From 3d48ae355b64957c184e220161f5efb1fe51f5ad Mon Sep 17 00:00:00 2001 From: Sam Arnold Date: Sat, 10 Aug 2019 12:06:50 -0400 Subject: [PATCH] strconv: Speed improvement to number parsing MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit Run underscore validation only if we have seen underscores. Some performance results on my laptop: name old time/op new time/op delta Atof64Decimal-12 30.5ns ± 0% 23.8ns ± 0% -22.02% (p=0.016 n=5+4) Atof64Float-12 39.0ns ± 0% 28.7ns ± 0% -26.39% (p=0.002 n=6+6) Atof64FloatExp-12 64.4ns ± 1% 54.4ns ± 1% -15.65% (p=0.002 n=6+6) Atof64Big-12 115ns ± 1% 87ns ± 1% -24.45% (p=0.002 n=6+6) Atof64RandomBits-12 187ns ±14% 156ns ±19% -16.46% (p=0.032 n=6+6) Atof64RandomFloats-12 126ns ± 0% 105ns ± 1% -16.65% (p=0.000 n=6+5) Atof32Decimal-12 32.0ns ± 1% 24.0ns ± 1% -24.97% (p=0.002 n=6+6) Atof32Float-12 37.1ns ± 1% 27.0ns ± 1% -27.42% (p=0.002 n=6+6) Atof32FloatExp-12 68.4ns ± 1% 54.2ns ± 1% -20.77% (p=0.002 n=6+6) Atof32Random-12 92.0ns ± 1% 77.4ns ± 0% -15.81% (p=0.000 n=6+5) ParseInt/Pos/7bit-12 19.4ns ± 1% 13.8ns ±10% -28.94% (p=0.002 n=6+6) ParseInt/Pos/26bit-12 29.1ns ± 1% 19.8ns ± 2% -31.92% (p=0.002 n=6+6) ParseInt/Pos/31bit-12 33.1ns ± 0% 22.3ns ± 3% -32.62% (p=0.004 n=5+6) ParseInt/Pos/56bit-12 47.8ns ± 1% 30.7ns ± 1% -35.78% (p=0.004 n=6+5) ParseInt/Pos/63bit-12 51.9ns ± 1% 33.4ns ± 2% -35.49% (p=0.002 n=6+6) ParseInt/Neg/7bit-12 18.5ns ± 4% 13.4ns ± 3% -27.88% (p=0.002 n=6+6) ParseInt/Neg/26bit-12 28.4ns ± 3% 19.7ns ± 3% -30.38% (p=0.002 n=6+6) ParseInt/Neg/31bit-12 31.9ns ± 1% 21.8ns ± 2% -31.56% (p=0.002 n=6+6) ParseInt/Neg/56bit-12 46.2ns ± 0% 30.6ns ± 1% -33.73% (p=0.004 n=5+6) ParseInt/Neg/63bit-12 50.2ns ± 1% 33.2ns ± 1% -33.96% (p=0.002 n=6+6) Fixes #33330 Change-Id: I119da66457c2fbaf6e88bb90cf56417a16df8f0e Reviewed-on: https://go-review.googlesource.com/c/go/+/187957 Run-TryBot: Robert Griesemer TryBot-Result: Gobot Gobot Reviewed-by: Robert Griesemer --- src/strconv/atof.go | 37 ++++++++++++++++++++++++------------- src/strconv/atoi.go | 9 +++++++-- 2 files changed, 31 insertions(+), 15 deletions(-) diff --git a/src/strconv/atof.go b/src/strconv/atof.go index 190b25fbf5..23de70b1c9 100644 --- a/src/strconv/atof.go +++ b/src/strconv/atof.go @@ -84,7 +84,7 @@ func (b *decimal) set(s string) (ok bool) { for ; i < len(s); i++ { switch { case s[i] == '_': - // underscoreOK already called + // readFloat already checked underscores continue case s[i] == '.': if sawdot { @@ -140,7 +140,7 @@ func (b *decimal) set(s string) (ok bool) { e := 0 for ; i < len(s) && ('0' <= s[i] && s[i] <= '9' || s[i] == '_'); i++ { if s[i] == '_' { - // underscoreOK already called + // readFloat already checked underscores continue } if e < 10000 { @@ -159,10 +159,11 @@ func (b *decimal) set(s string) (ok bool) { } // readFloat reads a decimal mantissa and exponent from a float -// string representation. It returns ok==false if the number could -// not fit return types or is invalid. +// string representation. It returns ok==false if the number +// is invalid. func readFloat(s string) (mantissa uint64, exp int, neg, trunc, hex, ok bool) { i := 0 + underscores := false // optional sign if i >= len(s) { @@ -195,7 +196,7 @@ func readFloat(s string) (mantissa uint64, exp int, neg, trunc, hex, ok bool) { for ; i < len(s); i++ { switch c := s[i]; true { case c == '_': - // underscoreOK already called + underscores = true continue case c == '.': @@ -271,7 +272,7 @@ func readFloat(s string) (mantissa uint64, exp int, neg, trunc, hex, ok bool) { e := 0 for ; i < len(s) && ('0' <= s[i] && s[i] <= '9' || s[i] == '_'); i++ { if s[i] == '_' { - // underscoreOK already called + underscores = true continue } if e < 10000 { @@ -291,6 +292,11 @@ func readFloat(s string) (mantissa uint64, exp int, neg, trunc, hex, ok bool) { if mantissa != 0 { exp = dp - ndMant } + + if underscores && !underscoreOK(s) { + return + } + ok = true return } @@ -554,12 +560,16 @@ func atof32(s string) (f float32, err error) { } mantissa, exp, neg, trunc, hex, ok := readFloat(s) - if hex && ok { + if !ok { + return 0, syntaxError(fnParseFloat, s) + } + + if hex { f, err := atofHex(s, &float32info, mantissa, exp, neg, trunc) return float32(f), err } - if optimize && ok { + if optimize { // Try pure floating-point arithmetic conversion. if !trunc { if f, ok := atof32exact(mantissa, exp, neg); ok { @@ -597,11 +607,15 @@ func atof64(s string) (f float64, err error) { } mantissa, exp, neg, trunc, hex, ok := readFloat(s) - if hex && ok { + if !ok { + return 0, syntaxError(fnParseFloat, s) + } + + if hex { return atofHex(s, &float64info, mantissa, exp, neg, trunc) } - if optimize && ok { + if optimize { // Try pure floating-point arithmetic conversion. if !trunc { if f, ok := atof64exact(mantissa, exp, neg); ok { @@ -658,9 +672,6 @@ func atof64(s string) (f float64, err error) { // ParseFloat recognizes the strings "NaN", "+Inf", and "-Inf" as their // respective special floating point values. It ignores case when matching. func ParseFloat(s string, bitSize int) (float64, error) { - if !underscoreOK(s) { - return 0, syntaxError(fnParseFloat, s) - } if bitSize == 32 { f, err := atof32(s) return float64(f), err diff --git a/src/strconv/atoi.go b/src/strconv/atoi.go index e811bc4df1..131b088e31 100644 --- a/src/strconv/atoi.go +++ b/src/strconv/atoi.go @@ -58,7 +58,7 @@ const maxUint64 = 1<<64 - 1 func ParseUint(s string, base int, bitSize int) (uint64, error) { const fnParseUint = "ParseUint" - if s == "" || !underscoreOK(s) { + if s == "" { return 0, syntaxError(fnParseUint, s) } @@ -113,12 +113,13 @@ func ParseUint(s string, base int, bitSize int) (uint64, error) { maxVal := uint64(1)<