Make rustdoc not include self-by-value methods from Deref target

[fhartwig]

Fixes #39550

[rust-highfive]

r? @frewsxcv

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

[kennytm]

Two rustdoc test cases need to be updated, removing the count_ones methods. Or maybe rustdoc should keep the self-by-value methods if the target is Copy (but that would get us into #33396).

[00:56:43] ---- [rustdoc] rustdoc/issue-19190-3.rs stdout ----
[00:56:43] 	
[00:56:43] error: htmldocck failed!
[00:56:43] status: exit code: 1
[00:56:43] command: "/usr/bin/python2.7" "/checkout/src/etc/htmldocck.py" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/rustdoc/issue-19190-3.stage2-x86_64-unknown-linux-gnu" "/checkout/src/test/rustdoc/issue-19190-3.rs"
[00:56:43] stdout:
[00:56:43] ------------------------------------------
[00:56:43] 
[00:56:43] ------------------------------------------
[00:56:43] stderr:
[00:56:43] ------------------------------------------
[00:56:43] 20: @has check failed
[00:56:43] 	`XPATH PATTERN` did not match
[00:56:43] 	// @has - '//*[@id="method.count_ones"]' 'fn count_ones(self) -> u32'
[00:56:43] 
[00:56:43] Encountered 1 errors
[00:56:43] 
[00:56:43] ------------------------------------------
[00:56:43] 
[00:56:43] failures:
[00:56:43]     [rustdoc] rustdoc/issue-19190-2.rs
[00:56:43]     [rustdoc] rustdoc/issue-19190-3.rs

[bors]

☔ The latest upstream changes (presumably #45187) made this pull request unmergeable. Please resolve the merge conflicts.

[frewsxcv]

r? @QuietMisdreavus

[QuietMisdreavus]

What i would suggest to fix the tests, rather than trying to wrangle with Copy, is to instead change the Derefs in those tests to point to String instead, and use fn as_str(&self) -> &str and fn new() -> String instead. That way the test is still checking for the proper thing. Note that for issue-19190-3.rs the Deref impls are actually in src/test/rustdoc/auxiliary/issue-19190-3.rs - that test is checking for cross-crate imports.

[shepmaster]

We haven't heard from you in a week or so, @fhartwig — will you have time to address the most recent feedback?

[fhartwig]

@shepmaster Sorry for the silence, I've been pretty busy. I should be able to finish this in the next 2-3 days.

[fhartwig]

I've rebased the changes and adapted the failing tests as @QuietMisdreavus suggested (thanks!).

Sorry for the delay.

[QuietMisdreavus]

@bors r+

Thanks!

[bors]

📌 Commit 32af136 has been approved by QuietMisdreavus

[bors]

⌛ Testing commit 32af136 with merge e061383...

[bors]

☀️ Test successful - status-appveyor, status-travis
Approved by: QuietMisdreavus
Pushing e061383 to master...