Re: SQL pre-processor (was RPG IV release levels and complexity)

[Hans Boldt <boldt@xxxxxxxxxx>]

Jon Paris wrote:

>  >> Basically, just as it's unacceptable for SQL in Rochester to depend on
> the compiler for SQL processing, we in compiler development simply don't
> want that responsibility either!
>
> I know that Hans - and I didn't suggest anything of the sort.  The only
> dependency placed upon the compiler should be for it to recognize the SQL
> block.  The syntax checkers do that today so it is hardly arduous.
>
> It would take changes in the compiler and changes in methodology by the
> pre-compiler, but there is no need for either party to know the others job.

Jon: The problem is that if we were to adopt that OS/2 style 
"inner/outer" prep architecture, which is what was on the table, there 
/would/ be a requirement for the compiler to understand at least some of 
the SQL language.

That of course doesn't preclude the possibility of other architectures. 
But considering the complexities of the task, I'm not aware of any. It 
is a subject I've given some thought to, and indeed I think it is 
possible to come up with a generalized scheme in the compiler for 
processing embedded statements that are "pre-compiled" by some 3rd party 
program. But even in such "pie in the sky" schemes, it's hard to 
shoe-horn in all the nitty gritties of embedded SQL.

Cheers! Hans