Scrapy框架自定义pipeline两层下载路径去下载图片，关于item传值的问题

自定义两层路径的时候，item是需要经过传值的，爬虫函数如下

import scrapy
from urllib.request import urljoin
from ..items import OffmymindspiderItem
class OffmymindSpider(scrapy.Spider):
    name = 'offmymind'
    allowed_domains = ['www.biaobaiju.com']
    start_urls = ['http://www.biaobaiju.com/']

    def parse(self, response):
        """
        获取每个分类的地址和分类的名称
        :param response:
        :return:
        """
        a_list = response.xpath("//ul[@class='nav clearfix']/li/a")
        for a in a_list:
            img_type_url = a.xpath("@href").extract_first("")
            img_type_name = a.xpath("text()").extract_first("")
            yield scrapy.Request(url=img_type_url, dont_filter=False, callback=self.parse_img_type_info, meta={"img_type_name":img_type_name})

    def parse_img_