Flood fill fills diagonally & errors if edges are to be filled

[appgurueu]

1f17076
It:

1. Fills diagonally, which isn't how flood fill usually works;
2. Lacks a check for x < width && y < height: I assume it crashes if that happens (with an AIOOB Ex.)

[siriak]

IMO filling diagonally is fine, it's one of the ways it can be implemented. As for the check, could you fix that and maybe add a test for that?

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[appgurueu]

What the heck, issues don't resolve themselves. The bot should be disabled.

[nikslyon19]

Should this not be categorised as a Backtracking problem, rather than dynamic programming?

[appgurueu]

Should this not be categorised as a Backtracking problem, rather than dynamic programming?

Flood fill is indeed often categorized as backtracing. That's a different issue though: This issue is about a possible crash in the implementation.

[nikslyon19]

Flood fill is indeed often categorized as backtracing. That's a different issue though: This issue is about a possible crash in the implementation.

Agreed. The corner case is not covered for the function going beyond the dimensions of the 2D array.
In the example given in the code, if you were to replace '3' with '1' at any of the right/bottom most position where it is adjacent to '1', you will get index out of bounds exception.

[nikslyon19]

IMO filling diagonally is fine, it's one of the ways it can be implemented. As for the check, could you fix that and maybe add a test for that?

Since this issue is still unresolved for months, would be happy to take over this one.

Would be covering the missing corner case, add a few test cases (asserts I believe?), and if it's okay, would like to move this code under the backtracking directory/category.