chore(codegen): error to prevent undesired behavior

logos-codegen/src/mir.rs

@@ -110,9 +110,13 @@ impl TryFrom<Hir> for Mir {
                 }
 
                 let kind = repetition.kind;
+                let is_dot = *repetition.hir == Hir::dot(!repetition.hir.is_always_utf8());
                 let mir = Mir::try_from(*repetition.hir)?;
 
                 match kind {
+                    RepetitionKind::ZeroOrMore | RepetitionKind::OneOrMore if is_dot => {
+                Err(r#"#[regex]: ".+" and ".*" regexes will always match everything, which is most propably not what you want. If you are looking to match everything until a specific character, you should use a capturing group. E.g., use regex r"\([^\}]\)" to match anything in between two parentheses. Read more about that here: https://github.com/maciejhirsz/logos/issues/302#issuecomment-1521342541."#.into())
+            }
                     RepetitionKind::ZeroOrOne => Ok(Mir::Maybe(Box::new(mir))),
                     RepetitionKind::ZeroOrMore => Ok(Mir::Loop(Box::new(mir))),
                     RepetitionKind::OneOrMore => {

tests/src/lib.rs

@@ -1,3 +1,87 @@
+//! ```compile_fail
+//! use logos::Logos;
+//! use logos_derive::Logos;
+//!
+//! #[derive(Logos)]
+//! enum Token {
+//!     #[token(b"\xFF")]
+//!     NonUtf8,
+//! }
+//!
+//! fn main() {
+//!     Token::lexer("This shouldn't work with a string literal!");
+//! }
+//! ```
+//!
+//! Same, but with regex:
+//!
+//! ```compile_fail
+//! use logos::Logos;
+//! use logos_derive::Logos;
+//!
+//! #[derive(Logos)]
+//! enum Token {
+//!     #[regex(b"\xFF")]
+//!     NonUtf8,
+//! }
+//!
+//! fn main() {
+//!     Token::lexer("This shouldn't work with a string literal!");
+//! }
+//! ```
+//!
+//! Matching against .* (or .+) should fail to compile:
+//!
+//! ```compile_fail
+//! use logos::Logos;
+//! use logos_derive::Logos;
+//!
+//! #[derive(Logos, Debug, PartialEq)]
+//! enum Token {
+//!     #[regex(r"\(.*\)")]
+//!     BetweenParen,
+//!
+//! }
+//! ```
+//!
+//! ```compile_fail
+//! use logos::Logos;
+//! use logos_derive::Logos;
+//!
+//! #[derive(Logos, Debug, PartialEq)]
+//! enum Token {
+//!     #[regex(r"\(.+\)")]
+//!     BetweenParen,
+//!
+//! }
+//! ```
+//!
+//! And also when working with bytes:
+//!
+//! ```compile_fail
+//! use logos::Logos;
+//! use logos_derive::Logos;
+//!
+//! #[derive(Logos, Debug, PartialEq)]
+//! enum Token {
+//!     #[regex(b"\x00.*")]
+//!     NonUtf8,
+//!
+//! }
+//! ```
+//!
+//! ```compile_fail
+//! use logos::Logos;
+//! use logos_derive::Logos;
+//!
+//! #[derive(Logos, Debug, PartialEq)]
+//! enum Token {
+//!     #[regex(b"\x00.+")]
+//!     NonUtf8,
+//!
+//! }
+//! ```
+
 use logos::source::Source;
 use logos::Logos;
 

tests/tests/binary.rs

@@ -1,34 +1,3 @@
-//! ```compile_fail
-//! use logos::Logos;
-//! use logos_derive::Logos;
-//!
-//! #[derive(Logos)]
-//! enum Token {
-//!     #[token(b"\xFF")]
-//!     NonUtf8,
-//! }
-//!
-//! fn main() {
-//!     Token::lexer("This shouldn't work with a string literal!");
-//! }
-//! ```
-//! Same, but with regex:
-//!
-//! ```compile_fail
-//! use logos::Logos;
-//! use logos_derive::Logos;
-//!
-//! #[derive(Logos)]
-//! enum Token {
-//!     #[regex(b"\xFF")]
-//!     NonUtf8,
-//! }
-//!
-//! fn main() {
-//!     Token::lexer("This shouldn't work with a string literal!");
-//! }
-//! ```
-
 use logos_derive::Logos;
 use tests::assert_lex;