refactor: update StringToBytes logic

Co-authored-by: zhuhebin <zhuhebin@fengtaisec.com>

random/random.go

@@ -8,20 +8,25 @@ import (
 	crand "crypto/rand"
 	"fmt"
 	"io"
+	"math"
 	"math/rand"
 	"time"
+	"unsafe"
 
 	"github.com/duke-git/lancet/v2/mathutil"
 )
 
 const (
-	Numeral       = "0123456789"
-	LowwerLetters = "abcdefghijklmnopqrstuvwxyz"
-	UpperLetters  = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
-	Letters       = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"
-	SymbolChars   = "!@#$%^&*()_+-=[]{}|;':\",./<>?"
+	MaximumCapacity = 1 << 31
+	Numeral         = "0123456789"
+	LowwerLetters   = "abcdefghijklmnopqrstuvwxyz"
+	UpperLetters    = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+	Letters         = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"
+	SymbolChars     = "!@#$%^&*()_+-=[]{}|;':\",./<>?"
 )
 
+var rn = rand.NewSource(time.Now().UnixNano())
+
 func init() {
 	rand.Seed(time.Now().UnixNano())
 }
@@ -108,18 +113,61 @@ func RandSymbolChar(length int) string {
 	return random(SymbolChars, length)
 }
 
+// nearestPowerOfTwo 返回一个大于等于cap的最近的2的整数次幂，参考java8的hashmap的tableSizeFor函数
+func nearestPowerOfTwo(cap int) int {
+	n := cap - 1
+	n |= n >> 1
+	n |= n >> 2
+	n |= n >> 4
+	n |= n >> 8
+	n |= n >> 16
+	if n < 0 {
+		return 1
+	} else if n >= MaximumCapacity {
+		return MaximumCapacity
+	}
+	return n + 1
+}
+
 // random generate a random string based on given string range.
 func random(s string, length int) string {
-	b := make([]byte, length)
-
-	// fix: https://github.com/duke-git/lancet/issues/75
-	// r := rand.New(rand.NewSource(time.Now().UnixNano()))
-
-	for i := range b {
-		b[i] = s[rand.Int63()%int64(len(s))]
+	// 仿照strings.Builder
+	// 创建一个长度为 length 的字节切片
+	bytes := make([]byte, length)
+	strLength := len(s)
+	if strLength <= 0 {
+		return ""
+	} else if strLength == 1 {
+		for i := 0; i < length; i++ {
+			bytes[i] = s[0]
+		}
+		return *(*string)(unsafe.Pointer(&bytes))
 	}
-
-	return string(b)
+	// s的字符需要使用多少个比特位数才能表示完
+	// letterIdBits := int(math.Ceil(math.Log2(strLength))),下面比上面的代码快
+	letterIdBits := int(math.Log2(float64(nearestPowerOfTwo(strLength))))
+	// 最大的字母id掩码
+	var letterIdMask int64 = 1<<letterIdBits - 1
+	// 可用次数的最大值
+	letterIdMax := 63 / letterIdBits
+	// 循环生成随机字符串
+	for i, cache, remain := length-1, rn.Int63(), letterIdMax; i >= 0; {
+		// 检查随机数生成器是否用尽所有随机数
+		if remain == 0 {
+			cache, remain = rn.Int63(), letterIdMax
+		}
+		// 从可用字符的字符串中随机选择一个字符
+		if idx := int(cache & letterIdMask); idx < len(s) {
+			bytes[i] = s[idx]
+			i--
+		}
+		// 右移比特位数，为下次选择字符做准备
+		cache >>= letterIdBits
+		remain--
+	}
+	// 仿照strings.Builder用unsafe包返回一个字符串，避免拷贝
+	// 将字节切片转换为字符串并返回
+	return *(*string)(unsafe.Pointer(&bytes))
 }
 
 // UUIdV4 generate a random UUID of version 4 according to RFC 4122.

strutil/string.go

@@ -4,7 +4,6 @@
 package strutil
 
 import (
-	"reflect"
 	"regexp"
 	"strings"
 	"unicode"
@@ -380,10 +379,7 @@ func RemoveNonPrintable(str string) string {
 // StringToBytes converts a string to byte slice without a memory allocation.
 // Play: https://go.dev/play/p/7OyFBrf9AxA
 func StringToBytes(str string) (b []byte) {
-	sh := *(*reflect.StringHeader)(unsafe.Pointer(&str))
-	bh := (*reflect.SliceHeader)(unsafe.Pointer(&b))
-	bh.Data, bh.Len, bh.Cap = sh.Data, sh.Len, sh.Len
-	return b
+	return *(*[]byte)(unsafe.Pointer(&str))
 }
 
 // BytesToString converts a byte slice to string without a memory allocation.