Deleting Data from Existing MU and PE Fields
It is not possible to delete occurrences of PE fields using Adabas direct calls, and therefore we cannot offer this capability through SQL.
The rest of this section describes how to delete occurrences of MU fields.
When using flattened version of the table, deletes of MU fields are accomplished by using the update SQL statement. Simply blank (if the data type is text) or zero (if the data type is numeric) out the field occurrence you want to delete.
For example, if a record in the employees table currently has two occurrences in the LANG MU field, and you want to delete the 2nd occurrence, use the following SQL statement (you must know the ISN number of the existing record you want to modify):
update employees set LANG_2 = '' where ISN_12_11 = 1004
When using the rotated and root version of the table, a SQL delete is used to remove the data for MU fields. The desired occurrence for deletion must be specified using the cnxarraycolumn pseudo field. This pseudo field is zero-based. So if you want to delete the first occurrence of a MU field, you would specify cnxarraycolumn=0 in the Where clause of the delete statement. If you want to delete the second occurrence of a MU field, you would specify cnxarraycolumn=1 in the where clause of the delete statement.
Using the same example, deleting the second occurrence of the LANG MU field would require the following SQL statement.
delete from employees_lang
where ISN_12_11 = 1004 and cnxarraycolumn = 1