name = 'sky'
start_urls = ['http://ivsky.com/']

def parse(self, response):

    selector = Selector(response)
    # print(response.text)
    types = selector.xpath("//div[@class='kw']/a")
    for type in types:
        typeUrl = type.xpath("@href").extract()[0]　　　　#分类地址
        typeName = type.xpath("text()").extract()[0] 　　#分类名称
        # print(typeUrl+" "+typeName)
        yield Request(self.start_urls[0]+typeUrl,callback=self.parseTotalPage,meta={'typeName':typeName})

    def parseTotalPage(self,response):

        typeName = response.meta["typeName"]
        # print(typeName)
        selector = Selector(response)
        # print(response.text)
        pageList = selector.xpath("//div[@class='pagelist']//a//@href").extract() #每一页的地址
        for page in pageList:
            yield Request(self.start_urls[0]+page,callback=self.parseGetImg,meta={'typeName':typeName})

def parseGetImg(self,response):

    typeName = response.meta["typeName"]
    selector = Selector(response)
    imgs = selector.xpath("//div[@class='il_img']//a")
    for img in imgs:
        imgUrl = img.xpath("@href").extract()[0]　　#一类图的地址
        # print(imgUrl+" "+imgName)
        yield Request(self.start_urls[0]+imgUrl,callback=self.parseGetMoreImg)

 def parseGetMoreImg(self,response):
        # / html / body / div[3] / div[4] / ul / li[3] / div / a / img
        selector = Selector(response)
        # print(response.text)
        items = IvskyItem()
        items["imgName"] = response.meta["imgName"]
        items["imgUrl"] = selector.xpath("//div[@class='il_img']//a//img//@src").extract()
        # print(items)
        yield items

ITEM_PIPELINES = {
   'ivsky.pipelines.IvskyPipeline': 300,
}