1 use clippy_utils::diagnostics::span_lint_and_sugg;
2 use clippy_utils::is_lint_allowed;
3 use clippy_utils::source::snippet;
4 use rustc_ast::ast::LitKind;
5 use rustc_errors::Applicability;
6 use rustc_hir::{Expr, ExprKind, HirId};
7 use rustc_lint::{LateContext, LateLintPass};
8 use rustc_session::{declare_lint_pass, declare_tool_lint};
9 use rustc_span::source_map::Span;
10 use unicode_normalization::UnicodeNormalization;
12 declare_clippy_lint! {
14 /// Checks for invisible Unicode characters in the code.
16 /// ### Why is this bad?
17 /// Having an invisible character in the code makes for all
18 /// sorts of April fools, but otherwise is very much frowned upon.
21 /// You don't see it, but there may be a zero-width space or soft hyphen
22 /// somewhere in this text.
23 pub INVISIBLE_CHARACTERS,
25 "using an invisible character in a string literal, which is confusing"
28 declare_clippy_lint! {
30 /// Checks for non-ASCII characters in string literals.
32 /// ### Why is this bad?
33 /// Yeah, we know, the 90's called and wanted their charset
34 /// back. Even so, there still are editors and other programs out there that
35 /// don't work well with Unicode. So if the code is meant to be used
36 /// internationally, on multiple operating systems, or has other portability
37 /// requirements, activating this lint could be useful.
41 /// let x = String::from("€");
43 /// Could be written as:
45 /// let x = String::from("\u{20ac}");
47 pub NON_ASCII_LITERAL,
49 "using any literal non-ASCII chars in a string literal instead of using the `\\u` escape"
52 declare_clippy_lint! {
54 /// Checks for string literals that contain Unicode in a form
55 /// that is not equal to its
56 /// [NFC-recomposition](http://www.unicode.org/reports/tr15/#Norm_Forms).
58 /// ### Why is this bad?
59 /// If such a string is compared to another, the results
60 /// may be surprising.
63 /// You may not see it, but "à"" and "à"" aren't the same string. The
64 /// former when escaped is actually `"a\u{300}"` while the latter is `"\u{e0}"`.
67 "using a Unicode literal not in NFC normal form (see [Unicode tr15](http://www.unicode.org/reports/tr15/) for further information)"
70 declare_lint_pass!(Unicode => [INVISIBLE_CHARACTERS, NON_ASCII_LITERAL, UNICODE_NOT_NFC]);
72 impl LateLintPass<'_> for Unicode {
73 fn check_expr(&mut self, cx: &LateContext<'_>, expr: &'_ Expr<'_>) {
74 if let ExprKind::Lit(ref lit) = expr.kind {
75 if let LitKind::Str(_, _) = lit.node {
76 check_str(cx, lit.span, expr.hir_id);
82 fn escape<T: Iterator<Item = char>>(s: T) -> String {
83 let mut result = String::new();
86 for d in c.escape_unicode() {
96 fn check_str(cx: &LateContext<'_>, span: Span, id: HirId) {
97 let string = snippet(cx, span, "");
98 if string.chars().any(|c| ['\u{200B}', '\u{ad}', '\u{2060}'].contains(&c)) {
101 INVISIBLE_CHARACTERS,
103 "invisible character detected",
104 "consider replacing the string with",
106 .replace("\u{200B}", "\\u{200B}")
107 .replace("\u{ad}", "\\u{AD}")
108 .replace("\u{2060}", "\\u{2060}"),
109 Applicability::MachineApplicable,
112 if string.chars().any(|c| c as u32 > 0x7F) {
117 "literal non-ASCII character detected",
118 "consider replacing the string with",
119 if is_lint_allowed(cx, UNICODE_NOT_NFC, id) {
120 escape(string.chars())
124 Applicability::MachineApplicable,
127 if is_lint_allowed(cx, NON_ASCII_LITERAL, id) && string.chars().zip(string.nfc()).any(|(a, b)| a != b) {
132 "non-NFC Unicode sequence detected",
133 "consider replacing the string with",
134 string.nfc().collect::<String>(),
135 Applicability::MachineApplicable,