bors[bot] [Sun, 23 May 2021 12:40:14 +0000 (12:40 +0000)]
Merge #8940
8940: Give ‘unsafe’ semantic token modifier to unsafe traits r=Veykril a=arzg
Hi! This is my first pull request that touches rust-analyzer itself beyond a search-and-replace, so please tell me if I should change anything or do anything differently. :)
bors[bot] [Sun, 23 May 2021 11:12:45 +0000 (11:12 +0000)]
Merge #8936 #8939
8936: fix: Improve nightly downloads with better local state management r=matklad a=SomeoneToIgnore
When any nightly is downloaded, we store its GitHub release id in the local cache and never invalidate that cache.
Due to this, it was possible to do the following sequence:
* have the nightly locally
* downgrade the extension to any stable version
* observe that despite the `"rust-analyzer.updates.channel": "nightly",` setting, no nightly updates are happening
* on the next day, the actual update happens (given the new nightly is released)
Since it's impossible to install nightly version directly through the VSCode marketplace, any fiddling with dev version results in the same situation: one have to wait for the next nightly release to happen in order to restore the nightly.
This PR
* invalidates the cache eagerly during bootstrap if the current plugin is not nightly
* enforces the release id check for nightly versions only
* fixes the `ctx.globalStoragePath` deprecated API usage
Hopefully, it also helps mysterious non-updated plugins that we encounter from time to time, but hard to tell for sure.
8939: internal: disable debug symbols due to failing windows build r=matklad a=matklad
Florian Diebold [Sun, 23 May 2021 11:00:14 +0000 (13:00 +0200)]
Paper over #8931 a bit more
The problem was the skipping of binders in
`resolve_method_call_as_callable`; this still doesn't use the _correct_
substitution, but at least it doesn't return a type with free variables
in it.
bors[bot] [Sat, 22 May 2021 12:32:12 +0000 (12:32 +0000)]
Merge #8921
8921: Resolve any lifetime variables to 'static after inference r=flodiebold a=flodiebold
Chalk's unification can sometimes create lifetime variables, which we
currently don't really deal with, but at least we don't want to leak
them outside of inference.
Florian Diebold [Sat, 22 May 2021 12:25:58 +0000 (14:25 +0200)]
Resolve any lifetime variables to 'static after inference
Chalk's unification can sometimes create lifetime variables, which we
currently don't really deal with, but at least we don't want to leak
them outside of inference.
This adds a printer for `ItemTree` contents, and a few tests to ensure that `ItemTree` lowering works like we expect it to. It also adds a new "Debug ItemTree" command that can be used to see the `ItemTree` of the currently open file. The pretty-printed output is usually close enough to Rust syntax that we can even use Rust syntax highlighting.
This is similar to the old `ItemTree` tests we had, but produces significantly more readable output, so these should actually carry their weight.
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
bors[bot] [Fri, 21 May 2021 17:51:53 +0000 (17:51 +0000)]
Merge #8856
8856: Use Chalk for unification r=flodiebold a=flodiebold
- use Chalk's unification, get rid of our own `unify`
- rewrite coercion to not use unification internals and to be more analogous to rustc
- fix various coercion bugs
- rewrite handling of obligations, since the old hacky optimization where we noted when an inference variable changes wasn't possible anymore
- stop trying to deeply resolve types all the time during inference, instead only do it shallowly where necessary
bors[bot] [Fri, 21 May 2021 16:29:31 +0000 (16:29 +0000)]
Merge #8910
8910: Don't lower extern block in the ItemTree r=jonas-schievink a=jonas-schievink
The ItemTree lowering code used to attach attributes on an `extern {}` block to all the children. This is wrong and causes problems with attribute resolution that manifested as a hang.
This PR treats extern blocks as first-class items in the ItemTree and lowers its contents in the `ModCollector` instead.
Florian Diebold [Sun, 16 May 2021 13:50:28 +0000 (15:50 +0200)]
Rework obligation handling
We can't do the easy hack that we did before anymore, where we kept
track of whether any inference variables changed since the last time we
rechecked obligations. Instead, we store the obligations in
canonicalized form; that way we can easily check the inference variables
to see whether they have changed since the goal was canonicalized.
This renames our `MergeBehavior` to `ImportGranularity` as rustfmt has it as the purpose of them are basically the same. `ImportGranularity::Preserve` currently has no specific purpose for us as we don't have an organize imports assist yet, so it currently acts the same as `ImportGranularity::Item`.
We now try to guess the import style on a per file basis and fall back to the user granularity setting if the file has no specific style yet or where it is ambiguous. This can be turned off by setting `import.enforceGranularity` to `true`.