Rename intra_doc_link_resolution_failure

[Manishearth]

It should be plural to follow the conventions in https://github.com/rust-lang/rfcs/blob/master/text/0344-conventions-galore.md#lints

[rust-highfive]

Some changes occurred in intra-doc-links.

cc @jyn514

[rust-highfive]

r? @davidtwco

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

[Manishearth]

Done!

[jyn514]

BTW, the failure before the second rename was

error: lint `intra_doc_link_resolution_failure` has been renamed to `intra_doc_link_resolution_failures`
  --> library/core/src/lib.rs:64:9
   |
64 | #![deny(intra_doc_link_resolution_failure)] // rustdoc is run without -D warnings
   |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use the new name: `intra_doc_link_resolution_failures`
   |
   = note: `-D renamed-and-removed-lints` implied by `-D warnings`

error: lint `intra_doc_link_resolution_failure` has been renamed to `intra_doc_link_resolution_failures`
   --> library/core/src/lib.rs:152:9
    |
152 | #![deny(intra_doc_link_resolution_failure)]
    |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use the new name: `intra_doc_link_resolution_failures`

So you probably have to update those still.

[Manishearth]

Ah, I ran my replaces in src/

[jyn514]

Ugh, looks like you need cfg_attr(not(bootstrap)) in core.

error: unknown lint: `intra_doc_resolution_failures`
  --> library/core/src/lib.rs:64:9
   |
64 | #![deny(intra_doc_resolution_failures)] // rustdoc is run without -D warnings
   |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: did you mean: `intra_doc_link_resolution_failure`
   |
   = note: `-D unknown-lints` implied by `-D warnings`

error: unknown lint: `intra_doc_resolution_failures`
   --> library/core/src/lib.rs:152:9
    |
152 | #![deny(intra_doc_resolution_failures)]
    |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: did you mean: `intra_doc_link_resolution_failure`

r=me once CI is fixed.

[Mark-Simulacrum]

Noticed in passing - do we have an issue about rustdoc not denying warnings? If that's still true today, please file one and cc me on it. We should almost certainly not have any deny attributes here.

[jyn514]

I think this was added shortly before we denied rustdoc warnings (#71670, #74300). So it can probably be removed now.

[jyn514]

@Manishearth you need to run x.py fmt.

[Manishearth]

Aaah I thought I'd done this, but not each time

[jyn514]

You also need to run x.py test src/test/rustdoc --bless 😆

[bors]

📌 Commit c17eb56 has been approved by jyn514

[Manishearth]

@bors rollup=iffy

still haven't been able to locally test it

[bors]

⌛ Testing commit c17eb56 with merge ffa80f0...

[bors]

☀️ Test successful - checks-actions, checks-azure
Approved by: jyn514
Pushing ffa80f0 to master...

[lopopolo]

Thanks for naming the lint to something that's both consistent with the API guidelines and also more memorable. One side effect is that it is now difficult to build existing projects on the current stable and latest nightly.

Running on nightly with -D warnings in CI for generating rustdoc results in renamed-and-removed-lints warnings. Updating the lint to the new name results in unknown_lints warnings on stable.

The only options are to either blanket allow one of the failing lints (since allow attributes can't be applied to allow attributes) or do a dance with -A lint flags on CI and unwind them once the new name hits stable.

[jyn514]

@lopopolo can't you allow(unknown_lints) on stable? But in general warnings are not guaranteed to be stable, see #43466 (comment)

[jyn514]

Or in other words, I don't know a way to prevent the situation you mentioned other than never renaming lints ever.

[Manishearth]

Renaming lints is not uncommon, library authors are expected to apply allow(unknown_lints) for that duration.