unicode: speed-up is16/is32

TocarIP · TocarIP · commit bc8bb5b27a5a · 2017-09-12T21:32:50.000Z
Avoid division in common case. There are 5438 ranges in unicode/tables.go 4110 of them have stride 1. Stride 1 case got significantly faster. Other stride is a bit slower. Measured by import ( "testing" "unicode" ) func BenchmarkDiv1(b *testing.B) { rtb := &unicode.RangeTable{ R16: []unicode.Range16{ {0xa800, 0xdfff, 1}, // or 3 }, } for i := 0; i < b.N; i++ { unicode.Is(rtb, rune(0xc700)) } } Div1-6 15.6ns ± 1% 9.9ns ± 1% -36.54% (p=0.000 n=10+10) Div3-6 15.5ns ± 1% 16.1ns ± 1% +3.67% (p=0.000 n=10+10) Helps a bit with xml parsing from issue golang#21823 XMLsax-6 30.9s ± 0% 29.6s ± 0% -4.15% (p=0.000 n=10+9) Change-Id: Ibac1a91d7b9474d0c134b0add83e56caa62daa20 Reviewed-on: https://go-review.googlesource.com/63390 Run-TryBot: Ilya Tocar <ilya.tocar@intel.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
diff --git a/src/unicode/letter.go b/src/unicode/letter.go
@@ -97,7 +97,7 @@ func is16(ranges []Range16, r uint16) bool {
 				return false
 			}
 			if r <= range_.Hi {
-				return (r-range_.Lo)%range_.Stride == 0
+				return range_.Stride == 1 || (r-range_.Lo)%range_.Stride == 0
 			}
 		}
 		return false
@@ -110,7 +110,7 @@ func is16(ranges []Range16, r uint16) bool {
 		m := lo + (hi-lo)/2
 		range_ := &ranges[m]
 		if range_.Lo <= r && r <= range_.Hi {
-			return (r-range_.Lo)%range_.Stride == 0
+			return range_.Stride == 1 || (r-range_.Lo)%range_.Stride == 0
 		}
 		if r < range_.Lo {
 			hi = m
@@ -130,7 +130,7 @@ func is32(ranges []Range32, r uint32) bool {
 				return false
 			}
 			if r <= range_.Hi {
-				return (r-range_.Lo)%range_.Stride == 0
+				return range_.Stride == 1 || (r-range_.Lo)%range_.Stride == 0
 			}
 		}
 		return false
@@ -143,7 +143,7 @@ func is32(ranges []Range32, r uint32) bool {
 		m := lo + (hi-lo)/2
 		range_ := ranges[m]
 		if range_.Lo <= r && r <= range_.Hi {
-			return (r-range_.Lo)%range_.Stride == 0
+			return range_.Stride == 1 || (r-range_.Lo)%range_.Stride == 0
 		}
 		if r < range_.Lo {
 			hi = m

Original file line number	Diff line number	Diff line change
`@@ -97,7 +97,7 @@ func is16(ranges []Range16, r uint16) bool {`
`97`	`97`	`return false`
`98`	`98`	`}`
`99`	`99`	`if r <= range_.Hi {`
`100`		`- return (r-range_.Lo)%range_.Stride == 0`
	`100`	`+ return range_.Stride == 1 \|\| (r-range_.Lo)%range_.Stride == 0`
`101`	`101`	`}`
`102`	`102`	`}`
`103`	`103`	`return false`
`@@ -110,7 +110,7 @@ func is16(ranges []Range16, r uint16) bool {`
`110`	`110`	`m := lo + (hi-lo)/2`
`111`	`111`	`range_ := &ranges[m]`
`112`	`112`	`if range_.Lo <= r && r <= range_.Hi {`
`113`		`- return (r-range_.Lo)%range_.Stride == 0`
	`113`	`+ return range_.Stride == 1 \|\| (r-range_.Lo)%range_.Stride == 0`
`114`	`114`	`}`
`115`	`115`	`if r < range_.Lo {`
`116`	`116`	`hi = m`
`@@ -130,7 +130,7 @@ func is32(ranges []Range32, r uint32) bool {`
`130`	`130`	`return false`
`131`	`131`	`}`
`132`	`132`	`if r <= range_.Hi {`
`133`		`- return (r-range_.Lo)%range_.Stride == 0`
	`133`	`+ return range_.Stride == 1 \|\| (r-range_.Lo)%range_.Stride == 0`
`134`	`134`	`}`
`135`	`135`	`}`
`136`	`136`	`return false`
`@@ -143,7 +143,7 @@ func is32(ranges []Range32, r uint32) bool {`
`143`	`143`	`m := lo + (hi-lo)/2`
`144`	`144`	`range_ := ranges[m]`
`145`	`145`	`if range_.Lo <= r && r <= range_.Hi {`
`146`		`- return (r-range_.Lo)%range_.Stride == 0`
	`146`	`+ return range_.Stride == 1 \|\| (r-range_.Lo)%range_.Stride == 0`
`147`	`147`	`}`
`148`	`148`	`if r < range_.Lo {`
`149`	`149`	`hi = m`