Python環境準備

1. 安裝pipx。如已經安裝，可跳過本步驟：

    python -m pip install --user pipxpython -m pipx ensurepath##  驗證安裝pipx --version

2. 安裝 uv。如已經安裝，可跳過本步驟：

pipx install uv
## 設置為阿里云 PyPI 鏡像源
set UV_INDEX=https://mirrors.aliyun.com/pypi/simple

3. 克隆 RAGFlow

git clone https://github.com/infiniflow/ragflow.git
cd ragflow/

4.安裝 Python 依賴項：

slim:

# install RAGFlow dependent python modules
uv sync --python 3.11 

slim不用full

full:

--all-extras # install RAGFlow dependent python modules
uv sync --python 3.11 

注2：會存在無法安裝pyicu得問題，本地運行若只需要聊天/知識庫可以先忽略
pyicu (國際化依賴源包)版本 ==2.14 安裝可能會不成功，需要注釋pyproject.toml中的pyicu，對于現階段開發無影響，

注釋掉 docker/entrypoint.sh 中的這一行。nginx

# /usr/sbin/nginx

激活 Python 虛擬環境：

source .venv/bin/activate
export PYTHONPATH=$(pwd)

5.下載項目必備模型

找到目錄下得download_depts.py文件，開啟加速器運行，當前目錄下會下載這些文件

注：如果存在虛擬環境則需要把nltk_data文件夾移到虛擬環境文件目錄下

6. 修改配置文件（service_conf.yml）

編輯/conf/service_conf.yml

ragflow:host: 0.0.0.0http_port: 9380
mysql:# 需要先在數據庫創建該庫name: '庫名'user: 'xx'password: 'xxx'host: 'ip'port: portmax_connections: 100stale_timeout: 30
minio:user: 'access_key'password: 'secret_key'host: 'ip:port'
es:hosts: 'http://ip:port'username: 'xx'password: 'xxx'
infinity:uri: 'localhost:23817'db_name: 'default_db'
redis:db: 1password: 'x'host: 'ip:port'# postgres:
#   name: 'rag_flow'
#   user: 'rag_flow'
#   password: 'infini_rag_flow'
#   host: 'postgres'
#   port: 5432
#   max_connections: 100
#   stale_timeout: 30
# s3:
#   access_key: 'access_key'
#   secret_key: 'secret_key'
#   region: 'region'
# oss:
#   access_key: 'access_key'
#   secret_key: 'secret_key'
#   endpoint_url: 'http://oss-cn-hangzhou.aliyuncs.com'
#   region: 'cn-hangzhou'
#   bucket: 'bucket_name'
# azure:
#   auth_type: 'sas'
#   container_url: 'container_url'
#   sas_token: 'sas_token'
# azure:
#   auth_type: 'spn'
#   account_url: 'account_url'
#   client_id: 'client_id'
#   secret: 'secret'
#   tenant_id: 'tenant_id'
#   container_name: 'container_name'
# user_default_llm:
#   factory: 'Tongyi-Qianwen'
#   api_key: 'sk-xxxxxxxxxxxxx'
#   base_url: ''
# oauth:
#   github:
#     client_id: xxxxxxxxxxxxxxxxxxxxxxxxx
#     secret_key: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
#     url: https://github.com/login/oauth/access_token
#   feishu:
#     app_id: cli_xxxxxxxxxxxxxxxxxxx
#     app_secret: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
#     app_access_token_url: https://open.feishu.cn/open-apis/auth/v3/app_access_token/internal
#     user_access_token_url: https://open.feishu.cn/open-apis/authen/v1/oidc/access_token
#     grant_type: 'authorization_code'
# authentication:
#   client:
#     switch: false
#     http_app_key:
#     http_secret_key:
#   site:
#     switch: false
# permission:
#   switch: false
#   component: false
#   dataset: false

6.下載項目必備模型

用git bash進入D:\code\python\ragflow

 export HF_ENDPOINT=https://hf-mirror.com

找到目錄下得download_depts.py文件，開啟加速器運行，當前目錄下會下載這些文件

7.啟動服務

接口服務python路徑：api/ragflow_server.py

python api/ragflow_server.py

任務服務python路徑：rag/svr/task_executor.py

python rag/svr/task_executor.py

8.前端

準備：安裝node>16
idea打開web的Terminal窗口執行：
依賴安裝

npm run install

程序啟動

npm run dev

9.pycharm配置虛擬環境

參考：
https://ragflow.io/docs/dev/launch_ragflow_from_source
https://blog.csdn.net/university96/article/details/146361237
https://blog.csdn.net/qq_33407429/article/details/146182253
https://blog.csdn.net/weixin_45535519/article/details/146199830
http://mp.weixin.qq.com/s?__biz=MjM5NzQxMTAyNw==&mid=2247486291&idx=1&sn=2b5cd4d0e884683b7597d9efa5faaac1&chksm=a7e28746e040f049c53b0cf8558f969cc1699ff35dc17a9a04726c1bc59d22f2741ddc59510a&mpshare=1&scene=1&srcid=0416xjiIyCzsKwumLEySgZw3&sharer_shareinfo=30a70149dc5a708922287838404bcf5a&sharer_shareinfo_first=30a70149dc5a708922287838404bcf5a#rd