Constructor Detail

• LZWStringTable

  public LZWStringTable()

  Constructor allocate memory for string store data

Method Detail

• AddCharString

  public int AddCharString(short index,
                           byte b)

  Parameters:

  index - value of -1 indicates no predecessor [used in initialisation]

  b - the byte [character] to add to the string store which follows the predecessor string specified the index.

  Returns:

  0xFFFF if no space in table left for addition of predecesor index and byte b. Else return the code allocated for combination index + b.

• FindCharString

  public short FindCharString(short index,
                              byte b)

  Parameters:

  index - index to prefix string

  b - the character that follws the index prefix

  Returns:

  b if param index is HASH_FREE. Else return the code for this prefix and byte successor

• ClearTable

  public void ClearTable(int codesize)

  Parameters:

  codesize - the size of code to be preallocated for the string store.

• Hash

  public static int Hash(short index,
                         byte lastbyte)

• expandCode

  public int expandCode(byte[] buf,
                        int offset,
                        short code,
                        int skipHead)

  If expanded data doesnt fit into array only what will fit is written to buf and the return value indicates how much of the expanded code has been written to the buf. The next call to expandCode() should be with the same code and have the skip parameter set the negated value of the previous return. Succesive negative return values should be negated and added together for next skip parameter value with same code.

  Parameters:

  buf - buffer to place expanded data into

  offset - offset to place expanded data

  code - the code to expand to the byte array it represents. PRECONDITION This code must allready be in the LZSS

  skipHead - is the number of bytes at the start of the expanded code to be skipped before data is written to buf. It is possible that skipHead is equal to codeLen.

  Returns:

  the length of data expanded into buf. If the expanded code is longer than space left in buf then the value returned is a negative number which when negated is equal to the number of bytes that were used of the code being expanded. This negative value also indicates the buffer is full.

• dump

  public void dump(PrintStream out)