Re: Generating fake databases

On 13 okt, 17:48, Roy Hann <specia..._at_processed.almost.meat> wrote:
> Derek Asirvadem wrote:
> > On Oct 11, 10:54 pm, Roy Hann <specia..._at_processed.almost.meat> wrote:
>
> [Much good and clever advice for solving a different problem snipped.]
>
> >> and the "data" must satisfy my database constraints
>
> > Well assuming you means database, when you uses the word "database",
> > the data in the database will *always* satisfy the database
> > constraints, regardless of whether it is obfuscated or transformed or
> > hammered. (maybe you mean something else by that statement ?).  You
> > *have* implemented the C in ACID, haven't you ?  If you have not, then
> > it is not a database, it is a bucket of fish.  In that case, you will
> > have to expend some effort transforming it into a database, before you
> > can expect database capabilities from it.
>
> I fear you may be somewhat dazzled by the clarity of your own vision of
> the solution.
>
> Recall I want to generate entirely synthetic data that satisfies my
> constraints, and do it efficiently.  SQL table definitions and
> constraints are all I expect to have available.  I do not expect to have
> a real database (i.e. logically consistent set of real data).
>
> I didn't spell it out, but I want to be able to use this technique to
> create data for any arbitrary database design.  
>
> --
> Roy

Roy,

How far does the scope go of what you mean by "constraints" here ?

Really just any arbitrary constraint (such as "the sum of the percentages in this table must add up to 100", and which are at best enforced by trigger code buried inside the database definition, and for which you are unlikely to have any machine-readable declarative definition), or only the classes of constraint for which SQL has declarative support ? Received on Thu Oct 13 2011 - 19:57:36 CEST