NonStop SQL/MP Reference Manual

NonStop SQL/MP Reference Manual—142115
D-3
Data Types
column is a single-byte or double-byte character set, and on whether the data type
declaration for the column includes the VARYING clause.
CHAR[ACTER] SET { ISO88591 | ISO88592 | ... UNKNOWN }
associates a character set with the column or indicates that the character set
associated with the column is unknown.
The character set can be one of the single-byte character sets ISO 8859/1 through
ISO 8859/9 or one of the double-byte character sets Kanji (KANJI) or KS C5601
(KSC5601). If you specify one of the double-byte character sets, however, the data
type declaration for the item cannot include the UPSHIFT or COLLATION clause.
(SQL cannot upshift double-byte characters or associate collations with double-byte
characters. SQL always collates double-byte characters according to their binary
values.) See Character Sets
on page C-16 for more information about specific
character sets.
UNKNOWN specifies that the data has an unknown character set. Specifying
UNKNOWN is equivalent to omitting the character set specification. SQL treats the
data as 8-bit data.
UPSHIFT
directs SQL to upshift characters before storing them in the column.
You cannot specify UPSHIFT for a column associated with a double-byte character
set. If you specify both the UPSHIFT clause and the COLLATE clause for a column,
the rules for upshifting depend on those specified in the collation named in the
COLLATE clause.
COLLATE { collation | CHARACTER SET }
specifies a collating sequence for the column. The collating sequence determines the
default ordering of data returned by a SELECT and the default ordering for
comparison predicates, although you can override these defaults for specific
statements. For a key column, the collating sequence also determines storage order
within the table.
Data Type Key-Sequenced Relative or Entry-Sequenced
Single-byte unvarying 4061 4072
Single-byte VARYING 4059 4070
Double-byte unvarying 2030 2036
Double-byte
VA RY IN G
2029 2035
collation is the name of an existing collation (or an equivalent DEFINE)
that specifies a collating sequence that is associated with the same
character set as the column
CHARACTER SET specifies a collating sequence based on the binary value of
characters in the column