----------  Create or Drop Index  ----------

An index on one or more attributes may be created  using  the  "create  index"
command.  This usually speeds up the retrieval of attribute values.

Syntax:

create [| distinct                         |]
        | unique                           |
        | [(]PRIMARY[,] OVERFLOW[)]        |

        | [|btree     |] index [NAME] [on] |
           |timeseries|

        | TABLE (ATTR_NAME {, ATTR_NAME})  |;
        | TABLE set ATTR_NAME {, ATTR_NAME}|

If the keyword "distinct" is specified, then there must not be  any  duplicate
or  null attribute values.  Using "distinct" has the same effect as specifying
PRIMARY as 1 and OVERFLOW as 0.  Note that "unique" is a  synonym  for  "dis-
tinct".

Values specified for PRIMARY and OVERFLOW  control  the  number  of  duplicate
values  allowed  for  each  attribute.   PRIMARY  sets the number of duplicate
values stored in the main index file; OVERFLOW sets the  number  of  duplicate
values stored per line in the secondary index file.

If OVERFLOW is non-zero, unlimited duplicates and nulls are allowed.  If OVER-
FLOW is zero, PRIMARY controls the maximum number of duplicates.

If index method is not specified, it defaults to btree.

To remove an index and its associated files from one or more  attributes,  the
"drop index" command is used.

Syntax:

     drop index | NAME                                                 |;
                | NAME [on] TABLE                                      |
                | [NAME] [on] |TABLE set ATTR_NAME {, ATTR_NAME} |     |
                | [NAME] [on] |TABLE (ATTR_NAME {, ATTR_NAME})   |     |
