Re: gnubol: Problem 10

[RKRayhawk@aol.com]

In a message dated 12/27/99 0:41:36 AM EST, wmklein@ix.netcom.com writes:

<< Just for those who are thinking about "knowing" their data division before
 they get to the Procedure Division - please do think (in advance) of the
 implications of GLOBAL data on a nested program.  I *think* that what you are
 talking about will still work with this - but did think I should alert you to
 those issues before you go off on a "dead-end" path.
 
 Bill Klein
   wmklein <at> ix.netcom.com
  >>

Actually I am a complete radical, and think that the symbol table should be 
instantiated BEFORE the data division is parsed!  By means a sneak preview in 
the outbound data buffer of the preprocessor. That same mechanism would have 
as its primary chore the delimitting of programs and divisions for the task 
master of parses; which obviously would put it in a position to see some 
aspects of nesting.

But your comments make me curious. Can data within a nested program be 
declared GLOBAL and thereby be seen 'above' as in by the programs which 
preceed it?  COBOL is still lexically scoped, right?

So as long as we have parsed a program's data division and that of every 
enclosing program, we should be able to consider driving all necessary data 
division attributes right into manifest tokens within the parser of the 
procedure division. Agreed?

Best Wishes
Bob Rayhawk
RKRayhawk@aol.com




--
This message was sent through the gnu-cobol mailing list.  To remove yourself
from this mailing list, send a message to majordomo@lusars.net with the
words "unsubscribe gnu-cobol" in the message body.  For more information on
the GNU COBOL project, send mail to gnu-cobol-owner@lusars.net.