[cfe-dev] Tidy vs Warnings: missing override

[Matthieu Brucher via cfe-dev]

I think that the override is not an after-the-bug tool, but really a
proactive checker. It doesn't tell you of a bug, it prevents you for
introducing one.
On an old code base, that's where I find it the most helpful, when you have
to work on it. Without the override, you don't have the certainty of
properly updating the APIs (old code base also often miss proper test
coverage). With override, before your code changes, you can ensure that you
are refactoring the full API.

Cheers,

Matthieu

Le lun. 13 août 2018 à 01:59, David Blaikie <dblaikie at gmail.com> a écrit :

> When I say a "false positive" in this context it's not a case where a
> warning would suggest override where it would be invalid/incorrect to put
> ovterride - but that it would suggest the 'override' keyword when it would
> be benign/when it isn't telling the user there's a bug/mistake here.
>
> On an old codebase, likely /many/ more missing overrides are there because
> there was no override keyword when the code was written - most of those
> missing overrides aren't buggy code. So this warning would cause a lot of
> cleanup for relatively little bug finding.
>
> - Dave
>
> On Fri, Aug 10, 2018 at 8:36 AM Matthieu Brucher via cfe-dev <
> cfe-dev at lists.llvm.org> wrote:
>
>> Yes, I have to write an example case!
>> I think the odds of an error with a wrong override are very low. Ont he
>> contrary, I noticed several times errors with missing overrides and APIs
>> moving and wrong arguments.
>>
>> Cheers
>>
>> Le ven. 10 août 2018 à 15:41, Jonas Toth <development at jonas-toth.eu> a
>> écrit :
>>
>>> That would be worth a bug-report :)
>>>
>>> Am 10.08.2018 um 15:03 schrieb Matthieu Brucher:
>>>
>>> Hi,
>>>
>>> I have noticed the missing override warning when using virtual diamond
>>> inheritance. In that case, there was no warning generated.
>>>
>>> Regards,
>>>
>>> Matthieu
>>>
>>> Le ven. 10 août 2018 à 10:23, Jonas Toth via cfe-dev <
>>> cfe-dev at lists.llvm.org> a écrit :
>>>
>>>> Hi,
>>>>
>>>> I experience the override-warning when it was inconsistent within the
>>>> class (meaning it has been used at least once and some places missed it).
>>>> This measure is probably to not warn on old code bases.
>>>>
>>>> Could this be the case?
>>>>
>>>> Best, Jonas
>>>>
>>>> Am 09.08.2018 um 19:26 schrieb JVApen via cfe-dev:
>>>>
>>>> Hello all,
>>>>
>>>> Today we have discovered a case where we did not have a compiler
>>>> warning for a missing override. To verify, we explicitly enabled the
>>>> warning with the pragma.
>>>>
>>>> While investigating this in the IDE, we noticed that the Visual Assist
>>>> plugin did notice that missing override. As this uses clang-tidy to
>>>> determine this, we were wondering if these differences we noticed are
>>>> intentional. If so, is their any documentation on these differences?
>>>>
>>>> Tnx
>>>>
>>>>
>>>> _______________________________________________
>>>> cfe-dev mailing listcfe-dev at lists.llvm.orghttp://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>>>
>>>>
>>>> _______________________________________________
>>>> cfe-dev mailing list
>>>> cfe-dev at lists.llvm.org
>>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>>>
>>>
>>>
>>> --
>>> Quantitative analyst, Ph.D.
>>> Blog: http://blog.audio-tk.com/
>>> LinkedIn: http://www.linkedin.com/in/matthieubrucher
>>>
>>>
>>>
>>
>> --
>> Quantitative analyst, Ph.D.
>> Blog: http://blog.audio-tk.com/
>> LinkedIn: http://www.linkedin.com/in/matthieubrucher
>> _______________________________________________
>> cfe-dev mailing list
>> cfe-dev at lists.llvm.org
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>
>

-- 
Quantitative analyst, Ph.D.
Blog: http://blog.audio-tk.com/
LinkedIn: http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-dev/attachments/20180813/6d4b9987/attachment.html>