1 | //! A UTF-8βencoded, growable string. |
2 | //! |
3 | //! This module contains the [`String`] type, the [`ToString`] trait for |
4 | //! converting to strings, and several error types that may result from |
5 | //! working with [`String`]s. |
6 | //! |
7 | //! # Examples |
8 | //! |
9 | //! There are multiple ways to create a new [`String`] from a string literal: |
10 | //! |
11 | //! ``` |
12 | //! let s = "Hello" .to_string(); |
13 | //! |
14 | //! let s = String::from("world" ); |
15 | //! let s: String = "also this" .into(); |
16 | //! ``` |
17 | //! |
18 | //! You can create a new [`String`] from an existing one by concatenating with |
19 | //! `+`: |
20 | //! |
21 | //! ``` |
22 | //! let s = "Hello" .to_string(); |
23 | //! |
24 | //! let message = s + " world!" ; |
25 | //! ``` |
26 | //! |
27 | //! If you have a vector of valid UTF-8 bytes, you can make a [`String`] out of |
28 | //! it. You can do the reverse too. |
29 | //! |
30 | //! ``` |
31 | //! let sparkle_heart = vec![240, 159, 146, 150]; |
32 | //! |
33 | //! // We know these bytes are valid, so we'll use `unwrap()`. |
34 | //! let sparkle_heart = String::from_utf8(sparkle_heart).unwrap(); |
35 | //! |
36 | //! assert_eq!("π" , sparkle_heart); |
37 | //! |
38 | //! let bytes = sparkle_heart.into_bytes(); |
39 | //! |
40 | //! assert_eq!(bytes, [240, 159, 146, 150]); |
41 | //! ``` |
42 | |
43 | #![stable (feature = "rust1" , since = "1.0.0" )] |
44 | |
45 | use core::error::Error; |
46 | use core::fmt; |
47 | use core::hash; |
48 | #[cfg (not(no_global_oom_handling))] |
49 | use core::iter::from_fn; |
50 | use core::iter::FusedIterator; |
51 | #[cfg (not(no_global_oom_handling))] |
52 | use core::ops::Add; |
53 | #[cfg (not(no_global_oom_handling))] |
54 | use core::ops::AddAssign; |
55 | #[cfg (not(no_global_oom_handling))] |
56 | use core::ops::Bound::{Excluded, Included, Unbounded}; |
57 | use core::ops::{self, Range, RangeBounds}; |
58 | use core::ptr; |
59 | use core::slice; |
60 | use core::str::pattern::Pattern; |
61 | #[cfg (not(no_global_oom_handling))] |
62 | use core::str::Utf8Chunks; |
63 | |
64 | #[cfg (not(no_global_oom_handling))] |
65 | use crate::borrow::{Cow, ToOwned}; |
66 | use crate::boxed::Box; |
67 | use crate::collections::TryReserveError; |
68 | use crate::str::{self, from_utf8_unchecked_mut, Chars, Utf8Error}; |
69 | #[cfg (not(no_global_oom_handling))] |
70 | use crate::str::{from_boxed_utf8_unchecked, FromStr}; |
71 | use crate::vec::Vec; |
72 | |
73 | /// A UTF-8βencoded, growable string. |
74 | /// |
75 | /// `String` is the most common string type. It has ownership over the contents |
76 | /// of the string, stored in a heap-allocated buffer (see [Representation](#representation)). |
77 | /// It is closely related to its borrowed counterpart, the primitive [`str`]. |
78 | /// |
79 | /// # Examples |
80 | /// |
81 | /// You can create a `String` from [a literal string][`&str`] with [`String::from`]: |
82 | /// |
83 | /// [`String::from`]: From::from |
84 | /// |
85 | /// ``` |
86 | /// let hello = String::from("Hello, world!" ); |
87 | /// ``` |
88 | /// |
89 | /// You can append a [`char`] to a `String` with the [`push`] method, and |
90 | /// append a [`&str`] with the [`push_str`] method: |
91 | /// |
92 | /// ``` |
93 | /// let mut hello = String::from("Hello, " ); |
94 | /// |
95 | /// hello.push('w' ); |
96 | /// hello.push_str("orld!" ); |
97 | /// ``` |
98 | /// |
99 | /// [`push`]: String::push |
100 | /// [`push_str`]: String::push_str |
101 | /// |
102 | /// If you have a vector of UTF-8 bytes, you can create a `String` from it with |
103 | /// the [`from_utf8`] method: |
104 | /// |
105 | /// ``` |
106 | /// // some bytes, in a vector |
107 | /// let sparkle_heart = vec![240, 159, 146, 150]; |
108 | /// |
109 | /// // We know these bytes are valid, so we'll use `unwrap()`. |
110 | /// let sparkle_heart = String::from_utf8(sparkle_heart).unwrap(); |
111 | /// |
112 | /// assert_eq!("π" , sparkle_heart); |
113 | /// ``` |
114 | /// |
115 | /// [`from_utf8`]: String::from_utf8 |
116 | /// |
117 | /// # UTF-8 |
118 | /// |
119 | /// `String`s are always valid UTF-8. If you need a non-UTF-8 string, consider |
120 | /// [`OsString`]. It is similar, but without the UTF-8 constraint. Because UTF-8 |
121 | /// is a variable width encoding, `String`s are typically smaller than an array of |
122 | /// the same `chars`: |
123 | /// |
124 | /// ``` |
125 | /// use std::mem; |
126 | /// |
127 | /// // `s` is ASCII which represents each `char` as one byte |
128 | /// let s = "hello" ; |
129 | /// assert_eq!(s.len(), 5); |
130 | /// |
131 | /// // A `char` array with the same contents would be longer because |
132 | /// // every `char` is four bytes |
133 | /// let s = ['h' , 'e' , 'l' , 'l' , 'o' ]; |
134 | /// let size: usize = s.into_iter().map(|c| mem::size_of_val(&c)).sum(); |
135 | /// assert_eq!(size, 20); |
136 | /// |
137 | /// // However, for non-ASCII strings, the difference will be smaller |
138 | /// // and sometimes they are the same |
139 | /// let s = "πππππ" ; |
140 | /// assert_eq!(s.len(), 20); |
141 | /// |
142 | /// let s = ['π' , 'π' , 'π' , 'π' , 'π' ]; |
143 | /// let size: usize = s.into_iter().map(|c| mem::size_of_val(&c)).sum(); |
144 | /// assert_eq!(size, 20); |
145 | /// ``` |
146 | /// |
147 | /// This raises interesting questions as to how `s[i]` should work. |
148 | /// What should `i` be here? Several options include byte indices and |
149 | /// `char` indices but, because of UTF-8 encoding, only byte indices |
150 | /// would provide constant time indexing. Getting the `i`th `char`, for |
151 | /// example, is available using [`chars`]: |
152 | /// |
153 | /// ``` |
154 | /// let s = "hello" ; |
155 | /// let third_character = s.chars().nth(2); |
156 | /// assert_eq!(third_character, Some('l' )); |
157 | /// |
158 | /// let s = "πππππ" ; |
159 | /// let third_character = s.chars().nth(2); |
160 | /// assert_eq!(third_character, Some('π' )); |
161 | /// ``` |
162 | /// |
163 | /// Next, what should `s[i]` return? Because indexing returns a reference |
164 | /// to underlying data it could be `&u8`, `&[u8]`, or something else similar. |
165 | /// Since we're only providing one index, `&u8` makes the most sense but that |
166 | /// might not be what the user expects and can be explicitly achieved with |
167 | /// [`as_bytes()`]: |
168 | /// |
169 | /// ``` |
170 | /// // The first byte is 104 - the byte value of `'h'` |
171 | /// let s = "hello" ; |
172 | /// assert_eq!(s.as_bytes()[0], 104); |
173 | /// // or |
174 | /// assert_eq!(s.as_bytes()[0], b'h' ); |
175 | /// |
176 | /// // The first byte is 240 which isn't obviously useful |
177 | /// let s = "πππππ" ; |
178 | /// assert_eq!(s.as_bytes()[0], 240); |
179 | /// ``` |
180 | /// |
181 | /// Due to these ambiguities/restrictions, indexing with a `usize` is simply |
182 | /// forbidden: |
183 | /// |
184 | /// ```compile_fail,E0277 |
185 | /// let s = "hello" ; |
186 | /// |
187 | /// // The following will not compile! |
188 | /// println!("The first letter of s is {}" , s[0]); |
189 | /// ``` |
190 | /// |
191 | /// It is more clear, however, how `&s[i..j]` should work (that is, |
192 | /// indexing with a range). It should accept byte indices (to be constant-time) |
193 | /// and return a `&str` which is UTF-8 encoded. This is also called "string slicing". |
194 | /// Note this will panic if the byte indices provided are not character |
195 | /// boundaries - see [`is_char_boundary`] for more details. See the implementations |
196 | /// for [`SliceIndex<str>`] for more details on string slicing. For a non-panicking |
197 | /// version of string slicing, see [`get`]. |
198 | /// |
199 | /// [`OsString`]: ../../std/ffi/struct.OsString.html "ffi::OsString" |
200 | /// [`SliceIndex<str>`]: core::slice::SliceIndex |
201 | /// [`as_bytes()`]: str::as_bytes |
202 | /// [`get`]: str::get |
203 | /// [`is_char_boundary`]: str::is_char_boundary |
204 | /// |
205 | /// The [`bytes`] and [`chars`] methods return iterators over the bytes and |
206 | /// codepoints of the string, respectively. To iterate over codepoints along |
207 | /// with byte indices, use [`char_indices`]. |
208 | /// |
209 | /// [`bytes`]: str::bytes |
210 | /// [`chars`]: str::chars |
211 | /// [`char_indices`]: str::char_indices |
212 | /// |
213 | /// # Deref |
214 | /// |
215 | /// `String` implements <code>[Deref]<Target = [str]></code>, and so inherits all of [`str`]'s |
216 | /// methods. In addition, this means that you can pass a `String` to a |
217 | /// function which takes a [`&str`] by using an ampersand (`&`): |
218 | /// |
219 | /// ``` |
220 | /// fn takes_str(s: &str) { } |
221 | /// |
222 | /// let s = String::from("Hello" ); |
223 | /// |
224 | /// takes_str(&s); |
225 | /// ``` |
226 | /// |
227 | /// This will create a [`&str`] from the `String` and pass it in. This |
228 | /// conversion is very inexpensive, and so generally, functions will accept |
229 | /// [`&str`]s as arguments unless they need a `String` for some specific |
230 | /// reason. |
231 | /// |
232 | /// In certain cases Rust doesn't have enough information to make this |
233 | /// conversion, known as [`Deref`] coercion. In the following example a string |
234 | /// slice [`&'a str`][`&str`] implements the trait `TraitExample`, and the function |
235 | /// `example_func` takes anything that implements the trait. In this case Rust |
236 | /// would need to make two implicit conversions, which Rust doesn't have the |
237 | /// means to do. For that reason, the following example will not compile. |
238 | /// |
239 | /// ```compile_fail,E0277 |
240 | /// trait TraitExample {} |
241 | /// |
242 | /// impl<'a> TraitExample for &'a str {} |
243 | /// |
244 | /// fn example_func<A: TraitExample>(example_arg: A) {} |
245 | /// |
246 | /// let example_string = String::from("example_string" ); |
247 | /// example_func(&example_string); |
248 | /// ``` |
249 | /// |
250 | /// There are two options that would work instead. The first would be to |
251 | /// change the line `example_func(&example_string);` to |
252 | /// `example_func(example_string.as_str());`, using the method [`as_str()`] |
253 | /// to explicitly extract the string slice containing the string. The second |
254 | /// way changes `example_func(&example_string);` to |
255 | /// `example_func(&*example_string);`. In this case we are dereferencing a |
256 | /// `String` to a [`str`], then referencing the [`str`] back to |
257 | /// [`&str`]. The second way is more idiomatic, however both work to do the |
258 | /// conversion explicitly rather than relying on the implicit conversion. |
259 | /// |
260 | /// # Representation |
261 | /// |
262 | /// A `String` is made up of three components: a pointer to some bytes, a |
263 | /// length, and a capacity. The pointer points to the internal buffer which `String` |
264 | /// uses to store its data. The length is the number of bytes currently stored |
265 | /// in the buffer, and the capacity is the size of the buffer in bytes. As such, |
266 | /// the length will always be less than or equal to the capacity. |
267 | /// |
268 | /// This buffer is always stored on the heap. |
269 | /// |
270 | /// You can look at these with the [`as_ptr`], [`len`], and [`capacity`] |
271 | /// methods: |
272 | /// |
273 | /// ``` |
274 | /// use std::mem; |
275 | /// |
276 | /// let story = String::from("Once upon a time..." ); |
277 | /// |
278 | // FIXME Update this when vec_into_raw_parts is stabilized |
279 | /// // Prevent automatically dropping the String's data |
280 | /// let mut story = mem::ManuallyDrop::new(story); |
281 | /// |
282 | /// let ptr = story.as_mut_ptr(); |
283 | /// let len = story.len(); |
284 | /// let capacity = story.capacity(); |
285 | /// |
286 | /// // story has nineteen bytes |
287 | /// assert_eq!(19, len); |
288 | /// |
289 | /// // We can re-build a String out of ptr, len, and capacity. This is all |
290 | /// // unsafe because we are responsible for making sure the components are |
291 | /// // valid: |
292 | /// let s = unsafe { String::from_raw_parts(ptr, len, capacity) } ; |
293 | /// |
294 | /// assert_eq!(String::from("Once upon a time..." ), s); |
295 | /// ``` |
296 | /// |
297 | /// [`as_ptr`]: str::as_ptr |
298 | /// [`len`]: String::len |
299 | /// [`capacity`]: String::capacity |
300 | /// |
301 | /// If a `String` has enough capacity, adding elements to it will not |
302 | /// re-allocate. For example, consider this program: |
303 | /// |
304 | /// ``` |
305 | /// let mut s = String::new(); |
306 | /// |
307 | /// println!("{}" , s.capacity()); |
308 | /// |
309 | /// for _ in 0..5 { |
310 | /// s.push_str("hello" ); |
311 | /// println!("{}" , s.capacity()); |
312 | /// } |
313 | /// ``` |
314 | /// |
315 | /// This will output the following: |
316 | /// |
317 | /// ```text |
318 | /// 0 |
319 | /// 8 |
320 | /// 16 |
321 | /// 16 |
322 | /// 32 |
323 | /// 32 |
324 | /// ``` |
325 | /// |
326 | /// At first, we have no memory allocated at all, but as we append to the |
327 | /// string, it increases its capacity appropriately. If we instead use the |
328 | /// [`with_capacity`] method to allocate the correct capacity initially: |
329 | /// |
330 | /// ``` |
331 | /// let mut s = String::with_capacity(25); |
332 | /// |
333 | /// println!("{}" , s.capacity()); |
334 | /// |
335 | /// for _ in 0..5 { |
336 | /// s.push_str("hello" ); |
337 | /// println!("{}" , s.capacity()); |
338 | /// } |
339 | /// ``` |
340 | /// |
341 | /// [`with_capacity`]: String::with_capacity |
342 | /// |
343 | /// We end up with a different output: |
344 | /// |
345 | /// ```text |
346 | /// 25 |
347 | /// 25 |
348 | /// 25 |
349 | /// 25 |
350 | /// 25 |
351 | /// 25 |
352 | /// ``` |
353 | /// |
354 | /// Here, there's no need to allocate more memory inside the loop. |
355 | /// |
356 | /// [str]: prim@str "str" |
357 | /// [`str`]: prim@str "str" |
358 | /// [`&str`]: prim@str "&str" |
359 | /// [Deref]: core::ops::Deref "ops::Deref" |
360 | /// [`Deref`]: core::ops::Deref "ops::Deref" |
361 | /// [`as_str()`]: String::as_str |
362 | #[derive (PartialEq, PartialOrd, Eq, Ord)] |
363 | #[stable (feature = "rust1" , since = "1.0.0" )] |
364 | #[cfg_attr (not(test), lang = "String" )] |
365 | pub struct String { |
366 | vec: Vec<u8>, |
367 | } |
368 | |
369 | /// A possible error value when converting a `String` from a UTF-8 byte vector. |
370 | /// |
371 | /// This type is the error type for the [`from_utf8`] method on [`String`]. It |
372 | /// is designed in such a way to carefully avoid reallocations: the |
373 | /// [`into_bytes`] method will give back the byte vector that was used in the |
374 | /// conversion attempt. |
375 | /// |
376 | /// [`from_utf8`]: String::from_utf8 |
377 | /// [`into_bytes`]: FromUtf8Error::into_bytes |
378 | /// |
379 | /// The [`Utf8Error`] type provided by [`std::str`] represents an error that may |
380 | /// occur when converting a slice of [`u8`]s to a [`&str`]. In this sense, it's |
381 | /// an analogue to `FromUtf8Error`, and you can get one from a `FromUtf8Error` |
382 | /// through the [`utf8_error`] method. |
383 | /// |
384 | /// [`Utf8Error`]: str::Utf8Error "std::str::Utf8Error" |
385 | /// [`std::str`]: core::str "std::str" |
386 | /// [`&str`]: prim@str "&str" |
387 | /// [`utf8_error`]: FromUtf8Error::utf8_error |
388 | /// |
389 | /// # Examples |
390 | /// |
391 | /// ``` |
392 | /// // some invalid bytes, in a vector |
393 | /// let bytes = vec![0, 159]; |
394 | /// |
395 | /// let value = String::from_utf8(bytes); |
396 | /// |
397 | /// assert!(value.is_err()); |
398 | /// assert_eq!(vec![0, 159], value.unwrap_err().into_bytes()); |
399 | /// ``` |
400 | #[stable (feature = "rust1" , since = "1.0.0" )] |
401 | #[cfg_attr (not(no_global_oom_handling), derive(Clone))] |
402 | #[derive (Debug, PartialEq, Eq)] |
403 | pub struct FromUtf8Error { |
404 | bytes: Vec<u8>, |
405 | error: Utf8Error, |
406 | } |
407 | |
408 | /// A possible error value when converting a `String` from a UTF-16 byte slice. |
409 | /// |
410 | /// This type is the error type for the [`from_utf16`] method on [`String`]. |
411 | /// |
412 | /// [`from_utf16`]: String::from_utf16 |
413 | /// |
414 | /// # Examples |
415 | /// |
416 | /// ``` |
417 | /// // πmu<invalid>ic |
418 | /// let v = &[0xD834, 0xDD1E, 0x006d, 0x0075, |
419 | /// 0xD800, 0x0069, 0x0063]; |
420 | /// |
421 | /// assert!(String::from_utf16(v).is_err()); |
422 | /// ``` |
423 | #[stable (feature = "rust1" , since = "1.0.0" )] |
424 | #[derive (Debug)] |
425 | pub struct FromUtf16Error(()); |
426 | |
427 | impl String { |
428 | /// Creates a new empty `String`. |
429 | /// |
430 | /// Given that the `String` is empty, this will not allocate any initial |
431 | /// buffer. While that means that this initial operation is very |
432 | /// inexpensive, it may cause excessive allocation later when you add |
433 | /// data. If you have an idea of how much data the `String` will hold, |
434 | /// consider the [`with_capacity`] method to prevent excessive |
435 | /// re-allocation. |
436 | /// |
437 | /// [`with_capacity`]: String::with_capacity |
438 | /// |
439 | /// # Examples |
440 | /// |
441 | /// ``` |
442 | /// let s = String::new(); |
443 | /// ``` |
444 | #[inline ] |
445 | #[rustc_const_stable (feature = "const_string_new" , since = "1.39.0" )] |
446 | #[stable (feature = "rust1" , since = "1.0.0" )] |
447 | #[must_use ] |
448 | pub const fn new() -> String { |
449 | String { vec: Vec::new() } |
450 | } |
451 | |
452 | /// Creates a new empty `String` with at least the specified capacity. |
453 | /// |
454 | /// `String`s have an internal buffer to hold their data. The capacity is |
455 | /// the length of that buffer, and can be queried with the [`capacity`] |
456 | /// method. This method creates an empty `String`, but one with an initial |
457 | /// buffer that can hold at least `capacity` bytes. This is useful when you |
458 | /// may be appending a bunch of data to the `String`, reducing the number of |
459 | /// reallocations it needs to do. |
460 | /// |
461 | /// [`capacity`]: String::capacity |
462 | /// |
463 | /// If the given capacity is `0`, no allocation will occur, and this method |
464 | /// is identical to the [`new`] method. |
465 | /// |
466 | /// [`new`]: String::new |
467 | /// |
468 | /// # Examples |
469 | /// |
470 | /// ``` |
471 | /// let mut s = String::with_capacity(10); |
472 | /// |
473 | /// // The String contains no chars, even though it has capacity for more |
474 | /// assert_eq!(s.len(), 0); |
475 | /// |
476 | /// // These are all done without reallocating... |
477 | /// let cap = s.capacity(); |
478 | /// for _ in 0..10 { |
479 | /// s.push('a' ); |
480 | /// } |
481 | /// |
482 | /// assert_eq!(s.capacity(), cap); |
483 | /// |
484 | /// // ...but this may make the string reallocate |
485 | /// s.push('a' ); |
486 | /// ``` |
487 | #[cfg (not(no_global_oom_handling))] |
488 | #[inline ] |
489 | #[stable (feature = "rust1" , since = "1.0.0" )] |
490 | #[must_use ] |
491 | pub fn with_capacity(capacity: usize) -> String { |
492 | String { vec: Vec::with_capacity(capacity) } |
493 | } |
494 | |
495 | /// Creates a new empty `String` with at least the specified capacity. |
496 | /// |
497 | /// # Errors |
498 | /// |
499 | /// Returns [`Err`] if the capacity exceeds `isize::MAX` bytes, |
500 | /// or if the memory allocator reports failure. |
501 | /// |
502 | #[inline ] |
503 | #[unstable (feature = "try_with_capacity" , issue = "91913" )] |
504 | pub fn try_with_capacity(capacity: usize) -> Result<String, TryReserveError> { |
505 | Ok(String { vec: Vec::try_with_capacity(capacity)? }) |
506 | } |
507 | |
508 | // HACK(japaric): with cfg(test) the inherent `[T]::to_vec` method, which is |
509 | // required for this method definition, is not available. Since we don't |
510 | // require this method for testing purposes, I'll just stub it |
511 | // NB see the slice::hack module in slice.rs for more information |
512 | #[inline ] |
513 | #[cfg (test)] |
514 | pub fn from_str(_: &str) -> String { |
515 | panic!("not available with cfg(test)" ); |
516 | } |
517 | |
518 | /// Converts a vector of bytes to a `String`. |
519 | /// |
520 | /// A string ([`String`]) is made of bytes ([`u8`]), and a vector of bytes |
521 | /// ([`Vec<u8>`]) is made of bytes, so this function converts between the |
522 | /// two. Not all byte slices are valid `String`s, however: `String` |
523 | /// requires that it is valid UTF-8. `from_utf8()` checks to ensure that |
524 | /// the bytes are valid UTF-8, and then does the conversion. |
525 | /// |
526 | /// If you are sure that the byte slice is valid UTF-8, and you don't want |
527 | /// to incur the overhead of the validity check, there is an unsafe version |
528 | /// of this function, [`from_utf8_unchecked`], which has the same behavior |
529 | /// but skips the check. |
530 | /// |
531 | /// This method will take care to not copy the vector, for efficiency's |
532 | /// sake. |
533 | /// |
534 | /// If you need a [`&str`] instead of a `String`, consider |
535 | /// [`str::from_utf8`]. |
536 | /// |
537 | /// The inverse of this method is [`into_bytes`]. |
538 | /// |
539 | /// # Errors |
540 | /// |
541 | /// Returns [`Err`] if the slice is not UTF-8 with a description as to why the |
542 | /// provided bytes are not UTF-8. The vector you moved in is also included. |
543 | /// |
544 | /// # Examples |
545 | /// |
546 | /// Basic usage: |
547 | /// |
548 | /// ``` |
549 | /// // some bytes, in a vector |
550 | /// let sparkle_heart = vec![240, 159, 146, 150]; |
551 | /// |
552 | /// // We know these bytes are valid, so we'll use `unwrap()`. |
553 | /// let sparkle_heart = String::from_utf8(sparkle_heart).unwrap(); |
554 | /// |
555 | /// assert_eq!("π" , sparkle_heart); |
556 | /// ``` |
557 | /// |
558 | /// Incorrect bytes: |
559 | /// |
560 | /// ``` |
561 | /// // some invalid bytes, in a vector |
562 | /// let sparkle_heart = vec![0, 159, 146, 150]; |
563 | /// |
564 | /// assert!(String::from_utf8(sparkle_heart).is_err()); |
565 | /// ``` |
566 | /// |
567 | /// See the docs for [`FromUtf8Error`] for more details on what you can do |
568 | /// with this error. |
569 | /// |
570 | /// [`from_utf8_unchecked`]: String::from_utf8_unchecked |
571 | /// [`Vec<u8>`]: crate::vec::Vec "Vec" |
572 | /// [`&str`]: prim@str "&str" |
573 | /// [`into_bytes`]: String::into_bytes |
574 | #[inline ] |
575 | #[stable (feature = "rust1" , since = "1.0.0" )] |
576 | pub fn from_utf8(vec: Vec<u8>) -> Result<String, FromUtf8Error> { |
577 | match str::from_utf8(&vec) { |
578 | Ok(..) => Ok(String { vec }), |
579 | Err(e) => Err(FromUtf8Error { bytes: vec, error: e }), |
580 | } |
581 | } |
582 | |
583 | /// Converts a slice of bytes to a string, including invalid characters. |
584 | /// |
585 | /// Strings are made of bytes ([`u8`]), and a slice of bytes |
586 | /// ([`&[u8]`][byteslice]) is made of bytes, so this function converts |
587 | /// between the two. Not all byte slices are valid strings, however: strings |
588 | /// are required to be valid UTF-8. During this conversion, |
589 | /// `from_utf8_lossy()` will replace any invalid UTF-8 sequences with |
590 | /// [`U+FFFD REPLACEMENT CHARACTER`][U+FFFD], which looks like this: οΏ½ |
591 | /// |
592 | /// [byteslice]: prim@slice |
593 | /// [U+FFFD]: core::char::REPLACEMENT_CHARACTER |
594 | /// |
595 | /// If you are sure that the byte slice is valid UTF-8, and you don't want |
596 | /// to incur the overhead of the conversion, there is an unsafe version |
597 | /// of this function, [`from_utf8_unchecked`], which has the same behavior |
598 | /// but skips the checks. |
599 | /// |
600 | /// [`from_utf8_unchecked`]: String::from_utf8_unchecked |
601 | /// |
602 | /// This function returns a [`Cow<'a, str>`]. If our byte slice is invalid |
603 | /// UTF-8, then we need to insert the replacement characters, which will |
604 | /// change the size of the string, and hence, require a `String`. But if |
605 | /// it's already valid UTF-8, we don't need a new allocation. This return |
606 | /// type allows us to handle both cases. |
607 | /// |
608 | /// [`Cow<'a, str>`]: crate::borrow::Cow "borrow::Cow" |
609 | /// |
610 | /// # Examples |
611 | /// |
612 | /// Basic usage: |
613 | /// |
614 | /// ``` |
615 | /// // some bytes, in a vector |
616 | /// let sparkle_heart = vec![240, 159, 146, 150]; |
617 | /// |
618 | /// let sparkle_heart = String::from_utf8_lossy(&sparkle_heart); |
619 | /// |
620 | /// assert_eq!("π" , sparkle_heart); |
621 | /// ``` |
622 | /// |
623 | /// Incorrect bytes: |
624 | /// |
625 | /// ``` |
626 | /// // some invalid bytes |
627 | /// let input = b"Hello \xF0\x90\x80World" ; |
628 | /// let output = String::from_utf8_lossy(input); |
629 | /// |
630 | /// assert_eq!("Hello οΏ½World" , output); |
631 | /// ``` |
632 | #[must_use ] |
633 | #[cfg (not(no_global_oom_handling))] |
634 | #[stable (feature = "rust1" , since = "1.0.0" )] |
635 | pub fn from_utf8_lossy(v: &[u8]) -> Cow<'_, str> { |
636 | let mut iter = Utf8Chunks::new(v); |
637 | |
638 | let first_valid = if let Some(chunk) = iter.next() { |
639 | let valid = chunk.valid(); |
640 | if chunk.invalid().is_empty() { |
641 | debug_assert_eq!(valid.len(), v.len()); |
642 | return Cow::Borrowed(valid); |
643 | } |
644 | valid |
645 | } else { |
646 | return Cow::Borrowed("" ); |
647 | }; |
648 | |
649 | const REPLACEMENT: &str = " \u{FFFD}" ; |
650 | |
651 | let mut res = String::with_capacity(v.len()); |
652 | res.push_str(first_valid); |
653 | res.push_str(REPLACEMENT); |
654 | |
655 | for chunk in iter { |
656 | res.push_str(chunk.valid()); |
657 | if !chunk.invalid().is_empty() { |
658 | res.push_str(REPLACEMENT); |
659 | } |
660 | } |
661 | |
662 | Cow::Owned(res) |
663 | } |
664 | |
665 | /// Decode a UTF-16βencoded vector `v` into a `String`, returning [`Err`] |
666 | /// if `v` contains any invalid data. |
667 | /// |
668 | /// # Examples |
669 | /// |
670 | /// ``` |
671 | /// // πmusic |
672 | /// let v = &[0xD834, 0xDD1E, 0x006d, 0x0075, |
673 | /// 0x0073, 0x0069, 0x0063]; |
674 | /// assert_eq!(String::from("πmusic" ), |
675 | /// String::from_utf16(v).unwrap()); |
676 | /// |
677 | /// // πmu<invalid>ic |
678 | /// let v = &[0xD834, 0xDD1E, 0x006d, 0x0075, |
679 | /// 0xD800, 0x0069, 0x0063]; |
680 | /// assert!(String::from_utf16(v).is_err()); |
681 | /// ``` |
682 | #[cfg (not(no_global_oom_handling))] |
683 | #[stable (feature = "rust1" , since = "1.0.0" )] |
684 | pub fn from_utf16(v: &[u16]) -> Result<String, FromUtf16Error> { |
685 | // This isn't done via collect::<Result<_, _>>() for performance reasons. |
686 | // FIXME: the function can be simplified again when #48994 is closed. |
687 | let mut ret = String::with_capacity(v.len()); |
688 | for c in char::decode_utf16(v.iter().cloned()) { |
689 | if let Ok(c) = c { |
690 | ret.push(c); |
691 | } else { |
692 | return Err(FromUtf16Error(())); |
693 | } |
694 | } |
695 | Ok(ret) |
696 | } |
697 | |
698 | /// Decode a UTF-16βencoded slice `v` into a `String`, replacing |
699 | /// invalid data with [the replacement character (`U+FFFD`)][U+FFFD]. |
700 | /// |
701 | /// Unlike [`from_utf8_lossy`] which returns a [`Cow<'a, str>`], |
702 | /// `from_utf16_lossy` returns a `String` since the UTF-16 to UTF-8 |
703 | /// conversion requires a memory allocation. |
704 | /// |
705 | /// [`from_utf8_lossy`]: String::from_utf8_lossy |
706 | /// [`Cow<'a, str>`]: crate::borrow::Cow "borrow::Cow" |
707 | /// [U+FFFD]: core::char::REPLACEMENT_CHARACTER |
708 | /// |
709 | /// # Examples |
710 | /// |
711 | /// ``` |
712 | /// // πmus<invalid>ic<invalid> |
713 | /// let v = &[0xD834, 0xDD1E, 0x006d, 0x0075, |
714 | /// 0x0073, 0xDD1E, 0x0069, 0x0063, |
715 | /// 0xD834]; |
716 | /// |
717 | /// assert_eq!(String::from("πmus \u{FFFD}ic \u{FFFD}" ), |
718 | /// String::from_utf16_lossy(v)); |
719 | /// ``` |
720 | #[cfg (not(no_global_oom_handling))] |
721 | #[must_use ] |
722 | #[inline ] |
723 | #[stable (feature = "rust1" , since = "1.0.0" )] |
724 | pub fn from_utf16_lossy(v: &[u16]) -> String { |
725 | char::decode_utf16(v.iter().cloned()) |
726 | .map(|r| r.unwrap_or(char::REPLACEMENT_CHARACTER)) |
727 | .collect() |
728 | } |
729 | |
730 | /// Decode a UTF-16LEβencoded vector `v` into a `String`, returning [`Err`] |
731 | /// if `v` contains any invalid data. |
732 | /// |
733 | /// # Examples |
734 | /// |
735 | /// Basic usage: |
736 | /// |
737 | /// ``` |
738 | /// #![feature(str_from_utf16_endian)] |
739 | /// // πmusic |
740 | /// let v = &[0x34, 0xD8, 0x1E, 0xDD, 0x6d, 0x00, 0x75, 0x00, |
741 | /// 0x73, 0x00, 0x69, 0x00, 0x63, 0x00]; |
742 | /// assert_eq!(String::from("πmusic" ), |
743 | /// String::from_utf16le(v).unwrap()); |
744 | /// |
745 | /// // πmu<invalid>ic |
746 | /// let v = &[0x34, 0xD8, 0x1E, 0xDD, 0x6d, 0x00, 0x75, 0x00, |
747 | /// 0x00, 0xD8, 0x69, 0x00, 0x63, 0x00]; |
748 | /// assert!(String::from_utf16le(v).is_err()); |
749 | /// ``` |
750 | #[cfg (not(no_global_oom_handling))] |
751 | #[unstable (feature = "str_from_utf16_endian" , issue = "116258" )] |
752 | pub fn from_utf16le(v: &[u8]) -> Result<String, FromUtf16Error> { |
753 | if v.len() % 2 != 0 { |
754 | return Err(FromUtf16Error(())); |
755 | } |
756 | match (cfg!(target_endian = "little" ), unsafe { v.align_to::<u16>() }) { |
757 | (true, ([], v, [])) => Self::from_utf16(v), |
758 | _ => char::decode_utf16(v.array_chunks::<2>().copied().map(u16::from_le_bytes)) |
759 | .collect::<Result<_, _>>() |
760 | .map_err(|_| FromUtf16Error(())), |
761 | } |
762 | } |
763 | |
764 | /// Decode a UTF-16LEβencoded slice `v` into a `String`, replacing |
765 | /// invalid data with [the replacement character (`U+FFFD`)][U+FFFD]. |
766 | /// |
767 | /// Unlike [`from_utf8_lossy`] which returns a [`Cow<'a, str>`], |
768 | /// `from_utf16le_lossy` returns a `String` since the UTF-16 to UTF-8 |
769 | /// conversion requires a memory allocation. |
770 | /// |
771 | /// [`from_utf8_lossy`]: String::from_utf8_lossy |
772 | /// [`Cow<'a, str>`]: crate::borrow::Cow "borrow::Cow" |
773 | /// [U+FFFD]: core::char::REPLACEMENT_CHARACTER |
774 | /// |
775 | /// # Examples |
776 | /// |
777 | /// Basic usage: |
778 | /// |
779 | /// ``` |
780 | /// #![feature(str_from_utf16_endian)] |
781 | /// // πmus<invalid>ic<invalid> |
782 | /// let v = &[0x34, 0xD8, 0x1E, 0xDD, 0x6d, 0x00, 0x75, 0x00, |
783 | /// 0x73, 0x00, 0x1E, 0xDD, 0x69, 0x00, 0x63, 0x00, |
784 | /// 0x34, 0xD8]; |
785 | /// |
786 | /// assert_eq!(String::from("πmus \u{FFFD}ic \u{FFFD}" ), |
787 | /// String::from_utf16le_lossy(v)); |
788 | /// ``` |
789 | #[cfg (not(no_global_oom_handling))] |
790 | #[unstable (feature = "str_from_utf16_endian" , issue = "116258" )] |
791 | pub fn from_utf16le_lossy(v: &[u8]) -> String { |
792 | match (cfg!(target_endian = "little" ), unsafe { v.align_to::<u16>() }) { |
793 | (true, ([], v, [])) => Self::from_utf16_lossy(v), |
794 | (true, ([], v, [_remainder])) => Self::from_utf16_lossy(v) + " \u{FFFD}" , |
795 | _ => { |
796 | let mut iter = v.array_chunks::<2>(); |
797 | let string = char::decode_utf16(iter.by_ref().copied().map(u16::from_le_bytes)) |
798 | .map(|r| r.unwrap_or(char::REPLACEMENT_CHARACTER)) |
799 | .collect(); |
800 | if iter.remainder().is_empty() { string } else { string + " \u{FFFD}" } |
801 | } |
802 | } |
803 | } |
804 | |
805 | /// Decode a UTF-16BEβencoded vector `v` into a `String`, returning [`Err`] |
806 | /// if `v` contains any invalid data. |
807 | /// |
808 | /// # Examples |
809 | /// |
810 | /// Basic usage: |
811 | /// |
812 | /// ``` |
813 | /// #![feature(str_from_utf16_endian)] |
814 | /// // πmusic |
815 | /// let v = &[0xD8, 0x34, 0xDD, 0x1E, 0x00, 0x6d, 0x00, 0x75, |
816 | /// 0x00, 0x73, 0x00, 0x69, 0x00, 0x63]; |
817 | /// assert_eq!(String::from("πmusic" ), |
818 | /// String::from_utf16be(v).unwrap()); |
819 | /// |
820 | /// // πmu<invalid>ic |
821 | /// let v = &[0xD8, 0x34, 0xDD, 0x1E, 0x00, 0x6d, 0x00, 0x75, |
822 | /// 0xD8, 0x00, 0x00, 0x69, 0x00, 0x63]; |
823 | /// assert!(String::from_utf16be(v).is_err()); |
824 | /// ``` |
825 | #[cfg (not(no_global_oom_handling))] |
826 | #[unstable (feature = "str_from_utf16_endian" , issue = "116258" )] |
827 | pub fn from_utf16be(v: &[u8]) -> Result<String, FromUtf16Error> { |
828 | if v.len() % 2 != 0 { |
829 | return Err(FromUtf16Error(())); |
830 | } |
831 | match (cfg!(target_endian = "big" ), unsafe { v.align_to::<u16>() }) { |
832 | (true, ([], v, [])) => Self::from_utf16(v), |
833 | _ => char::decode_utf16(v.array_chunks::<2>().copied().map(u16::from_be_bytes)) |
834 | .collect::<Result<_, _>>() |
835 | .map_err(|_| FromUtf16Error(())), |
836 | } |
837 | } |
838 | |
839 | /// Decode a UTF-16BEβencoded slice `v` into a `String`, replacing |
840 | /// invalid data with [the replacement character (`U+FFFD`)][U+FFFD]. |
841 | /// |
842 | /// Unlike [`from_utf8_lossy`] which returns a [`Cow<'a, str>`], |
843 | /// `from_utf16le_lossy` returns a `String` since the UTF-16 to UTF-8 |
844 | /// conversion requires a memory allocation. |
845 | /// |
846 | /// [`from_utf8_lossy`]: String::from_utf8_lossy |
847 | /// [`Cow<'a, str>`]: crate::borrow::Cow "borrow::Cow" |
848 | /// [U+FFFD]: core::char::REPLACEMENT_CHARACTER |
849 | /// |
850 | /// # Examples |
851 | /// |
852 | /// Basic usage: |
853 | /// |
854 | /// ``` |
855 | /// #![feature(str_from_utf16_endian)] |
856 | /// // πmus<invalid>ic<invalid> |
857 | /// let v = &[0xD8, 0x34, 0xDD, 0x1E, 0x00, 0x6d, 0x00, 0x75, |
858 | /// 0x00, 0x73, 0xDD, 0x1E, 0x00, 0x69, 0x00, 0x63, |
859 | /// 0xD8, 0x34]; |
860 | /// |
861 | /// assert_eq!(String::from("πmus \u{FFFD}ic \u{FFFD}" ), |
862 | /// String::from_utf16be_lossy(v)); |
863 | /// ``` |
864 | #[cfg (not(no_global_oom_handling))] |
865 | #[unstable (feature = "str_from_utf16_endian" , issue = "116258" )] |
866 | pub fn from_utf16be_lossy(v: &[u8]) -> String { |
867 | match (cfg!(target_endian = "big" ), unsafe { v.align_to::<u16>() }) { |
868 | (true, ([], v, [])) => Self::from_utf16_lossy(v), |
869 | (true, ([], v, [_remainder])) => Self::from_utf16_lossy(v) + " \u{FFFD}" , |
870 | _ => { |
871 | let mut iter = v.array_chunks::<2>(); |
872 | let string = char::decode_utf16(iter.by_ref().copied().map(u16::from_be_bytes)) |
873 | .map(|r| r.unwrap_or(char::REPLACEMENT_CHARACTER)) |
874 | .collect(); |
875 | if iter.remainder().is_empty() { string } else { string + " \u{FFFD}" } |
876 | } |
877 | } |
878 | } |
879 | |
880 | /// Decomposes a `String` into its raw components: `(pointer, length, capacity)`. |
881 | /// |
882 | /// Returns the raw pointer to the underlying data, the length of |
883 | /// the string (in bytes), and the allocated capacity of the data |
884 | /// (in bytes). These are the same arguments in the same order as |
885 | /// the arguments to [`from_raw_parts`]. |
886 | /// |
887 | /// After calling this function, the caller is responsible for the |
888 | /// memory previously managed by the `String`. The only way to do |
889 | /// this is to convert the raw pointer, length, and capacity back |
890 | /// into a `String` with the [`from_raw_parts`] function, allowing |
891 | /// the destructor to perform the cleanup. |
892 | /// |
893 | /// [`from_raw_parts`]: String::from_raw_parts |
894 | /// |
895 | /// # Examples |
896 | /// |
897 | /// ``` |
898 | /// #![feature(vec_into_raw_parts)] |
899 | /// let s = String::from("hello" ); |
900 | /// |
901 | /// let (ptr, len, cap) = s.into_raw_parts(); |
902 | /// |
903 | /// let rebuilt = unsafe { String::from_raw_parts(ptr, len, cap) }; |
904 | /// assert_eq!(rebuilt, "hello" ); |
905 | /// ``` |
906 | #[must_use = "`self` will be dropped if the result is not used" ] |
907 | #[unstable (feature = "vec_into_raw_parts" , reason = "new API" , issue = "65816" )] |
908 | pub fn into_raw_parts(self) -> (*mut u8, usize, usize) { |
909 | self.vec.into_raw_parts() |
910 | } |
911 | |
912 | /// Creates a new `String` from a pointer, a length and a capacity. |
913 | /// |
914 | /// # Safety |
915 | /// |
916 | /// This is highly unsafe, due to the number of invariants that aren't |
917 | /// checked: |
918 | /// |
919 | /// * The memory at `buf` needs to have been previously allocated by the |
920 | /// same allocator the standard library uses, with a required alignment of exactly 1. |
921 | /// * `length` needs to be less than or equal to `capacity`. |
922 | /// * `capacity` needs to be the correct value. |
923 | /// * The first `length` bytes at `buf` need to be valid UTF-8. |
924 | /// |
925 | /// Violating these may cause problems like corrupting the allocator's |
926 | /// internal data structures. For example, it is normally **not** safe to |
927 | /// build a `String` from a pointer to a C `char` array containing UTF-8 |
928 | /// _unless_ you are certain that array was originally allocated by the |
929 | /// Rust standard library's allocator. |
930 | /// |
931 | /// The ownership of `buf` is effectively transferred to the |
932 | /// `String` which may then deallocate, reallocate or change the |
933 | /// contents of memory pointed to by the pointer at will. Ensure |
934 | /// that nothing else uses the pointer after calling this |
935 | /// function. |
936 | /// |
937 | /// # Examples |
938 | /// |
939 | /// ``` |
940 | /// use std::mem; |
941 | /// |
942 | /// unsafe { |
943 | /// let s = String::from("hello" ); |
944 | /// |
945 | // FIXME Update this when vec_into_raw_parts is stabilized |
946 | /// // Prevent automatically dropping the String's data |
947 | /// let mut s = mem::ManuallyDrop::new(s); |
948 | /// |
949 | /// let ptr = s.as_mut_ptr(); |
950 | /// let len = s.len(); |
951 | /// let capacity = s.capacity(); |
952 | /// |
953 | /// let s = String::from_raw_parts(ptr, len, capacity); |
954 | /// |
955 | /// assert_eq!(String::from("hello" ), s); |
956 | /// } |
957 | /// ``` |
958 | #[inline ] |
959 | #[stable (feature = "rust1" , since = "1.0.0" )] |
960 | pub unsafe fn from_raw_parts(buf: *mut u8, length: usize, capacity: usize) -> String { |
961 | unsafe { String { vec: Vec::from_raw_parts(buf, length, capacity) } } |
962 | } |
963 | |
964 | /// Converts a vector of bytes to a `String` without checking that the |
965 | /// string contains valid UTF-8. |
966 | /// |
967 | /// See the safe version, [`from_utf8`], for more details. |
968 | /// |
969 | /// [`from_utf8`]: String::from_utf8 |
970 | /// |
971 | /// # Safety |
972 | /// |
973 | /// This function is unsafe because it does not check that the bytes passed |
974 | /// to it are valid UTF-8. If this constraint is violated, it may cause |
975 | /// memory unsafety issues with future users of the `String`, as the rest of |
976 | /// the standard library assumes that `String`s are valid UTF-8. |
977 | /// |
978 | /// # Examples |
979 | /// |
980 | /// ``` |
981 | /// // some bytes, in a vector |
982 | /// let sparkle_heart = vec![240, 159, 146, 150]; |
983 | /// |
984 | /// let sparkle_heart = unsafe { |
985 | /// String::from_utf8_unchecked(sparkle_heart) |
986 | /// }; |
987 | /// |
988 | /// assert_eq!("π" , sparkle_heart); |
989 | /// ``` |
990 | #[inline ] |
991 | #[must_use ] |
992 | #[stable (feature = "rust1" , since = "1.0.0" )] |
993 | pub unsafe fn from_utf8_unchecked(bytes: Vec<u8>) -> String { |
994 | String { vec: bytes } |
995 | } |
996 | |
997 | /// Converts a `String` into a byte vector. |
998 | /// |
999 | /// This consumes the `String`, so we do not need to copy its contents. |
1000 | /// |
1001 | /// # Examples |
1002 | /// |
1003 | /// ``` |
1004 | /// let s = String::from("hello" ); |
1005 | /// let bytes = s.into_bytes(); |
1006 | /// |
1007 | /// assert_eq!(&[104, 101, 108, 108, 111][..], &bytes[..]); |
1008 | /// ``` |
1009 | #[inline ] |
1010 | #[must_use = "`self` will be dropped if the result is not used" ] |
1011 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1012 | pub fn into_bytes(self) -> Vec<u8> { |
1013 | self.vec |
1014 | } |
1015 | |
1016 | /// Extracts a string slice containing the entire `String`. |
1017 | /// |
1018 | /// # Examples |
1019 | /// |
1020 | /// ``` |
1021 | /// let s = String::from("foo" ); |
1022 | /// |
1023 | /// assert_eq!("foo" , s.as_str()); |
1024 | /// ``` |
1025 | #[inline ] |
1026 | #[must_use ] |
1027 | #[stable (feature = "string_as_str" , since = "1.7.0" )] |
1028 | pub fn as_str(&self) -> &str { |
1029 | self |
1030 | } |
1031 | |
1032 | /// Converts a `String` into a mutable string slice. |
1033 | /// |
1034 | /// # Examples |
1035 | /// |
1036 | /// ``` |
1037 | /// let mut s = String::from("foobar" ); |
1038 | /// let s_mut_str = s.as_mut_str(); |
1039 | /// |
1040 | /// s_mut_str.make_ascii_uppercase(); |
1041 | /// |
1042 | /// assert_eq!("FOOBAR" , s_mut_str); |
1043 | /// ``` |
1044 | #[inline ] |
1045 | #[must_use ] |
1046 | #[stable (feature = "string_as_str" , since = "1.7.0" )] |
1047 | pub fn as_mut_str(&mut self) -> &mut str { |
1048 | self |
1049 | } |
1050 | |
1051 | /// Appends a given string slice onto the end of this `String`. |
1052 | /// |
1053 | /// # Examples |
1054 | /// |
1055 | /// ``` |
1056 | /// let mut s = String::from("foo" ); |
1057 | /// |
1058 | /// s.push_str("bar" ); |
1059 | /// |
1060 | /// assert_eq!("foobar" , s); |
1061 | /// ``` |
1062 | #[cfg (not(no_global_oom_handling))] |
1063 | #[inline ] |
1064 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1065 | #[rustc_confusables ("append" , "push" )] |
1066 | pub fn push_str(&mut self, string: &str) { |
1067 | self.vec.extend_from_slice(string.as_bytes()) |
1068 | } |
1069 | |
1070 | /// Copies elements from `src` range to the end of the string. |
1071 | /// |
1072 | /// # Panics |
1073 | /// |
1074 | /// Panics if the starting point or end point do not lie on a [`char`] |
1075 | /// boundary, or if they're out of bounds. |
1076 | /// |
1077 | /// # Examples |
1078 | /// |
1079 | /// ``` |
1080 | /// #![feature(string_extend_from_within)] |
1081 | /// let mut string = String::from("abcde" ); |
1082 | /// |
1083 | /// string.extend_from_within(2..); |
1084 | /// assert_eq!(string, "abcdecde" ); |
1085 | /// |
1086 | /// string.extend_from_within(..2); |
1087 | /// assert_eq!(string, "abcdecdeab" ); |
1088 | /// |
1089 | /// string.extend_from_within(4..8); |
1090 | /// assert_eq!(string, "abcdecdeabecde" ); |
1091 | /// ``` |
1092 | #[cfg (not(no_global_oom_handling))] |
1093 | #[unstable (feature = "string_extend_from_within" , issue = "103806" )] |
1094 | pub fn extend_from_within<R>(&mut self, src: R) |
1095 | where |
1096 | R: RangeBounds<usize>, |
1097 | { |
1098 | let src @ Range { start, end } = slice::range(src, ..self.len()); |
1099 | |
1100 | assert!(self.is_char_boundary(start)); |
1101 | assert!(self.is_char_boundary(end)); |
1102 | |
1103 | self.vec.extend_from_within(src); |
1104 | } |
1105 | |
1106 | /// Returns this `String`'s capacity, in bytes. |
1107 | /// |
1108 | /// # Examples |
1109 | /// |
1110 | /// ``` |
1111 | /// let s = String::with_capacity(10); |
1112 | /// |
1113 | /// assert!(s.capacity() >= 10); |
1114 | /// ``` |
1115 | #[inline ] |
1116 | #[must_use ] |
1117 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1118 | pub fn capacity(&self) -> usize { |
1119 | self.vec.capacity() |
1120 | } |
1121 | |
1122 | /// Reserves capacity for at least `additional` bytes more than the |
1123 | /// current length. The allocator may reserve more space to speculatively |
1124 | /// avoid frequent allocations. After calling `reserve`, |
1125 | /// capacity will be greater than or equal to `self.len() + additional`. |
1126 | /// Does nothing if capacity is already sufficient. |
1127 | /// |
1128 | /// # Panics |
1129 | /// |
1130 | /// Panics if the new capacity overflows [`usize`]. |
1131 | /// |
1132 | /// # Examples |
1133 | /// |
1134 | /// Basic usage: |
1135 | /// |
1136 | /// ``` |
1137 | /// let mut s = String::new(); |
1138 | /// |
1139 | /// s.reserve(10); |
1140 | /// |
1141 | /// assert!(s.capacity() >= 10); |
1142 | /// ``` |
1143 | /// |
1144 | /// This might not actually increase the capacity: |
1145 | /// |
1146 | /// ``` |
1147 | /// let mut s = String::with_capacity(10); |
1148 | /// s.push('a' ); |
1149 | /// s.push('b' ); |
1150 | /// |
1151 | /// // s now has a length of 2 and a capacity of at least 10 |
1152 | /// let capacity = s.capacity(); |
1153 | /// assert_eq!(2, s.len()); |
1154 | /// assert!(capacity >= 10); |
1155 | /// |
1156 | /// // Since we already have at least an extra 8 capacity, calling this... |
1157 | /// s.reserve(8); |
1158 | /// |
1159 | /// // ... doesn't actually increase. |
1160 | /// assert_eq!(capacity, s.capacity()); |
1161 | /// ``` |
1162 | #[cfg (not(no_global_oom_handling))] |
1163 | #[inline ] |
1164 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1165 | pub fn reserve(&mut self, additional: usize) { |
1166 | self.vec.reserve(additional) |
1167 | } |
1168 | |
1169 | /// Reserves the minimum capacity for at least `additional` bytes more than |
1170 | /// the current length. Unlike [`reserve`], this will not |
1171 | /// deliberately over-allocate to speculatively avoid frequent allocations. |
1172 | /// After calling `reserve_exact`, capacity will be greater than or equal to |
1173 | /// `self.len() + additional`. Does nothing if the capacity is already |
1174 | /// sufficient. |
1175 | /// |
1176 | /// [`reserve`]: String::reserve |
1177 | /// |
1178 | /// # Panics |
1179 | /// |
1180 | /// Panics if the new capacity overflows [`usize`]. |
1181 | /// |
1182 | /// # Examples |
1183 | /// |
1184 | /// Basic usage: |
1185 | /// |
1186 | /// ``` |
1187 | /// let mut s = String::new(); |
1188 | /// |
1189 | /// s.reserve_exact(10); |
1190 | /// |
1191 | /// assert!(s.capacity() >= 10); |
1192 | /// ``` |
1193 | /// |
1194 | /// This might not actually increase the capacity: |
1195 | /// |
1196 | /// ``` |
1197 | /// let mut s = String::with_capacity(10); |
1198 | /// s.push('a' ); |
1199 | /// s.push('b' ); |
1200 | /// |
1201 | /// // s now has a length of 2 and a capacity of at least 10 |
1202 | /// let capacity = s.capacity(); |
1203 | /// assert_eq!(2, s.len()); |
1204 | /// assert!(capacity >= 10); |
1205 | /// |
1206 | /// // Since we already have at least an extra 8 capacity, calling this... |
1207 | /// s.reserve_exact(8); |
1208 | /// |
1209 | /// // ... doesn't actually increase. |
1210 | /// assert_eq!(capacity, s.capacity()); |
1211 | /// ``` |
1212 | #[cfg (not(no_global_oom_handling))] |
1213 | #[inline ] |
1214 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1215 | pub fn reserve_exact(&mut self, additional: usize) { |
1216 | self.vec.reserve_exact(additional) |
1217 | } |
1218 | |
1219 | /// Tries to reserve capacity for at least `additional` bytes more than the |
1220 | /// current length. The allocator may reserve more space to speculatively |
1221 | /// avoid frequent allocations. After calling `try_reserve`, capacity will be |
1222 | /// greater than or equal to `self.len() + additional` if it returns |
1223 | /// `Ok(())`. Does nothing if capacity is already sufficient. This method |
1224 | /// preserves the contents even if an error occurs. |
1225 | /// |
1226 | /// # Errors |
1227 | /// |
1228 | /// If the capacity overflows, or the allocator reports a failure, then an error |
1229 | /// is returned. |
1230 | /// |
1231 | /// # Examples |
1232 | /// |
1233 | /// ``` |
1234 | /// use std::collections::TryReserveError; |
1235 | /// |
1236 | /// fn process_data(data: &str) -> Result<String, TryReserveError> { |
1237 | /// let mut output = String::new(); |
1238 | /// |
1239 | /// // Pre-reserve the memory, exiting if we can't |
1240 | /// output.try_reserve(data.len())?; |
1241 | /// |
1242 | /// // Now we know this can't OOM in the middle of our complex work |
1243 | /// output.push_str(data); |
1244 | /// |
1245 | /// Ok(output) |
1246 | /// } |
1247 | /// # process_data("rust" ).expect("why is the test harness OOMing on 4 bytes?" ); |
1248 | /// ``` |
1249 | #[stable (feature = "try_reserve" , since = "1.57.0" )] |
1250 | pub fn try_reserve(&mut self, additional: usize) -> Result<(), TryReserveError> { |
1251 | self.vec.try_reserve(additional) |
1252 | } |
1253 | |
1254 | /// Tries to reserve the minimum capacity for at least `additional` bytes |
1255 | /// more than the current length. Unlike [`try_reserve`], this will not |
1256 | /// deliberately over-allocate to speculatively avoid frequent allocations. |
1257 | /// After calling `try_reserve_exact`, capacity will be greater than or |
1258 | /// equal to `self.len() + additional` if it returns `Ok(())`. |
1259 | /// Does nothing if the capacity is already sufficient. |
1260 | /// |
1261 | /// Note that the allocator may give the collection more space than it |
1262 | /// requests. Therefore, capacity can not be relied upon to be precisely |
1263 | /// minimal. Prefer [`try_reserve`] if future insertions are expected. |
1264 | /// |
1265 | /// [`try_reserve`]: String::try_reserve |
1266 | /// |
1267 | /// # Errors |
1268 | /// |
1269 | /// If the capacity overflows, or the allocator reports a failure, then an error |
1270 | /// is returned. |
1271 | /// |
1272 | /// # Examples |
1273 | /// |
1274 | /// ``` |
1275 | /// use std::collections::TryReserveError; |
1276 | /// |
1277 | /// fn process_data(data: &str) -> Result<String, TryReserveError> { |
1278 | /// let mut output = String::new(); |
1279 | /// |
1280 | /// // Pre-reserve the memory, exiting if we can't |
1281 | /// output.try_reserve_exact(data.len())?; |
1282 | /// |
1283 | /// // Now we know this can't OOM in the middle of our complex work |
1284 | /// output.push_str(data); |
1285 | /// |
1286 | /// Ok(output) |
1287 | /// } |
1288 | /// # process_data("rust" ).expect("why is the test harness OOMing on 4 bytes?" ); |
1289 | /// ``` |
1290 | #[stable (feature = "try_reserve" , since = "1.57.0" )] |
1291 | pub fn try_reserve_exact(&mut self, additional: usize) -> Result<(), TryReserveError> { |
1292 | self.vec.try_reserve_exact(additional) |
1293 | } |
1294 | |
1295 | /// Shrinks the capacity of this `String` to match its length. |
1296 | /// |
1297 | /// # Examples |
1298 | /// |
1299 | /// ``` |
1300 | /// let mut s = String::from("foo" ); |
1301 | /// |
1302 | /// s.reserve(100); |
1303 | /// assert!(s.capacity() >= 100); |
1304 | /// |
1305 | /// s.shrink_to_fit(); |
1306 | /// assert_eq!(3, s.capacity()); |
1307 | /// ``` |
1308 | #[cfg (not(no_global_oom_handling))] |
1309 | #[inline ] |
1310 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1311 | pub fn shrink_to_fit(&mut self) { |
1312 | self.vec.shrink_to_fit() |
1313 | } |
1314 | |
1315 | /// Shrinks the capacity of this `String` with a lower bound. |
1316 | /// |
1317 | /// The capacity will remain at least as large as both the length |
1318 | /// and the supplied value. |
1319 | /// |
1320 | /// If the current capacity is less than the lower limit, this is a no-op. |
1321 | /// |
1322 | /// # Examples |
1323 | /// |
1324 | /// ``` |
1325 | /// let mut s = String::from("foo" ); |
1326 | /// |
1327 | /// s.reserve(100); |
1328 | /// assert!(s.capacity() >= 100); |
1329 | /// |
1330 | /// s.shrink_to(10); |
1331 | /// assert!(s.capacity() >= 10); |
1332 | /// s.shrink_to(0); |
1333 | /// assert!(s.capacity() >= 3); |
1334 | /// ``` |
1335 | #[cfg (not(no_global_oom_handling))] |
1336 | #[inline ] |
1337 | #[stable (feature = "shrink_to" , since = "1.56.0" )] |
1338 | pub fn shrink_to(&mut self, min_capacity: usize) { |
1339 | self.vec.shrink_to(min_capacity) |
1340 | } |
1341 | |
1342 | /// Appends the given [`char`] to the end of this `String`. |
1343 | /// |
1344 | /// # Examples |
1345 | /// |
1346 | /// ``` |
1347 | /// let mut s = String::from("abc" ); |
1348 | /// |
1349 | /// s.push('1' ); |
1350 | /// s.push('2' ); |
1351 | /// s.push('3' ); |
1352 | /// |
1353 | /// assert_eq!("abc123" , s); |
1354 | /// ``` |
1355 | #[cfg (not(no_global_oom_handling))] |
1356 | #[inline ] |
1357 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1358 | pub fn push(&mut self, ch: char) { |
1359 | match ch.len_utf8() { |
1360 | 1 => self.vec.push(ch as u8), |
1361 | _ => self.vec.extend_from_slice(ch.encode_utf8(&mut [0; 4]).as_bytes()), |
1362 | } |
1363 | } |
1364 | |
1365 | /// Returns a byte slice of this `String`'s contents. |
1366 | /// |
1367 | /// The inverse of this method is [`from_utf8`]. |
1368 | /// |
1369 | /// [`from_utf8`]: String::from_utf8 |
1370 | /// |
1371 | /// # Examples |
1372 | /// |
1373 | /// ``` |
1374 | /// let s = String::from("hello" ); |
1375 | /// |
1376 | /// assert_eq!(&[104, 101, 108, 108, 111], s.as_bytes()); |
1377 | /// ``` |
1378 | #[inline ] |
1379 | #[must_use ] |
1380 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1381 | pub fn as_bytes(&self) -> &[u8] { |
1382 | &self.vec |
1383 | } |
1384 | |
1385 | /// Shortens this `String` to the specified length. |
1386 | /// |
1387 | /// If `new_len` is greater than the string's current length, this has no |
1388 | /// effect. |
1389 | /// |
1390 | /// Note that this method has no effect on the allocated capacity |
1391 | /// of the string |
1392 | /// |
1393 | /// # Panics |
1394 | /// |
1395 | /// Panics if `new_len` does not lie on a [`char`] boundary. |
1396 | /// |
1397 | /// # Examples |
1398 | /// |
1399 | /// ``` |
1400 | /// let mut s = String::from("hello" ); |
1401 | /// |
1402 | /// s.truncate(2); |
1403 | /// |
1404 | /// assert_eq!("he" , s); |
1405 | /// ``` |
1406 | #[inline ] |
1407 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1408 | pub fn truncate(&mut self, new_len: usize) { |
1409 | if new_len <= self.len() { |
1410 | assert!(self.is_char_boundary(new_len)); |
1411 | self.vec.truncate(new_len) |
1412 | } |
1413 | } |
1414 | |
1415 | /// Removes the last character from the string buffer and returns it. |
1416 | /// |
1417 | /// Returns [`None`] if this `String` is empty. |
1418 | /// |
1419 | /// # Examples |
1420 | /// |
1421 | /// ``` |
1422 | /// let mut s = String::from("abΔ" ); |
1423 | /// |
1424 | /// assert_eq!(s.pop(), Some('Δ' )); |
1425 | /// assert_eq!(s.pop(), Some('b' )); |
1426 | /// assert_eq!(s.pop(), Some('a' )); |
1427 | /// |
1428 | /// assert_eq!(s.pop(), None); |
1429 | /// ``` |
1430 | #[inline ] |
1431 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1432 | pub fn pop(&mut self) -> Option<char> { |
1433 | let ch = self.chars().rev().next()?; |
1434 | let newlen = self.len() - ch.len_utf8(); |
1435 | unsafe { |
1436 | self.vec.set_len(newlen); |
1437 | } |
1438 | Some(ch) |
1439 | } |
1440 | |
1441 | /// Removes a [`char`] from this `String` at a byte position and returns it. |
1442 | /// |
1443 | /// This is an *O*(*n*) operation, as it requires copying every element in the |
1444 | /// buffer. |
1445 | /// |
1446 | /// # Panics |
1447 | /// |
1448 | /// Panics if `idx` is larger than or equal to the `String`'s length, |
1449 | /// or if it does not lie on a [`char`] boundary. |
1450 | /// |
1451 | /// # Examples |
1452 | /// |
1453 | /// ``` |
1454 | /// let mut s = String::from("abç" ); |
1455 | /// |
1456 | /// assert_eq!(s.remove(0), 'a' ); |
1457 | /// assert_eq!(s.remove(1), 'Γ§' ); |
1458 | /// assert_eq!(s.remove(0), 'b' ); |
1459 | /// ``` |
1460 | #[inline ] |
1461 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1462 | #[rustc_confusables ("delete" , "take" )] |
1463 | pub fn remove(&mut self, idx: usize) -> char { |
1464 | let ch = match self[idx..].chars().next() { |
1465 | Some(ch) => ch, |
1466 | None => panic!("cannot remove a char from the end of a string" ), |
1467 | }; |
1468 | |
1469 | let next = idx + ch.len_utf8(); |
1470 | let len = self.len(); |
1471 | unsafe { |
1472 | ptr::copy(self.vec.as_ptr().add(next), self.vec.as_mut_ptr().add(idx), len - next); |
1473 | self.vec.set_len(len - (next - idx)); |
1474 | } |
1475 | ch |
1476 | } |
1477 | |
1478 | /// Remove all matches of pattern `pat` in the `String`. |
1479 | /// |
1480 | /// # Examples |
1481 | /// |
1482 | /// ``` |
1483 | /// #![feature(string_remove_matches)] |
1484 | /// let mut s = String::from("Trees are not green, the sky is not blue." ); |
1485 | /// s.remove_matches("not " ); |
1486 | /// assert_eq!("Trees are green, the sky is blue." , s); |
1487 | /// ``` |
1488 | /// |
1489 | /// Matches will be detected and removed iteratively, so in cases where |
1490 | /// patterns overlap, only the first pattern will be removed: |
1491 | /// |
1492 | /// ``` |
1493 | /// #![feature(string_remove_matches)] |
1494 | /// let mut s = String::from("banana" ); |
1495 | /// s.remove_matches("ana" ); |
1496 | /// assert_eq!("bna" , s); |
1497 | /// ``` |
1498 | #[cfg (not(no_global_oom_handling))] |
1499 | #[unstable (feature = "string_remove_matches" , reason = "new API" , issue = "72826" )] |
1500 | pub fn remove_matches<'a, P>(&'a mut self, pat: P) |
1501 | where |
1502 | P: for<'x> Pattern<'x>, |
1503 | { |
1504 | use core::str::pattern::Searcher; |
1505 | |
1506 | let rejections = { |
1507 | let mut searcher = pat.into_searcher(self); |
1508 | // Per Searcher::next: |
1509 | // |
1510 | // A Match result needs to contain the whole matched pattern, |
1511 | // however Reject results may be split up into arbitrary many |
1512 | // adjacent fragments. Both ranges may have zero length. |
1513 | // |
1514 | // In practice the implementation of Searcher::next_match tends to |
1515 | // be more efficient, so we use it here and do some work to invert |
1516 | // matches into rejections since that's what we want to copy below. |
1517 | let mut front = 0; |
1518 | let rejections: Vec<_> = from_fn(|| { |
1519 | let (start, end) = searcher.next_match()?; |
1520 | let prev_front = front; |
1521 | front = end; |
1522 | Some((prev_front, start)) |
1523 | }) |
1524 | .collect(); |
1525 | rejections.into_iter().chain(core::iter::once((front, self.len()))) |
1526 | }; |
1527 | |
1528 | let mut len = 0; |
1529 | let ptr = self.vec.as_mut_ptr(); |
1530 | |
1531 | for (start, end) in rejections { |
1532 | let count = end - start; |
1533 | if start != len { |
1534 | // SAFETY: per Searcher::next: |
1535 | // |
1536 | // The stream of Match and Reject values up to a Done will |
1537 | // contain index ranges that are adjacent, non-overlapping, |
1538 | // covering the whole haystack, and laying on utf8 |
1539 | // boundaries. |
1540 | unsafe { |
1541 | ptr::copy(ptr.add(start), ptr.add(len), count); |
1542 | } |
1543 | } |
1544 | len += count; |
1545 | } |
1546 | |
1547 | unsafe { |
1548 | self.vec.set_len(len); |
1549 | } |
1550 | } |
1551 | |
1552 | /// Retains only the characters specified by the predicate. |
1553 | /// |
1554 | /// In other words, remove all characters `c` such that `f(c)` returns `false`. |
1555 | /// This method operates in place, visiting each character exactly once in the |
1556 | /// original order, and preserves the order of the retained characters. |
1557 | /// |
1558 | /// # Examples |
1559 | /// |
1560 | /// ``` |
1561 | /// let mut s = String::from("f_o_ob_ar" ); |
1562 | /// |
1563 | /// s.retain(|c| c != '_' ); |
1564 | /// |
1565 | /// assert_eq!(s, "foobar" ); |
1566 | /// ``` |
1567 | /// |
1568 | /// Because the elements are visited exactly once in the original order, |
1569 | /// external state may be used to decide which elements to keep. |
1570 | /// |
1571 | /// ``` |
1572 | /// let mut s = String::from("abcde" ); |
1573 | /// let keep = [false, true, true, false, true]; |
1574 | /// let mut iter = keep.iter(); |
1575 | /// s.retain(|_| *iter.next().unwrap()); |
1576 | /// assert_eq!(s, "bce" ); |
1577 | /// ``` |
1578 | #[inline ] |
1579 | #[stable (feature = "string_retain" , since = "1.26.0" )] |
1580 | pub fn retain<F>(&mut self, mut f: F) |
1581 | where |
1582 | F: FnMut(char) -> bool, |
1583 | { |
1584 | struct SetLenOnDrop<'a> { |
1585 | s: &'a mut String, |
1586 | idx: usize, |
1587 | del_bytes: usize, |
1588 | } |
1589 | |
1590 | impl<'a> Drop for SetLenOnDrop<'a> { |
1591 | fn drop(&mut self) { |
1592 | let new_len = self.idx - self.del_bytes; |
1593 | debug_assert!(new_len <= self.s.len()); |
1594 | unsafe { self.s.vec.set_len(new_len) }; |
1595 | } |
1596 | } |
1597 | |
1598 | let len = self.len(); |
1599 | let mut guard = SetLenOnDrop { s: self, idx: 0, del_bytes: 0 }; |
1600 | |
1601 | while guard.idx < len { |
1602 | let ch = |
1603 | // SAFETY: `guard.idx` is positive-or-zero and less that len so the `get_unchecked` |
1604 | // is in bound. `self` is valid UTF-8 like string and the returned slice starts at |
1605 | // a unicode code point so the `Chars` always return one character. |
1606 | unsafe { guard.s.get_unchecked(guard.idx..len).chars().next().unwrap_unchecked() }; |
1607 | let ch_len = ch.len_utf8(); |
1608 | |
1609 | if !f(ch) { |
1610 | guard.del_bytes += ch_len; |
1611 | } else if guard.del_bytes > 0 { |
1612 | // SAFETY: `guard.idx` is in bound and `guard.del_bytes` represent the number of |
1613 | // bytes that are erased from the string so the resulting `guard.idx - |
1614 | // guard.del_bytes` always represent a valid unicode code point. |
1615 | // |
1616 | // `guard.del_bytes` >= `ch.len_utf8()`, so taking a slice with `ch.len_utf8()` len |
1617 | // is safe. |
1618 | ch.encode_utf8(unsafe { |
1619 | crate::slice::from_raw_parts_mut( |
1620 | guard.s.as_mut_ptr().add(guard.idx - guard.del_bytes), |
1621 | ch.len_utf8(), |
1622 | ) |
1623 | }); |
1624 | } |
1625 | |
1626 | // Point idx to the next char |
1627 | guard.idx += ch_len; |
1628 | } |
1629 | |
1630 | drop(guard); |
1631 | } |
1632 | |
1633 | /// Inserts a character into this `String` at a byte position. |
1634 | /// |
1635 | /// This is an *O*(*n*) operation as it requires copying every element in the |
1636 | /// buffer. |
1637 | /// |
1638 | /// # Panics |
1639 | /// |
1640 | /// Panics if `idx` is larger than the `String`'s length, or if it does not |
1641 | /// lie on a [`char`] boundary. |
1642 | /// |
1643 | /// # Examples |
1644 | /// |
1645 | /// ``` |
1646 | /// let mut s = String::with_capacity(3); |
1647 | /// |
1648 | /// s.insert(0, 'f' ); |
1649 | /// s.insert(1, 'o' ); |
1650 | /// s.insert(2, 'o' ); |
1651 | /// |
1652 | /// assert_eq!("foo" , s); |
1653 | /// ``` |
1654 | #[cfg (not(no_global_oom_handling))] |
1655 | #[inline ] |
1656 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1657 | #[rustc_confusables ("set" )] |
1658 | pub fn insert(&mut self, idx: usize, ch: char) { |
1659 | assert!(self.is_char_boundary(idx)); |
1660 | let mut bits = [0; 4]; |
1661 | let bits = ch.encode_utf8(&mut bits).as_bytes(); |
1662 | |
1663 | unsafe { |
1664 | self.insert_bytes(idx, bits); |
1665 | } |
1666 | } |
1667 | |
1668 | #[cfg (not(no_global_oom_handling))] |
1669 | unsafe fn insert_bytes(&mut self, idx: usize, bytes: &[u8]) { |
1670 | let len = self.len(); |
1671 | let amt = bytes.len(); |
1672 | self.vec.reserve(amt); |
1673 | |
1674 | unsafe { |
1675 | ptr::copy(self.vec.as_ptr().add(idx), self.vec.as_mut_ptr().add(idx + amt), len - idx); |
1676 | ptr::copy_nonoverlapping(bytes.as_ptr(), self.vec.as_mut_ptr().add(idx), amt); |
1677 | self.vec.set_len(len + amt); |
1678 | } |
1679 | } |
1680 | |
1681 | /// Inserts a string slice into this `String` at a byte position. |
1682 | /// |
1683 | /// This is an *O*(*n*) operation as it requires copying every element in the |
1684 | /// buffer. |
1685 | /// |
1686 | /// # Panics |
1687 | /// |
1688 | /// Panics if `idx` is larger than the `String`'s length, or if it does not |
1689 | /// lie on a [`char`] boundary. |
1690 | /// |
1691 | /// # Examples |
1692 | /// |
1693 | /// ``` |
1694 | /// let mut s = String::from("bar" ); |
1695 | /// |
1696 | /// s.insert_str(0, "foo" ); |
1697 | /// |
1698 | /// assert_eq!("foobar" , s); |
1699 | /// ``` |
1700 | #[cfg (not(no_global_oom_handling))] |
1701 | #[inline ] |
1702 | #[stable (feature = "insert_str" , since = "1.16.0" )] |
1703 | pub fn insert_str(&mut self, idx: usize, string: &str) { |
1704 | assert!(self.is_char_boundary(idx)); |
1705 | |
1706 | unsafe { |
1707 | self.insert_bytes(idx, string.as_bytes()); |
1708 | } |
1709 | } |
1710 | |
1711 | /// Returns a mutable reference to the contents of this `String`. |
1712 | /// |
1713 | /// # Safety |
1714 | /// |
1715 | /// This function is unsafe because the returned `&mut Vec` allows writing |
1716 | /// bytes which are not valid UTF-8. If this constraint is violated, using |
1717 | /// the original `String` after dropping the `&mut Vec` may violate memory |
1718 | /// safety, as the rest of the standard library assumes that `String`s are |
1719 | /// valid UTF-8. |
1720 | /// |
1721 | /// # Examples |
1722 | /// |
1723 | /// ``` |
1724 | /// let mut s = String::from("hello" ); |
1725 | /// |
1726 | /// unsafe { |
1727 | /// let vec = s.as_mut_vec(); |
1728 | /// assert_eq!(&[104, 101, 108, 108, 111][..], &vec[..]); |
1729 | /// |
1730 | /// vec.reverse(); |
1731 | /// } |
1732 | /// assert_eq!(s, "olleh" ); |
1733 | /// ``` |
1734 | #[inline ] |
1735 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1736 | pub unsafe fn as_mut_vec(&mut self) -> &mut Vec<u8> { |
1737 | &mut self.vec |
1738 | } |
1739 | |
1740 | /// Returns the length of this `String`, in bytes, not [`char`]s or |
1741 | /// graphemes. In other words, it might not be what a human considers the |
1742 | /// length of the string. |
1743 | /// |
1744 | /// # Examples |
1745 | /// |
1746 | /// ``` |
1747 | /// let a = String::from("foo" ); |
1748 | /// assert_eq!(a.len(), 3); |
1749 | /// |
1750 | /// let fancy_f = String::from("Ζoo" ); |
1751 | /// assert_eq!(fancy_f.len(), 4); |
1752 | /// assert_eq!(fancy_f.chars().count(), 3); |
1753 | /// ``` |
1754 | #[inline ] |
1755 | #[must_use ] |
1756 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1757 | #[rustc_confusables ("length" , "size" )] |
1758 | pub fn len(&self) -> usize { |
1759 | self.vec.len() |
1760 | } |
1761 | |
1762 | /// Returns `true` if this `String` has a length of zero, and `false` otherwise. |
1763 | /// |
1764 | /// # Examples |
1765 | /// |
1766 | /// ``` |
1767 | /// let mut v = String::new(); |
1768 | /// assert!(v.is_empty()); |
1769 | /// |
1770 | /// v.push('a' ); |
1771 | /// assert!(!v.is_empty()); |
1772 | /// ``` |
1773 | #[inline ] |
1774 | #[must_use ] |
1775 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1776 | pub fn is_empty(&self) -> bool { |
1777 | self.len() == 0 |
1778 | } |
1779 | |
1780 | /// Splits the string into two at the given byte index. |
1781 | /// |
1782 | /// Returns a newly allocated `String`. `self` contains bytes `[0, at)`, and |
1783 | /// the returned `String` contains bytes `[at, len)`. `at` must be on the |
1784 | /// boundary of a UTF-8 code point. |
1785 | /// |
1786 | /// Note that the capacity of `self` does not change. |
1787 | /// |
1788 | /// # Panics |
1789 | /// |
1790 | /// Panics if `at` is not on a `UTF-8` code point boundary, or if it is beyond the last |
1791 | /// code point of the string. |
1792 | /// |
1793 | /// # Examples |
1794 | /// |
1795 | /// ``` |
1796 | /// # fn main() { |
1797 | /// let mut hello = String::from("Hello, World!" ); |
1798 | /// let world = hello.split_off(7); |
1799 | /// assert_eq!(hello, "Hello, " ); |
1800 | /// assert_eq!(world, "World!" ); |
1801 | /// # } |
1802 | /// ``` |
1803 | #[cfg (not(no_global_oom_handling))] |
1804 | #[inline ] |
1805 | #[stable (feature = "string_split_off" , since = "1.16.0" )] |
1806 | #[must_use = "use `.truncate()` if you don't need the other half" ] |
1807 | pub fn split_off(&mut self, at: usize) -> String { |
1808 | assert!(self.is_char_boundary(at)); |
1809 | let other = self.vec.split_off(at); |
1810 | unsafe { String::from_utf8_unchecked(other) } |
1811 | } |
1812 | |
1813 | /// Truncates this `String`, removing all contents. |
1814 | /// |
1815 | /// While this means the `String` will have a length of zero, it does not |
1816 | /// touch its capacity. |
1817 | /// |
1818 | /// # Examples |
1819 | /// |
1820 | /// ``` |
1821 | /// let mut s = String::from("foo" ); |
1822 | /// |
1823 | /// s.clear(); |
1824 | /// |
1825 | /// assert!(s.is_empty()); |
1826 | /// assert_eq!(0, s.len()); |
1827 | /// assert_eq!(3, s.capacity()); |
1828 | /// ``` |
1829 | #[inline ] |
1830 | #[stable (feature = "rust1" , since = "1.0.0" )] |
1831 | pub fn clear(&mut self) { |
1832 | self.vec.clear() |
1833 | } |
1834 | |
1835 | /// Removes the specified range from the string in bulk, returning all |
1836 | /// removed characters as an iterator. |
1837 | /// |
1838 | /// The returned iterator keeps a mutable borrow on the string to optimize |
1839 | /// its implementation. |
1840 | /// |
1841 | /// # Panics |
1842 | /// |
1843 | /// Panics if the starting point or end point do not lie on a [`char`] |
1844 | /// boundary, or if they're out of bounds. |
1845 | /// |
1846 | /// # Leaking |
1847 | /// |
1848 | /// If the returned iterator goes out of scope without being dropped (due to |
1849 | /// [`core::mem::forget`], for example), the string may still contain a copy |
1850 | /// of any drained characters, or may have lost characters arbitrarily, |
1851 | /// including characters outside the range. |
1852 | /// |
1853 | /// # Examples |
1854 | /// |
1855 | /// ``` |
1856 | /// let mut s = String::from("Ξ± is alpha, Ξ² is beta" ); |
1857 | /// let beta_offset = s.find('Ξ²' ).unwrap_or(s.len()); |
1858 | /// |
1859 | /// // Remove the range up until the Ξ² from the string |
1860 | /// let t: String = s.drain(..beta_offset).collect(); |
1861 | /// assert_eq!(t, "Ξ± is alpha, " ); |
1862 | /// assert_eq!(s, "Ξ² is beta" ); |
1863 | /// |
1864 | /// // A full range clears the string, like `clear()` does |
1865 | /// s.drain(..); |
1866 | /// assert_eq!(s, "" ); |
1867 | /// ``` |
1868 | #[stable (feature = "drain" , since = "1.6.0" )] |
1869 | pub fn drain<R>(&mut self, range: R) -> Drain<'_> |
1870 | where |
1871 | R: RangeBounds<usize>, |
1872 | { |
1873 | // Memory safety |
1874 | // |
1875 | // The String version of Drain does not have the memory safety issues |
1876 | // of the vector version. The data is just plain bytes. |
1877 | // Because the range removal happens in Drop, if the Drain iterator is leaked, |
1878 | // the removal will not happen. |
1879 | let Range { start, end } = slice::range(range, ..self.len()); |
1880 | assert!(self.is_char_boundary(start)); |
1881 | assert!(self.is_char_boundary(end)); |
1882 | |
1883 | // Take out two simultaneous borrows. The &mut String won't be accessed |
1884 | // until iteration is over, in Drop. |
1885 | let self_ptr = self as *mut _; |
1886 | // SAFETY: `slice::range` and `is_char_boundary` do the appropriate bounds checks. |
1887 | let chars_iter = unsafe { self.get_unchecked(start..end) }.chars(); |
1888 | |
1889 | Drain { start, end, iter: chars_iter, string: self_ptr } |
1890 | } |
1891 | |
1892 | /// Removes the specified range in the string, |
1893 | /// and replaces it with the given string. |
1894 | /// The given string doesn't need to be the same length as the range. |
1895 | /// |
1896 | /// # Panics |
1897 | /// |
1898 | /// Panics if the starting point or end point do not lie on a [`char`] |
1899 | /// boundary, or if they're out of bounds. |
1900 | /// |
1901 | /// # Examples |
1902 | /// |
1903 | /// ``` |
1904 | /// let mut s = String::from("Ξ± is alpha, Ξ² is beta" ); |
1905 | /// let beta_offset = s.find('Ξ²' ).unwrap_or(s.len()); |
1906 | /// |
1907 | /// // Replace the range up until the Ξ² from the string |
1908 | /// s.replace_range(..beta_offset, "Ξ is capital alpha; " ); |
1909 | /// assert_eq!(s, "Ξ is capital alpha; Ξ² is beta" ); |
1910 | /// ``` |
1911 | #[cfg (not(no_global_oom_handling))] |
1912 | #[stable (feature = "splice" , since = "1.27.0" )] |
1913 | pub fn replace_range<R>(&mut self, range: R, replace_with: &str) |
1914 | where |
1915 | R: RangeBounds<usize>, |
1916 | { |
1917 | // Memory safety |
1918 | // |
1919 | // Replace_range does not have the memory safety issues of a vector Splice. |
1920 | // of the vector version. The data is just plain bytes. |
1921 | |
1922 | // WARNING: Inlining this variable would be unsound (#81138) |
1923 | let start = range.start_bound(); |
1924 | match start { |
1925 | Included(&n) => assert!(self.is_char_boundary(n)), |
1926 | Excluded(&n) => assert!(self.is_char_boundary(n + 1)), |
1927 | Unbounded => {} |
1928 | }; |
1929 | // WARNING: Inlining this variable would be unsound (#81138) |
1930 | let end = range.end_bound(); |
1931 | match end { |
1932 | Included(&n) => assert!(self.is_char_boundary(n + 1)), |
1933 | Excluded(&n) => assert!(self.is_char_boundary(n)), |
1934 | Unbounded => {} |
1935 | }; |
1936 | |
1937 | // Using `range` again would be unsound (#81138) |
1938 | // We assume the bounds reported by `range` remain the same, but |
1939 | // an adversarial implementation could change between calls |
1940 | unsafe { self.as_mut_vec() }.splice((start, end), replace_with.bytes()); |
1941 | } |
1942 | |
1943 | /// Converts this `String` into a <code>[Box]<[str]></code>. |
1944 | /// |
1945 | /// This will drop any excess capacity. |
1946 | /// |
1947 | /// [str]: prim@str "str" |
1948 | /// |
1949 | /// # Examples |
1950 | /// |
1951 | /// ``` |
1952 | /// let s = String::from("hello" ); |
1953 | /// |
1954 | /// let b = s.into_boxed_str(); |
1955 | /// ``` |
1956 | #[cfg (not(no_global_oom_handling))] |
1957 | #[stable (feature = "box_str" , since = "1.4.0" )] |
1958 | #[must_use = "`self` will be dropped if the result is not used" ] |
1959 | #[inline ] |
1960 | pub fn into_boxed_str(self) -> Box<str> { |
1961 | let slice = self.vec.into_boxed_slice(); |
1962 | unsafe { from_boxed_utf8_unchecked(slice) } |
1963 | } |
1964 | |
1965 | /// Consumes and leaks the `String`, returning a mutable reference to the contents, |
1966 | /// `&'a mut str`. |
1967 | /// |
1968 | /// The caller has free choice over the returned lifetime, including `'static`. Indeed, |
1969 | /// this function is ideally used for data that lives for the remainder of the program's life, |
1970 | /// as dropping the returned reference will cause a memory leak. |
1971 | /// |
1972 | /// It does not reallocate or shrink the `String`, |
1973 | /// so the leaked allocation may include unused capacity that is not part |
1974 | /// of the returned slice. If you don't want that, call [`into_boxed_str`], |
1975 | /// and then [`Box::leak`]. |
1976 | /// |
1977 | /// [`into_boxed_str`]: Self::into_boxed_str |
1978 | /// |
1979 | /// # Examples |
1980 | /// |
1981 | /// ``` |
1982 | /// let x = String::from("bucket" ); |
1983 | /// let static_ref: &'static mut str = x.leak(); |
1984 | /// assert_eq!(static_ref, "bucket" ); |
1985 | /// ``` |
1986 | #[stable (feature = "string_leak" , since = "1.72.0" )] |
1987 | #[inline ] |
1988 | pub fn leak<'a>(self) -> &'a mut str { |
1989 | let slice = self.vec.leak(); |
1990 | unsafe { from_utf8_unchecked_mut(slice) } |
1991 | } |
1992 | } |
1993 | |
1994 | impl FromUtf8Error { |
1995 | /// Returns a slice of [`u8`]s bytes that were attempted to convert to a `String`. |
1996 | /// |
1997 | /// # Examples |
1998 | /// |
1999 | /// ``` |
2000 | /// // some invalid bytes, in a vector |
2001 | /// let bytes = vec![0, 159]; |
2002 | /// |
2003 | /// let value = String::from_utf8(bytes); |
2004 | /// |
2005 | /// assert_eq!(&[0, 159], value.unwrap_err().as_bytes()); |
2006 | /// ``` |
2007 | #[must_use ] |
2008 | #[stable (feature = "from_utf8_error_as_bytes" , since = "1.26.0" )] |
2009 | pub fn as_bytes(&self) -> &[u8] { |
2010 | &self.bytes[..] |
2011 | } |
2012 | |
2013 | /// Returns the bytes that were attempted to convert to a `String`. |
2014 | /// |
2015 | /// This method is carefully constructed to avoid allocation. It will |
2016 | /// consume the error, moving out the bytes, so that a copy of the bytes |
2017 | /// does not need to be made. |
2018 | /// |
2019 | /// # Examples |
2020 | /// |
2021 | /// ``` |
2022 | /// // some invalid bytes, in a vector |
2023 | /// let bytes = vec![0, 159]; |
2024 | /// |
2025 | /// let value = String::from_utf8(bytes); |
2026 | /// |
2027 | /// assert_eq!(vec![0, 159], value.unwrap_err().into_bytes()); |
2028 | /// ``` |
2029 | #[must_use = "`self` will be dropped if the result is not used" ] |
2030 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2031 | pub fn into_bytes(self) -> Vec<u8> { |
2032 | self.bytes |
2033 | } |
2034 | |
2035 | /// Fetch a `Utf8Error` to get more details about the conversion failure. |
2036 | /// |
2037 | /// The [`Utf8Error`] type provided by [`std::str`] represents an error that may |
2038 | /// occur when converting a slice of [`u8`]s to a [`&str`]. In this sense, it's |
2039 | /// an analogue to `FromUtf8Error`. See its documentation for more details |
2040 | /// on using it. |
2041 | /// |
2042 | /// [`std::str`]: core::str "std::str" |
2043 | /// [`&str`]: prim@str "&str" |
2044 | /// |
2045 | /// # Examples |
2046 | /// |
2047 | /// ``` |
2048 | /// // some invalid bytes, in a vector |
2049 | /// let bytes = vec![0, 159]; |
2050 | /// |
2051 | /// let error = String::from_utf8(bytes).unwrap_err().utf8_error(); |
2052 | /// |
2053 | /// // the first byte is invalid here |
2054 | /// assert_eq!(1, error.valid_up_to()); |
2055 | /// ``` |
2056 | #[must_use ] |
2057 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2058 | pub fn utf8_error(&self) -> Utf8Error { |
2059 | self.error |
2060 | } |
2061 | } |
2062 | |
2063 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2064 | impl fmt::Display for FromUtf8Error { |
2065 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
2066 | fmt::Display::fmt(&self.error, f) |
2067 | } |
2068 | } |
2069 | |
2070 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2071 | impl fmt::Display for FromUtf16Error { |
2072 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
2073 | fmt::Display::fmt(self:"invalid utf-16: lone surrogate found" , f) |
2074 | } |
2075 | } |
2076 | |
2077 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2078 | impl Error for FromUtf8Error { |
2079 | #[allow (deprecated)] |
2080 | fn description(&self) -> &str { |
2081 | "invalid utf-8" |
2082 | } |
2083 | } |
2084 | |
2085 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2086 | impl Error for FromUtf16Error { |
2087 | #[allow (deprecated)] |
2088 | fn description(&self) -> &str { |
2089 | "invalid utf-16" |
2090 | } |
2091 | } |
2092 | |
2093 | #[cfg (not(no_global_oom_handling))] |
2094 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2095 | impl Clone for String { |
2096 | fn clone(&self) -> Self { |
2097 | String { vec: self.vec.clone() } |
2098 | } |
2099 | |
2100 | /// Clones the contents of `source` into `self`. |
2101 | /// |
2102 | /// This method is preferred over simply assigning `source.clone()` to `self`, |
2103 | /// as it avoids reallocation if possible. |
2104 | fn clone_from(&mut self, source: &Self) { |
2105 | self.vec.clone_from(&source.vec); |
2106 | } |
2107 | } |
2108 | |
2109 | #[cfg (not(no_global_oom_handling))] |
2110 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2111 | impl FromIterator<char> for String { |
2112 | fn from_iter<I: IntoIterator<Item = char>>(iter: I) -> String { |
2113 | let mut buf: String = String::new(); |
2114 | buf.extend(iter); |
2115 | buf |
2116 | } |
2117 | } |
2118 | |
2119 | #[cfg (not(no_global_oom_handling))] |
2120 | #[stable (feature = "string_from_iter_by_ref" , since = "1.17.0" )] |
2121 | impl<'a> FromIterator<&'a char> for String { |
2122 | fn from_iter<I: IntoIterator<Item = &'a char>>(iter: I) -> String { |
2123 | let mut buf: String = String::new(); |
2124 | buf.extend(iter); |
2125 | buf |
2126 | } |
2127 | } |
2128 | |
2129 | #[cfg (not(no_global_oom_handling))] |
2130 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2131 | impl<'a> FromIterator<&'a str> for String { |
2132 | fn from_iter<I: IntoIterator<Item = &'a str>>(iter: I) -> String { |
2133 | let mut buf: String = String::new(); |
2134 | buf.extend(iter); |
2135 | buf |
2136 | } |
2137 | } |
2138 | |
2139 | #[cfg (not(no_global_oom_handling))] |
2140 | #[stable (feature = "extend_string" , since = "1.4.0" )] |
2141 | impl FromIterator<String> for String { |
2142 | fn from_iter<I: IntoIterator<Item = String>>(iter: I) -> String { |
2143 | let mut iterator: ::IntoIter = iter.into_iter(); |
2144 | |
2145 | // Because we're iterating over `String`s, we can avoid at least |
2146 | // one allocation by getting the first string from the iterator |
2147 | // and appending to it all the subsequent strings. |
2148 | match iterator.next() { |
2149 | None => String::new(), |
2150 | Some(mut buf: String) => { |
2151 | buf.extend(iter:iterator); |
2152 | buf |
2153 | } |
2154 | } |
2155 | } |
2156 | } |
2157 | |
2158 | #[cfg (not(no_global_oom_handling))] |
2159 | #[stable (feature = "box_str2" , since = "1.45.0" )] |
2160 | impl FromIterator<Box<str>> for String { |
2161 | fn from_iter<I: IntoIterator<Item = Box<str>>>(iter: I) -> String { |
2162 | let mut buf: String = String::new(); |
2163 | buf.extend(iter); |
2164 | buf |
2165 | } |
2166 | } |
2167 | |
2168 | #[cfg (not(no_global_oom_handling))] |
2169 | #[stable (feature = "herd_cows" , since = "1.19.0" )] |
2170 | impl<'a> FromIterator<Cow<'a, str>> for String { |
2171 | fn from_iter<I: IntoIterator<Item = Cow<'a, str>>>(iter: I) -> String { |
2172 | let mut iterator: ::IntoIter = iter.into_iter(); |
2173 | |
2174 | // Because we're iterating over CoWs, we can (potentially) avoid at least |
2175 | // one allocation by getting the first item and appending to it all the |
2176 | // subsequent items. |
2177 | match iterator.next() { |
2178 | None => String::new(), |
2179 | Some(cow: Cow<'_, str>) => { |
2180 | let mut buf: String = cow.into_owned(); |
2181 | buf.extend(iter:iterator); |
2182 | buf |
2183 | } |
2184 | } |
2185 | } |
2186 | } |
2187 | |
2188 | #[cfg (not(no_global_oom_handling))] |
2189 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2190 | impl Extend<char> for String { |
2191 | fn extend<I: IntoIterator<Item = char>>(&mut self, iter: I) { |
2192 | let iterator: ::IntoIter = iter.into_iter(); |
2193 | let (lower_bound: usize, _) = iterator.size_hint(); |
2194 | self.reserve(additional:lower_bound); |
2195 | iterator.for_each(move |c: char| self.push(ch:c)); |
2196 | } |
2197 | |
2198 | #[inline ] |
2199 | fn extend_one(&mut self, c: char) { |
2200 | self.push(ch:c); |
2201 | } |
2202 | |
2203 | #[inline ] |
2204 | fn extend_reserve(&mut self, additional: usize) { |
2205 | self.reserve(additional); |
2206 | } |
2207 | } |
2208 | |
2209 | #[cfg (not(no_global_oom_handling))] |
2210 | #[stable (feature = "extend_ref" , since = "1.2.0" )] |
2211 | impl<'a> Extend<&'a char> for String { |
2212 | fn extend<I: IntoIterator<Item = &'a char>>(&mut self, iter: I) { |
2213 | self.extend(iter:iter.into_iter().cloned()); |
2214 | } |
2215 | |
2216 | #[inline ] |
2217 | fn extend_one(&mut self, &c: char: &'a char) { |
2218 | self.push(ch:c); |
2219 | } |
2220 | |
2221 | #[inline ] |
2222 | fn extend_reserve(&mut self, additional: usize) { |
2223 | self.reserve(additional); |
2224 | } |
2225 | } |
2226 | |
2227 | #[cfg (not(no_global_oom_handling))] |
2228 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2229 | impl<'a> Extend<&'a str> for String { |
2230 | fn extend<I: IntoIterator<Item = &'a str>>(&mut self, iter: I) { |
2231 | iter.into_iter().for_each(move |s: &str| self.push_str(string:s)); |
2232 | } |
2233 | |
2234 | #[inline ] |
2235 | fn extend_one(&mut self, s: &'a str) { |
2236 | self.push_str(string:s); |
2237 | } |
2238 | } |
2239 | |
2240 | #[cfg (not(no_global_oom_handling))] |
2241 | #[stable (feature = "box_str2" , since = "1.45.0" )] |
2242 | impl Extend<Box<str>> for String { |
2243 | fn extend<I: IntoIterator<Item = Box<str>>>(&mut self, iter: I) { |
2244 | iter.into_iter().for_each(move |s: Box| self.push_str(&s)); |
2245 | } |
2246 | } |
2247 | |
2248 | #[cfg (not(no_global_oom_handling))] |
2249 | #[stable (feature = "extend_string" , since = "1.4.0" )] |
2250 | impl Extend<String> for String { |
2251 | fn extend<I: IntoIterator<Item = String>>(&mut self, iter: I) { |
2252 | iter.into_iter().for_each(move |s: String| self.push_str(&s)); |
2253 | } |
2254 | |
2255 | #[inline ] |
2256 | fn extend_one(&mut self, s: String) { |
2257 | self.push_str(&s); |
2258 | } |
2259 | } |
2260 | |
2261 | #[cfg (not(no_global_oom_handling))] |
2262 | #[stable (feature = "herd_cows" , since = "1.19.0" )] |
2263 | impl<'a> Extend<Cow<'a, str>> for String { |
2264 | fn extend<I: IntoIterator<Item = Cow<'a, str>>>(&mut self, iter: I) { |
2265 | iter.into_iter().for_each(move |s: Cow<'_, str>| self.push_str(&s)); |
2266 | } |
2267 | |
2268 | #[inline ] |
2269 | fn extend_one(&mut self, s: Cow<'a, str>) { |
2270 | self.push_str(&s); |
2271 | } |
2272 | } |
2273 | |
2274 | /// A convenience impl that delegates to the impl for `&str`. |
2275 | /// |
2276 | /// # Examples |
2277 | /// |
2278 | /// ``` |
2279 | /// assert_eq!(String::from("Hello world" ).find("world" ), Some(6)); |
2280 | /// ``` |
2281 | #[unstable ( |
2282 | feature = "pattern" , |
2283 | reason = "API not fully fleshed out and ready to be stabilized" , |
2284 | issue = "27721" |
2285 | )] |
2286 | impl<'a, 'b> Pattern<'a> for &'b String { |
2287 | type Searcher = <&'b str as Pattern<'a>>::Searcher; |
2288 | |
2289 | fn into_searcher(self, haystack: &'a str) -> <&'b str as Pattern<'a>>::Searcher { |
2290 | self[..].into_searcher(haystack) |
2291 | } |
2292 | |
2293 | #[inline ] |
2294 | fn is_contained_in(self, haystack: &'a str) -> bool { |
2295 | self[..].is_contained_in(haystack) |
2296 | } |
2297 | |
2298 | #[inline ] |
2299 | fn is_prefix_of(self, haystack: &'a str) -> bool { |
2300 | self[..].is_prefix_of(haystack) |
2301 | } |
2302 | |
2303 | #[inline ] |
2304 | fn strip_prefix_of(self, haystack: &'a str) -> Option<&'a str> { |
2305 | self[..].strip_prefix_of(haystack) |
2306 | } |
2307 | |
2308 | #[inline ] |
2309 | fn is_suffix_of(self, haystack: &'a str) -> bool { |
2310 | self[..].is_suffix_of(haystack) |
2311 | } |
2312 | |
2313 | #[inline ] |
2314 | fn strip_suffix_of(self, haystack: &'a str) -> Option<&'a str> { |
2315 | self[..].strip_suffix_of(haystack) |
2316 | } |
2317 | } |
2318 | |
2319 | macro_rules! impl_eq { |
2320 | ($lhs:ty, $rhs: ty) => { |
2321 | #[stable(feature = "rust1" , since = "1.0.0" )] |
2322 | #[allow(unused_lifetimes)] |
2323 | impl<'a, 'b> PartialEq<$rhs> for $lhs { |
2324 | #[inline] |
2325 | fn eq(&self, other: &$rhs) -> bool { |
2326 | PartialEq::eq(&self[..], &other[..]) |
2327 | } |
2328 | #[inline] |
2329 | fn ne(&self, other: &$rhs) -> bool { |
2330 | PartialEq::ne(&self[..], &other[..]) |
2331 | } |
2332 | } |
2333 | |
2334 | #[stable(feature = "rust1" , since = "1.0.0" )] |
2335 | #[allow(unused_lifetimes)] |
2336 | impl<'a, 'b> PartialEq<$lhs> for $rhs { |
2337 | #[inline] |
2338 | fn eq(&self, other: &$lhs) -> bool { |
2339 | PartialEq::eq(&self[..], &other[..]) |
2340 | } |
2341 | #[inline] |
2342 | fn ne(&self, other: &$lhs) -> bool { |
2343 | PartialEq::ne(&self[..], &other[..]) |
2344 | } |
2345 | } |
2346 | }; |
2347 | } |
2348 | |
2349 | impl_eq! { String, str } |
2350 | impl_eq! { String, &'a str } |
2351 | #[cfg (not(no_global_oom_handling))] |
2352 | impl_eq! { Cow<'a, str>, str } |
2353 | #[cfg (not(no_global_oom_handling))] |
2354 | impl_eq! { Cow<'a, str>, &'b str } |
2355 | #[cfg (not(no_global_oom_handling))] |
2356 | impl_eq! { Cow<'a, str>, String } |
2357 | |
2358 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2359 | impl Default for String { |
2360 | /// Creates an empty `String`. |
2361 | #[inline ] |
2362 | fn default() -> String { |
2363 | String::new() |
2364 | } |
2365 | } |
2366 | |
2367 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2368 | impl fmt::Display for String { |
2369 | #[inline ] |
2370 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
2371 | fmt::Display::fmt(&**self, f) |
2372 | } |
2373 | } |
2374 | |
2375 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2376 | impl fmt::Debug for String { |
2377 | #[inline ] |
2378 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
2379 | fmt::Debug::fmt(&**self, f) |
2380 | } |
2381 | } |
2382 | |
2383 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2384 | impl hash::Hash for String { |
2385 | #[inline ] |
2386 | fn hash<H: hash::Hasher>(&self, hasher: &mut H) { |
2387 | (**self).hash(state:hasher) |
2388 | } |
2389 | } |
2390 | |
2391 | /// Implements the `+` operator for concatenating two strings. |
2392 | /// |
2393 | /// This consumes the `String` on the left-hand side and re-uses its buffer (growing it if |
2394 | /// necessary). This is done to avoid allocating a new `String` and copying the entire contents on |
2395 | /// every operation, which would lead to *O*(*n*^2) running time when building an *n*-byte string by |
2396 | /// repeated concatenation. |
2397 | /// |
2398 | /// The string on the right-hand side is only borrowed; its contents are copied into the returned |
2399 | /// `String`. |
2400 | /// |
2401 | /// # Examples |
2402 | /// |
2403 | /// Concatenating two `String`s takes the first by value and borrows the second: |
2404 | /// |
2405 | /// ``` |
2406 | /// let a = String::from("hello" ); |
2407 | /// let b = String::from(" world" ); |
2408 | /// let c = a + &b; |
2409 | /// // `a` is moved and can no longer be used here. |
2410 | /// ``` |
2411 | /// |
2412 | /// If you want to keep using the first `String`, you can clone it and append to the clone instead: |
2413 | /// |
2414 | /// ``` |
2415 | /// let a = String::from("hello" ); |
2416 | /// let b = String::from(" world" ); |
2417 | /// let c = a.clone() + &b; |
2418 | /// // `a` is still valid here. |
2419 | /// ``` |
2420 | /// |
2421 | /// Concatenating `&str` slices can be done by converting the first to a `String`: |
2422 | /// |
2423 | /// ``` |
2424 | /// let a = "hello" ; |
2425 | /// let b = " world" ; |
2426 | /// let c = a.to_string() + b; |
2427 | /// ``` |
2428 | #[cfg (not(no_global_oom_handling))] |
2429 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2430 | impl Add<&str> for String { |
2431 | type Output = String; |
2432 | |
2433 | #[inline ] |
2434 | fn add(mut self, other: &str) -> String { |
2435 | self.push_str(string:other); |
2436 | self |
2437 | } |
2438 | } |
2439 | |
2440 | /// Implements the `+=` operator for appending to a `String`. |
2441 | /// |
2442 | /// This has the same behavior as the [`push_str`][String::push_str] method. |
2443 | #[cfg (not(no_global_oom_handling))] |
2444 | #[stable (feature = "stringaddassign" , since = "1.12.0" )] |
2445 | impl AddAssign<&str> for String { |
2446 | #[inline ] |
2447 | fn add_assign(&mut self, other: &str) { |
2448 | self.push_str(string:other); |
2449 | } |
2450 | } |
2451 | |
2452 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2453 | impl<I> ops::Index<I> for String |
2454 | where |
2455 | I: slice::SliceIndex<str>, |
2456 | { |
2457 | type Output = I::Output; |
2458 | |
2459 | #[inline ] |
2460 | fn index(&self, index: I) -> &I::Output { |
2461 | index.index(self.as_str()) |
2462 | } |
2463 | } |
2464 | |
2465 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2466 | impl<I> ops::IndexMut<I> for String |
2467 | where |
2468 | I: slice::SliceIndex<str>, |
2469 | { |
2470 | #[inline ] |
2471 | fn index_mut(&mut self, index: I) -> &mut I::Output { |
2472 | index.index_mut(self.as_mut_str()) |
2473 | } |
2474 | } |
2475 | |
2476 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2477 | impl ops::Deref for String { |
2478 | type Target = str; |
2479 | |
2480 | #[inline ] |
2481 | fn deref(&self) -> &str { |
2482 | unsafe { str::from_utf8_unchecked(&self.vec) } |
2483 | } |
2484 | } |
2485 | |
2486 | #[unstable (feature = "deref_pure_trait" , issue = "87121" )] |
2487 | unsafe impl ops::DerefPure for String {} |
2488 | |
2489 | #[stable (feature = "derefmut_for_string" , since = "1.3.0" )] |
2490 | impl ops::DerefMut for String { |
2491 | #[inline ] |
2492 | fn deref_mut(&mut self) -> &mut str { |
2493 | unsafe { str::from_utf8_unchecked_mut(&mut *self.vec) } |
2494 | } |
2495 | } |
2496 | |
2497 | /// A type alias for [`Infallible`]. |
2498 | /// |
2499 | /// This alias exists for backwards compatibility, and may be eventually deprecated. |
2500 | /// |
2501 | /// [`Infallible`]: core::convert::Infallible "convert::Infallible" |
2502 | #[stable (feature = "str_parse_error" , since = "1.5.0" )] |
2503 | pub type ParseError = core::convert::Infallible; |
2504 | |
2505 | #[cfg (not(no_global_oom_handling))] |
2506 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2507 | impl FromStr for String { |
2508 | type Err = core::convert::Infallible; |
2509 | #[inline ] |
2510 | fn from_str(s: &str) -> Result<String, Self::Err> { |
2511 | Ok(String::from(s)) |
2512 | } |
2513 | } |
2514 | |
2515 | /// A trait for converting a value to a `String`. |
2516 | /// |
2517 | /// This trait is automatically implemented for any type which implements the |
2518 | /// [`Display`] trait. As such, `ToString` shouldn't be implemented directly: |
2519 | /// [`Display`] should be implemented instead, and you get the `ToString` |
2520 | /// implementation for free. |
2521 | /// |
2522 | /// [`Display`]: fmt::Display |
2523 | #[cfg_attr (not(test), rustc_diagnostic_item = "ToString" )] |
2524 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2525 | pub trait ToString { |
2526 | /// Converts the given value to a `String`. |
2527 | /// |
2528 | /// # Examples |
2529 | /// |
2530 | /// ``` |
2531 | /// let i = 5; |
2532 | /// let five = String::from("5" ); |
2533 | /// |
2534 | /// assert_eq!(five, i.to_string()); |
2535 | /// ``` |
2536 | #[rustc_conversion_suggestion ] |
2537 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2538 | #[cfg_attr (not(test), rustc_diagnostic_item = "to_string_method" )] |
2539 | fn to_string(&self) -> String; |
2540 | } |
2541 | |
2542 | /// # Panics |
2543 | /// |
2544 | /// In this implementation, the `to_string` method panics |
2545 | /// if the `Display` implementation returns an error. |
2546 | /// This indicates an incorrect `Display` implementation |
2547 | /// since `fmt::Write for String` never returns an error itself. |
2548 | #[cfg (not(no_global_oom_handling))] |
2549 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2550 | impl<T: fmt::Display + ?Sized> ToString for T { |
2551 | // A common guideline is to not inline generic functions. However, |
2552 | // removing `#[inline]` from this method causes non-negligible regressions. |
2553 | // See <https://github.com/rust-lang/rust/pull/74852>, the last attempt |
2554 | // to try to remove it. |
2555 | #[inline ] |
2556 | default fn to_string(&self) -> String { |
2557 | let mut buf: String = String::new(); |
2558 | let mut formatter: Formatter<'_> = core::fmt::Formatter::new(&mut buf); |
2559 | // Bypass format_args!() to avoid write_str with zero-length strs |
2560 | fmt::Display::fmt(self, &mut formatter) |
2561 | .expect(msg:"a Display implementation returned an error unexpectedly" ); |
2562 | buf |
2563 | } |
2564 | } |
2565 | |
2566 | #[doc (hidden)] |
2567 | #[cfg (not(no_global_oom_handling))] |
2568 | #[unstable (feature = "ascii_char" , issue = "110998" )] |
2569 | impl ToString for core::ascii::Char { |
2570 | #[inline ] |
2571 | fn to_string(&self) -> String { |
2572 | self.as_str().to_owned() |
2573 | } |
2574 | } |
2575 | |
2576 | #[doc (hidden)] |
2577 | #[cfg (not(no_global_oom_handling))] |
2578 | #[stable (feature = "char_to_string_specialization" , since = "1.46.0" )] |
2579 | impl ToString for char { |
2580 | #[inline ] |
2581 | fn to_string(&self) -> String { |
2582 | String::from(self.encode_utf8(&mut [0; 4])) |
2583 | } |
2584 | } |
2585 | |
2586 | #[doc (hidden)] |
2587 | #[cfg (not(no_global_oom_handling))] |
2588 | #[stable (feature = "bool_to_string_specialization" , since = "1.68.0" )] |
2589 | impl ToString for bool { |
2590 | #[inline ] |
2591 | fn to_string(&self) -> String { |
2592 | String::from(if *self { "true" } else { "false" }) |
2593 | } |
2594 | } |
2595 | |
2596 | #[doc (hidden)] |
2597 | #[cfg (not(no_global_oom_handling))] |
2598 | #[stable (feature = "u8_to_string_specialization" , since = "1.54.0" )] |
2599 | impl ToString for u8 { |
2600 | #[inline ] |
2601 | fn to_string(&self) -> String { |
2602 | let mut buf: String = String::with_capacity(3); |
2603 | let mut n: u8 = *self; |
2604 | if n >= 10 { |
2605 | if n >= 100 { |
2606 | buf.push((b'0' + n / 100) as char); |
2607 | n %= 100; |
2608 | } |
2609 | buf.push((b'0' + n / 10) as char); |
2610 | n %= 10; |
2611 | } |
2612 | buf.push((b'0' + n) as char); |
2613 | buf |
2614 | } |
2615 | } |
2616 | |
2617 | #[doc (hidden)] |
2618 | #[cfg (not(no_global_oom_handling))] |
2619 | #[stable (feature = "i8_to_string_specialization" , since = "1.54.0" )] |
2620 | impl ToString for i8 { |
2621 | #[inline ] |
2622 | fn to_string(&self) -> String { |
2623 | let mut buf: String = String::with_capacity(4); |
2624 | if self.is_negative() { |
2625 | buf.push(ch:'-' ); |
2626 | } |
2627 | let mut n: u8 = self.unsigned_abs(); |
2628 | if n >= 10 { |
2629 | if n >= 100 { |
2630 | buf.push(ch:'1' ); |
2631 | n -= 100; |
2632 | } |
2633 | buf.push((b'0' + n / 10) as char); |
2634 | n %= 10; |
2635 | } |
2636 | buf.push((b'0' + n) as char); |
2637 | buf |
2638 | } |
2639 | } |
2640 | |
2641 | #[doc (hidden)] |
2642 | #[cfg (not(no_global_oom_handling))] |
2643 | #[stable (feature = "str_to_string_specialization" , since = "1.9.0" )] |
2644 | impl ToString for str { |
2645 | #[inline ] |
2646 | fn to_string(&self) -> String { |
2647 | String::from(self) |
2648 | } |
2649 | } |
2650 | |
2651 | #[doc (hidden)] |
2652 | #[cfg (not(no_global_oom_handling))] |
2653 | #[stable (feature = "cow_str_to_string_specialization" , since = "1.17.0" )] |
2654 | impl ToString for Cow<'_, str> { |
2655 | #[inline ] |
2656 | fn to_string(&self) -> String { |
2657 | self[..].to_owned() |
2658 | } |
2659 | } |
2660 | |
2661 | #[doc (hidden)] |
2662 | #[cfg (not(no_global_oom_handling))] |
2663 | #[stable (feature = "string_to_string_specialization" , since = "1.17.0" )] |
2664 | impl ToString for String { |
2665 | #[inline ] |
2666 | fn to_string(&self) -> String { |
2667 | self.to_owned() |
2668 | } |
2669 | } |
2670 | |
2671 | #[doc (hidden)] |
2672 | #[cfg (not(no_global_oom_handling))] |
2673 | #[stable (feature = "fmt_arguments_to_string_specialization" , since = "1.71.0" )] |
2674 | impl ToString for fmt::Arguments<'_> { |
2675 | #[inline ] |
2676 | fn to_string(&self) -> String { |
2677 | crate::fmt::format(*self) |
2678 | } |
2679 | } |
2680 | |
2681 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2682 | impl AsRef<str> for String { |
2683 | #[inline ] |
2684 | fn as_ref(&self) -> &str { |
2685 | self |
2686 | } |
2687 | } |
2688 | |
2689 | #[stable (feature = "string_as_mut" , since = "1.43.0" )] |
2690 | impl AsMut<str> for String { |
2691 | #[inline ] |
2692 | fn as_mut(&mut self) -> &mut str { |
2693 | self |
2694 | } |
2695 | } |
2696 | |
2697 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2698 | impl AsRef<[u8]> for String { |
2699 | #[inline ] |
2700 | fn as_ref(&self) -> &[u8] { |
2701 | self.as_bytes() |
2702 | } |
2703 | } |
2704 | |
2705 | #[cfg (not(no_global_oom_handling))] |
2706 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2707 | impl From<&str> for String { |
2708 | /// Converts a `&str` into a [`String`]. |
2709 | /// |
2710 | /// The result is allocated on the heap. |
2711 | #[inline ] |
2712 | fn from(s: &str) -> String { |
2713 | s.to_owned() |
2714 | } |
2715 | } |
2716 | |
2717 | #[cfg (not(no_global_oom_handling))] |
2718 | #[stable (feature = "from_mut_str_for_string" , since = "1.44.0" )] |
2719 | impl From<&mut str> for String { |
2720 | /// Converts a `&mut str` into a [`String`]. |
2721 | /// |
2722 | /// The result is allocated on the heap. |
2723 | #[inline ] |
2724 | fn from(s: &mut str) -> String { |
2725 | s.to_owned() |
2726 | } |
2727 | } |
2728 | |
2729 | #[cfg (not(no_global_oom_handling))] |
2730 | #[stable (feature = "from_ref_string" , since = "1.35.0" )] |
2731 | impl From<&String> for String { |
2732 | /// Converts a `&String` into a [`String`]. |
2733 | /// |
2734 | /// This clones `s` and returns the clone. |
2735 | #[inline ] |
2736 | fn from(s: &String) -> String { |
2737 | s.clone() |
2738 | } |
2739 | } |
2740 | |
2741 | // note: test pulls in std, which causes errors here |
2742 | #[cfg (not(test))] |
2743 | #[stable (feature = "string_from_box" , since = "1.18.0" )] |
2744 | impl From<Box<str>> for String { |
2745 | /// Converts the given boxed `str` slice to a [`String`]. |
2746 | /// It is notable that the `str` slice is owned. |
2747 | /// |
2748 | /// # Examples |
2749 | /// |
2750 | /// ``` |
2751 | /// let s1: String = String::from("hello world" ); |
2752 | /// let s2: Box<str> = s1.into_boxed_str(); |
2753 | /// let s3: String = String::from(s2); |
2754 | /// |
2755 | /// assert_eq!("hello world" , s3) |
2756 | /// ``` |
2757 | fn from(s: Box<str>) -> String { |
2758 | s.into_string() |
2759 | } |
2760 | } |
2761 | |
2762 | #[cfg (not(no_global_oom_handling))] |
2763 | #[stable (feature = "box_from_str" , since = "1.20.0" )] |
2764 | impl From<String> for Box<str> { |
2765 | /// Converts the given [`String`] to a boxed `str` slice that is owned. |
2766 | /// |
2767 | /// # Examples |
2768 | /// |
2769 | /// ``` |
2770 | /// let s1: String = String::from("hello world" ); |
2771 | /// let s2: Box<str> = Box::from(s1); |
2772 | /// let s3: String = String::from(s2); |
2773 | /// |
2774 | /// assert_eq!("hello world" , s3) |
2775 | /// ``` |
2776 | fn from(s: String) -> Box<str> { |
2777 | s.into_boxed_str() |
2778 | } |
2779 | } |
2780 | |
2781 | #[cfg (not(no_global_oom_handling))] |
2782 | #[stable (feature = "string_from_cow_str" , since = "1.14.0" )] |
2783 | impl<'a> From<Cow<'a, str>> for String { |
2784 | /// Converts a clone-on-write string to an owned |
2785 | /// instance of [`String`]. |
2786 | /// |
2787 | /// This extracts the owned string, |
2788 | /// clones the string if it is not already owned. |
2789 | /// |
2790 | /// # Example |
2791 | /// |
2792 | /// ``` |
2793 | /// # use std::borrow::Cow; |
2794 | /// // If the string is not owned... |
2795 | /// let cow: Cow<'_, str> = Cow::Borrowed("eggplant" ); |
2796 | /// // It will allocate on the heap and copy the string. |
2797 | /// let owned: String = String::from(cow); |
2798 | /// assert_eq!(&owned[..], "eggplant" ); |
2799 | /// ``` |
2800 | fn from(s: Cow<'a, str>) -> String { |
2801 | s.into_owned() |
2802 | } |
2803 | } |
2804 | |
2805 | #[cfg (not(no_global_oom_handling))] |
2806 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2807 | impl<'a> From<&'a str> for Cow<'a, str> { |
2808 | /// Converts a string slice into a [`Borrowed`] variant. |
2809 | /// No heap allocation is performed, and the string |
2810 | /// is not copied. |
2811 | /// |
2812 | /// # Example |
2813 | /// |
2814 | /// ``` |
2815 | /// # use std::borrow::Cow; |
2816 | /// assert_eq!(Cow::from("eggplant" ), Cow::Borrowed("eggplant" )); |
2817 | /// ``` |
2818 | /// |
2819 | /// [`Borrowed`]: crate::borrow::Cow::Borrowed "borrow::Cow::Borrowed" |
2820 | #[inline ] |
2821 | fn from(s: &'a str) -> Cow<'a, str> { |
2822 | Cow::Borrowed(s) |
2823 | } |
2824 | } |
2825 | |
2826 | #[cfg (not(no_global_oom_handling))] |
2827 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2828 | impl<'a> From<String> for Cow<'a, str> { |
2829 | /// Converts a [`String`] into an [`Owned`] variant. |
2830 | /// No heap allocation is performed, and the string |
2831 | /// is not copied. |
2832 | /// |
2833 | /// # Example |
2834 | /// |
2835 | /// ``` |
2836 | /// # use std::borrow::Cow; |
2837 | /// let s = "eggplant" .to_string(); |
2838 | /// let s2 = "eggplant" .to_string(); |
2839 | /// assert_eq!(Cow::from(s), Cow::<'static, str>::Owned(s2)); |
2840 | /// ``` |
2841 | /// |
2842 | /// [`Owned`]: crate::borrow::Cow::Owned "borrow::Cow::Owned" |
2843 | #[inline ] |
2844 | fn from(s: String) -> Cow<'a, str> { |
2845 | Cow::Owned(s) |
2846 | } |
2847 | } |
2848 | |
2849 | #[cfg (not(no_global_oom_handling))] |
2850 | #[stable (feature = "cow_from_string_ref" , since = "1.28.0" )] |
2851 | impl<'a> From<&'a String> for Cow<'a, str> { |
2852 | /// Converts a [`String`] reference into a [`Borrowed`] variant. |
2853 | /// No heap allocation is performed, and the string |
2854 | /// is not copied. |
2855 | /// |
2856 | /// # Example |
2857 | /// |
2858 | /// ``` |
2859 | /// # use std::borrow::Cow; |
2860 | /// let s = "eggplant" .to_string(); |
2861 | /// assert_eq!(Cow::from(&s), Cow::Borrowed("eggplant" )); |
2862 | /// ``` |
2863 | /// |
2864 | /// [`Borrowed`]: crate::borrow::Cow::Borrowed "borrow::Cow::Borrowed" |
2865 | #[inline ] |
2866 | fn from(s: &'a String) -> Cow<'a, str> { |
2867 | Cow::Borrowed(s.as_str()) |
2868 | } |
2869 | } |
2870 | |
2871 | #[cfg (not(no_global_oom_handling))] |
2872 | #[stable (feature = "cow_str_from_iter" , since = "1.12.0" )] |
2873 | impl<'a> FromIterator<char> for Cow<'a, str> { |
2874 | fn from_iter<I: IntoIterator<Item = char>>(it: I) -> Cow<'a, str> { |
2875 | Cow::Owned(FromIterator::from_iter(it)) |
2876 | } |
2877 | } |
2878 | |
2879 | #[cfg (not(no_global_oom_handling))] |
2880 | #[stable (feature = "cow_str_from_iter" , since = "1.12.0" )] |
2881 | impl<'a, 'b> FromIterator<&'b str> for Cow<'a, str> { |
2882 | fn from_iter<I: IntoIterator<Item = &'b str>>(it: I) -> Cow<'a, str> { |
2883 | Cow::Owned(FromIterator::from_iter(it)) |
2884 | } |
2885 | } |
2886 | |
2887 | #[cfg (not(no_global_oom_handling))] |
2888 | #[stable (feature = "cow_str_from_iter" , since = "1.12.0" )] |
2889 | impl<'a> FromIterator<String> for Cow<'a, str> { |
2890 | fn from_iter<I: IntoIterator<Item = String>>(it: I) -> Cow<'a, str> { |
2891 | Cow::Owned(FromIterator::from_iter(it)) |
2892 | } |
2893 | } |
2894 | |
2895 | #[stable (feature = "from_string_for_vec_u8" , since = "1.14.0" )] |
2896 | impl From<String> for Vec<u8> { |
2897 | /// Converts the given [`String`] to a vector [`Vec`] that holds values of type [`u8`]. |
2898 | /// |
2899 | /// # Examples |
2900 | /// |
2901 | /// ``` |
2902 | /// let s1 = String::from("hello world" ); |
2903 | /// let v1 = Vec::from(s1); |
2904 | /// |
2905 | /// for b in v1 { |
2906 | /// println!("{b}" ); |
2907 | /// } |
2908 | /// ``` |
2909 | fn from(string: String) -> Vec<u8> { |
2910 | string.into_bytes() |
2911 | } |
2912 | } |
2913 | |
2914 | #[cfg (not(no_global_oom_handling))] |
2915 | #[stable (feature = "rust1" , since = "1.0.0" )] |
2916 | impl fmt::Write for String { |
2917 | #[inline ] |
2918 | fn write_str(&mut self, s: &str) -> fmt::Result { |
2919 | self.push_str(string:s); |
2920 | Ok(()) |
2921 | } |
2922 | |
2923 | #[inline ] |
2924 | fn write_char(&mut self, c: char) -> fmt::Result { |
2925 | self.push(ch:c); |
2926 | Ok(()) |
2927 | } |
2928 | } |
2929 | |
2930 | /// A draining iterator for `String`. |
2931 | /// |
2932 | /// This struct is created by the [`drain`] method on [`String`]. See its |
2933 | /// documentation for more. |
2934 | /// |
2935 | /// [`drain`]: String::drain |
2936 | #[stable (feature = "drain" , since = "1.6.0" )] |
2937 | pub struct Drain<'a> { |
2938 | /// Will be used as &'a mut String in the destructor |
2939 | string: *mut String, |
2940 | /// Start of part to remove |
2941 | start: usize, |
2942 | /// End of part to remove |
2943 | end: usize, |
2944 | /// Current remaining range to remove |
2945 | iter: Chars<'a>, |
2946 | } |
2947 | |
2948 | #[stable (feature = "collection_debug" , since = "1.17.0" )] |
2949 | impl fmt::Debug for Drain<'_> { |
2950 | fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { |
2951 | f.debug_tuple(name:"Drain" ).field(&self.as_str()).finish() |
2952 | } |
2953 | } |
2954 | |
2955 | #[stable (feature = "drain" , since = "1.6.0" )] |
2956 | unsafe impl Sync for Drain<'_> {} |
2957 | #[stable (feature = "drain" , since = "1.6.0" )] |
2958 | unsafe impl Send for Drain<'_> {} |
2959 | |
2960 | #[stable (feature = "drain" , since = "1.6.0" )] |
2961 | impl Drop for Drain<'_> { |
2962 | fn drop(&mut self) { |
2963 | unsafe { |
2964 | // Use Vec::drain. "Reaffirm" the bounds checks to avoid |
2965 | // panic code being inserted again. |
2966 | let self_vec: &mut Vec = (*self.string).as_mut_vec(); |
2967 | if self.start <= self.end && self.end <= self_vec.len() { |
2968 | self_vec.drain(self.start..self.end); |
2969 | } |
2970 | } |
2971 | } |
2972 | } |
2973 | |
2974 | impl<'a> Drain<'a> { |
2975 | /// Returns the remaining (sub)string of this iterator as a slice. |
2976 | /// |
2977 | /// # Examples |
2978 | /// |
2979 | /// ``` |
2980 | /// let mut s = String::from("abc" ); |
2981 | /// let mut drain = s.drain(..); |
2982 | /// assert_eq!(drain.as_str(), "abc" ); |
2983 | /// let _ = drain.next().unwrap(); |
2984 | /// assert_eq!(drain.as_str(), "bc" ); |
2985 | /// ``` |
2986 | #[must_use ] |
2987 | #[stable (feature = "string_drain_as_str" , since = "1.55.0" )] |
2988 | pub fn as_str(&self) -> &str { |
2989 | self.iter.as_str() |
2990 | } |
2991 | } |
2992 | |
2993 | #[stable (feature = "string_drain_as_str" , since = "1.55.0" )] |
2994 | impl<'a> AsRef<str> for Drain<'a> { |
2995 | fn as_ref(&self) -> &str { |
2996 | self.as_str() |
2997 | } |
2998 | } |
2999 | |
3000 | #[stable (feature = "string_drain_as_str" , since = "1.55.0" )] |
3001 | impl<'a> AsRef<[u8]> for Drain<'a> { |
3002 | fn as_ref(&self) -> &[u8] { |
3003 | self.as_str().as_bytes() |
3004 | } |
3005 | } |
3006 | |
3007 | #[stable (feature = "drain" , since = "1.6.0" )] |
3008 | impl Iterator for Drain<'_> { |
3009 | type Item = char; |
3010 | |
3011 | #[inline ] |
3012 | fn next(&mut self) -> Option<char> { |
3013 | self.iter.next() |
3014 | } |
3015 | |
3016 | fn size_hint(&self) -> (usize, Option<usize>) { |
3017 | self.iter.size_hint() |
3018 | } |
3019 | |
3020 | #[inline ] |
3021 | fn last(mut self) -> Option<char> { |
3022 | self.next_back() |
3023 | } |
3024 | } |
3025 | |
3026 | #[stable (feature = "drain" , since = "1.6.0" )] |
3027 | impl DoubleEndedIterator for Drain<'_> { |
3028 | #[inline ] |
3029 | fn next_back(&mut self) -> Option<char> { |
3030 | self.iter.next_back() |
3031 | } |
3032 | } |
3033 | |
3034 | #[stable (feature = "fused" , since = "1.26.0" )] |
3035 | impl FusedIterator for Drain<'_> {} |
3036 | |
3037 | #[cfg (not(no_global_oom_handling))] |
3038 | #[stable (feature = "from_char_for_string" , since = "1.46.0" )] |
3039 | impl From<char> for String { |
3040 | /// Allocates an owned [`String`] from a single character. |
3041 | /// |
3042 | /// # Example |
3043 | /// ```rust |
3044 | /// let c: char = 'a' ; |
3045 | /// let s: String = String::from(c); |
3046 | /// assert_eq!("a" , &s[..]); |
3047 | /// ``` |
3048 | #[inline ] |
3049 | fn from(c: char) -> Self { |
3050 | c.to_string() |
3051 | } |
3052 | } |
3053 | |