duplicate issue (see below)

@eb110 I actually agree with you. It should be explained (or at least displayed in sample tests) that:

• The input must be trimmed
• The character range is 0-9 for digits between 0-9, and A-J for digits between 10-19
• The function is case insensitive
• If the beginning of the string is convertable, you convert that and ignore whatever is left

Explaining why 'K' is invalid and '4jzaaa' is valid??

What would be a better wording, if you would propose one? ^^

If a character is not within the valid range to generate its base-20 representation, then it is considered invalid, so should return -1

For a 7-kyu kata, simple 'If it cannot be done, return -1.' is too ambigous to be treated as valid description. 81% of rating score confirms that.

It looks like you were working hard :-) First thing first - the author of the kata is from Poland and this is why i have included the comment in the same language. The content is rather thrivial - simple -> 'As usually you have given interesting task to solve'. Regards your research - '£' does not exist in polish alphabet, but you can treat it as polish 'Ł' as both looks similar.

I have tested these keys without success to decrypt your message: agedyropulik politykarenu regulaminowy kaceminutowy koniecmatury arewybuhoki bawoletykiju

Should your message contain the Pound-Sign?

Let's assume the language is English and the alphabet consists of 18 ( = 26 (a, b, ..., z) - 8 (=p r o b l e m t) ) unknown characters. This results in 1.5e12 ( =Binomial(18*18, 6) ) keys. Then around 1.0e13 (=7*1.5e12) words would have to be checked to see whether they are valid English words. Rather difficult, isn't it?

The best way! I keep forgetting that I don't need to explicity write true and false because the comparison operator always does that automatically for me.

The upper bound is not inclusive, so.... The 2nd change must take place to match the description.

https://stackoverflow.com/questions/5063269/c-sharp-random-next-never-returns-the-upper-bound

Random tests are fine - were made upon specific measurement. Not going to change it according to your suggestion.

Removed System.Linq - btw, this is arbitrary decission but don't want to waste time for argue. The range for random test is correct - not gonna change it.

Random tests should be generated as such: (use 51, 21 and 11 as the upper bound since Python's randint upper bound is inclusive whilst C# is exclusive)

n,m,s=randint(1,50),randint(1,50),[randint(1,20),randint(1,10),randint(1,10),randint(1,20)]

• using System should be in initial solution setup to ease debugging

• should be

int now = rnd.Next(0, 24);
int start = rnd.Next(0, 24);
int end = (start + rnd.Next(1, 24)) % 24;

• using System.Linq; should be removed from initial solution setup

• should be int l = rnd.Next(2, 51);