mikros/rust
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
16,063 Commits 2 Branches 0 Tags
Commit Graph

1158 Commits

Author SHA1 Message Date
Tim Chevalier
ef833d4101 Introduce a T_err type for type errors 
This allows more errors to be non-fatal, as per #1871.

I only went through and started changing span_fatal to span_err in
check.rs. There are probably more errors that could be made
non-fatal. So if you see derived type errors appearing from now on,
file a bug!

r=graydon

Closes #1871
 2012-11-21 12:44:51 -08:00
Patrick Walton
318e534895 rustc: Implement explicit self for Eq and Ord. r=graydon 2012-11-19 15:33:11 -08:00
Brian Anderson
74b2e99797 Report errors better when failing to open files for sub-parsers 2012-11-18 18:09:41 -08:00
Brian Anderson
72cc1aca17 Parse file mods from .rs files 2012-11-18 18:09:41 -08:00
Tim Chevalier
b9b21b2614 Update test and un-xfail it 2012-11-15 19:36:48 -08:00
Tim Chevalier
9d6ab9ae23 Update error message and un-xfail test 2012-11-15 19:36:48 -08:00
Tim Chevalier
39c306eeef Un-xfail working tests 2012-11-15 19:36:44 -08:00
Tim Chevalier
2fd2a56063 Update error message and un-xfail test 2012-11-15 18:41:17 -08:00
Tim Chevalier
f8bd95589f Add comments explaining why these tests are xfailed 2012-11-15 18:41:17 -08:00
Tim Chevalier
d4cc7db138 Un-xfail working test 2012-11-15 18:41:17 -08:00
Tim Chevalier
2f83baf857 Check for duplicate supertraits and forbid them 
As per #3953
 2012-11-15 18:41:17 -08:00
Tim Chevalier
82017b8416 Make staticness mismatch a fatal error 
Exit with a fatal error, instead of recording a non-fatal error,
when we encounter an impl method that's static when its corresponding
trait method isn't (or vice versa). This is because code later on in
the typechecker will expect the staticness of the two methods to be
consistent and ICE otherwise.

r=nmatsakis

Closes #3969
 2012-11-15 18:14:13 -08:00
Tim Chevalier
cb355bf7ad Update error message and un-xfail test 2012-11-15 17:14:28 -08:00
Tim Chevalier
fb47e41411 Un-xfail working tests 2012-11-15 16:59:07 -08:00
Tim Chevalier
6b65aed350 Remove test that tested behavior of unchecked blocks 
We don't have unchecked blocks anymore.
 2012-11-15 16:59:07 -08:00
Tim Chevalier
d58e0e30ed Update error messages and un-xfail 2012-11-15 16:59:07 -08:00
Tim Chevalier
35b2960d00 Add another test for #3177 2012-11-15 16:59:07 -08:00
Tim Chevalier
1f81e4cbc6 Remove typestate test 2012-11-15 16:59:07 -08:00
Patrick Walton
adc4bed773 librustc: Implement (and require) explicit self for derivable traits. r=nmatsakis 2012-11-15 10:26:49 -08:00
Ben Striegel
f4a5a76aa4 Convert the test suite to use the Drop trait 2012-11-14 19:26:37 -08:00
Brian Anderson
8336cad95d Move two fn assignability tests from compile-fail to run-pass 2012-11-14 12:40:23 -08:00
Brian Anderson
f14240b1c8 Remove obsolete const pattern test 2012-11-14 11:43:28 -08:00
Patrick Walton
32ad4ae4cd librustc: Require the #[derivable] attribute, remove the significance of "impl Foo : Bar;", and allow only a subset of methods in a trait to be derived. r=brson 2012-11-14 11:36:55 -08:00
Brian Anderson
cf002e9d32 Tools should not be build prereqs of tests. Closes #3946 2012-11-13 18:35:02 -08:00
Graydon Hoare
4e1ce014c8 cleanup: convert some remaining #foo invocations to foo! form. 2012-11-13 08:57:31 -08:00
Patrick Walton
fe02814a63 rustc: Implement floating point literal inference. r=nmatsakis 2012-11-12 10:39:08 -08:00
Brian Anderson
f05e2da709 Add a test for multiple main functions 2012-11-09 12:13:55 -08:00
Patrick Walton
0fc952372a rustc: Support irrefutable patterns in function arguments. r=nmatsakis 2012-11-07 19:29:30 -08:00
Patrick Walton
2904095570 rustc: Implement the Drop trait. r=brson 2012-11-07 12:34:32 -08:00
Niko Matsakis
b0ed151539 Cleanup how we handle proto in types, remove unsound subtyping 
Fixes #1896 which was never truly fixed, just masked.
The given tests would have failed had they used `~fn()` and
not `@fn()`.  They now result in compilation errors.

Fixes #2978.

Necessary first step for #2202, #2263.
 2012-11-06 08:56:29 -08:00
Patrick Walton
be93b29d30 rustc: Implement parsing and typechecking for "once fn" 2012-11-05 13:17:02 -08:00
Patrick Walton
caf68038dd test: Fix error messages in compile-fail tests. rs=rustbot 2012-11-02 10:43:28 -07:00
Patrick Walton
0c2e6fda73 rustc: Implement ~Trait. r=nmatsakis 2012-11-02 10:15:19 -07:00
Patrick Walton
106f9976ab rustc: Implement typechecking, exhaustiveness checking, and borrow checking for pattern matching of tuple structs. r=nmatsakis 
Conflicts:

	src/rustc/middle/typeck/check/alt.rs
 2012-11-02 09:56:09 -07:00
Tim Chevalier
768247f393 Tests for #1896 2012-11-01 15:14:52 -07:00
Brian Anderson
cb4de738cf Fix checking of duplicate and missing struct field initializers. Closes #3486. Closes #3892 2012-10-31 13:56:41 -07:00
Tim Chevalier
165ce14f68 Remove xfail-pretty from tests that pretty-print correctly now 2012-10-30 15:06:10 -07:00
Tim Chevalier
62f98c8ff8 Preserve parenthesization in the AST 
Maintain explicit "paren" nodes in the AST so we can pretty-print
without having to guess where parens should go. We may revisit this
in the future.

r=graydon
 2012-10-30 15:05:32 -07:00
Patrick Walton
a369a7881f rustc: Implement typechecking for automatically-derived enums 2012-10-29 14:11:56 -07:00
Tim Chevalier
5e5ea04608 Make class-cast-to-trait compile (not sure why this worked before) 2012-10-27 17:38:57 -07:00
Niko Matsakis
2093952847 Partial fix for #2687---impl method must only be subtype of trait method, not exact match. 2012-10-26 19:41:17 -07:00
Patrick Walton
57cd6b3e3f rustc: Translate and check exhaustiveness of struct-like enum variant patterns. r=nmatsakis 2012-10-25 13:59:10 -07:00
Niko Matsakis
1a3a70760b Implement proper subtyping for region fn types (part of #2263) 2012-10-24 18:56:31 -07:00
Viktor Dahl
0707780eac Added compile-fail tests 2012-10-24 19:14:50 +02:00
Tim Chevalier
11e92f37c1 Remove uses of binary move - <- - from tests and libraries 2012-10-23 12:10:03 -07:00
Tim Chevalier
46d4bbbae4 Simplify the AST representation of ty param bounds 
Change ast::ty_param_bound so that all ty param bounds are represented
as traits, with no special cases for Copy/Send/Owned/Const.
typeck::collect generates the special cases.

A consequence of this is that code using the #[no_core] attribute
can't use the Copy kind/trait. Probably not a big deal?

As a side effect, any user-defined traits that happen to be called
Copy, etc. in the same module override the built-in Copy trait.

r=nmatsakis

Closes #2284
 2012-10-22 09:01:12 -07:00
Niko Matsakis
42c05fe642 Correct propagation of mutability from components to base in borrowck 
Fixes #3828.
 2012-10-21 23:52:36 -07:00
Brian Anderson
837875711a Remove obsolete fixed-length string test 2012-10-20 17:50:46 -07:00
Brian Anderson
f3df50f67f Parse and report obsolete fixed-length vector syntax 2012-10-20 17:50:46 -07:00
Ben Striegel
ac81fff229 Remove old fixed-length vector syntax 2012-10-20 17:50:46 -07:00