爬虫-scrapy框架
Scrapy是一个为了爬取网站数据,提取结构性数据而编写的应用框架。 其可以应用在数据挖掘,信息处理或存储历史数据等一系列的程序中。
其最初是为了页面抓取 (更确切来说, 网络抓取 )所设计的, 也可以应用在获取API所返回的数据(例如 Amazon Associates Web Services ) 或者通用的网络爬虫。Scrapy用途广泛,可以用于数据挖掘、监测和自动化测试。
Scrapy 使用了 Twisted异步网络库来处理网络通讯。整体架构大致如下
Scrapy主要包括了以下组件:
- 引擎(Scrapy)
用来处理整个系统的数据流处理, 触发事务(框架核心) - 调度器(Scheduler)
用来接受引擎发过来的请求, 压入队列中, 并在引擎再次请求的时候返回. 可以想像成一个URL(抓取网页的网址或者说是链接)的优先队列, 由它来决定下一个要抓取的网址是什么, 同时去除重复的网址 - 下载器(Downloader)
用于下载网页内容, 并将网页内容返回给蜘蛛(Scrapy下载器是建立在twisted这个高效的异步模型上的) - 爬虫(Spiders)
爬虫是主要干活的, 用于从特定的网页中提取自己需要的信息, 即所谓的实体(Item)。用户也可以从中提取出链接,让Scrapy继续抓取下一个页面 - 项目管道(Pipeline)
负责处理爬虫从网页中抽取的实体,主要的功能是持久化实体、验证实体的有效性、清除不需要的信息。当页面被爬虫解析后,将被发送到项目管道,并经过几个特定的次序处理数据。 - 下载器中间件(Downloader Middlewares)
位于Scrapy引擎和下载器之间的框架,主要是处理Scrapy引擎与下载器之间的请求及响应。 - 爬虫中间件(Spider Middlewares)
介于Scrapy引擎和爬虫之间的框架,主要工作是处理蜘蛛的响应输入和请求输出。 - 调度中间件(Scheduler Middewares)
介于Scrapy引擎和调度之间的中间件,从Scrapy引擎发送到调度的请求和响应。
Scrapy运行流程大概如下:
- 引擎从调度器中取出一个链接(URL)用于接下来的抓取
- 引擎把URL封装成一个请求(Request)传给下载器
- 下载器把资源下载下来,并封装成应答包(Response)
- 爬虫解析Response
- 解析出实体(Item),则交给实体管道进行进一步的处理
- 解析出的是链接(URL),则把URL交给调度器等待抓取
一、安装
pip install Scrapy
注:windows平台需要依赖pywin32,请根据自己系统32/64位选择下载安装,https://sourceforge.net/projects/pywin32/
二、基本使用
1、创建项目
运行命令:
scrapy startproject your_project_name
自动创建目录:
project_name/ scrapy.cfg project_name/ __init__.py items.py pipelines.py settings.py spiders/ __init__.py
文件说明:
- scrapy.cfg 项目的配置信息,主要为Scrapy命令行工具提供一个基础的配置信息。(真正爬虫相关的配置信息在settings.py文件中)
- items.py 设置数据存储模板,用于结构化数据,如:Django的Model
- pipelines 数据处理行为,如:一般结构化的数据持久化
- settings.py 配置文件,如:递归的层数、并发数,延迟下载等
- spiders 爬虫目录,如:创建文件,编写爬虫规则
注意:一般创建爬虫文件时,以网站域名命名
2、编写爬虫
在spiders目录中新建 xiaohuar_spider.py 文件
#!/usr/bin/env python # -*- coding:utf-8 -*- import scrapy class XiaoHuarSpider(scrapy.spiders.Spider): name = "xiaohuar" allowed_domains = ["xiaohuar.com"] start_urls = [ "http://www.xiaohuar.com/hua/", ] def parse(self, response): # print(response, type(response)) # from scrapy.http.response.html import HtmlResponse # print(response.body_as_unicode()) current_url = response.url body = response.body unicode_body = response.body_as_unicode()
3、运行
进入project_name目录,运行命令
import sys,os sys.stdout=io.TextIOWrapper(sys.stdout.buffer,encoding='gb18030')
scrapy crawl spider_name --nolog
=============选择器
#!/usr/bin/env python # -*- coding:utf-8 -*- from scrapy.selector import Selector, HtmlXPathSelector from scrapy.http import HtmlResponse html = """<!DOCTYPE html> <html> <head lang="en"> <meta charset="UTF-8"> <title></title> </head> <body> <ul> <li class="item-"><a id='i1' href="link.html">first item</a></li> <li class="item-0"><a id='i2' href="llink.html">first item</a></li> <li class="item-1"><a href="llink2.html">second item<span>vv</span></a></li> </ul> <div><a href="llink2.html">second item</a></div> </body> </html> """ response = HtmlResponse(url='http://example.com', body=html,encoding='utf-8') # hxs = HtmlXPathSelector(response) # print(hxs) # hxs = Selector(response=response).xpath('//a') # print(hxs) # hxs = Selector(response=response).xpath('//a[2]') # print(hxs) # hxs = Selector(response=response).xpath('//a[@id]') # print(hxs) # hxs = Selector(response=response).xpath('//a[@id="i1"]') # print(hxs) # hxs = Selector(response=response).xpath('//a[@href="link.html"][@id="i1"]') # print(hxs) # hxs = Selector(response=response).xpath('//a[contains(@href, "link")]') # print(hxs) # hxs = Selector(response=response).xpath('//a[starts-with(@href, "link")]') # print(hxs) # hxs = Selector(response=response).xpath('//a[re:test(@id, "i\d+")]') # print(hxs) # hxs = Selector(response=response).xpath('//a[re:test(@id, "i\d+")]/text()').extract() # print(hxs) # hxs = Selector(response=response).xpath('//a[re:test(@id, "i\d+")]/@href').extract() # print(hxs) # hxs = Selector(response=response).xpath('/html/body/ul/li/a/@href').extract() # print(hxs) # hxs = Selector(response=response).xpath('//body/ul/li/a/@href').extract_first() # print(hxs) # ul_list = Selector(response=response).xpath('//body/ul/li') # for item in ul_list: # v = item.xpath('./a/span') # # 或 # # v = item.xpath('a/span') # # 或 # # v = item.xpath('*/a/span') # print(v)
4、递归的访问
以上的爬虫仅仅是爬去初始页,而我们爬虫是需要源源不断的执行下去,直到所有的网页被执行完毕
#!/usr/bin/env python # -*- coding:utf-8 -*- import scrapy from scrapy.http import Request from scrapy.selector import HtmlXPathSelector import re import urllib import os class XiaoHuarSpider(scrapy.spiders.Spider): name = "xiaohuar" allowed_domains = ["xiaohuar.com"] start_urls = [ "http://www.xiaohuar.com/list-1-1.html", ] def parse(self, response): # 分析页面 # 找到页面中符合规则的内容(校花图片),保存 # 找到所有的a标签,再访问其他a标签,一层一层的搞下去 hxs = HtmlXPathSelector(response) # 如果url是 http://www.xiaohuar.com/list-1-\d+.html if re.match('http://www.xiaohuar.com/list-1-\d+.html', response.url): items = hxs.select('//div[@class="item_list infinite_scroll"]/div') for i in range(len(items)): src = hxs.select('//div[@class="item_list infinite_scroll"]/div[%d]//div[@class="img"]/a/img/@src' % i).extract() name = hxs.select('//div[@class="item_list infinite_scroll"]/div[%d]//div[@class="img"]/span/text()' % i).extract() school = hxs.select('//div[@class="item_list infinite_scroll"]/div[%d]//div[@class="img"]/div[@class="btns"]/a/text()' % i).extract() if src: ab_src = "http://www.xiaohuar.com" + src[0] file_name = "%s_%s.jpg" % (school[0].encode('utf-8'), name[0].encode('utf-8')) file_path = os.path.join("/Users/wupeiqi/PycharmProjects/beauty/pic", file_name) urllib.urlretrieve(ab_src, file_path) # 获取所有的url,继续访问,并在其中寻找相同的url all_urls = hxs.select('//a/@href').extract() for url in all_urls: if url.startswith('http://www.xiaohuar.com/list-1-'): yield Request(url, callback=self.parse)
以上代码将符合规则的页面中的图片保存在指定目录,并且在HTML源码中找到所有的其他 a 标签的href属性,从而“递归”的执行下去,直到所有的页面都被访问过为止。以上代码之所以可以进行“递归”的访问相关URL,关键在于parse方法使用了 yield Request对象。
注:可以修改settings.py 中的配置文件,以此来指定“递归”的层数,如: DEPTH_LIMIT = 1
from scrapy.selector import Selector from scrapy.http import HtmlResponse html = """<!DOCTYPE html> <html> <head lang="en"> <meta charset="UTF-8"> <title></title> </head> <body> <li class="item-"><a href="link.html">first item</a></li> <li class="item-0"><a href="link1.html">first item</a></li> <li class="item-1"><a href="link2.html">second item</a></li> </body> </html> """ response = HtmlResponse(url='http://example.com', body=html,encoding='utf-8') ret = Selector(response=response).xpath('//li[re:test(@class, "item-\d*")]//@href').extract() print(ret) 正则选择器
#!/usr/bin/env python # -*- coding:utf-8 -*- import scrapy import hashlib from tutorial.items import JinLuoSiItem from scrapy.http import Request from scrapy.selector import HtmlXPathSelector class JinLuoSiSpider(scrapy.spiders.Spider): count = 0 url_set = set() name = "jluosi" domain = 'http://www.jluosi.com' allowed_domains = ["jluosi.com"] start_urls = [ "http://www.jluosi.com:80/ec/goodsDetail.action?jls=QjRDNEIzMzAzOEZFNEE3NQ==", ] def parse(self, response): md5_obj = hashlib.md5() md5_obj.update(response.url) md5_url = md5_obj.hexdigest() if md5_url in JinLuoSiSpider.url_set: pass else: JinLuoSiSpider.url_set.add(md5_url) hxs = HtmlXPathSelector(response) if response.url.startswith('http://www.jluosi.com:80/ec/goodsDetail.action'): item = JinLuoSiItem() item['company'] = hxs.select('//div[@class="ShopAddress"]/ul/li[1]/text()').extract() item['link'] = hxs.select('//div[@class="ShopAddress"]/ul/li[2]/text()').extract() item['qq'] = hxs.select('//div[@class="ShopAddress"]//a/@href').re('.*uin=(?P<qq>\d*)&') item['address'] = hxs.select('//div[@class="ShopAddress"]/ul/li[4]/text()').extract() item['title'] = hxs.select('//h1[@class="goodsDetail_goodsName"]/text()').extract() item['unit'] = hxs.select('//table[@class="R_WebDetail_content_tab"]//tr[1]//td[3]/text()').extract() product_list = [] product_tr = hxs.select('//table[@class="R_WebDetail_content_tab"]//tr') for i in range(2,len(product_tr)): temp = { 'standard':hxs.select('//table[@class="R_WebDetail_content_tab"]//tr[%d]//td[2]/text()' %i).extract()[0].strip(), 'price':hxs.select('//table[@class="R_WebDetail_content_tab"]//tr[%d]//td[3]/text()' %i).extract()[0].strip(), } product_list.append(temp) item['product_list'] = product_list yield item current_page_urls = hxs.select('//a/@href').extract() for i in range(len(current_page_urls)): url = current_page_urls[i] if url.startswith('http://www.jluosi.com'): url_ab = url yield Request(url_ab, callback=self.parse) 选择器规则Demo
def parse(self, response): from scrapy.http.cookies import CookieJar cookieJar = CookieJar() cookieJar.extract_cookies(response, response.request) print(cookieJar._cookies)
更多选择器规则:http://scrapy-chs.readthedocs.io/zh_CN/latest/topics/selectors.html
5、格式化处理
上述实例只是简单的处理,所以在parse方法中直接处理。如果对于想要获取更多的数据处理,则可以利用Scrapy的items将数据格式化,然后统一交由pipelines来处理。
import scrapy from scrapy.selector import HtmlXPathSelector from scrapy.http.request import Request from scrapy.http.cookies import CookieJar from scrapy import FormRequest class XiaoHuarSpider(scrapy.Spider): # 爬虫应用的名称,通过此名称启动爬虫命令 name = "xiaohuar" # 允许的域名 allowed_domains = ["xiaohuar.com"] start_urls = [ "http://www.xiaohuar.com/list-1-1.html", ] # custom_settings = { # 'ITEM_PIPELINES':{ # 'spider1.pipelines.JsonPipeline': 100 # } # } has_request_set = {} def parse(self, response): # 分析页面 # 找到页面中符合规则的内容(校花图片),保存 # 找到所有的a标签,再访问其他a标签,一层一层的搞下去 hxs = HtmlXPathSelector(response) items = hxs.select('//div[@class="item_list infinite_scroll"]/div') for item in items: src = item.select('.//div[@class="img"]/a/img/@src').extract_first() name = item.select('.//div[@class="img"]/span/text()').extract_first() school = item.select('.//div[@class="img"]/div[@class="btns"]/a/text()').extract_first() url = "http://www.xiaohuar.com%s" % src from ..items import XiaoHuarItem obj = XiaoHuarItem(name=name, school=school, url=url) yield obj urls = hxs.select('//a[re:test(@href, "http://www.xiaohuar.com/list-1-\d+.html")]/@href') for url in urls: key = self.md5(url) if key in self.has_request_set: pass else: self.has_request_set[key] = url req = Request(url=url,method='GET',callback=self.parse) yield req @staticmethod def md5(val): import hashlib ha = hashlib.md5() ha.update(bytes(val, encoding='utf-8')) key = ha.hexdigest() return key spiders/xiahuar.py
import scrapy class XiaoHuarItem(scrapy.Item): name = scrapy.Field() school = scrapy.Field() url = scrapy.Field() items
import json import os import requests class JsonPipeline(object): def __init__(self): self.file = open('xiaohua.txt', 'w') def process_item(self, item, spider): v = json.dumps(dict(item), ensure_ascii=False) self.file.write(v) self.file.write('\n') self.file.flush() return item class FilePipeline(object): def __init__(self): if not os.path.exists('imgs'): os.makedirs('imgs') def process_item(self, item, spider): response = requests.get(item['url'], stream=True) file_name = '%s_%s.jpg' % (item['name'], item['school']) with open(os.path.join('imgs', file_name), mode='wb') as f: f.write(response.content) return item pipelines
ITEM_PIPELINES = { 'spider1.pipelines.JsonPipeline': 100, 'spider1.pipelines.FilePipeline': 300, } # 每行后面的整型值,确定了他们运行的顺序,item按数字从低到高的顺序,通过pipeline,通常将这些数字定义在0-1000范围内。
对于pipeline可以做更多,如下:
from scrapy.exceptions import DropItem class CustomPipeline(object): def __init__(self,v): self.value = v def process_item(self, item, spider): # 操作并进行持久化 # return表示会被后续的pipeline继续处理 return item # 表示将item丢弃,不会被后续pipeline处理 # raise DropItem() @classmethod def from_crawler(cls, crawler): """ 初始化时候,用于创建pipeline对象 :param crawler: :return: """ val = crawler.settings.getint('MMMM') return cls(val) def open_spider(self,spider): """ 爬虫开始执行时,调用 :param spider: :return: """ print('000000') def close_spider(self,spider): """ 爬虫关闭时,被调用 :param spider: :return: """ print('111111') 自定义pipeline
# -*- coding: utf-8 -*- # Define here the models for your scraped items # # See documentation in: # http://doc.scrapy.org/en/latest/topics/items.html import scrapy class JieYiCaiItem(scrapy.Item): company = scrapy.Field() title = scrapy.Field() qq = scrapy.Field() info = scrapy.Field() more = scrapy.Field()
上述定义模板,以后对于从请求的源码中获取的数据同意按照此结构来获取,所以在spider中需要有一下操作:
#!/usr/bin/env python # -*- coding:utf-8 -*- import scrapy import hashlib from beauty.items import JieYiCaiItem from scrapy.http import Request from scrapy.selector import HtmlXPathSelector from scrapy.spiders import CrawlSpider, Rule from scrapy.linkextractors import LinkExtractor class JieYiCaiSpider(scrapy.spiders.Spider): count = 0 url_set = set() name = "jieyicai" domain = 'http://www.jieyicai.com' allowed_domains = ["jieyicai.com"] start_urls = [ "http://www.jieyicai.com", ] rules = [ #下面是符合规则的网址,但是不抓取内容,只是提取该页的链接(这里网址是虚构的,实际使用时请替换) #Rule(SgmlLinkExtractor(allow=(r'http://test_url/test?page_index=\d+'))), #下面是符合规则的网址,提取内容,(这里网址是虚构的,实际使用时请替换) #Rule(LinkExtractor(allow=(r'http://www.jieyicai.com/Product/Detail.aspx?pid=\d+')), callback="parse"), ] def parse(self, response): md5_obj = hashlib.md5() md5_obj.update(response.url) md5_url = md5_obj.hexdigest() if md5_url in JieYiCaiSpider.url_set: pass else: JieYiCaiSpider.url_set.add(md5_url) hxs = HtmlXPathSelector(response) if response.url.startswith('http://www.jieyicai.com/Product/Detail.aspx'): item = JieYiCaiItem() item['company'] = hxs.select('//span[@class="username g-fs-14"]/text()').extract() item['qq'] = hxs.select('//span[@class="g-left bor1qq"]/a/@href').re('.*uin=(?P<qq>\d*)&') item['info'] = hxs.select('//div[@class="padd20 bor1 comard"]/text()').extract() item['more'] = hxs.select('//li[@class="style4"]/a/@href').extract() item['title'] = hxs.select('//div[@class="g-left prodetail-text"]/h2/text()').extract() yield item current_page_urls = hxs.select('//a/@href').extract() for i in range(len(current_page_urls)): url = current_page_urls[i] if url.startswith('/'): url_ab = JieYiCaiSpider.domain + url yield Request(url_ab, callback=self.parse) spider
此处代码的关键在于:
- 将获取的数据封装在了Item对象中
- yield Item对象 (一旦parse中执行yield Item对象,则自动将该对象交个pipelines的类来处理)
# -*- coding: utf-8 -*- # Define your item pipelines here # # Don't forget to add your pipeline to the ITEM_PIPELINES setting # See: http://doc.scrapy.org/en/latest/topics/item-pipeline.html import json from twisted.enterprise import adbapi import MySQLdb.cursors import re mobile_re = re.compile(r'(13[0-9]|15[012356789]|17[678]|18[0-9]|14[57])[0-9]{8}') phone_re = re.compile(r'(\d+-\d+|\d+)') class JsonPipeline(object): def __init__(self): self.file = open('/Users/wupeiqi/PycharmProjects/beauty/beauty/jieyicai.json', 'wb') def process_item(self, item, spider): line = "%s %s\n" % (item['company'][0].encode('utf-8'), item['title'][0].encode('utf-8')) self.file.write(line) return item class DBPipeline(object): def __init__(self): self.db_pool = adbapi.ConnectionPool('MySQLdb', db='DbCenter', user='root', passwd='123', cursorclass=MySQLdb.cursors.DictCursor, use_unicode=True) def process_item(self, item, spider): query = self.db_pool.runInteraction(self._conditional_insert, item) query.addErrback(self.handle_error) return item def _conditional_insert(self, tx, item): tx.execute("select nid from company where company = %s", (item['company'][0], )) result = tx.fetchone() if result: pass else: phone_obj = phone_re.search(item['info'][0].strip()) phone = phone_obj.group() if phone_obj else ' ' mobile_obj = mobile_re.search(item['info'][1].strip()) mobile = mobile_obj.group() if mobile_obj else ' ' values = ( item['company'][0], item['qq'][0], phone, mobile, item['info'][2].strip(), item['more'][0]) tx.execute("insert into company(company,qq,phone,mobile,address,more) values(%s,%s,%s,%s,%s,%s)", values) def handle_error(self, e): print 'error',e pipelines
上述中的pipelines中有多个类,到底Scapy会自动执行那个?哈哈哈哈,当然需要先配置了,不然Scapy就蒙逼了。。。
在settings.py中做如下配置:
ITEM_PIPELINES = { 'beauty.pipelines.DBPipeline': 300, 'beauty.pipelines.JsonPipeline': 100, } # 每行后面的整型值,确定了他们运行的顺序,item按数字从低到高的顺序,通过pipeline,通常将这些数字定义在0-1000范围内。