Question

我目前有以下规则：

# Matches all comments page under user overview,
# http://lookbook.nu/user/50784-Adam-G/comments/
Rule(SgmlLinkExtractor(allow=('/user/\d+[^/]+/comments/?$'), deny=('\?locale=')),
    callback='parse_model_comments'),
# http://lookbook.nu/user/50784-Adam-G/comments?page=2
Rule(SgmlLinkExtractor(allow=('/user/\d+[^/]+/comments\?page=\d+$'), deny=('\?locale=')),
    callback='parse_model_comments'),

在我的功能定义中，

def parse_model_comments(self, response):
  log.msg("Inside parse_model_comments")
  hxs = HtmlXPathSelector(response)
  model_url = hxs.select('//div[@id="userheader"]/h1/a/@href').extract()[0]
  comments_hxs = hxs.select(
      '//div[@id="profile_comments"]/div[@id="comments"]/div[@class="comment"]')
  if comments_hxs:
    log.msg("Yielding next page." + LookbookSpider.next_page(response.url))
    yield Request(LookbookSpider.next_page(response.url))

这是实际的运行日志：

2012-11-26 18:52:46-0800 [lookbook] DEBUG: Crawled (200) <GET http://lookbook.nu/user/1363501-Rachael-Jane-H/comments> (referer: None)
2012-11-26 18:52:46-0800 [lookbook] DEBUG: Crawled (200) <GET http://lookbook.nu/user/1363501-Rachael-Jane-H/comments> (referer: http://lookbook.nu/user/1363501-Rachael-Jane-H/comments)
2012-11-26 18:52:46-0800 [scrapy] INFO: Inside parse_model_comments
2012-11-26 18:52:46-0800 [scrapy] INFO: Yielding next page.http://lookbook.nu/user/1363501-Rachael-Jane-H/comments?page=2
2012-11-26 18:52:46-0800 [lookbook] DEBUG: Scraped from <200 http://lookbook.nu/user/1363501-Rachael-Jane-H/comments>
    {'model_url': u'http://lookbook.nu/rachinald',
     'posted_at': u'2012-11-26T13:21:49-05:00',
     'target_url': u'http://lookbook.nu/look/4290423-Blackout-Challenge-One',
     'text': u"Thanks Justina :) They're actually purple - the whole premise is to not wear black all week ^^",
     'type': 2}
...
2012-11-26 18:52:47-0800 [lookbook] DEBUG: Crawled (200) <GET http://lookbook.nu/user/1363501-Rachael-Jane-H/comments?page=2> (referer: http://lookbook.nu/user/1363501-Rachael-Jane-H/comments)
2012-11-26 18:52:48-0800 [lookbook] INFO: Closing spider (finished)
2012-11-26 18:52:48-0800 [lookbook] INFO: Dumping Scrapy stats:
    {'downloader/request_bytes': 2072,
     'downloader/request_count': 3,
     'downloader/request_method_count/GET': 3,
     'downloader/response_bytes': 51499,
     'downloader/response_count': 3,
     'downloader/response_status_count/200': 3,
     'finish_reason': 'finished',
     'finish_time': datetime.datetime(2012, 11, 27, 2, 52, 48, 43058),
     'item_scraped_count': 14,
     'log_count/DEBUG': 23,
     'log_count/INFO': 6,
     'request_depth_max': 3,
     'response_received_count': 3,
     'scheduler/dequeued': 3,
     'scheduler/dequeued/memory': 3,
     'scheduler/enqueued': 3,
     'scheduler/enqueued/memory': 3,
     'start_time': datetime.datetime(2012, 11, 27, 2, 52, 44, 446851)}
2012-11-26 18:52:48-0800 [lookbook] INFO: Spider closed (finished)

即使抓取了page = 2，也没有调用parse_model_comments，因为未记录“Inside parse_model_comments”。

我检查了re.search('/user/\d+[^/]+/comments\?page=\d+$', 'http://lookbook.nu/user/1363501-Rachael-Jane-H/comments?page=2')并确认确实有效。

抓取了为什么页面= 2的任何想法但未调用该函数？

Answer 1

当你产生一个Request时，你必须在CrawlSpider中手动指定回调。

如果请求对象没有设置回调，则将调用默认的parse（）。

Crawlspider的parse（）函数只返回[] 检查源

Answer 2

我知道它看起来很奇怪，但回调不应该是一个生成器（产生一些东西）

我建议：

def parse_model_comments(self, response):
   return list(_iter_parse_model_comments(self, response))

def _iter_parse_model_comments(self, response):
    # place your current code here

即使请求已排队，Scrapy也不会爬行？

2 个答案: