The runtime configuration variable CODE_SYSTEM tells the runtime if double-byte character data is being accepted or displayed, and which code system (that is, which standard for encoding Japanese and other Asian character sets, for example) is being used. Each code system has a range of values that it allows within each byte of a two-byte character, so identifying the code system allows the runtime to recognize character boundaries when it is processing double-byte data for ACCEPT and DISPLAY statements.
Setting CODE_SYSTEM to the proper value allows your COBOL applications to handle input and display of double-byte character data without source program changes. The syntax is:
CODE_SYSTEM setting
The table below shows the possible settings of the CODE_SYSTEM variable, the code system to which each setting refers, and some examples of operating systems to which the particular code system applies:
| Setting | Code System | Op. System Examples |
| BIG5 | Big Five (placecountry-regionTaiwan) | Chinese DOS, Windows |
| DBC | Acucorp Generic Double-byte Coding Scheme | other double-byte machines |
| EUC | Extended UNIX | Most UNIX machines |
| GB | Code of Chinese Graphic Character Set (People's Republic of placecountry-regionChina) | Chinese DOS, Windows |
| KSC | Korean Character Standard | Korean DOS |
| SJC | Shift JIS Code (Japanese Industrial Standard) | DOS/V, Windows, some UNIX machines |
The default 0 means ASCII or EBCDIC single-byte characters.
The following table shows the decimal values that the respective code systems allow for each byte of the two-byte character:
| Code System Setting | 1st byte | 2nd byte |
| BIG5
(second format) |
161 - 254
161 - 254 |
64 - 126
161 - 254 |
| DBC | 128 - 255 | 128 - 255 |
| EUC
(second format) |
142
161 - 254 |
161 - 223
161 - 254 |
| GB and KSC | 161- 254 | 161 - 254 |
| SJC
(second format) |
129 - 159
224 - 239 |
64 - 252 (not 127)
64 - 252 (not 127) |