Spaces:
Sleeping
Sleeping
File size: 11,905 Bytes
2cf467c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 | """
布局系统核心节点定义
基于 template.ebnf 实现的布局系统
支持:
- Flow Layout (ROW/COLUMN)
- Non-Flow Layout (Z-layer)
"""
from abc import ABC, abstractmethod
from typing import List, Optional, Tuple, Any
from dataclasses import dataclass, field
from enum import Enum
import numpy as np
# ============= 枚举定义 =============
class NodeType(Enum):
"""节点类型"""
GROUP = "GROUP"
TEXT = "TEXT"
IMAGE = "IMAGE"
CHART = "CHART"
SHAPE = "SHAPE"
class LayoutType(Enum):
"""布局类型"""
FLOW = "FLOW"
NON_FLOW = "NON_FLOW"
class FlowDirection(Enum):
"""Flow布局方向"""
ROW = "ROW" # 水平排列
COLUMN = "COLUMN" # 垂直排列
# 未来扩展:
# GRID = "GRID"
# CIRCULAR = "CIRCULAR"
# IRREGULAR = "IRREGULAR"
class MainAlignment(Enum):
"""主轴对齐方式"""
START = "START"
CENTER = "CENTER"
END = "END"
class CrossAlignment(Enum):
"""交叉轴对齐方式"""
START = "START"
CENTER = "CENTER"
END = "END"
STRETCH = "STRETCH"
class PositionAlign(Enum):
"""非流式布局位置对齐"""
TOP_LEFT = "top-left"
TOP_CENTER = "top-center"
TOP_RIGHT = "top-right"
CENTER_LEFT = "left"
CENTER = "center"
CENTER_RIGHT = "right"
BOTTOM_LEFT = "bottom-left"
BOTTOM_CENTER = "bottom-center"
BOTTOM_RIGHT = "bottom-right"
class Alignment(Enum):
"""对齐方式(用于每个节点自己的对齐)"""
START = "START"
CENTER = "CENTER"
END = "END"
LEFT = "LEFT"
RIGHT = "RIGHT"
# 特殊值(用于 layer 子节点)
BACKGROUND = "BACKGROUND"
TOP_LEFT = "TOP_LEFT"
TOP_CENTER = "TOP_CENTER"
TOP_RIGHT = "TOP_RIGHT"
BOTTOM_LEFT = "BOTTOM_LEFT"
BOTTOM_CENTER = "BOTTOM_CENTER"
BOTTOM_RIGHT = "BOTTOM_RIGHT"
# ============= 数据类定义 =============
@dataclass
class BoundingBox:
"""边界框 (x, y, width, height)"""
x: float
y: float
width: float
height: float
@property
def left(self) -> float:
return self.x
@property
def right(self) -> float:
return self.x + self.width
@property
def top(self) -> float:
return self.y
@property
def bottom(self) -> float:
return self.y + self.height
@property
def center_x(self) -> float:
return self.x + self.width / 2
@property
def center_y(self) -> float:
return self.y + self.height / 2
@dataclass
class Padding:
"""内边距"""
top: float = 0
right: float = 0
bottom: float = 0
left: float = 0
@classmethod
def uniform(cls, value: float) -> 'Padding':
"""创建统一的内边距"""
return cls(value, value, value, value)
@property
def horizontal(self) -> float:
"""水平方向总内边距"""
return self.left + self.right
@property
def vertical(self) -> float:
"""垂直方向总内边距"""
return self.top + self.bottom
@dataclass
class FlowAlignment:
"""Flow布局对齐配置"""
main: MainAlignment = MainAlignment.START
cross: CrossAlignment = CrossAlignment.START
# ============= 抽象基类 =============
class Node(ABC):
"""节点抽象基类"""
def __init__(self, node_id: str, node_type: NodeType, alignment: Optional[str] = None, parent: Optional['Node'] = None):
self.id = node_id
self.type = node_type
self.alignment = alignment # 每个节点自己的对齐方式
self.bbox: Optional[BoundingBox] = None # 布局计算后的边界框
self.parent: Optional['Node'] = parent # 父节点引用
@abstractmethod
def compute_intrinsic_size(self) -> Tuple[float, float]:
"""
计算节点的固有尺寸 (width, height)
对于 Leaf Node: 返回内容的实际尺寸
对于 Non-Leaf Node: 根据子节点和布局规则计算
"""
pass
@abstractmethod
def layout(self, x: float, y: float, available_width: Optional[float] = None,
available_height: Optional[float] = None) -> BoundingBox:
"""
执行布局计算
Args:
x: 起始 x 坐标
y: 起始 y 坐标
available_width: 可用宽度(可选)
available_height: 可用高度(可选)
Returns:
计算后的边界框
"""
pass
@abstractmethod
def to_dict(self) -> dict:
"""序列化为字典"""
pass
# ============= Leaf Node =============
class LeafNode(Node):
"""叶子节点
叶子节点表示具体的视觉元素,具有固定的尺寸和可选的遮罩
"""
def __init__(self, node_id: str, node_type: NodeType,
width: float, height: float,
mask: Optional[np.ndarray] = None,
metadata: Optional[dict] = None,
alignment: Optional[str] = None,
parent: Optional[Node] = None):
super().__init__(node_id, node_type, alignment, parent)
self.width = width
self.height = height
self.mask = mask # 可选的二值遮罩,用于不规则形状
self.metadata = metadata or {} # 存储额外的元数据(如content, role, src等)
def compute_intrinsic_size(self) -> Tuple[float, float]:
"""叶子节点的固有尺寸就是其宽高"""
return (self.width, self.height)
def layout(self, x: float, y: float, available_width: Optional[float] = None,
available_height: Optional[float] = None) -> BoundingBox:
"""叶子节点的布局很简单,直接放置在指定位置"""
self.bbox = BoundingBox(x, y, self.width, self.height)
return self.bbox
def to_dict(self) -> dict:
result = {
"id": self.id,
"type": self.type.value,
"bbox": {
"x": self.bbox.x if self.bbox else 0,
"y": self.bbox.y if self.bbox else 0,
"width": self.width,
"height": self.height
}
}
# 添加元数据
if self.metadata:
result.update(self.metadata)
return result
# ============= Non-Leaf Node =============
class GroupNode(Node):
"""组节点(非叶子节点)
组节点包含多个子节点,并根据布局类型对子节点进行排列
"""
def __init__(self, node_id: str, layout_type: LayoutType,
children: List[Node], padding: Optional[Padding] = None,
alignment: Optional[str] = None,
parent: Optional[Node] = None):
super().__init__(node_id, NodeType.GROUP, alignment, parent)
self.layout_type = layout_type
self.children = children
# 设置每个子节点的 parent
for child in self.children:
child.parent = self
self.padding = padding or Padding()
self.mask: Optional[np.ndarray] = None # 根据子节点mask合并得到
# 布局特定属性(子类设置)
self.layout_attrs = {}
def compute_intrinsic_size(self) -> Tuple[float, float]:
"""
根据子节点和布局规则计算固有尺寸
这个方法会在具体的布局子类中实现
"""
raise NotImplementedError("Subclass must implement compute_intrinsic_size")
def layout(self, x: float, y: float, available_width: Optional[float] = None,
available_height: Optional[float] = None) -> BoundingBox:
"""
执行组节点的布局
具体的布局算法由子类实现
"""
raise NotImplementedError("Subclass must implement layout")
def _compute_mask_from_children(self):
"""
根据所有子节点的mask计算父节点的mask
将每个子节点的mask根据其bbox位置转换到父节点坐标系,
然后合并所有mask(使用OR操作)
"""
if not self.bbox or not self.children:
self.mask = None
return
# 收集所有有mask的子节点
children_with_mask = [
child for child in self.children
if child.mask is not None and child.bbox is not None
]
if not children_with_mask:
self.mask = None
return
# 创建父节点的mask(全零)
parent_width = int(self.bbox.width)
parent_height = int(self.bbox.height)
parent_mask = np.zeros((parent_height, parent_width), dtype=np.uint8)
# 将每个子节点的mask转换到父节点坐标系并合并
for child in children_with_mask:
child_mask = child.mask
child_bbox = child.bbox
# 计算子节点在父节点坐标系中的位置(相对于父节点左上角)
child_x_in_parent = int(child_bbox.x - self.bbox.x)
child_y_in_parent = int(child_bbox.y - self.bbox.y)
child_width = int(child_bbox.width)
child_height = int(child_bbox.height)
# 确保子节点mask的尺寸与bbox一致
if child_mask.shape != (child_height, child_width):
# 如果尺寸不匹配,调整mask尺寸(使用最近邻插值)
mask_h, mask_w = child_mask.shape
# 计算源坐标
y_coords = np.clip(
(np.arange(child_height) * mask_h / child_height).astype(int),
0, mask_h - 1
)
x_coords = np.clip(
(np.arange(child_width) * mask_w / child_width).astype(int),
0, mask_w - 1
)
# 使用numpy的高级索引进行最近邻插值
y_indices, x_indices = np.meshgrid(y_coords, x_coords, indexing='ij')
child_mask = child_mask[y_indices, x_indices]
# 计算在父节点mask中的位置范围
y_start = max(0, child_y_in_parent)
y_end = min(parent_height, child_y_in_parent + child_height)
x_start = max(0, child_x_in_parent)
x_end = min(parent_width, child_x_in_parent + child_width)
# 计算在子节点mask中的对应范围
child_y_start = max(0, -child_y_in_parent)
child_y_end = child_y_start + (y_end - y_start)
child_x_start = max(0, -child_x_in_parent)
child_x_end = child_x_start + (x_end - x_start)
# 将子节点mask复制到父节点mask的对应位置(OR操作)
if (y_end > y_start and x_end > x_start and
child_y_end > child_y_start and child_x_end > child_x_start):
parent_mask[y_start:y_end, x_start:x_end] = np.maximum(
parent_mask[y_start:y_end, x_start:x_end],
child_mask[child_y_start:child_y_end, child_x_start:child_x_end]
)
self.mask = parent_mask
def to_dict(self) -> dict:
return {
"id": self.id,
"type": self.type.value,
"layout": self.layout_type.value,
"layoutAttrs": self.layout_attrs,
"children": [child.to_dict() for child in self.children],
"bbox": {
"x": self.bbox.x if self.bbox else 0,
"y": self.bbox.y if self.bbox else 0,
"width": self.bbox.width if self.bbox else 0,
"height": self.bbox.height if self.bbox else 0
}
}
|