Re: [Patch, Fortran] PR 36426

[Daniel Kraft <d at domob dot eu>]

Hi Janus,

Janus Weil wrote:
> About comment #0 in the PR: I'm assuming that a statement like
>
>    character(len=len(x)) :: x
>
> is invalid, since the symbol x is used in it's own declaration. Can
> someone confirm that? This line gives the somewhat strange error
> message:
>
>    character(len=len(x)) :: x
>                             1
> Error: Symbol 'x' at (1) already has basic type of REAL
>
> Using "implicit none" the error transforms into
>
>    character(len=len(x)) :: x
>                     1
> Error: Symbol 'x' is used before it is typed at (1)
>
> which is much clearer.

this is a fix I did some months ago.  The basic type REAL stems from the 
fact that x is implicitelly typed (and thus REAL) once it is used in the 
LEN call.

I hope the error is at least somewhat clear and ok this way, but I don't 
know how to make it better for implicit typing.

Cheers,
Daniel

--
Done:  Arc-Bar-Cav-Rog-Sam-Val-Wiz
To go: Hea-Kni-Mon-Pri-Ran-Tou