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 /// use std::ascii::AsciiExt;
395 /// let mut v = String::from("hello");
396 /// // correct length
397 /// assert!(v.get_mut(0..5).is_some());
399 /// assert!(v.get_mut(..42).is_none());
400 /// assert_eq!(Some("he"), v.get_mut(0..2).map(|v| &*v));
402 /// assert_eq!("hello", v);
404 /// let s = v.get_mut(0..2);
405 /// let s = s.map(|s| {
406 /// s.make_ascii_uppercase();
409 /// assert_eq!(Some("HE"), s);
411 /// assert_eq!("HEllo", v);
413 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
415 pub fn get_mut<I: SliceIndex<str>>(&mut self, i: I) -> Option<&mut I::Output> {
416 core_str::StrExt::get_mut(self, i)
419 /// Returns a unchecked subslice of `str`.
421 /// This is the unchecked alternative to indexing the `str`.
425 /// Callers of this function are responsible that these preconditions are
428 /// * The starting index must come before the ending index;
429 /// * Indexes must be within bounds of the original slice;
430 /// * Indexes must lie on UTF-8 sequence boundaries.
432 /// Failing that, the returned string slice may reference invalid memory or
433 /// violate the invariants communicated by the `str` type.
440 /// assert_eq!("🗻", v.get_unchecked(0..4));
441 /// assert_eq!("∈", v.get_unchecked(4..7));
442 /// assert_eq!("🌏", v.get_unchecked(7..11));
445 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
447 pub unsafe fn get_unchecked<I: SliceIndex<str>>(&self, i: I) -> &I::Output {
448 core_str::StrExt::get_unchecked(self, i)
451 /// Returns a mutable, unchecked subslice of `str`.
453 /// This is the unchecked alternative to indexing the `str`.
457 /// Callers of this function are responsible that these preconditions are
460 /// * The starting index must come before the ending index;
461 /// * Indexes must be within bounds of the original slice;
462 /// * Indexes must lie on UTF-8 sequence boundaries.
464 /// Failing that, the returned string slice may reference invalid memory or
465 /// violate the invariants communicated by the `str` type.
470 /// let mut v = String::from("🗻∈🌏");
472 /// assert_eq!("🗻", v.get_unchecked_mut(0..4));
473 /// assert_eq!("∈", v.get_unchecked_mut(4..7));
474 /// assert_eq!("🌏", v.get_unchecked_mut(7..11));
477 #[stable(feature = "str_checked_slicing", since = "1.20.0")]
479 pub unsafe fn get_unchecked_mut<I: SliceIndex<str>>(&mut self, i: I) -> &mut I::Output {
480 core_str::StrExt::get_unchecked_mut(self, i)
483 /// Creates a string slice from another string slice, bypassing safety
486 /// This is generally not recommended, use with caution! For a safe
487 /// alternative see [`str`] and [`Index`].
489 /// [`str`]: primitive.str.html
490 /// [`Index`]: ops/trait.Index.html
492 /// This new slice goes from `begin` to `end`, including `begin` but
495 /// To get a mutable string slice instead, see the
496 /// [`slice_mut_unchecked`] method.
498 /// [`slice_mut_unchecked`]: #method.slice_mut_unchecked
502 /// Callers of this function are responsible that three preconditions are
505 /// * `begin` must come before `end`.
506 /// * `begin` and `end` must be byte positions within the string slice.
507 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
514 /// let s = "Löwe 老虎 Léopard";
517 /// assert_eq!("Löwe 老虎 Léopard", s.slice_unchecked(0, 21));
520 /// let s = "Hello, world!";
523 /// assert_eq!("world", s.slice_unchecked(7, 12));
526 #[stable(feature = "rust1", since = "1.0.0")]
528 pub unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
529 core_str::StrExt::slice_unchecked(self, begin, end)
532 /// Creates a string slice from another string slice, bypassing safety
534 /// This is generally not recommended, use with caution! For a safe
535 /// alternative see [`str`] and [`IndexMut`].
537 /// [`str`]: primitive.str.html
538 /// [`IndexMut`]: ops/trait.IndexMut.html
540 /// This new slice goes from `begin` to `end`, including `begin` but
543 /// To get an immutable string slice instead, see the
544 /// [`slice_unchecked`] method.
546 /// [`slice_unchecked`]: #method.slice_unchecked
550 /// Callers of this function are responsible that three preconditions are
553 /// * `begin` must come before `end`.
554 /// * `begin` and `end` must be byte positions within the string slice.
555 /// * `begin` and `end` must lie on UTF-8 sequence boundaries.
556 #[stable(feature = "str_slice_mut", since = "1.5.0")]
558 pub unsafe fn slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str {
559 core_str::StrExt::slice_mut_unchecked(self, begin, end)
562 /// Divide one string slice into two at an index.
564 /// The argument, `mid`, should be a byte offset from the start of the
565 /// string. It must also be on the boundary of a UTF-8 code point.
567 /// The two slices returned go from the start of the string slice to `mid`,
568 /// and from `mid` to the end of the string slice.
570 /// To get mutable string slices instead, see the [`split_at_mut`]
573 /// [`split_at_mut`]: #method.split_at_mut
577 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
578 /// beyond the last code point of the string slice.
585 /// let s = "Per Martin-Löf";
587 /// let (first, last) = s.split_at(3);
589 /// assert_eq!("Per", first);
590 /// assert_eq!(" Martin-Löf", last);
593 #[stable(feature = "str_split_at", since = "1.4.0")]
594 pub fn split_at(&self, mid: usize) -> (&str, &str) {
595 core_str::StrExt::split_at(self, mid)
598 /// Divide one mutable string slice into two at an index.
600 /// The argument, `mid`, should be a byte offset from the start of the
601 /// string. It must also be on the boundary of a UTF-8 code point.
603 /// The two slices returned go from the start of the string slice to `mid`,
604 /// and from `mid` to the end of the string slice.
606 /// To get immutable string slices instead, see the [`split_at`] method.
608 /// [`split_at`]: #method.split_at
612 /// Panics if `mid` is not on a UTF-8 code point boundary, or if it is
613 /// beyond the last code point of the string slice.
620 /// use std::ascii::AsciiExt;
622 /// let mut s = "Per Martin-Löf".to_string();
624 /// let (first, last) = s.split_at_mut(3);
625 /// first.make_ascii_uppercase();
626 /// assert_eq!("PER", first);
627 /// assert_eq!(" Martin-Löf", last);
629 /// assert_eq!("PER Martin-Löf", s);
632 #[stable(feature = "str_split_at", since = "1.4.0")]
633 pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {
634 core_str::StrExt::split_at_mut(self, mid)
637 /// Returns an iterator over the [`char`]s of a string slice.
639 /// As a string slice consists of valid UTF-8, we can iterate through a
640 /// string slice by [`char`]. This method returns such an iterator.
642 /// It's important to remember that [`char`] represents a Unicode Scalar
643 /// Value, and may not match your idea of what a 'character' is. Iteration
644 /// over grapheme clusters may be what you actually want.
646 /// [`char`]: primitive.char.html
653 /// let word = "goodbye";
655 /// let count = word.chars().count();
656 /// assert_eq!(7, count);
658 /// let mut chars = word.chars();
660 /// assert_eq!(Some('g'), chars.next());
661 /// assert_eq!(Some('o'), chars.next());
662 /// assert_eq!(Some('o'), chars.next());
663 /// assert_eq!(Some('d'), chars.next());
664 /// assert_eq!(Some('b'), chars.next());
665 /// assert_eq!(Some('y'), chars.next());
666 /// assert_eq!(Some('e'), chars.next());
668 /// assert_eq!(None, chars.next());
671 /// Remember, [`char`]s may not match your human intuition about characters:
676 /// let mut chars = y.chars();
678 /// assert_eq!(Some('y'), chars.next()); // not 'y̆'
679 /// assert_eq!(Some('\u{0306}'), chars.next());
681 /// assert_eq!(None, chars.next());
683 #[stable(feature = "rust1", since = "1.0.0")]
685 pub fn chars(&self) -> Chars {
686 core_str::StrExt::chars(self)
688 /// Returns an iterator over the [`char`]s of a string slice, and their
691 /// As a string slice consists of valid UTF-8, we can iterate through a
692 /// string slice by [`char`]. This method returns an iterator of both
693 /// these [`char`]s, as well as their byte positions.
695 /// The iterator yields tuples. The position is first, the [`char`] is
698 /// [`char`]: primitive.char.html
705 /// let word = "goodbye";
707 /// let count = word.char_indices().count();
708 /// assert_eq!(7, count);
710 /// let mut char_indices = word.char_indices();
712 /// assert_eq!(Some((0, 'g')), char_indices.next());
713 /// assert_eq!(Some((1, 'o')), char_indices.next());
714 /// assert_eq!(Some((2, 'o')), char_indices.next());
715 /// assert_eq!(Some((3, 'd')), char_indices.next());
716 /// assert_eq!(Some((4, 'b')), char_indices.next());
717 /// assert_eq!(Some((5, 'y')), char_indices.next());
718 /// assert_eq!(Some((6, 'e')), char_indices.next());
720 /// assert_eq!(None, char_indices.next());
723 /// Remember, [`char`]s may not match your human intuition about characters:
728 /// let mut char_indices = y.char_indices();
730 /// assert_eq!(Some((0, 'y')), char_indices.next()); // not (0, 'y̆')
731 /// assert_eq!(Some((1, '\u{0306}')), char_indices.next());
733 /// assert_eq!(None, char_indices.next());
735 #[stable(feature = "rust1", since = "1.0.0")]
737 pub fn char_indices(&self) -> CharIndices {
738 core_str::StrExt::char_indices(self)
741 /// An iterator over the bytes of a string slice.
743 /// As a string slice consists of a sequence of bytes, we can iterate
744 /// through a string slice by byte. This method returns such an iterator.
751 /// let mut bytes = "bors".bytes();
753 /// assert_eq!(Some(b'b'), bytes.next());
754 /// assert_eq!(Some(b'o'), bytes.next());
755 /// assert_eq!(Some(b'r'), bytes.next());
756 /// assert_eq!(Some(b's'), bytes.next());
758 /// assert_eq!(None, bytes.next());
760 #[stable(feature = "rust1", since = "1.0.0")]
762 pub fn bytes(&self) -> Bytes {
763 core_str::StrExt::bytes(self)
766 /// Split a string slice by whitespace.
768 /// The iterator returned will return string slices that are sub-slices of
769 /// the original string slice, separated by any amount of whitespace.
771 /// 'Whitespace' is defined according to the terms of the Unicode Derived
772 /// Core Property `White_Space`.
779 /// let mut iter = "A few words".split_whitespace();
781 /// assert_eq!(Some("A"), iter.next());
782 /// assert_eq!(Some("few"), iter.next());
783 /// assert_eq!(Some("words"), iter.next());
785 /// assert_eq!(None, iter.next());
788 /// All kinds of whitespace are considered:
791 /// let mut iter = " Mary had\ta\u{2009}little \n\t lamb".split_whitespace();
792 /// assert_eq!(Some("Mary"), iter.next());
793 /// assert_eq!(Some("had"), iter.next());
794 /// assert_eq!(Some("a"), iter.next());
795 /// assert_eq!(Some("little"), iter.next());
796 /// assert_eq!(Some("lamb"), iter.next());
798 /// assert_eq!(None, iter.next());
800 #[stable(feature = "split_whitespace", since = "1.1.0")]
802 pub fn split_whitespace(&self) -> SplitWhitespace {
803 UnicodeStr::split_whitespace(self)
806 /// An iterator over the lines of a string, as string slices.
808 /// Lines are ended with either a newline (`\n`) or a carriage return with
809 /// a line feed (`\r\n`).
811 /// The final line ending is optional.
818 /// let text = "foo\r\nbar\n\nbaz\n";
819 /// let mut lines = text.lines();
821 /// assert_eq!(Some("foo"), lines.next());
822 /// assert_eq!(Some("bar"), lines.next());
823 /// assert_eq!(Some(""), lines.next());
824 /// assert_eq!(Some("baz"), lines.next());
826 /// assert_eq!(None, lines.next());
829 /// The final line ending isn't required:
832 /// let text = "foo\nbar\n\r\nbaz";
833 /// let mut lines = text.lines();
835 /// assert_eq!(Some("foo"), lines.next());
836 /// assert_eq!(Some("bar"), lines.next());
837 /// assert_eq!(Some(""), lines.next());
838 /// assert_eq!(Some("baz"), lines.next());
840 /// assert_eq!(None, lines.next());
842 #[stable(feature = "rust1", since = "1.0.0")]
844 pub fn lines(&self) -> Lines {
845 core_str::StrExt::lines(self)
848 /// An iterator over the lines of a string.
849 #[stable(feature = "rust1", since = "1.0.0")]
850 #[rustc_deprecated(since = "1.4.0", reason = "use lines() instead now")]
853 pub fn lines_any(&self) -> LinesAny {
854 core_str::StrExt::lines_any(self)
857 /// Returns an iterator of `u16` over the string encoded as UTF-16.
864 /// let text = "Zażółć gęślą jaźń";
866 /// let utf8_len = text.len();
867 /// let utf16_len = text.encode_utf16().count();
869 /// assert!(utf16_len <= utf8_len);
871 #[stable(feature = "encode_utf16", since = "1.8.0")]
872 pub fn encode_utf16(&self) -> EncodeUtf16 {
873 EncodeUtf16 { encoder: Utf16Encoder::new(self[..].chars()) }
876 /// Returns `true` if the given pattern matches a sub-slice of
877 /// this string slice.
879 /// Returns `false` if it does not.
886 /// let bananas = "bananas";
888 /// assert!(bananas.contains("nana"));
889 /// assert!(!bananas.contains("apples"));
891 #[stable(feature = "rust1", since = "1.0.0")]
893 pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
894 core_str::StrExt::contains(self, pat)
897 /// Returns `true` if the given pattern matches a prefix of this
900 /// Returns `false` if it does not.
907 /// let bananas = "bananas";
909 /// assert!(bananas.starts_with("bana"));
910 /// assert!(!bananas.starts_with("nana"));
912 #[stable(feature = "rust1", since = "1.0.0")]
913 pub fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
914 core_str::StrExt::starts_with(self, pat)
917 /// Returns `true` if the given pattern matches a suffix of this
920 /// Returns `false` if it does not.
927 /// let bananas = "bananas";
929 /// assert!(bananas.ends_with("anas"));
930 /// assert!(!bananas.ends_with("nana"));
932 #[stable(feature = "rust1", since = "1.0.0")]
933 pub fn ends_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool
934 where P::Searcher: ReverseSearcher<'a>
936 core_str::StrExt::ends_with(self, pat)
939 /// Returns the byte index of the first character of this string slice that
940 /// matches the pattern.
942 /// Returns [`None`] if the pattern doesn't match.
944 /// The pattern can be a `&str`, [`char`], or a closure that determines if
945 /// a character matches.
947 /// [`char`]: primitive.char.html
948 /// [`None`]: option/enum.Option.html#variant.None
955 /// let s = "Löwe 老虎 Léopard";
957 /// assert_eq!(s.find('L'), Some(0));
958 /// assert_eq!(s.find('é'), Some(14));
959 /// assert_eq!(s.find("Léopard"), Some(13));
962 /// More complex patterns using point-free style and closures:
965 /// let s = "Löwe 老虎 Léopard";
967 /// assert_eq!(s.find(char::is_whitespace), Some(5));
968 /// assert_eq!(s.find(char::is_lowercase), Some(1));
969 /// assert_eq!(s.find(|c: char| c.is_whitespace() || c.is_lowercase()), Some(1));
970 /// assert_eq!(s.find(|c: char| (c < 'o') && (c > 'a')), Some(4));
973 /// Not finding the pattern:
976 /// let s = "Löwe 老虎 Léopard";
977 /// let x: &[_] = &['1', '2'];
979 /// assert_eq!(s.find(x), None);
981 #[stable(feature = "rust1", since = "1.0.0")]
983 pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
984 core_str::StrExt::find(self, pat)
987 /// Returns the byte index of the last character of this string slice that
988 /// matches the pattern.
990 /// Returns [`None`] if the pattern doesn't match.
992 /// The pattern can be a `&str`, [`char`], or a closure that determines if
993 /// a character matches.
995 /// [`char`]: primitive.char.html
996 /// [`None`]: option/enum.Option.html#variant.None
1000 /// Simple patterns:
1003 /// let s = "Löwe 老虎 Léopard";
1005 /// assert_eq!(s.rfind('L'), Some(13));
1006 /// assert_eq!(s.rfind('é'), Some(14));
1009 /// More complex patterns with closures:
1012 /// let s = "Löwe 老虎 Léopard";
1014 /// assert_eq!(s.rfind(char::is_whitespace), Some(12));
1015 /// assert_eq!(s.rfind(char::is_lowercase), Some(20));
1018 /// Not finding the pattern:
1021 /// let s = "Löwe 老虎 Léopard";
1022 /// let x: &[_] = &['1', '2'];
1024 /// assert_eq!(s.rfind(x), None);
1026 #[stable(feature = "rust1", since = "1.0.0")]
1028 pub fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
1029 where P::Searcher: ReverseSearcher<'a>
1031 core_str::StrExt::rfind(self, pat)
1034 /// An iterator over substrings of this string slice, separated by
1035 /// characters matched by a pattern.
1037 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1040 /// # Iterator behavior
1042 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1043 /// allows a reverse search and forward/reverse search yields the same
1044 /// elements. This is true for, eg, [`char`] but not for `&str`.
1046 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1048 /// If the pattern allows a reverse search but its results might differ
1049 /// from a forward search, the [`rsplit`] method can be used.
1051 /// [`char`]: primitive.char.html
1052 /// [`rsplit`]: #method.rsplit
1056 /// Simple patterns:
1059 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
1060 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
1062 /// let v: Vec<&str> = "".split('X').collect();
1063 /// assert_eq!(v, [""]);
1065 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
1066 /// assert_eq!(v, ["lion", "", "tiger", "leopard"]);
1068 /// let v: Vec<&str> = "lion::tiger::leopard".split("::").collect();
1069 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1071 /// let v: Vec<&str> = "abc1def2ghi".split(char::is_numeric).collect();
1072 /// assert_eq!(v, ["abc", "def", "ghi"]);
1074 /// let v: Vec<&str> = "lionXtigerXleopard".split(char::is_uppercase).collect();
1075 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1078 /// A more complex pattern, using a closure:
1081 /// let v: Vec<&str> = "abc1defXghi".split(|c| c == '1' || c == 'X').collect();
1082 /// assert_eq!(v, ["abc", "def", "ghi"]);
1085 /// If a string contains multiple contiguous separators, you will end up
1086 /// with empty strings in the output:
1089 /// let x = "||||a||b|c".to_string();
1090 /// let d: Vec<_> = x.split('|').collect();
1092 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1095 /// Contiguous separators are separated by the empty string.
1098 /// let x = "(///)".to_string();
1099 /// let d: Vec<_> = x.split('/').collect();
1101 /// assert_eq!(d, &["(", "", "", ")"]);
1104 /// Separators at the start or end of a string are neighbored
1105 /// by empty strings.
1108 /// let d: Vec<_> = "010".split("0").collect();
1109 /// assert_eq!(d, &["", "1", ""]);
1112 /// When the empty string is used as a separator, it separates
1113 /// every character in the string, along with the beginning
1114 /// and end of the string.
1117 /// let f: Vec<_> = "rust".split("").collect();
1118 /// assert_eq!(f, &["", "r", "u", "s", "t", ""]);
1121 /// Contiguous separators can lead to possibly surprising behavior
1122 /// when whitespace is used as the separator. This code is correct:
1125 /// let x = " a b c".to_string();
1126 /// let d: Vec<_> = x.split(' ').collect();
1128 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
1131 /// It does _not_ give you:
1134 /// assert_eq!(d, &["a", "b", "c"]);
1137 /// Use [`split_whitespace`] for this behavior.
1139 /// [`split_whitespace`]: #method.split_whitespace
1140 #[stable(feature = "rust1", since = "1.0.0")]
1142 pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
1143 core_str::StrExt::split(self, pat)
1146 /// An iterator over substrings of the given string slice, separated by
1147 /// characters matched by a pattern and yielded in reverse order.
1149 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1152 /// [`char`]: primitive.char.html
1154 /// # Iterator behavior
1156 /// The returned iterator requires that the pattern supports a reverse
1157 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1158 /// search yields the same elements.
1160 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1162 /// For iterating from the front, the [`split`] method can be used.
1164 /// [`split`]: #method.split
1168 /// Simple patterns:
1171 /// let v: Vec<&str> = "Mary had a little lamb".rsplit(' ').collect();
1172 /// assert_eq!(v, ["lamb", "little", "a", "had", "Mary"]);
1174 /// let v: Vec<&str> = "".rsplit('X').collect();
1175 /// assert_eq!(v, [""]);
1177 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplit('X').collect();
1178 /// assert_eq!(v, ["leopard", "tiger", "", "lion"]);
1180 /// let v: Vec<&str> = "lion::tiger::leopard".rsplit("::").collect();
1181 /// assert_eq!(v, ["leopard", "tiger", "lion"]);
1184 /// A more complex pattern, using a closure:
1187 /// let v: Vec<&str> = "abc1defXghi".rsplit(|c| c == '1' || c == 'X').collect();
1188 /// assert_eq!(v, ["ghi", "def", "abc"]);
1190 #[stable(feature = "rust1", since = "1.0.0")]
1192 pub fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
1193 where P::Searcher: ReverseSearcher<'a>
1195 core_str::StrExt::rsplit(self, pat)
1198 /// An iterator over substrings of the given string slice, separated by
1199 /// characters matched by a pattern.
1201 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1204 /// Equivalent to [`split`], except that the trailing substring
1205 /// is skipped if empty.
1207 /// [`split`]: #method.split
1209 /// This method can be used for string data that is _terminated_,
1210 /// rather than _separated_ by a pattern.
1212 /// # Iterator behavior
1214 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1215 /// allows a reverse search and forward/reverse search yields the same
1216 /// elements. This is true for, eg, [`char`] but not for `&str`.
1218 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1219 /// [`char`]: primitive.char.html
1221 /// If the pattern allows a reverse search but its results might differ
1222 /// from a forward search, the [`rsplit_terminator`] method can be used.
1224 /// [`rsplit_terminator`]: #method.rsplit_terminator
1231 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
1232 /// assert_eq!(v, ["A", "B"]);
1234 /// let v: Vec<&str> = "A..B..".split_terminator(".").collect();
1235 /// assert_eq!(v, ["A", "", "B", ""]);
1237 #[stable(feature = "rust1", since = "1.0.0")]
1239 pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
1240 core_str::StrExt::split_terminator(self, pat)
1243 /// An iterator over substrings of `self`, separated by characters
1244 /// matched by a pattern and yielded in reverse order.
1246 /// The pattern can be a simple `&str`, [`char`], or a closure that
1247 /// determines the split.
1248 /// Additional libraries might provide more complex patterns like
1249 /// regular expressions.
1251 /// [`char`]: primitive.char.html
1253 /// Equivalent to [`split`], except that the trailing substring is
1254 /// skipped if empty.
1256 /// [`split`]: #method.split
1258 /// This method can be used for string data that is _terminated_,
1259 /// rather than _separated_ by a pattern.
1261 /// # Iterator behavior
1263 /// The returned iterator requires that the pattern supports a
1264 /// reverse search, and it will be double ended if a forward/reverse
1265 /// search yields the same elements.
1267 /// For iterating from the front, the [`split_terminator`] method can be
1270 /// [`split_terminator`]: #method.split_terminator
1275 /// let v: Vec<&str> = "A.B.".rsplit_terminator('.').collect();
1276 /// assert_eq!(v, ["B", "A"]);
1278 /// let v: Vec<&str> = "A..B..".rsplit_terminator(".").collect();
1279 /// assert_eq!(v, ["", "B", "", "A"]);
1281 #[stable(feature = "rust1", since = "1.0.0")]
1283 pub fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1284 where P::Searcher: ReverseSearcher<'a>
1286 core_str::StrExt::rsplit_terminator(self, pat)
1289 /// An iterator over substrings of the given string slice, separated by a
1290 /// pattern, restricted to returning at most `n` items.
1292 /// If `n` substrings are returned, the last substring (the `n`th substring)
1293 /// will contain the remainder of the string.
1295 /// The pattern can be a `&str`, [`char`], or a closure that determines the
1298 /// [`char`]: primitive.char.html
1300 /// # Iterator behavior
1302 /// The returned iterator will not be double ended, because it is
1303 /// not efficient to support.
1305 /// If the pattern allows a reverse search, the [`rsplitn`] method can be
1308 /// [`rsplitn`]: #method.rsplitn
1312 /// Simple patterns:
1315 /// let v: Vec<&str> = "Mary had a little lambda".splitn(3, ' ').collect();
1316 /// assert_eq!(v, ["Mary", "had", "a little lambda"]);
1318 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(3, "X").collect();
1319 /// assert_eq!(v, ["lion", "", "tigerXleopard"]);
1321 /// let v: Vec<&str> = "abcXdef".splitn(1, 'X').collect();
1322 /// assert_eq!(v, ["abcXdef"]);
1324 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1325 /// assert_eq!(v, [""]);
1328 /// A more complex pattern, using a closure:
1331 /// let v: Vec<&str> = "abc1defXghi".splitn(2, |c| c == '1' || c == 'X').collect();
1332 /// assert_eq!(v, ["abc", "defXghi"]);
1334 #[stable(feature = "rust1", since = "1.0.0")]
1336 pub fn splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P> {
1337 core_str::StrExt::splitn(self, n, pat)
1340 /// An iterator over substrings of this string slice, separated by a
1341 /// pattern, starting from the end of the string, restricted to returning
1342 /// at most `n` items.
1344 /// If `n` substrings are returned, the last substring (the `n`th substring)
1345 /// will contain the remainder of the string.
1347 /// The pattern can be a `&str`, [`char`], or a closure that
1348 /// determines the split.
1350 /// [`char`]: primitive.char.html
1352 /// # Iterator behavior
1354 /// The returned iterator will not be double ended, because it is not
1355 /// efficient to support.
1357 /// For splitting from the front, the [`splitn`] method can be used.
1359 /// [`splitn`]: #method.splitn
1363 /// Simple patterns:
1366 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(3, ' ').collect();
1367 /// assert_eq!(v, ["lamb", "little", "Mary had a"]);
1369 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(3, 'X').collect();
1370 /// assert_eq!(v, ["leopard", "tiger", "lionX"]);
1372 /// let v: Vec<&str> = "lion::tiger::leopard".rsplitn(2, "::").collect();
1373 /// assert_eq!(v, ["leopard", "lion::tiger"]);
1376 /// A more complex pattern, using a closure:
1379 /// let v: Vec<&str> = "abc1defXghi".rsplitn(2, |c| c == '1' || c == 'X').collect();
1380 /// assert_eq!(v, ["ghi", "abc1def"]);
1382 #[stable(feature = "rust1", since = "1.0.0")]
1384 pub fn rsplitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> RSplitN<'a, P>
1385 where P::Searcher: ReverseSearcher<'a>
1387 core_str::StrExt::rsplitn(self, n, pat)
1390 /// An iterator over the disjoint matches of a pattern within the given string
1393 /// The pattern can be a `&str`, [`char`], or a closure that
1394 /// determines if a character matches.
1396 /// [`char`]: primitive.char.html
1398 /// # Iterator behavior
1400 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1401 /// allows a reverse search and forward/reverse search yields the same
1402 /// elements. This is true for, eg, [`char`] but not for `&str`.
1404 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1405 /// [`char`]: primitive.char.html
1407 /// If the pattern allows a reverse search but its results might differ
1408 /// from a forward search, the [`rmatches`] method can be used.
1410 /// [`rmatches`]: #method.rmatches
1417 /// let v: Vec<&str> = "abcXXXabcYYYabc".matches("abc").collect();
1418 /// assert_eq!(v, ["abc", "abc", "abc"]);
1420 /// let v: Vec<&str> = "1abc2abc3".matches(char::is_numeric).collect();
1421 /// assert_eq!(v, ["1", "2", "3"]);
1423 #[stable(feature = "str_matches", since = "1.2.0")]
1425 pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1426 core_str::StrExt::matches(self, pat)
1429 /// An iterator over the disjoint matches of a pattern within this string slice,
1430 /// yielded in reverse order.
1432 /// The pattern can be a `&str`, [`char`], or a closure that determines if
1433 /// a character matches.
1435 /// [`char`]: primitive.char.html
1437 /// # Iterator behavior
1439 /// The returned iterator requires that the pattern supports a reverse
1440 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1441 /// search yields the same elements.
1443 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1445 /// For iterating from the front, the [`matches`] method can be used.
1447 /// [`matches`]: #method.matches
1454 /// let v: Vec<&str> = "abcXXXabcYYYabc".rmatches("abc").collect();
1455 /// assert_eq!(v, ["abc", "abc", "abc"]);
1457 /// let v: Vec<&str> = "1abc2abc3".rmatches(char::is_numeric).collect();
1458 /// assert_eq!(v, ["3", "2", "1"]);
1460 #[stable(feature = "str_matches", since = "1.2.0")]
1462 pub fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
1463 where P::Searcher: ReverseSearcher<'a>
1465 core_str::StrExt::rmatches(self, pat)
1468 /// An iterator over the disjoint matches of a pattern within this string
1469 /// slice as well as the index that the match starts at.
1471 /// For matches of `pat` within `self` that overlap, only the indices
1472 /// corresponding to the first match are returned.
1474 /// The pattern can be a `&str`, [`char`], or a closure that determines
1475 /// if a character matches.
1477 /// [`char`]: primitive.char.html
1479 /// # Iterator behavior
1481 /// The returned iterator will be a [`DoubleEndedIterator`] if the pattern
1482 /// allows a reverse search and forward/reverse search yields the same
1483 /// elements. This is true for, eg, [`char`] but not for `&str`.
1485 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1487 /// If the pattern allows a reverse search but its results might differ
1488 /// from a forward search, the [`rmatch_indices`] method can be used.
1490 /// [`rmatch_indices`]: #method.rmatch_indices
1497 /// let v: Vec<_> = "abcXXXabcYYYabc".match_indices("abc").collect();
1498 /// assert_eq!(v, [(0, "abc"), (6, "abc"), (12, "abc")]);
1500 /// let v: Vec<_> = "1abcabc2".match_indices("abc").collect();
1501 /// assert_eq!(v, [(1, "abc"), (4, "abc")]);
1503 /// let v: Vec<_> = "ababa".match_indices("aba").collect();
1504 /// assert_eq!(v, [(0, "aba")]); // only the first `aba`
1506 #[stable(feature = "str_match_indices", since = "1.5.0")]
1508 pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1509 core_str::StrExt::match_indices(self, pat)
1512 /// An iterator over the disjoint matches of a pattern within `self`,
1513 /// yielded in reverse order along with the index of the match.
1515 /// For matches of `pat` within `self` that overlap, only the indices
1516 /// corresponding to the last match are returned.
1518 /// The pattern can be a `&str`, [`char`], or a closure that determines if a
1519 /// character matches.
1521 /// [`char`]: primitive.char.html
1523 /// # Iterator behavior
1525 /// The returned iterator requires that the pattern supports a reverse
1526 /// search, and it will be a [`DoubleEndedIterator`] if a forward/reverse
1527 /// search yields the same elements.
1529 /// [`DoubleEndedIterator`]: iter/trait.DoubleEndedIterator.html
1531 /// For iterating from the front, the [`match_indices`] method can be used.
1533 /// [`match_indices`]: #method.match_indices
1540 /// let v: Vec<_> = "abcXXXabcYYYabc".rmatch_indices("abc").collect();
1541 /// assert_eq!(v, [(12, "abc"), (6, "abc"), (0, "abc")]);
1543 /// let v: Vec<_> = "1abcabc2".rmatch_indices("abc").collect();
1544 /// assert_eq!(v, [(4, "abc"), (1, "abc")]);
1546 /// let v: Vec<_> = "ababa".rmatch_indices("aba").collect();
1547 /// assert_eq!(v, [(2, "aba")]); // only the last `aba`
1549 #[stable(feature = "str_match_indices", since = "1.5.0")]
1551 pub fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
1552 where P::Searcher: ReverseSearcher<'a>
1554 core_str::StrExt::rmatch_indices(self, pat)
1557 /// Returns a string slice with leading and trailing whitespace removed.
1559 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1560 /// Core Property `White_Space`.
1567 /// let s = " Hello\tworld\t";
1569 /// assert_eq!("Hello\tworld", s.trim());
1571 #[stable(feature = "rust1", since = "1.0.0")]
1572 pub fn trim(&self) -> &str {
1573 UnicodeStr::trim(self)
1576 /// Returns a string slice with leading whitespace removed.
1578 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1579 /// Core Property `White_Space`.
1581 /// # Text directionality
1583 /// A string is a sequence of bytes. 'Left' in this context means the first
1584 /// position of that byte string; for a language like Arabic or Hebrew
1585 /// which are 'right to left' rather than 'left to right', this will be
1586 /// the _right_ side, not the left.
1593 /// let s = " Hello\tworld\t";
1595 /// assert_eq!("Hello\tworld\t", s.trim_left());
1601 /// let s = " English";
1602 /// assert!(Some('E') == s.trim_left().chars().next());
1604 /// let s = " עברית";
1605 /// assert!(Some('ע') == s.trim_left().chars().next());
1607 #[stable(feature = "rust1", since = "1.0.0")]
1608 pub fn trim_left(&self) -> &str {
1609 UnicodeStr::trim_left(self)
1612 /// Returns a string slice with trailing whitespace removed.
1614 /// 'Whitespace' is defined according to the terms of the Unicode Derived
1615 /// Core Property `White_Space`.
1617 /// # Text directionality
1619 /// A string is a sequence of bytes. 'Right' in this context means the last
1620 /// position of that byte string; for a language like Arabic or Hebrew
1621 /// which are 'right to left' rather than 'left to right', this will be
1622 /// the _left_ side, not the right.
1629 /// let s = " Hello\tworld\t";
1631 /// assert_eq!(" Hello\tworld", s.trim_right());
1637 /// let s = "English ";
1638 /// assert!(Some('h') == s.trim_right().chars().rev().next());
1640 /// let s = "עברית ";
1641 /// assert!(Some('ת') == s.trim_right().chars().rev().next());
1643 #[stable(feature = "rust1", since = "1.0.0")]
1644 pub fn trim_right(&self) -> &str {
1645 UnicodeStr::trim_right(self)
1648 /// Returns a string slice with all prefixes and suffixes that match a
1649 /// pattern repeatedly removed.
1651 /// The pattern can be a [`char`] or a closure that determines if a
1652 /// character matches.
1654 /// [`char`]: primitive.char.html
1658 /// Simple patterns:
1661 /// assert_eq!("11foo1bar11".trim_matches('1'), "foo1bar");
1662 /// assert_eq!("123foo1bar123".trim_matches(char::is_numeric), "foo1bar");
1664 /// let x: &[_] = &['1', '2'];
1665 /// assert_eq!("12foo1bar12".trim_matches(x), "foo1bar");
1668 /// A more complex pattern, using a closure:
1671 /// assert_eq!("1foo1barXX".trim_matches(|c| c == '1' || c == 'X'), "foo1bar");
1673 #[stable(feature = "rust1", since = "1.0.0")]
1674 pub fn trim_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1675 where P::Searcher: DoubleEndedSearcher<'a>
1677 core_str::StrExt::trim_matches(self, pat)
1680 /// Returns a string slice with all prefixes that match a pattern
1681 /// repeatedly removed.
1683 /// The pattern can be a `&str`, [`char`], or a closure that determines if
1684 /// a character matches.
1686 /// [`char`]: primitive.char.html
1688 /// # Text directionality
1690 /// A string is a sequence of bytes. 'Left' in this context means the first
1691 /// position of that byte string; for a language like Arabic or Hebrew
1692 /// which are 'right to left' rather than 'left to right', this will be
1693 /// the _right_ side, not the left.
1700 /// assert_eq!("11foo1bar11".trim_left_matches('1'), "foo1bar11");
1701 /// assert_eq!("123foo1bar123".trim_left_matches(char::is_numeric), "foo1bar123");
1703 /// let x: &[_] = &['1', '2'];
1704 /// assert_eq!("12foo1bar12".trim_left_matches(x), "foo1bar12");
1706 #[stable(feature = "rust1", since = "1.0.0")]
1707 pub fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
1708 core_str::StrExt::trim_left_matches(self, pat)
1711 /// Returns a string slice with all suffixes that match a pattern
1712 /// repeatedly removed.
1714 /// The pattern can be a `&str`, [`char`], or a closure that
1715 /// determines if a character matches.
1717 /// [`char`]: primitive.char.html
1719 /// # Text directionality
1721 /// A string is a sequence of bytes. 'Right' in this context means the last
1722 /// position of that byte string; for a language like Arabic or Hebrew
1723 /// which are 'right to left' rather than 'left to right', this will be
1724 /// the _left_ side, not the right.
1728 /// Simple patterns:
1731 /// assert_eq!("11foo1bar11".trim_right_matches('1'), "11foo1bar");
1732 /// assert_eq!("123foo1bar123".trim_right_matches(char::is_numeric), "123foo1bar");
1734 /// let x: &[_] = &['1', '2'];
1735 /// assert_eq!("12foo1bar12".trim_right_matches(x), "12foo1bar");
1738 /// A more complex pattern, using a closure:
1741 /// assert_eq!("1fooX".trim_left_matches(|c| c == '1' || c == 'X'), "fooX");
1743 #[stable(feature = "rust1", since = "1.0.0")]
1744 pub fn trim_right_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1745 where P::Searcher: ReverseSearcher<'a>
1747 core_str::StrExt::trim_right_matches(self, pat)
1750 /// Parses this string slice into another type.
1752 /// Because `parse` is so general, it can cause problems with type
1753 /// inference. As such, `parse` is one of the few times you'll see
1754 /// the syntax affectionately known as the 'turbofish': `::<>`. This
1755 /// helps the inference algorithm understand specifically which type
1756 /// you're trying to parse into.
1758 /// `parse` can parse any type that implements the [`FromStr`] trait.
1760 /// [`FromStr`]: str/trait.FromStr.html
1764 /// Will return [`Err`] if it's not possible to parse this string slice into
1765 /// the desired type.
1767 /// [`Err`]: str/trait.FromStr.html#associatedtype.Err
1774 /// let four: u32 = "4".parse().unwrap();
1776 /// assert_eq!(4, four);
1779 /// Using the 'turbofish' instead of annotating `four`:
1782 /// let four = "4".parse::<u32>();
1784 /// assert_eq!(Ok(4), four);
1787 /// Failing to parse:
1790 /// let nope = "j".parse::<u32>();
1792 /// assert!(nope.is_err());
1795 #[stable(feature = "rust1", since = "1.0.0")]
1796 pub fn parse<F: FromStr>(&self) -> Result<F, F::Err> {
1797 core_str::StrExt::parse(self)
1800 /// Converts a `Box<str>` into a `Box<[u8]>` without copying or allocating.
1807 /// let s = "this is a string";
1808 /// let boxed_str = s.to_owned().into_boxed_str();
1809 /// let boxed_bytes = boxed_str.into_boxed_bytes();
1810 /// assert_eq!(*boxed_bytes, *s.as_bytes());
1812 #[stable(feature = "str_box_extras", since = "1.20.0")]
1813 pub fn into_boxed_bytes(self: Box<str>) -> Box<[u8]> {
1817 /// Replaces all matches of a pattern with another string.
1819 /// `replace` creates a new [`String`], and copies the data from this string slice into it.
1820 /// While doing so, it attempts to find matches of a pattern. If it finds any, it
1821 /// replaces them with the replacement string slice.
1823 /// [`String`]: string/struct.String.html
1830 /// let s = "this is old";
1832 /// assert_eq!("this is new", s.replace("old", "new"));
1835 /// When the pattern doesn't match:
1838 /// let s = "this is old";
1839 /// assert_eq!(s, s.replace("cookie monster", "little lamb"));
1841 #[stable(feature = "rust1", since = "1.0.0")]
1843 pub fn replace<'a, P: Pattern<'a>>(&'a self, from: P, to: &str) -> String {
1844 let mut result = String::new();
1845 let mut last_end = 0;
1846 for (start, part) in self.match_indices(from) {
1847 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1848 result.push_str(to);
1849 last_end = start + part.len();
1851 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1855 /// Replaces first N matches of a pattern with another string.
1857 /// `replacen` creates a new [`String`], and copies the data from this string slice into it.
1858 /// While doing so, it attempts to find matches of a pattern. If it finds any, it
1859 /// replaces them with the replacement string slice at most `count` times.
1861 /// [`String`]: string/struct.String.html
1868 /// let s = "foo foo 123 foo";
1869 /// assert_eq!("new new 123 foo", s.replacen("foo", "new", 2));
1870 /// assert_eq!("faa fao 123 foo", s.replacen('o', "a", 3));
1871 /// assert_eq!("foo foo new23 foo", s.replacen(char::is_numeric, "new", 1));
1874 /// When the pattern doesn't match:
1877 /// let s = "this is old";
1878 /// assert_eq!(s, s.replacen("cookie monster", "little lamb", 10));
1880 #[stable(feature = "str_replacen", since = "1.16.0")]
1881 pub fn replacen<'a, P: Pattern<'a>>(&'a self, pat: P, to: &str, count: usize) -> String {
1882 // Hope to reduce the times of re-allocation
1883 let mut result = String::with_capacity(32);
1884 let mut last_end = 0;
1885 for (start, part) in self.match_indices(pat).take(count) {
1886 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1887 result.push_str(to);
1888 last_end = start + part.len();
1890 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1894 /// Returns the lowercase equivalent of this string slice, as a new [`String`].
1896 /// 'Lowercase' is defined according to the terms of the Unicode Derived Core Property
1899 /// Since some characters can expand into multiple characters when changing
1900 /// the case, this function returns a [`String`] instead of modifying the
1901 /// parameter in-place.
1903 /// [`String`]: string/struct.String.html
1910 /// let s = "HELLO";
1912 /// assert_eq!("hello", s.to_lowercase());
1915 /// A tricky example, with sigma:
1918 /// let sigma = "Σ";
1920 /// assert_eq!("σ", sigma.to_lowercase());
1922 /// // but at the end of a word, it's ς, not σ:
1923 /// let odysseus = "ὈΔΥΣΣΕΎΣ";
1925 /// assert_eq!("ὀδυσσεύς", odysseus.to_lowercase());
1928 /// Languages without case are not changed:
1931 /// let new_year = "农历新年";
1933 /// assert_eq!(new_year, new_year.to_lowercase());
1935 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
1936 pub fn to_lowercase(&self) -> String {
1937 let mut s = String::with_capacity(self.len());
1938 for (i, c) in self[..].char_indices() {
1940 // Σ maps to σ, except at the end of a word where it maps to ς.
1941 // This is the only conditional (contextual) but language-independent mapping
1942 // in `SpecialCasing.txt`,
1943 // so hard-code it rather than have a generic "condition" mechanism.
1944 // See https://github.com/rust-lang/rust/issues/26035
1945 map_uppercase_sigma(self, i, &mut s)
1947 s.extend(c.to_lowercase());
1952 fn map_uppercase_sigma(from: &str, i: usize, to: &mut String) {
1953 // See http://www.unicode.org/versions/Unicode7.0.0/ch03.pdf#G33992
1954 // for the definition of `Final_Sigma`.
1955 debug_assert!('Σ'.len_utf8() == 2);
1956 let is_word_final = case_ignoreable_then_cased(from[..i].chars().rev()) &&
1957 !case_ignoreable_then_cased(from[i + 2..].chars());
1958 to.push_str(if is_word_final { "ς" } else { "σ" });
1961 fn case_ignoreable_then_cased<I: Iterator<Item = char>>(iter: I) -> bool {
1962 use std_unicode::derived_property::{Cased, Case_Ignorable};
1963 match iter.skip_while(|&c| Case_Ignorable(c)).next() {
1964 Some(c) => Cased(c),
1970 /// Returns the uppercase equivalent of this string slice, as a new [`String`].
1972 /// 'Uppercase' is defined according to the terms of the Unicode Derived Core Property
1975 /// Since some characters can expand into multiple characters when changing
1976 /// the case, this function returns a [`String`] instead of modifying the
1977 /// parameter in-place.
1979 /// [`String`]: string/struct.String.html
1986 /// let s = "hello";
1988 /// assert_eq!("HELLO", s.to_uppercase());
1991 /// Scripts without case are not changed:
1994 /// let new_year = "农历新年";
1996 /// assert_eq!(new_year, new_year.to_uppercase());
1998 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
1999 pub fn to_uppercase(&self) -> String {
2000 let mut s = String::with_capacity(self.len());
2001 s.extend(self.chars().flat_map(|c| c.to_uppercase()));
2005 /// Escapes each char in `s` with [`char::escape_debug`].
2007 /// [`char::escape_debug`]: primitive.char.html#method.escape_debug
2008 #[unstable(feature = "str_escape",
2009 reason = "return type may change to be an iterator",
2011 pub fn escape_debug(&self) -> String {
2012 self.chars().flat_map(|c| c.escape_debug()).collect()
2015 /// Escapes each char in `s` with [`char::escape_default`].
2017 /// [`char::escape_default`]: primitive.char.html#method.escape_default
2018 #[unstable(feature = "str_escape",
2019 reason = "return type may change to be an iterator",
2021 pub fn escape_default(&self) -> String {
2022 self.chars().flat_map(|c| c.escape_default()).collect()
2025 /// Escapes each char in `s` with [`char::escape_unicode`].
2027 /// [`char::escape_unicode`]: primitive.char.html#method.escape_unicode
2028 #[unstable(feature = "str_escape",
2029 reason = "return type may change to be an iterator",
2031 pub fn escape_unicode(&self) -> String {
2032 self.chars().flat_map(|c| c.escape_unicode()).collect()
2035 /// Converts a [`Box<str>`] into a [`String`] without copying or allocating.
2037 /// [`String`]: string/struct.String.html
2038 /// [`Box<str>`]: boxed/struct.Box.html
2045 /// let string = String::from("birthday gift");
2046 /// let boxed_str = string.clone().into_boxed_str();
2048 /// assert_eq!(boxed_str.into_string(), string);
2050 #[stable(feature = "box_str", since = "1.4.0")]
2051 pub fn into_string(self: Box<str>) -> String {
2052 let slice = Box::<[u8]>::from(self);
2053 unsafe { String::from_utf8_unchecked(slice.into_vec()) }
2056 /// Create a [`String`] by repeating a string `n` times.
2058 /// [`String`]: string/struct.String.html
2065 /// assert_eq!("abc".repeat(4), String::from("abcabcabcabc"));
2067 #[stable(feature = "repeat_str", since = "1.16.0")]
2068 pub fn repeat(&self, n: usize) -> String {
2069 let mut s = String::with_capacity(self.len() * n);
2070 s.extend((0..n).map(|_| self));
2075 /// Converts a boxed slice of bytes to a boxed string slice without checking
2076 /// that the string contains valid UTF-8.
2083 /// let smile_utf8 = Box::new([226, 152, 186]);
2084 /// let smile = unsafe { std::str::from_boxed_utf8_unchecked(smile_utf8) };
2086 /// assert_eq!("☺", &*smile);
2088 #[stable(feature = "str_box_extras", since = "1.20.0")]
2089 pub unsafe fn from_boxed_utf8_unchecked(v: Box<[u8]>) -> Box<str> {
2090 Box::from_raw(Box::into_raw(v) as *mut str)