lxml 通常都是用etree解析，书上却用的是lxml.html，用我浅薄的英文看了下，貌似etree功能更多，lxml.html专解析html,带了点特殊的方法....

etree

from lxml import etree html = etree.HTML(sample) result1 = etree.tostring(html,pretty_print =True) print(result1)

lxml.html

import lxml.html html = lxml.html.fromstring(sample) result2 = lxml.html.tostring(html,pretty_print =True) print(result2)

带编码的etree.HTML

import lxml.etree as le with open(‘books.xml’,’r’,encoding=‘utf-8’) as b: contents=b.read() contents_html=le.HTML(contents.encode(‘utf-8’)) co_ht_xpath=contents_html.xpath(‘/*’) print(co_ht_xpath)

得到的结果基本一致,etree结果多了个html、body标签,虽然我都加了pretty_print ,实际输出还是乱作一团,根本没有整齐输出的效果,网上好多人都是这样,还没找到解决方法,还是Beautiful Soup4的prettify() 好用 link[0].attrib {'herf':'link1.html'} 使用attrib属性得到一个字典 link[0].attrib['herf'] 结果 和 html.xpath('//a/@href')[0] 一样 <wbr> <wbr> link[0].text 'aa' link[0].text_content() <wbr> (etree 没有这个属性,lxml.html独有) 'aa' <wbr> lxml.html.diff可以比较2个文件的差别

lxml.html.clean clean_html 用来清除未被解析文件中所有可疑?内容

参考: lxml.html 与 etree