statsmailer.py 1.2 KB

12345678910111213141516171819202122232425262728293031323334
  1. """
  2. StatsMailer extension sends an email when a spider finishes scraping.
  3. Use STATSMAILER_RCPTS setting to enable and give the recipient mail address
  4. """
  5. from scrapy import signals
  6. from scrapy.mail import MailSender
  7. from scrapy.exceptions import NotConfigured
  8. class StatsMailer(object):
  9. def __init__(self, stats, recipients, mail):
  10. self.stats = stats
  11. self.recipients = recipients
  12. self.mail = mail
  13. @classmethod
  14. def from_crawler(cls, crawler):
  15. recipients = crawler.settings.getlist("STATSMAILER_RCPTS")
  16. if not recipients:
  17. raise NotConfigured
  18. mail = MailSender.from_settings(crawler.settings)
  19. o = cls(crawler.stats, recipients, mail)
  20. crawler.signals.connect(o.spider_closed, signal=signals.spider_closed)
  21. return o
  22. def spider_closed(self, spider):
  23. spider_stats = self.stats.get_stats(spider)
  24. body = "Global stats\n\n"
  25. body += "\n".join("%-50s : %s" % i for i in self.stats.get_stats().items())
  26. body += "\n\n%s stats\n\n" % spider.name
  27. body += "\n".join("%-50s : %s" % i for i in spider_stats.items())
  28. return self.mail.send(self.recipients, "Scrapy stats for: %s" % spider.name, body)