From 31a4e547640844f4a6332df807dd8ca03da475a3 Mon Sep 17 00:00:00 2001 From: Wenjie Zhang Date: Tue, 2 Dec 2025 21:17:32 +0800 Subject: [PATCH] =?UTF-8?q?refactor:=20=E4=BC=98=E5=8C=96=E6=96=87?= =?UTF-8?q?=E4=BB=B6=E4=B8=8B=E8=BD=BD=EF=BC=8C=E7=A7=BB=E9=99=A4=E5=8E=9F?= =?UTF-8?q?=E6=9C=89=E9=80=BB=E8=BE=91?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 清理minio客户端导入顺序 移除已处理的TODO注释和废弃代码 --- server/routers/knowledge_router.py | 23 +---------------------- src/knowledge/indexing.py | 5 ----- src/knowledge/utils/kb_utils.py | 2 +- src/storage/minio/client.py | 4 ++-- 4 files changed, 4 insertions(+), 30 deletions(-) diff --git a/server/routers/knowledge_router.py b/server/routers/knowledge_router.py index a109e5f7..118188af 100644 --- a/server/routers/knowledge_router.py +++ b/server/routers/knowledge_router.py @@ -516,29 +516,10 @@ async def rechunks_documents( @knowledge.get("/databases/{db_id}/documents/{doc_id}/download") async def download_document(db_id: str, doc_id: str, request: Request, current_user: User = Depends(get_admin_user)): - # TODO: 可以考虑修改为minio下载，将文件相关逻辑完全迁移到minio """下载原始文件""" logger.debug(f"Download document {doc_id} from {db_id}") try: file_info = await knowledge_base.get_file_basic_info(db_id, doc_id) - if not file_info: - raise HTTPException(status_code=404, detail="File not found") - - file_path = file_info.get("meta", {}).get("path") - if not file_path: - raise HTTPException(status_code=404, detail="File path not found in metadata") - - # 安全检查：验证文件路径 - from src.knowledge.utils.kb_utils import validate_file_path - - try: - normalized_path = validate_file_path(file_path, db_id) - except ValueError as e: - raise HTTPException(status_code=403, detail=str(e)) - - if not os.path.exists(normalized_path): - raise HTTPException(status_code=404, detail=f"File not found on disk: {file_info=}") - # 获取文件扩展名和MIME类型，解码URL编码的文件名 filename = file_info.get("meta", {}).get("filename", "file") logger.debug(f"Original filename from database: {filename}") @@ -1075,9 +1056,7 @@ async def upload_file( basename, ext = os.path.splitext(file.filename) # TODO: - # 如果知识库中的文件多了，上传了内容修改过的同名文件应当把旧的文件删除掉 - # 否则会保存两份相同的文档，建议固定salt，上传逻辑是： - # 若上传了同名文件时且hash相同则报错，不同则直接替换同名文件 + # 后续修改为遇到同名文件则在上传区域提示，是否删除旧文件，同时 filename name 也就不用添加 hash 了 filename = f"{basename}_{hashstr(basename, 4, with_salt=True, salt='fixed_salt')}{ext}".lower() file_path = os.path.join(upload_dir, filename) diff --git a/src/knowledge/indexing.py b/src/knowledge/indexing.py index 996277c0..98b703ea 100644 --- a/src/knowledge/indexing.py +++ b/src/knowledge/indexing.py @@ -371,11 +371,6 @@ async def process_file_to_markdown(file_path: str, params: dict | None = None) - return markdown_content.strip() - # TODO: - # 此处修改了excel的处理逻辑，原本的excel是转为markdown后切分 - # 但是实际使用时发现，对于excel这种结构化数据，保留表头非常有必要 - # 因此改为了每10行重复保存一次表头 - # 目前前端显示有点问题，不知道为啥不能换行 elif file_ext in [".xls", ".xlsx"]: # 处理 Excel 文件 import pandas as pd diff --git a/src/knowledge/utils/kb_utils.py b/src/knowledge/utils/kb_utils.py index 93f334c7..0b06459f 100644 --- a/src/knowledge/utils/kb_utils.py +++ b/src/knowledge/utils/kb_utils.py @@ -90,7 +90,7 @@ def split_text_into_chunks(text: str, file_id: str, filename: str, params: dict chunks.append( { "id": f"{file_id}_chunk_{chunk_index}", - "content": chunk_content, # .strip(), + "content": chunk_content, # .strip(), "file_id": file_id, "filename": filename, "chunk_index": chunk_index, diff --git a/src/storage/minio/client.py b/src/storage/minio/client.py index afb899d1..63154161 100644 --- a/src/storage/minio/client.py +++ b/src/storage/minio/client.py @@ -10,10 +10,10 @@ import uuid from datetime import timedelta from io import BytesIO -from minio import Minio -from minio.error import S3Error from urllib3 import BaseHTTPResponse +from minio import Minio +from minio.error import S3Error from src.utils import logger