cnnindonesia.py 文件源码

def parse(self, response):
        self.logger.info('parse: {}'.format(response))
        is_no_update = False

        # Collect list of news from current page
        # Note: no next page button on cnnindonesia, all is loaded here
        article_selectors = response.css('a.list_kontribusi');
        if not article_selectors:
            raise CloseSpider('article_selectors not found')

        for article in article_selectors:
            url_selectors = article.css('a::attr(href)')
            if not url_selectors:
                raise CloseSpider('url_selectors not found')
            url = url_selectors.extract()[0]

            # Example: Jumat, 23/09/2016 21:17
            info_selectors = article.css('div.text > div > span.tanggal::text')
            if not info_selectors:
                raise CloseSpider('info_selectors not found')
            info = info_selectors.extract()[0]
            info_time = info.split(',')[1].strip()

            # Parse date information
            try:
                # Example: 23/09/2016 21:17
                published_at_wib = datetime.strptime(info_time, '%d/%m/%Y %H:%M')
            except ValueError as err:
                raise CloseSpider('cannot_parse_date: {}'.format(err))
            published_at = wib_to_utc(published_at_wib)

            if self.media['last_scraped_at'] >= published_at:
                is_no_update = True
                break
            # For each url we create new scrapy Request
            yield Request(url, callback=self.parse_news)

        if is_no_update:
            self.logger.info('Media have no update')
            return