Rollup merge of #111543 - Urgau:uplift_invalid_utf8_in_unchecked, r=WaffleLapkin

Uplift `clippy::invalid_utf8_in_unchecked` lint

This PR aims at uplifting the `clippy::invalid_utf8_in_unchecked` lint into two lints.

## `invalid_from_utf8_unchecked`

(deny-by-default)

The `invalid_from_utf8_unchecked` lint checks for calls to `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut` with an invalid UTF-8 literal.

### Example

```rust
unsafe {
    std::str::from_utf8_unchecked(b"cl\x82ippy");
}
```

### Explanation

Creating such a `str` would result in undefined behavior as per documentation for `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut`.

## `invalid_from_utf8`

(warn-by-default)

The `invalid_from_utf8` lint checks for calls to `std::str::from_utf8` and `std::str::from_utf8_mut` with an invalid UTF-8 literal.

### Example

```rust
std::str::from_utf8(b"ru\x82st");
```

### Explanation

Trying to create such a `str` would always return an error as per documentation for `std::str::from_utf8` and `std::str::from_utf8_mut`.

-----

Mostly followed the instructions for uplifting a clippy lint described here: https://github.com/rust-lang/rust/pull/99696#pullrequestreview-1134072751

````@rustbot```` label: +I-lang-nominated
r? compiler

-----

For Clippy:

changelog: Moves: Uplifted `clippy::invalid_utf8_in_unchecked` into rustc

4 files changed, 148 insertions, 0 deletions

diff --git a/compiler/rustc_lint/messages.ftl b/compiler/rustc_lint/messages.ftl
index d34a3afcba5..e707ac41a05 100644
--- a/compiler/rustc_lint/messages.ftl
+++ b/compiler/rustc_lint/messages.ftl
@@ -304,6 +304,14 @@ lint_improper_ctypes_union_layout_help = consider adding a `#[repr(C)]` or `#[re
 lint_improper_ctypes_union_layout_reason = this union has unspecified layout
 lint_improper_ctypes_union_non_exhaustive = this union is non-exhaustive
 
+# FIXME: we should ordinalize $valid_up_to when we add support for doing so
+lint_invalid_from_utf8_checked = calls to `{$method}` with a invalid literal always return an error
+    .label = the literal was valid UTF-8 up to the {$valid_up_to} bytes
+
+# FIXME: we should ordinalize $valid_up_to when we add support for doing so
+lint_invalid_from_utf8_unchecked = calls to `{$method}` with a invalid literal are undefined behavior
+    .label = the literal was valid UTF-8 up to the {$valid_up_to} bytes
+
 lint_lintpass_by_hand = implementing `LintPass` by hand
     .help = try using `declare_lint_pass!` or `impl_lint_pass!` instead
 
diff --git a/compiler/rustc_lint/src/invalid_from_utf8.rs b/compiler/rustc_lint/src/invalid_from_utf8.rs
new file mode 100644
index 00000000000..3291286ad67
--- /dev/null
+++ b/compiler/rustc_lint/src/invalid_from_utf8.rs
@@ -0,0 +1,118 @@
+use std::str::Utf8Error;
+
+use rustc_ast::{BorrowKind, LitKind};
+use rustc_hir::{Expr, ExprKind};
+use rustc_span::source_map::Spanned;
+use rustc_span::sym;
+
+use crate::lints::InvalidFromUtf8Diag;
+use crate::{LateContext, LateLintPass, LintContext};
+
+declare_lint! {
+    /// The `invalid_from_utf8_unchecked` lint checks for calls to
+    /// `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut`
+    /// with an invalid UTF-8 literal.
+    ///
+    /// ### Example
+    ///
+    /// ```rust,compile_fail
+    /// # #[allow(unused)]
+    /// unsafe {
+    ///     std::str::from_utf8_unchecked(b"Ru\x82st");
+    /// }
+    /// ```
+    ///
+    /// {{produces}}
+    ///
+    /// ### Explanation
+    ///
+    /// Creating such a `str` would result in undefined behavior as per documentation
+    /// for `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut`.
+    pub INVALID_FROM_UTF8_UNCHECKED,
+    Deny,
+    "using a non UTF-8 literal in `std::str::from_utf8_unchecked`"
+}
+
+declare_lint! {
+    /// The `invalid_from_utf8` lint checks for calls to
+    /// `std::str::from_utf8` and `std::str::from_utf8_mut`
+    /// with an invalid UTF-8 literal.
+    ///
+    /// ### Example
+    ///
+    /// ```rust
+    /// # #[allow(unused)]
+    /// std::str::from_utf8(b"Ru\x82st");
+    /// ```
+    ///
+    /// {{produces}}
+    ///
+    /// ### Explanation
+    ///
+    /// Trying to create such a `str` would always return an error as per documentation
+    /// for `std::str::from_utf8` and `std::str::from_utf8_mut`.
+    pub INVALID_FROM_UTF8,
+    Warn,
+    "using a non UTF-8 literal in `std::str::from_utf8`"
+}
+
+declare_lint_pass!(InvalidFromUtf8 => [INVALID_FROM_UTF8_UNCHECKED, INVALID_FROM_UTF8]);
+
+impl<'tcx> LateLintPass<'tcx> for InvalidFromUtf8 {
+    fn check_expr(&mut self, cx: &LateContext<'tcx>, expr: &'tcx Expr<'tcx>) {
+        if let ExprKind::Call(path, [arg]) = expr.kind
+            && let ExprKind::Path(ref qpath) = path.kind
+            && let Some(def_id) = cx.qpath_res(qpath, path.hir_id).opt_def_id()
+            && let Some(diag_item) = cx.tcx.get_diagnostic_name(def_id)
+            && [sym::str_from_utf8, sym::str_from_utf8_mut,
+                sym::str_from_utf8_unchecked, sym::str_from_utf8_unchecked_mut].contains(&diag_item)
+        {
+            let lint = |utf8_error: Utf8Error| {
+                let label = arg.span;
+                let method = diag_item.as_str().strip_prefix("str_").unwrap();
+                let method = format!("std::str::{method}");
+                let valid_up_to = utf8_error.valid_up_to();
+                let is_unchecked_variant = diag_item.as_str().contains("unchecked");
+
+                cx.emit_spanned_lint(
+                    if is_unchecked_variant { INVALID_FROM_UTF8_UNCHECKED } else { INVALID_FROM_UTF8 },
+                    expr.span,
+                    if is_unchecked_variant {
+                        InvalidFromUtf8Diag::Unchecked { method,  valid_up_to, label }
+                    } else {
+                        InvalidFromUtf8Diag::Checked { method,  valid_up_to, label }
+                    }
+                )
+            };
+
+            match &arg.kind {
+                ExprKind::Lit(Spanned { node: lit, .. }) => {
+                    if let LitKind::ByteStr(bytes, _) = &lit
+                        && let Err(utf8_error) = std::str::from_utf8(bytes)
+                    {
+                        lint(utf8_error);
+                    }
+                },
+                ExprKind::AddrOf(BorrowKind::Ref, _, Expr { kind: ExprKind::Array(args), .. }) => {
+                    let elements = args.iter().map(|e|{
+                        match &e.kind {
+                            ExprKind::Lit(Spanned { node: lit, .. }) => match lit {
+                                LitKind::Byte(b) => Some(*b),
+                                LitKind::Int(b, _) => Some(*b as u8),
+                                _ => None
+                            }
+                            _ => None
+                        }
+                    }).collect::<Option<Vec<_>>>();
+
+                    if let Some(elements) = elements
+                        && let Err(utf8_error) = std::str::from_utf8(&elements)
+                    {
+                        lint(utf8_error);
+                    }
+                }
+                _ => {}
+            }
+        }
+    }
+}
diff --git a/compiler/rustc_lint/src/lib.rs b/compiler/rustc_lint/src/lib.rs
index dfddfe09ab3..c62109b2986 100644
--- a/compiler/rustc_lint/src/lib.rs
+++ b/compiler/rustc_lint/src/lib.rs
@@ -60,6 +60,7 @@ mod expect;
 mod for_loops_over_fallibles;
 pub mod hidden_unicode_codepoints;
 mod internal;
+mod invalid_from_utf8;
 mod late;
 mod let_underscore;
 mod levels;
@@ -102,6 +103,7 @@ use enum_intrinsics_non_enums::EnumIntrinsicsNonEnums;
 use for_loops_over_fallibles::*;
 use hidden_unicode_codepoints::*;
 use internal::*;
+use invalid_from_utf8::*;
 use let_underscore::*;
 use map_unit_fn::*;
 use methods::*;
@@ -207,6 +209,7 @@ late_lint_methods!(
             HardwiredLints: HardwiredLints,
             ImproperCTypesDeclarations: ImproperCTypesDeclarations,
             ImproperCTypesDefinitions: ImproperCTypesDefinitions,
+            InvalidFromUtf8: InvalidFromUtf8,
             VariantSizeDifferences: VariantSizeDifferences,
             BoxPointers: BoxPointers,
             PathStatements: PathStatements,
diff --git a/compiler/rustc_lint/src/lints.rs b/compiler/rustc_lint/src/lints.rs
index 6b4d01551ae..746abebeb37 100644
--- a/compiler/rustc_lint/src/lints.rs
+++ b/compiler/rustc_lint/src/lints.rs
@@ -699,6 +699,25 @@ pub struct ForgetCopyDiag<'a> {
     pub label: Span,
 }
 
+// invalid_from_utf8.rs
+#[derive(LintDiagnostic)]
+pub enum InvalidFromUtf8Diag {
+    #[diag(lint_invalid_from_utf8_unchecked)]
+    Unchecked {
+        method: String,
+        valid_up_to: usize,
+        #[label]
+        label: Span,
+    },
+    #[diag(lint_invalid_from_utf8_checked)]
+    Checked {
+        method: String,
+        valid_up_to: usize,
+        #[label]
+        label: Span,
+    },
+}
+
 // hidden_unicode_codepoints.rs
 #[derive(LintDiagnostic)]
 #[diag(lint_hidden_unicode_codepoints)]