将字节转换为字符串的方法

Question

I am using the below code to read messages from IBM MQ using Java.

public void asyncReadMessage() 
    {
        System.out.println("inside asyncReadmessage");
        
            
        try{
            
            com.ibm.mq.MQQueue defaultLocalQueue;
            MQQueueManager qManager=null;
            String strMessage=null;
            MQEnvironment.hostname = "reese.int.westgroup.com";
            MQEnvironment.channel = "CLIENTCONNECTION";
            MQEnvironment.port = 1414;
            String qMngrStr = "";
            qManager = new MQQueueManager(qMngrStr); 
            int openOptions = MQC.MQOO_FAIL_IF_QUIESCING |
                    MQC.MQOO_INPUT_SHARED | MQC.MQOO_BROWSE;
                    //| MQC.MQOO_INQUIRE ; 
            String queueName="COMSERV.SRCHEXT.EVENTS.PUBLISH.QA.Q01";
            System.out.println("accessing::"+queueName);
            defaultLocalQueue = qManager.accessQueue(queueName, openOptions);
            //set transport properties.
            System.out.println("set MQ props");
            MQEnvironment.properties.put(MQC.TRANSPORT_PROPERTY, MQC.TRANSPORT_MQSERIES_CLIENT);
            System.out.println("new Queuemanager");
            ArrayList<String> msglist=new ArrayList<String>();
            boolean getMore = true;
            String messageText = null;
            while(getMore)
            {  
            MQMessage getMessages = new MQMessage();
            MQGetMessageOptions gmo = new MQGetMessageOptions();
            gmo.options=MQC.MQGMO_WAIT | MQC.MQGMO_BROWSE_FIRST;
            gmo.matchOptions=MQC.MQMO_NONE;
            gmo.waitInterval=100000;
            defaultLocalQueue.get(getMessages, gmo);
            int length=getMessages.getMessageLength();
            byte[] bMsg = new byte[length];
            getMessages.readFully(bMsg);
            String retrievedmsg = new String(bMsg);
            System.out.println("retrievedmsg::"+retrievedmsg);
           
            if(!(retrievedmsg.equals("")))
            {
                MessageHandler mh = new MessageHandler();
                System.out.println("messagehandler java file:"+retrievedmsg);
                mh.processMessages(retrievedmsg);
            }
            msglist.add(retrievedmsg);
            System.out.println("msglist::"+msglist.toString());
            }
            if(defaultLocalQueue.getCurrentDepth()==0)
             {
                getMore=false;
             }
            /*for (String message : msglist)
            {
                System.out.println("message in for loop::"+message);
                MessageHandler mh = new MessageHandler();
                System.out.println("messagehandler java file:"+message);
                mh.processMessages(message);
            }*/
            }
            
            catch(Exception e)
            {
                e.printStackTrace();
            }
    }

When I print the "retrievedmsg" variable, it's getting printed as "< maestro . proces s. event". Instead it should populate as "<maestro.process.event". And also if I try to copy "< maestro . process . event" from the Catalina log file and paste it in a notepad, then only the first letter of it will be copied and other letters will not be copied. Looks like it is reading as a byte. Correct me if I am wrong. Please help how to achieve this.

Answer 1

尝试String retrievedmsg = new String(bMsg, StandardCharsets.UTF_8)

Answer 2

There are many ways bytes can be interpreted as text. Each scheme for converting between bytes and text is called a Charset .

When you write new String(bMsg) you are converting the bytes of bMsg using the system's default Charset. It would appear that the bytes you received were not in that Charset.

Whenever you try to convert between bytes and text, you need to know the Charset. Using new String(byte[]) does not make the problem go away, it just means you're ignoring the issue at your peril, as you can see in this case.

My guess is that your system's default charset is a one-byte Charset, like ISO-8859-1 or windows-1252, which expect one byte to represent one character, but the text you're receiving is little-endian UTF-16, which uses two bytes to represent each character. If that is the case, you can address it by specifying that Charset:

String retrievedmsg = new String(bMsg, StandardCharsets.UTF_16LE);

However, this alone is not a solution. You need to find out from the provider of the message whether they will always will send messages in UTF-16LE. You simply cannot know how to convert bytes to text without knowing the Charset used to transmit the data.

Answer 3

MQEnvironment.hostname = "reese.int.westgroup.com"; String queueName="COMSERV.SRCHEXT.EVENTS.PUBLISH.QA.Q01";

You should always replace private internal company values with dummy values.

Secondly, you should be using a Hashtable and not MQEnvironment Class. Just read any of the Java/MQ samples that I have post here on StackOverflow. ie MQTest11.java

"< maestro . proces s. event"

That looks like Windows double byte aka character set 1252. This is not an MQ issue but rather a Java conversion issue. What did you try? Did you try specifying the character set on the "new String" as per VGR or rsampaths16 suggested.

ie

String retrievedMsg = new String(bMsg, "cp1252");

Finally, why don't you output the message payload in a HEX dump? That way you would know exactly what you are dealing with.

If you don't know how to do that then here is some sample Java code that will do it for you. Put it in your toolkit, so that you can easily add it to any project

/**
 * Dump a byte array as a standard HEX dump output
 * @param data the entire byte array
 * @param widthSize width of the HEX dump. Must be in multiples of 16
 */
public void dumpBuffer(byte[] data, int widthSize)
{
   int endPt = widthSize;
   int len = data.length;
   byte[] tempB = new byte[widthSize];

   if (widthSize == 16)
   {
      System.out.println("Address  0 1 2 3  4 5 6 7  8 9 A B  C D E F  0123456789ABCDEF");
      System.out.println("------- -------- -------- -------- --------  ----------------");
   }

   for (int i=0; i < len; i+=widthSize)
   {
      if (i+widthSize >= len)
         endPt = len - i;

      for (int j=0; j < endPt; j++)
         tempB[j] = data[i+j];

      System.out.println(formatHex(tempB, (i+widthSize < len?widthSize:len-i), i, widthSize ));
   }
}

/**
 * Format an array of bytes into a hex display
 * @param src a portion of the byte array
 * @param len length of this part of the byte array
 * @param index location of current position in data
 * @param width width of the HEX dump
 * @return
 */
private String formatHex(byte[] src, int lenSrc, int index, int width)
{
   String HEX = "0123456789ABCDEF";
   int i,j;
   int g = width / 4; /* number of groups of 4 bytes */
   int d = g * 9;     /* hex display width */
   StringBuffer sb = new StringBuffer();

   if ( (src == null) ||
        (lenSrc < 1) || (lenSrc > width) ||
        (index < 0) ||
        (g % 4 != 0) ||   /* only allow width of 16 / 32 / 48 etc. */
        (d < 36) )
   {
      return "";
   }

   String hdr = Integer.toHexString(index);
   if (hdr.length() <= 6)
      sb.append("000000".substring(0, 6 - hdr.length()) + hdr + ": ");
   else
      sb.append(hdr + ": ");

   /* hex display 4 by 4 */
   for(i=0; i < lenSrc; i++)
   {
      sb.append(""+HEX.charAt((src[i]) >> 4) + HEX.charAt((src[i] & 0x0F)));

      if (((i+1) % 4) == 0)
         sb.append(" ");
   }

   /* blank fill hex area if we do not have "width" bytes */
   if (lenSrc < width)
   {
      j = d - (lenSrc*2) - (lenSrc / 4);
      for(i=0; i < j; i++)
         sb.append(" ");
   }

   /* character display */
   sb.append(" ");
   for (i=0; i < lenSrc; i++)
   {
      if(Character.isISOControl((char)src[i]))
         sb.append(".");
      else
         sb.append((char)src[i]);
   }

   /* blank fill character area if we do not have "width" bytes */
   if (lenSrc < width)
   {
      j = width - lenSrc;
      for(i=0; i < j; i++)
         sb.append(" ");
   }

   return sb.toString();
}

To use this code, simply do the following:

dumpBuffer(bMsg, 16);

where bMsg is the message the code retrieved from the queue. And the hex dump will look something like:

Address  0 1 2 3  4 5 6 7  8 9 A B  C D E F  0123456789ABCDEF
------- -------- -------- -------- --------  ----------------
000000: 54686973 20697320 61207465 7374206D  This is a test m
000010: 65737361 67652061 6E642061 6E6F7468  essage and anoth
000020: 65722065 7863656C 6C656E74 20657861  er excellent exa
000030: 6D706C65 20627920 526F6765 722C2069  mple by Roger, i
000040: 66204920 646F2073 61792073 6F206D79  f I do say so my
000050: 73656C66 21212020 3A29               self!!  :)