Labels:book | box | bulletin board | cabinet | plant | reckoner | shelf | sky OCR: Memory Start of string 10301 Once the string length be searched known, every byte up 10302 10303 and including the first match for the byte 10304 re searching for (or 10305 every byte in the string. 10306 if no match exists must 10307 be reao again First match tor 10308 search byte 10309 1030 1030B 1030 1030D 1030F FIGURE 11.10 continued) possibly run fast enough t0 allow the double search approach to outperform an approach that accesses memory only one third as many times? The answer is ves conditionally The double -scarch approach actually is slightly faster than lodsb based single search string searching approach for the average case The double search approach perfe orms relatively more poorly matches tend to occur most frequently in the first half of the strings being searched pue rela ...
(.png)
