def __unlock_sogou(self,
url,
resp,
session,
unlock_callback=None,
identify_image_callback=None):
if unlock_callback is None:
unlock_callback = unlock_sogou_callback_example
millis = int(round(time.time() * 1000))
r_captcha = session.get(
'http://weixin.sogou.com/antispider/util/seccode.php?tc={}'.format(
millis),
headers={
'Referer': url,
})
if not r_captcha.ok:
raise WechatSogouRequestsException('WechatSogouAPI get img',
r_captcha)
r_unlock = unlock_callback(url, session, resp, r_captcha.content,
identify_image_callback)
if r_unlock['code'] != 0:
raise WechatSogouVcodeOcrException(
'[WechatSogouAPI identify image] code: {code}, msg: {msg}'.
format(code=r_unlock.get('code'), msg=r_unlock.get('msg')))
else:
self.__set_cache(session.cookies.get('SUID'), r_unlock['id'])
def __unlock_sogou(self,
url,
resp,
session,
unlock_callback=None,
identify_image_callback=None):
if unlock_callback is None:
unlock_callback = unlock_sogou_callback_example
millis = int(round(time.time() * 1000))
#这个r_captcha.content就是img图片
r_captcha = session.get(
'http://weixin.sogou.com/antispider/util/seccode.php?tc={}'.format(
millis))
if not r_captcha.ok:
raise WechatSogouRequestsException('WechatSogouAPI get img', resp)
#unlock_callback为unlock_sogou_callback_example,r_unlock返回的为手动输入的验证码，identify_image_callback为identify_image_callback_by_hand
r_unlock = unlock_callback(url, session, resp, r_captcha.content,
identify_image_callback)
if r_unlock['code'] != 0:
raise WechatSogouVcodeOcrException(
'[WechatSogouAPI identify image] code: {code}, msg: {msg}'.
format(code=r_unlock.get('code'), msg=r_unlock.get('msg')))
else:
print('suid', session.cookies.get('SUID'))
print('SNUID', r_unlock['id'])
self.__set_cache(session.cookies.get('SUID'), r_unlock['id'])
def get_by_unlock(self,
url,
referer=None,
unlock_platform=None,
unlock_callback=None,
identify_image_callback=None):
#保证uvlock_platform可用，这里调用的__unlock_sougou
assert unlock_platform is None or callable(unlock_platform)
#验证码识别用identify_image_callback_by_hand即手动识别
if identify_image_callback is None:
identify_image_callback = identify_image_callback_by_hand
#断言验证码识别可用
assert callable(identify_image_callback)
#断言unlock_callback可以调用或者为None，这里为None
assert unlock_callback is None or callable(unlock_callback)
#建立一个session的请求
session = requests.session()
#对搜索页面的网页进行请求，但是要设置headers,这里的refer还是为None
resp = self.__get(url=url,
session=session,
headers=self.__set_cookie(referer=referer))
#如果弹出验证码那个url中就带有antispider就识别验证码,在这里才需要unlock_callback和identify_image_callback
if 'antispider' in resp.url or '请输入验证码' in resp.text:
#for i in range(self.captcha_break_times):
print('要输入验证码')
for i in range(2):
try:
#这个地方进入了打码平台，会进行验证码验证__unlock_sougou，unlock_callback还是None，idenfify_image_callback是identify_image_callback_by_hand
#resp为进入了验证码那个界面
unlock_platform(url, resp, session, unlock_callback,
identify_image_callback)
break
except WechatSogouVcodeOcrException as e:
if i == 1:
raise WechatSogouVcodeOcrException(e)
if '请输入验证码' in resp.text:
resp = session.get(url)
else:
resp = self.__get(url,
session,
headers=self.__set_cookie(referer=referer))
#没有验证码就直接返回得到的网页了
return resp
def unlock_weixin_callback_example(url, req, resp, img,
identify_image_callback):
"""手动打码解锁
Parameters
----------
url : str or unicode
验证码页面 之前的 url
req : requests.sessions.Session
requests.Session() 供调用解锁
resp : requests.models.Response
requests 访问页面返回的，已经跳转了
img : bytes
验证码图片二进制数据
identify_image_callback : callable
处理验证码函数，输入验证码二进制数据，输出文字，参见 identify_image_callback_example
Returns
-------
dict
{
'ret': '',
'errmsg': '',
'cookie_count': '',
}
"""
# no use resp
unlock_url = 'https://mp.weixin.qq.com/mp/verifycode'
data = {'cert': time.time() * 1000, 'input': identify_image_callback(img)}
headers = {
'Host': 'mp.weixin.qq.com',
'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
'Referer': url
}
r_unlock = req.post(unlock_url, data, headers=headers)
if not r_unlock.ok:
raise WechatSogouVcodeOcrException('unlock[{}] failed: {}[{}]'.format(
unlock_url, r_unlock.text, r_unlock.status_code))
return r_unlock.json()
def unlock_sogou_callback_example(url, req, resp, img,
identify_image_callback):
"""手动打码解锁
Parameters
----------
url : str or unicode
验证码页面 之前的 url
req : requests.sessions.Session
requests.Session() 供调用解锁
resp : requests.models.Response
requests 访问页面返回的，已经跳转了
img : bytes
验证码图片二进制数据
identify_image_callback : callable
处理验证码函数，输入验证码二进制数据，输出文字，参见 identify_image_callback_example
Returns
-------
dict
{
'code': '',
'msg': '',
}
"""
# no use resp
url_quote = url.split('weixin.sogou.com/')[-1]
unlock_url = 'http://weixin.sogou.com/antispider/thank.php'
data = {'c': identify_image_callback(img), 'r': '%2F' + url_quote, 'v': 5}
headers = {
'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
'Referer': 'http://weixin.sogou.com/antispider/?from=%2f' + url_quote
}
print(data)
r_unlock = req.post(unlock_url, data, headers=headers)
if not r_unlock.ok:
raise WechatSogouVcodeOcrException('unlock[{}] failed: {}'.format(
unlock_url, r_unlock.text, r_unlock.status_code))
print(r_unlock.json())
return r_unlock.json()
def __get_by_unlock(self,
url,
referer=None,
unlock_platform=None,
unlock_callback=None,
identify_image_callback=None,
session=None):
assert unlock_platform is None or callable(unlock_platform)
if identify_image_callback is None:
identify_image_callback = identify_image_callback_by_hand
assert unlock_callback is None or callable(unlock_callback)
assert callable(identify_image_callback)
if not session:
session = requests.session()
resp = self.__get(url,
session,
headers=self.__set_cookie(referer=referer))
resp.encoding = 'utf-8'
if 'antispider' in resp.url or '请输入验证码' in resp.text:
for i in range(self.captcha_break_times):
try:
unlock_platform(
url=url,
resp=resp,
session=session,
unlock_callback=unlock_callback,
identify_image_callback=identify_image_callback)
break
except WechatSogouVcodeOcrException as e:
if i == self.captcha_break_times - 1:
raise WechatSogouVcodeOcrException(e)
if '请输入验证码' in resp.text:
headers = self.__set_cookie(referer=referer)
headers['User-Agent'] = 'Mozilla/5.0 (Windows NT 6.1; WOW64)'
resp = self.__get(url, session, headers)
resp.encoding = 'utf-8'
return resp