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 //! String manipulation
13 //! For more details, see std::str
15 #![doc(primitive = "str")]
17 use self::OldSearcher::{TwoWay, TwoWayLong};
18 use self::pattern::Pattern;
19 use self::pattern::{Searcher, ReverseSearcher, DoubleEndedSearcher};
27 use iter::ExactSizeIterator;
28 use iter::{Map, Iterator, DoubleEndedIterator};
30 use ops::{Fn, FnMut, FnOnce};
31 use option::Option::{self, None, Some};
32 use raw::{Repr, Slice};
33 use result::Result::{self, Ok, Err};
34 use slice::{self, SliceExt};
39 /// A trait to abstract the idea of creating a new instance of a type from a
41 #[stable(feature = "rust1", since = "1.0.0")]
43 /// The associated error which can be returned from parsing.
44 #[stable(feature = "rust1", since = "1.0.0")]
47 /// Parses a string `s` to return a value of this type.
49 /// If parsing succeeds, return the value inside `Ok`, otherwise
50 /// when the string is ill-formatted return an error specific to the
51 /// inside `Err`. The error type is specific to implementation of the trait.
52 #[stable(feature = "rust1", since = "1.0.0")]
53 fn from_str(s: &str) -> Result<Self, Self::Err>;
56 #[stable(feature = "rust1", since = "1.0.0")]
57 impl FromStr for bool {
58 type Err = ParseBoolError;
60 /// Parse a `bool` from a string.
62 /// Yields a `Result<bool, ParseBoolError>`, because `s` may or may not
63 /// actually be parseable.
68 /// use std::str::FromStr;
70 /// assert_eq!(FromStr::from_str("true"), Ok(true));
71 /// assert_eq!(FromStr::from_str("false"), Ok(false));
72 /// assert!(<bool as FromStr>::from_str("not even a boolean").is_err());
75 /// Note, in many cases, the `.parse()` method on `str` is more proper.
78 /// assert_eq!("true".parse(), Ok(true));
79 /// assert_eq!("false".parse(), Ok(false));
80 /// assert!("not even a boolean".parse::<bool>().is_err());
83 fn from_str(s: &str) -> Result<bool, ParseBoolError> {
87 _ => Err(ParseBoolError { _priv: () }),
92 /// An error returned when parsing a `bool` from a string fails.
93 #[derive(Debug, Clone, PartialEq)]
94 #[stable(feature = "rust1", since = "1.0.0")]
95 pub struct ParseBoolError { _priv: () }
97 #[stable(feature = "rust1", since = "1.0.0")]
98 impl fmt::Display for ParseBoolError {
99 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
100 "provided string was not `true` or `false`".fmt(f)
105 Section: Creating a string
108 /// Errors which can occur when attempting to interpret a byte slice as a `str`.
109 #[derive(Copy, Eq, PartialEq, Clone, Debug)]
110 #[stable(feature = "rust1", since = "1.0.0")]
111 pub struct Utf8Error {
116 /// Returns the index in the given string up to which valid UTF-8 was
119 /// Starting at the index provided, but not necessarily at it precisely, an
120 /// invalid UTF-8 encoding sequence was found.
121 #[unstable(feature = "utf8_error", reason = "method just added")]
122 pub fn valid_up_to(&self) -> usize { self.valid_up_to }
125 /// Converts a slice of bytes to a string slice without performing any
128 /// Once the slice has been validated as utf-8, it is transmuted in-place and
129 /// returned as a '&str' instead of a '&[u8]'
133 /// Returns `Err` if the slice is not utf-8 with a description as to why the
134 /// provided slice is not utf-8.
135 #[stable(feature = "rust1", since = "1.0.0")]
136 pub fn from_utf8(v: &[u8]) -> Result<&str, Utf8Error> {
137 try!(run_utf8_validation_iterator(&mut v.iter()));
138 Ok(unsafe { from_utf8_unchecked(v) })
141 /// Converts a slice of bytes to a string slice without checking
142 /// that the string contains valid UTF-8.
144 #[stable(feature = "rust1", since = "1.0.0")]
145 pub unsafe fn from_utf8_unchecked<'a>(v: &'a [u8]) -> &'a str {
149 #[stable(feature = "rust1", since = "1.0.0")]
150 impl fmt::Display for Utf8Error {
151 fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
152 write!(f, "invalid utf-8: invalid byte near index {}", self.valid_up_to)
160 /// Iterator for the char (representing *Unicode Scalar Values*) of a string
162 /// Created with the method `.chars()`.
164 #[stable(feature = "rust1", since = "1.0.0")]
165 pub struct Chars<'a> {
166 iter: slice::Iter<'a, u8>
169 /// Return the initial codepoint accumulator for the first byte.
170 /// The first byte is special, only want bottom 5 bits for width 2, 4 bits
171 /// for width 3, and 3 bits for width 4.
173 fn utf8_first_byte(byte: u8, width: u32) -> u32 { (byte & (0x7F >> width)) as u32 }
175 /// Return the value of `ch` updated with continuation byte `byte`.
177 fn utf8_acc_cont_byte(ch: u32, byte: u8) -> u32 { (ch << 6) | (byte & CONT_MASK) as u32 }
179 /// Checks whether the byte is a UTF-8 continuation byte (i.e. starts with the
182 fn utf8_is_cont_byte(byte: u8) -> bool { (byte & !CONT_MASK) == TAG_CONT_U8 }
185 fn unwrap_or_0(opt: Option<&u8>) -> u8 {
192 /// Reads the next code point out of a byte iterator (assuming a
193 /// UTF-8-like encoding).
194 #[unstable(feature = "core")]
196 pub fn next_code_point(bytes: &mut slice::Iter<u8>) -> Option<u32> {
198 let x = match bytes.next() {
200 Some(&next_byte) if next_byte < 128 => return Some(next_byte as u32),
201 Some(&next_byte) => next_byte,
204 // Multibyte case follows
205 // Decode from a byte combination out of: [[[x y] z] w]
206 // NOTE: Performance is sensitive to the exact formulation here
207 let init = utf8_first_byte(x, 2);
208 let y = unwrap_or_0(bytes.next());
209 let mut ch = utf8_acc_cont_byte(init, y);
212 // 5th bit in 0xE0 .. 0xEF is always clear, so `init` is still valid
213 let z = unwrap_or_0(bytes.next());
214 let y_z = utf8_acc_cont_byte((y & CONT_MASK) as u32, z);
215 ch = init << 12 | y_z;
218 // use only the lower 3 bits of `init`
219 let w = unwrap_or_0(bytes.next());
220 ch = (init & 7) << 18 | utf8_acc_cont_byte(y_z, w);
227 /// Reads the last code point out of a byte iterator (assuming a
228 /// UTF-8-like encoding).
229 #[unstable(feature = "core")]
231 pub fn next_code_point_reverse(bytes: &mut slice::Iter<u8>) -> Option<u32> {
233 let w = match bytes.next_back() {
235 Some(&next_byte) if next_byte < 128 => return Some(next_byte as u32),
236 Some(&back_byte) => back_byte,
239 // Multibyte case follows
240 // Decode from a byte combination out of: [x [y [z w]]]
242 let z = unwrap_or_0(bytes.next_back());
243 ch = utf8_first_byte(z, 2);
244 if utf8_is_cont_byte(z) {
245 let y = unwrap_or_0(bytes.next_back());
246 ch = utf8_first_byte(y, 3);
247 if utf8_is_cont_byte(y) {
248 let x = unwrap_or_0(bytes.next_back());
249 ch = utf8_first_byte(x, 4);
250 ch = utf8_acc_cont_byte(ch, y);
252 ch = utf8_acc_cont_byte(ch, z);
254 ch = utf8_acc_cont_byte(ch, w);
259 #[stable(feature = "rust1", since = "1.0.0")]
260 impl<'a> Iterator for Chars<'a> {
264 fn next(&mut self) -> Option<char> {
265 next_code_point(&mut self.iter).map(|ch| {
266 // str invariant says `ch` is a valid Unicode Scalar Value
274 fn size_hint(&self) -> (usize, Option<usize>) {
275 let (len, _) = self.iter.size_hint();
276 // `(len + 3)` can't overflow, because we know that the `slice::Iter`
277 // belongs to a slice in memory which has a maximum length of
278 // `isize::MAX` (that's well below `usize::MAX`).
279 ((len + 3) / 4, Some(len))
283 #[stable(feature = "rust1", since = "1.0.0")]
284 impl<'a> DoubleEndedIterator for Chars<'a> {
286 fn next_back(&mut self) -> Option<char> {
287 next_code_point_reverse(&mut self.iter).map(|ch| {
288 // str invariant says `ch` is a valid Unicode Scalar Value
296 /// Iterator for a string's characters and their byte offsets.
298 #[stable(feature = "rust1", since = "1.0.0")]
299 pub struct CharIndices<'a> {
304 #[stable(feature = "rust1", since = "1.0.0")]
305 impl<'a> Iterator for CharIndices<'a> {
306 type Item = (usize, char);
309 fn next(&mut self) -> Option<(usize, char)> {
310 let (pre_len, _) = self.iter.iter.size_hint();
311 match self.iter.next() {
314 let index = self.front_offset;
315 let (len, _) = self.iter.iter.size_hint();
316 self.front_offset += pre_len - len;
323 fn size_hint(&self) -> (usize, Option<usize>) {
324 self.iter.size_hint()
328 #[stable(feature = "rust1", since = "1.0.0")]
329 impl<'a> DoubleEndedIterator for CharIndices<'a> {
331 fn next_back(&mut self) -> Option<(usize, char)> {
332 match self.iter.next_back() {
335 let (len, _) = self.iter.iter.size_hint();
336 let index = self.front_offset + len;
343 /// External iterator for a string's bytes.
344 /// Use with the `std::iter` module.
346 /// Created with the method `.bytes()`.
347 #[stable(feature = "rust1", since = "1.0.0")]
349 pub struct Bytes<'a>(Map<slice::Iter<'a, u8>, BytesDeref>);
351 /// A nameable, clonable fn type
355 impl<'a> Fn<(&'a u8,)> for BytesDeref {
357 extern "rust-call" fn call(&self, (ptr,): (&'a u8,)) -> u8 {
362 impl<'a> FnMut<(&'a u8,)> for BytesDeref {
364 extern "rust-call" fn call_mut(&mut self, (ptr,): (&'a u8,)) -> u8 {
365 Fn::call(&*self, (ptr,))
369 impl<'a> FnOnce<(&'a u8,)> for BytesDeref {
373 extern "rust-call" fn call_once(self, (ptr,): (&'a u8,)) -> u8 {
374 Fn::call(&self, (ptr,))
378 #[stable(feature = "rust1", since = "1.0.0")]
379 impl<'a> Iterator for Bytes<'a> {
383 fn next(&mut self) -> Option<u8> {
388 fn size_hint(&self) -> (usize, Option<usize>) {
393 #[stable(feature = "rust1", since = "1.0.0")]
394 impl<'a> DoubleEndedIterator for Bytes<'a> {
396 fn next_back(&mut self) -> Option<u8> {
401 #[stable(feature = "rust1", since = "1.0.0")]
402 impl<'a> ExactSizeIterator for Bytes<'a> {
404 fn len(&self) -> usize {
409 /// This macro generates a Clone impl for string pattern API
410 /// wrapper types of the form X<'a, P>
411 macro_rules! derive_pattern_clone {
412 (clone $t:ident with |$s:ident| $e:expr) => {
413 impl<'a, P: Pattern<'a>> Clone for $t<'a, P>
414 where P::Searcher: Clone
416 fn clone(&self) -> Self {
424 /// This macro generates two public iterator structs
425 /// wrapping an private internal one that makes use of the `Pattern` API.
427 /// For all patterns `P: Pattern<'a>` the following items will be
428 /// generated (generics omitted):
430 /// struct $forward_iterator($internal_iterator);
431 /// struct $reverse_iterator($internal_iterator);
433 /// impl Iterator for $forward_iterator
434 /// { /* internal ends up calling Searcher::next_match() */ }
436 /// impl DoubleEndedIterator for $forward_iterator
437 /// where P::Searcher: DoubleEndedSearcher
438 /// { /* internal ends up calling Searcher::next_match_back() */ }
440 /// impl Iterator for $reverse_iterator
441 /// where P::Searcher: ReverseSearcher
442 /// { /* internal ends up calling Searcher::next_match_back() */ }
444 /// impl DoubleEndedIterator for $reverse_iterator
445 /// where P::Searcher: DoubleEndedSearcher
446 /// { /* internal ends up calling Searcher::next_match() */ }
448 /// The internal one is defined outside the macro, and has almost the same
449 /// semantic as a DoubleEndedIterator by delegating to `pattern::Searcher` and
450 /// `pattern::ReverseSearcher` for both forward and reverse iteration.
452 /// "Almost", because a `Searcher` and a `ReverseSearcher` for a given
453 /// `Pattern` might not return the same elements, so actually implementing
454 /// `DoubleEndedIterator` for it would be incorrect.
455 /// (See the docs in `str::pattern` for more details)
457 /// However, the internal struct still represents a single ended iterator from
458 /// either end, and depending on pattern is also a valid double ended iterator,
459 /// so the two wrapper structs implement `Iterator`
460 /// and `DoubleEndedIterator` depending on the concrete pattern type, leading
461 /// to the complex impls seen above.
462 macro_rules! generate_pattern_iterators {
466 $(#[$forward_iterator_attribute:meta])*
467 struct $forward_iterator:ident;
471 $(#[$reverse_iterator_attribute:meta])*
472 struct $reverse_iterator:ident;
474 // Stability of all generated items
476 $(#[$common_stability_attribute:meta])*
478 // Internal almost-iterator that is being delegated to
480 $internal_iterator:ident yielding ($iterty:ty);
482 // Kind of delgation - either single ended or double ended
485 $(#[$forward_iterator_attribute])*
486 $(#[$common_stability_attribute])*
487 pub struct $forward_iterator<'a, P: Pattern<'a>>($internal_iterator<'a, P>);
489 $(#[$common_stability_attribute])*
490 impl<'a, P: Pattern<'a>> Iterator for $forward_iterator<'a, P> {
494 fn next(&mut self) -> Option<$iterty> {
499 $(#[$common_stability_attribute])*
500 impl<'a, P: Pattern<'a>> Clone for $forward_iterator<'a, P>
501 where P::Searcher: Clone
503 fn clone(&self) -> Self {
504 $forward_iterator(self.0.clone())
508 $(#[$reverse_iterator_attribute])*
509 $(#[$common_stability_attribute])*
510 pub struct $reverse_iterator<'a, P: Pattern<'a>>($internal_iterator<'a, P>);
512 $(#[$common_stability_attribute])*
513 impl<'a, P: Pattern<'a>> Iterator for $reverse_iterator<'a, P>
514 where P::Searcher: ReverseSearcher<'a>
519 fn next(&mut self) -> Option<$iterty> {
524 $(#[$common_stability_attribute])*
525 impl<'a, P: Pattern<'a>> Clone for $reverse_iterator<'a, P>
526 where P::Searcher: Clone
528 fn clone(&self) -> Self {
529 $reverse_iterator(self.0.clone())
533 generate_pattern_iterators!($($t)* with $(#[$common_stability_attribute])*,
535 $reverse_iterator, $iterty);
538 double ended; with $(#[$common_stability_attribute:meta])*,
539 $forward_iterator:ident,
540 $reverse_iterator:ident, $iterty:ty
542 $(#[$common_stability_attribute])*
543 impl<'a, P: Pattern<'a>> DoubleEndedIterator for $forward_iterator<'a, P>
544 where P::Searcher: DoubleEndedSearcher<'a>
547 fn next_back(&mut self) -> Option<$iterty> {
552 $(#[$common_stability_attribute])*
553 impl<'a, P: Pattern<'a>> DoubleEndedIterator for $reverse_iterator<'a, P>
554 where P::Searcher: DoubleEndedSearcher<'a>
557 fn next_back(&mut self) -> Option<$iterty> {
563 single ended; with $(#[$common_stability_attribute:meta])*,
564 $forward_iterator:ident,
565 $reverse_iterator:ident, $iterty:ty
569 derive_pattern_clone!{
571 with |s| SplitInternal { matcher: s.matcher.clone(), ..*s }
573 struct SplitInternal<'a, P: Pattern<'a>> {
576 matcher: P::Searcher,
577 allow_trailing_empty: bool,
581 impl<'a, P: Pattern<'a>> SplitInternal<'a, P> {
583 fn get_end(&mut self) -> Option<&'a str> {
584 if !self.finished && (self.allow_trailing_empty || self.end - self.start > 0) {
585 self.finished = true;
587 let string = self.matcher.haystack().slice_unchecked(self.start, self.end);
596 fn next(&mut self) -> Option<&'a str> {
597 if self.finished { return None }
599 let haystack = self.matcher.haystack();
600 match self.matcher.next_match() {
601 Some((a, b)) => unsafe {
602 let elt = haystack.slice_unchecked(self.start, a);
606 None => self.get_end(),
611 fn next_back(&mut self) -> Option<&'a str>
612 where P::Searcher: ReverseSearcher<'a>
614 if self.finished { return None }
616 if !self.allow_trailing_empty {
617 self.allow_trailing_empty = true;
618 match self.next_back() {
619 Some(elt) if !elt.is_empty() => return Some(elt),
620 _ => if self.finished { return None }
624 let haystack = self.matcher.haystack();
625 match self.matcher.next_match_back() {
626 Some((a, b)) => unsafe {
627 let elt = haystack.slice_unchecked(b, self.end);
632 self.finished = true;
633 Some(haystack.slice_unchecked(self.start, self.end))
639 generate_pattern_iterators! {
641 #[doc="Created with the method `.split()`."]
644 #[doc="Created with the method `.rsplit()`."]
647 #[stable(feature = "rust1", since = "1.0.0")]
649 SplitInternal yielding (&'a str);
650 delegate double ended;
653 generate_pattern_iterators! {
655 #[doc="Created with the method `.split_terminator()`."]
656 struct SplitTerminator;
658 #[doc="Created with the method `.rsplit_terminator()`."]
659 struct RSplitTerminator;
661 #[stable(feature = "rust1", since = "1.0.0")]
663 SplitInternal yielding (&'a str);
664 delegate double ended;
667 derive_pattern_clone!{
669 with |s| SplitNInternal { iter: s.iter.clone(), ..*s }
671 struct SplitNInternal<'a, P: Pattern<'a>> {
672 iter: SplitInternal<'a, P>,
673 /// The number of splits remaining
677 impl<'a, P: Pattern<'a>> SplitNInternal<'a, P> {
679 fn next(&mut self) -> Option<&'a str> {
682 1 => { self.count = 0; self.iter.get_end() }
683 _ => { self.count -= 1; self.iter.next() }
688 fn next_back(&mut self) -> Option<&'a str>
689 where P::Searcher: ReverseSearcher<'a>
693 1 => { self.count = 0; self.iter.get_end() }
694 _ => { self.count -= 1; self.iter.next_back() }
699 generate_pattern_iterators! {
701 #[doc="Created with the method `.splitn()`."]
704 #[doc="Created with the method `.rsplitn()`."]
707 #[stable(feature = "rust1", since = "1.0.0")]
709 SplitNInternal yielding (&'a str);
710 delegate single ended;
713 derive_pattern_clone!{
714 clone MatchIndicesInternal
715 with |s| MatchIndicesInternal(s.0.clone())
717 struct MatchIndicesInternal<'a, P: Pattern<'a>>(P::Searcher);
719 impl<'a, P: Pattern<'a>> MatchIndicesInternal<'a, P> {
721 fn next(&mut self) -> Option<(usize, usize)> {
726 fn next_back(&mut self) -> Option<(usize, usize)>
727 where P::Searcher: ReverseSearcher<'a>
729 self.0.next_match_back()
733 generate_pattern_iterators! {
735 #[doc="Created with the method `.match_indices()`."]
738 #[doc="Created with the method `.rmatch_indices()`."]
739 struct RMatchIndices;
741 #[unstable(feature = "core",
742 reason = "type may be removed or have its iterator impl changed")]
744 MatchIndicesInternal yielding ((usize, usize));
745 delegate double ended;
748 derive_pattern_clone!{
749 clone MatchesInternal
750 with |s| MatchesInternal(s.0.clone())
752 struct MatchesInternal<'a, P: Pattern<'a>>(P::Searcher);
754 impl<'a, P: Pattern<'a>> MatchesInternal<'a, P> {
756 fn next(&mut self) -> Option<&'a str> {
757 self.0.next_match().map(|(a, b)| unsafe {
758 // Indices are known to be on utf8 boundaries
759 self.0.haystack().slice_unchecked(a, b)
764 fn next_back(&mut self) -> Option<&'a str>
765 where P::Searcher: ReverseSearcher<'a>
767 self.0.next_match_back().map(|(a, b)| unsafe {
768 // Indices are known to be on utf8 boundaries
769 self.0.haystack().slice_unchecked(a, b)
774 generate_pattern_iterators! {
776 #[doc="Created with the method `.matches()`."]
779 #[doc="Created with the method `.rmatches()`."]
782 #[unstable(feature = "core", reason = "type got recently added")]
784 MatchesInternal yielding (&'a str);
785 delegate double ended;
788 /// Created with the method `.lines()`.
789 #[stable(feature = "rust1", since = "1.0.0")]
791 pub struct Lines<'a>(SplitTerminator<'a, char>);
793 #[stable(feature = "rust1", since = "1.0.0")]
794 impl<'a> Iterator for Lines<'a> {
798 fn next(&mut self) -> Option<&'a str> {
803 fn size_hint(&self) -> (usize, Option<usize>) {
808 #[stable(feature = "rust1", since = "1.0.0")]
809 impl<'a> DoubleEndedIterator for Lines<'a> {
811 fn next_back(&mut self) -> Option<&'a str> {
816 /// Created with the method `.lines_any()`.
817 #[stable(feature = "rust1", since = "1.0.0")]
819 pub struct LinesAny<'a>(Map<Lines<'a>, LinesAnyMap>);
821 /// A nameable, clonable fn type
825 impl<'a> Fn<(&'a str,)> for LinesAnyMap {
827 extern "rust-call" fn call(&self, (line,): (&'a str,)) -> &'a str {
829 if l > 0 && line.as_bytes()[l - 1] == b'\r' { &line[0 .. l - 1] }
834 impl<'a> FnMut<(&'a str,)> for LinesAnyMap {
836 extern "rust-call" fn call_mut(&mut self, (line,): (&'a str,)) -> &'a str {
837 Fn::call(&*self, (line,))
841 impl<'a> FnOnce<(&'a str,)> for LinesAnyMap {
842 type Output = &'a str;
845 extern "rust-call" fn call_once(self, (line,): (&'a str,)) -> &'a str {
846 Fn::call(&self, (line,))
850 #[stable(feature = "rust1", since = "1.0.0")]
851 impl<'a> Iterator for LinesAny<'a> {
855 fn next(&mut self) -> Option<&'a str> {
860 fn size_hint(&self) -> (usize, Option<usize>) {
865 #[stable(feature = "rust1", since = "1.0.0")]
866 impl<'a> DoubleEndedIterator for LinesAny<'a> {
868 fn next_back(&mut self) -> Option<&'a str> {
873 /// The internal state of an iterator that searches for matches of a substring
874 /// within a larger string using two-way search
876 struct TwoWaySearcher {
888 This is the Two-Way search algorithm, which was introduced in the paper:
889 Crochemore, M., Perrin, D., 1991, Two-way string-matching, Journal of the ACM 38(3):651-675.
891 Here's some background information.
893 A *word* is a string of symbols. The *length* of a word should be a familiar
894 notion, and here we denote it for any word x by |x|.
895 (We also allow for the possibility of the *empty word*, a word of length zero).
897 If x is any non-empty word, then an integer p with 0 < p <= |x| is said to be a
898 *period* for x iff for all i with 0 <= i <= |x| - p - 1, we have x[i] == x[i+p].
899 For example, both 1 and 2 are periods for the string "aa". As another example,
900 the only period of the string "abcd" is 4.
902 We denote by period(x) the *smallest* period of x (provided that x is non-empty).
903 This is always well-defined since every non-empty word x has at least one period,
904 |x|. We sometimes call this *the period* of x.
906 If u, v and x are words such that x = uv, where uv is the concatenation of u and
907 v, then we say that (u, v) is a *factorization* of x.
909 Let (u, v) be a factorization for a word x. Then if w is a non-empty word such
910 that both of the following hold
912 - either w is a suffix of u or u is a suffix of w
913 - either w is a prefix of v or v is a prefix of w
915 then w is said to be a *repetition* for the factorization (u, v).
917 Just to unpack this, there are four possibilities here. Let w = "abc". Then we
920 - w is a suffix of u and w is a prefix of v. ex: ("lolabc", "abcde")
921 - w is a suffix of u and v is a prefix of w. ex: ("lolabc", "ab")
922 - u is a suffix of w and w is a prefix of v. ex: ("bc", "abchi")
923 - u is a suffix of w and v is a prefix of w. ex: ("bc", "a")
925 Note that the word vu is a repetition for any factorization (u,v) of x = uv,
926 so every factorization has at least one repetition.
928 If x is a string and (u, v) is a factorization for x, then a *local period* for
929 (u, v) is an integer r such that there is some word w such that |w| = r and w is
930 a repetition for (u, v).
932 We denote by local_period(u, v) the smallest local period of (u, v). We sometimes
933 call this *the local period* of (u, v). Provided that x = uv is non-empty, this
934 is well-defined (because each non-empty word has at least one factorization, as
937 It can be proven that the following is an equivalent definition of a local period
938 for a factorization (u, v): any positive integer r such that x[i] == x[i+r] for
939 all i such that |u| - r <= i <= |u| - 1 and such that both x[i] and x[i+r] are
940 defined. (i.e. i > 0 and i + r < |x|).
942 Using the above reformulation, it is easy to prove that
944 1 <= local_period(u, v) <= period(uv)
946 A factorization (u, v) of x such that local_period(u,v) = period(x) is called a
947 *critical factorization*.
949 The algorithm hinges on the following theorem, which is stated without proof:
951 **Critical Factorization Theorem** Any word x has at least one critical
952 factorization (u, v) such that |u| < period(x).
954 The purpose of maximal_suffix is to find such a critical factorization.
957 impl TwoWaySearcher {
959 fn new(needle: &[u8]) -> TwoWaySearcher {
960 let (crit_pos_false, period_false) = TwoWaySearcher::maximal_suffix(needle, false);
961 let (crit_pos_true, period_true) = TwoWaySearcher::maximal_suffix(needle, true);
963 let (crit_pos, period) =
964 if crit_pos_false > crit_pos_true {
965 (crit_pos_false, period_false)
967 (crit_pos_true, period_true)
970 // This isn't in the original algorithm, as far as I'm aware.
971 let byteset = needle.iter()
972 .fold(0, |a, &b| (1 << ((b & 0x3f) as usize)) | a);
974 // A particularly readable explanation of what's going on here can be found
975 // in Crochemore and Rytter's book "Text Algorithms", ch 13. Specifically
976 // see the code for "Algorithm CP" on p. 323.
978 // What's going on is we have some critical factorization (u, v) of the
979 // needle, and we want to determine whether u is a suffix of
980 // &v[..period]. If it is, we use "Algorithm CP1". Otherwise we use
981 // "Algorithm CP2", which is optimized for when the period of the needle
983 if &needle[..crit_pos] == &needle[period.. period + crit_pos] {
995 period: cmp::max(crit_pos, needle.len() - crit_pos) + 1,
999 memory: usize::MAX // Dummy value to signify that the period is long
1004 // One of the main ideas of Two-Way is that we factorize the needle into
1005 // two halves, (u, v), and begin trying to find v in the haystack by scanning
1006 // left to right. If v matches, we try to match u by scanning right to left.
1007 // How far we can jump when we encounter a mismatch is all based on the fact
1008 // that (u, v) is a critical factorization for the needle.
1010 fn next(&mut self, haystack: &[u8], needle: &[u8], long_period: bool)
1011 -> Option<(usize, usize)> {
1013 // Check that we have room to search in
1014 if self.position + needle.len() > haystack.len() {
1018 // Quickly skip by large portions unrelated to our substring
1020 ((haystack[self.position + needle.len() - 1] & 0x3f)
1021 as usize)) & 1 == 0 {
1022 self.position += needle.len();
1029 // See if the right part of the needle matches
1030 let start = if long_period { self.crit_pos }
1031 else { cmp::max(self.crit_pos, self.memory) };
1032 for i in start..needle.len() {
1033 if needle[i] != haystack[self.position + i] {
1034 self.position += i - self.crit_pos + 1;
1042 // See if the left part of the needle matches
1043 let start = if long_period { 0 } else { self.memory };
1044 for i in (start..self.crit_pos).rev() {
1045 if needle[i] != haystack[self.position + i] {
1046 self.position += self.period;
1048 self.memory = needle.len() - self.period;
1054 // We have found a match!
1055 let match_pos = self.position;
1056 self.position += needle.len(); // add self.period for all matches
1058 self.memory = 0; // set to needle.len() - self.period for all matches
1060 return Some((match_pos, match_pos + needle.len()));
1064 // Computes a critical factorization (u, v) of `arr`.
1065 // Specifically, returns (i, p), where i is the starting index of v in some
1066 // critical factorization (u, v) and p = period(v)
1069 #[allow(deprecated)]
1070 fn maximal_suffix(arr: &[u8], reversed: bool) -> (usize, usize) {
1071 let mut left: usize = !0; // Corresponds to i in the paper
1072 let mut right = 0; // Corresponds to j in the paper
1073 let mut offset = 1; // Corresponds to k in the paper
1074 let mut period = 1; // Corresponds to p in the paper
1076 while right + offset < arr.len() {
1080 a = arr[left.wrapping_add(offset)];
1081 b = arr[right + offset];
1083 a = arr[right + offset];
1084 b = arr[left.wrapping_add(offset)];
1087 // Suffix is smaller, period is entire prefix so far.
1090 period = right.wrapping_sub(left);
1092 // Advance through repetition of the current period.
1093 if offset == period {
1100 // Suffix is larger, start over from current location.
1107 (left.wrapping_add(1), period)
1111 /// The internal state of an iterator that searches for matches of a substring
1112 /// within a larger string using a dynamically chosen search algorithm
1114 // NB: This is kept around for convenience because
1115 // it is planned to be used again in the future
1117 TwoWay(TwoWaySearcher),
1118 TwoWayLong(TwoWaySearcher),
1123 fn new(haystack: &[u8], needle: &[u8]) -> OldSearcher {
1124 if needle.is_empty() {
1127 // FIXME: Tune this.
1128 // FIXME(#16715): This unsigned integer addition will probably not
1129 // overflow because that would mean that the memory almost solely
1130 // consists of the needle. Needs #16715 to be formally fixed.
1131 } else if needle.len() + 20 > haystack.len() {
1132 // Use naive searcher
1135 let searcher = TwoWaySearcher::new(needle);
1136 if searcher.memory == usize::MAX { // If the period is long
1137 TwoWayLong(searcher)
1146 // NB: This is kept around for convenience because
1147 // it is planned to be used again in the future
1148 struct OldMatchIndices<'a, 'b> {
1152 searcher: OldSearcher
1155 impl<'a, 'b> OldMatchIndices<'a, 'b> {
1158 fn next(&mut self) -> Option<(usize, usize)> {
1159 match self.searcher {
1160 TwoWay(ref mut searcher)
1161 => searcher.next(self.haystack.as_bytes(), self.needle.as_bytes(), false),
1162 TwoWayLong(ref mut searcher)
1163 => searcher.next(self.haystack.as_bytes(), self.needle.as_bytes(), true),
1169 Section: Comparing strings
1172 // share the implementation of the lang-item vs. non-lang-item
1174 /// NOTE: This function is (ab)used in rustc::middle::trans::_match
1175 /// to compare &[u8] byte slices that are not necessarily valid UTF-8.
1177 fn eq_slice_(a: &str, b: &str) -> bool {
1178 // NOTE: In theory n should be libc::size_t and not usize, but libc is not available here
1179 #[allow(improper_ctypes)]
1180 extern { fn memcmp(s1: *const i8, s2: *const i8, n: usize) -> i32; }
1181 a.len() == b.len() && unsafe {
1182 memcmp(a.as_ptr() as *const i8,
1183 b.as_ptr() as *const i8,
1188 /// Bytewise slice equality
1189 /// NOTE: This function is (ab)used in rustc::middle::trans::_match
1190 /// to compare &[u8] byte slices that are not necessarily valid UTF-8.
1193 fn eq_slice(a: &str, b: &str) -> bool {
1201 /// Walk through `iter` checking that it's a valid UTF-8 sequence,
1202 /// returning `true` in that case, or, if it is invalid, `false` with
1203 /// `iter` reset such that it is pointing at the first byte in the
1204 /// invalid sequence.
1206 fn run_utf8_validation_iterator(iter: &mut slice::Iter<u8>)
1207 -> Result<(), Utf8Error> {
1208 let whole = iter.as_slice();
1210 // save the current thing we're pointing at.
1211 let old = iter.clone();
1213 // restore the iterator we had at the start of this codepoint.
1214 macro_rules! err { () => {{
1215 *iter = old.clone();
1216 return Err(Utf8Error {
1217 valid_up_to: whole.len() - iter.as_slice().len()
1221 macro_rules! next { () => {
1224 // we needed data, but there was none: error!
1229 let first = match iter.next() {
1231 // we're at the end of the iterator and a codepoint
1232 // boundary at the same time, so this string is valid.
1233 None => return Ok(())
1236 // ASCII characters are always valid, so only large
1237 // bytes need more examination.
1239 let w = UTF8_CHAR_WIDTH[first as usize];
1240 let second = next!();
1241 // 2-byte encoding is for codepoints \u{0080} to \u{07ff}
1242 // first C2 80 last DF BF
1243 // 3-byte encoding is for codepoints \u{0800} to \u{ffff}
1244 // first E0 A0 80 last EF BF BF
1245 // excluding surrogates codepoints \u{d800} to \u{dfff}
1246 // ED A0 80 to ED BF BF
1247 // 4-byte encoding is for codepoints \u{1000}0 to \u{10ff}ff
1248 // first F0 90 80 80 last F4 8F BF BF
1250 // Use the UTF-8 syntax from the RFC
1252 // https://tools.ietf.org/html/rfc3629
1254 // UTF8-2 = %xC2-DF UTF8-tail
1255 // UTF8-3 = %xE0 %xA0-BF UTF8-tail / %xE1-EC 2( UTF8-tail ) /
1256 // %xED %x80-9F UTF8-tail / %xEE-EF 2( UTF8-tail )
1257 // UTF8-4 = %xF0 %x90-BF 2( UTF8-tail ) / %xF1-F3 3( UTF8-tail ) /
1258 // %xF4 %x80-8F 2( UTF8-tail )
1260 2 => if second & !CONT_MASK != TAG_CONT_U8 {err!()},
1262 match (first, second, next!() & !CONT_MASK) {
1263 (0xE0 , 0xA0 ... 0xBF, TAG_CONT_U8) |
1264 (0xE1 ... 0xEC, 0x80 ... 0xBF, TAG_CONT_U8) |
1265 (0xED , 0x80 ... 0x9F, TAG_CONT_U8) |
1266 (0xEE ... 0xEF, 0x80 ... 0xBF, TAG_CONT_U8) => {}
1271 match (first, second, next!() & !CONT_MASK, next!() & !CONT_MASK) {
1272 (0xF0 , 0x90 ... 0xBF, TAG_CONT_U8, TAG_CONT_U8) |
1273 (0xF1 ... 0xF3, 0x80 ... 0xBF, TAG_CONT_U8, TAG_CONT_U8) |
1274 (0xF4 , 0x80 ... 0x8F, TAG_CONT_U8, TAG_CONT_U8) => {}
1284 // https://tools.ietf.org/html/rfc3629
1285 static UTF8_CHAR_WIDTH: [u8; 256] = [
1286 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1287 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x1F
1288 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1289 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x3F
1290 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1291 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x5F
1292 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1293 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, // 0x7F
1294 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
1295 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, // 0x9F
1296 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
1297 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, // 0xBF
1298 0,0,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
1299 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, // 0xDF
1300 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, // 0xEF
1301 4,4,4,4,4,0,0,0,0,0,0,0,0,0,0,0, // 0xFF
1304 /// Struct that contains a `char` and the index of the first byte of
1305 /// the next `char` in a string. This can be used as a data structure
1306 /// for iterating over the UTF-8 bytes of a string.
1307 #[derive(Copy, Clone)]
1308 #[unstable(feature = "str_char",
1309 reason = "existence of this struct is uncertain as it is frequently \
1310 able to be replaced with char.len_utf8() and/or \
1311 char/char_indices iterators")]
1312 pub struct CharRange {
1315 /// Index of the first byte of the next `char`
1319 /// Mask of the value bits of a continuation byte
1320 const CONT_MASK: u8 = 0b0011_1111;
1321 /// Value of the tag bits (tag mask is !CONT_MASK) of a continuation byte
1322 const TAG_CONT_U8: u8 = 0b1000_0000;
1325 Section: Trait implementations
1329 use cmp::{Ordering, Ord, PartialEq, PartialOrd, Eq};
1330 use cmp::Ordering::{Less, Equal, Greater};
1333 use option::Option::Some;
1335 use str::{StrExt, eq_slice};
1337 #[stable(feature = "rust1", since = "1.0.0")]
1340 fn cmp(&self, other: &str) -> Ordering {
1341 for (s_b, o_b) in self.bytes().zip(other.bytes()) {
1342 match s_b.cmp(&o_b) {
1343 Greater => return Greater,
1344 Less => return Less,
1349 self.len().cmp(&other.len())
1353 #[stable(feature = "rust1", since = "1.0.0")]
1354 impl PartialEq for str {
1356 fn eq(&self, other: &str) -> bool {
1357 eq_slice(self, other)
1360 fn ne(&self, other: &str) -> bool { !(*self).eq(other) }
1363 #[stable(feature = "rust1", since = "1.0.0")]
1366 #[stable(feature = "rust1", since = "1.0.0")]
1367 impl PartialOrd for str {
1369 fn partial_cmp(&self, other: &str) -> Option<Ordering> {
1370 Some(self.cmp(other))
1374 /// Returns a slice of the given string from the byte range
1375 /// [`begin`..`end`).
1377 /// This operation is `O(1)`.
1379 /// Panics when `begin` and `end` do not point to valid characters
1380 /// or point beyond the last character of the string.
1385 /// let s = "Löwe 老虎 Léopard";
1386 /// assert_eq!(&s[0 .. 1], "L");
1388 /// assert_eq!(&s[1 .. 9], "öwe 老");
1390 /// // these will panic:
1391 /// // byte 2 lies within `ö`:
1394 /// // byte 8 lies within `老`
1397 /// // byte 100 is outside the string
1398 /// // &s[3 .. 100];
1400 #[stable(feature = "rust1", since = "1.0.0")]
1401 impl ops::Index<ops::Range<usize>> for str {
1404 fn index(&self, index: ops::Range<usize>) -> &str {
1405 // is_char_boundary checks that the index is in [0, .len()]
1406 if index.start <= index.end &&
1407 self.is_char_boundary(index.start) &&
1408 self.is_char_boundary(index.end) {
1409 unsafe { self.slice_unchecked(index.start, index.end) }
1411 super::slice_error_fail(self, index.start, index.end)
1416 /// Returns a slice of the string from the beginning to byte
1419 /// Equivalent to `self[0 .. end]`.
1421 /// Panics when `end` does not point to a valid character, or is
1423 #[stable(feature = "rust1", since = "1.0.0")]
1424 impl ops::Index<ops::RangeTo<usize>> for str {
1428 fn index(&self, index: ops::RangeTo<usize>) -> &str {
1429 // is_char_boundary checks that the index is in [0, .len()]
1430 if self.is_char_boundary(index.end) {
1431 unsafe { self.slice_unchecked(0, index.end) }
1433 super::slice_error_fail(self, 0, index.end)
1438 /// Returns a slice of the string from `begin` to its end.
1440 /// Equivalent to `self[begin .. self.len()]`.
1442 /// Panics when `begin` does not point to a valid character, or is
1444 #[stable(feature = "rust1", since = "1.0.0")]
1445 impl ops::Index<ops::RangeFrom<usize>> for str {
1449 fn index(&self, index: ops::RangeFrom<usize>) -> &str {
1450 // is_char_boundary checks that the index is in [0, .len()]
1451 if self.is_char_boundary(index.start) {
1452 unsafe { self.slice_unchecked(index.start, self.len()) }
1454 super::slice_error_fail(self, index.start, self.len())
1459 #[stable(feature = "rust1", since = "1.0.0")]
1460 impl ops::Index<ops::RangeFull> for str {
1464 fn index(&self, _index: ops::RangeFull) -> &str {
1470 /// Methods for string slices
1471 #[allow(missing_docs)]
1474 // NB there are no docs here are they're all located on the StrExt trait in
1475 // libcollections, not here.
1477 fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool;
1478 fn contains_char<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool;
1479 fn chars<'a>(&'a self) -> Chars<'a>;
1480 fn bytes<'a>(&'a self) -> Bytes<'a>;
1481 fn char_indices<'a>(&'a self) -> CharIndices<'a>;
1482 fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P>;
1483 fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
1484 where P::Searcher: ReverseSearcher<'a>;
1485 fn splitn<'a, P: Pattern<'a>>(&'a self, count: usize, pat: P) -> SplitN<'a, P>;
1486 fn rsplitn<'a, P: Pattern<'a>>(&'a self, count: usize, pat: P) -> RSplitN<'a, P>
1487 where P::Searcher: ReverseSearcher<'a>;
1488 fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P>;
1489 fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1490 where P::Searcher: ReverseSearcher<'a>;
1491 fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P>;
1492 fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
1493 where P::Searcher: ReverseSearcher<'a>;
1494 fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P>;
1495 fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
1496 where P::Searcher: ReverseSearcher<'a>;
1497 fn lines<'a>(&'a self) -> Lines<'a>;
1498 fn lines_any<'a>(&'a self) -> LinesAny<'a>;
1499 fn char_len(&self) -> usize;
1500 fn slice_chars<'a>(&'a self, begin: usize, end: usize) -> &'a str;
1501 unsafe fn slice_unchecked<'a>(&'a self, begin: usize, end: usize) -> &'a str;
1502 fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool;
1503 fn ends_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool
1504 where P::Searcher: ReverseSearcher<'a>;
1505 fn trim_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1506 where P::Searcher: DoubleEndedSearcher<'a>;
1507 fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str;
1508 fn trim_right_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1509 where P::Searcher: ReverseSearcher<'a>;
1510 fn is_char_boundary(&self, index: usize) -> bool;
1511 fn char_range_at(&self, start: usize) -> CharRange;
1512 fn char_range_at_reverse(&self, start: usize) -> CharRange;
1513 fn char_at(&self, i: usize) -> char;
1514 fn char_at_reverse(&self, i: usize) -> char;
1515 fn as_bytes<'a>(&'a self) -> &'a [u8];
1516 fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>;
1517 fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
1518 where P::Searcher: ReverseSearcher<'a>;
1519 fn find_str<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>;
1520 fn split_at(&self, mid: usize) -> (&str, &str);
1521 fn slice_shift_char<'a>(&'a self) -> Option<(char, &'a str)>;
1522 fn subslice_offset(&self, inner: &str) -> usize;
1523 fn as_ptr(&self) -> *const u8;
1524 fn len(&self) -> usize;
1525 fn is_empty(&self) -> bool;
1526 fn parse<T: FromStr>(&self) -> Result<T, T::Err>;
1530 fn slice_error_fail(s: &str, begin: usize, end: usize) -> ! {
1531 assert!(begin <= end);
1532 panic!("index {} and/or {} in `{}` do not lie on character boundary",
1536 impl StrExt for str {
1538 fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
1539 pat.is_contained_in(self)
1543 fn contains_char<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
1544 pat.is_contained_in(self)
1548 fn chars(&self) -> Chars {
1549 Chars{iter: self.as_bytes().iter()}
1553 fn bytes(&self) -> Bytes {
1554 Bytes(self.as_bytes().iter().map(BytesDeref))
1558 fn char_indices(&self) -> CharIndices {
1559 CharIndices { front_offset: 0, iter: self.chars() }
1563 fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
1564 Split(SplitInternal {
1567 matcher: pat.into_searcher(self),
1568 allow_trailing_empty: true,
1574 fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
1575 where P::Searcher: ReverseSearcher<'a>
1577 RSplit(self.split(pat).0)
1581 fn splitn<'a, P: Pattern<'a>>(&'a self, count: usize, pat: P) -> SplitN<'a, P> {
1582 SplitN(SplitNInternal {
1583 iter: self.split(pat).0,
1589 fn rsplitn<'a, P: Pattern<'a>>(&'a self, count: usize, pat: P) -> RSplitN<'a, P>
1590 where P::Searcher: ReverseSearcher<'a>
1592 RSplitN(self.splitn(count, pat).0)
1596 fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
1597 SplitTerminator(SplitInternal {
1598 allow_trailing_empty: false,
1604 fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
1605 where P::Searcher: ReverseSearcher<'a>
1607 RSplitTerminator(self.split_terminator(pat).0)
1611 fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
1612 Matches(MatchesInternal(pat.into_searcher(self)))
1616 fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
1617 where P::Searcher: ReverseSearcher<'a>
1619 RMatches(self.matches(pat).0)
1623 fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
1624 MatchIndices(MatchIndicesInternal(pat.into_searcher(self)))
1628 fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
1629 where P::Searcher: ReverseSearcher<'a>
1631 RMatchIndices(self.match_indices(pat).0)
1634 fn lines(&self) -> Lines {
1635 Lines(self.split_terminator('\n'))
1639 fn lines_any(&self) -> LinesAny {
1640 LinesAny(self.lines().map(LinesAnyMap))
1644 fn char_len(&self) -> usize { self.chars().count() }
1646 fn slice_chars(&self, begin: usize, end: usize) -> &str {
1647 assert!(begin <= end);
1649 let mut begin_byte = None;
1650 let mut end_byte = None;
1652 // This could be even more efficient by not decoding,
1653 // only finding the char boundaries
1654 for (idx, _) in self.char_indices() {
1655 if count == begin { begin_byte = Some(idx); }
1656 if count == end { end_byte = Some(idx); break; }
1659 if begin_byte.is_none() && count == begin { begin_byte = Some(self.len()) }
1660 if end_byte.is_none() && count == end { end_byte = Some(self.len()) }
1662 match (begin_byte, end_byte) {
1663 (None, _) => panic!("slice_chars: `begin` is beyond end of string"),
1664 (_, None) => panic!("slice_chars: `end` is beyond end of string"),
1665 (Some(a), Some(b)) => unsafe { self.slice_unchecked(a, b) }
1670 unsafe fn slice_unchecked(&self, begin: usize, end: usize) -> &str {
1671 mem::transmute(Slice {
1672 data: self.as_ptr().offset(begin as isize),
1678 fn starts_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
1679 pat.is_prefix_of(self)
1683 fn ends_with<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool
1684 where P::Searcher: ReverseSearcher<'a>
1686 pat.is_suffix_of(self)
1690 fn trim_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1691 where P::Searcher: DoubleEndedSearcher<'a>
1695 let mut matcher = pat.into_searcher(self);
1696 if let Some((a, b)) = matcher.next_reject() {
1698 j = b; // Rember earliest known match, correct it below if
1699 // last match is different
1701 if let Some((_, b)) = matcher.next_reject_back() {
1705 // Searcher is known to return valid indices
1706 self.slice_unchecked(i, j)
1711 fn trim_left_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str {
1712 let mut i = self.len();
1713 let mut matcher = pat.into_searcher(self);
1714 if let Some((a, _)) = matcher.next_reject() {
1718 // Searcher is known to return valid indices
1719 self.slice_unchecked(i, self.len())
1724 fn trim_right_matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> &'a str
1725 where P::Searcher: ReverseSearcher<'a>
1728 let mut matcher = pat.into_searcher(self);
1729 if let Some((_, b)) = matcher.next_reject_back() {
1733 // Searcher is known to return valid indices
1734 self.slice_unchecked(0, j)
1739 fn is_char_boundary(&self, index: usize) -> bool {
1740 if index == self.len() { return true; }
1741 match self.as_bytes().get(index) {
1743 Some(&b) => b < 128 || b >= 192,
1748 fn char_range_at(&self, i: usize) -> CharRange {
1749 let (c, n) = char_range_at_raw(self.as_bytes(), i);
1750 CharRange { ch: unsafe { mem::transmute(c) }, next: n }
1754 fn char_range_at_reverse(&self, start: usize) -> CharRange {
1755 let mut prev = start;
1757 prev = prev.saturating_sub(1);
1758 if self.as_bytes()[prev] < 128 {
1759 return CharRange{ch: self.as_bytes()[prev] as char, next: prev}
1762 // Multibyte case is a fn to allow char_range_at_reverse to inline cleanly
1763 fn multibyte_char_range_at_reverse(s: &str, mut i: usize) -> CharRange {
1764 // while there is a previous byte == 10......
1765 while i > 0 && s.as_bytes()[i] & !CONT_MASK == TAG_CONT_U8 {
1769 let first= s.as_bytes()[i];
1770 let w = UTF8_CHAR_WIDTH[first as usize];
1773 let mut val = utf8_first_byte(first, w as u32);
1774 val = utf8_acc_cont_byte(val, s.as_bytes()[i + 1]);
1775 if w > 2 { val = utf8_acc_cont_byte(val, s.as_bytes()[i + 2]); }
1776 if w > 3 { val = utf8_acc_cont_byte(val, s.as_bytes()[i + 3]); }
1778 return CharRange {ch: unsafe { mem::transmute(val) }, next: i};
1781 return multibyte_char_range_at_reverse(self, prev);
1785 fn char_at(&self, i: usize) -> char {
1786 self.char_range_at(i).ch
1790 fn char_at_reverse(&self, i: usize) -> char {
1791 self.char_range_at_reverse(i).ch
1795 fn as_bytes(&self) -> &[u8] {
1796 unsafe { mem::transmute(self) }
1799 fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
1800 pat.into_searcher(self).next_match().map(|(i, _)| i)
1803 fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
1804 where P::Searcher: ReverseSearcher<'a>
1806 pat.into_searcher(self).next_match_back().map(|(i, _)| i)
1809 fn find_str<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
1813 fn split_at(&self, mid: usize) -> (&str, &str) {
1814 // is_char_boundary checks that the index is in [0, .len()]
1815 if self.is_char_boundary(mid) {
1817 (self.slice_unchecked(0, mid),
1818 self.slice_unchecked(mid, self.len()))
1821 slice_error_fail(self, 0, mid)
1826 fn slice_shift_char(&self) -> Option<(char, &str)> {
1827 if self.is_empty() {
1830 let ch = self.char_at(0);
1831 let next_s = unsafe { self.slice_unchecked(ch.len_utf8(), self.len()) };
1836 fn subslice_offset(&self, inner: &str) -> usize {
1837 let a_start = self.as_ptr() as usize;
1838 let a_end = a_start + self.len();
1839 let b_start = inner.as_ptr() as usize;
1840 let b_end = b_start + inner.len();
1842 assert!(a_start <= b_start);
1843 assert!(b_end <= a_end);
1848 fn as_ptr(&self) -> *const u8 {
1853 fn len(&self) -> usize { self.repr().len }
1856 fn is_empty(&self) -> bool { self.len() == 0 }
1859 fn parse<T: FromStr>(&self) -> Result<T, T::Err> { FromStr::from_str(self) }
1862 #[stable(feature = "rust1", since = "1.0.0")]
1863 impl AsRef<[u8]> for str {
1865 fn as_ref(&self) -> &[u8] {
1870 /// Pluck a code point out of a UTF-8-like byte slice and return the
1871 /// index of the next code point.
1873 #[unstable(feature = "core")]
1874 pub fn char_range_at_raw(bytes: &[u8], i: usize) -> (u32, usize) {
1876 return (bytes[i] as u32, i + 1);
1879 // Multibyte case is a fn to allow char_range_at to inline cleanly
1880 fn multibyte_char_range_at(bytes: &[u8], i: usize) -> (u32, usize) {
1881 let first = bytes[i];
1882 let w = UTF8_CHAR_WIDTH[first as usize];
1885 let mut val = utf8_first_byte(first, w as u32);
1886 val = utf8_acc_cont_byte(val, bytes[i + 1]);
1887 if w > 2 { val = utf8_acc_cont_byte(val, bytes[i + 2]); }
1888 if w > 3 { val = utf8_acc_cont_byte(val, bytes[i + 3]); }
1890 return (val, i + w as usize);
1893 multibyte_char_range_at(bytes, i)
1896 #[stable(feature = "rust1", since = "1.0.0")]
1897 impl<'a> Default for &'a str {
1898 #[stable(feature = "rust1", since = "1.0.0")]
1899 fn default() -> &'a str { "" }