Take a look at MINUS
"slim" <sfar_Selim_at_excite.com> wrote in message
news:16e3a9da.0204230724.603dc538_at_posting.google.com...
> Hi,
>
> suppose that we have 4 tables:
>
> --> t1 : t1_id (pk), t1_info
> --> t2 : t2_id (pk), t2_info, t2_t1_id (foreign key)
> --> t3 : t3_id (pk), t3_info, t3_t1_id (foreign key)
>
>
> Suppose that we have 1 millions of records in t1,
> 5 millions of records in t2,
> 4 millions of records in t3
> the goal is to compress these data in this case we process like this
>
> If n ( n > 1) records in t1 have the same t1_info, and the same set of
> info in both tables
> t2 and t3 then we will delete cascade (n-1) rows from t1
>
> e.g. :
> t1 : t1_id t1_info
> 1 'car'
> 2 'car'
> 3 'car'
> 4 'van'
> t2 : t2_id t2_info t2_t1_id
> 1 'tyre' 1
> 2 'brake' 1
> 3 'tyre' 2
> 4 'brake' 2
> 5 'tyre' 3
> 6 'brake' 3
> 7 'capacity' 4
>
> t3 : t3_id t3_info t2_t1_id
> 1 'price' 1
> 2 'power' 1
> 3 'price' 2
> 4 'power' 2
> 5 'price 3
> 6 'power' 3
> 7 'power' 4
>
> In this case we delete on cascade only t1_id = 2 and t1_id = 3 from
> t1
>
> I am writing the function and I have a big problem of performance,
> You are helpful for each idea you give to me!
> thanks ..
Received on Wed Apr 24 2002 - 14:39:12 CDT
