SocketChannel how to deal with cutted UTF-8 chars and lines in ByteBuffer

Question

My system is receiving UTF-8 lines ended by 0x0D 0x0A through a SocketChannel. I read data in a ByteBuffer like this:

final ByteBuffer buffer = ByteBuffer.allocate(bufsize);
final int bytesRead = _mySocket.read(buffer);

The buffer is big enough to contain a lot of lines. But I can't have a buffer big enough to contain all lines, lines are processed on the fly.

The buffer can't contain all the lines, so lines may be splitted, a character encoded in more that one byte may also be splitted.

I thought about a solution:

1. process all lines that I can extract by analyzing the ByteBuffer byte by byte (I search 0x0D 0x0A)
2. identify unprocessed bytes
3. store them in a temporary variable
4. clear the ByteBuffer
5. add unprocessed bytes to the ByteBuffer
6. read again the SocketChannel.
7. Repeat this until I have found the line indicating I received all the lines.

Does java or any library already provide this functionnality ?

Thanks for any answer, Mickaël

Answer 1

我建议使用java.util.Scanner，它知道如何处理SocketChannel，并且知道如何读取行

    Scanner sc = new Scanner(_mySocket);