Indexes are identified in the ABAP Dictionary via the table name and an index ID of up to three digits. Together they make up a unique index name when you create the index in the database. The possibilities for index names are as follows:
Table name TAB456789 and the index ID "A2" make up the index name TAB456789~A2 in the database.
Table names with a maximum of 18 digits are used. For tables with names 15 and 16 digits long, only the first two digits are relevant to the index ID, as only these are used for the index name in the database. If a name has 16 digits and a two-digit index ID, the separator is left out.
Table name TAB4567890123456 and the index ID "A23" therefore make up the index name TAB4567890123456A2 in the database.
From the table name TAB456789 and the index ID "A2", you get the index name TAB456789_A2 in the database.
Shorter table names are padded to ten digits with underscores. Likewise one-digit and two-digit index IDs are padded to three places with underscores. These alternative names are necessary for particular tables because of the upgrade procedure technology. New indexes are then only created according to this naming convention if all other indexes of the table follow this convention.
From the table name TAB456789 and the index ID "A2"; you get the index name TAB456789_A2_X in the database.
Longer table names are truncated to seven digits; shorter table names are padded to seven digits with underscores. This naming convention is only found for indexes that were created in the database before Release 3.0. When you delete and recreate this kind of index, it immediately follows one of the first two naming conventions. (Before Release 3.0, the index ID consisted of one digit and table names often had seven digits.)