[cfe-commits] [PATCH] Implements support to run standalone tools

[Manuel Klimek]

On Wed, Apr 4, 2012 at 1:14 AM, Douglas Gregor <dgregor at apple.com> wrote:
> Sent from my iPhone
>
> On Mar 30, 2012, at 2:20 AM, Manuel Klimek <klimek at google.com> wrote:
>
>> On Thu, Mar 29, 2012 at 5:25 PM, Marshall Clow <mclow.lists at gmail.com> wrote:
>>> 2) runSyntaxOnlyToolOnCode
>>> I don't understand why "input.cc" is hard wired here.
>>> Is this just some dummy file name - supposed to be unique?
>>> Maybe a comment to say what's going on?
>>
>> Done.
>>
>> Also, Chandler pointed out that he'd like to see this function renamed
>> to runToolOnCode, as users of the tooling would want syntax-only runs
>> to be the default (and not think about it unless they start wondering
>> how to run code generation as part of a map reduction).
>
> I think the name runToolOnCode better, too. Also, how about making the file name a parameter? That way, users can simply provide a different file name if it conflicts or if they want an easy way to say it's code that isn't C++ (e.g, by naming it input.c).

Changed, and done. Chandler also pointed out that he'd like us to use
a Twine there, so it's more convenient to use in tests. Let me know if
you're unhappy about that.

Also updated the api docs for the mapVirtualFile calls, to document
the necessity of zero termination.

Checked it in. Let me know if there's anything urgent that needs
fixing, otherwise the next step is writing some documentation.

Cheers,
/Manuel