def clear_cart(self):
"""清空购物车
包括两个请求：
1.选中购物车中所有的商品
2.批量删除
:return: 清空购物车结果 True/False
"""
# 1.select all items 2.batch remove items
select_url = 'https://cart.jd.com/selectAllItem.action'
remove_url = 'https://cart.jd.com/batchRemoveSkusFromCart.action'
data = {
't': 0,
'outSkus': '',
'random': random.random(),
}
try:
select_resp = self.sess.post(url=select_url, data=data)
remove_resp = self.sess.post(url=remove_url, data=data)
if (not response_status(select_resp)) or (
not response_status(remove_resp)):
print(get_current_time(), '购物车清空失败')
return False
print(get_current_time(), '购物车清空成功')
return True
except Exception as e:
print(get_current_time(), e)
return False
def _get_qrcode(self):
"""
缓存并展示登录二维码
:return:
"""
url = 'https://passport.suning.com/ids/qrLoginUuidGenerate.htm?image=true&yys=%s' % str(
time.time() * 1000)[:-5]
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://passport.suning.com/ids/login',
}
logger.info('登录二维码地址: %s' % url)
resp = self.session.get(url=url, headers=headers)
if not response_status(resp):
logger.info('获取二维码失败')
return False
save_image(resp, self.qrcode_img_file)
logger.info('二维码获取成功，请打开苏宁APP扫描')
open_image(self.qrcode_img_file)
if global_config.getRaw('messenger', 'email_enable') == 'true':
email.send('二维码获取成功，请打开京东APP扫描', "<img src='cid:qr_code.png'>",
[email.mail_user], 'qr_code.png')
return True
def _get_qrcode(self):
"""
缓存并展示登录二维码
:return:
"""
url = 'https://qr.m.jd.com/show'
payload = {
'appid': 133,
'size': 147,
't': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://passport.jd.com/new/login.aspx',
}
resp = self.session.get(url=url, headers=headers, params=payload)
if not response_status(resp):
logger.info('获取二维码失败')
return False
img_file_name = '{}{}'.format('./cookies/', self.qrcode_img_file)
save_image(resp, img_file_name)
logger.info('二维码获取成功，请打开京东APP扫描')
open_image(img_file_name)
if global_config.getRaw('messenger', 'email_enable') == 'true':
receivers = global_config.getRaw('messenger', 'email_receiver')
email.send('二维码获取成功，请打开京东APP扫描',
"<img src='cid:./cookies/qr_code.png'>",
receivers.split(','), img_file_name)
return True
def get_cart_detail(self):
"""获取购物车商品详情
:return:
"""
url = 'https://cart.jd.com/cart.action'
cart_detail_format = '商品名称:{0}----单价:{1}----数量:{2}----总价:{3}'
try:
resp = self.sess.get(url=url)
if not response_status(resp):
print(get_current_time(), '获取购物车信息失败')
return
soup = BeautifulSoup(resp.text, "html.parser")
print('************************购物车商品详情************************')
for item in soup.select('div.item-form'):
name = get_tag_value(item.select('div.p-name a'))
price = get_tag_value(item.select('div.p-price strong'))
quantity = get_tag_value(
item.select('div.quantity-form input'), 'value')
total_price = get_tag_value(item.select('div.p-sum strong'))
print(
cart_detail_format.format(name, price, quantity,
total_price))
except Exception as e:
print(get_current_time(), e)
def _get_qrcode(self):
"""
缓存并展示登录二维码
:return:
"""
url = 'https://qr.m.jd.com/show'
payload = {
'appid': 133,
'size': 300,
't': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://passport.jd.com/new/login.aspx',
}
resp = self.session.get(url=url, headers=headers, params=payload)
if not response_status(resp):
logger.info('获取二维码失败')
return False
save_image(resp, self.qrcode_img_file)
logger.info('二维码获取成功，请打开京东APP扫描')
open_image(add_bg_for_qr(self.qrcode_img_file))
if global_config.getRaw('messenger', 'email_enable') == 'true':
email.send('二维码获取成功，请打开京东APP扫描', "<img src='cid:qr_code.png'>",
[email.mail_user], 'qr_code.png')
return True
def _get_qrcode_ticket(self):
"""
通过 token 获取票据
:return:
"""
# 'https://passport.suning.com/ids/qrLoginStateProbe?callback=jQuery17207490926503827557_1609730767292'
url = 'https://passport.suning.com/ids/qrLoginStateProbe?callback=jQuery{}_{}'.format(
random.randint(1000000, 9999999), int(round(time.time() * 1000)))
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://order.suning.com/order/orderList.do',
'X-Requested-With': 'XMLHttpRequest'
}
data = {
'uuid': self.session.cookies.get_dict().get('ids_qr_uuid'),
'service': '',
'terminal': 'PC'
}
resp = self.session.post(url=url, data=data, headers=headers)
if not response_status(resp):
logger.error('获取二维码扫描结果异常')
return False
if 'Set-Cookie' in resp.headers:
if resp.headers['Set-Cookie'].find('nick') > -1:
logger.info('已完成手机客户端确认')
return True
logger.info('Result: %s' % resp.text)
return False
def _get_QRcode_ticket(self):
url = 'https://qr.m.jd.com/check'
payload = {
'appid': '133',
'callback': 'jQuery{}'.format(random.randint(1000000, 9999999)),
'token': self.sess.cookies.get('wlfstk_smdl'),
'_': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': USER_AGENT,
'Referer': 'https://passport.jd.com/new/login.aspx',
}
resp = self.sess.get(url=url, headers=headers, params=payload)
if not response_status(resp):
print(get_current_time(), '获取二维码扫描结果出错')
return False
resp_json = parse_json(resp.text)
if resp_json['code'] != 200:
print(
get_current_time(),
'Code: {0}, Message: {1}'.format(resp_json['code'],
resp_json['msg']))
return None
else:
print(get_current_time(), '已完成手机客户端确认')
return resp_json['ticket']
def make_reserve(self):
"""商品预约"""
logger.info('商品名称:{}'.format(self.get_sku_title()))
url = 'https://yushou.jd.com/youshouinfo.action?'
payload = {
'callback': 'fetchJSON',
'sku': self.sku_id,
'_': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': self.user_agent,
'Referer': 'https://item.jd.com/{}.html'.format(self.sku_id),
}
resp = self.session.get(url=url, params=payload, headers=headers)
resp_json = parse_json(resp.text)
reserve_url = resp_json.get('url')
# self.timers.start()
while True:
try:
res = self.session.get(url='https:' + reserve_url)
if response_status(res):
logger.info('预约成功，已获得抢购资格 / 您已成功预约过了，无需重复预约')
if global_config.getRaw('messenger', 'enable') == 'true':
success_message = "预约成功，已获得抢购资格 / 您已成功预约过了，无需重复预约"
send_wechat(success_message)
break
except Exception as e:
logger.error('预约失败正在重试...')
def _get_qrcode_ticket(self):
"""
通过 token 获取票据
:return:
"""
url = 'https://qr.m.jd.com/check'
payload = {
'appid': '133',
'callback': 'jQuery{}'.format(random.randint(1000000, 9999999)),
'token': self.session.cookies.get('wlfstk_smdl'),
'_': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://passport.jd.com/new/login.aspx',
}
resp = self.session.get(url=url, headers=headers, params=payload)
if not response_status(resp):
logger.error('获取二维码扫描结果异常')
return False
resp_json = parse_json(resp.text)
if resp_json['code'] != 200:
logger.info('Code: %s, Message: %s', resp_json['code'], resp_json['msg'])
return None
else:
logger.info('已完成手机客户端确认')
return resp_json['ticket']
def page_detail_captcha(session, isId):
url = 'https://captcha.jd.com/verify/image'
acid = '{}_{}'.format(random.random(), random.random())
payload = {
'acid': acid,
'srcid': 'trackWeb',
'is': isId,
}
headers = {
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/531.36",
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3",
"Referer": "https://trade.jd.com/shopping/order/getOrderInfo.action",
"Connection": "keep-alive",
'Host': 'captcha.jd.com',
}
try:
resp = session.get(url=url, params=payload, headers=headers)
if not response_status(resp):
logger.error('获取订单验证码失败')
return ''
logger.info('解析验证码开始')
image = Image.open(BytesIO(resp.content))
image.save('captcha.jpg')
result = analysis_captcha(resp.content)
if not result:
logger.error('解析订单验证码失败')
return ''
global submit_captcha_text, submit_captcha_rid
submit_captcha_text = result
submit_captcha_rid = acid
return result
except Exception as e:
logger.error('订单验证码获取异常：%s', e)
return ''
def _get_qrcode(self):
"""
缓存并展示登录二维码
:return:
"""
url = 'https://qr.m.jd.com/show'
payload = {
'appid': 133,
'size': 147,
't': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://passport.jd.com/new/login.aspx',
}
resp = self.session.get(url=url, headers=headers, params=payload)
if not response_status(resp):
logger.info('获取二维码失败')
return False
save_image(resp, self.qrcode_img_file)
logger.info('二维码获取成功，请打开京东APP扫描')
open_image(self.qrcode_img_file)
return True
def login_by_username(self):
if self.is_login:
print(get_current_time(), '登录成功')
return True
username = input('账号:')
password = input('密码:')
if (not username) or (not password):
print(get_current_time(), '用户名或密码不能为空')
return False
self.username = username
data = self._get_login_data()
uuid = data['uuid']
auth_code = ''
if self._need_auth_code(username):
print(get_current_time(), '本次登录需要验证码')
auth_code = self._get_auth_code(uuid)
else:
print(get_current_time(), '本次登录不需要验证码')
login_url = "https://passport.jd.com/uc/loginService"
payload = {
'uuid': uuid,
'version': 2015,
'r': random.random(),
}
data['authcode'] = auth_code
data['loginname'] = username
data['nloginpwd'] = encrypt_pwd(password)
headers = {
'User-Agent': USER_AGENT,
'Origin': 'https://passport.jd.com',
}
resp = self.sess.post(url=login_url,
data=data,
headers=headers,
params=payload)
if not response_status(resp):
print(get_current_time(), '登录失败')
return False
if not self._get_login_result(resp):
return False
# login success
print(get_current_time(), '登录成功')
self.nick_name = self.get_user_info()
self._save_cookies()
self.is_login = True
return True
def analysis_captcha(session, captchaUrl, pic):
for i in range(1, 10):
try:
url = captchaUrl
resp = session.post(url, pic)
if not response_status(resp):
logger.error('解析验证码失败')
continue
logger.info('解析验证码[%s]', resp.text)
return resp.text
except Exception as e:
print(traceback.format_exc())
continue
return ''
def get_checkout_page_detail(self):
"""访问订单结算页面
该方法会打印出订单结算页面的详细信息：商品名称、价格、数量、库存状态等。
如果只是想下单商品，可以不调用该方法。
:return:
"""
url = 'http://trade.jd.com/shopping/order/getOrderInfo.action'
# url = 'https://cart.jd.com/gotoOrder.action'
payload = {
'rid': str(int(time.time() * 1000)),
}
try:
resp = self.sess.get(url=url, params=payload)
if not response_status(resp):
print(get_current_time(), '获取订单结算页信息失败')
return
soup = BeautifulSoup(resp.text, "html.parser")
self.risk_control = get_tag_value(soup.select('input#riskControl'),
'value')
print('************************订单结算页详情************************')
items = soup.select('div.goods-list div.goods-items')[1:]
checkout_item_detail = '商品名称:{0}----单价:{1}----数量:{2}----库存状态:{3}'
for item in items:
name = get_tag_value(item.select('div.p-name a'))
div_tag = item.select('div.p-price')[0]
price = get_tag_value(div_tag.select('strong.jd-price'))[
2:] # remove '￥ ' from the begin of price
quantity = get_tag_value(div_tag.select('span.p-num'))[
1:] # remove 'x' from the begin of quantity
state = get_tag_value(
div_tag.select('span.p-state')) # in stock or out of stock
print(checkout_item_detail.format(name, price, quantity,
state))
sum_price = soup.find('span', id='sumPayPriceId').text[
1:] # remove '￥' from the begin of sum price
address = soup.find('span', id='sendAddr').text[
5:] # remove '收件人:' from the begin of receiver
receiver = soup.find('span', id='sendMobile').text[
4:] # remove '寄送至： ' from the begin of address
print('应付总额:{0}'.format(sum_price))
print('收货地址:{0}----收件人:{1}'.format(address, receiver))
except Exception as e:
print(get_current_time(), e)
def _validate_QRcode_ticket(self, ticket):
url = 'https://passport.jd.com/uc/qrCodeTicketValidation'
headers = {
'User-Agent': USER_AGENT,
'Referer': 'https://passport.jd.com/uc/login?ltype=logout',
}
resp = self.sess.get(url=url, headers=headers, params={'t': ticket})
if not response_status(resp):
return False
resp_json = json.loads(resp.text)
if resp_json['returnCode'] == 0:
return True
else:
print(get_current_time(), resp_json)
return False
def _need_auth_code(self, username):
url = 'https://passport.jd.com/uc/showAuthCode'
data = {
'loginName': username,
}
payload = {
'version': 2015,
'r': random.random(),
}
resp = self.sess.post(url,
params=payload,
data=data,
headers=self.headers)
if not response_status(resp):
print('获取是否需要验证码失败')
return False
resp_json = json.loads(resp.text[1:-1]) # ({"verifycode":true})
return resp_json['verifycode']
def _get_QRcode(self):
url = 'https://qr.m.jd.com/show'
payload = {
'appid': 133,
'size': 147,
't': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': USER_AGENT,
'Referer': 'https://passport.jd.com/new/login.aspx',
}
resp = self.sess.get(url=url, headers=headers, params=payload)
if not response_status(resp):
print(get_current_time(), '获取二维码失败')
return False
QRCode_file = 'QRcode.png'
save_image(resp, QRCode_file)
print(get_current_time(), '二维码获取成功，请打开京东APP扫描')
open_image(QRCode_file)
return True
def _validate_qrcode_ticket(self, ticket):
"""
通过已获取的票据进行校验
:param ticket: 已获取的票据
:return:
"""
url = 'https://passport.jd.com/uc/qrCodeTicketValidation'
headers = {
'User-Agent': self.spider_session.get_user_agent(),
'Referer': 'https://passport.jd.com/uc/login?ltype=logout',
}
resp = self.session.get(url=url, headers=headers, params={'t': ticket})
if not response_status(resp):
return False
resp_json = json.loads(resp.text)
if resp_json['returnCode'] == 0:
return True
else:
logger.info(resp_json)
return False
def _get_auth_code(self, uuid):
image_file = os.path.join(os.getcwd(), 'jd_authcode.jpg')
url = 'https://authcode.jd.com/verify/image'
payload = {
'a': 1,
'acid': uuid,
'uid': uuid,
'yys': str(int(time.time() * 1000)),
}
headers = {
'User-Agent': USER_AGENT,
'Referer': 'https://passport.jd.com/uc/login',
}
resp = self.sess.get(url, params=payload, headers=headers)
if not response_status(resp):
print('获取验证码失败')
return ''
save_image(resp, image_file)
open_image(image_file)
return input('验证码:')
def get_checkout_page_detail(session):
"""获取订单结算页面信息
该方法会返回订单结算页面的详细信息：商品名称、价格、数量、库存状态等。
:return: 结算信息 dict
"""
url = 'http://trade.jd.com/shopping/order/getOrderInfo.action'
# url = 'https://cart.jd.com/gotoOrder.action'
payload = {
'rid': str(int(time.time() * 1000)),
}
headers = {
"User-Agent":
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/531.36",
"Accept":
"text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3",
"Referer": "https://cart.jd.com/cart.action",
"Connection": "keep-alive",
'Host': 'trade.jd.com',
}
try:
resp = session.get(url=url, params=payload, headers=headers)
if not response_status(resp):
logger.error('获取订单结算页信息失败')
return ''
if '刷新太频繁了' in resp.text:
return '刷新太频繁了'
soup = BeautifulSoup(resp.text, "html.parser")
risk_control = get_tag_value(soup.select('input#riskControl'), 'value')
showCheckCode = get_tag_value(soup.select('input#showCheckCode'),
'value')
if showCheckCode in ['false', 'False', 'FALSE']:
pass
else:
if showCheckCode == 'true':
logger.info('提交订单需要验证码')
global is_Submit_captcha, encryptClientInfo
encryptClientInfo = get_tag_value(
soup.select('input#encryptClientInfo'), 'value')
is_Submit_captcha = True
order_detail = {
'address':
soup.find('span',
id='sendAddr').text[5:], # remove '寄送至： ' from the begin
'receiver': soup.find(
'span',
id='sendMobile').text[4:], # remove '收件人:' from the begin
'total_price': soup.find(
'span',
id='sumPayPriceId').text[1:], # remove '￥' from the begin
'items': []
}
logger.info("下单信息：%s", order_detail)
return risk_control
except requests.exceptions.RequestException as e:
logger.error('订单结算页面获取异常：%s' % e)
except Exception as e:
logger.error('下单页面数据解析异常：%s', e)
return ''
def get_order_info(self, unpaid=True):
"""查询订单信息
:param unpaid: 只显示未付款订单，可选参数，默认为True
:return:
"""
if not self.is_login:
print(get_current_time(), '请先登录再查询订单！')
return
url = 'https://order.jd.com/center/list.action'
payload = {
'search': 0,
'd': 1,
's': 4096,
} # Orders for nearly three months
headers = {
'User-Agent': USER_AGENT,
'Referer': 'https://passport.jd.com/uc/login?ltype=logout',
}
try:
resp = self.sess.get(url=url, params=payload, headers=headers)
if not response_status(resp):
print(get_current_time(), '获取订单页信息失败')
return
soup = BeautifulSoup(resp.text, "html.parser")
print('************************订单列表页查询************************')
order_table = soup.find('table', {'class': 'order-tb'})
table_bodies = order_table.select('tbody')
exist_order = False
for table_body in table_bodies:
# get order status
order_status = get_tag_value(
table_body.select('span.order-status')).replace(
"订单状态：", "")
# check if order is waiting for payment
# wait_payment = bool(table_body.select('a.btn-pay'))
wait_payment = "等待付款" in order_status
# only show unpaid orders if unpaid=True
if unpaid and (not wait_payment):
continue
exist_order = True
# get order_time, order_id
tr_th = table_body.select('tr.tr-th')[0]
order_time = get_tag_value(tr_th.select('span.dealtime'))
order_id = get_tag_value(tr_th.select('span.number a'))
# get sum_price, pay_method
sum_price = ''
pay_method = ''
amount_div = table_body.find('div', {'class': 'amount'})
if amount_div:
spans = amount_div.select('span')
pay_method = get_tag_value(spans, index=1)
# if the order is waiting for payment, the price after the discount is shown.
sum_price = get_tag_value(amount_div.select('strong'), index=1)[1:] if wait_payment \
else get_tag_value(spans, index=0)[4:]
# get name and quantity of items in order
items_dict = dict(
) # {'item_id_1': quantity_1, 'item_id_2': quantity_2, ...}
tr_bds = table_body.select('tr.tr-bd')
for tr_bd in tr_bds:
item = tr_bd.find('div', {'class': 'goods-item'})
if not item:
break
item_id = item.get('class')[1][2:]
quantity = get_tag_value(
tr_bd.select('div.goods-number'))[1:]
items_dict[item_id] = quantity
order_info_format = '下单时间:{0}----订单号:{1}----商品列表:{2}----订单状态:{3}----总金额:{4}元----付款方式:{5}'
print(
order_info_format.format(order_time, order_id,
parse_items_dict(items_dict),
order_status, sum_price,
pay_method))
if not exist_order:
print(get_current_time(), '订单查询为空')
except Exception as e:
print(get_current_time(), e)
