1 //! String manipulation.
3 //! For more details, see the [`std::str`] module.
5 //! [`std::str`]: ../../std/str/index.html
7 #![stable(feature = "rust1", since = "1.0.0")]
15 use self::pattern::Pattern;
16 use self::pattern::{DoubleEndedSearcher, ReverseSearcher, Searcher};
20 use crate::slice::{self, SliceIndex};
24 #[unstable(feature = "str_internals", issue = "none")]
25 #[allow(missing_docs)]
28 #[stable(feature = "rust1", since = "1.0.0")]
29 pub use converts::{from_utf8, from_utf8_unchecked};
31 #[stable(feature = "str_mut_extras", since = "1.20.0")]
32 pub use converts::{from_utf8_mut, from_utf8_unchecked_mut};
34 #[stable(feature = "rust1", since = "1.0.0")]
35 pub use error::{ParseBoolError, Utf8Error};
37 #[stable(feature = "rust1", since = "1.0.0")]
38 pub use traits::FromStr;
40 #[stable(feature = "rust1", since = "1.0.0")]
41 pub use iter::{Bytes, CharIndices, Chars, Lines, SplitWhitespace};
43 #[stable(feature = "rust1", since = "1.0.0")]
45 pub use iter::LinesAny;
47 #[stable(feature = "rust1", since = "1.0.0")]
48 pub use iter::{RSplit, RSplitTerminator, Split, SplitTerminator};
50 #[stable(feature = "rust1", since = "1.0.0")]
51 pub use iter::{RSplitN, SplitN};
53 #[stable(feature = "str_matches", since = "1.2.0")]
54 pub use iter::{Matches, RMatches};
56 #[stable(feature = "str_match_indices", since = "1.5.0")]
57 pub use iter::{MatchIndices, RMatchIndices};
59 #[stable(feature = "encode_utf16", since = "1.8.0")]
60 pub use iter::EncodeUtf16;
62 #[stable(feature = "str_escape", since = "1.34.0")]
63 pub use iter::{EscapeDebug, EscapeDefault, EscapeUnicode};
65 #[stable(feature = "split_ascii_whitespace", since = "1.34.0")]
66 pub use iter::SplitAsciiWhitespace;
68 #[unstable(feature = "split_inclusive", issue = "72360")]
69 use iter::SplitInclusive;
71 #[unstable(feature = "str_internals", issue = "none")]
72 pub use validations::next_code_point;
74 use iter::MatchIndicesInternal;
75 use iter::SplitInternal;
76 use iter::{MatchesInternal, SplitNInternal};
78 use validations::truncate_to_char_boundary;
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 { "" };
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);
97 "begin <= end ({} <= {}) when slicing `{}`{}",
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) {
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();
115 "byte index {} is not a char boundary; it is inside {:?} (bytes {:?}) of `{}`{}",
116 index, ch, char_range, s_trunc, ellipsis
123 /// Returns the length of `self`.
125 /// This length is in bytes, not [`char`]s or graphemes. In other words,
126 /// it may not be what a human considers the length of the string.
128 /// [`char`]: prim@char
135 /// let len = "foo".len();
136 /// assert_eq!(3, len);
138 /// assert_eq!("ƒoo".len(), 4); // fancy f!
139 /// assert_eq!("ƒoo".chars().count(), 3);
141 #[doc(alias = "length")]
142 #[stable(feature = "rust1", since = "1.0.0")]
143 #[rustc_const_stable(feature = "const_str_len", since = "1.32.0")]
145 pub const fn len(&self) -> usize {
146 self.as_bytes().len()
149 /// Returns `true` if `self` has a length of zero bytes.
157 /// assert!(s.is_empty());
159 /// let s = "not empty";
160 /// assert!(!s.is_empty());
163 #[stable(feature = "rust1", since = "1.0.0")]
164 #[rustc_const_stable(feature = "const_str_is_empty", since = "1.32.0")]
165 pub const fn is_empty(&self) -> bool {
169 /// Checks that `index`-th byte is the first byte in a UTF-8 code point
170 /// sequence or the end of the string.
172 /// The start and end of the string (when `index == self.len()`) are
173 /// considered to be boundaries.
175 /// Returns `false` if `index` is greater than `self.len()`.
180 /// let s = "Löwe 老虎 Léopard";
181 /// assert!(s.is_char_boundary(0));
183 /// assert!(s.is_char_boundary(6));
184 /// assert!(s.is_char_boundary(s.len()));
186 /// // second byte of `ö`
187 /// assert!(!s.is_char_boundary(2));
189 /// // third byte of `老`
190 /// assert!(!s.is_char_boundary(8));
192 #[stable(feature = "is_char_boundary", since = "1.9.0")]
194 pub fn is_char_boundary(&self, index: usize) -> bool {
195 // 0 and len are always ok.
196 // Test for 0 explicitly so that it can optimize out the check
197 // easily and skip reading string data for that case.
198 if index == 0 || index == self.len() {
201 match self.as_bytes().get(index) {
203 // This is bit magic equivalent to: b < 128 || b >= 192
204 Some(&b) => (b as i8) >= -0x40,
208 /// Converts a string slice to a byte slice. To convert the byte slice back
209 /// into a string slice, use the [`from_utf8`] function.
216 /// let bytes = "bors".as_bytes();
217 /// assert_eq!(b"bors", bytes);
219 #[stable(feature = "rust1", since = "1.0.0")]
220 #[rustc_const_stable(feature = "str_as_bytes", since = "1.32.0")]
222 #[allow(unused_attributes)]
223 #[rustc_allow_const_fn_unstable(const_fn_transmute)]
224 pub const fn as_bytes(&self) -> &[u8] {
225 // SAFETY: const sound because we transmute two types with the same layout
226 unsafe { mem::transmute(self) }
229 /// Converts a mutable string slice to a mutable byte slice.
233 /// The caller must ensure that the content of the slice is valid UTF-8
234 /// before the borrow ends and the underlying `str` is used.
236 /// Use of a `str` whose contents are not valid UTF-8 is undefined behavior.
243 /// let mut s = String::from("Hello");
244 /// let bytes = unsafe { s.as_bytes_mut() };
246 /// assert_eq!(b"Hello", bytes);
252 /// let mut s = String::from("🗻∈🌏");
255 /// let bytes = s.as_bytes_mut();
263 /// assert_eq!("🍔∈🌏", s);
265 #[stable(feature = "str_mut_extras", since = "1.20.0")]
267 pub unsafe fn as_bytes_mut(&mut self) -> &mut [u8] {
268 // SAFETY: the cast from `&str` to `&[u8]` is safe since `str`
269 // has the same layout as `&[u8]` (only libstd can make this guarantee).
270 // The pointer dereference is safe since it comes from a mutable reference which
271 // is guaranteed to be valid for writes.
272 unsafe { &mut *(self as *mut str as *mut [u8]) }
275 /// Converts a string slice to a raw pointer.
277 /// As string slices are a slice of bytes, the raw pointer points to a
278 /// [`u8`]. This pointer will be pointing to the first byte of the string
281 /// The caller must ensure that the returned pointer is never written to.
282 /// If you need to mutate the contents of the string slice, use [`as_mut_ptr`].
284 /// [`as_mut_ptr`]: str::as_mut_ptr
292 /// let ptr = s.as_ptr();
294 #[stable(feature = "rust1", since = "1.0.0")]
295 #[rustc_const_stable(feature = "rustc_str_as_ptr", since = "1.32.0")]
297 pub const fn as_ptr(&self) -> *const u8 {
298 self as *const str as *const u8
301 /// Converts a mutable string slice to a raw pointer.
303 /// As string slices are a slice of bytes, the raw pointer points to a
304 /// [`u8`]. This pointer will be pointing to the first byte of the string
307 /// It is your responsibility to make sure that the string slice only gets
308 /// modified in a way that it remains valid UTF-8.
309 #[stable(feature = "str_as_mut_ptr", since = "1.36.0")]
311 pub fn as_mut_ptr(&mut self) -> *mut u8 {
312 self as *mut str as *mut u8
315 /// Returns a subslice of `str`.
317 /// This is the non-panicking alternative to indexing the `str`. Returns
318 /// [`None`] whenever equivalent indexing operation would panic.
323 /// let v = String::from("🗻∈🌏");
325 /// assert_eq!(Some("🗻"), v.get(0..4));
327 /// // indices not on UTF-8 sequence boundaries
328 /// assert!(v.get(1..).is_none());
329 /// assert!(v.get(..8).is_none());
332 /// assert!(v.get(..42).is_none());
334 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
336 pub fn get<I: SliceIndex<str>>(&self, i: I) -> Option<&I::Output> {
340 /// Returns a mutable subslice of `str`.
342 /// This is the non-panicking alternative to indexing the `str`. Returns
343 /// [`None`] whenever equivalent indexing operation would panic.
348 /// let mut v = String::from("hello");
349 /// // correct length
350 /// assert!(v.get_mut(0..5).is_some());
352 /// assert!(v.get_mut(..42).is_none());
353 /// assert_eq!(Some("he"), v.get_mut(0..2).map(|v| &*v));
355 /// assert_eq!("hello", v);
357 /// let s = v.get_mut(0..2);
358 /// let s = s.map(|s| {
359 /// s.make_ascii_uppercase();
362 /// assert_eq!(Some("HE"), s);
364 /// assert_eq!("HEllo", v);
366 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
368 pub fn get_mut<I: SliceIndex<str>>(&mut self, i: I) -> Option<&mut I::Output> {
372 /// Returns an unchecked subslice of `str`.
374 /// This is the unchecked alternative to indexing the `str`.
378 /// Callers of this function are responsible that these preconditions are
381 /// * The starting index must not exceed the ending index;
382 /// * Indexes must be within bounds of the original slice;
383 /// * Indexes must lie on UTF-8 sequence boundaries.
385 /// Failing that, the returned string slice may reference invalid memory or
386 /// violate the invariants communicated by the `str` type.
393 /// assert_eq!("🗻", v.get_unchecked(0..4));
394 /// assert_eq!("∈", v.get_unchecked(4..7));
395 /// assert_eq!("🌏", v.get_unchecked(7..11));
398 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
400 pub unsafe fn get_unchecked<I: SliceIndex<str>>(&self, i: I) -> &I::Output {
401 // SAFETY: the caller must uphold the safety contract for `get_unchecked`;
402 // the slice is dereferencable because `self` is a safe reference.
403 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
404 unsafe { &*i.get_unchecked(self) }
407 /// Returns a mutable, unchecked subslice of `str`.
409 /// This is the unchecked alternative to indexing the `str`.
413 /// Callers of this function are responsible that these preconditions are
416 /// * The starting index must not exceed the ending index;
417 /// * Indexes must be within bounds of the original slice;
418 /// * Indexes must lie on UTF-8 sequence boundaries.
420 /// Failing that, the returned string slice may reference invalid memory or
421 /// violate the invariants communicated by the `str` type.
426 /// let mut v = String::from("🗻∈🌏");
428 /// assert_eq!("🗻", v.get_unchecked_mut(0..4));
429 /// assert_eq!("∈", v.get_unchecked_mut(4..7));
430 /// assert_eq!("🌏", v.get_unchecked_mut(7..11));
433 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
435 pub unsafe fn get_unchecked_mut<I: SliceIndex<str>>(&mut self, i: I) -> &mut I::Output {
436 // SAFETY: the caller must uphold the safety contract for `get_unchecked_mut`;
437 // the slice is dereferencable because `self` is a safe reference.
438 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
439 unsafe { &mut *i.get_unchecked_mut(self) }
442 /// Creates a string slice from another string slice, bypassing safety
445 /// This is generally not recommended, use with caution! For a safe
446 /// alternative see [`str`] and [`Index`].
448 /// [`Index`]: crate::ops::Index
450 /// This new slice goes from `begin` to `end`, including `begin` but
453 /// To get a mutable string slice instead, see the
454 /// [`slice_mut_unchecked`] method.
456 /// [`slice_mut_unchecked`]: str::slice_mut_unchecked
460 /// Callers of this function are responsible that three preconditions are
463 /// * `begin` must not exceed `end`.
464 /// * `begin` and `end` must be byte positions within the string slice.
465 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
472 /// let s = "Löwe 老虎 Léopard";
475 /// assert_eq!("Löwe 老虎 Léopard", s.slice_unchecked(0, 21));
478 /// let s = "Hello, world!";
481 /// assert_eq!("world", s.slice_unchecked(7, 12));
484 #[stable(feature = "rust1", since = "1.0.0")]
485 #[rustc_deprecated(since = "1.29.0", reason = "use `get_unchecked(begin..end)` instead")]
487 pub unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
488 // SAFETY: the caller must uphold the safety contract for `get_unchecked`;
489 // the slice is dereferencable because `self` is a safe reference.
490 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
491 unsafe { &*(begin..end).get_unchecked(self) }
494 /// Creates a string slice from another string slice, bypassing safety
496 /// This is generally not recommended, use with caution! For a safe
497 /// alternative see [`str`] and [`IndexMut`].
499 /// [`IndexMut`]: crate::ops::IndexMut
501 /// This new slice goes from `begin` to `end`, including `begin` but
504 /// To get an immutable string slice instead, see the
505 /// [`slice_unchecked`] method.
507 /// [`slice_unchecked`]: str::slice_unchecked
511 /// Callers of this function are responsible that three preconditions are
514 /// * `begin` must not exceed `end`.
515 /// * `begin` and `end` must be byte positions within the string slice.
516 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
517 #[stable(feature = "str_slice_mut", since = "1.5.0")]
518 #[rustc_deprecated(since = "1.29.0", reason = "use `get_unchecked_mut(begin..end)` instead")]
520 pub unsafe fn slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str {
521 // SAFETY: the caller must uphold the safety contract for `get_unchecked_mut`;
522 // the slice is dereferencable because `self` is a safe reference.
523 // The returned pointer is safe because impls of `SliceIndex` have to guarantee that it is.
524 unsafe { &mut *(begin..end).get_unchecked_mut(self) }
527 /// Divide one string slice into two at an index.
529 /// The argument, `mid`, should be a byte offset from the start of the
530 /// string. It must also be on the boundary of a UTF-8 code point.
532 /// The two slices returned go from the start of the string slice to `mid`,
533 /// and from `mid` to the end of the string slice.
535 /// To get mutable string slices instead, see the [`split_at_mut`]
538 /// [`split_at_mut`]: str::split_at_mut
542 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
543 /// past the end of the last code point of the string slice.
550 /// let s = "Per Martin-Löf";
552 /// let (first, last) = s.split_at(3);
554 /// assert_eq!("Per", first);
555 /// assert_eq!(" Martin-Löf", last);
558 #[stable(feature = "str_split_at", since = "1.4.0")]
559 pub fn split_at(&self, mid: usize) -> (&str, &str) {
560 // is_char_boundary checks that the index is in [0, .len()]
561 if self.is_char_boundary(mid) {
562 // SAFETY: just checked that `mid` is on a char boundary.
563 unsafe { (self.get_unchecked(0..mid), self.get_unchecked(mid..self.len())) }
565 slice_error_fail(self, 0, mid)
569 /// Divide one mutable string slice into two at an index.
571 /// The argument, `mid`, should be a byte offset from the start of the
572 /// string. It must also be on the boundary of a UTF-8 code point.
574 /// The two slices returned go from the start of the string slice to `mid`,
575 /// and from `mid` to the end of the string slice.
577 /// To get immutable string slices instead, see the [`split_at`] method.
579 /// [`split_at`]: str::split_at
583 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
584 /// past the end of the last code point of the string slice.
591 /// let mut s = "Per Martin-Löf".to_string();
593 /// let (first, last) = s.split_at_mut(3);
594 /// first.make_ascii_uppercase();
595 /// assert_eq!("PER", first);
596 /// assert_eq!(" Martin-Löf", last);
598 /// assert_eq!("PER Martin-Löf", s);
601 #[stable(feature = "str_split_at", since = "1.4.0")]
602 pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {
603 // is_char_boundary checks that the index is in [0, .len()]
604 if self.is_char_boundary(mid) {
605 let len = self.len();
606 let ptr = self.as_mut_ptr();
607 // SAFETY: just checked that `mid` is on a char boundary.
610 from_utf8_unchecked_mut(slice::from_raw_parts_mut(ptr, mid)),
611 from_utf8_unchecked_mut(slice::from_raw_parts_mut(ptr.add(mid), len - mid)),
615 slice_error_fail(self, 0, mid)
619 /// Returns an iterator over the [`char`]s of a string slice.
621 /// As a string slice consists of valid UTF-8, we can iterate through a
622 /// string slice by [`char`]. This method returns such an iterator.
624 /// It's important to remember that [`char`] represents a Unicode Scalar
625 /// Value, and may not match your idea of what a 'character' is. Iteration
626 /// over grapheme clusters may be what you actually want. This functionality
627 /// is not provided by Rust's standard library, check crates.io instead.
634 /// let word = "goodbye";
636 /// let count = word.chars().count();
637 /// assert_eq!(7, count);
639 /// let mut chars = word.chars();
641 /// assert_eq!(Some('g'), chars.next());
642 /// assert_eq!(Some('o'), chars.next());
643 /// assert_eq!(Some('o'), chars.next());
644 /// assert_eq!(Some('d'), chars.next());
645 /// assert_eq!(Some('b'), chars.next());
646 /// assert_eq!(Some('y'), chars.next());
647 /// assert_eq!(Some('e'), chars.next());
649 /// assert_eq!(None, chars.next());
652 /// Remember, [`char`]s may not match your intuition about characters:
654 /// [`char`]: prim@char
659 /// let mut chars = y.chars();
661 /// assert_eq!(Some('y'), chars.next()); // not 'y̆'
662 /// assert_eq!(Some('\u{0306}'), chars.next());
664 /// assert_eq!(None, chars.next());
666 #[stable(feature = "rust1", since = "1.0.0")]
668 pub fn chars(&self) -> Chars<'_> {
669 Chars { iter: self.as_bytes().iter() }
672 /// Returns an iterator over the [`char`]s of a string slice, and their
675 /// As a string slice consists of valid UTF-8, we can iterate through a
676 /// string slice by [`char`]. This method returns an iterator of both
677 /// these [`char`]s, as well as their byte positions.
679 /// The iterator yields tuples. The position is first, the [`char`] is
687 /// let word = "goodbye";
689 /// let count = word.char_indices().count();
690 /// assert_eq!(7, count);
692 /// let mut char_indices = word.char_indices();
694 /// assert_eq!(Some((0, 'g')), char_indices.next());
695 /// assert_eq!(Some((1, 'o')), char_indices.next());
696 /// assert_eq!(Some((2, 'o')), char_indices.next());
697 /// assert_eq!(Some((3, 'd')), char_indices.next());
698 /// assert_eq!(Some((4, 'b')), char_indices.next());
699 /// assert_eq!(Some((5, 'y')), char_indices.next());
700 /// assert_eq!(Some((6, 'e')), char_indices.next());
702 /// assert_eq!(None, char_indices.next());
705 /// Remember, [`char`]s may not match your intuition about characters:
707 /// [`char`]: prim@char
710 /// let yes = "y̆es";
712 /// let mut char_indices = yes.char_indices();
714 /// assert_eq!(Some((0, 'y')), char_indices.next()); // not (0, 'y̆')
715 /// assert_eq!(Some((1, '\u{0306}')), char_indices.next());
717 /// // note the 3 here - the last character took up two bytes
718 /// assert_eq!(Some((3, 'e')), char_indices.next());
719 /// assert_eq!(Some((4, 's')), char_indices.next());
721 /// assert_eq!(None, char_indices.next());
723 #[stable(feature = "rust1", since = "1.0.0")]
725 pub fn char_indices(&self) -> CharIndices<'_> {
726 CharIndices { front_offset: 0, iter: self.chars() }
729 /// An iterator over the bytes of a string slice.
731 /// As a string slice consists of a sequence of bytes, we can iterate
732 /// through a string slice by byte. This method returns such an iterator.
739 /// let mut bytes = "bors".bytes();
741 /// assert_eq!(Some(b'b'), bytes.next());
742 /// assert_eq!(Some(b'o'), bytes.next());
743 /// assert_eq!(Some(b'r'), bytes.next());
744 /// assert_eq!(Some(b's'), bytes.next());
746 /// assert_eq!(None, bytes.next());
748 #[stable(feature = "rust1", since = "1.0.0")]
750 pub fn bytes(&self) -> Bytes<'_> {
751 Bytes(self.as_bytes().iter().copied())
754 /// Splits a string slice by whitespace.
756 /// The iterator returned will return string slices that are sub-slices of
757 /// the original string slice, separated by any amount of whitespace.
759 /// 'Whitespace' is defined according to the terms of the Unicode Derived
760 /// Core Property `White_Space`. If you only want to split on ASCII whitespace
761 /// instead, use [`split_ascii_whitespace`].
763 /// [`split_ascii_whitespace`]: str::split_ascii_whitespace
770 /// let mut iter = "A few words".split_whitespace();
772 /// assert_eq!(Some("A"), iter.next());
773 /// assert_eq!(Some("few"), iter.next());
774 /// assert_eq!(Some("words"), iter.next());
776 /// assert_eq!(None, iter.next());
779 /// All kinds of whitespace are considered:
782 /// let mut iter = " Mary had\ta\u{2009}little \n\t lamb".split_whitespace();
783 /// assert_eq!(Some("Mary"), iter.next());
784 /// assert_eq!(Some("had"), iter.next());
785 /// assert_eq!(Some("a"), iter.next());
786 /// assert_eq!(Some("little"), iter.next());
787 /// assert_eq!(Some("lamb"), iter.next());
789 /// assert_eq!(None, iter.next());
791 #[stable(feature = "split_whitespace", since = "1.1.0")]
793 pub fn split_whitespace(&self) -> SplitWhitespace<'_> {
794 SplitWhitespace { inner: self.split(IsWhitespace).filter(IsNotEmpty) }
797 /// Splits a string slice by ASCII whitespace.
799 /// The iterator returned will return string slices that are sub-slices of
800 /// the original string slice, separated by any amount of ASCII whitespace.
802 /// To split by Unicode `Whitespace` instead, use [`split_whitespace`].
804 /// [`split_whitespace`]: str::split_whitespace
811 /// let mut iter = "A few words".split_ascii_whitespace();
813 /// assert_eq!(Some("A"), iter.next());
814 /// assert_eq!(Some("few"), iter.next());
815 /// assert_eq!(Some("words"), iter.next());
817 /// assert_eq!(None, iter.next());
820 /// All kinds of ASCII whitespace are considered:
823 /// let mut iter = " Mary had\ta little \n\t lamb".split_ascii_whitespace();
824 /// assert_eq!(Some("Mary"), iter.next());
825 /// assert_eq!(Some("had"), iter.next());
826 /// assert_eq!(Some("a"), iter.next());
827 /// assert_eq!(Some("little"), iter.next());
828 /// assert_eq!(Some("lamb"), iter.next());
830 /// assert_eq!(None, iter.next());
832 #[stable(feature = "split_ascii_whitespace", since = "1.34.0")]
834 pub fn split_ascii_whitespace(&self) -> SplitAsciiWhitespace<'_> {
836 self.as_bytes().split(IsAsciiWhitespace).filter(BytesIsNotEmpty).map(UnsafeBytesToStr);
837 SplitAsciiWhitespace { inner }
840 /// An iterator over the lines of a string, as string slices.
842 /// Lines are ended with either a newline (`\n`) or a carriage return with
843 /// a line feed (`\r\n`).
845 /// The final line ending is optional. A string that ends with a final line
846 /// ending will return the same lines as an otherwise identical string
847 /// without a final line ending.
854 /// let text = "foo\r\nbar\n\nbaz\n";
855 /// let mut lines = text.lines();
857 /// assert_eq!(Some("foo"), lines.next());
858 /// assert_eq!(Some("bar"), lines.next());
859 /// assert_eq!(Some(""), lines.next());
860 /// assert_eq!(Some("baz"), lines.next());
862 /// assert_eq!(None, lines.next());
865 /// The final line ending isn't required:
868 /// let text = "foo\nbar\n\r\nbaz";
869 /// let mut lines = text.lines();
871 /// assert_eq!(Some("foo"), lines.next());
872 /// assert_eq!(Some("bar"), lines.next());
873 /// assert_eq!(Some(""), lines.next());
874 /// assert_eq!(Some("baz"), lines.next());
876 /// assert_eq!(None, lines.next());
878 #[stable(feature = "rust1", since = "1.0.0")]
880 pub fn lines(&self) -> Lines<'_> {
881 Lines(self.split_terminator('\n').map(LinesAnyMap))
884 /// An iterator over the lines of a string.
885 #[stable(feature = "rust1", since = "1.0.0")]
886 #[rustc_deprecated(since = "1.4.0", reason = "use lines() instead now")]
889 pub fn lines_any(&self) -> LinesAny<'_> {
890 LinesAny(self.lines())
893 /// Returns an iterator of `u16` over the string encoded as UTF-16.
900 /// let text = "Zażółć gęślą jaźń";
902 /// let utf8_len = text.len();
903 /// let utf16_len = text.encode_utf16().count();
905 /// assert!(utf16_len <= utf8_len);
907 #[stable(feature = "encode_utf16", since = "1.8.0")]
908 pub fn encode_utf16(&self) -> EncodeUtf16<'_> {
909 EncodeUtf16 { chars: self.chars(), extra: 0 }
912 /// Returns `true` if the given pattern matches a sub-slice of
913 /// this string slice.
915 /// Returns `false` if it does not.
917 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
918 /// function or closure that determines if a character matches.
920 /// [`char`]: prim@char
921 /// [pattern]: self::pattern
928 /// let bananas = "bananas";
930 /// assert!(bananas.contains("nana"));
931 /// assert!(!bananas.contains("apples"));
933 #[stable(feature = "rust1", since = "1.0.0")]
935 pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
936 pat.is_contained_in(self)
939 /// Returns `true` if the given pattern matches a prefix of this
942 /// Returns `false` if it does not.
944 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
945 /// function or closure that determines if a character matches.
947 /// [`char`]: prim@char
948 /// [pattern]: self::pattern
955 /// let bananas = "bananas";
957 /// assert!(bananas.starts_with("bana"));
958 /// assert!(!bananas.starts_with("nana"));
960 #[stable(feature = "rust1", since = "1.0.0")]
961 pub fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
962 pat.is_prefix_of(self)
965 /// Returns `true` if the given pattern matches a suffix of this
968 /// Returns `false` if it does not.
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.
973 /// [`char`]: prim@char
974 /// [pattern]: self::pattern
981 /// let bananas = "bananas";
983 /// assert!(bananas.ends_with("anas"));
984 /// assert!(!bananas.ends_with("nana"));
986 #[stable(feature = "rust1", since = "1.0.0")]
987 pub fn ends_with<'a, P>(&'a self, pat: P) -> bool
989 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
991 pat.is_suffix_of(self)
994 /// Returns the byte index of the first character of this string slice that
995 /// matches the pattern.
997 /// Returns [`None`] if the pattern doesn't match.
999 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1000 /// function or closure that determines if a character matches.
1002 /// [`char`]: prim@char
1003 /// [pattern]: self::pattern
1007 /// Simple patterns:
1010 /// let s = "Löwe 老虎 Léopard Gepardi";
1012 /// assert_eq!(s.find('L'), Some(0));
1013 /// assert_eq!(s.find('é'), Some(14));
1014 /// assert_eq!(s.find("pard"), Some(17));
1017 /// More complex patterns using point-free style and closures:
1020 /// let s = "Löwe 老虎 Léopard";
1022 /// assert_eq!(s.find(char::is_whitespace), Some(5));
1023 /// assert_eq!(s.find(char::is_lowercase), Some(1));
1024 /// assert_eq!(s.find(|c: char| c.is_whitespace() || c.is_lowercase()), Some(1));
1025 /// assert_eq!(s.find(|c: char| (c < 'o') && (c > 'a')), Some(4));
1028 /// Not finding the pattern:
1031 /// let s = "Löwe 老虎 Léopard";
1032 /// let x: &[_] = &['1', '2'];
1034 /// assert_eq!(s.find(x), None);
1036 #[stable(feature = "rust1", since = "1.0.0")]
1038 pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
1039 pat.into_searcher(self).next_match().map(|(i, _)| i)
1042 /// Returns the byte index for the first character of the rightmost match of the pattern in
1043 /// this string slice.
1045 /// Returns [`None`] if the pattern doesn't match.
1047 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1048 /// function or closure that determines if a character matches.
1050 /// [`char`]: prim@char
1051 /// [pattern]: self::pattern
1055 /// Simple patterns:
1058 /// let s = "Löwe 老虎 Léopard Gepardi";
1060 /// assert_eq!(s.rfind('L'), Some(13));
1061 /// assert_eq!(s.rfind('é'), Some(14));
1062 /// assert_eq!(s.rfind("pard"), Some(24));
1065 /// More complex patterns with closures:
1068 /// let s = "Löwe 老虎 Léopard";
1070 /// assert_eq!(s.rfind(char::is_whitespace), Some(12));
1071 /// assert_eq!(s.rfind(char::is_lowercase), Some(20));
1074 /// Not finding the pattern:
1077 /// let s = "Löwe 老虎 Léopard";
1078 /// let x: &[_] = &['1', '2'];
1080 /// assert_eq!(s.rfind(x), None);
1082 #[stable(feature = "rust1", since = "1.0.0")]
1084 pub fn rfind<'a, P>(&'a self, pat: P) -> Option<usize>
1086 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1088 pat.into_searcher(self).next_match_back().map(|(i, _)| i)
1091 /// An iterator over substrings of this string slice, separated by
1092 /// characters matched by a pattern.
1094 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1095 /// function or closure that determines if a character matches.
1097 /// [`char`]: prim@char
1098 /// [pattern]: self::pattern
1100 /// # Iterator behavior
1102 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1103 /// allows a reverse search and forward/reverse search yields the same
1104 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1106 /// If the pattern allows a reverse search but its results might differ
1107 /// from a forward search, the [`rsplit`] method can be used.
1109 /// [`rsplit`]: str::rsplit
1113 /// Simple patterns:
1116 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
1117 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
1119 /// let v: Vec<&str> = "".split('X').collect();
1120 /// assert_eq!(v, [""]);
1122 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
1123 /// assert_eq!(v, ["lion", "", "tiger", "leopard"]);
1125 /// let v: Vec<&str> = "lion::tiger::leopard".split("::").collect();
1126 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1128 /// let v: Vec<&str> = "abc1def2ghi".split(char::is_numeric).collect();
1129 /// assert_eq!(v, ["abc", "def", "ghi"]);
1131 /// let v: Vec<&str> = "lionXtigerXleopard".split(char::is_uppercase).collect();
1132 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1135 /// If the pattern is a slice of chars, split on each occurrence of any of the characters:
1138 /// let v: Vec<&str> = "2020-11-03 23:59".split(&['-', ' ', ':', '@'][..]).collect();
1139 /// assert_eq!(v, ["2020", "11", "03", "23", "59"]);
1142 /// A more complex pattern, using a closure:
1145 /// let v: Vec<&str> = "abc1defXghi".split(|c| c == '1' || c == 'X').collect();
1146 /// assert_eq!(v, ["abc", "def", "ghi"]);
1149 /// If a string contains multiple contiguous separators, you will end up
1150 /// with empty strings in the output:
1153 /// let x = "||||a||b|c".to_string();
1154 /// let d: Vec<_> = x.split('|').collect();
1156 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1159 /// Contiguous separators are separated by the empty string.
1162 /// let x = "(///)".to_string();
1163 /// let d: Vec<_> = x.split('/').collect();
1165 /// assert_eq!(d, &["(", "", "", ")"]);
1168 /// Separators at the start or end of a string are neighbored
1169 /// by empty strings.
1172 /// let d: Vec<_> = "010".split("0").collect();
1173 /// assert_eq!(d, &["", "1", ""]);
1176 /// When the empty string is used as a separator, it separates
1177 /// every character in the string, along with the beginning
1178 /// and end of the string.
1181 /// let f: Vec<_> = "rust".split("").collect();
1182 /// assert_eq!(f, &["", "r", "u", "s", "t", ""]);
1185 /// Contiguous separators can lead to possibly surprising behavior
1186 /// when whitespace is used as the separator. This code is correct:
1189 /// let x = " a b c".to_string();
1190 /// let d: Vec<_> = x.split(' ').collect();
1192 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1195 /// It does _not_ give you:
1198 /// assert_eq!(d, &["a", "b", "c"]);
1201 /// Use [`split_whitespace`] for this behavior.
1203 /// [`split_whitespace`]: str::split_whitespace
1204 #[stable(feature = "rust1", since = "1.0.0")]
1206 pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
1207 Split(SplitInternal {
1210 matcher: pat.into_searcher(self),
1211 allow_trailing_empty: true,
1216 /// An iterator over substrings of this string slice, separated by
1217 /// characters matched by a pattern. Differs from the iterator produced by
1218 /// `split` in that `split_inclusive` leaves the matched part as the
1219 /// terminator of the substring.
1221 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1222 /// function or closure that determines if a character matches.
1224 /// [`char`]: prim@char
1225 /// [pattern]: self::pattern
1230 /// #![feature(split_inclusive)]
1231 /// let v: Vec<&str> = "Mary had a little lamb\nlittle lamb\nlittle lamb."
1232 /// .split_inclusive('\n').collect();
1233 /// assert_eq!(v, ["Mary had a little lamb\n", "little lamb\n", "little lamb."]);
1236 /// If the last element of the string is matched,
1237 /// that element will be considered the terminator of the preceding substring.
1238 /// That substring will be the last item returned by the iterator.
1241 /// #![feature(split_inclusive)]
1242 /// let v: Vec<&str> = "Mary had a little lamb\nlittle lamb\nlittle lamb.\n"
1243 /// .split_inclusive('\n').collect();
1244 /// assert_eq!(v, ["Mary had a little lamb\n", "little lamb\n", "little lamb.\n"]);
1246 #[unstable(feature = "split_inclusive", issue = "72360")]
1248 pub fn split_inclusive<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitInclusive<'a, P> {
1249 SplitInclusive(SplitInternal {
1252 matcher: pat.into_searcher(self),
1253 allow_trailing_empty: false,
1258 /// An iterator over substrings of the given string slice, separated by
1259 /// characters matched by a pattern and yielded in reverse order.
1261 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1262 /// function or closure that determines if a character matches.
1264 /// [`char`]: prim@char
1265 /// [pattern]: self::pattern
1267 /// # Iterator behavior
1269 /// The returned iterator requires that the pattern supports a reverse
1270 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1271 /// search yields the same elements.
1273 /// For iterating from the front, the [`split`] method can be used.
1275 /// [`split`]: str::split
1279 /// Simple patterns:
1282 /// let v: Vec<&str> = "Mary had a little lamb".rsplit(' ').collect();
1283 /// assert_eq!(v, ["lamb", "little", "a", "had", "Mary"]);
1285 /// let v: Vec<&str> = "".rsplit('X').collect();
1286 /// assert_eq!(v, [""]);
1288 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplit('X').collect();
1289 /// assert_eq!(v, ["leopard", "tiger", "", "lion"]);
1291 /// let v: Vec<&str> = "lion::tiger::leopard".rsplit("::").collect();
1292 /// assert_eq!(v, ["leopard", "tiger", "lion"]);
1295 /// A more complex pattern, using a closure:
1298 /// let v: Vec<&str> = "abc1defXghi".rsplit(|c| c == '1' || c == 'X').collect();
1299 /// assert_eq!(v, ["ghi", "def", "abc"]);
1301 #[stable(feature = "rust1", since = "1.0.0")]
1303 pub fn rsplit<'a, P>(&'a self, pat: P) -> RSplit<'a, P>
1305 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1307 RSplit(self.split(pat).0)
1310 /// An iterator over substrings of the given string slice, separated by
1311 /// characters matched by a pattern.
1313 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1314 /// function or closure that determines if a character matches.
1316 /// [`char`]: prim@char
1317 /// [pattern]: self::pattern
1319 /// Equivalent to [`split`], except that the trailing substring
1320 /// is skipped if empty.
1322 /// [`split`]: str::split
1324 /// This method can be used for string data that is _terminated_,
1325 /// rather than _separated_ by a pattern.
1327 /// # Iterator behavior
1329 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1330 /// allows a reverse search and forward/reverse search yields the same
1331 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1333 /// If the pattern allows a reverse search but its results might differ
1334 /// from a forward search, the [`rsplit_terminator`] method can be used.
1336 /// [`rsplit_terminator`]: str::rsplit_terminator
1343 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
1344 /// assert_eq!(v, ["A", "B"]);
1346 /// let v: Vec<&str> = "A..B..".split_terminator(".").collect();
1347 /// assert_eq!(v, ["A", "", "B", ""]);
1349 #[stable(feature = "rust1", since = "1.0.0")]
1351 pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
1352 SplitTerminator(SplitInternal { allow_trailing_empty: false, ..self.split(pat).0 })
1355 /// An iterator over substrings of `self`, separated by characters
1356 /// matched by a pattern and yielded in reverse order.
1358 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1359 /// function or closure that determines if a character matches.
1361 /// [`char`]: prim@char
1362 /// [pattern]: self::pattern
1364 /// Equivalent to [`split`], except that the trailing substring is
1365 /// skipped if empty.
1367 /// [`split`]: str::split
1369 /// This method can be used for string data that is _terminated_,
1370 /// rather than _separated_ by a pattern.
1372 /// # Iterator behavior
1374 /// The returned iterator requires that the pattern supports a
1375 /// reverse search, and it will be double ended if a forward/reverse
1376 /// search yields the same elements.
1378 /// For iterating from the front, the [`split_terminator`] method can be
1381 /// [`split_terminator`]: str::split_terminator
1386 /// let v: Vec<&str> = "A.B.".rsplit_terminator('.').collect();
1387 /// assert_eq!(v, ["B", "A"]);
1389 /// let v: Vec<&str> = "A..B..".rsplit_terminator(".").collect();
1390 /// assert_eq!(v, ["", "B", "", "A"]);
1392 #[stable(feature = "rust1", since = "1.0.0")]
1394 pub fn rsplit_terminator<'a, P>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1396 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1398 RSplitTerminator(self.split_terminator(pat).0)
1401 /// An iterator over substrings of the given string slice, separated by a
1402 /// pattern, restricted to returning at most `n` items.
1404 /// If `n` substrings are returned, the last substring (the `n`th substring)
1405 /// will contain the remainder of the string.
1407 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1408 /// function or closure that determines if a character matches.
1410 /// [`char`]: prim@char
1411 /// [pattern]: self::pattern
1413 /// # Iterator behavior
1415 /// The returned iterator will not be double ended, because it is
1416 /// not efficient to support.
1418 /// If the pattern allows a reverse search, the [`rsplitn`] method can be
1421 /// [`rsplitn`]: str::rsplitn
1425 /// Simple patterns:
1428 /// let v: Vec<&str> = "Mary had a little lambda".splitn(3, ' ').collect();
1429 /// assert_eq!(v, ["Mary", "had", "a little lambda"]);
1431 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(3, "X").collect();
1432 /// assert_eq!(v, ["lion", "", "tigerXleopard"]);
1434 /// let v: Vec<&str> = "abcXdef".splitn(1, 'X').collect();
1435 /// assert_eq!(v, ["abcXdef"]);
1437 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1438 /// assert_eq!(v, [""]);
1441 /// A more complex pattern, using a closure:
1444 /// let v: Vec<&str> = "abc1defXghi".splitn(2, |c| c == '1' || c == 'X').collect();
1445 /// assert_eq!(v, ["abc", "defXghi"]);
1447 #[stable(feature = "rust1", since = "1.0.0")]
1449 pub fn splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P> {
1450 SplitN(SplitNInternal { iter: self.split(pat).0, count: n })
1453 /// An iterator over substrings of this string slice, separated by a
1454 /// pattern, starting from the end of the string, restricted to returning
1455 /// at most `n` items.
1457 /// If `n` substrings are returned, the last substring (the `n`th substring)
1458 /// will contain the remainder of the string.
1460 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1461 /// function or closure that determines if a character matches.
1463 /// [`char`]: prim@char
1464 /// [pattern]: self::pattern
1466 /// # Iterator behavior
1468 /// The returned iterator will not be double ended, because it is not
1469 /// efficient to support.
1471 /// For splitting from the front, the [`splitn`] method can be used.
1473 /// [`splitn`]: str::splitn
1477 /// Simple patterns:
1480 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(3, ' ').collect();
1481 /// assert_eq!(v, ["lamb", "little", "Mary had a"]);
1483 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(3, 'X').collect();
1484 /// assert_eq!(v, ["leopard", "tiger", "lionX"]);
1486 /// let v: Vec<&str> = "lion::tiger::leopard".rsplitn(2, "::").collect();
1487 /// assert_eq!(v, ["leopard", "lion::tiger"]);
1490 /// A more complex pattern, using a closure:
1493 /// let v: Vec<&str> = "abc1defXghi".rsplitn(2, |c| c == '1' || c == 'X').collect();
1494 /// assert_eq!(v, ["ghi", "abc1def"]);
1496 #[stable(feature = "rust1", since = "1.0.0")]
1498 pub fn rsplitn<'a, P>(&'a self, n: usize, pat: P) -> RSplitN<'a, P>
1500 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1502 RSplitN(self.splitn(n, pat).0)
1505 /// Splits the string on the first occurrence of the specified delimiter and
1506 /// returns prefix before delimiter and suffix after delimiter.
1511 /// #![feature(str_split_once)]
1513 /// assert_eq!("cfg".split_once('='), None);
1514 /// assert_eq!("cfg=foo".split_once('='), Some(("cfg", "foo")));
1515 /// assert_eq!("cfg=foo=bar".split_once('='), Some(("cfg", "foo=bar")));
1517 #[unstable(feature = "str_split_once", reason = "newly added", issue = "74773")]
1519 pub fn split_once<'a, P: Pattern<'a>>(&'a self, delimiter: P) -> Option<(&'a str, &'a str)> {
1520 let (start, end) = delimiter.into_searcher(self).next_match()?;
1521 Some((&self[..start], &self[end..]))
1524 /// Splits the string on the last occurrence of the specified delimiter and
1525 /// returns prefix before delimiter and suffix after delimiter.
1530 /// #![feature(str_split_once)]
1532 /// assert_eq!("cfg".rsplit_once('='), None);
1533 /// assert_eq!("cfg=foo".rsplit_once('='), Some(("cfg", "foo")));
1534 /// assert_eq!("cfg=foo=bar".rsplit_once('='), Some(("cfg=foo", "bar")));
1536 #[unstable(feature = "str_split_once", reason = "newly added", issue = "74773")]
1538 pub fn rsplit_once<'a, P>(&'a self, delimiter: P) -> Option<(&'a str, &'a str)>
1540 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1542 let (start, end) = delimiter.into_searcher(self).next_match_back()?;
1543 Some((&self[..start], &self[end..]))
1546 /// An iterator over the disjoint matches of a pattern within the given string
1549 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1550 /// function or closure that determines if a character matches.
1552 /// [`char`]: prim@char
1553 /// [pattern]: self::pattern
1555 /// # Iterator behavior
1557 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1558 /// allows a reverse search and forward/reverse search yields the same
1559 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1561 /// If the pattern allows a reverse search but its results might differ
1562 /// from a forward search, the [`rmatches`] method can be used.
1564 /// [`rmatches`]: str::matches
1571 /// let v: Vec<&str> = "abcXXXabcYYYabc".matches("abc").collect();
1572 /// assert_eq!(v, ["abc", "abc", "abc"]);
1574 /// let v: Vec<&str> = "1abc2abc3".matches(char::is_numeric).collect();
1575 /// assert_eq!(v, ["1", "2", "3"]);
1577 #[stable(feature = "str_matches", since = "1.2.0")]
1579 pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1580 Matches(MatchesInternal(pat.into_searcher(self)))
1583 /// An iterator over the disjoint matches of a pattern within this string slice,
1584 /// yielded in reverse order.
1586 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1587 /// function or closure that determines if a character matches.
1589 /// [`char`]: prim@char
1590 /// [pattern]: self::pattern
1592 /// # Iterator behavior
1594 /// The returned iterator requires that the pattern supports a reverse
1595 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1596 /// search yields the same elements.
1598 /// For iterating from the front, the [`matches`] method can be used.
1600 /// [`matches`]: str::matches
1607 /// let v: Vec<&str> = "abcXXXabcYYYabc".rmatches("abc").collect();
1608 /// assert_eq!(v, ["abc", "abc", "abc"]);
1610 /// let v: Vec<&str> = "1abc2abc3".rmatches(char::is_numeric).collect();
1611 /// assert_eq!(v, ["3", "2", "1"]);
1613 #[stable(feature = "str_matches", since = "1.2.0")]
1615 pub fn rmatches<'a, P>(&'a self, pat: P) -> RMatches<'a, P>
1617 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1619 RMatches(self.matches(pat).0)
1622 /// An iterator over the disjoint matches of a pattern within this string
1623 /// slice as well as the index that the match starts at.
1625 /// For matches of `pat` within `self` that overlap, only the indices
1626 /// corresponding to the first match are returned.
1628 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1629 /// function or closure that determines if a character matches.
1631 /// [`char`]: prim@char
1632 /// [pattern]: self::pattern
1634 /// # Iterator behavior
1636 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1637 /// allows a reverse search and forward/reverse search yields the same
1638 /// elements. This is true for, e.g., [`char`], but not for `&str`.
1640 /// If the pattern allows a reverse search but its results might differ
1641 /// from a forward search, the [`rmatch_indices`] method can be used.
1643 /// [`rmatch_indices`]: str::match_indices
1650 /// let v: Vec<_> = "abcXXXabcYYYabc".match_indices("abc").collect();
1651 /// assert_eq!(v, [(0, "abc"), (6, "abc"), (12, "abc")]);
1653 /// let v: Vec<_> = "1abcabc2".match_indices("abc").collect();
1654 /// assert_eq!(v, [(1, "abc"), (4, "abc")]);
1656 /// let v: Vec<_> = "ababa".match_indices("aba").collect();
1657 /// assert_eq!(v, [(0, "aba")]); // only the first `aba`
1659 #[stable(feature = "str_match_indices", since = "1.5.0")]
1661 pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1662 MatchIndices(MatchIndicesInternal(pat.into_searcher(self)))
1665 /// An iterator over the disjoint matches of a pattern within `self`,
1666 /// yielded in reverse order along with the index of the match.
1668 /// For matches of `pat` within `self` that overlap, only the indices
1669 /// corresponding to the last match are returned.
1671 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1672 /// function or closure that determines if a character matches.
1674 /// [`char`]: prim@char
1675 /// [pattern]: self::pattern
1677 /// # Iterator behavior
1679 /// The returned iterator requires that the pattern supports a reverse
1680 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1681 /// search yields the same elements.
1683 /// For iterating from the front, the [`match_indices`] method can be used.
1685 /// [`match_indices`]: str::match_indices
1692 /// let v: Vec<_> = "abcXXXabcYYYabc".rmatch_indices("abc").collect();
1693 /// assert_eq!(v, [(12, "abc"), (6, "abc"), (0, "abc")]);
1695 /// let v: Vec<_> = "1abcabc2".rmatch_indices("abc").collect();
1696 /// assert_eq!(v, [(4, "abc"), (1, "abc")]);
1698 /// let v: Vec<_> = "ababa".rmatch_indices("aba").collect();
1699 /// assert_eq!(v, [(2, "aba")]); // only the last `aba`
1701 #[stable(feature = "str_match_indices", since = "1.5.0")]
1703 pub fn rmatch_indices<'a, P>(&'a self, pat: P) -> RMatchIndices<'a, P>
1705 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
1707 RMatchIndices(self.match_indices(pat).0)
1710 /// Returns a string slice with leading and trailing whitespace removed.
1712 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1713 /// Core Property `White_Space`.
1720 /// let s = " Hello\tworld\t";
1722 /// assert_eq!("Hello\tworld", s.trim());
1725 #[must_use = "this returns the trimmed string as a slice, \
1726 without modifying the original"]
1727 #[stable(feature = "rust1", since = "1.0.0")]
1728 pub fn trim(&self) -> &str {
1729 self.trim_matches(|c: char| c.is_whitespace())
1732 /// Returns a string slice with leading whitespace removed.
1734 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1735 /// Core Property `White_Space`.
1737 /// # Text directionality
1739 /// A string is a sequence of bytes. `start` in this context means the first
1740 /// position of that byte string; for a left-to-right language like English or
1741 /// Russian, this will be left side, and for right-to-left languages like
1742 /// Arabic or Hebrew, this will be the right side.
1749 /// let s = " Hello\tworld\t";
1750 /// assert_eq!("Hello\tworld\t", s.trim_start());
1756 /// let s = " English ";
1757 /// assert!(Some('E') == s.trim_start().chars().next());
1759 /// let s = " עברית ";
1760 /// assert!(Some('ע') == s.trim_start().chars().next());
1763 #[must_use = "this returns the trimmed string as a new slice, \
1764 without modifying the original"]
1765 #[stable(feature = "trim_direction", since = "1.30.0")]
1766 pub fn trim_start(&self) -> &str {
1767 self.trim_start_matches(|c: char| c.is_whitespace())
1770 /// Returns a string slice with trailing whitespace removed.
1772 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1773 /// Core Property `White_Space`.
1775 /// # Text directionality
1777 /// A string is a sequence of bytes. `end` in this context means the last
1778 /// position of that byte string; for a left-to-right language like English or
1779 /// Russian, this will be right side, and for right-to-left languages like
1780 /// Arabic or Hebrew, this will be the left side.
1787 /// let s = " Hello\tworld\t";
1788 /// assert_eq!(" Hello\tworld", s.trim_end());
1794 /// let s = " English ";
1795 /// assert!(Some('h') == s.trim_end().chars().rev().next());
1797 /// let s = " עברית ";
1798 /// assert!(Some('ת') == s.trim_end().chars().rev().next());
1801 #[must_use = "this returns the trimmed string as a new slice, \
1802 without modifying the original"]
1803 #[stable(feature = "trim_direction", since = "1.30.0")]
1804 pub fn trim_end(&self) -> &str {
1805 self.trim_end_matches(|c: char| c.is_whitespace())
1808 /// Returns a string slice with leading whitespace removed.
1810 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1811 /// Core Property `White_Space`.
1813 /// # Text directionality
1815 /// A string is a sequence of bytes. 'Left' in this context means the first
1816 /// position of that byte string; for a language like Arabic or Hebrew
1817 /// which are 'right to left' rather than 'left to right', this will be
1818 /// the _right_ side, not the left.
1825 /// let s = " Hello\tworld\t";
1827 /// assert_eq!("Hello\tworld\t", s.trim_left());
1833 /// let s = " English";
1834 /// assert!(Some('E') == s.trim_left().chars().next());
1836 /// let s = " עברית";
1837 /// assert!(Some('ע') == s.trim_left().chars().next());
1840 #[stable(feature = "rust1", since = "1.0.0")]
1843 reason = "superseded by `trim_start`",
1844 suggestion = "trim_start"
1846 pub fn trim_left(&self) -> &str {
1850 /// Returns a string slice with trailing whitespace removed.
1852 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1853 /// Core Property `White_Space`.
1855 /// # Text directionality
1857 /// A string is a sequence of bytes. 'Right' in this context means the last
1858 /// position of that byte string; for a language like Arabic or Hebrew
1859 /// which are 'right to left' rather than 'left to right', this will be
1860 /// the _left_ side, not the right.
1867 /// let s = " Hello\tworld\t";
1869 /// assert_eq!(" Hello\tworld", s.trim_right());
1875 /// let s = "English ";
1876 /// assert!(Some('h') == s.trim_right().chars().rev().next());
1878 /// let s = "עברית ";
1879 /// assert!(Some('ת') == s.trim_right().chars().rev().next());
1882 #[stable(feature = "rust1", since = "1.0.0")]
1885 reason = "superseded by `trim_end`",
1886 suggestion = "trim_end"
1888 pub fn trim_right(&self) -> &str {
1892 /// Returns a string slice with all prefixes and suffixes that match a
1893 /// pattern repeatedly removed.
1895 /// The [pattern] can be a [`char`], a slice of [`char`]s, or a function
1896 /// or closure that determines if a character matches.
1898 /// [`char`]: prim@char
1899 /// [pattern]: self::pattern
1903 /// Simple patterns:
1906 /// assert_eq!("11foo1bar11".trim_matches('1'), "foo1bar");
1907 /// assert_eq!("123foo1bar123".trim_matches(char::is_numeric), "foo1bar");
1909 /// let x: &[_] = &['1', '2'];
1910 /// assert_eq!("12foo1bar12".trim_matches(x), "foo1bar");
1913 /// A more complex pattern, using a closure:
1916 /// assert_eq!("1foo1barXX".trim_matches(|c| c == '1' || c == 'X'), "foo1bar");
1918 #[must_use = "this returns the trimmed string as a new slice, \
1919 without modifying the original"]
1920 #[stable(feature = "rust1", since = "1.0.0")]
1921 pub fn trim_matches<'a, P>(&'a self, pat: P) -> &'a str
1923 P: Pattern<'a, Searcher: DoubleEndedSearcher<'a>>,
1927 let mut matcher = pat.into_searcher(self);
1928 if let Some((a, b)) = matcher.next_reject() {
1930 j = b; // Remember earliest known match, correct it below if
1931 // last match is different
1933 if let Some((_, b)) = matcher.next_reject_back() {
1936 // SAFETY: `Searcher` is known to return valid indices.
1937 unsafe { self.get_unchecked(i..j) }
1940 /// Returns a string slice with all prefixes that match a pattern
1941 /// repeatedly removed.
1943 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1944 /// function or closure that determines if a character matches.
1946 /// [`char`]: prim@char
1947 /// [pattern]: self::pattern
1949 /// # Text directionality
1951 /// A string is a sequence of bytes. `start` in this context means the first
1952 /// position of that byte string; for a left-to-right language like English or
1953 /// Russian, this will be left side, and for right-to-left languages like
1954 /// Arabic or Hebrew, this will be the right side.
1961 /// assert_eq!("11foo1bar11".trim_start_matches('1'), "foo1bar11");
1962 /// assert_eq!("123foo1bar123".trim_start_matches(char::is_numeric), "foo1bar123");
1964 /// let x: &[_] = &['1', '2'];
1965 /// assert_eq!("12foo1bar12".trim_start_matches(x), "foo1bar12");
1967 #[must_use = "this returns the trimmed string as a new slice, \
1968 without modifying the original"]
1969 #[stable(feature = "trim_direction", since = "1.30.0")]
1970 pub fn trim_start_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
1971 let mut i = self.len();
1972 let mut matcher = pat.into_searcher(self);
1973 if let Some((a, _)) = matcher.next_reject() {
1976 // SAFETY: `Searcher` is known to return valid indices.
1977 unsafe { self.get_unchecked(i..self.len()) }
1980 /// Returns a string slice with the prefix removed.
1982 /// If the string starts with the pattern `prefix`, returns substring after the prefix, wrapped
1983 /// in `Some`. Unlike `trim_start_matches`, this method removes the prefix exactly once.
1985 /// If the string does not start with `prefix`, returns `None`.
1987 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
1988 /// function or closure that determines if a character matches.
1990 /// [`char`]: prim@char
1991 /// [pattern]: self::pattern
1996 /// assert_eq!("foo:bar".strip_prefix("foo:"), Some("bar"));
1997 /// assert_eq!("foo:bar".strip_prefix("bar"), None);
1998 /// assert_eq!("foofoo".strip_prefix("foo"), Some("foo"));
2000 #[must_use = "this returns the remaining substring as a new slice, \
2001 without modifying the original"]
2002 #[stable(feature = "str_strip", since = "1.45.0")]
2003 pub fn strip_prefix<'a, P: Pattern<'a>>(&'a self, prefix: P) -> Option<&'a str> {
2004 prefix.strip_prefix_of(self)
2007 /// Returns a string slice with the suffix removed.
2009 /// If the string ends with the pattern `suffix`, returns the substring before the suffix,
2010 /// wrapped in `Some`. Unlike `trim_end_matches`, this method removes the suffix exactly once.
2012 /// If the string does not end with `suffix`, returns `None`.
2014 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2015 /// function or closure that determines if a character matches.
2017 /// [`char`]: prim@char
2018 /// [pattern]: self::pattern
2023 /// assert_eq!("bar:foo".strip_suffix(":foo"), Some("bar"));
2024 /// assert_eq!("bar:foo".strip_suffix("bar"), None);
2025 /// assert_eq!("foofoo".strip_suffix("foo"), Some("foo"));
2027 #[must_use = "this returns the remaining substring as a new slice, \
2028 without modifying the original"]
2029 #[stable(feature = "str_strip", since = "1.45.0")]
2030 pub fn strip_suffix<'a, P>(&'a self, suffix: P) -> Option<&'a str>
2033 <P as Pattern<'a>>::Searcher: ReverseSearcher<'a>,
2035 suffix.strip_suffix_of(self)
2038 /// Returns a string slice with all suffixes that match a pattern
2039 /// repeatedly removed.
2041 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2042 /// function or closure that determines if a character matches.
2044 /// [`char`]: prim@char
2045 /// [pattern]: self::pattern
2047 /// # Text directionality
2049 /// A string is a sequence of bytes. `end` in this context means the last
2050 /// position of that byte string; for a left-to-right language like English or
2051 /// Russian, this will be right side, and for right-to-left languages like
2052 /// Arabic or Hebrew, this will be the left side.
2056 /// Simple patterns:
2059 /// assert_eq!("11foo1bar11".trim_end_matches('1'), "11foo1bar");
2060 /// assert_eq!("123foo1bar123".trim_end_matches(char::is_numeric), "123foo1bar");
2062 /// let x: &[_] = &['1', '2'];
2063 /// assert_eq!("12foo1bar12".trim_end_matches(x), "12foo1bar");
2066 /// A more complex pattern, using a closure:
2069 /// assert_eq!("1fooX".trim_end_matches(|c| c == '1' || c == 'X'), "1foo");
2071 #[must_use = "this returns the trimmed string as a new slice, \
2072 without modifying the original"]
2073 #[stable(feature = "trim_direction", since = "1.30.0")]
2074 pub fn trim_end_matches<'a, P>(&'a self, pat: P) -> &'a str
2076 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
2079 let mut matcher = pat.into_searcher(self);
2080 if let Some((_, b)) = matcher.next_reject_back() {
2083 // SAFETY: `Searcher` is known to return valid indices.
2084 unsafe { self.get_unchecked(0..j) }
2087 /// Returns a string slice with all prefixes that match a pattern
2088 /// repeatedly removed.
2090 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2091 /// function or closure that determines if a character matches.
2093 /// [`char`]: prim@char
2094 /// [pattern]: self::pattern
2096 /// # Text directionality
2098 /// A string is a sequence of bytes. 'Left' in this context means the first
2099 /// position of that byte string; for a language like Arabic or Hebrew
2100 /// which are 'right to left' rather than 'left to right', this will be
2101 /// the _right_ side, not the left.
2108 /// assert_eq!("11foo1bar11".trim_left_matches('1'), "foo1bar11");
2109 /// assert_eq!("123foo1bar123".trim_left_matches(char::is_numeric), "foo1bar123");
2111 /// let x: &[_] = &['1', '2'];
2112 /// assert_eq!("12foo1bar12".trim_left_matches(x), "foo1bar12");
2114 #[stable(feature = "rust1", since = "1.0.0")]
2117 reason = "superseded by `trim_start_matches`",
2118 suggestion = "trim_start_matches"
2120 pub fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
2121 self.trim_start_matches(pat)
2124 /// Returns a string slice with all suffixes that match a pattern
2125 /// repeatedly removed.
2127 /// The [pattern] can be a `&str`, [`char`], a slice of [`char`]s, or a
2128 /// function or closure that determines if a character matches.
2130 /// [`char`]: prim@char
2131 /// [pattern]: self::pattern
2133 /// # Text directionality
2135 /// A string is a sequence of bytes. 'Right' in this context means the last
2136 /// position of that byte string; for a language like Arabic or Hebrew
2137 /// which are 'right to left' rather than 'left to right', this will be
2138 /// the _left_ side, not the right.
2142 /// Simple patterns:
2145 /// assert_eq!("11foo1bar11".trim_right_matches('1'), "11foo1bar");
2146 /// assert_eq!("123foo1bar123".trim_right_matches(char::is_numeric), "123foo1bar");
2148 /// let x: &[_] = &['1', '2'];
2149 /// assert_eq!("12foo1bar12".trim_right_matches(x), "12foo1bar");
2152 /// A more complex pattern, using a closure:
2155 /// assert_eq!("1fooX".trim_right_matches(|c| c == '1' || c == 'X'), "1foo");
2157 #[stable(feature = "rust1", since = "1.0.0")]
2160 reason = "superseded by `trim_end_matches`",
2161 suggestion = "trim_end_matches"
2163 pub fn trim_right_matches<'a, P>(&'a self, pat: P) -> &'a str
2165 P: Pattern<'a, Searcher: ReverseSearcher<'a>>,
2167 self.trim_end_matches(pat)
2170 /// Parses this string slice into another type.
2172 /// Because `parse` is so general, it can cause problems with type
2173 /// inference. As such, `parse` is one of the few times you'll see
2174 /// the syntax affectionately known as the 'turbofish': `::<>`. This
2175 /// helps the inference algorithm understand specifically which type
2176 /// you're trying to parse into.
2178 /// `parse` can parse any type that implements the [`FromStr`] trait.
2183 /// Will return [`Err`] if it's not possible to parse this string slice into
2184 /// the desired type.
2186 /// [`Err`]: FromStr::Err
2193 /// let four: u32 = "4".parse().unwrap();
2195 /// assert_eq!(4, four);
2198 /// Using the 'turbofish' instead of annotating `four`:
2201 /// let four = "4".parse::<u32>();
2203 /// assert_eq!(Ok(4), four);
2206 /// Failing to parse:
2209 /// let nope = "j".parse::<u32>();
2211 /// assert!(nope.is_err());
2214 #[stable(feature = "rust1", since = "1.0.0")]
2215 pub fn parse<F: FromStr>(&self) -> Result<F, F::Err> {
2216 FromStr::from_str(self)
2219 /// Checks if all characters in this string are within the ASCII range.
2224 /// let ascii = "hello!\n";
2225 /// let non_ascii = "Grüße, Jürgen ❤";
2227 /// assert!(ascii.is_ascii());
2228 /// assert!(!non_ascii.is_ascii());
2230 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2232 pub fn is_ascii(&self) -> bool {
2233 // We can treat each byte as character here: all multibyte characters
2234 // start with a byte that is not in the ascii range, so we will stop
2236 self.as_bytes().is_ascii()
2239 /// Checks that two strings are an ASCII case-insensitive match.
2241 /// Same as `to_ascii_lowercase(a) == to_ascii_lowercase(b)`,
2242 /// but without allocating and copying temporaries.
2247 /// assert!("Ferris".eq_ignore_ascii_case("FERRIS"));
2248 /// assert!("Ferrös".eq_ignore_ascii_case("FERRöS"));
2249 /// assert!(!"Ferrös".eq_ignore_ascii_case("FERRÖS"));
2251 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2253 pub fn eq_ignore_ascii_case(&self, other: &str) -> bool {
2254 self.as_bytes().eq_ignore_ascii_case(other.as_bytes())
2257 /// Converts this string to its ASCII upper case equivalent in-place.
2259 /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z',
2260 /// but non-ASCII letters are unchanged.
2262 /// To return a new uppercased value without modifying the existing one, use
2263 /// [`to_ascii_uppercase()`].
2265 /// [`to_ascii_uppercase()`]: #method.to_ascii_uppercase
2270 /// let mut s = String::from("Grüße, Jürgen ❤");
2272 /// s.make_ascii_uppercase();
2274 /// assert_eq!("GRüßE, JüRGEN ❤", s);
2276 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2278 pub fn make_ascii_uppercase(&mut self) {
2279 // SAFETY: safe because we transmute two types with the same layout.
2280 let me = unsafe { self.as_bytes_mut() };
2281 me.make_ascii_uppercase()
2284 /// Converts this string to its ASCII lower case equivalent in-place.
2286 /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z',
2287 /// but non-ASCII letters are unchanged.
2289 /// To return a new lowercased value without modifying the existing one, use
2290 /// [`to_ascii_lowercase()`].
2292 /// [`to_ascii_lowercase()`]: #method.to_ascii_lowercase
2297 /// let mut s = String::from("GRÜßE, JÜRGEN ❤");
2299 /// s.make_ascii_lowercase();
2301 /// assert_eq!("grÜße, jÜrgen ❤", s);
2303 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
2305 pub fn make_ascii_lowercase(&mut self) {
2306 // SAFETY: safe because we transmute two types with the same layout.
2307 let me = unsafe { self.as_bytes_mut() };
2308 me.make_ascii_lowercase()
2311 /// Return an iterator that escapes each char in `self` with [`char::escape_debug`].
2313 /// Note: only extended grapheme codepoints that begin the string will be
2321 /// for c in "❤\n!".escape_debug() {
2322 /// print!("{}", c);
2327 /// Using `println!` directly:
2330 /// println!("{}", "❤\n!".escape_debug());
2334 /// Both are equivalent to:
2337 /// println!("❤\\n!");
2340 /// Using `to_string`:
2343 /// assert_eq!("❤\n!".escape_debug().to_string(), "❤\\n!");
2345 #[stable(feature = "str_escape", since = "1.34.0")]
2346 pub fn escape_debug(&self) -> EscapeDebug<'_> {
2347 let mut chars = self.chars();
2351 .map(|first| first.escape_debug_ext(true))
2354 .chain(chars.flat_map(CharEscapeDebugContinue)),
2358 /// Return an iterator that escapes each char in `self` with [`char::escape_default`].
2365 /// for c in "❤\n!".escape_default() {
2366 /// print!("{}", c);
2371 /// Using `println!` directly:
2374 /// println!("{}", "❤\n!".escape_default());
2378 /// Both are equivalent to:
2381 /// println!("\\u{{2764}}\\n!");
2384 /// Using `to_string`:
2387 /// assert_eq!("❤\n!".escape_default().to_string(), "\\u{2764}\\n!");
2389 #[stable(feature = "str_escape", since = "1.34.0")]
2390 pub fn escape_default(&self) -> EscapeDefault<'_> {
2391 EscapeDefault { inner: self.chars().flat_map(CharEscapeDefault) }
2394 /// Return an iterator that escapes each char in `self` with [`char::escape_unicode`].
2401 /// for c in "❤\n!".escape_unicode() {
2402 /// print!("{}", c);
2407 /// Using `println!` directly:
2410 /// println!("{}", "❤\n!".escape_unicode());
2414 /// Both are equivalent to:
2417 /// println!("\\u{{2764}}\\u{{a}}\\u{{21}}");
2420 /// Using `to_string`:
2423 /// assert_eq!("❤\n!".escape_unicode().to_string(), "\\u{2764}\\u{a}\\u{21}");
2425 #[stable(feature = "str_escape", since = "1.34.0")]
2426 pub fn escape_unicode(&self) -> EscapeUnicode<'_> {
2427 EscapeUnicode { inner: self.chars().flat_map(CharEscapeUnicode) }
2431 #[stable(feature = "rust1", since = "1.0.0")]
2432 impl AsRef<[u8]> for str {
2434 fn as_ref(&self) -> &[u8] {
2439 #[stable(feature = "rust1", since = "1.0.0")]
2440 impl Default for &str {
2441 /// Creates an empty str
2443 fn default() -> Self {
2448 #[stable(feature = "default_mut_str", since = "1.28.0")]
2449 impl Default for &mut str {
2450 /// Creates an empty mutable str
2452 fn default() -> Self {
2453 // SAFETY: The empty string is valid UTF-8.
2454 unsafe { from_utf8_unchecked_mut(&mut []) }
2459 /// A nameable, cloneable fn type
2461 struct LinesAnyMap impl<'a> Fn = |line: &'a str| -> &'a str {
2463 if l > 0 && line.as_bytes()[l - 1] == b'\r' { &line[0 .. l - 1] }
2468 struct CharEscapeDebugContinue impl Fn = |c: char| -> char::EscapeDebug {
2469 c.escape_debug_ext(false)
2473 struct CharEscapeUnicode impl Fn = |c: char| -> char::EscapeUnicode {
2477 struct CharEscapeDefault impl Fn = |c: char| -> char::EscapeDefault {
2482 struct IsWhitespace impl Fn = |c: char| -> bool {
2487 struct IsAsciiWhitespace impl Fn = |byte: &u8| -> bool {
2488 byte.is_ascii_whitespace()
2492 struct IsNotEmpty impl<'a, 'b> Fn = |s: &'a &'b str| -> bool {
2497 struct BytesIsNotEmpty impl<'a, 'b> Fn = |s: &'a &'b [u8]| -> bool {
2502 struct UnsafeBytesToStr impl<'a> Fn = |bytes: &'a [u8]| -> &'a str {
2504 unsafe { from_utf8_unchecked(bytes) }