can run playwright in multi-thread

This commit is contained in:
2022-03-28 14:17:12 +02:00
parent b486aa75d9
commit cf2af4bf7b
+61 -46
View File
@@ -1,7 +1,11 @@
import asyncio
import random import random
import threading
from concurrent.futures import ThreadPoolExecutor
from enum import Enum from enum import Enum
from typing import Union from typing import Union
from playwright.async_api import async_playwright
from playwright.sync_api import sync_playwright from playwright.sync_api import sync_playwright
import params import params
@@ -22,49 +26,16 @@ class ResultEnum(Enum):
PENDING = "PENDING" PENDING = "PENDING"
# check result with playright class Tls(threading.local):
def check_result_page(url, phone_number: str) -> ResultEnum: def __init__(self) -> None:
with sync_playwright() as p: self.playwright = sync_playwright().start()
return run(p, url, phone_number) print("Create playwright instance in Thread", threading.current_thread().name)
def run(playwright, url, phone_number: str) -> ResultEnum: class Worker:
browser = playwright.firefox.launch(headless=False) tls = Tls()
url_to_check = url.replace("register/", "")
url_to_check = url_to_check + "?lang=fr"
content = None
random_id_number = str(phone_number)[1:len(str(phone_number))]
proxy_username = "panleicim-stc-fr-" + random_id_number
print("proxy_username is " + proxy_username)
proxy = {
"server": params.PROXY_SERVER,
"username": proxy_username,
"password": params.PROXY_PASSWORD
}
while content is None:
content = load_page(browser, url_to_check, proxy)
random_id_number = get_random_id_number_for_proxy()
proxy_username = "panleicim-stc-fr-" + random_id_number
print("proxy_username is " + proxy_username)
proxy = {
"server": params.PROXY_SERVER,
"username": proxy_username,
"password": params.PROXY_PASSWORD
}
print(content)
browser.close()
if SORRY_SENTENCE in content:
print("status is REFUSED")
return ResultEnum.REFUSED
elif PENDING_SENTENCE in content:
print("status is PENDING")
return ResultEnum.PENDING
else:
print("status is ACCEPTED")
return ResultEnum.ACCEPTED
def load_page(self, browser, url, proxy) -> Union[str, None]:
def load_page(browser, url, proxy) -> Union[str, None]:
try: try:
firefox_user_agents = filter(lambda user_agent: "firefox" in user_agent.lower(), user_agent_list) firefox_user_agents = filter(lambda user_agent: "firefox" in user_agent.lower(), user_agent_list)
firefox_user_agents_list = list(firefox_user_agents) firefox_user_agents_list = list(firefox_user_agents)
@@ -79,21 +50,65 @@ def load_page(browser, url, proxy) -> Union[str, None]:
print(error) print(error)
return None return None
def run(self, reserve_pojo: ReserveResultPojo, collection):
print("Launched worker in ", threading.current_thread().name)
browser = self.tls.playwright.firefox.launch(headless=False)
url = reserve_pojo.url
phone_number = reserve_pojo.phone
url_to_check = url.replace("register/", "")
url_to_check = url_to_check + "?lang=fr"
content = None
random_id_number = str(phone_number)[1:len(str(phone_number))]
proxy_username = "panleicim-stc-fr-" + random_id_number
print("proxy_username is " + proxy_username)
proxy = {
"server": params.PROXY_SERVER,
"username": proxy_username,
"password": params.PROXY_PASSWORD
}
while content is None:
content = self.load_page(browser, url_to_check, proxy)
random_id_number = get_random_id_number_for_proxy()
proxy_username = "panleicim-stc-fr-" + random_id_number
print("proxy_username is " + proxy_username)
proxy = {
"server": params.PROXY_SERVER,
"username": proxy_username,
"password": params.PROXY_PASSWORD
}
print(content)
browser.close()
print("Stopped worker in ", threading.current_thread().name)
status = None
if SORRY_SENTENCE in content:
print("status is REFUSED")
status = ResultEnum.REFUSED
elif PENDING_SENTENCE in content:
print("status is PENDING")
status = ResultEnum.PENDING
else:
print("status is ACCEPTED")
status = ResultEnum.ACCEPTED
collection.document(reserve_pojo.id).update({u'accepted': status.name})
# need to start at 21h00 # need to start at 21h00
if __name__ == '__main__': if __name__ == '__main__':
# get the list # get the list
params.oracle_log_sender.send_log(msg="开始检查约会结果", subject=LOG_SUBJECT_EVENT, type=TYPE_EVENT_CHECK_RESULTS) params.oracle_log_sender.send_log(msg="开始检查约会结果", subject=LOG_SUBJECT_EVENT, type=TYPE_EVENT_CHECK_RESULTS)
db_manager = params.firebase_store_manager db_manager = params.firebase_store_manager
collection = db_manager.get_all_successful_items() collection = db_manager.get_all_successful_items_for_day("2022-03-16")
count = 0 count = 0
task_list = list()
with ThreadPoolExecutor(max_workers=6) as executor:
for appointment in collection.stream(): for appointment in collection.stream():
worker = Worker()
count = count + 1 count = count + 1
reserve_pojo = ReserveResultPojo.from_firestore_dict(appointment.to_dict()) reserve_pojo = ReserveResultPojo.from_firestore_dict(appointment.to_dict())
if reserve_pojo.accepted: # if reserve_pojo.accepted:
print("status is " + reserve_pojo.accepted) # print("status is " + reserve_pojo.accepted)
if reserve_pojo.accepted is None: # if reserve_pojo.accepted is None:
# or ResultEnum.PENDING.value == reserve_pojo.accepted: # or ResultEnum.PENDING.value == reserve_pojo.accepted:
result = check_result_page(reserve_pojo.url, reserve_pojo.phone) # tasks = check_result_page(reserve_pojo.url, reserve_pojo.phone)
collection.document(reserve_pojo.id).update({u'accepted': result.name}) executor.submit(worker.run, reserve_pojo, collection)
print(count) print(count)