Thank you, fixed

Have fixed test cases

Python random test cases have bugs, that VIOLATE kata description

The strings will not be the same length, but they may be empty (length0).

e.g.:
'U5y', 'QAt';
'lRuCSFUkYf', 'b4Lzkjz9sE';
etc.

I believe, there is a slight bug in Python random test cases, e.g.:

I get

Testing for F0x3d
It should work for random inputs too: None should equal 3901

but the right hex representation of 3901 is 0xf3d (case insensitive) and F0x3d is an invalid hex number;

More example:

Testing for a0x43
It should work for random inputs too: None should equal 2627

the right hex representation of 2627 - 0xa43, a0x43 is an invalid hex number;

Testing for B0x5967
It should work for random inputs too: None should equal 743783

the right hex representation of 743783 - 0xb5967, B0x5967 is an invalid hex number.

P.S. I had to run solution about 20 times before getting another buggy test case.

Agree
Thank you for your remarks and suggestions
Fixed?

Agree
Fixed

Sorry for the delayed reply.
I had a bug in my code - confused class field with instance field.
Issue resolved.

Got it, sorry for misunderstanding

I believe, for some n there are more than one solution!
I.e.:

n = 65
11, 10, 9, 8, 7, 6, 5, 4, 3, 2
15, 14, 13, 12, 11
33, 32 (in test case)

n = 18
6, 5, 4, 3
7, 6, 5 (in test case)

and there are more

There is definitely a bug in Python translation in random tests.
Example logs:

Testing for h.
 Log
Words in dictionary: [code, co, ab, at, vkwgu, ldx, cod, hrb, wd, codewars, hh, ejfi, ear, a, ate, qxd, dqyg, j, ko, hqhr, r, tpczl, w, iyz]
Pattern: 'h.' 
It should work for random inputs too: True should equal False

'h.' should match hh

Testing for ....
 Log
Words in dictionary: [lkvu, code, co, vp, at, yke, orkrl, po, alda, codewars, wr, ear, fm, a, c, ate, rr, oi, fxjvu, h, cod, v, y, kxxh]
Pattern: '....'
It should work for random inputs too: True should equal False

'....' should match lkvu, code, alda, kxxh

Adding DISTINCT keywork solves the constraint to use IN without actually using it)

Definitely true, but in case of few number of iterations, like in the cata above, you won't notice a difference in time

It should be mentioned in the description that spaces in the begining/end of test cases are VALID or change the test cases to return 0

Correct wrong type of parameter - should be long
In other way the compiler produces errors because of type overflowing