Centralize and clean type error reporting #34907

arielb1 · 2016-07-18T22:12:25Z

Refactors the code that handles type errors to be cleaner and fixes various edge cases.

This made the already-bad "type mismatch resolving" error message somewhat uglier. I want to fix that in another commit before this PR is merged.

Fixes #31173

r? @jonathandturner, cc @nikomatsakis

arielb1 · 2016-07-18T22:46:29Z

Few more errors to modernize:

expected constant integer for repeat count, but mismatched types
array length constant evaluation error: mismatched types
implemented const has an incompatible type for trait

Will fix these tomorrow.

arielb1 · 2016-07-18T22:47:38Z

All type errors now look like:

// newskool
error: mismatched types [--explain E0308]
 --> ../build-debug-assertions/fail.rs:3:32
  |>
3 |>     let x: Option<Option<_>> = y;
  |>                                ^ expected enum `std::option::Option`, found enum `std::result::Result`
note: expected type `std::option::Option<std::option::Option<_>>`
note:    found type `std::option::Option<std::result::Result<_, ()>>`




// oldskool
../build-debug-assertions/fail.rs:3:32: 3:33 error: mismatched types [E0308]
../build-debug-assertions/fail.rs:3     let x: Option<Option<_>> = y;
                                                                   ^
../build-debug-assertions/fail.rs:3:32: 3:33 help: run `rustc --explain E0308` to see a detailed explanation
../build-debug-assertions/fail.rs:3:32: 3:33 note: expected type `std::option::Option<std::option::Option<_>>` 
../build-debug-assertions/fail.rs:3:32: 3:33 note:    found type `std::option::Option<std::result::Result<_, ()>>` 
../build-debug-assertions/fail.rs:3:32: 3:33 note: expected enum `std::option::Option`, found enum `std::result::Result` 
../build-debug-assertions/fail.rs:3     let x: Option<Option<_>> = y;

nikomatsakis · 2016-07-19T15:22:35Z

src/test/compile-fail/issue-13033.rs

@@ -16,7 +16,9 @@ struct Baz;

 impl Foo for Baz {
    fn bar(&mut self, other: &Foo) {}
-    //~^ ERROR method `bar` has an incompatible type for trait: values differ in mutability [E0053]


it's hard for me to tell from this diff: has the "values differ in mutability" part of the message been entirely removed?

See the error message above ^. It is now a label in new-skool and a single note in old-skool.

nikomatsakis · 2016-07-19T15:27:49Z

So I'm 👍 on removing the hack around newlines and centralizing the error reporting infrastructure here to go through a more common set of methods. That said, it seems like the new messages may be less specific and quite possibly less helpful than the old ones.

It might be though that if we combined this patch with something that gives intelligent diffs, that would be the best of both worlds.

Basically I think we want to avoid presenting people with two types and having them find the difference: so if we used to point out that the problem has to do with mutability, but now we make you figure it out for yourself, that could be a regression in usability. I also think that we calling everything "mismatched types" might be confused for people, if we can give a more specific indicator of where the type problem is.

Also, @arielb1, you mention this:

This made the already-bad "type mismatch resolving" error message somewhat uglier. I want to fix that in another commit before this PR is merged.

How does it look now? Can you give an example of what got uglier?

arielb1 · 2016-07-20T23:16:40Z

ready for review. I would like @nikomatsakis to look at the last commit (efe1a68).

arielb1 · 2016-07-21T15:15:11Z

added a commit to fix the RFC1592 warnings (#33242, #33243) - they are there since 1.10 and I am planning to make more modifications to traits::error_reporting.

Unfortunately, projection errors do not come with a nice set of mismatched types. This is because the type equality check occurs within a higher-ranked context. Therefore, only the type error is reported. This is ugly but was always the situation. I will introduce better errors for the lower-ranked case in another commit. Fixes the last known occurence of rust-lang#31173

Refactor constant evaluation to use a single error reporting function that reports a type-error-like message. Also, unify all error codes with the "constant evaluation error" message to just E0080, and similarly for a few other duplicate codes. The old situation was a total mess, and now that we have *something* we can further iterate on the UX.

The type equation in projection takes place under a binder and a snapshot, which we can't easily take types out of. Instead, when encountering a projection error, try to re-do the projection and find the type error then. This fails to produce a sane type error when the failure was a "leak_check" failure. I can't think of a sane way to show *these*, so I just left them use the old crappy representation, and added a test to make sure we don't break them.

arielb1 · 2016-07-24T07:59:50Z

@bors r=nikomatsakis

bors · 2016-07-24T07:59:52Z

📌 Commit 717e392 has been approved by nikomatsakis

arielb1 · 2016-07-24T08:00:19Z

@bors retry

arielb1 · 2016-07-24T17:12:05Z

@bors r-

arielb1 · 2016-07-24T17:12:18Z

@bors retry r=nikomatsakis

bors · 2016-07-24T17:12:19Z

📌 Commit 717e392 has been approved by nikomatsakis

bors · 2016-07-25T04:37:11Z

⌛ Testing commit 717e392 with merge d398d9c...

bors · 2016-07-25T05:08:08Z

💔 Test failed - auto-linux-cross-opt

alexcrichton · 2016-07-25T05:10:58Z

@bors: retry

weeps

On Sun, Jul 24, 2016 at 10:08 PM, bors notifications@github.com wrote:

💔 Test failed - auto-linux-cross-opt
https://buildbot.rust-lang.org/builders/auto-linux-cross-opt/builds/3104

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#34907 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAD95PxFHLDN_ToYfmLuFhGWjiCn64Olks5qZES7gaJpZM4JPMZs
.

bors · 2016-07-25T05:37:04Z

⌛ Testing commit 717e392 with merge eccf954...

bors · 2016-07-25T05:43:21Z

💔 Test failed - auto-win-gnu-32-opt-rustbuild

alexcrichton · 2016-07-25T06:42:23Z

@bors: retry

On Sun, Jul 24, 2016 at 10:43 PM, bors notifications@github.com wrote:

💔 Test failed - auto-win-gnu-32-opt-rustbuild
https://buildbot.rust-lang.org/builders/auto-win-gnu-32-opt-rustbuild/builds/1924

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#34907 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAD95HAJjINvrKQNlfGpLA7U_uCqR80kks5qZEz8gaJpZM4JPMZs
.

alexcrichton · 2016-07-25T18:38:16Z

@bors: retry

bors · 2016-07-25T21:49:06Z

⌛ Testing commit 717e392 with merge 084ebe7...

bors · 2016-07-25T21:55:01Z

💔 Test failed - auto-win-gnu-32-opt

alexcrichton · 2016-07-25T22:14:30Z

@bors: retry

On Mon, Jul 25, 2016 at 2:55 PM, bors notifications@github.com wrote:

💔 Test failed - auto-win-gnu-32-opt
https://buildbot.rust-lang.org/builders/auto-win-gnu-32-opt/builds/5069

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#34907 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAD95Es23zS8rmNWCtgWwy2c_g4tY609ks5qZTC5gaJpZM4JPMZs
.

bors · 2016-07-26T10:09:06Z

⌛ Testing commit 717e392 with merge 727c4dc...

bors · 2016-07-26T10:13:43Z

💔 Test failed - auto-win-msvc-64-opt

alexcrichton · 2016-07-26T17:20:24Z

@bors: retry

On Tue, Jul 26, 2016 at 3:13 AM, bors notifications@github.com wrote:

💔 Test failed - auto-win-msvc-64-opt
https://buildbot.rust-lang.org/builders/auto-win-msvc-64-opt/builds/5067

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#34907 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAD95D0fOiaZG2vDG9bH59DTjnEGTYosks5qZd3ZgaJpZM4JPMZs
.

bors · 2016-07-27T21:28:07Z

⌛ Testing commit 717e392 with merge f2e59cc...

@nikomatsakis

Centralize and clean type error reporting Refactors the code that handles type errors to be cleaner and fixes various edge cases. This made the already-bad "type mismatch resolving" error message somewhat uglier. I want to fix that in another commit before this PR is merged. Fixes #31173 r? @jonathandturner, cc @nikomatsakis

bors · 2016-07-28T00:35:13Z

rust-highfive assigned sophiajt Jul 18, 2016

arielb1 force-pushed the found-parse-error branch from e09f9d0 to fdf8335 Compare July 18, 2016 22:16

nikomatsakis reviewed Jul 19, 2016
View reviewed changes

arielb1 force-pushed the found-parse-error branch 2 times, most recently from a44b5bf to 7966739 Compare July 20, 2016 22:15

arielb1 force-pushed the found-parse-error branch 8 times, most recently from 32fd204 to 5b2cf69 Compare July 21, 2016 15:14

arielb1 mentioned this pull request Jul 21, 2016

Non existent field err code #34954

Closed

arielb1 force-pushed the found-parse-error branch from ad37be1 to 031bc32 Compare July 21, 2016 23:35

arielb1 added 8 commits July 22, 2016 14:32

remove never-called type-error reporting functions

b2422ab

remove rustc_typeck::same_type_err

8eb12d9

refactor type error reporting

cea88eb

switch compare_method to new-style trait error reporting

b7b2db4

remove the now-unused multiline error code

712c5ca

arielb1 force-pushed the found-parse-error branch from 031bc32 to 521ada1 Compare July 22, 2016 12:08

bors merged commit 717e392 into rust-lang:master Jul 28, 2016

This was referenced Jul 28, 2016

extended information for slice pattern errors (E0527 through E0529) #34319

Merged

Enable reuse of .o files if nothing has changed #34956

Merged

run rustfmt on librustc_passes folder #33951

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Centralize and clean type error reporting #34907

Centralize and clean type error reporting #34907

arielb1 commented Jul 18, 2016

arielb1 commented Jul 18, 2016

arielb1 commented Jul 18, 2016 •

edited

Loading

nikomatsakis Jul 19, 2016

arielb1 Jul 19, 2016

nikomatsakis commented Jul 19, 2016

arielb1 commented Jul 20, 2016 •

edited

Loading

arielb1 commented Jul 21, 2016 •

edited

Loading

arielb1 commented Jul 24, 2016

bors commented Jul 24, 2016

arielb1 commented Jul 24, 2016

arielb1 commented Jul 24, 2016

arielb1 commented Jul 24, 2016

bors commented Jul 24, 2016

bors commented Jul 25, 2016

bors commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

bors commented Jul 25, 2016

bors commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

bors commented Jul 25, 2016

bors commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

bors commented Jul 26, 2016

bors commented Jul 26, 2016

alexcrichton commented Jul 26, 2016

bors commented Jul 27, 2016

bors commented Jul 28, 2016

Centralize and clean type error reporting #34907

Centralize and clean type error reporting #34907

Conversation

arielb1 commented Jul 18, 2016

arielb1 commented Jul 18, 2016

arielb1 commented Jul 18, 2016 • edited Loading

nikomatsakis Jul 19, 2016

Choose a reason for hiding this comment

arielb1 Jul 19, 2016

Choose a reason for hiding this comment

nikomatsakis commented Jul 19, 2016

arielb1 commented Jul 20, 2016 • edited Loading

arielb1 commented Jul 21, 2016 • edited Loading

arielb1 commented Jul 24, 2016

bors commented Jul 24, 2016

arielb1 commented Jul 24, 2016

arielb1 commented Jul 24, 2016

arielb1 commented Jul 24, 2016

bors commented Jul 24, 2016

bors commented Jul 25, 2016

bors commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

bors commented Jul 25, 2016

bors commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

bors commented Jul 25, 2016

bors commented Jul 25, 2016

alexcrichton commented Jul 25, 2016

bors commented Jul 26, 2016

bors commented Jul 26, 2016

alexcrichton commented Jul 26, 2016

bors commented Jul 27, 2016

bors commented Jul 28, 2016

arielb1 commented Jul 18, 2016 •

edited

Loading

arielb1 commented Jul 20, 2016 •

edited

Loading

arielb1 commented Jul 21, 2016 •

edited

Loading