1 // Copyright 2012-2017 The Rust Project Developers. See the COPYRIGHT
2 // file at the top-level directory of this distribution and at
3 // http://rust-lang.org/COPYRIGHT.
5 // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
6 // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
7 // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
8 // option. This file may not be copied, modified, or distributed
9 // except according to those terms.
11 //! Unicode string slices.
13 //! The `&str` type is one of the two main string types, the other being `String`.
14 //! Unlike its `String` counterpart, its contents are borrowed.
18 //! A basic string declaration of `&str` type:
21 //! let hello_world = "Hello, World!";
24 //! Here we have declared a string literal, also known as a string slice.
25 //! String literals have a static lifetime, which means the string `hello_world`
26 //! is guaranteed to be valid for the duration of the entire program.
27 //! We can explicitly specify `hello_world`'s lifetime as well:
30 //! let hello_world: &'static str = "Hello, world!";
33 //! *[See also the `str` primitive type](../../std/primitive.str.html).*
35 #![stable(feature = "rust1", since = "1.0.0")]
37 // Many of the usings in this module are only used in the test configuration.
38 // It's cleaner to just turn off the unused_imports warning than to fix them.
39 #![allow(unused_imports)]
42 use core::str as core_str;
43 use core::str::pattern::Pattern;
44 use core::str::pattern::{Searcher, ReverseSearcher, DoubleEndedSearcher};
46 use core::iter::FusedIterator;
47 use std_unicode::str::{UnicodeStr, Utf16Encoder};
49 use vec_deque::VecDeque;
50 use borrow::{Borrow, ToOwned};
54 use slice::{SliceConcatExt, SliceIndex};
57 #[stable(feature = "rust1", since = "1.0.0")]
58 pub use core::str::{FromStr, Utf8Error};
60 #[stable(feature = "rust1", since = "1.0.0")]
61 pub use core::str::{Lines, LinesAny};
62 #[stable(feature = "rust1", since = "1.0.0")]
63 pub use core::str::{Split, RSplit};
64 #[stable(feature = "rust1", since = "1.0.0")]
65 pub use core::str::{SplitN, RSplitN};
66 #[stable(feature = "rust1", since = "1.0.0")]
67 pub use core::str::{SplitTerminator, RSplitTerminator};
68 #[stable(feature = "rust1", since = "1.0.0")]
69 pub use core::str::{Matches, RMatches};
70 #[stable(feature = "rust1", since = "1.0.0")]
71 pub use core::str::{MatchIndices, RMatchIndices};
72 #[stable(feature = "rust1", since = "1.0.0")]
73 pub use core::str::{from_utf8, from_utf8_mut, Chars, CharIndices, Bytes};
74 #[stable(feature = "rust1", since = "1.0.0")]
75 pub use core::str::{from_utf8_unchecked, from_utf8_unchecked_mut, ParseBoolError};
76 #[stable(feature = "rust1", since = "1.0.0")]
77 pub use std_unicode::str::SplitWhitespace;
78 #[stable(feature = "rust1", since = "1.0.0")]
79 pub use core::str::pattern;
82 #[unstable(feature = "slice_concat_ext",
83 reason = "trait should not have to exist",
85 impl<S: Borrow<str>> SliceConcatExt<str> for [S] {
88 fn concat(&self) -> String {
93 // `len` calculation may overflow but push_str will check boundaries
94 let len = self.iter().map(|s| s.borrow().len()).sum();
95 let mut result = String::with_capacity(len);
98 result.push_str(s.borrow())
104 fn join(&self, sep: &str) -> String {
106 return String::new();
111 return self.concat();
114 // this is wrong without the guarantee that `self` is non-empty
115 // `len` calculation may overflow but push_str but will check boundaries
116 let len = sep.len() * (self.len() - 1) +
117 self.iter().map(|s| s.borrow().len()).sum::<usize>();
118 let mut result = String::with_capacity(len);
119 let mut first = true;
125 result.push_str(sep);
127 result.push_str(s.borrow());
132 fn connect(&self, sep: &str) -> String {
137 /// An iterator of [`u16`] over the string encoded as UTF-16.
139 /// [`u16`]: ../../std/primitive.u16.html
141 /// This struct is created by the [`encode_utf16`] method on [`str`].
142 /// See its documentation for more.
144 /// [`encode_utf16`]: ../../std/primitive.str.html#method.encode_utf16
145 /// [`str`]: ../../std/primitive.str.html
147 #[stable(feature = "encode_utf16", since = "1.8.0")]
148 pub struct EncodeUtf16<'a> {
149 encoder: Utf16Encoder<Chars<'a>>,
152 #[stable(feature = "collection_debug", since = "1.17.0")]
153 impl<'a> fmt::Debug for EncodeUtf16<'a> {
154 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
155 f.pad("EncodeUtf16 { .. }")
159 #[stable(feature = "encode_utf16", since = "1.8.0")]
160 impl<'a> Iterator for EncodeUtf16<'a> {
164 fn next(&mut self) -> Option<u16> {
169 fn size_hint(&self) -> (usize, Option<usize>) {
170 self.encoder.size_hint()
174 #[unstable(feature = "fused", issue = "35602")]
175 impl<'a> FusedIterator for EncodeUtf16<'a> {}
177 #[stable(feature = "rust1", since = "1.0.0")]
178 impl Borrow<str> for String {
180 fn borrow(&self) -> &str {
185 #[stable(feature = "rust1", since = "1.0.0")]
186 impl ToOwned for str {
188 fn to_owned(&self) -> String {
189 unsafe { String::from_utf8_unchecked(self.as_bytes().to_owned()) }
192 fn clone_into(&self, target: &mut String) {
193 let mut b = mem::replace(target, String::new()).into_bytes();
194 self.as_bytes().clone_into(&mut b);
195 *target = unsafe { String::from_utf8_unchecked(b) }
199 /// Methods for string slices.
203 /// Returns the length of `self`.
205 /// This length is in bytes, not [`char`]s or graphemes. In other words,
206 /// it may not be what a human considers the length of the string.
208 /// [`char`]: primitive.char.html
215 /// let len = "foo".len();
216 /// assert_eq!(3, len);
218 /// let len = "ƒoo".len(); // fancy f!
219 /// assert_eq!(4, len);
221 #[stable(feature = "rust1", since = "1.0.0")]
223 pub fn len(&self) -> usize {
224 core_str::StrExt::len(self)
227 /// Returns `true` if `self` has a length of zero bytes.
235 /// assert!(s.is_empty());
237 /// let s = "not empty";
238 /// assert!(!s.is_empty());
241 #[stable(feature = "rust1", since = "1.0.0")]
242 pub fn is_empty(&self) -> bool {
243 core_str::StrExt::is_empty(self)
246 /// Checks that `index`-th byte lies at the start and/or end of a
247 /// UTF-8 code point sequence.
249 /// The start and end of the string (when `index == self.len()`) are
253 /// Returns `false` if `index` is greater than `self.len()`.
258 /// let s = "Löwe 老虎 Léopard";
259 /// assert!(s.is_char_boundary(0));
261 /// assert!(s.is_char_boundary(6));
262 /// assert!(s.is_char_boundary(s.len()));
264 /// // second byte of `ö`
265 /// assert!(!s.is_char_boundary(2));
267 /// // third byte of `老`
268 /// assert!(!s.is_char_boundary(8));
270 #[stable(feature = "is_char_boundary", since = "1.9.0")]
272 pub fn is_char_boundary(&self, index: usize) -> bool {
273 core_str::StrExt::is_char_boundary(self, index)
276 /// Converts a string slice to a byte slice. To convert the byte slice back
277 /// into a string slice, use the [`str::from_utf8`] function.
279 /// [`str::from_utf8`]: ./str/fn.from_utf8.html
286 /// let bytes = "bors".as_bytes();
287 /// assert_eq!(b"bors", bytes);
289 #[stable(feature = "rust1", since = "1.0.0")]
291 pub fn as_bytes(&self) -> &[u8] {
292 core_str::StrExt::as_bytes(self)
295 /// Converts a mutable string slice to a mutable byte slice. To convert the
296 /// mutable byte slice back into a mutable string slice, use the
297 /// [`str::from_utf8_mut`] function.
299 /// [`str::from_utf8_mut`]: ./str/fn.from_utf8_mut.html
306 /// let mut s = String::from("Hello");
307 /// let bytes = unsafe { s.as_bytes_mut() };
309 /// assert_eq!(b"Hello", bytes);
315 /// let mut s = String::from("🗻∈🌏");
318 /// let bytes = s.as_bytes_mut();
326 /// assert_eq!("🍔∈🌏", s);
328 #[stable(feature = "str_mut_extras", since = "1.20.0")]
330 pub unsafe fn as_bytes_mut(&mut self) -> &mut [u8] {
331 core_str::StrExt::as_bytes_mut(self)
334 /// Converts a string slice to a raw pointer.
336 /// As string slices are a slice of bytes, the raw pointer points to a
337 /// [`u8`]. This pointer will be pointing to the first byte of the string
340 /// [`u8`]: primitive.u8.html
348 /// let ptr = s.as_ptr();
350 #[stable(feature = "rust1", since = "1.0.0")]
352 pub fn as_ptr(&self) -> *const u8 {
353 core_str::StrExt::as_ptr(self)
356 /// Returns a subslice of `str`.
358 /// This is the non-panicking alternative to indexing the `str`. Returns
359 /// [`None`] whenever equivalent indexing operation would panic.
361 /// [`None`]: option/enum.Option.html#variant.None
366 /// let mut v = String::from("🗻∈🌏");
368 /// assert_eq!(Some("🗻"), v.get(0..4));
370 /// // indices not on UTF-8 sequence boundaries
371 /// assert!(v.get_mut(1..).is_none());
372 /// assert!(v.get_mut(..8).is_none());
375 /// assert!(v.get_mut(..42).is_none());
377 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
379 pub fn get<I: SliceIndex<str>>(&self, i: I) -> Option<&I::Output> {
380 core_str::StrExt::get(self, i)
383 /// Returns a mutable subslice of `str`.
385 /// This is the non-panicking alternative to indexing the `str`. Returns
386 /// [`None`] whenever equivalent indexing operation would panic.
388 /// [`None`]: option/enum.Option.html#variant.None
393 /// let mut v = String::from("🗻∈🌏");
395 /// assert_eq!(Some("🗻"), v.get_mut(0..4).map(|v| &*v));
397 /// // indices not on UTF-8 sequence boundaries
398 /// assert!(v.get_mut(1..).is_none());
399 /// assert!(v.get_mut(..8).is_none());
402 /// assert!(v.get_mut(..42).is_none());
404 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
406 pub fn get_mut<I: SliceIndex<str>>(&mut self, i: I) -> Option<&mut I::Output> {
407 core_str::StrExt::get_mut(self, i)
410 /// Returns a unchecked subslice of `str`.
412 /// This is the unchecked alternative to indexing the `str`.
416 /// Callers of this function are responsible that these preconditions are
419 /// * The starting index must come before the ending index;
420 /// * Indexes must be within bounds of the original slice;
421 /// * Indexes must lie on UTF-8 sequence boundaries.
423 /// Failing that, the returned string slice may reference invalid memory or
424 /// violate the invariants communicated by the `str` type.
431 /// assert_eq!("🗻", v.get_unchecked(0..4));
432 /// assert_eq!("∈", v.get_unchecked(4..7));
433 /// assert_eq!("🌏", v.get_unchecked(7..11));
436 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
438 pub unsafe fn get_unchecked<I: SliceIndex<str>>(&self, i: I) -> &I::Output {
439 core_str::StrExt::get_unchecked(self, i)
442 /// Returns a mutable, unchecked subslice of `str`.
444 /// This is the unchecked alternative to indexing the `str`.
448 /// Callers of this function are responsible that these preconditions are
451 /// * The starting index must come before the ending index;
452 /// * Indexes must be within bounds of the original slice;
453 /// * Indexes must lie on UTF-8 sequence boundaries.
455 /// Failing that, the returned string slice may reference invalid memory or
456 /// violate the invariants communicated by the `str` type.
461 /// let mut v = String::from("🗻∈🌏");
463 /// assert_eq!("🗻", v.get_unchecked_mut(0..4));
464 /// assert_eq!("∈", v.get_unchecked_mut(4..7));
465 /// assert_eq!("🌏", v.get_unchecked_mut(7..11));
468 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
470 pub unsafe fn get_unchecked_mut<I: SliceIndex<str>>(&mut self, i: I) -> &mut I::Output {
471 core_str::StrExt::get_unchecked_mut(self, i)
474 /// Creates a string slice from another string slice, bypassing safety
477 /// This is generally not recommended, use with caution! For a safe
478 /// alternative see [`str`] and [`Index`].
480 /// [`str`]: primitive.str.html
481 /// [`Index`]: ops/trait.Index.html
483 /// This new slice goes from `begin` to `end`, including `begin` but
486 /// To get a mutable string slice instead, see the
487 /// [`slice_mut_unchecked`] method.
489 /// [`slice_mut_unchecked`]: #method.slice_mut_unchecked
493 /// Callers of this function are responsible that three preconditions are
496 /// * `begin` must come before `end`.
497 /// * `begin` and `end` must be byte positions within the string slice.
498 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
505 /// let s = "Löwe 老虎 Léopard";
508 /// assert_eq!("Löwe 老虎 Léopard", s.slice_unchecked(0, 21));
511 /// let s = "Hello, world!";
514 /// assert_eq!("world", s.slice_unchecked(7, 12));
517 #[stable(feature = "rust1", since = "1.0.0")]
519 pub unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
520 core_str::StrExt::slice_unchecked(self, begin, end)
523 /// Creates a string slice from another string slice, bypassing safety
525 /// This is generally not recommended, use with caution! For a safe
526 /// alternative see [`str`] and [`IndexMut`].
528 /// [`str`]: primitive.str.html
529 /// [`IndexMut`]: ops/trait.IndexMut.html
531 /// This new slice goes from `begin` to `end`, including `begin` but
534 /// To get an immutable string slice instead, see the
535 /// [`slice_unchecked`] method.
537 /// [`slice_unchecked`]: #method.slice_unchecked
541 /// Callers of this function are responsible that three preconditions are
544 /// * `begin` must come before `end`.
545 /// * `begin` and `end` must be byte positions within the string slice.
546 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
547 #[stable(feature = "str_slice_mut", since = "1.5.0")]
549 pub unsafe fn slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str {
550 core_str::StrExt::slice_mut_unchecked(self, begin, end)
553 /// Divide one string slice into two at an index.
555 /// The argument, `mid`, should be a byte offset from the start of the
556 /// string. It must also be on the boundary of a UTF-8 code point.
558 /// The two slices returned go from the start of the string slice to `mid`,
559 /// and from `mid` to the end of the string slice.
561 /// To get mutable string slices instead, see the [`split_at_mut`]
564 /// [`split_at_mut`]: #method.split_at_mut
568 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
569 /// beyond the last code point of the string slice.
576 /// let s = "Per Martin-Löf";
578 /// let (first, last) = s.split_at(3);
580 /// assert_eq!("Per", first);
581 /// assert_eq!(" Martin-Löf", last);
584 #[stable(feature = "str_split_at", since = "1.4.0")]
585 pub fn split_at(&self, mid: usize) -> (&str, &str) {
586 core_str::StrExt::split_at(self, mid)
589 /// Divide one mutable string slice into two at an index.
591 /// The argument, `mid`, should be a byte offset from the start of the
592 /// string. It must also be on the boundary of a UTF-8 code point.
594 /// The two slices returned go from the start of the string slice to `mid`,
595 /// and from `mid` to the end of the string slice.
597 /// To get immutable string slices instead, see the [`split_at`] method.
599 /// [`split_at`]: #method.split_at
603 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
604 /// beyond the last code point of the string slice.
611 /// use std::ascii::AsciiExt;
613 /// let mut s = "Per Martin-Löf".to_string();
615 /// let (first, last) = s.split_at_mut(3);
616 /// first.make_ascii_uppercase();
617 /// assert_eq!("PER", first);
618 /// assert_eq!(" Martin-Löf", last);
620 /// assert_eq!("PER Martin-Löf", s);
623 #[stable(feature = "str_split_at", since = "1.4.0")]
624 pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {
625 core_str::StrExt::split_at_mut(self, mid)
628 /// Returns an iterator over the [`char`]s of a string slice.
630 /// As a string slice consists of valid UTF-8, we can iterate through a
631 /// string slice by [`char`]. This method returns such an iterator.
633 /// It's important to remember that [`char`] represents a Unicode Scalar
634 /// Value, and may not match your idea of what a 'character' is. Iteration
635 /// over grapheme clusters may be what you actually want.
637 /// [`char`]: primitive.char.html
644 /// let word = "goodbye";
646 /// let count = word.chars().count();
647 /// assert_eq!(7, count);
649 /// let mut chars = word.chars();
651 /// assert_eq!(Some('g'), chars.next());
652 /// assert_eq!(Some('o'), chars.next());
653 /// assert_eq!(Some('o'), chars.next());
654 /// assert_eq!(Some('d'), chars.next());
655 /// assert_eq!(Some('b'), chars.next());
656 /// assert_eq!(Some('y'), chars.next());
657 /// assert_eq!(Some('e'), chars.next());
659 /// assert_eq!(None, chars.next());
662 /// Remember, [`char`]s may not match your human intuition about characters:
667 /// let mut chars = y.chars();
669 /// assert_eq!(Some('y'), chars.next()); // not 'y̆'
670 /// assert_eq!(Some('\u{0306}'), chars.next());
672 /// assert_eq!(None, chars.next());
674 #[stable(feature = "rust1", since = "1.0.0")]
676 pub fn chars(&self) -> Chars {
677 core_str::StrExt::chars(self)
679 /// Returns an iterator over the [`char`]s of a string slice, and their
682 /// As a string slice consists of valid UTF-8, we can iterate through a
683 /// string slice by [`char`]. This method returns an iterator of both
684 /// these [`char`]s, as well as their byte positions.
686 /// The iterator yields tuples. The position is first, the [`char`] is
689 /// [`char`]: primitive.char.html
696 /// let word = "goodbye";
698 /// let count = word.char_indices().count();
699 /// assert_eq!(7, count);
701 /// let mut char_indices = word.char_indices();
703 /// assert_eq!(Some((0, 'g')), char_indices.next());
704 /// assert_eq!(Some((1, 'o')), char_indices.next());
705 /// assert_eq!(Some((2, 'o')), char_indices.next());
706 /// assert_eq!(Some((3, 'd')), char_indices.next());
707 /// assert_eq!(Some((4, 'b')), char_indices.next());
708 /// assert_eq!(Some((5, 'y')), char_indices.next());
709 /// assert_eq!(Some((6, 'e')), char_indices.next());
711 /// assert_eq!(None, char_indices.next());
714 /// Remember, [`char`]s may not match your human intuition about characters:
719 /// let mut char_indices = y.char_indices();
721 /// assert_eq!(Some((0, 'y')), char_indices.next()); // not (0, 'y̆')
722 /// assert_eq!(Some((1, '\u{0306}')), char_indices.next());
724 /// assert_eq!(None, char_indices.next());
726 #[stable(feature = "rust1", since = "1.0.0")]
728 pub fn char_indices(&self) -> CharIndices {
729 core_str::StrExt::char_indices(self)
732 /// An iterator over the bytes of a string slice.
734 /// As a string slice consists of a sequence of bytes, we can iterate
735 /// through a string slice by byte. This method returns such an iterator.
742 /// let mut bytes = "bors".bytes();
744 /// assert_eq!(Some(b'b'), bytes.next());
745 /// assert_eq!(Some(b'o'), bytes.next());
746 /// assert_eq!(Some(b'r'), bytes.next());
747 /// assert_eq!(Some(b's'), bytes.next());
749 /// assert_eq!(None, bytes.next());
751 #[stable(feature = "rust1", since = "1.0.0")]
753 pub fn bytes(&self) -> Bytes {
754 core_str::StrExt::bytes(self)
757 /// Split a string slice by whitespace.
759 /// The iterator returned will return string slices that are sub-slices of
760 /// the original string slice, separated by any amount of whitespace.
762 /// 'Whitespace' is defined according to the terms of the Unicode Derived
763 /// Core Property `White_Space`.
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 UnicodeStr::split_whitespace(self)
797 /// An iterator over the lines of a string, as string slices.
799 /// Lines are ended with either a newline (`\n`) or a carriage return with
800 /// a line feed (`\r\n`).
802 /// The final line ending is optional.
809 /// let text = "foo\r\nbar\n\nbaz\n";
810 /// let mut lines = text.lines();
812 /// assert_eq!(Some("foo"), lines.next());
813 /// assert_eq!(Some("bar"), lines.next());
814 /// assert_eq!(Some(""), lines.next());
815 /// assert_eq!(Some("baz"), lines.next());
817 /// assert_eq!(None, lines.next());
820 /// The final line ending isn't required:
823 /// let text = "foo\nbar\n\r\nbaz";
824 /// let mut lines = text.lines();
826 /// assert_eq!(Some("foo"), lines.next());
827 /// assert_eq!(Some("bar"), lines.next());
828 /// assert_eq!(Some(""), lines.next());
829 /// assert_eq!(Some("baz"), lines.next());
831 /// assert_eq!(None, lines.next());
833 #[stable(feature = "rust1", since = "1.0.0")]
835 pub fn lines(&self) -> Lines {
836 core_str::StrExt::lines(self)
839 /// An iterator over the lines of a string.
840 #[stable(feature = "rust1", since = "1.0.0")]
841 #[rustc_deprecated(since = "1.4.0", reason = "use lines() instead now")]
844 pub fn lines_any(&self) -> LinesAny {
845 core_str::StrExt::lines_any(self)
848 /// Returns an iterator of `u16` over the string encoded as UTF-16.
849 #[stable(feature = "encode_utf16", since = "1.8.0")]
850 pub fn encode_utf16(&self) -> EncodeUtf16 {
851 EncodeUtf16 { encoder: Utf16Encoder::new(self[..].chars()) }
854 /// Returns `true` if the given pattern matches a sub-slice of
855 /// this string slice.
857 /// Returns `false` if it does not.
864 /// let bananas = "bananas";
866 /// assert!(bananas.contains("nana"));
867 /// assert!(!bananas.contains("apples"));
869 #[stable(feature = "rust1", since = "1.0.0")]
871 pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
872 core_str::StrExt::contains(self, pat)
875 /// Returns `true` if the given pattern matches a prefix of this
878 /// Returns `false` if it does not.
885 /// let bananas = "bananas";
887 /// assert!(bananas.starts_with("bana"));
888 /// assert!(!bananas.starts_with("nana"));
890 #[stable(feature = "rust1", since = "1.0.0")]
891 pub fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
892 core_str::StrExt::starts_with(self, pat)
895 /// Returns `true` if the given pattern matches a suffix of this
898 /// Returns `false` if it does not.
905 /// let bananas = "bananas";
907 /// assert!(bananas.ends_with("anas"));
908 /// assert!(!bananas.ends_with("nana"));
910 #[stable(feature = "rust1", since = "1.0.0")]
911 pub fn ends_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool
912 where P::Searcher: ReverseSearcher<'a>
914 core_str::StrExt::ends_with(self, pat)
917 /// Returns the byte index of the first character of this string slice that
918 /// matches the pattern.
920 /// Returns [`None`] if the pattern doesn't match.
922 /// The pattern can be a `&str`, [`char`], or a closure that determines if
923 /// a character matches.
925 /// [`char`]: primitive.char.html
926 /// [`None`]: option/enum.Option.html#variant.None
933 /// let s = "Löwe 老虎 Léopard";
935 /// assert_eq!(s.find('L'), Some(0));
936 /// assert_eq!(s.find('é'), Some(14));
937 /// assert_eq!(s.find("Léopard"), Some(13));
940 /// More complex patterns with closures:
943 /// let s = "Löwe 老虎 Léopard";
945 /// assert_eq!(s.find(char::is_whitespace), Some(5));
946 /// assert_eq!(s.find(char::is_lowercase), Some(1));
949 /// Not finding the pattern:
952 /// let s = "Löwe 老虎 Léopard";
953 /// let x: &[_] = &['1', '2'];
955 /// assert_eq!(s.find(x), None);
957 #[stable(feature = "rust1", since = "1.0.0")]
959 pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
960 core_str::StrExt::find(self, pat)
963 /// Returns the byte index of the last character of this string slice that
964 /// matches the pattern.
966 /// Returns [`None`] if the pattern doesn't match.
968 /// The pattern can be a `&str`, [`char`], or a closure that determines if
969 /// a character matches.
971 /// [`char`]: primitive.char.html
972 /// [`None`]: option/enum.Option.html#variant.None
979 /// let s = "Löwe 老虎 Léopard";
981 /// assert_eq!(s.rfind('L'), Some(13));
982 /// assert_eq!(s.rfind('é'), Some(14));
985 /// More complex patterns with closures:
988 /// let s = "Löwe 老虎 Léopard";
990 /// assert_eq!(s.rfind(char::is_whitespace), Some(12));
991 /// assert_eq!(s.rfind(char::is_lowercase), Some(20));
994 /// Not finding the pattern:
997 /// let s = "Löwe 老虎 Léopard";
998 /// let x: &[_] = &['1', '2'];
1000 /// assert_eq!(s.rfind(x), None);
1002 #[stable(feature = "rust1", since = "1.0.0")]
1004 pub fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
1005 where P::Searcher: ReverseSearcher<'a>
1007 core_str::StrExt::rfind(self, pat)
1010 /// An iterator over substrings of this string slice, separated by
1011 /// characters matched by a pattern.
1013 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1016 /// # Iterator behavior
1018 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1019 /// allows a reverse search and forward/reverse search yields the same
1020 /// elements. This is true for, eg, [`char`] but not for `&str`.
1022 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1024 /// If the pattern allows a reverse search but its results might differ
1025 /// from a forward search, the [`rsplit`] method can be used.
1027 /// [`char`]: primitive.char.html
1028 /// [`rsplit`]: #method.rsplit
1032 /// Simple patterns:
1035 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
1036 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
1038 /// let v: Vec<&str> = "".split('X').collect();
1039 /// assert_eq!(v, [""]);
1041 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
1042 /// assert_eq!(v, ["lion", "", "tiger", "leopard"]);
1044 /// let v: Vec<&str> = "lion::tiger::leopard".split("::").collect();
1045 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1047 /// let v: Vec<&str> = "abc1def2ghi".split(char::is_numeric).collect();
1048 /// assert_eq!(v, ["abc", "def", "ghi"]);
1050 /// let v: Vec<&str> = "lionXtigerXleopard".split(char::is_uppercase).collect();
1051 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1054 /// A more complex pattern, using a closure:
1057 /// let v: Vec<&str> = "abc1defXghi".split(|c| c == '1' || c == 'X').collect();
1058 /// assert_eq!(v, ["abc", "def", "ghi"]);
1061 /// If a string contains multiple contiguous separators, you will end up
1062 /// with empty strings in the output:
1065 /// let x = "||||a||b|c".to_string();
1066 /// let d: Vec<_> = x.split('|').collect();
1068 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1071 /// Contiguous separators are separated by the empty string.
1074 /// let x = "(///)".to_string();
1075 /// let d: Vec<_> = x.split('/').collect();
1077 /// assert_eq!(d, &["(", "", "", ")"]);
1080 /// Separators at the start or end of a string are neighbored
1081 /// by empty strings.
1084 /// let d: Vec<_> = "010".split("0").collect();
1085 /// assert_eq!(d, &["", "1", ""]);
1088 /// When the empty string is used as a separator, it separates
1089 /// every character in the string, along with the beginning
1090 /// and end of the string.
1093 /// let f: Vec<_> = "rust".split("").collect();
1094 /// assert_eq!(f, &["", "r", "u", "s", "t", ""]);
1097 /// Contiguous separators can lead to possibly surprising behavior
1098 /// when whitespace is used as the separator. This code is correct:
1101 /// let x = " a b c".to_string();
1102 /// let d: Vec<_> = x.split(' ').collect();
1104 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1107 /// It does _not_ give you:
1110 /// assert_eq!(d, &["a", "b", "c"]);
1113 /// Use [`split_whitespace`] for this behavior.
1115 /// [`split_whitespace`]: #method.split_whitespace
1116 #[stable(feature = "rust1", since = "1.0.0")]
1118 pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
1119 core_str::StrExt::split(self, pat)
1122 /// An iterator over substrings of the given string slice, separated by
1123 /// characters matched by a pattern and yielded in reverse order.
1125 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1128 /// [`char`]: primitive.char.html
1130 /// # Iterator behavior
1132 /// The returned iterator requires that the pattern supports a reverse
1133 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1134 /// search yields the same elements.
1136 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1138 /// For iterating from the front, the [`split`] method can be used.
1140 /// [`split`]: #method.split
1144 /// Simple patterns:
1147 /// let v: Vec<&str> = "Mary had a little lamb".rsplit(' ').collect();
1148 /// assert_eq!(v, ["lamb", "little", "a", "had", "Mary"]);
1150 /// let v: Vec<&str> = "".rsplit('X').collect();
1151 /// assert_eq!(v, [""]);
1153 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplit('X').collect();
1154 /// assert_eq!(v, ["leopard", "tiger", "", "lion"]);
1156 /// let v: Vec<&str> = "lion::tiger::leopard".rsplit("::").collect();
1157 /// assert_eq!(v, ["leopard", "tiger", "lion"]);
1160 /// A more complex pattern, using a closure:
1163 /// let v: Vec<&str> = "abc1defXghi".rsplit(|c| c == '1' || c == 'X').collect();
1164 /// assert_eq!(v, ["ghi", "def", "abc"]);
1166 #[stable(feature = "rust1", since = "1.0.0")]
1168 pub fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
1169 where P::Searcher: ReverseSearcher<'a>
1171 core_str::StrExt::rsplit(self, pat)
1174 /// An iterator over substrings of the given string slice, separated by
1175 /// characters matched by a pattern.
1177 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1180 /// Equivalent to [`split`], except that the trailing substring
1181 /// is skipped if empty.
1183 /// [`split`]: #method.split
1185 /// This method can be used for string data that is _terminated_,
1186 /// rather than _separated_ by a pattern.
1188 /// # Iterator behavior
1190 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1191 /// allows a reverse search and forward/reverse search yields the same
1192 /// elements. This is true for, eg, [`char`] but not for `&str`.
1194 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1195 /// [`char`]: primitive.char.html
1197 /// If the pattern allows a reverse search but its results might differ
1198 /// from a forward search, the [`rsplit_terminator`] method can be used.
1200 /// [`rsplit_terminator`]: #method.rsplit_terminator
1207 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
1208 /// assert_eq!(v, ["A", "B"]);
1210 /// let v: Vec<&str> = "A..B..".split_terminator(".").collect();
1211 /// assert_eq!(v, ["A", "", "B", ""]);
1213 #[stable(feature = "rust1", since = "1.0.0")]
1215 pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
1216 core_str::StrExt::split_terminator(self, pat)
1219 /// An iterator over substrings of `self`, separated by characters
1220 /// matched by a pattern and yielded in reverse order.
1222 /// The pattern can be a simple `&str`, [`char`], or a closure that
1223 /// determines the split.
1224 /// Additional libraries might provide more complex patterns like
1225 /// regular expressions.
1227 /// [`char`]: primitive.char.html
1229 /// Equivalent to [`split`], except that the trailing substring is
1230 /// skipped if empty.
1232 /// [`split`]: #method.split
1234 /// This method can be used for string data that is _terminated_,
1235 /// rather than _separated_ by a pattern.
1237 /// # Iterator behavior
1239 /// The returned iterator requires that the pattern supports a
1240 /// reverse search, and it will be double ended if a forward/reverse
1241 /// search yields the same elements.
1243 /// For iterating from the front, the [`split_terminator`] method can be
1246 /// [`split_terminator`]: #method.split_terminator
1251 /// let v: Vec<&str> = "A.B.".rsplit_terminator('.').collect();
1252 /// assert_eq!(v, ["B", "A"]);
1254 /// let v: Vec<&str> = "A..B..".rsplit_terminator(".").collect();
1255 /// assert_eq!(v, ["", "B", "", "A"]);
1257 #[stable(feature = "rust1", since = "1.0.0")]
1259 pub fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1260 where P::Searcher: ReverseSearcher<'a>
1262 core_str::StrExt::rsplit_terminator(self, pat)
1265 /// An iterator over substrings of the given string slice, separated by a
1266 /// pattern, restricted to returning at most `n` items.
1268 /// If `n` substrings are returned, the last substring (the `n`th substring)
1269 /// will contain the remainder of the string.
1271 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1274 /// [`char`]: primitive.char.html
1276 /// # Iterator behavior
1278 /// The returned iterator will not be double ended, because it is
1279 /// not efficient to support.
1281 /// If the pattern allows a reverse search, the [`rsplitn`] method can be
1284 /// [`rsplitn`]: #method.rsplitn
1288 /// Simple patterns:
1291 /// let v: Vec<&str> = "Mary had a little lambda".splitn(3, ' ').collect();
1292 /// assert_eq!(v, ["Mary", "had", "a little lambda"]);
1294 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(3, "X").collect();
1295 /// assert_eq!(v, ["lion", "", "tigerXleopard"]);
1297 /// let v: Vec<&str> = "abcXdef".splitn(1, 'X').collect();
1298 /// assert_eq!(v, ["abcXdef"]);
1300 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1301 /// assert_eq!(v, [""]);
1304 /// A more complex pattern, using a closure:
1307 /// let v: Vec<&str> = "abc1defXghi".splitn(2, |c| c == '1' || c == 'X').collect();
1308 /// assert_eq!(v, ["abc", "defXghi"]);
1310 #[stable(feature = "rust1", since = "1.0.0")]
1312 pub fn splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P> {
1313 core_str::StrExt::splitn(self, n, pat)
1316 /// An iterator over substrings of this string slice, separated by a
1317 /// pattern, starting from the end of the string, restricted to returning
1318 /// at most `n` items.
1320 /// If `n` substrings are returned, the last substring (the `n`th substring)
1321 /// will contain the remainder of the string.
1323 /// The pattern can be a `&str`, [`char`], or a closure that
1324 /// determines the split.
1326 /// [`char`]: primitive.char.html
1328 /// # Iterator behavior
1330 /// The returned iterator will not be double ended, because it is not
1331 /// efficient to support.
1333 /// For splitting from the front, the [`splitn`] method can be used.
1335 /// [`splitn`]: #method.splitn
1339 /// Simple patterns:
1342 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(3, ' ').collect();
1343 /// assert_eq!(v, ["lamb", "little", "Mary had a"]);
1345 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(3, 'X').collect();
1346 /// assert_eq!(v, ["leopard", "tiger", "lionX"]);
1348 /// let v: Vec<&str> = "lion::tiger::leopard".rsplitn(2, "::").collect();
1349 /// assert_eq!(v, ["leopard", "lion::tiger"]);
1352 /// A more complex pattern, using a closure:
1355 /// let v: Vec<&str> = "abc1defXghi".rsplitn(2, |c| c == '1' || c == 'X').collect();
1356 /// assert_eq!(v, ["ghi", "abc1def"]);
1358 #[stable(feature = "rust1", since = "1.0.0")]
1360 pub fn rsplitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> RSplitN<'a, P>
1361 where P::Searcher: ReverseSearcher<'a>
1363 core_str::StrExt::rsplitn(self, n, pat)
1366 /// An iterator over the disjoint matches of a pattern within the given string
1369 /// The pattern can be a `&str`, [`char`], or a closure that
1370 /// determines if a character matches.
1372 /// [`char`]: primitive.char.html
1374 /// # Iterator behavior
1376 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1377 /// allows a reverse search and forward/reverse search yields the same
1378 /// elements. This is true for, eg, [`char`] but not for `&str`.
1380 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1381 /// [`char`]: primitive.char.html
1383 /// If the pattern allows a reverse search but its results might differ
1384 /// from a forward search, the [`rmatches`] method can be used.
1386 /// [`rmatches`]: #method.rmatches
1393 /// let v: Vec<&str> = "abcXXXabcYYYabc".matches("abc").collect();
1394 /// assert_eq!(v, ["abc", "abc", "abc"]);
1396 /// let v: Vec<&str> = "1abc2abc3".matches(char::is_numeric).collect();
1397 /// assert_eq!(v, ["1", "2", "3"]);
1399 #[stable(feature = "str_matches", since = "1.2.0")]
1401 pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1402 core_str::StrExt::matches(self, pat)
1405 /// An iterator over the disjoint matches of a pattern within this string slice,
1406 /// yielded in reverse order.
1408 /// The pattern can be a `&str`, [`char`], or a closure that determines if
1409 /// a character matches.
1411 /// [`char`]: primitive.char.html
1413 /// # Iterator behavior
1415 /// The returned iterator requires that the pattern supports a reverse
1416 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1417 /// search yields the same elements.
1419 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1421 /// For iterating from the front, the [`matches`] method can be used.
1423 /// [`matches`]: #method.matches
1430 /// let v: Vec<&str> = "abcXXXabcYYYabc".rmatches("abc").collect();
1431 /// assert_eq!(v, ["abc", "abc", "abc"]);
1433 /// let v: Vec<&str> = "1abc2abc3".rmatches(char::is_numeric).collect();
1434 /// assert_eq!(v, ["3", "2", "1"]);
1436 #[stable(feature = "str_matches", since = "1.2.0")]
1438 pub fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
1439 where P::Searcher: ReverseSearcher<'a>
1441 core_str::StrExt::rmatches(self, pat)
1444 /// An iterator over the disjoint matches of a pattern within this string
1445 /// slice as well as the index that the match starts at.
1447 /// For matches of `pat` within `self` that overlap, only the indices
1448 /// corresponding to the first match are returned.
1450 /// The pattern can be a `&str`, [`char`], or a closure that determines
1451 /// if a character matches.
1453 /// [`char`]: primitive.char.html
1455 /// # Iterator behavior
1457 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1458 /// allows a reverse search and forward/reverse search yields the same
1459 /// elements. This is true for, eg, [`char`] but not for `&str`.
1461 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1463 /// If the pattern allows a reverse search but its results might differ
1464 /// from a forward search, the [`rmatch_indices`] method can be used.
1466 /// [`rmatch_indices`]: #method.rmatch_indices
1473 /// let v: Vec<_> = "abcXXXabcYYYabc".match_indices("abc").collect();
1474 /// assert_eq!(v, [(0, "abc"), (6, "abc"), (12, "abc")]);
1476 /// let v: Vec<_> = "1abcabc2".match_indices("abc").collect();
1477 /// assert_eq!(v, [(1, "abc"), (4, "abc")]);
1479 /// let v: Vec<_> = "ababa".match_indices("aba").collect();
1480 /// assert_eq!(v, [(0, "aba")]); // only the first `aba`
1482 #[stable(feature = "str_match_indices", since = "1.5.0")]
1484 pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1485 core_str::StrExt::match_indices(self, pat)
1488 /// An iterator over the disjoint matches of a pattern within `self`,
1489 /// yielded in reverse order along with the index of the match.
1491 /// For matches of `pat` within `self` that overlap, only the indices
1492 /// corresponding to the last match are returned.
1494 /// The pattern can be a `&str`, [`char`], or a closure that determines if a
1495 /// character matches.
1497 /// [`char`]: primitive.char.html
1499 /// # Iterator behavior
1501 /// The returned iterator requires that the pattern supports a reverse
1502 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1503 /// search yields the same elements.
1505 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1507 /// For iterating from the front, the [`match_indices`] method can be used.
1509 /// [`match_indices`]: #method.match_indices
1516 /// let v: Vec<_> = "abcXXXabcYYYabc".rmatch_indices("abc").collect();
1517 /// assert_eq!(v, [(12, "abc"), (6, "abc"), (0, "abc")]);
1519 /// let v: Vec<_> = "1abcabc2".rmatch_indices("abc").collect();
1520 /// assert_eq!(v, [(4, "abc"), (1, "abc")]);
1522 /// let v: Vec<_> = "ababa".rmatch_indices("aba").collect();
1523 /// assert_eq!(v, [(2, "aba")]); // only the last `aba`
1525 #[stable(feature = "str_match_indices", since = "1.5.0")]
1527 pub fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
1528 where P::Searcher: ReverseSearcher<'a>
1530 core_str::StrExt::rmatch_indices(self, pat)
1533 /// Returns a string slice with leading and trailing whitespace removed.
1535 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1536 /// Core Property `White_Space`.
1543 /// let s = " Hello\tworld\t";
1545 /// assert_eq!("Hello\tworld", s.trim());
1547 #[stable(feature = "rust1", since = "1.0.0")]
1548 pub fn trim(&self) -> &str {
1549 UnicodeStr::trim(self)
1552 /// Returns a string slice with leading whitespace removed.
1554 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1555 /// Core Property `White_Space`.
1557 /// # Text directionality
1559 /// A string is a sequence of bytes. 'Left' in this context means the first
1560 /// position of that byte string; for a language like Arabic or Hebrew
1561 /// which are 'right to left' rather than 'left to right', this will be
1562 /// the _right_ side, not the left.
1569 /// let s = " Hello\tworld\t";
1571 /// assert_eq!("Hello\tworld\t", s.trim_left());
1577 /// let s = " English";
1578 /// assert!(Some('E') == s.trim_left().chars().next());
1580 /// let s = " עברית";
1581 /// assert!(Some('ע') == s.trim_left().chars().next());
1583 #[stable(feature = "rust1", since = "1.0.0")]
1584 pub fn trim_left(&self) -> &str {
1585 UnicodeStr::trim_left(self)
1588 /// Returns a string slice with trailing whitespace removed.
1590 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1591 /// Core Property `White_Space`.
1593 /// # Text directionality
1595 /// A string is a sequence of bytes. 'Right' in this context means the last
1596 /// position of that byte string; for a language like Arabic or Hebrew
1597 /// which are 'right to left' rather than 'left to right', this will be
1598 /// the _left_ side, not the right.
1605 /// let s = " Hello\tworld\t";
1607 /// assert_eq!(" Hello\tworld", s.trim_right());
1613 /// let s = "English ";
1614 /// assert!(Some('h') == s.trim_right().chars().rev().next());
1616 /// let s = "עברית ";
1617 /// assert!(Some('ת') == s.trim_right().chars().rev().next());
1619 #[stable(feature = "rust1", since = "1.0.0")]
1620 pub fn trim_right(&self) -> &str {
1621 UnicodeStr::trim_right(self)
1624 /// Returns a string slice with all prefixes and suffixes that match a
1625 /// pattern repeatedly removed.
1627 /// The pattern can be a [`char`] or a closure that determines if a
1628 /// character matches.
1630 /// [`char`]: primitive.char.html
1634 /// Simple patterns:
1637 /// assert_eq!("11foo1bar11".trim_matches('1'), "foo1bar");
1638 /// assert_eq!("123foo1bar123".trim_matches(char::is_numeric), "foo1bar");
1640 /// let x: &[_] = &['1', '2'];
1641 /// assert_eq!("12foo1bar12".trim_matches(x), "foo1bar");
1644 /// A more complex pattern, using a closure:
1647 /// assert_eq!("1foo1barXX".trim_matches(|c| c == '1' || c == 'X'), "foo1bar");
1649 #[stable(feature = "rust1", since = "1.0.0")]
1650 pub fn trim_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1651 where P::Searcher: DoubleEndedSearcher<'a>
1653 core_str::StrExt::trim_matches(self, pat)
1656 /// Returns a string slice with all prefixes that match a pattern
1657 /// repeatedly removed.
1659 /// The pattern can be a `&str`, [`char`], or a closure that determines if
1660 /// a character matches.
1662 /// [`char`]: primitive.char.html
1664 /// # Text directionality
1666 /// A string is a sequence of bytes. 'Left' in this context means the first
1667 /// position of that byte string; for a language like Arabic or Hebrew
1668 /// which are 'right to left' rather than 'left to right', this will be
1669 /// the _right_ side, not the left.
1676 /// assert_eq!("11foo1bar11".trim_left_matches('1'), "foo1bar11");
1677 /// assert_eq!("123foo1bar123".trim_left_matches(char::is_numeric), "foo1bar123");
1679 /// let x: &[_] = &['1', '2'];
1680 /// assert_eq!("12foo1bar12".trim_left_matches(x), "foo1bar12");
1682 #[stable(feature = "rust1", since = "1.0.0")]
1683 pub fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
1684 core_str::StrExt::trim_left_matches(self, pat)
1687 /// Returns a string slice with all suffixes that match a pattern
1688 /// repeatedly removed.
1690 /// The pattern can be a `&str`, [`char`], or a closure that
1691 /// determines if a character matches.
1693 /// [`char`]: primitive.char.html
1695 /// # Text directionality
1697 /// A string is a sequence of bytes. 'Right' in this context means the last
1698 /// position of that byte string; for a language like Arabic or Hebrew
1699 /// which are 'right to left' rather than 'left to right', this will be
1700 /// the _left_ side, not the right.
1704 /// Simple patterns:
1707 /// assert_eq!("11foo1bar11".trim_right_matches('1'), "11foo1bar");
1708 /// assert_eq!("123foo1bar123".trim_right_matches(char::is_numeric), "123foo1bar");
1710 /// let x: &[_] = &['1', '2'];
1711 /// assert_eq!("12foo1bar12".trim_right_matches(x), "12foo1bar");
1714 /// A more complex pattern, using a closure:
1717 /// assert_eq!("1fooX".trim_left_matches(|c| c == '1' || c == 'X'), "fooX");
1719 #[stable(feature = "rust1", since = "1.0.0")]
1720 pub fn trim_right_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1721 where P::Searcher: ReverseSearcher<'a>
1723 core_str::StrExt::trim_right_matches(self, pat)
1726 /// Parses this string slice into another type.
1728 /// Because `parse` is so general, it can cause problems with type
1729 /// inference. As such, `parse` is one of the few times you'll see
1730 /// the syntax affectionately known as the 'turbofish': `::<>`. This
1731 /// helps the inference algorithm understand specifically which type
1732 /// you're trying to parse into.
1734 /// `parse` can parse any type that implements the [`FromStr`] trait.
1736 /// [`FromStr`]: str/trait.FromStr.html
1740 /// Will return [`Err`] if it's not possible to parse this string slice into
1741 /// the desired type.
1743 /// [`Err`]: str/trait.FromStr.html#associatedtype.Err
1750 /// let four: u32 = "4".parse().unwrap();
1752 /// assert_eq!(4, four);
1755 /// Using the 'turbofish' instead of annotating `four`:
1758 /// let four = "4".parse::<u32>();
1760 /// assert_eq!(Ok(4), four);
1763 /// Failing to parse:
1766 /// let nope = "j".parse::<u32>();
1768 /// assert!(nope.is_err());
1771 #[stable(feature = "rust1", since = "1.0.0")]
1772 pub fn parse<F: FromStr>(&self) -> Result<F, F::Err> {
1773 core_str::StrExt::parse(self)
1776 /// Converts a `Box<str>` into a `Box<[u8]>` without copying or allocating.
1777 #[stable(feature = "str_box_extras", since = "1.20.0")]
1778 pub fn into_boxed_bytes(self: Box<str>) -> Box<[u8]> {
1782 /// Replaces all matches of a pattern with another string.
1784 /// `replace` creates a new [`String`], and copies the data from this string slice into it.
1785 /// While doing so, it attempts to find matches of a pattern. If it finds any, it
1786 /// replaces them with the replacement string slice.
1788 /// [`String`]: string/struct.String.html
1795 /// let s = "this is old";
1797 /// assert_eq!("this is new", s.replace("old", "new"));
1800 /// When the pattern doesn't match:
1803 /// let s = "this is old";
1804 /// assert_eq!(s, s.replace("cookie monster", "little lamb"));
1806 #[stable(feature = "rust1", since = "1.0.0")]
1808 pub fn replace<'a, P: Pattern<'a>>(&'a self, from: P, to: &str) -> String {
1809 let mut result = String::new();
1810 let mut last_end = 0;
1811 for (start, part) in self.match_indices(from) {
1812 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1813 result.push_str(to);
1814 last_end = start + part.len();
1816 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1820 /// Replaces first N matches of a pattern with another string.
1822 /// `replacen` creates a new [`String`], and copies the data from this string slice into it.
1823 /// While doing so, it attempts to find matches of a pattern. If it finds any, it
1824 /// replaces them with the replacement string slice at most `count` times.
1826 /// [`String`]: string/struct.String.html
1833 /// let s = "foo foo 123 foo";
1834 /// assert_eq!("new new 123 foo", s.replacen("foo", "new", 2));
1835 /// assert_eq!("faa fao 123 foo", s.replacen('o', "a", 3));
1836 /// assert_eq!("foo foo new23 foo", s.replacen(char::is_numeric, "new", 1));
1839 /// When the pattern doesn't match:
1842 /// let s = "this is old";
1843 /// assert_eq!(s, s.replacen("cookie monster", "little lamb", 10));
1845 #[stable(feature = "str_replacen", since = "1.16.0")]
1846 pub fn replacen<'a, P: Pattern<'a>>(&'a self, pat: P, to: &str, count: usize) -> String {
1847 // Hope to reduce the times of re-allocation
1848 let mut result = String::with_capacity(32);
1849 let mut last_end = 0;
1850 for (start, part) in self.match_indices(pat).take(count) {
1851 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1852 result.push_str(to);
1853 last_end = start + part.len();
1855 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1859 /// Returns the lowercase equivalent of this string slice, as a new [`String`].
1861 /// 'Lowercase' is defined according to the terms of the Unicode Derived Core Property
1864 /// Since some characters can expand into multiple characters when changing
1865 /// the case, this function returns a [`String`] instead of modifying the
1866 /// parameter in-place.
1868 /// [`String`]: string/struct.String.html
1875 /// let s = "HELLO";
1877 /// assert_eq!("hello", s.to_lowercase());
1880 /// A tricky example, with sigma:
1883 /// let sigma = "Σ";
1885 /// assert_eq!("σ", sigma.to_lowercase());
1887 /// // but at the end of a word, it's ς, not σ:
1888 /// let odysseus = "ὈΔΥΣΣΕΎΣ";
1890 /// assert_eq!("ὀδυσσεύς", odysseus.to_lowercase());
1893 /// Languages without case are not changed:
1896 /// let new_year = "农历新年";
1898 /// assert_eq!(new_year, new_year.to_lowercase());
1900 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
1901 pub fn to_lowercase(&self) -> String {
1902 let mut s = String::with_capacity(self.len());
1903 for (i, c) in self[..].char_indices() {
1905 // Σ maps to σ, except at the end of a word where it maps to ς.
1906 // This is the only conditional (contextual) but language-independent mapping
1907 // in `SpecialCasing.txt`,
1908 // so hard-code it rather than have a generic "condition" mechanism.
1909 // See https://github.com/rust-lang/rust/issues/26035
1910 map_uppercase_sigma(self, i, &mut s)
1912 s.extend(c.to_lowercase());
1917 fn map_uppercase_sigma(from: &str, i: usize, to: &mut String) {
1918 // See http://www.unicode.org/versions/Unicode7.0.0/ch03.pdf#G33992
1919 // for the definition of `Final_Sigma`.
1920 debug_assert!('Σ'.len_utf8() == 2);
1921 let is_word_final = case_ignoreable_then_cased(from[..i].chars().rev()) &&
1922 !case_ignoreable_then_cased(from[i + 2..].chars());
1923 to.push_str(if is_word_final { "ς" } else { "σ" });
1926 fn case_ignoreable_then_cased<I: Iterator<Item = char>>(iter: I) -> bool {
1927 use std_unicode::derived_property::{Cased, Case_Ignorable};
1928 match iter.skip_while(|&c| Case_Ignorable(c)).next() {
1929 Some(c) => Cased(c),
1935 /// Returns the uppercase equivalent of this string slice, as a new [`String`].
1937 /// 'Uppercase' is defined according to the terms of the Unicode Derived Core Property
1940 /// Since some characters can expand into multiple characters when changing
1941 /// the case, this function returns a [`String`] instead of modifying the
1942 /// parameter in-place.
1944 /// [`String`]: string/struct.String.html
1951 /// let s = "hello";
1953 /// assert_eq!("HELLO", s.to_uppercase());
1956 /// Scripts without case are not changed:
1959 /// let new_year = "农历新年";
1961 /// assert_eq!(new_year, new_year.to_uppercase());
1963 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
1964 pub fn to_uppercase(&self) -> String {
1965 let mut s = String::with_capacity(self.len());
1966 s.extend(self.chars().flat_map(|c| c.to_uppercase()));
1970 /// Escapes each char in `s` with [`char::escape_debug`].
1972 /// [`char::escape_debug`]: primitive.char.html#method.escape_debug
1973 #[unstable(feature = "str_escape",
1974 reason = "return type may change to be an iterator",
1976 pub fn escape_debug(&self) -> String {
1977 self.chars().flat_map(|c| c.escape_debug()).collect()
1980 /// Escapes each char in `s` with [`char::escape_default`].
1982 /// [`char::escape_default`]: primitive.char.html#method.escape_default
1983 #[unstable(feature = "str_escape",
1984 reason = "return type may change to be an iterator",
1986 pub fn escape_default(&self) -> String {
1987 self.chars().flat_map(|c| c.escape_default()).collect()
1990 /// Escapes each char in `s` with [`char::escape_unicode`].
1992 /// [`char::escape_unicode`]: primitive.char.html#method.escape_unicode
1993 #[unstable(feature = "str_escape",
1994 reason = "return type may change to be an iterator",
1996 pub fn escape_unicode(&self) -> String {
1997 self.chars().flat_map(|c| c.escape_unicode()).collect()
2000 /// Converts a [`Box<str>`] into a [`String`] without copying or allocating.
2002 /// [`String`]: string/struct.String.html
2003 /// [`Box<str>`]: boxed/struct.Box.html
2010 /// let string = String::from("birthday gift");
2011 /// let boxed_str = string.clone().into_boxed_str();
2013 /// assert_eq!(boxed_str.into_string(), string);
2015 #[stable(feature = "box_str", since = "1.4.0")]
2016 pub fn into_string(self: Box<str>) -> String {
2018 let slice = mem::transmute::<Box<str>, Box<[u8]>>(self);
2019 String::from_utf8_unchecked(slice.into_vec())
2023 /// Create a [`String`] by repeating a string `n` times.
2025 /// [`String`]: string/struct.String.html
2032 /// assert_eq!("abc".repeat(4), String::from("abcabcabcabc"));
2034 #[stable(feature = "repeat_str", since = "1.16.0")]
2035 pub fn repeat(&self, n: usize) -> String {
2036 let mut s = String::with_capacity(self.len() * n);
2037 s.extend((0..n).map(|_| self));
2042 /// Converts a boxed slice of bytes to a boxed string slice without checking
2043 /// that the string contains valid UTF-8.
2044 #[stable(feature = "str_box_extras", since = "1.20.0")]
2045 pub unsafe fn from_boxed_utf8_unchecked(v: Box<[u8]>) -> Box<str> {