http://t.cn/RS8uxtE

mp.py

# coding=utf-8
import re
import ast
import time
from datetime import datetime
from collections import defaultdict
from http.cookies import SimpleCookie

import requests
from bs4 import BeautifulSoup

url = 'https://mp.weixin.qq.com/misc/appmsgcomment?action=list_comment&type={type}&begin={start}&count=10&comment_id=3196803350&token=708943601&lang=zh_CN&mp_version=7'
ua = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36'
js_regex = 'list : (.*),\W+total_count'

c = SimpleCookie()
with open('cookies.txt') as f:
    c.load(f.read())

cookies = {}
for key, morsel in c.items():
    cookies[key] = morsel.value

comments = []

for type in (0, 1):
    start = 0
    while 1:
        r = requests.get(url.format(type=type, start=start*10),
                         cookies=cookies, headers={'user-agent': ua})
        soup = BeautifulSoup(r.text, 'lxml')
        js = [js for js in soup.find_all('script')
              if 'wx.cgiData' in str(js)][0]
        match = re.compile(js_regex).search(str(js))
        data = ast.literal_eval(match.group(1).replace(
            'false', 'False').replace('true', 'True'))
        comments_ = data['comment']
        comments.extend(comments_)
        if len(comments_) < 10:
            break
        start += 1
        time.sleep(0.3)

sorted(comments, key=lambda x:x['post_time'])
NUMBER = int(input('输入幸运数字❯ '))
COUNT = 5
NUMERIC_REGEX = re.compile(r'(\d){2}')
END_TIMESTAMP = datetime(2017, 6, 9)
SEEN_USERS = set()
NUMBER_MAP = defaultdict(list)

for comment in comments:
    created = datetime.fromtimestamp(comment['post_time'])
    if created >= END_TIMESTAMP:
        break
    content = re.sub('<[^<]+?>', '', comment['content'])
    match = NUMERIC_REGEX.search(content)
    if not match:
        continue
    number = match.group()
    _content = content.replace(number, '', 1)
    if NUMERIC_REGEX.search(_content):
        continue

    nickname = comment['nick_name']
    if nickname in SEEN_USERS:
        continue
    SEEN_USERS.add(nickname)

    NUMBER_MAP[int(number)].append(
        (nickname, comment['id'], str(created), content))

values = NUMBER_MAP.get(int(NUMBER), [])
if not len(values) >= COUNT:
    offset = 1
    while 1:
        values += sorted(sum([NUMBER_MAP.get(NUMBER + o, [])
                              for o in (offset, -offset)], []),
                         key=lambda x: x[1])
        if len(values) > COUNT:
            break
        offset += 1
values = values[:COUNT]

print('\n如下{}位同学获得《流畅的Python》:\n'.format(COUNT))

for val in values:
    print('用户: {:<6}\t楼层: {:<4}  参与时间: {:<10}  \n'
          '评论内容：{}'.format(*val))