPower10TableMinInclusive; }
// Returns true if Power10Mantissa(n) * 2**Power10Exponent(n) is exactly equal // Returns true if Power10Mantissa(n) * 2**Power10Exponent(n) is exactly equal
// to 10**n numerically. Put another way, this returns true if there is no // to 10**n numerically. Put another way, this returns true if there is no
@ -656,11 +656,11 @@ bool EiselLemire(const strings_internal::ParsedFloat& input, bool negative,
FloatType* value, std::errc* ec) { FloatType* value, std::errc* ec) {
uint64_t man = input.mantissa; uint64_t man = input.mantissa;
int exp10 = input.exponent; int exp10 = input.exponent;
if (exp10 < FloatTraits<FloatType>::kEiselLemireMinInclExp10) { if (exp10 < FloatTraits<FloatType>::kEiselLemireMinInclusiveExp10) {
*value = negative ? -0.0 : 0.0; *value = negative ? -0.0 : 0.0;
*ec = std::errc::result_out_of_range; *ec = std::errc::result_out_of_range;
return true; return true;
} else if (exp10 >= FloatTraits<FloatType>::kEiselLemireMaxExclExp10) { } else if (exp10 >= FloatTraits<FloatType>::kEiselLemireMaxExclusiveExp10) {
// Return max (a finite value) consistent with from_chars and DR 3081. For // Return max (a finite value) consistent with from_chars and DR 3081. For
// SimpleAtod and SimpleAtof, post-processing will return infinity. // SimpleAtod and SimpleAtof, post-processing will return infinity.
*value = negative ? -std::numeric_limits<FloatType>::max() *value = negative ? -std::numeric_limits<FloatType>::max()
@ -669,17 +669,16 @@ bool EiselLemire(const strings_internal::ParsedFloat& input, bool negative,
return true; return true;
} }
// Assert that (kPower10TableMin <= exp10) and (exp10 <= kPower10TableMax). // Assert kPower10TableMinInclusive <= exp10 < kPower10TableMaxExclusive.
// Equivalently, !Power10Underflow(exp10) and !Power10Overflow(exp10). // Equivalently, !Power10Underflow(exp10) and !Power10Overflow(exp10).
//
// The +1 is because kEiselLemireMaxExclExp10 is an exclusive bound but
// kPower10TableMax is inclusive.
static_assert( static_assert(
FloatTraits<FloatType>::kEiselLemireMinInclExp10 >= kPower10TableMin, FloatTraits<FloatType>::kEiselLemireMinInclusiveExp10 >=
"(exp10 - kPower10TableMin) within kPower10MantissaHighTable bounds"); kPower10TableMinInclusive,
"(exp10-kPower10TableMinInclusive) in kPower10MantissaHighTable bounds");
static_assert( static_assert(
FloatTraits<FloatType>::kEiselLemireMaxExclExp10 <= kPower10TableMax + 1, FloatTraits<FloatType>::kEiselLemireMaxExclusiveExp10 <=
"(exp10 - kPower10TableMin) within kPower10MantissaHighTable bounds"); kPower10TableMaxExclusive,
"(exp10-kPower10TableMinInclusive) in kPower10MantissaHighTable bounds");
// The terse (+) comments in this function body refer to sections of the // The terse (+) comments in this function body refer to sections of the
// https://nigeltao.github.io/blog/2020/eisel-lemire.html blog post. // https://nigeltao.github.io/blog/2020/eisel-lemire.html blog post.
@ -704,18 +703,19 @@ bool EiselLemire(const strings_internal::ParsedFloat& input, bool negative,
FloatTraits<FloatType>::kExponentBias - clz); FloatTraits<FloatType>::kExponentBias - clz);
// (+) Multiplication. // (+) Multiplication.
uint128 x = uint128 x = static_cast<uint128>(man) *
static_cast<uint128>(man) * static_cast<uint128>(
static_cast<uint128>(kPower10MantissaHighTable[exp10 - kPower10TableMin]); kPower10MantissaHighTable[exp10 - kPower10TableMinInclusive]);
// (+) Wider Approximation. // (+) Wider Approximation.
static constexpr uint64_t high64_mask = static constexpr uint64_t high64_mask =
FloatTraits<FloatType>::kEiselLemireMask; FloatTraits<FloatType>::kEiselLemireMask;
if (((Uint128High64(x) & high64_mask) == high64_mask) && if (((Uint128High64(x) & high64_mask) == high64_mask) &&
(man > (std::numeric_limits<uint64_t>::max() - Uint128Low64(x)))) { (man > (std::numeric_limits<uint64_t>::max() - Uint128Low64(x)))) {
uint128 y = static_cast<uint128>(man) * uint128 y =
static_cast<uint128>( static_cast<uint128>(man) *
kPower10MantissaLowTable[exp10 - kPower10TableMin]); static_cast<uint128>(
kPower10MantissaLowTable[exp10 - kPower10TableMinInclusive]);
x += Uint128High64(y); x += Uint128High64(y);
// For example, parsing "4503599627370497.5" will take the if-true // For example, parsing "4503599627370497.5" will take the if-true
// branch here (for double precision), since: // branch here (for double precision), since:
@ -926,13 +926,17 @@ from_chars_result from_chars(const char* first, const char* last, float& value,
namespace { namespace {
// Table of powers of 10, from kPower10TableMin to kPower10TableMax. // Table of powers of 10, from kPower10TableMinInclusive to
// kPower10TableMaxExclusive.
//
// kPower10MantissaHighTable[i - kPower10TableMinInclusive] stores the 64-bit
// mantissa. The high bit is always on.
//
// kPower10ExponentTable[i - kPower10TableMinInclusive] stores the power-of-two
// exponent.
// //
// kPower10MantissaHighTable[i - kPower10TableMin] stores the 64-bit mantissa // For a given number i, this gives the unique mantissa and exponent such that
// (high bit always on), and kPower10ExponentTable[i - kPower10TableMin] stores // (mantissa * 2**exponent) <= 10**i < ((mantissa + 1) * 2**exponent).
// the power-of-two exponent. For a given number i, this gives the unique
// mantissa and exponent such that mantissa * 2**exponent <= 10**i < (mantissa
// + 1) * 2**exponent.
// //
// kPower10MantissaLowTable extends that 64-bit mantissa to 128 bits. // kPower10MantissaLowTable extends that 64-bit mantissa to 128 bits.
// //

Write Preview
Loading…
Cancel
Save