Re: C++ cleanup : get rid of some compilation warnings (issue 353880043 by address@hidden)

[dak]

https://codereview.appspot.com/353880043/diff/2/lily/beam-quanting.cc
File lily/beam-quanting.cc (right):

https://codereview.appspot.com/353880043/diff/2/lily/beam-quanting.cc#newcode1039
lily/beam-quanting.cc:1039: string card = best->score_card_ + to_string
(" c%d/%lu", completed, configs.size ());
I think that %lu is not guaranteed platform-independently to correspond
to size_t (or whatever we have here).  There may be some special letter
for it, or maybe it is only a GNU extension.

[Initial thought, later revised:] Maybe bite the bullet and use
string card = (std::stringstream (best->score_card_) << " c" <<
completed << "/" << configs.size ()).str ();
instead?  It's not what we use elsewhere (yet) but it should be
comparatively future-proof and is what C++ is supposed to use.

It would need #include <sstream> to work.
[Revision: use %zu, see later for rationale]

https://codereview.appspot.com/353880043/diff/2/lily/main.cc
File lily/main.cc (right):

https://codereview.appspot.com/353880043/diff/2/lily/main.cc#newcode635
lily/main.cc:635: { }; // ignore option for backwards compatibility
This code is insane.  Maybe just outcomment the whole else if branch
instead?

https://codereview.appspot.com/353880043/diff/2/lily/text-interface.cc
File lily/text-interface.cc (right):

https://codereview.appspot.com/353880043/diff/2/lily/text-interface.cc#newcode176
lily/text-interface.cc:176: non_fatal_error (_f ("Markup depth exceeds
maximal value of %lu; "
Ugh.  stringstream does not really work sensibly with _f, does it?
Scratch the previous idea.  Wouldn't this then need %zu instead?
Assuming that C++ is following C99 in that regard?

https://codereview.appspot.com/353880043/