單字符匹配
. 除換行符之外的任意字符
\d 表示數(shù)字
\D 匹配非數(shù)字
\w 匹配單詞字符[a-z,A-Z,0-9]
\W 匹配非單詞字符
\s 匹配空白字符,空格,\n \t…
\S 匹配非空白字符
^ 匹配以…開頭
$ 匹配以…結(jié)尾
[0-9] => \d 匹配0-9
多字符匹配(貪婪匹配)
- 匹配*前面的字符任意次數(shù)
- 匹配+前面的字符至少一次
?匹配?前面的字符0-1次
{n,m}匹配{n,m}前面的字符n-m次
多字符匹配(非貪婪匹配)
*?
+?
??
其他
()分組
|邏輯或
\轉(zhuǎn)義字符
re模塊下的方法
re.compile():構(gòu)建正則表達(dá)式對象
re.match():從起始位開始匹配,單次匹配,如果匹配到結(jié)果立即返回,反之,返回None
re.search():在整個字符串中進(jìn)行匹配,單次匹配,如果匹配到結(jié)果立即返回,反之,返回None
re.findall():匹配出整個字符串中,所有符合正則規(guī)則的結(jié)果,返回一個列表
re.finditer():匹配出整個字符串中,所有符合正則規(guī)則的結(jié)果,返回的是一個可迭代對象
re.sub():根據(jù)正則表達(dá)式進(jìn)行字符串替換
re.split():根據(jù)正則表達(dá)式進(jìn)行分割
正則的用法
def get_rank_data(url='http://top.hengyan.com/dianji/default.aspx?p=1'):
#構(gòu)建請求頭
headers = {
'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/537.36',
}
# url, \目標(biāo)url
# data=None, \默認(rèn)為None表示是get請求,如果不為None說明是get請求
# timeout 設(shè)置請求的超時時間
# cafile=None, capath=None, cadefault=False,:證書相關(guān)參數(shù)
# context=None :忽略證書認(rèn)證
#urlopen不能添加請求頭
# response = request.urlopen(url=url,timeout=10)
#添加請求頭
req = request.Request(url=url,headers=headers)
response = request.urlopen(req,timeout=10)
#響應(yīng)狀態(tài)碼
code = response.status
#當(dāng)前請求的url地址
url = response.url
print(code,url)
b_content = response.read()
# bytes -> str: decode
# str -> bytes: encode
# print(b_content)
html = b_content.decode('utf-8')
# print(html)
# #文件操作
# """
# w: w+: wb: wb+ a: a+: ab: ab+: r: rb:
# """
# with open('hengyan.html','w') as file:
# file.write(html)
#證據(jù)正則表達(dá)式解析數(shù)據(jù)
# re.S 修飾:表示.可以匹配換行符
pattern = re.compile('<div\sclass="list">(.*?)</div>',re.S)
ul_str = re.findall(pattern,html)[0]
pattern1 = re.compile('<ul.*?>(.*?)</ul>',re.S)
li_strs = re.findall(pattern1,ul_str)[1:]
for li_str in li_strs:
# print(li_str)
pattern = re.compile(
'<li\sclass="num">(.*?)</li>'+
'.*?<a.*?>(.*?)</a>'+
'.*?<li.*?>(.*?)</li>'+
'.*?<li.*?>(.*?)</li>'+
'.*?<li.*?>(.*?)</li>'+
'.*?<li.*?>(.*?)</li>',
re.S
)
data = re.findall(pattern=pattern,string=li_str)[0]
print(data)
#提取下一頁:
if '下一頁' in html:
#說明還存在下一頁
pattern = re.compile('<span\sclass="pageBarCurrentStyle">(.*?)</span>',re.S)
current_page = int(re.findall(pattern,html)[0])
next_page = current_page+1
#構(gòu)造下一頁的URL地址
next_page_url = re.sub('\d+',str(next_page),url)
print(next_page_url)
get_rank_data(next_page_url)
else:
print('數(shù)據(jù)提取完畢')
if __name__ == '__main__':
get_rank_data()
xpath
安裝:pip install lxml
引用:from lxml import etree
創(chuàng)建etree對象進(jìn)行指定數(shù)據(jù)解析
1.本地
etree = etree.parse(‘本地路徑’)
etree.xpath(‘xpath表達(dá)式’)
2.網(wǎng)絡(luò)
etree = etree.HTML(‘網(wǎng)絡(luò)請求到頁面的數(shù)據(jù)’)
etree.xpath(‘xpath表達(dá)式’)
常用的xpath表達(dá)式:
1.屬性定位:
找到class屬性值為song的div標(biāo)簽
//div[@class=‘song’]
2.層級索引定位
找到class屬性值為tang的div的直系子標(biāo)簽ul下的第二個子標(biāo)簽li下的直系子標(biāo)簽a
//div[@class=‘tang’]/ul/li[2]/a
3.邏輯運(yùn)算
找到href屬性值為空且class屬性值為du的a標(biāo)簽
//a[@href=’’ and @class=‘du’]
4.模糊匹配
/表示獲取某個標(biāo)簽下的文本內(nèi)容 //div[@class=‘song’]/p[1]/text()
//表示獲取某個標(biāo)簽下的文本內(nèi)容和所有子標(biāo)簽下的文本內(nèi)容 //div[@class=‘tang’]//text()
5.取屬性
//div[@class=‘tang’]//li[2]/a/@href
class HengYanSpider(object):
def __init__(self):
self.first_url = 'http://all.hengyan.com/1/0_0_0_0_0_0_0_0_0_1.aspx'
self.default_headers = {
'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/537.36'
}
def get_noval_url(self, url=None):
url = self.first_url if not url else url
"""獲取小說詳情的url地址"""
html = self.send_request(url)
if html:
# 解析數(shù)據(jù)(獲取xpath解析器)
etree_html = etree.HTML(html)
noval_urls = etree_html.xpath('//li[@class="bookname"]/a[1]/@href')
for noval_url in noval_urls:
self.get_noval_detail(noval_url)
# 獲取下一頁
if '下一頁' in html:
# 說明還存在下一頁
current_page = int(self.extract_first(etree_html.xpath('//span[@class="pageBarCurrentStyle"]/text()]')))
next_page = current_page + 1
next_page_url = re.sub('\d+.aspx', str(next_page)+'.aspx', url)
self.get_noval_url(next_page_url)
else:
print('數(shù)據(jù)提取完畢')
else:
print('數(shù)據(jù)獲取失敗')
def get_noval_detail(self, noval_url):
"""獲取書籍詳情的頁面內(nèi)容,解析數(shù)據(jù)"""
html = self.send_request(noval_url)
if html:
# 解析數(shù)據(jù)(獲取xpath解析器)
etree_html = etree.HTML(html)
# print('得到了詳情頁面')
noval_dict = {}
# 書號
book_id = self.extract_first(etree_html.xpath('//div[@class="dh"]/p/label/text()'))
noval_dict['book_id'] = re.search('\d+', book_id).group()
# 熱度
noval_dict['hot'] = self.extract_first(etree_html.xpath('//p[@class="wendu"]/b/text()'))
# 火車票
noval_dict['hot_track'] = self.extract_first(
etree_html.xpath('//div[@class="piao"]/p[2]/span[@class="huocolor"]/text()'))
# 冰票
noval_dict['bing_track'] = self.extract_first(
etree_html.xpath('//div[@class="piao"]/p[2]/span[@class="bingcolor"]/text()'))
# 金筆
noval_dict['jingbi'] = self.extract_first(etree_html.xpath('//div[@class="jinbi"]//li[1]/p[2]/text()'))
# 標(biāo)題
noval_dict['title'] = self.extract_first(etree_html.xpath('//h2/text()'))
# 簡介
noval_dict['content'] = self.extract_first(
etree_html.xpath('//p[@class="intro ih1"]/text()|//p[@class="intro ih2"]/text()'))
# 作者
noval_dict['author'] = self.extract_first(etree_html.xpath('//div[@id="ainfo"]/p/span/a[2]/text()'))
print(noval_dict)
self.save_data(noval_dict)
def save_data(self, noval_dict):
"""保存數(shù)據(jù)"""
pass
def extract_first(self, data, default=''):
if len(data) > 0:
return data[0]
return default
def send_request(self, url, header=None, data=None, method="GET"):
"""發(fā)送請求"""
header = self.default_headers if not header else header
if method == 'GET':
# 發(fā)送get請求
response = requests.get(url=url, params=data, headers=header)
else:
# 發(fā)送post請求
response = requests.post(url=url, data=data, headers=header)
if response.status_code == 200:
# 請求成功,返回頁面源碼
return response.text
if __name__ == '__main__':
spider = HengYanSpider()
spider.get_noval_url()
bs4(python獨(dú)有簡單便捷和高效)
環(huán)境安裝:pip install lxml bs4用到lxml庫,如果沒有安裝過lxml庫的時候,需要安裝一下
代碼使用流程:
核心思想:可以將html文檔可以轉(zhuǎn)換成BeautifulSoup對象,調(diào)用該對象中的屬性和方法進(jìn)行
1.導(dǎo)包
from bs4 import BeautifulSoup
2.創(chuàng)建BeautifulSoup對象
a.本地
Beautiful(‘open(‘本地的html文件’)’,‘lxml’)
b.網(wǎng)絡(luò)
Beautiful(‘網(wǎng)絡(luò)請求到的頁面數(shù)據(jù)’,‘lxml’)
屬性和方法:
1.根據(jù)標(biāo)簽名查找
soup.a 只能找到第一個符合要求的標(biāo)簽
2.獲取屬性
soup.a.attrs 獲取a所有的屬性和屬性值,返回一個字典
soup.a.attrs[‘href’] 獲取href屬性
soup.a[‘href’] 也可簡寫為這種形式
3.獲取內(nèi)容
soup.a.string /text()
soup.a.text //text()
soup.a.get_text() //text()
如果標(biāo)簽還是標(biāo)簽,那么string獲取到的結(jié)果為none,而其他兩個,可以獲取文本內(nèi)容
4.find:找到第一個符合要求的標(biāo)簽
soup.find(‘a(chǎn)’) 找到第一個符合要求的
soup.find(‘a(chǎn)’,title=‘xxx’)
soup.find(‘a(chǎn)’,alt=‘xxx’)
soup.find(‘a(chǎn)’,class=‘xxx’)
soup.find(‘a(chǎn)’,id=‘xxx’)
5.find_All:找到所有符合要求的標(biāo)簽
soup.find_All(‘a(chǎn)’)
soup.find_All([‘a(chǎn)’,‘b’]) 找到所有的a和b標(biāo)簽
soup.find_All(‘a(chǎn)’,limit=2) 限制前兩個
6.根據(jù)選擇器選擇指定的內(nèi)容
select:soup.select(’#feng’)
常見的選擇器:標(biāo)簽選擇器(a)、類選擇器(.)、id選擇器(#)、層級選擇器
層級選擇器:
div .dudu #lala .name .xixi 下面好多級 div//img
div > p > a > .lala 只能是下面一級 div/img
select選擇器返回永遠(yuǎn)是列表,需要通過下標(biāo)提取指定對象
class HengYanSpider(object):
def __init__(self):
self.first_url = 'http://all.hengyan.com/1/0_0_0_0_0_0_0_0_0_1.aspx'
self.default_headers = {
'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/537.36'
}
def get_noval_url(self, url=None):
url = self.first_url if not url else url
"""獲取小說詳情的url地址"""
html = self.send_request(url)
if html:
bs_soup = BeautifulSoup(html, 'lxml')
lis = bs_soup.find_all(name='li', attrs={'class': 'bookname'})
for li in lis:
# a_list = li.find_all(name='a')
# if len(a_list) > 0:
# url = a_list[0].attrs['href']
a_list = li.select('a')
if len(a_list) > 0:
url = a_list[0].attrs['href']
self.get_noval_detail(url)
def get_noval_detail(self, noval_url):
"""獲取書籍詳情的頁面內(nèi)容,解析數(shù)據(jù)"""
html = self.send_request(noval_url)
if html:
# 解析數(shù)據(jù)(獲取xpath解析器)
bs_soup = BeautifulSoup(html, 'lxml')
# print('得到了詳情頁面')
noval_dict = {}
# 書號
book_id = bs_soup.select('div.dh p label')[0].get_text()
noval_dict['book_id'] = re.search('\d+', book_id).group()
# 熱度
noval_dict['hot'] = bs_soup.select('p.wendu b')[0].get_text()
# 火車票
noval_dict['hot_track'] = bs_soup.select('div.piao p')[1].select('.huocolor')[0].get_text()
# 冰票
noval_dict['bing_track'] = bs_soup.select('div.piao p')[1].select('.bingcolor')[0].get_text()
print(noval_dict)
# self.save_data(noval_dict)
def save_data(self, noval_dict):
"""保存數(shù)據(jù)"""
pass
def extract_first(self, data, default=''):
if len(data) > 0:
return data[0]
return default
def send_request(self, url, header=None, data=None, method="GET"):
"""發(fā)送請求"""
header = self.default_headers if not header else header
if method == 'GET':
# 發(fā)送get請求
response = requests.get(url=url, params=data, headers=header)
else:
# 發(fā)送post請求
response = requests.post(url=url, data=data, headers=header)
if response.status_code == 200:
# 請求成功,返回頁面源碼
return response.text
if __name__ == '__main__':
spider = HengYanSpider()
spider.get_noval_url()