#295 Number literals: let leading zeroes affect data type

[dkl]

On 32bit, number literals default to being 32bit Integers, even if they have leading zeroes that raise the number of digits to more than what a 32bit integer allows. For example:

print &h00000000FFFFFFFF

It could be useful to automatically make the number literal a LongInt in that case.
This is probably most useful for binary and hex number literals. Should it also apply to octal and decimal ones?

On the other hand: C/C++ ignores leading zeroes too, and to be safe FB users may use type suffixes already anyways.

[dkl]

Some discussions about this:
http://www.freebasic.net/forum/viewtopic.php?p=143015#p143015
http://www.freebasic.net/forum/viewtopic.php?p=206783#p206783