rustdoc: update pulldown + fix spurious rendering difference around footnotes

[QuietMisdreavus]

fixes #45420

[rust-highfive]

r? @GuillaumeGomez

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

[steveklabnik]

This does not fix the issue:

WARNING: documentation for this crate may be rendered differently using the new Pulldown renderer.
    See https://github.com/rust-lang/rust/issues/44229 for details.
WARNING: rendering difference in `# The Rust Core Library`
   --> src\libcore\lib.rs:11:0
    /html[0]/body[1]/p[1] Attributes differ in `sup`: expected: `{"id": "supref1"}`, found: `{"id": "fnref1"}`
WARNING: rendering difference in `# The Rust Core Library`
   --> src\libcore\lib.rs:11:0
    /html[0]/body[1]/div[7]/ol[1] Attributes differ in `li`: expected: `{"id": "ref1"}`, found: `{"id": "fn1"}`

[steveklabnik]

(We should still land this, as I believe it does fix rendering issues, or so I hear; I haven't validated that myself yet. But it doesn't fix the warning)

[QuietMisdreavus]

I've just pushed an update that adds a check to bypass reporting a warning if the difference was with id attributes that follow the pattern pulldown and hoedown use for footnotes. Some structural differences were still fixed with the version bump, but the last rendering difference surrounding footnotes should be gone with that check.

[steveklabnik]

With my PR + this PR, we have no warnings 🎊

@bors: r+

[bors]

📌 Commit cee1991 has been approved by steveklabnik

[ollie27]

The ids are generated by rustdoc so they could have been changed to match hoedown:

rust/src/librustdoc/html/markdown.rs

Line 374 in cc03782

let reference = format!("<sup id=\"supref{0}\"><a href=\"#ref{0}\">{0}\

[frewsxcv]

@bors rollup

[QuietMisdreavus]

@ollie27 D'oh! Since this is already in the middle of a rollup, i'll grab that in a separate PR. Nice catch!

[bors]

☔ The latest upstream changes (presumably #45430) made this pull request unmergeable. Please resolve the merge conflicts.

[kennytm]

@QuietMisdreavus No it failed to make into the rollup 😄, so that change could be put in this PR.

[QuietMisdreavus]

@kennytm Oh, didn't notice that, oops. >_>;;

Take two! I just force-pushed a different update that uses @ollie27's suggestion rather than writing in a warning suppression. I ran it locally and it fixed the rendering differences in libcore the same as the other way, only this time it's because it's actually the same.

[kennytm]

@GuillaumeGomez @ollie27 Is this PR good to go?

[GuillaumeGomez]

Seems good for me.

[QuietMisdreavus]

r? @steveklabnik

Since the footnote mitigation changed after the initial approval, this is worth another look before i blindly re-r+ it.

[steveklabnik]

@bors: r+

[bors]

📌 Commit af09ba8 has been approved by steveklabnik