Tobias,
Thanks for checking this. I will wait until the next patch has stabilized and
is applied to the trunk, then build a new version.
I think many will want to use this feature in conjunction with optional
arguments. Hence it might be a good idea to add a unit test for the use case.
Here is a simple example - is working fine:
program cond_expr
implicit none
integer :: ival
ival = init ()
! print *, (ival == -99 ? "pass" : "fail") ! char type not supported yet
print *, merge ("pass", "fail", ival == -99)
ival = init (42)
! print *, (ival == 42 ? "pass" : "fail") ! char type not supported yet
print *, merge ("pass", "fail", ival == 42)
contains
integer function init (i)
integer, intent(in), optional :: i
init = (present (i) ? i : -99)
end function
end program
When the next patch is ready, and character strings are supported, those two
print statements could be switched from using MERGE to conditional expressions.
Walter
-----Original Message-----
From: Tobias Burnus <[email protected]>
Sent: Sep 22, 2025 11:13 AM
To: Walter Spector <[email protected]>, <[email protected]>
Subject: Re: [PATCH] fortran: implement conditional expression for fortran 2023
Hi Walter,
Walter Spector wrote:
> I'm really looking forward to conditional expressions because I like > using
> them in other languages. So I built a gfortran from the trunk a > few days
> ago to test the mod.
>
> Here is a simple test case that compiles, but SEGVs at run-time:
This seems to be fixed by Yuao's latest patch. In one of my local branches, I
have it applied – in the other not. And the one that has it works, in the
other, I get the segfault. Thus, it really must be that patch.
The patch is
https://gcc.gnu.org/pipermail/gcc-patches/2025-September/695638.html and adds
additionally support for the character type. Unfortunately, I found some
follow-up issues that need to be fixed before character support is ready for
prime time.
Thanks for testing!
Tobias
