use rustc::lint::*; use rustc_front::hir::*; use syntax::codemap::Span; use syntax::ast::Lit_::*; use unicode_normalization::UnicodeNormalization; use utils::{snippet, span_help_and_lint}; declare_lint!{ pub ZERO_WIDTH_SPACE, Deny, "using a zero-width space in a string literal, which is confusing" } declare_lint!{ pub NON_ASCII_LITERAL, Allow, "using any literal non-ASCII chars in a string literal; suggests \ using the \\u escape instead" } declare_lint!{ pub UNICODE_NOT_NFC, Allow, "using a unicode literal not in NFC normal form (see \ http://www.unicode.org/reports/tr15/ for further information)" } #[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) } fn check_expr(&mut self, cx: &Context, expr: &Expr) { if let ExprLit(ref lit) = expr.node { if let LitStr(_, _) = lit.node { check_str(cx, lit.span) } } } } fn escape>(s: T) -> String { let mut result = String::new(); for c in s { if c as u32 > 0x7F { for d in c.escape_unicode() { result.push(d) }; } else { result.push(c); } } result } fn check_str(cx: &Context, span: Span) { let string = snippet(cx, span, ""); if string.contains('\u{200B}') { 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) { span_help_and_lint(cx, NON_ASCII_LITERAL, span, "literal non-ASCII character detected", &format!("Consider replacing the string with:\n\"{}\"", if cx.current_level(UNICODE_NOT_NFC) == Level::Allow { escape(string.chars()) } else { escape(string.nfc()) })); } if cx.current_level(NON_ASCII_LITERAL) == Level::Allow && string.chars().zip(string.nfc()).any(|(a, b)| a != b) { span_help_and_lint(cx, UNICODE_NOT_NFC, span, "non-nfc unicode sequence detected", &format!("Consider replacing the string with:\n\"{}\"", string.nfc().collect::())); } }