I'm opposed to changing int so that int('123_456') ignores the _ as that will change the behavior of existing code and could break apps.
On 06/05/2011 23:51, Cameron Simpson wrote:As far as I remember, Ada also permits it, but has the rule that it can
On 06May2011 15:40, Ethan Furman<ethan@stoneleaf.us> wrote:
| Bruce Leban wrote:
|>Is _ just ignored in numbers or are there more complex rules?
|>
|> * 1_2345_6789 (can I use groups of other sizes instead?)
|> * 1_2_3_4_5 (ditto)
|> * 1_234_6789 (do all the groups need to be the same size?)
|> * 1_ (must the _ only be in between 2 digits?)
|> * 1__234 (what about multiple _s?)
|> * 9.876_543_210 (can it be used to the right of the decimal point?)
|> * 0xFEFF_0042 (can it be used in hex, octal or binary numbers?)
|> * int('123_456') (do other functions accept this syntax too?)
|
| I would say it's ignored. Have the rule be something like
| number_string.replace('_','').
|
| The only wrinkle is that currently '_1' is usable name, and that
| should probably be disallowed if the above change took place.
|
| I'm +1 on the idea.
Personally I'm be for ignoring the _ also, save that I would forbid it
at the start or end, so no _1 or 1_.
And I would permit it in hex code etc.
I'm +0.5, myself.
occur only between digits. If we follow that, then:
1_2345_6789 => Yes
1_2_3_4_5 => Yes
1_234_6789 => Yes
1_ => No
_1 => No
1__234 => No
9.876_543_210 => Yes
9._876_543_210 => No
9_.876_543_210 => No
0xFEFF_0042 => Yes
int('123_456') => Yes
_______________________________________________
Python-ideas mailing list
Python-ideas@python.org
http://mail.python.org/mailman/listinfo/python-ideas