2032 Commits

Author	SHA1	Message	Date
Nicholas Nethercote	9d31f86f0d	Overhaul token collection. ... This commit does the following. - Renames `collect_tokens_trailing_token` as `collect_tokens`, because (a) it's annoying long, and (b) the `_trailing_token` bit is less accurate now that its types have changed. - In `collect_tokens`, adds a `Option<CollectPos>` argument and a `UsePreAttrPos` in the return type of `f`. These are used in `parse_expr_force_collect` (for vanilla expressions) and in `parse_stmt_without_recovery` (for two different cases of expression statements). Together these ensure are enough to fix all the problems with token collection and assoc expressions. The changes to the `stringify.rs` test demonstrate some of these. - Adds a new test. The code in this test was causing an assertion failure prior to this commit, due to an invalid `NodeRange`. The extra complexity is annoying, but necessary to fix the existing problems.	2024-08-16 09:07:55 +10:00
Nicholas Nethercote	5aaa2f92ee	Add an assertion to NodeRange::new.	2024-08-16 09:07:31 +10:00
Nicholas Nethercote	c8098be41f	Convert a bool to Trailing. ... This pre-existing type is suitable for use with the return value of the `f` parameter in `collect_tokens_trailing_token`. The more descriptive name will be useful because the next commit will add another boolean value to the return value of `f`.	2024-08-16 09:07:29 +10:00
Nicholas Nethercote	55906aa240	Make visibilities minimal and consistent in attr_wrapper.rs.	2024-08-16 09:06:15 +10:00
Nicholas Nethercote	af0093a6b8	Remove size assertion on AttrWrapper. ... It's not an important type when it comes to memory use.	2024-08-16 09:06:15 +10:00
Nicholas Nethercote	7923b20dd9	Use impl PartialEq<TokenKind> for Token more. ... This lets us compare a `Token` with a `TokenKind`. It's used a lot, but can be used even more, avoiding the need for some `.kind` uses.	2024-08-14 16:37:09 +10:00
Nicholas Nethercote	bbcfd90cd1	Convert a &mut self to &self.	2024-08-14 13:06:57 +10:00
Guillaume Gomez	99a785d62d	Rollup merge of #128994 - nnethercote:fix-Parser-look_ahead-more, r=compiler-errors ... Fix bug in `Parser::look_ahead`. The special case was failing to handle invisible delimiters on one path. Fixes (but doesn't close until beta backported) #128895. r? `@davidtwco`	2024-08-12 17:09:20 +02:00
Guillaume Gomez	7c6dca9050	Rollup merge of #128978 - compiler-errors:assert-matches, r=jieyouxu ... Use `assert_matches` around the compiler more It's a useful assertion, especially since it actually prints out the LHS.	2024-08-12 17:09:19 +02:00
Nicholas Nethercote	46b4c5adc5	Fix bug in Parser::look_ahead. ... The special case was failing to handle invisible delimiters on one path. Fixes #128895.	2024-08-12 13:00:12 +10:00
Michael Goulet	c361c924a0	Use assert_matches around the compiler	2024-08-11 12:25:39 -04:00
Matthias Krüger	32e0fe129d	Rollup merge of #128762 - fmease:use-more-slice-pats, r=compiler-errors ... Use more slice patterns inside the compiler Nothing super noteworthy. Just replacing the common 'fragile' pattern of "length check followed by indexing or unwrap" with slice patterns for legibility and 'robustness'. r? ghost	2024-08-11 07:51:51 +02:00
许杰友 Jieyou Xu (Joe)	d65f1316bb	parser: ensure let stmt compound assignment removal suggestion respect codepoint boundaries ... Previously we would try to issue a suggestion for `let x <op>= 1`, i.e. a compound assignment within a `let` binding, to remove the `<op>`. The suggestion code unfortunately incorrectly assumed that the `<op>` is an exactly-1-byte ASCII character, but this assumption is incorrect because we also recover Unicode-confusables like `➖=` as `-=`. In this example, the suggestion code used a `+ BytePos(1)` to calculate the span of the `<op>` codepoint that looks like `-` but the mult-byte Unicode look-alike would cause the suggested removal span to be inside a multi-byte codepoint boundary, triggering a codepoint boundary assertion. Issue: <https://github.com/rust-lang/rust/issues/128845>	2024-08-09 05:56:50 +00:00
León Orell Valerian Liehr	c4c518d2d4	Use more slice patterns inside the compiler	2024-08-07 13:37:52 +02:00
Matthias Krüger	7d9ed2a864	Rollup merge of #127921 - spastorino:stabilize-unsafe-extern-blocks, r=compiler-errors ... Stabilize unsafe extern blocks (RFC 3484) # Stabilization report ## Summary This is a tracking issue for the RFC 3484: Unsafe Extern Blocks We are stabilizing `#![feature(unsafe_extern_blocks)]`, as described in [Unsafe Extern Blocks RFC 3484](https://github.com/rust-lang/rfcs/pull/3484). This feature makes explicit that declaring an extern block is unsafe. Starting in Rust 2024, all extern blocks must be marked as unsafe. In all editions, items within unsafe extern blocks may be marked as safe to use. RFC: https://github.com/rust-lang/rfcs/pull/3484 Tracking issue: #123743 ## What is stabilized ### Summary of stabilization We now need extern blocks to be marked as unsafe and items inside can also have safety modifiers (unsafe or safe), by default items with no modifiers are unsafe to offer easy migration without surprising results. ```rust unsafe extern { // sqrt (from libm) may be called with any `f64` pub safe fn sqrt(x: f64) -> f64; // strlen (from libc) requires a valid pointer, // so we mark it as being an unsafe fn pub unsafe fn strlen(p: *const c_char) -> usize; // this function doesn't say safe or unsafe, so it defaults to unsafe pub fn free(p: *mut core::ffi::c_void); pub safe static IMPORTANT_BYTES: [u8; 256]; pub safe static LINES: SyncUnsafeCell<i32>; } ``` ## Tests The relevant tests are in `tests/ui/rust-2024/unsafe-extern-blocks`. ## History - https://github.com/rust-lang/rust/pull/124482 - https://github.com/rust-lang/rust/pull/124455 - https://github.com/rust-lang/rust/pull/125077 - https://github.com/rust-lang/rust/pull/125522 - https://github.com/rust-lang/rust/issues/126738 - https://github.com/rust-lang/rust/issues/126749 - https://github.com/rust-lang/rust/issues/126755 - https://github.com/rust-lang/rust/pull/126757 - https://github.com/rust-lang/rust/pull/126758 - https://github.com/rust-lang/rust/issues/126756 - https://github.com/rust-lang/rust/pull/126973 - https://github.com/rust-lang/rust/pull/127535 - https://github.com/rust-lang/rustfmt/pull/6204 ## Unresolved questions I am not aware of any unresolved questions.	2024-08-03 20:51:51 +02:00
Matthias Krüger	dee57ce043	Rollup merge of #128483 - nnethercote:still-more-cfg-cleanups, r=petrochenkov ... Still more `cfg` cleanups Found while looking closely at `cfg`/`cfg_attr` processing code. r? `````````@petrochenkov`````````	2024-08-03 11:17:44 +02:00
Matthias Krüger	29cd3103a1	Rollup merge of #128496 - clubby789:box-syntax-multipart, r=compiler-errors ... Fix removed `box_syntax` diagnostic if source isn't available Fix #128442	2024-08-01 18:43:41 +02:00
clubby789	e157954cce	Fix removed box_syntax diagnostic if source isn't available	2024-08-01 13:11:24 +00:00
bors	c0e32983f5	Auto merge of #127543 - carbotaniuman:more_unsafe_attr_verification, r=estebank,traviscross ... More unsafe attr verification This code denies unsafe on attributes such as `#[test]` and `#[ignore]`, while also changing the `MetaItem` parsing so `unsafe` in args like `#[allow(unsafe(dead_code))]` is not accidentally allowed. Tracking: - https://github.com/rust-lang/rust/issues/123757	2024-08-01 10:40:45 +00:00
Nicholas Nethercote	d1f05fd184	Distinguish the two kinds of token range. ... When collecting tokens there are two kinds of range: - a range relative to the parser's full token stream (which we get when we are parsing); - a range relative to a single AST node's token stream (which we use within `LazyAttrTokenStreamImpl` when replacing tokens). These are currently both represented with `Range<u32>` and it's easy to mix them up -- until now I hadn't properly understood the difference. This commit introduces `ParserRange` and `NodeRange` to distinguish them. This also requires splitting `ReplaceRange` in two, giving the new types `ParserReplacement` and `NodeReplacement`. (These latter two names reduce the overloading of the word "range".) The commit also rewrites some comments to be clearer. The end result is a little more verbose, but much clearer.	2024-08-01 19:30:40 +10:00
Nicholas Nethercote	9d77d17f71	Move a comment to a better spot.	2024-08-01 19:30:39 +10:00
Nicholas Nethercote	2eb2ef1684	Streamline attribute stitching on AST nodes. ... It can be done more concisely.	2024-08-01 19:30:32 +10:00
Matthias Krüger	3acd910036	Rollup merge of #126697 - vincenzopalazzo:macros/find_the_expression_tok, r=eholk,compiler-errors ... [RFC] mbe: consider the `_` in 2024 an expression This commit is adding the possibility to parse the `_` as an expression inside the esition 2024. Link: https://rust-lang.zulipchat.com/#narrow/stream/404510-wg-macros/topic/supporting.20.60_.60.20expressions Issue https://github.com/rust-lang/rust/issues/123742 r? `@eholk`	2024-07-31 23:20:09 +02:00
Michael Goulet	79ef91e879	tweak comment on NonterminalKind::Expr ... Co-authored-by: Eric Holk <eric@theincredibleholk.org>	2024-07-31 15:37:55 -04:00
Vincenzo Palazzo	276fa19c0a	rustc_parser: consider the in 2024 an expression ... This commit is adding the possibility to parse the `_` as an expression inside the esition 2024. Link: https://rust-lang.zulipchat.com/#narrow/stream/404510-wg-macros/topic/supporting.20.60_.60.20expressions Co-authored-by: Eric Holk <eric@theincredibleholk.org> Signed-off-by: Vincenzo Palazzo <vincenzopalazzodev@gmail.com>	2024-07-31 18:34:22 +00:00
Nicholas Nethercote	fe647f0538	Remove LhsExpr. ... `parse_expr_assoc_with` has an awkward structure -- sometimes the lhs is already parsed. This commit splits the post-lhs part into a new method `parse_expr_assoc_rest_with`, which makes everything shorter and simpler.	2024-07-31 12:56:25 +10:00
Nicholas Nethercote	281c2fd5bf	Inline and remove parse_local_mk. ... It has a single use. This makes the `let` handling case in `parse_stmt_without_recovery` more similar to the statement path and statement expression cases.	2024-07-31 12:08:55 +10:00
carbotaniuman	49db8a5a99	Add toggle for parse_meta_item unsafe parsing ... This makes it possible for the `unsafe(...)` syntax to only be valid at the top level, and the `NestedMetaItem`s will automatically reject `unsafe(...)`.	2024-07-30 18:28:43 -05:00
Matthias Krüger	6f0b237c72	Rollup merge of #128376 - compiler-errors:finish-ur-vegetables, r=jieyouxu ... Mark `Parser::eat`/`check` methods as `#[must_use]` These methods return a `bool`, but we probably should either use these values or explicitly throw them away (e.g. when we just want to unconditionally eat a token if it exists). I changed a few places from `eat` to `expect`, but otherwise I tried to leave a comment explaining why the `eat` was okay. This also adds a test for the `pattern_type!` macro, which used to silently accept a missing `is` token.	2024-07-30 22:51:38 +02:00
bors	f8060d282d	Auto merge of #128083 - Mark-Simulacrum:bump-bootstrap, r=albertlarsan68 ... Bump bootstrap compiler to new beta https://forge.rust-lang.org/release/process.html#master-bootstrap-update-t-2-day-tuesday	2024-07-30 17:49:08 +00:00
bors	595316b400	Auto merge of #127955 - chenyukang:yukang-fix-mismatched-delimiter-issue-12786, r=nnethercote ... Add limit for unclosed delimiters in lexer diagnostic Fixes #127868 The first commit shows the original diagnostic, and the second commit shows the changes.	2024-07-30 13:02:16 +00:00
carbotaniuman	d8bc8761a5	Deny unsafe on more builtin attributes	2024-07-29 21:00:09 -05:00
Michael Goulet	e4076e34f8	Mark Parser::eat/check methods as must_use	2024-07-29 21:29:08 -04:00
Nicholas Nethercote	84ac80f192	Reformat use declarations. ... The previous commit updated `rustfmt.toml` appropriately. This commit is the outcome of running `x fmt --all` with the new formatting options.	2024-07-29 08:26:52 +10:00
Mark Rousskov	5eca36d27a	step cfg(bootstrap)	2024-07-28 14:46:29 -04:00
Trevor Gross	9164dbd48c	Rollup merge of #128207 - folkertdev:asm-parser-generalize, r=Amanieu ... improve error message when `global_asm!` uses `asm!` options specifically, what was error: expected one of `)`, `att_syntax`, or `raw`, found `preserves_flags` --> $DIR/bad-options.rs:45:25 | LL | global_asm!("", options(preserves_flags)); | ^^^^^^^^^^^^^^^ expected one of `)`, `att_syntax`, or `raw` is now error: the `preserves_flags` option cannot be used with `global_asm!` --> $DIR/bad-options.rs:45:25 | LL | global_asm!("", options(preserves_flags)); | ^^^^^^^^^^^^^^^ the `preserves_flags` option is not meaningful for global-scoped inline assembly mirroring the phrasing of the [reference](https://doc.rust-lang.org/reference/inline-assembly.html#options). This is also a bit of a refactor for a future `naked_asm!` macro (for use in `#[naked]` functions). Currently this sort of error can come up when switching from inline to global asm, or when a user just isn't that experienced with assembly. With `naked_asm!` added to the mix hitting this error is more likely.	2024-07-27 13:32:56 -04:00
Trevor Gross	7eaf74743b	Rollup merge of #128229 - tdittr:unsafe-extern-abi-error, r=compiler-errors ... Improve `extern "<abi>" unsafe fn()` error message These errors were already reported in #87217, and fixed by #87235 but missed the case of an explicit ABI. This PR does not cover multiple keywords like `extern "C" pub const unsafe fn()`, but I don't know what a good way to cover this would be. It also seems rarer than `extern "C" unsafe` which I saw happen a few times in workshops.	2024-07-26 19:03:08 -04:00
Trevor Gross	af52be2cea	Rollup merge of #128224 - nnethercote:fewer-replace_ranges, r=petrochenkov ... Remove unnecessary range replacements This PR removes an unnecessary range replacement in `collect_tokens_trailing_token`, and does a couple of other small cleanups. r? ````@petrochenkov````	2024-07-26 19:03:06 -04:00
Trevor Gross	553a64f412	Rollup merge of #128223 - nnethercote:refactor-collect_tokens, r=petrochenkov ... Refactor complex conditions in `collect_tokens_trailing_token` More readability improvements for this complicated function. r? ````@petrochenkov````	2024-07-26 19:03:06 -04:00
Tamme Dittrich	3fdc99193e	Improve error message for extern "C" unsafe fn() ... This was handled correctly already for `extern unsafe fn()`. Co-authored-by: Folkert <folkert@folkertdev.nl>	2024-07-26 15:14:05 +02:00
Nicholas Nethercote	55d37ae711	Remove an unnecessary block.	2024-07-26 17:37:03 +10:00
Nicholas Nethercote	6ea2da5a28	Tweak a loop. ... A fully imperative style is easier to read than a half-iterator, half-imperative style. Also, rename `inner_attr` as `attr` because it might be an outer attribute.	2024-07-26 17:37:03 +10:00
Nicholas Nethercote	6e87858f26	Fix a comment. ... Imagine you have replace ranges (2..20,X) and (5..15,Y), and these tokens: ``` a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x ``` If we replace (5..15,Y) first, then (2..20,X) we get this sequence ``` a,b,c,d,e,Y,_,_,_,_,_,_,_,_,_,p,q,r,s,t,u,v,w,x a,b,X,_,_,_,_,_,_,_,_,_,_,_,_,_,_,_,_,_,u,v,w,x ``` which is what we want. If we do it in the other order, we get this: ``` a,b,X,_,_,_,_,_,_,_,_,_,_,_,_,p,q,r,s,t,u,v,w,x a,b,X,_,_,Y,_,_,_,_,_,_,_,_,_,_,_,_,_,_,u,v,w,x ``` which is wrong. So it's true that we need the `.rev()` but the comment is wrong about why.	2024-07-26 17:37:03 +10:00
Nicholas Nethercote	a560810a69	Don't include inner attribute ranges in CaptureState. ... The current code is this: ``` self.capture_state.replace_ranges.push((start_pos..end_pos, Some(target))); self.capture_state.replace_ranges.extend(inner_attr_replace_ranges); ``` What's not obvious is that every range in `inner_attr_replace_ranges` must be a strict sub-range of `start_pos..end_pos`. Which means, in `LazyAttrTokenStreamImpl::to_attr_token_stream`, they will be done first, and then the `start_pos..end_pos` replacement will just overwrite them. So they aren't needed.	2024-07-26 14:18:20 +10:00
Nicholas Nethercote	e631b1ebfa	Invert the sense of is_complete and rename it as needs_tokens. ... I have always found `is_complete` an unhelpful name. The new name (and inverted sense) fits in better with the conditions at its call sites.	2024-07-26 09:58:34 +10:00
Nicholas Nethercote	3d363c3d99	Move is_complete to the module that uses it. ... And make it non-`pub`.	2024-07-26 09:44:39 +10:00
Nicholas Nethercote	4288edb219	Inline and remove AttrWrapper::is_complete. ... It has a single call site. This change makes the two `needs_collect` conditions more similar to each other, and therefore easier to understand.	2024-07-26 09:44:07 +10:00
Nicholas Nethercote	caee195bdd	Invert early exit conditions in collect_tokens_trailing_token. ... This has been bugging me for a while. I find complex "if any of these are true" conditions easier to think about than complex "if all of these are true" conditions, because you can stop as soon as one is true.	2024-07-26 09:43:41 +10:00
Folkert	d3858f7465	improve error message when global_asm! uses asm! options	2024-07-25 22:33:52 +02:00
surechen	4ac60601d3	Fix a span error when parsing a wrong param of function. ... fixes #128042	2024-07-25 22:33:45 +08:00