-
Notifications
You must be signed in to change notification settings - Fork 1.5k
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Auto merge of #11919 - y21:manual_c_str_literals, r=xFrednet
new lint: `manual_c_str_literals` With rust-lang/rust#117472 merged and `c""` syntax stabilized, I think it'd be nice to have a lint for using `CStr::from_ptr` (and similar constructors) with a string literal as an argument. We can probably also lint `"foo\0".as_ptr()` and suggest `c"foo".as_ptr()`. I might add that to this PR tomorrow if I find the time. The byte string literal to c string literal rewriting is ugly but oh well. changelog: new lint: `manual_c_str_literals`
- Loading branch information
Showing
12 changed files
with
444 additions
and
3 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,197 @@ | ||
| use clippy_config::msrvs::{self, Msrv}; | ||
| use clippy_utils::diagnostics::span_lint_and_sugg; | ||
| use clippy_utils::get_parent_expr; | ||
| use clippy_utils::source::snippet; | ||
| use rustc_ast::{LitKind, StrStyle}; | ||
| use rustc_errors::Applicability; | ||
| use rustc_hir::{Expr, ExprKind, Node, QPath, TyKind}; | ||
| use rustc_lint::LateContext; | ||
| use rustc_span::{sym, Span, Symbol}; | ||
|
|
||
| use super::MANUAL_C_STR_LITERALS; | ||
|
|
||
| /// Checks: | ||
| /// - `b"...".as_ptr()` | ||
| /// - `b"...".as_ptr().cast()` | ||
| /// - `"...".as_ptr()` | ||
| /// - `"...".as_ptr().cast()` | ||
| /// | ||
| /// Iff the parent call of `.cast()` isn't `CStr::from_ptr`, to avoid linting twice. | ||
| pub(super) fn check_as_ptr<'tcx>( | ||
| cx: &LateContext<'tcx>, | ||
| expr: &'tcx Expr<'tcx>, | ||
| receiver: &'tcx Expr<'tcx>, | ||
| msrv: &Msrv, | ||
| ) { | ||
| if let ExprKind::Lit(lit) = receiver.kind | ||
| && let LitKind::ByteStr(_, StrStyle::Cooked) | LitKind::Str(_, StrStyle::Cooked) = lit.node | ||
| && let casts_removed = peel_ptr_cast_ancestors(cx, expr) | ||
| && !get_parent_expr(cx, casts_removed).is_some_and( | ||
| |parent| matches!(parent.kind, ExprKind::Call(func, _) if is_c_str_function(cx, func).is_some()), | ||
| ) | ||
| && let Some(sugg) = rewrite_as_cstr(cx, lit.span) | ||
| && msrv.meets(msrvs::C_STR_LITERALS) | ||
| { | ||
| span_lint_and_sugg( | ||
| cx, | ||
| MANUAL_C_STR_LITERALS, | ||
| receiver.span, | ||
| "manually constructing a nul-terminated string", | ||
| r#"use a `c""` literal"#, | ||
| sugg, | ||
| // an additional cast may be needed, since the type of `CStr::as_ptr` and | ||
| // `"".as_ptr()` can differ and is platform dependent | ||
| Applicability::HasPlaceholders, | ||
| ); | ||
| } | ||
| } | ||
|
|
||
| /// Checks if the callee is a "relevant" `CStr` function considered by this lint. | ||
| /// Returns the function name. | ||
| fn is_c_str_function(cx: &LateContext<'_>, func: &Expr<'_>) -> Option<Symbol> { | ||
| if let ExprKind::Path(QPath::TypeRelative(cstr, fn_name)) = &func.kind | ||
| && let TyKind::Path(QPath::Resolved(_, ty_path)) = &cstr.kind | ||
| && cx.tcx.lang_items().c_str() == ty_path.res.opt_def_id() | ||
| { | ||
| Some(fn_name.ident.name) | ||
| } else { | ||
| None | ||
| } | ||
| } | ||
|
|
||
| /// Checks calls to the `CStr` constructor functions: | ||
| /// - `CStr::from_bytes_with_nul(..)` | ||
| /// - `CStr::from_bytes_with_nul_unchecked(..)` | ||
| /// - `CStr::from_ptr(..)` | ||
| pub(super) fn check(cx: &LateContext<'_>, expr: &Expr<'_>, func: &Expr<'_>, args: &[Expr<'_>], msrv: &Msrv) { | ||
| if let Some(fn_name) = is_c_str_function(cx, func) | ||
| && let [arg] = args | ||
| && msrv.meets(msrvs::C_STR_LITERALS) | ||
| { | ||
| match fn_name.as_str() { | ||
| name @ ("from_bytes_with_nul" | "from_bytes_with_nul_unchecked") | ||
| if !arg.span.from_expansion() | ||
| && let ExprKind::Lit(lit) = arg.kind | ||
| && let LitKind::ByteStr(_, StrStyle::Cooked) | LitKind::Str(_, StrStyle::Cooked) = lit.node => | ||
| { | ||
| check_from_bytes(cx, expr, arg, name); | ||
| }, | ||
| "from_ptr" => check_from_ptr(cx, expr, arg), | ||
| _ => {}, | ||
| } | ||
| } | ||
| } | ||
|
|
||
| /// Checks `CStr::from_ptr(b"foo\0".as_ptr().cast())` | ||
| fn check_from_ptr(cx: &LateContext<'_>, expr: &Expr<'_>, arg: &Expr<'_>) { | ||
| if let ExprKind::MethodCall(method, lit, ..) = peel_ptr_cast(arg).kind | ||
| && method.ident.name == sym::as_ptr | ||
| && !lit.span.from_expansion() | ||
| && let ExprKind::Lit(lit) = lit.kind | ||
| && let LitKind::ByteStr(_, StrStyle::Cooked) = lit.node | ||
| && let Some(sugg) = rewrite_as_cstr(cx, lit.span) | ||
| { | ||
| span_lint_and_sugg( | ||
| cx, | ||
| MANUAL_C_STR_LITERALS, | ||
| expr.span, | ||
| "calling `CStr::from_ptr` with a byte string literal", | ||
| r#"use a `c""` literal"#, | ||
| sugg, | ||
| Applicability::MachineApplicable, | ||
| ); | ||
| } | ||
| } | ||
| /// Checks `CStr::from_bytes_with_nul(b"foo\0")` | ||
| fn check_from_bytes(cx: &LateContext<'_>, expr: &Expr<'_>, arg: &Expr<'_>, method: &str) { | ||
| let (span, applicability) = if let Some(parent) = get_parent_expr(cx, expr) | ||
| && let ExprKind::MethodCall(method, ..) = parent.kind | ||
| && [sym::unwrap, sym::expect].contains(&method.ident.name) | ||
| { | ||
| (parent.span, Applicability::MachineApplicable) | ||
| } else if method == "from_bytes_with_nul_unchecked" { | ||
| // `*_unchecked` returns `&CStr` directly, nothing needs to be changed | ||
| (expr.span, Applicability::MachineApplicable) | ||
| } else { | ||
| // User needs to remove error handling, can't be machine applicable | ||
| (expr.span, Applicability::HasPlaceholders) | ||
| }; | ||
|
|
||
| let Some(sugg) = rewrite_as_cstr(cx, arg.span) else { | ||
| return; | ||
| }; | ||
|
|
||
| span_lint_and_sugg( | ||
| cx, | ||
| MANUAL_C_STR_LITERALS, | ||
| span, | ||
| "calling `CStr::new` with a byte string literal", | ||
| r#"use a `c""` literal"#, | ||
| sugg, | ||
| applicability, | ||
| ); | ||
| } | ||
|
|
||
| /// Rewrites a byte string literal to a c-str literal. | ||
| /// `b"foo\0"` -> `c"foo"` | ||
| /// | ||
| /// Returns `None` if it doesn't end in a NUL byte. | ||
| fn rewrite_as_cstr(cx: &LateContext<'_>, span: Span) -> Option<String> { | ||
| let mut sugg = String::from("c") + snippet(cx, span.source_callsite(), "..").trim_start_matches('b'); | ||
|
|
||
| // NUL byte should always be right before the closing quote. | ||
| if let Some(quote_pos) = sugg.rfind('"') { | ||
| // Possible values right before the quote: | ||
| // - literal NUL value | ||
| if sugg.as_bytes()[quote_pos - 1] == b'\0' { | ||
| sugg.remove(quote_pos - 1); | ||
| } | ||
| // - \x00 | ||
| else if sugg[..quote_pos].ends_with("\\x00") { | ||
| sugg.replace_range(quote_pos - 4..quote_pos, ""); | ||
| } | ||
| // - \0 | ||
| else if sugg[..quote_pos].ends_with("\\0") { | ||
| sugg.replace_range(quote_pos - 2..quote_pos, ""); | ||
| } | ||
| // No known suffix, so assume it's not a C-string. | ||
| else { | ||
| return None; | ||
| } | ||
| } | ||
|
|
||
| Some(sugg) | ||
| } | ||
|
|
||
| fn get_cast_target<'tcx>(e: &'tcx Expr<'tcx>) -> Option<&'tcx Expr<'tcx>> { | ||
| match &e.kind { | ||
| ExprKind::MethodCall(method, receiver, [], _) if method.ident.as_str() == "cast" => Some(receiver), | ||
| ExprKind::Cast(expr, _) => Some(expr), | ||
| _ => None, | ||
| } | ||
| } | ||
|
|
||
| /// `x.cast()` -> `x` | ||
| /// `x as *const _` -> `x` | ||
| /// `x` -> `x` (returns the same expression for non-cast exprs) | ||
| fn peel_ptr_cast<'tcx>(e: &'tcx Expr<'tcx>) -> &'tcx Expr<'tcx> { | ||
| get_cast_target(e).map_or(e, peel_ptr_cast) | ||
| } | ||
|
|
||
| /// Same as `peel_ptr_cast`, but the other way around, by walking up the ancestor cast expressions: | ||
| /// | ||
| /// `foo(x.cast() as *const _)` | ||
| /// ^ given this `x` expression, returns the `foo(...)` expression | ||
| fn peel_ptr_cast_ancestors<'tcx>(cx: &LateContext<'tcx>, e: &'tcx Expr<'tcx>) -> &'tcx Expr<'tcx> { | ||
| let mut prev = e; | ||
| for (_, node) in cx.tcx.hir().parent_iter(e.hir_id) { | ||
| if let Node::Expr(e) = node | ||
| && get_cast_target(e).is_some() | ||
| { | ||
| prev = e; | ||
| } else { | ||
| break; | ||
| } | ||
| } | ||
| prev | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,60 @@ | ||
| #![warn(clippy::manual_c_str_literals)] | ||
| #![allow(clippy::no_effect)] | ||
|
|
||
| use std::ffi::CStr; | ||
|
|
||
| macro_rules! cstr { | ||
| ($s:literal) => { | ||
| CStr::from_bytes_with_nul(concat!($s, "\0").as_bytes()).unwrap() | ||
| }; | ||
| } | ||
|
|
||
| macro_rules! macro_returns_c_str { | ||
| () => { | ||
| CStr::from_bytes_with_nul(b"foo\0").unwrap(); | ||
| }; | ||
| } | ||
|
|
||
| macro_rules! macro_returns_byte_string { | ||
| () => { | ||
| b"foo\0" | ||
| }; | ||
| } | ||
|
|
||
| #[clippy::msrv = "1.76.0"] | ||
| fn pre_stabilization() { | ||
| CStr::from_bytes_with_nul(b"foo\0"); | ||
| } | ||
|
|
||
| #[clippy::msrv = "1.77.0"] | ||
| fn post_stabilization() { | ||
| c"foo"; | ||
| } | ||
|
|
||
| fn main() { | ||
| c"foo"; | ||
| c"foo"; | ||
| c"foo"; | ||
| c"foo\\0sdsd"; | ||
| CStr::from_bytes_with_nul(br"foo\\0sdsd\0").unwrap(); | ||
| CStr::from_bytes_with_nul(br"foo\x00").unwrap(); | ||
| CStr::from_bytes_with_nul(br##"foo#a\0"##).unwrap(); | ||
|
|
||
| unsafe { c"foo" }; | ||
| unsafe { c"foo" }; | ||
| let _: *const _ = c"foo".as_ptr(); | ||
| let _: *const _ = c"foo".as_ptr(); | ||
| let _: *const _ = "foo".as_ptr(); // not a C-string | ||
| let _: *const _ = "".as_ptr(); | ||
| let _: *const _ = c"foo".as_ptr().cast::<i8>(); | ||
| let _ = "电脑".as_ptr(); | ||
| let _ = "电脑\\".as_ptr(); | ||
| let _ = c"电脑\\".as_ptr(); | ||
| let _ = c"电脑".as_ptr(); | ||
| let _ = c"电脑".as_ptr(); | ||
|
|
||
| // Macro cases, don't lint: | ||
| cstr!("foo"); | ||
| macro_returns_c_str!(); | ||
| CStr::from_bytes_with_nul(macro_returns_byte_string!()).unwrap(); | ||
| } |
Oops, something went wrong.