pymrm.ibm.IBM

class IBM:
    """Consolidated IBM crossing data for both sides of an immersed interface.

    Each entry (index ``k``) corresponds to one *wall crossing* -- a face
    between a fluid cell (outside) and a solid cell (inside).  The wall
    position ``coords[k]`` is shared by both sides.

    Indices are **spatial** flat indices (C-order in ``spatial_shape``), not
    full-field indices.  :func:`apply_ibm` expands them to the full field using
    the ``axes`` / ``shape`` information.

    Parameters indexed by crossing (length ``n_crossings``)
    --------------------------------------------------------
    n_crossings : int
        Number of wall crossings.
    coords : ndarray, shape (n_crossings, ndim_spatial)
        Physical coordinates of each wall crossing (spatial dimensions only).
    crossing_key : ndarray of int
        Canonical face identifier.
    axis : ndarray of int
        Spatial axis of each crossing.
    direction : ndarray of int
        Direction (+1 or -1) from ``row_out`` to ``ghost_out``.

    Outside (fluid cut-cell) fields
    --------------------------------
    row_out, ghost_out : ndarray of int
        Spatial flat indices of the fluid cut cell and its solid ghost
        neighbour.
    opp_out : ndarray of int
        Spatial flat index of the opposite fluid neighbour (``-1`` when
        unavailable).
    coef_c_out, coef_o_out, coef_w_out, coef_w_sib_out : ndarray of float
        Lagrange coefficients.
    sib_out : ndarray of int
        Crossing index of the sandwich sibling (``-1`` if not a sandwich).
    row_scale_out : ndarray of float, shape (n_spatial_cells,)
        Per-spatial-cell conditioning scale for the outside (fluid cut) rows.

    Inside (solid cut-cell) fields
    --------------------------------
    row_in, ghost_in, opp_in, coef_c_in, coef_o_in, coef_w_in,
    coef_w_sib_in, sib_in, row_scale_in : analogous inside fields.

    Shape information
    -----------------
    spatial_shape : tuple
        Shape of the SDF / spatial grid.
    shape : tuple
        Full field shape (spatial + non-spatial axes).
    axes : tuple of int
        Which axes of ``shape`` are spatial.
    ns_shape : tuple
        Non-spatial dimensions of ``shape``.
    ns_size : int
        Product of non-spatial dims (1 when purely spatial).
    n_cells : int
        Total number of cells, ``math.prod(shape)``.
    n_spatial_cells : int
        Number of spatial cells, ``math.prod(spatial_shape)``.
    """

    n_crossings: int
    coords: np.ndarray
    crossing_key: np.ndarray
    axis: np.ndarray
    direction: np.ndarray

    row_out: np.ndarray
    ghost_out: np.ndarray
    opp_out: np.ndarray
    coef_c_out: np.ndarray
    coef_o_out: np.ndarray
    coef_w_out: np.ndarray
    coef_w_sib_out: np.ndarray
    sib_out: np.ndarray
    row_scale_out: np.ndarray

    row_in: np.ndarray
    ghost_in: np.ndarray
    opp_in: np.ndarray
    coef_c_in: np.ndarray
    coef_o_in: np.ndarray
    coef_w_in: np.ndarray
    coef_w_sib_in: np.ndarray
    sib_in: np.ndarray
    row_scale_in: np.ndarray

    spatial_shape: tuple
    shape: tuple
    axes: tuple
    ns_shape: tuple
    ns_size: int
    n_cells: int
    n_spatial_cells: int