Yes thanks for the advice...I've already combined the "raw" data i.e. vectors straight out of the annual reports and the calculated vectors e.g.. "noplat" derived from it which I've given a type "calcd" in the type field of the same table "itms". There is some market wide data however, which might not fit into this vector table Even though the individual items are stored as individual records.I'm just at the stage of integrating SQLITE with forth and everything's up for grabs at the moment.
On 23 June 2013 12:42, Simon Slavin <slav...@bigfraud.org> wrote: > > On 23 Jun 2013, at 12:26pm, e-mail mgbg25171 <mgbg25...@blueyonder.co.uk> > wrote: > > > The "table as a number" idea is necessary to fit my model > > where word definitions are stored as streams of code pointers (NUMBERS) > > that just get called. > > i.e. EVERYTHING has to be a number > > That makes perfect sense. > > If these tables all have the same fields in you might want to refactor > your SQLite schema so that these tables are actually all one big table. > The thing you currently consider to be a table name would just be a field > in the big table. > > This would give you a single rowid which would indicate both the table and > the (currently) row in that table. > > Simon. > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
