python请求POST错误，会话问题？答案

【问题标题】：python requests POST error, session issue?python请求POST错误，会话问题？
【发布时间】：2020-12-13 08:19:30
【问题描述】：

我正在尝试通过 python 的requests 模拟以下浏览器操作：

登陆https://www.bundesanzeiger.de/pub/en/to_nlp_start
点击“更多搜索选项”
点击复选框“同时查找历史数据”（对应于 POST 参数：isHistorical: true）
点击“搜索净空头头寸”按钮
点击“Als CSV herunterladen”按钮下载csv文件

这是我必须模拟的代码：

import requests
import re

s = requests.Session()
r = s.get("https://www.bundesanzeiger.de/pub/en/to_nlp_start", verify=False, allow_redirects=True)

matches = re.search(
        r'form class="search-form" id=".*" method="post" action="\.(?P<appendtxt>.*)"',
        r.text
    )
request_url = f"https://www.bundesanzeiger.de/pub/en{matches.group('appendtxt')}"
sr = session.post(request_url, data={'isHistorical': 'true', 'nlp-search-button': 'Search net short positions'}, allow_redirects=True)

然而，即使sr 给了我一个status_code 200，当我检查sr.url 时它确实是一个错误，它显示https://www.bundesanzeiger.de/pub/en/error-404?9

再深入一点，我注意到上面的 request_url 解析为类似

https://www.bundesanzeiger.de/pub/en/nlp;wwwsid=EFEB15CD4ADC8932A91BA88B561A50E9.web07-pub?0-1.-nlp~filter~form~panel-form

但是当我在 Chrome 中检查请求 url 时，它实际上是

https://www.bundesanzeiger.de/pub/en/nlp?87-1.-nlp~filter~form~panel-form`

这里的87 似乎发生了变化，表明它是某个会话 ID，但是当我使用 requests 执行此操作时，它似乎无法正确解析。

知道我在这里缺少什么吗？

【问题讨论】：

标签： python session web-scraping python-requests session-cookies

【解决方案1】：

如果你勾选https://www.bundesanzeiger.de/robots.txt，这个网站不喜欢被索引。该网站可能拒绝访问机器人使用的默认用户代理。这可能会有所帮助：Python requests vs. robots.txt

【讨论】：

我尝试添加：HEADERS = { "User-Agent": "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko)" " Chrome/50.0.2661.75 Safari/537.36", "X-Requested-With": "XMLHttpRequest", "Host": 'www.bundesanzeiger.de' } ，然后添加s.headers.update(HEADERS)，但仍然遇到同样的问题。

【解决方案2】：

您可以尝试使用此脚本下载 CSV 文件：

import requests
from bs4 import BeautifulSoup


url = 'https://www.bundesanzeiger.de/pub/en/to_nlp_start'

data = {
    'fulltext': '',
    'positionsinhaber': '',
    'ermittent': '',
    'isin': '',
    'positionVon': '',
    'positionBis': '',
    'datumVon': '',
    'datumBis': '',
    'isHistorical': 'true',
    'nlp-search-button': 'Search+net+short+positions'
}

headers = {
    'Referer': 'https://www.bundesanzeiger.de/'
}

with requests.session() as s:
    soup = BeautifulSoup(s.get(url).content, 'html.parser')

    action = soup.find('form', action=lambda t: 'nlp~filter~form~panel-for' in t)['action']
    u = 'https://www.bundesanzeiger.de/pub/en' + action.strip('.')    

    soup = BeautifulSoup( s.post(u, data=data, headers=headers).content, 'html.parser' )

    a = soup.select_one('a[title="Download as CSV"]')['href']
    a = 'https://www.bundesanzeiger.de/pub/en' + a.strip('.')    

    print( s.get(a, headers=headers).content.decode('utf-8-sig') )

打印：

"Positionsinhaber","Emittent","ISIN","Position","Datum"
"Citadel Advisors LLC","LEONI AG","DE0005408884","0,62","2020-08-21"
"AQR Capital Management, LLC","Evotec SE","DE0005664809","1,10","2020-08-21"
"BlackRock Investment Management (UK) Limited","thyssenkrupp AG","DE0007500001","1,50","2020-08-21"
"BlackRock Investment Management (UK) Limited","Deutsche Lufthansa Aktiengesellschaft","DE0008232125","0,75","2020-08-21"
"Citadel Europe LLP","TAG Immobilien AG","DE0008303504","0,70","2020-08-21"
"Davidson Kempner European Partners, LLP","TAG Immobilien AG","DE0008303504","0,36","2020-08-21"
"Maplelane Capital, LLC","VARTA AKTIENGESELLSCHAFT","DE000A0TGJ55","1,15","2020-08-21"


...and so on.

【讨论】：