# LICENSE HEADER MANAGED BY add-license-header
#
# Copyright 2018 Kornia Team
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#

"""Based from the original code from Meta Platforms, Inc. and affiliates.

https://github.com/facebookresearch/segment-
anything/blob/3518c86b78b3bc9cf4fbe3d18e682fad1c79dc51/segment_anything/modeling/image_encoder.py
"""

from __future__ import annotations

from typing import Optional

import torch
import torch.nn.functional as F
from torch import nn

from kornia.contrib.models.common import LayerNorm2d, window_partition, window_unpartition
from kornia.contrib.models.sam.architecture.common import MLPBlock
from kornia.core import Module, Tensor, zeros


# This class and its supporting functions below lightly adapted from the ViTDet backbone available at: https://github.com/facebookresearch/detectron2/blob/main/detectron2/modeling/backbone/vit.py # noqa
class ImageEncoderViT(Module):
    def __init__(
        self,
        img_size: int = 1024,
        *,
        patch_size: int = 16,
        in_chans: int = 3,
        embed_dim: int = 768,
        depth: int = 12,
        num_heads: int = 12,
        mlp_ratio: float = 4.0,
        out_chans: int = 256,
        qkv_bias: bool = True,
        norm_layer: type[Module] = nn.LayerNorm,
        act_layer: type[Module] = nn.GELU,
        use_abs_pos: bool = True,
        use_rel_pos: bool = False,
        rel_pos_zero_init: bool = True,
        window_size: int = 0,
        global_attn_indexes: tuple[int, ...] = (),
    ) -> None:
        """Construct Image Encoder ViT.

        Args:
        img_size: Input image size.
        patch_size: Patch size.
        in_chans: Number of input image channels.
        embed_dim: Patch embedding dimension.
        depth: Depth of ViT.
        num_heads: Number of attention heads in each ViT block.
        mlp_ratio: Ratio of mlp hidden dim to embedding dim.
        out_chans: Number of output channels.
        qkv_bias: If True, add a learnable bias to query, key, value.
        norm_layer: Normalization layer.
        act_layer: Activation layer.
        use_abs_pos: If True, use absolute positional embeddings.
        use_rel_pos: If True, add relative positional embeddings to the attention map.
        rel_pos_zero_init: If True, zero initialize relative positional parameters.
        window_size: Window size for window attention blocks.
        global_attn_indexes: Indexes for blocks using global attention.

        """
        super().__init__()
        self.img_size = img_size

        self.patch_embed = PatchEmbed(
            kernel_size=(patch_size, patch_size),
            stride=(patch_size, patch_size),
            in_chans=in_chans,
            embed_dim=embed_dim,
        )

        self.pos_embed: Optional[nn.Parameter] = None
        if use_abs_pos:
            # Initialize absolute positional embedding with pretrain image size.
            self.pos_embed = nn.Parameter(zeros(1, img_size // patch_size, img_size // patch_size, embed_dim))

        self.blocks = nn.ModuleList()
        for i in range(depth):
            block = Block(
                dim=embed_dim,
                num_heads=num_heads,
                mlp_ratio=mlp_ratio,
                qkv_bias=qkv_bias,
                norm_layer=norm_layer,
                act_layer=act_layer,
                use_rel_pos=use_rel_pos,
                rel_pos_zero_init=rel_pos_zero_init,
                window_size=window_size if i not in global_attn_indexes else 0,
                input_size=(img_size // patch_size, img_size // patch_size),
            )
            self.blocks.append(block)

        self.neck = nn.Sequential(
            nn.Conv2d(embed_dim, out_chans, kernel_size=1, bias=False),
            LayerNorm2d(out_chans),
            nn.Conv2d(out_chans, out_chans, kernel_size=3, padding=1, bias=False),
            LayerNorm2d(out_chans),
        )

    def forward(self, x: Tensor) -> Tensor:
        x = self.patch_embed(x)
        if self.pos_embed is not None:
            x = x + self.pos_embed

        for blk in self.blocks:
            x = blk(x)

        x = self.neck(x.permute(0, 3, 1, 2))

        return x


class Block(Module):
    """Transformer blocks with support of window attention and residual propagation blocks."""

    def __init__(
        self,
        dim: int,
        num_heads: int,
        mlp_ratio: float = 4.0,
        qkv_bias: bool = True,
        norm_layer: type[Module] = nn.LayerNorm,
        act_layer: type[Module] = nn.GELU,
        use_rel_pos: bool = False,
        rel_pos_zero_init: bool = True,
        window_size: int = 0,
        input_size: Optional[tuple[int, int]] = None,
    ) -> None:
        """Construct transformer block.

        Args:
        dim: Number of input channels.
        num_heads: Number of attention heads in each ViT block.
        mlp_ratio: Ratio of mlp hidden dim to embedding dim.
        qkv_bias: If True, add a learnable bias to query, key, value.
        norm_layer: Normalization layer.
        act_layer: Activation layer.
        use_rel_pos: If True, add relative positional embeddings to the attention map.
        rel_pos_zero_init: If True, zero initialize relative positional parameters.
        window_size: Window size for window attention blocks. If it equals 0, then use global attention.
        input_size: Input resolution for calculating the relative positional parameter size.

        """
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.attn = Attention(
            dim,
            num_heads=num_heads,
            qkv_bias=qkv_bias,
            use_rel_pos=use_rel_pos,
            rel_pos_zero_init=rel_pos_zero_init,
            input_size=input_size if window_size == 0 else (window_size, window_size),
        )

        self.norm2 = norm_layer(dim)
        self.mlp = MLPBlock(embedding_dim=dim, mlp_dim=int(dim * mlp_ratio), act=act_layer)

        self.window_size = window_size

    def forward(self, x: Tensor) -> Tensor:
        shortcut = x
        x = self.norm1(x)
        # Window partition
        if self.window_size > 0:
            H, W = x.shape[1], x.shape[2]
            x, pad_hw = window_partition(x, self.window_size)

        x = self.attn(x)
        # Reverse window partition
        if self.window_size > 0:
            x = window_unpartition(x, self.window_size, pad_hw, (H, W))

        x = shortcut + x
        x = x + self.mlp(self.norm2(x))

        return x


class Attention(Module):
    """Multi-head Attention block with relative position embeddings."""

    def __init__(
        self,
        dim: int,
        num_heads: int = 8,
        qkv_bias: bool = True,
        use_rel_pos: bool = False,
        rel_pos_zero_init: bool = True,
        input_size: Optional[tuple[int, int]] = None,
    ) -> None:
        """Construct attention block.

        Args:
        dim: Number of input channels.
        num_heads: Number of attention heads.
        qkv_bias:  If True, add a learnable bias to query, key, value.
        use_rel_pos: If True, add relative positional embeddings to the attention map.
        rel_pos_zero_init: If True, zero initialize relative positional parameters.
        input_size: Input resolution for calculating the relative positional parameter size.

        """
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = head_dim**-0.5

        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.proj = nn.Linear(dim, dim)

        self.use_rel_pos = use_rel_pos
        if self.use_rel_pos and isinstance(input_size, tuple):
            # initialize relative positional embeddings
            self.rel_pos_h = nn.Parameter(zeros(2 * input_size[0] - 1, head_dim))
            self.rel_pos_w = nn.Parameter(zeros(2 * input_size[1] - 1, head_dim))

    def forward(self, x: Tensor) -> Tensor:
        B, H, W, _ = x.shape
        # qkv with shape (3, B, nHead, H * W, C)
        qkv = self.qkv(x).reshape(B, H * W, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)
        # q, k, v with shape (B * nHead, H * W, C)
        q, k, v = qkv.reshape(3, B * self.num_heads, H * W, -1).unbind(0)

        attn = (q * self.scale) @ k.transpose(-2, -1)

        if self.use_rel_pos:
            attn = add_decomposed_rel_pos(attn, q, self.rel_pos_h, self.rel_pos_w, (H, W), (H, W))

        attn = attn.softmax(dim=-1)
        x = (attn @ v).view(B, self.num_heads, H, W, -1).permute(0, 2, 3, 1, 4).reshape(B, H, W, -1)
        x = self.proj(x)

        return x


def get_rel_pos(q_size: int, k_size: int, rel_pos: Tensor) -> Tensor:
    """Get relative positional embeddings according to the relative positions of query and key sizes.

    Args:
        q_size: size of query q.
        k_size: size of key k.
        rel_pos: relative position embeddings (L, C).

    Returns:
        Extracted positional embeddings according to relative positions.

    """
    max_rel_dist = int(2 * max(q_size, k_size) - 1)
    # Interpolate rel pos if needed.
    if rel_pos.shape[0] != max_rel_dist:
        # Interpolate rel pos.
        rel_pos_resized = F.interpolate(
            rel_pos.reshape(1, rel_pos.shape[0], -1).permute(0, 2, 1), size=max_rel_dist, mode="linear"
        )
        rel_pos_resized = rel_pos_resized.reshape(-1, max_rel_dist).permute(1, 0)
    else:
        rel_pos_resized = rel_pos

    # Scale the coords with short length if shapes for q and k are different.
    q_coords = torch.arange(q_size)[:, None] * max(k_size / q_size, 1.0)
    k_coords = torch.arange(k_size)[None, :] * max(q_size / k_size, 1.0)
    relative_coords = (q_coords - k_coords) + (k_size - 1) * max(q_size / k_size, 1.0)

    return rel_pos_resized[relative_coords.long()]


def add_decomposed_rel_pos(
    attn: Tensor, q: Tensor, rel_pos_h: Tensor, rel_pos_w: Tensor, q_size: tuple[int, int], k_size: tuple[int, int]
) -> Tensor:
    """Calculate decomposed Relative Positional Embeddings.

    from :paper:`mvitv2`.
    https://github.com/facebookresearch/mvit/blob/19786631e330df9f3622e5402b4a419a263a2c80/mvit/models/attention.py


    Args:
        attn: attention map.
        q: query q in the attention layer with shape (B, q_h * q_w, C).
        rel_pos_h: relative position embeddings (Lh, C) for height axis.
        rel_pos_w: relative position embeddings (Lw, C) for width axis.
        q_size: spatial sequence size of query q with (q_h, q_w).
        k_size: spatial sequence size of key k with (k_h, k_w).

    Returns:
        att: attention map with added relative positional embeddings.

    """
    q_h, q_w = q_size
    k_h, k_w = k_size
    Rh = get_rel_pos(q_h, k_h, rel_pos_h)
    Rw = get_rel_pos(q_w, k_w, rel_pos_w)

    B, _, dim = q.shape
    r_q = q.reshape(B, q_h, q_w, dim)
    rel_h = torch.einsum("bhwc,hkc->bhwk", r_q, Rh)
    rel_w = torch.einsum("bhwc,wkc->bhwk", r_q, Rw)

    attn = (attn.view(B, q_h, q_w, k_h, k_w) + rel_h[:, :, :, :, None] + rel_w[:, :, :, None, :]).view(
        B, q_h * q_w, k_h * k_w
    )

    return attn


class PatchEmbed(Module):
    """Image to Patch Embedding."""

    def __init__(
        self,
        kernel_size: tuple[int, int] = (16, 16),
        stride: tuple[int, int] = (16, 16),
        padding: tuple[int, int] = (0, 0),
        in_chans: int = 3,
        embed_dim: int = 768,
    ) -> None:
        """Construct Patch Embedding.

        Args:
        kernel_size: kernel size of the projection layer.
        stride: stride of the projection layer.
        padding: padding size of the projection layer.
        in_chans: Number of input image channels.
        embed_dim: Patch embedding dimension.

        """
        super().__init__()

        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=kernel_size, stride=stride, padding=padding)

    def forward(self, x: Tensor) -> Tensor:
        x = self.proj(x)
        # B C H W -> B H W C
        x = x.permute(0, 2, 3, 1)
        return x