python - 爬行时清空输出文件

Question

我知道我已经问过类似的问题，但它是一个新的蜘蛛，我也有同样的问题( Crawling data successfully but cannot scraped or write it into csv )...我将我的另一个蜘蛛放在这里，并提供了我应该拥有的输出示例和所有信息我通常需要获取输出文件...有人可以帮助我吗？我必须在周五完成这只蜘蛛...所以，我很着急!!

奇怪的是，我的 Fnac.csv 已创建但始终为空...所以我尝试直接在我想要抓取的页面示例上运行我的蜘蛛，并且我拥有我需要的所有信息...所以，我不明白...也许问题只是来 self 的规则或其他什么？

我的蜘蛛:

# -*- coding: utf-8 -*-
# Every import is done for a specific use
import scrapy                                       # Once you downloaded scrapy, you have to import it in your code to use it.
import re                                           # To use the .re() function, which extracts just a part of the text you crawl. It's using regex (regular expressions)
import numbers                                      # To use mathematics things, in this case : numbers.
from fnac.items import FnacItem                     # To return the items you want. Each item has a space allocated in the momery, created in the items.py file, which is in the second cdiscount_test directory.
from urllib.request import urlopen                  # To use urlopen, which allow the spider to find the links in a page that is in the actual page.
from scrapy.spiders import CrawlSpider, Rule        # To use rules and LinkExtractor, which allowed the spider to follow every url on the page you crawl.
from scrapy.linkextractors import LinkExtractor     # Look above.
from bs4 import BeautifulSoup                       # To crawl an iframe, which is a page in a page in web prgrammation.

# Your spider
class Fnac(CrawlSpider):
    name = 'FnacCom'                             # Name of your spider. You call it in the anaconda prompt.
    allowed_domains = ['fnac.com']               # Web domains allowed by you, your spider cannot enter on a page which is not in that domain.
    start_urls = ['https://www.fnac.com/Index-Vendeurs-MarketPlace/A/']        # The first link you crawl.

    # To allow your spider to follow the urls that are on the actual page.
    rules = (
        Rule(LinkExtractor(), callback='parse_start_url'),
    )

    # Your function that crawl the actual page you're on.
    def parse_start_url(self, response):
        item = FnacItem() # The spider now knowws that the items you want have to be stored in the item variable.

        # First data you want which are on the actual page.
        nb_sales = response.xpath('//body//table[@summary="données détaillée du vendeur"]/tbody/tr/td/span/text()').re(r'([\d]*) ventes')
        country = response.xpath('//body//table[@summary="données détaillée du vendeur"]/tbody/tr/td/text()').re(r'([A-Z].*)')

        # To store the data in their right places.
        item['nb_sales'] = ''.join(nb_sales).strip()
        item['country'] = ''.join(country).strip()

        # Find a specific link on the actual page and launch this function on it. It's the place where you will find your two first data.
        test_list = response.xpath('//a/@href')
        for test_list in response.xpath('.//div[@class="ProductPriceBox-item detail"]'):
            temporary = response.xpath('//div[@class="ProductPriceBox-item detail"]/div/a/@href').extract()
            for i in range(len(temporary)):
                scrapy.Request(temporary[i], callback=self.parse_start_url, meta={'dont_redirect': True, 'item': item})

        # To find the iframe on a page, launch the next function.
        yield scrapy.Request(response.url, callback=self.parse_iframe, meta={'dont_redirect': True, 'item': item})

    # Your function that crawl the iframe on a page
    def parse_iframe(self, response):
        f_item1 = response.meta['item'] # Just to use the same item location you used above.

        # Find all the iframe on a page.
        soup = BeautifulSoup(urlopen(response.url), "lxml")
        iframexx = soup.find_all('iframe')

        # If there's at least one iframe, launch the next function on it
        if (len(iframexx) != 0):
            for iframe in iframexx:
                yield scrapy.Request(iframe.attrs['src'], callback=self.extract_or_loop, meta={'dont_redirect': True, 'item': f_item1})

        # If there's no iframe, launch the next function on the link of the page where you looked after the potential iframe.
        else:
            yield scrapy.Request(response.url, callback=self.extract_or_loop, meta={'dont_redirect': True, 'item': f_item1})

    # Function to find the other data.
    def extract_or_loop(self, response):
        f_item2 = response.meta['item'] # Just to use the same item location you used above.

        # The rest of the data you want.
        address = response.xpath('//body//div/p/text()').re(r'.*Adresse \: (.*)\n?.*')
        email = response.xpath('//body//div/ul/li[contains(text(),"@")]/text()').extract()
        name = response.xpath('//body//div/p[@class="customer-policy-label"]/text()').re(r'Infos sur la boutique \: ([a-zA-Z0-9]*\s*)')
        phone = response.xpath('//body//div/p/text()').re(r'.*Tél \: ([\d]*)\n?.*')
        siret = response.xpath('//body//div/p/text()').re(r'.*Siret \: ([\d]*)\n?.*')
        vat = response.xpath('//body//div/text()').re(r'.*TVA \: (.*)')

        # If the name of the seller exist, then return the data.
        if (len(name) != 0):
            f_item2['name'] = ''.join(name).strip()
            f_item2['address'] = ''.join(address).strip()
            f_item2['phone'] = ''.join(phone).strip()
            f_item2['email'] = ''.join(email).strip()
            f_item2['vat'] = ''.join(vat).strip()
            f_item2['siret'] = ''.join(siret).strip()
            yield f_item2

        # If not, there was no data on the page and you have to find all the links on your page and launch the first function on them.
        else:
            for sel in response.xpath('//html/body'):
                list_urls = sel.xpath('//a/@href').extract()
                list_iframe = response.xpath('//div[@class="ProductPriceBox-item detail"]/div/a/@href').extract()
                if (len(list_iframe) != 0):
                    for list_iframe in list_urls:
                        yield scrapy.Request(list_iframe, callback=self.parse_start_url, meta={'dont_redirect': True})
                for url in list_urls:
                    yield scrapy.Request(response.urljoin(url), callback=self.parse_start_url, meta={'dont_redirect': True})

我的设置:

BOT_NAME = 'fnac'

SPIDER_MODULES = ['fnac.spiders']
NEWSPIDER_MODULE = 'fnac.spiders'
DOWNLOAD_DELAY = 2
COOKIES_ENABLED = False
ITEM_PIPELINES = {
   'fnac.pipelines.FnacPipeline': 300,
}

我的管道:

# -*- coding: utf-8 -*-
from scrapy import signals
from scrapy.exporters import CsvItemExporter

# Define your item pipelines here
#
# Don't forget to add your pipeline to the ITEM_PIPELINES setting
# See: http://doc.scrapy.org/en/latest/topics/item-pipeline.html

# Define your output file.
class FnacPipeline(CsvItemExporter):
    def __init__(self):
        self.files = {}

    @classmethod
    def from_crawler(cls, crawler):
        pipeline = cls()
        crawler.signals.connect(pipeline.spider_opened, signals.spider_opened)
        crawler.signals.connect(pipeline.spider_closed, signals.spider_closed)
        return pipeline

    def spider_opened(self, spider):
        f = open('..\\..\\..\\..\\Fnac.csv', 'w').close()
        file = open('..\\..\\..\\..\\Fnac.csv', 'wb')
        self.files[spider] = file
        self.exporter = CsvItemExporter(file)
        self.exporter.start_exporting()

    def spider_closed(self, spider):
        self.exporter.finish_exporting()
        file = self.files.pop(spider)
        file.close()

    def process_item(self, item, spider):
        self.exporter.export_item(item)
        return item

我的元素:

# -*- coding: utf-8 -*-
import scrapy

# Define here the models for your scraped items

# See documentation in:
# http://doc.scrapy.org/en/latest/topics/items.html

class FnacItem(scrapy.Item):
    # define the fields for your items :
    # name = scrapy.Field()
    name = scrapy.Field()
    nb_sales = scrapy.Field()
    country = scrapy.Field()
    address = scrapy.Field()
    siret = scrapy.Field()
    vat = scrapy.Field()
    phone = scrapy.Field()
    email = scrapy.Field()

我在提示符中编写的运行蜘蛛的命令是:

scrapy爬取FnacCom

输出示例如下:

2017-08-08 10:21:54 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Panasonic/TV-par-marque/nsh474980/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:21:56 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Philips/TV-par-marque/nsh474981/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:21:58 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Sony/TV-par-marque/nsh475001/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:01 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-LG/TV-par-marque/nsh474979/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:03 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Samsung/TV-par-marque/nsh474984/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:06 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Television/TV-par-marque/shi474972/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:08 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Television/TV-par-prix/shi474946/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:11 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Television/TV-par-taille-d-ecran/shi474945/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:12 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Television/TV-par-Technologie/shi474944/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:15 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Smart-TV-TV-connectee/TV-par-Technologie/nsh474953/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:18 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-QLED/TV-par-Technologie/nsh474948/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:21 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-4K-UHD/TV-par-Technologie/nsh474947/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:23 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Toutes-les-TV/TV-Television/nsh474940/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:26 [scrapy.extensions.logstats] INFO: Crawled 459 pages (at 24 pages/min), scraped 0 items (at 0 items/min)
2017-08-08 10:22:26 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-Television/shi474914/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:28 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/partner/canalplus#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:34 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Meilleures-ventes-TV/TV-Television/nsh474942/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:37 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Toutes-nos-Offres/Offres-de-remboursement/shi159784/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:38 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Offres-Adherents/Toutes-nos-Offres/nsh81745/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:41 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/labofnac#bl=MMtvh#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:44 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Lecteur-et-Enregistreur-DVD-Blu-Ray/Lecteur-DVD-Blu-Ray/shi475063/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:46 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/TV-OLED/TV-par-Technologie/nsh474949/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:49 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Lecteur-DVD-Portable/Lecteur-et-Enregistreur-DVD-Blu-Ray/nsh475064/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:52 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Home-Cinema/Home-Cinema-par-marque/shi475116/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:52 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Univers-TV/Univers-Ecran-plat/cl179/w-4#bl=MMtvh> (referer: https://www.fnac.com)
2017-08-08 10:22:55 [scrapy.core.engine] DEBUG: Crawled (200) <GET https://www.fnac.com/Casque-TV-HiFi/Casque-par-usage/nsh450507/w-4#bl=MMtvh> (referer: https://www.fnac.com)

非常感谢您的帮助!!!

Answer 1

我编写了一个小代码重构来展示如何在不使用crawlspider和使用常见的scrapy习惯用法的情况下显式地编写spider:

class Fnac(Spider):
    name = 'fnac.com' 
    allowed_domains = ['fnac.com'] 
    start_urls = ['https://www.fnac.com/Index-Vendeurs-MarketPlace/0/']  # The first link you crawl.

    def parse(self, response):
        # parse sellers
        sellers = response.xpath("//h1[contains(selftext(),'MarketPlace')]/following-sibling::ul/li/a/@href").extract()
        for url in sellers:
            yield Request(url, callback=self.parse_seller)

        # parse other pages A-Z
        pages = response.css('.pagerletter a::attr(href)').extract()
        for url in pages:
            yield Request(url, callback=self.parse)

    def parse_seller(self, response):
        nb_sales = response.xpath('//body//table[@summary="données détaillée du vendeur"]/tbody/tr/td/span/text()').re(r'([\d]*) ventes')
        country = response.xpath('//body//table[@summary="données détaillée du vendeur"]/tbody/tr/td/text()').re(r'([A-Z].*)')
        item = FnacItem()
        # To store the data in their right places.
        item['nb_sales'] = ''.join(nb_sales).strip()
        item['country'] = ''.join(country).strip()
        # go to details page now
        details_url = response.xpath("//iframe/@src[contains(.,'retour')]").extract_first()
        yield Request(details_url, self.parse_seller_details,
                      meta={'item': item})  # carry over our item to next response

    def parse_seller_details(self, response):
        item = response.meta['item']  # get item that's got filled in `parse_seller`
        address = response.xpath('//body//div/p/text()').re(r'.*Adresse \: (.*)\n?.*')
        email = response.xpath('//body//div/ul/li[contains(text(),"@")]/text()').extract()
        # parse here
        yield item