from urlparse import urljoin from urllib import urlencode from lxml import html from cgi import escape categories = ['social media'] base_url = 'https://twitter.com/' search_url = base_url+'search?' title_xpath = './/span[@class="username js-action-profile-name"]//text()' content_xpath = './/p[@class="js-tweet-text tweet-text"]//text()' def request(query, params): global search_url params['url'] = search_url + urlencode({'q': query}) return params def response(resp): global base_url results = [] dom = html.fromstring(resp.text) for tweet in dom.xpath('//li[@data-item-type="tweet"]'): link = tweet.xpath('.//small[@class="time"]//a')[0] url = urljoin(base_url, link.attrib.get('href')) title = ''.join(tweet.xpath(title_xpath)) content = escape(''.join(tweet.xpath(content_xpath))) results.append({'url': url, 'title': title, 'content': content}) return results