优化生成随机字符串

[pigwantacat]

优化随机字符串生成

背景

原来的随机字符串生成代码在Windows环境下就存在BUG，可参考Function RandomChars has bug · Issue #121 · gookit/goutil (github.com)

在Windows本地测试发现如果使用for循环重复生成随机字符串存在相同的情况,代码如下：

func main() {
	for i := 0; i < 10; i++ {
		println(strutil.RandomCharsV3(10))
	}
}

输出打印如下：

GOROOT=C:\Users\zhb15\sdk\go1.21.3 #gosetup
GOPATH=D:\Go\WorkSpace #gosetup
C:\Users\zhb15\sdk\go1.21.3\bin\go.exe build -o C:\Users\zhb15\AppData\Local\JetBrains\GoLand2023.3\tmp\GoLand\___go_build_mytest_com_zhb_utils.exe mytest/com/zhb/utils #gosetup
C:\Users\zhb15\AppData\Local\JetBrains\GoLand2023.3\tmp\GoLand\___go_build_mytest_com_zhb_utils.exe
p35cdB4ntT
p35cdB4ntT
p35cdB4ntT
p35cdB4ntT
p35cdB4ntT
p35cdB4ntT
p35cdB4ntT
qgHpMOu2R6
qgHpMOu2R6
qgHpMOu2R6

进程 已完成，退出代码为 0

正好最近看到一篇关于随机字符串生成的文章，原文链接：Golang生成随机字符串的八种方式与性能测试-云社区-华为云 (huaweicloud.com)，How to generate a random string of a fixed length in Go? - Stack Overflow

所以打算根据文章中提到的方法优化strutil的随机字符串生成方法

实现

新增的方法有2个nearestPowerOfTwo和buildRandomString，nearestPowerOfTwo方法是求大于等于字符模板的最近的2的整数次幂，这个方法参考java8的hashmap的tableSizeFor函数，可参考文章：HashMap部分源码简析 | 柳门竹巷 (zhbblog.top)

/**
 * Returns a power of two size for the given target capacity.
 */
//返回大于输入参数且最近的2的整数次幂的数。比如10，则返回16
static final int tableSizeFor(int cap) {
    int n = cap - 1;
    n |= n >>> 1;
    n |= n >>> 2;
    n |= n >>> 4;
    n |= n >>> 8;
    n |= n >>> 16;
    return (n < 0) ? 1 : (n >= MAXIMUM_CAPACITY) ? MAXIMUM_CAPACITY : n + 1;
}

buildRandomString主要用于根据输入的字符串模板，生成指定长度的随机字符串，具体原理请参考上诉链接

结果

测试结果

现在在Windows本地测试结果如下,代码：

func main() {
	for i := 0; i < 10; i++ {
		println(testUtil.RandomCharsV3(10))
	}
}

输出打印如下：

GOROOT=C:\Users\zhb15\sdk\go1.21.3 #gosetup
GOPATH=D:\Go\WorkSpace #gosetup
C:\Users\zhb15\sdk\go1.21.3\bin\go.exe build -o C:\Users\zhb15\AppData\Local\JetBrains\GoLand2023.3\tmp\GoLand\___go_build_mytest_com_zhb_utils.exe mytest/com/zhb/utils #gosetup
C:\Users\zhb15\AppData\Local\JetBrains\GoLand2023.3\tmp\GoLand\___go_build_mytest_com_zhb_utils.exe
e4dKh62ENr
cZfsXSTxhB
n8lGE8rE7f
ftiDJX8NHU
viz89Ur6Bb
j469ymTSQr
fyfcHHeHvq
wDoZelPwlL
BzRHQSgPIN
LzybA6Jz1e

进程 已完成，退出代码为 0

nearestPowerOfTwo方法

返回大于输入参数且最近的2的整数次幂的数方法测试如下，代码：

func BenchmarkApproach1(b *testing.B) {
    for i := 0; i < b.N; i++ {
       _ = int(math.Log2(float64(nearestPowerOfTwo(b.N))))
    }
}

func BenchmarkApproach2(b *testing.B) {
    for i := 0; i < b.N; i++ {
       _ = math.Ceil(math.Log2(float64(b.N)))
    }
}

输入打印如下：

$ go test -benchmem -bench .
goos: windows
goarch: amd64
pkg: mytest/com/zhb/utils
cpu: 12th Gen Intel(R) Core(TM) i5-12600KF
BenchmarkApproach1-16           277866668                4.158 ns/op           0 B/op          0 allocs/op
BenchmarkApproach2-16           151868052                7.809 ns/op           0 B/op          0 allocs/op
PASS
ok      mytest/com/zhb/utils    4.316s

优化前后性能对比

优化前后性能对比，代码如下：

func BenchmarkApproach3(b *testing.B) {
	for i := 0; i < b.N; i++ {
		_ = testUtil.RandomChars(1 << 10)
	}
}

func BenchmarkApproach4(b *testing.B) {
	for i := 0; i < b.N; i++ {
		_ = strutil.RandomChars(1 << 10)
	}
}

func BenchmarkApproach5(b *testing.B) {
	for i := 0; i < b.N; i++ {
		_ = testUtil.RandomCharsV2(1 << 10)
	}
}

func BenchmarkApproach6(b *testing.B) {
	for i := 0; i < b.N; i++ {
		_ = strutil.RandomCharsV2(1 << 10)
	}
}

func BenchmarkApproach7(b *testing.B) {
	for i := 0; i < b.N; i++ {
		_ = testUtil.RandomCharsV3(1 << 10)
	}
}

func BenchmarkApproach8(b *testing.B) {
	for i := 0; i < b.N; i++ {
		_ = strutil.RandomCharsV3(1 << 10)
	}
}

输出打印如下：

$ go test -benchmem -bench .
goos: windows
goarch: amd64
pkg: mytest/com/zhb/utils
cpu: 12th Gen Intel(R) Core(TM) i5-12600KF
BenchmarkApproach3-16             288414              3954 ns/op            1024 B/op          1 allocs/op
BenchmarkApproach4-16             108159             11052 ns/op            7472 B/op          4 allocs/op
BenchmarkApproach5-16             136956              8588 ns/op            1024 B/op          1 allocs/op
BenchmarkApproach6-16             107840             11156 ns/op            7472 B/op          4 allocs/op
BenchmarkApproach7-16             749732              1783 ns/op            1024 B/op          1 allocs/op
BenchmarkApproach8-16             107077             11216 ns/op            7472 B/op          4 allocs/op
PASS
ok      mytest/com/zhb/utils    8.544s

优化前后准确性对比

优化前后准确性对比，代码如下：

func TestApproach1(t *testing.T) {
	println("Approach1-testUtil")
	for i := 0; i < 10; i++ {
		println(testUtil.RandomChars(32))
	}
	println("Approach1-strUtil")
	for i := 0; i < 10; i++ {
		println(strutil.RandomChars(32))
	}
}

func TestApproach2(t *testing.T) {
	println("Approach2-testUtil")
	for i := 0; i < 10; i++ {
		println(testUtil.RandomCharsV2(32))
	}
	println("Approach2-strUtil")
	for i := 0; i < 10; i++ {
		println(strutil.RandomCharsV2(32))
	}
}

func TestApproach3(t *testing.T) {
	println("Approach3-testUtil")
	for i := 0; i < 10; i++ {
		println(testUtil.RandomCharsV3(32))
	}
	println("Approach3-strUtil")
	for i := 0; i < 10; i++ {
		println(strutil.RandomCharsV3(32))
	}
}

输出打印如下：

$ go test -benchmem -bench .
Approach1-testUtil
qfradxxxugivfeqdndkmwablgblbpmrq
iawlwilkjsyfbfyibiesrgcvxcktihqx
uvlimlwfvheezhoajjmugufdnijfonrw
heotinzkjwydwefjxxjrlbwajstherqo
klzwlmbmoqnhesinxduvnwarcdztwfzo
qtxknpgwkaecthmkcgpylqkqhblnwghv
qrmkukeiigoogfnuglhnelhoodctnnor
ozwjggaoccrujwsydzurhppvutcotaok
ezymohbmmkgfioujstqddlumydtoktji
ucsisvhejomsefnkmmfjwwidyxmifycm
Approach1-strUtil
nwjhtxwrojjrpumddjfmkuxxksxpturm
nwjhtxwrojjrpumddjfmkuxxksxpturm
nwjhtxwrojjrpumddjfmkuxxksxpturm
cxtldieyfuevrlqbrhomdsyahfdfbhml
cxtldieyfuevrlqbrhomdsyahfdfbhml
cxtldieyfuevrlqbrhomdsyahfdfbhml
cxtldieyfuevrlqbrhomdsyahfdfbhml
cxtldieyfuevrlqbrhomdsyahfdfbhml
cxtldieyfuevrlqbrhomdsyahfdfbhml
cxtldieyfuevrlqbrhomdsyahfdfbhml
Approach2-testUtil
fut0tk6qyg6zas3w428bg9xo79a8apek
2xd7779ws7dn3kto63vfycwiva9kwhf2
c5oi8awgiuyza7cba8s4sjw0qmygciu9
8w9d2ng9gndb6o99ymbcfcakjehtauxk
c0zsh48oirvzvq8u0q6z5sxonmw4w89y
4so6b8fqlyhhiulqz4czfckn9mv5foxf
qzigv9vbk1byyew2k22n9on7qg0xtr7g
8bsnlqcy12qk6m2ls83xf6ro6x2m1tz1
9ielerjzg75ugjhq10lr0u8rwer6hwpf
3zcl0w0q0c10vqyicef2mpsx2dgwfvy3
Approach2-strUtil
yhjgeb38po4d5k1511bl6x5osxtx0vnj
yhjgeb38po4d5k1511bl6x5osxtx0vnj
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
4vlcpg5j75j0qrc2haocto40tfs30nah
Approach3-testUtil
gPcjM6IBQ9CsrMqcHNGlSGHJS9sG7jgG
8ZqiXVkryXPmxq0lIXw03gyKMVxz1u89
gk2qzEN1GNQ57vkBOy6jWQs3VhlhHCBQ
ZtpmUIfuhmNE6X9jrmZumFpvEl8hAmzh
oA5trh14GyS59O0spbdVrqFSKdaGmcg7
UhqZTGllUjA28AVpUmiajH61y0aNNmpT
rDwSgIQPJauBzzLO2Ka9el5MgoP4nNso
CEoqbiQfT8M1pEIB9QTYzR5r0LCpQCtn
2fNUKyRiq7Zl3f3bfIw16EKiyHSZDG3J
SpM0gUnE31p01r6Ixth8HXFyZnkLQneB
Approach3-strUtil
taNxbaWkYLwfFF50ko3mLtll1BX1gW2K
taNxbaWkYLwfFF50ko3mLtll1BX1gW2K
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
By4WjX8QiNPz10b38DIHcFn0KJQJsMrB
PASS
ok      mytest/com/zhb/utils    0.044s

优化后实际效果仍有待完整测试