How to Automate WeChat Public Account Metrics Scraping with Python and MitmProxy

What is MitmProxy

Mitmproxy is an HTTP/HTTPS proxy with features similar to Fiddler, providing a command‑line interface (mitmdump) and a web UI (mitmweb) for inspecting traffic.

Installation and Setup

Install via pip install mitmproxy and start the web UI with mitmweb. Configure your mobile device to use the PC’s IP address and port 8080.

Scraping Read, Like, and View Counts

After installing the certificate on the phone, open any WeChat article. The request URL for metrics is https://mp.weixin.qq.com/mp/getappmsgext. The request requires the article URL, User‑Agent, cookie, and a body containing __biz, mid, idx, sn, plus fixed parameters is_only_read=1, is_temp_url=0, appmsg_type=9, and a time‑limited appmsg_token.

Python Implementation

The Articles class encapsulates the logic. Its read_like_nums method posts to the getappmsgext endpoint and returns read, old_like, and like numbers.

# articles.py
import html, requests, utils

class Articles(object):
    """Article information"""
    def __init__(self, appmsg_token, cookie):
        self.appmsg_token = appmsg_token
        self.headers = {
            "User-Agent": "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/57.0.2987.132 MQQBrowser/6.2 Mobile",
            "Cookie": cookie
        }
        self.data = {
            "is_only_read": "1",
            "is_temp_url": "0",
            "appmsg_type": "9",
        }
    def read_like_nums(self, article_url):
        """Get data"""
        appmsgstat = self.get_appmsgext(article_url)["appmsgstat"]
        return appmsgstat["read_num"], appmsgstat["old_like_num"], appmsgstat["like_num"]
    def get_params(self, article_url):
        """Extract query parameters from article URL"""
        article_url = html.unescape(article_url)
        url_params = utils.str_to_dict(urlsplit(article_url).query, "&", "=")
        return url_params
    def get_appmsgext(self, article_url):
        """Request read numbers"""
        url_params = self.get_params(article_url)
        appmsgext_url = f"https://mp.weixin.qq.com/mp/getappmsgext?appmsg_token={self.appmsg_token}&x5=0"
        self.data.update(url_params)
        appmsgext_json = requests.post(appmsgext_url, headers=self.headers, data=self.data).json()
        if "appmsgstat" not in appmsgext_json:
            raise Exception(appmsgext_json)
        return appmsgext_json

if __name__ == '__main__':
    info = Articles('YOUR_APPMSG_TOKEN', 'YOUR_COOKIE')
    a, b, c = info.read_like_nums('ARTICLE_URL')
    print(a, b, c)

Dynamic Cookie and Token Retrieval

Because appmsg_token and cookies expire, a Mitmproxy script ( write_cookie.py) captures the getappmsgext URL and cookies, writes them to a file, and exits. ReadCookie parses this file to extract appmsg_token, biz, and the cookie string.

# write_cookie.py
import urllib, sys
from mitmproxy import http

class WriterCookie:
    """Mitmproxy script that writes URL and cookies to a file"""
    def __init__(self, outfile: str) -> None:
        self.f = open(outfile, "w")
    def response(self, flow: http.HTTPFlow) -> None:
        url = urllib.parse.unquote(flow.request.url)
        if "mp.weixin.qq.com/mp/getappmsgext" in url:
            self.f.write(url + '
')
            self.f.write(str(flow.request.cookies))
            self.f.close()
            exit()

addons = [WriterCookie(sys.argv[4])]

# read_cookie.py
import re, os

class ReadCookie(object):
    """Parse the file generated by write_cookie.py"""
    def __init__(self, outfile):
        self.outfile = outfile
    def parse_cookie(self):
        f = open(self.outfile)
        lines = f.readlines()
        appmsg_token = re.findall("appmsg_token.+?&", lines[0])[0].split('=')[1][:-1]
        biz = re.findall('__biz.+?&', lines[0])[0].split('__biz=')[1][:-1]
        cookie_str = '; '.join(lines[1][15:-2].split('], ['))
        cookie_str = cookie_str.replace("'", '').replace(', ', '=')
        return appmsg_token, biz, cookie_str
    def write_cookie(self, outfile):
        path = os.path.split(os.path.realpath(__file__))[0]
        command = f"mitmdump -s {path}/write_cookie.py -w {outfile} mp.weixin.qq.com/mp/getappmsgext"
        os.system(command)

Batch Fetching Articles

Pagination is performed by calling

https://mp.weixin.qq.com/mp/profile_ext?action=getmsg&__biz=…&offset=10&count=10

. The WxCrawler class builds the request URL, converts the cookie to a dict, and iterates through pages, printing each article’s title and URL.

# wxCrawler.py
import os, requests, json, urllib3, utils

class WxCrawler(object):
    urllib3.disable_warnings()
    def __init__(self, appmsg_token, biz, cookie, begin_page_index=0, end_page_index=100):
        self.begin_page_index = begin_page_index
        self.end_page_index = end_page_index
        self.num = 1
        self.appmsg_token = appmsg_token
        self.biz = biz
        self.headers = {
            "User-Agent": "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/57.0.2987.132 MQQBrowser/6.2 Mobile",
            "Cookie": cookie
        }
        self.cookie = cookie
    def article_list(self, context):
        articles = json.loads(context).get('general_msg_list')
        return json.loads(articles)
    def run(self):
        page_url = f"https://mp.weixin.qq.com/mp/profile_ext?action=getmsg&__biz={self.biz}&f=json&offset={{}}&count=10&is_ok=1&scene=&uin=777&key=777&pass_ticket={{}}&wxtoken=&appmsg_token={self.appmsg_token}&x5=0f=json"
        wx_dict = utils.str_to_dict(self.cookie, '; ', '=')
        response = requests.get(page_url.format(self.begin_page_index*10, wx_dict['pass_ticket']), headers=self.headers, verify=False)
        articles = self.article_list(response.text)
        for a in articles['list']:
            if 'app_msg_ext_info' in a and a['app_msg_ext_info'].get('content_url'):
                print(f"{self.num}条", a['app_msg_ext_info']['title'], a['app_msg_ext_info']['content_url'])
            if 'app_msg_ext_info' in a:
                for m in a['app_msg_ext_info'].get('multi_app_msg_item_list', []):
                    print(f"{self.num}条", m['title'], a['content_url'])
            self.num += 1
        self.is_exit_or_continue()
        self.run()
    def is_exit_or_continue(self):
        self.begin_page_index += 1
        if self.begin_page_index > self.end_page_index:
            os._exit(0)

Running the Whole Pipeline

The main script launches ReadCookie to refresh the token, then starts WxCrawler to collect all historical articles.

# main.py
from read_cookie import ReadCookie
from wxCrawler import WxCrawler

if __name__ == '__main__':
    rc = ReadCookie('cookie.txt')
    rc.write_cookie('cookie.txt')
    appmsg_token, biz, cookie_str = rc.parse_cookie()
    wx = WxCrawler(appmsg_token, biz, cookie_str)
    wx.run()

Conclusion

The method automates most of the data‑collection steps, though a manual refresh of the public account page is still required when the token expires.