I do this all the time; In fact, I've written a dynamic aggregate engine
that uses a sudo bind variable technique & dynamic joins with dependency
injection because the table names and query logic are not known at run
time - all in plpgsql.
sb
On 7/5/2013 9:26 AM, Andrew Dunstan wrote:
On 07/
2013/7/5 Greg Jaskiewicz
> Hey,
>
> We have a search method that depending on search params will join 3-5
> tables, craft the joins and where section. Only problem is, this is done in
> rather horrible java code. So using pgtap for tests is not feasible.
> I want to move the database complexity b
On 07/05/2013 09:57 AM, Greg Jaskiewicz wrote:
Hey,
We have a search method that depending on search params will join 3-5 tables,
craft the joins and where section. Only problem is, this is done in rather
horrible java code. So using pgtap for tests is not feasible.
I want to move the databas
Hey,
We have a search method that depending on search params will join 3-5 tables,
craft the joins and where section. Only problem is, this is done in rather
horrible java code. So using pgtap for tests is not feasible.
I want to move the database complexity back to database, almost writing the
