17.2.3 Format of column structure information options

(a) Set format

• adb_import_input_record_size = input-record-length

~<integer>((1 to 536,870,912))(bytes)

You cannot specify this option if the file format of the input data files is CSV.

When the file format of the input data files is fixed-length, this option specifies the record length (data length per row) in the input data files.

If the records contain a line break code, specify the value in bytes including the length of the line break code.

(b) Command format

• adbcolumninfo

Specifies how to handle the input data in units of the columns defined in the table to be processed.

You can omit this operand if the file format of the input data files is CSV.

Specify this operand at least once if the file format of the input data files is fixed-length.

Make sure that the number of times this operand is specified does not exceed the number of columns defined for the table to be processed.

There is no need to use this operand to specify all columns defined in the table to be processed. For any columns not specified by this operand, the following data is assumed as the input data:

• If the file format of the input data files is CSV:

  Data with a field data number that is the same as the column definition order is assumed.

  A field data number is a number that is assigned according to the order in which field data items within a single record are arranged. The first field data item in the input record is assigned number 1, and the remaining field data items are assigned successive numbers. Even if a field data item is an empty string, it is counted as one field data item.

• If the file format of the input data files is fixed-length:

  An empty string is assumed.

-n column-name

~<character string>((1 to 100 bytes))

Specifies a column name defined for the table to be processed.

Note the following:

• If a column name contains a space, enclose the column name between double-quotation marks (").

  Example: "CΔ1"

  Legend: Δ: Space

• If a column name contains any lowercase letters, enclose the column name between sets of a backslash and a double-quotation mark (\"column-name\"). If the column name is not enclosed in sets of \", all characters in the column name will be treated as uppercase letters.

  Example: \"c1\"

  If a column name contains both lowercase letters and spaces, enclose the column name between sets of a backslash and a double-quotation mark and then between double-quotation marks (").

  Example: "\"cΔ1\""

  Legend: Δ: Space

-p start-position,length

You cannot specify this option if the file format of the input data files is CSV.

Specification of this option is required if the file format of the input data files is fixed-length.

• start-position

  ~<integer>((1 to 536,870,912))(bytes)

  Specifies the start position of the data that is to be stored in the column specified in the -n option. Specify the position relative to the beginning of the input record (where the beginning of the input record is 1).

• length

  ~<integer>((1 to 536,870,912))(bytes)

  Specifies the length in the input record of the data that is to be stored in the column specified in the -n option.

  The following cannot be specified for length in the -p option:

  • Any value that is greater than the maximum number of characters in the column with the column name specified in the -n option in character format

  • An odd value if the column with the column name specified in the -n option contains binary data

  • A value of 7 or smaller if bin is specified for the -b option

Important

• No error occurs if the range specified in this option overlaps the range of values specified in the -p option in any other adbcolumninfo operand.

• When input data is enclosed in enclosing characters, the enclosing characters are also stored as part of the data. If you do not want the enclosing characters to be treated as data to be stored, specify this option so that ranges do not include enclosing characters.

-c empty-string-comparison-value

~<character string>((1 to 255 bytes))

You cannot specify this option if the file format of the input data files is CSV.

When the file format of the input data files is fixed-length, this option specifies data that is to be treated as an empty string. Any input data that matches this option value is treated as an empty string.

Make sure that the length of this option's value in characters matches the length specified in the -p option. If the length specified in the -p option is greater than 255 bytes and the first 255 bytes of data from the start position specified in the -p option matches the value specified in this option, the data is treated as an empty string.

If the specified value contains any spaces, enclose the value between double-quotation marks (").

-r {field-data-number|empty_string}

• field-data-number

  ~<integer>((1 to 536,870,912))

  Specifies the field data number of the input data when the format of input data files is CSV.

  You cannot specify this option if the file format of the input data files is fixed-length.

  A field data number is a number that is assigned according to the order in which field data items within a single record are arranged. The first field data item in the input record is assigned number 1, and the remaining field data items are assigned successive numbers. Even if a field data item is an empty string, it is counted as one field data item.

• empty_string

  Specify this option to assume an empty string for the input data.

If this option is omitted, data with a field data number that is the same as the definition order of the column corresponding to the column name specified in the -n option is assumed as the input data.

Important

• Even if the field data number specified in this option matches the field data number of the -r option of another adbcolumninfo operand, no error occurs. The same field data is stored in multiple columns of the table to be processed.

• If you specify a field data number in this option and no applicable field data exists in the input record, an error occurs. If you omit this option and no data that corresponds to the assumed field number exists in the input record, an error occurs.

-b {hex|bin}

Specifies the format of the binary data in the input record when a binary type is defined for the column name specified in the -n option of the adbcolumninfo operand.

• hex

  Specifies that binary data in the input record is to be treated as binary data in hexadecimal format.

• bin

  Specifies that binary data in the input record is to be treated as binary data in binary format.

If this option is omitted, hex is assumed.

The following table shows the relationship between the value to be specified in this option and the data type defined for a column with a column name specified in the -n option of the adbcolumninfo operand.

Table 17‒7: Relationship between the data type defined for a column and the value to be specified in the -b option

No.	Data type defined for a column with a column name specified in the -n option	Value specified in the -b option	Data format of the input data
1	Binary	--	Treated as binary data in hexadecimal format
2		hex	Treated as binary data in hexadecimal format
3		bin	Treated as binary data in binary format
4	Data type other than binary	--	Treated as the data type defined for the column
5		hex	An error occurs (the KFAA50265-E message is issued).
6		bin

Legend:

--: Specification of the -b option is omitted.

(a) When an invalid value is specified for a column structure information option

The table below lists and describes the message IDs that are displayed when an invalid value is specified for a column structure information option and the possible errors. Correct the error and then re-execute the adbimport command.

(b) When there is a problem with consistency between a column structure information option value and the table to be processed

The table below lists and describes the message IDs that are displayed when there is a problem with consistency between a column structure information option value and the table to be processed, and the possible errors. Correct the error and then re-execute the adbimport command.