Examples of com.google.opengse.iobuffer.IOBuffer

com.google.opengse.iobuffer.IOBuffer
The IOBuffer class stores an arbitrary amount of data in a {@link java.util.LinkedList} of {@link java.nio.ByteBuffer} objects.Buffers are added to the list during calls to the IOBuffer write methods (i.e. write(char[]), writeBytes(byte[]), etc.). Buffers are consumed from the start of the list (FIFO) during calls to the IOBuffer read methods (i.e. read(char[]), readBytes(byte[]), etc.). There are three stages in the life of data in an IOBuffer. There's the current write buffer, the buffer list, and the current read buffer. All data contained in an IOBuffer arrives through calls to the write methods (there are also transfer, prepend & append methods). The write methods store data to the current write buffer, which is available through a call to {@link #getWriteBuffer()}. Obviously, the current write buffer has a limited capacity. Once full, however, the {@link #getWriteBuffer()} method will automatically appendit to the internal buffer list and allocate a new buffer for writing. The internal buffer list can grow arbitrarily large and cannot be accessed externally. Instead, the current read buffer is used to retrieve data from the IOBuffer. Initially null, the current read buffer is set to the first buffer in the internal buffer list on calls to the read methods. When all of the data in the current read buffer is consumed, it is discarded and the next buffer is popped from the internal list. Any current write buffer will not be automatically available for reading until is is manually flushed with a call to flush(). This is necessary to support concurrent reads and writes. There are also current read and write buffers for character-based read/write operations. Note that you cannot mix character reads with byte reads and character writes with byte writes. For example, it is illegal to call write(' ') and then writeByte(32) without an explicit {@code flush()}call between the invocations. This class is safe for a concurrent reader and writer. That is, one thread can read while another writes. Multiple readers and multiple writers are not supported. IMPORTANT: It is necessary to call flush() to ensure that data written to the IOBuffer is available for reading. A maximum IOBuffer size limit can be set by invoking the {@link #setSizeLimit} method. When the limit isexceeded, the contents of the IOBuffer are consumed via a call to the {@link ConsumeCallback#handleConsume(IOBuffer,boolean)}callback. This provides a mechanism to dispense with the contents of the IOBuffer in an application-dependent way. By default, the size limit is set to -1, which allows the IOBuffer to grow without bound. @see com.google.opengse.iobuffer.IOBufferInputStream @see com.google.opengse.iobuffer.IOBufferOutputStream @see com.google.opengse.iobuffer.IOBufferReader @see com.google.opengse.iobuffer.IOBufferWriter @author Spencer Kimball @author Peter Mattis

        // CRLF.
        if (buf.availableBytes() >= chunkSize + 2) {
          // As long as this isn't the last chunk (0 length), transfer bytes.
          if (chunkSize > 0) {
            if (post_body_ == null) {
              post_body_ = new IOBuffer();
            }
            post_body_.transfer(buf, chunkSize);
          }
          // Pull off trailing CRLF.
          if (!buf.readLine(line_buf, READLINE_LIMIT)) {

View Full Code Here

    post_body_.flush();
    InputStream inflator_input_stream =
        new GZIPInputStream(new IOBufferInputStream(post_body_), 1024);


    // decompress by writing to a new IO buffer, update body
    IOBuffer inflatedBodyBuffer = new IOBuffer();
    final byte[] byteBuffer = new byte[1024];
    int numBytes;
    while (-1 != (numBytes = inflator_input_stream.read(byteBuffer))) {
      inflatedBodyBuffer.writeBytes(byteBuffer, 0, numBytes);
    }
    inflatedBodyBuffer.flush();


    int inflated_body_length = inflatedBodyBuffer.totalBytes();
    _setPostBody(inflatedBodyBuffer, inflated_body_length);


    // Remove content encoding, set new length, so request forwarding will work
    requestContext.getHeaders().removeHeader("Content-Encoding");
    requestContext.getHeaders().setIntHeader("Content-Length",

View Full Code Here

  protected ParamMap parseParameters(String encoding) {
    ParamMap map = new ParamMap();
    FormUrlDecoder.parse(getQueryString(), map, encoding);


    if (post_body_ != null) {
      IOBuffer params = getPostBody();
      if ("application/x-www-form-urlencoded".equals(getContentType())) {
        // the character encoding on the post body is guaranteed to be US-ASCII
        params.setCharacterEncoding("US-ASCII");


        StringBuilder sb = new StringBuilder();
        char[] char_buf = new char[1024];
        int res;


        while ((res = params.read(char_buf)) > 0) {
          sb.append(char_buf, 0, res);
        }


        FormUrlDecoder.parse(sb.toString(), map, encoding);
      }

View Full Code Here

  public BufferedReader getReader() {
    if (input_stream_ != null) {
      throw new IllegalStateException("getInputStream() called previously");
    }
    if (input_reader_ == null) {
      IOBuffer body = getPostBody();
      body.setCharacterEncoding(getInternalCharacterEncoding());
      input_reader_ = new BufferedReader(new IOBufferReader(body));
    }
    return input_reader_;
  }

View Full Code Here

0 1

TOP

Related Classes of com.google.opengse.iobuffer.IOBuffer

com.google.opengse.core.HttpRequestImpl

com.google.opengse.core.HttpResponseImpl

java.io.IOException

java.nio.charset.CharsetEncoder

java.nio.charset.Charset

java.nio.charset.CoderResult

java.nio.CharBuffer

java.nio.ByteBuffer

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.