In principle, I like the patch. However, I think one should
replace
gfc_error ("Attribute at %L is not allowed in a %s definition",
…, state_name
by something like:
bool is_type = gfc_current_state () == COMP_DERIVED;
gfc_error (is_type ? G_("Attribute at %L is not allowed in a TYPE definition")
: G_("Attribute at %L is not allowed in a STRUCTURE
definition"),
…
Reason: (a) This makes translation simpler; e.g. 'structure' and 'type' have
different gender in several European languages. (Albeit in this case the
gender of 'definition' dominates in the cases I checked.)
(b) For TYPE, the string won't change such that the existing translations
still work – even if the update for STRUCTURE won't make it for the release.
Otherwise it looks good to me, including the test case in your follow-up email.
Cheers,
Tobias
On 4/1/20 7:19 PM, Fritz Reese via Fortran wrote:
This simple patch was submitted some time ago (over 1 year), but got
lost without review. I have lately rebased and tested, and the patch
is still good. Is this OK to commit to trunk and for backport? I'd
like to port as far back as 7.
---
Fritz Reese
gcc/ChangeLog:
2020-04-01 Fritz Reese <fore...@gcc.gnu.org>
PR fortran/85982
* fortran/decl.c (match_attr_spec): Lump COMP_STRUCTURE/COMP_MAP into
attribute checking used by TYPE.
gcc/testsuite/ChangeLog:
2020-04-01 Fritz Reese <fore...@gcc.gnu.org>
PR fortran/85982
* gfortran.dg/dec_structure_28.f90: New test.
-----------------
Mentor Graphics (Deutschland) GmbH, Arnulfstraße 201, 80634 München / Germany
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Alexander
Walter
