done by OP

Approved

python new test framework is required. updated in this fork

Python:
The example tests and test suite aren't using the test framework, missing the imports (user's solution + framework).

C#: method name should be PascalCase (Please refer to implementation of backward compatibility here )

Awesome kata actually,
was amazed by the solution page,
Nice simple beauiful solutions, mine was rather complicated!! :-(
Thanks Nancy27

Java requires an array with an empty string, other translations an empty array. This should be fixed.

"If you get back to beginning, return an empty array." - there is wrong test written in Sample Tests. It assumes that when you go to the beggining, you return not an empty array but an empty string so that the array length is = 1... Please correct that.

Fixed.

I also prevented input modification changing the test results and made the random tests into a loop (the author wrote really crappy Python code there).

Python test results seem reversed. ['S'] should equal [] should equal [] should equal ['S']...

sorry the issue is when I run this code in my com it's give the right output
but here it gave me wrong output
and every time I try it gives another output
the wrong output exactly in the sorting of the list

That's a question, not an issue.

Don't mutate the input object. If you must, copy it first.

Closed the previous issue while I shouldn"t have, so here it goes again...:

Java tests are not consistent with the python version:

For python:
goal=["N","W","S","E"]
test.assert_equals(directions(goal), [])

For java:
String[] goal3 = {"N","W","S","E"};
assertArrayEquals(new String[] {""}, ShortestPath.directions(goal3));

Python does the right move, here. Should be new String[0] for java.

And same problem for the inverted test: the input with new String[] {""} is not an empty array so is invalid. Python sends a real empty list [] instead.

And DM comments:

Agree. Irrespective of python, it is just not sensible

• it means all solutions have to code for a special case of 0.
• if means the size of the array does not always mean the number of steps

Damn me!! Not corrected. Reopening, so... :/