python爬虫爬取汽车页面信息,并附带分析(静态爬虫)

时间:2021-11-03 16:19:27

环境:

windows,python3.4

参考链接:

https://blog.csdn.net/weixin_36604953/article/details/78156605

代码:(亲测可以运行)

 import requests
from bs4 import BeautifulSoup
import re
import random
import time # 爬虫主函数
def mm(url):
# 设置目标url,使用requests创建请求
header = {
"User-Agent": "Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36"}
req0 = requests.get(url=url, headers=header)
req0.encoding = "gb18030" # 解决乱码问题
html0 = req0.text # 使用BeautifulSoup创建html代码的BeautifulSoup实例,存为soup0
soup0 = BeautifulSoup(html0, "html.parser") # 获取最后一页数字,对应-122(对照前一小节获取尾页的内容看你就明白了)
total_page = int(soup0.find("div", class_="pagers").findAll("a")[-2].get_text())
myfile = open("aika_qc_gn_1_1_1.txt", "a", encoding='gb18030', errors='ignore') # 解决乱码问题
print("user", " 来源", " 认为有用人数", " 类型", " comment")
NAME = "user" + " 来源" + " 认为有用人数" + " 类型" + " comment"
myfile.write(NAME + "\n")
for i in list(range(1, total_page + 1)):
# 设置随机暂停时间
stop = random.uniform(1, 3) url = "http://newcar.xcar.com.cn/257/review/0/0_" + str(i) + ".htm"
req = requests.get(url=url, headers=header)
req.encoding = "gb18030" # 解决乱码问题
html = req.text soup = BeautifulSoup(html, "html.parser")
contents = soup.find('div', class_="review_comments").findAll("dl")
l = len(contents)
for content in contents:
tiaoshu = contents.index(content)
try:
ss = "正在爬取第%d页的第%d的评论,网址为%s" % (i, tiaoshu + 1, url)
print(ss) # 正在爬取的条数
try: # 点评角度
comment_jiaodu = content.find("dt").find("em").find("a").get_text().strip().replace("\n",
"").replace(
"\t", "").replace("\r", "")
except:
comment_jiaodu = "sunny"
try: # 点评类型
comment_type0 = content.find("dt").get_text().strip().replace("\n", "").replace("\t", "").replace(
"\r",
"")
comment_type1 = comment_type0.split("【")[1]
comment_type = comment_type1.split("】")[0]
except:
comment_type = "sunny" # 认为该条评价有用的人数
try:
useful = int(
content.find("dd").find("div", class_="useful").find("i").find(
"span").get_text().strip().replace(
"\n", "").replace("\t", "").replace("\r", ""))
except:
useful = "sunny" # 评论来源
try:
comment_region = content.find("dd").find("p").find("a").get_text().strip().replace("\n",
"").replace(
"\t", "").replace("\r", "")
except:
comment_region = "sunny" # 评论者名称
try:
user = \
content.find("dd").find("p").get_text().strip().replace("\n", "").replace("\t", "").replace(
"\r",
"").split(
":")[-1]
except:
user = "sunny" # 评论内容
try:
comment_url = content.find('dt').findAll('a')[-1]['href']
urlc = comment_url
headerc = {
"User-Agent": "Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36"}
reqc = requests.get(urlc, headers=headerc)
htmlc = reqc.text
soupc = BeautifulSoup(htmlc, "html.parser") comment0 = \
soupc.find('div', id='mainNew').find('div', class_='maintable').findAll('form')[1].find('table',
class_='t_msg').findAll(
'tr')[1]
try:
comment = comment0.find('font').get_text().strip().replace("\n", "").replace("\t", "")
except:
comment = "sunny"
try:
comment_time = soupc.find('div', id='mainNew').find('div', class_='maintable').findAll('form')[
1].find('table', class_='t_msg').find('div',
style='padding-top: 4px;float:left').get_text().strip().replace(
"\n", "").replace(
"\t", "")[4:]
except:
comment_time = "sunny"
except:
try:
comment = \
content.find("dd").get_text().split("\n")[-1].split('\r')[-1].strip().replace("\n",
"").replace(
"\t", "").replace("\r", "").split(":")[-1]
except:
comment = "sunny" time.sleep(stop)
print(user, comment_region, useful, comment_type, comment) tt = user + " " + comment_region + " " + str(useful) + " " + comment_type + " " + comment
myfile.write(tt + "\n")
except Exception as e:
print(e)
s = "爬取第%d页的第%d的评论失败,网址为%s" % (i, tiaoshu + 1, url)
print(s)
pass
myfile.close() # 统计评论分布
def fenxi():
myfile = open("aika_qc_gn_1_1_1.txt", "r")
good = 0
middle = 0
bad = 0
nn = 0
for line in myfile:
commit = line.split(" ")[3]
if commit == "好评":
good = good + 1
elif commit == "中评":
middle = middle + 1
elif commit == "差评":
bad = bad + 1
else:
nn = nn + 1
count = good + middle + bad + nn
g = round(good / (count - nn) * 100, 2)
m = round(middle / (count - nn) * 100, 2)
b = round(bad / (count - nn) * 100, 2)
n = round(nn / (count - nn) * 100, 2)
print("好评占比:", g)
print("中评占比:", m)
print("差评占比:", b)
print ("未评论:", n) url = "http://newcar.xcar.com.cn/257/review/0.htm"
mm(url)
fenxi()

BeautifulSoup神器

Python一个第三方库bs4中有一个BeautifulSoup库,是用于解析html代码的,换句话说就是可以帮助你更方便的通过标签定位你需要的信息。这里只介绍两个比较关键的方法:

1、find方法和findAll方法: 
首先,BeautifulSoup会先将整个html或者你所指定的html代码编程一个BeautifulSoup对象的实例(不懂对象和实例不要紧,你只要把它当作是一套你使用F12看到的树形html代码代码就好),这个实例可以使用很多方法,最常用的就是find和findAll,二者的功能是相同的,通过find( )的参数,即find( )括号中指定的标签名,属性名,属性值去搜索对应的标签,并获取它,不过find只获取搜索到的第一个标签,而findAll将会获取搜索到的所有符合条件的标签,放入一个迭代器(实际上是将所有符合条件的标签放入一个list),findAll常用于兄弟标签的定位,如刚才定位口碑信息,口碑都在dl标签下,而同一页的10条口碑对应于10个dl标签,这时候用find方法只能获取第一个,而findAll会获取全部的10个标签,存入一个列表,想要获取每个标签的内容,只需对这个列表使用一个for循环遍历一遍即可。

2、get_text()方法: 
使用find获取的内容不仅仅是我们需要的内容,而且包括标签名、属性名、属性值等,比如使用find方法获取"<Y yy='aaa'>xxxx</Y>" 的内容xxxx,使用find后,我们会得到整个"<Y yy='aaa'>xxxx</Y>",十分冗长,实际我们想要的仅仅是这个标签的内容xxxx,因此,对使用find方法后的对象再使用get_text( )方法,就可以得到标签的内容了,对应到这里,我们通过get_text( )方法就可以得到xxxx了。