【unitrix】 4.3 左移运算（＜＜）的实现(shl.rs)
一、源码
这段代码实现了一个类型级别的左移运算（<<）系统，使用 Rust 的类型系统来表示和执行二进制数的左移操作。它支持不同类型的数值（零、正一、负一、二进制数、定点数、浮点数和变量）以及不同的移位量（零、非零、变量）。
use core::ops::Shl;
use crate::number::{FixedPoint, Float, IfB0, IfB1, NonZero, Primitive, PrimitiveInt, Sub1, TypedInt, Unsigned, Var, B0, B1, N1, P1, Z0};// ==================== Left Shift Operation (<<) ====================
// ==================== 左移运算（<<） ====================// ==================== Z0 << All ====================
// Zero left shifted by any amount is still zero
// 零左移任何位数仍然是零
impl<R: Unsigned> Shl<R> for Z0 {type Output = Z0;fn shl(self, _: R) -> Self::Output {Z0}
}// Zero left shifted by a variable amount is still zero
// 零左移可变位数仍然是零
impl<T: Primitive> Shl<Var<T>> for Z0 {type Output = Z0;fn shl(self, _: Var<T>) -> Self::Output {Z0}
}// ==================== P1 << U ====================
// Positive one left shifted by zero is itself
// 正一左移零位是其本身
impl Shl<Z0> for P1 {type Output = Self;fn shl(self, _: Z0) -> Self::Output {self}
}// Positive one left shifted by more than one
// Recursively shifts left by one until the shift amount is zero
// (equivalent to adding zeros at the least significant side)
// 正一左移多于一位
// 递归地左移一位直到移位量为零（相当于在最低有效位侧添加零）
impl<R: Unsigned + NonZero + Sub1> Shl<R> for P1
whereP1: Shl<<R as Sub1>::Output>,
{type Output = B0<<P1 as Shl<<R as Sub1>::Output>>::Output>;fn shl(self, _r: R) -> Self::Output {B0::new()}
}// Positive one left shifted by a variable amount
// 正一左移可变位数
impl<T: PrimitiveInt + From<P1> + Shl<Output=T>> Shl<Var<T>> for P1 {type Output = Var<T>;fn shl(self, rhs: Var<T>) -> Self::Output {Var(T::from(self) << rhs.0)}
}// ==================== N1 << U ====================
// Negative one left shifted by zero is itself
// 负一左移零位是其本身
impl Shl<Z0> for N1 {type Output = Self;fn shl(self, _: Z0) -> Self::Output {self}
}// Negative one left shifted by more than one
// Recursively shifts left by one until the shift amount is zero
// (maintains sign in two's complement arithmetic)
// 负一左移多于一位
// 递归地左移一位直到移位量为零（在二进制补码算术中保持符号）
impl<R: Unsigned + NonZero + Sub1> Shl<R> for N1
whereN1: Shl<<R as Sub1>::Output>,
{type Output = B0<<N1 as Shl<<R as Sub1>::Output>>::Output>;fn shl(self, _r: R) -> Self::Output {B0::new()}
}// Negative one left shifted by a variable amount
// 负一左移可变位数
impl<T: PrimitiveInt + From<N1> + Shl<Output=T>> Shl<Var<T>> for N1 {type Output = Var<T>;fn shl(self, rhs: Var<T>) -> Self::Output {Var(T::from(self) << rhs.0)}
}// ==================== B0 << U ====================
// Binary number ending with 0 left shifted by zero is itself
// 以0结尾的二进制数左移零位是其本身
impl<H: NonZero> Shl<Z0> for B0<H> {type Output = Self;fn shl(self, _: Z0) -> Self::Output {self}
}// Binary number ending with 0 left shifted by more than one
// Recursively shifts left by one until the shift amount is zero
// 以0结尾的二进制数左移多于一位
// 递归地左移一位直到移位量为零
impl<H: NonZero, R: Unsigned + NonZero + Sub1> Shl<R> for B0<H>
where B0<H>: Shl<<R as Sub1>::Output>,
{type Output = B0<<B0<H> as Shl<<R as Sub1>::Output>>::Output>;fn shl(self, _r: R) -> Self::Output {B0::new()}
}// Binary number ending with 0 left shifted by a variable amount
// 以0结尾的二进制数左移可变位数
impl<H: NonZero, T: PrimitiveInt + From<B0<H>> + Shl<Output=T>> Shl<Var<T>> for B0<H> {type Output = Var<T>;fn shl(self, rhs: Var<T>) -> Self::Output {Var(T::from(self) << rhs.0)}
}// ==================== B1 << U ====================
// Binary number ending with 1 left shifted by zero is itself
// 以1结尾的二进制数左移零位是其本身
impl<H: NonZero> Shl<Z0> for B1<H> {type Output = Self;fn shl(self, _: Z0) -> Self::Output {self}
}// Binary number ending with 1 left shifted by more than one
// Recursively shifts left by one until the shift amount is zero
// 以1结尾的二进制数左移多于一位
// 递归地左移一位直到移位量为零
impl<H: NonZero, R: Unsigned + NonZero + Sub1> Shl<R> for B1<H>
where B1<H>: Shl<<R as Sub1>::Output>,
{type Output = B0<<B1<H> as Shl<<R as Sub1>::Output>>::Output>;fn shl(self, _r: R) -> Self::Output {B0::new()}
}// Binary number ending with 1 left shifted by a variable amount
// 以1结尾的二进制数左移可变位数
impl<H: NonZero, T: PrimitiveInt + From<B1<H>> + Shl<Output=T>> Shl<Var<T>> for B1<H> {type Output = Var<T>;fn shl(self, rhs: Var<T>) -> Self::Output {Var(T::from(self) << rhs.0)}
}// ==================== FixedPoint<IntPart, FracPart> << All ====================
// ==================== 定点数左移实现 ====================// Fixed-point number left shifted by zero - no change
// 定点数左移零位 - 无变化
impl<IntPart: TypedInt, FracPart: TypedInt> Shl<Z0> for FixedPoint<IntPart, FracPart> {type Output = Self;fn shl(self, _: Z0) -> Self::Output {self}
}// Fixed-point with zero fractional part left shift
// Shifting affects only the integer part
// 小数部分为零的定点数左移
// 移位仅影响整数部分
impl<IntPart: TypedInt + Shl<R>, R: Unsigned + NonZero> Shl<R> for FixedPoint<IntPart, Z0>
{type Output = FixedPoint<<IntPart as Shl<R>>::Output,Z0>;fn shl(self, _r: R) -> Self::Output {FixedPoint::new()}
}// Fixed-point with one fractional part left shift
// Shifting affects only the integer part
// 小数部分为P1的定点数左移
impl<IntPart: IfB1, R: Unsigned + NonZero + Sub1> Shl<R> for FixedPoint<IntPart, P1>
where<IntPart as IfB1>::Output: Shl<<R as Sub1>::Output>, // 递归条件
{type Output = FixedPoint<<<IntPart as IfB1>::Output as Shl<<R as Sub1>::Output> >::Output,Z0>;fn shl(self, _r: R) -> Self::Output {FixedPoint::new()}
}// Fixed-point with fractional part ending with 0 left shift
// Shifting moves bit from fractional to integer part
// 小数部分以0结尾的定点数左移
// 移位将位从小数部分移动到整数部分
impl<IntPart: TypedInt + IfB0, L: NonZero, R: Unsigned + NonZero + Sub1> Shl<R> for FixedPoint<IntPart, B0<L>>
where FixedPoint< <IntPart as IfB0>::Output,L >: Shl<<R as Sub1>::Output>,
{type Output = < FixedPoint< <IntPart as IfB0>::Output,L > as Shl<<R as Sub1>::Output> >::Output;fn shl(self, r: R) -> Self::Output {FixedPoint::new() << r.sub1()}
}// Fixed-point with fractional part ending with 1 left shift
// Shifting moves bit from fractional to integer part
// 小数部分以1结尾的定点数左移
// 移位将位从小数部分移动到整数部分
impl<IntPart: IfB1, L: NonZero, R: Unsigned + NonZero + Sub1> Shl<R> for FixedPoint<IntPart, B1<L>>
where FixedPoint< <IntPart as IfB1>::Output,L >: Shl<<R as Sub1>::Output>,
{type Output = < FixedPoint< <IntPart as IfB1>::Output,L > as Shl<<R as Sub1>::Output> >::Output;fn shl(self, r: R) -> Self::Output {FixedPoint::new() << r.sub1()}
}/* // Fixed-point left shifted by a variable amount
// 定点数左移可变位数（无意义，取消）
impl<IntPart, FracPart, T> Shl<Var<T>> for FixedPoint<IntPart, FracPart>
whereT: PrimitiveInt,IntPart: Into<T>,FracPart: Into<T>,Self: Into<T>,
{type Output = Var<T>;fn shl(self, rhs: Var<T>) -> Self::Output {Var(self.into() << rhs.0)}
} */// ==================== Float << U ====================
// Floating-point number left shift
// Effectively increases the exponent (equivalent to multiplication by 2^r)
// 浮点数左移
// 实际上是增加指数（相当于乘以2^r）
impl<Significand, Exponent, R> Shl<R> for Float<Significand, Exponent>
whereR: Unsigned + NonZero,Exponent: Shl<R>,
{type Output = Float<Significand, <Exponent as Shl<R>>::Output>;fn shl(self, _r: R) -> Self::Output {Float::new()}
}// ==================== Var << U ====================
impl<T, R> Shl<R> for Var<T>
whereT: PrimitiveInt + From<R> + Shl<Output=T>,R: Unsigned,
{type Output = Var<T>;fn shl(self, rhs: R) -> Self::Output {Var(self.0 << T::from(rhs))}
}// ==================== 辅助类型 ====================
/// 类型别名：左移一位的结果
#[allow(dead_code)]
pub type Shl1<I> = <I as Shl<P1>>::Output;
二、源码分析
- 基础类型和导入
use core::ops::Shl;
use crate::number::{FixedPoint, Float, IfB0, IfB1, NonZero, Primitive, PrimitiveInt, Sub1, TypedInt, Unsigned, Var, B0, B1, N1, P1, Z0};
-
Shl 是 Rust 的左移运算符 trait -
其他导入的类型用于表示各种数值和操作：-
Z0 表示零-
P1 表示正一
-
-
N1 表示负一 -
B0 和 B1 表示二进制数的位 -
FixedPoint 和 Float 表示定点数和浮点数 -
Var 表示变量值 -
其他 trait 用于类型操作
-
- 零的左移 (Z0 << All)
impl<R: Unsigned> Shl<R> for Z0 {type Output = Z0;fn shl(self, _: R) -> Self::Output { Z0 }
}
-
零左移任何位数仍然是零
-
适用于固定位数和变量位数
- 正一的左移 (P1 << U)
impl Shl<Z0> for P1 {type Output = Self;fn shl(self, _: Z0) -> Self::Output { self }
}
-
正一左移零位保持不变 -
左移多于一位时递归地添加零（B0）
- 负一的左移 (N1 << U)
impl Shl<Z0> for N1 {type Output = Self;fn shl(self, _: Z0) -> Self::Output { self }
}
-
负一左移零位保持不变
-
左移多于一位时递归地添加零（B0），保持二进制补码表示
- 以0结尾的二进制数的左移 (B0 << U)
impl<H: NonZero> Shl<Z0> for B0<H> {type Output = Self;fn shl(self, _: Z0) -> Self::Output { self }
}
-
左移零位保持不变 -
左移多于一位时递归地添加零
- 以1结尾的二进制数的左移 (B1 << U)
impl<H: NonZero> Shl<Z0> for B1<H> {type Output = Self;fn shl(self, _: Z0) -> Self::Output { self }
}
-
左移零位保持不变
-
左移多于一位时递归地添加零
- 定点数的左移 (FixedPoint << U)
impl<IntPart: TypedInt, FracPart: TypedInt> Shl<Z0> for FixedPoint<IntPart, FracPart> {type Output = Self;fn shl(self, _: Z0) -> Self::Output { self }
}
-
处理不同情况的定点数左移：
-
零位移位：保持不变 -
整数部分移位 -
小数部分移位（将位从小数部分移动到整数部分）
- 浮点数的左移 (Float << U)
impl<Significand, Exponent, R> Shl<R> for Float<Significand, Exponent>
whereR: Unsigned + NonZero,Exponent: Shl<R>,
{type Output = Float<Significand, <Exponent as Shl<R>>::Output>;fn shl(self, _r: R) -> Self::Output { Float::new() }
}
- 浮点数左移通过增加指数实现（相当于乘以2^r）
- 变量的左移 (Var << U)
impl<T, R> Shl<R> for Var<T>
whereT: PrimitiveInt + From<R> + Shl<Output=T>,R: Unsigned,
{type Output = Var<T>;fn shl(self, rhs: R) -> Self::Output { Var(self.0 << T::from(rhs)) }
}
- 对变量值执行运行时左移操作
- 辅助类型
pub type Shl1<I> = <I as Shl<P1>>::Output;
- 定义类型别名表示左移一位的结果
三、特点总结
+ 类型级编程：所有操作都在编译时通过类型系统完成+ 递归实现：多位左移通过递归的单位移位实现+ 全面覆盖：支持各种数值类型和移位量+ 变量支持：提供运行时变量移位能力+ 数学正确性：正确处理符号（特别是负数的二进制补码表示）
这个系统展示了 Rust 类型系统的强大能力，能够在编译时完成复杂的数学运算类型检查。