1 // Copyright 2012-2014 The Rust Project Developers. See the COPYRIGHT
2 // file at the top-level directory of this distribution and at
3 // http://rust-lang.org/COPYRIGHT.
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 //! String manipulation
15 //! For more details, see std::str
17 #![doc(primitive = "str")]
24 use iter::{Map, Iterator};
25 use iter::{DoubleEndedIterator, ExactSize};
28 use num::{CheckedMul, Saturating};
29 use option::{Option, None, Some};
31 use slice::{mod, SlicePrelude};
35 Section: Creating a string
38 /// Converts a vector to a string slice without performing any allocations.
40 /// Once the slice has been validated as utf-8, it is transmuted in-place and
41 /// returned as a '&str' instead of a '&[u8]'
43 /// Returns None if the slice is not utf-8.
44 pub fn from_utf8<'a>(v: &'a [u8]) -> Option<&'a str> {
46 Some(unsafe { raw::from_utf8(v) })
50 /// Something that can be used to compare against a character
52 /// Determine if the splitter should split at the given character
53 fn matches(&mut self, char) -> bool;
54 /// Indicate if this is only concerned about ASCII characters,
55 /// which can allow for a faster implementation.
56 fn only_ascii(&self) -> bool;
59 impl CharEq for char {
61 fn matches(&mut self, c: char) -> bool { *self == c }
64 fn only_ascii(&self) -> bool { (*self as uint) < 128 }
67 impl<'a> CharEq for |char|: 'a -> bool {
69 fn matches(&mut self, c: char) -> bool { (*self)(c) }
72 fn only_ascii(&self) -> bool { false }
75 impl CharEq for extern "Rust" fn(char) -> bool {
77 fn matches(&mut self, c: char) -> bool { (*self)(c) }
80 fn only_ascii(&self) -> bool { false }
83 impl<'a> CharEq for &'a [char] {
85 fn matches(&mut self, c: char) -> bool {
86 self.iter().any(|&mut m| m.matches(c))
90 fn only_ascii(&self) -> bool {
91 self.iter().all(|m| m.only_ascii())
99 /// Iterator for the char (representing *Unicode Scalar Values*) of a string
101 /// Created with the method `.chars()`.
103 pub struct Chars<'a> {
104 iter: slice::Items<'a, u8>
107 // Return the initial codepoint accumulator for the first byte.
108 // The first byte is special, only want bottom 5 bits for width 2, 4 bits
109 // for width 3, and 3 bits for width 4
110 macro_rules! utf8_first_byte(
111 ($byte:expr, $width:expr) => (($byte & (0x7F >> $width)) as u32)
114 // return the value of $ch updated with continuation byte $byte
115 macro_rules! utf8_acc_cont_byte(
116 ($ch:expr, $byte:expr) => (($ch << 6) | ($byte & CONT_MASK) as u32)
119 macro_rules! utf8_is_cont_byte(
120 ($byte:expr) => (($byte & !CONT_MASK) == TAG_CONT_U8)
124 fn unwrap_or_0(opt: Option<&u8>) -> u8 {
131 impl<'a> Iterator<char> for Chars<'a> {
133 fn next(&mut self) -> Option<char> {
134 // Decode UTF-8, using the valid UTF-8 invariant
135 let x = match self.iter.next() {
137 Some(&next_byte) if next_byte < 128 => return Some(next_byte as char),
138 Some(&next_byte) => next_byte,
141 // Multibyte case follows
142 // Decode from a byte combination out of: [[[x y] z] w]
143 // NOTE: Performance is sensitive to the exact formulation here
144 let init = utf8_first_byte!(x, 2);
145 let y = unwrap_or_0(self.iter.next());
146 let mut ch = utf8_acc_cont_byte!(init, y);
149 // 5th bit in 0xE0 .. 0xEF is always clear, so `init` is still valid
150 let z = unwrap_or_0(self.iter.next());
151 let y_z = utf8_acc_cont_byte!((y & CONT_MASK) as u32, z);
152 ch = init << 12 | y_z;
155 // use only the lower 3 bits of `init`
156 let w = unwrap_or_0(self.iter.next());
157 ch = (init & 7) << 18 | utf8_acc_cont_byte!(y_z, w);
161 // str invariant says `ch` is a valid Unicode Scalar Value
163 Some(mem::transmute(ch))
168 fn size_hint(&self) -> (uint, Option<uint>) {
169 let (len, _) = self.iter.size_hint();
170 (len.saturating_add(3) / 4, Some(len))
174 impl<'a> DoubleEndedIterator<char> for Chars<'a> {
176 fn next_back(&mut self) -> Option<char> {
177 let w = match self.iter.next_back() {
179 Some(&back_byte) if back_byte < 128 => return Some(back_byte as char),
180 Some(&back_byte) => back_byte,
183 // Multibyte case follows
184 // Decode from a byte combination out of: [x [y [z w]]]
186 let z = unwrap_or_0(self.iter.next_back());
187 ch = utf8_first_byte!(z, 2);
188 if utf8_is_cont_byte!(z) {
189 let y = unwrap_or_0(self.iter.next_back());
190 ch = utf8_first_byte!(y, 3);
191 if utf8_is_cont_byte!(y) {
192 let x = unwrap_or_0(self.iter.next_back());
193 ch = utf8_first_byte!(x, 4);
194 ch = utf8_acc_cont_byte!(ch, y);
196 ch = utf8_acc_cont_byte!(ch, z);
198 ch = utf8_acc_cont_byte!(ch, w);
200 // str invariant says `ch` is a valid Unicode Scalar Value
202 Some(mem::transmute(ch))
207 /// External iterator for a string's characters and their byte offsets.
208 /// Use with the `std::iter` module.
210 pub struct CharOffsets<'a> {
215 impl<'a> Iterator<(uint, char)> for CharOffsets<'a> {
217 fn next(&mut self) -> Option<(uint, char)> {
218 let (pre_len, _) = self.iter.iter.size_hint();
219 match self.iter.next() {
222 let index = self.front_offset;
223 let (len, _) = self.iter.iter.size_hint();
224 self.front_offset += pre_len - len;
231 fn size_hint(&self) -> (uint, Option<uint>) {
232 self.iter.size_hint()
236 impl<'a> DoubleEndedIterator<(uint, char)> for CharOffsets<'a> {
238 fn next_back(&mut self) -> Option<(uint, char)> {
239 match self.iter.next_back() {
242 let (len, _) = self.iter.iter.size_hint();
243 let index = self.front_offset + len;
250 /// External iterator for a string's bytes.
251 /// Use with the `std::iter` module.
253 Map<'a, &'a u8, u8, slice::Items<'a, u8>>;
255 /// An iterator over the substrings of a string, separated by `sep`.
257 pub struct CharSplits<'a, Sep> {
258 /// The slice remaining to be iterated
261 /// Whether an empty string at the end is allowed
262 allow_trailing_empty: bool,
267 /// An iterator over the substrings of a string, separated by `sep`,
268 /// splitting at most `count` times.
270 pub struct CharSplitsN<'a, Sep> {
271 iter: CharSplits<'a, Sep>,
272 /// The number of splits remaining
277 /// An iterator over the lines of a string, separated by either `\n` or (`\r\n`).
278 pub type AnyLines<'a> =
279 Map<'a, &'a str, &'a str, CharSplits<'a, char>>;
281 impl<'a, Sep> CharSplits<'a, Sep> {
283 fn get_end(&mut self) -> Option<&'a str> {
284 if !self.finished && (self.allow_trailing_empty || self.string.len() > 0) {
285 self.finished = true;
293 impl<'a, Sep: CharEq> Iterator<&'a str> for CharSplits<'a, Sep> {
295 fn next(&mut self) -> Option<&'a str> {
296 if self.finished { return None }
298 let mut next_split = None;
300 for (idx, byte) in self.string.bytes().enumerate() {
301 if self.sep.matches(byte as char) && byte < 128u8 {
302 next_split = Some((idx, idx + 1));
307 for (idx, ch) in self.string.char_indices() {
308 if self.sep.matches(ch) {
309 next_split = Some((idx, self.string.char_range_at(idx).next));
315 Some((a, b)) => unsafe {
316 let elt = raw::slice_unchecked(self.string, 0, a);
317 self.string = raw::slice_unchecked(self.string, b, self.string.len());
320 None => self.get_end(),
325 impl<'a, Sep: CharEq> DoubleEndedIterator<&'a str>
326 for CharSplits<'a, Sep> {
328 fn next_back(&mut self) -> Option<&'a str> {
329 if self.finished { return None }
331 if !self.allow_trailing_empty {
332 self.allow_trailing_empty = true;
333 match self.next_back() {
334 Some(elt) if !elt.is_empty() => return Some(elt),
335 _ => if self.finished { return None }
338 let len = self.string.len();
339 let mut next_split = None;
342 for (idx, byte) in self.string.bytes().enumerate().rev() {
343 if self.sep.matches(byte as char) && byte < 128u8 {
344 next_split = Some((idx, idx + 1));
349 for (idx, ch) in self.string.char_indices().rev() {
350 if self.sep.matches(ch) {
351 next_split = Some((idx, self.string.char_range_at(idx).next));
357 Some((a, b)) => unsafe {
358 let elt = raw::slice_unchecked(self.string, b, len);
359 self.string = raw::slice_unchecked(self.string, 0, a);
362 None => { self.finished = true; Some(self.string) }
367 impl<'a, Sep: CharEq> Iterator<&'a str> for CharSplitsN<'a, Sep> {
369 fn next(&mut self) -> Option<&'a str> {
372 if self.invert { self.iter.next_back() } else { self.iter.next() }
379 /// The internal state of an iterator that searches for matches of a substring
380 /// within a larger string using naive search
382 struct NaiveSearcher {
387 fn new() -> NaiveSearcher {
388 NaiveSearcher { position: 0 }
391 fn next(&mut self, haystack: &[u8], needle: &[u8]) -> Option<(uint, uint)> {
392 while self.position + needle.len() <= haystack.len() {
393 if haystack[self.position .. self.position + needle.len()] == needle {
394 let match_pos = self.position;
395 self.position += needle.len(); // add 1 for all matches
396 return Some((match_pos, match_pos + needle.len()));
405 /// The internal state of an iterator that searches for matches of a substring
406 /// within a larger string using two-way search
408 struct TwoWaySearcher {
420 This is the Two-Way search algorithm, which was introduced in the paper:
421 Crochemore, M., Perrin, D., 1991, Two-way string-matching, Journal of the ACM 38(3):651-675.
423 Here's some background information.
425 A *word* is a string of symbols. The *length* of a word should be a familiar
426 notion, and here we denote it for any word x by |x|.
427 (We also allow for the possibility of the *empty word*, a word of length zero).
429 If x is any non-empty word, then an integer p with 0 < p <= |x| is said to be a
430 *period* for x iff for all i with 0 <= i <= |x| - p - 1, we have x[i] == x[i+p].
431 For example, both 1 and 2 are periods for the string "aa". As another example,
432 the only period of the string "abcd" is 4.
434 We denote by period(x) the *smallest* period of x (provided that x is non-empty).
435 This is always well-defined since every non-empty word x has at least one period,
436 |x|. We sometimes call this *the period* of x.
438 If u, v and x are words such that x = uv, where uv is the concatenation of u and
439 v, then we say that (u, v) is a *factorization* of x.
441 Let (u, v) be a factorization for a word x. Then if w is a non-empty word such
442 that both of the following hold
444 - either w is a suffix of u or u is a suffix of w
445 - either w is a prefix of v or v is a prefix of w
447 then w is said to be a *repetition* for the factorization (u, v).
449 Just to unpack this, there are four possibilities here. Let w = "abc". Then we
452 - w is a suffix of u and w is a prefix of v. ex: ("lolabc", "abcde")
453 - w is a suffix of u and v is a prefix of w. ex: ("lolabc", "ab")
454 - u is a suffix of w and w is a prefix of v. ex: ("bc", "abchi")
455 - u is a suffix of w and v is a prefix of w. ex: ("bc", "a")
457 Note that the word vu is a repetition for any factorization (u,v) of x = uv,
458 so every factorization has at least one repetition.
460 If x is a string and (u, v) is a factorization for x, then a *local period* for
461 (u, v) is an integer r such that there is some word w such that |w| = r and w is
462 a repetition for (u, v).
464 We denote by local_period(u, v) the smallest local period of (u, v). We sometimes
465 call this *the local period* of (u, v). Provided that x = uv is non-empty, this
466 is well-defined (because each non-empty word has at least one factorization, as
469 It can be proven that the following is an equivalent definition of a local period
470 for a factorization (u, v): any positive integer r such that x[i] == x[i+r] for
471 all i such that |u| - r <= i <= |u| - 1 and such that both x[i] and x[i+r] are
472 defined. (i.e. i > 0 and i + r < |x|).
474 Using the above reformulation, it is easy to prove that
476 1 <= local_period(u, v) <= period(uv)
478 A factorization (u, v) of x such that local_period(u,v) = period(x) is called a
479 *critical factorization*.
481 The algorithm hinges on the following theorem, which is stated without proof:
483 **Critical Factorization Theorem** Any word x has at least one critical
484 factorization (u, v) such that |u| < period(x).
486 The purpose of maximal_suffix is to find such a critical factorization.
489 impl TwoWaySearcher {
490 fn new(needle: &[u8]) -> TwoWaySearcher {
491 let (crit_pos1, period1) = TwoWaySearcher::maximal_suffix(needle, false);
492 let (crit_pos2, period2) = TwoWaySearcher::maximal_suffix(needle, true);
496 if crit_pos1 > crit_pos2 {
497 crit_pos = crit_pos1;
500 crit_pos = crit_pos2;
504 // This isn't in the original algorithm, as far as I'm aware.
505 let byteset = needle.iter()
506 .fold(0, |a, &b| (1 << ((b & 0x3f) as uint)) | a);
508 // A particularly readable explanation of what's going on here can be found
509 // in Crochemore and Rytter's book "Text Algorithms", ch 13. Specifically
510 // see the code for "Algorithm CP" on p. 323.
512 // What's going on is we have some critical factorization (u, v) of the
513 // needle, and we want to determine whether u is a suffix of
514 // v[..period]. If it is, we use "Algorithm CP1". Otherwise we use
515 // "Algorithm CP2", which is optimized for when the period of the needle
517 if needle[..crit_pos] == needle[period.. period + crit_pos] {
529 period: cmp::max(crit_pos, needle.len() - crit_pos) + 1,
533 memory: uint::MAX // Dummy value to signify that the period is long
538 // One of the main ideas of Two-Way is that we factorize the needle into
539 // two halves, (u, v), and begin trying to find v in the haystack by scanning
540 // left to right. If v matches, we try to match u by scanning right to left.
541 // How far we can jump when we encounter a mismatch is all based on the fact
542 // that (u, v) is a critical factorization for the needle.
544 fn next(&mut self, haystack: &[u8], needle: &[u8], long_period: bool) -> Option<(uint, uint)> {
546 // Check that we have room to search in
547 if self.position + needle.len() > haystack.len() {
551 // Quickly skip by large portions unrelated to our substring
553 ((haystack[self.position + needle.len() - 1] & 0x3f)
555 self.position += needle.len();
562 // See if the right part of the needle matches
563 let start = if long_period { self.crit_pos }
564 else { cmp::max(self.crit_pos, self.memory) };
565 for i in range(start, needle.len()) {
566 if needle[i] != haystack[self.position + i] {
567 self.position += i - self.crit_pos + 1;
575 // See if the left part of the needle matches
576 let start = if long_period { 0 } else { self.memory };
577 for i in range(start, self.crit_pos).rev() {
578 if needle[i] != haystack[self.position + i] {
579 self.position += self.period;
581 self.memory = needle.len() - self.period;
587 // We have found a match!
588 let match_pos = self.position;
589 self.position += needle.len(); // add self.period for all matches
591 self.memory = 0; // set to needle.len() - self.period for all matches
593 return Some((match_pos, match_pos + needle.len()));
597 // Computes a critical factorization (u, v) of `arr`.
598 // Specifically, returns (i, p), where i is the starting index of v in some
599 // critical factorization (u, v) and p = period(v)
601 fn maximal_suffix(arr: &[u8], reversed: bool) -> (uint, uint) {
602 let mut left = -1; // Corresponds to i in the paper
603 let mut right = 0; // Corresponds to j in the paper
604 let mut offset = 1; // Corresponds to k in the paper
605 let mut period = 1; // Corresponds to p in the paper
607 while right + offset < arr.len() {
611 a = arr[left + offset];
612 b = arr[right + offset];
614 a = arr[right + offset];
615 b = arr[left + offset];
618 // Suffix is smaller, period is entire prefix so far.
621 period = right - left;
623 // Advance through repetition of the current period.
624 if offset == period {
631 // Suffix is larger, start over from current location.
642 /// The internal state of an iterator that searches for matches of a substring
643 /// within a larger string using a dynamically chosen search algorithm
646 Naive(NaiveSearcher),
647 TwoWay(TwoWaySearcher),
648 TwoWayLong(TwoWaySearcher)
652 fn new(haystack: &[u8], needle: &[u8]) -> Searcher {
654 // FIXME(#16715): This unsigned integer addition will probably not
655 // overflow because that would mean that the memory almost solely
656 // consists of the needle. Needs #16715 to be formally fixed.
657 if needle.len() + 20 > haystack.len() {
658 Naive(NaiveSearcher::new())
660 let searcher = TwoWaySearcher::new(needle);
661 if searcher.memory == uint::MAX { // If the period is long
670 /// An iterator over the start and end indices of the matches of a
671 /// substring within a larger string
673 pub struct MatchIndices<'a> {
680 /// An iterator over the substrings of a string separated by a given
683 pub struct StrSplits<'a> {
684 it: MatchIndices<'a>,
689 impl<'a> Iterator<(uint, uint)> for MatchIndices<'a> {
691 fn next(&mut self) -> Option<(uint, uint)> {
692 match self.searcher {
693 Naive(ref mut searcher)
694 => searcher.next(self.haystack.as_bytes(), self.needle.as_bytes()),
695 TwoWay(ref mut searcher)
696 => searcher.next(self.haystack.as_bytes(), self.needle.as_bytes(), false),
697 TwoWayLong(ref mut searcher)
698 => searcher.next(self.haystack.as_bytes(), self.needle.as_bytes(), true)
703 impl<'a> Iterator<&'a str> for StrSplits<'a> {
705 fn next(&mut self) -> Option<&'a str> {
706 if self.finished { return None; }
708 match self.it.next() {
709 Some((from, to)) => {
710 let ret = Some(self.it.haystack.slice(self.last_end, from));
715 self.finished = true;
716 Some(self.it.haystack.slice(self.last_end, self.it.haystack.len()))
722 /// External iterator for a string's UTF16 codeunits.
723 /// Use with the `std::iter` module.
725 pub struct Utf16CodeUnits<'a> {
730 impl<'a> Iterator<u16> for Utf16CodeUnits<'a> {
732 fn next(&mut self) -> Option<u16> {
734 let tmp = self.extra;
739 let mut buf = [0u16, ..2];
740 self.chars.next().map(|ch| {
741 let n = ch.encode_utf16(buf[mut]).unwrap_or(0);
742 if n == 2 { self.extra = buf[1]; }
748 fn size_hint(&self) -> (uint, Option<uint>) {
749 let (low, high) = self.chars.size_hint();
750 // every char gets either one u16 or two u16,
751 // so this iterator is between 1 or 2 times as
752 // long as the underlying iterator.
753 (low, high.and_then(|n| n.checked_mul(&2)))
758 Section: Comparing strings
761 // share the implementation of the lang-item vs. non-lang-item
763 /// NOTE: This function is (ab)used in rustc::middle::trans::_match
764 /// to compare &[u8] byte slices that are not necessarily valid UTF-8.
766 fn eq_slice_(a: &str, b: &str) -> bool {
767 #[allow(improper_ctypes)]
768 extern { fn memcmp(s1: *const i8, s2: *const i8, n: uint) -> i32; }
769 a.len() == b.len() && unsafe {
770 memcmp(a.as_ptr() as *const i8,
771 b.as_ptr() as *const i8,
776 /// Bytewise slice equality
777 /// NOTE: This function is (ab)used in rustc::middle::trans::_match
778 /// to compare &[u8] byte slices that are not necessarily valid UTF-8.
781 pub fn eq_slice(a: &str, b: &str) -> bool {
789 /// Walk through `iter` checking that it's a valid UTF-8 sequence,
790 /// returning `true` in that case, or, if it is invalid, `false` with
791 /// `iter` reset such that it is pointing at the first byte in the
792 /// invalid sequence.
794 fn run_utf8_validation_iterator(iter: &mut slice::Items<u8>) -> bool {
796 // save the current thing we're pointing at.
799 // restore the iterator we had at the start of this codepoint.
800 macro_rules! err ( () => { {*iter = old; return false} });
801 macro_rules! next ( () => {
804 // we needed data, but there was none: error!
809 let first = match iter.next() {
811 // we're at the end of the iterator and a codepoint
812 // boundary at the same time, so this string is valid.
816 // ASCII characters are always valid, so only large
817 // bytes need more examination.
819 let w = utf8_char_width(first);
820 let second = next!();
821 // 2-byte encoding is for codepoints \u0080 to \u07ff
822 // first C2 80 last DF BF
823 // 3-byte encoding is for codepoints \u0800 to \uffff
824 // first E0 A0 80 last EF BF BF
825 // excluding surrogates codepoints \ud800 to \udfff
826 // ED A0 80 to ED BF BF
827 // 4-byte encoding is for codepoints \u10000 to \u10ffff
828 // first F0 90 80 80 last F4 8F BF BF
830 // Use the UTF-8 syntax from the RFC
832 // https://tools.ietf.org/html/rfc3629
834 // UTF8-2 = %xC2-DF UTF8-tail
835 // UTF8-3 = %xE0 %xA0-BF UTF8-tail / %xE1-EC 2( UTF8-tail ) /
836 // %xED %x80-9F UTF8-tail / %xEE-EF 2( UTF8-tail )
837 // UTF8-4 = %xF0 %x90-BF 2( UTF8-tail ) / %xF1-F3 3( UTF8-tail ) /
838 // %xF4 %x80-8F 2( UTF8-tail )
840 2 => if second & !CONT_MASK != TAG_CONT_U8 {err!()},
842 match (first, second, next!() & !CONT_MASK) {
843 (0xE0 , 0xA0 ... 0xBF, TAG_CONT_U8) |
844 (0xE1 ... 0xEC, 0x80 ... 0xBF, TAG_CONT_U8) |
845 (0xED , 0x80 ... 0x9F, TAG_CONT_U8) |
846 (0xEE ... 0xEF, 0x80 ... 0xBF, TAG_CONT_U8) => {}
851 match (first, second, next!() & !CONT_MASK, next!() & !CONT_MASK) {
852 (0xF0 , 0x90 ... 0xBF, TAG_CONT_U8, TAG_CONT_U8) |
853 (0xF1 ... 0xF3, 0x80 ... 0xBF, TAG_CONT_U8, TAG_CONT_U8) |
854 (0xF4 , 0x80 ... 0x8F, TAG_CONT_U8, TAG_CONT_U8) => {}
864 /// Determines if a vector of bytes contains valid UTF-8.
865 pub fn is_utf8(v: &[u8]) -> bool {
866 run_utf8_validation_iterator(&mut v.iter())
869 /// Determines if a vector of `u16` contains valid UTF-16
870 pub fn is_utf16(v: &[u16]) -> bool {
871 let mut it = v.iter();
872 macro_rules! next ( ($ret:expr) => {
873 match it.next() { Some(u) => *u, None => return $ret }
879 match char::from_u32(u as u32) {
882 let u2 = next!(false);
883 if u < 0xD7FF || u > 0xDBFF ||
884 u2 < 0xDC00 || u2 > 0xDFFF { return false; }
890 /// An iterator that decodes UTF-16 encoded codepoints from a vector
893 pub struct Utf16Items<'a> {
894 iter: slice::Items<'a, u16>
896 /// The possibilities for values decoded from a `u16` stream.
897 #[deriving(PartialEq, Eq, Clone, Show)]
899 /// A valid codepoint.
901 /// An invalid surrogate without its pair.
906 /// Convert `self` to a `char`, taking `LoneSurrogate`s to the
907 /// replacement character (U+FFFD).
909 pub fn to_char_lossy(&self) -> char {
912 LoneSurrogate(_) => '\uFFFD'
917 impl<'a> Iterator<Utf16Item> for Utf16Items<'a> {
918 fn next(&mut self) -> Option<Utf16Item> {
919 let u = match self.iter.next() {
924 if u < 0xD800 || 0xDFFF < u {
926 Some(ScalarValue(unsafe {mem::transmute(u as u32)}))
927 } else if u >= 0xDC00 {
928 // a trailing surrogate
929 Some(LoneSurrogate(u))
931 // preserve state for rewinding.
934 let u2 = match self.iter.next() {
937 None => return Some(LoneSurrogate(u))
939 if u2 < 0xDC00 || u2 > 0xDFFF {
940 // not a trailing surrogate so we're not a valid
941 // surrogate pair, so rewind to redecode u2 next time.
943 return Some(LoneSurrogate(u))
946 // all ok, so lets decode it.
947 let c = ((u - 0xD800) as u32 << 10 | (u2 - 0xDC00) as u32) + 0x1_0000;
948 Some(ScalarValue(unsafe {mem::transmute(c)}))
953 fn size_hint(&self) -> (uint, Option<uint>) {
954 let (low, high) = self.iter.size_hint();
955 // we could be entirely valid surrogates (2 elements per
956 // char), or entirely non-surrogates (1 element per char)
961 /// Create an iterator over the UTF-16 encoded codepoints in `v`,
962 /// returning invalid surrogates as `LoneSurrogate`s.
968 /// use std::str::{ScalarValue, LoneSurrogate};
970 /// // 𝄞mus<invalid>ic<invalid>
971 /// let v = [0xD834, 0xDD1E, 0x006d, 0x0075,
972 /// 0x0073, 0xDD1E, 0x0069, 0x0063,
975 /// assert_eq!(str::utf16_items(v).collect::<Vec<_>>(),
976 /// vec![ScalarValue('𝄞'),
977 /// ScalarValue('m'), ScalarValue('u'), ScalarValue('s'),
978 /// LoneSurrogate(0xDD1E),
979 /// ScalarValue('i'), ScalarValue('c'),
980 /// LoneSurrogate(0xD834)]);
982 pub fn utf16_items<'a>(v: &'a [u16]) -> Utf16Items<'a> {
983 Utf16Items { iter : v.iter() }
986 /// Return a slice of `v` ending at (and not including) the first NUL
995 /// let mut v = ['a' as u16, 'b' as u16, 'c' as u16, 'd' as u16];
996 /// // no NULs so no change
997 /// assert_eq!(str::truncate_utf16_at_nul(v), v.as_slice());
1001 /// let b: &[_] = &['a' as u16, 'b' as u16];
1002 /// assert_eq!(str::truncate_utf16_at_nul(v), b);
1004 pub fn truncate_utf16_at_nul<'a>(v: &'a [u16]) -> &'a [u16] {
1005 match v.iter().position(|c| *c == 0) {
1006 // don't include the 0
1012 // https://tools.ietf.org/html/rfc3629
1013 static UTF8_CHAR_WIDTH: [u8, ..256] = [
1014 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1015 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x1F
1016 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1017 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x3F
1018 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1019 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x5F
1020 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1021 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x7F
1022 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
1023 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, // 0x9F
1024 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
1025 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, // 0xBF
1026 0,0,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
1027 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, // 0xDF
1028 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, // 0xEF
1029 4,4,4,4,4,0,0,0,0,0,0,0,0,0,0,0, // 0xFF
1032 /// Given a first byte, determine how many bytes are in this UTF-8 character
1034 pub fn utf8_char_width(b: u8) -> uint {
1035 return UTF8_CHAR_WIDTH[b as uint] as uint;
1038 /// Struct that contains a `char` and the index of the first byte of
1039 /// the next `char` in a string. This can be used as a data structure
1040 /// for iterating over the UTF-8 bytes of a string.
1041 pub struct CharRange {
1044 /// Index of the first byte of the next `char`
1048 /// Mask of the value bits of a continuation byte
1049 const CONT_MASK: u8 = 0b0011_1111u8;
1050 /// Value of the tag bits (tag mask is !CONT_MASK) of a continuation byte
1051 const TAG_CONT_U8: u8 = 0b1000_0000u8;
1053 /// Unsafe operations
1058 use slice::SlicePrelude;
1059 use str::{is_utf8, StrPrelude};
1061 /// Converts a slice of bytes to a string slice without checking
1062 /// that the string contains valid UTF-8.
1063 pub unsafe fn from_utf8<'a>(v: &'a [u8]) -> &'a str {
1067 /// Form a slice from a C string. Unsafe because the caller must ensure the
1068 /// C string has the static lifetime, or else the return value may be
1069 /// invalidated later.
1070 pub unsafe fn c_str_to_static_slice(s: *const i8) -> &'static str {
1071 let s = s as *const u8;
1074 while *curr != 0u8 {
1076 curr = s.offset(len as int);
1078 let v = Slice { data: s, len: len };
1079 assert!(is_utf8(::mem::transmute(v)));
1083 /// Takes a bytewise (not UTF-8) slice from a string.
1085 /// Returns the substring from [`begin`..`end`).
1089 /// If begin is greater than end.
1090 /// If end is greater than the length of the string.
1092 pub unsafe fn slice_bytes<'a>(s: &'a str, begin: uint, end: uint) -> &'a str {
1093 assert!(begin <= end);
1094 assert!(end <= s.len());
1095 slice_unchecked(s, begin, end)
1098 /// Takes a bytewise (not UTF-8) slice from a string.
1100 /// Returns the substring from [`begin`..`end`).
1102 /// Caller must check slice boundaries!
1104 pub unsafe fn slice_unchecked<'a>(s: &'a str, begin: uint, end: uint) -> &'a str {
1105 mem::transmute(Slice {
1106 data: s.as_ptr().offset(begin as int),
1113 Section: Trait implementations
1116 #[allow(missing_docs)]
1118 use cmp::{Ord, Ordering, Less, Equal, Greater, PartialEq, PartialOrd, Equiv, Eq};
1120 use option::{Option, Some};
1122 use str::{Str, StrPrelude, eq_slice};
1124 // NOTE(stage0): remove impl after a snapshot
1126 impl<'a> Ord for &'a str {
1128 fn cmp(&self, other: & &'a str) -> Ordering {
1129 for (s_b, o_b) in self.bytes().zip(other.bytes()) {
1130 match s_b.cmp(&o_b) {
1131 Greater => return Greater,
1132 Less => return Less,
1137 self.len().cmp(&other.len())
1141 #[cfg(not(stage0))] // NOTE(stage0): remove cfg after a snapshot
1144 fn cmp(&self, other: &str) -> Ordering {
1145 for (s_b, o_b) in self.bytes().zip(other.bytes()) {
1146 match s_b.cmp(&o_b) {
1147 Greater => return Greater,
1148 Less => return Less,
1153 self.len().cmp(&other.len())
1157 // NOTE(stage0): remove impl after a snapshot
1159 impl<'a> PartialEq for &'a str {
1161 fn eq(&self, other: & &'a str) -> bool {
1162 eq_slice((*self), (*other))
1165 fn ne(&self, other: & &'a str) -> bool { !(*self).eq(other) }
1168 #[cfg(not(stage0))] // NOTE(stage0): remove cfg after a snapshot
1169 impl PartialEq for str {
1171 fn eq(&self, other: &str) -> bool {
1172 eq_slice(self, other)
1175 fn ne(&self, other: &str) -> bool { !(*self).eq(other) }
1178 // NOTE(stage0): remove impl after a snapshot
1180 impl<'a> Eq for &'a str {}
1182 #[cfg(not(stage0))] // NOTE(stage0): remove cfg after a snapshot
1185 // NOTE(stage0): remove impl after a snapshot
1187 impl<'a> PartialOrd for &'a str {
1189 fn partial_cmp(&self, other: &&'a str) -> Option<Ordering> {
1190 Some(self.cmp(other))
1194 #[cfg(not(stage0))] // NOTE(stage0): remove cfg after a snapshot
1195 impl PartialOrd for str {
1197 fn partial_cmp(&self, other: &str) -> Option<Ordering> {
1198 Some(self.cmp(other))
1202 impl<S: Str> Equiv<S> for str {
1204 fn equiv(&self, other: &S) -> bool { eq_slice(self, other.as_slice()) }
1207 impl ops::Slice<uint, str> for str {
1209 fn as_slice_<'a>(&'a self) -> &'a str {
1214 fn slice_from_or_fail<'a>(&'a self, from: &uint) -> &'a str {
1215 self.slice_from(*from)
1219 fn slice_to_or_fail<'a>(&'a self, to: &uint) -> &'a str {
1224 fn slice_or_fail<'a>(&'a self, from: &uint, to: &uint) -> &'a str {
1225 self.slice(*from, *to)
1230 /// Any string that can be represented as a slice
1232 /// Work with `self` as a slice.
1233 fn as_slice<'a>(&'a self) -> &'a str;
1236 impl<'a> Str for &'a str {
1238 fn as_slice<'a>(&'a self) -> &'a str { *self }
1241 /// Methods for string slices
1242 pub trait StrPrelude for Sized? {
1243 /// Returns true if one string contains another
1247 /// - needle - The string to look for
1252 /// assert!("bananas".contains("nana"));
1254 fn contains(&self, needle: &str) -> bool;
1256 /// Returns true if a string contains a char.
1260 /// - needle - The char to look for
1265 /// assert!("hello".contains_char('e'));
1267 fn contains_char(&self, needle: char) -> bool;
1269 /// An iterator over the characters of `self`. Note, this iterates
1270 /// over Unicode code-points, not Unicode graphemes.
1275 /// let v: Vec<char> = "abc åäö".chars().collect();
1276 /// assert_eq!(v, vec!['a', 'b', 'c', ' ', 'å', 'ä', 'ö']);
1278 fn chars<'a>(&'a self) -> Chars<'a>;
1280 /// An iterator over the bytes of `self`
1285 /// let v: Vec<u8> = "bors".bytes().collect();
1286 /// assert_eq!(v, b"bors".to_vec());
1288 fn bytes<'a>(&'a self) -> Bytes<'a>;
1290 /// An iterator over the characters of `self` and their byte offsets.
1291 fn char_indices<'a>(&'a self) -> CharOffsets<'a>;
1293 /// An iterator over substrings of `self`, separated by characters
1294 /// matched by `sep`.
1299 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').collect();
1300 /// assert_eq!(v, vec!["Mary", "had", "a", "little", "lamb"]);
1302 /// let v: Vec<&str> = "abc1def2ghi".split(|c: char| c.is_digit()).collect();
1303 /// assert_eq!(v, vec!["abc", "def", "ghi"]);
1305 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').collect();
1306 /// assert_eq!(v, vec!["lion", "", "tiger", "leopard"]);
1308 /// let v: Vec<&str> = "".split('X').collect();
1309 /// assert_eq!(v, vec![""]);
1311 fn split<'a, Sep: CharEq>(&'a self, sep: Sep) -> CharSplits<'a, Sep>;
1313 /// An iterator over substrings of `self`, separated by characters
1314 /// matched by `sep`, restricted to splitting at most `count`
1320 /// let v: Vec<&str> = "Mary had a little lambda".splitn(2, ' ').collect();
1321 /// assert_eq!(v, vec!["Mary", "had", "a little lambda"]);
1323 /// let v: Vec<&str> = "abc1def2ghi".splitn(1, |c: char| c.is_digit()).collect();
1324 /// assert_eq!(v, vec!["abc", "def2ghi"]);
1326 /// let v: Vec<&str> = "lionXXtigerXleopard".splitn(2, 'X').collect();
1327 /// assert_eq!(v, vec!["lion", "", "tigerXleopard"]);
1329 /// let v: Vec<&str> = "abcXdef".splitn(0, 'X').collect();
1330 /// assert_eq!(v, vec!["abcXdef"]);
1332 /// let v: Vec<&str> = "".splitn(1, 'X').collect();
1333 /// assert_eq!(v, vec![""]);
1335 fn splitn<'a, Sep: CharEq>(&'a self, count: uint, sep: Sep) -> CharSplitsN<'a, Sep>;
1337 /// An iterator over substrings of `self`, separated by characters
1338 /// matched by `sep`.
1340 /// Equivalent to `split`, except that the trailing substring
1341 /// is skipped if empty (terminator semantics).
1346 /// let v: Vec<&str> = "A.B.".split_terminator('.').collect();
1347 /// assert_eq!(v, vec!["A", "B"]);
1349 /// let v: Vec<&str> = "A..B..".split_terminator('.').collect();
1350 /// assert_eq!(v, vec!["A", "", "B", ""]);
1352 /// let v: Vec<&str> = "Mary had a little lamb".split(' ').rev().collect();
1353 /// assert_eq!(v, vec!["lamb", "little", "a", "had", "Mary"]);
1355 /// let v: Vec<&str> = "abc1def2ghi".split(|c: char| c.is_digit()).rev().collect();
1356 /// assert_eq!(v, vec!["ghi", "def", "abc"]);
1358 /// let v: Vec<&str> = "lionXXtigerXleopard".split('X').rev().collect();
1359 /// assert_eq!(v, vec!["leopard", "tiger", "", "lion"]);
1361 fn split_terminator<'a, Sep: CharEq>(&'a self, sep: Sep) -> CharSplits<'a, Sep>;
1363 /// An iterator over substrings of `self`, separated by characters
1364 /// matched by `sep`, starting from the end of the string.
1365 /// Restricted to splitting at most `count` times.
1370 /// let v: Vec<&str> = "Mary had a little lamb".rsplitn(2, ' ').collect();
1371 /// assert_eq!(v, vec!["lamb", "little", "Mary had a"]);
1373 /// let v: Vec<&str> = "abc1def2ghi".rsplitn(1, |c: char| c.is_digit()).collect();
1374 /// assert_eq!(v, vec!["ghi", "abc1def"]);
1376 /// let v: Vec<&str> = "lionXXtigerXleopard".rsplitn(2, 'X').collect();
1377 /// assert_eq!(v, vec!["leopard", "tiger", "lionX"]);
1379 fn rsplitn<'a, Sep: CharEq>(&'a self, count: uint, sep: Sep) -> CharSplitsN<'a, Sep>;
1381 /// An iterator over the start and end indices of the disjoint
1382 /// matches of `sep` within `self`.
1384 /// That is, each returned value `(start, end)` satisfies
1385 /// `self.slice(start, end) == sep`. For matches of `sep` within
1386 /// `self` that overlap, only the indices corresponding to the
1387 /// first match are returned.
1392 /// let v: Vec<(uint, uint)> = "abcXXXabcYYYabc".match_indices("abc").collect();
1393 /// assert_eq!(v, vec![(0,3), (6,9), (12,15)]);
1395 /// let v: Vec<(uint, uint)> = "1abcabc2".match_indices("abc").collect();
1396 /// assert_eq!(v, vec![(1,4), (4,7)]);
1398 /// let v: Vec<(uint, uint)> = "ababa".match_indices("aba").collect();
1399 /// assert_eq!(v, vec![(0, 3)]); // only the first `aba`
1401 fn match_indices<'a>(&'a self, sep: &'a str) -> MatchIndices<'a>;
1403 /// An iterator over the substrings of `self` separated by `sep`.
1408 /// let v: Vec<&str> = "abcXXXabcYYYabc".split_str("abc").collect();
1409 /// assert_eq!(v, vec!["", "XXX", "YYY", ""]);
1411 /// let v: Vec<&str> = "1abcabc2".split_str("abc").collect();
1412 /// assert_eq!(v, vec!["1", "", "2"]);
1414 fn split_str<'a>(&'a self, &'a str) -> StrSplits<'a>;
1416 /// An iterator over the lines of a string (subsequences separated
1417 /// by `\n`). This does not include the empty string after a
1423 /// let four_lines = "foo\nbar\n\nbaz\n";
1424 /// let v: Vec<&str> = four_lines.lines().collect();
1425 /// assert_eq!(v, vec!["foo", "bar", "", "baz"]);
1427 fn lines<'a>(&'a self) -> CharSplits<'a, char>;
1429 /// An iterator over the lines of a string, separated by either
1430 /// `\n` or `\r\n`. As with `.lines()`, this does not include an
1431 /// empty trailing line.
1436 /// let four_lines = "foo\r\nbar\n\r\nbaz\n";
1437 /// let v: Vec<&str> = four_lines.lines_any().collect();
1438 /// assert_eq!(v, vec!["foo", "bar", "", "baz"]);
1440 fn lines_any<'a>(&'a self) -> AnyLines<'a>;
1442 /// Returns the number of Unicode code points (`char`) that a
1445 /// This does not perform any normalization, and is `O(n)`, since
1446 /// UTF-8 is a variable width encoding of code points.
1448 /// *Warning*: The number of code points in a string does not directly
1449 /// correspond to the number of visible characters or width of the
1450 /// visible text due to composing characters, and double- and
1451 /// zero-width ones.
1453 /// See also `.len()` for the byte length.
1458 /// // composed forms of `ö` and `é`
1459 /// let c = "Löwe 老虎 Léopard"; // German, Simplified Chinese, French
1460 /// // decomposed forms of `ö` and `é`
1461 /// let d = "Lo\u0308we 老虎 Le\u0301opard";
1463 /// assert_eq!(c.char_len(), 15);
1464 /// assert_eq!(d.char_len(), 17);
1466 /// assert_eq!(c.len(), 21);
1467 /// assert_eq!(d.len(), 23);
1469 /// // the two strings *look* the same
1470 /// println!("{}", c);
1471 /// println!("{}", d);
1473 fn char_len(&self) -> uint;
1475 /// Returns a slice of the given string from the byte range
1476 /// [`begin`..`end`).
1478 /// This operation is `O(1)`.
1480 /// Panics when `begin` and `end` do not point to valid characters
1481 /// or point beyond the last character of the string.
1483 /// See also `slice_to` and `slice_from` for slicing prefixes and
1484 /// suffixes of strings, and `slice_chars` for slicing based on
1485 /// code point counts.
1490 /// let s = "Löwe 老虎 Léopard";
1491 /// assert_eq!(s.slice(0, 1), "L");
1493 /// assert_eq!(s.slice(1, 9), "öwe 老");
1495 /// // these will panic:
1496 /// // byte 2 lies within `ö`:
1497 /// // s.slice(2, 3);
1499 /// // byte 8 lies within `老`
1500 /// // s.slice(1, 8);
1502 /// // byte 100 is outside the string
1503 /// // s.slice(3, 100);
1505 fn slice<'a>(&'a self, begin: uint, end: uint) -> &'a str;
1507 /// Returns a slice of the string from `begin` to its end.
1509 /// Equivalent to `self.slice(begin, self.len())`.
1511 /// Panics when `begin` does not point to a valid character, or is
1514 /// See also `slice`, `slice_to` and `slice_chars`.
1515 fn slice_from<'a>(&'a self, begin: uint) -> &'a str;
1517 /// Returns a slice of the string from the beginning to byte
1520 /// Equivalent to `self.slice(0, end)`.
1522 /// Panics when `end` does not point to a valid character, or is
1525 /// See also `slice`, `slice_from` and `slice_chars`.
1526 fn slice_to<'a>(&'a self, end: uint) -> &'a str;
1528 /// Returns a slice of the string from the character range
1529 /// [`begin`..`end`).
1531 /// That is, start at the `begin`-th code point of the string and
1532 /// continue to the `end`-th code point. This does not detect or
1533 /// handle edge cases such as leaving a combining character as the
1534 /// first code point of the string.
1536 /// Due to the design of UTF-8, this operation is `O(end)`.
1537 /// See `slice`, `slice_to` and `slice_from` for `O(1)`
1538 /// variants that use byte indices rather than code point
1541 /// Panics if `begin` > `end` or the either `begin` or `end` are
1542 /// beyond the last character of the string.
1547 /// let s = "Löwe 老虎 Léopard";
1548 /// assert_eq!(s.slice_chars(0, 4), "Löwe");
1549 /// assert_eq!(s.slice_chars(5, 7), "老虎");
1551 fn slice_chars<'a>(&'a self, begin: uint, end: uint) -> &'a str;
1553 /// Returns true if `needle` is a prefix of the string.
1558 /// assert!("banana".starts_with("ba"));
1560 fn starts_with(&self, needle: &str) -> bool;
1562 /// Returns true if `needle` is a suffix of the string.
1567 /// assert!("banana".ends_with("nana"));
1569 fn ends_with(&self, needle: &str) -> bool;
1571 /// Returns a string with characters that match `to_trim` removed from the left and the right.
1575 /// * to_trim - a character matcher
1580 /// assert_eq!("11foo1bar11".trim_chars('1'), "foo1bar")
1581 /// let x: &[_] = &['1', '2'];
1582 /// assert_eq!("12foo1bar12".trim_chars(x), "foo1bar")
1583 /// assert_eq!("123foo1bar123".trim_chars(|c: char| c.is_digit()), "foo1bar")
1585 fn trim_chars<'a, C: CharEq>(&'a self, to_trim: C) -> &'a str;
1587 /// Returns a string with leading `chars_to_trim` removed.
1591 /// * to_trim - a character matcher
1596 /// assert_eq!("11foo1bar11".trim_left_chars('1'), "foo1bar11")
1597 /// let x: &[_] = &['1', '2'];
1598 /// assert_eq!("12foo1bar12".trim_left_chars(x), "foo1bar12")
1599 /// assert_eq!("123foo1bar123".trim_left_chars(|c: char| c.is_digit()), "foo1bar123")
1601 fn trim_left_chars<'a, C: CharEq>(&'a self, to_trim: C) -> &'a str;
1603 /// Returns a string with trailing `chars_to_trim` removed.
1607 /// * to_trim - a character matcher
1612 /// assert_eq!("11foo1bar11".trim_right_chars('1'), "11foo1bar")
1613 /// let x: &[_] = &['1', '2'];
1614 /// assert_eq!("12foo1bar12".trim_right_chars(x), "12foo1bar")
1615 /// assert_eq!("123foo1bar123".trim_right_chars(|c: char| c.is_digit()), "123foo1bar")
1617 fn trim_right_chars<'a, C: CharEq>(&'a self, to_trim: C) -> &'a str;
1619 /// Check that `index`-th byte lies at the start and/or end of a
1620 /// UTF-8 code point sequence.
1622 /// The start and end of the string (when `index == self.len()`)
1623 /// are considered to be boundaries.
1625 /// Panics if `index` is greater than `self.len()`.
1630 /// let s = "Löwe 老虎 Léopard";
1631 /// assert!(s.is_char_boundary(0));
1633 /// assert!(s.is_char_boundary(6));
1634 /// assert!(s.is_char_boundary(s.len()));
1636 /// // second byte of `ö`
1637 /// assert!(!s.is_char_boundary(2));
1639 /// // third byte of `老`
1640 /// assert!(!s.is_char_boundary(8));
1642 fn is_char_boundary(&self, index: uint) -> bool;
1644 /// Pluck a character out of a string and return the index of the next
1647 /// This function can be used to iterate over the Unicode characters of a
1652 /// This example manually iterates through the characters of a
1653 /// string; this should normally be done by `.chars()` or
1654 /// `.char_indices`.
1657 /// use std::str::CharRange;
1659 /// let s = "中华Việt Nam";
1661 /// while i < s.len() {
1662 /// let CharRange {ch, next} = s.char_range_at(i);
1663 /// println!("{}: {}", i, ch);
1685 /// * s - The string
1686 /// * i - The byte offset of the char to extract
1690 /// A record {ch: char, next: uint} containing the char value and the byte
1691 /// index of the next Unicode character.
1695 /// If `i` is greater than or equal to the length of the string.
1696 /// If `i` is not the index of the beginning of a valid UTF-8 character.
1697 fn char_range_at(&self, start: uint) -> CharRange;
1699 /// Given a byte position and a str, return the previous char and its position.
1701 /// This function can be used to iterate over a Unicode string in reverse.
1703 /// Returns 0 for next index if called on start index 0.
1707 /// If `i` is greater than the length of the string.
1708 /// If `i` is not an index following a valid UTF-8 character.
1709 fn char_range_at_reverse(&self, start: uint) -> CharRange;
1711 /// Plucks the character starting at the `i`th byte of a string.
1717 /// assert_eq!(s.char_at(1), 'b');
1718 /// assert_eq!(s.char_at(2), 'π');
1719 /// assert_eq!(s.char_at(4), 'c');
1724 /// If `i` is greater than or equal to the length of the string.
1725 /// If `i` is not the index of the beginning of a valid UTF-8 character.
1726 fn char_at(&self, i: uint) -> char;
1728 /// Plucks the character ending at the `i`th byte of a string.
1732 /// If `i` is greater than the length of the string.
1733 /// If `i` is not an index following a valid UTF-8 character.
1734 fn char_at_reverse(&self, i: uint) -> char;
1736 /// Work with the byte buffer of a string as a byte slice.
1741 /// assert_eq!("bors".as_bytes(), b"bors");
1743 fn as_bytes<'a>(&'a self) -> &'a [u8];
1745 /// Returns the byte index of the first character of `self` that
1746 /// matches `search`.
1750 /// `Some` containing the byte index of the last matching character
1751 /// or `None` if there is no match
1756 /// let s = "Löwe 老虎 Léopard";
1758 /// assert_eq!(s.find('L'), Some(0));
1759 /// assert_eq!(s.find('é'), Some(14));
1761 /// // the first space
1762 /// assert_eq!(s.find(|c: char| c.is_whitespace()), Some(5));
1764 /// // neither are found
1765 /// let x: &[_] = &['1', '2'];
1766 /// assert_eq!(s.find(x), None);
1768 fn find<C: CharEq>(&self, search: C) -> Option<uint>;
1770 /// Returns the byte index of the last character of `self` that
1771 /// matches `search`.
1775 /// `Some` containing the byte index of the last matching character
1776 /// or `None` if there is no match.
1781 /// let s = "Löwe 老虎 Léopard";
1783 /// assert_eq!(s.rfind('L'), Some(13));
1784 /// assert_eq!(s.rfind('é'), Some(14));
1786 /// // the second space
1787 /// assert_eq!(s.rfind(|c: char| c.is_whitespace()), Some(12));
1789 /// // searches for an occurrence of either `1` or `2`, but neither are found
1790 /// let x: &[_] = &['1', '2'];
1791 /// assert_eq!(s.rfind(x), None);
1793 fn rfind<C: CharEq>(&self, search: C) -> Option<uint>;
1795 /// Returns the byte index of the first matching substring
1799 /// * `needle` - The string to search for
1803 /// `Some` containing the byte index of the first matching substring
1804 /// or `None` if there is no match.
1809 /// let s = "Löwe 老虎 Léopard";
1811 /// assert_eq!(s.find_str("老虎 L"), Some(6));
1812 /// assert_eq!(s.find_str("muffin man"), None);
1814 fn find_str(&self, &str) -> Option<uint>;
1816 /// Retrieves the first character from a string slice and returns
1817 /// it. This does not allocate a new string; instead, it returns a
1818 /// slice that point one character beyond the character that was
1819 /// shifted. If the string does not contain any characters,
1820 /// a tuple of None and an empty string is returned instead.
1825 /// let s = "Löwe 老虎 Léopard";
1826 /// let (c, s1) = s.slice_shift_char();
1827 /// assert_eq!(c, Some('L'));
1828 /// assert_eq!(s1, "öwe 老虎 Léopard");
1830 /// let (c, s2) = s1.slice_shift_char();
1831 /// assert_eq!(c, Some('ö'));
1832 /// assert_eq!(s2, "we 老虎 Léopard");
1834 fn slice_shift_char<'a>(&'a self) -> (Option<char>, &'a str);
1836 /// Returns the byte offset of an inner slice relative to an enclosing outer slice.
1838 /// Panics if `inner` is not a direct slice contained within self.
1843 /// let string = "a\nb\nc";
1844 /// let lines: Vec<&str> = string.lines().collect();
1845 /// let lines = lines.as_slice();
1847 /// assert!(string.subslice_offset(lines[0]) == 0); // &"a"
1848 /// assert!(string.subslice_offset(lines[1]) == 2); // &"b"
1849 /// assert!(string.subslice_offset(lines[2]) == 4); // &"c"
1851 fn subslice_offset(&self, inner: &str) -> uint;
1853 /// Return an unsafe pointer to the strings buffer.
1855 /// The caller must ensure that the string outlives this pointer,
1856 /// and that it is not reallocated (e.g. by pushing to the
1858 fn as_ptr(&self) -> *const u8;
1860 /// Return an iterator of `u16` over the string encoded as UTF-16.
1861 fn utf16_units<'a>(&'a self) -> Utf16CodeUnits<'a>;
1863 /// Return the number of bytes in this string
1868 /// assert_eq!("foo".len(), 3);
1869 /// assert_eq!("ƒoo".len(), 4);
1871 #[experimental = "not triaged yet"]
1872 fn len(&self) -> uint;
1874 /// Returns true if this slice contains no bytes
1879 /// assert!("".is_empty());
1882 #[experimental = "not triaged yet"]
1883 fn is_empty(&self) -> bool { self.len() == 0 }
1887 fn slice_error_fail(s: &str, begin: uint, end: uint) -> ! {
1888 assert!(begin <= end);
1889 panic!("index {} and/or {} in `{}` do not lie on character boundary",
1893 impl StrPrelude for str {
1895 fn contains(&self, needle: &str) -> bool {
1896 self.find_str(needle).is_some()
1900 fn contains_char(&self, needle: char) -> bool {
1901 self.find(needle).is_some()
1905 fn chars(&self) -> Chars {
1906 Chars{iter: self.as_bytes().iter()}
1910 fn bytes(&self) -> Bytes {
1911 self.as_bytes().iter().map(|&b| b)
1915 fn char_indices(&self) -> CharOffsets {
1916 CharOffsets{front_offset: 0, iter: self.chars()}
1920 fn split<Sep: CharEq>(&self, sep: Sep) -> CharSplits<Sep> {
1923 only_ascii: sep.only_ascii(),
1925 allow_trailing_empty: true,
1931 fn splitn<Sep: CharEq>(&self, count: uint, sep: Sep)
1932 -> CharSplitsN<Sep> {
1934 iter: self.split(sep),
1941 fn split_terminator<Sep: CharEq>(&self, sep: Sep)
1942 -> CharSplits<Sep> {
1944 allow_trailing_empty: false,
1950 fn rsplitn<Sep: CharEq>(&self, count: uint, sep: Sep)
1951 -> CharSplitsN<Sep> {
1953 iter: self.split(sep),
1960 fn match_indices<'a>(&'a self, sep: &'a str) -> MatchIndices<'a> {
1961 assert!(!sep.is_empty())
1965 searcher: Searcher::new(self.as_bytes(), sep.as_bytes())
1970 fn split_str<'a>(&'a self, sep: &'a str) -> StrSplits<'a> {
1972 it: self.match_indices(sep),
1979 fn lines(&self) -> CharSplits<char> {
1980 self.split_terminator('\n')
1983 fn lines_any(&self) -> AnyLines {
1984 self.lines().map(|line| {
1986 if l > 0 && line.as_bytes()[l - 1] == b'\r' { line.slice(0, l - 1) }
1992 fn char_len(&self) -> uint { self.chars().count() }
1995 fn slice(&self, begin: uint, end: uint) -> &str {
1996 // is_char_boundary checks that the index is in [0, .len()]
1998 self.is_char_boundary(begin) &&
1999 self.is_char_boundary(end) {
2000 unsafe { raw::slice_unchecked(self, begin, end) }
2002 slice_error_fail(self, begin, end)
2007 fn slice_from(&self, begin: uint) -> &str {
2008 // is_char_boundary checks that the index is in [0, .len()]
2009 if self.is_char_boundary(begin) {
2010 unsafe { raw::slice_unchecked(self, begin, self.len()) }
2012 slice_error_fail(self, begin, self.len())
2017 fn slice_to(&self, end: uint) -> &str {
2018 // is_char_boundary checks that the index is in [0, .len()]
2019 if self.is_char_boundary(end) {
2020 unsafe { raw::slice_unchecked(self, 0, end) }
2022 slice_error_fail(self, 0, end)
2026 fn slice_chars(&self, begin: uint, end: uint) -> &str {
2027 assert!(begin <= end);
2029 let mut begin_byte = None;
2030 let mut end_byte = None;
2032 // This could be even more efficient by not decoding,
2033 // only finding the char boundaries
2034 for (idx, _) in self.char_indices() {
2035 if count == begin { begin_byte = Some(idx); }
2036 if count == end { end_byte = Some(idx); break; }
2039 if begin_byte.is_none() && count == begin { begin_byte = Some(self.len()) }
2040 if end_byte.is_none() && count == end { end_byte = Some(self.len()) }
2042 match (begin_byte, end_byte) {
2043 (None, _) => panic!("slice_chars: `begin` is beyond end of string"),
2044 (_, None) => panic!("slice_chars: `end` is beyond end of string"),
2045 (Some(a), Some(b)) => unsafe { raw::slice_bytes(self, a, b) }
2050 fn starts_with(&self, needle: &str) -> bool {
2051 let n = needle.len();
2052 self.len() >= n && needle.as_bytes() == self.as_bytes()[..n]
2056 fn ends_with(&self, needle: &str) -> bool {
2057 let (m, n) = (self.len(), needle.len());
2058 m >= n && needle.as_bytes() == self.as_bytes()[m-n..]
2062 fn trim_chars<C: CharEq>(&self, mut to_trim: C) -> &str {
2063 let cur = match self.find(|c: char| !to_trim.matches(c)) {
2065 Some(i) => unsafe { raw::slice_bytes(self, i, self.len()) }
2067 match cur.rfind(|c: char| !to_trim.matches(c)) {
2070 let right = cur.char_range_at(i).next;
2071 unsafe { raw::slice_bytes(cur, 0, right) }
2077 fn trim_left_chars<C: CharEq>(&self, mut to_trim: C) -> &str {
2078 match self.find(|c: char| !to_trim.matches(c)) {
2080 Some(first) => unsafe { raw::slice_bytes(self, first, self.len()) }
2085 fn trim_right_chars<C: CharEq>(&self, mut to_trim: C) -> &str {
2086 match self.rfind(|c: char| !to_trim.matches(c)) {
2089 let next = self.char_range_at(last).next;
2090 unsafe { raw::slice_bytes(self, 0u, next) }
2096 fn is_char_boundary(&self, index: uint) -> bool {
2097 if index == self.len() { return true; }
2098 match self.as_bytes().get(index) {
2100 Some(&b) => b < 128u8 || b >= 192u8,
2105 fn char_range_at(&self, i: uint) -> CharRange {
2106 if self.as_bytes()[i] < 128u8 {
2107 return CharRange {ch: self.as_bytes()[i] as char, next: i + 1 };
2110 // Multibyte case is a fn to allow char_range_at to inline cleanly
2111 fn multibyte_char_range_at(s: &str, i: uint) -> CharRange {
2112 let mut val = s.as_bytes()[i] as u32;
2113 let w = UTF8_CHAR_WIDTH[val as uint] as uint;
2116 val = utf8_first_byte!(val, w);
2117 val = utf8_acc_cont_byte!(val, s.as_bytes()[i + 1]);
2118 if w > 2 { val = utf8_acc_cont_byte!(val, s.as_bytes()[i + 2]); }
2119 if w > 3 { val = utf8_acc_cont_byte!(val, s.as_bytes()[i + 3]); }
2121 return CharRange {ch: unsafe { mem::transmute(val) }, next: i + w};
2124 return multibyte_char_range_at(self, i);
2128 fn char_range_at_reverse(&self, start: uint) -> CharRange {
2129 let mut prev = start;
2131 prev = prev.saturating_sub(1);
2132 if self.as_bytes()[prev] < 128 {
2133 return CharRange{ch: self.as_bytes()[prev] as char, next: prev}
2136 // Multibyte case is a fn to allow char_range_at_reverse to inline cleanly
2137 fn multibyte_char_range_at_reverse(s: &str, mut i: uint) -> CharRange {
2138 // while there is a previous byte == 10......
2139 while i > 0 && s.as_bytes()[i] & !CONT_MASK == TAG_CONT_U8 {
2143 let mut val = s.as_bytes()[i] as u32;
2144 let w = UTF8_CHAR_WIDTH[val as uint] as uint;
2147 val = utf8_first_byte!(val, w);
2148 val = utf8_acc_cont_byte!(val, s.as_bytes()[i + 1]);
2149 if w > 2 { val = utf8_acc_cont_byte!(val, s.as_bytes()[i + 2]); }
2150 if w > 3 { val = utf8_acc_cont_byte!(val, s.as_bytes()[i + 3]); }
2152 return CharRange {ch: unsafe { mem::transmute(val) }, next: i};
2155 return multibyte_char_range_at_reverse(self, prev);
2159 fn char_at(&self, i: uint) -> char {
2160 self.char_range_at(i).ch
2164 fn char_at_reverse(&self, i: uint) -> char {
2165 self.char_range_at_reverse(i).ch
2169 fn as_bytes(&self) -> &[u8] {
2170 unsafe { mem::transmute(self) }
2173 fn find<C: CharEq>(&self, mut search: C) -> Option<uint> {
2174 if search.only_ascii() {
2175 self.bytes().position(|b| search.matches(b as char))
2177 for (index, c) in self.char_indices() {
2178 if search.matches(c) { return Some(index); }
2184 fn rfind<C: CharEq>(&self, mut search: C) -> Option<uint> {
2185 if search.only_ascii() {
2186 self.bytes().rposition(|b| search.matches(b as char))
2188 for (index, c) in self.char_indices().rev() {
2189 if search.matches(c) { return Some(index); }
2195 fn find_str(&self, needle: &str) -> Option<uint> {
2196 if needle.is_empty() {
2199 self.match_indices(needle)
2201 .map(|(start, _end)| start)
2206 fn slice_shift_char(&self) -> (Option<char>, &str) {
2207 if self.is_empty() {
2208 return (None, self);
2210 let CharRange {ch, next} = self.char_range_at(0u);
2211 let next_s = unsafe { raw::slice_bytes(self, next, self.len()) };
2212 return (Some(ch), next_s);
2216 fn subslice_offset(&self, inner: &str) -> uint {
2217 let a_start = self.as_ptr() as uint;
2218 let a_end = a_start + self.len();
2219 let b_start = inner.as_ptr() as uint;
2220 let b_end = b_start + inner.len();
2222 assert!(a_start <= b_start);
2223 assert!(b_end <= a_end);
2228 fn as_ptr(&self) -> *const u8 {
2233 fn utf16_units(&self) -> Utf16CodeUnits {
2234 Utf16CodeUnits{ chars: self.chars(), extra: 0}
2238 fn len(&self) -> uint { self.repr().len }
2241 impl<'a> Default for &'a str {
2242 fn default() -> &'a str { "" }