[{{ restruct DBM-name.dataset-name
    {{ field redefined-field-name [position data-start-position]
       attr redefined-field-attribute(length)[not null]}} ...}} ...]

The specified DBM name, dataset name, and redefined field names must correspond to the target HiRDB's authorization identifier, table name, and column names, respectively.

• restruct DBM-name.dataset-name
  ~ <symbolic name of 1-6 characters>
  In the case of the PDM2 E2 log format, specify the DBM name and dataset name. In the case of a variable dataset using the coded record facility, specify the dataset name followed by the value of the record code.
  
• field redefined-field-name
  ~ <symbolic name of 1-30 characters>
  In the case of the PDM2 E2 log format, specify a name for a redefined field.
  
• position data-start-position
  ~ <<0>>
  Specify the position relative to the beginning of the row data (which is 0) that is the beginning of the redefined field.
  
• attr redefined-field-attribute(length) [not null]
  Specify a field attribute for the redefined field and the field's length. If the update data is 4040..., specify not null to handle this data as real data, rather than as the initial value (null data). However, this operand is not applicable to packed or zoned data. The following table lists the permitted field attributes and lengths.
  

  Table 5-23 Field attributes and lengths permitted for redefined fields

  Field attribute (length)	Description	Field (bytes)
  CHAR(n)	Character string data n: Number of characters (1 <= n <= 30000)	n
  NCHAR(n)	Japanese-language character string data n: Number of characters (1 <= n <= 15000)	n x 2
  PACK(m[,n])	Signed packed decimal data m: Integer section. n: Fraction part. (0 <= m <= 29,0 <= n <= 29,1 <= m + n <= 29)	(m + [n] + 1) / 2 Value is rounded up.
  PACKNS(m[,n])	Unsigned packed decimal data m: Integer section. n: Fraction part. (0 <= m <= 29,0 <= n <= 29,1 <= m + n <= 29)	(m + [n] + 1) / 2 Value is rounded up.
  UNPACK(m[,n])	Signed zoned decimal data m: Integer section. n: Fraction part. (0 <= m <= 29,0 <= n <= 29,1 <= m + n <= 29)	m + [n]
  UNPACKNS(m[,n])	Unsigned zoned decimal data m: Integer section, n: Fraction part (0 <= m <= 29,0 <= n <= 29,1 <= m + n <= 29)	m + [n]
  COMP(m[,n])	COBOL fixed-point binary data m: Integer section. n: Fraction part. (0 <= m <= 18,0 <= n <= 18,1 <= m + n <= 18)	1 <= m [ + n] <= 4: 2 5 <= m [ + n] <= 9: 4 10 <= m [ + n] <= 18: 8
  BINARY(m[,n])	PL/I fixed-point binary data m: Precision. n: Scaling. (1 <= m <= 31,0 <= n <= 31,m >=n)	1 <= m <= 15: 2 16 <= m <= 31: 4
  FLOAT(n)	Floating-point data n: Length in bytes (n = 4,8)	n