Following OK from Steve, with minor locus changes in access specification
errors, pushed to mainline as r16-3099.
Thanks for all the help with polishing the patch up to a good state.
Paul
On Thu, 7 Aug 2025 at 22:02, Paul Richard Thomas <
paul.richard.tho...@gmail.com> wrote:
> Hi Jerry et al.,
>
> I attached an earlier version of the patch - apologies!
>
> Please find the latest attached.
>
> Paul
>
>
> On Thu, 7 Aug 2025 at 17:07, Paul Richard Thomas <
> paul.richard.tho...@gmail.com> wrote:
>
>> Hi Jerry, Steve and co.,
>>
>> Thanks for the reviews.
>>
>> I have made extensive changes to this patch. Relative to the original
>> submission, the changes are:
>> (i) Cleaned up the comments in decl.cc. 'binding' removed from generic
>> statement comments;
>> (ii) 'generic_spec_name' used for formatted name instead of binding_name;
>> (iii) Handling of access specification corrected and tidied;
>> (iv) Warnings turned into errors, where appropriate;
>> (v) In generic_stmt_1.f90 have added checks that specific functions are
>> correctly identified;
>> (vi) Also in generic_stmt_1.f90 have added DTIO and intrinsic operator
>> tests;
>> (vii) In generic_stmt_2.f90 have changed error messages as appropriate
>> and checks on user defined and intrinsic operator errors; and
>> (viii) Have added generic_stmt_3.f90, which incorporates Steve's tests
>> and erroneous typing and name clashes with other entities.
>> (ix) Also, the problem with the public interface for 'bak' not exposing
>> the specific interfaces in 'snooze' has been resolved and is tested in
>> generic_stmt_4.f90.
>>
>> Regtests OK. The patch adds 46 expected passes.
>>
>> Good for mainline?
>>
>> Paul
>>
>>
>> On Mon, 4 Aug 2025 at 18:20, Jerry D <jvdelis...@gmail.com> wrote:
>>
>>> With your updated patch addressing Steve's comments OK.
>>>
>>> We have time for minor tweaks if necessary.
>>>
>>> On 8/3/25 11:06 AM, Steve Kargl wrote:
>>> > On Sun, Aug 03, 2025 at 12:20:24PM +0100, Paul Richard Thomas wrote:
>>> >
>>> > First, the easy one:
>>> >
>>> >> + /* Match the binding name; depending on type (operator / generic)
>>> format
>>> >> + it for future error messages into bind_name. */
>>> >
>>> > This comment looks muddled.
>>> >
>>> >> + if (op_type == INTERFACE_GENERIC &&
>>> !generic_spec->attr.generic)
>>> >> + {
>>> >> + gfc_error ("There's already a non-generic procedure with "
>>> >> + "binding name %qs at %C", generic_spec->name);
>>> >> + goto error;
>>> >
>>> > The contraction should probably be expanded to make
>>> > translation to other languages easier. I also find
>>> > the error message to be unclear for
>>> >
>>> > module foo
>>> > implicit none
>>> > private
>>> >
>>> > ! F2023:C815 An entity shall not be explicitly given any
>>> > ! attribute more than once in a scoping unit.
>>> > public bar
>>> > generic, public :: bar => bah, bak
>>> >
>>> > contains
>>> > integer function bah(i)
>>> > integer, intent(in) :: i
>>> > bah = i
>>> > end function bah
>>> > real function bak(x)
>>> > real, intent(in) :: x
>>> > bak = x
>>> > end function bak
>>> > end module foo
>>> >
>>> > % gfcx -c -fmax-errors=1 generic_stmt_3.F90
>>> > generic_stmt_3.F90:7:32:
>>> >
>>> > 7 | generic, public :: bar => bah, bak
>>> > | 1
>>> > Error: There's already a non-generic procedure with binding name 'bar'
>>> at (1)
>>> >
>>> > If you invert the statements to
>>> >
>>> > generic, public :: bar => bah, bak
>>> > public bar
>>> >
>>> > % gfcx -c -fmax-errors=1 generic_stmt_3.F90
>>> > generic_stmt_3.F90:7:16:
>>> >
>>> > 7 | public bar
>>> > | 1
>>> > Error: ACCESS specification at (1) was already specified
>>> >
>>> > The error is quite clear. This also generates an error
>>> >
>>> > public bar
>>> > generic :: bar => bah, bak
>>> >
>>> > while
>>> >
>>> > generic :: bar => bah, bak
>>> > public bar
>>> >
>>> > compiles without error.
>>> >
>>>
>>>
