Python中五种不同解析库的应用与提取速度对比
作者：失败又激情的man
这篇文章主要为大家详细介绍了Python中五种不同解析库的应用与提取速度对比,文中的示例代码讲解详细,感兴趣的小伙伴可以跟随小编一起学习一下
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
到此这篇关于Python中五种不同解析库的应用与提取速度对比的文章就介绍到这了,更多相关Python解析库应用内容请搜索脚本之家以前的文章或继续浏览下面的相关文章希望大家以后多多支持脚本之家！