Save This Page
Home » openjdk-7 » sun » misc » [javadoc | source]
public class: UCEncoder [javadoc | source]
This class implements a robust character encoder. The encoder is designed to convert binary data into printable characters. The characters are assumed to exist but they are not assumed to be ASCII, the complete set is 0-9, A-Z, a-z, "(", and ")". The basic encoding unit is a 3 character atom. It encodes two bytes of data. Bytes are encoded into a 64 character set, the characters were chosen specifically because they appear in all codesets. We don't care what their numerical equivalent is because we use a character array to map them. This is like UUencoding with the dependency on ASCII removed. The three chars that make up an atom are encoded as follows:
     00xxxyyy 00axxxxx 00byyyyy
     00 = leading zeros, all values are 0 - 63
     xxxyyy - Top 3 bits of X, Top 3 bits of Y
     axxxxx - a = X parity bit, xxxxx lower 5 bits of X
     byyyyy - b = Y parity bit, yyyyy lower 5 bits of Y
The atoms are arranged into lines suitable for inclusion into an email message or text file. The number of bytes that are encoded per line is 48 which keeps the total line length under 80 chars) Each line has the form(
 Where each (xxx) represents a three character atom.
 (LLSS) - 8 bit length (high byte), and sequence number
          modulo 256;
 (DDDD) - Data byte atoms, if length is odd, last data
          atom has (DD00) (high byte data, low byte 0)
 (CRC)  - 16 bit CRC for the line, includes length,
          sequence, and all data bytes. If there is a
          zero pad byte (odd length) it is _NOT_
          included in the CRC.
Fields inherited from sun.misc.CharacterEncoder:
Method from sun.misc.UCEncoder Summary:
bytesPerAtom,   bytesPerLine,   encodeAtom,   encodeBufferPrefix,   encodeLinePrefix,   encodeLineSuffix
Methods from sun.misc.CharacterEncoder:
bytesPerAtom,   bytesPerLine,   encode,   encode,   encode,   encode,   encode,   encodeAtom,   encodeBuffer,   encodeBuffer,   encodeBuffer,   encodeBuffer,   encodeBuffer,   encodeBufferPrefix,   encodeBufferSuffix,   encodeLinePrefix,   encodeLineSuffix,   readFully
Methods from java.lang.Object:
clone,   equals,   finalize,   getClass,   hashCode,   notify,   notifyAll,   toString,   wait,   wait,   wait
Method from sun.misc.UCEncoder Detail:
 protected int bytesPerAtom() 
    this clase encodes two bytes per atom
 protected int bytesPerLine() 
    this class encodes 48 bytes per line
 protected  void encodeAtom(OutputStream outStream,
    byte[] data,
    int offset,
    int len) throws IOException 
    encodeAtom - take two bytes and encode them into the correct three characters. If only one byte is to be encoded, the other must be zero. The padding byte is not included in the CRC computation.
 protected  void encodeBufferPrefix(OutputStream a) throws IOException 
    The buffer prefix code is used to initialize the sequence number to zero.
 protected  void encodeLinePrefix(OutputStream outStream,
    int length) throws IOException 
    Each UCE encoded line starts with a prefix of '*[XXX]', where the sequence number and the length are encoded in the first atom.
 protected  void encodeLineSuffix(OutputStream outStream) throws IOException 
    each UCE encoded line ends with YYY and encoded version of the 16 bit checksum. The most significant byte of the check sum is always encoded FIRST.