1 //! String manipulation.
2 //!
3 //! For more details, see the [`std::str`] module.
4 //!
5 //! [`std::str`]: ../../std/str/index.html
6
7 #![stable(feature = "rust1", since = "1.0.0")]
8
9 mod converts;
10 mod error;
11 mod iter;
12 mod traits;
13 mod validations;
14
15 use self::pattern::Pattern;
16 use self::pattern::{DoubleEndedSearcher, ReverseSearcher, Searcher};
17
18 use crate::char::{self, EscapeDebugExtArgs};
19 use crate::mem;
20 use crate::slice::{self, SliceIndex};
21
22 pub mod pattern;
23
24 #[unstable(feature = "str_internals", issue = "none")]
25 #[allow(missing_docs)]
26 pub mod lossy;
27
28 #[stable(feature = "rust1", since = "1.0.0")]
29 pub use converts::{from_utf8, from_utf8_unchecked};
30
31 #[stable(feature = "str_mut_extras", since = "1.20.0")]
32 pub use converts::{from_utf8_mut, from_utf8_unchecked_mut};
33
34 #[stable(feature = "rust1", since = "1.0.0")]
35 pub use error::{ParseBoolError, Utf8Error};
36
37 #[stable(feature = "rust1", since = "1.0.0")]
38 pub use traits::FromStr;
39
40 #[stable(feature = "rust1", since = "1.0.0")]
41 pub use iter::{Bytes, CharIndices, Chars, Lines, SplitWhitespace};
42
43 #[stable(feature = "rust1", since = "1.0.0")]
44 #[allow(deprecated)]
45 pub use iter::LinesAny;
46
47 #[stable(feature = "rust1", since = "1.0.0")]
48 pub use iter::{RSplit, RSplitTerminator, Split, SplitTerminator};
49
50 #[stable(feature = "rust1", since = "1.0.0")]
51 pub use iter::{RSplitN, SplitN};
52
53 #[stable(feature = "str_matches", since = "1.2.0")]
54 pub use iter::{Matches, RMatches};
55
56 #[stable(feature = "str_match_indices", since = "1.5.0")]
57 pub use iter::{MatchIndices, RMatchIndices};
58
59 #[stable(feature = "encode_utf16", since = "1.8.0")]
60 pub use iter::EncodeUtf16;
61
62 #[stable(feature = "str_escape", since = "1.34.0")]
63 pub use iter::{EscapeDebug, EscapeDefault, EscapeUnicode};
64
65 #[stable(feature = "split_ascii_whitespace", since = "1.34.0")]
66 pub use iter::SplitAsciiWhitespace;
67
68 #[stable(feature = "split_inclusive", since = "1.51.0")]
69 pub use iter::SplitInclusive;
70
71 #[unstable(feature = "str_internals", issue = "none")]
72 pub use validations::{next_code_point, utf8_char_width};
73
74 use iter::MatchIndicesInternal;
75 use iter::SplitInternal;
76 use iter::{MatchesInternal, SplitNInternal};
77
78 use validations::truncate_to_char_boundary;
79
80 #[inline(never)]
81 #[cold]
82 #[track_caller]
slice_error_fail(s: &str, begin: usize, end: usize) -> !83 fn slice_error_fail(s: &str, begin: usize, end: usize) -> ! {
84 const MAX_DISPLAY_LENGTH: usize = 256;
85 let (truncated, s_trunc) = truncate_to_char_boundary(s, MAX_DISPLAY_LENGTH);
86 let ellipsis = if truncated { "[...]" } else { "" };
87
88 // 1. out of bounds
89 if begin > s.len() || end > s.len() {
90 let oob_index = if begin > s.len() { begin } else { end };
91 panic!("byte index {} is out of bounds of `{}`{}", oob_index, s_trunc, ellipsis);
92 }
93
94 // 2. begin <= end
95 assert!(
96 begin <= end,
97 "begin <= end ({} <= {}) when slicing `{}`{}",
98 begin,
99 end,
100 s_trunc,
101 ellipsis
102 );
103
104 // 3. character boundary
105 let index = if !s.is_char_boundary(begin) { begin } else { end };
106 // find the character
107 let mut char_start = index;
108 while !s.is_char_boundary(char_start) {
109 char_start -= 1;
110 }
111 // `char_start` must be less than len and a char boundary
112 let ch = s[char_start..].chars().next().unwrap();
113 let char_range = char_start..char_start + ch.len_utf8();
114 panic!(
115 "byte index {} is not a char boundary; it is inside {:?} (bytes {:?}) of `{}`{}",
116 index, ch, char_range, s_trunc, ellipsis
117 );
118 }
119
120 #[lang = "str"]
121 #[cfg(not(test))]
122 impl str {
123 /// Returns the length of `self`.
124 ///
125 /// This length is in bytes, not [`char`]s or graphemes. In other words,
126 /// it might not be what a human considers the length of the string.
127 ///
128 /// [`char`]: prim@char
129 ///
130 /// # Examples
131 ///
132 /// Basic usage:
133 ///
134 /// ```
135 /// let len = "foo".len();
136 /// assert_eq!(3, len);
137 ///
138 /// assert_eq!("ƒoo".len(), 4); // fancy f!
139 /// assert_eq!("ƒoo".chars().count(), 3);
140 /// ```
141 #[stable(feature = "rust1", since = "1.0.0")]
142 #[rustc_const_stable(feature = "const_str_len", since = "1.39.0")]
143 #[must_use]
144 #[inline]
len(&self) -> usize145 pub const fn len(&self) -> usize {
146 self.as_bytes().len()
147 }
148
149 /// Returns `true` if `self` has a length of zero bytes.
150 ///
151 /// # Examples
152 ///
153 /// Basic usage:
154 ///
155 /// ```
156 /// let s = "";
157 /// assert!(s.is_empty());
158 ///
159 /// let s = "not empty";
160 /// assert!(!s.is_empty());
161 /// ```
162 #[stable(feature = "rust1", since = "1.0.0")]
163 #[rustc_const_stable(feature = "const_str_is_empty", since = "1.39.0")]
164 #[must_use]
165 #[inline]
is_empty(&self) -> bool166 pub const fn is_empty(&self) -> bool {
167 self.len() == 0
168 }
169
170 /// Checks that `index`-th byte is the first byte in a UTF-8 code point
171 /// sequence or the end of the string.
172 ///
173 /// The start and end of the string (when `index == self.len()`) are
174 /// considered to be boundaries.
175 ///
176 /// Returns `false` if `index` is greater than `self.len()`.
177 ///
178 /// # Examples
179 ///
180 /// ```
181 /// let s = "Löwe 老虎 Léopard";
182 /// assert!(s.is_char_boundary(0));
183 /// // start of `老`
184 /// assert!(s.is_char_boundary(6));
185 /// assert!(s.is_char_boundary(s.len()));
186 ///
187 /// // second byte of `ö`
188 /// assert!(!s.is_char_boundary(2));
189 ///
190 /// // third byte of `老`
191 /// assert!(!s.is_char_boundary(8));
192 /// ```
193 #[must_use]
194 #[stable(feature = "is_char_boundary", since = "1.9.0")]
195 #[inline]
is_char_boundary(&self, index: usize) -> bool196 pub fn is_char_boundary(&self, index: usize) -> bool {
197 // 0 is always ok.
198 // Test for 0 explicitly so that it can optimize out the check
199 // easily and skip reading string data for that case.
200 // Note that optimizing `self.get(..index)` relies on this.
201 if index == 0 {
202 return true;
203 }
204
205 match self.as_bytes().get(index) {
206 // For `None` we have two options:
207 //
208 // - index == self.len()
209 // Empty strings are valid, so return true
210 // - index > self.len()
211 // In this case return false
212 //
213 // The check is placed exactly here, because it improves generated
214 // code on higher opt-levels. See PR #84751 for more details.
215 None => index == self.len(),
216
217 // This is bit magic equivalent to: b < 128 || b >= 192
218 Some(&b) => (b as i8) >= -0x40,
219 }
220 }
221
222 /// Converts a string slice to a byte slice. To convert the byte slice back
223 /// into a string slice, use the [`from_utf8`] function.
224 ///
225 /// # Examples
226 ///
227 /// Basic usage:
228 ///
229 /// ```
230 /// let bytes = "bors".as_bytes();
231 /// assert_eq!(b"bors", bytes);
232 /// ```
233 #[stable(feature = "rust1", since = "1.0.0")]
234 #[rustc_const_stable(feature = "str_as_bytes", since = "1.39.0")]
235 #[must_use]
236 #[inline(always)]
237 #[allow(unused_attributes)]
as_bytes(&self) -> &[u8]238 pub const fn as_bytes(&self) -> &[u8] {
239 // SAFETY: const sound because we transmute two types with the same layout
240 unsafe { mem::transmute(self) }
241 }
242
243 /// Converts a mutable string slice to a mutable byte slice.
244 ///
245 /// # Safety
246 ///
247 /// The caller must ensure that the content of the slice is valid UTF-8
248 /// before the borrow ends and the underlying `str` is used.
249 ///
250 /// Use of a `str` whose contents are not valid UTF-8 is undefined behavior.
251 ///
252 /// # Examples
253 ///
254 /// Basic usage:
255 ///
256 /// ```
257 /// let mut s = String::from("Hello");
258 /// let bytes = unsafe { s.as_bytes_mut() };
259 ///
260 /// assert_eq!(b"Hello", bytes);
261 /// ```
262 ///
263 /// Mutability:
264 ///
265 /// ```
266 /// let mut s = String::from("∈");
267 ///
268 /// unsafe {
269 /// let bytes = s.as_bytes_mut();
270 ///
271 /// bytes[0] = 0xF0;
272 /// bytes[1] = 0x9F;
273 /// bytes[2] = 0x8D;
274 /// bytes[3] = 0x94;
275 /// }
276 ///
277 /// assert_eq!("∈", s);
278 /// ```
279 #[stable(feature = "str_mut_extras", since = "1.20.0")]
280 #[must_use]
281 #[inline(always)]
as_bytes_mut(&mut self) -> &mut [u8]282 pub unsafe fn as_bytes_mut(&mut self) -> &mut [u8] {
283 // SAFETY: the cast from `&str` to `&[u8]` is safe since `str`
284 // has the same layout as `&[u8]` (only libstd can make this guarantee).
285 // The pointer dereference is safe since it comes from a mutable reference which
286 // is guaranteed to be valid for writes.
287 unsafe { &mut *(self as *mut str as *mut [u8]) }
288 }
289
290 /// Converts a string slice to a raw pointer.
291 ///
292 /// As string slices are a slice of bytes, the raw pointer points to a
293 /// [`u8`]. This pointer will be pointing to the first byte of the string
294 /// slice.
295 ///
296 /// The caller must ensure that the returned pointer is never written to.
297 /// If you need to mutate the contents of the string slice, use [`as_mut_ptr`].
298 ///
299 /// [`as_mut_ptr`]: str::as_mut_ptr
300 ///
301 /// # Examples
302 ///
303 /// Basic usage:
304 ///
305 /// ```
306 /// let s = "Hello";
307 /// let ptr = s.as_ptr();
308 /// ```
309 #[stable(feature = "rust1", since = "1.0.0")]
310 #[rustc_const_stable(feature = "rustc_str_as_ptr", since = "1.32.0")]
311 #[must_use]
312 #[inline]
as_ptr(&self) -> *const u8313 pub const fn as_ptr(&self) -> *const u8 {
314 self as *const str as *const u8
315 }
316
317 /// Converts a mutable string slice to a raw pointer.
318 ///
319 /// As string slices are a slice of bytes, the raw pointer points to a
320 /// [`u8`]. This pointer will be pointing to the first byte of the string
321 /// slice.
322 ///
323 /// It is your responsibility to make sure that the string slice only gets
324 /// modified in a way that it remains valid UTF-8.
325 #[stable(feature = "str_as_mut_ptr", since = "1.36.0")]
326 #[must_use]
327 #[inline]
as_mut_ptr(&mut self) -> *mut u8328 pub fn as_mut_ptr(&mut self) -> *mut u8 {
329 self as *mut str as *mut u8
330 }
331
332 /// Returns a subslice of `str`.
333 ///
334 /// This is the non-panicking alternative to indexing the `str`. Returns
335 /// [`None`] whenever equivalent indexing operation would panic.
336 ///
337 /// # Examples
338 ///
339 /// ```
340 /// let v = String::from("∈");
341 ///
342 /// assert_eq!(Some(""), v.get(0..4));
343 ///
344 /// // indices not on UTF-8 sequence boundaries
345 /// assert!(v.get(1..).is_none());
346 /// assert!(v.get(..8).is_none());
347 ///
348 /// // out of bounds
349 /// assert!(v.get(..42).is_none());
350 /// ```
351 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
352 #[inline]
get<I: SliceIndex<str>>(&self, i: I) -> Option<&I::Output>353 pub fn get<I: SliceIndex<str>>(&self, i: I) -> Option<&I::Output> {
354 i.get(self)
355 }
356
357 /// Returns a mutable subslice of `str`.
358 ///
359 /// This is the non-panicking alternative to indexing the `str`. Returns
360 /// [`None`] whenever equivalent indexing operation would panic.
361 ///
362 /// # Examples
363 ///
364 /// ```
365 /// let mut v = String::from("hello");
366 /// // correct length
367 /// assert!(v.get_mut(0..5).is_some());
368 /// // out of bounds
369 /// assert!(v.get_mut(..42).is_none());
370 /// assert_eq!(Some("he"), v.get_mut(0..2).map(|v| &*v));
371 ///
372 /// assert_eq!("hello", v);
373 /// {
374 /// let s = v.get_mut(0..2);
375 /// let s = s.map(|s| {
376 /// s.make_ascii_uppercase();
377 /// &*s
378 /// });
379 /// assert_eq!(Some("HE"), s);
380 /// }
381 /// assert_eq!("HEllo", v);
382 /// ```
383 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
384 #[inline]
get_mut<I: SliceIndex<str>>(&mut self, i: I) -> Option<&mut I::Output>385 pub fn get_mut<I: SliceIndex<str>>(&mut self, i: I) -> Option<&mut I::Output> {
386 i.get_mut(self)
387 }
388
389 /// Returns an unchecked subslice of `str`.
390 ///
391 /// This is the unchecked alternative to indexing the `str`.
392 ///
393 /// # Safety
394 ///
395 /// Callers of this function are responsible that these preconditions are
396 /// satisfied:
397 ///
398 /// * The starting index must not exceed the ending index;
399 /// * Indexes must be within bounds of the original slice;
400 /// * Indexes must lie on UTF-8 sequence boundaries.
401 ///
402 /// Failing that, the returned string slice may reference invalid memory or
403 /// violate the invariants communicated by the `str` type.
404 ///
405 /// # Examples
406 ///
407 /// ```
408 /// let v = "∈";
409 /// unsafe {
410 /// assert_eq!("", v.get_unchecked(0..4));
411 /// assert_eq!("∈", v.get_unchecked(4..7));
412 /// assert_eq!("", v.get_unchecked(7..11));
413 /// }
414 /// ```
415 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
416 #[inline]
get_unchecked<I: SliceIndex<str>>(&self, i: I) -> &I::Output417 pub unsafe fn get_unchecked<I: SliceIndex<str>>(&self, i: I) -> &I::Output {
418 // SAFETY: the caller must uphold the safety contract for `get_unchecked`;
419 // the slice is dereferencable because `self` is a safe reference.
420 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
421 unsafe { &*i.get_unchecked(self) }
422 }
423
424 /// Returns a mutable, unchecked subslice of `str`.
425 ///
426 /// This is the unchecked alternative to indexing the `str`.
427 ///
428 /// # Safety
429 ///
430 /// Callers of this function are responsible that these preconditions are
431 /// satisfied:
432 ///
433 /// * The starting index must not exceed the ending index;
434 /// * Indexes must be within bounds of the original slice;
435 /// * Indexes must lie on UTF-8 sequence boundaries.
436 ///
437 /// Failing that, the returned string slice may reference invalid memory or
438 /// violate the invariants communicated by the `str` type.
439 ///
440 /// # Examples
441 ///
442 /// ```
443 /// let mut v = String::from("∈");
444 /// unsafe {
445 /// assert_eq!("", v.get_unchecked_mut(0..4));
446 /// assert_eq!("∈", v.get_unchecked_mut(4..7));
447 /// assert_eq!("", v.get_unchecked_mut(7..11));
448 /// }
449 /// ```
450 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
451 #[inline]
get_unchecked_mut<I: SliceIndex<str>>(&mut self, i: I) -> &mut I::Output452 pub unsafe fn get_unchecked_mut<I: SliceIndex<str>>(&mut self, i: I) -> &mut I::Output {
453 // SAFETY: the caller must uphold the safety contract for `get_unchecked_mut`;
454 // the slice is dereferencable because `self` is a safe reference.
455 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
456 unsafe { &mut *i.get_unchecked_mut(self) }
457 }
458
459 /// Creates a string slice from another string slice, bypassing safety
460 /// checks.
461 ///
462 /// This is generally not recommended, use with caution! For a safe
463 /// alternative see [`str`] and [`Index`].
464 ///
465 /// [`Index`]: crate::ops::Index
466 ///
467 /// This new slice goes from `begin` to `end`, including `begin` but
468 /// excluding `end`.
469 ///
470 /// To get a mutable string slice instead, see the
471 /// [`slice_mut_unchecked`] method.
472 ///
473 /// [`slice_mut_unchecked`]: str::slice_mut_unchecked
474 ///
475 /// # Safety
476 ///
477 /// Callers of this function are responsible that three preconditions are
478 /// satisfied:
479 ///
480 /// * `begin` must not exceed `end`.
481 /// * `begin` and `end` must be byte positions within the string slice.
482 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
483 ///
484 /// # Examples
485 ///
486 /// Basic usage:
487 ///
488 /// ```
489 /// let s = "Löwe 老虎 Léopard";
490 ///
491 /// unsafe {
492 /// assert_eq!("Löwe 老虎 Léopard", s.slice_unchecked(0, 21));
493 /// }
494 ///
495 /// let s = "Hello, world!";
496 ///
497 /// unsafe {
498 /// assert_eq!("world", s.slice_unchecked(7, 12));
499 /// }
500 /// ```
501 #[stable(feature = "rust1", since = "1.0.0")]
502 #[rustc_deprecated(since = "1.29.0", reason = "use `get_unchecked(begin..end)` instead")]
503 #[must_use]
504 #[inline]
slice_unchecked(&self, begin: usize, end: usize) -> &str505 pub unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
506 // SAFETY: the caller must uphold the safety contract for `get_unchecked`;
507 // the slice is dereferencable because `self` is a safe reference.
508 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
509 unsafe { &*(begin..end).get_unchecked(self) }
510 }
511
512 /// Creates a string slice from another string slice, bypassing safety
513 /// checks.
514 /// This is generally not recommended, use with caution! For a safe
515 /// alternative see [`str`] and [`IndexMut`].
516 ///
517 /// [`IndexMut`]: crate::ops::IndexMut
518 ///
519 /// This new slice goes from `begin` to `end`, including `begin` but
520 /// excluding `end`.
521 ///
522 /// To get an immutable string slice instead, see the
523 /// [`slice_unchecked`] method.
524 ///
525 /// [`slice_unchecked`]: str::slice_unchecked
526 ///
527 /// # Safety
528 ///
529 /// Callers of this function are responsible that three preconditions are
530 /// satisfied:
531 ///
532 /// * `begin` must not exceed `end`.
533 /// * `begin` and `end` must be byte positions within the string slice.
534 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
535 #[stable(feature = "str_slice_mut", since = "1.5.0")]
536 #[rustc_deprecated(since = "1.29.0", reason = "use `get_unchecked_mut(begin..end)` instead")]
537 #[inline]
slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str538 pub unsafe fn slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str {
539 // SAFETY: the caller must uphold the safety contract for `get_unchecked_mut`;
540 // the slice is dereferencable because `self` is a safe reference.
541 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
542 unsafe { &mut *(begin..end).get_unchecked_mut(self) }
543 }
544
545 /// Divide one string slice into two at an index.
546 ///
547 /// The argument, `mid`, should be a byte offset from the start of the
548 /// string. It must also be on the boundary of a UTF-8 code point.
549 ///
550 /// The two slices returned go from the start of the string slice to `mid`,
551 /// and from `mid` to the end of the string slice.
552 ///
553 /// To get mutable string slices instead, see the [`split_at_mut`]
554 /// method.
555 ///
556 /// [`split_at_mut`]: str::split_at_mut
557 ///
558 /// # Panics
559 ///
560 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
561 /// past the end of the last code point of the string slice.
562 ///
563 /// # Examples
564 ///
565 /// Basic usage:
566 ///
567 /// ```
568 /// let s = "Per Martin-Löf";
569 ///
570 /// let (first, last) = s.split_at(3);
571 ///
572 /// assert_eq!("Per", first);
573 /// assert_eq!(" Martin-Löf", last);
574 /// ```
575 #[inline]
576 #[must_use]
577 #[stable(feature = "str_split_at", since = "1.4.0")]
split_at(&self, mid: usize) -> (&str, &str)578 pub fn split_at(&self, mid: usize) -> (&str, &str) {
579 // is_char_boundary checks that the index is in [0, .len()]
580 if self.is_char_boundary(mid) {
581 // SAFETY: just checked that `mid` is on a char boundary.
582 unsafe { (self.get_unchecked(0..mid), self.get_unchecked(mid..self.len())) }
583 } else {
584 slice_error_fail(self, 0, mid)
585 }
586 }
587
588 /// Divide one mutable string slice into two at an index.
589 ///
590 /// The argument, `mid`, should be a byte offset from the start of the
591 /// string. It must also be on the boundary of a UTF-8 code point.
592 ///
593 /// The two slices returned go from the start of the string slice to `mid`,
594 /// and from `mid` to the end of the string slice.
595 ///
596 /// To get immutable string slices instead, see the [`split_at`] method.
597 ///
598 /// [`split_at`]: str::split_at
599 ///
600 /// # Panics
601 ///
602 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
603 /// past the end of the last code point of the string slice.
604 ///
605 /// # Examples
606 ///
607 /// Basic usage:
608 ///
609 /// ```
610 /// let mut s = "Per Martin-Löf".to_string();
611 /// {
612 /// let (first, last) = s.split_at_mut(3);
613 /// first.make_ascii_uppercase();
614 /// assert_eq!("PER", first);
615 /// assert_eq!(" Martin-Löf", last);
616 /// }
617 /// assert_eq!("PER Martin-Löf", s);
618 /// ```
619 #[inline]
620 #[must_use]
621 #[stable(feature = "str_split_at", since = "1.4.0")]
split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str)622 pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {
623 // is_char_boundary checks that the index is in [0, .len()]
624 if self.is_char_boundary(mid) {
625 let len = self.len();
626 let ptr = self.as_mut_ptr();
627 // SAFETY: just checked that `mid` is on a char boundary.
628 unsafe {
629 (
630 from_utf8_unchecked_mut(slice::from_raw_parts_mut(ptr, mid)),
631 from_utf8_unchecked_mut(slice::from_raw_parts_mut(ptr.add(mid), len - mid)),
632 )
633 }
634 } else {
635 slice_error_fail(self, 0, mid)
636 }
637 }
638
639 /// Returns an iterator over the [`char`]s of a string slice.
640 ///
641 /// As a string slice consists of valid UTF-8, we can iterate through a
642 /// string slice by [`char`]. This method returns such an iterator.
643 ///
644 /// It's important to remember that [`char`] represents a Unicode Scalar
645 /// Value, and might not match your idea of what a 'character' is. Iteration
646 /// over grapheme clusters may be what you actually want. This functionality
647 /// is not provided by Rust's standard library, check crates.io instead.
648 ///
649 /// # Examples
650 ///
651 /// Basic usage:
652 ///
653 /// ```
654 /// let word = "goodbye";
655 ///
656 /// let count = word.chars().count();
657 /// assert_eq!(7, count);
658 ///
659 /// let mut chars = word.chars();
660 ///
661 /// assert_eq!(Some('g'), chars.next());
662 /// assert_eq!(Some('o'), chars.next());
663 /// assert_eq!(Some('o'), chars.next());
664 /// assert_eq!(Some('d'), chars.next());
665 /// assert_eq!(Some('b'), chars.next());
666 /// assert_eq!(Some('y'), chars.next());
667 /// assert_eq!(Some('e'), chars.next());
668 ///
669 /// assert_eq!(None, chars.next());
670 /// ```
671 ///
672 /// Remember, [`char`]s might not match your intuition about characters:
673 ///
674 /// [`char`]: prim@char
675 ///
676 /// ```
677 /// let y = "y̆";
678 ///
679 /// let mut chars = y.chars();
680 ///
681 /// assert_eq!(Some('y'), chars.next()); // not 'y̆'
682 /// assert_eq!(Some('\u{0306}'), chars.next());
683 ///
684 /// assert_eq!(None, chars.next());
685 /// ```
686 #[stable(feature = "rust1", since = "1.0.0")]
687 #[inline]
chars(&self) -> Chars<'_>688 pub fn chars(&self) -> Chars<'_> {
689 Chars { iter: self.as_bytes().iter() }
690 }
691
692 /// Returns an iterator over the [`char`]s of a string slice, and their
693 /// positions.
694 ///
695 /// As a string slice consists of valid UTF-8, we can iterate through a
696 /// string slice by [`char`]. This method returns an iterator of both
697 /// these [`char`]s, as well as their byte positions.
698 ///
699 /// The iterator yields tuples. The position is first, the [`char`] is
700 /// second.
701 ///
702 /// # Examples
703 ///
704 /// Basic usage:
705 ///
706 /// ```
707 /// let word = "goodbye";
708 ///
709 /// let count = word.char_indices().count();
710 /// assert_eq!(7, count);
711 ///
712 /// let mut char_indices = word.char_indices();
713 ///
714 /// assert_eq!(Some((0, 'g')), char_indices.next());
715 /// assert_eq!(Some((1, 'o')), char_indices.next());
716 /// assert_eq!(Some((2, 'o')), char_indices.next());
717 /// assert_eq!(Some((3, 'd')), char_indices.next());
718 /// assert_eq!(Some((4, 'b')), char_indices.next());
719 /// assert_eq!(Some((5, 'y')), char_indices.next());
720 /// assert_eq!(Some((6, 'e')), char_indices.next());
721 ///
722 /// assert_eq!(None, char_indices.next());
723 /// ```
724 ///
725 /// Remember, [`char`]s might not match your intuition about characters:
726 ///
727 /// [`char`]: prim@char
728 ///
729 /// ```
730 /// let yes = "y̆es";
731 ///
732 /// let mut char_indices = yes.char_indices();
733 ///
734 /// assert_eq!(Some((0, 'y')), char_indices.next()); // not (0, 'y̆')
735 /// assert_eq!(Some((1, '\u{0306}')), char_indices.next());
736 ///
737 /// // note the 3 here - the last character took up two bytes
738 /// assert_eq!(Some((3, 'e')), char_indices.next());
739 /// assert_eq!(Some((4, 's')), char_indices.next());
740 ///
741 /// assert_eq!(None, char_indices.next());
742 /// ```
743 #[stable(feature = "rust1", since = "1.0.0")]
744 #[inline]
char_indices(&self) -> CharIndices<'_>745 pub fn char_indices(&self) -> CharIndices<'_> {
746 CharIndices { front_offset: 0, iter: self.chars() }
747 }
748
749 /// An iterator over the bytes of a string slice.
750 ///
751 /// As a string slice consists of a sequence of bytes, we can iterate
752 /// through a string slice by byte. This method returns such an iterator.
753 ///
754 /// # Examples
755 ///
756 /// Basic usage:
757 ///
758 /// ```
759 /// let mut bytes = "bors".bytes();
760 ///
761 /// assert_eq!(Some(b'b'), bytes.next());
762 /// assert_eq!(Some(b'o'), bytes.next());
763 /// assert_eq!(Some(b'r'), bytes.next());
764 /// assert_eq!(Some(b's'), bytes.next());
765 ///
766 /// assert_eq!(None, bytes.next());
767 /// ```
768 #[stable(feature = "rust1", since = "1.0.0")]
769 #[inline]
bytes(&self) -> Bytes<'_>770 pub fn bytes(&self) -> Bytes<'_> {
771 Bytes(self.as_bytes().iter().copied())
772 }
773
774 /// Splits a string slice by whitespace.
775 ///
776 /// The iterator returned will return string slices that are sub-slices of
777 /// the original string slice, separated by any amount of whitespace.
778 ///
779 /// 'Whitespace' is defined according to the terms of the Unicode Derived
780 /// Core Property `White_Space`. If you only want to split on ASCII whitespace
781 /// instead, use [`split_ascii_whitespace`].
782 ///
783 /// [`split_ascii_whitespace`]: str::split_ascii_whitespace
784 ///
785 /// # Examples
786 ///
787 /// Basic usage:
788 ///
789 /// ```
790 /// let mut iter = "A few words".split_whitespace();
791 ///
792 /// assert_eq!(Some("A"), iter.next());
793 /// assert_eq!(Some("few"), iter.next());
794 /// assert_eq!(Some("words"), iter.next());
795 ///
796 /// assert_eq!(None, iter.next());
797 /// ```
798 ///
799 /// All kinds of whitespace are considered:
800 ///
801 /// ```
802 /// let mut iter = " Mary had\ta\u{2009}little \n\t lamb".split_whitespace();
803 /// assert_eq!(Some("Mary"), iter.next());
804 /// assert_eq!(Some("had"), iter.next());
805 /// assert_eq!(Some("a"), iter.next());
806 /// assert_eq!(Some("little"), iter.next());
807 /// assert_eq!(Some("lamb"), iter.next());
808 ///
809 /// assert_eq!(None, iter.next());
810 /// ```
811 #[must_use = "this returns the split string as an iterator, \
812 without modifying the original"]
813 #[stable(feature = "split_whitespace", since = "1.1.0")]
814 #[inline]
split_whitespace(&self) -> SplitWhitespace<'_>815 pub fn split_whitespace(&self) -> SplitWhitespace<'_> {
816 SplitWhitespace { inner: self.split(IsWhitespace).filter(IsNotEmpty) }
817 }
818
819 /// Splits a string slice by ASCII whitespace.
820 ///
821 /// The iterator returned will return string slices that are sub-slices of
822 /// the original string slice, separated by any amount of ASCII whitespace.
823 ///
824 /// To split by Unicode `Whitespace` instead, use [`split_whitespace`].
825 ///
826 /// [`split_whitespace`]: str::split_whitespace
827 ///
828 /// # Examples
829 ///
830 /// Basic usage:
831 ///
832 /// ```
833 /// let mut iter = "A few words".split_ascii_whitespace();
834 ///
835 /// assert_eq!(Some("A"), iter.next());
836 /// assert_eq!(Some("few"), iter.next());
837 /// assert_eq!(Some("words"), iter.next());
838 ///
839 /// assert_eq!(None, iter.next());
840 /// ```
841 ///
842 /// All kinds of ASCII whitespace are considered:
843 ///
844 /// ```
845 /// let mut iter = " Mary had\ta little \n\t lamb".split_ascii_whitespace();
846 /// assert_eq!(Some("Mary"), iter.next());
847 /// assert_eq!(Some("had"), iter.next());
848 /// assert_eq!(Some("a"), iter.next());
849 /// assert_eq!(Some("little"), iter.next());
850 /// assert_eq!(Some("lamb"), iter.next());
851 ///
852 /// assert_eq!(None, iter.next());
853 /// ```
854 #[must_use = "this returns the split string as an iterator, \
855 without modifying the original"]
856 #[stable(feature = "split_ascii_whitespace", since = "1.34.0")]
857 #[inline]
split_ascii_whitespace(&self) -> SplitAsciiWhitespace<'_>858 pub fn split_ascii_whitespace(&self) -> SplitAsciiWhitespace<'_> {
859 let inner =
860 self.as_bytes().split(IsAsciiWhitespace).filter(BytesIsNotEmpty).map(UnsafeBytesToStr);
861 SplitAsciiWhitespace { inner }
862 }
863
864 /// An iterator over the lines of a string, as string slices.
865 ///
866 /// Lines are ended with either a newline (`\n`) or a carriage return with
867 /// a line feed (`\r\n`).
868 ///
869 /// The final line ending is optional. A string that ends with a final line
870 /// ending will return the same lines as an otherwise identical string
871 /// without a final line ending.
872 ///
873 /// # Examples
874 ///
875 /// Basic usage:
876 ///
877 /// ```
878 /// let text = "foo\r\nbar\n\nbaz\n";
879 /// let mut lines = text.lines();
880 ///
881 /// assert_eq!(Some("foo"), lines.next());
882 /// assert_eq!(Some("bar"), lines.next());
883 /// assert_eq!(Some(""), lines.next());
884 /// assert_eq!(Some("baz"), lines.next());
885 ///
886 /// assert_eq!(None, lines.next());
887 /// ```
888 ///
889 /// The final line ending isn't required:
890 ///
891 /// ```
892 /// let text = "foo\nbar\n\r\nbaz";
893 /// let mut lines = text.lines();
894 ///
895 /// assert_eq!(Some("foo"), lines.next());
896 /// assert_eq!(Some("bar"), lines.next());
897 /// assert_eq!(Some(""), lines.next());
898 /// assert_eq!(Some("baz"), lines.next());
899 ///
900 /// assert_eq!(None, lines.next());
901 /// ```
902 #[stable(feature = "rust1", since = "1.0.0")]
903 #[inline]
lines(&self) -> Lines<'_>904 pub fn lines(&self) -> Lines<'_> {
905 Lines(self.split_terminator('\n').map(LinesAnyMap))
906 }
907
908 /// An iterator over the lines of a string.
909 #[stable(feature = "rust1", since = "1.0.0")]
910 #[rustc_deprecated(since = "1.4.0", reason = "use lines() instead now")]
911 #[inline]
912 #[allow(deprecated)]
lines_any(&self) -> LinesAny<'_>913 pub fn lines_any(&self) -> LinesAny<'_> {
914 LinesAny(self.lines())
915 }
916
917 /// Returns an iterator of `u16` over the string encoded as UTF-16.
918 ///
919 /// # Examples
920 ///
921 /// Basic usage:
922 ///
923 /// ```
924 /// let text = "Zażółć gęślą jaźń";
925 ///
926 /// let utf8_len = text.len();
927 /// let utf16_len = text.encode_utf16().count();
928 ///
929 /// assert!(utf16_len <= utf8_len);
930 /// ```
931 #[must_use = "this returns the encoded string as an iterator, \
932 without modifying the original"]
933 #[stable(feature = "encode_utf16", since = "1.8.0")]
encode_utf16(&self) -> EncodeUtf16<'_>934 pub fn encode_utf16(&self) -> EncodeUtf16<'_> {
935 EncodeUtf16 { chars: self.chars(), extra: 0 }
936 }
937
938 /// Returns `true` if the given pattern matches a sub-slice of
939 /// this string slice.
940 ///
941 /// Returns `false` if it does not.
942 ///
943 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
944 /// function or closure that determines if a character matches.
945 ///
946 /// [`char`]: prim@char
947 /// [pattern]: self::pattern
948 ///
949 /// # Examples
950 ///
951 /// Basic usage:
952 ///
953 /// ```
954 /// let bananas = "bananas";
955 ///
956 /// assert!(bananas.contains("nana"));
957 /// assert!(!bananas.contains("apples"));
958 /// ```
959 #[stable(feature = "rust1", since = "1.0.0")]
960 #[inline]
contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool961 pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
962 pat.is_contained_in(self)
963 }
964
965 /// Returns `true` if the given pattern matches a prefix of this
966 /// string slice.
967 ///
968 /// Returns `false` if it does not.
969 ///
970 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
971 /// function or closure that determines if a character matches.
972 ///
973 /// [`char`]: prim@char
974 /// [pattern]: self::pattern
975 ///
976 /// # Examples
977 ///
978 /// Basic usage:
979 ///
980 /// ```
981 /// let bananas = "bananas";
982 ///
983 /// assert!(bananas.starts_with("bana"));
984 /// assert!(!bananas.starts_with("nana"));
985 /// ```
986 #[stable(feature = "rust1", since = "1.0.0")]
starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool987 pub fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
988 pat.is_prefix_of(self)
989 }
990
991 /// Returns `true` if the given pattern matches a suffix of this
992 /// string slice.
993 ///
994 /// Returns `false` if it does not.
995 ///
996 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
997 /// function or closure that determines if a character matches.
998 ///
999 /// [`char`]: prim@char
1000 /// [pattern]: self::pattern
1001 ///
1002 /// # Examples
1003 ///
1004 /// Basic usage:
1005 ///
1006 /// ```
1007 /// let bananas = "bananas";
1008 ///
1009 /// assert!(bananas.ends_with("anas"));
1010 /// assert!(!bananas.ends_with("nana"));
1011 /// ```
1012 #[stable(feature = "rust1", since = "1.0.0")]
ends_with<'a, P>(&'a self, pat: P) -> bool where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1013 pub fn ends_with<'a, P>(&'a self, pat: P) -> bool
1014 where
1015 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1016 {
1017 pat.is_suffix_of(self)
1018 }
1019
1020 /// Returns the byte index of the first character of this string slice that
1021 /// matches the pattern.
1022 ///
1023 /// Returns [`None`] if the pattern doesn't match.
1024 ///
1025 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1026 /// function or closure that determines if a character matches.
1027 ///
1028 /// [`char`]: prim@char
1029 /// [pattern]: self::pattern
1030 ///
1031 /// # Examples
1032 ///
1033 /// Simple patterns:
1034 ///
1035 /// ```
1036 /// let s = "Löwe 老虎 Léopard Gepardi";
1037 ///
1038 /// assert_eq!(s.find('L'), Some(0));
1039 /// assert_eq!(s.find('é'), Some(14));
1040 /// assert_eq!(s.find("pard"), Some(17));
1041 /// ```
1042 ///
1043 /// More complex patterns using point-free style and closures:
1044 ///
1045 /// ```
1046 /// let s = "Löwe 老虎 Léopard";
1047 ///
1048 /// assert_eq!(s.find(char::is_whitespace), Some(5));
1049 /// assert_eq!(s.find(char::is_lowercase), Some(1));
1050 /// assert_eq!(s.find(|c: char| c.is_whitespace() || c.is_lowercase()), Some(1));
1051 /// assert_eq!(s.find(|c: char| (c < 'o') && (c > 'a')), Some(4));
1052 /// ```
1053 ///
1054 /// Not finding the pattern:
1055 ///
1056 /// ```
1057 /// let s = "Löwe 老虎 Léopard";
1058 /// let x: &[_] = &['1', '2'];
1059 ///
1060 /// assert_eq!(s.find(x), None);
1061 /// ```
1062 #[stable(feature = "rust1", since = "1.0.0")]
1063 #[inline]
find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>1064 pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
1065 pat.into_searcher(self).next_match().map(|(i, _)| i)
1066 }
1067
1068 /// Returns the byte index for the first character of the rightmost match of the pattern in
1069 /// this string slice.
1070 ///
1071 /// Returns [`None`] if the pattern doesn't match.
1072 ///
1073 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1074 /// function or closure that determines if a character matches.
1075 ///
1076 /// [`char`]: prim@char
1077 /// [pattern]: self::pattern
1078 ///
1079 /// # Examples
1080 ///
1081 /// Simple patterns:
1082 ///
1083 /// ```
1084 /// let s = "Löwe 老虎 Léopard Gepardi";
1085 ///
1086 /// assert_eq!(s.rfind('L'), Some(13));
1087 /// assert_eq!(s.rfind('é'), Some(14));
1088 /// assert_eq!(s.rfind("pard"), Some(24));
1089 /// ```
1090 ///
1091 /// More complex patterns with closures:
1092 ///
1093 /// ```
1094 /// let s = "Löwe 老虎 Léopard";
1095 ///
1096 /// assert_eq!(s.rfind(char::is_whitespace), Some(12));
1097 /// assert_eq!(s.rfind(char::is_lowercase), Some(20));
1098 /// ```
1099 ///
1100 /// Not finding the pattern:
1101 ///
1102 /// ```
1103 /// let s = "Löwe 老虎 Léopard";
1104 /// let x: &[_] = &['1', '2'];
1105 ///
1106 /// assert_eq!(s.rfind(x), None);
1107 /// ```
1108 #[stable(feature = "rust1", since = "1.0.0")]
1109 #[inline]
rfind<'a, P>(&'a self, pat: P) -> Option<usize> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1110 pub fn rfind<'a, P>(&'a self, pat: P) -> Option<usize>
1111 where
1112 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1113 {
1114 pat.into_searcher(self).next_match_back().map(|(i, _)| i)
1115 }
1116
1117 /// An iterator over substrings of this string slice, separated by
1118 /// characters matched by a pattern.
1119 ///
1120 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1121 /// function or closure that determines if a character matches.
1122 ///
1123 /// [`char`]: prim@char
1124 /// [pattern]: self::pattern
1125 ///
1126 /// # Iterator behavior
1127 ///
1128 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1129 /// allows a reverse search and forward/reverse search yields the same
1130 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1131 ///
1132 /// If the pattern allows a reverse search but its results might differ
1133 /// from a forward search, the [`rsplit`] method can be used.
1134 ///
1135 /// [`rsplit`]: str::rsplit
1136 ///
1137 /// # Examples
1138 ///
1139 /// Simple patterns:
1140 ///
1141 /// ```
1142 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
1143 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
1144 ///
1145 /// let v: Vec<&str> = "".split('X').collect();
1146 /// assert_eq!(v, [""]);
1147 ///
1148 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
1149 /// assert_eq!(v, ["lion", "", "tiger", "leopard"]);
1150 ///
1151 /// let v: Vec<&str> = "lion::tiger::leopard".split("::").collect();
1152 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1153 ///
1154 /// let v: Vec<&str> = "abc1def2ghi".split(char::is_numeric).collect();
1155 /// assert_eq!(v, ["abc", "def", "ghi"]);
1156 ///
1157 /// let v: Vec<&str> = "lionXtigerXleopard".split(char::is_uppercase).collect();
1158 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1159 /// ```
1160 ///
1161 /// If the pattern is a slice of chars, split on each occurrence of any of the characters:
1162 ///
1163 /// ```
1164 /// let v: Vec<&str> = "2020-11-03 23:59".split(&['-', ' ', ':', '@'][..]).collect();
1165 /// assert_eq!(v, ["2020", "11", "03", "23", "59"]);
1166 /// ```
1167 ///
1168 /// A more complex pattern, using a closure:
1169 ///
1170 /// ```
1171 /// let v: Vec<&str> = "abc1defXghi".split(|c| c == '1' || c == 'X').collect();
1172 /// assert_eq!(v, ["abc", "def", "ghi"]);
1173 /// ```
1174 ///
1175 /// If a string contains multiple contiguous separators, you will end up
1176 /// with empty strings in the output:
1177 ///
1178 /// ```
1179 /// let x = "||||a||b|c".to_string();
1180 /// let d: Vec<_> = x.split('|').collect();
1181 ///
1182 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1183 /// ```
1184 ///
1185 /// Contiguous separators are separated by the empty string.
1186 ///
1187 /// ```
1188 /// let x = "(///)".to_string();
1189 /// let d: Vec<_> = x.split('/').collect();
1190 ///
1191 /// assert_eq!(d, &["(", "", "", ")"]);
1192 /// ```
1193 ///
1194 /// Separators at the start or end of a string are neighbored
1195 /// by empty strings.
1196 ///
1197 /// ```
1198 /// let d: Vec<_> = "010".split("0").collect();
1199 /// assert_eq!(d, &["", "1", ""]);
1200 /// ```
1201 ///
1202 /// When the empty string is used as a separator, it separates
1203 /// every character in the string, along with the beginning
1204 /// and end of the string.
1205 ///
1206 /// ```
1207 /// let f: Vec<_> = "rust".split("").collect();
1208 /// assert_eq!(f, &["", "r", "u", "s", "t", ""]);
1209 /// ```
1210 ///
1211 /// Contiguous separators can lead to possibly surprising behavior
1212 /// when whitespace is used as the separator. This code is correct:
1213 ///
1214 /// ```
1215 /// let x = " a b c".to_string();
1216 /// let d: Vec<_> = x.split(' ').collect();
1217 ///
1218 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1219 /// ```
1220 ///
1221 /// It does _not_ give you:
1222 ///
1223 /// ```,ignore
1224 /// assert_eq!(d, &["a", "b", "c"]);
1225 /// ```
1226 ///
1227 /// Use [`split_whitespace`] for this behavior.
1228 ///
1229 /// [`split_whitespace`]: str::split_whitespace
1230 #[stable(feature = "rust1", since = "1.0.0")]
1231 #[inline]
split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P>1232 pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
1233 Split(SplitInternal {
1234 start: 0,
1235 end: self.len(),
1236 matcher: pat.into_searcher(self),
1237 allow_trailing_empty: true,
1238 finished: false,
1239 })
1240 }
1241
1242 /// An iterator over substrings of this string slice, separated by
1243 /// characters matched by a pattern. Differs from the iterator produced by
1244 /// `split` in that `split_inclusive` leaves the matched part as the
1245 /// terminator of the substring.
1246 ///
1247 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1248 /// function or closure that determines if a character matches.
1249 ///
1250 /// [`char`]: prim@char
1251 /// [pattern]: self::pattern
1252 ///
1253 /// # Examples
1254 ///
1255 /// ```
1256 /// let v: Vec<&str> = "Mary had a little lamb\nlittle lamb\nlittle lamb."
1257 /// .split_inclusive('\n').collect();
1258 /// assert_eq!(v, ["Mary had a little lamb\n", "little lamb\n", "little lamb."]);
1259 /// ```
1260 ///
1261 /// If the last element of the string is matched,
1262 /// that element will be considered the terminator of the preceding substring.
1263 /// That substring will be the last item returned by the iterator.
1264 ///
1265 /// ```
1266 /// let v: Vec<&str> = "Mary had a little lamb\nlittle lamb\nlittle lamb.\n"
1267 /// .split_inclusive('\n').collect();
1268 /// assert_eq!(v, ["Mary had a little lamb\n", "little lamb\n", "little lamb.\n"]);
1269 /// ```
1270 #[stable(feature = "split_inclusive", since = "1.51.0")]
1271 #[inline]
split_inclusive<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitInclusive<'a, P>1272 pub fn split_inclusive<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitInclusive<'a, P> {
1273 SplitInclusive(SplitInternal {
1274 start: 0,
1275 end: self.len(),
1276 matcher: pat.into_searcher(self),
1277 allow_trailing_empty: false,
1278 finished: false,
1279 })
1280 }
1281
1282 /// An iterator over substrings of the given string slice, separated by
1283 /// characters matched by a pattern and yielded in reverse order.
1284 ///
1285 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1286 /// function or closure that determines if a character matches.
1287 ///
1288 /// [`char`]: prim@char
1289 /// [pattern]: self::pattern
1290 ///
1291 /// # Iterator behavior
1292 ///
1293 /// The returned iterator requires that the pattern supports a reverse
1294 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1295 /// search yields the same elements.
1296 ///
1297 /// For iterating from the front, the [`split`] method can be used.
1298 ///
1299 /// [`split`]: str::split
1300 ///
1301 /// # Examples
1302 ///
1303 /// Simple patterns:
1304 ///
1305 /// ```
1306 /// let v: Vec<&str> = "Mary had a little lamb".rsplit(' ').collect();
1307 /// assert_eq!(v, ["lamb", "little", "a", "had", "Mary"]);
1308 ///
1309 /// let v: Vec<&str> = "".rsplit('X').collect();
1310 /// assert_eq!(v, [""]);
1311 ///
1312 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplit('X').collect();
1313 /// assert_eq!(v, ["leopard", "tiger", "", "lion"]);
1314 ///
1315 /// let v: Vec<&str> = "lion::tiger::leopard".rsplit("::").collect();
1316 /// assert_eq!(v, ["leopard", "tiger", "lion"]);
1317 /// ```
1318 ///
1319 /// A more complex pattern, using a closure:
1320 ///
1321 /// ```
1322 /// let v: Vec<&str> = "abc1defXghi".rsplit(|c| c == '1' || c == 'X').collect();
1323 /// assert_eq!(v, ["ghi", "def", "abc"]);
1324 /// ```
1325 #[stable(feature = "rust1", since = "1.0.0")]
1326 #[inline]
rsplit<'a, P>(&'a self, pat: P) -> RSplit<'a, P> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1327 pub fn rsplit<'a, P>(&'a self, pat: P) -> RSplit<'a, P>
1328 where
1329 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1330 {
1331 RSplit(self.split(pat).0)
1332 }
1333
1334 /// An iterator over substrings of the given string slice, separated by
1335 /// characters matched by a pattern.
1336 ///
1337 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1338 /// function or closure that determines if a character matches.
1339 ///
1340 /// [`char`]: prim@char
1341 /// [pattern]: self::pattern
1342 ///
1343 /// Equivalent to [`split`], except that the trailing substring
1344 /// is skipped if empty.
1345 ///
1346 /// [`split`]: str::split
1347 ///
1348 /// This method can be used for string data that is _terminated_,
1349 /// rather than _separated_ by a pattern.
1350 ///
1351 /// # Iterator behavior
1352 ///
1353 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1354 /// allows a reverse search and forward/reverse search yields the same
1355 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1356 ///
1357 /// If the pattern allows a reverse search but its results might differ
1358 /// from a forward search, the [`rsplit_terminator`] method can be used.
1359 ///
1360 /// [`rsplit_terminator`]: str::rsplit_terminator
1361 ///
1362 /// # Examples
1363 ///
1364 /// Basic usage:
1365 ///
1366 /// ```
1367 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
1368 /// assert_eq!(v, ["A", "B"]);
1369 ///
1370 /// let v: Vec<&str> = "A..B..".split_terminator(".").collect();
1371 /// assert_eq!(v, ["A", "", "B", ""]);
1372 ///
1373 /// let v: Vec<&str> = "A.B:C.D".split_terminator(&['.', ':'][..]).collect();
1374 /// assert_eq!(v, ["A", "B", "C", "D"]);
1375 /// ```
1376 #[stable(feature = "rust1", since = "1.0.0")]
1377 #[inline]
split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P>1378 pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
1379 SplitTerminator(SplitInternal { allow_trailing_empty: false, ..self.split(pat).0 })
1380 }
1381
1382 /// An iterator over substrings of `self`, separated by characters
1383 /// matched by a pattern and yielded in reverse order.
1384 ///
1385 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1386 /// function or closure that determines if a character matches.
1387 ///
1388 /// [`char`]: prim@char
1389 /// [pattern]: self::pattern
1390 ///
1391 /// Equivalent to [`split`], except that the trailing substring is
1392 /// skipped if empty.
1393 ///
1394 /// [`split`]: str::split
1395 ///
1396 /// This method can be used for string data that is _terminated_,
1397 /// rather than _separated_ by a pattern.
1398 ///
1399 /// # Iterator behavior
1400 ///
1401 /// The returned iterator requires that the pattern supports a
1402 /// reverse search, and it will be double ended if a forward/reverse
1403 /// search yields the same elements.
1404 ///
1405 /// For iterating from the front, the [`split_terminator`] method can be
1406 /// used.
1407 ///
1408 /// [`split_terminator`]: str::split_terminator
1409 ///
1410 /// # Examples
1411 ///
1412 /// ```
1413 /// let v: Vec<&str> = "A.B.".rsplit_terminator('.').collect();
1414 /// assert_eq!(v, ["B", "A"]);
1415 ///
1416 /// let v: Vec<&str> = "A..B..".rsplit_terminator(".").collect();
1417 /// assert_eq!(v, ["", "B", "", "A"]);
1418 ///
1419 /// let v: Vec<&str> = "A.B:C.D".rsplit_terminator(&['.', ':'][..]).collect();
1420 /// assert_eq!(v, ["D", "C", "B", "A"]);
1421 /// ```
1422 #[stable(feature = "rust1", since = "1.0.0")]
1423 #[inline]
rsplit_terminator<'a, P>(&'a self, pat: P) -> RSplitTerminator<'a, P> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1424 pub fn rsplit_terminator<'a, P>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1425 where
1426 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1427 {
1428 RSplitTerminator(self.split_terminator(pat).0)
1429 }
1430
1431 /// An iterator over substrings of the given string slice, separated by a
1432 /// pattern, restricted to returning at most `n` items.
1433 ///
1434 /// If `n` substrings are returned, the last substring (the `n`th substring)
1435 /// will contain the remainder of the string.
1436 ///
1437 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1438 /// function or closure that determines if a character matches.
1439 ///
1440 /// [`char`]: prim@char
1441 /// [pattern]: self::pattern
1442 ///
1443 /// # Iterator behavior
1444 ///
1445 /// The returned iterator will not be double ended, because it is
1446 /// not efficient to support.
1447 ///
1448 /// If the pattern allows a reverse search, the [`rsplitn`] method can be
1449 /// used.
1450 ///
1451 /// [`rsplitn`]: str::rsplitn
1452 ///
1453 /// # Examples
1454 ///
1455 /// Simple patterns:
1456 ///
1457 /// ```
1458 /// let v: Vec<&str> = "Mary had a little lambda".splitn(3, ' ').collect();
1459 /// assert_eq!(v, ["Mary", "had", "a little lambda"]);
1460 ///
1461 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(3, "X").collect();
1462 /// assert_eq!(v, ["lion", "", "tigerXleopard"]);
1463 ///
1464 /// let v: Vec<&str> = "abcXdef".splitn(1, 'X').collect();
1465 /// assert_eq!(v, ["abcXdef"]);
1466 ///
1467 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1468 /// assert_eq!(v, [""]);
1469 /// ```
1470 ///
1471 /// A more complex pattern, using a closure:
1472 ///
1473 /// ```
1474 /// let v: Vec<&str> = "abc1defXghi".splitn(2, |c| c == '1' || c == 'X').collect();
1475 /// assert_eq!(v, ["abc", "defXghi"]);
1476 /// ```
1477 #[stable(feature = "rust1", since = "1.0.0")]
1478 #[inline]
splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P>1479 pub fn splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P> {
1480 SplitN(SplitNInternal { iter: self.split(pat).0, count: n })
1481 }
1482
1483 /// An iterator over substrings of this string slice, separated by a
1484 /// pattern, starting from the end of the string, restricted to returning
1485 /// at most `n` items.
1486 ///
1487 /// If `n` substrings are returned, the last substring (the `n`th substring)
1488 /// will contain the remainder of the string.
1489 ///
1490 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1491 /// function or closure that determines if a character matches.
1492 ///
1493 /// [`char`]: prim@char
1494 /// [pattern]: self::pattern
1495 ///
1496 /// # Iterator behavior
1497 ///
1498 /// The returned iterator will not be double ended, because it is not
1499 /// efficient to support.
1500 ///
1501 /// For splitting from the front, the [`splitn`] method can be used.
1502 ///
1503 /// [`splitn`]: str::splitn
1504 ///
1505 /// # Examples
1506 ///
1507 /// Simple patterns:
1508 ///
1509 /// ```
1510 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(3, ' ').collect();
1511 /// assert_eq!(v, ["lamb", "little", "Mary had a"]);
1512 ///
1513 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(3, 'X').collect();
1514 /// assert_eq!(v, ["leopard", "tiger", "lionX"]);
1515 ///
1516 /// let v: Vec<&str> = "lion::tiger::leopard".rsplitn(2, "::").collect();
1517 /// assert_eq!(v, ["leopard", "lion::tiger"]);
1518 /// ```
1519 ///
1520 /// A more complex pattern, using a closure:
1521 ///
1522 /// ```
1523 /// let v: Vec<&str> = "abc1defXghi".rsplitn(2, |c| c == '1' || c == 'X').collect();
1524 /// assert_eq!(v, ["ghi", "abc1def"]);
1525 /// ```
1526 #[stable(feature = "rust1", since = "1.0.0")]
1527 #[inline]
rsplitn<'a, P>(&'a self, n: usize, pat: P) -> RSplitN<'a, P> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1528 pub fn rsplitn<'a, P>(&'a self, n: usize, pat: P) -> RSplitN<'a, P>
1529 where
1530 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1531 {
1532 RSplitN(self.splitn(n, pat).0)
1533 }
1534
1535 /// Splits the string on the first occurrence of the specified delimiter and
1536 /// returns prefix before delimiter and suffix after delimiter.
1537 ///
1538 /// # Examples
1539 ///
1540 /// ```
1541 /// assert_eq!("cfg".split_once('='), None);
1542 /// assert_eq!("cfg=foo".split_once('='), Some(("cfg", "foo")));
1543 /// assert_eq!("cfg=foo=bar".split_once('='), Some(("cfg", "foo=bar")));
1544 /// ```
1545 #[stable(feature = "str_split_once", since = "1.52.0")]
1546 #[inline]
split_once<'a, P: Pattern<'a>>(&'a self, delimiter: P) -> Option<(&'a str, &'a str)>1547 pub fn split_once<'a, P: Pattern<'a>>(&'a self, delimiter: P) -> Option<(&'a str, &'a str)> {
1548 let (start, end) = delimiter.into_searcher(self).next_match()?;
1549 // SAFETY: `Searcher` is known to return valid indices.
1550 unsafe { Some((self.get_unchecked(..start), self.get_unchecked(end..))) }
1551 }
1552
1553 /// Splits the string on the last occurrence of the specified delimiter and
1554 /// returns prefix before delimiter and suffix after delimiter.
1555 ///
1556 /// # Examples
1557 ///
1558 /// ```
1559 /// assert_eq!("cfg".rsplit_once('='), None);
1560 /// assert_eq!("cfg=foo".rsplit_once('='), Some(("cfg", "foo")));
1561 /// assert_eq!("cfg=foo=bar".rsplit_once('='), Some(("cfg=foo", "bar")));
1562 /// ```
1563 #[stable(feature = "str_split_once", since = "1.52.0")]
1564 #[inline]
rsplit_once<'a, P>(&'a self, delimiter: P) -> Option<(&'a str, &'a str)> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1565 pub fn rsplit_once<'a, P>(&'a self, delimiter: P) -> Option<(&'a str, &'a str)>
1566 where
1567 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1568 {
1569 let (start, end) = delimiter.into_searcher(self).next_match_back()?;
1570 // SAFETY: `Searcher` is known to return valid indices.
1571 unsafe { Some((self.get_unchecked(..start), self.get_unchecked(end..))) }
1572 }
1573
1574 /// An iterator over the disjoint matches of a pattern within the given string
1575 /// slice.
1576 ///
1577 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1578 /// function or closure that determines if a character matches.
1579 ///
1580 /// [`char`]: prim@char
1581 /// [pattern]: self::pattern
1582 ///
1583 /// # Iterator behavior
1584 ///
1585 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1586 /// allows a reverse search and forward/reverse search yields the same
1587 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1588 ///
1589 /// If the pattern allows a reverse search but its results might differ
1590 /// from a forward search, the [`rmatches`] method can be used.
1591 ///
1592 /// [`rmatches`]: str::matches
1593 ///
1594 /// # Examples
1595 ///
1596 /// Basic usage:
1597 ///
1598 /// ```
1599 /// let v: Vec<&str> = "abcXXXabcYYYabc".matches("abc").collect();
1600 /// assert_eq!(v, ["abc", "abc", "abc"]);
1601 ///
1602 /// let v: Vec<&str> = "1abc2abc3".matches(char::is_numeric).collect();
1603 /// assert_eq!(v, ["1", "2", "3"]);
1604 /// ```
1605 #[stable(feature = "str_matches", since = "1.2.0")]
1606 #[inline]
matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P>1607 pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1608 Matches(MatchesInternal(pat.into_searcher(self)))
1609 }
1610
1611 /// An iterator over the disjoint matches of a pattern within this string slice,
1612 /// yielded in reverse order.
1613 ///
1614 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1615 /// function or closure that determines if a character matches.
1616 ///
1617 /// [`char`]: prim@char
1618 /// [pattern]: self::pattern
1619 ///
1620 /// # Iterator behavior
1621 ///
1622 /// The returned iterator requires that the pattern supports a reverse
1623 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1624 /// search yields the same elements.
1625 ///
1626 /// For iterating from the front, the [`matches`] method can be used.
1627 ///
1628 /// [`matches`]: str::matches
1629 ///
1630 /// # Examples
1631 ///
1632 /// Basic usage:
1633 ///
1634 /// ```
1635 /// let v: Vec<&str> = "abcXXXabcYYYabc".rmatches("abc").collect();
1636 /// assert_eq!(v, ["abc", "abc", "abc"]);
1637 ///
1638 /// let v: Vec<&str> = "1abc2abc3".rmatches(char::is_numeric).collect();
1639 /// assert_eq!(v, ["3", "2", "1"]);
1640 /// ```
1641 #[stable(feature = "str_matches", since = "1.2.0")]
1642 #[inline]
rmatches<'a, P>(&'a self, pat: P) -> RMatches<'a, P> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1643 pub fn rmatches<'a, P>(&'a self, pat: P) -> RMatches<'a, P>
1644 where
1645 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1646 {
1647 RMatches(self.matches(pat).0)
1648 }
1649
1650 /// An iterator over the disjoint matches of a pattern within this string
1651 /// slice as well as the index that the match starts at.
1652 ///
1653 /// For matches of `pat` within `self` that overlap, only the indices
1654 /// corresponding to the first match are returned.
1655 ///
1656 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1657 /// function or closure that determines if a character matches.
1658 ///
1659 /// [`char`]: prim@char
1660 /// [pattern]: self::pattern
1661 ///
1662 /// # Iterator behavior
1663 ///
1664 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1665 /// allows a reverse search and forward/reverse search yields the same
1666 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1667 ///
1668 /// If the pattern allows a reverse search but its results might differ
1669 /// from a forward search, the [`rmatch_indices`] method can be used.
1670 ///
1671 /// [`rmatch_indices`]: str::rmatch_indices
1672 ///
1673 /// # Examples
1674 ///
1675 /// Basic usage:
1676 ///
1677 /// ```
1678 /// let v: Vec<_> = "abcXXXabcYYYabc".match_indices("abc").collect();
1679 /// assert_eq!(v, [(0, "abc"), (6, "abc"), (12, "abc")]);
1680 ///
1681 /// let v: Vec<_> = "1abcabc2".match_indices("abc").collect();
1682 /// assert_eq!(v, [(1, "abc"), (4, "abc")]);
1683 ///
1684 /// let v: Vec<_> = "ababa".match_indices("aba").collect();
1685 /// assert_eq!(v, [(0, "aba")]); // only the first `aba`
1686 /// ```
1687 #[stable(feature = "str_match_indices", since = "1.5.0")]
1688 #[inline]
match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P>1689 pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1690 MatchIndices(MatchIndicesInternal(pat.into_searcher(self)))
1691 }
1692
1693 /// An iterator over the disjoint matches of a pattern within `self`,
1694 /// yielded in reverse order along with the index of the match.
1695 ///
1696 /// For matches of `pat` within `self` that overlap, only the indices
1697 /// corresponding to the last match are returned.
1698 ///
1699 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1700 /// function or closure that determines if a character matches.
1701 ///
1702 /// [`char`]: prim@char
1703 /// [pattern]: self::pattern
1704 ///
1705 /// # Iterator behavior
1706 ///
1707 /// The returned iterator requires that the pattern supports a reverse
1708 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1709 /// search yields the same elements.
1710 ///
1711 /// For iterating from the front, the [`match_indices`] method can be used.
1712 ///
1713 /// [`match_indices`]: str::match_indices
1714 ///
1715 /// # Examples
1716 ///
1717 /// Basic usage:
1718 ///
1719 /// ```
1720 /// let v: Vec<_> = "abcXXXabcYYYabc".rmatch_indices("abc").collect();
1721 /// assert_eq!(v, [(12, "abc"), (6, "abc"), (0, "abc")]);
1722 ///
1723 /// let v: Vec<_> = "1abcabc2".rmatch_indices("abc").collect();
1724 /// assert_eq!(v, [(4, "abc"), (1, "abc")]);
1725 ///
1726 /// let v: Vec<_> = "ababa".rmatch_indices("aba").collect();
1727 /// assert_eq!(v, [(2, "aba")]); // only the last `aba`
1728 /// ```
1729 #[stable(feature = "str_match_indices", since = "1.5.0")]
1730 #[inline]
rmatch_indices<'a, P>(&'a self, pat: P) -> RMatchIndices<'a, P> where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,1731 pub fn rmatch_indices<'a, P>(&'a self, pat: P) -> RMatchIndices<'a, P>
1732 where
1733 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1734 {
1735 RMatchIndices(self.match_indices(pat).0)
1736 }
1737
1738 /// Returns a string slice with leading and trailing whitespace removed.
1739 ///
1740 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1741 /// Core Property `White_Space`.
1742 ///
1743 /// # Examples
1744 ///
1745 /// Basic usage:
1746 ///
1747 /// ```
1748 /// let s = " Hello\tworld\t";
1749 ///
1750 /// assert_eq!("Hello\tworld", s.trim());
1751 /// ```
1752 #[inline]
1753 #[must_use = "this returns the trimmed string as a slice, \
1754 without modifying the original"]
1755 #[stable(feature = "rust1", since = "1.0.0")]
trim(&self) -> &str1756 pub fn trim(&self) -> &str {
1757 self.trim_matches(|c: char| c.is_whitespace())
1758 }
1759
1760 /// Returns a string slice with leading whitespace removed.
1761 ///
1762 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1763 /// Core Property `White_Space`.
1764 ///
1765 /// # Text directionality
1766 ///
1767 /// A string is a sequence of bytes. `start` in this context means the first
1768 /// position of that byte string; for a left-to-right language like English or
1769 /// Russian, this will be left side, and for right-to-left languages like
1770 /// Arabic or Hebrew, this will be the right side.
1771 ///
1772 /// # Examples
1773 ///
1774 /// Basic usage:
1775 ///
1776 /// ```
1777 /// let s = " Hello\tworld\t";
1778 /// assert_eq!("Hello\tworld\t", s.trim_start());
1779 /// ```
1780 ///
1781 /// Directionality:
1782 ///
1783 /// ```
1784 /// let s = " English ";
1785 /// assert!(Some('E') == s.trim_start().chars().next());
1786 ///
1787 /// let s = " עברית ";
1788 /// assert!(Some('ע') == s.trim_start().chars().next());
1789 /// ```
1790 #[inline]
1791 #[must_use = "this returns the trimmed string as a new slice, \
1792 without modifying the original"]
1793 #[stable(feature = "trim_direction", since = "1.30.0")]
trim_start(&self) -> &str1794 pub fn trim_start(&self) -> &str {
1795 self.trim_start_matches(|c: char| c.is_whitespace())
1796 }
1797
1798 /// Returns a string slice with trailing whitespace removed.
1799 ///
1800 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1801 /// Core Property `White_Space`.
1802 ///
1803 /// # Text directionality
1804 ///
1805 /// A string is a sequence of bytes. `end` in this context means the last
1806 /// position of that byte string; for a left-to-right language like English or
1807 /// Russian, this will be right side, and for right-to-left languages like
1808 /// Arabic or Hebrew, this will be the left side.
1809 ///
1810 /// # Examples
1811 ///
1812 /// Basic usage:
1813 ///
1814 /// ```
1815 /// let s = " Hello\tworld\t";
1816 /// assert_eq!(" Hello\tworld", s.trim_end());
1817 /// ```
1818 ///
1819 /// Directionality:
1820 ///
1821 /// ```
1822 /// let s = " English ";
1823 /// assert!(Some('h') == s.trim_end().chars().rev().next());
1824 ///
1825 /// let s = " עברית ";
1826 /// assert!(Some('ת') == s.trim_end().chars().rev().next());
1827 /// ```
1828 #[inline]
1829 #[must_use = "this returns the trimmed string as a new slice, \
1830 without modifying the original"]
1831 #[stable(feature = "trim_direction", since = "1.30.0")]
trim_end(&self) -> &str1832 pub fn trim_end(&self) -> &str {
1833 self.trim_end_matches(|c: char| c.is_whitespace())
1834 }
1835
1836 /// Returns a string slice with leading whitespace removed.
1837 ///
1838 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1839 /// Core Property `White_Space`.
1840 ///
1841 /// # Text directionality
1842 ///
1843 /// A string is a sequence of bytes. 'Left' in this context means the first
1844 /// position of that byte string; for a language like Arabic or Hebrew
1845 /// which are 'right to left' rather than 'left to right', this will be
1846 /// the _right_ side, not the left.
1847 ///
1848 /// # Examples
1849 ///
1850 /// Basic usage:
1851 ///
1852 /// ```
1853 /// let s = " Hello\tworld\t";
1854 ///
1855 /// assert_eq!("Hello\tworld\t", s.trim_left());
1856 /// ```
1857 ///
1858 /// Directionality:
1859 ///
1860 /// ```
1861 /// let s = " English";
1862 /// assert!(Some('E') == s.trim_left().chars().next());
1863 ///
1864 /// let s = " עברית";
1865 /// assert!(Some('ע') == s.trim_left().chars().next());
1866 /// ```
1867 #[must_use = "this returns the trimmed string as a new slice, \
1868 without modifying the original"]
1869 #[inline]
1870 #[stable(feature = "rust1", since = "1.0.0")]
1871 #[rustc_deprecated(
1872 since = "1.33.0",
1873 reason = "superseded by `trim_start`",
1874 suggestion = "trim_start"
1875 )]
trim_left(&self) -> &str1876 pub fn trim_left(&self) -> &str {
1877 self.trim_start()
1878 }
1879
1880 /// Returns a string slice with trailing whitespace removed.
1881 ///
1882 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1883 /// Core Property `White_Space`.
1884 ///
1885 /// # Text directionality
1886 ///
1887 /// A string is a sequence of bytes. 'Right' in this context means the last
1888 /// position of that byte string; for a language like Arabic or Hebrew
1889 /// which are 'right to left' rather than 'left to right', this will be
1890 /// the _left_ side, not the right.
1891 ///
1892 /// # Examples
1893 ///
1894 /// Basic usage:
1895 ///
1896 /// ```
1897 /// let s = " Hello\tworld\t";
1898 ///
1899 /// assert_eq!(" Hello\tworld", s.trim_right());
1900 /// ```
1901 ///
1902 /// Directionality:
1903 ///
1904 /// ```
1905 /// let s = "English ";
1906 /// assert!(Some('h') == s.trim_right().chars().rev().next());
1907 ///
1908 /// let s = "עברית ";
1909 /// assert!(Some('ת') == s.trim_right().chars().rev().next());
1910 /// ```
1911 #[must_use = "this returns the trimmed string as a new slice, \
1912 without modifying the original"]
1913 #[inline]
1914 #[stable(feature = "rust1", since = "1.0.0")]
1915 #[rustc_deprecated(
1916 since = "1.33.0",
1917 reason = "superseded by `trim_end`",
1918 suggestion = "trim_end"
1919 )]
trim_right(&self) -> &str1920 pub fn trim_right(&self) -> &str {
1921 self.trim_end()
1922 }
1923
1924 /// Returns a string slice with all prefixes and suffixes that match a
1925 /// pattern repeatedly removed.
1926 ///
1927 /// The [pattern] can be a [`char`], a slice of [`char`]s, or a function
1928 /// or closure that determines if a character matches.
1929 ///
1930 /// [`char`]: prim@char
1931 /// [pattern]: self::pattern
1932 ///
1933 /// # Examples
1934 ///
1935 /// Simple patterns:
1936 ///
1937 /// ```
1938 /// assert_eq!("11foo1bar11".trim_matches('1'), "foo1bar");
1939 /// assert_eq!("123foo1bar123".trim_matches(char::is_numeric), "foo1bar");
1940 ///
1941 /// let x: &[_] = &['1', '2'];
1942 /// assert_eq!("12foo1bar12".trim_matches(x), "foo1bar");
1943 /// ```
1944 ///
1945 /// A more complex pattern, using a closure:
1946 ///
1947 /// ```
1948 /// assert_eq!("1foo1barXX".trim_matches(|c| c == '1' || c == 'X'), "foo1bar");
1949 /// ```
1950 #[must_use = "this returns the trimmed string as a new slice, \
1951 without modifying the original"]
1952 #[stable(feature = "rust1", since = "1.0.0")]
trim_matches<'a, P>(&'a self, pat: P) -> &'a str where P: Pattern<'a, Searcher: DoubleEndedSearcher<'a>>,1953 pub fn trim_matches<'a, P>(&'a self, pat: P) -> &'a str
1954 where
1955 P: Pattern<'a, Searcher: DoubleEndedSearcher<'a>>,
1956 {
1957 let mut i = 0;
1958 let mut j = 0;
1959 let mut matcher = pat.into_searcher(self);
1960 if let Some((a, b)) = matcher.next_reject() {
1961 i = a;
1962 j = b; // Remember earliest known match, correct it below if
1963 // last match is different
1964 }
1965 if let Some((_, b)) = matcher.next_reject_back() {
1966 j = b;
1967 }
1968 // SAFETY: `Searcher` is known to return valid indices.
1969 unsafe { self.get_unchecked(i..j) }
1970 }
1971
1972 /// Returns a string slice with all prefixes that match a pattern
1973 /// repeatedly removed.
1974 ///
1975 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1976 /// function or closure that determines if a character matches.
1977 ///
1978 /// [`char`]: prim@char
1979 /// [pattern]: self::pattern
1980 ///
1981 /// # Text directionality
1982 ///
1983 /// A string is a sequence of bytes. `start` in this context means the first
1984 /// position of that byte string; for a left-to-right language like English or
1985 /// Russian, this will be left side, and for right-to-left languages like
1986 /// Arabic or Hebrew, this will be the right side.
1987 ///
1988 /// # Examples
1989 ///
1990 /// Basic usage:
1991 ///
1992 /// ```
1993 /// assert_eq!("11foo1bar11".trim_start_matches('1'), "foo1bar11");
1994 /// assert_eq!("123foo1bar123".trim_start_matches(char::is_numeric), "foo1bar123");
1995 ///
1996 /// let x: &[_] = &['1', '2'];
1997 /// assert_eq!("12foo1bar12".trim_start_matches(x), "foo1bar12");
1998 /// ```
1999 #[must_use = "this returns the trimmed string as a new slice, \
2000 without modifying the original"]
2001 #[stable(feature = "trim_direction", since = "1.30.0")]
trim_start_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str2002 pub fn trim_start_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
2003 let mut i = self.len();
2004 let mut matcher = pat.into_searcher(self);
2005 if let Some((a, _)) = matcher.next_reject() {
2006 i = a;
2007 }
2008 // SAFETY: `Searcher` is known to return valid indices.
2009 unsafe { self.get_unchecked(i..self.len()) }
2010 }
2011
2012 /// Returns a string slice with the prefix removed.
2013 ///
2014 /// If the string starts with the pattern `prefix`, returns substring after the prefix, wrapped
2015 /// in `Some`. Unlike `trim_start_matches`, this method removes the prefix exactly once.
2016 ///
2017 /// If the string does not start with `prefix`, returns `None`.
2018 ///
2019 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2020 /// function or closure that determines if a character matches.
2021 ///
2022 /// [`char`]: prim@char
2023 /// [pattern]: self::pattern
2024 ///
2025 /// # Examples
2026 ///
2027 /// ```
2028 /// assert_eq!("foo:bar".strip_prefix("foo:"), Some("bar"));
2029 /// assert_eq!("foo:bar".strip_prefix("bar"), None);
2030 /// assert_eq!("foofoo".strip_prefix("foo"), Some("foo"));
2031 /// ```
2032 #[must_use = "this returns the remaining substring as a new slice, \
2033 without modifying the original"]
2034 #[stable(feature = "str_strip", since = "1.45.0")]
strip_prefix<'a, P: Pattern<'a>>(&'a self, prefix: P) -> Option<&'a str>2035 pub fn strip_prefix<'a, P: Pattern<'a>>(&'a self, prefix: P) -> Option<&'a str> {
2036 prefix.strip_prefix_of(self)
2037 }
2038
2039 /// Returns a string slice with the suffix removed.
2040 ///
2041 /// If the string ends with the pattern `suffix`, returns the substring before the suffix,
2042 /// wrapped in `Some`. Unlike `trim_end_matches`, this method removes the suffix exactly once.
2043 ///
2044 /// If the string does not end with `suffix`, returns `None`.
2045 ///
2046 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2047 /// function or closure that determines if a character matches.
2048 ///
2049 /// [`char`]: prim@char
2050 /// [pattern]: self::pattern
2051 ///
2052 /// # Examples
2053 ///
2054 /// ```
2055 /// assert_eq!("bar:foo".strip_suffix(":foo"), Some("bar"));
2056 /// assert_eq!("bar:foo".strip_suffix("bar"), None);
2057 /// assert_eq!("foofoo".strip_suffix("foo"), Some("foo"));
2058 /// ```
2059 #[must_use = "this returns the remaining substring as a new slice, \
2060 without modifying the original"]
2061 #[stable(feature = "str_strip", since = "1.45.0")]
strip_suffix<'a, P>(&'a self, suffix: P) -> Option<&'a str> where P: Pattern<'a>, <P as Pattern<'a>>::Searcher: ReverseSearcher<'a>,2062 pub fn strip_suffix<'a, P>(&'a self, suffix: P) -> Option<&'a str>
2063 where
2064 P: Pattern<'a>,
2065 <P as Pattern<'a>>::Searcher: ReverseSearcher<'a>,
2066 {
2067 suffix.strip_suffix_of(self)
2068 }
2069
2070 /// Returns a string slice with all suffixes that match a pattern
2071 /// repeatedly removed.
2072 ///
2073 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2074 /// function or closure that determines if a character matches.
2075 ///
2076 /// [`char`]: prim@char
2077 /// [pattern]: self::pattern
2078 ///
2079 /// # Text directionality
2080 ///
2081 /// A string is a sequence of bytes. `end` in this context means the last
2082 /// position of that byte string; for a left-to-right language like English or
2083 /// Russian, this will be right side, and for right-to-left languages like
2084 /// Arabic or Hebrew, this will be the left side.
2085 ///
2086 /// # Examples
2087 ///
2088 /// Simple patterns:
2089 ///
2090 /// ```
2091 /// assert_eq!("11foo1bar11".trim_end_matches('1'), "11foo1bar");
2092 /// assert_eq!("123foo1bar123".trim_end_matches(char::is_numeric), "123foo1bar");
2093 ///
2094 /// let x: &[_] = &['1', '2'];
2095 /// assert_eq!("12foo1bar12".trim_end_matches(x), "12foo1bar");
2096 /// ```
2097 ///
2098 /// A more complex pattern, using a closure:
2099 ///
2100 /// ```
2101 /// assert_eq!("1fooX".trim_end_matches(|c| c == '1' || c == 'X'), "1foo");
2102 /// ```
2103 #[must_use = "this returns the trimmed string as a new slice, \
2104 without modifying the original"]
2105 #[stable(feature = "trim_direction", since = "1.30.0")]
trim_end_matches<'a, P>(&'a self, pat: P) -> &'a str where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,2106 pub fn trim_end_matches<'a, P>(&'a self, pat: P) -> &'a str
2107 where
2108 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
2109 {
2110 let mut j = 0;
2111 let mut matcher = pat.into_searcher(self);
2112 if let Some((_, b)) = matcher.next_reject_back() {
2113 j = b;
2114 }
2115 // SAFETY: `Searcher` is known to return valid indices.
2116 unsafe { self.get_unchecked(0..j) }
2117 }
2118
2119 /// Returns a string slice with all prefixes that match a pattern
2120 /// repeatedly removed.
2121 ///
2122 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2123 /// function or closure that determines if a character matches.
2124 ///
2125 /// [`char`]: prim@char
2126 /// [pattern]: self::pattern
2127 ///
2128 /// # Text directionality
2129 ///
2130 /// A string is a sequence of bytes. 'Left' in this context means the first
2131 /// position of that byte string; for a language like Arabic or Hebrew
2132 /// which are 'right to left' rather than 'left to right', this will be
2133 /// the _right_ side, not the left.
2134 ///
2135 /// # Examples
2136 ///
2137 /// Basic usage:
2138 ///
2139 /// ```
2140 /// assert_eq!("11foo1bar11".trim_left_matches('1'), "foo1bar11");
2141 /// assert_eq!("123foo1bar123".trim_left_matches(char::is_numeric), "foo1bar123");
2142 ///
2143 /// let x: &[_] = &['1', '2'];
2144 /// assert_eq!("12foo1bar12".trim_left_matches(x), "foo1bar12");
2145 /// ```
2146 #[stable(feature = "rust1", since = "1.0.0")]
2147 #[rustc_deprecated(
2148 since = "1.33.0",
2149 reason = "superseded by `trim_start_matches`",
2150 suggestion = "trim_start_matches"
2151 )]
trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str2152 pub fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
2153 self.trim_start_matches(pat)
2154 }
2155
2156 /// Returns a string slice with all suffixes that match a pattern
2157 /// repeatedly removed.
2158 ///
2159 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2160 /// function or closure that determines if a character matches.
2161 ///
2162 /// [`char`]: prim@char
2163 /// [pattern]: self::pattern
2164 ///
2165 /// # Text directionality
2166 ///
2167 /// A string is a sequence of bytes. 'Right' in this context means the last
2168 /// position of that byte string; for a language like Arabic or Hebrew
2169 /// which are 'right to left' rather than 'left to right', this will be
2170 /// the _left_ side, not the right.
2171 ///
2172 /// # Examples
2173 ///
2174 /// Simple patterns:
2175 ///
2176 /// ```
2177 /// assert_eq!("11foo1bar11".trim_right_matches('1'), "11foo1bar");
2178 /// assert_eq!("123foo1bar123".trim_right_matches(char::is_numeric), "123foo1bar");
2179 ///
2180 /// let x: &[_] = &['1', '2'];
2181 /// assert_eq!("12foo1bar12".trim_right_matches(x), "12foo1bar");
2182 /// ```
2183 ///
2184 /// A more complex pattern, using a closure:
2185 ///
2186 /// ```
2187 /// assert_eq!("1fooX".trim_right_matches(|c| c == '1' || c == 'X'), "1foo");
2188 /// ```
2189 #[stable(feature = "rust1", since = "1.0.0")]
2190 #[rustc_deprecated(
2191 since = "1.33.0",
2192 reason = "superseded by `trim_end_matches`",
2193 suggestion = "trim_end_matches"
2194 )]
trim_right_matches<'a, P>(&'a self, pat: P) -> &'a str where P: Pattern<'a, Searcher: ReverseSearcher<'a>>,2195 pub fn trim_right_matches<'a, P>(&'a self, pat: P) -> &'a str
2196 where
2197 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
2198 {
2199 self.trim_end_matches(pat)
2200 }
2201
2202 /// Parses this string slice into another type.
2203 ///
2204 /// Because `parse` is so general, it can cause problems with type
2205 /// inference. As such, `parse` is one of the few times you'll see
2206 /// the syntax affectionately known as the 'turbofish': `::<>`. This
2207 /// helps the inference algorithm understand specifically which type
2208 /// you're trying to parse into.
2209 ///
2210 /// `parse` can parse into any type that implements the [`FromStr`] trait.
2211
2212 ///
2213 /// # Errors
2214 ///
2215 /// Will return [`Err`] if it's not possible to parse this string slice into
2216 /// the desired type.
2217 ///
2218 /// [`Err`]: FromStr::Err
2219 ///
2220 /// # Examples
2221 ///
2222 /// Basic usage
2223 ///
2224 /// ```
2225 /// let four: u32 = "4".parse().unwrap();
2226 ///
2227 /// assert_eq!(4, four);
2228 /// ```
2229 ///
2230 /// Using the 'turbofish' instead of annotating `four`:
2231 ///
2232 /// ```
2233 /// let four = "4".parse::<u32>();
2234 ///
2235 /// assert_eq!(Ok(4), four);
2236 /// ```
2237 ///
2238 /// Failing to parse:
2239 ///
2240 /// ```
2241 /// let nope = "j".parse::<u32>();
2242 ///
2243 /// assert!(nope.is_err());
2244 /// ```
2245 #[inline]
2246 #[stable(feature = "rust1", since = "1.0.0")]
parse<F: FromStr>(&self) -> Result<F, F::Err>2247 pub fn parse<F: FromStr>(&self) -> Result<F, F::Err> {
2248 FromStr::from_str(self)
2249 }
2250
2251 /// Checks if all characters in this string are within the ASCII range.
2252 ///
2253 /// # Examples
2254 ///
2255 /// ```
2256 /// let ascii = "hello!\n";
2257 /// let non_ascii = "Grüße, Jürgen ❤";
2258 ///
2259 /// assert!(ascii.is_ascii());
2260 /// assert!(!non_ascii.is_ascii());
2261 /// ```
2262 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2263 #[must_use]
2264 #[inline]
is_ascii(&self) -> bool2265 pub fn is_ascii(&self) -> bool {
2266 // We can treat each byte as character here: all multibyte characters
2267 // start with a byte that is not in the ascii range, so we will stop
2268 // there already.
2269 self.as_bytes().is_ascii()
2270 }
2271
2272 /// Checks that two strings are an ASCII case-insensitive match.
2273 ///
2274 /// Same as `to_ascii_lowercase(a) == to_ascii_lowercase(b)`,
2275 /// but without allocating and copying temporaries.
2276 ///
2277 /// # Examples
2278 ///
2279 /// ```
2280 /// assert!("Ferris".eq_ignore_ascii_case("FERRIS"));
2281 /// assert!("Ferrös".eq_ignore_ascii_case("FERRöS"));
2282 /// assert!(!"Ferrös".eq_ignore_ascii_case("FERRÖS"));
2283 /// ```
2284 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2285 #[must_use]
2286 #[inline]
eq_ignore_ascii_case(&self, other: &str) -> bool2287 pub fn eq_ignore_ascii_case(&self, other: &str) -> bool {
2288 self.as_bytes().eq_ignore_ascii_case(other.as_bytes())
2289 }
2290
2291 /// Converts this string to its ASCII upper case equivalent in-place.
2292 ///
2293 /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z',
2294 /// but non-ASCII letters are unchanged.
2295 ///
2296 /// To return a new uppercased value without modifying the existing one, use
2297 /// [`to_ascii_uppercase()`].
2298 ///
2299 /// [`to_ascii_uppercase()`]: #method.to_ascii_uppercase
2300 ///
2301 /// # Examples
2302 ///
2303 /// ```
2304 /// let mut s = String::from("Grüße, Jürgen ❤");
2305 ///
2306 /// s.make_ascii_uppercase();
2307 ///
2308 /// assert_eq!("GRüßE, JüRGEN ❤", s);
2309 /// ```
2310 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2311 #[inline]
make_ascii_uppercase(&mut self)2312 pub fn make_ascii_uppercase(&mut self) {
2313 // SAFETY: safe because we transmute two types with the same layout.
2314 let me = unsafe { self.as_bytes_mut() };
2315 me.make_ascii_uppercase()
2316 }
2317
2318 /// Converts this string to its ASCII lower case equivalent in-place.
2319 ///
2320 /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z',
2321 /// but non-ASCII letters are unchanged.
2322 ///
2323 /// To return a new lowercased value without modifying the existing one, use
2324 /// [`to_ascii_lowercase()`].
2325 ///
2326 /// [`to_ascii_lowercase()`]: #method.to_ascii_lowercase
2327 ///
2328 /// # Examples
2329 ///
2330 /// ```
2331 /// let mut s = String::from("GRÜßE, JÜRGEN ❤");
2332 ///
2333 /// s.make_ascii_lowercase();
2334 ///
2335 /// assert_eq!("grÜße, jÜrgen ❤", s);
2336 /// ```
2337 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2338 #[inline]
make_ascii_lowercase(&mut self)2339 pub fn make_ascii_lowercase(&mut self) {
2340 // SAFETY: safe because we transmute two types with the same layout.
2341 let me = unsafe { self.as_bytes_mut() };
2342 me.make_ascii_lowercase()
2343 }
2344
2345 /// Return an iterator that escapes each char in `self` with [`char::escape_debug`].
2346 ///
2347 /// Note: only extended grapheme codepoints that begin the string will be
2348 /// escaped.
2349 ///
2350 /// # Examples
2351 ///
2352 /// As an iterator:
2353 ///
2354 /// ```
2355 /// for c in "❤\n!".escape_debug() {
2356 /// print!("{}", c);
2357 /// }
2358 /// println!();
2359 /// ```
2360 ///
2361 /// Using `println!` directly:
2362 ///
2363 /// ```
2364 /// println!("{}", "❤\n!".escape_debug());
2365 /// ```
2366 ///
2367 ///
2368 /// Both are equivalent to:
2369 ///
2370 /// ```
2371 /// println!("❤\\n!");
2372 /// ```
2373 ///
2374 /// Using `to_string`:
2375 ///
2376 /// ```
2377 /// assert_eq!("❤\n!".escape_debug().to_string(), "❤\\n!");
2378 /// ```
2379 #[must_use = "this returns the escaped string as an iterator, \
2380 without modifying the original"]
2381 #[stable(feature = "str_escape", since = "1.34.0")]
escape_debug(&self) -> EscapeDebug<'_>2382 pub fn escape_debug(&self) -> EscapeDebug<'_> {
2383 let mut chars = self.chars();
2384 EscapeDebug {
2385 inner: chars
2386 .next()
2387 .map(|first| first.escape_debug_ext(EscapeDebugExtArgs::ESCAPE_ALL))
2388 .into_iter()
2389 .flatten()
2390 .chain(chars.flat_map(CharEscapeDebugContinue)),
2391 }
2392 }
2393
2394 /// Return an iterator that escapes each char in `self` with [`char::escape_default`].
2395 ///
2396 /// # Examples
2397 ///
2398 /// As an iterator:
2399 ///
2400 /// ```
2401 /// for c in "❤\n!".escape_default() {
2402 /// print!("{}", c);
2403 /// }
2404 /// println!();
2405 /// ```
2406 ///
2407 /// Using `println!` directly:
2408 ///
2409 /// ```
2410 /// println!("{}", "❤\n!".escape_default());
2411 /// ```
2412 ///
2413 ///
2414 /// Both are equivalent to:
2415 ///
2416 /// ```
2417 /// println!("\\u{{2764}}\\n!");
2418 /// ```
2419 ///
2420 /// Using `to_string`:
2421 ///
2422 /// ```
2423 /// assert_eq!("❤\n!".escape_default().to_string(), "\\u{2764}\\n!");
2424 /// ```
2425 #[must_use = "this returns the escaped string as an iterator, \
2426 without modifying the original"]
2427 #[stable(feature = "str_escape", since = "1.34.0")]
escape_default(&self) -> EscapeDefault<'_>2428 pub fn escape_default(&self) -> EscapeDefault<'_> {
2429 EscapeDefault { inner: self.chars().flat_map(CharEscapeDefault) }
2430 }
2431
2432 /// Return an iterator that escapes each char in `self` with [`char::escape_unicode`].
2433 ///
2434 /// # Examples
2435 ///
2436 /// As an iterator:
2437 ///
2438 /// ```
2439 /// for c in "❤\n!".escape_unicode() {
2440 /// print!("{}", c);
2441 /// }
2442 /// println!();
2443 /// ```
2444 ///
2445 /// Using `println!` directly:
2446 ///
2447 /// ```
2448 /// println!("{}", "❤\n!".escape_unicode());
2449 /// ```
2450 ///
2451 ///
2452 /// Both are equivalent to:
2453 ///
2454 /// ```
2455 /// println!("\\u{{2764}}\\u{{a}}\\u{{21}}");
2456 /// ```
2457 ///
2458 /// Using `to_string`:
2459 ///
2460 /// ```
2461 /// assert_eq!("❤\n!".escape_unicode().to_string(), "\\u{2764}\\u{a}\\u{21}");
2462 /// ```
2463 #[must_use = "this returns the escaped string as an iterator, \
2464 without modifying the original"]
2465 #[stable(feature = "str_escape", since = "1.34.0")]
escape_unicode(&self) -> EscapeUnicode<'_>2466 pub fn escape_unicode(&self) -> EscapeUnicode<'_> {
2467 EscapeUnicode { inner: self.chars().flat_map(CharEscapeUnicode) }
2468 }
2469 }
2470
2471 #[stable(feature = "rust1", since = "1.0.0")]
2472 impl AsRef<[u8]> for str {
2473 #[inline]
as_ref(&self) -> &[u8]2474 fn as_ref(&self) -> &[u8] {
2475 self.as_bytes()
2476 }
2477 }
2478
2479 #[stable(feature = "rust1", since = "1.0.0")]
2480 #[rustc_const_unstable(feature = "const_default_impls", issue = "87864")]
2481 impl const Default for &str {
2482 /// Creates an empty str
2483 #[inline]
default() -> Self2484 fn default() -> Self {
2485 ""
2486 }
2487 }
2488
2489 #[stable(feature = "default_mut_str", since = "1.28.0")]
2490 impl Default for &mut str {
2491 /// Creates an empty mutable str
2492 #[inline]
default() -> Self2493 fn default() -> Self {
2494 // SAFETY: The empty string is valid UTF-8.
2495 unsafe { from_utf8_unchecked_mut(&mut []) }
2496 }
2497 }
2498
2499 impl_fn_for_zst! {
2500 /// A nameable, cloneable fn type
2501 #[derive(Clone)]
2502 struct LinesAnyMap impl<'a> Fn = |line: &'a str| -> &'a str {
2503 let l = line.len();
2504 if l > 0 && line.as_bytes()[l - 1] == b'\r' { &line[0 .. l - 1] }
2505 else { line }
2506 };
2507
2508 #[derive(Clone)]
2509 struct CharEscapeDebugContinue impl Fn = |c: char| -> char::EscapeDebug {
2510 c.escape_debug_ext(EscapeDebugExtArgs {
2511 escape_grapheme_extended: false,
2512 escape_single_quote: true,
2513 escape_double_quote: true
2514 })
2515 };
2516
2517 #[derive(Clone)]
2518 struct CharEscapeUnicode impl Fn = |c: char| -> char::EscapeUnicode {
2519 c.escape_unicode()
2520 };
2521 #[derive(Clone)]
2522 struct CharEscapeDefault impl Fn = |c: char| -> char::EscapeDefault {
2523 c.escape_default()
2524 };
2525
2526 #[derive(Clone)]
2527 struct IsWhitespace impl Fn = |c: char| -> bool {
2528 c.is_whitespace()
2529 };
2530
2531 #[derive(Clone)]
2532 struct IsAsciiWhitespace impl Fn = |byte: &u8| -> bool {
2533 byte.is_ascii_whitespace()
2534 };
2535
2536 #[derive(Clone)]
2537 struct IsNotEmpty impl<'a, 'b> Fn = |s: &'a &'b str| -> bool {
2538 !s.is_empty()
2539 };
2540
2541 #[derive(Clone)]
2542 struct BytesIsNotEmpty impl<'a, 'b> Fn = |s: &'a &'b [u8]| -> bool {
2543 !s.is_empty()
2544 };
2545
2546 #[derive(Clone)]
2547 struct UnsafeBytesToStr impl<'a> Fn = |bytes: &'a [u8]| -> &'a str {
2548 // SAFETY: not safe
2549 unsafe { from_utf8_unchecked(bytes) }
2550 };
2551 }
2552