[cfe-dev] Getting fully qualified names of random qualtypes.

[Vassil Vassilev via cfe-dev]

On 26/10/15 09:34, Sterling Augustine wrote:
> On Sat, Oct 24, 2015 at 10:40 AM, Vassil Vassilev <vvasilev at cern.ch> wrote:
>> On 23/10/15 21:06, Sterling Augustine wrote:
>>> Also, I want to be sure the Cling project is OK with the way it is
>>> attributed. Happy to work with you all however you like there.
>>>
>> I am more than happy to attribute this part of the code. I'd like as many
>> parts of cling as possible to go upstream.
>>
>> Could you add Phillipe as a co-author? Eg.
>>
>> +// author:  Vassil Vassilev<vasil.georgiev.vasilev at cern.ch>
>> +// author:  Philippe Canal <pcanal at fnal.gov>
>>
>> Out of curiosity: what do you need this for?
> I'm working on a tool that examines a header file and describes
> exactly how one would call the functions it declares--detects the
> namespaces and classes they are in. As part of that, it needs to
> determine how to declare the types of the parameters they take.
Thanks for the info, sounds cool ;)
>
> Apparently Clang doesn't allow file-level authorship comments, so I'll
> remove that detail if it is OK with you. If it isn't, I'll just
> incorporate cling/AST.[cpp|h]  directly into my own project--that will
> be quicker than the reformatting and refactoring rabbit hole anyway.
The authorship comments could be transformed as Philippe suggested.

I think we should try to aim for an "official" patch review. IMO 
investing some extra time in refactoring and reformatting to put the 
work mainline will definitely pay off in long term (if you plan your 
tool to be around for a long time). Maybe Philippe and I could help...