Re: coercion to strings and numbers (was: recursively traverse...)

[Rici Lake <lua@...>]

On 29-Jan-07, at 8:12 AM, Doug Rogers wrote:

> Rici Lake wrote:
>
> > Mark Hamburg wrote:
> > > I have to say that the automatic coercion between strings and numbers
> > > in the Lua API is one of my least favorite features. ...
> > > Furthermore, it would be nice if luaL_tostring was essentially the
> > > same as invoking tostring ... so that it could handle arbitrary 
> > > objects.
>
> > I agree, wholeheartedly.
>
> Sorry for the late entry, but I just wanted to cast my vote along with
> Mark and Rici.
>
> Perhaps there are some subtle side effects I'm missing. I have enjoyed
> the benefits of the coercion when I'm typing at the interpreter, but
> often that amounts to simple things like:
>
> > print('n = ' .. n)
>
> I would not mind having to type:
>
> > print('n = ' .. tostring(n))
>
> Wouldn't the __concat metamethod for strings handle that, though?

.. on strings is a primitive; there is no __concat metamethod for 
strings. (And Lua takes advantage of that to implement multiple .. 
efficiently, so that  a .. b .. c is a single operation if a, b and c 
are all strings, avoiding the creation of a temporary string b .. c)

However, one of the annoyances of automatic coercion, which has been 
mentioned a couple of times, is that if you attempt to implement a 
__concat metamethod you will discover that numbers are coerced to 
strings before your metamethod is called.

The __tostring metamethod is not used by the Lua VM or the API; it is 
implemented by the tostring() base library function (and consequently 
used by print(), which calls tostring(), but not by io.write().) Mark's 
suggestion of adding a luaL_tostring() function which invokes the 
__tostring metamethod might encourage more consistent behaviour.