Oracle SQL，单插入语句，多列

Question

Few weeks ago I saw an article about oracle sql optimization. There was some way to insert fe 200 rows in one INSERT statement, but it was not INSERT ALL clause like:

INSERT ALL
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date, sales_sun)
INTO sales (prod_id, cust_id, time_id, amount)
VALUES (product_id, customer_id, weekly_start_date+1, sales_mon)
SELECT product_id, customer_id, weekly_start_date, sales_sun,
sales_mon, sales_tue, sales_wed, sales_thu, sales_fri, sales_sat
FROM sales_input_table;

I am pretty sure (99%) it has been done with USING Clause in PL/SQL(?). Main problem was to "save" one insert statement and executing it 200 times instead of "saving" 200 inserts and executing every of each of them.

I tried to search in Google but always INSERT ALL came up. Anyone have an idea or have similar article about "this"? Thanks for help.

Answer 1

I would generate a bunch of numbers and do this in a single insert . This version uses union all to handle each of the weekday columns:

with n as (
     select level as n
     from dual
     connect by level <= 52
    )
insert into sales (prod_id, cust_id, time_id, amount)
    select product_id, customer_id, weekly_start_date + n.n - 1, sales_sun
    from sales_input_table cross join n
    union all
    select product_id, customer_id, weekly_start_date + n.n - 1, sales_mon
    from sales_input_table cross join n
    . . .

Answer 2

What you have seen is (sorry... MAY HAVE BEEN - see my comment to your original question though) the insert into (select ...) syntax. For example, suppose I have a table t, and I want to clone it. I create a table s with the same number, name and data type (and order) of columns at t, and now I want to copy all the data from t to s. The simple way to do that is

insert into s (select * from t);

What's in parentheses (the subquery) can be a full-blown subquery; for example, I can use a WHERE clause to limit which rows are copied to s; or I can use aggregate functions and GROUP BY in the subquery to insert summaries in s, instead of cloning the data; the only condition is that the number and data type of resulting columns in the subquery must match those of s.