sql-plus-plus

a unitive interface for operating database. Direct or Prepare-statement.
All operation can be done with just one interface--query.

Mysql

1、Clusetr model. Mysql MGR clusetr, automaticly adapt clusetr member changed, Master/slave connection update.
2、Single model.

Sqlserver/Oracle

single model.

Usage

For sqlserver/odbc single mode:

#include "db.hpp"
#include "sqlserver_connection_pool.hpp"
using namespace sqlcpp;

//for win
std::string driver_name = "Driver={ODBC Driver 17 for SQL Server}"; //"DRIVER={SQL Server}"
//for linux
std::string driver_name = "Driver=ODBC Driver 17 for SQL Server"; //"DRIVER={SQL Server}"
auto db_ptr = std::make_shared<db<model::single, sqlserver::connection_pool>>
		(std::vector<node_info>{ {"10.10.10.8"}}, "user", "pwd", driver_name);

auto conn = db_ptr->get_conn<conn_type::general>();
//create
conn->query<void>(R"(
CREATE TABLE [dbo].[user] (
  [name] varchar(255) COLLATE Chinese_PRC_CI_AS  NOT NULL,
  [sex] int COLLATE Chinese_PRC_CI_AS  NOT NULL
)  
ON [PRIMARY]
GO

ALTER TABLE [dbo].[user] SET (LOCK_ESCALATION = TABLE)
)");
conn->query<void>("insert into [dbo].[user] values(?,?)", "xixi", 1);

//read
std::vector<std::string> names = conn->query<std::string>("select name from [dbo].[user]");

struct info {
	std::string name;
	int sex;
	REFLECT(info, name, sex);
};
auto cs1 = conn->query<std::tuple<std::string, int>>("select * from [dbo].[user]");
std::vector<info> cs2 = conn->query<info>("select * from [dbo].[user]");

auto ns = conn->query<int>("select sex from [dbo].[user] where name = ?", "xixi");
std::vector<int> sn = conn->query<int>("select sex from [dbo].[user] where name = ? and sex = ?", "xixi", 1);

//update
conn->query<void>("update [dbo].[user] set sex = ? where name = ?", 2, "xixi");

//delete
conn->query<void>("delete from [dbo].[user] where name = ?", "xixi");

//transaction
conn->begin_transaction();
conn->commit_transaction();
conn->rollback();

//deal null column. use std::optional.
//if sex is null, std::optional<int> is empty, otherwise has value
std::vector<std::tuple<std::string, std::optional<int>>> cs11 =
 conn->query<std::tuple<std::string, std::optional<int>>>("select * from [dbo].[user]");

//if sex is empty, then the sex column will be null after inserting into
std::optional<int> sex;
conn->query<void>("insert into [dbo].[user] ([name],[sex]) values(?,?)", "xixi", sex);

For mysql single mode just:

#include "db.hpp"
#include "mysql_connection_pool.hpp"
using namespace sqlcpp;

auto db_ptr = std::make_shared<db<model::single, mysql::connection_pool>>
		(std::vector<node_info>{ {"10.10.10.8"}}, "user", "pwd");
auto conn = db_ptr->get_conn<conn_type::general>();
//then crud is same like above.

For mysql cluster mode:
just one seed node ip is ok, sqlpp will find all mgr cluster nodes and distinguish master and slave.

#include "db.hpp"
#include "mysql_connection_pool.hpp"
using namespace sqlcpp;

auto db_ptr = std::make_shared<db<model::cluster, mysql::connection_pool>>
        (std::vector<node_info>{ {"10.10.10.8"}}, "user", "pwd");
//this is a master node connection, use for writing and reading
auto conn = db_ptr->get_conn<sqlcpp::conn_type::master>();
//this is a slave node connection, just can use for reading
auto conn = db_ptr->get_conn<sqlcpp::conn_type::slave>(); 
//then crud is same like above.

Maybe do

1、postgresql
2、sqlite