[cfe-commits] [PATCH] Remove 'PrintInstantiation' feature from DeclPrinter

[Philip Craig]

On 8 January 2013 08:36, Richard Trieu <rtrieu at google.com> wrote:
> On Mon, Jan 7, 2013 at 1:44 PM, Dmitri Gribenko <gribozavr at gmail.com> wrote:
>>
>> On Mon, Jan 7, 2013 at 11:38 PM, jahanian <fjahanian at apple.com> wrote:
>> >
>> > On Jan 7, 2013, at 1:06 PM, Dmitri Gribenko <gribozavr at gmail.com> wrote:
>> >
>> >> On Mon, Jan 7, 2013 at 11:05 PM, Richard Trieu <rtrieu at google.com>
>> >> wrote:
>> >>> I don't understand your reasons for removing this code.  It was
>> >>> specifically
>> >>> added so that template instantiations would be outputted with
>> >>> -ast-dump.
>> >>> Why do you consider printing instantiations misleading?
>> >>
>> >> It prints something that is not written in the source code.
>> >
>> > I think you intend to remove it from -ast-print not -ast-dump. But since
>> > it is under an optional print policy
>> > why do we want to remove this capability entirely?
>
>
> That's a good point, I only care about -ast-dump while Dmitri was talking
> about -ast-print.
>>
>>
>> Thanks Fariborz!  I read Richard's email incorrectly: I want to remove
>> it from -ast-print.
>>
>> My motivation is that I expect -ast-print to print what I wrote in the
>> source code.
>>
>> Richard: if this is still useful for you, I will not push this further.
>>
> So, there's been a little confusion about two similarly named flags,
> -ast-dump and -ast-print.  I only care about -ast-dump while you were
> talking about -ast-print.  It also doesn't help that the test case
> Misc/ast-dump-templates.cpp was changed to be about use -ast-print instead
> of -ast-dump.

That change is my fault. I changed it so that DeclPrinter.cpp is only
used by -ast-print. I should have renamed the file too though.

> For -ast-dump, I would like all instantiations to be displayed.

I think the tests I added in Misc/ast-dump-decl.cpp are sufficient for
testing that all instantiations are displayed.

>
> For -ast-print, I don't have a strong preference either way.
>
> _______________________________________________
> cfe-commits mailing list
> cfe-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
>