def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
activity = "" # 行为
content = "" # 内容
key_words = [] # 内容关键词（自然语言做的分词）
retweet = 0 # 转推数
like = 0 # 喜欢数
reply_num = 0 # 回复数
row = data.iloc[i] # 数据元组
# print("type:", type(row), "row: ", row)
for j in data.columns:
j = str(j)
# if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
# self.nan_list.append(row)
# continue
element = str(data.iloc[i][j])
if j == "用户ID":
user_id = element
print(j, ":", user_id)
if j == "服务ID":
service_id = element
print(j, ":", service_id)
if j == "时间":
if str(data.iloc[i][j]) == "nan":
self.nan_list.append(row)
continue
if self.match_timestamp(element):
print("10位数字的时间戳")
time_stamp = int(self.match_timestamp(element))
# 将时间戳timestamp转换成格式化的字符串Datetime
l_time = time.localtime(time_stamp)
date_time = time.strftime("%Y-%m-%d %H:%M:%S", l_time)
print("date_time:", date_time, ", timestamp:",
time_stamp)
else:
if "上午" in element:
element = element.replace("上午", "")
# 格式化的字符串转换成Datetime
print("时间：", element)
dt = datetime.datetime.strptime(
element, "%H:%M - %Y年%m月%d日")
date_time = str(dt)
print("时间-：", date_time)
# 转化成时间戳
timeArray = time.strptime(date_time,
"%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if "下午" in element:
element = element.replace("下午", "")
# 格式化的字符串转换成Datetime
dt = datetime.datetime.strptime(
element, "%H:%M - %Y年%m月%d日")
print("dt:", dt)
# 加上12小时
aDay = timedelta(days=0.5)
now = dt + aDay
print("new now:", now)
element = str(now)
date_time = element
# 再变成时间戳
timeArray = time.strptime(element,
"%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if j == "行为":
activity = element
print(j, ":", activity)
if j == "内容":
# 判断是否有空值
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
element = element.replace("Twitter", "")
element = element.replace("的", "")
element = element.replace(' “@', "")
content = element
print(j, ":", content)
if activity is "":
if user_id in content:
activity = "Post"
print("行为：", activity)
else:
activity = "Reply"
print("行为：", activity)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
if j == "转推":
retweet = element
print(j, ":", retweet)
if j == "喜欢":
like = element
print(j, ":", like)
if j == "回复":
reply_num = element
print(j, ":", reply_num)
# end if
_id = self.get_next_counter()
print("_id:", _id)
# 输出每一行的input_list
insert_text = {
"uid": self.collection_name,
"用户ID": user_id,
"服务ID": service_id,
"时间": date_time,
"timestamp": time_stamp,
"activity": activity,
"内容": content,
"keywords": key_words,
"转推": retweet,
"喜欢": like,
"回复": reply_num,
"_id": _id
}
print("row_input_list:", insert_text)
# 插入数据库
if self.flag_insert == "1":
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))
def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
content = "" # 内容
key_words = [] # 内容关键词（自然语言做的分词）
row = data.iloc[i] # 数据元组
# print("type:", type(row), "row: ", row)
for j in data.columns:
j = str(j)
# if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
# self.nan_list.append(row)
# continue
element = str(data.iloc[i][j])
if j == "用户ID":
user_id = element
print(j, ":", user_id)
if j == "服务ID":
service_id = element
print(j, ":", service_id)
if j == "时间":
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
print(j, ":", element)
# 格式化时间
date_time = self.format_datetime(element)
print("格式化时间date_time:", date_time)
# 转化成时间戳
timeArray = time.strptime(date_time, "%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if j == "行为":
activity = element
print(j, ":", activity)
if j == "内容" or j == "title":
# 判断是否有空值
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
content = element
print(j, ":", content)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
# end if
# 输出每一行的input_list
_id = self.get_next_counter()
insert_text = {"uid": self.collection_name, "用户ID": user_id,
"服务ID": service_id, "时间": date_time,
"timestamp": time_stamp, "内容": content,
"keywords": key_words, "_id": _id
}
print("row_input_list:", insert_text)
# 插入数据库
if self.flag_insert == "1":
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))
def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
activity = "" # 行为
content = "" # 内容
key_words = [] # 内容关键词（自然语言做的分词）
title_text = "" # 标题
star_badge = "" # 奖励
row = data.iloc[i] # 数据元组
# print("type:", type(row), "row: ", row)
for j in data.columns:
j = str(j)
# if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
# self.nan_list.append(row)
# continue
element = str(data.iloc[i][j])
if j == "用户ID":
user_id = element
print(j, ":", user_id)
if j == "服务ID":
service_id = element
print(j, ":", service_id)
if j == "时间":
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
element = element.replace("Z", "")
element = element.replace("T", " ")
date_time = element
print(j, ":", date_time)
# 将格式化时间转换成时间戳10位
# 1中间过程，一般都需要将字符串转化为时间数组
try:
timeArray = time.strptime(element, "%Y-%m-%d %H:%M:%S")
except ValueError:
element = element.replace(" GM", "")
element = element.replace("Mon, ", "")
element = element.replace("Tue, ", "")
element = element.replace("Wed, ", "")
element = element.replace("Thu, ", "")
element = element.replace("Fri, ", "")
element = element.replace("Sat, ", "")
element = element.replace("Sun, ", "")
timeArray = time.strptime(element, "%Y-%m-%d")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp, " type:", type(timestamp))
time_stamp = timestamp
if j == "行为":
activity = element
print(j, ":", activity)
if j == "内容":
content = element
print(j, ":", content)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
if j == "title":
title_text = element
print(j, ":", title_text)
temp_keywords = dpt4.main(title_text) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
if j == "标记":
star_badge = element
print(j, ":", star_badge)
if j == "repository":
content = element
print(j, ":", content)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
# end if
# 调用Switch结构
# 输出每一行的input_list
_id = self.get_next_counter()
insert_text = {
"uid": self.collection_name,
"用户ID": user_id,
"服务ID": service_id,
"时间": date_time,
"timestamp": time_stamp,
"activity": activity,
"内容": content,
"keywords": key_words,
"_id": _id
}
print("row_input_list:", insert_text)
# 插入数据库
if self.flag_insert == "1":
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))
def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
activity = "" # 行为
content = "" # 内容
title = "" # 标题
key_words = [] # 内容关键词（自然语言做的分词）
row = data.iloc[i] # 数据元组
for j in data.columns:
j = str(j)
element = str(data.iloc[i][j])
if j == "用户ID":
user_id = element
print(j, ":", user_id)
if j == "服务ID":
service_id = element
print(j, ":", service_id)
if j == "时间":
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
element = self.match_date_time(element)
print(j, ":", element)
# 将格式化时间转换成时间戳10位
# 1中间过程，一般都需要将字符串转化为时间数组
timeArray = time.strptime(element, "%Y年%m月%d日")
# 2将"2011-09-28 10:00:00"转化为时间戳
time_stamp = int(time.mktime(timeArray))
# 将时间戳timestamp转换成格式化的字符串Datetime
l_time = time.localtime(time_stamp)
date_time = time.strftime("%Y-%m-%d %H:%M:%S", l_time)
print("date_time:", date_time, ", timestamp:", time_stamp)
if j == "行为":
activity = element
print(j, ":", activity)
if j == "内容":
content = element
print(j, ":", content)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words += temp_keywords
if j == "title" or j == "Title":
title = element
print(j, ":", title)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words += temp_keywords
# end if
# 调用Switch结构
# 输出每一行的input_list
_id = self.get_next_counter()
insert_text = {
"uid": self.collection_name,
"用户ID": user_id,
"服务ID": service_id,
"时间": date_time,
"timestamp": time_stamp,
"activity": activity,
"内容": content,
"keywords": key_words,
"_id": _id,
"title": title
}
print("row_input_list:", insert_text)
# 插入数据库
if self.flag_insert == "1":
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))
def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
title = "" # 标题
content = "" # 内容
key_words = [] # 内容关键词（自然语言做的分词）
retweet = 0 # 转推数
like = 0 # 喜欢数
reply_num = 0 # 回复数
row = data.iloc[i] # 数据元组
# print("type:", type(row), "row: ", row)
for j in data.columns:
j = str(j)
# if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
# self.nan_list.append(row)
# continue
element = str(data.iloc[i][j])
if j == "用户ID":
user_id = element
print(j, ":", user_id)
if j == "服务ID":
service_id = element
print(j, ":", service_id)
if j == "时间":
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
element = self.match_timestamp(element)
element = self.format_datetime(element)
print(j, ":", element)
if "am" in element:
element = element.replace("am", "")
# 格式化的字符串转换成Datetime
dt = datetime.datetime.strptime(
element, "%d %m %Y, %H:%M")
date_time = str(dt)
print("时间：", date_time)
# 转化成时间戳
timeArray = time.strptime(date_time,
"%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if "pm" in element:
element = element.replace("pm", "")
# 格式化的字符串转换成Datetime
dt = datetime.datetime.strptime(
element, "%d %m %Y, %H:%M")
print("dt:", dt)
# 加上12小时
aDay = timedelta(days=0.5)
now = dt + aDay
print("new now:", now)
element = str(now)
date_time = element
# 再变成时间戳
timeArray = time.strptime(element, "%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if j == "title":
title = element
print(j, ":", title)
temp_keywords = dpt4.main(title) # keywords是一个List结构
key_words += temp_keywords
print("keywords:", temp_keywords)
if j == "content_time":
# 判断是否有空值
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
list_ele = element.split("_")
content = list_ele[0]
date_time = list_ele[1]
print("内容：", content)
print("时间：", date_time)
timeArray = time.strptime(str(date_time),
"%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
temp_keywords = dpt4.main(content) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
# end if
# 调用Switch结构
# 输出每一行的input_list
insert_text = {
"uid": self.collection_name,
"用户ID": user_id,
"服务ID": service_id,
"时间": date_time,
"timestamp": time_stamp,
"内容": content,
"title": title,
"keywords": key_words
}
print("row_input_list:", insert_text)
# 插入数据库
if self.flag_insert == "1":
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))
def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
temp_time = "" # 临时时间
content = "" # 内容
key_words = [] # 内容关键词（自然语言做的分词）
row = data.iloc[i] # 数据元组
# print("type:", type(row), "row: ", row)
for j in data.columns:
j = str(j)
# if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
# self.nan_list.append(row)
# continue
element = str(data.iloc[i][j])
if j == "ATT":
con_list = element.split(",")
for item in range(len(con_list)):
print("item:", item, "element:", con_list[item])
if item == 0:
user_id = con_list[item]
if item == 1:
service_id = con_list[item]
if item == 2:
content = con_list[item]
temp_keywords = dpt4.main(
content) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
if item == 3:
temp_time = con_list[item]
temp_time = temp_time.replace("T", " ")
element = temp_time.replace("+00:00", "")
# 格式化的字符串转换成Datetime
dt = datetime.datetime.strptime(
element, "%Y-%m-%d %H:%M:%S")
date_time = str(dt)
print("时间：", date_time)
# 转化成时间戳
timeArray = time.strptime(date_time,
"%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if j == "时间":
if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
self.nan_list.append(row)
continue
print(j, ":", element)
if "上午" in element:
element = element.replace("上午", "")
# 格式化的字符串转换成Datetime
dt = datetime.datetime.strptime(
element, "%H:%M - %Y年%m月%d日")
date_time = str(dt)
print("时间：", date_time)
# 转化成时间戳
timeArray = time.strptime(date_time,
"%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
if "下午" in element:
element = element.replace("下午", "")
# 格式化的字符串转换成Datetime
dt = datetime.datetime.strptime(
element, "%H:%M - %Y年%m月%d日")
print("dt:", dt)
# 加上12小时
aDay = timedelta(days=0.5)
now = dt + aDay
print("new now:", now)
element = str(now)
# 再变成时间戳
timeArray = time.strptime(element, "%Y-%m-%d %H:%M:%S")
# 2将"2011-09-28 10:00:00"转化为时间戳
timestamp = int(time.mktime(timeArray))
print("timestamp:", timestamp)
time_stamp = timestamp
# end if
# 调用Switch结构
# 输出每一行的input_list
insert_text = {
"uid": self.collection_name,
"用户ID": user_id,
"服务ID": service_id,
"时间": date_time,
"timestamp": time_stamp,
"内容": content,
"keywords": key_words
}
print("row_input_list:", insert_text)
# 插入数据库
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))
def read_file(self):
data = pd.read_csv(self.file_path, encoding='utf-8') # 读取数据
print("data:")
# 逐个元素判断是否为空值,将空值行，放入一个队列中
for i in range(len(data)):
print("i:", i)
user_id = "" # 用户ID
service_id = "" # 服务ID
date_time = "" # 格式化时间
time_stamp = "" # 时间戳
activity = "" # 行为
content = "" # 内容
key_words = [] # 内容关键词（自然语言做的分词）
row = data.iloc[i] # 数据元组
# print("type:", type(row), "row: ", row)
for j in data.columns:
j = str(j)
# if str(data.iloc[i][j]) == "nan":
# print("空值", row, j, "element：", data.iloc[i][j])
# self.nan_list.append(row)
# continue
element = str(data.iloc[i][j])
if j == "用户ID":
user_id = element
print(j, ":", user_id)
if j == "服务ID":
service_id = element
print(j, ":", service_id)
if j == "timestamp":
if str(data.iloc[i][j]) == "nan":
self.nan_list.append(row)
continue
element = element.replace(".0", "")
print("element:", element, ", type:", type(element))
time_stamp = int(element)
# 将时间戳timestamp转换成格式化的字符串Datetime
l_time = time.localtime(time_stamp)
date_time = time.strftime("%Y-%m-%d %H:%M:%S", l_time)
print("date_time:", date_time, ", timestamp:", time_stamp)
if j == "内容":
element = element.replace("Idan Adar", "")
element = element.replace("的", "")
element = element.replace(' “@', "")
content = element
print(j, ":", content)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
if j == "title":
element = element.replace("Idan Adar", "")
element = element.replace("的", "")
element = element.replace(' “@', "")
content = element
print(j, ":", content)
temp_keywords = dpt4.main(element) # keywords是一个List结构
print("keywords:", temp_keywords)
key_words = key_words + temp_keywords
# end if
_id = self.get_next_counter()
print("_id:", _id)
# 输出每一行的input_list
insert_text = {
"uid": self.collection_name,
"用户ID": user_id,
"服务ID": service_id,
"时间": date_time,
"timestamp": time_stamp,
"内容": content,
"keywords": key_words,
"_id": _id
}
print("row_input_list:", insert_text)
# 插入数据库
if self.flag_insert == "1":
self.input_database(insert_text)
print()
# end for, 判断是不是有空值的元组
if self.nan_list:
for var_nan in self.nan_list:
print("NaN row:", var_nan)
print("空值的个数：", len(self.nan_list))