>Wouldn't it make more sense for "i in 1..5" to expand to "i >= 1 and i ><= 5"? > >Then it would also work for ordered types that aren't ordinal, such as >rationals >and strings and blobs and dates etc, and it would work for very large >ranges, >since there's no conceptual need to generate all the individual values. > >Of course, you'd want to support all 4 variants: 1..5, 1..^5, 1^..5, >1^..^5, >where a ^ means exclude that endpoint and its absence means include. > >This is more flexible than SQL's BETWEEN, which I believe only covers >one of >those 4 options.
That's getting into generic interval support. I find this interesting even if its really much more ambitious than my simple-minded (and highly optional) initial need/question. I might dig further in this direction someday. Good point. Thanks Darren. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users