python - Scrapy:根据条件停止先前的解析功能

Question

我现在正在开发的一个爬虫有一个非常特殊的情况。第一个函数 parse_posts_pages 遍历特定论坛页面的所有页面，并为每个页面调用第二个函数 parse_posts。

def parse_posts_pages(self, response):
    thread_id = response.meta['thread_id']
    thread_link = response.meta['thread_link']
    thread_name = response.meta['thread_name']
    if len(response.xpath('//*[@id="postpagestats_above"]/text()').re(r'(\d+)')) == 3:
        posts_per_page = int(response.xpath('//*[@id="postpagestats_above"]/text()').re(r'(\d+)')[1])
        total_posts = int(response.xpath('//*[@id="postpagestats_above"]/text()').re(r'(\d+)')[2])
        if posts_per_page > 0:
            post_mod = total_posts % posts_per_page
            pages = total_posts / posts_per_page
            if post_mod > 0: pages += 1
        else: pages = 1

    for page in range(pages, 0, -1):
        cur_page = '' if page == 1 else '/page' + str(page)
        post_page_link = thread_link + cur_page
        return scrapy.Request(post_page_link, self.parse_posts, meta={'thread_id': thread_id, 'thread_name': thread_name})


def parse_posts(self, response):
    global maxPostIDByThread, executeFullSpider
    thread_id = response.meta['thread_id']
    thread_name = response.meta['thread_name']
    for post in response.xpath('//*[@id="posts"]/li'):
        post_id = post.xpath('@id').re(r'(\d.*)')[0]
        if not executeFullSpider and post_id in maxPostIDByThread:
            break #<- I need this break to also cancel the for from parse_posts_pages function
        ...

在第二个函数中有一个 if 条件。当此条件为真时，我需要中断当前的 for 循环以及来自 parse_posts_pages 的 for 循环，因为不需要继续分页。

有没有办法从第二个函数停止第一个函数中的for循环？

Answer 1

按照手册中的描述，只需引发 CloseSpider

How can I instruct a spider to stop itself?

Raise the CloseSpider from a callback.

from scrapy.exceptions import CloseSpider

def parse_page(self, response):
    if 'Bandwidth exceeded' in response.body:
        raise CloseSpider('bandwidth_exceeded')

http://doc.scrapy.org/en/latest/faq.html#how-can-i-instruct-a-spider-to-stop-itself http://doc.scrapy.org/en/latest/topics/exceptions.html#scrapy.exceptions.CloseSpider

Note that requests that are still in progress (HTTP request sent, response not yet received) will still be parsed. No new request will be processed though.

https://stackoverflow.com/a/23895143/5041915

更新: 实际上我发现了一些有趣的东西 If stop spider in main function.

可能会出现新的有效 worker 没有时间启动的情况，因为引发异常的速度更快。

我建议在回调函数中检查条件并尽早引发异常。