Recursive derived type specifications - PR21458 - a standardquestion.

[Richard E Maine]

On Oct 1, 2005, at 10:26 AM, Paul Thomas wrote:

> I have scoured the standard and cannot find any mention of the need 
> for ordering of the derived type specifications.  I am going to submit 
> my recursion detector, together with the patch for PR24092.

It is specific to derived-type components and yes, it is to break 
recursion. (There could be other more sophisticated ways to do that, 
but this is the simple way that the standard does).

In f95, since that is the particular standard of most relevance here, 
it is the third constraint after R427.

  "If the POINTER attribute is not specified for a component, a 
type-spec in the component-def-stmt shall specify an intrinsic type or 
a previously defined derived type."

The "previously defined" is the important part here. Compare with the 
POINTER case in the next constraint, where any accessible derived type 
is allowed.

-- 
Richard Maine                |  Good judgment comes from experience;
Richard.Maine@nasa.gov       |  experience comes from bad judgment.
                             |        -- Mark Twain