def _get_new_urls(self, page_url, soup):
    new_urls=set()
    #/view/123.com
    links=soup.find_all('a',herf=re._compile(r"/view/\d+\.htm"))
    for link in links:
        new_url=link['herf']
        new_full_url=
        new_urls.add(new_full_url)
    return new_urls