参考链接

https://blog.csdn.net/glDemo/article/details/46886001

https://blog.csdn.net/kxw1994/article/details/76465152

https://stackoverflow.com/questions/10154633/load-csv-data-into-mysql-in-python

https://blog.csdn.net/qq_35246620/article/details/78148505

https://www.mydatahack.com/how-to-bulk-load-data-into-mysql-with-python/

根据 mongo 结构在 mysql 中建表

查看 mongo 结构：

mongos> db.mins.find().next()
{
	"_id" : ObjectId("59ce1e1d6e6dc7768c7140dc"),
	"code" : "SH900955",
	"time" : ISODate("1999-07-26T09:59:00Z"),
	"open" : 0.462,
	"close" : 0.462,
	"low" : 0.462,
	"high" : 0.462,
	"volume" : 0,
	"amount" : 0
}

在mysql 中建表：

CREATE TABLE IF NOT EXISTS `inc_mins`(
   `_id` VARCHAR(100) NOT NULL,
   `code` VARCHAR(100) NOT NULL,
   `time` VARCHAR(100) NOT NULL,
   `open` FLOAT DEFAULT NULL,
   `low` FLOAT DEFAULT NULL,
   `high` FLOAT DEFAULT NULL,
   `close` FLOAT DEFAULT NULL,
   `volume` text,
   `amount` text,
   UNIQUE KEY `_id` ( `_id` )
)ENGINE=MyISAM DEFAULT CHARSET=utf8;

处理 mongo 增量数据到 csv 文件中

这一步在这篇中略过，假设处理后得到两个以日期命名的文件夹，里面针对每一只股票的 csv 增量数据。如图：

可以使用 ls -lht 来查看文件的大小，找一个合适大小的文件来进行测试。

使用 Python 程序生成

考虑经历前后两步：
（1） 拼接多个 csv 文件
（2） 将 csv 导入 mysql

拼接多个 csv 文件

import pandas as pd
import os
import sys

Folder_Path = r'/Users/furuiyang/codes/mins_daily_import_to_csv/exportdir/20190720'  # 要拼接的文件夹及其完整路径，注意不要包含中文
SaveFile_Path = r'/Users/furuiyang/codes/mins_daily_import_to_csv/savedir/20190720'  # 拼接后要保存的文件路径
SaveFile_Name = r'20190720.csv'  # 合并后要保存的文件名

# 修改当前工作目录
os.chdir(Folder_Path)
# 将该文件夹下的所有文件名存入一个列表
file_list = os.listdir()
# print(file_list)
# sys.exit(0)

# 读取第一个CSV文件并包含表头
df = pd.read_csv(os.path.join(Folder_Path, file_list[0]))
# print(df)
# sys.exit(0)

# 创建要保存的文件夹
os.makedirs(SaveFile_Path, exist_ok=True)

# 将读取的第一个CSV文件写入合并后的文件保存
save_file = os.path.join(SaveFile_Path, SaveFile_Name)
df.to_csv(save_file)
# sys.exit(0)

# 循环遍历列表中各个CSV文件名，并追加到合并后的文件
count = 0
try:
    for i in range(1, len(file_list)):
        print(os.path.join(Folder_Path, file_list[i]))
        df = pd.read_csv(os.path.join(Folder_Path, file_list[i]))
        print(df)
        print(df.shape[0])
        count += df.shape[0]

        print()
        print()
        try:
            df.to_csv(save_file, encoding="utf-8", index=False, header=False, mode='a+')
        except Exception as e:
            print(e)
            sys.exit(0)
except Exception:
    print("wrong.")
    pass

print(count)   # 2599680

将 csv 文件导入 MYSQL

逐条插入

import pymysql

conn = pymysql.Connect(host='localhost', port=3306, user='root', password='ruiyang', db='test01', charset='utf8')
cursor = conn.cursor()

# 建表
# sql1 = "create table coupon(id INT(4) NOT NULL auto_increment,\
# coupon_num VARCHAR(255) not null,primary key(id))"
# cursor.execute(sql1)

# 插入执行
# sql2 = "insert into inc_mins (_id,code,time,open,close,low,high,volume,amount) values {},{},{},{},{},{},{},{},{}"
sql2 = "insert into inc_mins (_id,code,time,open,close,low,high,volume,amount) values {};"
file = open('/Users/furuiyang/codes/mins_daily_import_to_csv/savedir/20190720/20190720_001.csv','r')
file.readline()

for i in file.readlines():
    data = i.strip().split(',')
    print(data)
    print(sql2.format(tuple(data)))
    cursor.execute(sql2.format(tuple(data)))

这样逐条插入的耗时有点不能忍：
插入 11w 左右的数据用了约 14 min

使用 load file 的模式

测试：
在 mysql 的终端执行 load data 导入语句：

load data infile '/Users/furuiyang/codes/mins_daily_import_to_csv/exportdir/20190720_10001718.csv' \
replace into table test01.inc_mins  \
fields terminated by ',' \
optionally enclosed by '"' \
lines terminated by '\n' \
ignore 1 lines(_id,code,time,open,low,high,close,volume,amount);

报错：

The MySQL server is running with the --secure-file-priv option so it cannot execute this statement

可能是涉及到相关的权限。在程序中执行简化的语句，代码如下：

import pymysql


def csv_to_mysql(load_sql, host, user, password):
    '''
    This function load a csv file to MySQL table according to
    the load_sql statement.
    '''
    try:
        con = pymysql.connect(host=host,
                              user=user,
                              password=password,
                              autocommit=True,
                              local_infile=1)
        print('Connected to DB: {}'.format(host))
        # Create cursor and execute Load SQL
        cursor = con.cursor()
        cursor.execute(load_sql)
        print('Succuessfully loaded the table from csv.')
        con.close()

    except Exception as e:
        print('Error: {}'.format(str(e)))
        sys.exit(1)


# Execution Example
load_sql = """LOAD DATA LOCAL INFILE '/Users/furuiyang/codes/mins_daily_import_to_csv/savedir/20190720/20190720_001.csv' INTO TABLE test01.inc_mins \
FIELDS TERMINATED BY ',' \
ENCLOSED BY '"' \
IGNORE 1 LINES;"""

print(load_sql)


"""
load data infile '/Users/furuiyang/codes/mins_daily_import_to_csv/exportdir/20190720_10001718.csv' \
replace into table test01.inc_mins  \
fields terminated by ',' \
optionally enclosed by '"' \
lines terminated by '\n' \
ignore 1 lines(_id,code,time,open,low,high,close,volume,amount);
"""


host = 'localhost'
user = 'root'
password = 'ruiyang'
csv_to_mysql(load_sql, host, user, password)

load data 方式的耗时显然变快很多，

在我们建表并且设置了唯一索引的情况下，会给出插入重复警告，但是重复的不会终止程序。

解决方案修改 load_sql 加入 replace 关键字：

load_sql = """LOAD DATA LOCAL INFILE '/Users/furuiyang/codes/mins_daily_import_to_csv/savedir/20190720/20190720_001.csv' \
REPLACE INTO TABLE test01.inc_mins \
FIELDS TERMINATED BY ',' \
ENCLOSED BY '"' \
IGNORE 1 LINES;"""

print(load_sql)

再试一下，不再有警告。

查询 mysql 中的插入数量也没有问题：