rust/clippy_lints/src/unicode.rs

136 lines
4.1 KiB
Rust
Raw Normal View History

2015-06-11 04:35:00 -05:00
use rustc::lint::*;
use rustc::hir::*;
2017-08-11 07:11:46 -05:00
use syntax::ast::{LitKind, NodeId};
2016-02-24 10:38:57 -06:00
use syntax::codemap::Span;
use unicode_normalization::UnicodeNormalization;
2017-09-05 04:33:04 -05:00
use utils::{is_allowed, snippet, span_help_and_lint};
2015-06-11 04:35:00 -05:00
/// **What it does:** Checks for the Unicode zero-width space in the code.
///
/// **Why is this bad?** Having an invisible character in the code makes for all
/// sorts of April fools, but otherwise is very much frowned upon.
///
/// **Known problems:** None.
///
2017-08-09 02:30:56 -05:00
/// **Example:** You don't see it, but there may be a zero-width space
/// somewhere in this text.
2018-03-28 08:24:26 -05:00
declare_clippy_lint! {
pub ZERO_WIDTH_SPACE,
2018-03-28 08:24:26 -05:00
correctness,
"using a zero-width space in a string literal, which is confusing"
}
/// **What it does:** Checks for non-ASCII characters in string literals.
///
/// **Why is this bad?** Yeah, we know, the 90's called and wanted their charset
/// back. Even so, there still are editors and other programs out there that
/// don't work well with Unicode. So if the code is meant to be used
/// internationally, on multiple operating systems, or has other portability
/// requirements, activating this lint could be useful.
///
/// **Known problems:** None.
///
2016-07-15 17:25:44 -05:00
/// **Example:**
/// ```rust
/// let x = "Hä?"
/// ```
2018-03-28 08:24:26 -05:00
declare_clippy_lint! {
pub NON_ASCII_LITERAL,
2018-03-28 08:24:26 -05:00
pedantic,
"using any literal non-ASCII chars in a string literal instead of \
using the `\\u` escape"
}
/// **What it does:** Checks for string literals that contain Unicode in a form
/// that is not equal to its
/// [NFC-recomposition](http://www.unicode.org/reports/tr15/#Norm_Forms).
///
/// **Why is this bad?** If such a string is compared to another, the results
/// may be surprising.
///
/// **Known problems** None.
///
/// **Example:** You may not see it, but “à” and “à” aren't the same string. The
/// former when escaped is actually `"a\u{300}"` while the latter is `"\u{e0}"`.
2018-03-28 08:24:26 -05:00
declare_clippy_lint! {
pub UNICODE_NOT_NFC,
2018-03-28 08:24:26 -05:00
pedantic,
"using a unicode literal not in NFC normal form (see \
2016-05-06 09:09:05 -05:00
[unicode tr15](http://www.unicode.org/reports/tr15/) for further information)"
}
2015-06-11 04:35:00 -05:00
#[derive(Copy, Clone)]
pub struct Unicode;
impl LintPass for Unicode {
fn get_lints(&self) -> LintArray {
lint_array!(ZERO_WIDTH_SPACE, NON_ASCII_LITERAL, UNICODE_NOT_NFC)
2015-06-11 04:35:00 -05:00
}
}
impl<'a, 'tcx> LateLintPass<'a, 'tcx> for Unicode {
fn check_expr(&mut self, cx: &LateContext<'a, 'tcx>, expr: &'tcx Expr) {
if let ExprLit(ref lit) = expr.node {
2016-02-12 11:35:44 -06:00
if let LitKind::Str(_, _) = lit.node {
2017-08-11 07:11:46 -05:00
check_str(cx, lit.span, expr.id)
}
}
}
2015-06-11 04:35:00 -05:00
}
2016-01-03 22:26:12 -06:00
fn escape<T: Iterator<Item = char>>(s: T) -> String {
let mut result = String::new();
for c in s {
if c as u32 > 0x7F {
2016-01-03 22:26:12 -06:00
for d in c.escape_unicode() {
result.push(d)
}
} else {
result.push(c);
}
}
result
2015-06-11 04:35:00 -05:00
}
2017-08-11 07:11:46 -05:00
fn check_str(cx: &LateContext, span: Span, id: NodeId) {
let string = snippet(cx, span, "");
if string.contains('\u{200B}') {
2017-08-09 02:30:56 -05:00
span_help_and_lint(
cx,
ZERO_WIDTH_SPACE,
span,
"zero-width space detected",
&format!(
"Consider replacing the string with:\n\"{}\"",
string.replace("\u{200B}", "\\u{200B}")
),
);
}
if string.chars().any(|c| c as u32 > 0x7F) {
2017-08-09 02:30:56 -05:00
span_help_and_lint(
cx,
NON_ASCII_LITERAL,
span,
"literal non-ASCII character detected",
&format!(
"Consider replacing the string with:\n\"{}\"",
2017-08-11 07:11:46 -05:00
if is_allowed(cx, UNICODE_NOT_NFC, id) {
2017-08-09 02:30:56 -05:00
escape(string.chars())
} else {
escape(string.nfc())
}
),
);
}
2017-08-11 07:11:46 -05:00
if is_allowed(cx, NON_ASCII_LITERAL, id) && string.chars().zip(string.nfc()).any(|(a, b)| a != b) {
2017-08-09 02:30:56 -05:00
span_help_and_lint(
cx,
UNICODE_NOT_NFC,
span,
"non-nfc unicode sequence detected",
&format!("Consider replacing the string with:\n\"{}\"", string.nfc().collect::<String>()),
);
}
2015-06-11 04:35:00 -05:00
}