Re: Why duplicate keys in table constructor isn't an error?

["Leigh McRae" <leigh.mcrae@...>]

I think he means specifically in the constructor and the compiler. I could 
see a small case for this but it would have to be a switch. I am guessing 
that there is code out there that uses this 'feature'. Generated code 
specifically.

Leigh

----- Original Message ----- 
From: "Hakki Dogusan" <dogusanh@tr.net>
To: "Lua list" <lua@bazar2.conectiva.com.br>
Sent: Tuesday, March 13, 2007 10:30 AM
Subject: Re: Why duplicate keys in table constructor isn't an error?


> Hi,
>
> Ivko Stanilov yazmış:
> > I've often made this mistake in my programs when copying/pasting table
> > elements, or just by oversight:
> >
> >  local table = { key = "value1", ..., key = "value2" }
> >
> > Shouldn't the compiler be able to detect this case and report a syntax
> > error? I know that you could possibly have expressions with side
> > effects in the place of "value1", but nevertheless I can't envision
> > a scenario in which having the same constants keys in the same table
> > constructor is intentional. This is a 100% error in the code and I
> > think it should be reported as such.
> >
> > Regards,
> > Ivko
>
>
> If you think it like following it makes sense, IMHO:
>
> local table = {}
> table.key = "value1"
> ...
> ...
> table.key = "value2"
>
>
>
> --
> Regards,
> Hakki Dogusan