I have two tables orders and customerpaymentnote, which keep denormalized
columns of the status in rows related by orderid. The column duplication
is intentional, to ease end-user ad-hoc queries. I don't understand the
UPDATE FROM clause at:
http://www.postgresql.org/docs/7.4/static/sql-update.html
I have a query working (slowly) in MS Access with the pgodbc driver, but I
need to know the Postgres version of the SQL.
Working (but slow) MS Access version:
UPDATE orders
INNER JOIN customerpaymentnote
ON orders.orderid=customerpaymentnote.orderid
SET orders.customerdatetimepaid = customerpaymentnote.customerdatetimepaid,
orders.customerchargeaspaid = customerpaymentnote.customerchargeaspaid,
orders.customerchargestatus = "Payment-Recieved",
orders.orderworkflowbillingstateid = "Payment-Recieved";
Postgres translation, which does not yet work:
UPDATE orders
SET customerdatetimepaid = customerpaymentnote.customerdatetimepaid,
customerchargeaspaid = customerpaymentnote.customerchargeaspaid,
customerchargestatus = "Payment-Recieved",
orderworkflowbillingstateid = "Payment-Recieved"
FROM orders INNER JOIN customerpaymentnote
ON orders.orderid=customerpaymentnote.orderid
Query result with 0 rows will be returned.
ERROR: table name "orders" specified more than once
Can anyone suggest a proper translation. I'm open to other query
strategies, if an IN(...) statement or something else will speed up this
slow query.
Thanks.