Python爬虫scrapy-redis分布式实例（一）

目标任务：将之前新浪网的Scrapy爬虫项目，修改为基于RedisSpider类的scrapy-redis分布式爬虫项目，将数据存入redis数据库。

 

一、item文件，和之前项目一样不需要改变

# -*- coding: utf-8 -*-
import scrapy
import sys
reload(sys)
sys.setdefaultencoding("utf-8")
class SinanewsItem(scrapy.Item):
# 大类的标题和url
parentTitle = scrapy.Field()
parentUrls = scrapy.Field()
</span><span style="color: #008000">#</span><span style="color: #008000"> 小类的标题和子url</span>
subTitle =<span style="color: #000000"> scrapy.Field()
subUrls </span>=<span style="color: #000000"> scrapy.Field()

</span><span style="color: #008000">#</span><span style="color: #008000"> 小类目录存储路径</span>
subFilename =<span style="color: #000000"> scrapy.Field()

</span><span style="color: #008000">#</span><span style="color: #008000"> 小类下的子链接</span>
sonUrls =<span style="color: #000000"> scrapy.Field()

</span><span style="color: #008000">#</span><span style="color: #008000"> 文章标题和内容</span>
head =<span style="color: #000000"> scrapy.Field()
content </span>= scrapy.Field()</pre>

 

二、spiders爬虫文件，使用RedisSpider类替换之前的Spider类，其余地方做些许改动即可，具体代码如下：

 

# -*- coding: utf-8 -*-
import scrapy
import os
from sinaNews.items import SinanewsItem
from scrapy_redis.spiders import RedisSpider
import sys
reload(sys)
sys.setdefaultencoding("utf-8")
class SinaSpider(RedisSpider):
name = "sina"
# 启动爬虫的命令
redis_key = "sinaspider:strat_urls"
　　# 动态定义爬虫爬取域范围
def init(self, args, **kwargs):
domain = kwargs.pop('domain', '')
self.allowed_domains = filter(None, domain.split(','))
super(SinaSpider, self).init(args, **kwargs)
</span><span style="color: #0000ff">def</span><span style="color: #000000"> parse(self, response):
    items</span>=<span style="color: #000000"> []
    </span><span style="color: #008000">#</span><span style="color: #008000"> 所有大类的url 和 标题</span>
    parentUrls = response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()
    parentTitle </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()

    </span><span style="color: #008000">#</span><span style="color: #008000"> 所有小类的ur 和 标题</span>
    subUrls  = response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()
    subTitle </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()

    </span><span style="color: #008000">#</span><span style="color: #008000">爬取所有大类</span>
    <span style="color: #0000ff">for</span> i <span style="color: #0000ff">in</span><span style="color: #000000"> range(0, len(parentTitle)):
        
        </span><span style="color: #008000">#</span><span style="color: #008000"> 爬取所有小类</span>
        <span style="color: #0000ff">for</span> j <span style="color: #0000ff">in</span><span style="color: #000000"> range(0, len(subUrls)):
            item </span>=<span style="color: #000000"> SinanewsItem()

            </span><span style="color: #008000">#</span><span style="color: #008000"> 保存大类的title和urls</span>
            item[<span style="color: #800000">'</span><span style="color: #800000">parentTitle</span><span style="color: #800000">'</span>] =<span style="color: #000000"> parentTitle[i]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span>] =<span style="color: #000000"> parentUrls[i]

            </span><span style="color: #008000">#</span><span style="color: #008000"> 检查小类的url是否以同类别大类url开头，如果是返回True (sports.sina.com.cn 和 sports.sina.com.cn/nba)</span>
            if_belong = subUrls[j].startswith(item[<span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span><span style="color: #000000">])

            </span><span style="color: #008000">#</span><span style="color: #008000"> 如果属于本大类，将存储目录放在本大类目录下</span>
            <span style="color: #0000ff">if</span><span style="color: #000000">(if_belong):
                
                </span><span style="color: #008000">#</span><span style="color: #008000"> 存储 小类url、title和filename字段数据</span>
                item[<span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span>] =<span style="color: #000000"> subUrls[j]
                item[</span><span style="color: #800000">'</span><span style="color: #800000">subTitle</span><span style="color: #800000">'</span>] =<span style="color: #000000">subTitle[j]
                items.append(item)

    </span><span style="color: #008000">#</span><span style="color: #008000">发送每个小类url的Request请求，得到Response连同包含meta数据 一同交给回调函数 second_parse 方法处理</span>
    <span style="color: #0000ff">for</span> item <span style="color: #0000ff">in</span><span style="color: #000000"> items:
        </span><span style="color: #0000ff">yield</span> scrapy.Request( url = item[<span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span>], meta={<span style="color: #800000">'</span><span style="color: #800000">meta_1</span><span style="color: #800000">'</span>: item}, callback=<span style="color: #000000">self.second_parse)

</span><span style="color: #008000">#</span><span style="color: #008000">对于返回的小类的url，再进行递归请求</span>
<span style="color: #0000ff">def</span><span style="color: #000000"> second_parse(self, response):
    </span><span style="color: #008000">#</span><span style="color: #008000"> 提取每次Response的meta数据</span>
    meta_1= response.meta[<span style="color: #800000">'</span><span style="color: #800000">meta_1</span><span style="color: #800000">'</span><span style="color: #000000">]

    </span><span style="color: #008000">#</span><span style="color: #008000"> 取出小类里所有子链接</span>
    sonUrls = response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//a/@href</span><span style="color: #800000">'</span><span style="color: #000000">).extract()

    items</span>=<span style="color: #000000"> []
    </span><span style="color: #0000ff">for</span> i <span style="color: #0000ff">in</span><span style="color: #000000"> range(0, len(sonUrls)):
        </span><span style="color: #008000">#</span><span style="color: #008000"> 检查每个链接是否以大类url开头、以.shtml结尾，如果是返回True</span>
        if_belong = sonUrls[i].endswith(<span style="color: #800000">'</span><span style="color: #800000">.shtml</span><span style="color: #800000">'</span>) <span style="color: #0000ff">and</span> sonUrls[i].startswith(meta_1[<span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span><span style="color: #000000">])

        </span><span style="color: #008000">#</span><span style="color: #008000"> 如果属于本大类，获取字段值放在同一个item下便于传输</span>
        <span style="color: #0000ff">if</span><span style="color: #000000">(if_belong):
            item </span>=<span style="color: #000000"> SinanewsItem()
            item[</span><span style="color: #800000">'</span><span style="color: #800000">parentTitle</span><span style="color: #800000">'</span>] =meta_1[<span style="color: #800000">'</span><span style="color: #800000">parentTitle</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span>] =meta_1[<span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span>] = meta_1[<span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">subTitle</span><span style="color: #800000">'</span>] = meta_1[<span style="color: #800000">'</span><span style="color: #800000">subTitle</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">sonUrls</span><span style="color: #800000">'</span>] =<span style="color: #000000"> sonUrls[i]
            items.append(item)

    </span><span style="color: #008000">#</span><span style="color: #008000">发送每个小类下子链接url的Request请求，得到Response后连同包含meta数据 一同交给回调函数 detail_parse 方法处理</span>
    <span style="color: #0000ff">for</span> item <span style="color: #0000ff">in</span><span style="color: #000000"> items:
            </span><span style="color: #0000ff">yield</span> scrapy.Request(url=item[<span style="color: #800000">'</span><span style="color: #800000">sonUrls</span><span style="color: #800000">'</span>], meta={<span style="color: #800000">'</span><span style="color: #800000">meta_2</span><span style="color: #800000">'</span>:item}, callback =<span style="color: #000000"> self.detail_parse)

</span><span style="color: #008000">#</span><span style="color: #008000"> 数据解析方法，获取文章标题和内容</span>
<span style="color: #0000ff">def</span><span style="color: #000000"> detail_parse(self, response):
    item </span>= response.meta[<span style="color: #800000">'</span><span style="color: #800000">meta_2</span><span style="color: #800000">'</span><span style="color: #000000">]
    content </span>= <span style="color: #800000">""</span><span style="color: #000000">
    head </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//h1[@>)
    content_list </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()

    </span><span style="color: #008000">#</span><span style="color: #008000"> 将p标签里的文本内容合并到一起</span>
    <span style="color: #0000ff">for</span> content_one <span style="color: #0000ff">in</span><span style="color: #000000"> content_list:
        content </span>+=<span style="color: #000000"> content_one

    item[</span><span style="color: #800000">'</span><span style="color: #800000">head</span><span style="color: #800000">'</span>]= head[0] <span style="color: #0000ff">if</span> len(head) &gt; 0 <span style="color: #0000ff">else</span> <span style="color: #800000">"</span><span style="color: #800000">NULL</span><span style="color: #800000">"</span><span style="color: #000000">
    item[</span><span style="color: #800000">'</span><span style="color: #800000">content</span><span style="color: #800000">'</span>]=<span style="color: #000000"> content

    </span><span style="color: #0000ff">yield</span> item</pre>

 

 

三、settings文件设置

SPIDER_MODULES = ['sinaNews.spiders']
NEWSPIDER_MODULE = 'sinaNews.spiders'
# 使用scrapy-redis里的去重组件，不使用scrapy默认的去重方式
DUPEFILTER_CLASS = "scrapy_redis.dupefilter.RFPDupeFilter"
# 使用scrapy-redis里的调度器组件，不使用默认的调度器
SCHEDULER = "scrapy_redis.scheduler.Scheduler"
# 允许暂停，redis请求记录不丢失
SCHEDULER_PERSIST = True
# 默认的scrapy-redis请求队列形式（按优先级）
SCHEDULER_QUEUE_CLASS = "scrapy_redis.queue.SpiderPriorityQueue"
# 队列形式，请求先进先出


 栈形式，请求先进后出


# 只是将数据放到redis数据库，不需要写pipelines文件
ITEM_PIPELINES = {
#    'Sina.pipelines.SinaPipeline': 300,
'scrapy_redis.pipelines.RedisPipeline': 400,
}
# LOG_LEVEL = 'DEBUG'
# Introduce an artifical delay to make use of parallelism. to speed up the
 crawl.
DOWNLOAD_DELAY = 1
# 指定数据库的主机IP
REDIS_HOST = "192.168.13.26"
# 指定数据库的端口号
REDIS_PORT = 6379

执行命令：

本次直接使用本地的redis数据库，将settings文件中的REDIS_HOST和REDIS_PORT注释掉。

启动爬虫程序

scrapy runspider sina.py

执行程序后终端窗口显示如下：

表示程序处于等待状态，此时在redis数据库端执行如下命令：

redis-cli> lpush sinaspider:start_urls http://news.sina.com.cn/guide/

http://news.sina.com.cn/guide/为起始url，此时程序开始执行。

目标任务：将之前新浪网的Scrapy爬虫项目，修改为基于RedisSpider类的scrapy-redis分布式爬虫项目，将数据存入redis数据库。

 

一、item文件，和之前项目一样不需要改变

# -*- coding: utf-8 -*-
import scrapy
import sys
reload(sys)
sys.setdefaultencoding("utf-8")
class SinanewsItem(scrapy.Item):
# 大类的标题和url
parentTitle = scrapy.Field()
parentUrls = scrapy.Field()
</span><span style="color: #008000">#</span><span style="color: #008000"> 小类的标题和子url</span>
subTitle =<span style="color: #000000"> scrapy.Field()
subUrls </span>=<span style="color: #000000"> scrapy.Field()

</span><span style="color: #008000">#</span><span style="color: #008000"> 小类目录存储路径</span>
subFilename =<span style="color: #000000"> scrapy.Field()

</span><span style="color: #008000">#</span><span style="color: #008000"> 小类下的子链接</span>
sonUrls =<span style="color: #000000"> scrapy.Field()

</span><span style="color: #008000">#</span><span style="color: #008000"> 文章标题和内容</span>
head =<span style="color: #000000"> scrapy.Field()
content </span>= scrapy.Field()</pre>

 

二、spiders爬虫文件，使用RedisSpider类替换之前的Spider类，其余地方做些许改动即可，具体代码如下：

 

# -*- coding: utf-8 -*-
import scrapy
import os
from sinaNews.items import SinanewsItem
from scrapy_redis.spiders import RedisSpider
import sys
reload(sys)
sys.setdefaultencoding("utf-8")
class SinaSpider(RedisSpider):
name = "sina"
# 启动爬虫的命令
redis_key = "sinaspider:strat_urls"
　　# 动态定义爬虫爬取域范围
def init(self, args, **kwargs):
domain = kwargs.pop('domain', '')
self.allowed_domains = filter(None, domain.split(','))
super(SinaSpider, self).init(args, **kwargs)
</span><span style="color: #0000ff">def</span><span style="color: #000000"> parse(self, response):
    items</span>=<span style="color: #000000"> []
    </span><span style="color: #008000">#</span><span style="color: #008000"> 所有大类的url 和 标题</span>
    parentUrls = response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()
    parentTitle </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()

    </span><span style="color: #008000">#</span><span style="color: #008000"> 所有小类的ur 和 标题</span>
    subUrls  = response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()
    subTitle </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()

    </span><span style="color: #008000">#</span><span style="color: #008000">爬取所有大类</span>
    <span style="color: #0000ff">for</span> i <span style="color: #0000ff">in</span><span style="color: #000000"> range(0, len(parentTitle)):
        
        </span><span style="color: #008000">#</span><span style="color: #008000"> 爬取所有小类</span>
        <span style="color: #0000ff">for</span> j <span style="color: #0000ff">in</span><span style="color: #000000"> range(0, len(subUrls)):
            item </span>=<span style="color: #000000"> SinanewsItem()

            </span><span style="color: #008000">#</span><span style="color: #008000"> 保存大类的title和urls</span>
            item[<span style="color: #800000">'</span><span style="color: #800000">parentTitle</span><span style="color: #800000">'</span>] =<span style="color: #000000"> parentTitle[i]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span>] =<span style="color: #000000"> parentUrls[i]

            </span><span style="color: #008000">#</span><span style="color: #008000"> 检查小类的url是否以同类别大类url开头，如果是返回True (sports.sina.com.cn 和 sports.sina.com.cn/nba)</span>
            if_belong = subUrls[j].startswith(item[<span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span><span style="color: #000000">])

            </span><span style="color: #008000">#</span><span style="color: #008000"> 如果属于本大类，将存储目录放在本大类目录下</span>
            <span style="color: #0000ff">if</span><span style="color: #000000">(if_belong):
                
                </span><span style="color: #008000">#</span><span style="color: #008000"> 存储 小类url、title和filename字段数据</span>
                item[<span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span>] =<span style="color: #000000"> subUrls[j]
                item[</span><span style="color: #800000">'</span><span style="color: #800000">subTitle</span><span style="color: #800000">'</span>] =<span style="color: #000000">subTitle[j]
                items.append(item)

    </span><span style="color: #008000">#</span><span style="color: #008000">发送每个小类url的Request请求，得到Response连同包含meta数据 一同交给回调函数 second_parse 方法处理</span>
    <span style="color: #0000ff">for</span> item <span style="color: #0000ff">in</span><span style="color: #000000"> items:
        </span><span style="color: #0000ff">yield</span> scrapy.Request( url = item[<span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span>], meta={<span style="color: #800000">'</span><span style="color: #800000">meta_1</span><span style="color: #800000">'</span>: item}, callback=<span style="color: #000000">self.second_parse)

</span><span style="color: #008000">#</span><span style="color: #008000">对于返回的小类的url，再进行递归请求</span>
<span style="color: #0000ff">def</span><span style="color: #000000"> second_parse(self, response):
    </span><span style="color: #008000">#</span><span style="color: #008000"> 提取每次Response的meta数据</span>
    meta_1= response.meta[<span style="color: #800000">'</span><span style="color: #800000">meta_1</span><span style="color: #800000">'</span><span style="color: #000000">]

    </span><span style="color: #008000">#</span><span style="color: #008000"> 取出小类里所有子链接</span>
    sonUrls = response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//a/@href</span><span style="color: #800000">'</span><span style="color: #000000">).extract()

    items</span>=<span style="color: #000000"> []
    </span><span style="color: #0000ff">for</span> i <span style="color: #0000ff">in</span><span style="color: #000000"> range(0, len(sonUrls)):
        </span><span style="color: #008000">#</span><span style="color: #008000"> 检查每个链接是否以大类url开头、以.shtml结尾，如果是返回True</span>
        if_belong = sonUrls[i].endswith(<span style="color: #800000">'</span><span style="color: #800000">.shtml</span><span style="color: #800000">'</span>) <span style="color: #0000ff">and</span> sonUrls[i].startswith(meta_1[<span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span><span style="color: #000000">])

        </span><span style="color: #008000">#</span><span style="color: #008000"> 如果属于本大类，获取字段值放在同一个item下便于传输</span>
        <span style="color: #0000ff">if</span><span style="color: #000000">(if_belong):
            item </span>=<span style="color: #000000"> SinanewsItem()
            item[</span><span style="color: #800000">'</span><span style="color: #800000">parentTitle</span><span style="color: #800000">'</span>] =meta_1[<span style="color: #800000">'</span><span style="color: #800000">parentTitle</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span>] =meta_1[<span style="color: #800000">'</span><span style="color: #800000">parentUrls</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span>] = meta_1[<span style="color: #800000">'</span><span style="color: #800000">subUrls</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">subTitle</span><span style="color: #800000">'</span>] = meta_1[<span style="color: #800000">'</span><span style="color: #800000">subTitle</span><span style="color: #800000">'</span><span style="color: #000000">]
            item[</span><span style="color: #800000">'</span><span style="color: #800000">sonUrls</span><span style="color: #800000">'</span>] =<span style="color: #000000"> sonUrls[i]
            items.append(item)

    </span><span style="color: #008000">#</span><span style="color: #008000">发送每个小类下子链接url的Request请求，得到Response后连同包含meta数据 一同交给回调函数 detail_parse 方法处理</span>
    <span style="color: #0000ff">for</span> item <span style="color: #0000ff">in</span><span style="color: #000000"> items:
            </span><span style="color: #0000ff">yield</span> scrapy.Request(url=item[<span style="color: #800000">'</span><span style="color: #800000">sonUrls</span><span style="color: #800000">'</span>], meta={<span style="color: #800000">'</span><span style="color: #800000">meta_2</span><span style="color: #800000">'</span>:item}, callback =<span style="color: #000000"> self.detail_parse)

</span><span style="color: #008000">#</span><span style="color: #008000"> 数据解析方法，获取文章标题和内容</span>
<span style="color: #0000ff">def</span><span style="color: #000000"> detail_parse(self, response):
    item </span>= response.meta[<span style="color: #800000">'</span><span style="color: #800000">meta_2</span><span style="color: #800000">'</span><span style="color: #000000">]
    content </span>= <span style="color: #800000">""</span><span style="color: #000000">
    head </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//h1[@>)
    content_list </span>= response.xpath(<span style="color: #800000">'</span><span style="color: #800000">//div[@>).extract()

    </span><span style="color: #008000">#</span><span style="color: #008000"> 将p标签里的文本内容合并到一起</span>
    <span style="color: #0000ff">for</span> content_one <span style="color: #0000ff">in</span><span style="color: #000000"> content_list:
        content </span>+=<span style="color: #000000"> content_one

    item[</span><span style="color: #800000">'</span><span style="color: #800000">head</span><span style="color: #800000">'</span>]= head[0] <span style="color: #0000ff">if</span> len(head) &gt; 0 <span style="color: #0000ff">else</span> <span style="color: #800000">"</span><span style="color: #800000">NULL</span><span style="color: #800000">"</span><span style="color: #000000">
    item[</span><span style="color: #800000">'</span><span style="color: #800000">content</span><span style="color: #800000">'</span>]=<span style="color: #000000"> content

    </span><span style="color: #0000ff">yield</span> item</pre>

 

 

三、settings文件设置

SPIDER_MODULES = ['sinaNews.spiders']
NEWSPIDER_MODULE = 'sinaNews.spiders'
# 使用scrapy-redis里的去重组件，不使用scrapy默认的去重方式
DUPEFILTER_CLASS = "scrapy_redis.dupefilter.RFPDupeFilter"
# 使用scrapy-redis里的调度器组件，不使用默认的调度器
SCHEDULER = "scrapy_redis.scheduler.Scheduler"
# 允许暂停，redis请求记录不丢失
SCHEDULER_PERSIST = True
# 默认的scrapy-redis请求队列形式（按优先级）
SCHEDULER_QUEUE_CLASS = "scrapy_redis.queue.SpiderPriorityQueue"
# 队列形式，请求先进先出


 栈形式，请求先进后出


# 只是将数据放到redis数据库，不需要写pipelines文件
ITEM_PIPELINES = {
#    'Sina.pipelines.SinaPipeline': 300,
'scrapy_redis.pipelines.RedisPipeline': 400,
}
# LOG_LEVEL = 'DEBUG'
# Introduce an artifical delay to make use of parallelism. to speed up the
 crawl.
DOWNLOAD_DELAY = 1
# 指定数据库的主机IP
REDIS_HOST = "192.168.13.26"
# 指定数据库的端口号
REDIS_PORT = 6379

执行命令：

本次直接使用本地的redis数据库，将settings文件中的REDIS_HOST和REDIS_PORT注释掉。

启动爬虫程序

scrapy runspider sina.py

执行程序后终端窗口显示如下：

表示程序处于等待状态，此时在redis数据库端执行如下命令：

redis-cli> lpush sinaspider:start_urls http://news.sina.com.cn/guide/

http://news.sina.com.cn/guide/为起始url，此时程序开始执行。