Uh oh!
There was an error while loading. Please reload this page.
rustc: catch non-trait methods before typeck. - #14974
Conversation
Ryman
commented
Jun 17, 2014
Not sure to remove the potentially dead code in typeck, or update it to a compiler bug (should never be reached), or just leave it? See here. |
alexcrichton
commented
Jun 17, 2014
It looks like travis found a test case that also needs to be altered. It'd ok to leave the "dead code" for now, but you can change it to |
Ryman
commented
Jun 20, 2014
@alexcrichton yep, fixed up the travis problems, had to move another error to resolve too, else it's confusing to the user of what they did wrong in some scenarios. Should be grand now other than a lame span for the typedef case. |
There was a problem hiding this comment.
Previously it would error after resolve on this line due to using a typedef for a trait. I've changed to a note //~^ NOTE: type aliases cannot be used for traits but the span is on the impl.
If I drop the additional note and leave just a 'x is not a trait' then a user might still be confused, so it helps.
The 'x is not a trait' error was moved to resolve, as otherwise the user would encounter the 'method is not a member of trait' error, which, if they've mixed up the syntax, will be difficult for them to figure out what's wrong. (If they have impl type for trait)
I believe the span could be improved, but I don't feel it's a blocker as the note should give enough information to correct (and is an improvement over the old error message).
But on those lines, the 'reference to trait' error should still popup if there's a typedef for a trait which isn't used in an impl, I can add a new test for that case, as the initial reason for testing that is that it used to ICE (according to the issue).
There was a problem hiding this comment.
Ok, could you ensure that there's a test case for type Foo = Trait yielding an error, and remove the FIXME annotations in the tests?
alexcrichton
commented
Jun 21, 2014
This looks pretty good, but the lack of |
Properly format documentation for `SignatureHelpRequest`s Properly formats function documentation instead of returning it raw when responding to `SignatureHelpRequest`s. I added a test in `crates/rust-analyzer/tests/slow-tests/main.rs` -- not sure if this is the best location given the relevant code is in `crates/rust-analyzer` or if it's possible to test in a less heavyweight manner. Closesrust-lang#14958
Closes#3973.