Data loading to a table for which a character set has been defined

When data is loaded to a table that contains a column for which a character set has been defined (character set-defined column), the character codes in the input data are converted into the character codes specified for the column. Whether character codes are to be converted depends on the combination of the input data file format and the table that is the subject to data loading. The table below shows whether character code conversion is performed.

Table 5-1 Whether character code conversion is performed

Format of input data file			Whether the table subject to data loading contains a character set-defined column
Format of input data file			Yes	No
DAT format			Y	N
Binary format			N^#1
Fixed-size data format			Y
File output by pdrorg	DAT format		Y
	Binary format	Unload file for pdload^#3	Y
		Unload file for pdrorg^#3	Y^#2
		Table transfer unload file^#3	Y

Legend:: Y: Character code conversion is performed.; N: Character code conversion is not performed.

#1: Data stored in the database is not guaranteed because the utility stores data in the database without performing character code conversion.

#2: When you execute pdrorg, unload the data with -k unld specified and then use the unload data file that is output with the original character codes of the predefined data type. If you perform reorganization with -k rorg specified and use the unload data file that is output in the codes for the column for which a character set has been defined, the utility displays the KFPL15047-E message and cancels the processing.

#3: For details, see 5.5.6 pdrorg-output binary files.

pdload does not perform character code conversion when you perform data loading to a table that contains a character set-defined column in a binary-format input data file created by the user. Therefore, you must handle conversion of the data in the input data file into the character codes for the character set-defined column. The table below shows the character codes used when input data files are created.

Table 5-2 Character codes used when input data files are created

Input data file format	Whether the table subject to data loading contains a character set-defined column
Input data file format	Yes	No
DAT format	Default character set	Default character set
Binary format	Character codes specified for the column
Fixed-size data format	Default character set

Organization of this subsection: (1) Notes

When you use pdrorg with the -W option specified to unload data from a FIX table, the format of the obtained unload data file is the same as the format of the user-created binary file. Therefore, even if you omit the -W option when you execute pdload, you can load data correctly to a table containing no character set-defined columns. On the other hand, if you omit the -W option and perform data loading to a table that does contain a character set-defined column, the utility will load data into the character set-defined column without converting the character codes because the unload data file uses character codes of the default character set. In this case, you must specify -W.
If a character set-defined column is used as the partitioning key, the partitioning condition is checked based on the values obtained after the character codes have been converted.

5.1.4 Data loading to a table for which a character set has been defined

(1) Notes