[PATCH] Enhance PPCallbacks::Defined callback to provide SourceRange

[Argyrios Kyrtzidis]

On Jul 18, 2013, at 3:55 PM, "Thompson, John" <John_Thompson at playstation.sony.com> wrote:

> Thanks for looking at it, Argyrios.
>  
> > Why do you need a SourceRange, can't you just use the location of the identifier from the token parameter ?
>  
> I thought it would be cleaner to have the preprocessor pass it along, rather than crawling the source or approximating it.  But if avoiding changes to the preprocessor is preferred, I’ll go that route, since I don’t need the exact range, just something unique.

The source range would be the range of the parens right ? Not sure why you'd need it for modularize tool (I thought the location of the identifier is the important one) but in general I'm buying the "consistency with other callbacks" argument so LGTM.

It would be great if you could also add some unit tests for the recent changes in unittests/Lex/PPCallbacksTest.cpp but this can be done post-commit.

>  
> -John
>  
> From: Argyrios Kyrtzidis [mailto:akyrtzi at gmail.com] 
> Sent: Thursday, July 18, 2013 3:16 PM
> To: Thompson, John
> Cc: cfe-commits@; Sean Silva
> Subject: Re: [PATCH] Enhance PPCallbacks::Defined callback to provide SourceRange
>  
> On Jul 18, 2013, at 1:09 PM, Thompson, John <John_Thompson at playstation.sony.com> wrote:
> 
> 
> In modularize, it turns out I need the SourceRange for the “defined()” expression. 
>  
> Why do you need a SourceRange, can't you just use the location of the identifier from the token parameter ?
>  
> 
> 
> The enclosed patch addresses this, making it more like the MacroExpands callback.
>  
> MacroExpands needs a source range due to function macro expansions.
>  
>  
> Thanks.
>  
> -John
>  
> <ppcallbacks2.patch>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20130719/a7c5e328/attachment.html>