Re: GOP-PROP 3: C++ formatting

[Graham Percival]

On Mon, Jun 27, 2011 at 08:46:17AM -0700, Keith OHara wrote:
> In defense of astyle, I should have written astyle -Udp
>  --unpad-paren
>  --pad-paren-out
>  --pad-oper

Huh.

> and if a human puts a space in "context ()" astyle -Udp leaves it in place.

?!  hmm, looks promising.


I've added my fix-astyle-fiddle.py (and fix-astyle-all.sh) scripts
to dev/gperciva-astyle.  When I look at
  git diff origin/dev/gperciva-fixcc origin/dev/gperciva-astyle
I'm still leery about some changes.

Don't get me wrong -- I really want us to use astyle.  Begin tied
to a specific version of a text editor is silly, and is bound to
cause more pain and confusion down the road.  But at the moment,
astyle+postprocessing still appears to produce inferior code to
emacs+postprocessing (fixcc.py).

One possibility that I still haven't tried is to use fixcc.py
exactly, except replace the emacs call with an astyle call.  I'm
not certain how many of the "nice" changes that fixcc.py does are
the result of emacs, and how many are due to the complicated regex
in the script itself.

Cheers,
- Graham