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};
20 use core::default::Default;
23 use core::iter::FromIterator;
25 use core::ops::{self, Deref, Add, Index};
27 use core::raw::Slice as RawSlice;
28 use unicode::str as unicode_str;
29 use unicode::str::Utf16Item;
31 use str::{self, CharRange, FromStr, Utf8Error};
32 use vec::{DerefVec, Vec, as_vec};
34 /// A growable string stored as a UTF-8 encoded buffer.
35 #[derive(Clone, PartialOrd, Eq, Ord)]
41 /// A possible error value from the `String::from_utf8` function.
43 pub struct FromUtf8Error {
48 /// A possible error value from the `String::from_utf16` function.
50 #[allow(missing_copy_implementations)]
51 pub struct FromUtf16Error(());
54 /// Creates a new string buffer initialized with the empty string.
59 /// let mut s = String::new();
63 pub fn new() -> String {
69 /// Creates a new string buffer with the given capacity.
70 /// The string will be able to hold exactly `capacity` bytes without
71 /// reallocating. If `capacity` is 0, the string will not allocate.
76 /// let mut s = String::with_capacity(10);
80 pub fn with_capacity(capacity: uint) -> String {
82 vec: Vec::with_capacity(capacity),
86 /// Creates a new string buffer from the given string.
91 /// let s = String::from_str("hello");
92 /// assert_eq!(s.as_slice(), "hello");
95 #[unstable = "needs investigation to see if to_string() can match perf"]
96 pub fn from_str(string: &str) -> String {
97 String { vec: ::slice::SliceExt::to_vec(string.as_bytes()) }
100 /// Returns the vector as a string buffer, if possible, taking care not to
105 /// If the given vector is not valid UTF-8, then the original vector and the
106 /// corresponding error is returned.
111 /// use std::str::Utf8Error;
113 /// let hello_vec = vec![104, 101, 108, 108, 111];
114 /// let s = String::from_utf8(hello_vec).unwrap();
115 /// assert_eq!(s, "hello");
117 /// let invalid_vec = vec![240, 144, 128];
118 /// let s = String::from_utf8(invalid_vec).err().unwrap();
119 /// assert_eq!(s.utf8_error(), Utf8Error::TooShort);
120 /// assert_eq!(s.into_bytes(), vec![240, 144, 128]);
124 pub fn from_utf8(vec: Vec<u8>) -> Result<String, FromUtf8Error> {
125 match str::from_utf8(vec.as_slice()) {
126 Ok(..) => Ok(String { vec: vec }),
127 Err(e) => Err(FromUtf8Error { bytes: vec, error: e })
131 /// Converts a vector of bytes to a new UTF-8 string.
132 /// Any invalid UTF-8 sequences are replaced with U+FFFD REPLACEMENT CHARACTER.
137 /// let input = b"Hello \xF0\x90\x80World";
138 /// let output = String::from_utf8_lossy(input);
139 /// assert_eq!(output.as_slice(), "Hello \u{FFFD}World");
142 pub fn from_utf8_lossy<'a>(v: &'a [u8]) -> CowString<'a> {
144 match str::from_utf8(v) {
145 Ok(s) => return Cow::Borrowed(s),
147 if let Utf8Error::InvalidByte(firstbad) = e {
153 static TAG_CONT_U8: u8 = 128u8;
154 static REPLACEMENT: &'static [u8] = b"\xEF\xBF\xBD"; // U+FFFD in UTF-8
156 fn unsafe_get(xs: &[u8], i: uint) -> u8 {
157 unsafe { *xs.get_unchecked(i) }
159 fn safe_get(xs: &[u8], i: uint, total: uint) -> u8 {
167 let mut res = String::with_capacity(total);
171 res.as_mut_vec().push_all(&v[..i])
175 // subseqidx is the index of the first byte of the subsequence we're looking at.
176 // It's used to copy a bunch of contiguous good codepoints at once instead of copying
178 let mut subseqidx = i;
182 let byte = unsafe_get(v, i);
185 macro_rules! error { () => ({
188 res.as_mut_vec().push_all(&v[subseqidx..i_]);
191 res.as_mut_vec().push_all(REPLACEMENT);
196 // subseqidx handles this
198 let w = unicode_str::utf8_char_width(byte);
202 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
209 match (byte, safe_get(v, i, total)) {
210 (0xE0 , 0xA0 ... 0xBF) => (),
211 (0xE1 ... 0xEC, 0x80 ... 0xBF) => (),
212 (0xED , 0x80 ... 0x9F) => (),
213 (0xEE ... 0xEF, 0x80 ... 0xBF) => (),
220 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
227 match (byte, safe_get(v, i, total)) {
228 (0xF0 , 0x90 ... 0xBF) => (),
229 (0xF1 ... 0xF3, 0x80 ... 0xBF) => (),
230 (0xF4 , 0x80 ... 0x8F) => (),
237 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
242 if safe_get(v, i, total) & 192u8 != TAG_CONT_U8 {
255 if subseqidx < total {
257 res.as_mut_vec().push_all(&v[subseqidx..total])
263 /// Decode a UTF-16 encoded vector `v` into a `String`, returning `None`
264 /// if `v` contains any invalid data.
270 /// let mut v = &mut [0xD834, 0xDD1E, 0x006d, 0x0075,
271 /// 0x0073, 0x0069, 0x0063];
272 /// assert_eq!(String::from_utf16(v).unwrap(),
273 /// "𝄞music".to_string());
275 /// // 𝄞mu<invalid>ic
277 /// assert!(String::from_utf16(v).is_err());
280 pub fn from_utf16(v: &[u16]) -> Result<String, FromUtf16Error> {
281 let mut s = String::with_capacity(v.len());
282 for c in unicode_str::utf16_items(v) {
284 Utf16Item::ScalarValue(c) => s.push(c),
285 Utf16Item::LoneSurrogate(_) => return Err(FromUtf16Error(())),
291 /// Decode a UTF-16 encoded vector `v` into a string, replacing
292 /// invalid data with the replacement character (U+FFFD).
297 /// // 𝄞mus<invalid>ic<invalid>
298 /// let v = &[0xD834, 0xDD1E, 0x006d, 0x0075,
299 /// 0x0073, 0xDD1E, 0x0069, 0x0063,
302 /// assert_eq!(String::from_utf16_lossy(v),
303 /// "𝄞mus\u{FFFD}ic\u{FFFD}".to_string());
307 pub fn from_utf16_lossy(v: &[u16]) -> String {
308 unicode_str::utf16_items(v).map(|c| c.to_char_lossy()).collect()
311 /// Creates a new `String` from a length, capacity, and pointer.
313 /// This is unsafe because:
314 /// * We call `Vec::from_raw_parts` to get a `Vec<u8>`;
315 /// * We assume that the `Vec` contains valid UTF-8.
318 pub unsafe fn from_raw_parts(buf: *mut u8, length: uint, capacity: uint) -> String {
320 vec: Vec::from_raw_parts(buf, length, capacity),
324 /// Converts a vector of bytes to a new `String` without checking if
325 /// it contains valid UTF-8. This is unsafe because it assumes that
326 /// the UTF-8-ness of the vector has already been validated.
329 pub unsafe fn from_utf8_unchecked(bytes: Vec<u8>) -> String {
330 String { vec: bytes }
333 /// Return the underlying byte buffer, encoded as UTF-8.
338 /// let s = String::from_str("hello");
339 /// let bytes = s.into_bytes();
340 /// assert_eq!(bytes, vec![104, 101, 108, 108, 111]);
344 pub fn into_bytes(self) -> Vec<u8> {
348 /// Pushes the given string onto this string buffer.
353 /// let mut s = String::from_str("foo");
354 /// s.push_str("bar");
355 /// assert_eq!(s.as_slice(), "foobar");
359 pub fn push_str(&mut self, string: &str) {
360 self.vec.push_all(string.as_bytes())
363 /// Returns the number of bytes that this string buffer can hold without
369 /// let s = String::with_capacity(10);
370 /// assert!(s.capacity() >= 10);
374 pub fn capacity(&self) -> uint {
378 /// Reserves capacity for at least `additional` more bytes to be inserted
379 /// in the given `String`. The collection may reserve more space to avoid
380 /// frequent reallocations.
384 /// Panics if the new capacity overflows `uint`.
389 /// let mut s = String::new();
391 /// assert!(s.capacity() >= 10);
395 pub fn reserve(&mut self, additional: uint) {
396 self.vec.reserve(additional)
399 /// Reserves the minimum capacity for exactly `additional` more bytes to be
400 /// inserted in the given `String`. Does nothing if the capacity is already
403 /// Note that the allocator may give the collection more space than it
404 /// requests. Therefore capacity can not be relied upon to be precisely
405 /// minimal. Prefer `reserve` if future insertions are expected.
409 /// Panics if the new capacity overflows `uint`.
414 /// let mut s = String::new();
416 /// assert!(s.capacity() >= 10);
420 pub fn reserve_exact(&mut self, additional: uint) {
421 self.vec.reserve_exact(additional)
424 /// Shrinks the capacity of this string buffer to match its length.
429 /// let mut s = String::from_str("foo");
431 /// assert!(s.capacity() >= 100);
432 /// s.shrink_to_fit();
433 /// assert_eq!(s.capacity(), 3);
437 pub fn shrink_to_fit(&mut self) {
438 self.vec.shrink_to_fit()
441 /// Adds the given character to the end of the string.
446 /// let mut s = String::from_str("abc");
450 /// assert_eq!(s.as_slice(), "abc123");
454 pub fn push(&mut self, ch: char) {
455 if (ch as u32) < 0x80 {
456 self.vec.push(ch as u8);
460 let cur_len = self.len();
461 // This may use up to 4 bytes.
465 // Attempt to not use an intermediate buffer by just pushing bytes
466 // directly onto this string.
467 let slice = RawSlice {
468 data: self.vec.as_ptr().offset(cur_len as int),
471 let used = ch.encode_utf8(mem::transmute(slice)).unwrap_or(0);
472 self.vec.set_len(cur_len + used);
476 /// Works with the underlying buffer as a byte slice.
481 /// let s = String::from_str("hello");
482 /// let b: &[_] = &[104, 101, 108, 108, 111];
483 /// assert_eq!(s.as_bytes(), b);
487 pub fn as_bytes<'a>(&'a self) -> &'a [u8] {
491 /// Shortens a string to the specified length.
495 /// Panics if `new_len` > current length,
496 /// or if `new_len` is not a character boundary.
501 /// let mut s = String::from_str("hello");
503 /// assert_eq!(s.as_slice(), "he");
507 pub fn truncate(&mut self, new_len: uint) {
508 assert!(self.is_char_boundary(new_len));
509 self.vec.truncate(new_len)
512 /// Removes the last character from the string buffer and returns it.
513 /// Returns `None` if this string buffer is empty.
518 /// let mut s = String::from_str("foo");
519 /// assert_eq!(s.pop(), Some('o'));
520 /// assert_eq!(s.pop(), Some('o'));
521 /// assert_eq!(s.pop(), Some('f'));
522 /// assert_eq!(s.pop(), None);
526 pub fn pop(&mut self) -> Option<char> {
527 let len = self.len();
532 let CharRange {ch, next} = self.char_range_at_reverse(len);
534 self.vec.set_len(next);
539 /// Removes the character from the string buffer at byte position `idx` and
544 /// This is an O(n) operation as it requires copying every element in the
549 /// If `idx` does not lie on a character boundary, or if it is out of
550 /// bounds, then this function will panic.
555 /// let mut s = String::from_str("foo");
556 /// assert_eq!(s.remove(0), 'f');
557 /// assert_eq!(s.remove(1), 'o');
558 /// assert_eq!(s.remove(0), 'o');
562 pub fn remove(&mut self, idx: uint) -> char {
563 let len = self.len();
566 let CharRange { ch, next } = self.char_range_at(idx);
568 ptr::copy_memory(self.vec.as_mut_ptr().offset(idx as int),
569 self.vec.as_ptr().offset(next as int),
571 self.vec.set_len(len - (next - idx));
576 /// Insert a character into the string buffer at byte position `idx`.
580 /// This is an O(n) operation as it requires copying every element in the
585 /// If `idx` does not lie on a character boundary or is out of bounds, then
586 /// this function will panic.
589 pub fn insert(&mut self, idx: uint, ch: char) {
590 let len = self.len();
592 assert!(self.is_char_boundary(idx));
594 let mut bits = [0; 4];
595 let amt = ch.encode_utf8(&mut bits).unwrap();
598 ptr::copy_memory(self.vec.as_mut_ptr().offset((idx + amt) as int),
599 self.vec.as_ptr().offset(idx as int),
601 ptr::copy_memory(self.vec.as_mut_ptr().offset(idx as int),
604 self.vec.set_len(len + amt);
608 /// Views the string buffer as a mutable sequence of bytes.
610 /// This is unsafe because it does not check
611 /// to ensure that the resulting string will be valid UTF-8.
616 /// let mut s = String::from_str("hello");
618 /// let vec = s.as_mut_vec();
619 /// assert!(vec == &mut vec![104, 101, 108, 108, 111]);
622 /// assert_eq!(s.as_slice(), "olleh");
626 pub unsafe fn as_mut_vec<'a>(&'a mut self) -> &'a mut Vec<u8> {
630 /// Return the number of bytes in this string.
635 /// let a = "foo".to_string();
636 /// assert_eq!(a.len(), 3);
640 pub fn len(&self) -> uint { self.vec.len() }
642 /// Returns true if the string contains no bytes
647 /// let mut v = String::new();
648 /// assert!(v.is_empty());
650 /// assert!(!v.is_empty());
654 pub fn is_empty(&self) -> bool { self.len() == 0 }
656 /// Truncates the string, returning it to 0 length.
661 /// let mut s = "foo".to_string();
663 /// assert!(s.is_empty());
667 pub fn clear(&mut self) {
673 /// Consume this error, returning the bytes that were attempted to make a
676 pub fn into_bytes(self) -> Vec<u8> { self.bytes }
678 /// Access the underlying UTF8-error that was the cause of this error.
680 pub fn utf8_error(&self) -> Utf8Error { self.error }
683 impl fmt::Show for FromUtf8Error {
684 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
685 fmt::String::fmt(self, f)
690 impl fmt::String for FromUtf8Error {
691 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
692 fmt::String::fmt(&self.error, f)
696 impl fmt::Show for FromUtf16Error {
697 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
698 fmt::String::fmt(self, f)
703 impl fmt::String for FromUtf16Error {
704 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
705 fmt::String::fmt("invalid utf-16: lone surrogate found", f)
710 impl FromIterator<char> for String {
711 fn from_iter<I:Iterator<Item=char>>(iterator: I) -> String {
712 let mut buf = String::new();
713 buf.extend(iterator);
719 impl<'a> FromIterator<&'a str> for String {
720 fn from_iter<I:Iterator<Item=&'a str>>(iterator: I) -> String {
721 let mut buf = String::new();
722 buf.extend(iterator);
727 #[unstable = "waiting on Extend stabilization"]
728 impl Extend<char> for String {
729 fn extend<I:Iterator<Item=char>>(&mut self, mut iterator: I) {
730 let (lower_bound, _) = iterator.size_hint();
731 self.reserve(lower_bound);
738 #[unstable = "waiting on Extend stabilization"]
739 impl<'a> Extend<&'a str> for String {
740 fn extend<I: Iterator<Item=&'a str>>(&mut self, mut iterator: I) {
741 // A guess that at least one byte per iterator element will be needed.
742 let (lower_bound, _) = iterator.size_hint();
743 self.reserve(lower_bound);
751 impl PartialEq for String {
753 fn eq(&self, other: &String) -> bool { PartialEq::eq(&**self, &**other) }
755 fn ne(&self, other: &String) -> bool { PartialEq::ne(&**self, &**other) }
758 macro_rules! impl_eq {
759 ($lhs:ty, $rhs: ty) => {
761 impl<'a> PartialEq<$rhs> for $lhs {
763 fn eq(&self, other: &$rhs) -> bool { PartialEq::eq(&**self, &**other) }
765 fn ne(&self, other: &$rhs) -> bool { PartialEq::ne(&**self, &**other) }
769 impl<'a> PartialEq<$lhs> for $rhs {
771 fn eq(&self, other: &$lhs) -> bool { PartialEq::eq(&**self, &**other) }
773 fn ne(&self, other: &$lhs) -> bool { PartialEq::ne(&**self, &**other) }
779 impl_eq! { String, &'a str }
780 impl_eq! { CowString<'a>, String }
783 impl<'a, 'b> PartialEq<&'b str> for CowString<'a> {
785 fn eq(&self, other: &&'b str) -> bool { PartialEq::eq(&**self, &**other) }
787 fn ne(&self, other: &&'b str) -> bool { PartialEq::ne(&**self, &**other) }
791 impl<'a, 'b> PartialEq<CowString<'a>> for &'b str {
793 fn eq(&self, other: &CowString<'a>) -> bool { PartialEq::eq(&**self, &**other) }
795 fn ne(&self, other: &CowString<'a>) -> bool { PartialEq::ne(&**self, &**other) }
798 #[unstable = "waiting on Str stabilization"]
799 impl Str for String {
802 fn as_slice<'a>(&'a self) -> &'a str {
803 unsafe { mem::transmute(self.vec.as_slice()) }
808 impl Default for String {
811 fn default() -> String {
817 impl fmt::String for String {
819 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
820 fmt::String::fmt(&**self, f)
824 #[unstable = "waiting on fmt stabilization"]
825 impl fmt::Show for String {
827 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
828 fmt::Show::fmt(&**self, f)
832 #[unstable = "waiting on Hash stabilization"]
833 impl<H: hash::Writer + hash::Hasher> hash::Hash<H> for String {
835 fn hash(&self, hasher: &mut H) {
836 (**self).hash(hasher)
840 #[unstable = "recent addition, needs more experience"]
841 impl<'a> Add<&'a str> for String {
842 type Output = String;
845 fn add(mut self, other: &str) -> String {
846 self.push_str(other);
851 impl ops::Index<ops::Range<uint>> for String {
854 fn index(&self, index: &ops::Range<uint>) -> &str {
858 impl ops::Index<ops::RangeTo<uint>> for String {
861 fn index(&self, index: &ops::RangeTo<uint>) -> &str {
865 impl ops::Index<ops::RangeFrom<uint>> for String {
868 fn index(&self, index: &ops::RangeFrom<uint>) -> &str {
872 impl ops::Index<ops::FullRange> for String {
875 fn index(&self, _index: &ops::FullRange) -> &str {
876 unsafe { mem::transmute(self.vec.as_slice()) }
881 impl ops::Deref for String {
885 fn deref<'a>(&'a self) -> &'a str {
886 unsafe { mem::transmute(&self.vec[]) }
890 /// Wrapper type providing a `&String` reference via `Deref`.
892 pub struct DerefString<'a> {
896 impl<'a> Deref for DerefString<'a> {
897 type Target = String;
900 fn deref<'b>(&'b self) -> &'b String {
901 unsafe { mem::transmute(&*self.x) }
905 /// Convert a string slice to a wrapper type providing a `&String` reference.
910 /// use std::string::as_string;
912 /// fn string_consumer(s: String) {
913 /// assert_eq!(s, "foo".to_string());
916 /// let string = as_string("foo").clone();
917 /// string_consumer(string);
920 pub fn as_string<'a>(x: &'a str) -> DerefString<'a> {
921 DerefString { x: as_vec(x.as_bytes()) }
924 impl FromStr for String {
926 fn from_str(s: &str) -> Option<String> {
927 Some(String::from_str(s))
931 /// A generic trait for converting a value to a string
933 /// Converts the value of `self` to an owned string
934 fn to_string(&self) -> String;
937 impl<T: fmt::String + ?Sized> ToString for T {
939 fn to_string(&self) -> String {
940 use core::fmt::Writer;
941 let mut buf = String::new();
942 let _ = buf.write_fmt(format_args!("{}", self));
948 impl IntoCow<'static, String, str> for String {
950 fn into_cow(self) -> CowString<'static> {
955 impl<'a> IntoCow<'a, String, str> for &'a str {
957 fn into_cow(self) -> CowString<'a> {
962 /// A clone-on-write string
964 pub type CowString<'a> = Cow<'a, String, str>;
966 impl<'a> Str for CowString<'a> {
968 fn as_slice<'b>(&'b self) -> &'b str {
973 impl fmt::Writer for String {
975 fn write_str(&mut self, s: &str) -> fmt::Result {
987 use core::iter::repeat;
988 use super::{as_string, CowString};
989 use core::ops::FullRange;
992 fn test_as_string() {
994 assert_eq!(x, as_string(x).as_slice());
999 let owned: Option<::std::string::String> = "string".parse();
1000 assert_eq!(owned.as_ref().map(|s| s.as_slice()), Some("string"));
1004 fn test_unsized_to_string() {
1005 let s: &str = "abc";
1006 let _: String = (*s).to_string();
1010 fn test_from_utf8() {
1011 let xs = b"hello".to_vec();
1012 assert_eq!(String::from_utf8(xs).unwrap(),
1013 String::from_str("hello"));
1015 let xs = "ศไทย中华Việt Nam".as_bytes().to_vec();
1016 assert_eq!(String::from_utf8(xs).unwrap(),
1017 String::from_str("ศไทย中华Việt Nam"));
1019 let xs = b"hello\xFF".to_vec();
1020 let err = String::from_utf8(xs).err().unwrap();
1021 assert_eq!(err.utf8_error(), Utf8Error::TooShort);
1022 assert_eq!(err.into_bytes(), b"hello\xff".to_vec());
1026 fn test_from_utf8_lossy() {
1028 let ys: CowString = "hello".into_cow();
1029 assert_eq!(String::from_utf8_lossy(xs), ys);
1031 let xs = "ศไทย中华Việt Nam".as_bytes();
1032 let ys: CowString = "ศไทย中华Việt Nam".into_cow();
1033 assert_eq!(String::from_utf8_lossy(xs), ys);
1035 let xs = b"Hello\xC2 There\xFF Goodbye";
1036 assert_eq!(String::from_utf8_lossy(xs),
1037 String::from_str("Hello\u{FFFD} There\u{FFFD} Goodbye").into_cow());
1039 let xs = b"Hello\xC0\x80 There\xE6\x83 Goodbye";
1040 assert_eq!(String::from_utf8_lossy(xs),
1041 String::from_str("Hello\u{FFFD}\u{FFFD} There\u{FFFD} Goodbye").into_cow());
1043 let xs = b"\xF5foo\xF5\x80bar";
1044 assert_eq!(String::from_utf8_lossy(xs),
1045 String::from_str("\u{FFFD}foo\u{FFFD}\u{FFFD}bar").into_cow());
1047 let xs = b"\xF1foo\xF1\x80bar\xF1\x80\x80baz";
1048 assert_eq!(String::from_utf8_lossy(xs),
1049 String::from_str("\u{FFFD}foo\u{FFFD}bar\u{FFFD}baz").into_cow());
1051 let xs = b"\xF4foo\xF4\x80bar\xF4\xBFbaz";
1052 assert_eq!(String::from_utf8_lossy(xs),
1053 String::from_str("\u{FFFD}foo\u{FFFD}bar\u{FFFD}\u{FFFD}baz").into_cow());
1055 let xs = b"\xF0\x80\x80\x80foo\xF0\x90\x80\x80bar";
1056 assert_eq!(String::from_utf8_lossy(xs), String::from_str("\u{FFFD}\u{FFFD}\u{FFFD}\u{FFFD}\
1057 foo\u{10000}bar").into_cow());
1060 let xs = b"\xED\xA0\x80foo\xED\xBF\xBFbar";
1061 assert_eq!(String::from_utf8_lossy(xs), String::from_str("\u{FFFD}\u{FFFD}\u{FFFD}foo\
1062 \u{FFFD}\u{FFFD}\u{FFFD}bar").into_cow());
1066 fn test_from_utf16() {
1068 [(String::from_str("𐍅𐌿𐌻𐍆𐌹𐌻𐌰\n"),
1069 vec![0xd800_u16, 0xdf45_u16, 0xd800_u16, 0xdf3f_u16,
1070 0xd800_u16, 0xdf3b_u16, 0xd800_u16, 0xdf46_u16,
1071 0xd800_u16, 0xdf39_u16, 0xd800_u16, 0xdf3b_u16,
1072 0xd800_u16, 0xdf30_u16, 0x000a_u16]),
1074 (String::from_str("𐐒𐑉𐐮𐑀𐐲𐑋 𐐏𐐲𐑍\n"),
1075 vec![0xd801_u16, 0xdc12_u16, 0xd801_u16,
1076 0xdc49_u16, 0xd801_u16, 0xdc2e_u16, 0xd801_u16,
1077 0xdc40_u16, 0xd801_u16, 0xdc32_u16, 0xd801_u16,
1078 0xdc4b_u16, 0x0020_u16, 0xd801_u16, 0xdc0f_u16,
1079 0xd801_u16, 0xdc32_u16, 0xd801_u16, 0xdc4d_u16,
1082 (String::from_str("𐌀𐌖𐌋𐌄𐌑𐌉·𐌌𐌄𐌕𐌄𐌋𐌉𐌑\n"),
1083 vec![0xd800_u16, 0xdf00_u16, 0xd800_u16, 0xdf16_u16,
1084 0xd800_u16, 0xdf0b_u16, 0xd800_u16, 0xdf04_u16,
1085 0xd800_u16, 0xdf11_u16, 0xd800_u16, 0xdf09_u16,
1086 0x00b7_u16, 0xd800_u16, 0xdf0c_u16, 0xd800_u16,
1087 0xdf04_u16, 0xd800_u16, 0xdf15_u16, 0xd800_u16,
1088 0xdf04_u16, 0xd800_u16, 0xdf0b_u16, 0xd800_u16,
1089 0xdf09_u16, 0xd800_u16, 0xdf11_u16, 0x000a_u16 ]),
1091 (String::from_str("𐒋𐒘𐒈𐒑𐒛𐒒 𐒕𐒓 𐒈𐒚𐒍 𐒏𐒜𐒒𐒖𐒆 𐒕𐒆\n"),
1092 vec![0xd801_u16, 0xdc8b_u16, 0xd801_u16, 0xdc98_u16,
1093 0xd801_u16, 0xdc88_u16, 0xd801_u16, 0xdc91_u16,
1094 0xd801_u16, 0xdc9b_u16, 0xd801_u16, 0xdc92_u16,
1095 0x0020_u16, 0xd801_u16, 0xdc95_u16, 0xd801_u16,
1096 0xdc93_u16, 0x0020_u16, 0xd801_u16, 0xdc88_u16,
1097 0xd801_u16, 0xdc9a_u16, 0xd801_u16, 0xdc8d_u16,
1098 0x0020_u16, 0xd801_u16, 0xdc8f_u16, 0xd801_u16,
1099 0xdc9c_u16, 0xd801_u16, 0xdc92_u16, 0xd801_u16,
1100 0xdc96_u16, 0xd801_u16, 0xdc86_u16, 0x0020_u16,
1101 0xd801_u16, 0xdc95_u16, 0xd801_u16, 0xdc86_u16,
1103 // Issue #12318, even-numbered non-BMP planes
1104 (String::from_str("\u{20000}"),
1105 vec![0xD840, 0xDC00])];
1107 for p in pairs.iter() {
1108 let (s, u) = (*p).clone();
1109 let s_as_utf16 = s.utf16_units().collect::<Vec<u16>>();
1110 let u_as_string = String::from_utf16(u.as_slice()).unwrap();
1112 assert!(::unicode::str::is_utf16(u.as_slice()));
1113 assert_eq!(s_as_utf16, u);
1115 assert_eq!(u_as_string, s);
1116 assert_eq!(String::from_utf16_lossy(u.as_slice()), s);
1118 assert_eq!(String::from_utf16(s_as_utf16.as_slice()).unwrap(), s);
1119 assert_eq!(u_as_string.utf16_units().collect::<Vec<u16>>(), u);
1124 fn test_utf16_invalid() {
1125 // completely positive cases tested above.
1127 assert!(String::from_utf16(&[0xD800]).is_err());
1129 assert!(String::from_utf16(&[0xD800, 0xD800]).is_err());
1132 assert!(String::from_utf16(&[0x0061, 0xDC00]).is_err());
1135 assert!(String::from_utf16(&[0xD800, 0xd801, 0xdc8b, 0xD800]).is_err());
1139 fn test_from_utf16_lossy() {
1140 // completely positive cases tested above.
1142 assert_eq!(String::from_utf16_lossy(&[0xD800]), String::from_str("\u{FFFD}"));
1144 assert_eq!(String::from_utf16_lossy(&[0xD800, 0xD800]),
1145 String::from_str("\u{FFFD}\u{FFFD}"));
1148 assert_eq!(String::from_utf16_lossy(&[0x0061, 0xDC00]), String::from_str("a\u{FFFD}"));
1151 assert_eq!(String::from_utf16_lossy(&[0xD800, 0xd801, 0xdc8b, 0xD800]),
1152 String::from_str("\u{FFFD}𐒋\u{FFFD}"));
1156 fn test_push_bytes() {
1157 let mut s = String::from_str("ABC");
1159 let mv = s.as_mut_vec();
1160 mv.push_all(&[b'D']);
1162 assert_eq!(s, "ABCD");
1166 fn test_push_str() {
1167 let mut s = String::new();
1169 assert_eq!(s.slice_from(0), "");
1171 assert_eq!(s.slice_from(0), "abc");
1172 s.push_str("ประเทศไทย中华Việt Nam");
1173 assert_eq!(s.slice_from(0), "abcประเทศไทย中华Việt Nam");
1178 let mut data = String::from_str("ประเทศไทย中");
1180 data.push('b'); // 1 byte
1181 data.push('¢'); // 2 byte
1182 data.push('€'); // 3 byte
1183 data.push('𤭢'); // 4 byte
1184 assert_eq!(data, "ประเทศไทย中华b¢€𤭢");
1189 let mut data = String::from_str("ประเทศไทย中华b¢€𤭢");
1190 assert_eq!(data.pop().unwrap(), '𤭢'); // 4 bytes
1191 assert_eq!(data.pop().unwrap(), '€'); // 3 bytes
1192 assert_eq!(data.pop().unwrap(), '¢'); // 2 bytes
1193 assert_eq!(data.pop().unwrap(), 'b'); // 1 bytes
1194 assert_eq!(data.pop().unwrap(), '华');
1195 assert_eq!(data, "ประเทศไทย中");
1199 fn test_str_truncate() {
1200 let mut s = String::from_str("12345");
1202 assert_eq!(s, "12345");
1204 assert_eq!(s, "123");
1208 let mut s = String::from_str("12345");
1212 let p_ = s.as_ptr();
1218 fn test_str_truncate_invalid_len() {
1219 let mut s = String::from_str("12345");
1225 fn test_str_truncate_split_codepoint() {
1226 let mut s = String::from_str("\u{FC}"); // ü
1231 fn test_str_clear() {
1232 let mut s = String::from_str("12345");
1234 assert_eq!(s.len(), 0);
1240 let a = String::from_str("12345");
1243 assert_eq!(b.len(), 7);
1244 assert_eq!(b, "1234522");
1249 let mut s = "ศไทย中华Việt Nam; foobar".to_string();;
1250 assert_eq!(s.remove(0), 'ศ');
1251 assert_eq!(s.len(), 33);
1252 assert_eq!(s, "ไทย中华Việt Nam; foobar");
1253 assert_eq!(s.remove(17), 'ệ');
1254 assert_eq!(s, "ไทย中华Vit Nam; foobar");
1257 #[test] #[should_fail]
1259 "ศ".to_string().remove(1);
1264 let mut s = "foobar".to_string();
1266 assert_eq!(s, "ệfoobar");
1268 assert_eq!(s, "ệfooยbar");
1271 #[test] #[should_fail] fn insert_bad1() { "".to_string().insert(1, 't'); }
1272 #[test] #[should_fail] fn insert_bad2() { "ệ".to_string().insert(1, 't'); }
1276 let s = "foobar".to_string();
1277 assert_eq!("foobar", &s[]);
1278 assert_eq!("foo", &s[..3]);
1279 assert_eq!("bar", &s[3..]);
1280 assert_eq!("oob", &s[1..4]);
1284 fn test_simple_types() {
1285 assert_eq!(1i.to_string(), "1");
1286 assert_eq!((-1i).to_string(), "-1");
1287 assert_eq!(200u.to_string(), "200");
1288 assert_eq!(2u8.to_string(), "2");
1289 assert_eq!(true.to_string(), "true");
1290 assert_eq!(false.to_string(), "false");
1291 assert_eq!(("hi".to_string()).to_string(), "hi");
1296 let x: Vec<int> = vec![];
1297 assert_eq!(format!("{:?}", x), "[]");
1298 assert_eq!(format!("{:?}", vec![1i]), "[1i]");
1299 assert_eq!(format!("{:?}", vec![1i, 2, 3]), "[1i, 2i, 3i]");
1300 assert!(format!("{:?}", vec![vec![], vec![1i], vec![1i, 1]]) ==
1301 "[[], [1i], [1i, 1i]]");
1305 fn test_from_iterator() {
1306 let s = "ศไทย中华Việt Nam".to_string();
1310 let a: String = s.chars().collect();
1313 let mut b = t.to_string();
1314 b.extend(u.chars());
1317 let c: String = vec![t, u].into_iter().collect();
1320 let mut d = t.to_string();
1321 d.extend(vec![u].into_iter());
1326 fn bench_with_capacity(b: &mut Bencher) {
1328 String::with_capacity(100)
1333 fn bench_push_str(b: &mut Bencher) {
1334 let s = "ศไทย中华Việt Nam; Mary had a little lamb, Little lamb";
1336 let mut r = String::new();
1341 const REPETITIONS: u64 = 10_000;
1344 fn bench_push_str_one_byte(b: &mut Bencher) {
1345 b.bytes = REPETITIONS;
1347 let mut r = String::new();
1348 for _ in range(0, REPETITIONS) {
1355 fn bench_push_char_one_byte(b: &mut Bencher) {
1356 b.bytes = REPETITIONS;
1358 let mut r = String::new();
1359 for _ in range(0, REPETITIONS) {
1366 fn bench_push_char_two_bytes(b: &mut Bencher) {
1367 b.bytes = REPETITIONS * 2;
1369 let mut r = String::new();
1370 for _ in range(0, REPETITIONS) {
1377 fn from_utf8_lossy_100_ascii(b: &mut Bencher) {
1378 let s = b"Hello there, the quick brown fox jumped over the lazy dog! \
1379 Lorem ipsum dolor sit amet, consectetur. ";
1381 assert_eq!(100, s.len());
1383 let _ = String::from_utf8_lossy(s);
1388 fn from_utf8_lossy_100_multibyte(b: &mut Bencher) {
1389 let s = "𐌀𐌖𐌋𐌄𐌑𐌉ปรدولة الكويتทศไทย中华𐍅𐌿𐌻𐍆𐌹𐌻𐌰".as_bytes();
1390 assert_eq!(100, s.len());
1392 let _ = String::from_utf8_lossy(s);
1397 fn from_utf8_lossy_invalid(b: &mut Bencher) {
1398 let s = b"Hello\xC0\x80 There\xE6\x83 Goodbye";
1400 let _ = String::from_utf8_lossy(s);
1405 fn from_utf8_lossy_100_invalid(b: &mut Bencher) {
1406 let s = repeat(0xf5u8).take(100).collect::<Vec<_>>();
1408 let _ = String::from_utf8_lossy(s.as_slice());