mirror of
https://github.com/NanmiCoder/MediaCrawler.git
synced 2026-06-06 09:57:25 +08:00
feat: 代理IP功能 Done
This commit is contained in:
@@ -9,7 +9,7 @@ from playwright.async_api import (BrowserContext, BrowserType, Page,
|
||||
import config
|
||||
from base.base_crawler import AbstractCrawler
|
||||
from models import douyin
|
||||
from proxy.proxy_account_pool import AccountPool
|
||||
from proxy.proxy_ip_pool import create_ip_pool, IpInfoModel
|
||||
from tools import utils
|
||||
from var import crawler_type_var
|
||||
|
||||
@@ -24,27 +24,30 @@ class DouYinCrawler(AbstractCrawler):
|
||||
crawler_type: str
|
||||
context_page: Page
|
||||
dy_client: DOUYINClient
|
||||
account_pool: AccountPool
|
||||
browser_context: BrowserContext
|
||||
|
||||
def __init__(self) -> None:
|
||||
self.user_agent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36" # fixed
|
||||
self.index_url = "https://www.douyin.com"
|
||||
|
||||
def init_config(self, platform: str, login_type: str, account_pool: AccountPool, crawler_type: str) -> None:
|
||||
def init_config(self, platform: str, login_type: str, crawler_type: str) -> None:
|
||||
self.platform = platform
|
||||
self.login_type = login_type
|
||||
self.account_pool = account_pool
|
||||
self.crawler_type = crawler_type
|
||||
|
||||
async def start(self) -> None:
|
||||
account_phone, playwright_proxy, httpx_proxy = self.create_proxy_info()
|
||||
playwright_proxy_format, httpx_proxy_format = None, None
|
||||
if config.ENABLE_IP_PROXY:
|
||||
ip_proxy_pool = await create_ip_pool(config.IP_PROXY_POOL_COUNT, enable_validate_ip=True)
|
||||
ip_proxy_info: IpInfoModel = await ip_proxy_pool.get_proxy()
|
||||
playwright_proxy_format, httpx_proxy_format = self.format_proxy_info(ip_proxy_info)
|
||||
|
||||
async with async_playwright() as playwright:
|
||||
# Launch a browser context.
|
||||
chromium = playwright.chromium
|
||||
self.browser_context = await self.launch_browser(
|
||||
chromium,
|
||||
playwright_proxy,
|
||||
None,
|
||||
self.user_agent,
|
||||
headless=config.HEADLESS
|
||||
)
|
||||
@@ -53,11 +56,11 @@ class DouYinCrawler(AbstractCrawler):
|
||||
self.context_page = await self.browser_context.new_page()
|
||||
await self.context_page.goto(self.index_url)
|
||||
|
||||
self.dy_client = await self.create_douyin_client(httpx_proxy)
|
||||
self.dy_client = await self.create_douyin_client(httpx_proxy_format)
|
||||
if not await self.dy_client.pong(browser_context=self.browser_context):
|
||||
login_obj = DouYinLogin(
|
||||
login_type=self.login_type,
|
||||
login_phone=account_phone,
|
||||
login_phone="", # you phone number
|
||||
browser_context=self.browser_context,
|
||||
context_page=self.context_page,
|
||||
cookie_str=config.COOKIES
|
||||
@@ -148,20 +151,18 @@ class DouYinCrawler(AbstractCrawler):
|
||||
except DataFetchError as e:
|
||||
utils.logger.error(f"aweme_id: {aweme_id} get comments failed, error: {e}")
|
||||
|
||||
def create_proxy_info(self) -> Tuple[Optional[str], Optional[Dict], Optional[str]]:
|
||||
"""Create proxy info for playwright and httpx"""
|
||||
if not config.ENABLE_IP_PROXY:
|
||||
return None, None, None
|
||||
|
||||
# phone: 13012345671 ip_proxy: 111.122.xx.xx1:8888
|
||||
phone, ip_proxy = self.account_pool.get_account() # type: ignore
|
||||
@staticmethod
|
||||
def format_proxy_info(ip_proxy_info: IpInfoModel) -> Tuple[Optional[Dict], Optional[Dict]]:
|
||||
"""format proxy info for playwright and httpx"""
|
||||
playwright_proxy = {
|
||||
"server": f"{config.IP_PROXY_PROTOCOL}{ip_proxy}",
|
||||
"username": config.IP_PROXY_USER,
|
||||
"password": config.IP_PROXY_PASSWORD,
|
||||
"server": f"{ip_proxy_info.protocol}{ip_proxy_info.ip}:{ip_proxy_info.port}",
|
||||
"username": ip_proxy_info.user,
|
||||
"password": ip_proxy_info.password,
|
||||
}
|
||||
httpx_proxy = f"{config.IP_PROXY_PROTOCOL}{config.IP_PROXY_USER}:{config.IP_PROXY_PASSWORD}@{ip_proxy}"
|
||||
return phone, playwright_proxy, httpx_proxy
|
||||
httpx_proxy = {
|
||||
f"{ip_proxy_info.protocol}{ip_proxy_info.ip}": f"{ip_proxy_info.protocol}{ip_proxy_info.user}:{ip_proxy_info.password}@{ip_proxy_info.ip}:{ip_proxy_info.port}"
|
||||
}
|
||||
return playwright_proxy, httpx_proxy
|
||||
|
||||
async def create_douyin_client(self, httpx_proxy: Optional[str]) -> DOUYINClient:
|
||||
"""Create douyin client"""
|
||||
|
||||
Reference in New Issue
Block a user