BS4
BeautifulSoup是用来从HTML or XML中提取数据的Python lib。BeautifulSoup将文档转化为树形结构（DOM），每个节点都是下述四种类型的Python对象：
- BeautifulSoup
<class 'bs4.BeatifulSoup'> - Tag
<class 'bs4.element.Tag'> - NavigableString
<class 'bs4.element.NavigableString'> - Comment
<class 'bs4.element.Comment'>
从集合角度理解以上4中类的关系（类概念上并不准确）
- BeautifulSoup 为全集（将Document以入参传入生成BeautifulSoup object）， 包含 Tag子集
- Tag 包含 NavigableString 子集
- Comment 为 NavigableString 特殊集合
Usage
BeautifulSoup的第一个入参是Document，第二个入参指定Document parser 类型.
from bs4 import BeautifulSoup
import requests, re
url = 'http://m.kdslife.com/club/'
# get whole HTTP response
response = requests.get(url)
# args[0] is HTML document, args[1] select LXML parser. returned BeautifulSoup object
soup = BeautifulSoup( response.text, 'lxml')
print soup.name
# [document]'
print type(soup)
# <class 'bs4.BeatifulSoup'>
Sample codes for Tag objects
# BeutifulSoup --> Tag
# get the Tag object(title)
res = soup.title
print res
# <title>KDS Life</title>
res = soup.title.name
print res
# title
# attribules of a Tag object
res = soup.section
print type(res)
# <class 'bs4.element.Tag'>
print res['class']
# ['forum-head-hot', 'clearfix']
# All the attributes of section Tag object, returned a dict
print res.attrs
#{'class': ['forum-head-hot', 'clearfix']}
Sample codes for NavigableString object
# NavigableString object describes the string in Tag object
res = soup.title
print res.string
# KDS Life
print type(res.string)
# <class 'bs4.element.NavigableString'>
Sample codes for Comment object
# Comment, is a special NavigableString object
markup = "<b><!--Hey, buddy. Want to buy a used parser?--></b>"
soup = BeautifulSoup(markup)
comment = soup.b.string
print type(comment)
# <class 'bs4.element.Comment'>
BS4 Parser
按照优先顺序自动解析，'lxml' --> 'html5lib' --> 'html.parser'
常用Tag对象方法
find_all()
find_all(name,attrs,recursive,text,**kwargs) 不解释，直接看代码
# filter, returned a matching list
# returned [] if matching nothing
title = soup.find_all('title')
print title
#[<title>Google</title>]
res = soup.find_all('div', 'topAd')
print res
# find all the elements whose id is 'gb-main'
res = soup.find_all(id='topAd')
print res
#[<div id="topAd">...</div>]
# find all the elements with 'img' tag and 'src' attribute matching the specific pattern
res = soup.find_all('img', src=re.compile(r'^http://club-img',re.I))
print res
# [,
#...]
select()
# css selector
# select those whose tag's id = wrapperto
res = soup.select('#wrapperto')
print res
# [<div class="swiper-wrapper clearfix" id="wrapperto"></div>]
# select those 'img' tags who have 'src' attribute
res = soup.select('img[src]')
print res
#[, <im
#g src="http://club-img.kdslife.com/attach/1k0/gs/a/o41gty-1coa.png@0o_1l_600w_90q.src"/>]
# select those 'img' tags whose 'src' attribute is
res = soup.select('img[src=http://icon.pch-img.net/kds/club_m/club/icon/user1.png]')
print res
# []
Other
# get_text()
markup = '<a href="http://example.com/">\n a link to <i>example.com</i>\n</a>'
soup = BeautifulSoup(markup,'lxml')
res = soup.get_text()
print res
# a link to example.com
res = soup.i.get_text()
print res
# example.com
# .stripped_string
res = soup.stripped_strings
print list(res)
# [u'a link to', u'example.com']
最后贴上一个简单的KDS图片爬虫
Note
- BeautifulSoup进行了编码检测并自动转为Unicode. soup.original_encoding属性来获取自动识别编码的结果。
- Input converts to unicode, output encodes with utf-8
- 在BS使用中，可配合 XPath expression使用
