New lints [manual_is_infinite] and [manual_is_finite]

CHANGELOG.md

@@ -4941,6 +4941,8 @@ Released 2018-09-13
 [`manual_flatten`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_flatten
 [`manual_instant_elapsed`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_instant_elapsed
 [`manual_is_ascii_check`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_is_ascii_check
+[`manual_is_finite`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_is_finite
+[`manual_is_infinite`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_is_infinite
 [`manual_let_else`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_let_else
 [`manual_main_separator_str`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_main_separator_str
 [`manual_map`]: https://rust-lang.github.io/rust-clippy/master/index.html#manual_map

README.md

@@ -5,7 +5,7 @@
 
 A collection of lints to catch common mistakes and improve your [Rust](https://github.com/rust-lang/rust) code.
 
-[There are over 600 lints included in this crate!](https://rust-lang.github.io/rust-clippy/master/index.html)
+[There are over 650 lints included in this crate!](https://rust-lang.github.io/rust-clippy/master/index.html)
 
 Lints are divided into categories, each with a default [lint level](https://doc.rust-lang.org/rustc/lints/levels.html).
 You can choose how much Clippy is supposed to ~~annoy~~ help you by changing the lint level by category.

book/src/README.md

@@ -6,7 +6,7 @@
 A collection of lints to catch common mistakes and improve your
 [Rust](https://github.com/rust-lang/rust) code.
 
-[There are over 600 lints included in this crate!](https://rust-lang.github.io/rust-clippy/master/index.html)
+[There are over 650 lints included in this crate!](https://rust-lang.github.io/rust-clippy/master/index.html)
 
 Lints are divided into categories, each with a default [lint
 level](https://doc.rust-lang.org/rustc/lints/levels.html). You can choose how

clippy_lints/src/declared_lints.rs

@@ -273,6 +273,8 @@ pub(crate) static LINTS: &[&crate::LintInfo] = &[
     crate::manual_async_fn::MANUAL_ASYNC_FN_INFO,
     crate::manual_bits::MANUAL_BITS_INFO,
     crate::manual_clamp::MANUAL_CLAMP_INFO,
+    crate::manual_float_methods::MANUAL_IS_FINITE_INFO,
+    crate::manual_float_methods::MANUAL_IS_INFINITE_INFO,
     crate::manual_is_ascii_check::MANUAL_IS_ASCII_CHECK_INFO,
     crate::manual_let_else::MANUAL_LET_ELSE_INFO,
     crate::manual_main_separator_str::MANUAL_MAIN_SEPARATOR_STR_INFO,

clippy_lints/src/lib.rs

@@ -184,6 +184,7 @@ mod manual_assert;
 mod manual_async_fn;
 mod manual_bits;
 mod manual_clamp;
+mod manual_float_methods;
 mod manual_is_ascii_check;
 mod manual_let_else;
 mod manual_main_separator_str;
@@ -1073,6 +1074,7 @@ pub fn register_plugins(store: &mut rustc_lint::LintStore, sess: &Session, conf:
     store.register_late_pass(|_| Box::new(manual_range_patterns::ManualRangePatterns));
     store.register_early_pass(|| Box::new(visibility::Visibility));
     store.register_late_pass(move |_| Box::new(tuple_array_conversions::TupleArrayConversions { msrv: msrv() }));
+    store.register_late_pass(|_| Box::new(manual_float_methods::ManualFloatMethods));
     // add lints here, do not remove this comment, it's used in `new_lint`
 }
 

clippy_lints/src/manual_float_methods.rs

@@ -0,0 +1,175 @@
+use clippy_utils::{
+    consts::{constant, Constant},
+    diagnostics::span_lint_and_then,
+    is_from_proc_macro, path_to_local,
+    source::snippet_opt,
+};
+use rustc_errors::Applicability;
+use rustc_hir::{BinOpKind, Expr, ExprKind};
+use rustc_lint::{LateContext, LateLintPass, Lint, LintContext};
+use rustc_middle::lint::in_external_macro;
+use rustc_session::{declare_lint_pass, declare_tool_lint};
+
+declare_clippy_lint! {
+    /// ### What it does
+    /// Checks for manual `is_infinite` reimplementations
+    /// (i.e., `x == <float>::INFINITY || x == <float>::NEG_INFINITY`).
+    ///
+    /// ### Why is this bad?
+    /// The method `is_infinite` is shorter and more readable.
+    ///
+    /// ### Example
+    /// ```rust
+    /// # let x = 1.0f32;
+    /// if x == f32::INFINITY || x == f32::NEG_INFINITY {}
+    /// ```
+    /// Use instead:
+    /// ```rust
+    /// # let x = 1.0f32;
+    /// if x.is_infinite() {}
+    /// ```
+    #[clippy::version = "1.72.0"]
+    pub MANUAL_IS_INFINITE,
+    style,
+    "use dedicated method to check if a float is infinite"
+}
+declare_clippy_lint! {
+    /// ### What it does
+    /// Checks for manual `is_finite` reimplementations
+    /// (i.e., `x != <float>::INFINITY && x != <float>::NEG_INFINITY`).
+    ///
+    /// ### Why is this bad?
+    /// The method `is_finite` is shorter and more readable.
+    ///
+    /// ### Example
+    /// ```rust
+    /// # let x = 1.0f32;
+    /// if x != f32::INFINITY && x != f32::NEG_INFINITY {}
+    /// if x.abs() < f32::INFINITY {}
+    /// ```
+    /// Use instead:
+    /// ```rust
+    /// # let x = 1.0f32;
+    /// if x.is_finite() {}
+    /// if x.is_finite() {}
+    /// ```
+    #[clippy::version = "1.72.0"]
+    pub MANUAL_IS_FINITE,
+    style,
+    "use dedicated method to check if a float is finite"
+}
+declare_lint_pass!(ManualFloatMethods => [MANUAL_IS_INFINITE, MANUAL_IS_FINITE]);
+
+#[derive(Clone, Copy)]
+enum Variant {
+    ManualIsInfinite,
+    ManualIsFinite,
+}
+
+impl Variant {
+    pub fn lint(self) -> &'static Lint {
+        match self {
+            Self::ManualIsInfinite => MANUAL_IS_INFINITE,
+            Self::ManualIsFinite => MANUAL_IS_FINITE,
+        }
+    }
+
+    pub fn msg(self) -> &'static str {
+        match self {
+            Self::ManualIsInfinite => "manually checking if a float is infinite",
+            Self::ManualIsFinite => "manually checking if a float is finite",
+        }
+    }
+}
+
+impl<'tcx> LateLintPass<'tcx> for ManualFloatMethods {
+    fn check_expr(&mut self, cx: &LateContext<'tcx>, expr: &'tcx Expr<'tcx>) {
+        if !in_external_macro(cx.sess(), expr.span)
+            && (!cx.param_env.is_const() || cx.tcx.features().active(sym!(const_float_classify)))
+            && let ExprKind::Binary(kind, lhs, rhs) = expr.kind
+            && let ExprKind::Binary(lhs_kind, lhs_lhs, lhs_rhs) = lhs.kind
+            && let ExprKind::Binary(rhs_kind, rhs_lhs, rhs_rhs) = rhs.kind
+            // Checking all possible scenarios using a function would be a hopeless task, as we have
+            // 16 possible alignments of constants/operands. For now, let's use `partition`.
+            && let (operands, constants) = [lhs_lhs, lhs_rhs, rhs_lhs, rhs_rhs]
+                .into_iter()
+                .partition::<Vec<&Expr<'_>>, _>(|i| path_to_local(i).is_some())
+            && let [first, second] = &*operands
+            && let Some([const_1, const_2]) = constants
+                .into_iter()
+                .map(|i| constant(cx, cx.typeck_results(), i))
+                .collect::<Option<Vec<_>>>()
+                .as_deref()
+            && path_to_local(first).is_some_and(|f| path_to_local(second).is_some_and(|s| f == s))
+            // The actual infinity check, we also allow `NEG_INFINITY` before` INFINITY` just in
+            // case somebody does that for some reason
+            && (is_infinity(const_1) && is_neg_infinity(const_2)
+                || is_neg_infinity(const_1) && is_infinity(const_2))
+            && !is_from_proc_macro(cx, expr)
+            && let Some(local_snippet) = snippet_opt(cx, first.span)
+        {
+            let variant = match (kind.node, lhs_kind.node, rhs_kind.node) {
+                (BinOpKind::Or, BinOpKind::Eq, BinOpKind::Eq) => Variant::ManualIsInfinite,
+                (BinOpKind::And, BinOpKind::Ne, BinOpKind::Ne) => Variant::ManualIsFinite,
+                _ => return,
+            };
+
+            span_lint_and_then(
+                cx,
+                variant.lint(),
+                expr.span,
+                variant.msg(),
+                |diag| {
+                    match variant {
+                        Variant::ManualIsInfinite => {
+                            diag.span_suggestion(
+                                expr.span,
+                                "use the dedicated method instead",
+                                format!("{local_snippet}.is_infinite()"),
+                                Applicability::MachineApplicable,
+                            );
+                        },
+                        Variant::ManualIsFinite => {
+                            // TODO: There's probably some better way to do this, i.e., create
+                            // multiple suggestions with notes between each of them
+                            diag.span_suggestion_verbose(
+                                expr.span,
+                                "use the dedicated method instead",
+                                format!("{local_snippet}.is_finite()"),
+                                Applicability::MaybeIncorrect,
+                            )
+                            .span_suggestion_verbose(
+                                expr.span,
+                                "this will alter how it handles NaN; if that is a problem, use instead",
+                                format!("{local_snippet}.is_finite() || {local_snippet}.is_nan()"),
+                                Applicability::MaybeIncorrect,
+                            )
+                            .span_suggestion_verbose(
+                                expr.span,
+                                "or, for conciseness",
+                                format!("!{local_snippet}.is_infinite()"),
+                                Applicability::MaybeIncorrect,
+                            );
+                        },
+                    }
+                },
+            );
+        }
+    }
+}
+
+fn is_infinity(constant: &Constant<'_>) -> bool {
+    match constant {
+        Constant::F32(float) => *float == f32::INFINITY,
+        Constant::F64(float) => *float == f64::INFINITY,
+        _ => false,
+    }
+}
+
+fn is_neg_infinity(constant: &Constant<'_>) -> bool {
+    match constant {
+        Constant::F32(float) => *float == f32::NEG_INFINITY,
+        Constant::F64(float) => *float == f64::NEG_INFINITY,
+        _ => false,
+    }
+}

tests/ui/manual_float_methods.rs

@@ -0,0 +1,55 @@
+//@aux-build:proc_macros.rs:proc-macro
+#![allow(clippy::needless_if, unused)]
+#![warn(clippy::manual_is_infinite, clippy::manual_is_finite)]
+#![feature(inline_const)]
+
+#[macro_use]
+extern crate proc_macros;
+
+const INFINITE: f32 = f32::INFINITY;
+const NEG_INFINITE: f32 = f32::NEG_INFINITY;
+
+fn fn_test() -> f64 {
+    f64::NEG_INFINITY
+}
+
+fn fn_test_not_inf() -> f64 {
+    112.0
+}
+
+fn main() {
+    let x = 1.0f32;
+    if x == f32::INFINITY || x == f32::NEG_INFINITY {}
+    if x != f32::INFINITY && x != f32::NEG_INFINITY {}
+    if x == INFINITE || x == NEG_INFINITE {}
+    if x != INFINITE && x != NEG_INFINITE {}
+    let x = 1.0f64;
+    if x == f64::INFINITY || x == f64::NEG_INFINITY {}
+    if x != f64::INFINITY && x != f64::NEG_INFINITY {}
+    // Don't lint
+    if x.is_infinite() {}
+    if x.is_finite() {}
+    if x.abs() < f64::INFINITY {}
+    if f64::INFINITY > x.abs() {}
+    if f64::abs(x) < f64::INFINITY {}
+    if f64::INFINITY > f64::abs(x) {}
+    // Is not evaluated by `clippy_utils::constant`
+    if x != f64::INFINITY && x != fn_test() {}
+    // Not -inf
+    if x != f64::INFINITY && x != fn_test_not_inf() {}
+    const X: f64 = 1.0f64;
+    // Will be linted if `const_float_classify` is enabled
+    if const { X == f64::INFINITY || X == f64::NEG_INFINITY } {}
+    if const { X != f64::INFINITY && X != f64::NEG_INFINITY } {}
+    external! {
+        let x = 1.0;
+        if x == f32::INFINITY || x == f32::NEG_INFINITY {}
+        if x != f32::INFINITY && x != f32::NEG_INFINITY {}
+    }
+    with_span! {
+        span
+        let x = 1.0;
+        if x == f32::INFINITY || x == f32::NEG_INFINITY {}
+        if x != f32::INFINITY && x != f32::NEG_INFINITY {}
+    }
+}

tests/ui/manual_float_methods.stderr

@@ -0,0 +1,80 @@
+error: manually checking if a float is infinite
+  --> $DIR/manual_float_methods.rs:22:8
+   |
+LL |     if x == f32::INFINITY || x == f32::NEG_INFINITY {}
+   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use the dedicated method instead: `x.is_infinite()`
+   |
+   = note: `-D clippy::manual-is-infinite` implied by `-D warnings`
+
+error: manually checking if a float is finite
+  --> $DIR/manual_float_methods.rs:23:8
+   |
+LL |     if x != f32::INFINITY && x != f32::NEG_INFINITY {}
+   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+   |
+   = note: `-D clippy::manual-is-finite` implied by `-D warnings`
+help: use the dedicated method instead
+   |
+LL |     if x.is_finite() {}
+   |        ~~~~~~~~~~~~~
+help: this will alter how it handles NaN; if that is a problem, use instead
+   |
+LL |     if x.is_finite() || x.is_nan() {}
+   |        ~~~~~~~~~~~~~~~~~~~~~~~~~~~
+help: or, for conciseness
+   |
+LL |     if !x.is_infinite() {}
+   |        ~~~~~~~~~~~~~~~~
+
+error: manually checking if a float is infinite
+  --> $DIR/manual_float_methods.rs:24:8
+   |
+LL |     if x == INFINITE || x == NEG_INFINITE {}
+   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use the dedicated method instead: `x.is_infinite()`
+
+error: manually checking if a float is finite
+  --> $DIR/manual_float_methods.rs:25:8
+   |
+LL |     if x != INFINITE && x != NEG_INFINITE {}
+   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+   |
+help: use the dedicated method instead
+   |
+LL |     if x.is_finite() {}
+   |        ~~~~~~~~~~~~~
+help: this will alter how it handles NaN; if that is a problem, use instead
+   |
+LL |     if x.is_finite() || x.is_nan() {}
+   |        ~~~~~~~~~~~~~~~~~~~~~~~~~~~
+help: or, for conciseness
+   |
+LL |     if !x.is_infinite() {}
+   |        ~~~~~~~~~~~~~~~~
+
+error: manually checking if a float is infinite
+  --> $DIR/manual_float_methods.rs:27:8
+   |
+LL |     if x == f64::INFINITY || x == f64::NEG_INFINITY {}
+   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use the dedicated method instead: `x.is_infinite()`
+
+error: manually checking if a float is finite
+  --> $DIR/manual_float_methods.rs:28:8
+   |
+LL |     if x != f64::INFINITY && x != f64::NEG_INFINITY {}
+   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+   |
+help: use the dedicated method instead
+   |
+LL |     if x.is_finite() {}
+   |        ~~~~~~~~~~~~~
+help: this will alter how it handles NaN; if that is a problem, use instead
+   |
+LL |     if x.is_finite() || x.is_nan() {}
+   |        ~~~~~~~~~~~~~~~~~~~~~~~~~~~
+help: or, for conciseness
+   |
+LL |     if !x.is_infinite() {}
+   |        ~~~~~~~~~~~~~~~~
+
+error: aborting due to 6 previous errors
+