# -*- coding: utf-8 -*-

"""
This python module implements the FORCE flux
for the shallow water equations

Copyright (C) 2016  SINTEF ICT

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <http://www.gnu.org/licenses/>.
"""

# Import packages we need
import numpy as np

from GPUSimulators.common import ArakawaA2D
from GPUSimulators.simulator import BaseSimulator, BoundaryCondition
from GPUSimulators.gpu import GPUHandler


class Force(BaseSimulator):
    """
    Class that solves the SW equations 
    """

    def __init__(self,
                 context,
                 h0, hu0, hv0,
                 nx, ny,
                 dx, dy,
                 g,
                 cfl_scale=0.9,
                 boundary_conditions=BoundaryCondition(),
                 block_width=16, block_height=16,
                 dt: float = None,
                 compile_opts: list[str] = []):
        """
        Initialization routine
        
        Args:
            h0: Water depth incl ghost cells, (nx+1)*(ny+1) cells
            hu0: Initial momentum along x-axis incl ghost cells, (nx+1)*(ny+1) cells
            hv0: Initial momentum along y-axis incl ghost cells, (nx+1)*(ny+1) cells
            nx: Number of cells along x-axis
            ny: Number of cells along y-axis
            dx: Grid cell spacing along x-axis (20 000 m)
            dy: Grid cell spacing along y-axis (20 000 m)
            dt: Size of each timestep (90 s)
            g: Gravitational accelleration (9.81 m/s^2)
            compile_opts: Pass a list of nvcc compiler options
        """

        # Call super constructor
        super().__init__(context,
                         nx, ny,
                         dx, dy,
                         boundary_conditions,
                         cfl_scale,
                         1,
                         block_width, block_height)
        self.g = np.float32(g)

        # Get kernels
        module = context.get_module("SWE2D_FORCE",
                                    "FORCEKernel",
                                    defines={
                                        'BLOCK_WIDTH': self.block_size[0],
                                        'BLOCK_HEIGHT': self.block_size[1]
                                    },
                                    compile_args={
                                        'cuda': {
                                            'no_extern_c': True,
                                            'options': ["--use_fast_math"] + compile_opts,
                                        },
                                        'hip': compile_opts,
                                    },
                                    jit_compile_args={})
        self.handler = GPUHandler(context, module, "FORCEKernel", "iiffffiPiPiPiPiPiPiPiiii", self.grid_size)

        # Create data by uploading to the device
        self.u0 = ArakawaA2D(self.stream,
                             nx, ny,
                             1, 1,
                             [h0, hu0, hv0])
        self.u1 = ArakawaA2D(self.stream,
                             nx, ny,
                             1, 1,
                             [None, None, None])

        if dt is None:
            dt_x = np.min(self.dx / (np.abs(hu0 / h0) + np.sqrt(g * h0)))
            dt_y = np.min(self.dy / (np.abs(hv0 / h0) + np.sqrt(g * h0)))
            self.dt = min(dt_x, dt_y)
        else:
            self.dt = dt

        self.handler.array_fill(self.dt, self.stream)

    def substep(self, dt, step_number):
        self.handler.prepared_call(self.grid_size, self.block_size, self.stream,
                                        [self.nx, self.ny,
                                        self.dx, self.dy, dt,
                                        self.g,
                                        self.boundary_conditions,
                                        self.u0[0].data, self.u0[0].get_pitch(),
                                        self.u0[1].data, self.u0[1].get_pitch(),
                                        self.u0[2].data, self.u0[2].get_pitch(),
                                        self.u1[0].data, self.u1[0].get_pitch(),
                                        self.u1[1].data, self.u1[1].get_pitch(),
                                        self.u1[2].data, self.u1[2].get_pitch(),
                                        self.handler.cfl_data,
                                        0, 0,
                                        self.nx, self.ny])

        self.u0, self.u1 = self.u1, self.u0

    def get_output(self):
        return self.u0

    def check(self):
        self.u0.check()
        self.u1.check()

    def compute_dt(self):
        max_dt = self.handler.array_min(self.stream)
        return max_dt