Re: It Hurts When I Do This

Richard Maine wrote:
> Beliavsky <beliavsky@[EMAIL PROTECTED]
> wrote:
> 
>> The blog entry mentions a problem that combines allocation upon
>> assignment (new in Fortran 2003) and array sections. Although some
>> compilers now sup****t a line of code
>>
>> A = B
>>
>> where A is declared ALLOCATABLE and is here allocated to the shape of
>> B and then assigned to it, I am wary of using this feature, because
>> assignments in Fortran, up to now, have been separated from
>> allocations (unless I am mistaken),
> 
> Depends what you mean by "up to now". That particular variant of
> allocate-on-assignment is new to f2003. But allocation of allocatable
> components on assignment is an integral part of the allocatable TR.
> Allocatable components pretty much don't work without that. Admitedly,
> the alocatable TR isn't exactly very old.
> 
> The fact that there was already such allocation for components was one
> of the arguments for adding the feature to objects that weren't
> components. It seemed inconsistent to have automatic reallocation for
> one, but not the other. It also seemed like it would invite people to
> add the superfluous complication of putting things in a derived type
> just to get the automatic allocation.
> 
>> I wonder what people think of this style suggestion.
> 
> I repeat my unanswered question to... (I forget who; someone else in
> this thread): what about allocatable strings? It has been my observation
> that the rather large majority of people intuitively expect strings to
> act like the new allocatable ones will. They expect that
> 
>   string = 'hello'
> 
> will end up with string being 4 characters in length. They have to be
> carefully trained to understand how Fortran fixed-length strings work.

C programmers might expect 4.  Most Fortran people will expect 5 ;).

Dick Hendrickson
> 
> I expect allocatable strings to eventually become almost the norm, with
> fixed-length strings reserved for special cases.
> 
> So I wonder whether you would also extend your style choice to
> allocatable strings. I personally think that the automatic reallocation
> is more intuitive than the current scheme, which is (a bit
> inconsistently)
> 
> 1. For strings, pad or truncate to make things work. It is never an
> error.
> 
> 2. For arrays, if the size wasn't pre-allocated just right, it is an
> error.
> 
> I think it is much more natural that, if someone writes
> 
>   X = something
> 
> then they probably want X to end up having the same size or length as
> the something. I think that the resistance to this reflects being
> trained in prior Fortran practice, as distinguished from ab initio
> reevaluations of natural useability. Alternatively, maybe it's just that
> my intuition is different.
>