def user_recommend(self, request, context):
# 接收web后端发送过来的请求
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
# 调用系统推荐方法获取推荐文章数据
# 返回推荐的文章响应对象
article_response = reco_pb2.ArticleResponse()
# 设置曝光数据
article_response.exposure = "exposure message"
# 设置时间戳
article_response.time_stamp = round(time.time() * 1000)
article_list = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i + 1
# 埋点数据
article.track.click = "click action {}".format(i + 1)
article.track.collect = "collect action {}".format(i + 1)
article.track.share = "share action {}".format(i + 1)
article.track.read = "read action {}".format(i + 1)
article_list.append(article)
article_response.recommends.extend(article_list)
return article_response
def user_recommend(self, request, context):
"""
伪造推荐系统
:param request: rpc调用的请求数据对象 UserRequest对象
:param context: 给予程序上下文环境，主要用于返回异常
context.set_code(grpc.StatusCode.UNIMPLEMENTED)
context.set_details('Method not implemented!')
:return:
"""
# 获取rpc调用的请求参数
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
# 伪造推荐系统返回对象ArticleResponse
response = reco_pb2.ArticleResponse()
response.exposure = 'exposure param'
response.time_stamp = round(time.time() * 1000)
articles_list = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i + 1
article.track.click = 'click param '
article.track.collect = 'collect param '
article.track.share = 'share param '
article.track.read = 'read param '
articles_list.append(article)
# response.recommends.append(Article())
# 注意 在grpc中 不使用append方法 使用extend
response.recommends.extend(articles_list)
return response
def user_recommend(self, request, context):
# 1.解析请求对象中的参数 request
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
print('接收到了rpc请求的参数')
# 2 .构造响应对象,并且返回
response = reco_pb2.ArticleResponse() #实例化
response.expousre = '曝光埋点数据'
response.time_stamp = round(time.time() *
1000) #round() python内置函数,四舍五入取整
recommends_list = []
for i in range(article_num):
article = reco_pb2.Article() #实例化
# article.track 不是间接关系,而是直接关系,所以track里面有的直接使用
article.article_id = i + 1
article.track.click = 'click'
article.track.collect = 'collect'
article.track.share = 'share'
article.track.read = 'read'
recommends_list.append(article)
print(recommends_list)
response.recommends.extend(recommends_list) #因为是列表,所以用extend添加一个构造列表对象
return response
def user_recommend(self, request, context):
"""
在接口中定义的用户推荐方法
:param request: 调用时的请求参数对象 UserRequest
:param context: 通过此对象可以设置调用返回的异常信息
:return:
"""
# 获取调用的参数
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
# 决定调用返回数据
resp = reco_pb2.ArticleResponse()
resp.exposure = 'exposure param'
resp.time_stamp = round(time.time() * 1000)
_recommends = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i+1
article.track.click = 'click param'
article.track.collect = 'collect param'
article.track.share = 'share param'
article.track.read = 'read param'
_recommends.append(article)
# 注意 对于列表类型的赋值使用extend
resp.recommends.extend(_recommends)
return resp
def user_recommend(self, request, context):
# request是调用的请求数据对象
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
response = reco_pb2.ArticleResponse()
# 手动构造推荐结果，后续对接真实推荐代码
response.exposure = 'exposure param'
response.time_stamp = round(time.time() * 1000)
recommends = []
for i in range(article_num):
article = reco_pb2.Article()
article.track.click = 'click param {}'.format(i + 1)
article.track.collect = 'collect param {}'.format(i + 1)
article.track.share = 'share param {}'.format(i + 1)
article.track.read = 'read param {}'.format(i + 1)
article.article_id = i + 1
recommends.append(article)
response.recommends.extend(recommends)
# 最终要返回一个调用结果
return response
def user_recommend(self, request, context):
# 1.解析请求对象中的参数
user_id = request.user_id # 用户id str类型
channel_id = request.channel_id # 频道id int类型
article_num = request.article_num # 推荐的文章数量 向推荐系统索要推荐文章的数量 int类型
time_stamp = request.time_stamp # 推荐的时间戳 单位秒 int类型
"""这里根据请求参数做了一些操作"""
# 2.构造相应对象,并返回
response = reco_pb2.ArticleResponse()
response.expousre = 'sssss'
response.time_stamp = round(time.time()*1000) # round() py内置函数 四舍五入取整
recommends_list = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i+1
article.track.click = 'click'
article.track.collect = 'collect'
article.track.share = 'share'
article.track.read = 'read'
recommends_list.append(article)
response.recommends.extend(recommends_list)
return response
def user_recommend(self, request, context):
"""
给用户推荐文章具体业务逻辑实现
:param request: UserRequest类的请求参数对象
:param context: 错误信息
:return:
"""
# 1.获取web后端的请求参数
user_id = request.user_id
channel_id = request.channel_id
# 推荐文章的数量
article_num = request.article_num
time_stamp = request.time_stamp
# 2.参数校验
# 3.逻辑处理
# TODO: 后续补充推荐系统的业务逻辑
# 伪推荐 造数据
# 4.返回值处理
article_response = reco_pb2.ArticleResponse()
# 曝光数据
article_response.exposure = "exposure message"
article_response.time_stamp = round(time.time() * 1000)
article_list = []
for i in range(article_num): # 0 1 2 3
# 创建文章对象
article = reco_pb2.Article()
# 给文章id赋值
article.article_id = i + 1
# 用户行为埋点数据
article.track.click = "click action {}".format(i+1)
article.track.collect = "collect action {}".format(i+1)
article.track.share = "share action {}".format(i+1)
article.track.read = "read action {}".format(i+1)
article_list.append(article)
# extend 在列表尾部补充元素
article_response.recommends.extend(article_list)
return article_response
def user_recommend(self, request, context):
"""
理应由推荐系统编写的 推荐业务逻辑
为了跑通整个rpc通讯流程，暂时由web来完成 伪推荐
:param request: rpc的请求对象 在当前的rpc 首页新闻推荐中，按照proto文件指明，应该是UserRequset类对象
:param context: 在当前被执行的函数中 如果出现问题，可以通过context设置 返回给调用一方的错误信息
context.set_code(grpc.StatusCode.UNIMPLEMENTED)
context.set_details('Method not implemented!')
:return:
"""
# 获取调用的请求参数
user_id = request.user_id # 用户id
channel_id = request.channel_id # 频道id
article_num = request.article_num # 文章数量
time_stamp = request.time_stamp # 时间戳
# 构建文章id和伪造埋点参数 返回
resp = reco_pb2.ArticleResponse()#构建响应对象
resp.exposure = 'exposure param'#曝光参数
resp.time_stamp = round(time.time() * 1000)#时间戳，毫秒整数
# 在grpc中 repeated 列表参数的构建 需要使用extend，不用append
#列表参数有埋点数
# resp.recommends.append()
articles = []
for i in range(article_num):
#Article对象
article = reco_pb2.Article()
article.article_id = i + 1
article.track.click = 'click param {}'.format(i + 1)
article.track.collect = 'collect param {}'.format(i + 1)
article.track.share = 'share param {}'.format(i + 1)
article.track.read = 'read param {}'.format(i + 1)
articles.append(article)
#扩展给列表属性 extend
resp.recommends.extend(articles)
return resp
def user_recommends(self, request, context):
# 用户推荐
# 获取请求信息
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
# 此处代码为伪推荐，后续系统推荐课程中，可以补全此处代码
resp = reco_pb2.ArticleResponse()
resp.exposure = 'exposure data'
resp.time_stamp = round(time.time() * 1000)
content = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i
article.track.click = 'click data {}'.format(i + 1)
article.track.collect = 'collect data {}'.format(i + 1)
article.track.share = 'share data {}'.format(i + 1)
article.track.read = 'read data {}'.format(i + 1)
content.append(article)
resp.recommends.extend(content)
# 返回响应
return resp
def user_recommend(self, request, context):
"""
用户文章推荐 实际是由推荐系统编写
:param request: rpc调用的请求数据对象 UserRequest对象
:param context: 如果被调用的这个方法出现了异常，可以通过context提供的方法来告知rpc调用的一方错误信息
context.set_code(grpc.StatusCode.UNIMPLEMENTED)
context.set_details('Method not implemented!')
:return:
"""
# 获取rpc调用的请求参数
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
# 为了看到效果 先伪推荐
# 构建调用返回值对象
response = reco_pb2.ArticleResponse()
response.exposure = 'exposure param'
response.time_stamp = round(time.time() * 1000)
articles_list = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i + 1
article.track.click = 'click param '
article.track.collect = 'collect param '
article.track.share = 'share param '
article.track.read = 'read param '
articles_list.append(article)
# response.recommends.append(Article())
# 注意 在grpc中 不使用append方法 使用extend
response.recommends.extend(articles_list)
return response
def user_recommend(self, request, context):
# 接收请求，request不是Flask请求上下文对象
user_id = request.user_id
channel_id = request.channel_id
article_num = request.article_num
time_stamp = request.time_stamp
# 返回响应,后续需要对接推荐系统，获取推荐的文章列表，现在模拟推荐
# 实例化文章对象
resp = reco_pb2.ArticleResponse()
resp.exposure = 'exposure params'
resp.time_stamp = round(time.time() * 1000)
# 默认返回的文章使用article_num
data_list = []
for i in range(article_num):
article = reco_pb2.Article()
article.article_id = i + 1
article.track.click = 'click params {}'.format(i + 1)
article.track.collect = 'collect params {}'.format(i + 1)
article.track.share = 'share params {}'.format(i + 1)
article.track.read = 'read params {}'.format(i + 1)
data_list.append(article)
resp.recommends.extend(data_list)
return resp