How do I allocate to a C# dictionary having a string key and an integer 2D array or transform for value?

Question

I'm working in C# on a Unity 2D game that spawns a shape that has only one correct landing place. There are 7 basic shapes but the component 4 squares of each shape are unique, with each bearing, say, a unique number. There are many more than 7 unique shapes because any component square is unique to the shape set. So the shapes have squares bearing a unique number (1,2,3,4 on a square; 5,6,7,8 on another square; 9,10,11,12 in a line (I-shape), 13,14,15,16 in another line (I-shape), etc. each of these shapes must be moved (by dragging, dropping or rotating) to its particular place in the grid.

I need to swiftly (1) indicate whether a shape has landed correctly and, if not, (2) move the misplaced (and possibly mis-rotated) shape to its proper location so that remaining falling pieces are not blocked from attaining their correct placements.

For (1), I'm using a dictionary with a string for the key that identifies the square component of the shape and an integer 2D array for the grid location values. Here is the declaration:

Dictionary<string, int[,]> dictOfProperLocations = new Dictionary<string, int[,]>();

Visual Studio is OK with that. But I am unable so far to allocate key-value pairs to the dictionary. I've tried the following and variants to indicate that a straight shape (an I-shape) bearing the numbers 1-4 belongs at the origin at its left end and stretches straight to the right:

dictOfProperLocations.Add("1", [0,0]);
dictOfProperLocations.Add("2", [1,0]);
dictOfProperLocations.Add("3", [2,0]);
dictOfProperLocations.Add("4", [3,0]);

But Visual Studio objects to that. I find many examples for C# dictionaries with string-keys and 1D-array-values, but no examples for this case with 2D-array-values.

Or might this be better handled with a 2D Transform[,] for the values representing the proper grid locations? Thank you.

Answer 1

First, you need to understand that just because you need two elements in your array does not make it a two-dimensional array.

We can define a one-dimensional array with two elements like this:

int[] myArray = new int[]{1, 2};

This array is one-dimensional: you need only one index input to select either myArray[0] == 1 or myArray[1] == 2

A two-dimensional array is an array that needs two indices to select an element. eg..

int[,] myMatrix = new int[,]{
 {0, 1, 2},
 {3, 4, 5},
 {6, 7, 8}
};

When we want to index this array, we need to specify both the row and the column we want to address: myMatrix[1, 0] == 3

But if you know in advance that you just need a pair of coordinates, an array is overkill. We can store our coordinates more efficiently like this:

in Int2.cs:

public struct Int2 {
    public int x;
    public int y;
    public Int2(int x, int y) {
        this.x = x;
        this.y = y;
    }
}

At the top of your Board class, where you define member variables:

public Dictionary<int, Int2> dictionaryOfProperLocations = new Dictionary<int, Int2>();

When you want to define a new location mapping inside that type:

dictionaryOfProperLocations.Add(0, new Int2(0, 0));

This way you have fewer heap allocations, array bounds checks, etc. Using an integer as a key instead of a string containing an integer makes the hashing work of the Dictionary a little easier too.

I'm not sure if using a dictionary in this way is necessarily the best way to implement the gameplay you're describing, but for now I'll just address the syntax. If you want advice on algorithms & data structures for doing spatial queries about tetrominos, that might be a separate question we could help you with.