Bert Regelink提出的准则根本不起作用。 试试下面的方法:

import javax.xml.bind.DatatypeConverter;
import java.io.*;

public class Test
{  
    @Test
    public void testObjectStreams( ) throws IOException, ClassNotFoundException
    {     
            ByteArrayOutputStream baos = new ByteArrayOutputStream();
            ObjectOutputStream oos = new ObjectOutputStream(baos);

            String stringTest = "TEST";
            oos.writeObject( stringTest );

            oos.close();
            baos.close();

            byte[] bytes = baos.toByteArray();
            String hexString = DatatypeConverter.printHexBinary( bytes);
            byte[] reconvertedBytes = DatatypeConverter.parseHexBinary(hexString);

            assertArrayEquals( bytes, reconvertedBytes );

            ByteArrayInputStream bais = new ByteArrayInputStream(reconvertedBytes);
            ObjectInputStream ois = new ObjectInputStream(bais);

            String readString = (String) ois.readObject();

            assertEquals( stringTest, readString);
        }
    }

common -codec中的Hex类应该为您做到这一点。

http://commons.apache.org/codec/

import org.apache.commons.codec.binary.Hex;
...
byte[] decoded = Hex.decodeHex("00A0BF");
// 0x00 0xA0 0xBF

我发现内核恐慌有解决方案对我最有用，但遇到问题，如果十六进制字符串是一个奇数。是这样解决的:

boolean isOdd(int value)
{
    return (value & 0x01) !=0;
}

private int hexToByte(byte[] out, int value)
{
    String hexVal = "0123456789ABCDEF"; 
    String hexValL = "0123456789abcdef";
    String st = Integer.toHexString(value);
    int len = st.length();
    if (isOdd(len))
        {
        len+=1; // need length to be an even number.
        st = ("0" + st);  // make it an even number of chars
        }
    out[0]=(byte)(len/2);
    for (int i =0;i<len;i+=2)
    {
        int hh = hexVal.indexOf(st.charAt(i));
            if (hh == -1)  hh = hexValL.indexOf(st.charAt(i));
        int lh = hexVal.indexOf(st.charAt(i+1));
            if (lh == -1)  lh = hexValL.indexOf(st.charAt(i+1));
        out[(i/2)+1] = (byte)((hh << 4)|lh);
    }
    return (len/2)+1;
}

我添加了一个十六进制数的数组，所以我通过引用我正在使用的数组，和int我需要转换和返回下一个十六进制数的相对位置。所以最后的字节数组有[0]个十六进制对，[1…十六进制对，然后是对的数目…

对我来说，这是解决方案，HEX="FF01"然后分裂到FF(255)和01(01)

private static byte[] BytesEncode(String encoded) {
    //System.out.println(encoded.length());
    byte result[] = new byte[encoded.length() / 2];
    char enc[] = encoded.toUpperCase().toCharArray();
    String curr = "";
    for (int i = 0; i < encoded.length(); i=i+2) {
        curr = encoded.substring(i,i+2);
        System.out.println(curr);
        if(i==0){
            result[i]=((byte) Integer.parseInt(curr, 16));
        }else{
            result[i/2]=((byte) Integer.parseInt(curr, 16));
        }

    }
    return result;
}

现在可以在guava中使用BaseEncoding来实现这一点。

BaseEncoding.base16().decode(string);

反过来使用

BaseEncoding.base16().encode(bytes);

我喜欢这个角色。数位解，我是这样解的

public byte[] hex2ByteArray( String hexString ) {
    String hexVal = "0123456789ABCDEF";
    byte[] out = new byte[hexString.length() / 2];

    int n = hexString.length();

    for( int i = 0; i < n; i += 2 ) {
        //make a bit representation in an int of the hex value 
        int hn = hexVal.indexOf( hexString.charAt( i ) );
        int ln = hexVal.indexOf( hexString.charAt( i + 1 ) );

        //now just shift the high order nibble and add them together
        out[i/2] = (byte)( ( hn << 4 ) | ln );
    }

    return out;
}