图片上传识别功能

backend/app/api/agents.py

@@ -76,6 +76,7 @@ class PreviewChatTurnResponse(BaseModel):
     created_at: datetime
     user_text: str
     agent_text: str
+    attachments: List[Dict[str, Any]] = Field(default_factory=list)
 
     class Config:
         from_attributes = True

backend/app/api/uploads.py

@@ -6,15 +6,17 @@ from __future__ import annotations
 import re
 import uuid
 import logging
+import mimetypes
 from pathlib import Path
 
-from fastapi import APIRouter, Depends, File, HTTPException, UploadFile, status
+from fastapi import APIRouter, Depends, File, HTTPException, UploadFile, status, Query
+from fastapi.responses import FileResponse
 from pydantic import BaseModel
 
 from app.api.auth import get_current_user
 from app.core.config import settings
 from app.models.user import User
-from app.services.builtin_tools import _local_file_workspace_root
+from app.services.builtin_tools import _local_file_workspace_root, _resolve_path_under_workspace
 
 logger = logging.getLogger(__name__)
 
@@ -110,3 +112,32 @@ async def upload_preview_file(
         size=total,
         content_type=file.content_type,
     )
+
+
+@router.get("/preview/file")
+async def get_preview_file(
+    file_path: str = Query(..., description="上传后返回的 relative_path"),
+    current_user: User = Depends(get_current_user),
+):
+    """
+    读取当前用户预览附件（用于前端历史回显图片缩略图）。
+    仅允许访问 uploads/preview/<current_user.id>/ 下文件。
+    """
+    path, err = _resolve_path_under_workspace(file_path)
+    if err or path is None:
+        raise HTTPException(status_code=400, detail=f"无效文件路径: {err or file_path}")
+    if not path.is_file():
+        raise HTTPException(status_code=404, detail="文件不存在")
+
+    root = _local_file_workspace_root()
+    try:
+        rel = path.relative_to(root).as_posix()
+    except ValueError:
+        raise HTTPException(status_code=403, detail="不允许访问该文件")
+
+    prefix = f"uploads/preview/{current_user.id}/"
+    if not rel.startswith(prefix):
+        raise HTTPException(status_code=403, detail="无权访问该文件")
+
+    media_type, _ = mimetypes.guess_type(str(path))
+    return FileResponse(path=str(path), media_type=media_type or "application/octet-stream")

backend/app/services/agent_workspace_chat_log.py

@@ -97,6 +97,31 @@ def _user_id_from_input(input_data: Optional[Dict[str, Any]]) -> Optional[str]:
     return None
 
 
+def _extract_attachments(input_data: Optional[Dict[str, Any]]) -> List[Dict[str, Any]]:
+    if not input_data:
+        return []
+    raw = input_data.get("attachments")
+    if not isinstance(raw, list):
+        return []
+    out: List[Dict[str, Any]] = []
+    for item in raw:
+        if not isinstance(item, dict):
+            continue
+        rel = str(item.get("relative_path") or "").strip()
+        name = str(item.get("filename") or "").strip()
+        if not rel:
+            continue
+        content_type = str(item.get("content_type") or "").strip() or None
+        out.append(
+            {
+                "relative_path": rel,
+                "filename": name or rel.rsplit("/", 1)[-1],
+                "content_type": content_type,
+            }
+        )
+    return out
+
+
 def fetch_agent_preview_chat_turns(
     db: Session,
     agent_id: str,
@@ -149,6 +174,7 @@ def fetch_agent_preview_chat_turns(
                 "created_at": ex.created_at,
                 "user_text": user_text or "（无文本）",
                 "agent_text": agent_text or "（无输出）",
+                "attachments": _extract_attachments(inp),
             }
         )
     return out

backend/scripts/check_ocr_env.py

@@ -0,0 +1,58 @@
+"""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())