# Tiberius Row
`tiberius_row` 是一个Rust过程宏库，用于简化从Tiberius SQL Server客户端获取的行数据到Rust结构体的转换过程。
## 功能特点
- 提供`Row_Ext`派生宏，自动为结构体实现从Tiberius行数据到Rust结构体的转换
- 支持多种SQL数据类型，包括整数、浮点数、字符串、日期时间等
- 使用serde进行数据反序列化，提供类型安全的数据访问
- 自动处理NULL值
- 提供友好的错误处理
## 安装
在你的`Cargo.toml`文件中添加以下依赖：
```toml
[dependencies]
tiberius_row = "0.2.1"
```
默认情况下，`tiberius_row` 会自动包含所需的依赖。如果你想使用特定版本的依赖，可以禁用默认特性并手动添加依赖：
```toml
[dependencies]
tiberius_row = { version = "0.2.1", default-features = false }
tiberius = { version = "0.12.3", features = ["chrono", "rust_decimal", "time"] }
serde = { version = "1.0", features = ["derive"] }
serde_json = "1.0"
base64 = "0.22"
anyhow = "1.0"
```
## 使用方法
1. 首先，为你的结构体派生必要的特性：
```rust
use serde::Deserialize;
use tiberius_row::Row_Ext;
#[derive(Deserialize, Row_Ext)]
struct User {
id: i32,
name: String,
email: Option<String>,
created_at: chrono::DateTime<chrono::Utc>,
}
```
2. 然后，你可以直接从Tiberius的Row转换为你的结构体：
```rust
use anyhow::Result;
use tiberius::Client;
async fn get_user(client: &mut Client<Connection>, user_id: i32) -> Result<User> {
let query = "SELECT id, name, email, created_at FROM users WHERE id = @P1";
let stream = client.query(query, &[&user_id]).await?;
let row = stream.into_row().await?.expect("No user found");
// 使用派生的from_row方法
let user = User::from_row(row)?;
Ok(user)
}
```
## 支持的数据类型
下表展示了SQL Server数据类型与Rust类型的对应关系：
| INT | i32 | 32位整数 |
| BIGINT | i64 或 i128 | 64位整数 |
| SMALLINT | i16 | 16位整数 |
| TINYINT | u8 | 8位无符号整数 |
| BIT | bool | 布尔值 |
| DECIMAL/NUMERIC | tiberius::numeric::Decimal | 高精度数值 |
| FLOAT | f64 | 64位浮点数 |
| REAL | f32 | 32位浮点数 |
| DATE | chrono::NaiveDate | 日期 |
| DATETIME | chrono::NaiveDateTime | 日期时间 |
| DATETIME2 | chrono::NaiveDateTime | 高精度日期时间 |
| TIME | chrono::NaiveTime | 时间 |
| CHAR/VARCHAR | String | 固定/可变长度字符串 |
| NCHAR/NVARCHAR | String | Unicode固定/可变长度字符串 |
| TEXT/NTEXT | String | 长文本 |
| BINARY/VARBINARY | String (Base64编码) | 二进制数据，转换为Base64字符串 |
| UNIQUEIDENTIFIER | String | GUID/UUID |
| XML | String | XML数据 |
## 错误处理
该库使用`anyhow`进行错误处理，当遇到不支持的列数据类型或反序列化失败时会返回适当的错误。
## 许可证
MIT
## 贡献
欢迎提交问题和拉取请求！
---
# Tiberius Row (English Version)
`tiberius_row` is a Rust procedural macro library designed to simplify the conversion of row data from the Tiberius SQL Server client to Rust structures.
## Features
- Provides a `Row_Ext` derive macro that automatically implements conversion from Tiberius row data to Rust structures
- Supports various SQL data types including integers, floating-point numbers, strings, date-time, etc.
- Uses serde for data deserialization, providing type-safe data access
- Automatically handles NULL values
- Provides friendly error handling
## Installation
Add the following dependencies to your `Cargo.toml` file:
```toml
[dependencies]
tiberius_row = "0.2.1"
```
By default, `tiberius_row` will automatically include the required dependencies. If you want to use a specific version of a dependency, you can disable the default features and manually add the dependencies:
```toml
[dependencies]
tiberius_row = { version = "0.2.1", default-features = false }
tiberius = { version = "0.12.3", features = ["chrono", "rust_decimal", "time"] }
serde = { version = "1.0", features = ["derive"] }
serde_json = "1.0"
base64 = "0.22"
anyhow = "1.0"
```
## Usage
1. First, derive the necessary traits for your structure:
```rust
use serde::Deserialize;
use tiberius_row::Row_Ext;
#[derive(Deserialize, Row_Ext)]
struct User {
id: i32,
name: String,
email: Option<String>,
created_at: chrono::DateTime<chrono::Utc>,
}
```
2. Then, you can convert directly from a Tiberius Row to your structure:
```rust
use anyhow::Result;
use tiberius::Client;
async fn get_user(client: &mut Client<Connection>, user_id: i32) -> Result<User> {
let query = "SELECT id, name, email, created_at FROM users WHERE id = @P1";
let stream = client.query(query, &[&user_id]).await?;
let row = stream.into_row().await?.expect("No user found");
// Use the derived from_row method
let user = User::from_row(row)?;
Ok(user)
}
```
## Supported Data Types
The following table shows the correspondence between SQL Server data types and Rust types:
| INT | i32 | 32-bit integer |
| BIGINT | i64 or i128 | 64-bit integer |
| SMALLINT | i16 | 16-bit integer |
| TINYINT | u8 | 8-bit unsigned integer |
| BIT | bool | Boolean value |
| DECIMAL/NUMERIC | tiberius::numeric::Decimal | High-precision numeric value |
| FLOAT | f64 | 64-bit floating-point number |
| REAL | f32 | 32-bit floating-point number |
| DATE | chrono::NaiveDate | Date |
| DATETIME | chrono::NaiveDateTime | Date and time |
| DATETIME2 | chrono::NaiveDateTime | High-precision date and time |
| TIME | chrono::NaiveTime | Time |
| CHAR/VARCHAR | String | Fixed/variable-length string |
| NCHAR/NVARCHAR | String | Unicode fixed/variable-length string |
| TEXT/NTEXT | String | Long text |
| BINARY/VARBINARY | String (Base64 encoded) | Binary data, converted to Base64 string |
| UNIQUEIDENTIFIER | String | GUID/UUID |
| XML | String | XML data |
## Error Handling
This library uses `anyhow` for error handling, returning appropriate errors when encountering unsupported column data types or deserialization failures.
## License
MIT
## Contributions
Issues and pull requests are welcome!