wiki_crawler/backend/main.py

# backend/main.py
from fastapi import FastAPI, APIRouter, BackgroundTasks
# 确保导入路径与你的文件名一致，如果文件名是 workflow.py 则用 workflow
from .services.crawler_sql_service import crawler_sql_service
from .services.automated_crawler import workflow 
from .schemas import (
    RegisterRequest, PendingRequest, SaveResultsRequest, AddUrlsRequest, SearchRequest,
    AutoMapRequest, AutoProcessRequest, TextSearchRequest
)
from .utils import make_response

app = FastAPI(title="Wiki Crawler API")

# ==========================================
# 工具函数
# ==========================================


# ==========================================
# V1 Router: 原始的底层接口 (Manual Control)
# ==========================================
router_v1 = APIRouter()

@router_v1.post("/register")
async def register(req: RegisterRequest):
    try:
        # Service 返回: {'task_id': 1, 'is_new_task': True, 'msg': '...'}
        res = crawler_sql_service.register_task(req.url)
        # 使用 pop 将 msg 提取出来作为响应的 msg，剩下的作为 data
        return make_response(1, res.pop("msg", "Success"), res)
    except Exception as e:
        return make_response(0, str(e))

@router_v1.post("/add_urls")
async def add_urls(req: AddUrlsRequest):
    try:
        urls = req.urls_obj["urls"]
        res = crawler_sql_service.add_urls(req.task_id, urls=urls)
        return make_response(1, res.pop("msg", "Success"), res)
    except Exception as e:
        return make_response(0, str(e))

@router_v1.post("/pending_urls")
async def pending_urls(req: PendingRequest):
    try:
        res = crawler_sql_service.get_pending_urls(req.task_id, req.limit)
        # 即使队列为空，Service 也会返回 msg="Queue is empty"
        return make_response(1, res.pop("msg", "Success"), res)
    except Exception as e:
        return make_response(0, str(e))

@router_v1.post("/save_results")
async def save_results(req: SaveResultsRequest):
    try:
        res = crawler_sql_service.save_results(req.task_id, req.results)
        return make_response(1, res.pop("msg", "Success"), res)
    except Exception as e:
        return make_response(0, str(e))

@router_v1.post("/search")
async def search_v1(req: SearchRequest):
    """V1 搜索：客户端手动传向量"""
    try:
        vector = req.query_embedding['vector']
        if not vector:
            return make_response(2, "Vector is empty", None)

        # Service 现在返回 {'results': [...], 'msg': 'Found ...'}
        res = crawler_sql_service.search_knowledge(
            query_embedding=vector,
            task_id=req.task_id,
            limit=req.limit
        )
        return make_response(1, res.pop("msg", "Search Done"), res)
    except Exception as e:
        return make_response(0, str(e))


# ==========================================
# V2 Router: 自动化工作流 (Automated Workflow)
# ==========================================
router_v2 = APIRouter()

@router_v2.post("/auto/map")
async def auto_map(req: AutoMapRequest):
    """
    [同步] 输入首页 URL，自动调用 Firecrawl Map 并入库
    """
    try:
        # Workflow 返回: {'task_id':..., 'msg': 'Task mapped...', ...}
        res = workflow.map_and_ingest(req.url)
        return make_response(1, res.pop("msg", "Mapping Started"), res)
    except Exception as e:
        return make_response(0, str(e))

@router_v2.post("/auto/process")
async def auto_process(req: AutoProcessRequest, background_tasks: BackgroundTasks):
    """
    [异步] 触发后台任务：消费队列 -> 抓取 -> Embedding -> 入库
    """
    try:
        # 将耗时操作放入后台任务
        background_tasks.add_task(workflow.process_task_queue, req.task_id, req.batch_size)
        
        # 因为是后台任务，无法立即获取 Service 的返回值 msg，只能返回通用消息
        return make_response(1, "Background processing started", {"task_id": req.task_id})
    except Exception as e:
        return make_response(0, str(e))

@router_v2.post("/search")
async def search_v2(req: TextSearchRequest):
    """
    [智能] 输入自然语言文本 -> 后端转向量 -> 搜索
    """
    try:
        # Workflow 返回 {'results': [...], 'msg': '...'}
        res = workflow.search_with_embedding(req.query, req.task_id, req.limit)
        return make_response(1, res.pop("msg", "Search Success"), res)
    except Exception as e:
        return make_response(0, f"Search Failed: {str(e)}")


# ==========================================
# 挂载路由
# ==========================================
app.include_router(router_v1, prefix="/api/v1", tags=["V1 Manual API"])
app.include_router(router_v2, prefix="/api/v2", tags=["V2 Automated Workflow"])

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=8000, reload=True)