The basic premise of AAM is that in order for a record to contain the desired search keys, it must contain ALL of the triplets generated by those search keys; alternately any record that does not contain all of those triplets can not possibly contain those search keys.
An AAM READ begins by converting the search keys into triplets and then into hash values. Using the AAM index, it then creates a list of all text records which contain those hash values and thus can contain the search keys. With very large indexes or with slow PCs an analysis which was saved by AAMexam can save the time of having to re-analyze the index.
For more information or to order a copy of PL/B Companion, contact
5959 Westheimer #135,
Houston, TX 77057 USA,
or order through
B4-U-BUY or Sunbelt Computer Systems
PL/B Companion Overview | ISIcheck | pcDATAck | DBhelp | Return to B4-U-BUY Main Page