Function equalStr

bytes_unsafe.go:22–24

The reason for using *[]byte rather than []byte in parameters is an optimization. As of Go 1.6, the compiler cannot perfectly inline the function when using a non-pointer slice. That is, the non-pointer []byte parameter version is slower than if its function body is manually inlined, whereas the po

(b *[]byte, s string)

// TODO: Remove hack after Go 1.7 release
//
func equalStr(b *[]byte, s string) bool {
	return *(*string)(unsafe.Pointer(b)) == s
}

func parseFloat(b *[]byte) (float64, error) {
	return strconv.ParseFloat(*(*string)(unsafe.Pointer(b)), 64)