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 v = String::from("🗻∈🌏");
368 /// assert_eq!(Some("🗻"), v.get(0..4));
370 /// // indices not on UTF-8 sequence boundaries
371 /// assert!(v.get(1..).is_none());
372 /// assert!(v.get(..8).is_none());
375 /// assert!(v.get(..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("hello");
394 /// // correct length
395 /// assert!(v.get_mut(0..5).is_some());
397 /// assert!(v.get_mut(..42).is_none());
398 /// assert_eq!(Some("he"), v.get_mut(0..2).map(|v| &*v));
400 /// assert_eq!("hello", v);
402 /// let s = v.get_mut(0..2);
403 /// let s = s.map(|s| {
404 /// s.make_ascii_uppercase();
407 /// assert_eq!(Some("HE"), s);
409 /// assert_eq!("HEllo", v);
411 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
413 pub fn get_mut<I: SliceIndex<str>>(&mut self, i: I) -> Option<&mut I::Output> {
414 core_str::StrExt::get_mut(self, i)
417 /// Returns a unchecked subslice of `str`.
419 /// This is the unchecked alternative to indexing the `str`.
423 /// Callers of this function are responsible that these preconditions are
426 /// * The starting index must come before the ending index;
427 /// * Indexes must be within bounds of the original slice;
428 /// * Indexes must lie on UTF-8 sequence boundaries.
430 /// Failing that, the returned string slice may reference invalid memory or
431 /// violate the invariants communicated by the `str` type.
438 /// assert_eq!("🗻", v.get_unchecked(0..4));
439 /// assert_eq!("∈", v.get_unchecked(4..7));
440 /// assert_eq!("🌏", v.get_unchecked(7..11));
443 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
445 pub unsafe fn get_unchecked<I: SliceIndex<str>>(&self, i: I) -> &I::Output {
446 core_str::StrExt::get_unchecked(self, i)
449 /// Returns a mutable, unchecked subslice of `str`.
451 /// This is the unchecked alternative to indexing the `str`.
455 /// Callers of this function are responsible that these preconditions are
458 /// * The starting index must come before the ending index;
459 /// * Indexes must be within bounds of the original slice;
460 /// * Indexes must lie on UTF-8 sequence boundaries.
462 /// Failing that, the returned string slice may reference invalid memory or
463 /// violate the invariants communicated by the `str` type.
468 /// let mut v = String::from("🗻∈🌏");
470 /// assert_eq!("🗻", v.get_unchecked_mut(0..4));
471 /// assert_eq!("∈", v.get_unchecked_mut(4..7));
472 /// assert_eq!("🌏", v.get_unchecked_mut(7..11));
475 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
477 pub unsafe fn get_unchecked_mut<I: SliceIndex<str>>(&mut self, i: I) -> &mut I::Output {
478 core_str::StrExt::get_unchecked_mut(self, i)
481 /// Creates a string slice from another string slice, bypassing safety
484 /// This is generally not recommended, use with caution! For a safe
485 /// alternative see [`str`] and [`Index`].
487 /// [`str`]: primitive.str.html
488 /// [`Index`]: ops/trait.Index.html
490 /// This new slice goes from `begin` to `end`, including `begin` but
493 /// To get a mutable string slice instead, see the
494 /// [`slice_mut_unchecked`] method.
496 /// [`slice_mut_unchecked`]: #method.slice_mut_unchecked
500 /// Callers of this function are responsible that three preconditions are
503 /// * `begin` must come before `end`.
504 /// * `begin` and `end` must be byte positions within the string slice.
505 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
512 /// let s = "Löwe 老虎 Léopard";
515 /// assert_eq!("Löwe 老虎 Léopard", s.slice_unchecked(0, 21));
518 /// let s = "Hello, world!";
521 /// assert_eq!("world", s.slice_unchecked(7, 12));
524 #[stable(feature = "rust1", since = "1.0.0")]
526 pub unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
527 core_str::StrExt::slice_unchecked(self, begin, end)
530 /// Creates a string slice from another string slice, bypassing safety
532 /// This is generally not recommended, use with caution! For a safe
533 /// alternative see [`str`] and [`IndexMut`].
535 /// [`str`]: primitive.str.html
536 /// [`IndexMut`]: ops/trait.IndexMut.html
538 /// This new slice goes from `begin` to `end`, including `begin` but
541 /// To get an immutable string slice instead, see the
542 /// [`slice_unchecked`] method.
544 /// [`slice_unchecked`]: #method.slice_unchecked
548 /// Callers of this function are responsible that three preconditions are
551 /// * `begin` must come before `end`.
552 /// * `begin` and `end` must be byte positions within the string slice.
553 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
554 #[stable(feature = "str_slice_mut", since = "1.5.0")]
556 pub unsafe fn slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str {
557 core_str::StrExt::slice_mut_unchecked(self, begin, end)
560 /// Divide one string slice into two at an index.
562 /// The argument, `mid`, should be a byte offset from the start of the
563 /// string. It must also be on the boundary of a UTF-8 code point.
565 /// The two slices returned go from the start of the string slice to `mid`,
566 /// and from `mid` to the end of the string slice.
568 /// To get mutable string slices instead, see the [`split_at_mut`]
571 /// [`split_at_mut`]: #method.split_at_mut
575 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
576 /// beyond the last code point of the string slice.
583 /// let s = "Per Martin-Löf";
585 /// let (first, last) = s.split_at(3);
587 /// assert_eq!("Per", first);
588 /// assert_eq!(" Martin-Löf", last);
591 #[stable(feature = "str_split_at", since = "1.4.0")]
592 pub fn split_at(&self, mid: usize) -> (&str, &str) {
593 core_str::StrExt::split_at(self, mid)
596 /// Divide one mutable string slice into two at an index.
598 /// The argument, `mid`, should be a byte offset from the start of the
599 /// string. It must also be on the boundary of a UTF-8 code point.
601 /// The two slices returned go from the start of the string slice to `mid`,
602 /// and from `mid` to the end of the string slice.
604 /// To get immutable string slices instead, see the [`split_at`] method.
606 /// [`split_at`]: #method.split_at
610 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
611 /// beyond the last code point of the string slice.
618 /// let mut s = "Per Martin-Löf".to_string();
620 /// let (first, last) = s.split_at_mut(3);
621 /// first.make_ascii_uppercase();
622 /// assert_eq!("PER", first);
623 /// assert_eq!(" Martin-Löf", last);
625 /// assert_eq!("PER Martin-Löf", s);
628 #[stable(feature = "str_split_at", since = "1.4.0")]
629 pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {
630 core_str::StrExt::split_at_mut(self, mid)
633 /// Returns an iterator over the [`char`]s of a string slice.
635 /// As a string slice consists of valid UTF-8, we can iterate through a
636 /// string slice by [`char`]. This method returns such an iterator.
638 /// It's important to remember that [`char`] represents a Unicode Scalar
639 /// Value, and may not match your idea of what a 'character' is. Iteration
640 /// over grapheme clusters may be what you actually want.
642 /// [`char`]: primitive.char.html
649 /// let word = "goodbye";
651 /// let count = word.chars().count();
652 /// assert_eq!(7, count);
654 /// let mut chars = word.chars();
656 /// assert_eq!(Some('g'), chars.next());
657 /// assert_eq!(Some('o'), chars.next());
658 /// assert_eq!(Some('o'), chars.next());
659 /// assert_eq!(Some('d'), chars.next());
660 /// assert_eq!(Some('b'), chars.next());
661 /// assert_eq!(Some('y'), chars.next());
662 /// assert_eq!(Some('e'), chars.next());
664 /// assert_eq!(None, chars.next());
667 /// Remember, [`char`]s may not match your human intuition about characters:
672 /// let mut chars = y.chars();
674 /// assert_eq!(Some('y'), chars.next()); // not 'y̆'
675 /// assert_eq!(Some('\u{0306}'), chars.next());
677 /// assert_eq!(None, chars.next());
679 #[stable(feature = "rust1", since = "1.0.0")]
681 pub fn chars(&self) -> Chars {
682 core_str::StrExt::chars(self)
684 /// Returns an iterator over the [`char`]s of a string slice, and their
687 /// As a string slice consists of valid UTF-8, we can iterate through a
688 /// string slice by [`char`]. This method returns an iterator of both
689 /// these [`char`]s, as well as their byte positions.
691 /// The iterator yields tuples. The position is first, the [`char`] is
694 /// [`char`]: primitive.char.html
701 /// let word = "goodbye";
703 /// let count = word.char_indices().count();
704 /// assert_eq!(7, count);
706 /// let mut char_indices = word.char_indices();
708 /// assert_eq!(Some((0, 'g')), char_indices.next());
709 /// assert_eq!(Some((1, 'o')), char_indices.next());
710 /// assert_eq!(Some((2, 'o')), char_indices.next());
711 /// assert_eq!(Some((3, 'd')), char_indices.next());
712 /// assert_eq!(Some((4, 'b')), char_indices.next());
713 /// assert_eq!(Some((5, 'y')), char_indices.next());
714 /// assert_eq!(Some((6, 'e')), char_indices.next());
716 /// assert_eq!(None, char_indices.next());
719 /// Remember, [`char`]s may not match your human intuition about characters:
724 /// let mut char_indices = y.char_indices();
726 /// assert_eq!(Some((0, 'y')), char_indices.next()); // not (0, 'y̆')
727 /// assert_eq!(Some((1, '\u{0306}')), char_indices.next());
729 /// assert_eq!(None, char_indices.next());
731 #[stable(feature = "rust1", since = "1.0.0")]
733 pub fn char_indices(&self) -> CharIndices {
734 core_str::StrExt::char_indices(self)
737 /// An iterator over the bytes of a string slice.
739 /// As a string slice consists of a sequence of bytes, we can iterate
740 /// through a string slice by byte. This method returns such an iterator.
747 /// let mut bytes = "bors".bytes();
749 /// assert_eq!(Some(b'b'), bytes.next());
750 /// assert_eq!(Some(b'o'), bytes.next());
751 /// assert_eq!(Some(b'r'), bytes.next());
752 /// assert_eq!(Some(b's'), bytes.next());
754 /// assert_eq!(None, bytes.next());
756 #[stable(feature = "rust1", since = "1.0.0")]
758 pub fn bytes(&self) -> Bytes {
759 core_str::StrExt::bytes(self)
762 /// Split a string slice by whitespace.
764 /// The iterator returned will return string slices that are sub-slices of
765 /// the original string slice, separated by any amount of whitespace.
767 /// 'Whitespace' is defined according to the terms of the Unicode Derived
768 /// Core Property `White_Space`.
775 /// let mut iter = "A few words".split_whitespace();
777 /// assert_eq!(Some("A"), iter.next());
778 /// assert_eq!(Some("few"), iter.next());
779 /// assert_eq!(Some("words"), iter.next());
781 /// assert_eq!(None, iter.next());
784 /// All kinds of whitespace are considered:
787 /// let mut iter = " Mary had\ta\u{2009}little \n\t lamb".split_whitespace();
788 /// assert_eq!(Some("Mary"), iter.next());
789 /// assert_eq!(Some("had"), iter.next());
790 /// assert_eq!(Some("a"), iter.next());
791 /// assert_eq!(Some("little"), iter.next());
792 /// assert_eq!(Some("lamb"), iter.next());
794 /// assert_eq!(None, iter.next());
796 #[stable(feature = "split_whitespace", since = "1.1.0")]
798 pub fn split_whitespace(&self) -> SplitWhitespace {
799 UnicodeStr::split_whitespace(self)
802 /// An iterator over the lines of a string, as string slices.
804 /// Lines are ended with either a newline (`\n`) or a carriage return with
805 /// a line feed (`\r\n`).
807 /// The final line ending is optional.
814 /// let text = "foo\r\nbar\n\nbaz\n";
815 /// let mut lines = text.lines();
817 /// assert_eq!(Some("foo"), lines.next());
818 /// assert_eq!(Some("bar"), lines.next());
819 /// assert_eq!(Some(""), lines.next());
820 /// assert_eq!(Some("baz"), lines.next());
822 /// assert_eq!(None, lines.next());
825 /// The final line ending isn't required:
828 /// let text = "foo\nbar\n\r\nbaz";
829 /// let mut lines = text.lines();
831 /// assert_eq!(Some("foo"), lines.next());
832 /// assert_eq!(Some("bar"), lines.next());
833 /// assert_eq!(Some(""), lines.next());
834 /// assert_eq!(Some("baz"), lines.next());
836 /// assert_eq!(None, lines.next());
838 #[stable(feature = "rust1", since = "1.0.0")]
840 pub fn lines(&self) -> Lines {
841 core_str::StrExt::lines(self)
844 /// An iterator over the lines of a string.
845 #[stable(feature = "rust1", since = "1.0.0")]
846 #[rustc_deprecated(since = "1.4.0", reason = "use lines() instead now")]
849 pub fn lines_any(&self) -> LinesAny {
850 core_str::StrExt::lines_any(self)
853 /// Returns an iterator of `u16` over the string encoded as UTF-16.
860 /// let text = "Zażółć gęślą jaźń";
862 /// let utf8_len = text.len();
863 /// let utf16_len = text.encode_utf16().count();
865 /// assert!(utf16_len <= utf8_len);
867 #[stable(feature = "encode_utf16", since = "1.8.0")]
868 pub fn encode_utf16(&self) -> EncodeUtf16 {
869 EncodeUtf16 { encoder: Utf16Encoder::new(self[..].chars()) }
872 /// Returns `true` if the given pattern matches a sub-slice of
873 /// this string slice.
875 /// Returns `false` if it does not.
882 /// let bananas = "bananas";
884 /// assert!(bananas.contains("nana"));
885 /// assert!(!bananas.contains("apples"));
887 #[stable(feature = "rust1", since = "1.0.0")]
889 pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
890 core_str::StrExt::contains(self, pat)
893 /// Returns `true` if the given pattern matches a prefix of this
896 /// Returns `false` if it does not.
903 /// let bananas = "bananas";
905 /// assert!(bananas.starts_with("bana"));
906 /// assert!(!bananas.starts_with("nana"));
908 #[stable(feature = "rust1", since = "1.0.0")]
909 pub fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
910 core_str::StrExt::starts_with(self, pat)
913 /// Returns `true` if the given pattern matches a suffix of this
916 /// Returns `false` if it does not.
923 /// let bananas = "bananas";
925 /// assert!(bananas.ends_with("anas"));
926 /// assert!(!bananas.ends_with("nana"));
928 #[stable(feature = "rust1", since = "1.0.0")]
929 pub fn ends_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool
930 where P::Searcher: ReverseSearcher<'a>
932 core_str::StrExt::ends_with(self, pat)
935 /// Returns the byte index of the first character of this string slice that
936 /// matches the pattern.
938 /// Returns [`None`] if the pattern doesn't match.
940 /// The pattern can be a `&str`, [`char`], or a closure that determines if
941 /// a character matches.
943 /// [`char`]: primitive.char.html
944 /// [`None`]: option/enum.Option.html#variant.None
951 /// let s = "Löwe 老虎 Léopard";
953 /// assert_eq!(s.find('L'), Some(0));
954 /// assert_eq!(s.find('é'), Some(14));
955 /// assert_eq!(s.find("Léopard"), Some(13));
958 /// More complex patterns using point-free style and closures:
961 /// let s = "Löwe 老虎 Léopard";
963 /// assert_eq!(s.find(char::is_whitespace), Some(5));
964 /// assert_eq!(s.find(char::is_lowercase), Some(1));
965 /// assert_eq!(s.find(|c: char| c.is_whitespace() || c.is_lowercase()), Some(1));
966 /// assert_eq!(s.find(|c: char| (c < 'o') && (c > 'a')), Some(4));
969 /// Not finding the pattern:
972 /// let s = "Löwe 老虎 Léopard";
973 /// let x: &[_] = &['1', '2'];
975 /// assert_eq!(s.find(x), None);
977 #[stable(feature = "rust1", since = "1.0.0")]
979 pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
980 core_str::StrExt::find(self, pat)
983 /// Returns the byte index of the last character of this string slice that
984 /// matches the pattern.
986 /// Returns [`None`] if the pattern doesn't match.
988 /// The pattern can be a `&str`, [`char`], or a closure that determines if
989 /// a character matches.
991 /// [`char`]: primitive.char.html
992 /// [`None`]: option/enum.Option.html#variant.None
999 /// let s = "Löwe 老虎 Léopard";
1001 /// assert_eq!(s.rfind('L'), Some(13));
1002 /// assert_eq!(s.rfind('é'), Some(14));
1005 /// More complex patterns with closures:
1008 /// let s = "Löwe 老虎 Léopard";
1010 /// assert_eq!(s.rfind(char::is_whitespace), Some(12));
1011 /// assert_eq!(s.rfind(char::is_lowercase), Some(20));
1014 /// Not finding the pattern:
1017 /// let s = "Löwe 老虎 Léopard";
1018 /// let x: &[_] = &['1', '2'];
1020 /// assert_eq!(s.rfind(x), None);
1022 #[stable(feature = "rust1", since = "1.0.0")]
1024 pub fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
1025 where P::Searcher: ReverseSearcher<'a>
1027 core_str::StrExt::rfind(self, pat)
1030 /// An iterator over substrings of this string slice, separated by
1031 /// characters matched by a pattern.
1033 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1036 /// # Iterator behavior
1038 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1039 /// allows a reverse search and forward/reverse search yields the same
1040 /// elements. This is true for, eg, [`char`] but not for `&str`.
1042 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1044 /// If the pattern allows a reverse search but its results might differ
1045 /// from a forward search, the [`rsplit`] method can be used.
1047 /// [`char`]: primitive.char.html
1048 /// [`rsplit`]: #method.rsplit
1052 /// Simple patterns:
1055 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
1056 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
1058 /// let v: Vec<&str> = "".split('X').collect();
1059 /// assert_eq!(v, [""]);
1061 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
1062 /// assert_eq!(v, ["lion", "", "tiger", "leopard"]);
1064 /// let v: Vec<&str> = "lion::tiger::leopard".split("::").collect();
1065 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1067 /// let v: Vec<&str> = "abc1def2ghi".split(char::is_numeric).collect();
1068 /// assert_eq!(v, ["abc", "def", "ghi"]);
1070 /// let v: Vec<&str> = "lionXtigerXleopard".split(char::is_uppercase).collect();
1071 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1074 /// A more complex pattern, using a closure:
1077 /// let v: Vec<&str> = "abc1defXghi".split(|c| c == '1' || c == 'X').collect();
1078 /// assert_eq!(v, ["abc", "def", "ghi"]);
1081 /// If a string contains multiple contiguous separators, you will end up
1082 /// with empty strings in the output:
1085 /// let x = "||||a||b|c".to_string();
1086 /// let d: Vec<_> = x.split('|').collect();
1088 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1091 /// Contiguous separators are separated by the empty string.
1094 /// let x = "(///)".to_string();
1095 /// let d: Vec<_> = x.split('/').collect();
1097 /// assert_eq!(d, &["(", "", "", ")"]);
1100 /// Separators at the start or end of a string are neighbored
1101 /// by empty strings.
1104 /// let d: Vec<_> = "010".split("0").collect();
1105 /// assert_eq!(d, &["", "1", ""]);
1108 /// When the empty string is used as a separator, it separates
1109 /// every character in the string, along with the beginning
1110 /// and end of the string.
1113 /// let f: Vec<_> = "rust".split("").collect();
1114 /// assert_eq!(f, &["", "r", "u", "s", "t", ""]);
1117 /// Contiguous separators can lead to possibly surprising behavior
1118 /// when whitespace is used as the separator. This code is correct:
1121 /// let x = " a b c".to_string();
1122 /// let d: Vec<_> = x.split(' ').collect();
1124 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1127 /// It does _not_ give you:
1130 /// assert_eq!(d, &["a", "b", "c"]);
1133 /// Use [`split_whitespace`] for this behavior.
1135 /// [`split_whitespace`]: #method.split_whitespace
1136 #[stable(feature = "rust1", since = "1.0.0")]
1138 pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
1139 core_str::StrExt::split(self, pat)
1142 /// An iterator over substrings of the given string slice, separated by
1143 /// characters matched by a pattern and yielded in reverse order.
1145 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1148 /// [`char`]: primitive.char.html
1150 /// # Iterator behavior
1152 /// The returned iterator requires that the pattern supports a reverse
1153 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1154 /// search yields the same elements.
1156 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1158 /// For iterating from the front, the [`split`] method can be used.
1160 /// [`split`]: #method.split
1164 /// Simple patterns:
1167 /// let v: Vec<&str> = "Mary had a little lamb".rsplit(' ').collect();
1168 /// assert_eq!(v, ["lamb", "little", "a", "had", "Mary"]);
1170 /// let v: Vec<&str> = "".rsplit('X').collect();
1171 /// assert_eq!(v, [""]);
1173 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplit('X').collect();
1174 /// assert_eq!(v, ["leopard", "tiger", "", "lion"]);
1176 /// let v: Vec<&str> = "lion::tiger::leopard".rsplit("::").collect();
1177 /// assert_eq!(v, ["leopard", "tiger", "lion"]);
1180 /// A more complex pattern, using a closure:
1183 /// let v: Vec<&str> = "abc1defXghi".rsplit(|c| c == '1' || c == 'X').collect();
1184 /// assert_eq!(v, ["ghi", "def", "abc"]);
1186 #[stable(feature = "rust1", since = "1.0.0")]
1188 pub fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
1189 where P::Searcher: ReverseSearcher<'a>
1191 core_str::StrExt::rsplit(self, pat)
1194 /// An iterator over substrings of the given string slice, separated by
1195 /// characters matched by a pattern.
1197 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1200 /// Equivalent to [`split`], except that the trailing substring
1201 /// is skipped if empty.
1203 /// [`split`]: #method.split
1205 /// This method can be used for string data that is _terminated_,
1206 /// rather than _separated_ by a pattern.
1208 /// # Iterator behavior
1210 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1211 /// allows a reverse search and forward/reverse search yields the same
1212 /// elements. This is true for, eg, [`char`] but not for `&str`.
1214 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1215 /// [`char`]: primitive.char.html
1217 /// If the pattern allows a reverse search but its results might differ
1218 /// from a forward search, the [`rsplit_terminator`] method can be used.
1220 /// [`rsplit_terminator`]: #method.rsplit_terminator
1227 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
1228 /// assert_eq!(v, ["A", "B"]);
1230 /// let v: Vec<&str> = "A..B..".split_terminator(".").collect();
1231 /// assert_eq!(v, ["A", "", "B", ""]);
1233 #[stable(feature = "rust1", since = "1.0.0")]
1235 pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
1236 core_str::StrExt::split_terminator(self, pat)
1239 /// An iterator over substrings of `self`, separated by characters
1240 /// matched by a pattern and yielded in reverse order.
1242 /// The pattern can be a simple `&str`, [`char`], or a closure that
1243 /// determines the split.
1244 /// Additional libraries might provide more complex patterns like
1245 /// regular expressions.
1247 /// [`char`]: primitive.char.html
1249 /// Equivalent to [`split`], except that the trailing substring is
1250 /// skipped if empty.
1252 /// [`split`]: #method.split
1254 /// This method can be used for string data that is _terminated_,
1255 /// rather than _separated_ by a pattern.
1257 /// # Iterator behavior
1259 /// The returned iterator requires that the pattern supports a
1260 /// reverse search, and it will be double ended if a forward/reverse
1261 /// search yields the same elements.
1263 /// For iterating from the front, the [`split_terminator`] method can be
1266 /// [`split_terminator`]: #method.split_terminator
1271 /// let v: Vec<&str> = "A.B.".rsplit_terminator('.').collect();
1272 /// assert_eq!(v, ["B", "A"]);
1274 /// let v: Vec<&str> = "A..B..".rsplit_terminator(".").collect();
1275 /// assert_eq!(v, ["", "B", "", "A"]);
1277 #[stable(feature = "rust1", since = "1.0.0")]
1279 pub fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1280 where P::Searcher: ReverseSearcher<'a>
1282 core_str::StrExt::rsplit_terminator(self, pat)
1285 /// An iterator over substrings of the given string slice, separated by a
1286 /// pattern, restricted to returning at most `n` items.
1288 /// If `n` substrings are returned, the last substring (the `n`th substring)
1289 /// will contain the remainder of the string.
1291 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1294 /// [`char`]: primitive.char.html
1296 /// # Iterator behavior
1298 /// The returned iterator will not be double ended, because it is
1299 /// not efficient to support.
1301 /// If the pattern allows a reverse search, the [`rsplitn`] method can be
1304 /// [`rsplitn`]: #method.rsplitn
1308 /// Simple patterns:
1311 /// let v: Vec<&str> = "Mary had a little lambda".splitn(3, ' ').collect();
1312 /// assert_eq!(v, ["Mary", "had", "a little lambda"]);
1314 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(3, "X").collect();
1315 /// assert_eq!(v, ["lion", "", "tigerXleopard"]);
1317 /// let v: Vec<&str> = "abcXdef".splitn(1, 'X').collect();
1318 /// assert_eq!(v, ["abcXdef"]);
1320 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1321 /// assert_eq!(v, [""]);
1324 /// A more complex pattern, using a closure:
1327 /// let v: Vec<&str> = "abc1defXghi".splitn(2, |c| c == '1' || c == 'X').collect();
1328 /// assert_eq!(v, ["abc", "defXghi"]);
1330 #[stable(feature = "rust1", since = "1.0.0")]
1332 pub fn splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P> {
1333 core_str::StrExt::splitn(self, n, pat)
1336 /// An iterator over substrings of this string slice, separated by a
1337 /// pattern, starting from the end of the string, restricted to returning
1338 /// at most `n` items.
1340 /// If `n` substrings are returned, the last substring (the `n`th substring)
1341 /// will contain the remainder of the string.
1343 /// The pattern can be a `&str`, [`char`], or a closure that
1344 /// determines the split.
1346 /// [`char`]: primitive.char.html
1348 /// # Iterator behavior
1350 /// The returned iterator will not be double ended, because it is not
1351 /// efficient to support.
1353 /// For splitting from the front, the [`splitn`] method can be used.
1355 /// [`splitn`]: #method.splitn
1359 /// Simple patterns:
1362 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(3, ' ').collect();
1363 /// assert_eq!(v, ["lamb", "little", "Mary had a"]);
1365 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(3, 'X').collect();
1366 /// assert_eq!(v, ["leopard", "tiger", "lionX"]);
1368 /// let v: Vec<&str> = "lion::tiger::leopard".rsplitn(2, "::").collect();
1369 /// assert_eq!(v, ["leopard", "lion::tiger"]);
1372 /// A more complex pattern, using a closure:
1375 /// let v: Vec<&str> = "abc1defXghi".rsplitn(2, |c| c == '1' || c == 'X').collect();
1376 /// assert_eq!(v, ["ghi", "abc1def"]);
1378 #[stable(feature = "rust1", since = "1.0.0")]
1380 pub fn rsplitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> RSplitN<'a, P>
1381 where P::Searcher: ReverseSearcher<'a>
1383 core_str::StrExt::rsplitn(self, n, pat)
1386 /// An iterator over the disjoint matches of a pattern within the given string
1389 /// The pattern can be a `&str`, [`char`], or a closure that
1390 /// determines if a character matches.
1392 /// [`char`]: primitive.char.html
1394 /// # Iterator behavior
1396 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1397 /// allows a reverse search and forward/reverse search yields the same
1398 /// elements. This is true for, eg, [`char`] but not for `&str`.
1400 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1401 /// [`char`]: primitive.char.html
1403 /// If the pattern allows a reverse search but its results might differ
1404 /// from a forward search, the [`rmatches`] method can be used.
1406 /// [`rmatches`]: #method.rmatches
1413 /// let v: Vec<&str> = "abcXXXabcYYYabc".matches("abc").collect();
1414 /// assert_eq!(v, ["abc", "abc", "abc"]);
1416 /// let v: Vec<&str> = "1abc2abc3".matches(char::is_numeric).collect();
1417 /// assert_eq!(v, ["1", "2", "3"]);
1419 #[stable(feature = "str_matches", since = "1.2.0")]
1421 pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1422 core_str::StrExt::matches(self, pat)
1425 /// An iterator over the disjoint matches of a pattern within this string slice,
1426 /// yielded in reverse order.
1428 /// The pattern can be a `&str`, [`char`], or a closure that determines if
1429 /// a character matches.
1431 /// [`char`]: primitive.char.html
1433 /// # Iterator behavior
1435 /// The returned iterator requires that the pattern supports a reverse
1436 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1437 /// search yields the same elements.
1439 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1441 /// For iterating from the front, the [`matches`] method can be used.
1443 /// [`matches`]: #method.matches
1450 /// let v: Vec<&str> = "abcXXXabcYYYabc".rmatches("abc").collect();
1451 /// assert_eq!(v, ["abc", "abc", "abc"]);
1453 /// let v: Vec<&str> = "1abc2abc3".rmatches(char::is_numeric).collect();
1454 /// assert_eq!(v, ["3", "2", "1"]);
1456 #[stable(feature = "str_matches", since = "1.2.0")]
1458 pub fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
1459 where P::Searcher: ReverseSearcher<'a>
1461 core_str::StrExt::rmatches(self, pat)
1464 /// An iterator over the disjoint matches of a pattern within this string
1465 /// slice as well as the index that the match starts at.
1467 /// For matches of `pat` within `self` that overlap, only the indices
1468 /// corresponding to the first match are returned.
1470 /// The pattern can be a `&str`, [`char`], or a closure that determines
1471 /// if a character matches.
1473 /// [`char`]: primitive.char.html
1475 /// # Iterator behavior
1477 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1478 /// allows a reverse search and forward/reverse search yields the same
1479 /// elements. This is true for, eg, [`char`] but not for `&str`.
1481 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1483 /// If the pattern allows a reverse search but its results might differ
1484 /// from a forward search, the [`rmatch_indices`] method can be used.
1486 /// [`rmatch_indices`]: #method.rmatch_indices
1493 /// let v: Vec<_> = "abcXXXabcYYYabc".match_indices("abc").collect();
1494 /// assert_eq!(v, [(0, "abc"), (6, "abc"), (12, "abc")]);
1496 /// let v: Vec<_> = "1abcabc2".match_indices("abc").collect();
1497 /// assert_eq!(v, [(1, "abc"), (4, "abc")]);
1499 /// let v: Vec<_> = "ababa".match_indices("aba").collect();
1500 /// assert_eq!(v, [(0, "aba")]); // only the first `aba`
1502 #[stable(feature = "str_match_indices", since = "1.5.0")]
1504 pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1505 core_str::StrExt::match_indices(self, pat)
1508 /// An iterator over the disjoint matches of a pattern within `self`,
1509 /// yielded in reverse order along with the index of the match.
1511 /// For matches of `pat` within `self` that overlap, only the indices
1512 /// corresponding to the last match are returned.
1514 /// The pattern can be a `&str`, [`char`], or a closure that determines if a
1515 /// character matches.
1517 /// [`char`]: primitive.char.html
1519 /// # Iterator behavior
1521 /// The returned iterator requires that the pattern supports a reverse
1522 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1523 /// search yields the same elements.
1525 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1527 /// For iterating from the front, the [`match_indices`] method can be used.
1529 /// [`match_indices`]: #method.match_indices
1536 /// let v: Vec<_> = "abcXXXabcYYYabc".rmatch_indices("abc").collect();
1537 /// assert_eq!(v, [(12, "abc"), (6, "abc"), (0, "abc")]);
1539 /// let v: Vec<_> = "1abcabc2".rmatch_indices("abc").collect();
1540 /// assert_eq!(v, [(4, "abc"), (1, "abc")]);
1542 /// let v: Vec<_> = "ababa".rmatch_indices("aba").collect();
1543 /// assert_eq!(v, [(2, "aba")]); // only the last `aba`
1545 #[stable(feature = "str_match_indices", since = "1.5.0")]
1547 pub fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
1548 where P::Searcher: ReverseSearcher<'a>
1550 core_str::StrExt::rmatch_indices(self, pat)
1553 /// Returns a string slice with leading and trailing whitespace removed.
1555 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1556 /// Core Property `White_Space`.
1563 /// let s = " Hello\tworld\t";
1565 /// assert_eq!("Hello\tworld", s.trim());
1567 #[stable(feature = "rust1", since = "1.0.0")]
1568 pub fn trim(&self) -> &str {
1569 UnicodeStr::trim(self)
1572 /// Returns a string slice with leading whitespace removed.
1574 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1575 /// Core Property `White_Space`.
1577 /// # Text directionality
1579 /// A string is a sequence of bytes. 'Left' in this context means the first
1580 /// position of that byte string; for a language like Arabic or Hebrew
1581 /// which are 'right to left' rather than 'left to right', this will be
1582 /// the _right_ side, not the left.
1589 /// let s = " Hello\tworld\t";
1591 /// assert_eq!("Hello\tworld\t", s.trim_left());
1597 /// let s = " English";
1598 /// assert!(Some('E') == s.trim_left().chars().next());
1600 /// let s = " עברית";
1601 /// assert!(Some('ע') == s.trim_left().chars().next());
1603 #[stable(feature = "rust1", since = "1.0.0")]
1604 pub fn trim_left(&self) -> &str {
1605 UnicodeStr::trim_left(self)
1608 /// Returns a string slice with trailing whitespace removed.
1610 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1611 /// Core Property `White_Space`.
1613 /// # Text directionality
1615 /// A string is a sequence of bytes. 'Right' in this context means the last
1616 /// position of that byte string; for a language like Arabic or Hebrew
1617 /// which are 'right to left' rather than 'left to right', this will be
1618 /// the _left_ side, not the right.
1625 /// let s = " Hello\tworld\t";
1627 /// assert_eq!(" Hello\tworld", s.trim_right());
1633 /// let s = "English ";
1634 /// assert!(Some('h') == s.trim_right().chars().rev().next());
1636 /// let s = "עברית ";
1637 /// assert!(Some('ת') == s.trim_right().chars().rev().next());
1639 #[stable(feature = "rust1", since = "1.0.0")]
1640 pub fn trim_right(&self) -> &str {
1641 UnicodeStr::trim_right(self)
1644 /// Returns a string slice with all prefixes and suffixes that match a
1645 /// pattern repeatedly removed.
1647 /// The pattern can be a [`char`] or a closure that determines if a
1648 /// character matches.
1650 /// [`char`]: primitive.char.html
1654 /// Simple patterns:
1657 /// assert_eq!("11foo1bar11".trim_matches('1'), "foo1bar");
1658 /// assert_eq!("123foo1bar123".trim_matches(char::is_numeric), "foo1bar");
1660 /// let x: &[_] = &['1', '2'];
1661 /// assert_eq!("12foo1bar12".trim_matches(x), "foo1bar");
1664 /// A more complex pattern, using a closure:
1667 /// assert_eq!("1foo1barXX".trim_matches(|c| c == '1' || c == 'X'), "foo1bar");
1669 #[stable(feature = "rust1", since = "1.0.0")]
1670 pub fn trim_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1671 where P::Searcher: DoubleEndedSearcher<'a>
1673 core_str::StrExt::trim_matches(self, pat)
1676 /// Returns a string slice with all prefixes that match a pattern
1677 /// repeatedly removed.
1679 /// The pattern can be a `&str`, [`char`], or a closure that determines if
1680 /// a character matches.
1682 /// [`char`]: primitive.char.html
1684 /// # Text directionality
1686 /// A string is a sequence of bytes. 'Left' in this context means the first
1687 /// position of that byte string; for a language like Arabic or Hebrew
1688 /// which are 'right to left' rather than 'left to right', this will be
1689 /// the _right_ side, not the left.
1696 /// assert_eq!("11foo1bar11".trim_left_matches('1'), "foo1bar11");
1697 /// assert_eq!("123foo1bar123".trim_left_matches(char::is_numeric), "foo1bar123");
1699 /// let x: &[_] = &['1', '2'];
1700 /// assert_eq!("12foo1bar12".trim_left_matches(x), "foo1bar12");
1702 #[stable(feature = "rust1", since = "1.0.0")]
1703 pub fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
1704 core_str::StrExt::trim_left_matches(self, pat)
1707 /// Returns a string slice with all suffixes that match a pattern
1708 /// repeatedly removed.
1710 /// The pattern can be a `&str`, [`char`], or a closure that
1711 /// determines if a character matches.
1713 /// [`char`]: primitive.char.html
1715 /// # Text directionality
1717 /// A string is a sequence of bytes. 'Right' in this context means the last
1718 /// position of that byte string; for a language like Arabic or Hebrew
1719 /// which are 'right to left' rather than 'left to right', this will be
1720 /// the _left_ side, not the right.
1724 /// Simple patterns:
1727 /// assert_eq!("11foo1bar11".trim_right_matches('1'), "11foo1bar");
1728 /// assert_eq!("123foo1bar123".trim_right_matches(char::is_numeric), "123foo1bar");
1730 /// let x: &[_] = &['1', '2'];
1731 /// assert_eq!("12foo1bar12".trim_right_matches(x), "12foo1bar");
1734 /// A more complex pattern, using a closure:
1737 /// assert_eq!("1fooX".trim_left_matches(|c| c == '1' || c == 'X'), "fooX");
1739 #[stable(feature = "rust1", since = "1.0.0")]
1740 pub fn trim_right_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1741 where P::Searcher: ReverseSearcher<'a>
1743 core_str::StrExt::trim_right_matches(self, pat)
1746 /// Parses this string slice into another type.
1748 /// Because `parse` is so general, it can cause problems with type
1749 /// inference. As such, `parse` is one of the few times you'll see
1750 /// the syntax affectionately known as the 'turbofish': `::<>`. This
1751 /// helps the inference algorithm understand specifically which type
1752 /// you're trying to parse into.
1754 /// `parse` can parse any type that implements the [`FromStr`] trait.
1756 /// [`FromStr`]: str/trait.FromStr.html
1760 /// Will return [`Err`] if it's not possible to parse this string slice into
1761 /// the desired type.
1763 /// [`Err`]: str/trait.FromStr.html#associatedtype.Err
1770 /// let four: u32 = "4".parse().unwrap();
1772 /// assert_eq!(4, four);
1775 /// Using the 'turbofish' instead of annotating `four`:
1778 /// let four = "4".parse::<u32>();
1780 /// assert_eq!(Ok(4), four);
1783 /// Failing to parse:
1786 /// let nope = "j".parse::<u32>();
1788 /// assert!(nope.is_err());
1791 #[stable(feature = "rust1", since = "1.0.0")]
1792 pub fn parse<F: FromStr>(&self) -> Result<F, F::Err> {
1793 core_str::StrExt::parse(self)
1796 /// Converts a `Box<str>` into a `Box<[u8]>` without copying or allocating.
1803 /// let s = "this is a string";
1804 /// let boxed_str = s.to_owned().into_boxed_str();
1805 /// let boxed_bytes = boxed_str.into_boxed_bytes();
1806 /// assert_eq!(*boxed_bytes, *s.as_bytes());
1808 #[stable(feature = "str_box_extras", since = "1.20.0")]
1809 pub fn into_boxed_bytes(self: Box<str>) -> Box<[u8]> {
1813 /// Replaces all matches of a pattern with another string.
1815 /// `replace` creates a new [`String`], and copies the data from this string slice into it.
1816 /// While doing so, it attempts to find matches of a pattern. If it finds any, it
1817 /// replaces them with the replacement string slice.
1819 /// [`String`]: string/struct.String.html
1826 /// let s = "this is old";
1828 /// assert_eq!("this is new", s.replace("old", "new"));
1831 /// When the pattern doesn't match:
1834 /// let s = "this is old";
1835 /// assert_eq!(s, s.replace("cookie monster", "little lamb"));
1837 #[stable(feature = "rust1", since = "1.0.0")]
1839 pub fn replace<'a, P: Pattern<'a>>(&'a self, from: P, to: &str) -> String {
1840 let mut result = String::new();
1841 let mut last_end = 0;
1842 for (start, part) in self.match_indices(from) {
1843 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1844 result.push_str(to);
1845 last_end = start + part.len();
1847 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1851 /// Replaces first N matches of a pattern with another string.
1853 /// `replacen` creates a new [`String`], and copies the data from this string slice into it.
1854 /// While doing so, it attempts to find matches of a pattern. If it finds any, it
1855 /// replaces them with the replacement string slice at most `count` times.
1857 /// [`String`]: string/struct.String.html
1864 /// let s = "foo foo 123 foo";
1865 /// assert_eq!("new new 123 foo", s.replacen("foo", "new", 2));
1866 /// assert_eq!("faa fao 123 foo", s.replacen('o', "a", 3));
1867 /// assert_eq!("foo foo new23 foo", s.replacen(char::is_numeric, "new", 1));
1870 /// When the pattern doesn't match:
1873 /// let s = "this is old";
1874 /// assert_eq!(s, s.replacen("cookie monster", "little lamb", 10));
1876 #[stable(feature = "str_replacen", since = "1.16.0")]
1877 pub fn replacen<'a, P: Pattern<'a>>(&'a self, pat: P, to: &str, count: usize) -> String {
1878 // Hope to reduce the times of re-allocation
1879 let mut result = String::with_capacity(32);
1880 let mut last_end = 0;
1881 for (start, part) in self.match_indices(pat).take(count) {
1882 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1883 result.push_str(to);
1884 last_end = start + part.len();
1886 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1890 /// Returns the lowercase equivalent of this string slice, as a new [`String`].
1892 /// 'Lowercase' is defined according to the terms of the Unicode Derived Core Property
1895 /// Since some characters can expand into multiple characters when changing
1896 /// the case, this function returns a [`String`] instead of modifying the
1897 /// parameter in-place.
1899 /// [`String`]: string/struct.String.html
1906 /// let s = "HELLO";
1908 /// assert_eq!("hello", s.to_lowercase());
1911 /// A tricky example, with sigma:
1914 /// let sigma = "Σ";
1916 /// assert_eq!("σ", sigma.to_lowercase());
1918 /// // but at the end of a word, it's ς, not σ:
1919 /// let odysseus = "ὈΔΥΣΣΕΎΣ";
1921 /// assert_eq!("ὀδυσσεύς", odysseus.to_lowercase());
1924 /// Languages without case are not changed:
1927 /// let new_year = "农历新年";
1929 /// assert_eq!(new_year, new_year.to_lowercase());
1931 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
1932 pub fn to_lowercase(&self) -> String {
1933 let mut s = String::with_capacity(self.len());
1934 for (i, c) in self[..].char_indices() {
1936 // Σ maps to σ, except at the end of a word where it maps to ς.
1937 // This is the only conditional (contextual) but language-independent mapping
1938 // in `SpecialCasing.txt`,
1939 // so hard-code it rather than have a generic "condition" mechanism.
1940 // See https://github.com/rust-lang/rust/issues/26035
1941 map_uppercase_sigma(self, i, &mut s)
1943 s.extend(c.to_lowercase());
1948 fn map_uppercase_sigma(from: &str, i: usize, to: &mut String) {
1949 // See http://www.unicode.org/versions/Unicode7.0.0/ch03.pdf#G33992
1950 // for the definition of `Final_Sigma`.
1951 debug_assert!('Σ'.len_utf8() == 2);
1952 let is_word_final = case_ignoreable_then_cased(from[..i].chars().rev()) &&
1953 !case_ignoreable_then_cased(from[i + 2..].chars());
1954 to.push_str(if is_word_final { "ς" } else { "σ" });
1957 fn case_ignoreable_then_cased<I: Iterator<Item = char>>(iter: I) -> bool {
1958 use std_unicode::derived_property::{Cased, Case_Ignorable};
1959 match iter.skip_while(|&c| Case_Ignorable(c)).next() {
1960 Some(c) => Cased(c),
1966 /// Returns the uppercase equivalent of this string slice, as a new [`String`].
1968 /// 'Uppercase' is defined according to the terms of the Unicode Derived Core Property
1971 /// Since some characters can expand into multiple characters when changing
1972 /// the case, this function returns a [`String`] instead of modifying the
1973 /// parameter in-place.
1975 /// [`String`]: string/struct.String.html
1982 /// let s = "hello";
1984 /// assert_eq!("HELLO", s.to_uppercase());
1987 /// Scripts without case are not changed:
1990 /// let new_year = "农历新年";
1992 /// assert_eq!(new_year, new_year.to_uppercase());
1994 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
1995 pub fn to_uppercase(&self) -> String {
1996 let mut s = String::with_capacity(self.len());
1997 s.extend(self.chars().flat_map(|c| c.to_uppercase()));
2001 /// Escapes each char in `s` with [`char::escape_debug`].
2003 /// [`char::escape_debug`]: primitive.char.html#method.escape_debug
2004 #[unstable(feature = "str_escape",
2005 reason = "return type may change to be an iterator",
2007 pub fn escape_debug(&self) -> String {
2008 self.chars().flat_map(|c| c.escape_debug()).collect()
2011 /// Escapes each char in `s` with [`char::escape_default`].
2013 /// [`char::escape_default`]: primitive.char.html#method.escape_default
2014 #[unstable(feature = "str_escape",
2015 reason = "return type may change to be an iterator",
2017 pub fn escape_default(&self) -> String {
2018 self.chars().flat_map(|c| c.escape_default()).collect()
2021 /// Escapes each char in `s` with [`char::escape_unicode`].
2023 /// [`char::escape_unicode`]: primitive.char.html#method.escape_unicode
2024 #[unstable(feature = "str_escape",
2025 reason = "return type may change to be an iterator",
2027 pub fn escape_unicode(&self) -> String {
2028 self.chars().flat_map(|c| c.escape_unicode()).collect()
2031 /// Converts a [`Box<str>`] into a [`String`] without copying or allocating.
2033 /// [`String`]: string/struct.String.html
2034 /// [`Box<str>`]: boxed/struct.Box.html
2041 /// let string = String::from("birthday gift");
2042 /// let boxed_str = string.clone().into_boxed_str();
2044 /// assert_eq!(boxed_str.into_string(), string);
2046 #[stable(feature = "box_str", since = "1.4.0")]
2047 pub fn into_string(self: Box<str>) -> String {
2048 let slice = Box::<[u8]>::from(self);
2049 unsafe { String::from_utf8_unchecked(slice.into_vec()) }
2052 /// Create a [`String`] by repeating a string `n` times.
2054 /// [`String`]: string/struct.String.html
2061 /// assert_eq!("abc".repeat(4), String::from("abcabcabcabc"));
2063 #[stable(feature = "repeat_str", since = "1.16.0")]
2064 pub fn repeat(&self, n: usize) -> String {
2065 let mut s = String::with_capacity(self.len() * n);
2066 s.extend((0..n).map(|_| self));
2070 /// Checks if all characters in this string are within the ASCII range.
2075 /// let ascii = "hello!\n";
2076 /// let non_ascii = "Grüße, Jürgen ❤";
2078 /// assert!(ascii.is_ascii());
2079 /// assert!(!non_ascii.is_ascii());
2081 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.21.0")]
2083 pub fn is_ascii(&self) -> bool {
2084 // We can treat each byte as character here: all multibyte characters
2085 // start with a byte that is not in the ascii range, so we will stop
2087 self.bytes().all(|b| b.is_ascii())
2090 /// Returns a copy of this string where each character is mapped to its
2091 /// ASCII upper case equivalent.
2093 /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z',
2094 /// but non-ASCII letters are unchanged.
2096 /// To uppercase the value in-place, use [`make_ascii_uppercase`].
2098 /// To uppercase ASCII characters in addition to non-ASCII characters, use
2099 /// [`to_uppercase`].
2104 /// let s = "Grüße, Jürgen ❤";
2106 /// assert_eq!("GRüßE, JüRGEN ❤", s.to_ascii_uppercase());
2109 /// [`make_ascii_uppercase`]: #method.make_ascii_uppercase
2110 /// [`to_uppercase`]: #method.to_uppercase
2111 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.21.0")]
2114 pub fn to_ascii_uppercase(&self) -> String {
2115 let mut bytes = self.as_bytes().to_vec();
2116 bytes.make_ascii_uppercase();
2117 // make_ascii_uppercase() preserves the UTF-8 invariant.
2118 unsafe { String::from_utf8_unchecked(bytes) }
2121 /// Returns a copy of this string where each character is mapped to its
2122 /// ASCII lower case equivalent.
2124 /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z',
2125 /// but non-ASCII letters are unchanged.
2127 /// To lowercase the value in-place, use [`make_ascii_lowercase`].
2129 /// To lowercase ASCII characters in addition to non-ASCII characters, use
2130 /// [`to_lowercase`].
2135 /// let s = "Grüße, Jürgen ❤";
2137 /// assert_eq!("grüße, jürgen ❤", s.to_ascii_lowercase());
2140 /// [`make_ascii_lowercase`]: #method.make_ascii_lowercase
2141 /// [`to_lowercase`]: #method.to_lowercase
2142 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.21.0")]
2145 pub fn to_ascii_lowercase(&self) -> String {
2146 let mut bytes = self.as_bytes().to_vec();
2147 bytes.make_ascii_lowercase();
2148 // make_ascii_lowercase() preserves the UTF-8 invariant.
2149 unsafe { String::from_utf8_unchecked(bytes) }
2152 /// Checks that two strings are an ASCII case-insensitive match.
2154 /// Same as `to_ascii_lowercase(a) == to_ascii_lowercase(b)`,
2155 /// but without allocating and copying temporaries.
2160 /// assert!("Ferris".eq_ignore_ascii_case("FERRIS"));
2161 /// assert!("Ferrös".eq_ignore_ascii_case("FERRöS"));
2162 /// assert!(!"Ferrös".eq_ignore_ascii_case("FERRÖS"));
2164 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.21.0")]
2167 pub fn eq_ignore_ascii_case(&self, other: &str) -> bool {
2168 self.as_bytes().eq_ignore_ascii_case(other.as_bytes())
2171 /// Converts this string to its ASCII upper case equivalent in-place.
2173 /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z',
2174 /// but non-ASCII letters are unchanged.
2176 /// To return a new uppercased value without modifying the existing one, use
2177 /// [`to_ascii_uppercase`].
2179 /// [`to_ascii_uppercase`]: #method.to_ascii_uppercase
2180 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.21.0")]
2182 pub fn make_ascii_uppercase(&mut self) {
2183 let me = unsafe { self.as_bytes_mut() };
2184 me.make_ascii_uppercase()
2187 /// Converts this string to its ASCII lower case equivalent in-place.
2189 /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z',
2190 /// but non-ASCII letters are unchanged.
2192 /// To return a new lowercased value without modifying the existing one, use
2193 /// [`to_ascii_lowercase`].
2195 /// [`to_ascii_lowercase`]: #method.to_ascii_lowercase
2196 #[stable(feature = "ascii_methods_on_intrinsics", since = "1.21.0")]
2198 pub fn make_ascii_lowercase(&mut self) {
2199 let me = unsafe { self.as_bytes_mut() };
2200 me.make_ascii_lowercase()
2204 /// Converts a boxed slice of bytes to a boxed string slice without checking
2205 /// that the string contains valid UTF-8.
2212 /// let smile_utf8 = Box::new([226, 152, 186]);
2213 /// let smile = unsafe { std::str::from_boxed_utf8_unchecked(smile_utf8) };
2215 /// assert_eq!("☺", &*smile);
2217 #[stable(feature = "str_box_extras", since = "1.20.0")]
2218 pub unsafe fn from_boxed_utf8_unchecked(v: Box<[u8]>) -> Box<str> {
2219 Box::from_raw(Box::into_raw(v) as *mut str)