Straight optimization proposal about texasholdemgameengine

[NikolayIT]

Will be fixed. Thanks! Create pull request next time.

[kiko81]

this way not the highest possible straight is returned but the lowest instead - will fix it

[NikolayIT]

Add unit tests (for regression)

[kiko81]

!!! line 134 - var straightCards = cards.Select(x => (int)x.Type).ToList(); - returns duplicates

renaming proposal: CompareTwoHandsWithHighCard => CompareKickers/CompareSideCards

Comparing straights - comparing any card[1-3] is enough - no Ace possible there -
bigger wins - no matter which way sorted - think will be faster O(1)

[NikolayIT]

Open new issues for every single bug.

[NikolayIT]

When comparing straights there are objects of type CardType so Ace is expected.

[kiko81]

Mean if compare e.g. 2nd cards only - no Ace possible there - Ace may be first or last only

[NikolayIT]

What about this case:

new[]
    {
        CardType.Six, CardType.Five,
        CardType.Four, CardType.Three,
        CardType.Two
    },
new[]
    {
        CardType.Five, CardType.Four,
        CardType.Three, CardType.Two,
        CardType.Ace
    }

[kiko81]

e.g.
2nd cards: five > four => first wins
3rd cards: four > three => first wins
4th cards.......
Aces can be in the beginning or the end of hand - not in the middle - don't check ends - any middle card will do the trick
Not wrong now, of course, but may be less complex

[NikolayIT]

I've got the idea but the current solution is O(n) and the one you suggest requires sorting which will be O(n * log(n).

[kiko81]

thought they come sorted from hand evaluator, from tests come shuffled - anyway U know better 👍