Today's Python introductory tutorial , with the main article tell you how to obtain all the public numbers!
We usually read the article numbers public will encounter a problem - a bad experience reading the article history.
We know the common ways crawling public numbers in two ways: to acquire by Sogou search, the disadvantage is only for the latest push Ten article. By micro-channel public number of material management, access to public article number. The disadvantage is the need to apply for their own public number.
Today introduced a method to obtain the number of articles by way of public Ethereal PC end micro letter. Compared to other methods is very convenient.
As shown above, the network packets captured by micro-information request letter, we find that every article is going to be the drop-down refresh request mp.weixin.qq.com/mp/xxx (No. public not to add the Home link, xxx represents profile_ext) this interface.
After several tests were used to analyze several parameters
- Unique id number between users and the public: __biz
- uin: user privacy id
- key: the secret key request, only some time will fail
- offset: Offset
- count: the number of requests per
Data are as follows
{
"ret": 0,
"errmsg": "ok", # 请求状态
"msg_count": 10, # 信息条数
"can_msg_continue": 1, # 是否还可以继续获取，1代表可以。0代表不可以，也就是最后一页
"general_msg_list": "{"list":[]}", # 公众号文本信息
"next_offset": 20,
"video_count": 1,
"use_video_tab": 1,
"real_type": 0,
"home_page_list": []
}
部分代码如下
params = {
'__biz': biz,
'uin': uin,
'key': key,
'offset': offset,
'count': count,
'action': 'getmsg',
'f': 'json'
}
headers = {
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.131 Safari/537.36'
}
response = requests.get(url=url, params=params, headers=headers)
resp_json = response.json()
if resp_json.get('errmsg') == 'ok':
resp_json = response.json()
# 是否还有分页数据， 用于判断return的值
can_msg_continue = resp_json['can_msg_continue']
# 当前分页文章数
msg_count = resp_json['msg_count']
general_msg_list = json.loads(resp_json['general_msg_list'])
list = general_msg_list.get('list')
print(list, "**************")
最后打印的list就是公众号的文章信息详情。包括标题(titile)、摘要(digest)、文章地址(content_url)、阅读原文地址(source_url)、封面图(cover)、作者(author)等等...
输出结果如下：
[{
"comm_msg_info": {
"id": 1000000038,
"type": 49,
"datetime": 1560474000,
"fakeid": "3881067844",
"status": 2,
"content": ""
},
"app_msg_ext_info": {
"title": "入门爬虫，这一篇就够了！！！",
"digest": "入门爬虫，这一篇就够了！！！",
"content": "",
"fileid": 0,
"content_url": "http:XXXXXX",
"source_url": "",
"cover": "I5kME6BVXeLibZDUhsiaEYiaX7zOoibxa9sb4stIwrfuqID5ttmiaoVAFyxKF6IjOCyl22vg8n2NPv98ibow\/0?wx_fmt=jpeg",
"subtype": 9,
"is_multi": 0,
"multi_app_msg_item_list": [],
"author": "Python3X",
"copyright_stat": 11,
"duration": 0,
"del_flag": 1,
"item_show_type": 0,
"audio_fileid": 0,
"play_url": "",
"malicious_title_reason_id": 0,
"malicious_content_type": 0
}
},{...},{...},{...},{...},{...},{...},{...},{...},{...}]
获取数据之后，可以保存到数据库中，也可以将文章保存在PDF中。
1、保存在Mongo中
# Mongo配置
conn = MongoClient('127.0.0.1', 27017)
db = conn.wx #连接wx数据库，没有则自动创建
mongo_wx = db.article #使用article集合，没有则自动创建
for i in list:
app_msg_ext_info = i['app_msg_ext_info']
# 标题
title = app_msg_ext_info['title']
# 文章地址
content_url = app_msg_ext_info['content_url']
# 封面图
cover = app_msg_ext_info['cover']
# 发布时间
datetime = i['comm_msg_info']['datetime']
datetime = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(datetime))
mongo_wx.insert({
'title': title,
'content_url': content_url,
'cover': cover,
'datetime': datetime
})
结果如下
2、导入到PDF文件中
Python3中常用的操作PDF的库有python-pdf和pdfkit。我用了pdfkit这个模块导出pdf文件。
pdfkit是工具包Wkhtmltopdf的封装类，因此需要安装Wkhtmltopdf才能使用。
可以访问 https://wkhtmltopdf.org/downloads.html 下载和操作系统匹配的工具包。
实现代码也比较简单，只需要传入导入文件的url即可。
安装pdfkit库
pip3 install pdfkit -i http://pypi.douban.com/simple --trusted-host pypi.douban.com
import pdfkit
pdfkit.from_url('公众号文章地址', 'out.pdf')
运行之后成功导出pdf文件。
完整代码
import requests
import json
import time
from pymongo import MongoClient
url = 'http://mp.weixin.qq.com/mp/xxx'（公众号不让添加主页链接，xxx表示profile_ext)
# Mongo配置
conn = MongoClient('127.0.0.1', 27017)
db = conn.wx #连接wx数据库，没有则自动创建
mongo_wx = db.article #使用article集合，没有则自动创建
def get_wx_article(biz, uin, key, index=0, count=10):
offset = (index + 1) * count
params = {
'__biz': biz,
'uin': uin,
'key': key,
'offset': offset,
'count': count,
'action': 'getmsg',
'f': 'json'
}
headers = {
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.131 Safari/537.36'
}
response = requests.get(url=url, params=params, headers=headers)
resp_json = response.json()
if resp_json.get('errmsg') == 'ok':
resp_json = response.json()
# 是否还有分页数据， 用于判断return的值
can_msg_continue = resp_json['can_msg_continue']
# 当前分页文章数
msg_count = resp_json['msg_count']
general_msg_list = json.loads(resp_json['general_msg_list'])
list = general_msg_list.get('list')
print(list, "**************")
for i in list:
app_msg_ext_info = i['app_msg_ext_info']
# 标题
title = app_msg_ext_info['title']
# 文章地址
content_url = app_msg_ext_info['content_url']
# 封面图
cover = app_msg_ext_info['cover']
# 发布时间
datetime = i['comm_msg_info']['datetime']
datetime = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(datetime))
mongo_wx.insert({
'title': title,
'content_url': content_url,
'cover': cover,
'datetime': datetime
})
if can_msg_continue == 1:
return True
return False
else:
print('获取文章异常...')
return False
if __name__ == '__main__':
biz = 'Mzg4MTA2Nzg0NA=='
uin = 'NDIyMTI5NDM1'
= Key '20a680e825f03f1e7f38f326772e54e7dc0fd02ffba17e92730ba3f0a0329c5ed310b0bd55b3c0b1f122e5896c6261df2eaea4036ab5a5d32dbdbcb0a638f5f3605cf1821decf486bb6eb4d92d36c620'
index = 0
the while. 1:
Print (F '. No. start crawling public {index + 1} p articles')
In Flag = get_wx_article (BIZ, UIN, Key, index = index)
# prevent harmony, suspended 8 seconds
the time.sleep (8)
index + = 1
IF not Flag:
Print ( '. No public posts have been crawling all finished, quit the program')
BREAK
Print (f '.......... ready to catch No. taken public {index + 1} page articles. ')