python3如何解析html

解析html是爬虫后的重要的一个处理数据的环节。一下记录解析html的几种方式。

先介绍基础的辅助函数，主要用于获取html并输入解析后的结束。

#把传递解析函数，便于下面的修改
def get_html(url, paraser=bs4_paraser):
 headers = {
  'Accept': '*/*',
  'Accept-Encoding': 'gzip, deflate, sdch',
  'Accept-Language': 'zh-CN,zh;q=0.8',
  'Host': 'www.360kan.com',
  'Proxy-Connection': 'keep-alive',
  'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/52.0.2743.116 
  Safari/537.36'
 }
 request = urllib2.Request(url, headers=headers)
 response = urllib2.urlopen(request)
 response.encoding = 'utf-8'
 if response.code == 200:
  data = StringIO.StringIO(response.read())
  gzipper = gzip.GzipFile(fileobj=data)
  data = gzipper.read()
  value = paraser(data) # open('E:/h5/haPkY0osd0r5UB.html').read()
  return value
 else:
  pass
  
  
value = get_html('http://www.360kan.com/m/haPkY0osd0r5UB.html', paraser=lxml_parser)
for row in value:
 print row

1、lxml.html的方式进行解析。

The lxml XML toolkit is a Pythonic binding for the C libraries libxml2 and libxslt. It is unique in that it combines the speed and XML feature completeness of these libraries with the simplicity of a native Python API, mostly compatible but superior to the well-known ElementTree API. The latest release works with all CPython versions from 2.6 to 3.5. See the introduction for more information about background and goals of the lxml project. Some common questions are answered in the FAQ. [官网](http://lxml.de/)

def lxml_parser(page):
 data = []
 doc = etree.HTML(page)
 all_div = doc.xpath('//div[@class="yingping-list-wrap"]')
 for row in all_div:
  # 获取每一个影评，即影评的item
  all_div_item = row.xpath('.//div[@class="item"]') # find_all('div', attrs={'class': 'item'})
  for r in all_div_item:
   value = {}
   # 获取影评的标题部分
   title = r.xpath('.//div[@class="g-clear title-wrap"][1]')
   value['title'] = title[0].xpath('./a/text()')[0]
   value['title_href'] = title[0].xpath('./a/@href')[0]
   score_text = title[0].xpath('./div/span/span/@style')[0]
   score_text = re.search(r'\d+', score_text).group()
   value['score'] = int(score_text) / 20
   # 时间
   value['time'] = title[0].xpath('./div/span[@class="time"]/text()')[0]
   # 多少人喜欢
   value['people'] = int(
     re.search(r'\d+', title[0].xpath('./div[@class="num"]/span/text()')[0]).group())
   data.append(value)
 return data

相关推荐：《Python入门教程》

2、使用BeautifulSoup，不多说了，大家网上找资料看看。

def bs4_paraser(html):
 all_value = []
 value = {}
 soup = BeautifulSoup(html, 'html.parser')
 # 获取影评的部分
 all_div = soup.find_all('div', attrs={'class': 'yingping-list-wrap'}, limit=1)
 for row in all_div:
  # 获取每一个影评，即影评的item
  all_div_item = row.find_all('div', attrs={'class': 'item'})
  for r in all_div_item:
   # 获取影评的标题部分
   title = r.find_all('div', attrs={'class': 'g-clear title-wrap'}, limit=1)
   if title is not None and len(title) > 0:
    value['title'] = title[0].a.string
    value['title_href'] = title[0].a['href']
    score_text = title[0].div.span.span['style']
    score_text = re.search(r'\d+', score_text).group()
    value['score'] = int(score_text) / 20
    # 时间
    value['time'] = title[0].div.find_all('span', attrs={'class': 'time'})[0].string
    # 多少人喜欢
    value['people'] = int(
      re.search(r'\d+', title[0].find_all('div', attrs={'class': 'num'})[0].span.string).group())
   # print r
   all_value.append(value)
   value = {}
 return all_value

3、使用SGMLParser，主要是通过start、end tag的方式进行，解析工程比较明朗，但是有点麻烦，而该案例的场景不太适合该方法。

class CommentParaser(SGMLParser):
 def __init__(self):
  SGMLParser.__init__(self)
  self.__start_div_yingping = False
  self.__start_div_item = False
  self.__start_div_gclear = False
  self.__start_div_ratingwrap = False
  self.__start_div_num = False
  # a
  self.__start_a = False
  # span 3中状态
  self.__span_state = 0
  # 数据
  self.__value = {}
  self.data = []
  
 def start_div(self, attrs):
  for k, v in attrs:
   if k == 'class' and v == 'yingping-list-wrap':
    self.__start_div_yingping = True
   elif k == 'class' and v == 'item':
    self.__start_div_item = True
   elif k == 'class' and v == 'g-clear title-wrap':
    self.__start_div_gclear = True
   elif k == 'class' and v == 'rating-wrap g-clear':
    self.__start_div_ratingwrap = True
   elif k == 'class' and v == 'num':
    self.__start_div_num = True
  
 def end_div(self):
  if self.__start_div_yingping:
   if self.__start_div_item:
    if self.__start_div_gclear:
     if self.__start_div_num or self.__start_div_ratingwrap:
      if self.__start_div_num:
       self.__start_div_num = False
      if self.__start_div_ratingwrap:
       self.__start_div_ratingwrap = False
     else:
      self.__start_div_gclear = False
    else:
     self.data.append(self.__value)
     self.__value = {}
     self.__start_div_item = False
   else:
    self.__start_div_yingping = False
  
 def start_a(self, attrs):
  if self.__start_div_yingping and self.__start_div_item and self.__start_div_gclear:
   self.__start_a = True
   for k, v in attrs:
    if k == 'href':
     self.__value['href'] = v
  
 def end_a(self):
  if self.__start_div_yingping and self.__start_div_item and self.__start_div_gclear and self.__start_a:
   self.__start_a = False
  
 def start_span(self, attrs):
  if self.__start_div_yingping and self.__start_div_item and self.__start_div_gclear:
   if self.__start_div_ratingwrap:
    if self.__span_state != 1:
     for k, v in attrs:
      if k == 'class' and v == 'rating':
       self.__span_state = 1
      elif k == 'class' and v == 'time':
       self.__span_state = 2
    else:
     for k, v in attrs:
      if k == 'style':
       score_text = re.search(r'\d+', v).group()
     self.__value['score'] = int(score_text) / 20
     self.__span_state = 3
   elif self.__start_div_num:
    self.__span_state = 4
  
 def end_span(self):
  self.__span_state = 0
  
 def handle_data(self, data):
  if self.__start_a:
   self.__value['title'] = data
  elif self.__span_state == 2:
   self.__value['time'] = data
  elif self.__span_state == 4:
   score_text = re.search(r'\d+', data).group()
   self.__value['people'] = int(score_text)
  pass
def sgl_parser(html):
 parser = CommentParaser()
 parser.feed(html)
 return parser.data

4、HTMLParaer，与3原理相识，就是调用的方法不太一样，基本上可以公用。

class CommentHTMLParser(HTMLParser.HTMLParser):
 def __init__(self):
  HTMLParser.HTMLParser.__init__(self)
  self.__start_div_yingping = False
  self.__start_div_item = False
  self.__start_div_gclear = False
  self.__start_div_ratingwrap = False
  self.__start_div_num = False
  # a
  self.__start_a = False
  # span 3中状态
  self.__span_state = 0
  # 数据
  self.__value = {}
  self.data = []
  
 def handle_starttag(self, tag, attrs):
  if tag == 'div':
   for k, v in attrs:
    if k == 'class' and v == 'yingping-list-wrap':
     self.__start_div_yingping = True
    elif k == 'class' and v == 'item':
     self.__start_div_item = True
    elif k == 'class' and v == 'g-clear title-wrap':
     self.__start_div_gclear = True
    elif k == 'class' and v == 'rating-wrap g-clear':
     self.__start_div_ratingwrap = True
    elif k == 'class' and v == 'num':
     self.__start_div_num = True
  elif tag == 'a':
   if self.__start_div_yingping and self.__start_div_item and self.__start_div_gclear:
    self.__start_a = True
    for k, v in attrs:
     if k == 'href':
      self.__value['href'] = v
  elif tag == 'span':
   if self.__start_div_yingping and self.__start_div_item and self.__start_div_gclear:
    if self.__start_div_ratingwrap:
     if self.__span_state != 1:
      for k, v in attrs:
       if k == 'class' and v == 'rating':
        self.__span_state = 1
       elif k == 'class' and v == 'time':
        self.__span_state = 2
     else:
      for k, v in attrs:
       if k == 'style':
        score_text = re.search(r'\d+', v).group()
      self.__value['score'] = int(score_text) / 20
      self.__span_state = 3
    elif self.__start_div_num:
     self.__span_state = 4
  
 def handle_endtag(self, tag):
  if tag == 'div':
   if self.__start_div_yingping:
    if self.__start_div_item:
     if self.__start_div_gclear:
      if self.__start_div_num or self.__start_div_ratingwrap:
       if self.__start_div_num:
        self.__start_div_num = False
       if self.__start_div_ratingwrap:
        self.__start_div_ratingwrap = False
      else:
       self.__start_div_gclear = False
     else:
      self.data.append(self.__value)
      self.__value = {}
      self.__start_div_item = False
    else:
     self.__start_div_yingping = False
  elif tag == 'a':
   if self.__start_div_yingping and self.__start_div_item and self.__start_div_gclear and self.__start_a:
    self.__start_a = False
  elif tag == 'span':
   self.__span_state = 0
  
 def handle_data(self, data):
  if self.__start_a:
   self.__value['title'] = data
  elif self.__span_state == 2:
   self.__value['time'] = data
  elif self.__span_state == 4:
   score_text = re.search(r'\d+', data).group()
   self.__value['people'] = int(score_text)
  pass
def html_parser(html):
 parser = CommentHTMLParser()
 parser.feed(html)
 return parser.data