| 摘要 |
<p>A class of lossless data compression algorithms use a RAM-based dictionary to facilitate the compression and decompression of data. The dictionary is initialized or reset at various events, such as data file boundaries, or at specific compression ratio thresholds. The average time to perform the dictionary reset is decreased by using a dictionary reset optimizer (12) to reset the dictionary without writing the DICT_VALID fields (19) of all dictionary locations (1,2,3,..L) immediately upon each reset. The DICT_VALID field in static RAM (16) is set up as an N-bit field and the optimizer includes a counter (20) for generating an N-bit DICT_NUMBER field (22) under control of a controller (24) responsive to reset signals. During operation to compress/decompress data after a reset, the DICT_VALID field of a dictionary location is compared to the current DICT_NUMBER(1). If their values are equal, the DICT_ENTRY field (18) in such location is deemed valid. Otherwise, the DICT_ENTRY field is rewritten with new compression data and the DICT_VALID field is reset equal to the current DICT_NUMBER(1). Upon each reset, the counter is incremented to provide a new DICT_NUMBER(2) whose value will no longer match current DICT_VALID values stored in the dictionary locations. <IMAGE></p> |
