Commit Graph

1799 Commits

Author SHA1 Message Date
mcarton f7cc6dc1ed Fix bad error message with ::< in types 2016-10-28 12:52:41 +02:00
Eduard Burtescu 9908711e5e Implement field shorthands in struct literal expressions. 2016-10-27 03:15:13 +03:00
Vadim Petrochenkov fea630ef9d Tweak path parsing logic 2016-10-20 20:28:10 +03:00
Vadim Petrochenkov 65ff4ca294 Refactor parser lookahead buffer and increase its size 2016-10-20 20:28:10 +03:00
Jeffrey Seyfried 8b0c292a72 Improve $crate. 2016-10-19 10:03:06 +00:00
Nicholas Nethercote 3c4c85947c Rename Parser::last_token_kind as prev_token_kind.
Likewise, rename LastTokenKind as PrevTokenKind.

This is a [breaking-change] for libsyntax.
2016-10-05 08:53:18 +11:00
Nicholas Nethercote 2747923c27 Rename Parser::last_span as prev_span.
This is a [breaking-change] for libsyntax.
2016-10-05 08:53:18 +11:00
Manish Goregaokar 406fe7e3c2 Rollup merge of #34764 - pnkfelix:attrs-on-generic-formals, r=eddyb
First step for #34761
2016-10-01 19:22:39 +05:30
Manish Goregaokar 259d1fcd47 Rollup merge of #36599 - jonas-schievink:whats-a-pirates-favorite-data-structure, r=pnkfelix
Contains a syntax-[breaking-change] as a separate commit (cc #31645).nnAlso renames slice patterns from `PatKind::Vec` to `PatKind::Slice`.
2016-10-01 19:22:12 +05:30
Jonathan Turner f1ea5cc273 Rollup merge of #36789 - jseyfried:non_inline_mod_in_block, r=nikomatsakis
Allow more non-inline modules in blocks

Currently, non-inline modules without a `#[path]` attribute are not allowed in blocks.
This PR allows non-inline modules that have an ancestor module with a `#[path]` attribute, provided there is not a nearer ancestor block.

For example,
```rust
fn main() {
    #[path = "..."] mod foo {
        mod bar; //< allowed by this PR
        fn f() {
            mod bar; //< still an error
        }
    }
}
```

Fixes #36772.
r? @nikomatsakis
2016-09-28 20:21:52 -07:00
Jeffrey Seyfried ce5ad1da12 Allow non-inline modules in more places. 2016-09-28 22:16:20 +00:00
Jonas Schievink 48e5199de3 libsyntax: clearer names for some AST parts
This applies the HIR changes from the previous commits to the AST, and
is thus a syntax-[breaking-change]

Renames `PatKind::Vec` to `PatKind::Slice`, since these are called slice
patterns, not vec patterns. Renames `TyKind::Vec`, which represents the
type `[T]`, to `TyKind::Slice`. Renames `TyKind::FixedLengthVec` to
`TyKind::Array`.
2016-09-28 22:31:18 +02:00
Jonathan Turner 5cc9fb9ed9 Rollup merge of #36669 - jseyfried:refactor_tok_result, r=nrc
Unify `TokResult` and `ResultAnyMacro`

Fixes #36641.
r? @nrc
2016-09-26 17:29:48 -07:00
Jonathan Turner 1d9646228d Rollup merge of #36662 - jseyfried:parse_macro_invoc_paths, r=nrc
parser: support paths in bang macro invocations (e.g. `path::to::macro!()`)

r? @nrc
2016-09-26 17:29:47 -07:00
Jeffrey Seyfried df0e4bf911 Move ensure_complete_parse into expand.rs. 2016-09-26 11:24:10 +00:00
Jeffrey Seyfried b90ceddcee Refactor ensure_complete_parse. 2016-09-26 04:29:30 +00:00
Felix S. Klock II 3a9b7be10b Added tests and fixed corner case for trailing attributes with no attached binding in generics. 2016-09-23 17:01:04 +02:00
Felix S. Klock II 4c37ad6607 Add attribute support to generic lifetime and type parameters.
I am using `ThinAttributes` rather than a vector for attributes
attached to generics, since I expect almost all lifetime and types
parameters to not carry any attributes.
2016-09-23 17:01:04 +02:00
Jeffrey Seyfried 2c85733521 Fix indents. 2016-09-23 04:27:08 +00:00
Jeffrey Seyfried 1e1804db18 Cleanup. 2016-09-23 04:27:08 +00:00
Jeffrey Seyfried 6c08d03039 Parse paths in item, trait item, and impl item macro invocations. 2016-09-23 04:26:59 +00:00
Jeffrey Seyfried a0e178db79 Parse paths in statement and pattern macro invocations. 2016-09-23 04:26:56 +00:00
Nick Cameron 3863834d9c reviewer comments and rebasing 2016-09-23 07:19:31 +12:00
Jeffrey Seyfried dc7ed303f7 Refactor out parse_struct_expr. 2016-09-22 07:34:03 +00:00
bors f39039e6e5 Auto merge of #36527 - nnethercote:last_token_kind, r=jseyfried
Optimize the parser's last token handling.

The parser currently makes a heap copy of the last token in four cases:
identifiers, paths, doc comments, and commas. The identifier and
interpolation cases are unused, and for doc comments and commas we only
need to record their presence, not their value.

This commit consolidates the last token handling and avoids the
unnecessary copies by replacing `last_token`, `last_token_eof`, and
`last_token_interpolated` with a new field `last_token_kind`. This
simplifies the parser slightly and speeds up parsing on some files by
3--4%.
2016-09-18 00:48:51 -07:00
bors 0b03ba1f55 Auto merge of #36502 - TimNN:correct-cancel, r=jseyfried
correctly cancel some errors

Fixes #36499.

I also (proactively) changed all other calls in `parser.rs` to use `Handler::cancel`.
2016-09-17 20:57:05 -07:00
bors 141012dd52 Auto merge of #36482 - jseyfried:dont_load_unconfigured_noninline_modules, r=nrc
Avoid loading and parsing unconfigured non-inline modules.

For example, `#[cfg(any())] mod foo;` will always compile after this PR, even if `foo.rs` and `foo/mod.rs` do not exist or do not contain valid Rust.

Fixes #36478 and fixes #27873.

r? @nrc
2016-09-16 23:29:15 -07:00
Nicholas Nethercote 8075d54606 Optimize the parser's last token handling.
The parser currently makes a heap copy of the last token in four cases:
identifiers, paths, doc comments, and commas. The identifier and
interpolation cases are unused, and for doc comments and commas we only
need to record their presence, not their value.

This commit consolidates the last token handling and avoids the
unnecessary copies by replacing `last_token`, `last_token_eof`, and
`last_token_interpolated` with a new field `last_token_kind`. This
simplifies the parser slightly and speeds up parsing on some files by
3--4%.
2016-09-16 15:46:40 +10:00
Tim Neumann 9f4e908360 correctly cancel some errors 2016-09-15 22:12:22 +02:00
Jeffrey Seyfried b232f6d9fe Avoid loading and parsing unconfigured non-inline modules. 2016-09-15 08:16:18 +00:00
Vadim Petrochenkov b57f1099b5 Remove parsing of obsolete pre-1.0 syntaxes 2016-09-13 23:33:50 +03:00
Jeffrey Seyfried e1e5c14bad In Parser and ExtCtxt, replace fields filename and mod_path_stack
with a single field `directory: PathBuf`.
2016-09-05 04:52:48 +00:00
Vadim Petrochenkov f3b41c18a8 Check fields in union patters/expressions
Make parsing of union items backward compatible
Add some tests
2016-09-03 13:39:34 +03:00
Vadim Petrochenkov a014323e45 Lower unions from AST to HIR and from HIR to types
Parse union items and add a feature for them
2016-09-03 13:39:33 +03:00
Jeffrey Seyfried bd38e890ee Rollup merge of #35480 - KiChjang:e0379-bonus, r=nikomatsakis
Move E0379 check from typeck to ast validation

Part of #35233.
Extension of #35338, #35364.
Fixes #35404.
2016-08-28 10:32:52 +00:00
Jeffrey Seyfried 37f30173a0 Rollup merge of #35618 - jseyfried:ast_view_path_refactor, r=eddyb
Refactor `PathListItem`s

This refactors away variant `Mod` of `ast::PathListItemKind` and refactors the remaining variant `Ident` to a struct `ast::PathListItem_`.
2016-08-28 10:31:24 +00:00
Jeffrey Seyfried b833e8d0a0 Rollup merge of #35591 - GuillaumeGomez:generics_span, r=jntrmr
Add Span field for Generics structs
2016-08-28 10:30:08 +00:00
Keith Yeung aa5c4bb05d Change Constness to Spanned<Constness> 2016-08-27 22:43:51 -07:00
James Miller 72d629caa5 Improve error message when failing to parse a block
We want to catch this error:

```
if (foo)
    bar;
```

as it's valid syntax in other languages, and say how to fix it.
Unfortunately it didn't care if the suggestion made sense and just
highlighted the unexpected token.

Now it attempts to parse a statement, and if it succeeds, it shows the
help message.

Fixes #35907
2016-08-23 18:23:31 +12:00
Jeffrey Seyfried 98ce875b58 Refactor away variant ast::PathListItemKind::Mod
and refactor `ast::PathListItemKind::Ident` -> `ast::PathListItem_`.
2016-08-21 22:59:36 +00:00
bors 38fa82a314 Auto merge of #33922 - estebank:doc-comment, r=alexcrichton
Specific error message for missplaced doc comments

Identify when documetation comments have been missplaced in the following places:

 * After a struct element:

    ```rust
    // file.rs:
    struct X {
        a: u8 /** document a */,
    }
    ```

    ```bash
    $ rustc file.rs
    file.rs:2:11: 2:28 error: found documentation comment that doesn't
    document anything
    file.rs:2     a: u8 /** document a */,
                        ^~~~~~~~~~~~~~~~~
    file.rs:2:11: 2:28 help: doc comments must come before what they document,
    maybe a comment was intended with `//`?
    ```

 * As the last line of a struct:

    ```rust
    // file.rs:
    struct X {
        a: u8,
        /// incorrect documentation
    }
    ```

    ```bash
    $ rustc file.rs
    file.rs:3:5: 3:27 error: found a documentation comment that doesn't
    document anything
    file.rs:3     /// incorrect documentation
                  ^~~~~~~~~~~~~~~~~~~~~~
    file.rs:3:5: 3:27 help: doc comments must come before what they document,
    maybe a comment was intended with `//`?
    ```

 * As the last line of a `fn`:

    ```rust
    // file.rs:
    fn main() {
        let x = 1;
        /// incorrect documentation
    }
    ```

    ```bash
    $ rustc file.rs
    file.rs:3:5: 3:27 error: found a documentation comment that doesn't
    document anything
    file.rs:3     /// incorrect documentation
                  ^~~~~~~~~~~~~~~~~~~~~~
    file.rs:3:5: 3:27 help: doc comments must come before what they document,
    maybe a comment was intended with `//`?
    ```

Fix #27429, #30322
2016-08-19 18:14:53 -07:00
Guillaume Gomez 5948182367 Add Span field for Generics structs 2016-08-18 18:23:36 +02:00
bors e25542cb02 Auto merge of #35162 - canndrew:bang_type_coerced, r=nikomatsakis
Implement the `!` type

This implements the never type (`!`) and hides it behind the feature gate `#[feature(never_type)]`. With the feature gate off, things should build as normal (although some error messages may be different). With the gate on, `!` is usable as a type and diverging type variables (ie. types that are unconstrained by anything in the code) will default to `!` instead of `()`.
2016-08-16 00:12:12 -07:00
Eduard-Mihai Burtescu c35d8dba7f Rollup merge of #35491 - sanxiyn:pub-restricted-span, r=nikomatsakis
Correct span for pub_restricted field

Fix #35435.
2016-08-14 20:29:48 +03:00
Andrew Cann fadabe08f5 Rename empty/bang to never
Split Ty::is_empty method into is_never and is_uninhabited
2016-08-13 21:37:09 +08:00
Andrew Cann 104963c539 Switch on TyEmpty
Parse -> ! as FnConverging(!)
Add AdjustEmptyToAny coercion to all ! expressions
Some fixes
2016-08-13 21:37:09 +08:00
Andrew Cann b0a9acd783 Parse ! as TyEmpty (except in fn return type) 2016-08-13 21:37:09 +08:00
Vadim Petrochenkov f6624782d4 Parse numeric fields in struct expressions and patterns 2016-08-13 00:08:14 +03:00
Seo Sanghyeon f76a737bae Correct span for pub_restricted field 2016-08-12 21:08:02 +09:00
Eduard Burtescu f0baec691f syntax: add anonymized type syntax, i.e. impl TraitA+TraitB. 2016-08-12 06:43:34 +03:00