vendor/rustc-ap-rustc_ast/src/tokenstream.rs

   1 //! # Token Streams
   2 //!
   3 //! `TokenStream`s represent syntactic objects before they are converted into ASTs.
   4 //! A `TokenStream` is, roughly speaking, a sequence of [`TokenTree`]s,
   5 //! which are themselves a single [`Token`] or a `Delimited` subsequence of tokens.
   6 //!
   7 //! ## Ownership
   8 //!
   9 //! `TokenStream`s are persistent data structures constructed as ropes with reference
  10 //! counted-children. In general, this means that calling an operation on a `TokenStream`
  11 //! (such as `slice`) produces an entirely new `TokenStream` from the borrowed reference to
  12 //! the original. This essentially coerces `TokenStream`s into "views" of their subparts,
  13 //! and a borrowed `TokenStream` is sufficient to build an owned `TokenStream` without taking
  14 //! ownership of the original.
  15
  16 use crate::token::{self, DelimToken, Token, TokenKind};
  17
  18 use rustc_data_structures::stable_hasher::{HashStable, StableHasher};
  19 use rustc_data_structures::sync::{self, Lrc};
  20 use rustc_macros::HashStable_Generic;
  21 use rustc_serialize::{Decodable, Decoder, Encodable, Encoder};
  22 use rustc_span::{Span, DUMMY_SP};
  23 use smallvec::{smallvec, SmallVec};
  24
  25 use std::{fmt, iter, mem};
  26
  27 /// When the main Rust parser encounters a syntax-extension invocation, it
  28 /// parses the arguments to the invocation as a token tree. This is a very
  29 /// loose structure, such that all sorts of different AST fragments can
  30 /// be passed to syntax extensions using a uniform type.
  31 ///
  32 /// If the syntax extension is an MBE macro, it will attempt to match its
  33 /// LHS token tree against the provided token tree, and if it finds a
  34 /// match, will transcribe the RHS token tree, splicing in any captured
  35 /// `macro_parser::matched_nonterminals` into the `SubstNt`s it finds.
  36 ///
  37 /// The RHS of an MBE macro is the only place `SubstNt`s are substituted.
  38 /// Nothing special happens to misnamed or misplaced `SubstNt`s.
  39 #[derive(Debug, Clone, PartialEq, Encodable, Decodable, HashStable_Generic)]
  40 pub enum TokenTree {
  41     /// A single token.
  42     Token(Token),
  43     /// A delimited sequence of token trees.
  44     Delimited(DelimSpan, DelimToken, TokenStream),
  45 }
  46
  47 #[derive(Copy, Clone)]
  48 pub enum CanSynthesizeMissingTokens {
  49     Yes,
  50     No,
  51 }
  52
  53 // Ensure all fields of `TokenTree` is `Send` and `Sync`.
  54 #[cfg(parallel_compiler)]
  55 fn _dummy()
  56 where
  57     Token: Send + Sync,
  58     DelimSpan: Send + Sync,
  59     DelimToken: Send + Sync,
  60     TokenStream: Send + Sync,
  61 {
  62 }
  63
  64 impl TokenTree {
  65     /// Checks if this `TokenTree` is equal to the other, regardless of span information.
  66     pub fn eq_unspanned(&self, other: &TokenTree) -> bool {
  67         match (self, other) {
  68             (TokenTree::Token(token), TokenTree::Token(token2)) => token.kind == token2.kind,
  69             (TokenTree::Delimited(_, delim, tts), TokenTree::Delimited(_, delim2, tts2)) => {
  70                 delim == delim2 && tts.eq_unspanned(&tts2)
  71             }
  72             _ => false,
  73         }
  74     }
  75
  76     /// Retrieves the `TokenTree`'s span.
  77     pub fn span(&self) -> Span {
  78         match self {
  79             TokenTree::Token(token) => token.span,
  80             TokenTree::Delimited(sp, ..) => sp.entire(),
  81         }
  82     }
  83
  84     /// Modify the `TokenTree`'s span in-place.
  85     pub fn set_span(&mut self, span: Span) {
  86         match self {
  87             TokenTree::Token(token) => token.span = span,
  88             TokenTree::Delimited(dspan, ..) => *dspan = DelimSpan::from_single(span),
  89         }
  90     }
  91
  92     pub fn joint(self) -> TokenStream {
  93         TokenStream::new(vec![(self, Spacing::Joint)])
  94     }
  95
  96     pub fn token(kind: TokenKind, span: Span) -> TokenTree {
  97         TokenTree::Token(Token::new(kind, span))
  98     }
  99
 100     /// Returns the opening delimiter as a token tree.
 101     pub fn open_tt(span: DelimSpan, delim: DelimToken) -> TokenTree {
 102         TokenTree::token(token::OpenDelim(delim), span.open)
 103     }
 104
 105     /// Returns the closing delimiter as a token tree.
 106     pub fn close_tt(span: DelimSpan, delim: DelimToken) -> TokenTree {
 107         TokenTree::token(token::CloseDelim(delim), span.close)
 108     }
 109
 110     pub fn uninterpolate(self) -> TokenTree {
 111         match self {
 112             TokenTree::Token(token) => TokenTree::Token(token.uninterpolate().into_owned()),
 113             tt => tt,
 114         }
 115     }
 116 }
 117
 118 impl<CTX> HashStable<CTX> for TokenStream
 119 where
 120     CTX: crate::HashStableContext,
 121 {
 122     fn hash_stable(&self, hcx: &mut CTX, hasher: &mut StableHasher) {
 123         for sub_tt in self.trees() {
 124             sub_tt.hash_stable(hcx, hasher);
 125         }
 126     }
 127 }
 128
 129 pub trait CreateTokenStream: sync::Send + sync::Sync {
 130     fn create_token_stream(&self) -> TokenStream;
 131 }
 132
 133 impl CreateTokenStream for TokenStream {
 134     fn create_token_stream(&self) -> TokenStream {
 135         self.clone()
 136     }
 137 }
 138
 139 /// A lazy version of [`TokenStream`], which defers creation
 140 /// of an actual `TokenStream` until it is needed.
 141 /// `Box` is here only to reduce the structure size.
 142 #[derive(Clone)]
 143 pub struct LazyTokenStream(Lrc<Box<dyn CreateTokenStream>>);
 144
 145 impl LazyTokenStream {
 146     pub fn new(inner: impl CreateTokenStream + 'static) -> LazyTokenStream {
 147         LazyTokenStream(Lrc::new(Box::new(inner)))
 148     }
 149
 150     pub fn create_token_stream(&self) -> TokenStream {
 151         self.0.create_token_stream()
 152     }
 153 }
 154
 155 impl fmt::Debug for LazyTokenStream {
 156     fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
 157         fmt::Debug::fmt("LazyTokenStream", f)
 158     }
 159 }
 160
 161 impl<S: Encoder> Encodable<S> for LazyTokenStream {
 162     fn encode(&self, s: &mut S) -> Result<(), S::Error> {
 163         // Used by AST json printing.
 164         Encodable::encode(&self.create_token_stream(), s)
 165     }
 166 }
 167
 168 impl<D: Decoder> Decodable<D> for LazyTokenStream {
 169     fn decode(_d: &mut D) -> Result<Self, D::Error> {
 170         panic!("Attempted to decode LazyTokenStream");
 171     }
 172 }
 173
 174 impl<CTX> HashStable<CTX> for LazyTokenStream {
 175     fn hash_stable(&self, _hcx: &mut CTX, _hasher: &mut StableHasher) {
 176         panic!("Attempted to compute stable hash for LazyTokenStream");
 177     }
 178 }
 179
 180 /// A `TokenStream` is an abstract sequence of tokens, organized into [`TokenTree`]s.
 181 ///
 182 /// The goal is for procedural macros to work with `TokenStream`s and `TokenTree`s
 183 /// instead of a representation of the abstract syntax tree.
 184 /// Today's `TokenTree`s can still contain AST via `token::Interpolated` for
 185 /// backwards compatability.
 186 #[derive(Clone, Debug, Default, Encodable, Decodable)]
 187 pub struct TokenStream(pub(crate) Lrc<Vec<TreeAndSpacing>>);
 188
 189 pub type TreeAndSpacing = (TokenTree, Spacing);
 190
 191 // `TokenStream` is used a lot. Make sure it doesn't unintentionally get bigger.
 192 #[cfg(target_arch = "x86_64")]
 193 rustc_data_structures::static_assert_size!(TokenStream, 8);
 194
 195 #[derive(Clone, Copy, Debug, PartialEq, Encodable, Decodable)]
 196 pub enum Spacing {
 197     Alone,
 198     Joint,
 199 }
 200
 201 impl TokenStream {
 202     /// Given a `TokenStream` with a `Stream` of only two arguments, return a new `TokenStream`
 203     /// separating the two arguments with a comma for diagnostic suggestions.
 204     pub fn add_comma(&self) -> Option<(TokenStream, Span)> {
 205         // Used to suggest if a user writes `foo!(a b);`
 206         let mut suggestion = None;
 207         let mut iter = self.0.iter().enumerate().peekable();
 208         while let Some((pos, ts)) = iter.next() {
 209             if let Some((_, next)) = iter.peek() {
 210                 let sp = match (&ts, &next) {
 211                     (_, (TokenTree::Token(Token { kind: token::Comma, .. }), _)) => continue,
 212                     (
 213                         (TokenTree::Token(token_left), Spacing::Alone),
 214                         (TokenTree::Token(token_right), _),
 215                     ) if ((token_left.is_ident() && !token_left.is_reserved_ident())
 216                         || token_left.is_lit())
 217                         && ((token_right.is_ident() && !token_right.is_reserved_ident())
 218                             || token_right.is_lit()) =>
 219                     {
 220                         token_left.span
 221                     }
 222                     ((TokenTree::Delimited(sp, ..), Spacing::Alone), _) => sp.entire(),
 223                     _ => continue,
 224                 };
 225                 let sp = sp.shrink_to_hi();
 226                 let comma = (TokenTree::token(token::Comma, sp), Spacing::Alone);
 227                 suggestion = Some((pos, comma, sp));
 228             }
 229         }
 230         if let Some((pos, comma, sp)) = suggestion {
 231             let mut new_stream = Vec::with_capacity(self.0.len() + 1);
 232             let parts = self.0.split_at(pos + 1);
 233             new_stream.extend_from_slice(parts.0);
 234             new_stream.push(comma);
 235             new_stream.extend_from_slice(parts.1);
 236             return Some((TokenStream::new(new_stream), sp));
 237         }
 238         None
 239     }
 240 }
 241
 242 impl From<TokenTree> for TokenStream {
 243     fn from(tree: TokenTree) -> TokenStream {
 244         TokenStream::new(vec![(tree, Spacing::Alone)])
 245     }
 246 }
 247
 248 impl From<TokenTree> for TreeAndSpacing {
 249     fn from(tree: TokenTree) -> TreeAndSpacing {
 250         (tree, Spacing::Alone)
 251     }
 252 }
 253
 254 impl iter::FromIterator<TokenTree> for TokenStream {
 255     fn from_iter<I: IntoIterator<Item = TokenTree>>(iter: I) -> Self {
 256         TokenStream::new(iter.into_iter().map(Into::into).collect::<Vec<TreeAndSpacing>>())
 257     }
 258 }
 259
 260 impl Eq for TokenStream {}
 261
 262 impl PartialEq<TokenStream> for TokenStream {
 263     fn eq(&self, other: &TokenStream) -> bool {
 264         self.trees().eq(other.trees())
 265     }
 266 }
 267
 268 impl TokenStream {
 269     pub fn new(streams: Vec<TreeAndSpacing>) -> TokenStream {
 270         TokenStream(Lrc::new(streams))
 271     }
 272
 273     pub fn is_empty(&self) -> bool {
 274         self.0.is_empty()
 275     }
 276
 277     pub fn len(&self) -> usize {
 278         self.0.len()
 279     }
 280
 281     pub fn span(&self) -> Option<Span> {
 282         match &**self.0 {
 283             [] => None,
 284             [(tt, _)] => Some(tt.span()),
 285             [(tt_start, _), .., (tt_end, _)] => Some(tt_start.span().to(tt_end.span())),
 286         }
 287     }
 288
 289     pub fn from_streams(mut streams: SmallVec<[TokenStream; 2]>) -> TokenStream {
 290         match streams.len() {
 291             0 => TokenStream::default(),
 292             1 => streams.pop().unwrap(),
 293             _ => {
 294                 // We are going to extend the first stream in `streams` with
 295                 // the elements from the subsequent streams. This requires
 296                 // using `make_mut()` on the first stream, and in practice this
 297                 // doesn't cause cloning 99.9% of the time.
 298                 //
 299                 // One very common use case is when `streams` has two elements,
 300                 // where the first stream has any number of elements within
 301                 // (often 1, but sometimes many more) and the second stream has
 302                 // a single element within.
 303
 304                 // Determine how much the first stream will be extended.
 305                 // Needed to avoid quadratic blow up from on-the-fly
 306                 // reallocations (#57735).
 307                 let num_appends = streams.iter().skip(1).map(|ts| ts.len()).sum();
 308
 309                 // Get the first stream. If it's `None`, create an empty
 310                 // stream.
 311                 let mut iter = streams.drain(..);
 312                 let mut first_stream_lrc = iter.next().unwrap().0;
 313
 314                 // Append the elements to the first stream, after reserving
 315                 // space for them.
 316                 let first_vec_mut = Lrc::make_mut(&mut first_stream_lrc);
 317                 first_vec_mut.reserve(num_appends);
 318                 for stream in iter {
 319                     first_vec_mut.extend(stream.0.iter().cloned());
 320                 }
 321
 322                 // Create the final `TokenStream`.
 323                 TokenStream(first_stream_lrc)
 324             }
 325         }
 326     }
 327
 328     pub fn trees_ref(&self) -> CursorRef<'_> {
 329         CursorRef::new(self)
 330     }
 331
 332     pub fn trees(&self) -> Cursor {
 333         self.clone().into_trees()
 334     }
 335
 336     pub fn into_trees(self) -> Cursor {
 337         Cursor::new(self)
 338     }
 339
 340     /// Compares two `TokenStream`s, checking equality without regarding span information.
 341     pub fn eq_unspanned(&self, other: &TokenStream) -> bool {
 342         let mut t1 = self.trees();
 343         let mut t2 = other.trees();
 344         for (t1, t2) in t1.by_ref().zip(t2.by_ref()) {
 345             if !t1.eq_unspanned(&t2) {
 346                 return false;
 347             }
 348         }
 349         t1.next().is_none() && t2.next().is_none()
 350     }
 351
 352     pub fn map_enumerated<F: FnMut(usize, &TokenTree) -> TokenTree>(self, mut f: F) -> TokenStream {
 353         TokenStream(Lrc::new(
 354             self.0
 355                 .iter()
 356                 .enumerate()
 357                 .map(|(i, (tree, is_joint))| (f(i, tree), *is_joint))
 358                 .collect(),
 359         ))
 360     }
 361 }
 362
 363 // 99.5%+ of the time we have 1 or 2 elements in this vector.
 364 #[derive(Clone)]
 365 pub struct TokenStreamBuilder(SmallVec<[TokenStream; 2]>);
 366
 367 impl TokenStreamBuilder {
 368     pub fn new() -> TokenStreamBuilder {
 369         TokenStreamBuilder(SmallVec::new())
 370     }
 371
 372     pub fn push<T: Into<TokenStream>>(&mut self, stream: T) {
 373         let mut stream = stream.into();
 374
 375         // If `self` is not empty and the last tree within the last stream is a
 376         // token tree marked with `Joint`...
 377         if let Some(TokenStream(ref mut last_stream_lrc)) = self.0.last_mut() {
 378             if let Some((TokenTree::Token(last_token), Spacing::Joint)) = last_stream_lrc.last() {
 379                 // ...and `stream` is not empty and the first tree within it is
 380                 // a token tree...
 381                 let TokenStream(ref mut stream_lrc) = stream;
 382                 if let Some((TokenTree::Token(token), spacing)) = stream_lrc.first() {
 383                     // ...and the two tokens can be glued together...
 384                     if let Some(glued_tok) = last_token.glue(&token) {
 385                         // ...then do so, by overwriting the last token
 386                         // tree in `self` and removing the first token tree
 387                         // from `stream`. This requires using `make_mut()`
 388                         // on the last stream in `self` and on `stream`,
 389                         // and in practice this doesn't cause cloning 99.9%
 390                         // of the time.
 391
 392                         // Overwrite the last token tree with the merged
 393                         // token.
 394                         let last_vec_mut = Lrc::make_mut(last_stream_lrc);
 395                         *last_vec_mut.last_mut().unwrap() = (TokenTree::Token(glued_tok), *spacing);
 396
 397                         // Remove the first token tree from `stream`. (This
 398                         // is almost always the only tree in `stream`.)
 399                         let stream_vec_mut = Lrc::make_mut(stream_lrc);
 400                         stream_vec_mut.remove(0);
 401
 402                         // Don't push `stream` if it's empty -- that could
 403                         // block subsequent token gluing, by getting
 404                         // between two token trees that should be glued
 405                         // together.
 406                         if !stream.is_empty() {
 407                             self.0.push(stream);
 408                         }
 409                         return;
 410                     }
 411                 }
 412             }
 413         }
 414         self.0.push(stream);
 415     }
 416
 417     pub fn build(self) -> TokenStream {
 418         TokenStream::from_streams(self.0)
 419     }
 420 }
 421
 422 /// By-reference iterator over a [`TokenStream`].
 423 #[derive(Clone)]
 424 pub struct CursorRef<'t> {
 425     stream: &'t TokenStream,
 426     index: usize,
 427 }
 428
 429 impl<'t> CursorRef<'t> {
 430     fn new(stream: &TokenStream) -> CursorRef<'_> {
 431         CursorRef { stream, index: 0 }
 432     }
 433
 434     fn next_with_spacing(&mut self) -> Option<&'t TreeAndSpacing> {
 435         self.stream.0.get(self.index).map(|tree| {
 436             self.index += 1;
 437             tree
 438         })
 439     }
 440 }
 441
 442 impl<'t> Iterator for CursorRef<'t> {
 443     type Item = &'t TokenTree;
 444
 445     fn next(&mut self) -> Option<&'t TokenTree> {
 446         self.next_with_spacing().map(|(tree, _)| tree)
 447     }
 448 }
 449
 450 /// Owning by-value iterator over a [`TokenStream`].
 451 // FIXME: Many uses of this can be replaced with by-reference iterator to avoid clones.
 452 #[derive(Clone)]
 453 pub struct Cursor {
 454     pub stream: TokenStream,
 455     index: usize,
 456 }
 457
 458 impl Iterator for Cursor {
 459     type Item = TokenTree;
 460
 461     fn next(&mut self) -> Option<TokenTree> {
 462         self.next_with_spacing().map(|(tree, _)| tree)
 463     }
 464 }
 465
 466 impl Cursor {
 467     fn new(stream: TokenStream) -> Self {
 468         Cursor { stream, index: 0 }
 469     }
 470
 471     pub fn next_with_spacing(&mut self) -> Option<TreeAndSpacing> {
 472         if self.index < self.stream.len() {
 473             self.index += 1;
 474             Some(self.stream.0[self.index - 1].clone())
 475         } else {
 476             None
 477         }
 478     }
 479
 480     pub fn append(&mut self, new_stream: TokenStream) {
 481         if new_stream.is_empty() {
 482             return;
 483         }
 484         let index = self.index;
 485         let stream = mem::take(&mut self.stream);
 486         *self = TokenStream::from_streams(smallvec![stream, new_stream]).into_trees();
 487         self.index = index;
 488     }
 489
 490     pub fn look_ahead(&self, n: usize) -> Option<&TokenTree> {
 491         self.stream.0[self.index..].get(n).map(|(tree, _)| tree)
 492     }
 493 }
 494
 495 #[derive(Debug, Copy, Clone, PartialEq, Encodable, Decodable, HashStable_Generic)]
 496 pub struct DelimSpan {
 497     pub open: Span,
 498     pub close: Span,
 499 }
 500
 501 impl DelimSpan {
 502     pub fn from_single(sp: Span) -> Self {
 503         DelimSpan { open: sp, close: sp }
 504     }
 505
 506     pub fn from_pair(open: Span, close: Span) -> Self {
 507         DelimSpan { open, close }
 508     }
 509
 510     pub fn dummy() -> Self {
 511         Self::from_single(DUMMY_SP)
 512     }
 513
 514     pub fn entire(self) -> Span {
 515         self.open.with_hi(self.close.hi())
 516     }
 517 }