写爬虫，只是为了下小说

整体思路

spider

大概兴趣是第一生产力吧

网上找的小说现成的txt排版都有问题，缺少章节，不过一些盗版网站章节倒是挺全

本来想着是找到index页，爬取所有章节连接，然后拼接成网址，再进行细节爬取，但是搞了一下发现不知道怎么弄

所以后来就变成了直接从第一章节开始爬，找到标题和正文，用xpath就可以了，然后找到下一章的地址

看着网上一个博客，用的  yield ，我的理解就是一个循环递归吧

因为所有页面都是相同处理的，所以就相当于一个调用，简单粗暴的解决方式，一个函数  process_item(self, item, spider) 就解决了

pipeline

下载下来处理写成html，<h1> 和title，至于content直接带<p>，直接写了就可以

本来还想写目录的主体<li>之类的

还好还好calibre够强大，可以识别不规则的html，也可以自动生成目录，所以免去了很多步骤

设置了一个global，用来生成文件名的，从1开始递增，所以保证文件每一章节是顺序的

四个文件

item.py

# -*- coding: utf-8 -*-
# Define here the models for your scraped items
#
# See documentation in:
# http://doc.scrapy.org/topics/items.html

from scrapy.item import Item, Field

class NoveldownloadItem(Item):
    # define the fields for your item here like:
    # name = Field()
    title = Field() # title 
    content = Field() #content
    pass

setting.py

# -*- coding: utf-8 -*-
# Scrapy settings for noveldownload project
#
# For simplicity, this file contains only the most important settings by
# default. All the other settings are documented here:
#
#     http://doc.scrapy.org/topics/settings.html
#

BOT_NAME = 'noveldownload'
BOT_VERSION = '1.0'

SPIDER_MODULES = ['noveldownload.spiders']
NEWSPIDER_MODULE = 'noveldownload.spiders'
USER_AGENT = '%s/%s' % (BOT_NAME, BOT_VERSION)

DOWNLOAD_DELAY = 0

ITEM_PIPELINES = {
    'noveldownload.pipelines.NoveldownloadPipeline':300,
}


DOWNLOAD_DELAY = 0.2
#RANDOMIZE_DOWNLOAD_DELAY = True


CONCURRENT_ITEMS = 128
CONCURRENT_REQUEST = 64
CONCURRENT_REQUEST_PER_DOMAIN = 64


LOG_ENABLED = False

COOKIES_ENABLES = False

pineline.py

# -*- 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/topics/item-pipeline.html
INDEX = 100

class NoveldownloadPipeline(object):
	

    def process_item(self, item, spider):
    	global INDEX
    	datapath = '../2/'+str(INDEX)+'.html'
    	INDEX +=1
    	#print INDEX
    	fd = open(datapath,'a')
    	line = '<h1> '+ str(item['title']) + ' </h1>\n'+str(item['content'])+'\n'
    	fd.write(line)
    	fd.close
        return item

novel_spider.py

# -*- coding: utf-8 -*-
from scrapy.selector import HtmlXPathSelector
from scrapy.contrib.spiders import CrawlSpider,Rule
from scrapy.contrib.linkextractors.sgml import SgmlLinkExtractor
from noveldownload.items import NoveldownloadItem
from scrapy.http import Request

class NovelSpider(CrawlSpider):
    #name="doubanmoive"
    #allowed_domains=["movie.douban.com"]
    #start_urls=["http://movie.douban.com/top250"]
    name="noveldownload"
    allowed_domains=["lwxs520.com"]
    #start_urls=["http://www.xxx.html"]
    start_urls=["http://www.xxxx.html"]
#    rules=[
#        Rule(SgmlLinkExtractor(allow=(r'http://movie.douban.com/top250\?start=\d+.*'))),
#        Rule(SgmlLinkExtractor(allow=(r'http://movie.douban.com/subject/\d+')),callback="parse_item"),      
#    ]
    def parse(self,response):       
        prefix = "http://www.xxx"
        item = NoveldownloadItem()
        sel = HtmlXPathSelector(response)
        title = sel.select('//*[@id="bgdiv"]/table[2]/tbody/tr[1]/td/div/h1/text()').extract()
        # //*[@id="content"]/p
        content = sel.select('//*[@id="content"]/p').extract()
        #print title
        #print content
        item['title'] = title[0].encode('utf-8')
        item['content'] = content[0].encode('utf-8')
        #print item['title']
        #print item['content']      

        yield item

        nexturl = sel.select('//*[@id="thumb"]/a[3]/@href').extract()
        for url in nexturl:
            url = prefix + url
            print url
            yield Request(url, callback=self.parse)

总结

额，配置那里不是很清楚

还有就是不知道是网络问题，还是其他问题

最开始总是爬取10几个网页就开始无法GET，所以只能手动改 start_urls 和 global

但是后来把，速度虽然还是有时快有时慢，但是也能几百个不中断了，也不知道原因

另外就是，数据结构啊

返回值是list，或者是什么HtmlXPathSelector 结构，所以要extract() 或者 [0]，才能取到正真的值

总之，虽然很粗糙，但是还是完成了下载

并且calibre也做成了mobi的书，而且应该很好调整了，毕竟每个章节都是分开，就算顺序错误或者章节不对也好调整了，并且正在更新的书也很好添加了

不过这个程度太粗糙了，只是简单功能实现