]> git.proxmox.com Git - rustc.git/blame - src/libcollections/str.rs
Imported Upstream version 1.4.0+dfsg1
[rustc.git] / src / libcollections / str.rs
CommitLineData
1a4d82fc
JJ
1// Copyright 2012-2014 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.
4//
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.
1a4d82fc 10
c1a9b12d 11//! Unicode string slices
1a4d82fc 12//!
c1a9b12d
SL
13//! *[See also the `str` primitive type](../primitive.str.html).*
14
1a4d82fc 15
85aaf69f 16#![stable(feature = "rust1", since = "1.0.0")]
1a4d82fc 17
62682a34
SL
18// Many of the usings in this module are only used in the test configuration.
19// It's cleaner to just turn off the unused_imports warning than to fix them.
20#![allow(unused_imports)]
21
1a4d82fc 22use core::clone::Clone;
c34b1796 23use core::iter::{Iterator, Extend};
1a4d82fc 24use core::option::Option::{self, Some, None};
85aaf69f 25use core::result::Result;
1a4d82fc 26use core::str as core_str;
9346a6ac
AL
27use core::str::pattern::Pattern;
28use core::str::pattern::{Searcher, ReverseSearcher, DoubleEndedSearcher};
c1a9b12d 29use core::mem;
d9579d0f 30use rustc_unicode::str::{UnicodeStr, Utf16Encoder};
1a4d82fc 31
85aaf69f
SL
32use vec_deque::VecDeque;
33use borrow::{Borrow, ToOwned};
1a4d82fc 34use string::String;
d9579d0f 35use rustc_unicode;
1a4d82fc
JJ
36use vec::Vec;
37use slice::SliceConcatExt;
c1a9b12d 38use boxed::Box;
1a4d82fc 39
9346a6ac
AL
40pub use core::str::{FromStr, Utf8Error};
41pub use core::str::{Lines, LinesAny, CharRange};
42pub use core::str::{Split, RSplit};
43pub use core::str::{SplitN, RSplitN};
44pub use core::str::{SplitTerminator, RSplitTerminator};
45pub use core::str::{Matches, RMatches};
46pub use core::str::{MatchIndices, RMatchIndices};
c34b1796
AL
47pub use core::str::{from_utf8, Chars, CharIndices, Bytes};
48pub use core::str::{from_utf8_unchecked, ParseBoolError};
e9174d1e 49pub use rustc_unicode::str::{SplitWhitespace};
9346a6ac 50pub use core::str::pattern;
1a4d82fc 51
d9579d0f
AL
52impl<S: Borrow<str>> SliceConcatExt<str> for [S] {
53 type Output = String;
54
1a4d82fc 55 fn concat(&self) -> String {
c34b1796 56 if self.is_empty() {
1a4d82fc
JJ
57 return String::new();
58 }
59
60 // `len` calculation may overflow but push_str will check boundaries
bd371182 61 let len = self.iter().map(|s| s.borrow().len()).sum();
1a4d82fc
JJ
62 let mut result = String::with_capacity(len);
63
c34b1796 64 for s in self {
bd371182 65 result.push_str(s.borrow())
1a4d82fc
JJ
66 }
67
68 result
69 }
70
c1a9b12d 71 fn join(&self, sep: &str) -> String {
c34b1796 72 if self.is_empty() {
1a4d82fc
JJ
73 return String::new();
74 }
75
76 // concat is faster
77 if sep.is_empty() {
c34b1796 78 return self.concat();
1a4d82fc
JJ
79 }
80
81 // this is wrong without the guarantee that `self` is non-empty
82 // `len` calculation may overflow but push_str but will check boundaries
c34b1796 83 let len = sep.len() * (self.len() - 1)
bd371182 84 + self.iter().map(|s| s.borrow().len()).sum::<usize>();
1a4d82fc
JJ
85 let mut result = String::with_capacity(len);
86 let mut first = true;
87
c34b1796 88 for s in self {
1a4d82fc
JJ
89 if first {
90 first = false;
91 } else {
92 result.push_str(sep);
93 }
bd371182 94 result.push_str(s.borrow());
1a4d82fc
JJ
95 }
96 result
97 }
1a4d82fc 98
c1a9b12d
SL
99 fn connect(&self, sep: &str) -> String {
100 self.join(sep)
101 }
102}
1a4d82fc 103
1a4d82fc 104/// External iterator for a string's UTF16 codeunits.
c34b1796
AL
105///
106/// For use with the `std::iter` module.
1a4d82fc 107#[derive(Clone)]
e9174d1e 108#[unstable(feature = "str_utf16", issue = "27714")]
1a4d82fc
JJ
109pub struct Utf16Units<'a> {
110 encoder: Utf16Encoder<Chars<'a>>
111}
112
85aaf69f 113#[stable(feature = "rust1", since = "1.0.0")]
1a4d82fc
JJ
114impl<'a> Iterator for Utf16Units<'a> {
115 type Item = u16;
116
117 #[inline]
118 fn next(&mut self) -> Option<u16> { self.encoder.next() }
119
120 #[inline]
85aaf69f 121 fn size_hint(&self) -> (usize, Option<usize>) { self.encoder.size_hint() }
1a4d82fc
JJ
122}
123
1a4d82fc
JJ
124// Return the initial codepoint accumulator for the first byte.
125// The first byte is special, only want bottom 5 bits for width 2, 4 bits
126// for width 3, and 3 bits for width 4
127macro_rules! utf8_first_byte {
128 ($byte:expr, $width:expr) => (($byte & (0x7F >> $width)) as u32)
129}
130
131// return the value of $ch updated with continuation byte $byte
132macro_rules! utf8_acc_cont_byte {
c34b1796 133 ($ch:expr, $byte:expr) => (($ch << 6) | ($byte & 63) as u32)
1a4d82fc
JJ
134}
135
85aaf69f
SL
136#[stable(feature = "rust1", since = "1.0.0")]
137impl Borrow<str> for String {
d9579d0f 138 #[inline]
85aaf69f 139 fn borrow(&self) -> &str { &self[..] }
1a4d82fc
JJ
140}
141
85aaf69f
SL
142#[stable(feature = "rust1", since = "1.0.0")]
143impl ToOwned for str {
144 type Owned = String;
1a4d82fc
JJ
145 fn to_owned(&self) -> String {
146 unsafe {
147 String::from_utf8_unchecked(self.as_bytes().to_owned())
148 }
149 }
150}
151
1a4d82fc 152/// Any string that can be represented as a slice.
c34b1796
AL
153#[lang = "str"]
154#[cfg(not(test))]
85aaf69f 155#[stable(feature = "rust1", since = "1.0.0")]
c34b1796 156impl str {
62682a34 157 /// Returns the length of `self` in bytes.
1a4d82fc
JJ
158 ///
159 /// # Examples
160 ///
c34b1796 161 /// ```
62682a34
SL
162 /// assert_eq!("foo".len(), 3);
163 /// assert_eq!("ƒoo".len(), 4); // fancy f!
1a4d82fc 164 /// ```
85aaf69f 165 #[stable(feature = "rust1", since = "1.0.0")]
1a4d82fc 166 #[inline]
62682a34
SL
167 pub fn len(&self) -> usize {
168 core_str::StrExt::len(self)
1a4d82fc
JJ
169 }
170
62682a34
SL
171 /// Returns true if this slice has a length of zero bytes.
172 ///
173 /// # Examples
174 ///
175 /// ```
176 /// assert!("".is_empty());
177 /// ```
1a4d82fc 178 #[inline]
62682a34
SL
179 #[stable(feature = "rust1", since = "1.0.0")]
180 pub fn is_empty(&self) -> bool {
181 core_str::StrExt::is_empty(self)
1a4d82fc
JJ
182 }
183
62682a34
SL
184 /// Checks that `index`-th byte lies at the start and/or end of a
185 /// UTF-8 code point sequence.
186 ///
187 /// The start and end of the string (when `index == self.len()`) are
188 /// considered to be
189 /// boundaries.
190 ///
c1a9b12d 191 /// Returns `false` if `index` is greater than `self.len()`.
1a4d82fc 192 ///
c34b1796 193 /// # Examples
1a4d82fc 194 ///
c34b1796 195 /// ```
c1a9b12d
SL
196 /// #![feature(str_char)]
197 ///
62682a34
SL
198 /// let s = "Löwe 老虎 Léopard";
199 /// assert!(s.is_char_boundary(0));
200 /// // start of `老`
201 /// assert!(s.is_char_boundary(6));
202 /// assert!(s.is_char_boundary(s.len()));
c34b1796 203 ///
62682a34
SL
204 /// // second byte of `ö`
205 /// assert!(!s.is_char_boundary(2));
206 ///
207 /// // third byte of `老`
208 /// assert!(!s.is_char_boundary(8));
1a4d82fc 209 /// ```
62682a34
SL
210 #[unstable(feature = "str_char",
211 reason = "it is unclear whether this method pulls its weight \
212 with the existence of the char_indices iterator or \
213 this method may want to be replaced with checked \
e9174d1e
SL
214 slicing",
215 issue = "27754")]
c1a9b12d 216 #[inline]
62682a34
SL
217 pub fn is_char_boundary(&self, index: usize) -> bool {
218 core_str::StrExt::is_char_boundary(self, index)
1a4d82fc
JJ
219 }
220
62682a34 221 /// Converts `self` to a byte slice.
1a4d82fc 222 ///
c34b1796 223 /// # Examples
1a4d82fc 224 ///
1a4d82fc 225 /// ```
62682a34 226 /// assert_eq!("bors".as_bytes(), b"bors");
1a4d82fc 227 /// ```
85aaf69f 228 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
229 #[inline(always)]
230 pub fn as_bytes(&self) -> &[u8] {
231 core_str::StrExt::as_bytes(self)
1a4d82fc
JJ
232 }
233
62682a34
SL
234 /// Returns a raw pointer to the `&str`'s buffer.
235 ///
236 /// The caller must ensure that the string outlives this pointer, and
237 /// that it is not
238 /// reallocated (e.g. by pushing to the string).
1a4d82fc 239 ///
c34b1796 240 /// # Examples
1a4d82fc 241 ///
c34b1796 242 /// ```
62682a34
SL
243 /// let s = "Hello";
244 /// let p = s.as_ptr();
1a4d82fc 245 /// ```
85aaf69f 246 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
247 #[inline]
248 pub fn as_ptr(&self) -> *const u8 {
249 core_str::StrExt::as_ptr(self)
1a4d82fc
JJ
250 }
251
62682a34
SL
252 /// Takes a bytewise slice from a string.
253 ///
254 /// Returns the substring from [`begin`..`end`).
255 ///
256 /// # Unsafety
257 ///
c1a9b12d 258 /// Caller must check both UTF-8 sequence boundaries and the boundaries
e9174d1e 259 /// of the entire slice as well.
c34b1796
AL
260 ///
261 /// # Examples
262 ///
263 /// ```
62682a34 264 /// let s = "Löwe 老虎 Léopard";
c34b1796 265 ///
62682a34
SL
266 /// unsafe {
267 /// assert_eq!(s.slice_unchecked(0, 21), "Löwe 老虎 Léopard");
268 /// }
c34b1796 269 /// ```
85aaf69f 270 #[stable(feature = "rust1", since = "1.0.0")]
c1a9b12d 271 #[inline]
62682a34
SL
272 pub unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
273 core_str::StrExt::slice_unchecked(self, begin, end)
1a4d82fc
JJ
274 }
275
c1a9b12d
SL
276 /// Takes a bytewise mutable slice from a string.
277 ///
278 /// Same as `slice_unchecked`, but works with `&mut str` instead of `&str`.
e9174d1e
SL
279 #[unstable(feature = "str_slice_mut", reason = "recently added",
280 issue = "27793")]
c1a9b12d
SL
281 #[inline]
282 pub unsafe fn slice_mut_unchecked(&mut self, begin: usize, end: usize) -> &mut str {
283 core_str::StrExt::slice_mut_unchecked(self, begin, end)
284 }
285
c1a9b12d 286 /// Given a byte position, return the next code point and its index.
9346a6ac 287 ///
c1a9b12d 288 /// This can be used to iterate over the Unicode code points of a string.
9346a6ac 289 ///
62682a34 290 /// # Panics
1a4d82fc 291 ///
62682a34 292 /// If `i` is greater than or equal to the length of the string.
c1a9b12d 293 /// If `i` is not the index of the beginning of a valid UTF-8 sequence.
c34b1796 294 ///
62682a34 295 /// # Examples
1a4d82fc 296 ///
c1a9b12d 297 /// This example manually iterates through the code points of a string;
62682a34
SL
298 /// this should normally be
299 /// done by `.chars()` or `.char_indices()`.
c34b1796
AL
300 ///
301 /// ```
c1a9b12d
SL
302 /// #![feature(str_char, core)]
303 ///
62682a34 304 /// use std::str::CharRange;
9346a6ac 305 ///
c1a9b12d 306 /// let s = "中华Việt Nam";
62682a34
SL
307 /// let mut i = 0;
308 /// while i < s.len() {
309 /// let CharRange {ch, next} = s.char_range_at(i);
310 /// println!("{}: {}", i, ch);
311 /// i = next;
312 /// }
313 /// ```
9346a6ac 314 ///
62682a34 315 /// This outputs:
9346a6ac 316 ///
62682a34
SL
317 /// ```text
318 /// 0: 中
319 /// 3: 华
320 /// 6: V
321 /// 7: i
c1a9b12d 322 /// 8: e
e9174d1e
SL
323 /// 9:
324 /// 11:
c1a9b12d
SL
325 /// 13: t
326 /// 14:
327 /// 15: N
328 /// 16: a
329 /// 17: m
1a4d82fc 330 /// ```
62682a34
SL
331 #[unstable(feature = "str_char",
332 reason = "often replaced by char_indices, this method may \
333 be removed in favor of just char_at() or eventually \
e9174d1e
SL
334 removed altogether",
335 issue = "27754")]
c1a9b12d 336 #[inline]
62682a34
SL
337 pub fn char_range_at(&self, start: usize) -> CharRange {
338 core_str::StrExt::char_range_at(self, start)
1a4d82fc
JJ
339 }
340
62682a34 341 /// Given a byte position, return the previous `char` and its position.
9346a6ac 342 ///
c1a9b12d
SL
343 /// This function can be used to iterate over a Unicode code points in reverse.
344 ///
345 /// Note that Unicode has many features, such as combining marks, ligatures,
346 /// and direction marks, that need to be taken into account to correctly reverse a string.
9346a6ac 347 ///
62682a34 348 /// Returns 0 for next index if called on start index 0.
9346a6ac 349 ///
62682a34 350 /// # Panics
9346a6ac 351 ///
62682a34 352 /// If `i` is greater than the length of the string.
c1a9b12d 353 /// If `i` is not an index following a valid UTF-8 sequence.
1a4d82fc 354 ///
c34b1796
AL
355 /// # Examples
356 ///
c1a9b12d 357 /// This example manually iterates through the code points of a string;
62682a34
SL
358 /// this should normally be
359 /// done by `.chars().rev()` or `.char_indices()`.
c34b1796
AL
360 ///
361 /// ```
c1a9b12d
SL
362 /// #![feature(str_char, core)]
363 ///
62682a34 364 /// use std::str::CharRange;
1a4d82fc 365 ///
c1a9b12d 366 /// let s = "中华Việt Nam";
62682a34
SL
367 /// let mut i = s.len();
368 /// while i > 0 {
369 /// let CharRange {ch, next} = s.char_range_at_reverse(i);
370 /// println!("{}: {}", i, ch);
371 /// i = next;
372 /// }
c34b1796 373 /// ```
1a4d82fc 374 ///
62682a34 375 /// This outputs:
1a4d82fc 376 ///
62682a34 377 /// ```text
c1a9b12d
SL
378 /// 18: m
379 /// 17: a
380 /// 16: N
381 /// 15:
382 /// 14: t
e9174d1e
SL
383 /// 13:
384 /// 11:
c1a9b12d 385 /// 9: e
62682a34
SL
386 /// 8: i
387 /// 7: V
388 /// 6: 华
389 /// 3: 中
1a4d82fc 390 /// ```
62682a34
SL
391 #[unstable(feature = "str_char",
392 reason = "often replaced by char_indices, this method may \
393 be removed in favor of just char_at_reverse() or \
e9174d1e
SL
394 eventually removed altogether",
395 issue = "27754")]
c1a9b12d 396 #[inline]
62682a34
SL
397 pub fn char_range_at_reverse(&self, start: usize) -> CharRange {
398 core_str::StrExt::char_range_at_reverse(self, start)
1a4d82fc
JJ
399 }
400
62682a34 401 /// Given a byte position, return the `char` at that position.
9346a6ac 402 ///
62682a34 403 /// # Panics
9346a6ac 404 ///
62682a34 405 /// If `i` is greater than or equal to the length of the string.
c1a9b12d 406 /// If `i` is not the index of the beginning of a valid UTF-8 sequence.
1a4d82fc 407 ///
c34b1796 408 /// # Examples
1a4d82fc 409 ///
c34b1796 410 /// ```
c1a9b12d
SL
411 /// #![feature(str_char)]
412 ///
62682a34
SL
413 /// let s = "abπc";
414 /// assert_eq!(s.char_at(1), 'b');
415 /// assert_eq!(s.char_at(2), 'π');
c1a9b12d 416 /// assert_eq!(s.char_at(4), 'c');
c34b1796 417 /// ```
62682a34
SL
418 #[unstable(feature = "str_char",
419 reason = "frequently replaced by the chars() iterator, this \
420 method may be removed or possibly renamed in the \
421 future; it is normally replaced by chars/char_indices \
422 iterators or by getting the first char from a \
e9174d1e
SL
423 subslice",
424 issue = "27754")]
c1a9b12d 425 #[inline]
62682a34
SL
426 pub fn char_at(&self, i: usize) -> char {
427 core_str::StrExt::char_at(self, i)
9346a6ac
AL
428 }
429
62682a34
SL
430 /// Given a byte position, return the `char` at that position, counting
431 /// from the end.
9346a6ac 432 ///
62682a34 433 /// # Panics
9346a6ac 434 ///
62682a34 435 /// If `i` is greater than the length of the string.
c1a9b12d 436 /// If `i` is not an index following a valid UTF-8 sequence.
9346a6ac
AL
437 ///
438 /// # Examples
439 ///
9346a6ac 440 /// ```
c1a9b12d
SL
441 /// #![feature(str_char)]
442 ///
62682a34
SL
443 /// let s = "abπc";
444 /// assert_eq!(s.char_at_reverse(1), 'a');
445 /// assert_eq!(s.char_at_reverse(2), 'b');
c1a9b12d 446 /// assert_eq!(s.char_at_reverse(3), 'π');
9346a6ac 447 /// ```
62682a34
SL
448 #[unstable(feature = "str_char",
449 reason = "see char_at for more details, but reverse semantics \
450 are also somewhat unclear, especially with which \
e9174d1e
SL
451 cases generate panics",
452 issue = "27754")]
c1a9b12d 453 #[inline]
62682a34
SL
454 pub fn char_at_reverse(&self, i: usize) -> char {
455 core_str::StrExt::char_at_reverse(self, i)
c34b1796
AL
456 }
457
c1a9b12d
SL
458 /// Retrieves the first code point from a `&str` and returns it.
459 ///
460 /// Note that a single Unicode character (grapheme cluster)
461 /// can be composed of multiple `char`s.
9346a6ac 462 ///
62682a34 463 /// This does not allocate a new string; instead, it returns a slice that
c1a9b12d 464 /// points one code point beyond the code point that was shifted.
9346a6ac 465 ///
c1a9b12d 466 /// `None` is returned if the slice is empty.
9346a6ac 467 ///
c34b1796
AL
468 /// # Examples
469 ///
c34b1796 470 /// ```
c1a9b12d
SL
471 /// #![feature(str_char)]
472 ///
473 /// let s = "Łódź"; // \u{141}o\u{301}dz\u{301}
62682a34 474 /// let (c, s1) = s.slice_shift_char().unwrap();
9346a6ac 475 ///
c1a9b12d
SL
476 /// assert_eq!(c, 'Ł');
477 /// assert_eq!(s1, "ódź");
c34b1796 478 ///
62682a34 479 /// let (c, s2) = s1.slice_shift_char().unwrap();
1a4d82fc 480 ///
c1a9b12d
SL
481 /// assert_eq!(c, 'o');
482 /// assert_eq!(s2, "\u{301}dz\u{301}");
c34b1796 483 /// ```
62682a34
SL
484 #[unstable(feature = "str_char",
485 reason = "awaiting conventions about shifting and slices and \
486 may not be warranted with the existence of the chars \
e9174d1e
SL
487 and/or char_indices iterators",
488 issue = "27754")]
c1a9b12d 489 #[inline]
62682a34
SL
490 pub fn slice_shift_char(&self) -> Option<(char, &str)> {
491 core_str::StrExt::slice_shift_char(self)
1a4d82fc
JJ
492 }
493
62682a34 494 /// Divide one string slice into two at an index.
9346a6ac 495 ///
62682a34 496 /// The index `mid` is a byte offset from the start of the string
c1a9b12d 497 /// that must be on a `char` boundary.
9346a6ac 498 ///
62682a34 499 /// Return slices `&self[..mid]` and `&self[mid..]`.
9346a6ac 500 ///
62682a34 501 /// # Panics
9346a6ac 502 ///
c1a9b12d
SL
503 /// Panics if `mid` is beyond the last code point of the string,
504 /// or if it is not on a `char` boundary.
9346a6ac
AL
505 ///
506 /// # Examples
9346a6ac 507 /// ```
c1a9b12d
SL
508 /// #![feature(str_split_at)]
509 ///
62682a34
SL
510 /// let s = "Löwe 老虎 Léopard";
511 /// let first_space = s.find(' ').unwrap_or(s.len());
512 /// let (a, b) = s.split_at(first_space);
9346a6ac 513 ///
62682a34
SL
514 /// assert_eq!(a, "Löwe");
515 /// assert_eq!(b, " 老虎 Léopard");
9346a6ac 516 /// ```
62682a34 517 #[inline]
e9174d1e 518 #[stable(feature = "str_split_at", since = "1.4.0")]
62682a34
SL
519 pub fn split_at(&self, mid: usize) -> (&str, &str) {
520 core_str::StrExt::split_at(self, mid)
9346a6ac
AL
521 }
522
c1a9b12d
SL
523 /// Divide one mutable string slice into two at an index.
524 #[inline]
e9174d1e 525 #[stable(feature = "str_split_at", since = "1.4.0")]
c1a9b12d
SL
526 pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {
527 core_str::StrExt::split_at_mut(self, mid)
528 }
529
530 /// An iterator over the code points of `self`.
531 ///
532 /// In Unicode relationship between code points and characters is complex.
533 /// A single character may be composed of multiple code points
534 /// (e.g. diacritical marks added to a letter), and a single code point
535 /// (e.g. Hangul syllable) may contain multiple characters.
536 ///
537 /// For iteration over human-readable characters a grapheme cluster iterator
538 /// may be more appropriate. See the [unicode-segmentation crate][1].
539 ///
540 /// [1]: https://crates.io/crates/unicode-segmentation
9346a6ac
AL
541 ///
542 /// # Examples
543 ///
544 /// ```
c1a9b12d 545 /// let v: Vec<char> = "ASCII żółć 🇨🇭 한".chars().collect();
9346a6ac 546 ///
c1a9b12d
SL
547 /// assert_eq!(v, ['A', 'S', 'C', 'I', 'I', ' ',
548 /// 'z', '\u{307}', 'o', '\u{301}', 'ł', 'c', '\u{301}', ' ',
549 /// '\u{1f1e8}', '\u{1f1ed}', ' ', '한']);
9346a6ac 550 /// ```
62682a34 551 #[stable(feature = "rust1", since = "1.0.0")]
c1a9b12d 552 #[inline]
62682a34
SL
553 pub fn chars(&self) -> Chars {
554 core_str::StrExt::chars(self)
9346a6ac
AL
555 }
556
c1a9b12d 557 /// An iterator over the `char`s of `self` and their byte offsets.
9346a6ac 558 ///
c34b1796 559 /// # Examples
1a4d82fc 560 ///
c34b1796 561 /// ```
c1a9b12d
SL
562 /// let v: Vec<(usize, char)> = "A🇨🇭".char_indices().collect();
563 /// let b = vec![(0, 'A'), (1, '\u{1f1e8}'), (5, '\u{1f1ed}')];
1a4d82fc 564 ///
62682a34 565 /// assert_eq!(v, b);
1a4d82fc 566 /// ```
62682a34 567 #[stable(feature = "rust1", since = "1.0.0")]
c1a9b12d 568 #[inline]
62682a34
SL
569 pub fn char_indices(&self) -> CharIndices {
570 core_str::StrExt::char_indices(self)
1a4d82fc
JJ
571 }
572
62682a34 573 /// An iterator over the bytes of `self`.
9346a6ac 574 ///
62682a34 575 /// # Examples
9346a6ac 576 ///
62682a34
SL
577 /// ```
578 /// let v: Vec<u8> = "bors".bytes().collect();
9346a6ac 579 ///
62682a34
SL
580 /// assert_eq!(v, b"bors".to_vec());
581 /// ```
582 #[stable(feature = "rust1", since = "1.0.0")]
c1a9b12d 583 #[inline]
62682a34
SL
584 pub fn bytes(&self) -> Bytes {
585 core_str::StrExt::bytes(self)
586 }
587
588 /// An iterator over the non-empty substrings of `self` which contain no whitespace,
589 /// and which are separated by any amount of whitespace.
9346a6ac
AL
590 ///
591 /// # Examples
592 ///
593 /// ```
c1a9b12d 594 /// let some_words = " Mary had\ta\u{2009}little \n\t lamb";
62682a34 595 /// let v: Vec<&str> = some_words.split_whitespace().collect();
9346a6ac 596 ///
62682a34
SL
597 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
598 /// ```
599 #[stable(feature = "split_whitespace", since = "1.1.0")]
c1a9b12d 600 #[inline]
62682a34
SL
601 pub fn split_whitespace(&self) -> SplitWhitespace {
602 UnicodeStr::split_whitespace(self)
603 }
604
e9174d1e 605 /// An iterator over the lines of a string, separated by `\n` or `\r\n`.
1a4d82fc 606 ///
e9174d1e 607 /// This does not include the empty string after a trailing newline or CRLF.
1a4d82fc 608 ///
c34b1796 609 /// # Examples
1a4d82fc 610 ///
1a4d82fc 611 /// ```
e9174d1e 612 /// let four_lines = "foo\nbar\n\r\nbaz";
c34b1796 613 /// let v: Vec<&str> = four_lines.lines().collect();
1a4d82fc 614 ///
c34b1796
AL
615 /// assert_eq!(v, ["foo", "bar", "", "baz"]);
616 /// ```
1a4d82fc 617 ///
c34b1796
AL
618 /// Leaving off the trailing character:
619 ///
620 /// ```
e9174d1e 621 /// let four_lines = "foo\r\nbar\n\nbaz\n";
1a4d82fc 622 /// let v: Vec<&str> = four_lines.lines().collect();
c34b1796
AL
623 ///
624 /// assert_eq!(v, ["foo", "bar", "", "baz"]);
1a4d82fc 625 /// ```
85aaf69f 626 #[stable(feature = "rust1", since = "1.0.0")]
c1a9b12d 627 #[inline]
c34b1796 628 pub fn lines(&self) -> Lines {
62682a34 629 core_str::StrExt::lines(self)
1a4d82fc
JJ
630 }
631
9346a6ac
AL
632 /// An iterator over the lines of a string, separated by either
633 /// `\n` or `\r\n`.
1a4d82fc 634 ///
c34b1796 635 /// As with `.lines()`, this does not include an empty trailing line.
1a4d82fc 636 ///
c34b1796
AL
637 /// # Examples
638 ///
639 /// ```
640 /// let four_lines = "foo\r\nbar\n\r\nbaz";
641 /// let v: Vec<&str> = four_lines.lines_any().collect();
642 ///
643 /// assert_eq!(v, ["foo", "bar", "", "baz"]);
644 /// ```
645 ///
646 /// Leaving off the trailing character:
647 ///
648 /// ```
1a4d82fc
JJ
649 /// let four_lines = "foo\r\nbar\n\r\nbaz\n";
650 /// let v: Vec<&str> = four_lines.lines_any().collect();
c34b1796
AL
651 ///
652 /// assert_eq!(v, ["foo", "bar", "", "baz"]);
1a4d82fc 653 /// ```
85aaf69f 654 #[stable(feature = "rust1", since = "1.0.0")]
e9174d1e 655 #[deprecated(since = "1.4.0", reason = "use lines() instead now")]
c1a9b12d 656 #[inline]
e9174d1e 657 #[allow(deprecated)]
c34b1796 658 pub fn lines_any(&self) -> LinesAny {
62682a34 659 core_str::StrExt::lines_any(self)
1a4d82fc 660 }
62682a34 661
62682a34
SL
662 /// Returns an iterator of `u16` over the string encoded as UTF-16.
663 #[unstable(feature = "str_utf16",
e9174d1e
SL
664 reason = "this functionality may only be provided by libunicode",
665 issue = "27714")]
62682a34
SL
666 pub fn utf16_units(&self) -> Utf16Units {
667 Utf16Units { encoder: Utf16Encoder::new(self[..].chars()) }
668 }
669
670 /// Returns `true` if `self` contains another `&str`.
c34b1796
AL
671 ///
672 /// # Examples
673 ///
674 /// ```
62682a34 675 /// assert!("bananas".contains("nana"));
c34b1796 676 ///
62682a34 677 /// assert!(!"bananas".contains("foobar"));
c34b1796 678 /// ```
85aaf69f 679 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
680 pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
681 core_str::StrExt::contains(self, pat)
1a4d82fc
JJ
682 }
683
c34b1796 684 /// Returns `true` if the given `&str` is a prefix of the string.
1a4d82fc 685 ///
c34b1796 686 /// # Examples
1a4d82fc 687 ///
c34b1796 688 /// ```
1a4d82fc
JJ
689 /// assert!("banana".starts_with("ba"));
690 /// ```
85aaf69f 691 #[stable(feature = "rust1", since = "1.0.0")]
c34b1796 692 pub fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
62682a34 693 core_str::StrExt::starts_with(self, pat)
1a4d82fc
JJ
694 }
695
c34b1796 696 /// Returns true if the given `&str` is a suffix of the string.
1a4d82fc 697 ///
c34b1796 698 /// # Examples
1a4d82fc
JJ
699 ///
700 /// ```rust
701 /// assert!("banana".ends_with("nana"));
702 /// ```
85aaf69f 703 #[stable(feature = "rust1", since = "1.0.0")]
c34b1796
AL
704 pub fn ends_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool
705 where P::Searcher: ReverseSearcher<'a>
706 {
62682a34 707 core_str::StrExt::ends_with(self, pat)
1a4d82fc
JJ
708 }
709
62682a34
SL
710 /// Returns the byte index of the first character of `self` that matches
711 /// the pattern, if it
712 /// exists.
713 ///
714 /// Returns `None` if it doesn't exist.
715 ///
716 /// The pattern can be a simple `&str`, `char`, or a closure that
717 /// determines the
718 /// split.
719 ///
720 /// # Examples
721 ///
722 /// Simple patterns:
723 ///
724 /// ```
725 /// let s = "Löwe 老虎 Léopard";
726 ///
727 /// assert_eq!(s.find('L'), Some(0));
728 /// assert_eq!(s.find('é'), Some(14));
729 /// assert_eq!(s.find("Léopard"), Some(13));
730 ///
731 /// ```
732 ///
733 /// More complex patterns with closures:
734 ///
735 /// ```
736 /// let s = "Löwe 老虎 Léopard";
737 ///
738 /// assert_eq!(s.find(char::is_whitespace), Some(5));
739 /// assert_eq!(s.find(char::is_lowercase), Some(1));
740 /// ```
741 ///
742 /// Not finding the pattern:
743 ///
744 /// ```
745 /// let s = "Löwe 老虎 Léopard";
746 /// let x: &[_] = &['1', '2'];
747 ///
748 /// assert_eq!(s.find(x), None);
749 /// ```
750 #[stable(feature = "rust1", since = "1.0.0")]
751 pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
752 core_str::StrExt::find(self, pat)
753 }
754
755 /// Returns the byte index of the last character of `self` that
756 /// matches the pattern, if it
757 /// exists.
758 ///
759 /// Returns `None` if it doesn't exist.
1a4d82fc 760 ///
62682a34
SL
761 /// The pattern can be a simple `&str`, `char`,
762 /// or a closure that determines the split.
1a4d82fc 763 ///
c34b1796 764 /// # Examples
1a4d82fc 765 ///
9346a6ac 766 /// Simple patterns:
1a4d82fc 767 ///
c34b1796 768 /// ```
62682a34 769 /// let s = "Löwe 老虎 Léopard";
c34b1796 770 ///
62682a34
SL
771 /// assert_eq!(s.rfind('L'), Some(13));
772 /// assert_eq!(s.rfind('é'), Some(14));
c34b1796
AL
773 /// ```
774 ///
9346a6ac 775 /// More complex patterns with closures:
c34b1796
AL
776 ///
777 /// ```
62682a34
SL
778 /// let s = "Löwe 老虎 Léopard";
779 ///
780 /// assert_eq!(s.rfind(char::is_whitespace), Some(12));
781 /// assert_eq!(s.rfind(char::is_lowercase), Some(20));
782 /// ```
783 ///
784 /// Not finding the pattern:
785 ///
786 /// ```
787 /// let s = "Löwe 老虎 Léopard";
788 /// let x: &[_] = &['1', '2'];
789 ///
790 /// assert_eq!(s.rfind(x), None);
1a4d82fc 791 /// ```
85aaf69f 792 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
793 pub fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
794 where P::Searcher: ReverseSearcher<'a>
c34b1796 795 {
62682a34 796 core_str::StrExt::rfind(self, pat)
1a4d82fc
JJ
797 }
798
62682a34
SL
799 /// An iterator over substrings of `self`, separated by characters
800 /// matched by a pattern.
1a4d82fc 801 ///
9346a6ac 802 /// The pattern can be a simple `&str`, `char`, or a closure that
62682a34
SL
803 /// determines the split. Additional libraries might provide more complex
804 /// patterns like regular expressions.
805 ///
806 /// # Iterator behavior
807 ///
808 /// The returned iterator will be double ended if the pattern allows a
809 /// reverse search and forward/reverse search yields the same elements.
810 /// This is true for, eg, `char` but not
811 /// for `&str`.
812 ///
813 /// If the pattern allows a reverse search but its results might differ
814 /// from a forward search, `rsplit()` can be used.
1a4d82fc 815 ///
c34b1796 816 /// # Examples
1a4d82fc 817 ///
9346a6ac 818 /// Simple patterns:
1a4d82fc 819 ///
c34b1796 820 /// ```
62682a34
SL
821 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
822 /// assert_eq!(v, ["Mary", "had", "a", "little", "lamb"]);
c34b1796 823 ///
62682a34
SL
824 /// let v: Vec<&str> = "".split('X').collect();
825 /// assert_eq!(v, [""]);
c34b1796 826 ///
62682a34
SL
827 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
828 /// assert_eq!(v, ["lion", "", "tiger", "leopard"]);
c34b1796 829 ///
62682a34
SL
830 /// let v: Vec<&str> = "lion::tiger::leopard".split("::").collect();
831 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
1a4d82fc 832 ///
62682a34
SL
833 /// let v: Vec<&str> = "abc1def2ghi".split(char::is_numeric).collect();
834 /// assert_eq!(v, ["abc", "def", "ghi"]);
1a4d82fc 835 ///
62682a34
SL
836 /// let v: Vec<&str> = "lionXtigerXleopard".split(char::is_uppercase).collect();
837 /// assert_eq!(v, ["lion", "tiger", "leopard"]);
838 /// ```
1a4d82fc 839 ///
62682a34 840 /// A more complex pattern, using a closure:
1a4d82fc 841 ///
c34b1796 842 /// ```
62682a34
SL
843 /// let v: Vec<&str> = "abc1defXghi".split(|c| c == '1' || c == 'X').collect();
844 /// assert_eq!(v, ["abc", "def", "ghi"]);
c34b1796
AL
845 /// ```
846 ///
62682a34
SL
847 /// If a string contains multiple contiguous separators, you will end up
848 /// with empty strings in the output:
c34b1796
AL
849 ///
850 /// ```
62682a34
SL
851 /// let x = "||||a||b|c".to_string();
852 /// let d: Vec<_> = x.split('|').collect();
c34b1796 853 ///
62682a34
SL
854 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
855 /// ```
1a4d82fc 856 ///
62682a34
SL
857 /// This can lead to possibly surprising behavior when whitespace is used
858 /// as the separator. This code is correct:
1a4d82fc 859 ///
62682a34
SL
860 /// ```
861 /// let x = " a b c".to_string();
862 /// let d: Vec<_> = x.split(' ').collect();
1a4d82fc 863 ///
62682a34 864 /// assert_eq!(d, &["", "", "", "", "a", "", "b", "c"]);
c34b1796 865 /// ```
1a4d82fc 866 ///
62682a34 867 /// It does _not_ give you:
1a4d82fc 868 ///
62682a34
SL
869 /// ```rust,ignore
870 /// assert_eq!(d, &["a", "b", "c"]);
1a4d82fc 871 /// ```
62682a34
SL
872 #[stable(feature = "rust1", since = "1.0.0")]
873 pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
874 core_str::StrExt::split(self, pat)
1a4d82fc
JJ
875 }
876
62682a34
SL
877 /// An iterator over substrings of `self`, separated by characters
878 /// matched by a pattern and yielded in reverse order.
1a4d82fc 879 ///
62682a34
SL
880 /// The pattern can be a simple `&str`, `char`, or a closure that
881 /// determines the split.
882 /// Additional libraries might provide more complex patterns like
883 /// regular expressions.
1a4d82fc 884 ///
62682a34 885 /// # Iterator behavior
1a4d82fc 886 ///
62682a34
SL
887 /// The returned iterator requires that the pattern supports a
888 /// reverse search,
889 /// and it will be double ended if a forward/reverse search yields
890 /// the same elements.
891 ///
892 /// For iterating from the front, `split()` can be used.
1a4d82fc 893 ///
c34b1796
AL
894 /// # Examples
895 ///
62682a34 896 /// Simple patterns:
c34b1796 897 ///
62682a34
SL
898 /// ```rust
899 /// let v: Vec<&str> = "Mary had a little lamb".rsplit(' ').collect();
900 /// assert_eq!(v, ["lamb", "little", "a", "had", "Mary"]);
1a4d82fc 901 ///
62682a34
SL
902 /// let v: Vec<&str> = "".rsplit('X').collect();
903 /// assert_eq!(v, [""]);
904 ///
905 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplit('X').collect();
906 /// assert_eq!(v, ["leopard", "tiger", "", "lion"]);
907 ///
908 /// let v: Vec<&str> = "lion::tiger::leopard".rsplit("::").collect();
909 /// assert_eq!(v, ["leopard", "tiger", "lion"]);
1a4d82fc
JJ
910 /// ```
911 ///
62682a34 912 /// A more complex pattern, using a closure:
1a4d82fc 913 ///
1a4d82fc 914 /// ```
62682a34
SL
915 /// let v: Vec<&str> = "abc1defXghi".rsplit(|c| c == '1' || c == 'X').collect();
916 /// assert_eq!(v, ["ghi", "def", "abc"]);
917 /// ```
918 #[stable(feature = "rust1", since = "1.0.0")]
919 pub fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
920 where P::Searcher: ReverseSearcher<'a>
921 {
922 core_str::StrExt::rsplit(self, pat)
1a4d82fc
JJ
923 }
924
62682a34
SL
925 /// An iterator over substrings of `self`, separated by characters
926 /// matched by a pattern.
1a4d82fc 927 ///
62682a34
SL
928 /// The pattern can be a simple `&str`, `char`, or a closure that
929 /// determines the split.
930 /// Additional libraries might provide more complex patterns
931 /// like regular expressions.
1a4d82fc 932 ///
62682a34
SL
933 /// Equivalent to `split`, except that the trailing substring
934 /// is skipped if empty.
1a4d82fc 935 ///
62682a34
SL
936 /// This method can be used for string data that is _terminated_,
937 /// rather than _separated_ by a pattern.
1a4d82fc 938 ///
62682a34 939 /// # Iterator behavior
1a4d82fc 940 ///
62682a34
SL
941 /// The returned iterator will be double ended if the pattern allows a
942 /// reverse search
943 /// and forward/reverse search yields the same elements. This is true
944 /// for, eg, `char` but not for `&str`.
c34b1796 945 ///
62682a34
SL
946 /// If the pattern allows a reverse search but its results might differ
947 /// from a forward search, `rsplit_terminator()` can be used.
1a4d82fc 948 ///
62682a34 949 /// # Examples
1a4d82fc 950 ///
c34b1796 951 /// ```
62682a34
SL
952 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
953 /// assert_eq!(v, ["A", "B"]);
1a4d82fc 954 ///
62682a34
SL
955 /// let v: Vec<&str> = "A..B..".split_terminator(".").collect();
956 /// assert_eq!(v, ["A", "", "B", ""]);
c34b1796 957 /// ```
62682a34
SL
958 #[stable(feature = "rust1", since = "1.0.0")]
959 pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
960 core_str::StrExt::split_terminator(self, pat)
1a4d82fc
JJ
961 }
962
62682a34
SL
963 /// An iterator over substrings of `self`, separated by characters
964 /// matched by a pattern and yielded in reverse order.
c34b1796 965 ///
62682a34
SL
966 /// The pattern can be a simple `&str`, `char`, or a closure that
967 /// determines the split.
968 /// Additional libraries might provide more complex patterns like
969 /// regular expressions.
970 ///
971 /// Equivalent to `split`, except that the trailing substring is
972 /// skipped if empty.
973 ///
974 /// This method can be used for string data that is _terminated_,
975 /// rather than _separated_ by a pattern.
976 ///
977 /// # Iterator behavior
978 ///
979 /// The returned iterator requires that the pattern supports a
980 /// reverse search, and it will be double ended if a forward/reverse
981 /// search yields the same elements.
c34b1796 982 ///
62682a34 983 /// For iterating from the front, `split_terminator()` can be used.
c34b1796
AL
984 ///
985 /// # Examples
986 ///
987 /// ```
62682a34
SL
988 /// let v: Vec<&str> = "A.B.".rsplit_terminator('.').collect();
989 /// assert_eq!(v, ["B", "A"]);
990 ///
991 /// let v: Vec<&str> = "A..B..".rsplit_terminator(".").collect();
992 /// assert_eq!(v, ["", "B", "", "A"]);
c34b1796 993 /// ```
62682a34
SL
994 #[stable(feature = "rust1", since = "1.0.0")]
995 pub fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
996 where P::Searcher: ReverseSearcher<'a>
997 {
998 core_str::StrExt::rsplit_terminator(self, pat)
c34b1796
AL
999 }
1000
62682a34
SL
1001 /// An iterator over substrings of `self`, separated by a pattern,
1002 /// restricted to returning
1003 /// at most `count` items.
1a4d82fc 1004 ///
62682a34
SL
1005 /// The last element returned, if any, will contain the remainder of the
1006 /// string.
1007 /// The pattern can be a simple `&str`, `char`, or a closure that
1008 /// determines the split.
1009 /// Additional libraries might provide more complex patterns like
1010 /// regular expressions.
1a4d82fc 1011 ///
62682a34
SL
1012 /// # Iterator behavior
1013 ///
1014 /// The returned iterator will not be double ended, because it is
1015 /// not efficient to support.
1016 ///
1017 /// If the pattern allows a reverse search, `rsplitn()` can be used.
c34b1796
AL
1018 ///
1019 /// # Examples
1020 ///
62682a34
SL
1021 /// Simple patterns:
1022 ///
c34b1796 1023 /// ```
62682a34
SL
1024 /// let v: Vec<&str> = "Mary had a little lambda".splitn(3, ' ').collect();
1025 /// assert_eq!(v, ["Mary", "had", "a little lambda"]);
1026 ///
1027 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(3, "X").collect();
1028 /// assert_eq!(v, ["lion", "", "tigerXleopard"]);
1029 ///
1030 /// let v: Vec<&str> = "abcXdef".splitn(1, 'X').collect();
1031 /// assert_eq!(v, ["abcXdef"]);
1032 ///
1033 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1034 /// assert_eq!(v, [""]);
c34b1796 1035 /// ```
1a4d82fc 1036 ///
62682a34 1037 /// A more complex pattern, using a closure:
1a4d82fc 1038 ///
c34b1796 1039 /// ```
62682a34
SL
1040 /// let v: Vec<&str> = "abc1defXghi".splitn(2, |c| c == '1' || c == 'X').collect();
1041 /// assert_eq!(v, ["abc", "defXghi"]);
1a4d82fc 1042 /// ```
85aaf69f 1043 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1044 pub fn splitn<'a, P: Pattern<'a>>(&'a self, count: usize, pat: P) -> SplitN<'a, P> {
1045 core_str::StrExt::splitn(self, count, pat)
1a4d82fc
JJ
1046 }
1047
62682a34
SL
1048 /// An iterator over substrings of `self`, separated by a pattern,
1049 /// starting from the end of the string, restricted to returning
1050 /// at most `count` items.
1a4d82fc 1051 ///
62682a34
SL
1052 /// The last element returned, if any, will contain the remainder of the
1053 /// string.
1a4d82fc 1054 ///
9346a6ac 1055 /// The pattern can be a simple `&str`, `char`, or a closure that
62682a34
SL
1056 /// determines the split.
1057 /// Additional libraries might provide more complex patterns like
1058 /// regular expressions.
1059 ///
1060 /// # Iterator behavior
1061 ///
1062 /// The returned iterator will not be double ended, because it is not
1063 /// efficient to support.
1064 ///
1065 /// `splitn()` can be used for splitting from the front.
1a4d82fc 1066 ///
c34b1796 1067 /// # Examples
1a4d82fc 1068 ///
9346a6ac 1069 /// Simple patterns:
c34b1796
AL
1070 ///
1071 /// ```
62682a34
SL
1072 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(3, ' ').collect();
1073 /// assert_eq!(v, ["lamb", "little", "Mary had a"]);
1a4d82fc 1074 ///
62682a34
SL
1075 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(3, 'X').collect();
1076 /// assert_eq!(v, ["leopard", "tiger", "lionX"]);
1a4d82fc 1077 ///
62682a34
SL
1078 /// let v: Vec<&str> = "lion::tiger::leopard".rsplitn(2, "::").collect();
1079 /// assert_eq!(v, ["leopard", "lion::tiger"]);
c34b1796
AL
1080 /// ```
1081 ///
62682a34 1082 /// A more complex pattern, using a closure:
c34b1796
AL
1083 ///
1084 /// ```
62682a34
SL
1085 /// let v: Vec<&str> = "abc1defXghi".rsplitn(2, |c| c == '1' || c == 'X').collect();
1086 /// assert_eq!(v, ["ghi", "abc1def"]);
c34b1796 1087 /// ```
62682a34
SL
1088 #[stable(feature = "rust1", since = "1.0.0")]
1089 pub fn rsplitn<'a, P: Pattern<'a>>(&'a self, count: usize, pat: P) -> RSplitN<'a, P>
1090 where P::Searcher: ReverseSearcher<'a>
1091 {
1092 core_str::StrExt::rsplitn(self, count, pat)
1093 }
1094
1095 /// An iterator over the matches of a pattern within `self`.
1a4d82fc 1096 ///
62682a34
SL
1097 /// The pattern can be a simple `&str`, `char`, or a closure that
1098 /// determines the split.
1099 /// Additional libraries might provide more complex patterns like
1100 /// regular expressions.
1101 ///
1102 /// # Iterator behavior
1103 ///
1104 /// The returned iterator will be double ended if the pattern allows
1105 /// a reverse search
1106 /// and forward/reverse search yields the same elements. This is true
1107 /// for, eg, `char` but not
1108 /// for `&str`.
1109 ///
1110 /// If the pattern allows a reverse search but its results might differ
1111 /// from a forward search, `rmatches()` can be used.
1112 ///
1113 /// # Examples
c34b1796
AL
1114 ///
1115 /// ```
62682a34
SL
1116 /// let v: Vec<&str> = "abcXXXabcYYYabc".matches("abc").collect();
1117 /// assert_eq!(v, ["abc", "abc", "abc"]);
c34b1796 1118 ///
62682a34
SL
1119 /// let v: Vec<&str> = "1abc2abc3".matches(char::is_numeric).collect();
1120 /// assert_eq!(v, ["1", "2", "3"]);
1a4d82fc 1121 /// ```
62682a34
SL
1122 #[stable(feature = "str_matches", since = "1.2.0")]
1123 pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1124 core_str::StrExt::matches(self, pat)
1a4d82fc
JJ
1125 }
1126
62682a34
SL
1127 /// An iterator over the matches of a pattern within `self`, yielded in
1128 /// reverse order.
1a4d82fc 1129 ///
62682a34
SL
1130 /// The pattern can be a simple `&str`, `char`, or a closure that
1131 /// determines the split.
1132 /// Additional libraries might provide more complex patterns like
1133 /// regular expressions.
1a4d82fc 1134 ///
62682a34 1135 /// # Iterator behavior
1a4d82fc 1136 ///
62682a34
SL
1137 /// The returned iterator requires that the pattern supports a
1138 /// reverse search,
1139 /// and it will be double ended if a forward/reverse search yields
1140 /// the same elements.
1a4d82fc 1141 ///
62682a34
SL
1142 /// For iterating from the front, `matches()` can be used.
1143 ///
1144 /// # Examples
c34b1796
AL
1145 ///
1146 /// ```
62682a34
SL
1147 /// let v: Vec<&str> = "abcXXXabcYYYabc".rmatches("abc").collect();
1148 /// assert_eq!(v, ["abc", "abc", "abc"]);
1a4d82fc 1149 ///
62682a34
SL
1150 /// let v: Vec<&str> = "1abc2abc3".rmatches(char::is_numeric).collect();
1151 /// assert_eq!(v, ["3", "2", "1"]);
c34b1796 1152 /// ```
62682a34
SL
1153 #[stable(feature = "str_matches", since = "1.2.0")]
1154 pub fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
1155 where P::Searcher: ReverseSearcher<'a>
1156 {
1157 core_str::StrExt::rmatches(self, pat)
1158 }
1159
1160 /// An iterator over the start and end indices of the disjoint matches
1161 /// of a pattern within `self`.
c34b1796 1162 ///
62682a34
SL
1163 /// For matches of `pat` within `self` that overlap, only the indices
1164 /// corresponding to the first
1165 /// match are returned.
c34b1796 1166 ///
62682a34
SL
1167 /// The pattern can be a simple `&str`, `char`, or a closure that
1168 /// determines
1169 /// the split.
1170 /// Additional libraries might provide more complex patterns like
1171 /// regular expressions.
1a4d82fc 1172 ///
62682a34 1173 /// # Iterator behavior
c34b1796 1174 ///
62682a34
SL
1175 /// The returned iterator will be double ended if the pattern allows a
1176 /// reverse search
1177 /// and forward/reverse search yields the same elements. This is true for,
1178 /// eg, `char` but not
1179 /// for `&str`.
1180 ///
1181 /// If the pattern allows a reverse search but its results might differ
1182 /// from a forward search, `rmatch_indices()` can be used.
1183 ///
1184 /// # Examples
1a4d82fc 1185 ///
c34b1796 1186 /// ```
c1a9b12d
SL
1187 /// #![feature(str_match_indices)]
1188 ///
62682a34
SL
1189 /// let v: Vec<(usize, usize)> = "abcXXXabcYYYabc".match_indices("abc").collect();
1190 /// assert_eq!(v, [(0, 3), (6, 9), (12, 15)]);
c34b1796 1191 ///
62682a34
SL
1192 /// let v: Vec<(usize, usize)> = "1abcabc2".match_indices("abc").collect();
1193 /// assert_eq!(v, [(1, 4), (4, 7)]);
1194 ///
1195 /// let v: Vec<(usize, usize)> = "ababa".match_indices("aba").collect();
1196 /// assert_eq!(v, [(0, 3)]); // only the first `aba`
1a4d82fc 1197 /// ```
62682a34 1198 #[unstable(feature = "str_match_indices",
e9174d1e
SL
1199 reason = "might have its iterator type changed",
1200 issue = "27743")]
62682a34
SL
1201 // NB: Right now MatchIndices yields `(usize, usize)`, but it would
1202 // be more consistent with `matches` and `char_indices` to return `(usize, &str)`
1203 pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1204 core_str::StrExt::match_indices(self, pat)
1a4d82fc
JJ
1205 }
1206
62682a34
SL
1207 /// An iterator over the start and end indices of the disjoint matches of
1208 /// a pattern within
1209 /// `self`, yielded in reverse order.
1a4d82fc 1210 ///
62682a34
SL
1211 /// For matches of `pat` within `self` that overlap, only the indices
1212 /// corresponding to the last
1213 /// match are returned.
1a4d82fc 1214 ///
62682a34
SL
1215 /// The pattern can be a simple `&str`, `char`, or a closure that
1216 /// determines
1217 /// the split.
1218 /// Additional libraries might provide more complex patterns like
1219 /// regular expressions.
1220 ///
1221 /// # Iterator behavior
1222 ///
1223 /// The returned iterator requires that the pattern supports a
1224 /// reverse search,
1225 /// and it will be double ended if a forward/reverse search yields
1226 /// the same elements.
1227 ///
1228 /// For iterating from the front, `match_indices()` can be used.
1a4d82fc 1229 ///
c34b1796 1230 /// # Examples
1a4d82fc 1231 ///
1a4d82fc 1232 /// ```
c1a9b12d
SL
1233 /// #![feature(str_match_indices)]
1234 ///
62682a34
SL
1235 /// let v: Vec<(usize, usize)> = "abcXXXabcYYYabc".rmatch_indices("abc").collect();
1236 /// assert_eq!(v, [(12, 15), (6, 9), (0, 3)]);
1a4d82fc 1237 ///
62682a34
SL
1238 /// let v: Vec<(usize, usize)> = "1abcabc2".rmatch_indices("abc").collect();
1239 /// assert_eq!(v, [(4, 7), (1, 4)]);
c34b1796 1240 ///
62682a34
SL
1241 /// let v: Vec<(usize, usize)> = "ababa".rmatch_indices("aba").collect();
1242 /// assert_eq!(v, [(2, 5)]); // only the last `aba`
1a4d82fc 1243 /// ```
62682a34 1244 #[unstable(feature = "str_match_indices",
e9174d1e
SL
1245 reason = "might have its iterator type changed",
1246 issue = "27743")]
62682a34
SL
1247 // NB: Right now RMatchIndices yields `(usize, usize)`, but it would
1248 // be more consistent with `rmatches` and `char_indices` to return `(usize, &str)`
1249 pub fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
1250 where P::Searcher: ReverseSearcher<'a>
1251 {
1252 core_str::StrExt::rmatch_indices(self, pat)
1a4d82fc
JJ
1253 }
1254
62682a34 1255 /// Returns a `&str` with leading and trailing whitespace removed.
1a4d82fc 1256 ///
c34b1796
AL
1257 /// # Examples
1258 ///
1259 /// ```
62682a34
SL
1260 /// let s = " Hello\tworld\t";
1261 /// assert_eq!(s.trim(), "Hello\tworld");
c34b1796 1262 /// ```
85aaf69f 1263 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1264 pub fn trim(&self) -> &str {
1265 UnicodeStr::trim(self)
1a4d82fc
JJ
1266 }
1267
62682a34 1268 /// Returns a `&str` with leading whitespace removed.
1a4d82fc 1269 ///
c34b1796 1270 /// # Examples
1a4d82fc
JJ
1271 ///
1272 /// ```
62682a34
SL
1273 /// let s = " Hello\tworld\t";
1274 /// assert_eq!(s.trim_left(), "Hello\tworld\t");
1a4d82fc 1275 /// ```
85aaf69f 1276 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1277 pub fn trim_left(&self) -> &str {
1278 UnicodeStr::trim_left(self)
1a4d82fc
JJ
1279 }
1280
62682a34 1281 /// Returns a `&str` with trailing whitespace removed.
1a4d82fc 1282 ///
c34b1796 1283 /// # Examples
1a4d82fc
JJ
1284 ///
1285 /// ```
62682a34
SL
1286 /// let s = " Hello\tworld\t";
1287 /// assert_eq!(s.trim_right(), " Hello\tworld");
1a4d82fc 1288 /// ```
85aaf69f 1289 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1290 pub fn trim_right(&self) -> &str {
1291 UnicodeStr::trim_right(self)
1a4d82fc
JJ
1292 }
1293
62682a34
SL
1294 /// Returns a string with all pre- and suffixes that match a pattern
1295 /// repeatedly removed.
c34b1796 1296 ///
62682a34
SL
1297 /// The pattern can be a simple `char`, or a closure that determines
1298 /// the split.
c34b1796 1299 ///
62682a34 1300 /// # Examples
1a4d82fc 1301 ///
62682a34 1302 /// Simple patterns:
1a4d82fc
JJ
1303 ///
1304 /// ```
62682a34
SL
1305 /// assert_eq!("11foo1bar11".trim_matches('1'), "foo1bar");
1306 /// assert_eq!("123foo1bar123".trim_matches(char::is_numeric), "foo1bar");
1307 ///
1308 /// let x: &[_] = &['1', '2'];
1309 /// assert_eq!("12foo1bar12".trim_matches(x), "foo1bar");
c34b1796
AL
1310 /// ```
1311 ///
62682a34 1312 /// A more complex pattern, using a closure:
c34b1796
AL
1313 ///
1314 /// ```
62682a34 1315 /// assert_eq!("1foo1barXX".trim_matches(|c| c == '1' || c == 'X'), "foo1bar");
1a4d82fc 1316 /// ```
85aaf69f 1317 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1318 pub fn trim_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1319 where P::Searcher: DoubleEndedSearcher<'a>
1320 {
1321 core_str::StrExt::trim_matches(self, pat)
1a4d82fc
JJ
1322 }
1323
62682a34
SL
1324 /// Returns a string with all prefixes that match a pattern
1325 /// repeatedly removed.
1a4d82fc 1326 ///
62682a34
SL
1327 /// The pattern can be a simple `&str`, `char`, or a closure that
1328 /// determines the split.
1a4d82fc 1329 ///
c34b1796 1330 /// # Examples
1a4d82fc 1331 ///
c34b1796 1332 /// ```
62682a34
SL
1333 /// assert_eq!("11foo1bar11".trim_left_matches('1'), "foo1bar11");
1334 /// assert_eq!("123foo1bar123".trim_left_matches(char::is_numeric), "foo1bar123");
c34b1796 1335 ///
62682a34
SL
1336 /// let x: &[_] = &['1', '2'];
1337 /// assert_eq!("12foo1bar12".trim_left_matches(x), "foo1bar12");
1a4d82fc 1338 /// ```
62682a34
SL
1339 #[stable(feature = "rust1", since = "1.0.0")]
1340 pub fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
1341 core_str::StrExt::trim_left_matches(self, pat)
1a4d82fc
JJ
1342 }
1343
62682a34
SL
1344 /// Returns a string with all suffixes that match a pattern
1345 /// repeatedly removed.
1346 ///
1347 /// The pattern can be a simple `&str`, `char`, or a closure that
1348 /// determines the split.
1a4d82fc 1349 ///
c34b1796 1350 /// # Examples
1a4d82fc 1351 ///
62682a34
SL
1352 /// Simple patterns:
1353 ///
c34b1796 1354 /// ```
62682a34
SL
1355 /// assert_eq!("11foo1bar11".trim_right_matches('1'), "11foo1bar");
1356 /// assert_eq!("123foo1bar123".trim_right_matches(char::is_numeric), "123foo1bar");
c34b1796 1357 ///
62682a34
SL
1358 /// let x: &[_] = &['1', '2'];
1359 /// assert_eq!("12foo1bar12".trim_right_matches(x), "12foo1bar");
1a4d82fc 1360 /// ```
1a4d82fc 1361 ///
62682a34 1362 /// A more complex pattern, using a closure:
c34b1796
AL
1363 ///
1364 /// ```
62682a34 1365 /// assert_eq!("1fooX".trim_left_matches(|c| c == '1' || c == 'X'), "fooX");
1a4d82fc 1366 /// ```
62682a34
SL
1367 #[stable(feature = "rust1", since = "1.0.0")]
1368 pub fn trim_right_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1369 where P::Searcher: ReverseSearcher<'a>
1370 {
1371 core_str::StrExt::trim_right_matches(self, pat)
1a4d82fc
JJ
1372 }
1373
62682a34 1374 /// Parses `self` into the specified type.
d9579d0f 1375 ///
62682a34 1376 /// # Failure
d9579d0f 1377 ///
62682a34 1378 /// Will return `Err` if it's not possible to parse `self` into the type.
1a4d82fc 1379 ///
62682a34 1380 /// # Example
c34b1796 1381 ///
62682a34
SL
1382 /// ```
1383 /// assert_eq!("4".parse::<u32>(), Ok(4));
1384 /// ```
c34b1796 1385 ///
62682a34 1386 /// Failing:
c34b1796
AL
1387 ///
1388 /// ```
62682a34 1389 /// assert!("j".parse::<u32>().is_err());
c34b1796 1390 /// ```
62682a34 1391 #[inline]
85aaf69f 1392 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1393 pub fn parse<F: FromStr>(&self) -> Result<F, F::Err> {
1394 core_str::StrExt::parse(self)
1a4d82fc
JJ
1395 }
1396
62682a34
SL
1397 /// Replaces all occurrences of one string with another.
1398 ///
1399 /// `replace` takes two arguments, a sub-`&str` to find in `self`, and a
1400 /// second `&str` to
1401 /// replace it with. If the original `&str` isn't found, no change occurs.
c34b1796
AL
1402 ///
1403 /// # Examples
1404 ///
1405 /// ```
62682a34
SL
1406 /// let s = "this is old";
1407 ///
1408 /// assert_eq!(s.replace("old", "new"), "this is new");
c34b1796 1409 /// ```
c34b1796 1410 ///
62682a34 1411 /// When a `&str` isn't found:
c34b1796
AL
1412 ///
1413 /// ```
62682a34
SL
1414 /// let s = "this is old";
1415 /// assert_eq!(s.replace("cookie monster", "little lamb"), s);
c34b1796 1416 /// ```
85aaf69f 1417 #[stable(feature = "rust1", since = "1.0.0")]
62682a34
SL
1418 pub fn replace(&self, from: &str, to: &str) -> String {
1419 let mut result = String::new();
1420 let mut last_end = 0;
1421 for (start, end) in self.match_indices(from) {
1422 result.push_str(unsafe { self.slice_unchecked(last_end, start) });
1423 result.push_str(to);
1424 last_end = end;
1425 }
1426 result.push_str(unsafe { self.slice_unchecked(last_end, self.len()) });
1427 result
1a4d82fc 1428 }
1a4d82fc 1429
c34b1796
AL
1430 /// Returns the lowercase equivalent of this string.
1431 ///
1432 /// # Examples
1433 ///
62682a34 1434 /// ```
c34b1796
AL
1435 /// let s = "HELLO";
1436 /// assert_eq!(s.to_lowercase(), "hello");
62682a34
SL
1437 /// ```
1438 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
c34b1796
AL
1439 pub fn to_lowercase(&self) -> String {
1440 let mut s = String::with_capacity(self.len());
62682a34
SL
1441 for (i, c) in self[..].char_indices() {
1442 if c == 'Σ' {
1443 // Σ maps to σ, except at the end of a word where it maps to ς.
1444 // This is the only conditional (contextual) but language-independent mapping
1445 // in `SpecialCasing.txt`,
1446 // so hard-code it rather than have a generic "condition" mechanim.
1447 // See https://github.com/rust-lang/rust/issues/26035
1448 map_uppercase_sigma(self, i, &mut s)
1449 } else {
1450 s.extend(c.to_lowercase());
1451 }
1452 }
c34b1796 1453 return s;
62682a34
SL
1454
1455 fn map_uppercase_sigma(from: &str, i: usize, to: &mut String) {
1456 // See http://www.unicode.org/versions/Unicode7.0.0/ch03.pdf#G33992
1457 // for the definition of `Final_Sigma`.
1458 debug_assert!('Σ'.len_utf8() == 2);
1459 let is_word_final =
1460 case_ignoreable_then_cased(from[..i].chars().rev()) &&
1461 !case_ignoreable_then_cased(from[i + 2..].chars());
1462 to.push_str(if is_word_final { "ς" } else { "σ" });
1463 }
1464
1465 fn case_ignoreable_then_cased<I: Iterator<Item=char>>(iter: I) -> bool {
1466 use rustc_unicode::derived_property::{Cased, Case_Ignorable};
1467 match iter.skip_while(|&c| Case_Ignorable(c)).next() {
1468 Some(c) => Cased(c),
1469 None => false,
1470 }
1471 }
1a4d82fc
JJ
1472 }
1473
c34b1796
AL
1474 /// Returns the uppercase equivalent of this string.
1475 ///
1476 /// # Examples
1477 ///
62682a34 1478 /// ```
c34b1796
AL
1479 /// let s = "hello";
1480 /// assert_eq!(s.to_uppercase(), "HELLO");
62682a34
SL
1481 /// ```
1482 #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
c34b1796
AL
1483 pub fn to_uppercase(&self) -> String {
1484 let mut s = String::with_capacity(self.len());
62682a34 1485 s.extend(self.chars().flat_map(|c| c.to_uppercase()));
c34b1796 1486 return s;
1a4d82fc 1487 }
62682a34
SL
1488
1489 /// Escapes each char in `s` with `char::escape_default`.
1490 #[unstable(feature = "str_escape",
e9174d1e
SL
1491 reason = "return type may change to be an iterator",
1492 issue = "27791")]
62682a34
SL
1493 pub fn escape_default(&self) -> String {
1494 self.chars().flat_map(|c| c.escape_default()).collect()
1495 }
1496
1497 /// Escapes each char in `s` with `char::escape_unicode`.
1498 #[unstable(feature = "str_escape",
e9174d1e
SL
1499 reason = "return type may change to be an iterator",
1500 issue = "27791")]
62682a34
SL
1501 pub fn escape_unicode(&self) -> String {
1502 self.chars().flat_map(|c| c.escape_unicode()).collect()
1503 }
c1a9b12d
SL
1504
1505 /// Converts the `Box<str>` into a `String` without copying or allocating.
e9174d1e 1506 #[stable(feature = "box_str", since = "1.4.0")]
c1a9b12d
SL
1507 pub fn into_string(self: Box<str>) -> String {
1508 unsafe {
1509 let slice = mem::transmute::<Box<str>, Box<[u8]>>(self);
1510 String::from_utf8_unchecked(slice.into_vec())
1511 }
1512 }
1a4d82fc 1513}