Commit graph

316 commits

Author SHA1 Message Date
Nicholas Nethercote
c60ed5d22c Share empty Vecs more within MatcherPos::matches.
`create_matches` creates a `Vec<Rc<Vec<NamedMatch>>>`. Even though all the
inner `Vec`s are empty, each one is created separately.

This commit changes `create_matches` so it instead creates one empty inner
`Vec`, and shares it.

The commit also changes `MatcherPos::matches` to a boxed slice, because its
length doesn't change.
2018-11-01 08:43:40 +11:00
ljedrz
130a32fa72 Remove OneVector 2018-09-26 10:43:37 +02:00
David Tolnay
57d6ada91d
Rename sp_lo to sp_open 2018-09-08 23:47:42 -07:00
David Tolnay
a1dd39e724
Track distinct spans for open and close delimiter 2018-09-08 19:01:48 -07:00
Eduard-Mihai Burtescu
93f3f5b155 Use FxHash{Map,Set} instead of the default Hash{Map,Set} everywhere in rustc. 2018-08-28 17:04:04 +03:00
Igor Gutorov
4d81fe9243 Use optimized SmallVec implementation 2018-08-23 10:45:53 +03:00
Matthias Krüger
71120ef1e5 Fix typos found by codespell. 2018-08-19 17:41:28 +02:00
ljedrz
e5e6375352 Move SmallVec and ThinVec out of libsyntax 2018-08-13 22:11:57 +02:00
Michael Lamparski
0467ae0cf5 cleanup unnecessary else 2018-07-16 16:15:34 -04:00
Michael Lamparski
191e76c5b8 fix perf issue in macro parser
For a fuller description of the performance issue fixed by this:

https://github.com/rust-lang/rust/issues/51754#issuecomment-403242159
2018-07-07 17:13:21 -04:00
David Tolnay
987020846c
Enable fall through past $:lifetime matcher 2018-06-10 14:39:16 -07:00
Takanori Ishibashi
f386cdae19 innacurate -> inaccurate 2018-05-27 09:47:04 +09:00
Vadim Petrochenkov
1e4269cb83 Add Ident::as_str helper 2018-05-26 15:20:23 +03:00
bors
4c26e2e3fb Auto merge of #50855 - nnethercote:fewer-macro_parser-allocs, r=petrochenkov
Speed up the macro parser

These three commits reduce the number of allocations done by the macro parser, in some cases dramatically. For example, for a clean check builds of html5ever, the number of allocations is reduced by 40%.

Here are the rustc-benchmarks that are sped up by at least 1%.
```
html5ever-check
        avg: -6.6%      min: -10.3%     max: -4.1%
html5ever
        avg: -5.2%      min: -9.5%      max: -2.8%
html5ever-opt
        avg: -4.3%      min: -9.3%      max: -1.6%
crates.io-check
        avg: -1.8%      min: -2.9%      max: -0.6%
crates.io-opt
        avg: -1.0%      min: -2.2%      max: -0.1%
crates.io
        avg: -1.1%      min: -2.2%      max: -0.2%
```
2018-05-20 08:55:50 +00:00
Nicholas Nethercote
fcf2b24e1b Introduce MatcherPosHandle.
This lets us store most `MatcherPos` instances on the stack. This speeds
up various runs of html5ever, the best by 3%.
2018-05-18 22:20:27 +10:00
Nicholas Nethercote
6872377357 Change TokenTreeOrTokenTreeVec to TokenTreeOrTokenTreeSlice.
This avoids a `to_owned` call that can be hot, speeding up the various
runs of html5ever by 1--5%, and some runs of crates.io by 2--3%.
2018-05-17 20:41:38 +10:00
Dan Aloni
37ed2ab910 Macros: Add a 'literal' fragment specifier
Implements RFC 1576.

See: https://github.com/rust-lang/rfcs/blob/master/text/1576-macros-literal-matcher.md

Changes are mostly in libsyntax, docs, and tests. Feature gate is
enabled for 1.27.0.

Many thanks to Vadim Petrochenkov for following through code reviews
and suggestions.

Example:

````rust

macro_rules! test_literal {
    ($l:literal) => {
        println!("literal: {}", $l);
    };
    ($e:expr) => {
        println!("expr: {}", $e);
    };
}

fn main() {
    let a = 1;
    test_literal!(a);
    test_literal!(2);
    test_literal!(-3);
}
```

Output:

```
expr: 1
literal: 2
literal: -3
```
2018-05-13 19:17:02 +03:00
rleungx
390c3cee6a check if the token is a lifetime before parsing 2018-05-03 19:12:39 +08:00
Vadim Petrochenkov
bfaf4180ae Make lifetime nonterminals closer to identifier nonterminals 2018-04-06 11:52:16 +03:00
Vadim Petrochenkov
b3b5ef186c Remove more duplicated spans 2018-04-06 11:50:49 +03:00
Vadim Petrochenkov
e2afefd80b Get rid of SpannedIdent 2018-04-06 11:48:19 +03:00
Lymia Aluysia
5c3d6320de
Return a is_raw parameter from Token::ident rather than having separate methods. 2018-03-18 12:16:02 -05:00
Lymia Aluysia
fad1648e0f
Initial implementation of RFC 2151, Raw Identifiers 2018-03-18 10:07:19 -05:00
Vadim Petrochenkov
ed5ea5c705 Reject _ in ident matcher 2018-03-17 22:08:18 +03:00
Vadim Petrochenkov
5d06c890fe syntax: Make _ an identifier 2018-03-17 22:08:07 +03:00
Mark Mansi
549534e438 Update a few comments 2018-01-30 16:20:46 -06:00
Mark Mansi
786b2ca155 Fix trailing whitespace 2018-01-30 12:45:35 -06:00
Mark Mansi
f59b821944 Attempted fix for ? kleene op 2018-01-30 12:41:04 -06:00
Mark Mansi
5ac48ec826 Run rustfmt on macro_parser.rs 2018-01-30 12:38:02 -06:00
Mark Mansi
bb8110c1fc Update the macro parser to allow at most once repetitions for ? Kleene 2018-01-30 12:30:41 -06:00
Mark Mansi
2184400be7 Update comment 2018-01-29 16:37:57 -06:00
Mark Mansi
e2d558ad56 A few more comments 2018-01-26 14:47:24 -06:00
Mark Mansi
02d1d92878 Still more comments 2018-01-26 14:47:24 -06:00
Mark Mansi
b01b481db3 Added/improved comments 2018-01-26 14:47:24 -06:00
Mark Mansi
6d4ed65585 Added lots of comments + minor reorganization 2018-01-26 14:47:24 -06:00
Mark Mansi
0d7f193dd3 Added a bunch of comments to macro_parser.rs 2018-01-26 14:47:24 -06:00
Mark Mansi
ac0c16d3b5 Run rustfmt on /libsyntax/ext/tt/macro_parser.rs 2018-01-26 14:47:24 -06:00
John Kåre Alsaker
9a8d6b8bb5 Do not capture stderr in the compiler. Instead just panic silently for fatal errors 2018-01-26 04:52:30 +01:00
Michael Hewson
e12b87096a replace parse_lifetime with expect_lifetime
made `parser::Parser::expect_lifetime` public, so it can be called from `macro_parser::parse_nt`
2017-12-28 11:32:05 -05:00
Michael Hewson
03a51019a4 Resurrecting #33135
Started rebasing @sgrif's PR #33135 off of current master. (Well, actually merging it into a new branch based off current master.)

The following files still need to be fixed or at least reviewed:

- `src/libsyntax/ext/tt/macro_parser.rs`: calls `Parser::parse_lifetime`, which doesn't exist anymore
- `src/libsyntax/parse/parser.rs`: @sgrif added an error message to `Parser::parse_lifetime`. Code has since been refactored, so I just took it out for now.
- `src/libsyntax/ext/tt/transcribe.rs`: This code has been refactored bigtime. Not sure whether @sgrif's changes here are still necessary. Took it out for this commit.
2017-12-28 11:32:05 -05:00
Vadim Petrochenkov
3da868dcb6 Make fields of Span private 2017-08-30 01:38:54 +03:00
bors
8df670b6a6 Auto merge of #43540 - petrochenkov:pathrelax, r=nikomatsakis
syntax: Relax path grammar

TLDR: Accept the disambiguator `::` in "type" paths (`Type::<Args>`), accept the disambiguator `::` before parenthesized generic arguments (`Fn::(Args)`).

The "turbofish" disambiguator `::<>` in expression paths is a necessary evil required for path parsing to be both simple and to give reasonable results.
Since paths in expressions usually refer to values (but not necessarily, e.g. `Struct::<u8> { field: 0 }` is disambiguated, but refers to a type), people often consider `::<>` to be inherent to *values*, and not *expressions* and want to write disambiguated paths for values even in contexts where disambiguation is not strictly necessary, for example when a path is passed to a macro `m!(Vec::<i32>::new)`.
The problem is that currently, if the disambiguator is not *required*, then it's *prohibited*. This results in confusion - see https://github.com/rust-lang/rust/issues/41740, https://internals.rust-lang.org/t/macro-path-uses-novel-syntax/5561.

This PR makes the disambiguator *optional* instead of prohibited in contexts where it's not strictly required, so people can pass paths to macros in whatever form they consider natural (e.g. disambiguated form for value paths).
This PR also accepts the disambiguator in paths with parenthesized arguments (`Fn::(Args)`) for consistency and to simplify testing of stuff like https://github.com/rust-lang/rust/pull/41856#issuecomment-301219194.

Closes https://github.com/rust-lang/rust/issues/41740

cc @rust-lang/lang
r? @nikomatsakis
2017-08-21 23:03:57 +00:00
Zack M. Davis
1b6c9605e4 use field init shorthand EVERYWHERE
Like #43008 (f668999), but _much more aggressive_.
2017-08-15 15:29:17 -07:00
Vadim Petrochenkov
804459bdca Issue warnings for unnecessary path disambiguators 2017-08-11 02:47:22 +03:00
Isaac van Bakel
400075d9d9 Fixed all unnecessary muts in language core 2017-08-01 23:01:24 +01:00
Piotr Czarnecki
346717686b Make the macro parser theory description more accurate 2017-07-24 13:57:08 +02:00
kennytm
600800480a
Only match a fragment specifier the if it starts with certain tokens.
Fixes #24189.
Fixes #26444.
Fixes #27832.
Fixes #34030.
Fixes #35650.
Fixes #39964.
Fixes the 4th comment in #40569.
Fixes the issue blocking #40984.
2017-07-07 14:12:12 +08:00
Jeffrey Seyfried
d4488b7df9 Simplify hygiene::Mark application, and
remove variant `Token::SubstNt` in favor of `quoted::TokenTree::MetaVar`.
2017-06-26 02:05:45 +00:00
Mark Simulacrum
3d9ebf2916 Speed up expansion.
This reduces duplication, thereby increasing expansion speed.
2017-06-08 08:53:31 -06:00
Nick Cameron
a2566301e1 Add an option to the parser to avoid parsing out of line modules
This is useful if parsing from stdin or a String and don't want to try and read in a module from another file. Instead we just leave a stub in the AST.
2017-05-18 11:03:07 +12:00