| 摘要 |
<p>The present invention relates to compression of values and bitmaps, and methods thereof. Such methods are configured for operating on a computer system having a word length architecture of length WL and are based on the observation that not all the bits used for the run-length counter—i.e., the fill length field (FL) inhere—are often used, since runs are seldom so long. Contrarily to other compression schemes (e.g., WAH), said methods may assign the unused bits to one or more position list fields (PL, PL1, PL2, PLs), thus boosting the compression ratio. Moreover, the total length (in terms of number of bits) of the uncompressed data—comprising values or bitmaps—may be stored just once, preferably at the beginning of the compression, thus dramatically diminishing the storage requirements for the compression scheme, since it is not required to keep track of the length of each bitmap word while performing the compression or the decompression.</p> |
