5.5.1. Reference Type Casting

Given a compile-time reference type S (source) and a compile-time reference type T (target), a casting conversion exists from S to T if no compile-time errors occur due to the following rules.

If S is a class type:

  • If T is a class type, then either |S| <: |T|, or |T| <: |S|. Otherwise, a compile-time error occurs.

    Furthermore, if there exists a supertype X of T, and a supertype Y of S, such that both X and Y are provably distinct parameterized types (§4.5), and that the erasures of X and Y are the same, a compile-time error occurs.

package test5;

interface II {}

class XY<T> {}

class T extends XY<Integer>implements II {}

class S extends XY<Number>implements II {}

public class main1 {

	public static void main(String args[]) {
		II a = new S();
		T t = (T) a; // test5.S cannot be cast to test5.T
	}
}
  • If T is an interface type:

    • If S is not a final class (§8.1.1), then, if there exists a supertype X of T, and a supertype Y of S, such that both X and Y are provably distinct parameterized types, and that the erasures of X and Y are the same, a compile-time error occurs.

      Otherwise, the cast is always legal at compile time (because even if S does not implement T, a subclass of S might).

    • If S is a final class (§8.1.1), then S must implement T, or a compile-time error occurs.

  • If T is a type variable, then this algorithm is applied recursively, using the upper bound of T in place of T.

interface I{}
class S implements I {}

public class main2<B extends I,T extends B> {
   public void test(){
	   T d = (T)new S(); // Type safety: Unchecked cast from S to D
   }
}
  • If T is an array type, then S must be the class Object, or a compile-time error occurs.

If S is an interface type:

  • If T is an array type, then S must be the type java.io.Serializable or Cloneable (the only interfaces implemented by arrays), or a compile-time error occurs.

  • If T is a type that is not final (§8.1.1), then if there exists a supertype X of T, and a supertype Y of S, such that both X and Y are provably distinct parameterized types, and that the erasures of X and Y are the same, a compile-time error occurs.

    Otherwise, the cast is always legal at compile time (because even if T does not implement S, a subclass of T might).

  • If T is a type that is final, then:

    • If S is not a parameterized type or a raw type, then T must implement S, or a compile-time error occurs.

    • Otherwise, S is either a parameterized type that is an invocation of some generic type declaration G, or a raw type corresponding to a generic type declaration G. Then there must exist a supertype X of T, such that X is an invocation of G, or a compile-time error occurs.

      Furthermore, if S and X are provably distinct parameterized types then a compile-time error occurs.

If S is a type variable,

then this algorithm is applied recursively, using the upper bound of S in place of S.

If S is an intersection type A1 & ... & An,

then it is a compile-time error if there exists an Ai (1 ≤ i ≤ n) such that S cannot be cast to Ai by this algorithm. That is, the success of the cast is determined by the most restrictive component of the intersection type.

If S is an array type SC[], that is, an array of components of type SC:

  • If T is a class type, then if T is not Object, then a compile-time error occurs (because Object is the only class type to which arrays can be assigned).

  • If T is an interface type, then a compile-time error occurs unless T is the type java.io.Serializable or the type Cloneable (the only interfaces implemented by arrays).

  • If T is a type variable, then:

    • If the upper bound of T is Object or java.io.Serializable or Cloneable, or a type variable that S could undergo casting conversion to, then the cast is legal (though unchecked).

    • If the upper bound of T is an array type TC[], then a compile-time error occurs unless the type SC[] can undergo casting conversion to TC[].

    • Otherwise, a compile-time error occurs.

  • If T is an array type TC[], that is, an array of components of type TC, then a compile-time error occurs unless one of the following is true:

    • TC and SC are the same primitive type.

    • TC and SC are reference types and type SC can undergo casting conversion to TC.

Reference Type Casting的更多相关文章

  1. A const field of a reference type other than string can only be initialized with null Error [duplicate]

    I'm trying to create a 2D array to store some values that don't change like this. const int[,] hiveI ...

  2. C#中的值类型(value type)与引用类型(reference type)的区别

    ylbtech- .NET-Basic:C#中的值类型与引用类型的区别 C#中的值类型(value type)与引用类型(reference type)的区别 1.A,相关概念返回顶部     C#中 ...

  3. Welcome-to-Swift-18类型转换(Type Casting)

    类型转换是一种检查类实例的方式,并且哦或者也是让实例作为它的父类或者子类的一种方式. Type casting is a way to check the type of an instance, a ...

  4. 引用类型 (Reference Type Matters)、扩展与派发方式

    引用类型 (Reference Type Matters) 引用的类型决定了派发的方式. 这很显而易见, 但也是决定性的差异. 一个比较常见的疑惑, 发生在一个协议拓展和类型拓展同时实现了同一个函数的 ...

  5. 快速了解C# 8.0中“可空引用类型(Nullable reference type)”语言特性

    Visual C# 8.0中引入了可空引用类型(Nullable reference type),通过编译器提供的强大功能,帮助开发人员尽可能地规避由空引用带来的代码问题.这里我大致介绍一下可空引用类 ...

  6. java中Number Type Casting(数字类型强转)的用法

    4.5 Number Type Casting(数字类型强转)隐式 casting(from small to big) byte a = 111; int b = a;显式 casting(from ...

  7. system verilog中的类型转换(type casting)、位宽转换(size casting)和符号转换(sign casting)

    类型转换 verilog中,任何类型的任何数值都用来给任何类型赋值.verilog使用赋值语句自动将一种类型的数值转换为另一种类型. 例如,当一个wire类型赋值给一个reg类型的变量时,wire类型 ...

  8. [Java in NetBeans] Lesson 03. More Variables / Type Casting

    这个课程的参考视频在youtube. 主要学到的知识点有: It is different from python, that "1" only present string &q ...

  9. C# Value type vs Reference type

    [MY NOTE]   [转载请注明出处] Reference Source: http://www.albahari.com/valuevsreftypes.aspx http://www.c-sh ...

随机推荐

  1. Android-Sqlite-OOP方式操作增删改查

    之前写的数据库增删改查,是使用SQL语句来实现的,Google 就为Android开发人员考虑,就算不会SQL语句也能实现增删改查,所以就有了OOP面向对象的增删改查方式 其实这种OOP面向对象的增删 ...

  2. Objective-C 学习笔记(二) 函数

    Objective-C 函数 定义一个方法 在Objective-C编程的方法定义的一般形式如下: - (return_type) method_name:( argumentType1 )argum ...

  3. Postgres 主从配置(四)

    Postgres 主从切换 数据库主从结构中由从库升级为主库较为容易些,但是主库恢复后重新加入到主从结构中就不那么容易了.以往的做法是当成一个全新的从库加入进来,数据需要重新从现有的主库中使用pg_b ...

  4. mybatis 使用tips - 使用多个参数

    执行如下命令: mvn -Dmybatis.generator.overwrite=true mybatis-generator:generate 可以使用mybatis generator myba ...

  5. 从0开始学习Unity的学习笔记(I 界面学习和简单模型拼装)

    先给一个大致今天学习的图,然后后面是细节 1.下载Unity :官网下载需要版本 2.Unity安装:一定不要有中文路径:一台电脑可以安装不同版本的Unity,但是要安装在不同的文件夹下: 3. 新建 ...

  6. 在VC++中执行VBS代码

    此代码来自https://blog.csdn.net/zhu2695/article/details/13770671 作者: zhu2695   时间:2013年10月31日 13:08:41 #i ...

  7. SpringMvc date数据传递处理

    1.form表单提交date数据 form表单提交的date数据要在接收的字段上加入@DateTimeFormat注解 @DateTimeFormat(pattern = "yyyy-MM- ...

  8. robot framework接口测试之二-四种常见的POST提交数据方式

    写接口测试用例时,遇到以json格式提交数据时,报错,Request如下图: Response如下图: 改成form格式提交,可以正常运行,如下图: 代码如下: ------------------- ...

  9. [CISCO] VLAN、TRUNK 和 VTP 简介

    VLAN.TRUNK 和 VTP 简介 VLAN 如图,虚拟局域网 VLAN ( Virtual LAN ) 是交换机端口的逻辑组合. VLAN 工作在 OSI 的第 2 层(数据链路层),一个 VL ...

  10. 【ASP.NET Core MVC 入门须知】Net Core和Net Framework 的区别

     1.简单介绍 从上面图中我们可以看到.net  主要分为三个部分 .net FrameWork,.net Core ,Xamarin XAMARIN  主要用来构建APP的主要用的是C#语言 .NE ...