1 | //! Iterators for `str` methods. |
---|---|
2 | |
3 | use super::pattern::{DoubleEndedSearcher, Pattern, ReverseSearcher, Searcher}; |
4 | use super::validations::{next_code_point, next_code_point_reverse}; |
5 | use super::{ |
6 | BytesIsNotEmpty, CharEscapeDebugContinue, CharEscapeDefault, CharEscapeUnicode, |
7 | IsAsciiWhitespace, IsNotEmpty, IsWhitespace, LinesMap, UnsafeBytesToStr, from_utf8_unchecked, |
8 | }; |
9 | use crate::fmt::{self, Write}; |
10 | use crate::iter::{ |
11 | Chain, Copied, Filter, FlatMap, Flatten, FusedIterator, Map, TrustedLen, TrustedRandomAccess, |
12 | TrustedRandomAccessNoCoerce, |
13 | }; |
14 | use crate::num::NonZero; |
15 | use crate::ops::Try; |
16 | use crate::slice::{self, Split as SliceSplit}; |
17 | use crate::{char as char_mod, option}; |
18 | |
19 | /// An iterator over the [`char`]s of a string slice. |
20 | /// |
21 | /// |
22 | /// This struct is created by the [`chars`] method on [`str`]. |
23 | /// See its documentation for more. |
24 | /// |
25 | /// [`char`]: prim@char |
26 | /// [`chars`]: str::chars |
27 | #[derive(Clone)] |
28 | #[must_use= "iterators are lazy and do nothing unless consumed"] |
29 | #[stable(feature = "rust1", since = "1.0.0")] |
30 | pub struct Chars<'a> { |
31 | pub(super) iter: slice::Iter<'a, u8>, |
32 | } |
33 | |
34 | #[stable(feature = "rust1", since = "1.0.0")] |
35 | impl<'a> Iterator for Chars<'a> { |
36 | type Item = char; |
37 | |
38 | #[inline] |
39 | fn next(&mut self) -> Option<char> { |
40 | // SAFETY: `str` invariant says `self.iter` is a valid UTF-8 string and |
41 | // the resulting `ch` is a valid Unicode Scalar Value. |
42 | unsafe { next_code_point(&mut self.iter).map(|ch| char::from_u32_unchecked(ch)) } |
43 | } |
44 | |
45 | #[inline] |
46 | fn count(self) -> usize { |
47 | super::count::count_chars(self.as_str()) |
48 | } |
49 | |
50 | #[inline] |
51 | fn advance_by(&mut self, mut remainder: usize) -> Result<(), NonZero<usize>> { |
52 | const CHUNK_SIZE: usize = 32; |
53 | |
54 | if remainder >= CHUNK_SIZE { |
55 | let mut chunks = self.iter.as_slice().array_chunks::<CHUNK_SIZE>(); |
56 | let mut bytes_skipped: usize = 0; |
57 | |
58 | while remainder > CHUNK_SIZE |
59 | && let Some(chunk) = chunks.next() |
60 | { |
61 | bytes_skipped += CHUNK_SIZE; |
62 | |
63 | let mut start_bytes = [false; CHUNK_SIZE]; |
64 | |
65 | for i in 0..CHUNK_SIZE { |
66 | start_bytes[i] = !super::validations::utf8_is_cont_byte(chunk[i]); |
67 | } |
68 | |
69 | remainder -= start_bytes.into_iter().map(|i| i as u8).sum::<u8>() as usize; |
70 | } |
71 | |
72 | // SAFETY: The amount of bytes exists since we just iterated over them, |
73 | // so advance_by will succeed. |
74 | unsafe { self.iter.advance_by(bytes_skipped).unwrap_unchecked() }; |
75 | |
76 | // skip trailing continuation bytes |
77 | while self.iter.len() > 0 { |
78 | let b = self.iter.as_slice()[0]; |
79 | if !super::validations::utf8_is_cont_byte(b) { |
80 | break; |
81 | } |
82 | // SAFETY: We just peeked at the byte, therefore it exists |
83 | unsafe { self.iter.advance_by(1).unwrap_unchecked() }; |
84 | } |
85 | } |
86 | |
87 | while (remainder > 0) && (self.iter.len() > 0) { |
88 | remainder -= 1; |
89 | let b = self.iter.as_slice()[0]; |
90 | let slurp = super::validations::utf8_char_width(b); |
91 | // SAFETY: utf8 validity requires that the string must contain |
92 | // the continuation bytes (if any) |
93 | unsafe { self.iter.advance_by(slurp).unwrap_unchecked() }; |
94 | } |
95 | |
96 | NonZero::new(remainder).map_or(Ok(()), Err) |
97 | } |
98 | |
99 | #[inline] |
100 | fn size_hint(&self) -> (usize, Option<usize>) { |
101 | let len = self.iter.len(); |
102 | // `(len + 3)` can't overflow, because we know that the `slice::Iter` |
103 | // belongs to a slice in memory which has a maximum length of |
104 | // `isize::MAX` (that's well below `usize::MAX`). |
105 | ((len + 3) / 4, Some(len)) |
106 | } |
107 | |
108 | #[inline] |
109 | fn last(mut self) -> Option<char> { |
110 | // No need to go through the entire string. |
111 | self.next_back() |
112 | } |
113 | } |
114 | |
115 | #[stable(feature = "chars_debug_impl", since = "1.38.0")] |
116 | impl fmt::Debug for Chars<'_> { |
117 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
118 | write!(f, "Chars(")?; |
119 | f.debug_list().entries(self.clone()).finish()?; |
120 | write!(f, ")")?; |
121 | Ok(()) |
122 | } |
123 | } |
124 | |
125 | #[stable(feature = "rust1", since = "1.0.0")] |
126 | impl<'a> DoubleEndedIterator for Chars<'a> { |
127 | #[inline] |
128 | fn next_back(&mut self) -> Option<char> { |
129 | // SAFETY: `str` invariant says `self.iter` is a valid UTF-8 string and |
130 | // the resulting `ch` is a valid Unicode Scalar Value. |
131 | unsafe { next_code_point_reverse(&mut self.iter).map(|ch: u32| char::from_u32_unchecked(ch)) } |
132 | } |
133 | } |
134 | |
135 | #[stable(feature = "fused", since = "1.26.0")] |
136 | impl FusedIterator for Chars<'_> {} |
137 | |
138 | impl<'a> Chars<'a> { |
139 | /// Views the underlying data as a subslice of the original data. |
140 | /// |
141 | /// This has the same lifetime as the original slice, and so the |
142 | /// iterator can continue to be used while this exists. |
143 | /// |
144 | /// # Examples |
145 | /// |
146 | /// ``` |
147 | /// let mut chars = "abc".chars(); |
148 | /// |
149 | /// assert_eq!(chars.as_str(), "abc"); |
150 | /// chars.next(); |
151 | /// assert_eq!(chars.as_str(), "bc"); |
152 | /// chars.next(); |
153 | /// chars.next(); |
154 | /// assert_eq!(chars.as_str(), ""); |
155 | /// ``` |
156 | #[stable(feature = "iter_to_slice", since = "1.4.0")] |
157 | #[must_use] |
158 | #[inline] |
159 | pub fn as_str(&self) -> &'a str { |
160 | // SAFETY: `Chars` is only made from a str, which guarantees the iter is valid UTF-8. |
161 | unsafe { from_utf8_unchecked(self.iter.as_slice()) } |
162 | } |
163 | } |
164 | |
165 | /// An iterator over the [`char`]s of a string slice, and their positions. |
166 | /// |
167 | /// This struct is created by the [`char_indices`] method on [`str`]. |
168 | /// See its documentation for more. |
169 | /// |
170 | /// [`char`]: prim@char |
171 | /// [`char_indices`]: str::char_indices |
172 | #[derive(Clone, Debug)] |
173 | #[must_use= "iterators are lazy and do nothing unless consumed"] |
174 | #[stable(feature = "rust1", since = "1.0.0")] |
175 | pub struct CharIndices<'a> { |
176 | pub(super) front_offset: usize, |
177 | pub(super) iter: Chars<'a>, |
178 | } |
179 | |
180 | #[stable(feature = "rust1", since = "1.0.0")] |
181 | impl<'a> Iterator for CharIndices<'a> { |
182 | type Item = (usize, char); |
183 | |
184 | #[inline] |
185 | fn next(&mut self) -> Option<(usize, char)> { |
186 | let pre_len = self.iter.iter.len(); |
187 | match self.iter.next() { |
188 | None => None, |
189 | Some(ch) => { |
190 | let index = self.front_offset; |
191 | let len = self.iter.iter.len(); |
192 | self.front_offset += pre_len - len; |
193 | Some((index, ch)) |
194 | } |
195 | } |
196 | } |
197 | |
198 | #[inline] |
199 | fn count(self) -> usize { |
200 | self.iter.count() |
201 | } |
202 | |
203 | #[inline] |
204 | fn size_hint(&self) -> (usize, Option<usize>) { |
205 | self.iter.size_hint() |
206 | } |
207 | |
208 | #[inline] |
209 | fn last(mut self) -> Option<(usize, char)> { |
210 | // No need to go through the entire string. |
211 | self.next_back() |
212 | } |
213 | } |
214 | |
215 | #[stable(feature = "rust1", since = "1.0.0")] |
216 | impl<'a> DoubleEndedIterator for CharIndices<'a> { |
217 | #[inline] |
218 | fn next_back(&mut self) -> Option<(usize, char)> { |
219 | self.iter.next_back().map(|ch: char| { |
220 | let index: usize = self.front_offset + self.iter.iter.len(); |
221 | (index, ch) |
222 | }) |
223 | } |
224 | } |
225 | |
226 | #[stable(feature = "fused", since = "1.26.0")] |
227 | impl FusedIterator for CharIndices<'_> {} |
228 | |
229 | impl<'a> CharIndices<'a> { |
230 | /// Views the underlying data as a subslice of the original data. |
231 | /// |
232 | /// This has the same lifetime as the original slice, and so the |
233 | /// iterator can continue to be used while this exists. |
234 | #[stable(feature = "iter_to_slice", since = "1.4.0")] |
235 | #[must_use] |
236 | #[inline] |
237 | pub fn as_str(&self) -> &'a str { |
238 | self.iter.as_str() |
239 | } |
240 | |
241 | /// Returns the byte position of the next character, or the length |
242 | /// of the underlying string if there are no more characters. |
243 | /// |
244 | /// This means that, when the iterator has not been fully consumed, |
245 | /// the returned value will match the index that will be returned |
246 | /// by the next call to [`next()`](Self::next). |
247 | /// |
248 | /// # Examples |
249 | /// |
250 | /// ``` |
251 | /// let mut chars = "a楽".char_indices(); |
252 | /// |
253 | /// // `next()` has not been called yet, so `offset()` returns the byte |
254 | /// // index of the first character of the string, which is always 0. |
255 | /// assert_eq!(chars.offset(), 0); |
256 | /// // As expected, the first call to `next()` also returns 0 as index. |
257 | /// assert_eq!(chars.next(), Some((0, 'a'))); |
258 | /// |
259 | /// // `next()` has been called once, so `offset()` returns the byte index |
260 | /// // of the second character ... |
261 | /// assert_eq!(chars.offset(), 1); |
262 | /// // ... which matches the index returned by the next call to `next()`. |
263 | /// assert_eq!(chars.next(), Some((1, '楽'))); |
264 | /// |
265 | /// // Once the iterator has been consumed, `offset()` returns the length |
266 | /// // in bytes of the string. |
267 | /// assert_eq!(chars.offset(), 4); |
268 | /// assert_eq!(chars.next(), None); |
269 | /// ``` |
270 | #[inline] |
271 | #[must_use] |
272 | #[stable(feature = "char_indices_offset", since = "1.82.0")] |
273 | pub fn offset(&self) -> usize { |
274 | self.front_offset |
275 | } |
276 | } |
277 | |
278 | /// An iterator over the bytes of a string slice. |
279 | /// |
280 | /// This struct is created by the [`bytes`] method on [`str`]. |
281 | /// See its documentation for more. |
282 | /// |
283 | /// [`bytes`]: str::bytes |
284 | #[must_use= "iterators are lazy and do nothing unless consumed"] |
285 | #[stable(feature = "rust1", since = "1.0.0")] |
286 | #[derive(Clone, Debug)] |
287 | pub struct Bytes<'a>(pub(super) Copied<slice::Iter<'a, u8>>); |
288 | |
289 | #[stable(feature = "rust1", since = "1.0.0")] |
290 | impl Iterator for Bytes<'_> { |
291 | type Item = u8; |
292 | |
293 | #[inline] |
294 | fn next(&mut self) -> Option<u8> { |
295 | self.0.next() |
296 | } |
297 | |
298 | #[inline] |
299 | fn size_hint(&self) -> (usize, Option<usize>) { |
300 | self.0.size_hint() |
301 | } |
302 | |
303 | #[inline] |
304 | fn count(self) -> usize { |
305 | self.0.count() |
306 | } |
307 | |
308 | #[inline] |
309 | fn last(self) -> Option<Self::Item> { |
310 | self.0.last() |
311 | } |
312 | |
313 | #[inline] |
314 | fn nth(&mut self, n: usize) -> Option<Self::Item> { |
315 | self.0.nth(n) |
316 | } |
317 | |
318 | #[inline] |
319 | fn all<F>(&mut self, f: F) -> bool |
320 | where |
321 | F: FnMut(Self::Item) -> bool, |
322 | { |
323 | self.0.all(f) |
324 | } |
325 | |
326 | #[inline] |
327 | fn any<F>(&mut self, f: F) -> bool |
328 | where |
329 | F: FnMut(Self::Item) -> bool, |
330 | { |
331 | self.0.any(f) |
332 | } |
333 | |
334 | #[inline] |
335 | fn find<P>(&mut self, predicate: P) -> Option<Self::Item> |
336 | where |
337 | P: FnMut(&Self::Item) -> bool, |
338 | { |
339 | self.0.find(predicate) |
340 | } |
341 | |
342 | #[inline] |
343 | fn position<P>(&mut self, predicate: P) -> Option<usize> |
344 | where |
345 | P: FnMut(Self::Item) -> bool, |
346 | { |
347 | self.0.position(predicate) |
348 | } |
349 | |
350 | #[inline] |
351 | fn rposition<P>(&mut self, predicate: P) -> Option<usize> |
352 | where |
353 | P: FnMut(Self::Item) -> bool, |
354 | { |
355 | self.0.rposition(predicate) |
356 | } |
357 | |
358 | #[inline] |
359 | unsafe fn __iterator_get_unchecked(&mut self, idx: usize) -> u8 { |
360 | // SAFETY: the caller must uphold the safety contract |
361 | // for `Iterator::__iterator_get_unchecked`. |
362 | unsafe { self.0.__iterator_get_unchecked(idx) } |
363 | } |
364 | } |
365 | |
366 | #[stable(feature = "rust1", since = "1.0.0")] |
367 | impl DoubleEndedIterator for Bytes<'_> { |
368 | #[inline] |
369 | fn next_back(&mut self) -> Option<u8> { |
370 | self.0.next_back() |
371 | } |
372 | |
373 | #[inline] |
374 | fn nth_back(&mut self, n: usize) -> Option<Self::Item> { |
375 | self.0.nth_back(n) |
376 | } |
377 | |
378 | #[inline] |
379 | fn rfind<P>(&mut self, predicate: P) -> Option<Self::Item> |
380 | where |
381 | P: FnMut(&Self::Item) -> bool, |
382 | { |
383 | self.0.rfind(predicate) |
384 | } |
385 | } |
386 | |
387 | #[stable(feature = "rust1", since = "1.0.0")] |
388 | impl ExactSizeIterator for Bytes<'_> { |
389 | #[inline] |
390 | fn len(&self) -> usize { |
391 | self.0.len() |
392 | } |
393 | |
394 | #[inline] |
395 | fn is_empty(&self) -> bool { |
396 | self.0.is_empty() |
397 | } |
398 | } |
399 | |
400 | #[stable(feature = "fused", since = "1.26.0")] |
401 | impl FusedIterator for Bytes<'_> {} |
402 | |
403 | #[unstable(feature = "trusted_len", issue = "37572")] |
404 | unsafe impl TrustedLen for Bytes<'_> {} |
405 | |
406 | #[doc(hidden)] |
407 | #[unstable(feature = "trusted_random_access", issue = "none")] |
408 | unsafe impl TrustedRandomAccess for Bytes<'_> {} |
409 | |
410 | #[doc(hidden)] |
411 | #[unstable(feature = "trusted_random_access", issue = "none")] |
412 | unsafe impl TrustedRandomAccessNoCoerce for Bytes<'_> { |
413 | const MAY_HAVE_SIDE_EFFECT: bool = false; |
414 | } |
415 | |
416 | /// This macro generates a Clone impl for string pattern API |
417 | /// wrapper types of the form X<'a, P> |
418 | macro_rules! derive_pattern_clone { |
419 | (clone $t:ident with |$s:ident| $e:expr) => { |
420 | impl<'a, P> Clone for $t<'a, P> |
421 | where |
422 | P: Pattern<Searcher<'a>: Clone>, |
423 | { |
424 | fn clone(&self) -> Self { |
425 | let $s = self; |
426 | $e |
427 | } |
428 | } |
429 | }; |
430 | } |
431 | |
432 | /// This macro generates two public iterator structs |
433 | /// wrapping a private internal one that makes use of the `Pattern` API. |
434 | /// |
435 | /// For all patterns `P: Pattern` the following items will be |
436 | /// generated (generics omitted): |
437 | /// |
438 | /// struct $forward_iterator($internal_iterator); |
439 | /// struct $reverse_iterator($internal_iterator); |
440 | /// |
441 | /// impl Iterator for $forward_iterator |
442 | /// { /* internal ends up calling Searcher::next_match() */ } |
443 | /// |
444 | /// impl DoubleEndedIterator for $forward_iterator |
445 | /// where P::Searcher: DoubleEndedSearcher |
446 | /// { /* internal ends up calling Searcher::next_match_back() */ } |
447 | /// |
448 | /// impl Iterator for $reverse_iterator |
449 | /// where P::Searcher: ReverseSearcher |
450 | /// { /* internal ends up calling Searcher::next_match_back() */ } |
451 | /// |
452 | /// impl DoubleEndedIterator for $reverse_iterator |
453 | /// where P::Searcher: DoubleEndedSearcher |
454 | /// { /* internal ends up calling Searcher::next_match() */ } |
455 | /// |
456 | /// The internal one is defined outside the macro, and has almost the same |
457 | /// semantic as a DoubleEndedIterator by delegating to `pattern::Searcher` and |
458 | /// `pattern::ReverseSearcher` for both forward and reverse iteration. |
459 | /// |
460 | /// "Almost", because a `Searcher` and a `ReverseSearcher` for a given |
461 | /// `Pattern` might not return the same elements, so actually implementing |
462 | /// `DoubleEndedIterator` for it would be incorrect. |
463 | /// (See the docs in `str::pattern` for more details) |
464 | /// |
465 | /// However, the internal struct still represents a single ended iterator from |
466 | /// either end, and depending on pattern is also a valid double ended iterator, |
467 | /// so the two wrapper structs implement `Iterator` |
468 | /// and `DoubleEndedIterator` depending on the concrete pattern type, leading |
469 | /// to the complex impls seen above. |
470 | macro_rules! generate_pattern_iterators { |
471 | { |
472 | // Forward iterator |
473 | forward: |
474 | $(#[$forward_iterator_attribute:meta])* |
475 | struct $forward_iterator:ident; |
476 | |
477 | // Reverse iterator |
478 | reverse: |
479 | $(#[$reverse_iterator_attribute:meta])* |
480 | struct $reverse_iterator:ident; |
481 | |
482 | // Stability of all generated items |
483 | stability: |
484 | $(#[$common_stability_attribute:meta])* |
485 | |
486 | // Internal almost-iterator that is being delegated to |
487 | internal: |
488 | $internal_iterator:ident yielding ($iterty:ty); |
489 | |
490 | // Kind of delegation - either single ended or double ended |
491 | delegate $($t:tt)* |
492 | } => { |
493 | $(#[$forward_iterator_attribute])* |
494 | $(#[$common_stability_attribute])* |
495 | pub struct $forward_iterator<'a, P: Pattern>(pub(super) $internal_iterator<'a, P>); |
496 | |
497 | $(#[$common_stability_attribute])* |
498 | impl<'a, P> fmt::Debug for $forward_iterator<'a, P> |
499 | where |
500 | P: Pattern<Searcher<'a>: fmt::Debug>, |
501 | { |
502 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
503 | f.debug_tuple(stringify!($forward_iterator)) |
504 | .field(&self.0) |
505 | .finish() |
506 | } |
507 | } |
508 | |
509 | $(#[$common_stability_attribute])* |
510 | impl<'a, P: Pattern> Iterator for $forward_iterator<'a, P> { |
511 | type Item = $iterty; |
512 | |
513 | #[inline] |
514 | fn next(&mut self) -> Option<$iterty> { |
515 | self.0.next() |
516 | } |
517 | } |
518 | |
519 | $(#[$common_stability_attribute])* |
520 | impl<'a, P> Clone for $forward_iterator<'a, P> |
521 | where |
522 | P: Pattern<Searcher<'a>: Clone>, |
523 | { |
524 | fn clone(&self) -> Self { |
525 | $forward_iterator(self.0.clone()) |
526 | } |
527 | } |
528 | |
529 | $(#[$reverse_iterator_attribute])* |
530 | $(#[$common_stability_attribute])* |
531 | pub struct $reverse_iterator<'a, P: Pattern>(pub(super) $internal_iterator<'a, P>); |
532 | |
533 | $(#[$common_stability_attribute])* |
534 | impl<'a, P> fmt::Debug for $reverse_iterator<'a, P> |
535 | where |
536 | P: Pattern<Searcher<'a>: fmt::Debug>, |
537 | { |
538 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
539 | f.debug_tuple(stringify!($reverse_iterator)) |
540 | .field(&self.0) |
541 | .finish() |
542 | } |
543 | } |
544 | |
545 | $(#[$common_stability_attribute])* |
546 | impl<'a, P> Iterator for $reverse_iterator<'a, P> |
547 | where |
548 | P: Pattern<Searcher<'a>: ReverseSearcher<'a>>, |
549 | { |
550 | type Item = $iterty; |
551 | |
552 | #[inline] |
553 | fn next(&mut self) -> Option<$iterty> { |
554 | self.0.next_back() |
555 | } |
556 | } |
557 | |
558 | $(#[$common_stability_attribute])* |
559 | impl<'a, P> Clone for $reverse_iterator<'a, P> |
560 | where |
561 | P: Pattern<Searcher<'a>: Clone>, |
562 | { |
563 | fn clone(&self) -> Self { |
564 | $reverse_iterator(self.0.clone()) |
565 | } |
566 | } |
567 | |
568 | #[stable(feature = "fused", since = "1.26.0")] |
569 | impl<'a, P: Pattern> FusedIterator for $forward_iterator<'a, P> {} |
570 | |
571 | #[stable(feature = "fused", since = "1.26.0")] |
572 | impl<'a, P> FusedIterator for $reverse_iterator<'a, P> |
573 | where |
574 | P: Pattern<Searcher<'a>: ReverseSearcher<'a>>, |
575 | {} |
576 | |
577 | generate_pattern_iterators!($($t)* with $(#[$common_stability_attribute])*, |
578 | $forward_iterator, |
579 | $reverse_iterator, $iterty); |
580 | }; |
581 | { |
582 | double ended; with $(#[$common_stability_attribute:meta])*, |
583 | $forward_iterator:ident, |
584 | $reverse_iterator:ident, $iterty:ty |
585 | } => { |
586 | $(#[$common_stability_attribute])* |
587 | impl<'a, P> DoubleEndedIterator for $forward_iterator<'a, P> |
588 | where |
589 | P: Pattern<Searcher<'a>: DoubleEndedSearcher<'a>>, |
590 | { |
591 | #[inline] |
592 | fn next_back(&mut self) -> Option<$iterty> { |
593 | self.0.next_back() |
594 | } |
595 | } |
596 | |
597 | $(#[$common_stability_attribute])* |
598 | impl<'a, P> DoubleEndedIterator for $reverse_iterator<'a, P> |
599 | where |
600 | P: Pattern<Searcher<'a>: DoubleEndedSearcher<'a>>, |
601 | { |
602 | #[inline] |
603 | fn next_back(&mut self) -> Option<$iterty> { |
604 | self.0.next() |
605 | } |
606 | } |
607 | }; |
608 | { |
609 | single ended; with $(#[$common_stability_attribute:meta])*, |
610 | $forward_iterator:ident, |
611 | $reverse_iterator:ident, $iterty:ty |
612 | } => {} |
613 | } |
614 | |
615 | derive_pattern_clone! { |
616 | clone SplitInternal |
617 | with |s| SplitInternal { matcher: s.matcher.clone(), ..*s } |
618 | } |
619 | |
620 | pub(super) struct SplitInternal<'a, P: Pattern> { |
621 | pub(super) start: usize, |
622 | pub(super) end: usize, |
623 | pub(super) matcher: P::Searcher<'a>, |
624 | pub(super) allow_trailing_empty: bool, |
625 | pub(super) finished: bool, |
626 | } |
627 | |
628 | impl<'a, P> fmt::Debug for SplitInternal<'a, P> |
629 | where |
630 | P: Pattern<Searcher<'a>: fmt::Debug>, |
631 | { |
632 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
633 | f&mut DebugStruct<'_, '_>.debug_struct("SplitInternal") |
634 | .field("start", &self.start) |
635 | .field("end", &self.end) |
636 | .field("matcher", &self.matcher) |
637 | .field("allow_trailing_empty", &self.allow_trailing_empty) |
638 | .field(name:"finished", &self.finished) |
639 | .finish() |
640 | } |
641 | } |
642 | |
643 | impl<'a, P: Pattern> SplitInternal<'a, P> { |
644 | #[inline] |
645 | fn get_end(&mut self) -> Option<&'a str> { |
646 | if !self.finished { |
647 | self.finished = true; |
648 | |
649 | if self.allow_trailing_empty || self.end - self.start > 0 { |
650 | // SAFETY: `self.start` and `self.end` always lie on unicode boundaries. |
651 | let string = unsafe { self.matcher.haystack().get_unchecked(self.start..self.end) }; |
652 | return Some(string); |
653 | } |
654 | } |
655 | |
656 | None |
657 | } |
658 | |
659 | #[inline] |
660 | fn next(&mut self) -> Option<&'a str> { |
661 | if self.finished { |
662 | return None; |
663 | } |
664 | |
665 | let haystack = self.matcher.haystack(); |
666 | match self.matcher.next_match() { |
667 | // SAFETY: `Searcher` guarantees that `a` and `b` lie on unicode boundaries. |
668 | Some((a, b)) => unsafe { |
669 | let elt = haystack.get_unchecked(self.start..a); |
670 | self.start = b; |
671 | Some(elt) |
672 | }, |
673 | None => self.get_end(), |
674 | } |
675 | } |
676 | |
677 | #[inline] |
678 | fn next_inclusive(&mut self) -> Option<&'a str> { |
679 | if self.finished { |
680 | return None; |
681 | } |
682 | |
683 | let haystack = self.matcher.haystack(); |
684 | match self.matcher.next_match() { |
685 | // SAFETY: `Searcher` guarantees that `b` lies on unicode boundary, |
686 | // and self.start is either the start of the original string, |
687 | // or `b` was assigned to it, so it also lies on unicode boundary. |
688 | Some((_, b)) => unsafe { |
689 | let elt = haystack.get_unchecked(self.start..b); |
690 | self.start = b; |
691 | Some(elt) |
692 | }, |
693 | None => self.get_end(), |
694 | } |
695 | } |
696 | |
697 | #[inline] |
698 | fn next_back(&mut self) -> Option<&'a str> |
699 | where |
700 | P::Searcher<'a>: ReverseSearcher<'a>, |
701 | { |
702 | if self.finished { |
703 | return None; |
704 | } |
705 | |
706 | if !self.allow_trailing_empty { |
707 | self.allow_trailing_empty = true; |
708 | match self.next_back() { |
709 | Some(elt) if !elt.is_empty() => return Some(elt), |
710 | _ => { |
711 | if self.finished { |
712 | return None; |
713 | } |
714 | } |
715 | } |
716 | } |
717 | |
718 | let haystack = self.matcher.haystack(); |
719 | match self.matcher.next_match_back() { |
720 | // SAFETY: `Searcher` guarantees that `a` and `b` lie on unicode boundaries. |
721 | Some((a, b)) => unsafe { |
722 | let elt = haystack.get_unchecked(b..self.end); |
723 | self.end = a; |
724 | Some(elt) |
725 | }, |
726 | // SAFETY: `self.start` and `self.end` always lie on unicode boundaries. |
727 | None => unsafe { |
728 | self.finished = true; |
729 | Some(haystack.get_unchecked(self.start..self.end)) |
730 | }, |
731 | } |
732 | } |
733 | |
734 | #[inline] |
735 | fn next_back_inclusive(&mut self) -> Option<&'a str> |
736 | where |
737 | P::Searcher<'a>: ReverseSearcher<'a>, |
738 | { |
739 | if self.finished { |
740 | return None; |
741 | } |
742 | |
743 | if !self.allow_trailing_empty { |
744 | self.allow_trailing_empty = true; |
745 | match self.next_back_inclusive() { |
746 | Some(elt) if !elt.is_empty() => return Some(elt), |
747 | _ => { |
748 | if self.finished { |
749 | return None; |
750 | } |
751 | } |
752 | } |
753 | } |
754 | |
755 | let haystack = self.matcher.haystack(); |
756 | match self.matcher.next_match_back() { |
757 | // SAFETY: `Searcher` guarantees that `b` lies on unicode boundary, |
758 | // and self.end is either the end of the original string, |
759 | // or `b` was assigned to it, so it also lies on unicode boundary. |
760 | Some((_, b)) => unsafe { |
761 | let elt = haystack.get_unchecked(b..self.end); |
762 | self.end = b; |
763 | Some(elt) |
764 | }, |
765 | // SAFETY: self.start is either the start of the original string, |
766 | // or start of a substring that represents the part of the string that hasn't |
767 | // iterated yet. Either way, it is guaranteed to lie on unicode boundary. |
768 | // self.end is either the end of the original string, |
769 | // or `b` was assigned to it, so it also lies on unicode boundary. |
770 | None => unsafe { |
771 | self.finished = true; |
772 | Some(haystack.get_unchecked(self.start..self.end)) |
773 | }, |
774 | } |
775 | } |
776 | |
777 | #[inline] |
778 | fn remainder(&self) -> Option<&'a str> { |
779 | // `Self::get_end` doesn't change `self.start` |
780 | if self.finished { |
781 | return None; |
782 | } |
783 | |
784 | // SAFETY: `self.start` and `self.end` always lie on unicode boundaries. |
785 | Some(unsafe { self.matcher.haystack().get_unchecked(self.start..self.end) }) |
786 | } |
787 | } |
788 | |
789 | generate_pattern_iterators! { |
790 | forward: |
791 | /// Created with the method [`split`]. |
792 | /// |
793 | /// [`split`]: str::split |
794 | struct Split; |
795 | reverse: |
796 | /// Created with the method [`rsplit`]. |
797 | /// |
798 | /// [`rsplit`]: str::rsplit |
799 | struct RSplit; |
800 | stability: |
801 | #[stable(feature = "rust1", since = "1.0.0")] |
802 | internal: |
803 | SplitInternal yielding (&'a str); |
804 | delegate double ended; |
805 | } |
806 | |
807 | impl<'a, P: Pattern> Split<'a, P> { |
808 | /// Returns remainder of the split string. |
809 | /// |
810 | /// If the iterator is empty, returns `None`. |
811 | /// |
812 | /// # Examples |
813 | /// |
814 | /// ``` |
815 | /// #![feature(str_split_remainder)] |
816 | /// let mut split = "Mary had a little lamb".split( ' '); |
817 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
818 | /// split.next(); |
819 | /// assert_eq!(split.remainder(), Some("had a little lamb")); |
820 | /// split.by_ref().for_each(drop); |
821 | /// assert_eq!(split.remainder(), None); |
822 | /// ``` |
823 | #[inline] |
824 | #[unstable(feature = "str_split_remainder", issue = "77998")] |
825 | pub fn remainder(&self) -> Option<&'a str> { |
826 | self.0.remainder() |
827 | } |
828 | } |
829 | |
830 | impl<'a, P: Pattern> RSplit<'a, P> { |
831 | /// Returns remainder of the split string. |
832 | /// |
833 | /// If the iterator is empty, returns `None`. |
834 | /// |
835 | /// # Examples |
836 | /// |
837 | /// ``` |
838 | /// #![feature(str_split_remainder)] |
839 | /// let mut split = "Mary had a little lamb".rsplit( ' '); |
840 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
841 | /// split.next(); |
842 | /// assert_eq!(split.remainder(), Some("Mary had a little")); |
843 | /// split.by_ref().for_each(drop); |
844 | /// assert_eq!(split.remainder(), None); |
845 | /// ``` |
846 | #[inline] |
847 | #[unstable(feature = "str_split_remainder", issue = "77998")] |
848 | pub fn remainder(&self) -> Option<&'a str> { |
849 | self.0.remainder() |
850 | } |
851 | } |
852 | |
853 | generate_pattern_iterators! { |
854 | forward: |
855 | /// Created with the method [`split_terminator`]. |
856 | /// |
857 | /// [`split_terminator`]: str::split_terminator |
858 | struct SplitTerminator; |
859 | reverse: |
860 | /// Created with the method [`rsplit_terminator`]. |
861 | /// |
862 | /// [`rsplit_terminator`]: str::rsplit_terminator |
863 | struct RSplitTerminator; |
864 | stability: |
865 | #[stable(feature = "rust1", since = "1.0.0")] |
866 | internal: |
867 | SplitInternal yielding (&'a str); |
868 | delegate double ended; |
869 | } |
870 | |
871 | impl<'a, P: Pattern> SplitTerminator<'a, P> { |
872 | /// Returns remainder of the split string. |
873 | /// |
874 | /// If the iterator is empty, returns `None`. |
875 | /// |
876 | /// # Examples |
877 | /// |
878 | /// ``` |
879 | /// #![feature(str_split_remainder)] |
880 | /// let mut split = "A..B..".split_terminator( '.'); |
881 | /// assert_eq!(split.remainder(), Some("A..B..")); |
882 | /// split.next(); |
883 | /// assert_eq!(split.remainder(), Some(".B..")); |
884 | /// split.by_ref().for_each(drop); |
885 | /// assert_eq!(split.remainder(), None); |
886 | /// ``` |
887 | #[inline] |
888 | #[unstable(feature = "str_split_remainder", issue = "77998")] |
889 | pub fn remainder(&self) -> Option<&'a str> { |
890 | self.0.remainder() |
891 | } |
892 | } |
893 | |
894 | impl<'a, P: Pattern> RSplitTerminator<'a, P> { |
895 | /// Returns remainder of the split string. |
896 | /// |
897 | /// If the iterator is empty, returns `None`. |
898 | /// |
899 | /// # Examples |
900 | /// |
901 | /// ``` |
902 | /// #![feature(str_split_remainder)] |
903 | /// let mut split = "A..B..".rsplit_terminator( '.'); |
904 | /// assert_eq!(split.remainder(), Some("A..B..")); |
905 | /// split.next(); |
906 | /// assert_eq!(split.remainder(), Some("A..B")); |
907 | /// split.by_ref().for_each(drop); |
908 | /// assert_eq!(split.remainder(), None); |
909 | /// ``` |
910 | #[inline] |
911 | #[unstable(feature = "str_split_remainder", issue = "77998")] |
912 | pub fn remainder(&self) -> Option<&'a str> { |
913 | self.0.remainder() |
914 | } |
915 | } |
916 | |
917 | derive_pattern_clone! { |
918 | clone SplitNInternal |
919 | with |s| SplitNInternal { iter: s.iter.clone(), ..*s } |
920 | } |
921 | |
922 | pub(super) struct SplitNInternal<'a, P: Pattern> { |
923 | pub(super) iter: SplitInternal<'a, P>, |
924 | /// The number of splits remaining |
925 | pub(super) count: usize, |
926 | } |
927 | |
928 | impl<'a, P> fmt::Debug for SplitNInternal<'a, P> |
929 | where |
930 | P: Pattern<Searcher<'a>: fmt::Debug>, |
931 | { |
932 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
933 | f&mut DebugStruct<'_, '_>.debug_struct("SplitNInternal") |
934 | .field("iter", &self.iter) |
935 | .field(name:"count", &self.count) |
936 | .finish() |
937 | } |
938 | } |
939 | |
940 | impl<'a, P: Pattern> SplitNInternal<'a, P> { |
941 | #[inline] |
942 | fn next(&mut self) -> Option<&'a str> { |
943 | match self.count { |
944 | 0 => None, |
945 | 1 => { |
946 | self.count = 0; |
947 | self.iter.get_end() |
948 | } |
949 | _ => { |
950 | self.count -= 1; |
951 | self.iter.next() |
952 | } |
953 | } |
954 | } |
955 | |
956 | #[inline] |
957 | fn next_back(&mut self) -> Option<&'a str> |
958 | where |
959 | P::Searcher<'a>: ReverseSearcher<'a>, |
960 | { |
961 | match self.count { |
962 | 0 => None, |
963 | 1 => { |
964 | self.count = 0; |
965 | self.iter.get_end() |
966 | } |
967 | _ => { |
968 | self.count -= 1; |
969 | self.iter.next_back() |
970 | } |
971 | } |
972 | } |
973 | |
974 | #[inline] |
975 | fn remainder(&self) -> Option<&'a str> { |
976 | self.iter.remainder() |
977 | } |
978 | } |
979 | |
980 | generate_pattern_iterators! { |
981 | forward: |
982 | /// Created with the method [`splitn`]. |
983 | /// |
984 | /// [`splitn`]: str::splitn |
985 | struct SplitN; |
986 | reverse: |
987 | /// Created with the method [`rsplitn`]. |
988 | /// |
989 | /// [`rsplitn`]: str::rsplitn |
990 | struct RSplitN; |
991 | stability: |
992 | #[stable(feature = "rust1", since = "1.0.0")] |
993 | internal: |
994 | SplitNInternal yielding (&'a str); |
995 | delegate single ended; |
996 | } |
997 | |
998 | impl<'a, P: Pattern> SplitN<'a, P> { |
999 | /// Returns remainder of the split string. |
1000 | /// |
1001 | /// If the iterator is empty, returns `None`. |
1002 | /// |
1003 | /// # Examples |
1004 | /// |
1005 | /// ``` |
1006 | /// #![feature(str_split_remainder)] |
1007 | /// let mut split = "Mary had a little lamb".splitn(3, ' '); |
1008 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
1009 | /// split.next(); |
1010 | /// assert_eq!(split.remainder(), Some("had a little lamb")); |
1011 | /// split.by_ref().for_each(drop); |
1012 | /// assert_eq!(split.remainder(), None); |
1013 | /// ``` |
1014 | #[inline] |
1015 | #[unstable(feature = "str_split_remainder", issue = "77998")] |
1016 | pub fn remainder(&self) -> Option<&'a str> { |
1017 | self.0.remainder() |
1018 | } |
1019 | } |
1020 | |
1021 | impl<'a, P: Pattern> RSplitN<'a, P> { |
1022 | /// Returns remainder of the split string. |
1023 | /// |
1024 | /// If the iterator is empty, returns `None`. |
1025 | /// |
1026 | /// # Examples |
1027 | /// |
1028 | /// ``` |
1029 | /// #![feature(str_split_remainder)] |
1030 | /// let mut split = "Mary had a little lamb".rsplitn(3, ' '); |
1031 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
1032 | /// split.next(); |
1033 | /// assert_eq!(split.remainder(), Some("Mary had a little")); |
1034 | /// split.by_ref().for_each(drop); |
1035 | /// assert_eq!(split.remainder(), None); |
1036 | /// ``` |
1037 | #[inline] |
1038 | #[unstable(feature = "str_split_remainder", issue = "77998")] |
1039 | pub fn remainder(&self) -> Option<&'a str> { |
1040 | self.0.remainder() |
1041 | } |
1042 | } |
1043 | |
1044 | derive_pattern_clone! { |
1045 | clone MatchIndicesInternal |
1046 | with |s| MatchIndicesInternal(s.0.clone()) |
1047 | } |
1048 | |
1049 | pub(super) struct MatchIndicesInternal<'a, P: Pattern>(pub(super) P::Searcher<'a>); |
1050 | |
1051 | impl<'a, P> fmt::Debug for MatchIndicesInternal<'a, P> |
1052 | where |
1053 | P: Pattern<Searcher<'a>: fmt::Debug>, |
1054 | { |
1055 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
1056 | f.debug_tuple(name:"MatchIndicesInternal").field(&self.0).finish() |
1057 | } |
1058 | } |
1059 | |
1060 | impl<'a, P: Pattern> MatchIndicesInternal<'a, P> { |
1061 | #[inline] |
1062 | fn next(&mut self) -> Option<(usize, &'a str)> { |
1063 | self.0 |
1064 | .next_match() |
1065 | // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries. |
1066 | .map(|(start: usize, end: usize)| unsafe { (start, self.0.haystack().get_unchecked(start..end)) }) |
1067 | } |
1068 | |
1069 | #[inline] |
1070 | fn next_back(&mut self) -> Option<(usize, &'a str)> |
1071 | where |
1072 | P::Searcher<'a>: ReverseSearcher<'a>, |
1073 | { |
1074 | self.0 |
1075 | .next_match_back() |
1076 | // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries. |
1077 | .map(|(start: usize, end: usize)| unsafe { (start, self.0.haystack().get_unchecked(start..end)) }) |
1078 | } |
1079 | } |
1080 | |
1081 | generate_pattern_iterators! { |
1082 | forward: |
1083 | /// Created with the method [`match_indices`]. |
1084 | /// |
1085 | /// [`match_indices`]: str::match_indices |
1086 | struct MatchIndices; |
1087 | reverse: |
1088 | /// Created with the method [`rmatch_indices`]. |
1089 | /// |
1090 | /// [`rmatch_indices`]: str::rmatch_indices |
1091 | struct RMatchIndices; |
1092 | stability: |
1093 | #[stable(feature = "str_match_indices", since = "1.5.0")] |
1094 | internal: |
1095 | MatchIndicesInternal yielding ((usize, &'a str)); |
1096 | delegate double ended; |
1097 | } |
1098 | |
1099 | derive_pattern_clone! { |
1100 | clone MatchesInternal |
1101 | with |s| MatchesInternal(s.0.clone()) |
1102 | } |
1103 | |
1104 | pub(super) struct MatchesInternal<'a, P: Pattern>(pub(super) P::Searcher<'a>); |
1105 | |
1106 | impl<'a, P> fmt::Debug for MatchesInternal<'a, P> |
1107 | where |
1108 | P: Pattern<Searcher<'a>: fmt::Debug>, |
1109 | { |
1110 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
1111 | f.debug_tuple(name:"MatchesInternal").field(&self.0).finish() |
1112 | } |
1113 | } |
1114 | |
1115 | impl<'a, P: Pattern> MatchesInternal<'a, P> { |
1116 | #[inline] |
1117 | fn next(&mut self) -> Option<&'a str> { |
1118 | // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries. |
1119 | self.0.next_match().map(|(a: usize, b: usize)| unsafe { |
1120 | // Indices are known to be on utf8 boundaries |
1121 | self.0.haystack().get_unchecked(a..b) |
1122 | }) |
1123 | } |
1124 | |
1125 | #[inline] |
1126 | fn next_back(&mut self) -> Option<&'a str> |
1127 | where |
1128 | P::Searcher<'a>: ReverseSearcher<'a>, |
1129 | { |
1130 | // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries. |
1131 | self.0.next_match_back().map(|(a: usize, b: usize)| unsafe { |
1132 | // Indices are known to be on utf8 boundaries |
1133 | self.0.haystack().get_unchecked(a..b) |
1134 | }) |
1135 | } |
1136 | } |
1137 | |
1138 | generate_pattern_iterators! { |
1139 | forward: |
1140 | /// Created with the method [`matches`]. |
1141 | /// |
1142 | /// [`matches`]: str::matches |
1143 | struct Matches; |
1144 | reverse: |
1145 | /// Created with the method [`rmatches`]. |
1146 | /// |
1147 | /// [`rmatches`]: str::rmatches |
1148 | struct RMatches; |
1149 | stability: |
1150 | #[stable(feature = "str_matches", since = "1.2.0")] |
1151 | internal: |
1152 | MatchesInternal yielding (&'a str); |
1153 | delegate double ended; |
1154 | } |
1155 | |
1156 | /// An iterator over the lines of a string, as string slices. |
1157 | /// |
1158 | /// This struct is created with the [`lines`] method on [`str`]. |
1159 | /// See its documentation for more. |
1160 | /// |
1161 | /// [`lines`]: str::lines |
1162 | #[stable(feature = "rust1", since = "1.0.0")] |
1163 | #[must_use= "iterators are lazy and do nothing unless consumed"] |
1164 | #[derive(Clone, Debug)] |
1165 | pub struct Lines<'a>(pub(super) Map<SplitInclusive<'a, char>, LinesMap>); |
1166 | |
1167 | #[stable(feature = "rust1", since = "1.0.0")] |
1168 | impl<'a> Iterator for Lines<'a> { |
1169 | type Item = &'a str; |
1170 | |
1171 | #[inline] |
1172 | fn next(&mut self) -> Option<&'a str> { |
1173 | self.0.next() |
1174 | } |
1175 | |
1176 | #[inline] |
1177 | fn size_hint(&self) -> (usize, Option<usize>) { |
1178 | self.0.size_hint() |
1179 | } |
1180 | |
1181 | #[inline] |
1182 | fn last(mut self) -> Option<&'a str> { |
1183 | self.next_back() |
1184 | } |
1185 | } |
1186 | |
1187 | #[stable(feature = "rust1", since = "1.0.0")] |
1188 | impl<'a> DoubleEndedIterator for Lines<'a> { |
1189 | #[inline] |
1190 | fn next_back(&mut self) -> Option<&'a str> { |
1191 | self.0.next_back() |
1192 | } |
1193 | } |
1194 | |
1195 | #[stable(feature = "fused", since = "1.26.0")] |
1196 | impl FusedIterator for Lines<'_> {} |
1197 | |
1198 | impl<'a> Lines<'a> { |
1199 | /// Returns the remaining lines of the split string. |
1200 | /// |
1201 | /// # Examples |
1202 | /// |
1203 | /// ``` |
1204 | /// #![feature(str_lines_remainder)] |
1205 | /// |
1206 | /// let mut lines = "a\n b\n c\n d".lines(); |
1207 | /// assert_eq!(lines.remainder(), Some("a\n b\n c\n d")); |
1208 | /// |
1209 | /// lines.next(); |
1210 | /// assert_eq!(lines.remainder(), Some("b\n c\n d")); |
1211 | /// |
1212 | /// lines.by_ref().for_each(drop); |
1213 | /// assert_eq!(lines.remainder(), None); |
1214 | /// ``` |
1215 | #[inline] |
1216 | #[must_use] |
1217 | #[unstable(feature = "str_lines_remainder", issue = "77998")] |
1218 | pub fn remainder(&self) -> Option<&'a str> { |
1219 | self.0.iter.remainder() |
1220 | } |
1221 | } |
1222 | |
1223 | /// Created with the method [`lines_any`]. |
1224 | /// |
1225 | /// [`lines_any`]: str::lines_any |
1226 | #[stable(feature = "rust1", since = "1.0.0")] |
1227 | #[deprecated(since = "1.4.0", note = "use lines()/Lines instead now")] |
1228 | #[must_use= "iterators are lazy and do nothing unless consumed"] |
1229 | #[derive(Clone, Debug)] |
1230 | #[allow(deprecated)] |
1231 | pub struct LinesAny<'a>(pub(super) Lines<'a>); |
1232 | |
1233 | #[stable(feature = "rust1", since = "1.0.0")] |
1234 | #[allow(deprecated)] |
1235 | impl<'a> Iterator for LinesAny<'a> { |
1236 | type Item = &'a str; |
1237 | |
1238 | #[inline] |
1239 | fn next(&mut self) -> Option<&'a str> { |
1240 | self.0.next() |
1241 | } |
1242 | |
1243 | #[inline] |
1244 | fn size_hint(&self) -> (usize, Option<usize>) { |
1245 | self.0.size_hint() |
1246 | } |
1247 | } |
1248 | |
1249 | #[stable(feature = "rust1", since = "1.0.0")] |
1250 | #[allow(deprecated)] |
1251 | impl<'a> DoubleEndedIterator for LinesAny<'a> { |
1252 | #[inline] |
1253 | fn next_back(&mut self) -> Option<&'a str> { |
1254 | self.0.next_back() |
1255 | } |
1256 | } |
1257 | |
1258 | #[stable(feature = "fused", since = "1.26.0")] |
1259 | #[allow(deprecated)] |
1260 | impl FusedIterator for LinesAny<'_> {} |
1261 | |
1262 | /// An iterator over the non-whitespace substrings of a string, |
1263 | /// separated by any amount of whitespace. |
1264 | /// |
1265 | /// This struct is created by the [`split_whitespace`] method on [`str`]. |
1266 | /// See its documentation for more. |
1267 | /// |
1268 | /// [`split_whitespace`]: str::split_whitespace |
1269 | #[stable(feature = "split_whitespace", since = "1.1.0")] |
1270 | #[derive(Clone, Debug)] |
1271 | pub struct SplitWhitespace<'a> { |
1272 | pub(super) inner: Filter<Split<'a, IsWhitespace>, IsNotEmpty>, |
1273 | } |
1274 | |
1275 | /// An iterator over the non-ASCII-whitespace substrings of a string, |
1276 | /// separated by any amount of ASCII whitespace. |
1277 | /// |
1278 | /// This struct is created by the [`split_ascii_whitespace`] method on [`str`]. |
1279 | /// See its documentation for more. |
1280 | /// |
1281 | /// [`split_ascii_whitespace`]: str::split_ascii_whitespace |
1282 | #[stable(feature = "split_ascii_whitespace", since = "1.34.0")] |
1283 | #[derive(Clone, Debug)] |
1284 | pub struct SplitAsciiWhitespace<'a> { |
1285 | pub(super) inner: |
1286 | Map<Filter<SliceSplit<'a, u8, IsAsciiWhitespace>, BytesIsNotEmpty>, UnsafeBytesToStr>, |
1287 | } |
1288 | |
1289 | /// An iterator over the substrings of a string, |
1290 | /// terminated by a substring matching to a predicate function |
1291 | /// Unlike `Split`, it contains the matched part as a terminator |
1292 | /// of the subslice. |
1293 | /// |
1294 | /// This struct is created by the [`split_inclusive`] method on [`str`]. |
1295 | /// See its documentation for more. |
1296 | /// |
1297 | /// [`split_inclusive`]: str::split_inclusive |
1298 | #[stable(feature = "split_inclusive", since = "1.51.0")] |
1299 | pub struct SplitInclusive<'a, P: Pattern>(pub(super) SplitInternal<'a, P>); |
1300 | |
1301 | #[stable(feature = "split_whitespace", since = "1.1.0")] |
1302 | impl<'a> Iterator for SplitWhitespace<'a> { |
1303 | type Item = &'a str; |
1304 | |
1305 | #[inline] |
1306 | fn next(&mut self) -> Option<&'a str> { |
1307 | self.inner.next() |
1308 | } |
1309 | |
1310 | #[inline] |
1311 | fn size_hint(&self) -> (usize, Option<usize>) { |
1312 | self.inner.size_hint() |
1313 | } |
1314 | |
1315 | #[inline] |
1316 | fn last(mut self) -> Option<&'a str> { |
1317 | self.next_back() |
1318 | } |
1319 | } |
1320 | |
1321 | #[stable(feature = "split_whitespace", since = "1.1.0")] |
1322 | impl<'a> DoubleEndedIterator for SplitWhitespace<'a> { |
1323 | #[inline] |
1324 | fn next_back(&mut self) -> Option<&'a str> { |
1325 | self.inner.next_back() |
1326 | } |
1327 | } |
1328 | |
1329 | #[stable(feature = "fused", since = "1.26.0")] |
1330 | impl FusedIterator for SplitWhitespace<'_> {} |
1331 | |
1332 | impl<'a> SplitWhitespace<'a> { |
1333 | /// Returns remainder of the split string |
1334 | /// |
1335 | /// # Examples |
1336 | /// |
1337 | /// ``` |
1338 | /// #![feature(str_split_whitespace_remainder)] |
1339 | /// |
1340 | /// let mut split = "Mary had a little lamb".split_whitespace(); |
1341 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
1342 | /// |
1343 | /// split.next(); |
1344 | /// assert_eq!(split.remainder(), Some("had a little lamb")); |
1345 | /// |
1346 | /// split.by_ref().for_each(drop); |
1347 | /// assert_eq!(split.remainder(), None); |
1348 | /// ``` |
1349 | #[inline] |
1350 | #[must_use] |
1351 | #[unstable(feature = "str_split_whitespace_remainder", issue = "77998")] |
1352 | pub fn remainder(&self) -> Option<&'a str> { |
1353 | self.inner.iter.remainder() |
1354 | } |
1355 | } |
1356 | |
1357 | #[stable(feature = "split_ascii_whitespace", since = "1.34.0")] |
1358 | impl<'a> Iterator for SplitAsciiWhitespace<'a> { |
1359 | type Item = &'a str; |
1360 | |
1361 | #[inline] |
1362 | fn next(&mut self) -> Option<&'a str> { |
1363 | self.inner.next() |
1364 | } |
1365 | |
1366 | #[inline] |
1367 | fn size_hint(&self) -> (usize, Option<usize>) { |
1368 | self.inner.size_hint() |
1369 | } |
1370 | |
1371 | #[inline] |
1372 | fn last(mut self) -> Option<&'a str> { |
1373 | self.next_back() |
1374 | } |
1375 | } |
1376 | |
1377 | #[stable(feature = "split_ascii_whitespace", since = "1.34.0")] |
1378 | impl<'a> DoubleEndedIterator for SplitAsciiWhitespace<'a> { |
1379 | #[inline] |
1380 | fn next_back(&mut self) -> Option<&'a str> { |
1381 | self.inner.next_back() |
1382 | } |
1383 | } |
1384 | |
1385 | #[stable(feature = "split_ascii_whitespace", since = "1.34.0")] |
1386 | impl FusedIterator for SplitAsciiWhitespace<'_> {} |
1387 | |
1388 | impl<'a> SplitAsciiWhitespace<'a> { |
1389 | /// Returns remainder of the split string. |
1390 | /// |
1391 | /// If the iterator is empty, returns `None`. |
1392 | /// |
1393 | /// # Examples |
1394 | /// |
1395 | /// ``` |
1396 | /// #![feature(str_split_whitespace_remainder)] |
1397 | /// |
1398 | /// let mut split = "Mary had a little lamb".split_ascii_whitespace(); |
1399 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
1400 | /// |
1401 | /// split.next(); |
1402 | /// assert_eq!(split.remainder(), Some("had a little lamb")); |
1403 | /// |
1404 | /// split.by_ref().for_each(drop); |
1405 | /// assert_eq!(split.remainder(), None); |
1406 | /// ``` |
1407 | #[inline] |
1408 | #[must_use] |
1409 | #[unstable(feature = "str_split_whitespace_remainder", issue = "77998")] |
1410 | pub fn remainder(&self) -> Option<&'a str> { |
1411 | if self.inner.iter.iter.finished { |
1412 | return None; |
1413 | } |
1414 | |
1415 | // SAFETY: Slice is created from str. |
1416 | Some(unsafe { crate::str::from_utf8_unchecked(&self.inner.iter.iter.v) }) |
1417 | } |
1418 | } |
1419 | |
1420 | #[stable(feature = "split_inclusive", since = "1.51.0")] |
1421 | impl<'a, P: Pattern> Iterator for SplitInclusive<'a, P> { |
1422 | type Item = &'a str; |
1423 | |
1424 | #[inline] |
1425 | fn next(&mut self) -> Option<&'a str> { |
1426 | self.0.next_inclusive() |
1427 | } |
1428 | } |
1429 | |
1430 | #[stable(feature = "split_inclusive", since = "1.51.0")] |
1431 | impl<'a, P: Pattern<Searcher<'a>: fmt::Debug>> fmt::Debug for SplitInclusive<'a, P> { |
1432 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
1433 | f.debug_struct("SplitInclusive").field(name: "0", &self.0).finish() |
1434 | } |
1435 | } |
1436 | |
1437 | // FIXME(#26925) Remove in favor of `#[derive(Clone)]` |
1438 | #[stable(feature = "split_inclusive", since = "1.51.0")] |
1439 | impl<'a, P: Pattern<Searcher<'a>: Clone>> Clone for SplitInclusive<'a, P> { |
1440 | fn clone(&self) -> Self { |
1441 | SplitInclusive(self.0.clone()) |
1442 | } |
1443 | } |
1444 | |
1445 | #[stable(feature = "split_inclusive", since = "1.51.0")] |
1446 | impl<'a, P: Pattern<Searcher<'a>: DoubleEndedSearcher<'a>>> DoubleEndedIterator |
1447 | for SplitInclusive<'a, P> |
1448 | { |
1449 | #[inline] |
1450 | fn next_back(&mut self) -> Option<&'a str> { |
1451 | self.0.next_back_inclusive() |
1452 | } |
1453 | } |
1454 | |
1455 | #[stable(feature = "split_inclusive", since = "1.51.0")] |
1456 | impl<'a, P: Pattern> FusedIterator for SplitInclusive<'a, P> {} |
1457 | |
1458 | impl<'a, P: Pattern> SplitInclusive<'a, P> { |
1459 | /// Returns remainder of the split string. |
1460 | /// |
1461 | /// If the iterator is empty, returns `None`. |
1462 | /// |
1463 | /// # Examples |
1464 | /// |
1465 | /// ``` |
1466 | /// #![feature(str_split_inclusive_remainder)] |
1467 | /// let mut split = "Mary had a little lamb".split_inclusive( ' '); |
1468 | /// assert_eq!(split.remainder(), Some("Mary had a little lamb")); |
1469 | /// split.next(); |
1470 | /// assert_eq!(split.remainder(), Some("had a little lamb")); |
1471 | /// split.by_ref().for_each(drop); |
1472 | /// assert_eq!(split.remainder(), None); |
1473 | /// ``` |
1474 | #[inline] |
1475 | #[unstable(feature = "str_split_inclusive_remainder", issue = "77998")] |
1476 | pub fn remainder(&self) -> Option<&'a str> { |
1477 | self.0.remainder() |
1478 | } |
1479 | } |
1480 | |
1481 | /// An iterator of [`u16`] over the string encoded as UTF-16. |
1482 | /// |
1483 | /// This struct is created by the [`encode_utf16`] method on [`str`]. |
1484 | /// See its documentation for more. |
1485 | /// |
1486 | /// [`encode_utf16`]: str::encode_utf16 |
1487 | #[derive(Clone)] |
1488 | #[stable(feature = "encode_utf16", since = "1.8.0")] |
1489 | pub struct EncodeUtf16<'a> { |
1490 | pub(super) chars: Chars<'a>, |
1491 | pub(super) extra: u16, |
1492 | } |
1493 | |
1494 | #[stable(feature = "collection_debug", since = "1.17.0")] |
1495 | impl fmt::Debug for EncodeUtf16<'_> { |
1496 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
1497 | f.debug_struct(name:"EncodeUtf16").finish_non_exhaustive() |
1498 | } |
1499 | } |
1500 | |
1501 | #[stable(feature = "encode_utf16", since = "1.8.0")] |
1502 | impl<'a> Iterator for EncodeUtf16<'a> { |
1503 | type Item = u16; |
1504 | |
1505 | #[inline] |
1506 | fn next(&mut self) -> Option<u16> { |
1507 | if self.extra != 0 { |
1508 | let tmp = self.extra; |
1509 | self.extra = 0; |
1510 | return Some(tmp); |
1511 | } |
1512 | |
1513 | let mut buf = [0; 2]; |
1514 | self.chars.next().map(|ch| { |
1515 | let n = ch.encode_utf16(&mut buf).len(); |
1516 | if n == 2 { |
1517 | self.extra = buf[1]; |
1518 | } |
1519 | buf[0] |
1520 | }) |
1521 | } |
1522 | |
1523 | #[inline] |
1524 | fn size_hint(&self) -> (usize, Option<usize>) { |
1525 | let len = self.chars.iter.len(); |
1526 | // The highest bytes:code units ratio occurs for 3-byte sequences, |
1527 | // since a 4-byte sequence results in 2 code units. The lower bound |
1528 | // is therefore determined by assuming the remaining bytes contain as |
1529 | // many 3-byte sequences as possible. The highest bytes:code units |
1530 | // ratio is for 1-byte sequences, so use this for the upper bound. |
1531 | // `(len + 2)` can't overflow, because we know that the `slice::Iter` |
1532 | // belongs to a slice in memory which has a maximum length of |
1533 | // `isize::MAX` (that's well below `usize::MAX`) |
1534 | if self.extra == 0 { |
1535 | ((len + 2) / 3, Some(len)) |
1536 | } else { |
1537 | // We're in the middle of a surrogate pair, so add the remaining |
1538 | // surrogate to the bounds. |
1539 | ((len + 2) / 3 + 1, Some(len + 1)) |
1540 | } |
1541 | } |
1542 | } |
1543 | |
1544 | #[stable(feature = "fused", since = "1.26.0")] |
1545 | impl FusedIterator for EncodeUtf16<'_> {} |
1546 | |
1547 | /// The return type of [`str::escape_debug`]. |
1548 | #[stable(feature = "str_escape", since = "1.34.0")] |
1549 | #[derive(Clone, Debug)] |
1550 | pub struct EscapeDebug<'a> { |
1551 | pub(super) inner: Chain< |
1552 | Flatten<option::IntoIter<char_mod::EscapeDebug>>, |
1553 | FlatMap<Chars<'a>, char_mod::EscapeDebug, CharEscapeDebugContinue>, |
1554 | >, |
1555 | } |
1556 | |
1557 | /// The return type of [`str::escape_default`]. |
1558 | #[stable(feature = "str_escape", since = "1.34.0")] |
1559 | #[derive(Clone, Debug)] |
1560 | pub struct EscapeDefault<'a> { |
1561 | pub(super) inner: FlatMap<Chars<'a>, char_mod::EscapeDefault, CharEscapeDefault>, |
1562 | } |
1563 | |
1564 | /// The return type of [`str::escape_unicode`]. |
1565 | #[stable(feature = "str_escape", since = "1.34.0")] |
1566 | #[derive(Clone, Debug)] |
1567 | pub struct EscapeUnicode<'a> { |
1568 | pub(super) inner: FlatMap<Chars<'a>, char_mod::EscapeUnicode, CharEscapeUnicode>, |
1569 | } |
1570 | |
1571 | macro_rules! escape_types_impls { |
1572 | ($( $Name: ident ),+) => {$( |
1573 | #[stable(feature = "str_escape", since = "1.34.0")] |
1574 | impl<'a> fmt::Display for $Name<'a> { |
1575 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
1576 | self.clone().try_for_each(|c| f.write_char(c)) |
1577 | } |
1578 | } |
1579 | |
1580 | #[stable(feature = "str_escape", since = "1.34.0")] |
1581 | impl<'a> Iterator for $Name<'a> { |
1582 | type Item = char; |
1583 | |
1584 | #[inline] |
1585 | fn next(&mut self) -> Option<char> { self.inner.next() } |
1586 | |
1587 | #[inline] |
1588 | fn size_hint(&self) -> (usize, Option<usize>) { self.inner.size_hint() } |
1589 | |
1590 | #[inline] |
1591 | fn try_fold<Acc, Fold, R>(&mut self, init: Acc, fold: Fold) -> R where |
1592 | Self: Sized, Fold: FnMut(Acc, Self::Item) -> R, R: Try<Output = Acc> |
1593 | { |
1594 | self.inner.try_fold(init, fold) |
1595 | } |
1596 | |
1597 | #[inline] |
1598 | fn fold<Acc, Fold>(self, init: Acc, fold: Fold) -> Acc |
1599 | where Fold: FnMut(Acc, Self::Item) -> Acc, |
1600 | { |
1601 | self.inner.fold(init, fold) |
1602 | } |
1603 | } |
1604 | |
1605 | #[stable(feature = "str_escape", since = "1.34.0")] |
1606 | impl<'a> FusedIterator for $Name<'a> {} |
1607 | )+} |
1608 | } |
1609 | |
1610 | escape_types_impls!(EscapeDebug, EscapeDefault, EscapeUnicode); |
1611 |
Definitions
- Chars
- iter
- Item
- next
- count
- advance_by
- size_hint
- last
- fmt
- next_back
- as_str
- CharIndices
- front_offset
- iter
- Item
- next
- count
- size_hint
- last
- next_back
- as_str
- offset
- Bytes
- Item
- next
- size_hint
- count
- last
- nth
- all
- any
- find
- position
- rposition
- __iterator_get_unchecked
- next_back
- nth_back
- rfind
- len
- is_empty
- derive_pattern_clone
- generate_pattern_iterators
- SplitInternal
- start
- end
- matcher
- allow_trailing_empty
- finished
- fmt
- get_end
- next
- next_inclusive
- next_back
- next_back_inclusive
- remainder
- Split
- RSplit
- remainder
- remainder
- SplitTerminator
- RSplitTerminator
- remainder
- remainder
- SplitNInternal
- iter
- count
- fmt
- next
- next_back
- remainder
- SplitN
- RSplitN
- remainder
- remainder
- MatchIndicesInternal
- fmt
- next
- next_back
- MatchIndices
- RMatchIndices
- MatchesInternal
- fmt
- next
- next_back
- Matches
- RMatches
- Lines
- Item
- next
- size_hint
- last
- next_back
- remainder
- LinesAny
- Item
- next
- size_hint
- next_back
- SplitWhitespace
- inner
- SplitAsciiWhitespace
- inner
- SplitInclusive
- Item
- next
- size_hint
- last
- next_back
- remainder
- Item
- next
- size_hint
- last
- next_back
- remainder
- Item
- next
- fmt
- clone
- next_back
- remainder
- EncodeUtf16
- chars
- extra
- fmt
- Item
- next
- size_hint
- EscapeDebug
- inner
- EscapeDefault
- inner
- EscapeUnicode
- inner
Learn Rust with the experts
Find out more