#### 1. 介绍
- 通过 `sync.Pool`, `strings.Builder` 等实现的较高性能 sql 拼接工具, 如果使用 `orm` 的此项目不满足使用(但还是希望大佬阅读, 指正下问题😃)
- 自动打印 sql 最终的 log
- 非法字符会自动转义，防注入
- 安装:
```
go get -u gitee.com/xuesongtao/spellsql
```
#### 2. 占位符
- 目前支持占位符 `?, ?d, ?v`, 说明如下:
##### 2.1 占位符 ?
- 直接根据 args 中类型来自动推动 arg 的类型, 使用如下:
1.第一种用法: 根据 args 中类型来自动推动 arg 的类型
```
如: NewCacheSql("SELECT username, password FROM sys_user WHERE username = ? AND password = ?", "test", 123).GetSqlStr()
=> SELECT username, password FROM sys_user WHERE username = "test" AND password = 123
```
2.第二种用法: 当 arg 为 []int, 暂时支持 []int, []int32, []int64
```
如: NewCacheSql("SELECT username, password FROM sys_user WHERE id IN (?)", []int{1, 2, 3}).GetSqlStr()
=> SELECT username, password FROM sys_user WHERE id IN (1,2,3)
```
##### 2.2 占位符 ?d
- 只会把数字型的字符串转为数字型, 如果是字母的话会被转义为 0, 如: `"123" => 123`; `[]string{"1", "2", "3"} => 1,2,3`, 如下:
第一种用法: 当 arg 为字符串时, 又想不加双引号就用这个
```
如: NewCacheSql("SELECT username, password FROM sys_user WHERE id = ?d", "123").GetSqlStr()
=> SELECT username, password FROM sys_user WHERE id = 123
```
第二种用法: 当 arg 为 []string, 又想把解析后的单个元素不加引号
```
如: NewCacheSql("SELECT username, password FROM sys_user WHERE id IN (?d)", []string{"1", "2", "3"}).GetSqlStr()
=> SELECT username, password FROM sys_user WHERE id IN (1,2,3)
```
##### 2.3 占位符为: ?v
- 这样会让字符串类型不加引号, 原样输出, 如: "test" => test;
第一种用法: 当 arg 为字符串时, 又想不加双引号就用这个, 注: 只支持 arg 为字符串类型
```
如: NewCacheSql("SELECT username, password FROM ?v WHERE id = ?d", "sys_user", "123").GetSqlStr()
=> SELECT username, password FROM sys_user WHERE id = 123
```
第二种用法: 子查询
```
如: NewCacheSql("SELECT u.username, u.password FROM sys_user su LEFT JOIN user u ON su.id = u.id WHERE u.id IN (?v)", FmtSqlStr("SELECT id FROM user WHERE name=?", "test").GetSqlStr()
=> SELECT u.username, u.password FROM sys_user su LEFT JOIN user u ON su.id = u.id WHERE u.id IN (SELECT id FROM user WHERE name="test");
```
- **注:** 由于这种不会进行转义处理, 所有这种不推荐用于请求输入(外部非法输入)的内容, 会出现 **SQL 注入风险**; 当我们明确知道参数是干什么的可以使用会简化我们代码, 这里就不进行演示.
#### 3. 使用
- 可以参考 `getsqlstr_test.go` 里的测试方法
##### 3.1 新增
```
s := NewCacheSql("INSERT INTO sys_user (username, password, name)")
s.SetInsertValues("xuesongtao", "123456", "阿桃")
s.SetInsertValues("xuesongtao", "123456", "阿桃")
s.GetSqlStr()
// Output:
// INSERT INTO sys_user (username, password, name) VALUES ("test", 123456, "阿涛"), ("xuesongtao", "123456", "阿桃"), ("xuesongtao", "123456", "阿桃");
```
##### 3.2 删除
```
s := NewCacheSql("DELETE FROM sys_user WHERE id = ?", 123)
if true {
s.SetWhere("name", "test")
}
s.GetSqlStr()
// Output:
// DELETE FROM sys_user WHERE id = 123 AND name = "test";
```
##### 3.3 查询
```
s := NewCacheSql("SELECT * FROM user u LEFT JOIN role r ON u.id = r.user_id")
s.SetOrWhere("u.name", "xue")
s.SetOrWhereArgs("(r.id IN (?d))", []string{"1", "2"})
s.SetWhere("u.age", ">", 20)
s.SetWhereArgs("u.addr = ?", "南部")
s.GetTotalSqlStr()
s.SetLimit(1, 10)
s.GetSqlStr()
// Output:
// sqlTotalStr: SELECT COUNT(*) FROM user u LEFT JOIN role r ON u.id = r.user_id WHERE u.name = "xue" OR (r.id IN (1,2)) AND u.age > 20 AND u.addr = "南部";
// sqlStr: SELECT * FROM user u LEFT JOIN role r ON u.id = r.user_id WHERE u.name = "xue" OR (r.id IN (1,2)) AND u.age > 20 AND u.addr = "南部" LIMIT 0, 10;
```
##### 3.4 修改
```
s := NewCacheSql("UPDATE sys_user SET")
idsStr := []string{"1", "2", "3", "4", "5"}
s.SetUpdateValue("name", "xue")
s.SetUpdateValueArgs("age = ?, score = ?", 18, 90.5)
s.SetWhereArgs("id IN (?d) AND name = ?", idsStr, "tao")
s.GetSqlStr()
// Output:
// UPDATE sys_user SET name = "xue", age = 18, score = 90.50 WHERE id IN (1,2,3,4,5) AND name = "tao";
```
##### 3.5 复用
- 1. `NewCacheSql()` 获取的对象在调用 `GetSqlStr()` 后会重置并放入内存池, 是不能对结果进行再进行 `GetSqlStr()`, 当然你是可以对结果作为 `NewCacheSql()` 的入参进行使用以此达到复用, 这样代码看起来不是多优雅, 分页处理案例如下:
```
sqlObj := NewCacheSql("SELECT * FROM user_info WHERE status = 1")
handleFn := func(obj *SqlStrObj, page, size int32) {
// 业务代码
fmt.Println(obj.SetLimit(page, size).SetPrintLog(false).GetSqlStr())
}
// 每次同步大小
var (
totalNum int32 = 30
page int32 = 1
size int32 = 10
totalPage int32 = int32(math.Ceil(float64(totalNum / size)))
)
sqlStr := sqlObj.SetPrintLog(false).GetSqlStr("", "")
for page <= totalPage {
handleFn(NewCacheSql(sqlStr), page, size)
page++
}
// Output:
// SELECT * FROM user_info WHERE u_status = 1 LIMIT 0, 10;
// SELECT * FROM user_info WHERE u_status = 1 LIMIT 10, 10;
// SELECT * FROM user_info WHERE u_status = 1 LIMIT 20, 10;
```
- `NewSql()` 的产生的对象不会放入内存池, 可以进行多次调用 `GetSqlStr()`, 对应上面的示例可以使用 `NewSql()` 再调用 `Clone()` 进行处理, 如下:
```
sqlObj := NewSql("SELECT u_name, phone, account_id FROM user_info WHERE u_status = 1")
handleFn := func(obj *SqlStrObj, page, size int32) {
// 业务代码
fmt.Println(obj.SetLimit(page, size).SetPrintLog(false).GetSqlStr())
}
// 每次同步大小
var (
totalNum int32 = 30
page int32 = 1
size int32 = 10
totalPage int32 = int32(math.Ceil(float64(totalNum / size)))
)
for page <= totalPage {
handleFn(sqlObj.Clone(), page, size)
page++
}
// Output:
// SELECT * FROM user_info WHERE u_status = 1 LIMIT 0, 10;
// SELECT * FROM user_info WHERE u_status = 1 LIMIT 10, 10;
// SELECT * FROM user_info WHERE u_status = 1 LIMIT 20, 10;
```
#### 其他
- 欢迎大佬们指正, 希望大佬给 **star**, [to gitee](https://gitee.com/xuesongtao/spellsql)
