英文:
About unused variable in golang
问题
为什么如果我们在切片中追加数据并且不使用它（程序中的变量mySlice），为什么不会抛出"golang未使用的变量"错误消息。谢谢
#示例
var mySlice []string
mySlice = append(mySlice, "halo")
并且与映射(map)具有相同的行为
#示例
var my = map[string]int{}
my["tes"] = 1
英文:
why if we append data in slice and not use it (variable mySlice in the program), why not throw the error message "unused variable" golang. Thankyou
#example
var mySlice []string
mySlice = append(mySlice,"halo")
and same behaviour with map
#example
var my = map[string]int{}
my["tes"]=1
答案1
得分: 5
> 实现限制：如果变量从未被使用，编译器可能会禁止在函数体内声明变量。
正如你所看到的，编译器可能会接受或不接受未使用的变量。当前的Go编译器通过检查变量是否被读取来实现这一点。_读取_变量意味着变量被使用。
如果你不读取一个变量，编译器会给出"未使用的变量"（或者更准确地说是"声明但未使用的变量"）的错误。
例如：
var i int
i = 3
尽管在上面的例子中给i赋了一个值，但由于它没有被读取，这是一个编译时错误。
使用append()向切片追加元素涉及到读取切片（切片被传递给append()）。
给映射的键赋值涉及到读取映射的值。
如果不使用append()，而是直接给切片元素赋值：
var i = []int{0}
i[0] = 1
这是可以的，因为给i[0]赋值涉及到读取切片头部，以确定我们要赋值的元素的地址。
但是，如果我们像这样给i本身赋值：
var i = []int{0}
i = []int{1}
这个赋值不涉及到读取i，所以也是一个编译时错误。
如果我们使用结构体，例如：
type Point struct{ x, y int }
var p Point
p.x = 1
这仍然可以编译通过，即使赋值实际上并没有涉及读取p结构体。我们可以将其视为"识别" p.x 字段涉及到访问p变量。
这是一个有点"灰色地带"，基于Go语言作者的观点来决定和实现的。
英文:
> Implementation restriction: A compiler may make it illegal to declare a variable inside a function body if the variable is never used.
As you can see, a compiler may or may not accept unused variables. The current Go compiler implements this by checking if the variable is read. Reading a variable means the variable is used.
The "unused variable" (or rather "declared but not used") is given by the compiler if you don't read a variable.
For example:
var i int
i = 3
Even though you are assigning a value to i in the above example, since it is not read, this is a compile-time error.
Appending to a slice using append() involves reading the slice (the slice is passed to append()).
Assigning a value to a map key involves reading the map value.
Same goes if you don't use append() but "directly" assign a value to a slice element:
var i = []int{0}
i[0] = 1
This is OK because assigning to i[0] involves reading the slice header to determine the (address of the) element to which we're assigning.
But if we were to assign to i itself like this:
var i = []int{0}
i = []int{1}
This assignment doesn't involve reading i so it is also a compile-time error.
If we were to use a struct for example:
type Point struct{ x, y int }
var p Point
p.x = 1
This still compiles, even though the assignment doesn't really involve reading the p struct. We might think of it as "identifying" the p.x field involves accessing the p variable.
This is a little bit of "gray area" and is based on opinion of the Go authors, but this is how they decided and implemented it.
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论