转移

2024-11-04 05:20:42 +08:00
parent e990473dcd
commit 07de4d5fd5
24 changed files with 1385 additions and 2 deletions
--- a/.gitignore
+++ b/.gitignore
@@ -0,0 +1 @@
+__pycache__
--- a/README.md
+++ b/README.md
@@ -1,3 +1,118 @@
-# reverse_image_search_gpu
+# reverse_image_search

-图片反向搜索, 利用GPU加速
+图片反向检索
+
+```urls
+https://www.gameui.net/api/default/create_vector?count=1024 # 预先生成指定数量的向量
+https://www.gameui.net/api/default/create_index             # 手动重建索引
+```
+
+## Intall
+
+```bash
+# 先安装 milvus v2.1.4 矢量数据库
+wget https://github.com/milvus-io/milvus/releases/download/v2.1.4/milvus_2.1.4-1_amd64.deb
+sudo apt-get update
+sudo dpkg -i milvus_2.1.4-1_amd64.deb
+sudo apt-get -f install
+
+# 查看 Milvus 及其依賴的狀態(etcd 和 MinIO)
+sudo systemctl status milvus
+sudo systemctl status milvus-etcd
+sudo systemctl status milvus-minio
+
+# 安装 etcdctl (管理工具)
+wget https://github.com/etcd-io/etcd/releases/download/v3.4.22/etcd-v3.4.22-linux-amd64.tar.gz
+# 解压 etcdctl 到 /bin/
+# 向 ~/.bashrc 添加 export ETCDCTL_API=3
+
+# 安装 python 包管理工具 pip
+sudo apt install python3-pip
+
+# 使用 venv 创建虚拟环境(注意安装路径是当前用户的 .local/bin)
+python3 -m venv venv
+source venv/bin/activate
+python -m pip install --upgrade pip
+
+# 为 .bashrc 设置快捷命令(使用echo或cat追加写)
+vim ~/.bashrc
+alias venv='source venv/bin/activate'
+source ~/.bashrc
+
+# 不使用虚拟环境的防止依赖冲突方法
+pip freeze > requirements.txt
+pip uninstall -r requirements.txt
+
+# 安装依赖
+pip3 install -r requirements.txt
+
+# 注意 python 要求 V3.9.7 版本以上
+python3 src/main.py
+
+```
+
+挂载对象存储服务作为存储盘
+
+## 批量处理工具
+
+- [x] 使项目不依赖平台提供的服务, 易于迁移
+- [x] 操作读写数据库, 不使用mysql
+- [x] /img/no_content.8dd8acce.png 冲突
+- [x] 在上海区新建OSS实例作为磁盘挂载到服务器(用于存储缓存图) ...登录
+- [x] 在服务器提供接口(设置CDN层分发) ...登录
+- [x] 获取前端项目(修改列表图片三类分辨率)
+- [x] 通过 path 获取缩略图
+- [x] 通过 articleDetails_id 获取缩略图
+- [x] 图片向量无缓存时重新生成
+- [ ] 提前生成原图webp以提高首次打开速度
+
+## RTT
+
+1. OSS 中 gameui-webp Bucket 作为图片缓存层
+2. 通过 /img/xxx.webp 接口提供图片, 由 OSS 作镜像回源
+
+NGINX config
+
+```config
+server {
+  location /docs {
+    proxy_pass     http://gameui_ai_server;
+    proxy_redirect  off;
+    proxy_set_header Host $host;
+    proxy_set_header X-Real-IP $remote_addr;
+    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+  }
+  
+  location /api/ {
+    proxy_pass     http://gameui_ai_server;
+    proxy_redirect  off;
+    proxy_set_header Host $host;
+    proxy_set_header X-Real-IP $remote_addr;
+    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+  }
+  
+  location ~* /img/([0-9]+)\.(webp|jpeg) {
+    proxy_pass     http://gameui_ai_server;
+    proxy_redirect  off;
+    proxy_set_header Host $host;
+    proxy_set_header X-Real-IP $remote_addr;
+    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+  }
+  
+  location ~* /img/([0-9]+).*\.(webp|jpeg) {
+    proxy_pass     http://gameui_ai_server;
+    proxy_redirect  off;
+    proxy_set_header Host $host;
+    proxy_set_header X-Real-IP $remote_addr;
+    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+  }
+  
+  location ~* /img/article-([0-9]+).*\.(webp|jpeg) {
+    proxy_pass     http://gameui_ai_server;
+    proxy_redirect  off;
+    proxy_set_header Host $host;
+    proxy_set_header X-Real-IP $remote_addr;
+    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+  }
+}
+```
--- a/configs/config.py
+++ b/configs/config.py
@@ -0,0 +1,61 @@
+import os
+import json
+import time
+
+from apscheduler.schedulers.background import BackgroundScheduler
+
+
+# 以 .env 文件中的环境变量为准, 检查是否存在 .env 文件, 没有则创建
+if not os.path.exists('.env'):
+    print('请输入环境变量参数, 将会写入 .env 文件中')
+    OSS_HOST = input('OSS_HOST: ')
+    MYSQL_HOST = input('MYSQL_HOST: ')
+    MILVUS_HOST = input('MILVUS_HOST: ')
+    with open('.env', 'w') as f:
+        f.write(f'OSS_HOST={OSS_HOST}\n')
+        f.write(f'MYSQL_HOST={MYSQL_HOST}\n')
+        f.write(f'MILVUS_HOST={MILVUS_HOST}\n')
+
+
+# 读取 .env 文件中的环境变量
+with open('.env', 'r') as f:
+    env = f.readlines()
+    env = list(filter(lambda x: not x.startswith('#') and not x.startswith('\n') and len(x.split('=')) == 2, env))
+    env = list(map(lambda x: x.replace('\n', '').split('='), env))
+    env = {k: v for k, v in env}
+    print(json.dumps(env, indent=4, ensure_ascii=False))
+    OSS_HOST = env.get('OSS_HOST')
+    MYSQL_HOST = env.get('MYSQL_HOST')
+    MILVUS_HOST = env.get('MILVUS_HOST')
+
+
+# 创建上传图片的临时目录
+UPLOAD_PATH = '/tmp/search-images'
+if not os.path.exists(UPLOAD_PATH):
+    os.makedirs(UPLOAD_PATH)
+
+
+# 创建压缩图片的临时目录
+IMAGES_PATH = '/tmp/images'
+if not os.path.exists(IMAGES_PATH):
+    os.makedirs(IMAGES_PATH)
+
+
+# 创建sqlite3数据库目录
+SQLITE3_PATH = '/tmp/sqlite3'
+if not os.path.exists(SQLITE3_PATH):
+    os.makedirs(SQLITE3_PATH)
+
+
+def clear_images():
+    print('开始清理创建时间大于30分钟的图片缓存')
+    for file in os.listdir(IMAGES_PATH):
+        if os.path.getmtime(os.path.join(IMAGES_PATH, file)) < time.time() - 30 * 60:
+            print('清理图片缓存:', file)
+            os.remove(os.path.join(IMAGES_PATH, file))
+
+
+# 构建一个定时任务每30分钟清理一次图片缓存
+scheduler = BackgroundScheduler()
+scheduler.add_job(clear_images, 'interval', minutes=30)
+scheduler.start()
--- a/demo.png
+++ b/demo.png
--- a/demo.py
+++ b/demo.py
@@ -0,0 +1,53 @@
+import timm
+import torch
+import torch.nn.functional as functional
+from PIL import Image
+from torchvision import transforms
+
+# 加载预训练模型
+model = timm.create_model('resnet50', pretrained=True)
+model = model.eval()
+
+# 定义图片处理流程
+preprocess = transforms.Compose([
+    transforms.Resize(256),
+    transforms.CenterCrop(224),
+    transforms.ToTensor(),
+    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
+])
+
+# 读取图片
+print("loading image")
+img = Image.open("demo.png").convert("RGB")
+tensor = preprocess(img).unsqueeze(0)
+
+# 检查是否有可用的GPU
+if torch.cuda.is_available():
+    input_batch = tensor.to('cuda')
+    model.to('cuda')
+
+print("start run model")
+with torch.no_grad():
+    output = model(input_batch)
+
+for x in output:
+    print(x.shape)
+
+# 输出2048维向量
+# print(output[0])
+
+'''
+from towhee import pipe, ops, DataCollection
+
+p = (
+    pipe.input('path')
+        .map('path', 'img', ops.image_decode())
+        .map('img', 'vec', ops.image_embedding.timm(model_name='resnet50'))
+        .output('img', 'vec')
+)
+
+ea = DataCollection(p('demo.png')).to_list()
+print(ea)
+
+'''
+
--- a/download.py
+++ b/download.py
@@ -0,0 +1,41 @@
+import io
+import os
+import requests
+
+from PIL import Image, ImageFile
+from models.oss import bucket_image2, bucket_webp
+from configs.config import IMAGES_PATH
+
+ImageFile.LOAD_TRUNCATED_IMAGES = True
+
+# 下载图片(使用OSS下载)
+def download_image(url:str) -> Image:
+    if url.startswith('http://image.gameuiux.cn/') or url.startswith('https://image.gameuiux.cn/'):
+        try:
+            url = url.replace('http://image.gameuiux.cn/', '').replace('https://image.gameuiux.cn/', '')
+            obj = bucket_image2.get_object(url).read()
+            return Image.open(io.BytesIO(obj))
+        except Exception:
+            return None
+    else:
+        try:
+            response = requests.get(url)
+            return Image.open(io.BytesIO(response.content))
+        except Exception:
+            print('图片下载失败:', url)
+            return None
+
+# 生成缩略图, 写入OSS
+def generate_thumbnail(image:Image, id:int, version:str, n:int, w:int, ext:str):
+    path = f"{id}-{version}@{n}x{w}.{ext}"
+    if bucket_webp.object_exists(path):
+        print('缩略图已经存在:', path)
+        return
+    # 将 image 对象复制一份, 防止影响原图
+    image = image.copy()
+    image.thumbnail((n*w, image.size[1]))
+    image.save(f"{IMAGES_PATH}/{path}", ext, save_all=True)
+    bucket_webp.put_object_from_file(path, f"{IMAGES_PATH}/{path}")
+    os.remove(f"{IMAGES_PATH}/{path}")
+    print('缩略图写入 OSS:', path)
+
--- a/log.yaml
+++ b/log.yaml
@@ -0,0 +1,19 @@
+version: 1
+formatters:
+  simple:
+    format: '%(asctime)s - %(name)s - %(levelname)s - %(message)s'
+handlers:
+  console:
+    class: logging.StreamHandler
+    level: DEBUG
+    formatter: simple
+    stream: ext://sys.stdout
+loggers:
+  simpleExample:
+    level: DEBUG
+    handlers: [console]
+    propagate: no
+root:
+  level: DEBUG
+  handlers: [console]
+
--- a/main.py
+++ b/main.py
@@ -0,0 +1,28 @@
+# -*- coding:utf-8 -*-
+
+import sys
+import uvicorn
+
+from fastapi import FastAPI
+from starlette.middleware.cors import CORSMiddleware
+from functools import lru_cache
+from routers import reverse, user, task, img, user_collect
+
+
+# 初始化 FastAPI
+app = FastAPI(title="GameUI", description="GameUI", version="1.5.0", openapi_url="/docs/openapi.json", docs_url="/docs", redoc_url="/redoc")
+app.add_middleware(CORSMiddleware, allow_origins=["*"], allow_credentials=True, allow_methods=["*"], allow_headers=["*"])
+
+
+# 导入路由
+app.include_router(user.router, prefix='/api/user', tags=['用户'])
+app.include_router(task.router, prefix='/api/task', tags=['任务'])
+app.include_router(reverse.router, prefix='/api/default', tags=['搜图'])
+app.include_router(user_collect.router, prefix='/api/user_collect', tags=['收藏'])
+app.include_router(img.router, prefix='/imgs', tags=['图片'])
+
+
+# 启动服务
+if __name__ == '__main__':
+    port = 5002 if len(sys.argv) < 2 else int(sys.argv[1])
+    uvicorn.run(app='main:app', host='0.0.0.0', port=port, reload=True, workers=1)
--- a/models/milvus.py
+++ b/models/milvus.py
@@ -0,0 +1,26 @@
+import pymilvus
+
+from configs.config import MILVUS_HOST
+
+# 连接 Milvus (开启 Milvus 服务)
+collection_name = 'default'
+
+
+# 获取 Milvus 连接
+def get_collection(collection_name):
+    pymilvus.connections.connect(host=MILVUS_HOST, port='19530')
+    if not pymilvus.utility.has_collection(collection_name):
+        field1 = pymilvus.FieldSchema(name="id", dtype=pymilvus.DataType.INT64, is_primary=True)
+        field2 = pymilvus.FieldSchema(name="embedding", dtype=pymilvus.DataType.FLOAT_VECTOR, dim=2048)
+        field3 = pymilvus.FieldSchema(name="article_id", dtype=pymilvus.DataType.INT64)
+        schema = pymilvus.CollectionSchema(fields=[field1, field2, field3])
+        return pymilvus.Collection(name=collection_name, schema=schema)
+    return pymilvus.Collection(name=collection_name)
+
+
+# 检查索引是否存在, 不存在则创建, 并加载
+#collection = get_collection(collection_name)
+#if not collection.has_index():
+#    default_index = {"index_type": "IVF_SQ8", "metric_type": 'L2', "params": {"nlist": 16384}}
+#    collection.create_index(field_name="embedding", index_params=default_index)
+#collection.load()
--- a/models/mysql.py
+++ b/models/mysql.py
@@ -0,0 +1,16 @@
+import pymysql
+from configs.config import MYSQL_HOST
+
+
+# 连接 MySQL (开启 MySQL 服务)
+conn = pymysql.connect(host=MYSQL_HOST, user="gameui", port=3306, password="gameui@2022", database='gameui', local_infile=True, cursorclass=pymysql.cursors.DictCursor)
+
+
+# 获取 MySQL 连接
+def get_cursor():
+    try:
+        dx = conn.ping()
+        return conn.cursor()
+    except Exception:
+        conn = pymysql.connect(host=MYSQL_HOST, user="gameui", port=3306, password="gameui@2022", database='gameui', local_infile=True, cursorclass=pymysql.cursors.DictCursor)
+        return conn.cursor()
--- a/models/oss.py
+++ b/models/oss.py
@@ -0,0 +1,9 @@
+import oss2
+from configs.config import OSS_HOST
+
+
+# 连接 OSS
+oss2.defaults.connection_pool_size = 100
+auth = oss2.Auth('LTAI4GH3qP6VA3QpmTYCgXEW', 'r2wz4bJty8iYfGIcFmEqlY1yon2Ruy')
+bucket_image2 = oss2.Bucket(auth, f'http://{OSS_HOST}', 'gameui-image2')
+bucket_webp = oss2.Bucket(auth, f'http://{OSS_HOST}', 'gameui-webp')
--- a/models/resnet.py
+++ b/models/resnet.py
@@ -0,0 +1,11 @@
+import towhee
+
+class Resnet50:
+    def resnet50_extract_feat(self, img_path):
+        feat = towhee.glob(img_path).image_decode().image_embedding.timm(model_name='resnet50').tensor_normalize().to_list()
+        print(feat[0])
+        return feat[0]
+
+if __name__ == '__main__':
+    print('This script is running as the main program.')
+    #resnet = 
--- a/models/task.py
+++ b/models/task.py
@@ -0,0 +1,184 @@
+import os
+import sqlite3
+import datetime
+from pydantic import BaseModel
+from configs.config import SQLITE3_PATH
+import uuid
+import time
+import requests
+
+
+__taskdb = sqlite3.connect(os.path.join(SQLITE3_PATH, 'tasks.db'), check_same_thread=False)
+__taskdb.row_factory = sqlite3.Row
+
+# 创建任务表
+__taskdb.execute("""
+CREATE TABLE IF NOT EXISTS tasks (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    name TEXT NOT NULL,
+    user_id INTEGER NOT NULL,
+    description TEXT NOT NULL,
+    created_at TEXT NOT NULL,
+    updated_at TEXT NOT NULL
+)
+""")
+
+
+# 任务表单模型
+class TaskForm(BaseModel):
+    name: str
+    description: str
+
+
+class Task(BaseModel):
+    id: str
+    name: str
+    user_id: int
+    description: str
+    created_at: str
+    updated_at: str
+    status: str = "pending"
+    progress: int = 0
+    websockets: list = []
+
+
+class TaskManager:
+    __tasks = {}
+    
+    # 使用pydantic模型的写法
+    def add_task(self, name: str, user_id: int, description: str):
+        id = str(uuid.uuid4())
+        date = str(datetime.datetime.now())
+        task = Task(id=id, name=name, user_id=user_id, description=description, created_at=date, updated_at=date)
+        self.__tasks[id] = task
+        return task
+
+    def get_task(self, task_id: str):
+        task = self.__tasks.get(task_id)
+        if task:
+            return task.dict()
+        else:
+            return None
+    
+    # 直接返回对象的写法
+    def get_tasks(self, user_id: int=None):
+        if user_id:
+            return [task for task in self.__tasks.values() if task.user_id == user_id]
+        else:
+            return [task for task in self.__tasks.values()]
+
+
+    def remove_task(self, task_id: str):
+        if self.__tasks.get(task_id):
+            self.__tasks.pop(task_id)
+            return True
+        else:
+            return False
+
+    # 正确的写法(不等待async函数直接返回结果)
+    async def add_websocket(self, task_id: str, websocket):
+        print(self.__tasks)
+        # 每次调用连接数都是1, 这是错误的
+        self.__tasks[task_id].websockets.append(websocket)
+        print(f"当前连接数: {len(self.__tasks[task_id].websockets)}")
+
+
+        #if len(task.websockets) > 0 and task.progress == 0 and task.status == "pending":
+        #    print("任务未被监听, 开启监听")
+        #    await self.start_listen_task(task)
+    
+    # 移除 websocket
+    def remove_websocket(self, task_id: str, websocket):
+        task = self.__tasks.get(task_id)
+        if task:
+            print("连接已断开, 移除 websocket")
+            task.websockets.remove(websocket)
+            print(f"当前剩余连接数: {len(task.websockets)}")
+    
+    # 正确的写法
+    async def start_listen_task(self, task: Task):
+        while True:
+            print(f"start listen task: {len(task.websockets)}")
+            time.sleep(2.5)
+            if len(task.websockets) == 0:
+                break
+            print(f"get task status:{task.id}")
+            break
+            #data = requests.get(f"http://localhost:5001/api/task/{task.id}").json()
+            #print(data)
+            ## 通过 http 请求获取任务的状态
+            ## 如果任务状态发生变化, 则将任务状态发送给所有监听该任务的 websocket
+            #if task.status != data.status:
+            #    task.status = data.status
+            #    print("send task status to websocket")
+            #    for websocket in task.websockets:
+            #        await websocket.send_json(task)
+
+
+
+## 任务实体模型
+#class TaskModel(BaseModel):
+#    id: str
+#    name: str
+#    user_id: int
+#    description: str
+#    created_at: str
+#    updated_at: str
+#    websockets: list=[]
+#
+## 向数据库中添加任务
+#def add_task(name, user_id, description):
+#    print(name, user_id, description)
+#    date = datetime.datetime.now()
+#    cursor = __taskdb.cursor()
+#    cursor.execute("INSERT INTO tasks(name,user_id,description,created_at,updated_at) VALUES (?, ?, ?, ?, ?)", (name,user_id,description,date,date))
+#    id = cursor.lastrowid
+#    cursor.close()
+#    __taskdb.commit()
+#    return {"id": id,"name": name,"user_id": user_id,"description": description,"created_at": date,"updated_at": date}
+#
+#
+## 从数据库中获取任务列表
+#def get_tasks(user_id: int=None):
+#    cursor = __taskdb.cursor()
+#    if user_id:
+#        cursor.execute("SELECT * FROM tasks WHERE user_id = ?", (user_id,))
+#    else:
+#        cursor.execute("SELECT * FROM tasks")
+#    tasks = cursor.fetchall()
+#    cursor.close()
+#    return [dict(task) for task in tasks]
+#
+#
+## 从数据库中获取指定任务
+#def get_task(task_id: int):
+#    cursor = __taskdb.cursor()
+#    cursor.execute("SELECT * FROM tasks WHERE id = ?", (task_id,))
+#    task = cursor.fetchone()
+#    cursor.close()
+#    return task
+#
+#
+## 更新数据库中的任务
+#def update_task(task: TaskModel):
+#    cursor = __taskdb.cursor()
+#    cursor.execute("UPDATE tasks SET name = ?, description = ?, updated_at = ? WHERE id = ?", (
+#        task.name,
+#        task.description,
+#        datetime.datetime.now(),
+#        task.id
+#    ))
+#    cursor.close()
+#    __taskdb.commit()
+#    return task
+#
+#
+## 从数据库中删除任务
+#def delete_task(task_id: int):
+#    cursor = __taskdb.cursor()
+#    cursor.execute("DELETE FROM tasks WHERE id = ?", (task_id,))
+#    cursor.close()
+#    __taskdb.commit()
+#    return task_id
+
+
--- a/requirements.txt
+++ b/requirements.txt
@@ -0,0 +1,63 @@
+aliyun-python-sdk-core==2.13.36
+aliyun-python-sdk-kms==2.16.0
+anyio==3.6.2
+bleach==6.0.0
+certifi==2022.12.7
+cffi==1.15.1
+charset-normalizer==3.0.1
+click==8.1.3
+crcmod==1.7
+cryptography==39.0.1
+docutils==0.19
+fastapi==0.92.0
+grpcio==1.37.1
+grpcio-tools==1.37.1
+h11==0.14.0
+idna==3.4
+importlib-metadata==6.0.0
+jaraco.classes==3.2.3
+jeepney==0.8.0
+jmespath==0.10.0
+keyring==23.13.1
+markdown-it-py==2.1.0
+mdurl==0.1.2
+mmh3==3.0.0
+more-itertools==9.0.0
+numpy==1.24.2
+opencv-python==4.6.0.66
+oss2==2.16.0
+pandas==1.5.3
+pgzip==0.3.4
+Pillow==9.4.0
+pkginfo==1.9.6
+protobuf==3.20.3
+pyarrow==11.0.0
+pycparser==2.21
+pycryptodome==3.17
+pydantic==1.10.5
+pygit2==1.10.1
+Pygments==2.14.0
+pymilvus==2.0.2
+PyMySQL==1.0.2
+python-dateutil==2.8.2
+python-multipart==0.0.5
+pytz==2022.7.1
+readme-renderer==37.3
+requests==2.28.2
+requests-toolbelt==0.10.1
+rfc3986==2.0.0
+rich==13.3.1
+SecretStorage==3.3.3
+six==1.16.0
+sniffio==1.3.0
+starlette==0.25.0
+tabulate==0.9.0
+towhee==0.9.0
+tqdm==4.64.1
+twine==4.0.2
+typing_extensions==4.5.0
+ujson==5.1.0
+urllib3==1.26.14
+uvicorn==0.20.0
+webencodings==0.5.1
+zipp==3.13.0
--- a/resize.sh
+++ b/resize.sh
@@ -0,0 +1,39 @@
+#!/bin/sh
+
+# 定时整理 etcd 空间防止 etcd 服务挂掉
+# 将此脚本加入定时任务(每日1次, 一次一片)
+
+# crontab -e
+# 15 * * * * /bin/sh /home/milvus/resize.sh
+
+# 查看 Milvus 及其依賴的狀態(etcd 和 MinIO)
+# sudo systemctl status milvus
+# sudo systemctl status milvus-etcd
+# sudo systemctl status milvus-minio
+
+echo "使用API3:"
+export ETCDCTL_API=3
+
+echo "查看空间占用:"
+etcdctl endpoint status --write-out table
+
+echo "查看告警状态:"
+etcdctl alarm list
+
+echo "获取当前版本:"
+rev=$(etcdctl --endpoints=http://127.0.0.1:2379 endpoint status --write-out="json" | egrep -o '"revision":[0-9]*' | egrep -o '[0-9].*')
+
+echo "压缩掉所有旧版本:"
+etcdctl --endpoints=http://127.0.0.1:2379 compact $rev
+
+echo "整理多余的空间:"
+etcdctl --endpoints=http://127.0.0.1:2379 defrag
+
+echo "取消告警信息:"
+etcdctl --endpoints=http://127.0.0.1:2379 alarm disarm
+
+echo "再次查看空间占用:"
+etcdctl endpoint status --write-out table
+
+echo "再次查看告警状态:"
+etcdctl alarm list
--- a/routers/img.py
+++ b/routers/img.py
@@ -0,0 +1,214 @@
+import os
+import time
+import base64
+import psutil
+import statistics
+import _thread as thread
+
+from fastapi import APIRouter, HTTPException, Response
+from urllib.parse import unquote
+from configs.config import IMAGES_PATH
+from models.mysql import get_cursor, conn
+from utilities.download import download_image, generate_thumbnail
+
+
+router = APIRouter()
+
+
+# 预热图片(获取一次图片, 遍历图片表, 检查OSS中所有被预定的尺寸是否存在, 不存在则生成)
+@router.get("/warm", summary="预热图片", description="预热图片")
+def warm_image(op:int=0, end:int=10, version:str='0'):
+    cursor = get_cursor()
+    cursor.execute(f"SELECT * FROM `web_images` limit {op}, {end}")
+    for img in cursor.fetchall():
+        # 如果CPU使用率大于50%, 则等待, 直到CPU使用率小于50%
+        while statistics.mean(psutil.cpu_percent(interval=1, percpu=True)) > 50:
+            print(statistics.mean(psutil.cpu_percent(interval=1, percpu=True)), '等待CPU释放...')
+            time.sleep(2)
+        
+        # 如果内存剩余小于1G, 则等待, 直到内存剩余大于1G
+        while psutil.virtual_memory().available < 1024 * 1024 * 1024:
+            print(psutil.virtual_memory().available, '等待内存释放...')
+            time.sleep(2)
+
+        # CPU使用率已降低, 开始处理图片
+        image = download_image(img['content']) # 从OSS下载原图
+        if not image:
+            print('跳过不存在的图片:', img['content'])
+            continue
+
+        # 创建新线程处理图片
+        try:
+            print('开始处理图片:', img['content'])
+            thread.start_new_thread(generate_thumbnail, (image, img['id'], version, 3, 328, 'webp'))
+            thread.start_new_thread(generate_thumbnail, (image, img['id'], version, 2, 328, 'webp'))
+            thread.start_new_thread(generate_thumbnail, (image, img['id'], version, 1, 328, 'webp'))
+        except:
+            print('无法启动线程')
+    cursor.close()
+    return Response('预热成功', status_code=200, media_type='text/plain', headers={'Content-Type': 'text/plain; charset=utf-8'})
+
+
+# 获取非标准类缩略图
+@router.get("/{type}-{id}-{version}@{n}x{w}.{ext}", summary="获取非标准类缩略图", description="/img/article-233-version@1x320.webp 通过webp后缀获取webp格式图片, 通过@1x320后缀获取1x缩略图, 通过@2x320后缀获取2x缩略图, 通过@3x320后缀获取3x缩略图")
+def get_image_type_thumbnail(type:str, id:str, version:str, n:int, w:int, ext:str):
+    img_path = f"{IMAGES_PATH}/{type}-{id}-{version}@{n}x{w}.{ext}"
+    if os.path.exists(img_path):
+        return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+    if type == 'ad' or type == 'article' or type == 'article_attribute':
+        cursor = get_cursor()
+        count = cursor.execute(f"SELECT * FROM `web_{type}` WHERE `id`={id}")
+        img = cursor.fetchone()
+        cursor.close()
+        if img is None:
+            print('图片不存在:', count)
+            return Response('图片不存在', status_code=404)
+        url = img['image']
+    elif type == 'url':
+        id = unquote(id, 'utf-8')
+        id = id.replace(' ','+')
+        url = unquote(base64.b64decode(id))
+        print(url)
+    elif type == 'avatar':
+        cursor = get_cursor()
+        count = cursor.execute(f"SELECT avatar FROM `web_member` WHERE `id`={id}")
+        user = cursor.fetchone()
+        cursor.close()
+        if user is None:
+            print('用户不存在:', count)
+            return Response('用户不存在', status_code=404)
+        url = user['avatar']
+    else:
+        print('图片类型不存在:', type)
+        return Response('图片类型不存在', status_code=404)
+    image = download_image(url)
+    if not image:
+        return Response('图片不存在', status_code=404)
+    # 如果是 avatar, 则裁剪为正方形
+    if type == 'avatar':
+        px = image.size[0] if image.size[0] < image.size[1] else image.size[1]
+        image = image.crop((0, 0, px, px))
+    image.thumbnail((n*w, image.size[1]))
+    image.save(img_path, ext, save_all=True)
+    return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+
+
+# 获取非标准类原尺寸图
+@router.get("/{type}-{id}-{version}.{ext}", summary="获取文章缩略图", description="/img/233@1x320.webp 通过webp后缀获取webp格式图片, 通过@1x320后缀获取1x缩略图, 通过@2x320后缀获取2x缩略图, 通过@3x320后缀获取3x缩略图")
+def get_image_type(type:str, id:str, version:str, ext:str):
+    img_path = f"{IMAGES_PATH}/{type}-{id}-{version}.{ext}"
+    if os.path.exists(img_path):
+        return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+    if type == 'ad' or type == 'article' or type == 'article_attribute':
+        cursor = get_cursor()
+        count = cursor.execute(f"SELECT * FROM `web_{type}` WHERE `id`={id}")
+        img = cursor.fetchone()
+        cursor.close()
+        if img is None:
+            print('图片不存在:', count)
+            return Response('图片不存在', status_code=404)
+        url = img['image']
+    elif type == 'url':
+        id = unquote(id, 'utf-8')
+        id = id.replace(' ','+')
+        url = unquote(base64.b64decode(id))
+        print("url:", url)
+    elif type == 'avatar':
+        cursor = get_cursor()
+        count = cursor.execute(f"SELECT avatar FROM `web_member` WHERE `id`={id}")
+        user = cursor.fetchone()
+        cursor.close()
+        if user is None:
+            print('用户不存在:', count)
+            return Response('用户不存在', status_code=404)
+        url = user['avatar']
+    else:
+       print('图片类型不存在:', type)
+       return Response('图片类型不存在', status_code=404)
+    image = download_image(url)
+    image.save(img_path, ext, save_all=True)
+    return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+
+
+# 通过url获取图片
+@router.get("/url-{url}@{n}x{w}.{ext}", summary="通过url获取图片", description="/img/article-233.webp")
+def get_image_url(url:str, n:int, w:int, ext:str):
+    img_path = f"{IMAGES_PATH}/{type}-{url}.{ext}"
+    if os.path.exists(img_path):
+        return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+    url = unquote(url, 'utf-8').replace(' ','+')
+    url = unquote(base64.b64decode(url))
+    image = download_image(url)
+    if not image:
+        return Response('图片不存在', status_code=404)
+    image.thumbnail((n*w, image.size[1]))
+    image.save(img_path, ext, save_all=True)
+    return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+
+
+# 获取标准缩略图(带版本号)
+@router.get("/{id}-{version}@{n}x{w}.{ext}", summary="获取缩略图(带版本号)", description="/img/233@1x320.webp 通过webp后缀获取webp格式图片, 通过@1x320后缀获取1x缩略图, 通过@2x320后缀获取2x缩略图, 通过@3x320后缀获取3x缩略图")
+def get_image_thumbnail(id:int, version:str, n:int, w:int, ext:str):
+    # 判断图片是否已经生成
+    img_path = f"{IMAGES_PATH}/{id}-{version}@{n}x{w}.{ext}"
+    if os.path.exists(img_path):
+        return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+    # 从数据库获取原图地址
+    cursor = get_cursor()
+    cursor.execute(f"SELECT * FROM `web_images` WHERE `id`={id}")
+    img = cursor.fetchone()
+    cursor.close()
+    if img is None:
+        print('图片不存在:', id)
+        return Response('图片不存在', status_code=404)
+    image = download_image(img['content'])
+    if not image:
+        return Response('图片不存在', status_code=404)
+    image.thumbnail((n*w, image.size[1]))
+    image.save(img_path, ext, save_all=True)
+    return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+
+
+# 获取标准缩略图
+@router.get("/{id}@{n}x{w}.{ext}", summary="获取缩略图", description="/img/233@1x320.webp 通过webp后缀获取webp格式图片, 通过@1x320后缀获取1x缩略图, 通过@2x320后缀获取2x缩略图, 通过@3x320后缀获取3x缩略图")
+def get_image_thumbnail(id:int, n:int, w:int, ext:str):
+    # 判断图片是否已经生成
+    img_path = f"{IMAGES_PATH}/{id}@{n}x{w}.{ext}"
+    if os.path.exists(img_path):
+        return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+    # 从数据库获取原图地址
+    cursor = get_cursor()
+    cursor.execute(f"SELECT * FROM `web_images` WHERE `id`={id}")
+    img = cursor.fetchone()
+    cursor.close()
+    if img is None:
+        print('图片不存在:', id)
+        return Response('图片不存在', status_code=404)
+    image = download_image(img['content'])
+    if not image:
+        return Response('图片不存在', status_code=404)
+    image.thumbnail((n*w, image.size[1]))
+    image.save(img_path, ext, save_all=True)
+    return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+
+
+# 获取标准原尺寸图
+@router.get("/{id}.{ext}", summary="获取标准原尺寸图", description="/img/233@1x320.webp 通过webp后缀获取webp格式图片, 无后缀获取原图")
+def get_image(id: int = 824, ext: str = 'webp'):
+    # 判断图片是否已经生成
+    img_path = f"{IMAGES_PATH}/{id}.{ext}"
+    if os.path.exists(img_path):
+        return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
+    # 从数据库获取原图地址
+    cursor = get_cursor()
+    cursor.execute(f"SELECT * FROM `web_images` WHERE `id`={id}")
+    img = cursor.fetchone()
+    cursor.close()
+    if img is None:
+        print('图片不存在:', id)
+        return Response('图片不存在', status_code=404)
+    image = download_image(img['content'])
+    if not image:
+        return Response('图片不存在', status_code=404)
+    image.save(img_path, ext, save_all=True)
+    return Response(content=open(img_path, 'rb').read(), media_type=f"image/{ext}")
--- a/routers/reverse.py
+++ b/routers/reverse.py
@@ -0,0 +1,231 @@
+import os
+import random
+import string
+import sqlite3
+import numpy as np
+import time
+import io
+from PIL import Image
+from towhee import pipe, ops
+
+from fastapi import APIRouter, HTTPException, File, UploadFile, Response, status, Header
+from models.milvus import get_collection, collection_name
+from models.mysql import get_cursor, conn
+#from models.resnet import Resnet50
+
+from configs.config import UPLOAD_PATH
+from utilities.download import download_image
+
+router = APIRouter()
+#MODEL = Resnet50()
+
+RESNET50 = (pipe.input('img').map('img', 'vec', ops.image_embedding.timm(model_name='resnet50')).output('vec'))
+
+# 获取状态统计
+@router.get('', summary='状态统计', description='通过表名获取状态统计')
+def count_images():
+    collection = get_collection(collection_name)
+    return {'status': True, 'count': collection.num_entities}
+
+
+# 手动重建索引
+@router.get('/create_index', summary='重建索引', description='手动重建索引', include_in_schema=False)
+async def create_index():
+    collection = get_collection(collection_name)
+    default_index = {"index_type": "IVF_SQ8", "metric_type": 'L2', "params": {"nlist": 16384}}
+    collection.create_index(field_name="embedding", index_params=default_index)
+    collection.load()
+    return {'status': True, 'count': collection.num_entities}
+
+'''
+# 批量生成向量
+@router.get('/create_vector', summary='生成向量', description='手动生成向量', include_in_schema=False)
+async def create_vector(count: int = 10):
+    cursor = get_cursor()
+    cursor.execute(f"SELECT id,thumbnail_image,article_id,milvus_id FROM `web_images` WHERE thumbnail_image IS NOT NULL AND article_id IS NOT NULL AND milvus_id != 2048 AND width IS NOT NULL LIMIT 0,{count}")
+    images = cursor.fetchall()
+    cursor.close()
+    for item in images:
+        print(item)
+        # 先查询 milvus 中是否存在
+        collection = get_collection(collection_name)
+        data = collection.query(expr=f'id in [{item["id"]}]', output_fields=None, partition_names=None, timeout=None) # offset, limit
+        if len(data) > 0:
+            cursor = get_cursor()
+            cursor.execute(f"UPDATE web_images SET milvus_id=2048 WHERE id={item['id']}")
+            conn.commit()
+            cursor.close()
+            continue
+        try:
+            img_path = os.path.join(UPLOAD_PATH, os.path.basename(item['thumbnail_image']))
+            download_image(item['thumbnail_image']).save(img_path, 'png', save_all=True)
+            feat = MODEL.resnet50_extract_feat(img_path)
+            collection.insert([[item['id']], [feat], [item['article_id']]])
+            cursor = get_cursor()
+            cursor.execute(f"UPDATE web_images SET milvus_id=2048 WHERE id={item['id']}")
+            conn.commit()
+            cursor.close()
+        except Exception as e:
+            print(e)
+    print('END')
+    return images
+'''
+
+@router.put('/{image_id}', summary='覆写向量', description='重建指定图像的向量')
+async def rewrite_image(image_id: int):
+    print('START', image_id, '重建向量')
+    cursor = get_cursor()
+    cursor.execute(f"SELECT * FROM `web_images` WHERE id={image_id}")
+    img = cursor.fetchone()
+    cursor.close()
+    if img is None:
+        print('mysql中原始图片不存在:', image_id)
+        return Response('图片不存在', status_code=404)
+    img_path = os.path.join(UPLOAD_PATH, os.path.basename(img['content']))
+    print('img_path', img_path)
+    image = download_image(img['content'])
+    if image is None:
+        print('图片下载失败:', img['content'])
+        return Response('图片下载失败', status_code=404)
+    image.save(img_path, 'png', save_all=True)
+
+    with Image.open(img_path) as imgx:
+        feat = RESNET50(imgx).get()[0]
+
+    collection = get_collection(collection_name)
+    collection.delete(expr=f'id in [{image_id}]')
+    rest = collection.insert([[image_id], [feat], [img['article_id']]])
+    os.remove(img_path)
+    print('END', image_id, '重建向量', rest.primary_keys)
+    return {"code": 0, "status": True, "message": "重建成功", "feature": feat.tolist()}
+
+
+# 获取相似(废弃)
+@router.get('/{image_id}', summary='获取相似', description='通过图片ID获取相似图片')
+async def similar_images(image_id: int, page: int = 1, pageSize: int = 20):
+    collection = get_collection(collection_name)
+    result = collection.query(expr=f'id in [{image_id}]', output_fields = ['id', 'article_id', 'embedding'], top_k=1)
+    # 如果没有结果, 则重新生成记录
+    if len(result) == 0:
+        cursor = get_cursor()
+        cursor.execute(f"SELECT * FROM `web_images` WHERE id={image_id}")
+        img = cursor.fetchone()
+        cursor.close()
+        if img is None:
+            print('mysql 中图片不存在:', image_id)
+            return Response('图片不存在', status_code=404)
+        img_path = os.path.join(UPLOAD_PATH, os.path.basename(img['content']))
+        image = download_image(img['content'])
+        if image is None:
+            print('图片下载失败:', img['content'])
+            return Response('图片下载失败', status_code=404)
+        image.save(img_path, 'png', save_all=True)
+        with Image.open(img_path) as imgx:
+            feat = RESNET50(imgx).get()[0]
+            # 移除可能存在的旧记录, 换上新的
+            collection.delete(expr=f'id in [{image_id}]')
+            collection.insert([[image_id], [feat], [img['article_id']]])
+        os.remove(img_path)
+        print('生成')
+    else:
+        print('通过')
+        feat = result[0]['embedding']
+    res = collection.search([feat],anns_field="embedding",param={"metric_type": 'L2', "params": {"nprobe": 16}}, output_fields=["id", "article_id"], limit=200)[0]
+    # 翻页(截取有效范围, page * pageize)
+    ope = page*pageSize-pageSize
+    end = page*pageSize
+    next = False
+    # 为数据附加信息
+    ids = [i.id for i in res] # 获取所有ID
+    if len(res) <= end:
+        ids = ids[ope:]
+        next = False
+    else:
+        ids = ids[ope:end]
+        next = True
+    str_ids = str(ids).replace('[', '').replace(']', '')
+    if str_ids == '':
+        print('没有更多数据了')
+        return {'code': 0, 'pageSize': pageSize, 'page': page, 'next': False, 'list': []}
+    cursor = get_cursor()
+    cursor.execute(f"SELECT id,user_id,article_id,width,height,content,larger_image,thumbnail_image,article_category_top_id,praise_count,collect_count,create_time,update_time FROM `web_images` WHERE id IN ({str_ids})")
+    imgs = cursor.fetchall()
+    if len(imgs) == 0:
+        return imgs
+    # 获取用户ID和文章ID
+    uids = list(set([x['user_id'] for x in imgs]))
+    tids = list(set([x['article_id'] for x in imgs]))
+    # 获取用户信息
+    cursor.execute(f"SELECT id,user_name,avatar FROM `web_member` WHERE id IN ({str(uids).replace('[', '').replace(']', '')})")
+    users = cursor.fetchall()
+    # 获取文章信息
+    cursor.execute(f"SELECT id,title,tags FROM `web_article` WHERE id IN ({str(tids).replace('[', '').replace(']', '')})")
+    articles = cursor.fetchall()
+    cursor.close()
+    # 合并信息
+    user, article = {}, {}
+    for x in users: user[x['id']] = x
+    for x in articles: article[x['id']] = x
+    for x in imgs:
+        x['article'] = article[x['article_id']]
+        x['user'] = user[x['user_id']]
+        x['distance'] = [i.distance for i in res if i.id == x['id']][0]
+        if x['praise_count'] == None: x['praise_count'] = 0
+        if x['collect_count'] == None: x['collect_count'] = 0
+        # 将字段名转换为驼峰
+        x['createTime'] = x.pop('create_time')
+        x['updateTime'] = x.pop('update_time')
+    # 对 imgs 重新排序(按照 distance 字段)
+    imgs = sorted(imgs, key=lambda x: x['distance'])
+    return {'code': 0, 'pageSize': pageSize, 'page': page, 'next': next, 'list': imgs}
+
+
+@router.post(path='', summary='以图搜图', description='上传图片进行搜索')
+async def search_imagex(image: UploadFile = File(...), page: int = 1, pageSize: int = 20):
+    content = await image.read()
+    img = Image.open(image.file)
+    embeddig = RESNET50(img).get()[0]
+    collection = get_collection('default')
+    res = collection.search([embeddig],anns_field="embedding",param={"metric_type": 'L2', "params": {"nprobe": 16}}, output_fields=["id", "article_id"], limit=500)[0]
+    ope, end = (page - 1) * pageSize, page * pageSize
+    ids, nextx = [x.id for x in res][ope:end], len(res) > end
+
+    if not ids:
+        return {'code': 0, 'pageSize': pageSize, 'page': page, 'next': False, 'list': []}
+
+    cursor = get_cursor()
+    cursor.execute(f"SELECT id,user_id,article_id,width,height,content,larger_image,thumbnail_image,article_category_top_id,praise_count,collect_count,create_time,update_time FROM `web_images` WHERE id IN ({','.join(map(str, ids))})")
+    imgs = cursor.fetchall()
+
+    if not imgs:
+        return imgs
+
+    uids, tids = list(set(x['user_id'] for x in imgs)), list(set(x['article_id'] for x in imgs))
+    cursor.execute(f"SELECT id,user_name,avatar FROM `web_member` WHERE id IN ({str(uids).replace('[', '').replace(']', '')})")
+    users = cursor.fetchall()
+    cursor.execute(f"SELECT id,title,tags FROM `web_article` WHERE id IN ({str(tids).replace('[', '').replace(']', '')})")
+    articles = cursor.fetchall()
+    cursor.close()
+
+    user, article = {x['id']: x for x in users}, {x['id']: x for x in articles}
+    for x in imgs:
+        x.update({
+            'article': article[x['article_id']],
+            'user': user[x['user_id']],
+            #'distance': next(i.distance for i in res if i.id == x['id'], 0),
+            'distance': [i.distance for i in res if i.id == x['id']][0],
+            'praise_count': x.get('praise_count', 0),
+            'collect_count': x.get('collect_count', 0)
+        })
+
+    imgs.sort(key=lambda x: x['distance'])
+    return {'code': 0, 'pageSize': pageSize, 'page': page, 'next': nextx, 'list': imgs}
+
+@router.delete('/{thread_id}', summary="删除主题", description="删除指定主题下的所有图像")
+async def delete_images(thread_id: str):
+    collection = get_collection(collection_name)
+    collection.delete(expr="article_id in ["+thread_id+"]")
+    collection.load()
+    default_index = {"index_type": "IVF_SQ8", "metric_type": 'L2', "params": {"nlist": 16384}}
+    collection.create_index(field_name="embedding", index_params=default_index)
+    return {"status": True, 'msg': '删除完毕'}
--- a/routers/task.py
+++ b/routers/task.py
@@ -0,0 +1,94 @@
+from fastapi import APIRouter, HTTPException, Request, WebSocket
+from fastapi.responses import HTMLResponse
+from fastapi.templating import Jinja2Templates
+from models.task import TaskForm, TaskManager
+
+
+router = APIRouter()
+task_manager = TaskManager()
+
+
+# 创建新任务
+@router.post("", summary="创建新任务")
+def create_task(form: TaskForm):
+    task = task_manager.add_task(name=form.name, user_id=123456, description=form.description)
+    return task
+
+
+# 获取任务列表
+@router.get("", summary="获取任务列表", description="可使用user_id参数筛选指定用户的任务列表")
+def get_task_list(user_id: int=None):
+    return task_manager.get_tasks(user_id)
+
+
+# websocket demo
+@router.get("/demo", response_class=HTMLResponse)
+async def websocket_demo(request: Request):
+    task_list = task_manager.get_tasks()
+    templates = Jinja2Templates(directory="templates")
+    return templates.TemplateResponse("websocket.html", {"request": request, "task_list": task_list})
+
+
+# 监听指定任务的变化事件, 通知前端(不使用pydantic模型, 正确的写法)
+@router.websocket("/{task_id}", name="监听任务变化")
+async def websocket_endpoint(task_id: str, websocket: WebSocket):
+    await websocket.accept()
+    await task_manager.add_websocket(task_id, websocket)
+    async for data in websocket.iter_text():
+        await websocket.send_text(f"Message text was: {data}")
+    task_manager.remove_websocket(task_id, websocket)
+    print("websocket 连接已自动关闭")
+
+    #await websocket.send_json({"message": "Hello WebSocket!"})
+    #task = task_manager.get_task(task_id)
+    #print(task)
+    #if not task:
+    #    print("task 不存在, 结束连接")
+    #    return await websocket.close()             # 任务不存在, 结束连接
+    #await websocket.send_json(task)                # 将任务的状态发送给客户端
+    #await task_manager.add_websocket(task_id, websocket)
+    # 正确的写法, 使用 async for, 并且处理意外断开的情况
+    #try:
+    #    async for data in websocket.iter_text():
+    #        if data == "close":
+    #            print("客户端主动关闭连接")
+    #            task_manager.remove_websocket(task_id, websocket)
+    #            await websocket.close()
+    #            break
+    #        else:
+    #            print(f"接收到客户端消息: {data}")
+    #            await websocket.send_text(f"Message text was: {data}")
+    #except Exception as e:
+    #    print(f"客户端意外断开连接: {e}")
+    #    task_manager.remove_websocket(task_id, websocket)
+    #    #await websocket.close()
+
+    # 监听客户端的状态, 如果客户端主动关闭连接或意外断开连接, 都从任务的websocket列表中移除
+    #while True:
+    #    try:
+    #        data = await websocket.receive_text()
+    #        if data == "close":
+    #            print("客户端主动关闭连接")
+    #            task_manager.remove_websocket(task_id, websocket)
+    #            await websocket.close()
+    #            break
+    #        else:
+    #            print(f"接收到客户端消息: {data}")
+    #            await websocket.send_text(f"Message text was: {data}")
+    #    except Exception as e:
+    #        print(f"客户端意外断开连接: {e}")
+    #        task_manager.remove_websocket(task_id, websocket)
+    #        #await websocket.close()
+    #        break
+
+
+# 获取任务详情
+@router.get("/{task_id}", summary="获取任务详情")
+def get_task(task_id: int):
+    return get_task(task_id)
+
+
+# 删除任务
+@router.delete("/{task_id}", summary="删除指定任务")
+def delete_task(task_id: int):
+    return delete_task(task_id)
--- a/routers/user.py
+++ b/routers/user.py
@@ -0,0 +1,19 @@
+from fastapi import APIRouter, HTTPException
+from models.mysql import conn, get_cursor
+
+router = APIRouter()
+
+@router.get('/{user_id}/collect', summary="用户收藏记录", description="获取指定用户收藏记录")
+def get_user_collect(user_id:int):
+    # TODO: 需要验证权限
+    cursor = get_cursor()
+    cursor.execute(f"SELECT content FROM web_collect  WHERE user_id={user_id} AND type='1'")
+    data = cursor.fetchall()
+    if not data:
+        return {'code': 0, 'user_id': user_id, 'total': 0, 'data': []}
+    data = [str(item['content']) for item in data]
+    cursor.execute(f"SELECT id FROM web_images WHERE content IN ({','.join(['%s'] * len(data))})", data)
+    data = cursor.fetchall()
+    data = [str(item['id']) for item in data]
+    return {'code': 0, 'user_id': user_id, 'total': len(data), 'data': data }
+
--- a/routers/user_collect.py
+++ b/routers/user_collect.py
@@ -0,0 +1,45 @@
+from typing import Optional
+from fastapi import APIRouter, HTTPException, Header
+from models.mysql import conn, get_cursor
+
+
+router = APIRouter()
+
+
+# 获取当前用户收藏记录
+@router.get('', summary='自己的收藏记录', description='获取自己的收藏记录, 用于判断是否收藏(headers中必须附带token)')
+def get_self_collect(token: Optional[str] = Header()):
+    print('token: ', token)
+    cursor = get_cursor()
+    # 查询用户ID
+    cursor.execute(f"SELECT user_id FROM web_auth WHERE token={token} limit 1")
+    data = cursor.fetchone()
+    print('auth: ', data)
+    if not data:
+        raise HTTPException(status_code=401, detail="用户未登录")
+    user_id = data['user_id']
+    # 查询收藏记录
+    cursor.execute(f"SELECT content FROM web_collect  WHERE user_id={user_id} AND type='1'")
+    data = cursor.fetchall()                       # 获取所有记录列表
+    data = [str(item['content']) for item in data] # 转换为数组
+    # 查询图片ID(对特殊字符安全转义)
+    cursor.execute(f"SELECT id FROM web_images WHERE content IN ({','.join(['%s'] * len(data))})", data)
+    data = cursor.fetchall()
+    data = [str(item['id']) for item in data]
+    return {'code': 0, 'user_id': user_id, 'total': len(data), 'data': data }
+
+
+# 获取指定用户收藏记录
+@router.get('/{user_id}', summary='指定用的户收藏记录', description='获取指定用户收藏记录(仅测试用)')
+def get_user_collect(user_id: int):
+    cursor = get_cursor()
+    cursor.execute(f"SELECT content FROM web_collect  WHERE user_id={user_id} AND type=1")
+    data = cursor.fetchall()                       # 获取所有记录列表
+    data = [str(item['content']) for item in data] # 转换为数组
+    if not data:
+        return {'code': 0, 'user_id': user_id, 'total': 0, 'data': []}
+    # 查询图片ID(对特殊字符安全转义)
+    cursor.execute(f"SELECT id FROM web_images WHERE content IN ({','.join(['%s'] * len(data))})", data)
+    data = cursor.fetchall()
+    data = [str(item['id']) for item in data]
+    return {'code': 0, 'user_id': user_id, 'total': len(data), 'data': data}
--- a/start.sh
+++ b/start.sh
@@ -0,0 +1,25 @@
+#!/bin/sh
+
+# pip freeze > requirements.txt
+#pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
+#python3 -m pip install -U pyOpenSSL cryptography
+
+# 使用 uvicorn 启动两个工作进程, 并支持自动重载
+#uvicorn main:app --host 0.0.0.0 --port 5001 --workers 2 --reload >/dev/null 2>&1 &
+#uvicorn main:app --host 0.0.0.0 --port 5002 >/dev/null 2>&1 & 
+
+python main.py >/dev/null 2>&1 & 
+
+# 为容器设定自动重启
+#docker update --restart=always 51bb94aa2726
+
+#cd ~/reverse_image_search
+#pm2 start python3 --name reverse-1 -- main.py 5001
+#pm2 start python3 --name reverse-2 -- main.py 5002
+#pm2 start python3 --name reverse-3 -- main.py 5003
+
+# 某些系统下，需要手动安装指定版本 opencv-python (4.7引用zlib错误)
+# pip install opencv-python==4.6.0.66 -i https://pypi.tuna.tsinghua.edu.cn/simple
+
+# 如果网络不通, 从自定义服务器下载模型参数
+#scp root@172.21.216.33:~/root/.cache/torch/hub/checkpoints/resnet50_a1_0-14fe96d1.pth resnet50_a1_0-14fe96d1.pth
--- a/templates/websocket.html
+++ b/templates/websocket.html
@@ -0,0 +1,44 @@
+<!DOCTYPE html>
+<html>
+    <head>
+        <title>WebSocket Demo</title>
+    </head>
+    <body>
+        <div>
+            <p>{{ task_list }}</p>
+            <button id="create_task">Create Task</button>
+        </div>
+        <h1>WebSocket Demo</h1>
+        <div id="message"></div>
+        <button id="create">Create WebSocket</button>
+        <button id="close">Close WebSocket</button>
+        <script>
+            function createWebSocket(id) {
+                var ws = new WebSocket("ws://localhost:5001/api/task/"+id);
+                ws.onmessage = function(event) {
+                    var message = event.data;
+                    document.getElementById("message").innerHTML += message + "<br>";
+                };
+                ws.onclose = function(event) {
+                    document.getElementById("message").innerHTML += "WebSocket closed";
+                };
+                return ws;
+            }
+            document.getElementById("create_task").onclick = function() {
+                fetch("/api/task", {
+                    method: "POST",
+                    headers: {
+                        "Content-Type": "application/json"
+                    },
+                    body: JSON.stringify({
+                        "name": "Task 1",
+                        "description": "Task 1 description"
+                    })
+                }).then(res => res.json()).then(data => {
+                    console.log(data)
+                    var ws = createWebSocket(data.id);
+                });
+            };
+        </script>
+    </body>
+</html>
--- a/update.sh
+++ b/update.sh
@@ -0,0 +1,4 @@
+#!/bin/sh
+
+# 服务器跟随更新
+ssh ai "cd ~/reverse_image_search_gpu; git pull;"
--- a/utilities/download.py
+++ b/utilities/download.py
@@ -0,0 +1,41 @@
+import io
+import os
+import requests
+
+from PIL import Image, ImageFile
+from models.oss import bucket_image2, bucket_webp
+from configs.config import IMAGES_PATH
+
+ImageFile.LOAD_TRUNCATED_IMAGES = True
+
+# 下载图片(使用OSS下载)
+def download_image(url:str) -> Image:
+    if url.startswith('http://image.gameuiux.cn/') or url.startswith('https://image.gameuiux.cn/'):
+        try:
+            url = url.replace('http://image.gameuiux.cn/', '').replace('https://image.gameuiux.cn/', '')
+            obj = bucket_image2.get_object(url).read()
+            return Image.open(io.BytesIO(obj))
+        except Exception:
+            return None
+    else:
+        try:
+            response = requests.get(url)
+            return Image.open(io.BytesIO(response.content))
+        except Exception:
+            print('图片下载失败:', url)
+            return None
+
+# 生成缩略图, 写入OSS
+def generate_thumbnail(image:Image, id:int, version:str, n:int, w:int, ext:str):
+    path = f"{id}-{version}@{n}x{w}.{ext}"
+    if bucket_webp.object_exists(path):
+        print('缩略图已经存在:', path)
+        return
+    # 将 image 对象复制一份, 防止影响原图
+    image = image.copy()
+    image.thumbnail((n*w, image.size[1]))
+    image.save(f"{IMAGES_PATH}/{path}", ext, save_all=True)
+    bucket_webp.put_object_from_file(path, f"{IMAGES_PATH}/{path}")
+    os.remove(f"{IMAGES_PATH}/{path}")
+    print('缩略图写入 OSS:', path)
+