Secondary keys are additional keys that can be defined for internal tables to optimize the access to the table using fields that are not part of the primary key. Secondary keys can be either sorted or hashed, depending on the table type and the uniqueness of the key. Secondary keys have the following characteristics1:
A. Secondary keys must be chosen explicitly when you actually read from an internal table. This means that when you use a READ TABLE or a LOOP AT statement to access an internal table, you have to specify the secondary key that you want to use with the USING KEY addition. For example, the following statement reads an internal table itab using a secondary key sec_key:
READ TABLE itab USING KEY sec_key INTO DATA(wa).
If you do not specify the secondary key, the system will use the primary key by default2.
B. Multiple secondary keys are allowed for any kind of internal table. This means that you can define more than one secondary key for an internal table, regardless of the table type. For example, the following statement defines an internal table itab with two secondary keys sec_key_1 and sec_key_2:
DATA itab TYPE SORTED TABLE OF ty_itab WITH NON-UNIQUE KEY sec_key_1 COMPONENTS field1 field2 sec_key_2 COMPONENTS field3 field4.
You can then choose which secondary key to use when you access the internal table1.
D. Sorted secondary keys do NOT have to be unique. This means that you can define a sorted secondary key for an internal table that allows duplicate values for the key fields. A sorted secondary key maintains a predefined sorting order for the internal table, which is defined by the key fields in the order in which they are specified. For example, the following statement defines a sorted secondary key sec_key for an internal table itab that sorts the table by field1 in ascending order and field2 in descending order:
DATA itab TYPE STANDARD TABLE OF ty_itab WITH NON-UNIQUE SORTED KEY sec_key COMPONENTS field1 ASCENDING field2 DESCENDING.
You can then access the internal table using the sorted secondary key with a binary search algorithm, which is faster than a linear search3.
The following are not characteristics of secondary keys for internal tables, because:
C. Hashed secondary keys do NOT have to be unique. This is false because hashed secondary keys must be unique. This means that you can only define a hashed secondary key for an internal table that does not allow duplicate values for the key fields. A hashed secondary key does not have a predefined sorting order for the internal table, but uses a hash algorithm to store and access the table rows. For example, the following statement defines a hashed secondary key sec_key for an internal table itab that hashes the table by field1 and field2:
DATA itab TYPE STANDARD TABLE OF ty_itab WITH UNIQUE HASHED KEY sec_key COMPONENTS field1 field2.
You can then access the internal table using the hashed secondary key with a direct access algorithm, which is very fast.
E. Secondary keys can only be created for standard tables. This is false because secondary keys can be created for any kind of internal table, such as standard tables, sorted tables, and hashed tables. However, the type of the secondary key depends on the type of the internal table. For example, a standard table can have sorted or hashed secondary keys, a sorted table can have sorted secondary keys, and a hashed table can have hashed secondary keys1.
References: 1: Secondary Table Key - ABAP Keyword Documentation 2: READ TABLE - ABAP Keyword Documentation 3: Sorted Tables - ABAP Keyword Documentation : Hashed Tables - ABAP Keyword Documentation
