Haskell translation

JavaScript translation

C translation

Thanks for the Kata

Multiple punctuation marks are not being stripped in the tests:

should equal 'snatch. . meddle powerful available crow guide adorable books science hate?'
should equal 'milk destroy guide? . borrow crow books round penitent plough whimsical drain brash?'
should equal 'activity wise. .'

My code passed all the sample tests then passed 99 of the test suite cases with 11 failed when I attempted it. Then 103 cases with 7 failed when I attempted it again with no changes to the code. How would I go about checking the errors when the test suite code is not available for me to debug my failed test cases against? Also why would the number of passed cases and failed cases change with no change to the code?

Firstly, thank you for this great challenge. It was fun!

I believe the instructions should be clarified. Saying the maze does not have to be a rectangle is not clear. It could mean either that you should treat blank spaces (e.g. last two chars in the last row, which has 8 columns where the others had 10) as walls or that you should treat them as edges where you've succeeded in escaping. I interpret it the first way, which caused me some confusion. Effectively, the maze always is a rectangle, it's just not always given as such in the input data. You have to extrapolate it into a rectangle.

My suggestion would be to say something like, "The maze input is not always a rectangle. Any spaces not defined by the input should be treated as open spaces (not walls)."

python: the "jagged maze" test should be done with the reversed version too: some implementations may use the length of the first string as a constant, and that will make succeed that test where it would fail on the other version:

maze = ["###########",
        "#k        #",
        "#########"]