(j3.2006) Did we overlook this intentionally?

[Malcolm Cohen]

Van Snyder asks:
> Did we overlook this intentionally?

!

> In 10-007 12.5.2.4p13 [294:28-31],

The complicated condition here combines two things: the default/C character 
exception (new to F2003), and the "normal" sequence association condition, and 
something about coarrays.

Let's cut it down to the "normal" sequence association permitting part:

   "If the actual argument is ... scalar, the corresponding dummy argument shall 
be scalar unless the actual argument is an element or substring of an element of 
an array that is not an assumed-shape, pointer, or polymorphic array."

Van comments:
> I assume the "pointer" caveat was to avoid discontiguous actual arguments.

The pointer caveat is not different from the assumed-shape caveat in that 
respect.

We also used a bigger hammer than necessary for polymorphic - these are not 
potentially discontiguous unless the dummy argument is not polymorphic.

> Did we intentionally not allow the case of contiguous pointers, or did we just 
> overlook this paragraph when we added the CONTIGUOUS attribute?

Extending sequence association was not (as far as I recall) part of the 
CONTIGUOUS attribute feature.

Whether the omission of sequence association here is a bad thing is another 
matter, but it's not obviously an error.

Cheers,
-- 
................................Malcolm Cohen, Nihon NAG, Tokyo.