IndexOutOfRangeException when decoding string with chars not in alphabet about hashids.net HOT 7 CLOSED

@manigandham Hi, I'll look into it although I think returning empty array seems most appropriate given we already do it for other errors.

from hashids.net.

Hi,
Yes, that's obviously not going to work, as you said, Z are not in the alphabet. The question is whether it should throw a more descriptive error or not. I guess that could be good at the cost of needing to enumerate the alphabet and the hash every time to make sure that all characters in the hash exist in the alphabet or not before trying to code.

checking for < 0 might be a solution. But throwing an "Hash Included character not in alphabet"-exception when that happens isn't necessarily true, so don't like that either.

I'll ask the authors of the other port to see if we should have a way of handling this 😄

from hashids.net.

... or maybe it shouldn't throw at all but just return an empty array...

from hashids.net.

Yes, I was not clear, I meant I thought it should return an empty array when the input chars are not in the alphabet! (or any decoded numbers are negative).

from hashids.net.

Personally I think it should throw an error and let the calling application handle that exception. That seems like a better design than arbitrarily returning an empty array.

from hashids.net.

@casserlyprogramming Well, it's not that arbitrary because the same method returns an empty array in other error cases. Better to be consistent!

from hashids.net.

It would be great to have TryDecode methods that follow the .NET idioms of attempting to parse without throwing an exception.

This is very useful in situations where the input can't be controlled or is unpredictable and avoids the overhead of throwing/catching exceptions.

from hashids.net.