scrapy 主动停止爬虫

"""CloseSpider is an extension that forces spiders to be closed after certain
conditions are met.

See documentation in docs/topics/extensions.rst
"""

class CloseSpider(object):

    def __init__(self, crawler):
        self.crawler = crawler

        self.close_on = {
            'timeout': crawler.settings.getfloat('CLOSESPIDER_TIMEOUT'),
            'itemcount': crawler.settings.getint('CLOSESPIDER_ITEMCOUNT'),
            'pagecount': crawler.settings.getint('CLOSESPIDER_PAGECOUNT'),
            'errorcount': crawler.settings.getint('CLOSESPIDER_ERRORCOUNT'),
            }

        if not any(self.close_on.values()):
            raise NotConfigured

        self.counter = defaultdict(int)

        if self.close_on.get('errorcount'):
            crawler.signals.connect(self.error_count, signal=signals.spider_error)
        if self.close_on.get('pagecount'):
            crawler.signals.connect(self.page_count, signal=signals.response_received)
        if self.close_on.get('timeout'):
            crawler.signals.connect(self.spider_opened, signal=signals.spider_opened)
        if self.close_on.get('itemcount'):
            crawler.signals.connect(self.item_scraped, signal=signals.item_scraped)
        crawler.signals.connect(self.spider_closed, signal=signals.spider_closed)

    @classmethod
    def from_crawler(cls, crawler):
        return cls(crawler)

    def error_count(self, failure, response, spider):
        self.counter['errorcount'] += 1
        if self.counter['errorcount'] == self.close_on['errorcount']:
            self.crawler.engine.close_spider(spider, 'closespider_errorcount')

    def page_count(self, response, request, spider):
        self.counter['pagecount'] += 1
        if self.counter['pagecount'] == self.close_on['pagecount']:
            self.crawler.engine.close_spider(spider, 'closespider_pagecount')

    def spider_opened(self, spider):
        self.task = reactor.callLater(self.close_on['timeout'], \
            self.crawler.engine.close_spider, spider, \
            reason='closespider_timeout')

    def item_scraped(self, item, spider):
        self.counter['itemcount'] += 1
        if self.counter['itemcount'] == self.close_on['itemcount']:
            self.crawler.engine.close_spider(spider, 'closespider_itemcount')

    def spider_closed(self, spider):
        task = getattr(self, 'task', False)
        if task and task.active():
            task.cancel()

 

1  上述代码是一个scrapy 关闭爬虫的一个的扩展类,从代码中可以看出主要是实现了timeout, itemcount, pagecount, errorcount 4种方式,因此可以在setting中设置这4种方式,当触发条件的时候会自动停止爬虫

在setting中设置

CLOSESPIDER_TIMEOUT  # 指定时间退出

CLOSESPIDER_ITEMCOUNT  # 生成了指定数量的item

CLOSESPIDER_PAGECOUNT  # 抓取了指定数量的响应

CLOSESPIDER_ERRORCOUNT  # 在发生指定数量的错误

# 打开EXTENSIONS扩展
EXTENSIONS = {
'scrapy.extensions.closespider.CloseSpider': 500,
}

 

 2  从CloseSpider类中可以了解到停止爬虫是使用了 self.crawler.engine.close_spider() 方法,因此当满足一定条件的时候我们也可以调用这个方法停止scrapy

# 在Spider文件中

self.crawler.engine.close_spider(self, 错误信息)

# 在middlewares文件中

spider.crawler.engine.close_spider(spider, 错误信息)