Don't leak the compiler's internal representation of scopes in error messages.

[eddyb]

Fixes #37884 (actually fixes #27942, which was made worse by #37412) by handling more node types.
Ideally we'd turn the unknown node type situations into ICEs and fix them as they show up in errors.
But we might want to backport this patch so I was less aggressive.

[rust-highfive]

r? @pnkfelix

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

[arielb1]

@bors r+

[bors]

📌 Commit 25b1d2c has been approved by arielb1

[nikomatsakis]

Marked as beta-accepted: regr in diagnostics, low risk change.

[alexcrichton]

@bors: p=1

(beta nominated)

[bors]

⌛ Testing commit 25b1d2c with merge a1708c2...

[bors]

💔 Test failed - status-travis

[alexcrichton]

@bors: retry * network failure

…

On Mon, Dec 26, 2016 at 11:50 PM, bors ***@***.***> wrote: 💔 Test failed - status-travis <https://travis-ci.org/rust-lang/rust/builds/186914248> — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#38552 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAD95NwjMHdvzY_Y-crCeMoQ08eIdYXGks5rMMNTgaJpZM4LUN6w> .

[bors]

⌛ Testing commit 25b1d2c with merge 9fc4005...

[bors]

💔 Test failed - status-travis

[eddyb]

How can that even happen? It almost sounds like one line was replaced with the previous line, musl bug?

[eddyb]

@bors retry

[bors]

⌛ Testing commit 25b1d2c with merge abd4aa7...

[bors]

💔 Test failed - status-travis

[nikomatsakis]

@eddyb

looks legitimate

   Compiling rustc v0.0.0 (file:///checkout/src/librustc)

error[E0531]: unresolved tuple struct/variant `hir::MethodTraitItem`

   --> /checkout/src/librustc/infer/error_reporting.rs:118:17

    |

118 |                 hir::MethodTraitItem(..) => "method body",

    |                 ^^^^^^^^^^^^^^^^^^^^

error[E0531]: unresolved tuple struct/variant `hir::ConstTraitItem`

   --> /checkout/src/librustc/infer/error_reporting.rs:119:17

    |

119 |                 hir::ConstTraitItem(..) |

    |                 ^^^^^^^^^^^^^^^^^^^

error[E0531]: unresolved tuple struct/variant `hir::TypeTraitItem`

   --> /checkout/src/librustc/infer/error_reporting.rs:120:17

    |

120 |                 hir::TypeTraitItem(..) => "associated item"

    |                 ^^^^^^^^^^^^^^^^^^

error: aborting due to 3 previous errors

[eddyb]

I think I'll keep changes that are needed solely to make the original patch on master, separate.
At least then the original commit should still apply on beta.

[eddyb]

@bors r=arielb1

[bors]

📌 Commit 987f52f has been approved by arielb1

[bors]

⌛ Testing commit 987f52f with merge e06ce71...

[bors]

☀️ Test successful - status-appveyor, status-travis
Approved by: arielb1
Pushing e06ce71 to master...

[alexcrichton]

@eddyb this PR backports cleanly to beta but then results in a build failure. Can you investigate backporting to beta?

[eddyb]

@alexcrichton Do you have the build failure?

[alexcrichton]

@eddyb unfortunately no