在抓取网页的时候只想抓取主要的文本框，例如 csdn 中的主要文本框为下图红色框：

抓取的思想是，利用 bs4 查找所有的 div，用正则筛选出每个 div 里面的中文，找到中文字数最多的 div 就是属于正文的 div 了。定义一个抓取的头部抓取网页内容：

import requests

headers = {

    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/47.0.2526.106 Safari/537.36',

    'Host': 'blog.csdn.net'}

session = requests.session()

def getHtmlByRequests(url):

    headers.update(

        dict(Referer=url, Accept="*/*", Connection="keep-alive"))

    htmlContent = session.get(url=url, headers=headers).content

    return htmlContent.decode("utf-8", "ignore")

识别每个 div 中文字的正则：

import re

# 统计中文字数

def countContent(string):

    pattern = re.compile(u'[\u1100-\uFFFD]+?')

    content = pattern.findall(string)

    return content

遍历每一个 div ，利用正则判断里面中文的字数长度，找到长度最长的 div ：

# 分析页面信息

def analyzeHtml(html):

    # 初始化网页

    soup = BeautifulSoup(html, "html.parser")

    part = soup.select('div')

    match = ""

    for paragraph in part:

        content = countContent(str(paragraph))

        if len(content) > len(match):

            match = str(paragraph)

    return match

得到主要的 div 后，提取里面的文字出来：

def main():

    url = "http://blog.csdn.net/"

    html = getHtmlByRequests(url)

    mainContent = analyzeHtml(html)

    soup = BeautifulSoup(mainContent, "html.parser")

    print(soup.select('div')[0].text)

完整的代码如下：

#!/usr/bin/env python

# -*- coding: utf-8 -*-

from bs4 import BeautifulSoup

import requests

import re

headers = {

    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/47.0.2526.106 Safari/537.36',

    'Host': 'blog.csdn.net'}

session = requests.session()

def getHtmlByRequests(url):

    headers.update(

        dict(Referer=url, Accept="*/*", Connection="keep-alive"))

    htmlContent = session.get(url=url, headers=headers).content

    return htmlContent.decode("utf-8", "ignore")

# 统计中文字数

def countContent(string):

    pattern = re.compile(u'[\u1100-\uFFFD]+?')

    content = pattern.findall(string)

    return content

# 分析页面信息

def analyzeHtml(html):

    # 初始化网页

    soup = BeautifulSoup(html, "html.parser")

    part = soup.select('div')

    match = ""

    for paragraph in part:

        content = countContent(str(paragraph))

        if len(content) > len(match):

            match = str(paragraph)

    return match

def main():

    url = "http://blog.csdn.net/"

    html = getHtmlByRequests(url)

    mainContent = analyzeHtml(html)

    soup = BeautifulSoup(mainContent, "html.parser")

    print(soup.select('div')[0].text)

if __name__ == '__main__':

    main()