"""Transformer block used by Flexibrain downstream heads.

Reference:
- Brain-Harmony / BrainHarmonix official codebase: https://github.com/hzlab/Brain-Harmony
- The official README marks the project license as CC BY-NC-SA 4.0.

Only the small Block/Attention/MLP subset required by the downstream head is
kept here; the rest of the Brain-Harmony repository is intentionally not
vendored into Flexibrain.
"""

import torch
import torch.nn as nn


def drop_path(x, drop_prob: float = 0.0, training: bool = False):
    if drop_prob == 0.0 or not training:
        return x
    keep_prob = 1 - drop_prob
    shape = (x.shape[0],) + (1,) * (x.ndim - 1)
    random_tensor = keep_prob + torch.rand(shape, dtype=x.dtype, device=x.device)
    random_tensor.floor_()
    return x.div(keep_prob) * random_tensor


class DropPath(nn.Module):
    def __init__(self, drop_prob=0.0):
        super().__init__()
        self.drop_prob = float(drop_prob)

    def forward(self, x):
        return drop_path(x, self.drop_prob, self.training)


class MLP(nn.Module):
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.0):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.drop = nn.Dropout(drop)

    def forward(self, x):
        x = self.fc1(x)
        x = self.act(x)
        x = self.drop(x)
        x = self.fc2(x)
        return self.drop(x)


class Attention(nn.Module):
    def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0.0, proj_drop=0.0):
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = qk_scale or head_dim ** -0.5
        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

    def forward(self, x, attention_mask=None, output_attentions=False):
        B, N, C = x.shape
        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
        q, k, v = qkv[0], qkv[1], qkv[2]
        attn = (q @ k.transpose(-2, -1)) * self.scale
        if attention_mask is not None:
            valid = attention_mask.bool()
            key_mask = ~valid[:, None, None, :]
            attn = attn.masked_fill(key_mask, torch.finfo(attn.dtype).min)
        attn = self.attn_drop(attn.softmax(dim=-1))
        x = (attn @ v).transpose(1, 2).reshape(B, N, C)
        x = self.proj_drop(self.proj(x))
        return (x, attn) if output_attentions else (x, None)


class Block(nn.Module):
    def __init__(self, dim, num_heads, mlp_ratio=4.0, qkv_bias=False, qk_scale=None, drop=0.0, attn_drop=0.0, drop_path=0.0, act_layer=nn.GELU, norm_layer=nn.LayerNorm, attn_mode=None):
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.norm2 = norm_layer(dim)
        self.attn = Attention(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
        self.mlp = MLP(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer, drop=drop)

    def forward(self, x, attention_mask=None, return_attention=False):
        y, attn = self.attn(self.norm1(x), attention_mask=attention_mask, output_attentions=return_attention)
        x = x + self.drop_path(y)
        x = x + self.drop_path(self.mlp(self.norm2(x)))
        return (x, attn) if return_attention else x