Re: Numeric literals, take 3

On 11/27/2002 7:54 PM, Angel Faus wrote:
> For example, the integer 30 can be written in base 16
> in two equivalent ways:
> 
>    my $x = 16#1D;
>    my $x = 16#1:14;
> 
> These two representations are incompatible, so writing
> something like C<16#D:13> will generate a compile-time
> error.
Ambiguity:
Is C<my $x = 16#14;> equivlent to the digit "14", base-sixteen, or is it 
equivlent to 0x14?  (This has been noted before, but I don't think 
anybody has decided.  My vote is 0x14, you should say my $x=16#0:14 if 
you want the other meaning.)

> =section *** Bin/Hex/Oct shorthands
> 
> Since writing in binary, hexadecimal and octal is a 
> moderately common operation, Perl offers a shorthand
> for them:
> 
>  0b0110   # bin
>  0c0123   # oct
>  0x00ff   # hex
>  0x00fF   # hex, == 0x00ff
>  0x00FF   # hex, == 0x00ff
Loose specification:
You should probably note that while the digits are case-insensitive, the 
shorthand-radix-specifier is case-sensitive -- IE 0B0110 isn't 2:110, 
it's a syntax error (invalid digit in decimal literal).

Possible useful note:
Should we note that the C and perl5 notation of 0777 being octal is 
/not/ supported?

> This won't work for bases greater than 36, so we
> have too:
Grammar: I think this should be "so we also have:", or possibly "so we 
also have...".

> In boolean context (see "Boolean Context") C<NaN> always
> evaluates to false so.
Missing end-of-setance:  "to false so those last two examples are 
false".  (IE both !(NaN!=NaN) and !(NaN==NaN).)

Useful node:
We should probably note that Inf and NaN are case sensitive (or are they)?

> NOTE: are we going to have +-Inf too?
Speculation: I don't think so.  I don't think we're going to have -0 or 
+0 notations either (for underflow).  Any parroters?

> If you need to create typed C<Int> numbers without risking 
> to lose precision, you should write them as a string literals:
I think Larry has mandated the oppisite -- that we Do The Right Thing 
with bigints automaticly -- IE that BigInt is a subclass of Int, with 
full automatic promotion.  (I think this isn't going to be true of 
little-i int, though.)

	-=- James Mastros