add Cell.xy for RasterLayer cell coordinates and clarify pos/indices semantics

[wang-boyu]

What's the problem this feature will solve?

Mesa-Geo's RasterLayer / Cell APIs expose grid coordinates (cell.pos) and raster indices (cell.indices), but there is no direct way to access a cell's real-world / CRS coordinates. This is confusing for users working with georeferenced rasters, because positions looks like "(x, y)" and the layer carries CRS/transform metadata, so it’s natural to expect Cell.pos to be physical coordinates.

Describe the solution you'd like

1. Add cell.xy returning the cell's geographic/projected (x, y) coordinates, which is the cell center.
2. Add new properties:

• cell.grid_pos for cell.pos, Mesa-like coordinates, that also use [x, y] (int) but for grid space. (mesa doc)
• cell.rowcol for cell.indices, (row, col) indexing.

3. Depreate current cell.pos and cell.indices with a warning stating that they will be removed in a future release.

Additional context

• Confusion example: Invalid pos/index within RasterLayer (geographic spatial location not exposed to GeoSpace / RasterLayer)? #267
• Rasterio docs on Transforms: https://rasterio.readthedocs.io/en/stable/topics/transforms.html where (x, y) conventionally referrs to real-world geographic/projected coordinates, and (row, col) are indices.

[wang-boyu]

Closed through #299. See #299 (comment) for updates to the PR.

Later if we want to add support to Hexagon layers, we can define a HexCell class with the same cell.xy property for its geographic/projected coordinages in CRS. There could be properties such as cell.ijk if it's using a H3 ijk coordinate system, or cell.qr for Axial coordinates, and so on.