Specifying the code page for character data

The code page that is in effect for a particular data item depends on the following aspects:

• Which USAGE clause you used
• Whether you used the NATIVE phrase with the USAGE clause
• Whether you used the CHAR(NATIVE) or CHAR(EBCDIC) compiler option
• The value of the EBCDIC_CODEPAGE environment variable
• Which locale is active

For USAGE NATIONAL data items, the code page is always UTF-16 in big-endian format.

For USAGE DISPLAY data items, COBOL for AIX chooses between ASCII, UTF-8, EUC, and EBCDIC code pages as follows:

• Data items that are described with the NATIVE phrase in the USAGE clause or that are compiled with the CHAR(NATIVE) option in effect are encoded in an ASCII, EUC, or UTF-8 code page.
• Data items that are described without the NATIVE phrase in the USAGE clause and that are compiled with the CHAR(EBCDIC) option in effect are encoded in an EBCDIC code page.

COBOL determines the appropriate code page as follows:

ASCII, UTF-8, EUC

From the active locale at run time

EBCDIC

From the EBCDIC_CODEPAGE environment variable, if set, otherwise the default EBCDIC code page from the current locale setting