[Binary Search Tree][c language] new test case

[damaxi]

Inspired by @siebenschlaefer comment https://exercism.io/my/solutions/b6f9db46186547ba9a7bbebf3f193078?iteration_idx=1

It looks like there is a missing test case:

   int tree_data[] = { 3, 1, 2 };
   node_t *tree = build_tree(
      tree_data, ARRAY_SIZE(tree_data));

   int expected[] = { 1, 2, 3 };
   int *actual = sorted_data(tree);
   TEST_ASSERT_EQUAL_INT_ARRAY(
      expected, actual,
      ARRAY_SIZE(expected));

   free_tree(tree);
   free(actual);

I had created initially a wrong algorithm for the left and right branches. Where the left branch passe because there was a missing test case, while the right branch I fixed. Adding this test case will help to write a proper implementation.

[wolf99]

I cannot see the comment on the exercise (probably because I am not a mentor). Can you copy/screenshot the comment here, for context?

[damaxi]

Sure here you are: sorted_data() calls fill_data() which itself calls left_branch() for the left sub-tree and right_branch() for the right sub-tree.

Those two functions left_branch() and right_branch() look slightly different, in fact one of them does not do what it should.

You could try to analyze them to find the issue, or you could try this (additional) test case:

static void test_for_damaxi(void)
{
   int tree_data[] = { 3, 1, 2 };
   node_t *tree = build_tree(
      tree_data, ARRAY_SIZE(tree_data));

   int expected[] = { 1, 2, 3 };
   int *actual = sorted_data(tree);
   TEST_ASSERT_EQUAL_INT_ARRAY(
      expected, actual,
      ARRAY_SIZE(expected));

   free_tree(tree);
   free(actual);
}

Let me know if you need another hint or some less cryptic help.

[siebenschlaefer]

@wolf99 Perhaps you might be able to access https://exercism.io/mentor/solutions/b6f9db46186547ba9a7bbebf3f193078?iteration_idx=1

There are five tests for "can sort data": one for 1 element, three for 2 elements, and one for a "complex tree" (see canonical-data.json). The tree for that last test looks like this:

  2
+-+-+
1   3
    +---+
        6
      +-+-+
      5   7

In that particular solution there were separate functions for the left and right subtree of the root. The right subtree was handled correctly and the left subtree was not complex enough to detect the error.

We could add some notes in that example, or we could add another test with a different "complex" tree.

[wolf99]

If you would like to propose another tree that should be tested, please share here and a PR can be made with that

[ErikSchierboom]

@damaxi See the comment by @wolf99 above: do you have another tree that would make for a good test case? If so, could you share it here?