Attempting to Decode short strings results in incorrect exceptions being produced about hashids.net

[crmeadowcroft]

Should have mentioned this was using the default arguments except a minimum hash length of 8. The exception with 1 character still happens using any minimum hash lengths including the default (0).

The System.ArgumentException: Destination is too short. (Parameter 'destination') happens whenever a string is passed in which is less than the minimum hash length.

[Clive321A]

Hi, I just came across this with a Unit test trying to decode the phrase "Junk"

Even the TryDecodeSingle fails? would have thought that one should have just returned false instead of throwing an Exception?
I ended up wrapping the whole thing in a Try Catch, which works, but is this intended behaviour?

[ullmark]

@CliveBennett I agree, a function named "TryBlabla" shouldn't ever produce an exception imo if we want to follow the C# naming "standard"...

We should fix that.

[crmeadowcroft]

Sorry I'm not able to work out a Pull request, but in the function "private long GetNumberFrom(string hash)" changing:

Span<char> resultBuffer = stackalloc char[guardedHash.Length];

to

var resultBufferLength = Math.Max(Math.Max(guardedHash.Length, 20), _minHashLength);`
Span<char> resultBuffer = stackalloc char[resultBufferLength];

seems to fix things. The constant length 20 seems to be used commonly through the algorithm though I don't know why. Also could add tests:

[Fact]
void Issue_75_1_character_hash_string_should_not_throw_out_of_range_exception()
{
    var hashids = new Hashids("salt");
    Assert.Throws<NoResultException>(() => hashids.DecodeSingle("a"));
}

[Fact]
void Issue_75_too_short_hash_string_should_not_throw_argument_exception()
{
    var hashids = new Hashids("salt", 40);
    Assert.Throws<NoResultException>(() => hashids.DecodeSingle("ab"));
}

[manigandham]

thanks @crmeadowcroft for pointing that out, added a new PR to fix

[manigandham]

fixed in 1.6.1: https://www.nuget.org/packages/Hashids.net/1.6.1

[jetersen]

Thanks for fixing the bug I introduced in #65