writertoutf8buffered.java

来自「JAVA 所有包」· Java 代码 · 共 502 行 · 第 1/2 页

    final byte[] buf_loc = m_outputBytes; // local reference for faster access
    int count_loc = count;      // local integer for faster access
    int i = start;
    {
        /* This block could be omitted and the code would produce
         * the same result. But this block exists to give the JIT
         * a better chance of optimizing a tight and common loop which
         * occurs when writing out ASCII characters. 
         */ 
        char c;
        for(; i < n && (c = chars[i])< 0x80 ; i++ )
            buf_loc[count_loc++] = (byte)c;
    }
    for (; i < n; i++)
    {

      final char c = chars[i];

      if (c < 0x80)
        buf_loc[count_loc++] = (byte) (c);
      else if (c < 0x800)
      {
        buf_loc[count_loc++] = (byte) (0xc0 + (c >> 6));
        buf_loc[count_loc++] = (byte) (0x80 + (c & 0x3f));
      }
      /**
        * The following else if condition is added to support XML 1.1 Characters for 
        * UTF-8:   [1111 0uuu] [10uu zzzz] [10yy yyyy] [10xx xxxx]*
        * Unicode: [1101 10ww] [wwzz zzyy] (high surrogate)
        *          [1101 11yy] [yyxx xxxx] (low surrogate)
        *          * uuuuu = wwww + 1
        */
      else if (c >= 0xD800 && c <= 0xDBFF) 
      {
          char high, low;
          high = c;
          i++;
          low = chars[i];

          buf_loc[count_loc++] = (byte) (0xF0 | (((high + 0x40) >> 8) & 0xf0));
          buf_loc[count_loc++] = (byte) (0x80 | (((high + 0x40) >> 2) & 0x3f));
          buf_loc[count_loc++] = (byte) (0x80 | ((low >> 6) & 0x0f) + ((high << 4) & 0x30));
          buf_loc[count_loc++] = (byte) (0x80 | (low & 0x3f));
      }
      else
      {
        buf_loc[count_loc++] = (byte) (0xe0 + (c >> 12));
        buf_loc[count_loc++] = (byte) (0x80 + ((c >> 6) & 0x3f));
        buf_loc[count_loc++] = (byte) (0x80 + (c & 0x3f));
      }
    }
    // Store the local integer back into the instance variable
    count = count_loc;

  }

  /**
   * Write a string.
   *
   * @param  s  String to be written
   *
   * @exception  IOException  If an I/O error occurs
   */
  public void write(final String s) throws IOException
  {

    // We multiply the length by three since this is the maximum length
    // of the characters that we can put into the buffer.  It is possible
    // for each Unicode character to expand to three bytes.
    final int length = s.length();
    int lengthx3 = 3*length;

    if (lengthx3 >= BYTES_MAX - count)
    {
      // The requested length is greater than the unused part of the buffer
      flushBuffer();

      if (lengthx3 > BYTES_MAX)
      {
        /*
         * The requested length exceeds the size of the buffer,
         * so break it up in chunks that don't exceed the buffer size.
         */
         final int start = 0;
         int split = length/CHARS_MAX; 
         final int chunks;
         if (split > 1)
             chunks = split;
         else
             chunks = 2;
         int end_chunk = 0;
         for (int chunk = 1; chunk <= chunks; chunk++)
         {
             int start_chunk = end_chunk;
             end_chunk = start + (int) ((((long) length) * chunk) / chunks);
             s.getChars(start_chunk,end_chunk, m_inputChars,0);
             int len_chunk = (end_chunk - start_chunk);

             // Adjust the end of the chunk if it ends on a high char 
             // of a Unicode surrogate pair and low char of the pair
             // is not going to be in the same chunk
             final char c = m_inputChars[len_chunk - 1];
             if (c >= 0xD800 && c <= 0xDBFF) {
                 // Exclude char in this chunk, 
                 // to avoid spanning a Unicode character 
                 // that is in two Java chars as a high/low surrogate
                 end_chunk--;
                 len_chunk--;
                 if (chunk == chunks) {
                     /* TODO: error message needed.
                      * The String incorrectly ends in a high char
                      * of a high/low surrogate pair, but there is
                      * no corresponding low as the high is the last char
                      * Recover by ignoring this last char.
                      */
                 }
             }

             this.write(m_inputChars,0, len_chunk);
         }
         return;
      }
    }


    s.getChars(0, length , m_inputChars, 0);
    final char[] chars = m_inputChars;
    final int n = length;
    final byte[] buf_loc = m_outputBytes; // local reference for faster access
    int count_loc = count;      // local integer for faster access
    int i = 0;
    {
        /* This block could be omitted and the code would produce
         * the same result. But this block exists to give the JIT
         * a better chance of optimizing a tight and common loop which
         * occurs when writing out ASCII characters. 
         */ 
        char c;
        for(; i < n && (c = chars[i])< 0x80 ; i++ )
            buf_loc[count_loc++] = (byte)c;
    }
    for (; i < n; i++)
    {

      final char c = chars[i];

      if (c < 0x80)
        buf_loc[count_loc++] = (byte) (c);
      else if (c < 0x800)
      {
        buf_loc[count_loc++] = (byte) (0xc0 + (c >> 6));
        buf_loc[count_loc++] = (byte) (0x80 + (c & 0x3f));
      }
    /**
      * The following else if condition is added to support XML 1.1 Characters for 
      * UTF-8:   [1111 0uuu] [10uu zzzz] [10yy yyyy] [10xx xxxx]*
      * Unicode: [1101 10ww] [wwzz zzyy] (high surrogate)
      *          [1101 11yy] [yyxx xxxx] (low surrogate)
      *          * uuuuu = wwww + 1
      */
    else if (c >= 0xD800 && c <= 0xDBFF) 
    {
        char high, low;
        high = c;
        i++;
        low = chars[i];

        buf_loc[count_loc++] = (byte) (0xF0 | (((high + 0x40) >> 8) & 0xf0));
        buf_loc[count_loc++] = (byte) (0x80 | (((high + 0x40) >> 2) & 0x3f));
        buf_loc[count_loc++] = (byte) (0x80 | ((low >> 6) & 0x0f) + ((high << 4) & 0x30));
        buf_loc[count_loc++] = (byte) (0x80 | (low & 0x3f));
    }
      else
      {
        buf_loc[count_loc++] = (byte) (0xe0 + (c >> 12));
        buf_loc[count_loc++] = (byte) (0x80 + ((c >> 6) & 0x3f));
        buf_loc[count_loc++] = (byte) (0x80 + (c & 0x3f));
      }
    }
    // Store the local integer back into the instance variable
    count = count_loc;

  }

  /**
   * Flush the internal buffer
   *
   * @throws IOException
   */
  public void flushBuffer() throws IOException
  {

    if (count > 0)
    {
      m_os.write(m_outputBytes, 0, count);

      count = 0;
    }
  }

  /**
   * Flush the stream.  If the stream has saved any characters from the
   * various write() methods in a buffer, write them immediately to their
   * intended destination.  Then, if that destination is another character or
   * byte stream, flush it.  Thus one flush() invocation will flush all the
   * buffers in a chain of Writers and OutputStreams.
   *
   * @exception  IOException  If an I/O error occurs
   *
   * @throws java.io.IOException
   */
  public void flush() throws java.io.IOException
  {
    flushBuffer();
    m_os.flush();
  }

  /**
   * Close the stream, flushing it first.  Once a stream has been closed,
   * further write() or flush() invocations will cause an IOException to be
   * thrown.  Closing a previously-closed stream, however, has no effect.
   *
   * @exception  IOException  If an I/O error occurs
   *
   * @throws java.io.IOException
   */
  public void close() throws java.io.IOException
  {
    flushBuffer();
    m_os.close();
  }

  /**
   * Get the output stream where the events will be serialized to.
   *
   * @return reference to the result stream, or null of only a writer was
   * set.
   */
  public OutputStream getOutputStream()
  {
    return m_os;
  }

  public Writer getWriter()
  {
    // Only one of getWriter() or getOutputStream() can return null
    // This type of writer wraps an OutputStream, not a Writer.
    return null;
  }
}