Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,14 +1,12 @@
|
|
| 1 |
import os, asyncio
|
| 2 |
from fastapi import FastAPI, HTTPException, Header, Response
|
| 3 |
from pydantic import BaseModel, HttpUrl
|
| 4 |
-
|
| 5 |
-
from selenium.webdriver.chrome.service import Service
|
| 6 |
-
from selenium.webdriver.chrome.options import Options
|
| 7 |
|
| 8 |
API_KEY = os.getenv("OPENER_API_KEY", "")
|
| 9 |
-
PROFILE = os.getenv("
|
| 10 |
UA = os.getenv("UA", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/129.0.0.0 Safari/537.36")
|
| 11 |
-
|
| 12 |
|
| 13 |
app = FastAPI()
|
| 14 |
drv = None
|
|
@@ -18,27 +16,7 @@ worker = None
|
|
| 18 |
class OpenPayload(BaseModel):
|
| 19 |
url: HttpUrl
|
| 20 |
|
| 21 |
-
def
|
| 22 |
-
os.makedirs(p, exist_ok=True)
|
| 23 |
-
for n in ("SingletonLock", "SingletonCookie", "SingletonSocket", "DevToolsActivePort"):
|
| 24 |
-
f = os.path.join(p, n)
|
| 25 |
-
if os.path.exists(f): os.remove(f)
|
| 26 |
-
|
| 27 |
-
def build_driver():
|
| 28 |
-
clean_profile(PROFILE)
|
| 29 |
-
opts = Options()
|
| 30 |
-
opts.add_argument("--headless=new")
|
| 31 |
-
opts.add_argument("--no-sandbox")
|
| 32 |
-
opts.add_argument("--disable-gpu")
|
| 33 |
-
opts.add_argument("--disable-dev-shm-usage")
|
| 34 |
-
opts.add_argument("--no-first-run")
|
| 35 |
-
opts.add_argument("--no-default-browser-check")
|
| 36 |
-
opts.add_argument(f"--user-agent={UA}")
|
| 37 |
-
opts.add_argument(f"--user-data-dir={PROFILE}")
|
| 38 |
-
svc = Service(executable_path=CHROMEDRIVER)
|
| 39 |
-
return webdriver.Chrome(service=svc, options=opts)
|
| 40 |
-
|
| 41 |
-
async def worker_fn():
|
| 42 |
while True:
|
| 43 |
url = await queue.get()
|
| 44 |
try:
|
|
@@ -50,11 +28,19 @@ async def worker_fn():
|
|
| 50 |
@app.on_event("startup")
|
| 51 |
async def startup():
|
| 52 |
global drv, queue, worker
|
| 53 |
-
os.
|
| 54 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
drv.get("data:text/html,ready")
|
| 56 |
queue = asyncio.Queue()
|
| 57 |
-
worker = asyncio.create_task(
|
| 58 |
print("[startup] browser ready")
|
| 59 |
|
| 60 |
@app.on_event("shutdown")
|
|
|
|
| 1 |
import os, asyncio
|
| 2 |
from fastapi import FastAPI, HTTPException, Header, Response
|
| 3 |
from pydantic import BaseModel, HttpUrl
|
| 4 |
+
import undetected_chromedriver as uc
|
|
|
|
|
|
|
| 5 |
|
| 6 |
API_KEY = os.getenv("OPENER_API_KEY", "")
|
| 7 |
+
PROFILE = os.getenv("UC_PROFILE_DIR", "/tmp/uc-profile")
|
| 8 |
UA = os.getenv("UA", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/129.0.0.0 Safari/537.36")
|
| 9 |
+
CHROME_BIN = os.getenv("CHROME_BIN", "/usr/bin/chromium")
|
| 10 |
|
| 11 |
app = FastAPI()
|
| 12 |
drv = None
|
|
|
|
| 16 |
class OpenPayload(BaseModel):
|
| 17 |
url: HttpUrl
|
| 18 |
|
| 19 |
+
async def _worker():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
while True:
|
| 21 |
url = await queue.get()
|
| 22 |
try:
|
|
|
|
| 28 |
@app.on_event("startup")
|
| 29 |
async def startup():
|
| 30 |
global drv, queue, worker
|
| 31 |
+
os.makedirs(PROFILE, exist_ok=True)
|
| 32 |
+
o = uc.ChromeOptions()
|
| 33 |
+
o.headless = True
|
| 34 |
+
o.binary_location = CHROME_BIN
|
| 35 |
+
o.add_argument(f"--user-data-dir={PROFILE}")
|
| 36 |
+
o.add_argument("--no-sandbox")
|
| 37 |
+
o.add_argument("--disable-dev-shm-usage")
|
| 38 |
+
o.add_argument("--disable-gpu")
|
| 39 |
+
o.add_argument(f"--user-agent={UA}")
|
| 40 |
+
drv = uc.Chrome(options=o, use_subprocess=True)
|
| 41 |
drv.get("data:text/html,ready")
|
| 42 |
queue = asyncio.Queue()
|
| 43 |
+
worker = asyncio.create_task(_worker())
|
| 44 |
print("[startup] browser ready")
|
| 45 |
|
| 46 |
@app.on_event("shutdown")
|