Ada 中的 Unchecked_Conversion
任何人都可以让我清楚地了解 Ada 语言中未经检查的转换的使用。我已经尝试过 pdf 和 net，但所有这些都没有给我一个清晰的图片。
现在我有一小段代码如下所示：
subtype Element4_Range is integer range 1..4;
subtype Element3_Range is integer range 1..3;
subtype Myarr_Range is integer range 1..10;
type Myarr3_Type is array (Myarr_Range) of Element3_Range;
type Myarr4_Type is array (Myarr_Range) of Element4_Range;
Myarr3 : Myarr3_Type;
Myarr4 : Myarr4_Type := (1, 2, 3, 3, 1, 3, 2, 1, 2, 1);
Count_1 : Integer := 0;
Count_2 : Integer := 0;
Count_3 : Integer := 0;
*function To_Myarr3 is new Unchecked_Conversion(Myarr4_type, Myarr3_type);*
现在我的疑问是函数 Myarr3 到底做了什么？
Can anyone please make me clear about the use of unchecked conversion in the Ada language.I had tried the pdf and net but all doesn't give me a clear picture to me.
Now I have a small piece of code shown below:
subtype Element4_Range is integer range 1..4;
subtype Element3_Range is integer range 1..3;
subtype Myarr_Range is integer range 1..10;
type Myarr3_Type is array (Myarr_Range) of Element3_Range;
type Myarr4_Type is array (Myarr_Range) of Element4_Range;
Myarr3 : Myarr3_Type;
Myarr4 : Myarr4_Type := (1, 2, 3, 3, 1, 3, 2, 1, 2, 1);
Count_1 : Integer := 0;
Count_2 : Integer := 0;
Count_3 : Integer := 0;
*function To_Myarr3 is new Unchecked_Conversion(Myarr4_type, Myarr3_type);*
Now my doubt here is what does the function Myarr3 exactly do?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（2）
Unchecked_Conversion的实例将源值的字节复制到目标，而不检查这是否合理。如果值的大小不同，一些编译器会发出警告（可能取决于编译选项）。Element3_Range和Element4_Range均基于Integer，并且将使用相同的字节数；因此，您的两个数组变量（Myarr3、Myarr4）将需要相同数量的字节（通常为 40）。您可以这样写
，就目前情况而言，不会发生任何不好的事情，因为您用于初始化
Myarr4的所有值都与Element3_Range的值一样合法。但是，如果您这样做，
最终会得到
Myarr3(4)包含超出Element3_Range合法范围的值，并且编译器没有理由相信它可能无效。这很可能会导致Constraint_Error的出现。您可以自己强制检查：
An instantiation of
Unchecked_Conversioncopies the bytes of the source value to the target without checking whether this is sensible. Some compilers will warn (depending maybe on compilation options) if the values are of different sizes.Element3_RangeandElement4_Rangeare both based onIntegerand will use the same number of bytes; so both of your array variables (Myarr3,Myarr4) will need the same number of bytes (40 typically).You could write
As it stands, nothing bad would happen because all the values you've used to initialise
Myarr4are legal as values ofElement3_Range.However, if you had
you'd end up with
Myarr3(4)containing a value outside the legal range ofElement3_Range, and with the compiler having no reason to believe that it might not be valid. This may well lead toConstraint_Errors down the line.You could force a check yourself:
我曾经有一位同事坚持认为
unchecked_conversion应该被命名为“Unchcked_Copy”。它所做的只是将一种类型的对象复制到另一种类型的对象中。因此，您的
To_Myarr3例程将接受 Myarr4 类型的数组作为参数，假装它是 Myarr3 类型的数组，然后将其中的每个元素复制到表达式的左侧。如果你想将对象的视图从一种类型更改为另一种类型而不复制整个该死的东西，你可以在它们的访问类型上使用 Unchecked_Conversion （所以你只复制指向它们的指针）。另一种方法是使用
for object_name'address use at将一个覆盖在另一个上（但是，它们可能都会被初始化，这可能很糟糕）。但实际上最好的方法是设计系统的类型，以便您永远不必使用 Unchecked_Conversion。I had an coworker once who instisted that
unchecked_conversionshould have been named "Unchcked_Copy" instead. All it does is copy an object of one type into an object of another type.So your
To_Myarr3routine will accept as a parameter an array of type Myarr4, pretend it is an array of type Myarr3, and then copy every element in it into the left-hand side of your expression.If you want to change your view of an object from one type to another without copying the whole darn thing around, you can instead use an Unchecked_Conversion on access types for them (so you are only copying the pointer to them). Another method is using
for object_name'address use atto overlay one on the other (however, they may both get initialized, which can be bad). But really the best way is to design your system's types so that you never have to use Unchecked_Conversion.