On Mon, Jul 18, 2011 at 9:30 AM, Igor Tandetnik <[email protected]> wrote: > On 7/18/2011 11:10 AM, > [email protected] wrote: >> Sorry the mail was sent prematurely. Continuing ... >> 2. In a loop, read each address, >> for that address find the next time this address was used >> ( to implement the replacement policy) > > Sort the list by address then sequence number, probably using some form > of the merge sort algorithm (which allows one to sort a data set larger > than the available RAM). Then do a single pass over the sorted list, > looking for sequences of repeated addresses.
ya - make the databae do your address comparison don't pull back each record - just the records you need. maybe add a timestamp or other number you can increment rather than delete the old records, just use the old record updated as appropriate > -- > Igor Tandetnik > > _______________________________________________ > sqlite-users mailing list > [email protected] > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list [email protected] http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
