Re: [PATCH v2 31/31] Support an "unlimited" number of user-defined arguments

[Philipp Rudo <prudo@linux.vnet.ibm.com>]

On Wed, 19 Oct 2016 13:47:22 +0100
Pedro Alves <palves@redhat.com> wrote:

> On 10/19/2016 12:33 PM, Philipp Rudo wrote:
> > Pedro Alves <palves@redhat.com> wrote:
> 
> >> I also needed a way to convert a number to a std::string, so I added a
> >> new utility for that, gdb::to_string.  Yet another thing that can go
> >> away with C++11.
> > [... snip ...]
> >> +/* Returns a string representation of VAL.  Replacement for
> >> +   std::to_string, which is only available in C++11 or later.  */
> >> +
> >> +namespace gdb {
> >> +
> >> +template <class T>
> >> +inline std::string
> >> +to_string (const T &val)
> >> +{
> >> +  std::stringstream ss;
> >> +
> >> +  ss << val;
> >> +  return ss.str ();
> >> +}
> >> +
> >> +}
> >> +
> >>  /* Make a copy of the string at PTR with LEN characters
> >>     (and add a null character at the end in the copy).
> >>     Uses malloc to get the space.  Returns the address of the copy.  */
> > 
> > Is this really necessary?
> > As far as I understood the discussion, we jump directly to C++11. Thus there is no need for an homemade to_string.
> 
> We haven't actually officially committed to requiring
> C++11, so I didn't want to be blocked by that.  It's
> trivial to remove that bit latter if this lands first.

True that C++11 is not official yet. But the "trivial to remove" is
exactly the reason I wrote, as those pieces of code tend to stay in the longest.
And in 10+ years nobody knows why there is such a messy mix of different
functions doing the same thing and why it hasn't been done right in the
first place ;)

So my mail should not push you to use std::to_string but be seen as a reminder,
that if we require C++11 your home made function should be deleted better sooner
than later.

Phil