Selenium识别验证码

url = ‘‘
driver = webdriver.PhantomJS()
driver.get(url)
driver.set_window_size(1200, 800) #此处一定要设置固定值，在其他的机器上面运行的时候可能会有问题
# 暂停10s确保登录界面加载完成
time.sleep(10)
# 截取登录界面的屏幕
screenshot_path = ‘screenshot.png‘
if os.path.exists(screenshot_path):
    os.remove(screenshot_path)
driver.save_screenshot(screenshot_path)

# 找到验证码元素，并获取到位置坐标
element = driver.find_element_by_xpath("//*[@id=\"login-content\"]/div[2]/div[4]/img")
left = int(element.location[‘x‘])
top = int(element.location[‘y‘])
right = int(element.location[‘x‘] + element.size[‘width‘])
bottom = int(element.location[‘y‘] + element.size[‘height‘])
# 从截图中抠出来验证码的区域
captcha_path = ‘captcha.png‘
if os.path.exists(captcha_path):
    os.remove(captcha_path)
img = Image.open(screenshot_path)
img = img.crop((left, top, right, bottom))
img.save(captcha_path)

def captcha_recognition(appkey, codeType, imagePath):
    """
    调用验证码在线识别的API
   :param appkey: 平台申请的appkey
   :param codeType: 验证码类型
   :param imagePath: 验证码图片路径
   :return: 查询结果
    """
    if not os.path.exists(imagePath):
        return ‘‘
    captcha_result = ‘‘
    submitUrl = ‘http://op.juhe.cn/vercode/index‘  # 接口地址
    # buld post body data
    boundary = ‘----------%s‘ % hex(int(time.time() * 1000))
    data = []
    data.append(‘--%s‘ % boundary)
    data.append(‘Content-Disposition: form-data; name="%s"\r\n‘ % ‘key‘)
    data.append(appkey)
    data.append(‘--%s‘ % boundary)
    data.append(‘Content-Disposition: form-data; name="%s"\r\n‘ % ‘codeType‘)
    data.append(codeType)
    data.append(‘--%s‘ % boundary)
    fr = open(imagePath, ‘rb‘)
    data.append(‘Content-Disposition: form-data; name="%s"; filename="b.png"‘ % ‘image‘)
    data.append(‘Content-Type: %s\r\n‘ % ‘image/png‘)
    data.append(fr.read())
    fr.close()
    data.append(‘--%s--\r\n‘ % boundary)
    http_body = ‘\r\n‘.join(data)
    try:
        req = urllib2.Request(submitUrl, data=http_body)
        req.add_header(‘Content-Type‘, ‘multipart/form-data; boundary=%s‘ % boundary)
        req.add_header(‘User-Agent‘, ‘Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2785.143 Safari/537.36‘)
        req.add_header(‘Referer‘, ‘http://op.juhe.cn/‘)
        resp = urllib2.urlopen(req, timeout=60)
        qrcont = resp.read()
        result = json.loads(qrcont, ‘utf-8‘)
        error_code = result[‘error_code‘]
        if (error_code == 0):
            data = result[‘result‘]
            captcha_result = data
        else:
            errorinfo = u"错误码:%s,描述:%s" % (result[‘error_code‘], result[‘reason‘])
            print errorinfo
    except Exception as e:
        print e
    return captcha_result