C语言中的强制类型转换分析

自动类型转换

常见的自动类型转换分两种情况：

1，不同类型的数据进行运算时，系统会自动将运算优先级别低的转换成级别高的，结果和运算级别高的数据类型一样。

2，赋值运算符两边类型不一样时，系统会自动将右侧表达式的类型转换成左侧的变量类型，然后再赋值

复制运算的时候
左边 <= 右边
大A <= 小a；

自动转化为：
大A <= 大A；

强制类型转换
小a <= 大A；

会有数据损失【蓝色】

下面设计程序做详细的分析：

【类型A】num0 = 【类型B】（强制类型转换）（num1 / num 2）

	 double a;
	 int A;

	 a = 2/3;
	 printf( "a = (2 / 3 f) = %f\n",a );
	 printf( "a = (2 / 3 d) = %d\n\n",a );

	 a = (double) 2/3;
	 printf( "a = (double)(2 / 3 f) = %f\n",a );
	 printf( "a = (double)(2 / 3 d) = %d\n\n",a );

	 a = 3/2;
	 printf( "a = (3 / 2 f) = %f\n",a );
	 printf( "a = (3 / 2 d) = %d\n\n",a );

	 a = (double) 3/2;
	 printf( "a = (double)(3 / 2 f) = %f\n",a );
	 printf( "a = (double)(3 / 2 d) = %d\n\n",a );

	 printf( "*-----------------*\n" );

	 A = 2/3;
	 printf( "A = (2 / 3 f) = %f\n",A );
	 printf( "A = (2 / 3 d) = %d\n\n",A );

	 A = (int)2/3.0;
	 printf( "A = (int)(2 / 3 f) = %f\n",A );
	 printf( "A = (int)(2 / 3 d) = %d\n\n",A );

	 A = 3/2;
	 printf( "A = (3 / 2 f) = %f\n",A );
	 printf( "A = (3 / 2 d) = %d\n\n",A );

	 A = (int)3/2.0;
	 printf( "A = (int)(3 / 2 f) = %f\n",A );
	 printf( "A = (int)(3 / 2 d) = %d\n\n",A );

	 printf( "*-----------------*\n" );

	 A = 2/1;
	 printf( "A = (2 / 1 f) = %f\n",A );
	 printf( "A = (2 / 1 d) = %d\n\n",A );

	 A = (int)2/1.0;
	 printf( "A = (int)(2 / 1.0 f) = %f\n",A );
	 printf( "A = (int)(2 / 1.0 d) = %d\n\n",A );

	 A = 3/1;
	 printf( "A = (2 / 1 f) = %f\n",A );
	 printf( "A = (2 / 1 d) = %d\n\n",A );

	 A = (int)2/1.0;
	 printf( "A = (int)(2 / 1.0 f) = %f\n",A );
	 printf( "A = (int)(2 / 1.0 d) = %d\n\n",A );

输出结果：
a = (2 / 3 f) = 0.000000
*//1（ 2 / 3 = 0，再转换为0.00000，赋给double_a,以 “ % f” 浮点型格式输出0.000000 ）
a = (2 / 3 d) = 0
*//2（ 2 / 3 = 0，再转换为0.00000，赋给double_a,以 “% d” 整型格式输出 0 ）

a = (double)(2 / 3 f) = 0.666667
*//3（ 先将 2 和 3 都转换为double型，2.0 / 3.0 = 0.666667，赋给double_a,以 “ % f” 浮点型格式输出0.666667 ）
a = (double)(2 / 3 d) = 1431655765
//*4(目前还不清楚)

a = (3 / 2 f) = 1.000000
//同 2
a = (3 / 2 d) = 0
//0不论计算结果非小数位是否为0，int型结果以 %f 和 double型结果以 %d 输出的时候，结果都是0，或者0.000000；

a = (double)(3 / 2 f) = 1.500000
//正确
a = (double)(3 / 2 d) = 0
//强制类型转换后 a = 1.5,但是以%d 整型输出，同0

-----------------
A = (2 / 3 f) = 0.000000
//同0
A = (2 / 3 d) = 0
//正确

A = (int)(2 / 3 f) = 0.000000
//同0
A = (int)(2 / 3 d) = 0
//同2

A = (3 / 2 f) = 0.000000
//同0
A = (3 / 2 d) = 1
//同1

A = (int)(3 / 2 f) = 0.000000
//同0
A = (int)(3 / 2 d) = 1
//和上面的对比

-----------------
A = (2 / 1 f) = 0.000000
//同0
A = (2 / 1 d) = 2
//正确

A = (int)(2 / 1.0 f) = 0.000000
//同0
A = (int)(2 / 1.0 d) = 2
//正确

A = (2 / 1 f) = 0.000000
//同0
A = (2 / 1 d) = 3
//正确

A = (int)(2 / 1.0 f) = 0.000000
//同0
A = (int)(2 / 1.0 d) = 2
/正确
Press any key to continue

补充
针对
a = (double)(2 / 3 f) = 0.666667
*//3（ 先将 2 和 3 都转换为double型，2.0 / 3.0 = 0.666667，赋给double_a,以 “ % f” 浮点型格式输出0.666667 ）

我们又验证了这样的一个情况
A = (int)21/2.9;
printf( “A = (int)(21 / 2.9 d) = %d\n”,A );

运算结果：
A = (int)(21 / 2.9 d) = 7

//如果按照 3 的逻辑，这里先将里面的数都转换为int类型，变为 （21 / 2 = 10），但是输出结果为 7（21 / 2.9 = 7.24…），证明为了保证精度，double型变量的计算比int型有更高的优先级，由于除数有小数位，先换算成（21.0 / 2.9 = 7.24…），再转换成int型 7，赋给intA；