python scrapy简单使用

最近因为项目需求，需要写个爬虫爬取一些题库。在这之前爬虫我都是用node或者php写的。一直听说python写爬虫有一手，便入手了python的爬虫框架scrapy.

下面简单的介绍一下scrapy的目录结构与使用：

首先我们得安装scrapy框架

pip install scrapy

接着使用scrapy命令创建一个爬虫项目：

scrapy startproject questions

相关文件简介：

• scrapy.cfg: 项目的配置文件
• questions/: 该项目的python模块。之后您将在此加入代码。
• questions/items.py: 项目中的item文件.
• questions/pipelines.py: 项目中的pipelines文件.
• questions/settings.py: 项目的设置文件.
• questions/spiders/: 放置spider代码的目录.
• questions/spiders/xueersi.py: 实现爬虫的主体代码.

xueersi.py  爬虫主体

# -*- coding: utf-8 -*-
import scrapy
import time
import numpy
import re
from questions.items import QuestionsItem


class xueersiSpider(scrapy.Spider):
    name = "xueersi" # 爬虫名字
    allowed_domains = ["tiku.xueersi.com"] # 目标的域名
　　# 爬取的目标地址
    start_urls = [
        "http://tiku.xueersi.com/shiti/list_1_1_0_0_4_0_1",
        "http://tiku.xueersi.com/shiti/list_1_2_0_0_4_0_1",
        "http://tiku.xueersi.com/shiti/list_1_3_0_0_4_0_1",
    ]
    levels = ['偏易','中档','偏难']
    subjects = ['英语','语文','数学']
　　 # 爬虫开始的时候，自动调用该方法，如果该方法不存在会自动调用parse方法
    # def start_requests(self):
    #     yield scrapy.Request('http://tiku.xueersi.com/shiti/list_1_2_0_0_4_0_39',callback=self.getquestion)
　　# start_requests方法不存在时，parse方法自动被调用
    def parse(self, response):
　　　　　# xpath的选择器语法不多介绍，可以直接查看官方文档
        arr = response.xpath("//ul[@class='pagination']/li/a/text()").extract()
        total_page = arr[3]
　　　　　# 获取分页
        for index in range(int(total_page)):
            yield scrapy.Request(response.url.replace('_0_0_4_0_1',"_0_0_4_0_"+str(index)),callback=self.getquestion) # 发出新的请求，获取每个分页所有题目
    # 获取题目
    def getquestion(self,response):
        for res in response.xpath('//div[@class="main-wrap"]/ul[@class="items"]/li'):
            item = QuestionsItem()  # 实例化Item类
            # 获取问题
            questions = res.xpath('./div[@class="content-area"]').re(r'<div class="content-area">?([\s\S]+?)<(table|\/td|div|br)')
            if len(questions):
                # 获取题目
                question = questions[0].strip()
                item['source'] = question
                dr = re.compile(r'<[^>]+>',re.S)
                question = dr.sub('',question)
                content = res.extract()
                item['content'] = question
                # 获取课目
                subject =  re.findall(ur'http:\/\/tiku\.xueersi\.com\/shiti\/list_1_(\d+)',response.url)
                item['subject'] = self.subjects[int(subject[0])-1]
                # 获取难度等级
                levels = res.xpath('//div[@class="info"]').re(ur'难度：([\s\S]+?)<')
                item['level'] = self.levels.index(levels[0])+1
                
                # 获取选项
                options =  re.findall(ur'[A-D][\.．]([\s\S]+?)<(\/td|\/p|br)',content)
                item['options'] = options
                if len(options):
                    url = res.xpath('./div[@class="info"]/a/@href').extract()[0]
                    request = scrapy.Request(url,callback=self.getanswer)
                    request.meta['item'] = item  # 缓存item数据，传递给下一个请求
                    yield request
            #for option in options:
    # 获取答案            
    def getanswer(self,response):
        
        res = response.xpath('//div[@class="part"]').re(ur'<td>([\s\S]+?)<\/td>')
        con = re.findall(ur'([\s\S]+?)<br>[\s\S]+?([A-D])',res[0]) # 获取含有解析的答案
        if con:
            answer = con[0][1]
            analysis = con[0][0] # 获取解析
        else:
            answer = res[0]
            analysis = ''
        if answer:
            item = response.meta['item'] # 获取item
            item['answer'] = answer.strip()
            item['analysis'] = analysis.strip()
            item['answer_url'] = response.url
            yield item # 返回item,输出管道（pipelines.py）会自动接收该数据

 

items.py 数据结构定义:

# -*- coding: utf-8 -*-

# Define here the models for your scraped items
#
# See documentation in:
# https://doc.scrapy.org/en/latest/topics/items.html

import scrapy


class QuestionsItem(scrapy.Item):
    content = scrapy.Field()
    subject = scrapy.Field()
    level = scrapy.Field()
    answer = scrapy.Field()
    options = scrapy.Field()
    analysis = scrapy.Field()
    source = scrapy.Field()
    answer_url = scrapy.Field()
    pass

pipelines.py 输出管道（本例子输出的数据写入本地数据库）：

# -*- coding: utf-8 -*-

# Define your item pipelines here
#
# Don't forget to add your pipeline to the ITEM_PIPELINES setting
# See: https://doc.scrapy.org/en/latest/topics/item-pipeline.html
import pymysql
import md5


class QuestionsPipeline(object):
    def __init__(self):  
        # 建立数据库连接  
        self.connect = pymysql.connect('localhost','root','','question',use_unicode=True,charset='utf8')  
        # 获取游标 
        self.cursor = self.connect.cursor()  
        print("connecting mysql success!")  
        self.answer = ['A','B','C','D']
    def process_item(self, item, spider):
        content = pymysql.escape_string(item['content'])
　　　　 # 获取题目hash值，使用该字段过滤重复的题目
        m1 = md5.new()   
        m1.update(content)
        hash = m1.hexdigest()
        selectstr = "select id from question where hash='%s'"%(hash)
        self.cursor.execute(selectstr)
        res = self.cursor.fetchone()
        # 过滤相同的题目
        if not res:
　　　　　　　# 插入题目
            sqlstr = "insert into question(content,source,subject,level,answer,analysis,hash,answer_url) VALUES('%s','%s','%s','%s','%s','%s','%s','%s')"%(content,pymysql.escape_string(item['source']),item['subject'],item['level'],item['answer'],pymysql.escape_string(item['analysis']),hash,item['answer_url'])
            self.cursor.execute(sqlstr)
            qid = self.cursor.lastrowid
　　　　　　　# 插入选项
            for index in  range(len(item['options'])):
                option = item['options'][index]
                answer = self.answer.index(item['answer'])
                if answer==index:
                    ans = '2'
                else:
                    ans = '1'
                sqlstr = "insert into options(content,qid,answer) VALUES('%s','%s','%s')"%(pymysql.escape_string(option[0]),qid,ans)
                self.cursor.execute(sqlstr)
            self.connect.commit() 
            #self.connect.close() 
        return item

 

爬虫构建完毕后，在项目的根目录下运行

scrapy crawl xueersi  # scrapy crawl 爬虫的名称