refactor: 项目配置重构及 OCR 功能增强

- 更新 .gitignore，忽略本地文件。
- 调整 docker-compose.yml，优化服务配置及端口设置。
- api.Dockerfile 更换为基于 NVIDIA CUDA 的镜像，支持 GPU。
- pyproject.toml 新增依赖 langchain-tavily，并升级 mineru。
- tools_factory.py 重构，采用 langchain_tavily 的 TavilySearch。
- knowledgebase.py 增强 Milvus 连接异常的错误日志。
- _ocr.py 中 OCRPlugin 升级，使用 mineru 新解析方法，并调整健康检查 URL。
- 新增 mineru.py，实现文档解析逻辑。
- logging_config.py 优化日志配置，采用 loguru 管理日志。
- DataBaseInfoView.vue 新增分块参数配置及文件上传相关弹窗和功能。

.gitignore

@@ -20,7 +20,6 @@ logs
 *.log.*
 *.db
 *.lock
-!uv.lock
 tmp
 cache
 
@@ -29,6 +28,7 @@ cache
 .idea
 *.nogit*
 *.private*
+*.local*
 *.secret*
 
 *.pdf

docker-compose.yml

@@ -3,6 +3,9 @@ services:
     build:
       context: .
       dockerfile: docker/api.Dockerfile
+      # args:
+      #   http_proxy: http://172.19.13.5:7890
+      #   https_proxy: http://172.19.13.5:7890
     image: yuxi-api:0.1.0
     container_name: api-dev
     working_dir: /app
@@ -16,7 +19,7 @@ services:
         reservations:
           devices:
             - driver: nvidia
-              device_ids: ['1']
+              device_ids: ['0']  # 使用GPU 1
               capabilities: [gpu]
     ports:
       - "5050:5050"
@@ -28,8 +31,8 @@ services:
       - NEO4J_URI=bolt://graph:7687
       - NEO4J_USERNAME=${NEO4J_USERNAME:-neo4j}
       - NEO4J_PASSWORD=${NEO4J_PASSWORD:-0123456789}
-      - MILVUS_URI=http://milvus:19530
-      - MINERU_OCR_URI=http://mineru-api:5051
+      - MILVUS_URI=http://127.0.0.1:19530
+      - MINERU_OCR_URI=http://127.0.0.1:30000
       - MODEL_DIR=/models
       - RUNNING_IN_DOCKER=true
     command: uv run uvicorn server.main:app --host 0.0.0.0 --port 5050 --reload
@@ -143,61 +146,37 @@ services:
     networks:
       - app-network
     restart: unless-stopped
-
-  mineru-api:
+  # lastest version: wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/docker/compose.yaml
+  mineru-sglang:
     build:
-      context: scripts/mineru-api
+      context: scripts/mineru-sglang
       dockerfile: Dockerfile
-    image: mineru-api:latest
-    container_name: mineru-api
-    profiles:
-      - all
+    image: mineru-sglang:latest
+    container_name: mineru-sglang
+    ports:
+      - 30000:30000
+    environment:
+      MINERU_MODEL_SOURCE: modelscope
+    entrypoint: mineru-sglang-server
+    command:
+      --host 0.0.0.0
+      --port 30000
+    ulimits:
+      memlock: -1
+      stack: 67108864
+    ipc: host
+    healthcheck:
+      test: ["CMD-SHELL", "curl -f http://localhost:30000/health || exit 1"]
     deploy:
       resources:
         reservations:
           devices:
             - driver: nvidia
-              device_ids: ['0']
+              device_ids: ["0"]
               capabilities: [gpu]
-    ports:
-      - "5051:5051"
     networks:
       - app-network
     restart: unless-stopped
-    command: python -m uvicorn app:app --host 0.0.0.0 --port 5051 --app-dir /app
-
-
-  # # lastest version: wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/docker/compose.yaml
-  # mineru-sglang:
-  #   build:
-  #     context: scripts/mineru-sglang
-  #     dockerfile: Dockerfile
-  #   image: mineru-sglang:latest
-  #   container_name: mineru-sglang
-  #   ports:
-  #     - 30000:30000
-  #   environment:
-  #     MINERU_MODEL_SOURCE: modelscope
-  #   entrypoint: mineru-sglang-server
-  #   command:
-  #     --host 0.0.0.0
-  #     --port 30000
-  #   ulimits:
-  #     memlock: -1
-  #     stack: 67108864
-  #   ipc: host
-  #   healthcheck:
-  #     test: ["CMD-SHELL", "curl -f http://localhost:30000/health || exit 1"]
-  #   deploy:
-  #     resources:
-  #       reservations:
-  #         devices:
-  #           - driver: nvidia
-  #             device_ids: ["0"]
-  #             capabilities: [gpu]
-  #   networks:
-  #     - app-network
-  #   restart: unless-stopped
 
 networks:
   app-network:

docker/api.Dockerfile

@@ -1,5 +1,5 @@
 # 使用基础镜像
-FROM python:3.12
+FROM nvidia/cuda:12.4.1-cudnn-devel-ubuntu22.04
 COPY --from=ghcr.io/astral-sh/uv:0.7.2 /uv /uvx /bin/
 
 # 设置工作目录

pyproject.toml

@@ -14,6 +14,7 @@ dependencies = [
     "langchain-deepseek>=0.1.3",
     "langchain-huggingface>=0.2.0",
     "langchain-openai>=0.3.14",
+    "langchain-tavily>=0.2.3",
     "langchain-together>=0.3.0",
     "langgraph>=0.3.34",
     "langgraph-checkpoint-sqlite>=2.0.7",
@@ -21,6 +22,7 @@ dependencies = [
     "langsmith>=0.3.37",
     "llama-index>=0.12.33",
     "llama-index-readers-file>=0.4.7",
+    "mineru[all]>=2.0.3",
     "neo4j>=5.28.1",
     "openai>=1.76.0",
     "opencv-python-headless>=4.11.0.86",

src/agents/tools_factory.py

@@ -3,7 +3,7 @@ import re
 from collections.abc import Callable
 from typing import Annotated, Any
 
-from langchain_community.tools.tavily_search import TavilySearchResults
+from langchain_tavily import TavilySearch
 from langchain_core.tools import BaseTool, StructuredTool, tool
 from pydantic import BaseModel, Field
 
@@ -151,4 +151,4 @@ _TOOLS_REGISTRY = {
 }
 
 if config.enable_web_search:
-    _TOOLS_REGISTRY["WebSearchWithTavily"] = TavilySearchResults(max_results=10)
+    _TOOLS_REGISTRY["WebSearchWithTavily"] = TavilySearch(max_results=10)

src/core/knowledgebase.py

@@ -668,11 +668,11 @@ class KnowledgeBase:
             logger.info(f"Successfully connected to Milvus at {uri}")
             return True
         except MilvusException as e:
-            logger.error(f"Failed to connect to Milvus: {e}，请检查 milvus 的容器是否正常运行。")
-            logger.error("如果已退出，请重新启动 `docker restart milvus-standalone-dev`。")
+            logger.error(f"Failed to connect to Milvus: {e}，请检查 milvus 的容器是否正常运行。{traceback.format_exc()}")
+            logger.error("如果已退出，请重新启动 `docker restart milvus`。")
             return False
         except Exception as e: # Catch other potential errors like requests.exceptions.ConnectionError
-            logger.error(f"An unexpected error occurred while connecting to Milvus at {uri}: {e}")
+            logger.error(f"An unexpected error occurred while connecting to Milvus at {uri}: {e}, {traceback.format_exc()}")
             return False
 
 

src/plugins/_ocr.py

@@ -159,53 +159,26 @@ class OCRPlugin:
         :param pdf_path: PDF文件路径
         :return: 提取的文本
         """
-        mineru_ocr_uri = os.getenv("MINERU_OCR_URI", "http://localhost:5051")
+        mineru_ocr_uri = os.getenv("MINERU_OCR_URI", "http://localhost:30000")
+        mineru_ocr_uri_health = f"{mineru_ocr_uri}/health"
         import requests
         import json
+        from .mineru import parse_doc
 
-        health_check_response = requests.get(f"{mineru_ocr_uri}/health", timeout=5)
-        if health_check_response.status_code != 200 or health_check_response.json().get("status") != "healthy":
-            logger.error("Mineru OCR service health check failed.")
+        health_check_response = requests.get(mineru_ocr_uri_health, timeout=5)
+        if health_check_response.status_code != 200:
+            logger.error(f"Mineru OCR service health check failed with {mineru_ocr_uri_health}: {health_check_response.json()}")
             raise RuntimeError("Mineru OCR service health check failed. Please check the log use `docker logs mineru-api`")
 
-        # 读取PDF文件
-        with open(pdf_path, 'rb') as f:
-            files = {'file': f}
-            data = {
-                'parse_method': 'ocr',  # 使用OCR模式
-                'is_json_md_dump': False,  # 不需要保存中间文件
-                'return_layout': False,  # 不需要返回布局信息
-                'return_info': False,  # 不需要返回额外信息
-                'return_content_list': False,  # 不需要返回内容列表
-                'return_images': False,  # 不需要返回图片
-            }
+        pdf_path_list = [pdf_path]
+        output_dir = os.path.join(os.getcwd(), "tmp", "mineru_ocr")
 
-            try:
-                # 发送POST请求到Mineru OCR服务
-                response = requests.post(
-                    f"{mineru_ocr_uri}/file_parse",
-                    files=files,
-                    data=data
-                )
-                response.raise_for_status()  # 检查响应状态
+        pdf_text = parse_doc(pdf_path_list, output_dir,
+                         backend="vlm-sglang-client",
+                         server_url=mineru_ocr_uri)[0]
 
-                # 解析响应
-                result = response.json()
-                if 'md_content' in result:
-                    return result['md_content']
-                else:
-                    logger.error("Mineru OCR response does not contain md_content")
-                    return ""
-
-            except requests.exceptions.RequestException as e:
-                logger.error(f"Mineru OCR request failed: {str(e)}")
-                return ""
-            except json.JSONDecodeError as e:
-                logger.error(f"Failed to parse Mineru OCR response: {str(e)}")
-                return ""
-            except Exception as e:
-                logger.error(f"Unexpected error in Mineru OCR processing: {str(e)}")
-                return ""
+        logger.debug(f"Mineru OCR result: {pdf_text[:50]}(...) total {len(pdf_text)} characters.")
+        return pdf_text
 
 def get_state(task_id):
     return GOLBAL_STATE.get(task_id, {})

src/plugins/mineru.py

@@ -0,0 +1,247 @@
+# Copyright (c) Opendatalab. All rights reserved.
+import copy
+import json
+import os
+from pathlib import Path
+from tqdm import tqdm
+
+from mineru.cli.common import convert_pdf_bytes_to_bytes_by_pypdfium2, prepare_env, read_fn
+from mineru.data.data_reader_writer import FileBasedDataWriter
+from mineru.utils.draw_bbox import draw_layout_bbox, draw_span_bbox
+from mineru.utils.enum_class import MakeMode
+from mineru.backend.vlm.vlm_analyze import doc_analyze as vlm_doc_analyze
+from mineru.backend.pipeline.pipeline_analyze import doc_analyze as pipeline_doc_analyze
+from mineru.backend.pipeline.pipeline_middle_json_mkcontent import union_make as pipeline_union_make
+from mineru.backend.pipeline.model_json_to_middle_json import result_to_middle_json as pipeline_result_to_middle_json
+from mineru.backend.vlm.vlm_middle_json_mkcontent import union_make as vlm_union_make
+
+from src.utils.logging_config import logger
+
+
+def do_parse(
+    output_dir,  # Output directory for storing parsing results
+    pdf_file_names: list[str],  # List of PDF file names to be parsed
+    pdf_bytes_list: list[bytes],  # List of PDF bytes to be parsed
+    p_lang_list: list[str],  # List of languages for each PDF, default is 'ch' (Chinese)
+    backend="pipeline",  # The backend for parsing PDF, default is 'pipeline'
+    parse_method="auto",  # The method for parsing PDF, default is 'auto'
+    p_formula_enable=True,  # Enable formula parsing
+    p_table_enable=True,  # Enable table parsing
+    server_url=None,  # Server URL for vlm-sglang-client backend
+    f_draw_layout_bbox=True,  # Whether to draw layout bounding boxes
+    f_draw_span_bbox=True,  # Whether to draw span bounding boxes
+    f_dump_md=True,  # Whether to dump markdown files
+    f_dump_middle_json=True,  # Whether to dump middle JSON files
+    f_dump_model_output=True,  # Whether to dump model output files
+    f_dump_orig_pdf=True,  # Whether to dump original PDF files
+    f_dump_content_list=True,  # Whether to dump content list files
+    f_make_md_mode=MakeMode.MM_MD,  # The mode for making markdown content, default is MM_MD
+    start_page_id=0,  # Start page ID for parsing, default is 0
+    end_page_id=None,  # End page ID for parsing, default is None (parse all pages until the end of the document)
+) -> list[str]:
+
+    if backend == "pipeline":
+        for idx, pdf_bytes in enumerate(pdf_bytes_list):
+            new_pdf_bytes = convert_pdf_bytes_to_bytes_by_pypdfium2(pdf_bytes, start_page_id, end_page_id)
+            pdf_bytes_list[idx] = new_pdf_bytes
+
+        infer_results, all_image_lists, all_pdf_docs, lang_list, ocr_enabled_list = pipeline_doc_analyze(pdf_bytes_list, p_lang_list, parse_method=parse_method, formula_enable=p_formula_enable,table_enable=p_table_enable)
+
+
+        md_results = []
+        for idx, model_list in enumerate(infer_results):
+            model_json = copy.deepcopy(model_list)
+            pdf_file_name = pdf_file_names[idx]
+            local_image_dir, local_md_dir = prepare_env(output_dir, pdf_file_name, parse_method)
+            image_writer, md_writer = FileBasedDataWriter(local_image_dir), FileBasedDataWriter(local_md_dir)
+
+            images_list = all_image_lists[idx]
+            pdf_doc = all_pdf_docs[idx]
+            _lang = lang_list[idx]
+            _ocr_enable = ocr_enabled_list[idx]
+            middle_json = pipeline_result_to_middle_json(model_list, images_list, pdf_doc, image_writer, _lang, _ocr_enable, p_formula_enable)
+
+            pdf_info = middle_json["pdf_info"]
+
+            pdf_bytes = pdf_bytes_list[idx]
+            if f_draw_layout_bbox:
+                draw_layout_bbox(pdf_info, pdf_bytes, local_md_dir, f"{pdf_file_name}_layout.pdf")
+
+            if f_draw_span_bbox:
+                draw_span_bbox(pdf_info, pdf_bytes, local_md_dir, f"{pdf_file_name}_span.pdf")
+
+            if f_dump_orig_pdf:
+                md_writer.write(
+                    f"{pdf_file_name}_origin.pdf",
+                    pdf_bytes,
+                )
+
+            if f_dump_md:
+                image_dir = str(os.path.basename(local_image_dir))
+                md_content_str = pipeline_union_make(pdf_info, f_make_md_mode, image_dir)
+                md_writer.write_string(
+                    f"{pdf_file_name}.md",
+                    md_content_str,
+                )
+                md_results.append(md_content_str)
+
+            if f_dump_content_list:
+                image_dir = str(os.path.basename(local_image_dir))
+                content_list = pipeline_union_make(pdf_info, MakeMode.CONTENT_LIST, image_dir)
+                md_writer.write_string(
+                    f"{pdf_file_name}_content_list.json",
+                    json.dumps(content_list, ensure_ascii=False, indent=4),
+                )
+
+            if f_dump_middle_json:
+                md_writer.write_string(
+                    f"{pdf_file_name}_middle.json",
+                    json.dumps(middle_json, ensure_ascii=False, indent=4),
+                )
+
+            if f_dump_model_output:
+                md_writer.write_string(
+                    f"{pdf_file_name}_model.json",
+                    json.dumps(model_json, ensure_ascii=False, indent=4),
+                )
+
+            logger.info(f"local output dir is {local_md_dir}")
+
+        return md_results
+
+    else:
+        if backend.startswith("vlm-"):
+            backend = backend[4:]
+
+        f_draw_span_bbox = False
+        parse_method = "vlm"
+        md_results = []
+
+        for idx, pdf_bytes in enumerate(tqdm(pdf_bytes_list, desc="Parsing documents bytes")):
+            pdf_file_name = pdf_file_names[idx]
+            pdf_bytes = convert_pdf_bytes_to_bytes_by_pypdfium2(pdf_bytes, start_page_id, end_page_id)
+            local_image_dir, local_md_dir = prepare_env(output_dir, pdf_file_name, parse_method)
+            image_writer, md_writer = FileBasedDataWriter(local_image_dir), FileBasedDataWriter(local_md_dir)
+            middle_json, infer_result = vlm_doc_analyze(pdf_bytes, image_writer=image_writer, backend=backend, server_url=server_url)
+
+            pdf_info = middle_json["pdf_info"]
+
+            if f_draw_layout_bbox:
+                draw_layout_bbox(pdf_info, pdf_bytes, local_md_dir, f"{pdf_file_name}_layout.pdf")
+
+            if f_draw_span_bbox:
+                draw_span_bbox(pdf_info, pdf_bytes, local_md_dir, f"{pdf_file_name}_span.pdf")
+
+            if f_dump_orig_pdf:
+                md_writer.write(
+                    f"{pdf_file_name}_origin.pdf",
+                    pdf_bytes,
+                )
+
+            if f_dump_md:
+                image_dir = str(os.path.basename(local_image_dir))
+                md_content_str = vlm_union_make(pdf_info, f_make_md_mode, image_dir)
+                md_writer.write_string(
+                    f"{pdf_file_name}.md",
+                    md_content_str,
+                )
+                md_results.append(md_content_str)
+
+            if f_dump_content_list:
+                image_dir = str(os.path.basename(local_image_dir))
+                content_list = vlm_union_make(pdf_info, MakeMode.CONTENT_LIST, image_dir)
+                md_writer.write_string(
+                    f"{pdf_file_name}_content_list.json",
+                    json.dumps(content_list, ensure_ascii=False, indent=4),
+                )
+
+            if f_dump_middle_json:
+                md_writer.write_string(
+                    f"{pdf_file_name}_middle.json",
+                    json.dumps(middle_json, ensure_ascii=False, indent=4),
+                )
+
+            if f_dump_model_output:
+                model_output = ("\n" + "-" * 50 + "\n").join(infer_result)
+                md_writer.write_string(
+                    f"{pdf_file_name}_model_output.txt",
+                    model_output,
+                )
+
+            logger.info(f"local output dir is {local_md_dir}")
+
+        return md_results
+
+
+def parse_doc(
+        path_list: list[Path],
+        output_dir,
+        lang="ch",
+        backend="pipeline",
+        method="auto",
+        server_url=None,
+        start_page_id=0,  # Start page ID for parsing, default is 0
+        end_page_id=None  # End page ID for parsing, default is None (parse all pages until the end of the document)
+) -> list[str]:
+    """
+        Parameter description:
+        path_list: List of document paths to be parsed, can be PDF or image files.
+        output_dir: Output directory for storing parsing results.
+        lang: Language option, default is 'ch', optional values include['ch', 'ch_server', 'ch_lite', 'en', 'korean', 'japan', 'chinese_cht', 'ta', 'te', 'ka']。
+            Input the languages in the pdf (if known) to improve OCR accuracy.  Optional.
+            Adapted only for the case where the backend is set to "pipeline"
+        backend: the backend for parsing pdf:
+            pipeline: More general.
+            vlm-transformers: More general.
+            vlm-sglang-engine: Faster(engine).
+            vlm-sglang-client: Faster(client).
+            without method specified, pipeline will be used by default.
+        method: the method for parsing pdf:
+            auto: Automatically determine the method based on the file type.
+            txt: Use text extraction method.
+            ocr: Use OCR method for image-based PDFs.
+            Without method specified, 'auto' will be used by default.
+            Adapted only for the case where the backend is set to "pipeline".
+        server_url: When the backend is `sglang-client`, you need to specify the server_url, for example:`http://127.0.0.1:30000`
+    """
+    try:
+        file_name_list = []
+        pdf_bytes_list = []
+        lang_list = []
+        for path in tqdm(path_list, desc="Parsing documents"):
+            file_name = str(Path(path).stem)
+            pdf_bytes = read_fn(path)
+            file_name_list.append(file_name)
+            pdf_bytes_list.append(pdf_bytes)
+            lang_list.append(lang)
+
+        result = do_parse(
+            output_dir=output_dir,
+            pdf_file_names=file_name_list,
+            pdf_bytes_list=pdf_bytes_list,
+            p_lang_list=lang_list,
+            backend=backend,
+            parse_method=method,
+            server_url=server_url,
+            start_page_id=start_page_id,
+            end_page_id=end_page_id
+        )
+        return result if result else [""]
+
+    except Exception as e:
+        logger.exception(e)
+        return [""]
+
+if __name__ == "__main__":
+
+    pdf_files_dir = "/home/zwj/workspace/projects/Yuxi-Know/test/struct_pdf"
+    output_dir = "/home/zwj/workspace/projects/Yuxi-Know/test/struct_pdf_output"
+    pdf_suffixes = [".pdf"]
+    image_suffixes = [".png", ".jpeg", ".jpg"]
+
+    doc_path_list = []
+    for doc_path in Path(pdf_files_dir).glob('*'):
+        if doc_path.suffix in pdf_suffixes + image_suffixes:
+            doc_path_list.append(doc_path)
+
+    parse_doc(doc_path_list, output_dir, backend="vlm-sglang-client", server_url="http://172.19.13.5:30000")  # faster(client).

src/utils/logging_config.py

@@ -3,52 +3,43 @@ import os
 from datetime import datetime
 
 import pytz
-from colorlog import ColoredFormatter
+
+from loguru import logger
 
 DATETIME = datetime.now(pytz.timezone('Asia/Shanghai')).strftime('%Y-%m-%d-%H%M%S')
 # DATETIME = "debug" # 为了方便，调试的时候输出到 debug.log 文件
 LOG_FILE = f'saves/log/project-{DATETIME}.log'
-
-def setup_logger(name, level=logging.DEBUG, console=True):
+def setup_logger(name, level="DEBUG", console=True):
+    """使用 loguru 设置日志记录器"""
     os.makedirs("saves/log", exist_ok=True)
-
-    """Function to setup logger with the given name and log file."""
-    logger = logging.getLogger(name)
-    logger.setLevel(level)
-
-    # 清除已有的 Handler，防止重复添加
-    if logger.hasHandlers():
-        logger.handlers.clear()
-
-    # 文件日志（无颜色）
-    file_handler = logging.FileHandler(LOG_FILE, encoding='utf-8')
-    file_handler.setLevel(level)
-    file_formatter = logging.Formatter('%(asctime)s - %(levelname)s - %(name)s - %(message)s')
-    file_handler.setFormatter(file_formatter)
-    logger.addHandler(file_handler)
-
-    # 控制台日志（有颜色）
+    
+    # 移除默认的 handler
+    logger.remove()
+    
+    # 添加文件日志（无颜色）
+    logger.add(
+        LOG_FILE,
+        level=level,
+        format="{time:YYYY-MM-DD HH:mm:ss} - {level} - {name} - {message}",
+        encoding="utf-8",
+        rotation="10 MB",  # 文件大小达到 10MB 时轮转
+        retention="30 days",  # 保留30天的日志
+        compression="zip"  # 压缩旧日志文件
+    )
+    
+    # 添加控制台日志（有颜色）
     if console:
-        console_handler = logging.StreamHandler()
-        console_handler.setLevel(level)
-        color_formatter = ColoredFormatter(
-            "%(log_color)s%(asctime)s - %(levelname)s - %(name)s - %(message)s",
-            datefmt="%Y-%m-%d %H:%M:%S",
-            log_colors={
-                'DEBUG': 'cyan',
-                'INFO': 'green',
-                'WARNING': 'yellow',
-                'ERROR': 'red',
-                'CRITICAL': 'bold_red',
-            }
+        logger.add(
+            lambda msg: print(msg, end=""),
+            level=level,
+            format="<green>{time:YYYY-MM-DD HH:mm:ss}</green> - <level>{level}</level> - <cyan>{name}</cyan> - <level>{message}</level>",
+            colorize=True
         )
-        console_handler.setFormatter(color_formatter)
-        logger.addHandler(console_handler)
-
+    
     return logger
 
 
-# Setup the root logger
+# 设置根日志记录器
 logger = setup_logger('Yuxi')
 
 # If you want to disable logging from external libraries

web/src/views/DataBaseInfoView.vue

@@ -41,6 +41,113 @@
     </a-form>
   </a-modal>
 
+  <!-- 分块参数配置弹窗 -->
+  <a-modal v-model:open="chunkConfigModalVisible" title="分块参数配置" width="500px">
+    <template #footer>
+      <a-button key="back" @click="chunkConfigModalVisible = false">取消</a-button>
+      <a-button key="submit" type="primary" @click="handleChunkConfigSubmit">确定</a-button>
+    </template>
+    <div class="chunk-config-content">
+      <div class="params-info">
+        <p>调整分块参数可以控制文本的切分方式，影响检索质量和文档加载效率。</p>
+      </div>
+      <a-form
+        :model="tempChunkParams"
+        name="chunkConfig"
+        autocomplete="off"
+        layout="vertical"
+      >
+        <a-form-item label="Chunk Size" name="chunk_size">
+          <a-input-number v-model:value="tempChunkParams.chunk_size" :min="100" :max="10000" style="width: 100%;" />
+          <p class="param-description">每个文本片段的最大字符数</p>
+        </a-form-item>
+        <a-form-item label="Chunk Overlap" name="chunk_overlap">
+          <a-input-number v-model:value="tempChunkParams.chunk_overlap" :min="0" :max="1000" style="width: 100%;" />
+          <p class="param-description">相邻文本片段间的重叠字符数</p>
+        </a-form-item>
+      </a-form>
+    </div>
+  </a-modal>
+
+  <!-- 添加文件弹窗 -->
+  <a-modal v-model:open="addFilesModalVisible" title="添加文件" width="800px">
+    <template #footer>
+      <a-button key="back" @click="addFilesModalVisible = false">取消</a-button>
+      <a-button
+        key="submit"
+        type="primary"
+        @click="chunkData"
+        :loading="state.chunkLoading"
+        :disabled="(uploadMode === 'file' && fileList.length === 0) || (uploadMode === 'url' && !urlList.trim())"
+      >
+        生成分块
+      </a-button>
+    </template>
+    <div class="add-files-content">
+      <div class="upload-header">
+        <div class="source-selector">
+          <div class="upload-mode-selector" @click="uploadMode = 'file'" :class="{ active: uploadMode === 'file' }">
+            <FileOutlined /> 上传文件
+          </div>
+          <div class="upload-mode-selector" @click="uploadMode = 'url'" :class="{ active: uploadMode === 'url' }">
+            <LinkOutlined /> 输入网址
+          </div>
+        </div>
+        <div class="config-controls">
+          <a-button type="dashed" @click="showChunkConfigModal">
+            <SettingOutlined /> 分块参数 ({{ chunkParams.chunk_size }}/{{ chunkParams.chunk_overlap }})
+          </a-button>
+        </div>
+      </div>
+      
+      <div class="ocr-config">
+        <a-form layout="horizontal">
+          <a-form-item label="使用OCR" name="enable_ocr">
+            <a-select v-model:value="chunkParams.enable_ocr" :options="enable_ocr_options" style="width: 200px;" />
+            <span class="param-description">启用OCR功能，支持PDF文件的文本提取</span>
+          </a-form-item>
+        </a-form>
+      </div>
+
+      <!-- 文件上传区域 -->
+      <div class="upload" v-if="uploadMode === 'file'">
+        <a-upload-dragger
+          class="upload-dragger"
+          v-model:fileList="fileList"
+          name="file"
+          :multiple="true"
+          :disabled="state.chunkLoading"
+          :action="'/api/data/upload?db_id=' + databaseId"
+          :headers="getAuthHeaders()"
+          @change="handleFileUpload"
+          @drop="handleDrop"
+        >
+          <p class="ant-upload-text">点击或者把文件拖拽到这里上传</p>
+          <p class="ant-upload-hint">
+            目前仅支持上传文本文件，如 .pdf, .txt, .md。且同名文件无法重复添加
+          </p>
+        </a-upload-dragger>
+      </div>
+
+      <!-- URL 输入区域 -->
+      <div class="url-input" v-else>
+        <a-form layout="vertical">
+          <a-form-item label="网页链接 (每行一个URL)">
+            <a-textarea
+              v-model:value="urlList"
+              placeholder="请输入网页链接，每行一个"
+              :rows="6"
+              :disabled="state.chunkLoading"
+            />
+          </a-form-item>
+        </a-form>
+        <p class="url-hint">
+          支持添加网页内容，系统会自动抓取网页文本并进行分块。请确保URL格式正确且可以公开访问。
+        </p>
+      </div>
+    </div>
+  </a-modal>
+
   <div class="db-main-container">
     <a-tabs v-model:activeKey="state.curPage" class="atab-container" type="card">
 
@@ -49,7 +156,7 @@
         <div class="db-tab-container">
           <div class="actions" style="display: flex; gap: 10px; justify-content: space-between;">
             <div class="left-actions" style="display: flex; gap: 10px;">
-              <a-button type="primary" @click="handleShowAddFilesBlock" :loading="state.refrashing" :icon="h(PlusOutlined)">添加文件</a-button>
+              <a-button type="primary" @click="showAddFilesModal" :loading="state.refrashing" :icon="h(PlusOutlined)">添加文件</a-button>
               <a-button @click="handleRefresh" :loading="state.refrashing">刷新</a-button>
             </div>
             <div class="batch-actions" style="display: flex; gap: 10px;" v-if="selectedRowKeys.length > 0">
@@ -73,93 +180,7 @@
               </a-button>
             </div>
           </div>
-          <div class="upload-section" v-if="state.showAddFilesBlock">
-            <div class="upload-sidebar">
-              <div class="chunking-params">
-                <div class="params-info">
-                  <p>调整分块参数可以控制文本的切分方式，影响检索质量和文档加载效率。</p>
-                </div>
-                <a-form
-                  :model="chunkParams"
-                  name="basic"
-                  autocomplete="off"
-                  layout="vertical"
-                >
-                  <a-form-item label="Chunk Size" name="chunk_size">
-                    <a-input-number v-model:value="chunkParams.chunk_size" :min="100" :max="10000" />
-                    <p class="param-description">每个文本片段的最大字符数</p>
-                  </a-form-item>
-                  <a-form-item label="Chunk Overlap" name="chunk_overlap">
-                    <a-input-number v-model:value="chunkParams.chunk_overlap" :min="0" :max="1000" />
-                    <p class="param-description">相邻文本片段间的重叠字符数</p>
-                  </a-form-item>
-                  <a-form-item label="使用OCR" name="enable_ocr">
-                    <a-select v-model:value="chunkParams.enable_ocr" :options="enable_ocr_options" />
-                    <p class="param-description">启用OCR功能，支持PDF文件的文本提取</p>
-                  </a-form-item>
-                </a-form>
-              </div>
-            </div>
-            <div class="upload-main">
-              <div class="source-selector">
-                <div class="upload-mode-selector" @click="uploadMode = 'file'" :class="{ active: uploadMode === 'file' }">
-                  <FileOutlined /> 上传文件
-                </div>
-                <div class="upload-mode-selector" @click="uploadMode = 'url'" :class="{ active: uploadMode === 'url' }">
-                  <LinkOutlined /> 输入网址
-                </div>
-              </div>
 
-              <!-- 文件上传区域 -->
-              <div class="upload" v-if="uploadMode === 'file'">
-                <a-upload-dragger
-                  class="upload-dragger"
-                  v-model:fileList="fileList"
-                  name="file"
-                  :multiple="true"
-                  :disabled="state.chunkLoading"
-                  :action="'/api/data/upload?db_id=' + databaseId"
-                  :headers="getAuthHeaders()"
-                  @change="handleFileUpload"
-                  @drop="handleDrop"
-                >
-                  <p class="ant-upload-text">点击或者把文件拖拽到这里上传</p>
-                  <p class="ant-upload-hint">
-                    目前仅支持上传文本文件，如 .pdf, .txt, .md。且同名文件无法重复添加
-                  </p>
-                </a-upload-dragger>
-              </div>
-
-              <!-- URL 输入区域 -->
-              <div class="url-input" v-else>
-                <a-form layout="vertical">
-                  <a-form-item label="网页链接 (每行一个URL)">
-                    <a-textarea
-                      v-model:value="urlList"
-                      placeholder="请输入网页链接，每行一个"
-                      :rows="6"
-                      :disabled="state.chunkLoading"
-                    />
-                  </a-form-item>
-                </a-form>
-                <p class="url-hint">
-                  支持添加网页内容，系统会自动抓取网页文本并进行分块。请确保URL格式正确且可以公开访问。
-                </p>
-              </div>
-
-              <div class="actions">
-                <a-button
-                  type="primary"
-                  @click="chunkData"
-                  :loading="state.chunkLoading"
-                  :disabled="(uploadMode === 'file' && fileList.length === 0) || (uploadMode === 'url' && !urlList.trim())"
-                  style="margin: 0px 20px 20px 0;"
-                >
-                  生成分块
-                </a-button>
-              </div>
-            </div>
-          </div>
           <a-table
             :columns="columns"
             :data-source="Object.values(database.files || {})"
@@ -374,6 +395,7 @@ import {
   EditOutlined,
   PlusOutlined,
   HddOutlined,
+  SettingOutlined,
 } from '@ant-design/icons-vue'
 import { h } from 'vue';
 
@@ -403,7 +425,6 @@ const state = reactive({
   refreshInterval: null,
   curPage: "files",
   indexingFile: null,
-  showAddFilesBlock: false,
   batchIndexing: false,
   batchDeleting: false,
   chunkLoading: false,
@@ -546,9 +567,7 @@ const handleRefresh = () => {
   })
 }
 
-const handleShowAddFilesBlock = () => {
-  state.showAddFilesBlock = !state.showAddFilesBlock
-}
+
 
 const deleteDatabse = () => {
   Modal.confirm({
@@ -794,8 +813,7 @@ const chunkFiles = () => {
     if (data.status === 'success') {
       message.success(data.message || '文件已提交处理，请稍后在列表刷新查看状态');
       fileList.value = []; // 清空已上传文件列表
-      // chunkResults.value = []; // 清空旧的预览结果
-      // activeFileKeys.value = []; // 清空旧的预览结果
+      addFilesModalVisible.value = false; // 关闭弹窗
       getDatabaseInfo(); // 刷新数据库信息以显示新文件及其状态
     } else {
       message.error(data.message || '文件处理失败');
@@ -833,8 +851,7 @@ const chunkUrls = () => {
     if (data.status === 'success') {
       message.success(data.message || 'URL已提交处理，请稍后在列表刷新查看状态');
       urlList.value = ''; // 清空URL输入
-      // chunkResults.value = []; // 清空旧的预览结果
-      // activeFileKeys.value = []; // 清空旧的预览结果
+      addFilesModalVisible.value = false; // 关闭弹窗
       getDatabaseInfo(); // 刷新数据库信息以显示新文件及其状态
     } else {
       message.error(data.message || 'URL处理失败');
@@ -928,6 +945,16 @@ const rules = {
   name: [{ required: true, message: '请输入知识库名称' }]
 };
 
+// 分块参数配置弹窗
+const chunkConfigModalVisible = ref(false);
+const tempChunkParams = ref({
+  chunk_size: 1000,
+  chunk_overlap: 200,
+});
+
+// 添加文件弹窗
+const addFilesModalVisible = ref(false);
+
 // 显示编辑对话框
 const showEditModal = () => {
   editForm.name = database.value.name || '';
@@ -964,6 +991,28 @@ const updateDatabaseInfo = async () => {
   }
 };
 
+// 显示分块参数配置弹窗
+const showChunkConfigModal = () => {
+  tempChunkParams.value = {
+    chunk_size: chunkParams.value.chunk_size,
+    chunk_overlap: chunkParams.value.chunk_overlap,
+  };
+  chunkConfigModalVisible.value = true;
+};
+
+// 处理分块参数配置提交
+const handleChunkConfigSubmit = () => {
+  chunkParams.value.chunk_size = tempChunkParams.value.chunk_size;
+  chunkParams.value.chunk_overlap = tempChunkParams.value.chunk_overlap;
+  chunkConfigModalVisible.value = false;
+  message.success('分块参数配置已更新');
+};
+
+// 显示添加文件弹窗
+const showAddFilesModal = () => {
+  addFilesModalVisible.value = true;
+};
+
 const handleIndexFile = (fileId) => {
   if (!fileId) {
     message.error('无效的文件ID');
@@ -1319,12 +1368,7 @@ const handleBatchIndex = async () => {
   }
 }
 
-.upload {
-  margin-bottom: 20px;
-  .upload-dragger {
-    margin: 0px;
-  }
-}
+
 
 .my-table {
   button.ant-btn-link {
@@ -1400,98 +1444,21 @@ const handleBatchIndex = async () => {
   }
 }
 
-.upload-section {
-  display: flex;
-  gap: 20px;
-
-  .upload-sidebar {
-    min-width: 280px;
-    height: fit-content;
-    padding: 20px;
-    background-color: var(--main-light-6);
-    border-radius: 8px;
-    border: 1px solid var(--main-light-3);
-    flex: 0 0 280px;
-    // box-shadow: 0 2px 8px rgba(0, 0, 0, 0.05);
-
-    .chunking-params {
-      h4 {
-        margin-top: 0;
-        margin-bottom: 16px;
-        color: var(--main-color);
-        font-size: 18px;
-        text-align: center;
-        font-weight: bold;
-        padding-bottom: 10px;
-        border-bottom: 1px dashed var(--main-light-3);
-      }
-
-      .params-info {
-        background-color: var(--main-light-4);
-        border-radius: 6px;
-        padding: 10px 12px;
-        margin-bottom: 16px;
-
-        p {
-          margin: 0;
-          font-size: 13px;
-          line-height: 1.5;
-          color: var(--gray-700);
-        }
-      }
-
-      .ant-form-item {
-        margin-bottom: 16px;
-
-        .ant-form-item-label {
-          padding-bottom: 6px;
-
-          label {
-            color: var(--gray-800);
-            font-weight: 500;
-            font-size: 15px;
-          }
-        }
-      }
-
-      .ant-input-number {
-        width: 100%;
-        border-radius: 6px;
-
-        &:hover, &:focus {
-          border-color: var(--main-color);
-        }
-      }
-
-      .ant-switch {
-        background-color: var(--gray-400);
-
-        &.ant-switch-checked {
-          background-color: var(--main-color);
-        }
-      }
-
-      // 添加参数说明
-      .param-description {
-        color: var(--gray-600);
-        font-size: 12px;
-        margin-top: 4px;
-        margin-bottom: 0;
-      }
-    }
-  }
-
-  .upload-main {
-    flex: 1;
+.add-files-content {
+  .upload-header {
+    display: flex;
+    justify-content: space-between;
+    align-items: center;
+    margin-bottom: 16px;
+    gap: 20px;
 
     .source-selector {
       display: flex;
       gap: 10px;
-      margin-bottom: 16px;
 
       .upload-mode-selector {
         cursor: pointer;
-        padding: 8px 16px;
+        padding: 4px 16px;
         border-radius: 8px;
         background-color: var(--main-light-4);
         border: 1px solid var(--main-light-3);
@@ -1503,6 +1470,116 @@ const handleBatchIndex = async () => {
         }
       }
     }
+
+    .config-controls {
+      .ant-btn {
+        border-color: var(--main-light-3);
+        color: var(--gray-700);
+        &:hover {
+          border-color: var(--main-color);
+          color: var(--main-color);
+        }
+      }
+    }
+  }
+
+  .ocr-config {
+    margin-bottom: 16px;
+    padding: 12px 16px;
+    background-color: var(--main-light-6);
+    border-radius: 8px;
+    border: 1px solid var(--main-light-3);
+
+    .ant-form-item {
+      margin-bottom: 0;
+
+      .ant-form-item-label {
+        color: var(--gray-800);
+        font-weight: 500;
+      }
+    }
+
+    .param-description {
+      color: var(--gray-600);
+      font-size: 12px;
+      margin-left: 12px;
+    }
+  }
+
+  .upload {
+    margin-bottom: 20px;
+    .upload-dragger {
+      margin: 0px;
+      min-height: 200px;
+    }
+  }
+
+  .url-input {
+    margin-bottom: 20px;
+
+    .ant-textarea {
+      border-color: var(--main-light-3);
+      background-color: #fff;
+      font-family: monospace;
+      resize: vertical;
+    }
+
+    .ant-textarea:hover,
+    .ant-textarea:focus {
+      border-color: var(--main-color);
+    }
+
+    .url-hint {
+      font-size: 13px;
+      color: var(--gray-600);
+      margin-top: 5px;
+      line-height: 1.5;
+    }
+  }
+}
+
+.chunk-config-content {
+  .params-info {
+    background-color: var(--main-light-4);
+    border-radius: 6px;
+    padding: 10px 12px;
+    margin-bottom: 16px;
+
+    p {
+      margin: 0;
+      font-size: 13px;
+      line-height: 1.5;
+      color: var(--gray-700);
+    }
+  }
+
+  .ant-form-item {
+    margin-bottom: 16px;
+
+    .ant-form-item-label {
+      padding-bottom: 6px;
+
+      label {
+        color: var(--gray-800);
+        font-weight: 500;
+        font-size: 15px;
+      }
+    }
+  }
+
+  .ant-input-number {
+    border-radius: 6px;
+
+    &:hover, &:focus {
+      border-color: var(--main-color);
+    }
+  }
+
+  .param-description {
+    color: var(--gray-600);
+    font-size: 12px;
+    margin-top: 4px;
+    margin-bottom: 0;
   }
 }
 
@@ -1560,24 +1637,7 @@ const handleBatchIndex = async () => {
   margin-bottom: 20px;
 }
 
-.url-input .ant-textarea {
-  border-color: var(--main-light-3);
-  background-color: #fff;
-  font-family: monospace;
-  resize: vertical;
-}
 
-.url-input .ant-textarea:hover,
-.url-input .ant-textarea:focus {
-  border-color: var(--main-color);
-}
-
-.url-hint {
-  font-size: 13px;
-  color: var(--gray-600);
-  margin-top: 5px;
-  line-height: 1.5;
-}
 
 .loading-container {
   display: flex;