init

utils/__init__.py

@@ -0,0 +1 @@
+

utils/config.py

@@ -0,0 +1,41 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Any, Optional
+
+import yaml
+
+
+@dataclass(frozen=True)
+class AppConfig:
+    umi_ocr_url: str
+    prefer_mss: bool = True
+    default_region: Optional[tuple[int, int, int, int]] = None  # left, top, width, height
+    click_pause: float = 0.05
+
+
+def load_config(path: str | Path) -> AppConfig:
+    p = Path(path)
+    data: dict[str, Any] = yaml.safe_load(p.read_text(encoding="utf-8")) or {}
+
+    umi_ocr = data.get("umi_ocr", {}) or {}
+    screenshot = data.get("screenshot", {}) or {}
+    click = data.get("click", {}) or {}
+
+    region = screenshot.get("default_region")
+    default_region: Optional[tuple[int, int, int, int]]
+    if region is None:
+        default_region = None
+    else:
+        if not (isinstance(region, (list, tuple)) and len(region) == 4):
+            raise ValueError("config.yaml screenshot.default_region 必须是 null 或 [left, top, width, height]")
+        default_region = (int(region[0]), int(region[1]), int(region[2]), int(region[3]))
+
+    return AppConfig(
+        umi_ocr_url=str(umi_ocr.get("url", "http://127.0.0.1:1224/api/ocr")),
+        prefer_mss=bool(screenshot.get("prefer_mss", True)),
+        default_region=default_region,
+        click_pause=float(click.get("pause", 0.05)),
+    )
+

utils/logger.py

@@ -0,0 +1,18 @@
+from __future__ import annotations
+
+import logging
+from typing import Optional
+
+
+def setup_logging(level: int = logging.INFO, log_file: Optional[str] = None) -> logging.Logger:
+    handlers: list[logging.Handler] = [logging.StreamHandler()]
+    if log_file:
+        handlers.append(logging.FileHandler(log_file, encoding="utf-8"))
+
+    logging.basicConfig(
+        level=level,
+        format="%(asctime)s | %(levelname)s | %(name)s | %(message)s",
+        handlers=handlers,
+    )
+    return logging.getLogger("yidaima")
+

utils/screenshot.py

@@ -0,0 +1,79 @@
+from __future__ import annotations
+
+import io
+from dataclasses import dataclass
+from typing import Optional
+
+
+@dataclass(frozen=True)
+class CaptureResult:
+    image_bytes: bytes
+    # 截图区域在“屏幕坐标系”中的位置（left, top, width, height）
+    region: tuple[int, int, int, int]
+    # 截图图片像素坐标 -> 屏幕坐标 的缩放系数
+    scale_x: float
+    scale_y: float
+
+
+def capture_screen(region: Optional[tuple[int, int, int, int]] = None, *, prefer_mss: bool = True) -> CaptureResult:
+    """
+    截取屏幕（全屏或局部），返回 PNG bytes + region + 像素到屏幕坐标的缩放。
+
+    - region: (left, top, width, height) 使用屏幕坐标
+    - scale_x/scale_y: 用于处理高 DPI 下 mss 取到的物理像素与 pyautogui 屏幕坐标不一致的问题
+    """
+    if prefer_mss:
+        try:
+            return _capture_with_mss(region)
+        except Exception:
+            # 回退到 pyautogui（更贴近“屏幕坐标系”）
+            return _capture_with_pyautogui(region)
+    return _capture_with_pyautogui(region)
+
+
+def _capture_with_pyautogui(region: Optional[tuple[int, int, int, int]]) -> CaptureResult:
+    import pyautogui
+
+    img = pyautogui.screenshot(region=region)  # type: ignore[arg-type]
+    buf = io.BytesIO()
+    img.save(buf, format="PNG")
+
+    if region is None:
+        w, h = img.size
+        region2 = (0, 0, int(w), int(h))
+    else:
+        region2 = (int(region[0]), int(region[1]), int(region[2]), int(region[3]))
+
+    return CaptureResult(image_bytes=buf.getvalue(), region=region2, scale_x=1.0, scale_y=1.0)
+
+
+def _capture_with_mss(region: Optional[tuple[int, int, int, int]]) -> CaptureResult:
+    import mss
+    import mss.tools
+    import pyautogui
+
+    with mss.mss() as sct:
+        if region is None:
+            mon = sct.monitors[1]  # 主屏
+            left, top, width, height = int(mon["left"]), int(mon["top"]), int(mon["width"]), int(mon["height"])
+        else:
+            left, top, width, height = (int(region[0]), int(region[1]), int(region[2]), int(region[3]))
+
+        shot = sct.grab({"left": left, "top": top, "width": width, "height": height})
+        png_bytes = mss.tools.to_png(shot.rgb, shot.size)
+
+        # DPI 缩放处理：mss 的像素尺寸可能与 pyautogui 的屏幕坐标尺寸不同
+        screen_w, screen_h = pyautogui.size()
+        mon = sct.monitors[1]
+        mon_w, mon_h = int(mon["width"]), int(mon["height"])
+        # 若 mss 返回的是物理像素，mon_w/mon_h 往往大于 pyautogui.size()
+        scale_x = float(screen_w) / float(mon_w) if mon_w else 1.0
+        scale_y = float(screen_h) / float(mon_h) if mon_h else 1.0
+
+        return CaptureResult(
+            image_bytes=png_bytes,
+            region=(left, top, width, height),
+            scale_x=scale_x,
+            scale_y=scale_y,
+        )
+