使用背景:
我们通常在爬去某个网站的时候都是爬去每个标签下的某些内容,往往一个网站的主页后面会包含很多物品或者信息的详细的内容,我们只提取某个大标签下的某些内容的话,会显的效率较低,大部分网站的都是按照固定套路(也就是固定模板,把各种信息展示给用户),LinkExtrator就非常适合整站抓取,为什么呢?因为你通过xpath、css等一些列参数设置,拿到整个网站的你想要的链接,而不是固定的某个标签下的一些链接内容,非常适合整站爬取。
1 import scrapy
2 from scrapy.linkextractor import LinkExtractor
3
4 class WeidsSpider(scrapy.Spider):
5 name = "weids"
6 allowed_domains = ["wds.modian.com"]
7 start_urls = ['http://www.gaosiedu.com/gsschool/']
8
9 def parse(self, response):
10 link = LinkExtractor(restrict_xpaths='//ul[@class="cont_xiaoqu"]/li')
11 links = link.extract_links(response)
12 print(links)
links是一个list
我们来迭代一下这个list
1 for link in links:
2 print(link)
links里面包含了我们要提取的url,那我们怎么才能拿到这个url呢?
直接在for循环里面link.url就能拿到我们要的url和text信息
1 for link in links:
2 print(link.url,link.text)
别着急,LinkExtrator里面不止一个xpath提取方法,还有很多参数。
>allow:接收一个正则表达式或一个正则表达式列表,提取绝对url于正则表达式匹配的链接,如果该参数为空,默认全部提取。
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4
5 class WeidsSpider(scrapy.Spider):
6 name = "weids"
7 allowed_domains = ["wds.modian.com"]
8 start_urls = ['http://www.gaosiedu.com/gsschool/']
9
10 def parse(self, response):
11 pattern = '/gsschool/.+\.shtml'
12 link = LinkExtractor(allow=pattern)
13 links = link.extract_links(response)
14 print(type(links))
15 for link in links:
16 print(link)
>deny:接收一个正则表达式或一个正则表达式列表,与allow相反,排除绝对url于正则表达式匹配的链接,换句话说,就是凡是跟正则表达式能匹配上的全部不提取。
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4
5 class WeidsSpider(scrapy.Spider):
6 name = "weids"
7 allowed_domains = ["wds.modian.com"]
8 start_urls = ['http://www.gaosiedu.com/gsschool/']
9
10 def parse(self, response):
11 pattern = '/gsschool/.+\.shtml'
12 link = LinkExtractor(deny=pattern)
13 links = link.extract_links(response)
14 print(type(links))
15 for link in links:
16 print(link)
>allow_domains:接收一个域名或一个域名列表,提取到指定域的链接。
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4
5 class WeidsSpider(scrapy.Spider):
6 name = "weids"
7 allowed_domains = ["wds.modian.com"]
8 start_urls = ['http://www.gaosiedu.com/gsschool/']
9
10 def parse(self, response):
11 domain = ['gaosivip.com','gaosiedu.com']
12 link = LinkExtractor(allow_domains=domain)
13 links = link.extract_links(response)
14 print(type(links))
15 for link in links:
16 print(link)
>deny_domains:和allow_doains相反,拒绝一个域名或一个域名列表,提取除被deny掉的所有匹配url。
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4
5 class WeidsSpider(scrapy.Spider):
6 name = "weids"
7 allowed_domains = ["wds.modian.com"]
8 start_urls = ['http://www.gaosiedu.com/gsschool/']
9
10 def parse(self, response):
11 domain = ['gaosivip.com','gaosiedu.com']
12 link = LinkExtractor(deny_domains=domain)
13 links = link.extract_links(response)
14 print(type(links))
15 for link in links:
16 print(link)
>restrict_xpaths:我们在最开始做那个那个例子,接收一个xpath表达式或一个xpath表达式列表,提取xpath表达式选中区域下的链接。
>restrict_css:这参数和restrict_xpaths参数经常能用到,所以同学必须掌握,个人更喜欢xpath。
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4
5 class WeidsSpider(scrapy.Spider):
6 name = "weids"
7 allowed_domains = ["wds.modian.com"]
8 start_urls = ['http://www.gaosiedu.com/gsschool/']
9
10 def parse(self, response):
11 link = LinkExtractor(restrict_css='ul.cont_xiaoqu > li')
12 links = link.extract_links(response)
13 print(type(links))
14 for link in links:
15 print(link)
>tags:接收一个标签(字符串)或一个标签列表,提取指定标签内的链接,默认为tags=(‘a’,‘area’)
>attrs:接收一个属性(字符串)或者一个属性列表,提取指定的属性内的链接,默认为attrs=(‘href’,),示例,按照这个中提取方法的话,这个页面上的某些标签的属性都会被提取出来,如下例所示,这个页面的a标签的href属性值都被提取到了。
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4
5 class WeidsSpider(scrapy.Spider):
6 name = "weids"
7 allowed_domains = ["wds.modian.com"]
8 start_urls = ['http://www.gaosiedu.com/gsschool/']
9
10 def parse(self, response):
11 link = LinkExtractor(tags='a',attrs='href')
12 links = link.extract_links(response)
13 print(type(links))
14 for link in links:
15 print(link)
前面我们讲了这么多LinkExtractor的基本用法,上面的只是为了快速试验,真正的基本用法是结合Crawler和Rule,代码如下
1 # -*- coding: utf-8 -*-
2 import scrapy
3 from scrapy.linkextractor import LinkExtractor
4 from scrapy.spiders.crawl import CrawlSpider,Rule
5
6
7 class GaosieduSpider(CrawlSpider):
8 name = "gaosiedu"
9 allowed_domains = ["www.gaosiedu.com"]
10 start_urls = ['http://www.gaosiedu.com/']
11 restrict_xpath = '//ul[@class="schoolList clearfix"]'
12 allow = '/gsschool/.+\.shtml'
13 rules = {
14 Rule(LinkExtractor(restrict_xpaths=restrict_xpath), callback="parse_item", follow=True)
15 }
16
17 def parse_item(self,response):
18 schooll_name = response.xpath('//div[@class="area_nav"]//h3/text()').extract_first()
19 print(schooll_name)
简单的说一下,上面我们本应该继承scrapy.Spider类,这里需要继承CrawlSpider类(因为CrawlSpider类也是继承了scrapy.Spider类),rules是基本写法,可不是随便哪个单词都可以的啊,而且注意rules必须是一个list或者dict,如果是tuple的话就会报错。里面的话Rule里面包含了几个参数,LinkExtractor就不在这里熬述了,看上面就行,至于其他的几个参数,可以看我们另外一篇文章:https://www.cdsy.xyz/computer/programme/Python/230521/cd43833.html