Re: >> IMP ESQL ...

2005-11-25 12:44:21 MET

"William M. Klein" wrote:

>I would certainly go with
>
> >>IMP ESQL ...
>
> as
>
> *>>IMP ESQL
>
> *should* be treated as a comment (while the first should be "meaningful"
> to the compiler/precompiler combination.
>
> Remember, it is POSSIBLE for the programmer to actually code in
>
> *>>IMP This is a comment
>
> and you shouldn't have to distinguish that from the "meaningful"
> >>IMP
>
> directives (IMHO) <G>

We now discussed it at length.

The thing is: The COBOL compiler "sees" the output of the PreCompiler.
SQL-statements, for instance, are commented out, and a few
COBOL-statement are generated (communication with the database).
So, in a listing, there is a different stuff from what the programmer
actually wrote anyway. The ESQL-COBOL-programmer knows that.

In a strict sense, COBOL with embedded SQL is no COBOL. An
implementor can do what he likes.

This implies the solution of making ESQL-directives into a comment.

Only problem: different dirctives (" >> IMP ...") may mean different
things to different vendors. - But this is the case anyway, even
without a precompiler.

Well - the thing is not urgent - we sleep over it.

If need be, in office hours.

Herwig



.