Question

因此，当我尝试从epinions.com中搜索产品评论信息时，如果主评论文本太长，则会有一个“阅读更多”链接到另一个页面。我从“http://www.epinions.com/reviews/samsung-galaxy-note-16-gb-cell-phone/pa_~1”中举了一个例子，如果你看一下第一篇评论，你会明白我的意思。

我想知道：是否有可能在for循环的每次迭代中都有一个小蜘蛛来抓取网址并从新链接中删除评论？我有以下代码，但它不适用于微小的“蜘蛛”。

这是我的代码：

from scrapy.spider import BaseSpider
from scrapy.selector import HtmlXPathSelector
from epinions_test.items import EpinionsTestItem
from scrapy.http import Response, HtmlResponse

class MySpider(BaseSpider):
    name = "epinions"
    allow_domains = ["epinions.com"]
    start_urls = ['http://www.epinions.com/reviews/samsung-galaxy-note-16-gb-cell-phone/pa_~1']

    def parse(self, response):
        hxs = HtmlXPathSelector(response)
        sites = hxs.select('//div[@class="review_info"]')

        items = []
        for sites in sites:
            item = EpinionsTestItem()
            item["title"] = sites.select('h2/a/text()').extract()
            item["star"] = sites.select('span/a/span/@title').extract()
            item["date"] = sites.select('span/span/span/@title').extract()
            item["review"] = sites.select('p/span/text()').extract()
# Everything works fine and i do have those four columns beautifully printed out, until....

            url2 = sites.select('p/span/a/@href').extract()
            url = str("http://www.epinions.com%s" %str(url2)[3:-2])
# This url is a string. when i print it out, it's like "http://www.epinions.com/review/samsung-galaxy-note-16-gb-cell-phone/content_624031731332", which looks legit.

            response2 = HtmlResponse(url)
# I tried in a scrapy shell, it shows that this is a htmlresponse...

            hxs2 = HtmlXPathSelector(response2)
            fullReview = hxs2.select('//div[@class = "user_review_full"]')
            item["url"] = fullReview.select('p/text()').extract()
# The three lines above works in an independent spider, where start_url is changed to the url just generated and everything.
# However, i got nothing from item["url"] in this code.

            items.append(item)
        return items

为什么item [“url”]什么都不返回？

谢谢！

Answer 1

您应该在回调中实例化新的Request并在item词典中传递meta：

from scrapy.http import Request
from scrapy.item import Item, Field
from scrapy.spider import BaseSpider
from scrapy.selector import HtmlXPathSelector


class EpinionsTestItem(Item):
    title = Field()
    star = Field()
    date = Field()
    review = Field()


class MySpider(BaseSpider):
    name = "epinions"
    allow_domains = ["epinions.com"]
    start_urls = ['http://www.epinions.com/reviews/samsung-galaxy-note-16-gb-cell-phone/pa_~1']

    def parse(self, response):
        hxs = HtmlXPathSelector(response)
        sites = hxs.select('//div[@class="review_info"]')

        for sites in sites:
            item = EpinionsTestItem()
            item["title"] = sites.select('h2/a/text()').extract()
            item["star"] = sites.select('span/a/span/@title').extract()
            item["date"] = sites.select('span/span/span/@title').extract()

            url = sites.select('p/span/a/@href').extract()
            url = str("http://www.epinions.com%s" % str(url)[3:-2])

            yield Request(url=url, callback=self.parse_url2, meta={'item': item})

    def parse_url2(self, response):
        hxs = HtmlXPathSelector(response)

        item = response.meta['item']
        fullReview = hxs.select('//div[@class = "user_review_full"]')
        item["review"] = fullReview.select('p/text()').extract()
        yield item

另见documentation。

希望有所帮助。

scrapy：蜘蛛中的一只小蜘蛛？

1 个答案: