59 lines
1.9 KiB
Python
59 lines
1.9 KiB
Python
"""检查图片 OCR 环境：Pillow、pytesseract、Tesseract 可执行文件、chi_sim 语言包。
|
||
|
||
在 backend 目录执行:
|
||
.\\venv\\Scripts\\python scripts\\check_ocr_env.py
|
||
"""
|
||
from __future__ import annotations
|
||
|
||
import sys
|
||
from pathlib import Path
|
||
|
||
# 保证能加载 app 配置
|
||
sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
|
||
|
||
from app.core.config import settings # noqa: E402
|
||
from app.services.builtin_tools import _local_file_workspace_root, _tessdata_dir_for_ocr # noqa: E402
|
||
|
||
|
||
def main() -> int:
|
||
print("TESSERACT_CMD (settings):", settings.TESSERACT_CMD or "(空，将尝试 PATH)")
|
||
print("TESSERACT_TESSDATA_DIR (settings):", settings.TESSERACT_TESSDATA_DIR or "(空，将尝试仓库 tessdata/)")
|
||
try:
|
||
import PIL # noqa: F401
|
||
|
||
print("Pillow: OK")
|
||
except ImportError as e:
|
||
print("Pillow: 缺失 —", e)
|
||
print(" 请执行: pip install Pillow")
|
||
return 2
|
||
try:
|
||
import pytesseract as pt
|
||
|
||
print("pytesseract: OK")
|
||
except ImportError as e:
|
||
print("pytesseract: 缺失 —", e)
|
||
print(" 请执行: pip install pytesseract")
|
||
return 3
|
||
cmd = (settings.TESSERACT_CMD or "").strip()
|
||
if cmd:
|
||
pt.pytesseract.tesseract_cmd = cmd
|
||
try:
|
||
ver = pt.get_tesseract_version()
|
||
print("Tesseract 版本:", ver)
|
||
except Exception as e:
|
||
print("Tesseract 可执行文件: 不可用 —", e)
|
||
print(" Windows 请安装 Tesseract，并在 .env 设置 TESSERACT_CMD=.../tesseract.exe")
|
||
return 4
|
||
td = _tessdata_dir_for_ocr()
|
||
print("解析到的 tessdata 目录:", td or "(未找到)")
|
||
root = _local_file_workspace_root()
|
||
loc = root / "tessdata"
|
||
if loc.is_dir():
|
||
has_chi = any(loc.glob("chi_sim.traineddata"))
|
||
print("仓库 tessdata/chi_sim.traineddata:", "存在" if has_chi else "缺失（中文识别差）")
|
||
return 0
|
||
|
||
|
||
if __name__ == "__main__":
|
||
raise SystemExit(main())
|