Specifies a set of columns as an index.
This element is part of the table element. The short-name identifier is only allowed in a CREATE TABLE DESCRIPTION or a CREATE CLUSTER DESCRIPTION statement.
Syntax:
index_identifier |
Represents the name of an index and must conform to the rules specified earlier in this section in section Identifiers. |
column_identifier |
Name of a column to be used in the forming of an index. |
start_position |
Start position within the column when defining an Adabas descriptor. |
end_position |
End position within the column when defining an Adabas descriptor. The end position must be greater than the start position. |
shortname_identifier |
Adabas short name. |
In order to improve an existing application's performance, establish an index for one or more base table column(s).
You can create an Adabas descriptor that reflects the capabilities of the Adabas database system's descriptors definition when you have the table index element. For a detailed discussion of Adabas descriptors, please refer to the Adabas documentation for your environment.
A ranges specification is when start and end positions are specified. This allows an index specification to be restricted to sub-elements of a column.
The following conventions hold true for the explanations below:
Let Cl = ( c1,... , cn ) be a column list of one or more columns for which this table index element is specified.
Let T be the table where the columns of Cl resides.
INDEX:
A
INDEX is used to allow more efficient base table access. The index is
based on one or more column(s) of a base table, when the listed columns'
are considered as an entity.
An Adabas Superdescriptor will be generated if the number of columns in
the column list is greater than one.
An Adabas Descriptor will be added if the index is in a single column with
no range specified.
An Adabas Subdescriptor will be generated if a range specification is in
a single column.
HAVING UNIQUE INDEX:
This feature is provided
for compatibility and will be removed in future versions. Use a UNIQUE
constraint instead.
UQINDEX:
A UQINDEX is an index that generates a unique Adabas sub- or superdescriptor on a sub-table column. SQL does not conisder this descriptor to be unique; it cannot be represented by a normal "unique constraint."
The shortname identifier is only used in a CREATE TABLE DESCRIPTION or a CREATE CLUSTER DESCRIPTION statement.
A specification of a UQINDEX is only valid for level 1 and level 2 base tables (subtables).
You are not allowed to specify a UNIQUE INDEX together with a UNIQUE constraint or a PRIMARY KEY.
When using a HAVING UNIQUE INDEX in conjunction with a SUPPRESSION attribute the attribute NOT NULL is not permitted.
When using a HAVING UNIQUE INDEX in conjunction with a DEFAULT Adabas attribute, the attribute NULL is not permitted.
The table index element is not part of the Standard.
None.