[PATCH] clang-modernize: Add SubprocessCompilationDatabase

[Manuel Klimek]

On Fri, Oct 11, 2013 at 9:19 AM, Kim Gräsman <kim.grasman at gmail.com> wrote:

> I totally misread the test lines, you're right, this is something else.
>
> I'm still keeping an eye on this thread, though, because I think there
> are synergies between the current FixedCompilationDatabase, what you
> want to do and a compilation database built from a build-system
> invocation, and that the current FixedCompilationDatabase could
> probably be made to support all three (not that I understand why
> multiprocessing requires a separate compilation db, anxious to hear
> your response to Manuel's question.)
>

I actually think those are different issues. Your issue is to drop-in run a
tool from a build system - by now I think you have convinced me enough that
this is a real need, that I'd vote for accepting patches to that end.
Here, it looks to me like the calling process has full control over how it
calls the subprocess, thus we have much more flexibility in what we require
for command line arguments...


>
> Cheers,
> - Kim
>
> On Fri, Oct 11, 2013 at 3:56 PM, Edwin Vane <revane at gmail.com> wrote:
> > I'm not sure if this is a step toward tools acting as CXX replacements
> but I
> > renamed the class to something a little more describe of what it does as
> > opposed to what it's used for.
> >
> >
> > On Fri, Oct 11, 2013 at 3:13 AM, Kim Gräsman <kim.grasman at gmail.com>
> wrote:
> >>
> >> Hi Edwin,
> >>
> >> On Fri, Oct 11, 2013 at 12:34 AM, Edwin Vane <edwin.vane at intel.com>
> wrote:
> >> >
> >> >   The problem is the FixedCompilationDatabase mangles the arguments
> you
> >> > pass. FixedCompilationDatabase as a result has some preconditions:
> the name
> >> > of the compiler should not be present nor should be the filename to
> operate
> >> > on. Check CompilationDatabase.h for a better description. As stated
> in the
> >> > patch summary we need a simpler compilation database that simply
> accepts a
> >> > full list of args and then provides those same args back upon request.
> >>
> >> This sounds nice, but the name feels a little funny... Isn't it just a
> >> LiteralCompilationDatabase? The subprocess forwarding is just one use
> >> case, unless I'm missing something. I'm still interested in allowing
> >> tools to act as CXX replacements in build systems, for example, and it
> >> feels like this is as step on the way there?
> >>
> >> - K
> >>
> >> _______________________________________________
> >> cfe-commits mailing list
> >> cfe-commits at cs.uiuc.edu
> >> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
> >
> >
> >
> >
> > --
> > Edwin V
>
> _______________________________________________
> cfe-commits mailing list
> cfe-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20131011/525edbb6/attachment.html>