[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#60397: 29.0.60; c++-ts-mode could report better defun names
From: |
Knut Anders Hatlen |
Subject: |
bug#60397: 29.0.60; c++-ts-mode could report better defun names |
Date: |
Sun, 08 Jan 2023 08:05:30 +0100 |
User-agent: |
Gnus/5.13 (Gnus v5.13) |
Yuan Fu <casouri@gmail.com> writes:
> Knut Anders Hatlen <kahatlen@gmail.com> writes:
>
>> The defun names reported by c++-ts-mode could still need a couple of
>> improvements:
>>
>> 1) In a buffer with c++-ts-mode and which-function-mode enabled, and
>> this content:
>>
>> struct S {
>> int f1(int x) {
>> return x + 1;
>> }
>> int g1(int x);
>> };
>>
>> int S::g1(int x) {
>> return x + 1;
>> }
>>
>> Inside the inline f1 function definition, which-function-mode shows
>> "S.f1". But inside the out-of-line g1 function definition, it shows
>> "n/a" instead of "S.g1". (Not limited to structs. Classes have the same
>> problem.)
>
> Now the second function is displayed as S::g1.
Looks good now. Classes seem to be handled fine too.
>> 2) Namespaces are not handled. Given this content:
>>
>> namespace n {
>> int f1(int x) {
>> return x + 1;
>> }
>> }
>>
>> namespace {
>> int f2(int x) {
>> return x + 1;
>> }
>> }
>>
>> Inside the f1 and f2 function bodies, which-function-mode shows "f1" and
>> "f2", respectively. It would be better if it showed "n.f1" for the
>> former, and perhaps something like "(anonymous).f2" for the latter.
>
> Now the first function is shown as n.f1, the second is shown as f2.
> Making it (anonymous).f2 isn’t necessarily better than f2 IMO, and
> requires some non-trivial change to the current code, so I didn’t do it.
Fair enough. Thanks!
--
Knut Anders