public class GridBufferedParser extends Object implements GridNioParser
GridNioServerBuffer
.
The rule for this parser is that every message sent over the stream is prepended with 4-byte integer header containing message size. So, the stream structure is as follows:
+--+--+--+--+--+--+...+--+--+--+--+--+--+--+...+--+ | MSG_SIZE | MESSAGE | MSG_SIZE | MESSAGE | +--+--+--+--+--+--+...+--+--+--+--+--+--+--+...+--+
Constructor and Description |
---|
GridBufferedParser(boolean directBuf,
ByteOrder order) |
Modifier and Type | Method and Description |
---|---|
byte[] |
decode(GridNioSession ses,
ByteBuffer buf)
This method is called when input bytes are available on the underlying network connection.
|
ByteBuffer |
encode(GridNioSession ses,
Object msg)
This method is called whenever a message should be sent to the network connection
and network buffer is ready to be filled with bytes.
|
String |
toString() |
public GridBufferedParser(boolean directBuf, ByteOrder order)
directBuf
- Direct buffer.order
- Byte order.public byte[] decode(GridNioSession ses, ByteBuffer buf) throws IOException, IgniteCheckedException
Method must read given buffer until either it reaches the end of buffer or a valid user message is encountered. In this case it must return parsed message.
decode
in interface GridNioParser
ses
- Session on which bytes are read.buf
- Buffer that contains input data.null
if complete message has not been received yet. Note
that in case of returning null
given buffer must be completely read.IOException
- If exception occurred while reading data.IgniteCheckedException
- If any user-specific error occurred.public ByteBuffer encode(GridNioSession ses, Object msg) throws IOException, IgniteCheckedException
Method must encode the complete message and return byte buffer from which data should be read.
encode
in interface GridNioParser
ses
- Session on which message is being sent.msg
- Message to encode.IOException
- If exception occurred while encoding data.IgniteCheckedException
- If any user-specific error occurred while encoding data.
Follow @ApacheIgnite
Ignite Fabric : ver. 2.0.0 Release Date : April 30 2017