On Fri, 2018-02-16 at 07:14 -0500, Trevor Saunders wrote:
> On Thu, Feb 15, 2018 at 04:04:45PM -0500, David Malcolm wrote:
> > On Fri, 2018-02-09 at 13:01 +0000, Nick Clifton wrote:
> > > +class escaped_string
> > > +{
> > > + public:
> > > + escaped_string () { m_owned = false; m_str = NULL; };
> > > + ~escaped_string () { if (m_owned) free (m_str); }
> > > + operator const char *() const { return (const char *) m_str; }
> > > + void escape (const char *);
> > > + private:
> > > + char * m_str;
> > > + bool m_owned;
> > > +};
> >
> > I'd hoped that instead of this we could have an escape_string
> > function
> > return an instance of a class that has responsibility for the "if
> > (ownership) free" dtor, but I don't think C++89 supports that (I
> > had a
> > go at implementing it, but I think we'd need C++11's move
> > semantics,
> > rather than just the NRVO).
>
> I think either gcc::unique_ptr does what you want here,
That would involve a second heap allocation; as-is, the object lives on
the stack.
> or you could
> use the same trick of having a copy ctor that takes a non constant
> reference to implement this in C++98.
Maybe. I was trying not to bash Nick over the head too much with C++
here :) If you want to implement it, go for it (I can think of
possible uses for it in the driver, for fixing leaks there when used by
libgccjit).
> Thanks!
>
> Trev
