Commit Graph

542 Commits

Author SHA1 Message Date
Jonas Schievink
278556f9fe Move ItemTree lowering into its own module 2020-06-24 16:53:16 +02:00
Jonas Schievink
f9e5ba7d40 Remove now-broken incremental test
The ItemTree tracks impl and trait items, so the CrateDefMap now needs
recomputing when those change
2020-06-24 16:53:16 +02:00
Jonas Schievink
0e2602f75e Remove raw item query 2020-06-24 16:53:16 +02:00
Jonas Schievink
7054e89d18 Fix test 2020-06-24 16:53:16 +02:00
Jonas Schievink
9d24412929 Move collector 2020-06-24 16:53:14 +02:00
Jonas Schievink
f9417bcf49 Draw the rest of the owl 2020-06-24 16:52:37 +02:00
Jonas Schievink
e24de2007f collector changes WIP 2020-06-24 16:52:37 +02:00
Jonas Schievink
696a7a5a41 lower traits 2020-06-24 16:52:37 +02:00
Jonas Schievink
34bc80650b Hook up query, add impls, lower moar 2020-06-24 16:52:37 +02:00
Aleksey Kladov
1d75d11d6c Start item tree 2020-06-24 16:52:37 +02:00
Paul Daniel Faria
43c6f6c039 Fix formatting issue with glob shadow test 2020-06-24 08:50:59 -04:00
Paul Daniel Faria
dfd79907b8 Use a macro in push_res to simplify code, simplify glob_shadowed_def assertion 2020-06-24 08:20:41 -04:00
Paul Daniel Faria
38af81f183 Account for updated module ids when determining whether a resolution is "changed" 2020-06-23 22:10:01 -04:00
Aleksey Kladov
d016cb4867 Use only one code-path for parsing fixtures
This removes leading newlines everywhere, shifting all ranges in tests
by one
2020-06-24 00:48:38 +02:00
Aleksey Kladov
6996ec860b Drop rarely used fixture functionality 2020-06-23 18:37:26 +02:00
Aleksey Kladov
dad1333b48 New VFS 2020-06-23 17:51:06 +02:00
Jonas Schievink
500d8b2f60 Handle bindings after @ in patterns 2020-06-21 15:19:38 +02:00
bors[bot]
04d64267de
Merge #4947
4947: Replace `impls_in_trait` query with smarter use of `CrateImplDefs` r=matklad a=jonas-schievink

`impls_in_trait` was allocating a whopping ~400 MB of RAM when running analysis-stats on r-a itself.

Remove it, instead adding a query that computes a summary `CrateImplDefs` map for all transitive dependencies. This can probably still be made more efficient, but this already reduces the peak memory usage by 25% without much performance impact on analysis-stats.

**Before**:

```
Total: 34.962107188s, 2083mb allocated 2141mb resident
   422mb ImplsForTraitQuery (deps)
   250mb CrateDefMapQueryQuery
   147mb MacroArgQuery
   140mb TraitSolveQuery (deps)
    68mb InferQueryQuery (deps)
    62mb ImplDatumQuery (deps)
```

**After**:

```
Total: 35.261100358s, 1520mb allocated 1569mb resident
   250mb CrateDefMapQueryQuery
   147mb MacroArgQuery
   144mb TraitSolveQuery (deps)
    68mb InferQueryQuery (deps)
    61mb ImplDatumQuery (deps)
    45mb BodyQuery
    45mb ImplDatumQuery
```

Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
2020-06-20 22:14:21 +00:00
Florian Diebold
584bddef0c Don't panic on crates depending on themselves
Fixes #3883.
2020-06-19 18:38:03 +02:00
Jonas Schievink
ebd8233b3e Replace impls_in_trait with CrateImplDefs 2020-06-19 01:29:34 +02:00
Aleksey Kladov
42be05a335 Simplify 2020-06-18 16:02:45 +02:00
Aleksey Kladov
3c72fc0573 Anchor file-system operations to the file, and not to the source root.
Anchoring to the SourceRoot wont' work if the path is absolute:

  #[path = "/tmp/foo.rs"]
  mod foo;

Anchoring to a file will.

However, we *should* anchor, instead of just producing an abs path.

I can imagine a situation where, for example, rust-analyzer processes
crates from different machines (or, for example, from in-memory git
branch), where the same absolute path in different crates might refer
to different files in the end!
2020-06-16 18:45:58 +02:00
Aleksey Kladov
52a220cece Deprecate hir::Path::from_ast 2020-06-15 10:55:48 +02:00
Jonas Schievink
0231e4ac77 find_path: return shorter paths for external items
If a containing module is already in scope, there's no need to
use the full path to the item.
2020-06-12 13:01:20 +02:00
Jonas Schievink
90331ea035 Make known paths use core instead of std 2020-06-11 16:23:20 +02:00
bors[bot]
32157d48f4
Merge #4843
4843: Don't guess macro expansion crate r=matklad a=matklad



bors r+
🤖

Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2020-06-11 10:18:53 +00:00
Aleksey Kladov
fac7b0e252 Don't guess macro expansion crate 2020-06-11 12:13:14 +02:00
Aleksey Kladov
d8a5d39c2d Make relevant_crates return a Set 2020-06-11 11:30:06 +02:00
Jonas Schievink
dd22657407 ImportMap: use IndexMap internally
It iterates in insertion order, which makes the ordering more
predictable.
2020-06-10 16:15:49 +02:00
Jonas Schievink
7e83ed99a8 Respect casing when searching for imports 2020-06-10 16:04:55 +02:00
Jonas Schievink
56c7145993 Limit import map queries 2020-06-10 12:40:33 +02:00
Jonas Schievink
bcf875f46a Clean up import_map.rs 2020-06-10 12:38:58 +02:00
Jonas Schievink
4bcf8c8c68 Add an FST index to ImportMap 2020-06-10 12:38:58 +02:00
bors[bot]
f133159ec0
Merge #4760
4760: Minimize FileLoader interface r=matklad a=matklad



bors r+
🤖

Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2020-06-05 15:23:22 +00:00
Aleksey Kladov
bbb40d7463 Minimize FileLoader interface 2020-06-05 17:22:56 +02:00
bors[bot]
4029628f15
Merge #4729 #4748
4729: Hover actions r=matklad a=vsrs

This PR adds a `hoverActions` LSP extension and a `Go to Implementations` action as an example:
![hover_actions_impl](https://user-images.githubusercontent.com/62505555/83335732-6d9de280-a2b7-11ea-8cc3-75253d062fe0.gif)



4748: Add an `ImportMap` and use it to resolve item paths in `find_path` r=matklad a=jonas-schievink

Removes the "go faster" queries I added in https://github.com/rust-analyzer/rust-analyzer/pull/4501 and https://github.com/rust-analyzer/rust-analyzer/pull/4506. I've checked this PR on the rustc code base and the assists are still fast.

This should fix https://github.com/rust-analyzer/rust-analyzer/issues/4515.

Note that this does introduce a change in behavior: We now always refer to items defined in external crates using paths through the external crate. Previously we could also use a local path (if for example the extern crate was reexported locally), as seen in the changed test. If that is undesired I can fix that, but the test didn't say why the previous behavior would be preferable.

Co-authored-by: vsrs <vit@conrlab.com>
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
Co-authored-by: Jonas Schievink <jonas.schievink@ferrous-systems.com>
2020-06-05 15:14:35 +00:00
Aleksey Kladov
bba374bab2 More direct signature for resolve_path 2020-06-05 15:07:30 +02:00
Aleksey Kladov
e63c00f100 Rename resolve_relative_path -> resolve_path
For things like `concant!(env!("OUT_DIR"))`, we need to support abs paths
2020-06-05 14:58:30 +02:00
Jonas Schievink
bc2d172995 Clarify when we visit modules multiple times 2020-06-05 14:24:51 +02:00
Jonas Schievink
2fb3d87bf7 impl Debug for ImportMap 2020-06-05 13:36:19 +02:00
Jonas Schievink
8395396782 Reorder imports 2020-06-05 13:15:16 +02:00
Jonas Schievink
5f23f8ca44 Make find_path_inner private again 2020-06-05 13:11:53 +02:00
Jonas Schievink
86fbd8cc2b defining_crate -> krate 2020-06-05 13:05:19 +02:00
Jonas Schievink
e0e9c6d1a4 Fix wrong comment 2020-06-05 13:04:35 +02:00
Jonas Schievink
3c496f7fa7 Use ImportMap in find_path, remove old queries 2020-06-04 19:33:01 +02:00
Jonas Schievink
d08c63cb9e Add an ImportMap 2020-06-04 19:33:01 +02:00
David Tolnay
4461796f33
Fix type inference failure when built with log/kv_unstable
This code is broken by an `impl From<kv::Error> for fmt::Error` in the
log crate when building in a codebase that has the log/kv_unstable
feature enabled.

    $ cargo check --manifest-path crates/ra_hir_def/Cargo.toml
        Checking ra_hir_def v0.1.0
        Finished dev [unoptimized] target(s) in 0.75s

    $ cargo check --manifest-path crates/ra_hir_def/Cargo.toml --features log/kv_unstable
        Checking ra_hir_def v0.1.0
    error[E0282]: type annotations needed for the closure `fn(&str) -> std::result::Result<(), _>`
       --> crates/ra_hir_def/src/path.rs:278:17
        |
    278 |                 f.write_str("::")?;
        |                 ^^^^^^^^^^^^^^^^^^ cannot infer type
        |
    help: give this closure an explicit return type without `_` placeholders
        |
    276 |         let mut add_segment = |s| -> std::result::Result<(), _> {
        |                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2020-06-03 19:06:02 -07:00
Aaron Loucks
85c4edb0af Consolidate documentation expansion and merging
Removes the duplicated `expand_doc_attrs` and `merge_doc_comments_and_attrs`
functions from `ra_ide` and exposes the same functionality via
`ra_hir::Documentation::from_ast`.
2020-06-03 06:54:41 -04:00
Aaron Loucks
5837acce53 Add basic hover and completion doc tests for macro generated items 2020-06-03 06:46:07 -04:00
Aaron Loucks
4c655c01f3 Enable hover and autocomplete docs on macro generated items 2020-06-03 06:46:07 -04:00