Hello Robert,
> [...] With your patch, you get different behavior depending on exactly
> how the input is malformed.
I understand that you require only one possible error message on malformed
input, instead of failing when converting to double if the input looked
like a double (there was a '.' clue int it, but that is not proof), or
something else if it was assumed to be an int.
So I'm going to assume that you do not like the type guessing.
> I'm not going to commit it this way, and frankly, neither is anyone
> else.
Hmmm. Probably my unconcious self is trying to reach 42.
Here is a v36 which inspect very carefully the string to decide whether
it is an int or a double. You may, or may not, find it to your taste, I
can't say.
--
Fabien.