Re: Compiling Reverse Polish Calculator Example

[Hans Aberg]

On 31 Oct 2012, at 09:32, Akim Demaille wrote:

> Le 29 oct. 2012 à 18:22, Akim Demaille a écrit :
> 
>> [Paul, this is about quotearg]
>> 
>> Hi Marc,
>> 
>> This message is not related directly to your complaint
>> (Hans explained well your issue).  This message is about
>> Bison not issuing the right message.
>> 
>> Le 26 oct. 2012 à 23:13, Marc Mendiola a écrit :
>> 
>>> rpcalc.y:1.1: invalid character: `\377'
>>> rpcalc.y:1.2: invalid character: `\376'
>> 
>> This is good.
>> 
>>> rpcalc.y:1.4: invalid character: `'
>>> rpcalc.y:1.5: invalid character: `'
>>> rpcalc.y:1.6: invalid character: `'
>>> rpcalc.y:1.7: invalid character: `'
>>> rpcalc.y:1.8: invalid character: `'
>> 
>> These last five are wrong: they should read
>> 
>> invalid character: `\0'
> 
> I now also realize that maybe Bison should ignore these
> BOM bytes.  Or even propagate them to the generated files?
> I see for instance GCC has been supporting them for four years
> (http://gcc.gnu.org/viewcvs?view=revision&revision=134507),
> and I guess more and more tools are supporting them.

It is pointless in UTF-8, and accepting it encourages a number of other 
problems.
  https://en.wikipedia.org/wiki/Byte_order_mark

Hans