We typically have options for these, going from the String#to_i args.

@RX14 I know but decided against it, because String#to_*methods are implemented directly within String, and Big* classes are re-opening String with their own #to_big_* impl - which means that additional options needs to be passed along to appropriate constructors, and none of them are implemented ATM so I'd suggest leaving it as an enhancement for a future PR.

You do implement the skip, using strip, so you can implement the options to allow whitespace or not.

If we are relaxing the treating of whitespaces we should probably allow them between the sign and the number itself. Please ensure the same rules applies for non-big numbers. If custom logic is used to ignore whitspaces we should ensure (adding specs) that some other whitspace than 0x20 is treated as whitespace.

Plus, all these string copies are expensive.

I see a valid use case for parsing numbers with stripping whitespace.

• But the implementation needs to be configurable. I'd be fine with making that behaviour default, but there should be an option to control it.
• The current implementation is not very performant because it creates string copies. Limiting the char array passed to the C function would avoid additional allocations.
• Non-Big number types should also be covered.
• In the current state, it's better to just call strip explicitly when you need this behaviour.

I propose to close this PR. If someone comes up with a new proposal, I'd be happy to accept it. But it's not a pressing issue, since the alternative is trivial.