Python中五种不同解析库的应用与提取速度对比_Python

今天闲着没事，突发异想，相对比不同解析库的解析速度，本文中对比五种解析库

一、概述

1.scrapy自带css选择

2.scrapy自带xpath

3.lxml自带xpath from lxml import etree

4.bs4自带美味的汤 from bs4 import beautifulsoup

5.re 正则表达式

以这个文本为例，在循环解析链接标题时间情况，再循环499次并打印输出，约循环7500次，为了避免误差，考虑到cpu资源波动，那么python程序能够得到的cpu时间就会或多或少，从而导致运行时间变长或变短，所以每种方式运行三次。

当cpu爆满时

运行时间非常慢

当cpu充足时，同样是程序就会非常快

本程序都在cpu充足情况下运行

text ='''<div class="news_list zy_listbox">
                        <!-- 展示15条数据后出现分页 -->
                        <ul>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2565" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-010号</a><span>2024-07-18</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2564" rel="external nofollow"  rel="external nofollow" >单一来源采购公示</a><span>2024-07-15</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2563" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-009号（第二次）</a><span>2024-07-10</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2562" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-009号</a><span>2024-07-05</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2561" rel="external nofollow"  rel="external nofollow" >等保测评询价函</a><span>2024-05-28</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2560" rel="external nofollow"  rel="external nofollow" >单一来源采购公示</a><span>2024-05-24</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2559" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-007号（第二次）</a><span>2024-05-20</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2558" rel="external nofollow"  rel="external nofollow" >单一来源采购公示</a><span>2024-05-16</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2557" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-008号</a><span>2024-05-06</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2556" rel="external nofollow"  rel="external nofollow" >询价函</a><span>2024-04-30</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2555" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-007号</a><span>2024-04-22</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2554" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2023-004号（第二次）</a><span>2024-04-12</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2553" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-006号</a><span>2024-03-29</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2552" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-005号</a><span>2024-03-20</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2551" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-004号</a><span>2024-03-16</span></li>
                            
                        </ul>
                    </div>'''

二、探讨css选择器解析速度

t= time.time()
selector = selector(text=text)
for i in range(1,500):
    for l in selector.css('div.news_list>ul>li'):
        print(l.css('a::text').get())
        print(l.css('span::text').get())
        print(l.css('a::attr(href)').get())
css_parse = time.time()-t
print('css_parse用时：',css_parse)

第一次

第二次

第三次

三、探讨scrapy里的xpath解析速度

t= time.time()
selector = selector(text=text)
for i in range(1,500):
    for l in selector.xpath('//div[@class = "news_list zy_listbox"]/ul/li'):
        print(l.xpath('.//a/text()').get())
        print(l.xpath('.//span/text()').get())
        print(l.xpath('.//a/@href').get())
scrapy_xpath_parse = time.time()-t
print('scrapy_xpath_parse用时：',scrapy_xpath_parse)

第一次

第二次

第三次

四、探讨lxml里的xpath解析速度

t= time.time()
selector = etree.html(text)
for i in range(1,500):
    for l in selector.xpath('//div[@class = "news_list zy_listbox"]/ul/li'):
        print(l.xpath('.//a/text()')[0])
        print(l.xpath('.//span/text()')[0])
        print(l.xpath('.//a/@href')[0])
xpath_parse = time.time()-t
print('xpath_parse用时：',xpath_parse)

第一次

第二次

第三次

五、探讨bs4自带beautifulsoup解析速度

t = time.time()
soup = beautifulsoup(text, 'html.parser')
news_list = soup.select('div.news_list.zy_listbox ul li')
for i in range(1,500):
    for l in news_list:
        print( l.select_one('a').get_text(strip=true))
        print(l.select_one('span').get_text(strip=true))
        print(l.select_one('a')['href'])
bs_parse = time.time() - t
print('bs_parse用时：',bs_parse )

第一次

第二次

第三次

六、探讨re正则解析速度

t = time.time()
li = re.findall(' <li><b>·</b>.*?</li>',text)
for l in range(1,500):
    for i in li:
        print(re.search('>(.*?)</a>',i).group(1))
        print(re.search('<span>(.*?)</span>',i).group(1))
        print(re.search('<a href="(.*?)" rel="external nofollow"  rel="external nofollow" >',i).group(1))
re_time = time.time()-t
print('re：',re_time )

第一次

第二次

第三次

七、总结

经上述程序实验证明，速度排序:

re正则 > lxml xpath > bs4 beautifulsoup > scrapy css ≈ scrapy xpath

在本实验中，并不严谨，除cpu波动外，只选取了一种特定的html网页进行速度对比，没有考虑到不同类型和结构的网页对解析速度的影响。例如，网页中若包含大量嵌套的表单、复杂的javascript生成的动态内容等，甚至有个html藏在动态接口的json里面，不同解析工具的表现可能会大不相同，这使得实验结果存在一定的局限性。实际应用中还需要根据具体的html网页结构、解析需求以及性能指标等多方面因素综合考量，选择最合适的解析工具。

总程序

from lxml import etree
from scrapy import selector
from bs4 import beautifulsoup
import re
import requests
import time
text ='''<div class="news_list zy_listbox">
                        <!-- 展示15条数据后出现分页 -->
                        <ul>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2565" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-010号</a><span>2024-07-18</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2564" rel="external nofollow"  rel="external nofollow" >单一来源采购公示</a><span>2024-07-15</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2563" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-009号（第二次）</a><span>2024-07-10</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2562" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-009号</a><span>2024-07-05</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2561" rel="external nofollow"  rel="external nofollow" >等保测评询价函</a><span>2024-05-28</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2560" rel="external nofollow"  rel="external nofollow" >单一来源采购公示</a><span>2024-05-24</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2559" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-007号（第二次）</a><span>2024-05-20</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2558" rel="external nofollow"  rel="external nofollow" >单一来源采购公示</a><span>2024-05-16</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2557" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-008号</a><span>2024-05-06</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2556" rel="external nofollow"  rel="external nofollow" >询价函</a><span>2024-04-30</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2555" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-007号</a><span>2024-04-22</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2554" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2023-004号（第二次）</a><span>2024-04-12</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2553" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-006号</a><span>2024-03-29</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2552" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-005号</a><span>2024-03-20</span></li>
                                                        <li><b>·</b><a href="/index/article/detail.html?id=2551" rel="external nofollow"  rel="external nofollow" >青海省生态环境监测中心自行采购2024-004号</a><span>2024-03-16</span></li>
                            
                        </ul>
                    </div>'''
t= time.time()
selector = selector(text=text)
for i in range(1,500):
    for l in selector.css('div.news_list>ul>li'):
        print(l.css('a::text').get())
        print(l.css('span::text').get())
        print(l.css('a::attr(href)').get())
css_parse = time.time()-t
t= time.time()
selector = selector(text=text)
for i in range(1,500):
    for l in selector.xpath('//div[@class = "news_list zy_listbox"]/ul/li'):
        print(l.xpath('.//a/text()').get())
        print(l.xpath('.//span/text()').get())
        print(l.xpath('.//a/@href').get())
scrapy_xpath_parse = time.time()-t
t= time.time()
selector = etree.html(text)
for i in range(1,500):
    for l in selector.xpath('//div[@class = "news_list zy_listbox"]/ul/li'):
        print(l.xpath('.//a/text()')[0])
        print(l.xpath('.//span/text()')[0])
        print(l.xpath('.//a/@href')[0])
xpath_parse = time.time()-t
t = time.time()
soup = beautifulsoup(text, 'html.parser')
news_list = soup.select('div.news_list.zy_listbox ul li')
for i in range(1,500):
    for l in news_list:
        print( l.select_one('a').get_text(strip=true))
        print(l.select_one('span').get_text(strip=true))
        print(l.select_one('a')['href'])
bs_parse = time.time() - t
t = time.time()
li = re.findall(' <li><b>·</b>.*?</li>',text)
for l in range(1,500):
    for i in li:
        print(re.search('>(.*?)</a>',i).group(1))
        print(re.search('<span>(.*?)</span>',i).group(1))
        print(re.search('<a href="(.*?)" rel="external nofollow"  rel="external nofollow" >',i).group(1))
re_time = time.time()-t
print('css_parse用时：',css_parse)
print('scrapy_xpath_parse用时：',scrapy_xpath_parse)
print('xpath_parse用时：',xpath_parse)
print('bs_parse用时：',bs_parse )
print('re用时：',re_time )

各时间对比