Hi,
Does anybody know if the autonumberhash256 function returns unique values for the same input fields? Its the definition of hashing, I know, but I seem to get sequential values!
I use it somewhat like this:
I have two tables (one for BUDGET and one for SALES) I store in two qvd, and before I store them I create this concatenated PK in both tables (this is done in different qvw):
autonumberhash256(MONTH_ID, STORE_ID, SUPPLIER_ID, BRAND_TYPE, CATEGORY_ID)as PK_BUDG
When I load them in another qvw I need to find out what exists in one table that doesn't exist in the other:
Budget_Temp:
LOAD PK_BUDG,
MONTH_ID,
STORE_ID,
SUPPLIER_ID,
BRAND_TYPE,
CATEGORY_ID
FROM $(srcFile_Budgets) (qvd)
WHERE NOT EXISTS (PK_BUDG); //everything that has budget but not sales
This fetches 0 lines (vs not using autonumberhash256() and just concatenating the fields for the PK, which fetched about 5M rows). What happens is: the SALES table has two years of sales and the PK begins in 2008-01 whilst the BUDGET table has only 1 year and the PK begins in 2009-01. I've peeked the first lines of each table stored to qvd and I reallized the PK seems sequential (beginning in 1, 2, 3,...(...) for 2008-01 in sales qvd and beginning in 1, 2, 3,...(...) for 2009-01 in budget qvd).
The sales table has about 130 million rows and the budget table has about 12 million rows so thats why it makes a huge difference when I use the autonumberhash!
Anyone has any ideas why this happens? Shouldn't I get an unique value instead of a sequential one? I've searched the forum and got nothing about this.
Thanks in advance,
Marina Carvalho
