// FIXME(f16_f128): const casting is not yet supported for these types. Add when available. #![warn(clippy::float_cmp)] #![allow( unused, clippy::no_effect, clippy::op_ref, clippy::unnecessary_operation, clippy::cast_lossless )] //@no-rustfix: suggestions have an error margin placeholder use std::ops::Add; const ZERO: f32 = 0.0; const ONE: f32 = ZERO + 1.0; fn twice(x: T) -> T where T: Add + Copy, { x + x } fn eq_fl(x: f32, y: f32) -> bool { if x.is_nan() { y.is_nan() } else { x == y } // no error, inside "eq" fn } fn fl_eq(x: f32, y: f32) -> bool { if x.is_nan() { y.is_nan() } else { x == y } // no error, inside "eq" fn } struct X { val: f32, } impl PartialEq for X { fn eq(&self, o: &X) -> bool { if self.val.is_nan() { o.val.is_nan() } else { self.val == o.val // no error, inside "eq" fn } } } impl PartialEq for X { fn eq(&self, o: &f32) -> bool { if self.val.is_nan() { o.is_nan() } else { self.val == *o // no error, inside "eq" fn } } } fn main() { ZERO == 0f32; //no error, comparison with zero is ok 1.0f32 != f32::INFINITY; // also comparison with infinity 1.0f32 != f32::NEG_INFINITY; // and negative infinity ZERO == 0.0; //no error, comparison with zero is ok ZERO + ZERO != 1.0; //no error, comparison with zero is ok let x = X { val: 1.0 }; x == 1.0; // no error, custom type that implement PartialOrder for float is not checked ONE == 1f32; ONE == 1.0 + 0.0; ONE + ONE == ZERO + ONE + ONE; ONE != 2.0; ONE != 0.0; // no error, comparison with zero is ok twice(ONE) != ONE; ONE as f64 != 2.0; //~^ ERROR: strict comparison of `f32` or `f64` ONE as f64 != 0.0; // no error, comparison with zero is ok let x: f64 = 1.0; x == 1.0; //~^ ERROR: strict comparison of `f32` or `f64` x != 0f64; // no error, comparison with zero is ok twice(x) != twice(ONE as f64); //~^ ERROR: strict comparison of `f32` or `f64` x < 0.0; // no errors, lower or greater comparisons need no fuzzyness x > 0.0; x <= 0.0; x >= 0.0; let xs: [f32; 1] = [0.0]; let a: *const f32 = xs.as_ptr(); let b: *const f32 = xs.as_ptr(); assert_eq!(a, b); // no errors const ZERO_ARRAY: [f32; 2] = [0.0, 0.0]; const NON_ZERO_ARRAY: [f32; 2] = [0.0, 0.1]; let i = 0; let j = 1; ZERO_ARRAY[i] == NON_ZERO_ARRAY[j]; // ok, because lhs is zero regardless of i NON_ZERO_ARRAY[i] == NON_ZERO_ARRAY[j]; //~^ ERROR: strict comparison of `f32` or `f64` let a1: [f32; 1] = [0.0]; let a2: [f32; 1] = [1.1]; a1 == a2; //~^ ERROR: strict comparison of `f32` or `f64` arrays a1[0] == a2[0]; //~^ ERROR: strict comparison of `f32` or `f64` // no errors - comparing signums is ok let x32 = 3.21f32; 1.23f32.signum() == x32.signum(); 1.23f32.signum() == -(x32.signum()); 1.23f32.signum() == 3.21f32.signum(); 1.23f32.signum() != x32.signum(); 1.23f32.signum() != -(x32.signum()); 1.23f32.signum() != 3.21f32.signum(); let x64 = 3.21f64; 1.23f64.signum() == x64.signum(); 1.23f64.signum() == -(x64.signum()); 1.23f64.signum() == 3.21f64.signum(); 1.23f64.signum() != x64.signum(); 1.23f64.signum() != -(x64.signum()); 1.23f64.signum() != 3.21f64.signum(); // the comparison should also look through references &0.0 == &ZERO; &&&&0.0 == &&&&ZERO; }