crawling check

upload
jaehwang 2026-05-11 14:02:17 +09:00
parent 0d3543d84d
commit 26cd946e1b
5 changed files with 14 additions and 2 deletions

View File

@ -1,5 +1,5 @@
import uuid6 import uuid6
from fastapi import APIRouter, BackgroundTasks, Depends, status from fastapi import APIRouter, BackgroundTasks, Depends, status, HTTPException
from common.deps import verify_api_key from common.deps import verify_api_key
from common.db import fetchone, insert_instagram_row, insert_facebook_row, insert_naver_blog_row, insert_youtube_row, insert_gangnam_unni_row, insert_analysis_run from common.db import fetchone, insert_instagram_row, insert_facebook_row, insert_naver_blog_row, insert_youtube_row, insert_gangnam_unni_row, insert_analysis_run
from models.analysis import AnalysisCreate, AnalysisStartResponse, AnalysisStatusResponse from models.analysis import AnalysisCreate, AnalysisStartResponse, AnalysisStatusResponse
@ -19,6 +19,8 @@ async def start_analysis(body: AnalysisCreate, background_tasks: BackgroundTasks
"SELECT owner_user_id FROM hospital_baseinfo WHERE hospital_id = %s", "SELECT owner_user_id FROM hospital_baseinfo WHERE hospital_id = %s",
(hospital_id,), (hospital_id,),
) )
if not hospital:
raise HTTPException(status_code=409, detail="Clinic not found")
owner_user_id = hospital["owner_user_id"] if hospital else 0 owner_user_id = hospital["owner_user_id"] if hospital else 0
ig_url = body.channels.instagram[0] if isinstance(body.channels.instagram, list) else body.channels.instagram ig_url = body.channels.instagram[0] if isinstance(body.channels.instagram, list) else body.channels.instagram
@ -33,7 +35,7 @@ async def start_analysis(body: AnalysisCreate, background_tasks: BackgroundTasks
if ig_id: if ig_id:
background_tasks.add_task(collect_instagram, analysis_run_id, ig_id, ig_url) background_tasks.add_task(collect_instagram, analysis_run_id, ig_id, ig_url)
if fb_id: if fb_id:
background_tasks.add_task(collect_facebook, analysis_run_id, fb_id, body.channels.facebook) background_tasks.add_task(collect_facebook, analysis_run_id, fb_id, f"https://www.facebook.com/{body.channels.facebook}")
if nb_id: if nb_id:
background_tasks.add_task(collect_naver_blog, analysis_run_id, nb_id, body.channels.naver_blog) background_tasks.add_task(collect_naver_blog, analysis_run_id, nb_id, body.channels.naver_blog)
if yt_id: if yt_id:

View File

@ -2,5 +2,7 @@ import os
from fastapi import Header, HTTPException from fastapi import Header, HTTPException
async def verify_api_key(x_api_key: str = Header(...)): async def verify_api_key(x_api_key: str = Header(...)):
print(x_api_key)
print(os.getenv("API_KEY"))
if x_api_key != os.getenv("API_KEY"): if x_api_key != os.getenv("API_KEY"):
raise HTTPException(status_code=401, detail="Invalid API Key") raise HTTPException(status_code=401, detail="Invalid API Key")

View File

@ -1,6 +1,7 @@
from fastapi import FastAPI from fastapi import FastAPI
from dotenv import load_dotenv from dotenv import load_dotenv
from api import routers from api import routers
import os
load_dotenv() load_dotenv()

View File

@ -14,6 +14,7 @@ from integrations.firecrawl import FirecrawlClient
async def collect_instagram(analysis_run_id: str, row_id: int, url: str) -> None: async def collect_instagram(analysis_run_id: str, row_id: int, url: str) -> None:
print("start a insta")
await set_instagram_status(row_id, "processing") await set_instagram_status(row_id, "processing")
data = await ApifyClient(get_env("APIFY_API_TOKEN")).fetch_instagram_profile(normalize_handle("instagram", url)) data = await ApifyClient(get_env("APIFY_API_TOKEN")).fetch_instagram_profile(normalize_handle("instagram", url))
await save_instagram_raw_data(row_id, data) await save_instagram_raw_data(row_id, data)
@ -21,6 +22,7 @@ async def collect_instagram(analysis_run_id: str, row_id: int, url: str) -> None
async def collect_facebook(analysis_run_id: str, row_id: int, url: str) -> None: async def collect_facebook(analysis_run_id: str, row_id: int, url: str) -> None:
print("start a facebook")
await set_facebook_status(row_id, "processing") await set_facebook_status(row_id, "processing")
data = await ApifyClient(get_env("APIFY_API_TOKEN")).fetch_facebook_page(url) data = await ApifyClient(get_env("APIFY_API_TOKEN")).fetch_facebook_page(url)
await save_facebook_raw_data(row_id, data) await save_facebook_raw_data(row_id, data)
@ -28,6 +30,7 @@ async def collect_facebook(analysis_run_id: str, row_id: int, url: str) -> None:
async def collect_naver_blog(analysis_run_id: str, row_id: int, url: str) -> None: async def collect_naver_blog(analysis_run_id: str, row_id: int, url: str) -> None:
print("start a blog")
await set_naver_blog_status(row_id, "processing") await set_naver_blog_status(row_id, "processing")
data = await NaverClient(get_env("NAVER_CLIENT_ID"), get_env("NAVER_CLIENT_SECRET")).fetch_blog_rss(normalize_handle("naver_blog", url)) data = await NaverClient(get_env("NAVER_CLIENT_ID"), get_env("NAVER_CLIENT_SECRET")).fetch_blog_rss(normalize_handle("naver_blog", url))
await save_naver_blog_raw_data(row_id, data) await save_naver_blog_raw_data(row_id, data)
@ -35,6 +38,7 @@ async def collect_naver_blog(analysis_run_id: str, row_id: int, url: str) -> Non
async def collect_youtube(analysis_run_id: str, row_id: int, url: str) -> None: async def collect_youtube(analysis_run_id: str, row_id: int, url: str) -> None:
print("start a youtube")
await set_youtube_status(row_id, "processing") await set_youtube_status(row_id, "processing")
data = await YouTubeClient(get_env("YOUTUBE_API_KEY")).fetch_channel(normalize_handle("youtube", url)) data = await YouTubeClient(get_env("YOUTUBE_API_KEY")).fetch_channel(normalize_handle("youtube", url))
await save_youtube_raw_data(row_id, data) await save_youtube_raw_data(row_id, data)
@ -42,6 +46,7 @@ async def collect_youtube(analysis_run_id: str, row_id: int, url: str) -> None:
async def collect_gangnam_unni(analysis_run_id: str, row_id: int, url: str) -> None: async def collect_gangnam_unni(analysis_run_id: str, row_id: int, url: str) -> None:
print("start a gangnam_unni")
await set_gangnam_unni_status(row_id, "processing") await set_gangnam_unni_status(row_id, "processing")
data = await FirecrawlClient(get_env("FIRECRAWL_API_KEY")).fetch_gangnam_unni(url) data = await FirecrawlClient(get_env("FIRECRAWL_API_KEY")).fetch_gangnam_unni(url)
await save_gangnam_unni_raw_data(row_id, data) await save_gangnam_unni_raw_data(row_id, data)

View File

@ -7,6 +7,8 @@ services:
container_name: o2o-infinith-backend container_name: o2o-infinith-backend
ports: ports:
- "8000:8000" - "8000:8000"
env_file:
- .env
volumes: volumes:
- ./app:/app - ./app:/app
restart: unless-stopped restart: unless-stopped