Fix failing type interference for floating point literal #1604
Conversation
|
Just a quick note: you don't need to go that route to test this. You can simply use the "Show syntax tree" command in VS Code. |
|
The lexer decides that 42f64 is an integer literal here: https://github.com/rust-lang/rust/blob/master/src/librustc_lexer/src/lib.rs#L419 Edit: and it looks like rustc might have a special case to handle this in https://github.com/rust-lang/rust/blob/a7f28678bbf4e16893bb6a718e427504167a9494/src/libsyntax/parse/literal.rs#L424 |
| Token { kind: FLOAT_NUMBER, len: TextUnit::from_usize(5)} | ||
| ); | ||
| } | ||
| } |
There was a problem hiding this comment.
Thanks, fixed (although it looks like this test might be incorrect anyway - see edit to my previous comment).
| // The lexer treats e.g. `1f64` as an integer literal. See | ||
| // https://github.com/rust-analyzer/rust-analyzer/issues/1592 | ||
| // and the comments on the linked PR. | ||
| let float_suffix_list = [ |
There was a problem hiding this comment.
Is this the best place to handle this? It appears roughly analogous to where the special case is handled in rustc itself - see https://github.com/rust-lang/rust/blob/a7f28678bbf4e16893bb6a718e427504167a9494/src/libsyntax/parse/literal.rs#L424
Edit: or maybe we should handle this in https://github.com/rust-analyzer/rust-analyzer/blob/master/crates/ra_syntax/src/parsing/lexer.rs#L128 ?
There was a problem hiding this comment.
I think this is indeed the right place to do this. 1 is indeed, lexically, an integer literal
| @@ -69,6 +69,21 @@ mod tests { | |||
| ); | |||
| } | |||
|
|
|||
| // https://github.com/rust-analyzer/rust-analyzer/issues/1592 | |||
| #[test] | |||
| fn add_explicit_type_infers_correct_type_for_floating_point_literal() { | |||
There was a problem hiding this comment.
I think this would be more fitting to test in the type inference tests, i.e. ra_hir/hy/tests.rs.
|
LGTM, but the test should be moved to |
|
Thanks. I've moved the test to infer, but marks don't seem to work here as the test and the code under test are in different crates (ra_hir and ra_syntax respectively). |
|
bors r+ Thanks! |
Build succeeded |
Fixes #1592