Implement completionItem/resolve, textDocument/codeAction, textDocument/codeLens and codeLens/resolve #107
Conversation
|
@ebkalderon I'm also working on Edit: Here's the branch on my repo with the implementation. |
There was a problem hiding this comment.
Looks correct to me, I think!
If you'd like to introduce support for the other three LSP request methods in this PR, feel free. I think that because #100 and #58 are now resolved, we no longer need to be overly wary about new methods introducing too much boilerplate. All of these items are somewhat related, so I don't think it's a big deal to group them together.
|
Also, this repository is rebasing merges and not squashing them, meaning we still get to preserve commit-by-commit history. So it's not like having them bundled in one PR ruins the readability of history in Ideally, we should have each newly added method be in a separate commit. Feel free to ping me again for a second review, if needed! Thanks for opening this pull request, @icsaszar. |
icsaszar
changed the title
|
@ebkalderon I've added the new methods, everything should be complete now. This also brings up a small new issue: now that methods are optional by default, it is easy to forget to implement them for Box in lib.rs. If you're ok with it, I can try to write a macro (I don't have too much experience with macros either) to automatically fill out the impls for Box. Edit: I just found auto_impl, which seems to be exactly what what we need. The only question is if it works with |
|
I agree that we should write a macro, but I doubt Otherwise, thanks a lot for the helpful PR! I think it's ready to merge. 👍 |
ebkalderon
changed the title
It turns out that `auto_impl` can work with `async-trait` as long as the invocation occurs first, before calling `#[async_trait]`, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the invocation occurs first, before calling `#[async_trait]`, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the invocation occurs first, before calling `#[async_trait]`, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
It turns out that `auto_impl` can work with `async-trait` as long as the `auto_impl` invocation occurs first, such that the `#[auto_impl]` macro receives the desugared output of `#[async_trait]`. Original suggestion can be found here: #107 (comment)
Resolves #9