util.py 文件源码

def get_parrafos(soup):
    prfs= soup.find_all(['li','table'])
    ps = soup.find_all('p')
    for p in ps:
        if not p.span:
            prfs.append(p)
            continue
        flag=False
        for c in p.contents:
            if ((isinstance(c, bs4.NavigableString) or isinstance(c, unicode)) and not is_vacio(c)):
                flag=True
                break
        if flag:
            prfs.append(p)
    return prfs