Implements RFC 1845, adding implementations of:
* `From<&[T]>` for `Rc<[T]>`
* `From<&str>` for `Rc<str>`
* `From<String>` for `Rc<str>`
* `From<Box<T: ?Sized>>` for `Rc<T>`
* `From<Vec<T>>` for `Rc<[T]>`
* and likewise for `Arc<_>`
Also removes now-obsolete internal methods `Rc::__from_array` and
`Rc::__from_str`, replacing their use with `Rc::from`.
Minor rewrite of char primitive unicode intro.
Opened primarily to address #36998.
Despite my love for emoji, the heart example is a little confusing because both heart characters start with the same code point and there can be stark rendering differences across browsers. I also spelled out what each of the code points is in the code block, which (hopefully) sheds light why one character is one code point while the other is two.
Very much open to suggestion and improvements. I'm pretty tired when I wrote this so I might wake up and realize that this is making things more confusing 😅
Redox paths are problematic. It would make sense to add a `Scheme`
variant to the `std::path::Component` enum; but that would presumably be
a breaking change due to exhaustive matching. Alternately it could use
the existing `Prefix` variant, like Windows, but none of the existing
types of prefix make sense, Redox only has one kind, and adding a new
variant to that enum has the same issue as `Component`.
When walking parents for lints we want to be sure to hit `let` statements which
can have attributes, so hook up these statements in the HIR map.
Closes#43910
Eat open paren when parsing list in libsyntax/parse/attr.rs
This PR adds a small refactoring:
```diff
pub fn parse_meta_item_kind(&mut self) -> PResult<'a, ast::MetaItemKind> {
Ok(if self.eat(&token::Eq) {
ast::MetaItemKind::NameValue(self.parse_unsuffixed_lit()?)
- } else if self.token == token::OpenDelim(token::Paren) {
+ } else if self.eat(&token::OpenDelim(token::Paren)) {
ast::MetaItemKind::List(self.parse_meta_seq()?)
} else {
- self.eat(&token::OpenDelim(token::Paren));
ast::MetaItemKind::Word
})
}
```
in `parse_meta_item_kind()`, the parser calls `self.eat(&token::OpenDelim(token::Paren));` before returning `ast::MetaItemKind::Word` just to add `(` to expected token. It seems more natural to eat the paren when parsing `ast::MetaItemKind::List`.
Refactoring: move net specific file descriptor methods
Move the implementations of net specific file descriptor methods from
io to net. This makes it easier to exclude net at all if it is not needed
for a target.
Implement CompilerDesugaringKind enum
This is the first step outlined in #35946. I think that the variants of `CompilerDesugaringKind` should be changed, I didn't know what the official names for `...` and `<-` are.
I'm not to sure how tests for the compiler work, but I would imagine that tests should be added such that
`Symbol::intern(s) == CompilerDesugaringKind::from(s).as_symbol()` for valid `s`.
On a case where an else conditional is missing, point this out
instead of the token immediately after the (incorrect) else block:
```
error: missing condition for `if` statemementt push fork -f
--> $DIR/issue-13483.rs:16:5
|
13 | } else if {
| ^ expected if condition here
```
instead of
```
error: expected `{`, found `else`
--> ../../src/test/ui/issue-13483.rs:14:7
|
14 | } else {
| ^^^^
```
Fix ES5 regression with shorthand names.
Reverts 1b6c9605e4.
I appreciate new features and syntax in Rust, but seriously, don't rewrite anything. Especially if this **breaks** documentation of language itself and every crate hosted at docs.rs.
Implement a temp redirect for cargo docs
As discussed in
https://github.com/rust-lang/cargo/issues/4040#issuecomment-321639074
This is a redirect meant to be replaced once cargo docs have been
converted to mdbook. We just want *a* URL to ride the trains for now so
that we can print doc.rust-lang.org/cargo in the paper book and
guarantee that it will go *somewhere* useful by the time the book is
printed.
Implemented as a meta redirect in HTML because we don't currently have
any google juice at doc.rust-lang.org/cargo to lose.
When I run `./x.py doc`, this creates a `build/x86_64-apple-darwin/doc/cargo/index.html` file that contains a meta redirect to doc.crates.io. As I understand rust-central-station to work, this should be what we need to make `doc.rust-lang.org/cargo` to work.
r? @alexcrichton and/or @steveklabnik
Document that `std:#️⃣:Hasher::finish()` does not reset the hasher.
Clarifies the fact that `finish()` doesn’t in fact end or reset the hasher. This was surprising to me …
Follows up on and fixes#43763
Remove useless help part
Part of #32658.
I think this error should be splitted into two parts, each more specific (`anonymous function` vs `type method`).
As discussed in
https://github.com/rust-lang/cargo/issues/4040#issuecomment-321639074
This is a redirect meant to be replaced once cargo docs have been
converted to mdbook. We just want *a* URL to ride the trains for now so
that we can print doc.rust-lang.org/cargo in the paper book and
guarantee that it will go *somewhere* useful by the time the book is
printed.
Implemented as a meta redirect in HTML because we don't currently have
any google juice at doc.rust-lang.org/cargo to lose.
MIR borrow check (under debug flag)
Here is the current state of MIR borrow check.
It consists of (1.) some refactoring, (2.) a dataflow analysis to identify the borrows themselves, and (3.) a mir "transform" that does the borrow check itself based on the aforementioned dataflow results.
(There's also a drive-by fix to dataflow that I can factor into a separate PR if necessary. Interestingly I could not find a way to observe the bug outside of MIR borrowck.)
To be clear, this branch is not ready to be used as the default borrow check. Thus the code is guarded: To get mir-borrowck to run, you need to either supply an attribute `#[rustc_mir_borrowck]` or a debug flag `-Z borrowck-mir`.
Here are the main issues with the current MIR borrowck as it stands in this PR:
* No Notes emitted yet, just errors. (So the feedback is definitely inferior compared to AST borrowck today)
* Lvalue rendering differs between Ast and Mir. (Mostly minor, but replacement of field names with indices is very bad; big priority for me to fix ASAP.)
* Lots of ICEs (presumably because some MIR operations used here have well-formedness assumptions that are violated in borrowck-broken code)
* Conflates lots of cases that are distinguished by AST-borrowck
* Conflates "uninitialized" with "moved" (special case of previous bullet, one that I think should be fixed ASAP)
(I am hoping to fix as many of the above issues as I can in the near term, but I also would like to land this even if they are *not* all fixed, because the rebasing effort is getting to be a real drag.)
Added two fixmes: The `SimplifyBranches` pass cannot stay where it is,
and `BorrowckMir` should be a query, not a pass. But I am going to
leave those changes to a future PR.
One can either use `-Z borrowck-mir` or add the `#[rustc_mir_borrowck]` attribute
to opt into MIR based borrow checking.
Note that regardless of whether one opts in or not, AST-based borrow
check will still run as well. The errors emitted from AST-based
borrow check will include a "(Ast)" suffix in their error message,
while the errors emitted from MIR-based borrow check will include a
"(Mir)" suffix.
post-rebase: removed check for intra-statement mutual conflict;
replaced with assertion checking that at most one borrow is generated
per statement.
post-rebase: removed dead code: `IdxSet::pairs` and supporting stuff.
post-rebase: Do not put "(Ast)" suffix in error msg unless passed `-Z borrowck-mir`.
(But unconditionally include "(Mir)" suffix for mir-borrowck errors.)
Move the implementations of net specific file descriptior implementations
to net. This makes it easier to exclude net at all if not needed for a
target.
rustc: Fix `unknown_lints` next to an unknown lint
The lint refactoring in #43522 didn't account for `#[allow(unknown_lints)]`
happening at the same node as an unknown lint itself, so this commit updates the
handling to ensure that the local set of lint configuration being built is
queried before looking at the chain of lint levels.
Closes#43809
Fix ICE with elided lifetimes in return type of foreign functions
cc https://github.com/rust-lang/rust/issues/43567
This is for a preliminary crater/cargobomb run.
Lifetime elision in foreign functions now works exactly like in other functions or function-like entities.
If the breakage is significant, I'll have to partially revert https://github.com/rust-lang/rust/pull/43543 (all the stuff that was required for dealing with late bound lifetimes in this position).
r? @eddyb
Add Vec::drain_filter
This implements the API proposed in #43244.
So I spent like half a day figuring out how to implement this in some awesome super-optimized unsafe way, which had me very confident this was worth putting into the stdlib.
Then I looked at the impl for `retain`, and was like "oh dang". I compared the two and they basically ended up being the same speed. And the `retain` impl probably translates to DoubleEndedIter a lot more cleanly if we ever want that.
So now I'm not totally confident this needs to go in the stdlib, but I've got two implementations and an amazingly robust test suite, so I figured I might as well toss it over the fence for discussion.
Make backtraces work on Redox, copying Unix implementation
The `backtrace/` directory here is the same as the Unix one, except for adding an implementation of `get_executable_filename`.
Add method `String::retain`
Behaves like `Vec::retain`, accepting a predicate `FnMut(char) -> bool`
and reducing the string to only characters for which the predicate
returns `true`.
emit StorageLive for box temporaries
We started emitting StorageDead, so we better emit the corrseponding
StorageLive to avoid problems.
cc #43772solson/miri#303
Behaves like `Vec::retain`, accepting a predicate `FnMut(char) -> bool`
and reducing the string to only characters for which the predicate
returns `true`.
Rewrite/reorganize docs for stack size/thread names for spawned threads.
* Moves docs about stack size and thread naming from `Builder` to the
`std::thread` module
* Adds more links to the new module-level documentation
* Mentions the 2 MiB stack size default, but indicate it's subject to
change
Fixes https://github.com/rust-lang/rust/issues/43805.
Instant is monotonically nondecreasing
We don't want to guarantee that `Instant::now() != Instant::now()` is
always true since that depends on the speed of the processor and the
resolution of the clock.
remove the "defaulted unit" type bit during writeback
The defaulted unit bit is only relevant for the surrounding inference
context, and can cause trouble, including spurious lints and ICEs,
outside of it.
Fixes#43853.
r? @eddyb
Fix "Mis-calculated spans" errors from `-Z save-analysis` + refactoring
Removed the path span extraction methods from `SpanUtils`:
* spans_with_brackets
* spans_for_path_segments
* spans_for_ty_params
Use the `span` fields in `PathSegment` and `TyParam` instead.
(Note that since it processes `ast::Path` not a qualified path (`hir::QPath` / `ast::QSelf`), UFCS path will be flattened: `<Foo as a:🅱️:c::Trait>::D::E::F::g` will be seen as `a:🅱️:c::Trait::D::E::F::g`.)
Fix#43796. Close#41478.
r? @nrc
Use hir::ItemLocalId as keys in TypeckTables.
This PR makes `TypeckTables` use `ItemLocalId` instead of `NodeId` as key. This is needed for incremental compilation -- for stable hashing and for being able to persist and reload these tables. The PR implements the most important part of https://github.com/rust-lang/rust/issues/40303.
Some notes on the implementation:
* The PR adds the `HirId` to HIR nodes where needed (`Expr`, `Local`, `Block`, `Pat`) which obviates the need to store a `NodeId -> HirId` mapping in crate metadata. Thanks @eddyb for the suggestion! In the future the `HirId` should completely replace the `NodeId` in HIR nodes.
* Before something is read or stored in one of the various `TypeckTables` subtables, the entry's key is validated via the new `TypeckTables::validate_hir_id()` method. This makes sure that we are not mixing information from different items in a single table.
That last part could be made a bit nicer by either (a) new-typing the table-key and making `validate_hir_id()` the only way to convert a `HirId` to the new-typed key, or (b) just encapsulate sub-table access a little better. This PR, however, contents itself with not making things significantly worse.
Also, there's quite a bit of switching around between `NodeId`, `HirId`, and `DefIndex`. These conversions are cheap except for `HirId -> NodeId`, so if the valued reviewer finds such an instance in a performance critical place, please let me know.
Ideally we convert more and more code from `NodeId` to `HirId` in the future so that there are no more `NodeId`s after HIR lowering anywhere. Then the amount of switching should be minimal again.
r? @eddyb, maybe?
The defaulted unit bit is only relevant for the surrounding inference
context, and can cause trouble, including spurious lints and ICEs,
outside of it.
Fixes#43853.
ast_validation: forbid "nonstandard" literal patterns
Since #42886, macros can create "nonstandard" PatKind::Lit patterns,
that contain path expressions instead of the usual literal expr. These
can cause trouble, including ICEs.
We *could* map these nonstandard patterns to PatKind::Path patterns
during HIR lowering, but that would be much effort for little gain, and
I think is too risky for beta. So let's just forbid them during AST
validation.
Fixes#43250.
beta-nominating because regression.
r? @eddyb
Historically many `Display` and `Debug` implementations for `OsStr`-like
abstractions have gone through `String::from_utf8_lossy`, but this was updated
in #42613 to use an internal `Utf8Lossy` abstraction instead. This had the
unfortunate side effect of causing a regression (#43765) in code which relied on
these `fmt` trait implementations respecting the various formatting flags
specified.
This commit opportunistically adds back interpretation of formatting trait flags
in the "common case" where where `OsStr`-like "thing" is all valid utf-8 and can
delegate to the formatting implementation for `str`. This doesn't entirely solve
the regression as non-utf8 paths will format differently than they did before
still (in that they will not respect formatting flags), but this should solve
the regression for all "real world" use cases of paths and such. The door's also
still open for handling these flags in the future!
Closes#43765
Cleanup for "Support compiling rustc without LLVM (try 2)"
This includes a small patch to allow running tests without llvm. Also check if you are not trying to compile a dylib.
cc #42932
r? @alexcrichton
Fix for issue #39827
*Cause of the issue*
While preparing for `trans_intrinsic_call()` invoke arguments are processed with `trans_argument()` method which excludes zero-sized types from argument list (to be more correct - all arguments for which `ArgKind` is `Ignore` are filtered out). As result `volatile_store()` intrinsic gets one argument instead of expected address and value.
*How it is fixed*
Modification of the `trans_argument()` method may cause side effects, therefore change was implemented in `volatile_store()` intrinsic building code itself. Now it checks function signature and if it was specialised with zero-sized type, then emits `C_nil()` instead of accessing non-existing second argument.
Optimize allocation paths in RawVec
Since the `Alloc` trait was introduced (https://github.com/rust-lang/rust/pull/42313) and it was integrated everywhere (https://github.com/rust-lang/rust/pull/42727) there's been some slowdowns and regressions that have slipped through. The intention of this PR is to try to tackle at least some of them, but they've been very difficult to quantify up to this point so it probably doesn't solve everything.
This PR primarily targets the `RawVec` type, specifically the `double` function. The codegen for this function is now much closer to what it was before #42313 landed as many runtime checks have been elided.
* Moves docs about stack size and thread naming from `Builder` to the
`std::thread` module
* Adds more links to the new module-level documentation
* Mentions the 2 MiB stack size default, but indicate it's subject to
change
Fixes https://github.com/rust-lang/rust/issues/43805.
Since #42886, macros can create "nonstandard" PatKind::Lit patterns,
that contain path expressions instead of the usual literal expr. These
can cause trouble, including ICEs.
We *could* map these nonstandard patterns to PatKind::Path patterns
during HIR lowering, but that would be much effort for little gain, and
I think is too risky for beta. So let's just forbid them during AST
validation.
Fixes#43250.
The lint refactoring in #43522 didn't account for `#[allow(unknown_lints)]`
happening at the same node as an unknown lint itself, so this commit updates the
handling to ensure that the local set of lint configuration being built is
queried before looking at the chain of lint levels.
Closes#43809
Currently the link on doc.rust-lang.org is semi-broken; it links to a
page that links to the exact page in the first edition in the book, or
to the index of the second edition of the book. If the second editions
is the recommended one now, we should point the links at that one. It
seems that in the mean time, the links have been updated to point
directly to the first edition of the book, but that hasn't made it onto
the stable channel yet. By the time this commit makes it onto the stable
channel, the second edition of the book should be complete enough. At
least the part about deref coercions is.
Fix include! in doc tests
By making the path relative to the current file.
Fixes#43153
[breaking-change] - if you use `include!` inside a doc test, you'll need to change the path to be relative to the current file rather than relative to the working directory.
Fix unused_result lint triggering when a function returns `()`, `!` or an empty enum
Also added a test to prevent this from happening again.
Fixes#43806
Rustbuild cleanups/fixes and improvements
Each commit is a standalone change, and can/should be reviewed separately.
This adds two new functionalities:
- `--target` and `--host` can be passed without changing config.toml, and we'll respect the users' wishes, instead of requiring that all possible targets are passed.
- Note that this means that `./x.py clean` won't be quite as wide-spread as before, since it limits itself to the configured hosts, not all hosts. This could be considered a feature as well.
- `ignore-git` field in `config.toml` which tells Rustbuild to not attempt to load git hashes from `.git`.
This is a precursor to eventual further simplification of the configuration system, but I want to get this merged first so that later work can be made in individual PRs.
r? @alexcrichton
Expose all OS-specific modules in libstd doc.
1. Uses the special `--cfg dox` configuration passed by rustbuild when running `rustdoc`. Changes the `#[cfg(platform)]` into `#[cfg(any(dox, platform))]` so that platform-specific API are visible to rustdoc.
2. Since platform-specific implementations often won't compile correctly on other platforms, `rustdoc` is changed to apply `everybody_loops` to the functions during documentation and doc-test harness.
3. Since platform-specific code are documented on all platforms now, it could confuse users who found a useful API but is non-portable. Also, their examples will be doc-tested, so must be excluded when not testing on the native platform. An undocumented attribute `#[doc(cfg(...))]` is introduced to serve the above purposed.
Fixes#24658 (Does _not_ fully implement #1998).
This fixes the bug we previously had where we'd build a libtest tool
after building a libstd tool and clear out the libstd tool. Since we
clear out all tools for a given stage on invocations of CleanTools after
lib{std, test, rustc} change, we need to make sure that all tools built
with that stage will be built after the clearing is done.
The fix contained here technically isn't perfect; there is still an edge
case of compiling a libstd tool, then compiling libtest, which will
clear out the libstd tool and it won't ever get rebuilt within that
session of rustbuild. This is where the caching system used today shows
it's problems -- in effect, all tools depend on a global counter of the
stage being cleared out. We can implement such a counter in a future
patch to ensure that tools are rebuilt as needed, but it is deemed
unlikely that it will be required in practice, since most if not all
tools are built after the relevant stage's std/test/rustc are built,
though this is only an opinion and hasn't been verified.
Some users of the build system change the git sha on every build due to
utilizing git to push changes to a remote server. This allows them to
simply configure that away instead of depending on custom patches to
rustbuild.
This introduces a slight change in behavior, where we unilaterally
respect the --host and --target parameters passed for all sanity
checking and runtime configuration.
Improve std::ops docs
Fixes#29365. (This fixes all but one point from @steveklabnik's list, but that point was referring to examples of implementing range traits, but there are no range traits in std::ops.)
The main changes are quite a bit of copyediting, adding more "real" examples for some of the traits, incorporating some guidance from the API docs, more linking (cross-docs and to the book & reference), cleaning up examples, moving things around, and so on. Refer to the commit messages for more details.
Note: I decided to link to the second edition of the book since I think it's more appropriate now for the sections I linked, if this is not okay, please say so!
The `RawVec` type has a number of invariants that it upholds throughout its
execution, and as a result many of the runtime checks imposed by using `Layout`
in a "raw" fashion aren't actually necessary. For example a `RawVec`'s capacity
is intended to always match the layout which "fits" the allocation, so we don't
need any runtime checks when retrieving the current `Layout` for a vector.
Consequently, this adds a safe `current_layout` function which internally uses
the `from_size_align_unchecked` function.
Along the same lines we know that most construction of new layouts will not
overflow. All allocations in `RawVec` are kept below `isize::MAX` and valid
alignments are also kept low enough that we're guaranteed that `Layout` for a
doubled vector will never overflow and will always succeed construction.
Consequently a few locations can use `from_size_align_unchecked` in addition
when constructing the *new* layout to allocate (or reallocate), which allows for
eliding some more runtime checks.
Overall this should significant improve performance for an important function,
`RawVec::double`. This commit removes four runtime jumps before `__rust_realloc`
is called, as well as one after it's called.
* fixed link typos and copy-paster errors
* rewrote Fn* explanations
* `RHS = Self` -> `RHS` is `Self` (added that to all applicable places as
well)
* fixed up some links
* s/MutDeref/DerefMut
* removed remaining superfluous `fn main()`s
* fixed some minor phrasings and factual errors and inaccuracies
std::ops docs: Fix phrasing and factual errors/inaccuracies
Check #[thread_local] statics correctly in the compiler.
Fixes#43733 by introducing `#[allow_internal_unsafe]` analogous to `#[allow_internal_unstable]`, for letting a macro expand to `unsafe` blocks and functions even in `#![forbid(unsafe_code)]` crates.
Fixes#17954 by not letting references to `#[thread_local]` statics escape the function they're taken in - we can't just use a magical lifetime because Rust has *lifetime parametrism*, so if we added the often-proposed `'thread` lifetime, we'd have no way to check it in generic code.
To avoid potential edge cases in the compiler, the lifetime is actually that of a temporary at the same position, i.e. `&TLS_STATIC` has the same lifetime `&non_const_fn()` would.
Referring to `#[thread_local]` `static`s at compile-time is banned now (as per PR discussion).
Additionally, to remove `unsafe impl Sync` from `std:🧵:local::fast::Key`, `#[thread_local]` statics are now not required to implement `Sync`, as they are not shared between threads.
- updates documentation on volatile memory intrinsics, now the case of
zero-sized types is mentioned explicitly.
Volatile memory operations which doesn't affect memory at all are omitted
in LLVM backend, e.g. if number of elements is zero or type used in
generic specialisation is zero-sized, then LLVM intrinsic or related code
is not generated. This was not explicitly documented before in Rust
documentation and potentially could cause issues.
- adds handling of zero-sized types for volatile_store.
- adds type size checks and warnigns for other volatile intrinsics.
- adds a test to check warnings emitting.
Cause of the issue
While preparing for trans_intrinsic_call() invoke arguments are
processed with trans_argument() method which excludes zero-sized types
from argument list (to be more correct - all arguments for which
ArgKind is Ignore are filtered out). As result volatile_store() intrinsic
gets one argument instead of expected address and value.
How it is fixed
Modification of the trans_argument() method may cause side effects,
therefore change was implemented in volatile_store() intrinsic building
code itself. Now it checks function signature and if it was specialised
with zero-sized type, then emits C_nil() instead of accessing
non-existing second argument.
Additionally warnings are added for all volatile operations which are
specialised with zero-sized arguments. In fact, those operations are omitted
in LLVM backend if no memory affected at all, e.g. number of elements
is zero or type is zero-sized. This was not explicitly documented before
and could lead to potential issues if developer expects volatile behaviour,
but type has degraded to zero-sized.
For box expressions, use NZ drop instead of a free block
This falls naturally out of making drop elaboration work with `box`
expressions, which is probably required for sane MIR borrow-checking.
This is a pure refactoring with no intentional functional effects.
r? @nagisa
None of these require a significant amount of code and using `#[inline]` will
allow constructors to get inlined, improving codegen at allocation callsites.
Improve LLVM/trans scheduling a bit
Currently it's possible that the main thread is waiting on LLVM threads to finish work while its implicit token is going to waste. This PR let's the main thread take over, so one of the running LLVM threads can free its token earlier.
r? @alexcrichton
Put `intrinsics::unreachable` on a possible path to stabilization
Mark it with the `unreachable` feature and put it into the `mem` module.
This is a pretty straight-forward API that can already be simulated in
stable Rust by using `transmute` to create an uninhabited enum that can
be matched.
AddValidation: handle Call terminators into blocks that have multiple incoming edges
The old code was just wrong: It would add validation on paths that don't even come from the call, and it would add multiple validations if multiple calls end return to the same block.
Unfortunately, the NodeId->HirId array is not sorted. Since this search is only
done right before calling bug!(), let's not waste time allocating a faster lookup.
Encode proper module spans in crate metadata.
The spans previously encoded only span the first token after the opening
brace, up to the closing brace of inline `mod` declarations. Thus, when
examining exports from an external crate, the spans don't include the
header of inline `mod` declarations.
r? @eddyb
Provide more explanation for Deref in String docs
While working on a different project I encountered a point of confusion where using `&String` to dereference a `String` into `&str` did not compile. I found the explanation of [String Deref](https://doc.rust-lang.org/std/string/struct.String.html#deref), thought that it matched what I was trying to do, and was confused as to why my program did not compile when the docs stated that it would work with 'any function which takes a `&str`'. At the bottom it is mentioned that this will 'generally' work, unless `String` is needed, but I found this statement confusing based on the previous claim of 'any'. Looking further into the docs I was able to find the function `as_str()` that works instead.
I thought it might be helpful to mention here deref coercion, an instance in which using `&String` does not work, to explain why it does not work, then direct users to a different option that should work in this instance. A user casually skimming the page will likely come to this explanation first, then find `as_str()` later, but be no the wiser as to what potentially went wrong.
r? @steveklabnik
Reexport all SyntaxExtension variants
This was previously done very inconsistently and made matches look weird since some variants had the `SyntaxExtension::` prefix while others didn't.
Detect relative urls in tidy check
This came up in #43631: there can be long relative urls in Markdown comments, that do not start with `http://` or `https://`, so the tidy check will not detect them as urls and complain about the line length. This PR adds detection of relative urls starting with `../`.
E0122: clarify wording
I *assume* the reason these constraints are not hard errors is backwards compatibility. If yes, I think the error explanation (at least the long form) should be clearer about that, which is what this PR does.
If not, the explanation should give some other suitable explanation.
