remoteco.py 文件源码

def parse_job(self, response):
        """Parse a joblink into a JobItem.
        """
        s = Selector(response)
        item = JobItem()
        item['url'] = response.url
        item['site'] = 'Remote.co'
        item['title'] = s.css('h1::text').extract_first()
        item['company'] = s.xpath(
            '//strong[@itemprop="name"]/text()').extract_first()
        job = s.css('.job-description')
        job.xpath('p[1]')
        item['text'] = s.xpath(
            '//div[@class="job_description"]//text()').extract()
        try:
            posted = s.xpath('//time//text()').extract_first()
            item['date_posted'] = utilities.naturaltime(
                posted.replace('Posted ', '')).isoformat()
        except Exception as e:
            self.logger.error(e)
        yield item