(j3.2006) Chaining type-bound procedure references

[Rafik Zurob]

Hello

While I use this feature extensively in my C++ programs, I (personally) 
strongly oppose it in today's Fortran.  C++ can get away with 
x.foo().bar() because inlining is part of the language and the definitions 
of inline member functions appear in the same translation unit as the 
call.  That's not the case in Fortran. 

val = x%foo()%bar()

would require a temp pointer

temp_ptr_dscr => x%foo()
val = temp_ptr_dscr%bar()

The most common usage would be for foo to be written as a module procedure 
and to be accessed via use-association.  i.e. It's going to be in a 
separate compilation unit and require inter-procedural analysis to inline. 
 It's a similar situation to ASSOCIATE, but ASSOCIATE encourages the user 
to reuse the temp since the temp is really the associate name.  Allowing 
the x%foo()%bar() syntax encourages many references to the same pointer 
function, thus resulting in many temps (and many pointer assignments). 
Even if you forget about inlining, you would expect the compiler to 
attempt to reuse the temp pointers / get rid of multiple calls to foo. 
That's not easy, even if you mandate that pointer functions like foo must 
be pure.  Note that Fortran has the notion of pure, but does not have the 
notion of const functions that are pure and don't read any use-associated 
or host-associated data.

Regards

Rafik
PS  These concerns are also why I don't like F2008's pointer function refs 
as vars feature.