The following example shows the behavior of the LANGUAGE compiler option in a batch environment. The default installation option is ENGLISH (abbreviated EN), and the invocation option is XX, a nonexistent language.
CBL LANG(JP),FLAG(I,I),APOST,SIZE(MAX) (1)
IDENTIFICATION DIVISION. (2)
PROGRAM-ID. COMPILE1.
. . .
END PROGRAM COMPILE1.
CBL LANGUAGE(YY) (3)
CBL SIZE(2048K),LANGUAGE(JP),LANG(!!) (4)
IDENTIFICATION DIVISION. (2)
PROGRAM-ID. COMPILE2.
. . .
END PROGRAM COMPILE2.
IDENTIFICATION DIVISION.
PROGRAM-ID. COMPILE3.
. . .
END PROGRAM COMPILE3.
CBL LANGUAGE(JP),LANGUAGE(YY) (5)
. . .
For the program COMPILE1, the default language English (EN) is in effect when the compiler scans the invocation options. A diagnostic message is issued in mixed-case English because XX is a nonexistent language identifier. The default EN remains in effect when the compiler scans the CBL statement. The unrecognized option APOST in the CBL statement is diagnosed in mixed-case English because the CBL statement has not completed processing and EN was the last valid language option. After the compiler processes the CBL options, the language in effect becomes Japanese (JP).
In the program COMPILE2, the compiler diagnoses CBL statement errors in mixed-case English because English is the language in effect before the first program is used. If more than one LANGUAGE option is specified, only the last valid language specified is used. In this example, the last valid language is Japanese (JP). Therefore Japanese becomes the language in effect when the compiler finishes processing the CBL options. If you want diagnostics in Japanese for the options in the CBL and PROCESS statements, the language in effect before COMPILE1 must be Japanese.
The program COMPILE3 has no CBL statement. It inherits the language in effect, Japanese (JP), from the previous compilation.
After compiling COMPILE3, the compiler resets the language in effect to English (EN) because of the CBL statement. The language option in the CBL statement resolves the last-specified two-character alphanumeric language identifier, YY. Because YY is nonexistent, the language in effect remains English.