Format

      { DUPLICATES                    }
      { CHANGES                       }
      { NULL_KEY                      }
      {                               }
      { NULL_VALUE null-value         }
      { DATA_AREA area-number         }
      { DATA_FILL numeric-literal     }
      {                               }
      { DATA_KEY_COMPRESSION          }
      { DATA_RECORD_COMPRESSION       }
      { INDEX_AREA area-number        }
      {                               }
      { INDEX_COMPRESSION             }
      { INDEX_FILL numeric-literal    }
      { LEVEL1_INDEX_AREA area-number }
      {                               }
      { NODATA_KEY_COMPRESSION        }
      { NODATA_RECORD_COMPRESSION     }
      { NOINDEX_COMPRESSION           }
      {                               }
      { PROLOG numeric-literal        }
      { SEGMENT within-name-clause    }
      { SORTED BY ASCENDING           }
      {                               }
      { SORTED BY DESCENDING          }

1  –  Parameters

1.1  –  null-value

    Specifies a null key value for an alternate index. This value is
    either an integer or a single character within quotation marks.

1.2  –  area-number

    Specifies a positive integer that corresponds to the number of a
    previously defined area.

1.3  –  numeric-literal

    Specifies a positive integer. See the OpenVMS documentation on
    RMS for the valid values for a particular keyword. For a mapping
    of CDO keywords to the corresponding RMS symbolic field offset,
    see RMS_Keyword_Mapping_Tables.

1.4  –  within-name-clause

    field-name { IN record-name } ...

    Specifies the field or record names that are part of a segmented
    key. A segmented key allows you to define a key that accesses
    noncontiguous fields in a record. The record name is the same
    name that defines the data structure of the RMS database in the
    DEFINE RMS_DATABASE command.

2  –  Description

    Key properties define the characteristics of a key in an indexed
    file. You define keys in numerical order, starting with 0. By
    default, Key 0 always has a value of DATA_AREA 0.

    A segmented key allows you to define a key that accesses non-
    contiguous fields in a record. You can specify from 2 to 8
    segments for each segmented key in an indexed file. For example:

    o  A field in a record

    o  A field in a record that is in a record

    o  A field in a structure in a record

    You must specify a segment for each key you define. A segmented
    key can contain up to 8 segments.

    The field element or elements specified by the keyword SEGMENT
    determine the data type of the key. A key with only 1 segment can
    point to a field that has any of the following RMS data types:

       DSC$K_DTYPE_B
       DSC$K_DTYPE_BU
       DSC$K_DTYPE_W
       DSC$K_DTYPE_WU
       DSC$K_DTYPE_L
       DSC$K_DTYPE_LU
       DSC$K_DTYPE_Q
       DSC$K_DTYPE_QU
       DSC$K_DTYPE_P
       DSC$K_DTYPE_T

    See the OpenVMS documentation on RMS for more information on
    valid RMS data types.

    Each segment of a key can contain up to 255 characters. The
    sum of all characters for all key segments must not exceed 255
    characters.

    When defining a segment, you cannot use a field that is in the
    variant portion of a record.

    If you specify NULL_VALUE without specifying NULL_KEY, CDO
    automatically specifies NULL_KEY. If you specify NULL_KEY without
    specifying NULL_VALUE, the default value for NULL_VALUE is
    0. When you display a RMS database element with a NULL_VALUE
    property, the null value appears as a decimal value.

    For a description and valid values for a particular keyword, see
    the OpenVMS documentation on RMS.

    For a mapping of CDO keywords to symbolics, see RMS_Keyword_
    Mapping_Tables.

3  –  Examples

  CDO>  DEFINE RMS_DATABASE EMPLOYEE_
 INFO DESCRIPTION IS "INFORMATION ON"
  cont> "CURRENT EMPLOYEE".
  cont>   RECORD EMPLOYEE_REC.
     .
     .
     .
  cont>   KEYS.
  cont>     KEY 0
  cont>       DATA_AREA 0
  cont>       INDEX_AREA 0
  cont>       SEGMENT EMP_ID IN EMPLOYEE_REC.
  cont>     KEY 1
  cont>       DUPLICATES
  cont>       DATA_AREA 1
  cont>       INDEX_AREA 2
  cont>       SEGMENT LAST_NAME IN EMPLOYEE_REC.
  cont>   END KEYS.
  cont> END EMPLOYEE_INFO RMS_DATABASE.
  CDO>

      This example shows the syntax for defining the DATA_AREA,
      INDEX_AREA, SEGMENT, and DUPLICATES key properties in the
      EMPLOYEE_INFO RMS database element.