Add new lint for byte char slices

This patch adds a new lint that checks for potentially harder to read
byte char slices: `&[b'a', b'b']` and suggests to replace them with the
easier to read `b"ab"` form.

Signed-Off-By: Marcel Müller <[email protected]>
Co-authored-by: Matthias Beyer <[email protected]>

Author: TheNeikos

CHANGELOG.md

@@ -4080,6 +4080,7 @@ Released 2018-09-13
 [`boxed_local`]: https://rust-lang.github.io/rust-clippy/master/index.html#boxed_local
 [`branches_sharing_code`]: https://rust-lang.github.io/rust-clippy/master/index.html#branches_sharing_code
 [`builtin_type_shadow`]: https://rust-lang.github.io/rust-clippy/master/index.html#builtin_type_shadow
+[`byte_char_slice`]: https://rust-lang.github.io/rust-clippy/master/index.html#byte_char_slice
 [`bytes_count_to_len`]: https://rust-lang.github.io/rust-clippy/master/index.html#bytes_count_to_len
 [`bytes_nth`]: https://rust-lang.github.io/rust-clippy/master/index.html#bytes_nth
 [`cargo_common_metadata`]: https://rust-lang.github.io/rust-clippy/master/index.html#cargo_common_metadata

clippy_lints/src/byte_char_slice.rs

@@ -0,0 +1,86 @@
+use clippy_utils::diagnostics::span_lint_and_sugg;
+use rustc_ast::ast::{BorrowKind, Expr, ExprKind, Mutability};
+use rustc_ast::token::{Lit, LitKind};
+use rustc_errors::Applicability;
+use rustc_lint::{EarlyContext, EarlyLintPass};
+use rustc_session::{declare_lint_pass, declare_tool_lint};
+
+declare_clippy_lint! {
+    /// ### What it does
+    /// Checks for hard to read slices of byte characters, that could be more easily expressed as a
+    /// byte string.
+    ///
+    /// ### Why is this bad?
+    ///
+    /// Potentially makes the string harder to read.
+    ///
+    /// ### Example
+    /// ```rust
+    /// &[b'H', b'e', b'l', b'l', b'o']
+    /// ```
+    /// Use instead:
+    /// ```rust
+    /// b"Hello"
+    /// ```
+    #[clippy::version = "1.68.0"]
+    pub BYTE_CHAR_SLICE,
+    style,
+    "hard to read byte char slice"
+}
+declare_lint_pass!(ByteCharSlice => [BYTE_CHAR_SLICE]);
+
+impl EarlyLintPass for ByteCharSlice {
+    fn check_expr(&mut self, cx: &EarlyContext<'_>, expr: &Expr) {
+        if let Some(slice) = is_byte_char_slice(expr) && !expr.span.from_expansion() {
+            span_lint_and_sugg(
+                cx,
+                BYTE_CHAR_SLICE,
+                expr.span,
+                "can be more succinctly written as a byte str",
+                "try",
+                format!("b\"{slice}\""),
+                Applicability::MaybeIncorrect,
+            );
+        }
+    }
+}
+
+fn is_byte_char_slice(expr: &Expr) -> Option<String> {
+    if let ExprKind::AddrOf(BorrowKind::Ref, Mutability::Not, expr) = &expr.kind {
+        match &expr.kind {
+            ExprKind::Array(members) => {
+                let all_member_bytes = members.iter().all(|member| {
+                    matches!(
+                        &member.kind,
+                        ExprKind::Lit(Lit {
+                            kind: LitKind::Byte,
+                            ..
+                        })
+                    )
+                });
+
+                if !all_member_bytes || members.is_empty() {
+                    return None;
+                }
+
+                Some(
+                    members
+                        .iter()
+                        .map(|member| match &member.kind {
+                            ExprKind::Lit(Lit {
+                                kind: LitKind::Byte,
+                                symbol,
+                                ..
+                            }) => symbol.as_str(),
+                            _ => unreachable!(),
+                        })
+                        .collect::<Vec<_>>()
+                        .join(""),
+                )
+            },
+            _ => None,
+        }
+    } else {
+        None
+    }
+}

clippy_lints/src/declared_lints.rs

@@ -61,6 +61,7 @@ pub(crate) static LINTS: &[&crate::LintInfo] = &[
     crate::booleans::OVERLY_COMPLEX_BOOL_EXPR_INFO,
     crate::borrow_deref_ref::BORROW_DEREF_REF_INFO,
     crate::box_default::BOX_DEFAULT_INFO,
+    crate::byte_char_slice::BYTE_CHAR_SLICE_INFO,
     crate::cargo::CARGO_COMMON_METADATA_INFO,
     crate::cargo::MULTIPLE_CRATE_VERSIONS_INFO,
     crate::cargo::NEGATIVE_FEATURE_NAMES_INFO,

clippy_lints/src/lib.rs

@@ -81,6 +81,7 @@ mod bool_to_int_with_if;
 mod booleans;
 mod borrow_deref_ref;
 mod box_default;
+mod byte_char_slice;
 mod cargo;
 mod casts;
 mod checked_conversions;
@@ -908,6 +909,7 @@ pub fn register_plugins(store: &mut rustc_lint::LintStore, sess: &Session, conf:
     store.register_late_pass(|_| Box::new(fn_null_check::FnNullCheck));
     store.register_late_pass(|_| Box::new(permissions_set_readonly_false::PermissionsSetReadonlyFalse));
     store.register_late_pass(|_| Box::new(size_of_ref::SizeOfRef));
+    store.register_early_pass(|| Box::new(byte_char_slice::ByteCharSlice));
     // add lints here, do not remove this comment, it's used in `new_lint`
 }
 

tests/ui/byte_char_slice.fixed

@@ -0,0 +1,10 @@
+// run-rustfix
+#![allow(unused)]
+#![warn(clippy::byte_char_slice)]
+
+fn main() {
+    let bad = b"abc";
+    let good = &[b'a', 0x42];
+    let good = vec![b'a', b'a'];
+    let good: u8 = [b'a', b'c'].into_iter().sum();
+}

tests/ui/byte_char_slice.rs

@@ -0,0 +1,10 @@
+// run-rustfix
+#![allow(unused)]
+#![warn(clippy::byte_char_slice)]
+
+fn main() {
+    let bad = &[b'a', b'b', b'c'];
+    let good = &[b'a', 0x42];
+    let good = vec![b'a', b'a'];
+    let good: u8 = [b'a', b'c'].into_iter().sum();
+}

tests/ui/byte_char_slice.stderr

@@ -0,0 +1,10 @@
+error: can be more succinctly written as a byte str
+  --> $DIR/byte_char_slice.rs:6:15
+   |
+LL |     let bad = &[b'a', b'b', b'c'];
+   |               ^^^^^^^^^^^^^^^^^^^ help: try: `b"abc"`
+   |
+   = note: `-D clippy::byte-char-slice` implied by `-D warnings`
+
+error: aborting due to previous error
+

tests/ui/invalid_utf8_in_unchecked.rs

@@ -1,4 +1,5 @@
 #![warn(clippy::invalid_utf8_in_unchecked)]
+#![allow(clippy::byte_char_slice)]
 
 fn main() {
     // Valid

tests/ui/invalid_utf8_in_unchecked.stderr

@@ -1,19 +1,19 @@
 error: non UTF-8 literal in `std::str::from_utf8_unchecked`
-  --> $DIR/invalid_utf8_in_unchecked.rs:16:9
+  --> $DIR/invalid_utf8_in_unchecked.rs:17:9
    |
 LL |         std::str::from_utf8_unchecked(&[99, 108, 130, 105, 112, 112, 121]);
    |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |
    = note: `-D clippy::invalid-utf8-in-unchecked` implied by `-D warnings`
 
 error: non UTF-8 literal in `std::str::from_utf8_unchecked`
-  --> $DIR/invalid_utf8_in_unchecked.rs:17:9
+  --> $DIR/invalid_utf8_in_unchecked.rs:18:9
    |
 LL |         std::str::from_utf8_unchecked(&[b'c', b'l', b'/x82', b'i', b'p', b'p', b'y']);
    |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 error: non UTF-8 literal in `std::str::from_utf8_unchecked`
-  --> $DIR/invalid_utf8_in_unchecked.rs:18:9
+  --> $DIR/invalid_utf8_in_unchecked.rs:19:9
    |
 LL |         std::str::from_utf8_unchecked(b"cl/x82ippy");
    |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

tests/ui/ptr_offset_with_cast.fixed

@@ -1,5 +1,5 @@
 // run-rustfix
-#![allow(clippy::unnecessary_cast)]
+#![allow(clippy::unnecessary_cast, clippy::byte_char_slice)]
 
 fn main() {
     let vec = vec![b'a', b'b', b'c'];

tests/ui/ptr_offset_with_cast.rs

@@ -1,5 +1,5 @@
 // run-rustfix
-#![allow(clippy::unnecessary_cast)]
+#![allow(clippy::unnecessary_cast, clippy::byte_char_slice)]
 
 fn main() {
     let vec = vec![b'a', b'b', b'c'];