Conference turris::fortran

1204.0. "# as continuation line is not a good idea." by BACHUS::SABLON (Mich�le Sablon, TP/IM Support Belgium 856-7238) Thu Feb 27 1997 12:21

OSF V3.2-* & 4.0-*
dfa401

	A customer, currently migration to Digital UNIX, got strange 
	compilation errors with some of his programs while using the option 
	-fpp. He finally identifies the problem to be caused by the use of 
	the '#' sign as continuation character (column 6).

	The fact is, I believe, that the C compiler (and therefore the fpp)
	recognizes as macro indicateur the # when it is the first non-blank 
	character.

	Conflicting with the legal use in FORTRAN of this character as
	continuation sign.

	A restriction to document, maybe ? And a usage to avoid, surely. But 
	who knew at the time they wrote the programs that C would become so
	central ?

	Best,
	Mich�le

-----

	Fyi, the small sample exercice:

tt.f
       program tt
       print *,
     #  1
       stop
       end

$ f77 -o toto1 -fpp toto1.f

------------------------------------------------------------------------
toto1.f:
    3:       #  1
                 ^
** warning  on line 3 in toto1.f: line directive is missing file name.

fort: Severe: toto1.f, line 2: Variable name, constant, or expression invalid it
       print *,
---------------^

Um...  if the customer doesn't want to run the program through the preprocessor,
then they shouldn't ask the driver to do so.  What's the issue?  If the customer
is going to use preprocessor directives in their source, then it's up to them
to make sure that anything that looks like a directive IS a directive.  # is
not typically used as continuation and is non-standard.

				Steve

	No real issue Steve. Just wanted to raise the point. 

	I'm also glad to ear it is a non-standard.

	Best,
	Mich�le.