某音用户信息获取

q838210720 · 发表于 2022-11-4 16:45:20

python源码

安装

 pip install playwright -i https://mirrors.bfsu.edu.cn/pypi/web/simple/

 python -m playwright install

python代码

# -*- coding: utf-8 -*-
"""
home.php?mod=space&uid=116177    : 2022/7/28 22:58
@Author  : superhero
home.php?mod=space&uid=270431   : 838210720@qq.com
home.php?mod=space&uid=702767    : dy.py
@IDE: PyCharm
"""

import asyncio
from playwright.async_api import async_playwright
import re

async def cancel_request(route, request):
    await route.abort()

async def get_userinfo(p, _url):
    # 获取用户信息，是否直播，IP属地等
    browser = await p.chromium.launch(headless=True)
    context = await browser.new_context()
    page = await context.new_page()
    await page.route(re.compile(r"(\.png)|(\.jpg)|(\.css)|(\.js)"), cancel_request)
    await page.goto(_url)

    data = {}
    ip = await page.query_selector_all('xpath=//*[@id="root"]/div/div[2]/div/div/div[2]/div[1]/p[1]')
    user_name = await page.locator('xpath=//*[@id="root"]/div/div[2]/div/div/div[2]/div[1]/div[2]/h1/span/span/span/span/span/span').text_content()
    try:
        is_live = await page.locator('xpath=//*[@id="root"]/div/div[2]/div/div/div[2]/div[1]/div[3]/div[2]/button').text_content(timeout=1000)
    except Exception:
        is_live = '未开播'

    try:
        is_re = await page.locator('xpath=//*[@id="root"]/div/div[2]/div/div/div[2]/div[1]/div[2]/div/span').text_content(timeout=1000)
    except Exception:
        is_re = '未认证'
    num_list = await page.query_selector_all('xpath=//*[@id="root"]/div/div[2]/div/div/div[2]/div[1]/div[1]/div[2]/div')
    text_list = await page.query_selector_all('xpath=//*[@id="root"]/div/div[2]/div/div/div[2]/div[1]/p[2]/span')

    if not is_re:
        is_re = '未认证'
    if not is_live:
        is_live = '未开播'

    data['认证信息'] = is_re
    data['是否开播'] = is_live
    data['昵称'] = user_name
    for i in ip:
        res_list = str(await i.text_content()).split('：')
        data['抖音号'] = res_list[1][:-4]
        data['IP属地'] = res_list[2]
    res_list = []
    for i in num_list:
        res_list.append(str(await i.text_content()))

    data.update({'num': res_list})
    for i in text_list:
        data['签名'] = await i.text_content()
    print({'data': data})

async def main():
    async with async_playwright() as playwright:
        await get_userinfo(playwright, 'https://www.douyin.com/user/MS4wLjABAAAAzzmS2TgIEvxGftMpWD13Ty8k5HmsjlGsLJ1yBUEm2Ew')

asyncio.run(main())