Have a look here: http://search.cpan.org/dist/SQL-Routine/
That's part of a project I'm working on. With it, each database table, as well as every other kind of thing you could possibly put in a schema or use with a database, is represented as a cross-referenced hierarchy of atomic-value nodes. You can create and manipulate your "create table statement" by adding, removing, and altering the nodes. Then the create can be generated from them when done. -- Darren Duncan jim wrote: Darren, this is cool. I can't wait to pass this along to a person at my job who works with DB's at his full-time job. I had thought wouldn't it be nice if people would agree on the table structue of the create table statement within the sqlite community for all these sqlite db tools popping up. In my own self serving way in part at least because I am not smart thought of the standardization because you just know I will mess up the table relations design. And like most projects who the hell wants the headache of maintaining all the changes. Why cant I borrow someone elses who did it before me who is smarter ;-) I know it sounds incredibly lazy but because of my brain it takes me longer to do things so I often think of things to cut corners to help save time. I never dreamed a whole language spanning multiple db systems would end up being developed. And if you look at some tools out there I believe the fancy structure change was omitted because of the headache to support my theory. Probably a feature creep decision. this looks like its required install OS is linux though??? thanks, jim
