1 // Copyright 2014 The Rust Project Developers. See the COPYRIGHT
2 // file at the top-level directory of this distribution and at
3 // http://rust-lang.org/COPYRIGHT.
5 // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
6 // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
7 // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
8 // option. This file may not be copied, modified, or distributed
9 // except according to those terms.
11 // ignore-lexer-test FIXME #15679
13 //! An owned, growable string that enforces that its contents are valid UTF-8.
19 use core::borrow::{Cow, IntoCow};
21 use core::default::Default;
24 use core::iter::FromIterator;
26 use core::ops::{mod, Deref, Add};
28 use core::raw::Slice as RawSlice;
29 use unicode::str as unicode_str;
30 use unicode::str::Utf16Item;
32 use str::{mod, CharRange, FromStr, Utf8Error};
33 use vec::{DerefVec, Vec, as_vec};
35 /// A growable string stored as a UTF-8 encoded buffer.
36 #[deriving(Clone, PartialOrd, Eq, Ord)]
42 /// A possible error value from the `String::from_utf8` function.
44 pub struct FromUtf8Error {
49 /// A possible error value from the `String::from_utf16` function.
51 #[allow(missing_copy_implementations)]
52 pub struct FromUtf16Error(());
55 /// Creates a new string buffer initialized with the empty string.
60 /// let mut s = String::new();
64 pub fn new() -> String {
70 /// Creates a new string buffer with the given capacity.
71 /// The string will be able to hold exactly `capacity` bytes without
72 /// reallocating. If `capacity` is 0, the string will not allocate.
77 /// let mut s = String::with_capacity(10);
81 pub fn with_capacity(capacity: uint) -> String {
83 vec: Vec::with_capacity(capacity),
87 /// Creates a new string buffer from the given string.
92 /// let s = String::from_str("hello");
93 /// assert_eq!(s.as_slice(), "hello");
96 #[experimental = "needs investigation to see if to_string() can match perf"]
97 pub fn from_str(string: &str) -> String {
98 String { vec: ::slice::SliceExt::to_vec(string.as_bytes()) }
101 /// Returns the vector as a string buffer, if possible, taking care not to
106 /// If the given vector is not valid UTF-8, then the original vector and the
107 /// corresponding error is returned.
112 /// # #![allow(deprecated)]
113 /// use std::str::Utf8Error;
115 /// let hello_vec = vec![104, 101, 108, 108, 111];
116 /// let s = String::from_utf8(hello_vec).unwrap();
117 /// assert_eq!(s, "hello");
119 /// let invalid_vec = vec![240, 144, 128];
120 /// let s = String::from_utf8(invalid_vec).err().unwrap();
121 /// assert_eq!(s.utf8_error(), Utf8Error::TooShort);
122 /// assert_eq!(s.into_bytes(), vec![240, 144, 128]);
126 pub fn from_utf8(vec: Vec<u8>) -> Result<String, FromUtf8Error> {
127 match str::from_utf8(vec.as_slice()) {
128 Ok(..) => Ok(String { vec: vec }),
129 Err(e) => Err(FromUtf8Error { bytes: vec, error: e })
133 /// Converts a vector of bytes to a new UTF-8 string.
134 /// Any invalid UTF-8 sequences are replaced with U+FFFD REPLACEMENT CHARACTER.
139 /// let input = b"Hello \xF0\x90\x80World";
140 /// let output = String::from_utf8_lossy(input);
141 /// assert_eq!(output.as_slice(), "Hello \u{FFFD}World");
144 pub fn from_utf8_lossy<'a>(v: &'a [u8]) -> CowString<'a> {
146 match str::from_utf8(v) {
147 Ok(s) => return Cow::Borrowed(s),
149 if let Utf8Error::InvalidByte(firstbad) = e {
155 static TAG_CONT_U8: u8 = 128u8;
156 static REPLACEMENT: &'static [u8] = b"\xEF\xBF\xBD"; // U+FFFD in UTF-8
158 fn unsafe_get(xs: &[u8], i: uint) -> u8 {
159 unsafe { *xs.get_unchecked(i) }
161 fn safe_get(xs: &[u8], i: uint, total: uint) -> u8 {
169 let mut res = String::with_capacity(total);
173 res.as_mut_vec().push_all(v[..i])
177 // subseqidx is the index of the first byte of the subsequence we're looking at.
178 // It's used to copy a bunch of contiguous good codepoints at once instead of copying
180 let mut subseqidx = i;
184 let byte = unsafe_get(v, i);
187 macro_rules! error(() => ({
190 res.as_mut_vec().push_all(v[subseqidx..i_]);
193 res.as_mut_vec().push_all(REPLACEMENT);
198 // subseqidx handles this
200 let w = unicode_str::utf8_char_width(byte);
204 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
211 match (byte, safe_get(v, i, total)) {
212 (0xE0 , 0xA0 ... 0xBF) => (),
213 (0xE1 ... 0xEC, 0x80 ... 0xBF) => (),
214 (0xED , 0x80 ... 0x9F) => (),
215 (0xEE ... 0xEF, 0x80 ... 0xBF) => (),
222 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
229 match (byte, safe_get(v, i, total)) {
230 (0xF0 , 0x90 ... 0xBF) => (),
231 (0xF1 ... 0xF3, 0x80 ... 0xBF) => (),
232 (0xF4 , 0x80 ... 0x8F) => (),
239 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
244 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
257 if subseqidx < total {
259 res.as_mut_vec().push_all(v[subseqidx..total])
265 /// Decode a UTF-16 encoded vector `v` into a `String`, returning `None`
266 /// if `v` contains any invalid data.
272 /// let mut v = &mut [0xD834, 0xDD1E, 0x006d, 0x0075,
273 /// 0x0073, 0x0069, 0x0063];
274 /// assert_eq!(String::from_utf16(v).unwrap(),
275 /// "𝄞music".to_string());
277 /// // 𝄞mu<invalid>ic
279 /// assert!(String::from_utf16(v).is_err());
282 pub fn from_utf16(v: &[u16]) -> Result<String, FromUtf16Error> {
283 let mut s = String::with_capacity(v.len());
284 for c in unicode_str::utf16_items(v) {
286 Utf16Item::ScalarValue(c) => s.push(c),
287 Utf16Item::LoneSurrogate(_) => return Err(FromUtf16Error(())),
293 /// Decode a UTF-16 encoded vector `v` into a string, replacing
294 /// invalid data with the replacement character (U+FFFD).
299 /// // 𝄞mus<invalid>ic<invalid>
300 /// let v = &[0xD834, 0xDD1E, 0x006d, 0x0075,
301 /// 0x0073, 0xDD1E, 0x0069, 0x0063,
304 /// assert_eq!(String::from_utf16_lossy(v),
305 /// "𝄞mus\u{FFFD}ic\u{FFFD}".to_string());
308 pub fn from_utf16_lossy(v: &[u16]) -> String {
309 unicode_str::utf16_items(v).map(|c| c.to_char_lossy()).collect()
312 /// Convert a vector of `char`s to a `String`.
317 /// # #![allow(deprecated)]
318 /// let chars = &['h', 'e', 'l', 'l', 'o'];
319 /// let s = String::from_chars(chars);
320 /// assert_eq!(s.as_slice(), "hello");
323 #[deprecated = "use .collect() instead"]
324 pub fn from_chars(chs: &[char]) -> String {
325 chs.iter().map(|c| *c).collect()
328 /// Creates a new `String` from a length, capacity, and pointer.
330 /// This is unsafe because:
331 /// * We call `Vec::from_raw_parts` to get a `Vec<u8>`;
332 /// * We assume that the `Vec` contains valid UTF-8.
335 pub unsafe fn from_raw_parts(buf: *mut u8, length: uint, capacity: uint) -> String {
337 vec: Vec::from_raw_parts(buf, length, capacity),
341 /// Creates a `String` from a null-terminated `*const u8` buffer.
343 /// This function is unsafe because we dereference memory until we find the
344 /// NUL character, which is not guaranteed to be present. Additionally, the
345 /// slice is not checked to see whether it contains valid UTF-8
346 #[unstable = "just renamed from `mod raw`"]
347 pub unsafe fn from_raw_buf(buf: *const u8) -> String {
348 String::from_str(str::from_c_str(buf as *const i8))
351 /// Creates a `String` from a `*const u8` buffer of the given length.
353 /// This function is unsafe because it blindly assumes the validity of the
354 /// pointer `buf` for `len` bytes of memory. This function will copy the
355 /// memory from `buf` into a new allocation (owned by the returned
358 /// This function is also unsafe because it does not validate that the
359 /// buffer is valid UTF-8 encoded data.
360 #[unstable = "just renamed from `mod raw`"]
361 pub unsafe fn from_raw_buf_len(buf: *const u8, len: uint) -> String {
362 String::from_utf8_unchecked(Vec::from_raw_buf(buf, len))
365 /// Converts a vector of bytes to a new `String` without checking if
366 /// it contains valid UTF-8. This is unsafe because it assumes that
367 /// the UTF-8-ness of the vector has already been validated.
370 pub unsafe fn from_utf8_unchecked(bytes: Vec<u8>) -> String {
371 String { vec: bytes }
374 /// Return the underlying byte buffer, encoded as UTF-8.
379 /// let s = String::from_str("hello");
380 /// let bytes = s.into_bytes();
381 /// assert_eq!(bytes, vec![104, 101, 108, 108, 111]);
385 pub fn into_bytes(self) -> Vec<u8> {
389 /// Creates a string buffer by repeating a character `length` times.
394 /// # #![allow(deprecated)]
395 /// let s = String::from_char(5, 'a');
396 /// assert_eq!(s.as_slice(), "aaaaa");
399 #[deprecated = "use repeat(ch).take(length).collect() instead"]
400 pub fn from_char(length: uint, ch: char) -> String {
405 let mut buf = String::new();
407 let size = buf.len() * (length - 1);
408 buf.reserve_exact(size);
409 for _ in range(1, length) {
415 /// Pushes the given string onto this string buffer.
420 /// let mut s = String::from_str("foo");
421 /// s.push_str("bar");
422 /// assert_eq!(s.as_slice(), "foobar");
426 pub fn push_str(&mut self, string: &str) {
427 self.vec.push_all(string.as_bytes())
430 /// Pushes `ch` onto the given string `count` times.
435 /// # #![allow(deprecated)]
436 /// let mut s = String::from_str("foo");
438 /// assert_eq!(s.as_slice(), "fooZZZZZ");
441 #[deprecated = "deprecated in favor of .extend(repeat(ch).take(count))"]
442 pub fn grow(&mut self, count: uint, ch: char) {
443 for _ in range(0, count) {
448 /// Returns the number of bytes that this string buffer can hold without
454 /// let s = String::with_capacity(10);
455 /// assert!(s.capacity() >= 10);
459 pub fn capacity(&self) -> uint {
463 /// Deprecated: Renamed to `reserve`.
464 #[deprecated = "Renamed to `reserve`"]
465 pub fn reserve_additional(&mut self, extra: uint) {
466 self.vec.reserve(extra)
469 /// Reserves capacity for at least `additional` more bytes to be inserted
470 /// in the given `String`. The collection may reserve more space to avoid
471 /// frequent reallocations.
475 /// Panics if the new capacity overflows `uint`.
480 /// let mut s = String::new();
482 /// assert!(s.capacity() >= 10);
486 pub fn reserve(&mut self, additional: uint) {
487 self.vec.reserve(additional)
490 /// Reserves the minimum capacity for exactly `additional` more bytes to be
491 /// inserted in the given `String`. Does nothing if the capacity is already
494 /// Note that the allocator may give the collection more space than it
495 /// requests. Therefore capacity can not be relied upon to be precisely
496 /// minimal. Prefer `reserve` if future insertions are expected.
500 /// Panics if the new capacity overflows `uint`.
505 /// let mut s = String::new();
507 /// assert!(s.capacity() >= 10);
511 pub fn reserve_exact(&mut self, additional: uint) {
512 self.vec.reserve_exact(additional)
515 /// Shrinks the capacity of this string buffer to match its length.
520 /// let mut s = String::from_str("foo");
522 /// assert!(s.capacity() >= 100);
523 /// s.shrink_to_fit();
524 /// assert_eq!(s.capacity(), 3);
528 pub fn shrink_to_fit(&mut self) {
529 self.vec.shrink_to_fit()
532 /// Adds the given character to the end of the string.
537 /// let mut s = String::from_str("abc");
541 /// assert_eq!(s.as_slice(), "abc123");
545 pub fn push(&mut self, ch: char) {
546 if (ch as u32) < 0x80 {
547 self.vec.push(ch as u8);
551 let cur_len = self.len();
552 // This may use up to 4 bytes.
556 // Attempt to not use an intermediate buffer by just pushing bytes
557 // directly onto this string.
558 let slice = RawSlice {
559 data: self.vec.as_ptr().offset(cur_len as int),
562 let used = ch.encode_utf8(mem::transmute(slice)).unwrap_or(0);
563 self.vec.set_len(cur_len + used);
567 /// Works with the underlying buffer as a byte slice.
572 /// let s = String::from_str("hello");
573 /// let b: &[_] = &[104, 101, 108, 108, 111];
574 /// assert_eq!(s.as_bytes(), b);
578 pub fn as_bytes<'a>(&'a self) -> &'a [u8] {
582 /// Shortens a string to the specified length.
586 /// Panics if `new_len` > current length,
587 /// or if `new_len` is not a character boundary.
592 /// let mut s = String::from_str("hello");
594 /// assert_eq!(s.as_slice(), "he");
598 pub fn truncate(&mut self, new_len: uint) {
599 assert!(self.is_char_boundary(new_len));
600 self.vec.truncate(new_len)
603 /// Removes the last character from the string buffer and returns it.
604 /// Returns `None` if this string buffer is empty.
609 /// let mut s = String::from_str("foo");
610 /// assert_eq!(s.pop(), Some('o'));
611 /// assert_eq!(s.pop(), Some('o'));
612 /// assert_eq!(s.pop(), Some('f'));
613 /// assert_eq!(s.pop(), None);
617 pub fn pop(&mut self) -> Option<char> {
618 let len = self.len();
623 let CharRange {ch, next} = self.char_range_at_reverse(len);
625 self.vec.set_len(next);
630 /// Removes the character from the string buffer at byte position `idx` and
635 /// This is an O(n) operation as it requires copying every element in the
640 /// If `idx` does not lie on a character boundary, or if it is out of
641 /// bounds, then this function will panic.
646 /// let mut s = String::from_str("foo");
647 /// assert_eq!(s.remove(0), 'f');
648 /// assert_eq!(s.remove(1), 'o');
649 /// assert_eq!(s.remove(0), 'o');
652 pub fn remove(&mut self, idx: uint) -> char {
653 let len = self.len();
656 let CharRange { ch, next } = self.char_range_at(idx);
658 ptr::copy_memory(self.vec.as_mut_ptr().offset(idx as int),
659 self.vec.as_ptr().offset(next as int),
661 self.vec.set_len(len - (next - idx));
666 /// Insert a character into the string buffer at byte position `idx`.
670 /// This is an O(n) operation as it requires copying every element in the
675 /// If `idx` does not lie on a character boundary or is out of bounds, then
676 /// this function will panic.
678 pub fn insert(&mut self, idx: uint, ch: char) {
679 let len = self.len();
681 assert!(self.is_char_boundary(idx));
683 let mut bits = [0; 4];
684 let amt = ch.encode_utf8(&mut bits).unwrap();
687 ptr::copy_memory(self.vec.as_mut_ptr().offset((idx + amt) as int),
688 self.vec.as_ptr().offset(idx as int),
690 ptr::copy_memory(self.vec.as_mut_ptr().offset(idx as int),
693 self.vec.set_len(len + amt);
697 /// Views the string buffer as a mutable sequence of bytes.
699 /// This is unsafe because it does not check
700 /// to ensure that the resulting string will be valid UTF-8.
705 /// let mut s = String::from_str("hello");
707 /// let vec = s.as_mut_vec();
708 /// assert!(vec == &mut vec![104, 101, 108, 108, 111]);
711 /// assert_eq!(s.as_slice(), "olleh");
714 pub unsafe fn as_mut_vec<'a>(&'a mut self) -> &'a mut Vec<u8> {
718 /// Return the number of bytes in this string.
723 /// let a = "foo".to_string();
724 /// assert_eq!(a.len(), 3);
728 pub fn len(&self) -> uint { self.vec.len() }
730 /// Returns true if the string contains no bytes
735 /// let mut v = String::new();
736 /// assert!(v.is_empty());
738 /// assert!(!v.is_empty());
741 pub fn is_empty(&self) -> bool { self.len() == 0 }
743 /// Truncates the string, returning it to 0 length.
748 /// let mut s = "foo".to_string();
750 /// assert!(s.is_empty());
754 pub fn clear(&mut self) {
760 /// Consume this error, returning the bytes that were attempted to make a
763 pub fn into_bytes(self) -> Vec<u8> { self.bytes }
765 /// Access the underlying UTF8-error that was the cause of this error.
767 pub fn utf8_error(&self) -> Utf8Error { self.error }
770 impl fmt::Show for FromUtf8Error {
771 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
776 impl fmt::Show for FromUtf16Error {
777 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
778 "invalid utf-16: lone surrogate found".fmt(f)
782 #[experimental = "waiting on FromIterator stabilization"]
783 impl FromIterator<char> for String {
784 fn from_iter<I:Iterator<Item=char>>(iterator: I) -> String {
785 let mut buf = String::new();
786 buf.extend(iterator);
791 #[experimental = "waiting on FromIterator stabilization"]
792 impl<'a> FromIterator<&'a str> for String {
793 fn from_iter<I:Iterator<Item=&'a str>>(iterator: I) -> String {
794 let mut buf = String::new();
795 buf.extend(iterator);
800 #[experimental = "waiting on Extend stabilization"]
801 impl Extend<char> for String {
802 fn extend<I:Iterator<Item=char>>(&mut self, mut iterator: I) {
803 let (lower_bound, _) = iterator.size_hint();
804 self.reserve(lower_bound);
811 #[experimental = "waiting on Extend stabilization"]
812 impl<'a> Extend<&'a str> for String {
813 fn extend<I: Iterator<Item=&'a str>>(&mut self, mut iterator: I) {
814 // A guess that at least one byte per iterator element will be needed.
815 let (lower_bound, _) = iterator.size_hint();
816 self.reserve(lower_bound);
824 impl PartialEq for String {
826 fn eq(&self, other: &String) -> bool { PartialEq::eq(&**self, &**other) }
828 fn ne(&self, other: &String) -> bool { PartialEq::ne(&**self, &**other) }
831 macro_rules! impl_eq {
832 ($lhs:ty, $rhs: ty) => {
834 impl<'a> PartialEq<$rhs> for $lhs {
836 fn eq(&self, other: &$rhs) -> bool { PartialEq::eq(&**self, &**other) }
838 fn ne(&self, other: &$rhs) -> bool { PartialEq::ne(&**self, &**other) }
842 impl<'a> PartialEq<$lhs> for $rhs {
844 fn eq(&self, other: &$lhs) -> bool { PartialEq::eq(&**self, &**other) }
846 fn ne(&self, other: &$lhs) -> bool { PartialEq::ne(&**self, &**other) }
852 impl_eq! { String, &'a str }
853 impl_eq! { CowString<'a>, String }
856 impl<'a, 'b> PartialEq<&'b str> for CowString<'a> {
858 fn eq(&self, other: &&'b str) -> bool { PartialEq::eq(&**self, &**other) }
860 fn ne(&self, other: &&'b str) -> bool { PartialEq::ne(&**self, &**other) }
864 impl<'a, 'b> PartialEq<CowString<'a>> for &'b str {
866 fn eq(&self, other: &CowString<'a>) -> bool { PartialEq::eq(&**self, &**other) }
868 fn ne(&self, other: &CowString<'a>) -> bool { PartialEq::ne(&**self, &**other) }
871 #[experimental = "waiting on Str stabilization"]
873 impl Str for String {
876 fn as_slice<'a>(&'a self) -> &'a str {
877 unsafe { mem::transmute(self.vec.as_slice()) }
882 impl Default for String {
884 fn default() -> String {
889 #[experimental = "waiting on Show stabilization"]
890 impl fmt::Show for String {
891 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
896 #[experimental = "waiting on Hash stabilization"]
897 impl<H: hash::Writer> hash::Hash<H> for String {
899 fn hash(&self, hasher: &mut H) {
900 (**self).hash(hasher)
905 #[deprecated = "Use overloaded `core::cmp::PartialEq`"]
906 impl<'a, S: Str> Equiv<S> for String {
908 fn equiv(&self, other: &S) -> bool {
909 self.as_slice() == other.as_slice()
913 #[experimental = "waiting on Add stabilization"]
914 impl<'a> Add<&'a str> for String {
915 type Output = String;
917 fn add(mut self, other: &str) -> String {
918 self.push_str(other);
923 impl ops::Slice<uint, str> for String {
925 fn as_slice_<'a>(&'a self) -> &'a str {
926 unsafe { mem::transmute(self.vec.as_slice()) }
930 fn slice_from_or_fail<'a>(&'a self, from: &uint) -> &'a str {
935 fn slice_to_or_fail<'a>(&'a self, to: &uint) -> &'a str {
940 fn slice_or_fail<'a>(&'a self, from: &uint, to: &uint) -> &'a str {
945 #[experimental = "waiting on Deref stabilization"]
946 impl ops::Deref for String {
949 fn deref<'a>(&'a self) -> &'a str {
950 unsafe { mem::transmute(self.vec[]) }
954 /// Wrapper type providing a `&String` reference via `Deref`.
956 pub struct DerefString<'a> {
960 impl<'a> Deref for DerefString<'a> {
961 type Target = String;
963 fn deref<'b>(&'b self) -> &'b String {
964 unsafe { mem::transmute(&*self.x) }
968 /// Convert a string slice to a wrapper type providing a `&String` reference.
973 /// use std::string::as_string;
975 /// fn string_consumer(s: String) {
976 /// assert_eq!(s, "foo".to_string());
979 /// let string = as_string("foo").clone();
980 /// string_consumer(string);
983 pub fn as_string<'a>(x: &'a str) -> DerefString<'a> {
984 DerefString { x: as_vec(x.as_bytes()) }
987 impl FromStr for String {
989 fn from_str(s: &str) -> Option<String> {
990 Some(String::from_str(s))
994 /// Trait for converting a type to a string, consuming it in the process.
995 #[deprecated = "trait will be removed"]
996 pub trait IntoString {
997 /// Consume and convert to a string.
998 fn into_string(self) -> String;
1001 /// A generic trait for converting a value to a string
1002 pub trait ToString {
1003 /// Converts the value of `self` to an owned string
1004 fn to_string(&self) -> String;
1007 impl<T: fmt::Show> ToString for T {
1008 fn to_string(&self) -> String {
1009 use core::fmt::Writer;
1010 let mut buf = String::new();
1011 let _ = buf.write_fmt(format_args!("{}", self));
1012 buf.shrink_to_fit();
1017 impl IntoCow<'static, String, str> for String {
1018 fn into_cow(self) -> CowString<'static> {
1023 impl<'a> IntoCow<'a, String, str> for &'a str {
1024 fn into_cow(self) -> CowString<'a> {
1029 /// Unsafe operations
1035 /// Creates a new `String` from a length, capacity, and pointer.
1037 /// This is unsafe because:
1038 /// * We call `Vec::from_raw_parts` to get a `Vec<u8>`;
1039 /// * We assume that the `Vec` contains valid UTF-8.
1041 #[deprecated = "renamed to String::from_raw_parts"]
1042 pub unsafe fn from_parts(buf: *mut u8, length: uint, capacity: uint) -> String {
1043 String::from_raw_parts(buf, length, capacity)
1046 /// Creates a `String` from a `*const u8` buffer of the given length.
1048 /// This function is unsafe because of two reasons:
1050 /// * A raw pointer is dereferenced and transmuted to `&[u8]`;
1051 /// * The slice is not checked to see whether it contains valid UTF-8.
1052 #[deprecated = "renamed to String::from_raw_buf_len"]
1053 pub unsafe fn from_buf_len(buf: *const u8, len: uint) -> String {
1054 String::from_raw_buf_len(buf, len)
1057 /// Creates a `String` from a null-terminated `*const u8` buffer.
1059 /// This function is unsafe because we dereference memory until we find the NUL character,
1060 /// which is not guaranteed to be present. Additionally, the slice is not checked to see
1061 /// whether it contains valid UTF-8
1062 #[deprecated = "renamed to String::from_raw_buf"]
1063 pub unsafe fn from_buf(buf: *const u8) -> String {
1064 String::from_raw_buf(buf)
1067 /// Converts a vector of bytes to a new `String` without checking if
1068 /// it contains valid UTF-8. This is unsafe because it assumes that
1069 /// the UTF-8-ness of the vector has already been validated.
1071 #[deprecated = "renamed to String::from_utf8_unchecked"]
1072 pub unsafe fn from_utf8(bytes: Vec<u8>) -> String {
1073 String::from_utf8_unchecked(bytes)
1077 /// A clone-on-write string
1079 pub type CowString<'a> = Cow<'a, String, str>;
1081 #[allow(deprecated)]
1082 impl<'a> Str for CowString<'a> {
1084 fn as_slice<'b>(&'b self) -> &'b str {
1089 impl fmt::Writer for String {
1090 fn write_str(&mut self, s: &str) -> fmt::Result {
1103 use super::as_string;
1106 fn test_as_string() {
1108 assert_eq!(x, as_string(x).as_slice());
1112 fn test_from_str() {
1113 let owned: Option<::std::string::String> = from_str("string");
1114 assert_eq!(owned.as_ref().map(|s| s.as_slice()), Some("string"));
1118 fn test_from_utf8() {
1119 let xs = b"hello".to_vec();
1120 assert_eq!(String::from_utf8(xs).unwrap(),
1121 String::from_str("hello"));
1123 let xs = "ศไทย中华Việt Nam".as_bytes().to_vec();
1124 assert_eq!(String::from_utf8(xs).unwrap(),
1125 String::from_str("ศไทย中华Việt Nam"));
1127 let xs = b"hello\xFF".to_vec();
1128 let err = String::from_utf8(xs).err().unwrap();
1129 assert_eq!(err.utf8_error(), Utf8Error::TooShort);
1130 assert_eq!(err.into_bytes(), b"hello\xff".to_vec());
1134 fn test_from_utf8_lossy() {
1136 let ys: str::CowString = "hello".into_cow();
1137 assert_eq!(String::from_utf8_lossy(xs), ys);
1139 let xs = "ศไทย中华Việt Nam".as_bytes();
1140 let ys: str::CowString = "ศไทย中华Việt Nam".into_cow();
1141 assert_eq!(String::from_utf8_lossy(xs), ys);
1143 let xs = b"Hello\xC2 There\xFF Goodbye";
1144 assert_eq!(String::from_utf8_lossy(xs),
1145 String::from_str("Hello\u{FFFD} There\u{FFFD} Goodbye").into_cow());
1147 let xs = b"Hello\xC0\x80 There\xE6\x83 Goodbye";
1148 assert_eq!(String::from_utf8_lossy(xs),
1149 String::from_str("Hello\u{FFFD}\u{FFFD} There\u{FFFD} Goodbye").into_cow());
1151 let xs = b"\xF5foo\xF5\x80bar";
1152 assert_eq!(String::from_utf8_lossy(xs),
1153 String::from_str("\u{FFFD}foo\u{FFFD}\u{FFFD}bar").into_cow());
1155 let xs = b"\xF1foo\xF1\x80bar\xF1\x80\x80baz";
1156 assert_eq!(String::from_utf8_lossy(xs),
1157 String::from_str("\u{FFFD}foo\u{FFFD}bar\u{FFFD}baz").into_cow());
1159 let xs = b"\xF4foo\xF4\x80bar\xF4\xBFbaz";
1160 assert_eq!(String::from_utf8_lossy(xs),
1161 String::from_str("\u{FFFD}foo\u{FFFD}bar\u{FFFD}\u{FFFD}baz").into_cow());
1163 let xs = b"\xF0\x80\x80\x80foo\xF0\x90\x80\x80bar";
1164 assert_eq!(String::from_utf8_lossy(xs), String::from_str("\u{FFFD}\u{FFFD}\u{FFFD}\u{FFFD}\
1165 foo\u{10000}bar").into_cow());
1168 let xs = b"\xED\xA0\x80foo\xED\xBF\xBFbar";
1169 assert_eq!(String::from_utf8_lossy(xs), String::from_str("\u{FFFD}\u{FFFD}\u{FFFD}foo\
1170 \u{FFFD}\u{FFFD}\u{FFFD}bar").into_cow());
1174 fn test_from_utf16() {
1176 [(String::from_str("𐍅𐌿𐌻𐍆𐌹𐌻𐌰\n"),
1177 vec![0xd800_u16, 0xdf45_u16, 0xd800_u16, 0xdf3f_u16,
1178 0xd800_u16, 0xdf3b_u16, 0xd800_u16, 0xdf46_u16,
1179 0xd800_u16, 0xdf39_u16, 0xd800_u16, 0xdf3b_u16,
1180 0xd800_u16, 0xdf30_u16, 0x000a_u16]),
1182 (String::from_str("𐐒𐑉𐐮𐑀𐐲𐑋 𐐏𐐲𐑍\n"),
1183 vec![0xd801_u16, 0xdc12_u16, 0xd801_u16,
1184 0xdc49_u16, 0xd801_u16, 0xdc2e_u16, 0xd801_u16,
1185 0xdc40_u16, 0xd801_u16, 0xdc32_u16, 0xd801_u16,
1186 0xdc4b_u16, 0x0020_u16, 0xd801_u16, 0xdc0f_u16,
1187 0xd801_u16, 0xdc32_u16, 0xd801_u16, 0xdc4d_u16,
1190 (String::from_str("𐌀𐌖𐌋𐌄𐌑𐌉·𐌌𐌄𐌕𐌄𐌋𐌉𐌑\n"),
1191 vec![0xd800_u16, 0xdf00_u16, 0xd800_u16, 0xdf16_u16,
1192 0xd800_u16, 0xdf0b_u16, 0xd800_u16, 0xdf04_u16,
1193 0xd800_u16, 0xdf11_u16, 0xd800_u16, 0xdf09_u16,
1194 0x00b7_u16, 0xd800_u16, 0xdf0c_u16, 0xd800_u16,
1195 0xdf04_u16, 0xd800_u16, 0xdf15_u16, 0xd800_u16,
1196 0xdf04_u16, 0xd800_u16, 0xdf0b_u16, 0xd800_u16,
1197 0xdf09_u16, 0xd800_u16, 0xdf11_u16, 0x000a_u16 ]),
1199 (String::from_str("𐒋𐒘𐒈𐒑𐒛𐒒 𐒕𐒓 𐒈𐒚𐒍 𐒏𐒜𐒒𐒖𐒆 𐒕𐒆\n"),
1200 vec![0xd801_u16, 0xdc8b_u16, 0xd801_u16, 0xdc98_u16,
1201 0xd801_u16, 0xdc88_u16, 0xd801_u16, 0xdc91_u16,
1202 0xd801_u16, 0xdc9b_u16, 0xd801_u16, 0xdc92_u16,
1203 0x0020_u16, 0xd801_u16, 0xdc95_u16, 0xd801_u16,
1204 0xdc93_u16, 0x0020_u16, 0xd801_u16, 0xdc88_u16,
1205 0xd801_u16, 0xdc9a_u16, 0xd801_u16, 0xdc8d_u16,
1206 0x0020_u16, 0xd801_u16, 0xdc8f_u16, 0xd801_u16,
1207 0xdc9c_u16, 0xd801_u16, 0xdc92_u16, 0xd801_u16,
1208 0xdc96_u16, 0xd801_u16, 0xdc86_u16, 0x0020_u16,
1209 0xd801_u16, 0xdc95_u16, 0xd801_u16, 0xdc86_u16,
1211 // Issue #12318, even-numbered non-BMP planes
1212 (String::from_str("\u{20000}"),
1213 vec![0xD840, 0xDC00])];
1215 for p in pairs.iter() {
1216 let (s, u) = (*p).clone();
1217 let s_as_utf16 = s.utf16_units().collect::<Vec<u16>>();
1218 let u_as_string = String::from_utf16(u.as_slice()).unwrap();
1220 assert!(::unicode::str::is_utf16(u.as_slice()));
1221 assert_eq!(s_as_utf16, u);
1223 assert_eq!(u_as_string, s);
1224 assert_eq!(String::from_utf16_lossy(u.as_slice()), s);
1226 assert_eq!(String::from_utf16(s_as_utf16.as_slice()).unwrap(), s);
1227 assert_eq!(u_as_string.utf16_units().collect::<Vec<u16>>(), u);
1232 fn test_utf16_invalid() {
1233 // completely positive cases tested above.
1235 assert!(String::from_utf16(&[0xD800]).is_err());
1237 assert!(String::from_utf16(&[0xD800, 0xD800]).is_err());
1240 assert!(String::from_utf16(&[0x0061, 0xDC00]).is_err());
1243 assert!(String::from_utf16(&[0xD800, 0xd801, 0xdc8b, 0xD800]).is_err());
1247 fn test_from_utf16_lossy() {
1248 // completely positive cases tested above.
1250 assert_eq!(String::from_utf16_lossy(&[0xD800]), String::from_str("\u{FFFD}"));
1252 assert_eq!(String::from_utf16_lossy(&[0xD800, 0xD800]),
1253 String::from_str("\u{FFFD}\u{FFFD}"));
1256 assert_eq!(String::from_utf16_lossy(&[0x0061, 0xDC00]), String::from_str("a\u{FFFD}"));
1259 assert_eq!(String::from_utf16_lossy(&[0xD800, 0xd801, 0xdc8b, 0xD800]),
1260 String::from_str("\u{FFFD}𐒋\u{FFFD}"));
1264 fn test_from_buf_len() {
1266 let a = vec![65u8, 65, 65, 65, 65, 65, 65, 0];
1267 assert_eq!(super::raw::from_buf_len(a.as_ptr(), 3), String::from_str("AAA"));
1272 fn test_from_buf() {
1274 let a = vec![65, 65, 65, 65, 65, 65, 65, 0];
1276 let c = super::raw::from_buf(b);
1277 assert_eq!(c, String::from_str("AAAAAAA"));
1282 fn test_push_bytes() {
1283 let mut s = String::from_str("ABC");
1285 let mv = s.as_mut_vec();
1286 mv.push_all(&[b'D']);
1288 assert_eq!(s, "ABCD");
1292 fn test_push_str() {
1293 let mut s = String::new();
1295 assert_eq!(s.slice_from(0), "");
1297 assert_eq!(s.slice_from(0), "abc");
1298 s.push_str("ประเทศไทย中华Việt Nam");
1299 assert_eq!(s.slice_from(0), "abcประเทศไทย中华Việt Nam");
1304 let mut data = String::from_str("ประเทศไทย中");
1306 data.push('b'); // 1 byte
1307 data.push('¢'); // 2 byte
1308 data.push('€'); // 3 byte
1309 data.push('𤭢'); // 4 byte
1310 assert_eq!(data, "ประเทศไทย中华b¢€𤭢");
1315 let mut data = String::from_str("ประเทศไทย中华b¢€𤭢");
1316 assert_eq!(data.pop().unwrap(), '𤭢'); // 4 bytes
1317 assert_eq!(data.pop().unwrap(), '€'); // 3 bytes
1318 assert_eq!(data.pop().unwrap(), '¢'); // 2 bytes
1319 assert_eq!(data.pop().unwrap(), 'b'); // 1 bytes
1320 assert_eq!(data.pop().unwrap(), '华');
1321 assert_eq!(data, "ประเทศไทย中");
1325 fn test_str_truncate() {
1326 let mut s = String::from_str("12345");
1328 assert_eq!(s, "12345");
1330 assert_eq!(s, "123");
1334 let mut s = String::from_str("12345");
1338 let p_ = s.as_ptr();
1344 fn test_str_truncate_invalid_len() {
1345 let mut s = String::from_str("12345");
1351 fn test_str_truncate_split_codepoint() {
1352 let mut s = String::from_str("\u{FC}"); // ü
1357 fn test_str_clear() {
1358 let mut s = String::from_str("12345");
1360 assert_eq!(s.len(), 0);
1366 let a = String::from_str("12345");
1369 assert_eq!(b.len(), 7);
1370 assert_eq!(b, "1234522");
1375 let mut s = "ศไทย中华Việt Nam; foobar".to_string();;
1376 assert_eq!(s.remove(0), 'ศ');
1377 assert_eq!(s.len(), 33);
1378 assert_eq!(s, "ไทย中华Việt Nam; foobar");
1379 assert_eq!(s.remove(17), 'ệ');
1380 assert_eq!(s, "ไทย中华Vit Nam; foobar");
1383 #[test] #[should_fail]
1385 "ศ".to_string().remove(1);
1390 let mut s = "foobar".to_string();
1392 assert_eq!(s, "ệfoobar");
1394 assert_eq!(s, "ệfooยbar");
1397 #[test] #[should_fail] fn insert_bad1() { "".to_string().insert(1, 't'); }
1398 #[test] #[should_fail] fn insert_bad2() { "ệ".to_string().insert(1, 't'); }
1402 let s = "foobar".to_string();
1403 assert_eq!("foobar", s[]);
1404 assert_eq!("foo", s[..3]);
1405 assert_eq!("bar", s[3..]);
1406 assert_eq!("oob", s[1..4]);
1410 fn test_simple_types() {
1411 assert_eq!(1i.to_string(), "1");
1412 assert_eq!((-1i).to_string(), "-1");
1413 assert_eq!(200u.to_string(), "200");
1414 assert_eq!(2u8.to_string(), "2");
1415 assert_eq!(true.to_string(), "true");
1416 assert_eq!(false.to_string(), "false");
1417 assert_eq!(().to_string(), "()");
1418 assert_eq!(("hi".to_string()).to_string(), "hi");
1423 let x: Vec<int> = vec![];
1424 assert_eq!(x.to_string(), "[]");
1425 assert_eq!((vec![1i]).to_string(), "[1]");
1426 assert_eq!((vec![1i, 2, 3]).to_string(), "[1, 2, 3]");
1427 assert!((vec![vec![], vec![1i], vec![1i, 1]]).to_string() ==
1428 "[[], [1], [1, 1]]");
1432 fn test_from_iterator() {
1433 let s = "ศไทย中华Việt Nam".to_string();
1437 let a: String = s.chars().collect();
1440 let mut b = t.to_string();
1441 b.extend(u.chars());
1444 let c: String = vec![t, u].into_iter().collect();
1447 let mut d = t.to_string();
1448 d.extend(vec![u].into_iter());
1453 fn bench_with_capacity(b: &mut Bencher) {
1455 String::with_capacity(100)
1460 fn bench_push_str(b: &mut Bencher) {
1461 let s = "ศไทย中华Việt Nam; Mary had a little lamb, Little lamb";
1463 let mut r = String::new();
1468 const REPETITIONS: u64 = 10_000;
1471 fn bench_push_str_one_byte(b: &mut Bencher) {
1472 b.bytes = REPETITIONS;
1474 let mut r = String::new();
1475 for _ in range(0, REPETITIONS) {
1482 fn bench_push_char_one_byte(b: &mut Bencher) {
1483 b.bytes = REPETITIONS;
1485 let mut r = String::new();
1486 for _ in range(0, REPETITIONS) {
1493 fn bench_push_char_two_bytes(b: &mut Bencher) {
1494 b.bytes = REPETITIONS * 2;
1496 let mut r = String::new();
1497 for _ in range(0, REPETITIONS) {
1504 fn from_utf8_lossy_100_ascii(b: &mut Bencher) {
1505 let s = b"Hello there, the quick brown fox jumped over the lazy dog! \
1506 Lorem ipsum dolor sit amet, consectetur. ";
1508 assert_eq!(100, s.len());
1510 let _ = String::from_utf8_lossy(s);
1515 fn from_utf8_lossy_100_multibyte(b: &mut Bencher) {
1516 let s = "𐌀𐌖𐌋𐌄𐌑𐌉ปรدولة الكويتทศไทย中华𐍅𐌿𐌻𐍆𐌹𐌻𐌰".as_bytes();
1517 assert_eq!(100, s.len());
1519 let _ = String::from_utf8_lossy(s);
1524 fn from_utf8_lossy_invalid(b: &mut Bencher) {
1525 let s = b"Hello\xC0\x80 There\xE6\x83 Goodbye";
1527 let _ = String::from_utf8_lossy(s);
1532 fn from_utf8_lossy_100_invalid(b: &mut Bencher) {
1533 let s = Vec::from_elem(100, 0xF5u8);
1535 let _ = String::from_utf8_lossy(s.as_slice());