Java：将浮点二进制转换为浮点十进制

我想转换表示 IEEE754 双精度数尾数部分的字符串。 找不到Java中是否有这样的转换方法，以避免手动添加1 + 1/2 + 1/4 + 1/8等。

|0100000011001010000111110000000000000000000000000000000000000000--> IEEE754 中的 13374|------------1010000111110000000000000000000000000000000000000000--> 尾数部分| 1.1010000111110000000000000000000000000000000000000000--> 恢复固定值1

String s = "1.1010000111110000000000000000000000000000000000000000"
double mant10 = Double.readFromFloatBinary(s); // does such method exists in Java?

是的，有多种方法可以读取二进制表示形式。但您没有 IEEE 格式的表示形式。

我会忽略句号并将其读为BigIntegerbase2，然后创建一个值来除以也使用BigInteger:

private static double binaryStringToDouble(String s) {
    return stringToDouble(s, 2);
}

private static double stringToDouble(String s, int base) {
    String withoutPeriod = s.replace(".", "");
    double value = new BigInteger(withoutPeriod, base).doubleValue();
    String binaryDivisor = "1" + s.split("\\.")[1].replace("1", "0");
    double divisor = new BigInteger(binaryDivisor, base).doubleValue();
    return value / divisor;
}

@Test
public void test_one_point_5() {
    String s = "1.1";
    double d = binaryStringToDouble(s);
    assertEquals(1.5, d, 0.0001);
}

@Test
public void test_6_8125() {
    String s = "110.1101";
    double d = binaryStringToDouble(s);
    assertEquals(6.8125, d, 0.0001);
}

@Test
public void test_yours() {
    String s = "1.1010000111110000000000000000000000000000000000000000";
    double d = binaryStringToDouble(s);
    assertEquals(1.632568359375, d, 0.000000000000000001);
}

@Test
public void test_yours_no_trailing_zeros() {
    String s = "1.101000011111";
    double d = binaryStringToDouble(s);
    assertEquals(1.632568359375, d, 0.000000000000000001);
}