Commit graph

193 commits

Author SHA1 Message Date
Aleksey Kladov
1834bae5b8 allow rustfmt to reorder imports
This wasn't a right decision in the first place, the feature flag was
broken in the last rustfmt release, and syntax highlighting of imports
is more important anyway
2019-07-04 23:09:09 +03:00
Florian Diebold
96c2b9c41d Simplifications / cleanup from review 2019-06-16 12:25:04 +02:00
Florian Diebold
3e78a6e3e0 Somewhat handle variables in the derefed type, and add another test 2019-06-15 18:21:23 +02:00
Florian Diebold
9c5e7dd849 Implement autoderef using the Deref trait
- add support for other lang item targets, since we need the Deref lang item
2019-06-15 18:21:23 +02:00
Florian Diebold
49489dc20c Add basic infrastructure for assoc type projection 2019-06-15 18:21:23 +02:00
Alan Du
b28ca32db2 Fix clippy::or_fun_call 2019-06-04 18:05:07 -04:00
Aleksey Kladov
5d54aa6781 add union to code_model 2019-05-23 20:18:47 +03:00
Aleksey Kladov
f63be06002 profile type inference 2019-05-21 16:26:13 +03:00
Edwin Cheng
d4dc879415 Add infer for generic default type 2019-05-19 21:08:16 +08:00
Florian Diebold
50bbf9eb09 Handle where clauses in trait solving 2019-05-11 16:21:20 +02:00
Florian Diebold
4083caa233 Add a HirDisplay implementation for TraitRef 2019-05-07 13:45:29 +02:00
Florian Diebold
a78228a39c Turn implements into a query again 2019-05-07 13:45:29 +02:00
Florian Diebold
a4eb1a546c Differentiate Tuple / FnPtr type constructors by cardinality
This is necessary because Chalk (reasonably) expects each 'struct' to know how
many type parameters it takes.
2019-05-04 19:11:21 +02:00
Florian Diebold
19fbd91998 Simplify subst / subst_bound_vars a bit 2019-05-04 18:41:48 +02:00
Florian Diebold
0ad7317b24 Canonicalize before doing method resolution 2019-05-04 18:18:30 +02:00
Florian Diebold
99492278ac Implement Deref<Target=[Ty]> for Substs 2019-05-04 18:18:30 +02:00
Florian Diebold
b9c0c2abb7 Chalk integration
- add proper canonicalization logic
 - add conversions from/to Chalk IR
2019-05-04 18:18:30 +02:00
Florian Diebold
6269791d36 Add Ty::Bound variant for use in Chalk integration 2019-05-04 18:18:30 +02:00
Florian Diebold
7650a44640 Make callable signature handling a bit nicer 2019-04-14 11:28:53 +02:00
Florian Diebold
a1ed53a4f1 More trait infrastructure
- make it possible to get parent trait from method
 - add 'obligation' machinery for checking that a type implements a
   trait (and inferring facts about type variables from that)
 - handle type parameters of traits (to a certain degree)
 - improve the hacky implements check to cover enough cases to exercise the
   handling of traits with type parameters
 - basic canonicalization (will probably also be done by Chalk)
2019-04-14 11:28:53 +02:00
Aleksey Kladov
505acc973b Make call info to use real name resolution 2019-04-11 16:29:33 +03:00
Lenard Pratt
e175921932 Added ArrayExprKind,
changed the  display for fixed array types,
Added Array Enum to ra_hir/expr
2019-04-07 13:23:14 +01:00
Lenard Pratt
2d73c909fe Added inference of array length 2019-04-07 13:23:12 +01:00
Florian Diebold
0f7e4a7d24 Implement a very naive implements check
... to make the infer_trait_method_simple test have the correct result.
2019-03-25 21:28:36 +01:00
Florian Diebold
1ee779d1f7 Assert in apply_substs that the number of parameters doesn't change
... and fix a small bug revealed by that.
2019-03-21 22:39:31 +01:00
Florian Diebold
cbb418ebb8 Rename name field to ctor as well 2019-03-21 22:29:12 +01:00
Florian Diebold
97be0e6c46 Some more doc comments 2019-03-21 22:23:52 +01:00
Florian Diebold
f10f5a81b3 TypeName => TypeCtor 2019-03-21 22:20:03 +01:00
Florian Diebold
8a5fbf4713 Remove the old variants replaced by Ty::Apply 2019-03-21 22:17:00 +01:00
Florian Diebold
7a8ba53542 Add Ty::Apply 2019-03-21 22:11:53 +01:00
Florian Diebold
bc7752e527 Represent FnPtr and Tuple using Substs 2019-03-21 22:11:53 +01:00
Florian Diebold
7faae12311 Remove FnSig from FnDef type
It doesn't need to be in there since it's just information from the def. Another
step towards aligning Ty with Chalk's representation.
2019-03-16 17:29:55 +01:00
Florian Diebold
a9ddaba905 Refactor FnSig a bit 2019-03-16 17:21:32 +01:00
Florian Diebold
628b530e92 Some more Ty displaying cleanup 2019-03-16 16:50:31 +01:00
Florian Diebold
c5ee60e05b Replace Display by a pretty printing trait for Ty
This allows removing the names from Adt and FnDef (and more later), as a first
step towards aligning more with chalk's Ty :)
2019-03-16 16:36:59 +01:00
Florian Diebold
affaf7700a Represent unknown types as {unknown} instead of [unknown]
Since the latter could actually be a real type...
2019-03-03 03:00:17 +01:00
Florian Diebold
862b663501 A bit of cleanup in ty.rs 2019-03-02 14:53:12 +01:00
Florian Diebold
dcfb4ee702 Split ty.rs into several modules
It was just getting too big. We now have:

 - ty: the `Ty` enum and helpers
 - ty::infer: actual type inference
 - ty::lower: lowering from HIR to `Ty`
 - ty::op: helpers for binary operations, currently
2019-02-23 15:36:38 +01:00
Florian Diebold
6a04d1f292 Fix resolution of associated method calls across crates
I think it'll be better to make the path resolution the number of unresolved
segments, not the first unresolved index; then this error could simply not have
happened. But I'll do that separately.
2019-02-23 12:37:29 +01:00
bors[bot]
3d8a0982a1 Merge #866
866: Implement basic support for Associated Methods r=flodiebold a=vipentti

This is my attempt at learning to understand how the type inference works by adding basic support for associated methods. Currently it does not resolve associated types or constants. 

The basic idea is that `Resolver::resolve_path` returns a new `PathResult` type, which has two variants, `FullyResolved` and `PartiallyResolved`, fully resolved matches the previous behavior, where as `PartiallyResolved` contains the `PerNs<Resolution` in addition to a `segment_index` which contains the index of the segment which we failed to resolve. This index can then be used to continue inference in `infer_path_expr` using the `Type` we managed to resolve.

This changes some of the previous apis, so looking for feedback and suggestions.

This should enable fixing #832

Co-authored-by: Ville Penttinen <villem.penttinen@gmail.com>
2019-02-22 19:58:22 +00:00
Ville Penttinen
247d1c17b3 Change resolve_path to return the fully resolved path or PerNs::none
This also adds new pub(crate) resolve_path_segments which returns the
PathResult, which may or may not be fully resolved. PathResult is also now
pub(crate) since it is an implementation detail.
2019-02-22 10:15:23 +02:00
Ville Penttinen
2e7bc905be Remove Const inference for now, refactor PathResult 2019-02-21 23:57:07 +02:00
Florian Diebold
db9a5a9ac0 Add an assert (and fix the other) 2019-02-21 21:25:27 +01:00
Ville Penttinen
816971ebc9 Implement basic support for Associated Methods and Constants
This is done in `infer_path_expr`. When `Resolver::resolve_path` returns
`PartiallyResolved`, we use the returned `Resolution` together with the given
`segment_index` to check if we can find something matching the segment at
segment_index in the impls for that particular type.
2019-02-21 12:25:55 +02:00
Florian Diebold
72712b8a42 Fix handling of generics in tuple variants and refactor a bit
Also make them display a tiny bit nicer.

Fixes #860.
2019-02-20 22:48:55 +01:00
bors[bot]
564ab84b78 Merge #852
852: Handle != r=flodiebold a=matklad

r? @flodiebold

Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2019-02-18 09:13:51 +00:00
Aleksey Kladov
4e8a3f565b handle != operator 2019-02-18 10:09:44 +03:00
Florian Diebold
b82db68400 Handle tuple structs / enum variants properly in type inference 2019-02-17 20:53:13 +01:00
Florian Diebold
795d718ba1 Unify with the autorefed/autoderefed receiver type during method resolution 2019-02-17 14:44:39 +01:00
Florian Diebold
a1bda3fc08 Handle generic args for method calls 2019-02-16 23:06:41 +01:00