Implement non-mod.rs mod statements #46531

cramertj · 2017-12-06T02:50:24Z

This will fail tidy right now because it doesn't recognize my UI tests as feature-gate tests. However, I'm not sure if compile-fail will work out either because compile-fail usually requires there to be error patterns in the top-level file, which isn't possible with this feature. What's the recommended way to handle this?

rust-highfive · 2017-12-06T02:50:35Z

r? @pnkfelix

(rust_highfive has picked a reviewer for you, use r? to override)

zackmdavis · 2017-12-06T06:51:04Z

because compile-fail usually requires there to be error patterns in the top-level file, which isn't possible with this feature. What's the recommended way to handle this?

There's a special rustc_error attribute (gated by rustc_attrs) that you can use to make the test artificially fail.

est31 · 2017-12-06T07:10:24Z

This will fail tidy right now because it doesn't recognize my UI tests as feature-gate tests.

You can also add the ui directory to the places where tidy searches for feature gate tests. Look into src/tools/tidy/src/features.rs. Then you need to name the files after the format that tidy recognizes.

nikomatsakis · 2017-12-06T15:01:15Z

There's a special rustc_error attribute (gated by rustc_attrs) that you can use to make the test artificially fail.

You can use // must-compile-successfully, which I believe is now the preferred approach.

nikomatsakis · 2017-12-06T15:01:45Z

That said, what @est31 said sounds better. =)

nikomatsakis · 2017-12-06T15:04:35Z

r? @nikomatsakis

nikomatsakis · 2017-12-06T18:13:50Z

src/libsyntax/parse/mod.rs

-    Owned,
+    Owned {
+        // None if `mod.rs`, `Some(path)` if we're in
+        // `foo.rs` and there's a corresponding `foo/...`


This comment makes it sound like we check for the existence of such a directory, but we don't right? Maybe "Some("foo") if we're in the top-level of foo.rs"?

est31 · 2017-12-06T18:17:07Z

btw this PR motivated me to move over some of the currently existing feature gate tests: #46532

nikomatsakis · 2017-12-06T18:24:23Z

src/libsyntax/feature_gate.rs

+        for &span in &*self.context.parse_sess.non_modrs_mods.borrow() {
+            gate_feature_post!(
+                &self, non_modrs_mods, span,
+                "mod declarations in non-mod.rs files are experimental");


maybe we want to add some extra notes here, given that this is a common source of confusion? Something like:

on stable builds, rename foo.rs to foo/mod.rs to enable submodules

not sure how hard that would be to do

nikomatsakis

looks pretty good to me -- I'd like to see a ui test for what happens when there is a mod foo with no matching file (what kind of error do we give)

cramertj · 2017-12-07T20:05:43Z

@nikomatsakis

I'd like to see a ui test

I've added one, but currently it just suggests to try creating the file, and once the file is created it will give you the stability error.

nikomatsakis · 2017-12-07T21:53:39Z

src/test/ui/invalid-module-declaration/invalid-module-declaration.stderr

-   |
-11 | pub mod baz;
-   |         ^^^
+   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "$DIR/auxiliary/foo"


nikomatsakis · 2017-12-07T21:54:07Z

@bors r+

bors · 2017-12-07T21:54:08Z

📌 Commit 5a821e3 has been approved by nikomatsakis

bors · 2017-12-08T07:49:16Z

⌛ Testing commit 5a821e3a7738ee9d04ac1fd8c91956309f425f7d with merge c41ec675c6e876c7f032cb9179af6750638f65be...

bors · 2017-12-08T08:47:56Z

💔 Test failed - status-appveyor

cramertj · 2017-12-08T08:57:27Z

failures:
---- [ui] ui\invalid-module-declaration\invalid-module-declaration.rs stdout ----
	normalized stderr:
error[E0583]: file not found for module `baz`
  --> $DIR/auxiliary/foo/bar.rs:11:9
   |
11 | pub mod baz;
   |         ^^^
   |
   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "C:/projects/rust/src/test/ui/invalid-module-declaration/auxiliary/foo"
error: aborting due to previous error
expected stderr:
error[E0583]: file not found for module `baz`
  --> $DIR/auxiliary/foo/bar.rs:11:9
   |
11 | pub mod baz;
   |         ^^^
   |
   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "$DIR/auxiliary/foo"
error: aborting due to previous error
diff of stderr:
 error[E0583]: file not found for module `baz`
   --> $DIR/auxiliary/foo/bar.rs:11:9
    |
 11 | pub mod baz;
    |         ^^^
    |
-   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "$DIR/auxiliary/foo"
+   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "C:/projects/rust/src/test/ui/invalid-module-declaration/auxiliary/foo"
 
 error: aborting due to previous error

Looks like the $DIR paths aren't handled correctly on Windows. Anyone know what's up here? The existing UI tests seem to all use $DIR with no problem. Is it maybe something to do with the fact that it's in quotes?

kennytm · 2017-12-08T14:47:53Z

@cramertj The quotes shouldn't matter. However, if the original output is C:/projects/rust/src/… instead of C:\projects\rust\src\… it won't perform the replacement.

cramertj · 2017-12-08T17:25:57Z

@kennytm Hm... it seems like the slashes are going the right direction, though. It's in MSYS. The diff is this part:

-   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "$DIR/auxiliary/foo"
+   = help: name the file either bar/baz.rs or bar/baz/mod.rs inside the directory "C:/projects/rust/src/test/ui/invalid-module-declaration/auxiliary/foo"

It looks like the replacement just isn't happening.

kennytm · 2017-12-08T17:56:14Z

@cramertj just to clarify, the original needs to be backslashes.

bors · 2017-12-10T21:35:07Z

⌛ Testing commit af87dd4131b44e5c0f94aacb3270d08794c9ed2f with merge 2b9c672a1d2cc46ceb916b6b3d164f754eb5a3b0...

bors · 2017-12-10T23:32:06Z

💔 Test failed - status-travis

cramertj · 2017-12-10T23:36:28Z

Failed pretty-printing tests:

[01:51:02] failures:
[01:51:02] 
[01:51:02] ---- [pretty] run-pass/non_modrs_mods/non_modrs_mods.rs stdout ----
[01:51:02] 	
[01:51:02] error: pretty-printed source does not typecheck
[01:51:02] status: exit code: 101
[01:51:02] command: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "-" "-Zno-trans" "--out-dir" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/run-pass/non_modrs_mods/non_modrs_mods.pretty-out" "--target=x86_64-unknown-linux-gnu" "-L" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/run-pass" "-L" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/run-pass/non_modrs_mods/non_modrs_mods.stage2-x86_64-unknown-linux-gnu.pretty.aux" "-Crpath" "-O" "-Lnative=/checkout/obj/build/x86_64-unknown-linux-gnu/native/rust-test-helpers"
[01:51:02] stdout:
[01:51:02] ------------------------------------------
[01:51:02] 
[01:51:02] ------------------------------------------
[01:51:02] stderr:
[01:51:02] ------------------------------------------
[01:51:02] error[E0433]: failed to resolve. Could not find `inner_modrs_mod` in `modrs_mod`
[01:51:02]   --> <anon>:19:16
[01:51:02]    |
[01:51:02] 19 |     modrs_mod::inner_modrs_mod::innest::foo();
[01:51:02]    |                ^^^^^^^^^^^^^^^ Could not find `inner_modrs_mod` in `modrs_mod`
[01:51:02] 
[01:51:02] error[E0433]: failed to resolve. Could not find `inner_foors_mod` in `modrs_mod`
[01:51:02]   --> <anon>:20:16
[01:51:02]    |
[01:51:02] 20 |     modrs_mod::inner_foors_mod::innest::foo();
[01:51:02]    |                ^^^^^^^^^^^^^^^ Could not find `inner_foors_mod` in `modrs_mod`
[01:51:02] 
[01:51:02] error[E0433]: failed to resolve. Could not find `inner_modrs_mod` in `foors_mod`
[01:51:02]   --> <anon>:21:16
[01:51:02]    |
[01:51:02] 21 |     foors_mod::inner_modrs_mod::innest::foo();
[01:51:02]    |                ^^^^^^^^^^^^^^^ Could not find `inner_modrs_mod` in `foors_mod`
[01:51:02] 
[01:51:02] error[E0433]: failed to resolve. Could not find `inner_foors_mod` in `foors_mod`
[01:51:02]   --> <anon>:22:16
[01:51:02]    |
[01:51:02] 22 |     foors_mod::inner_foors_mod::innest::foo();
[01:51:02]    |                ^^^^^^^^^^^^^^^ Could not find `inner_foors_mod` in `foors_mod`
[01:51:02] 
[01:51:02] error[E0433]: failed to resolve. Could not find `inner_modrs_mod` in `attr_mod`
[01:51:02]   --> <anon>:23:15
[01:51:02]    |
[01:51:02] 23 |     attr_mod::inner_modrs_mod::innest::foo();
[01:51:02]    |               ^^^^^^^^^^^^^^^ Could not find `inner_modrs_mod` in `attr_mod`
[01:51:02] 
[01:51:02] error: aborting due to 5 previous errors
[01:51:02] 
[01:51:02] 
[01:51:02] ------------------------------------------
[01:51:02] 
[01:51:02] thread '[pretty] run-pass/non_modrs_mods/non_modrs_mods.rs' panicked at 'explicit panic', /checkout/src/tools/compiletest/src/runtest.rs:2770:8
[01:51:02] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[01:51:02] 
[01:51:02] 
[01:51:02] failures:
[01:51:02]     [pretty] run-pass/non_modrs_mods/non_modrs_mods.rs
[01:51:02] 
[01:51:02] test result: �[31mFAILED�(B�[m. 2810 passed; 1 failed; 40 ignored; 0 measured; 0 filtered out

nikomatsakis · 2017-12-11T15:03:20Z

@cramertj do you know what might be causing that? (I don't know, not without looking into it)

cramertj · 2017-12-11T17:46:49Z

No, but I'll investigate this afternoon.

bors · 2017-12-13T04:22:15Z

☔ The latest upstream changes (presumably #46613) made this pull request unmergeable. Please resolve the merge conflicts.

nikomatsakis · 2017-12-18T20:22:42Z

After some discussion on IRC -- our hypothesis is that the pretty typeck runner is putting files in the wrong place. Most likely outcome is that we will skip pretty printing for this test, but @cramertj is going to confirm the hypothesis.

cramertj · 2017-12-19T23:25:11Z

Rebased and ignored pretty-printing, as this is issue #37195.

nikomatsakis · 2017-12-20T20:29:21Z

@bors r+

bors · 2017-12-20T20:29:22Z

📌 Commit 1d5977b has been approved by nikomatsakis

bors · 2017-12-21T05:24:52Z

⌛ Testing commit 1d5977b with merge 957dc8d...

Implement non-mod.rs mod statements Fixes #45385, cc #44660 This will fail tidy right now because it doesn't recognize my UI tests as feature-gate tests. However, I'm not sure if compile-fail will work out either because compile-fail usually requires there to be error patterns in the top-level file, which isn't possible with this feature. What's the recommended way to handle this?

bors · 2017-12-21T08:04:51Z

☀️ Test successful - status-appveyor, status-travis
Approved by: nikomatsakis
Pushing 957dc8d to master...

rust-highfive assigned pnkfelix Dec 6, 2017

kennytm added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Dec 6, 2017

rust-highfive assigned nikomatsakis and unassigned pnkfelix Dec 6, 2017

nikomatsakis reviewed Dec 6, 2017

View reviewed changes

cramertj force-pushed the no-mo-modrs branch from 232744c to 34e95b9 Compare December 7, 2017 20:03

cramertj force-pushed the no-mo-modrs branch from 34e95b9 to 5a821e3 Compare December 7, 2017 21:19

nikomatsakis reviewed Dec 7, 2017

View reviewed changes

nikomatsakis approved these changes Dec 7, 2017

View reviewed changes

kennytm added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Dec 8, 2017

cramertj added 4 commits December 19, 2017 14:58

Implement non-mod.rs mod statements

07f51fb

Break rustfmt

b82e2e9

Break rls

e3c2298

Ignore pretty printing tests for non_modrs_mods

1d5977b

cramertj force-pushed the no-mo-modrs branch from af87dd4 to 1d5977b Compare December 19, 2017 23:25

kennytm added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Dec 20, 2017

tcr mentioned this pull request Dec 21, 2017

Epoch.Next Tracking Issue #46889

Closed

bors merged commit 1d5977b into rust-lang:master Dec 21, 2017

This was referenced Dec 21, 2017

Add a feature gate for nested uses of impl Trait #46888

Merged

[auto-toolstate] Upload the toolstate result to an external git repository, and removes BuildExpectation #46554

Merged

briansmith mentioned this pull request Dec 22, 2017

file not found for module montgomery briansmith/ring#598

Closed

SergioBenitez mentioned this pull request Dec 22, 2017

module resolution regression on rustc 1.24.0-nightly (2017-12-21) #46936

Closed

cramertj mentioned this pull request Dec 28, 2017

legacy_directory_ownership future-compatibility warnings #37872

Closed

3 tasks

davidtwco mentioned this pull request Sep 8, 2018

invalid-module-declaration.rs test regression #53352

Closed

ehuss mentioned this pull request Aug 3, 2019

legacy_directory_ownership -> hard error #63241

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implement non-mod.rs mod statements #46531

Implement non-mod.rs mod statements #46531

cramertj commented Dec 6, 2017 •

edited

Loading

rust-highfive commented Dec 6, 2017

zackmdavis commented Dec 6, 2017

est31 commented Dec 6, 2017

nikomatsakis commented Dec 6, 2017

nikomatsakis commented Dec 6, 2017

nikomatsakis commented Dec 6, 2017

nikomatsakis Dec 6, 2017 •

edited

Loading

est31 commented Dec 6, 2017

nikomatsakis Dec 6, 2017

nikomatsakis left a comment

cramertj commented Dec 7, 2017

nikomatsakis Dec 7, 2017

nikomatsakis commented Dec 7, 2017

bors commented Dec 7, 2017

bors commented Dec 8, 2017

bors commented Dec 8, 2017

cramertj commented Dec 8, 2017

kennytm commented Dec 8, 2017

cramertj commented Dec 8, 2017

kennytm commented Dec 8, 2017

bors commented Dec 10, 2017

bors commented Dec 10, 2017

cramertj commented Dec 10, 2017

nikomatsakis commented Dec 11, 2017

cramertj commented Dec 11, 2017

bors commented Dec 13, 2017

nikomatsakis commented Dec 18, 2017

cramertj commented Dec 19, 2017

nikomatsakis commented Dec 20, 2017

bors commented Dec 20, 2017

bors commented Dec 21, 2017

bors commented Dec 21, 2017

Implement non-mod.rs mod statements #46531

Implement non-mod.rs mod statements #46531

Conversation

cramertj commented Dec 6, 2017 • edited Loading

rust-highfive commented Dec 6, 2017

zackmdavis commented Dec 6, 2017

est31 commented Dec 6, 2017

nikomatsakis commented Dec 6, 2017

nikomatsakis commented Dec 6, 2017

nikomatsakis commented Dec 6, 2017

nikomatsakis Dec 6, 2017 • edited Loading

Choose a reason for hiding this comment

est31 commented Dec 6, 2017

nikomatsakis Dec 6, 2017

Choose a reason for hiding this comment

nikomatsakis left a comment

Choose a reason for hiding this comment

cramertj commented Dec 7, 2017

nikomatsakis Dec 7, 2017

Choose a reason for hiding this comment

nikomatsakis commented Dec 7, 2017

bors commented Dec 7, 2017

bors commented Dec 8, 2017

bors commented Dec 8, 2017

cramertj commented Dec 8, 2017

kennytm commented Dec 8, 2017

cramertj commented Dec 8, 2017

kennytm commented Dec 8, 2017

bors commented Dec 10, 2017

bors commented Dec 10, 2017

cramertj commented Dec 10, 2017

nikomatsakis commented Dec 11, 2017

cramertj commented Dec 11, 2017

bors commented Dec 13, 2017

nikomatsakis commented Dec 18, 2017

cramertj commented Dec 19, 2017

nikomatsakis commented Dec 20, 2017

bors commented Dec 20, 2017

bors commented Dec 21, 2017

bors commented Dec 21, 2017

cramertj commented Dec 6, 2017 •

edited

Loading

nikomatsakis Dec 6, 2017 •

edited

Loading