全面架构重构：建立分层架构与高度可扩展的插件系统
后端重构： - 新增分层架构：API Routes -> Services -> Repositories -> Infrastructure - 彻底移除全局单例，全面采用 FastAPI 依赖注入 - 新增 api/ 目录拆分路由（proxies, plugins, scheduler, settings, stats） - 新增 services/ 业务逻辑层：ProxyService, PluginService, SchedulerService, ValidatorService, SettingsService - 新增 repositories/ 数据访问层：ProxyRepository, SettingsRepository, PluginSettingsRepository - 新增 models/ 层：Pydantic Schemas + Domain Models - 重写 core/config.py：采用 Pydantic Settings 管理配置 - 新增 core/db.py：基于 asynccontextmanager 的连接管理，支持数据库迁移 - 新增 core/exceptions.py：统一业务异常体系 插件系统重构（核心）： - 新增 core/plugin_system/：BaseCrawlerPlugin + PluginRegistry - 采用显式注册模式（装饰器 + plugins/__init__.py），类型安全、测试友好 - 新增 plugins/base.py：BaseHTTPPlugin 通用 HTTP 爬虫基类 - 迁移全部 7 个插件到新架构（fate0, proxylist_download, ip3366, ip89, kuaidaili, speedx, yundaili） - 插件状态持久化到 plugin_settings 表 任务调度重构： - 新增 core/tasks/queue.py：ValidationQueue + WorkerPool - 解耦爬取与验证：爬虫只负责爬取，代理提交队列后由 Worker 异步验证 - 调度器定时从数据库拉取存量代理并分批投入验证队列 前端调整： - 新增 frontend/src/services/ 层拆分 API 调用逻辑 - 调整 stores/ 和 views/ 使用 Service 层 - 保持 API 兼容性，页面无需大幅修改 其他： - 新增 main.py 作为新入口 - 新增 DESIGN.md 架构设计文档 - 更新 requirements.txt 增加 pydantic-settings
This commit is contained in:
祀梦
9
api/routes/__init__.py
Normal file
9
api/routes/__init__.py
Normal file
@@ -0,0 +1,9 @@
|
||||
from fastapi import APIRouter
|
||||
from . import stats, proxies, plugins, scheduler, settings
|
||||
|
||||
api_router = APIRouter()
|
||||
api_router.include_router(stats.router)
|
||||
api_router.include_router(proxies.router)
|
||||
api_router.include_router(plugins.router)
|
||||
api_router.include_router(scheduler.router)
|
||||
api_router.include_router(settings.router)
|
||||
139
api/routes/plugins.py
Normal file
139
api/routes/plugins.py
Normal file
@@ -0,0 +1,139 @@
|
||||
"""插件相关路由"""
|
||||
from fastapi import APIRouter, Depends
|
||||
from services.plugin_service import PluginService
|
||||
from services.scheduler_service import SchedulerService
|
||||
from models.schemas import PluginToggleRequest
|
||||
from api.deps import get_plugin_service, get_scheduler_service
|
||||
from core.log import logger
|
||||
|
||||
router = APIRouter(prefix="/api/plugins", tags=["plugins"])
|
||||
|
||||
|
||||
def success_response(message: str, data=None):
|
||||
return {"code": 200, "message": message, "data": data}
|
||||
|
||||
|
||||
def error_response(message: str, code: int = 500):
|
||||
return {"code": code, "message": message, "data": None}
|
||||
|
||||
|
||||
@router.get("")
|
||||
async def list_plugins(service: PluginService = Depends(get_plugin_service)):
|
||||
plugins = await service.list_plugins()
|
||||
return success_response(
|
||||
"获取插件列表成功",
|
||||
{
|
||||
"plugins": [
|
||||
{
|
||||
"id": p.id,
|
||||
"name": p.display_name, # 保持旧版本兼容：name 用于展示
|
||||
"display_name": p.display_name,
|
||||
"description": p.description,
|
||||
"enabled": p.enabled,
|
||||
"last_run": p.last_run.isoformat() if p.last_run else None,
|
||||
"success_count": p.success_count,
|
||||
"failure_count": p.failure_count,
|
||||
}
|
||||
for p in plugins
|
||||
]
|
||||
},
|
||||
)
|
||||
|
||||
|
||||
@router.put("/{plugin_id}/toggle")
|
||||
async def toggle_plugin(
|
||||
plugin_id: str,
|
||||
request: PluginToggleRequest,
|
||||
service: PluginService = Depends(get_plugin_service),
|
||||
):
|
||||
success = await service.toggle_plugin(plugin_id, request.enabled)
|
||||
if not success:
|
||||
return error_response("插件不存在", 404)
|
||||
return success_response(
|
||||
f"插件 {plugin_id} 已{'启用' if request.enabled else '禁用'}",
|
||||
{"plugin_id": plugin_id, "enabled": request.enabled},
|
||||
)
|
||||
|
||||
|
||||
@router.post("/{plugin_id}/crawl")
|
||||
async def crawl_plugin(
|
||||
plugin_id: str,
|
||||
plugin_service: PluginService = Depends(get_plugin_service),
|
||||
scheduler_service: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
plugin = plugin_service.get_plugin(plugin_id)
|
||||
if not plugin:
|
||||
return error_response("插件不存在", 404)
|
||||
|
||||
try:
|
||||
results = await plugin_service.run_plugin(plugin_id)
|
||||
if not results:
|
||||
return success_response(
|
||||
f"插件 {plugin_id} 爬取完成，未获取到代理",
|
||||
{"plugin_id": plugin_id, "proxy_count": 0, "valid_count": 0},
|
||||
)
|
||||
|
||||
logger.info(f"Plugin {plugin_id} crawled {len(results)} proxies, sending to validation queue")
|
||||
scheduler_service.validation_queue.reset_stats()
|
||||
await scheduler_service.validation_queue.submit(results)
|
||||
# 等待队列排空（最多等 30 秒，避免前端超时）
|
||||
try:
|
||||
await asyncio.wait_for(scheduler_service.validation_queue.drain(), timeout=30.0)
|
||||
except asyncio.TimeoutError:
|
||||
pass
|
||||
|
||||
valid_count = scheduler_service.validation_queue.valid_count
|
||||
invalid_count = scheduler_service.validation_queue.invalid_count
|
||||
|
||||
return success_response(
|
||||
f"插件 {plugin_id} 爬取并验证完成",
|
||||
{
|
||||
"plugin_id": plugin_id,
|
||||
"proxy_count": len(results),
|
||||
"valid_count": valid_count,
|
||||
"invalid_count": invalid_count,
|
||||
},
|
||||
)
|
||||
except Exception as e:
|
||||
logger.error(f"Crawl plugin {plugin_id} failed: {e}")
|
||||
return error_response(f"插件爬取失败: {str(e)}")
|
||||
|
||||
|
||||
@router.post("/crawl-all")
|
||||
async def crawl_all(
|
||||
plugin_service: PluginService = Depends(get_plugin_service),
|
||||
scheduler_service: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
try:
|
||||
results = await plugin_service.run_all_plugins()
|
||||
if not results:
|
||||
return success_response(
|
||||
"所有插件爬取完成，未获取到代理",
|
||||
{"total_crawled": 0, "valid_count": 0, "invalid_count": 0},
|
||||
)
|
||||
|
||||
logger.info(f"All plugins crawled {len(results)} unique proxies, sending to validation queue")
|
||||
scheduler_service.validation_queue.reset_stats()
|
||||
await scheduler_service.validation_queue.submit(results)
|
||||
try:
|
||||
await asyncio.wait_for(scheduler_service.validation_queue.drain(), timeout=60.0)
|
||||
except asyncio.TimeoutError:
|
||||
pass
|
||||
|
||||
valid_count = scheduler_service.validation_queue.valid_count
|
||||
invalid_count = scheduler_service.validation_queue.invalid_count
|
||||
|
||||
return success_response(
|
||||
"所有插件爬取并验证完成",
|
||||
{
|
||||
"total_crawled": len(results),
|
||||
"valid_count": valid_count,
|
||||
"invalid_count": invalid_count,
|
||||
},
|
||||
)
|
||||
except Exception as e:
|
||||
logger.error(f"Crawl all failed: {e}")
|
||||
return error_response(f"批量爬取失败: {str(e)}")
|
||||
|
||||
|
||||
import asyncio
|
||||
114
api/routes/proxies.py
Normal file
114
api/routes/proxies.py
Normal file
@@ -0,0 +1,114 @@
|
||||
"""代理相关路由"""
|
||||
from typing import Optional
|
||||
from fastapi import APIRouter, Depends, Query
|
||||
from services.proxy_service import ProxyService
|
||||
from models.schemas import ProxyListRequest, BatchDeleteRequest
|
||||
from api.deps import get_proxy_service
|
||||
|
||||
router = APIRouter(prefix="/api/proxies", tags=["proxies"])
|
||||
|
||||
|
||||
def success_response(message: str, data=None):
|
||||
return {"code": 200, "message": message, "data": data}
|
||||
|
||||
|
||||
def error_response(message: str, code: int = 500):
|
||||
return {"code": code, "message": message, "data": None}
|
||||
|
||||
|
||||
@router.post("")
|
||||
async def list_proxies(
|
||||
request: ProxyListRequest,
|
||||
service: ProxyService = Depends(get_proxy_service),
|
||||
):
|
||||
proxies, total = await service.list_proxies(
|
||||
page=request.page,
|
||||
page_size=request.page_size,
|
||||
protocol=request.protocol,
|
||||
min_score=request.min_score,
|
||||
max_score=request.max_score,
|
||||
sort_by=request.sort_by,
|
||||
sort_order=request.sort_order,
|
||||
)
|
||||
return success_response(
|
||||
"获取代理列表成功",
|
||||
{
|
||||
"list": [
|
||||
{
|
||||
"ip": p.ip,
|
||||
"port": p.port,
|
||||
"protocol": p.protocol,
|
||||
"score": p.score,
|
||||
"last_check": p.last_check.isoformat() if p.last_check else None,
|
||||
}
|
||||
for p in proxies
|
||||
],
|
||||
"total": total,
|
||||
"page": request.page,
|
||||
"page_size": request.page_size,
|
||||
},
|
||||
)
|
||||
|
||||
|
||||
@router.get("/random")
|
||||
async def get_random_proxy(service: ProxyService = Depends(get_proxy_service)):
|
||||
proxy = await service.get_random_proxy()
|
||||
if not proxy:
|
||||
return error_response("没有找到可用的代理", 404)
|
||||
return success_response(
|
||||
"获取随机代理成功",
|
||||
{
|
||||
"ip": proxy.ip,
|
||||
"port": proxy.port,
|
||||
"protocol": proxy.protocol,
|
||||
"score": proxy.score,
|
||||
"last_check": proxy.last_check.isoformat() if proxy.last_check else None,
|
||||
},
|
||||
)
|
||||
|
||||
|
||||
@router.get("/export/{fmt}")
|
||||
async def export_proxies(
|
||||
fmt: str,
|
||||
protocol: Optional[str] = None,
|
||||
limit: int = Query(default=10000, ge=1, le=100000),
|
||||
service: ProxyService = Depends(get_proxy_service),
|
||||
):
|
||||
if fmt not in ("csv", "txt", "json"):
|
||||
return error_response("不支持的导出格式", 400)
|
||||
|
||||
from fastapi.responses import StreamingResponse
|
||||
|
||||
media_types = {"csv": "text/csv", "txt": "text/plain", "json": "application/json"}
|
||||
|
||||
async def generate():
|
||||
async for chunk in service.export_proxies(fmt, protocol, limit):
|
||||
yield chunk
|
||||
|
||||
return StreamingResponse(
|
||||
generate(),
|
||||
media_type=media_types[fmt],
|
||||
headers={"Content-Disposition": f"attachment; filename=proxies.{fmt}"},
|
||||
)
|
||||
|
||||
|
||||
@router.delete("/{ip}/{port}")
|
||||
async def delete_proxy(ip: str, port: int, service: ProxyService = Depends(get_proxy_service)):
|
||||
await service.delete_proxy(ip, port)
|
||||
return success_response("删除代理成功")
|
||||
|
||||
|
||||
@router.post("/batch-delete")
|
||||
async def batch_delete(
|
||||
request: BatchDeleteRequest,
|
||||
service: ProxyService = Depends(get_proxy_service),
|
||||
):
|
||||
proxies = [(item.ip, item.port) for item in request.proxies]
|
||||
deleted = await service.batch_delete(proxies)
|
||||
return success_response(f"批量删除 {deleted} 个代理成功", {"deleted_count": deleted})
|
||||
|
||||
|
||||
@router.delete("/clean-invalid")
|
||||
async def clean_invalid(service: ProxyService = Depends(get_proxy_service)):
|
||||
count = await service.clean_invalid()
|
||||
return success_response(f"清理了 {count} 个无效代理", {"deleted_count": count})
|
||||
78
api/routes/scheduler.py
Normal file
78
api/routes/scheduler.py
Normal file
@@ -0,0 +1,78 @@
|
||||
"""调度器相关路由"""
|
||||
from fastapi import APIRouter, Depends
|
||||
from services.scheduler_service import SchedulerService
|
||||
from services.settings_service import SettingsService
|
||||
from api.deps import get_scheduler_service
|
||||
from core.log import logger
|
||||
|
||||
router = APIRouter(prefix="/api/scheduler", tags=["scheduler"])
|
||||
|
||||
|
||||
def success_response(message: str, data=None):
|
||||
return {"code": 200, "message": message, "data": data}
|
||||
|
||||
|
||||
def error_response(message: str, code: int = 500):
|
||||
return {"code": code, "message": message, "data": None}
|
||||
|
||||
|
||||
@router.post("/start")
|
||||
async def start_scheduler(
|
||||
scheduler: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
try:
|
||||
if scheduler.running:
|
||||
return success_response("验证调度器已在运行", {"running": True})
|
||||
await scheduler.start()
|
||||
# 持久化设置
|
||||
settings_service = SettingsService()
|
||||
settings = await settings_service.get_settings()
|
||||
settings["auto_validate"] = True
|
||||
from models.schemas import SettingsSchema
|
||||
await settings_service.save_settings(SettingsSchema(**settings))
|
||||
return success_response("验证调度器已启动", {"running": True})
|
||||
except Exception as e:
|
||||
logger.error(f"Start scheduler failed: {e}")
|
||||
return error_response(f"启动调度器失败: {str(e)}")
|
||||
|
||||
|
||||
@router.post("/stop")
|
||||
async def stop_scheduler(
|
||||
scheduler: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
try:
|
||||
if not scheduler.running:
|
||||
return success_response("验证调度器未运行", {"running": False})
|
||||
await scheduler.stop()
|
||||
# 持久化设置
|
||||
settings_service = SettingsService()
|
||||
settings = await settings_service.get_settings()
|
||||
settings["auto_validate"] = False
|
||||
from models.schemas import SettingsSchema
|
||||
await settings_service.save_settings(SettingsSchema(**settings))
|
||||
return success_response("验证调度器已停止", {"running": False})
|
||||
except Exception as e:
|
||||
logger.error(f"Stop scheduler failed: {e}")
|
||||
return error_response(f"停止调度器失败: {str(e)}")
|
||||
|
||||
|
||||
@router.post("/validate-now")
|
||||
async def validate_now(
|
||||
scheduler: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
try:
|
||||
scheduler.validate_all_now()
|
||||
return success_response("已开始全量验证", {"started": True})
|
||||
except Exception as e:
|
||||
logger.error(f"Validate now failed: {e}")
|
||||
return error_response(f"启动验证失败: {str(e)}")
|
||||
|
||||
|
||||
@router.get("/status")
|
||||
async def scheduler_status(
|
||||
scheduler: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
return success_response(
|
||||
"获取状态成功",
|
||||
{"running": scheduler.running, "interval_minutes": scheduler.interval_minutes},
|
||||
)
|
||||
41
api/routes/settings.py
Normal file
41
api/routes/settings.py
Normal file
@@ -0,0 +1,41 @@
|
||||
"""设置相关路由"""
|
||||
from fastapi import APIRouter, Depends
|
||||
from services.settings_service import SettingsService
|
||||
from models.schemas import SettingsSchema
|
||||
from api.deps import get_settings_service
|
||||
from core.log import logger
|
||||
|
||||
router = APIRouter(prefix="/api/settings", tags=["settings"])
|
||||
|
||||
|
||||
def success_response(message: str, data=None):
|
||||
return {"code": 200, "message": message, "data": data}
|
||||
|
||||
|
||||
def error_response(message: str, code: int = 500):
|
||||
return {"code": code, "message": message, "data": None}
|
||||
|
||||
|
||||
@router.get("")
|
||||
async def get_settings(service: SettingsService = Depends(get_settings_service)):
|
||||
try:
|
||||
settings = await service.get_settings()
|
||||
return success_response("获取设置成功", settings)
|
||||
except Exception as e:
|
||||
logger.error(f"Get settings failed: {e}")
|
||||
return error_response("获取设置失败")
|
||||
|
||||
|
||||
@router.post("")
|
||||
async def save_settings(
|
||||
request: SettingsSchema,
|
||||
service: SettingsService = Depends(get_settings_service),
|
||||
):
|
||||
try:
|
||||
success = await service.save_settings(request)
|
||||
if not success:
|
||||
return error_response("保存设置失败")
|
||||
return success_response("保存设置成功", request.model_dump())
|
||||
except Exception as e:
|
||||
logger.error(f"Save settings failed: {e}")
|
||||
return error_response(f"保存设置失败: {str(e)}")
|
||||
30
api/routes/stats.py
Normal file
30
api/routes/stats.py
Normal file
@@ -0,0 +1,30 @@
|
||||
"""统计信息路由"""
|
||||
from fastapi import APIRouter, Depends
|
||||
from services.proxy_service import ProxyService
|
||||
from services.scheduler_service import SchedulerService
|
||||
from api.deps import get_proxy_service, get_scheduler_service
|
||||
from core.log import logger
|
||||
|
||||
router = APIRouter(prefix="/api/stats", tags=["stats"])
|
||||
|
||||
|
||||
def success_response(message: str, data=None):
|
||||
return {"code": 200, "message": message, "data": data}
|
||||
|
||||
|
||||
def error_response(message: str, code: int = 500):
|
||||
return {"code": code, "message": message, "data": None}
|
||||
|
||||
|
||||
@router.get("")
|
||||
async def get_stats(
|
||||
proxy_service: ProxyService = Depends(get_proxy_service),
|
||||
scheduler_service: SchedulerService = Depends(get_scheduler_service),
|
||||
):
|
||||
try:
|
||||
stats = await proxy_service.get_stats()
|
||||
stats["scheduler_running"] = scheduler_service.running
|
||||
return success_response("获取统计信息成功", stats)
|
||||
except Exception as e:
|
||||
logger.error(f"Get stats failed: {e}")
|
||||
return error_response("获取统计信息失败")
|
||||
Reference in New Issue
Block a user