Beispiel #1
0
    def __init__(self, sort_mode=None, *args, **kwargs):
        from scrapy.conf import settings
        settings.overrides['DEPTH_PRIORITY'] = 1
        settings.overrides['SCHEDULER_DISK_QUEUE'] = 'scrapy.squeue.PickleFifoDiskQueue'
        settings.overrides['SCHEDULER_MEMORY_QUEUE'] = 'scrapy.squeue.FifoMemoryQueue'
        settings.overrides['CRAWLERA_ENABLED'] = True

        self.quantity = kwargs.get('quantity', 1000)  # default is 1000

        self.proxy = 'content.crawlera.com:8010'
        self.proxy_type = 'http'
        #self.user_agent = 'Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:32.0) Gecko/20100101 Firefox/32.0'
        self.user_agent = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_3) AppleWebKit/537.75.14 (KHTML, like Gecko) Version/7.0.3 Safari/7046A194A'

        self.br = BuyerReviewsBazaarApi(called_class=self)

        super(HayneedleProductSpider, self).__init__(
            site_name=self.allowed_domains[0],
            *args,
            **kwargs)
Beispiel #2
0
    def __init__(self, *args, **kwargs):
        self.br = BuyerReviewsBazaarApi(called_class=self)

        super(DebenhamsProductSpider, self).__init__(*args, **kwargs)
Beispiel #3
0
    def __init__(self, *args, **kwargs):
        self.br = BuyerReviewsBazaarApi(called_class=self)

        super(DockersProductsSpider, self).__init__(
            site_name=self.allowed_domains[0], *args, **kwargs)
Beispiel #4
0
 def __init__(self, *args, **kwargs):
     super(ChewyProductsSpider, self).__init__(*args, **kwargs)
     self.br = BuyerReviewsBazaarApi(called_class=self)
Beispiel #5
0
 def __init__(self, *args, **kwargs):
     super(PetcoProductsSpider, self).__init__(*args, **kwargs)
     self.br = BuyerReviewsBazaarApi(called_class=self)
     self.product_last_page = 0
Beispiel #6
0
    def __init__(self, *args, **kwargs):
        self.br = BuyerReviewsBazaarApi(called_class=self)
        self.start_index = 0

        super(MicrosoftStoreProductSpider, self).__init__(*args, **kwargs)