1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
use std::borrow::Cow;
use std::error;
use std::ffi::{OsStr, OsString};
use std::fmt;
use std::iter;
use std::ops;
use std::path::{Path, PathBuf};
use std::ptr;
use std::str;
use std::vec;

use crate::ext_slice::ByteSlice;
use crate::utf8::{self, Utf8Error};

/// Concatenate the elements given by the iterator together into a single
/// `Vec<u8>`.
///
/// The elements may be any type that can be cheaply converted into an `&[u8]`.
/// This includes, but is not limited to, `&str`, `&BStr` and `&[u8]` itself.
///
/// # Examples
///
/// Basic usage:
///
/// ```
/// use bstr;
///
/// let s = bstr::concat(&["foo", "bar", "baz"]);
/// assert_eq!(s, "foobarbaz".as_bytes());
/// ```
#[inline]
pub fn concat<T, I>(elements: I) -> Vec<u8>
where
    T: AsRef<[u8]>,
    I: IntoIterator<Item = T>,
{
    let mut dest = vec![];
    for element in elements {
        dest.push_str(element);
    }
    dest
}

/// Join the elements given by the iterator with the given separator into a
/// single `Vec<u8>`.
///
/// Both the separator and the elements may be any type that can be cheaply
/// converted into an `&[u8]`. This includes, but is not limited to,
/// `&str`, `&BStr` and `&[u8]` itself.
///
/// # Examples
///
/// Basic usage:
///
/// ```
/// use bstr;
///
/// let s = bstr::join(",", &["foo", "bar", "baz"]);
/// assert_eq!(s, "foo,bar,baz".as_bytes());
/// ```
#[inline]
pub fn join<B, T, I>(separator: B, elements: I) -> Vec<u8>
where
    B: AsRef<[u8]>,
    T: AsRef<[u8]>,
    I: IntoIterator<Item = T>,
{
    let mut it = elements.into_iter();
    let mut dest = vec![];
    match it.next() {
        None => return dest,
        Some(first) => {
            dest.push_str(first);
        }
    }
    for element in it {
        dest.push_str(&separator);
        dest.push_str(element);
    }
    dest
}

impl ByteVec for Vec<u8> {
    #[inline]
    fn as_vec(&self) -> &Vec<u8> {
        self
    }

    #[inline]
    fn as_vec_mut(&mut self) -> &mut Vec<u8> {
        self
    }

    #[inline]
    fn into_vec(self) -> Vec<u8> {
        self
    }
}

/// Ensure that callers cannot implement `ByteSlice` by making an
/// umplementable trait its super trait.
pub trait Sealed {}
impl Sealed for Vec<u8> {}

/// A trait that extends `Vec<u8>` with string oriented methods.
///
/// Note that when using the constructor methods, such as
/// `ByteVec::from_slice`, one should actually call them using the concrete
/// type. For example:
///
/// ```
/// use bstr::{B, ByteVec};
///
/// let s = Vec::from_slice(b"abc"); // NOT ByteVec::from_slice("...")
/// assert_eq!(s, B("abc"));
/// ```
pub trait ByteVec: Sealed {
    /// A method for accessing the raw vector bytes of this type. This is
    /// always a no-op and callers shouldn't care about it. This only exists
    /// for making the extension trait work.
    #[doc(hidden)]
    fn as_vec(&self) -> &Vec<u8>;

    /// A method for accessing the raw vector bytes of this type, mutably. This
    /// is always a no-op and callers shouldn't care about it. This only exists
    /// for making the extension trait work.
    #[doc(hidden)]
    fn as_vec_mut(&mut self) -> &mut Vec<u8>;

    /// A method for consuming ownership of this vector. This is always a no-op
    /// and callers shouldn't care about it. This only exists for making the
    /// extension trait work.
    #[doc(hidden)]
    fn into_vec(self) -> Vec<u8>
    where
        Self: Sized;

    /// Create a new owned byte string from the given byte slice.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::{B, ByteVec};
    ///
    /// let s = Vec::from_slice(b"abc");
    /// assert_eq!(s, B("abc"));
    /// ```
    #[inline]
    fn from_slice<B: AsRef<[u8]>>(bytes: B) -> Vec<u8> {
        bytes.as_ref().to_vec()
    }

    /// Create a new byte string from an owned OS string.
    ///
    /// On Unix, this always succeeds and is zero cost. On non-Unix systems,
    /// this returns the original OS string if it is not valid UTF-8.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use std::ffi::OsString;
    ///
    /// use bstr::{B, ByteVec};
    ///
    /// let os_str = OsString::from("foo");
    /// let bs = Vec::from_os_string(os_str).expect("valid UTF-8");
    /// assert_eq!(bs, B("foo"));
    /// ```
    #[inline]
    fn from_os_string(os_str: OsString) -> Result<Vec<u8>, OsString> {
        #[cfg(unix)]
        #[inline]
        fn imp(os_str: OsString) -> Result<Vec<u8>, OsString> {
            use std::os::unix::ffi::OsStringExt;

            Ok(Vec::from(os_str.into_vec()))
        }

        #[cfg(not(unix))]
        #[inline]
        fn imp(os_str: OsString) -> Result<Vec<u8>, OsString> {
            os_str.into_string().map(Vec::from)
        }

        imp(os_str)
    }

    /// Lossily create a new byte string from an OS string slice.
    ///
    /// On Unix, this always succeeds, is zero cost and always returns a slice.
    /// On non-Unix systems, this does a UTF-8 check. If the given OS string
    /// slice is not valid UTF-8, then it is lossily decoded into valid UTF-8
    /// (with invalid bytes replaced by the Unicode replacement codepoint).
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use std::ffi::OsStr;
    ///
    /// use bstr::{B, ByteVec};
    ///
    /// let os_str = OsStr::new("foo");
    /// let bs = Vec::from_os_str_lossy(os_str);
    /// assert_eq!(bs, B("foo"));
    /// ```
    #[inline]
    fn from_os_str_lossy<'a>(os_str: &'a OsStr) -> Cow<'a, [u8]> {
        #[cfg(unix)]
        #[inline]
        fn imp<'a>(os_str: &'a OsStr) -> Cow<'a, [u8]> {
            use std::os::unix::ffi::OsStrExt;

            Cow::Borrowed(os_str.as_bytes())
        }

        #[cfg(not(unix))]
        #[inline]
        fn imp<'a>(os_str: &'a OsStr) -> Cow<'a, [u8]> {
            match os_str.to_string_lossy() {
                Cow::Borrowed(x) => Cow::Borrowed(x.as_bytes()),
                Cow::Owned(x) => Cow::Owned(Vec::from(x)),
            }
        }

        imp(os_str)
    }

    /// Create a new byte string from an owned file path.
    ///
    /// On Unix, this always succeeds and is zero cost. On non-Unix systems,
    /// this returns the original path if it is not valid UTF-8.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use std::path::PathBuf;
    ///
    /// use bstr::{B, ByteVec};
    ///
    /// let path = PathBuf::from("foo");
    /// let bs = Vec::from_path_buf(path).expect("must be valid UTF-8");
    /// assert_eq!(bs, B("foo"));
    /// ```
    #[inline]
    fn from_path_buf(path: PathBuf) -> Result<Vec<u8>, PathBuf> {
        Vec::from_os_string(path.into_os_string()).map_err(PathBuf::from)
    }

    /// Lossily create a new byte string from a file path.
    ///
    /// On Unix, this always succeeds, is zero cost and always returns a slice.
    /// On non-Unix systems, this does a UTF-8 check. If the given path is not
    /// valid UTF-8, then it is lossily decoded into valid UTF-8 (with invalid
    /// bytes replaced by the Unicode replacement codepoint).
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use std::path::Path;
    ///
    /// use bstr::{B, ByteVec};
    ///
    /// let path = Path::new("foo");
    /// let bs = Vec::from_path_lossy(path);
    /// assert_eq!(bs, B("foo"));
    /// ```
    #[inline]
    fn from_path_lossy<'a>(path: &'a Path) -> Cow<'a, [u8]> {
        Vec::from_os_str_lossy(path.as_os_str())
    }

    /// Appends the given byte to the end of this byte string.
    ///
    /// Note that this is equivalent to the generic `Vec::push` method. This
    /// method is provided to permit callers to explicitly differentiate
    /// between pushing bytes, codepoints and strings.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = <Vec<u8>>::from("abc");
    /// s.push_byte(b'\xE2');
    /// s.push_byte(b'\x98');
    /// s.push_byte(b'\x83');
    /// assert_eq!(s, "abc☃".as_bytes());
    /// ```
    #[inline]
    fn push_byte(&mut self, byte: u8) {
        self.as_vec_mut().push(byte);
    }

    /// Appends the given `char` to the end of this byte string.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = <Vec<u8>>::from("abc");
    /// s.push_char('1');
    /// s.push_char('2');
    /// s.push_char('3');
    /// assert_eq!(s, "abc123".as_bytes());
    /// ```
    #[inline]
    fn push_char(&mut self, ch: char) {
        if ch.len_utf8() == 1 {
            self.push_byte(ch as u8);
            return;
        }
        self.as_vec_mut()
            .extend_from_slice(ch.encode_utf8(&mut [0; 4]).as_bytes());
    }

    /// Appends the given slice to the end of this byte string. This accepts
    /// any type that be converted to a `&[u8]`. This includes, but is not
    /// limited to, `&str`, `&BStr`, and of course, `&[u8]` itself.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = <Vec<u8>>::from("abc");
    /// s.push_str(b"123");
    /// assert_eq!(s, "abc123".as_bytes());
    /// ```
    #[inline]
    fn push_str<B: AsRef<[u8]>>(&mut self, bytes: B) {
        self.as_vec_mut().extend_from_slice(bytes.as_ref());
    }

    /// Converts a `Vec<u8>` into a `String` if and only if this byte string is
    /// valid UTF-8.
    ///
    /// If it is not valid UTF-8, then a
    /// [`FromUtf8Error`](struct.FromUtf8Error.html)
    /// is returned. (This error can be used to examine why UTF-8 validation
    /// failed, or to regain the original byte string.)
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// # fn example() -> Result<(), Box<dyn std::error::Error>> {
    /// let bytes = Vec::from("hello");
    /// let string = bytes.into_string()?;
    ///
    /// assert_eq!("hello", string);
    /// # Ok(()) }; example().unwrap()
    /// ```
    ///
    /// If this byte string is not valid UTF-8, then an error will be returned.
    /// That error can then be used to inspect the location at which invalid
    /// UTF-8 was found, or to regain the original byte string:
    ///
    /// ```
    /// use bstr::{B, ByteVec};
    ///
    /// let bytes = Vec::from_slice(b"foo\xFFbar");
    /// let err = bytes.into_string().unwrap_err();
    ///
    /// assert_eq!(err.utf8_error().valid_up_to(), 3);
    /// assert_eq!(err.utf8_error().error_len(), Some(1));
    ///
    /// // At no point in this example is an allocation performed.
    /// let bytes = Vec::from(err.into_vec());
    /// assert_eq!(bytes, B(b"foo\xFFbar"));
    /// ```
    #[inline]
    fn into_string(self) -> Result<String, FromUtf8Error>
    where
        Self: Sized,
    {
        match utf8::validate(self.as_vec()) {
            Err(err) => Err(FromUtf8Error { original: self.into_vec(), err }),
            Ok(()) => {
                // SAFETY: This is safe because of the guarantees provided by
                // utf8::validate.
                unsafe { Ok(self.into_string_unchecked()) }
            }
        }
    }

    /// Lossily converts a `Vec<u8>` into a `String`. If this byte string
    /// contains invalid UTF-8, then the invalid bytes are replaced with the
    /// Unicode replacement codepoint.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let bytes = Vec::from_slice(b"foo\xFFbar");
    /// let string = bytes.into_string_lossy();
    /// assert_eq!(string, "foo\u{FFFD}bar");
    /// ```
    #[inline]
    fn into_string_lossy(self) -> String
    where
        Self: Sized,
    {
        match self.as_vec().to_str_lossy() {
            Cow::Borrowed(_) => {
                // SAFETY: to_str_lossy() returning a Cow::Borrowed guarantees
                // the entire string is valid utf8.
                unsafe { self.into_string_unchecked() }
            }
            Cow::Owned(s) => s,
        }
    }

    /// Unsafely convert this byte string into a `String`, without checking for
    /// valid UTF-8.
    ///
    /// # Safety
    ///
    /// Callers *must* ensure that this byte string is valid UTF-8 before
    /// calling this method. Converting a byte string into a `String` that is
    /// not valid UTF-8 is considered undefined behavior.
    ///
    /// This routine is useful in performance sensitive contexts where the
    /// UTF-8 validity of the byte string is already known and it is
    /// undesirable to pay the cost of an additional UTF-8 validation check
    /// that [`into_string`](#method.into_string) performs.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// // SAFETY: This is safe because string literals are guaranteed to be
    /// // valid UTF-8 by the Rust compiler.
    /// let s = unsafe { Vec::from("☃βツ").into_string_unchecked() };
    /// assert_eq!("☃βツ", s);
    /// ```
    #[inline]
    unsafe fn into_string_unchecked(self) -> String
    where
        Self: Sized,
    {
        String::from_utf8_unchecked(self.into_vec())
    }

    /// Converts this byte string into an OS string, in place.
    ///
    /// On Unix, this always succeeds and is zero cost. On non-Unix systems,
    /// this returns the original byte string if it is not valid UTF-8.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use std::ffi::OsStr;
    ///
    /// use bstr::ByteVec;
    ///
    /// let bs = Vec::from("foo");
    /// let os_str = bs.into_os_string().expect("should be valid UTF-8");
    /// assert_eq!(os_str, OsStr::new("foo"));
    /// ```
    #[inline]
    fn into_os_string(self) -> Result<OsString, Vec<u8>>
    where
        Self: Sized,
    {
        #[cfg(unix)]
        #[inline]
        fn imp(v: Vec<u8>) -> Result<OsString, Vec<u8>> {
            use std::os::unix::ffi::OsStringExt;

            Ok(OsString::from_vec(v))
        }

        #[cfg(not(unix))]
        #[inline]
        fn imp(v: Vec<u8>) -> Result<OsString, Vec<u8>> {
            match v.into_string() {
                Ok(s) => Ok(OsString::from(s)),
                Err(err) => Err(err.into_vec()),
            }
        }

        imp(self.into_vec())
    }

    /// Lossily converts this byte string into an OS string, in place.
    ///
    /// On Unix, this always succeeds and is zero cost. On non-Unix systems,
    /// this will perform a UTF-8 check and lossily convert this byte string
    /// into valid UTF-8 using the Unicode replacement codepoint.
    ///
    /// Note that this can prevent the correct roundtripping of file paths on
    /// non-Unix systems such as Windows, where file paths are an arbitrary
    /// sequence of 16-bit integers.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let bs = Vec::from_slice(b"foo\xFFbar");
    /// let os_str = bs.into_os_string_lossy();
    /// assert_eq!(os_str.to_string_lossy(), "foo\u{FFFD}bar");
    /// ```
    #[inline]
    fn into_os_string_lossy(self) -> OsString
    where
        Self: Sized,
    {
        #[cfg(unix)]
        #[inline]
        fn imp(v: Vec<u8>) -> OsString {
            use std::os::unix::ffi::OsStringExt;

            OsString::from_vec(v)
        }

        #[cfg(not(unix))]
        #[inline]
        fn imp(v: Vec<u8>) -> OsString {
            OsString::from(v.into_string_lossy())
        }

        imp(self.into_vec())
    }

    /// Converts this byte string into an owned file path, in place.
    ///
    /// On Unix, this always succeeds and is zero cost. On non-Unix systems,
    /// this returns the original byte string if it is not valid UTF-8.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let bs = Vec::from("foo");
    /// let path = bs.into_path_buf().expect("should be valid UTF-8");
    /// assert_eq!(path.as_os_str(), "foo");
    /// ```
    #[inline]
    fn into_path_buf(self) -> Result<PathBuf, Vec<u8>>
    where
        Self: Sized,
    {
        self.into_os_string().map(PathBuf::from)
    }

    /// Lossily converts this byte string into an owned file path, in place.
    ///
    /// On Unix, this always succeeds and is zero cost. On non-Unix systems,
    /// this will perform a UTF-8 check and lossily convert this byte string
    /// into valid UTF-8 using the Unicode replacement codepoint.
    ///
    /// Note that this can prevent the correct roundtripping of file paths on
    /// non-Unix systems such as Windows, where file paths are an arbitrary
    /// sequence of 16-bit integers.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let bs = Vec::from_slice(b"foo\xFFbar");
    /// let path = bs.into_path_buf_lossy();
    /// assert_eq!(path.to_string_lossy(), "foo\u{FFFD}bar");
    /// ```
    #[inline]
    fn into_path_buf_lossy(self) -> PathBuf
    where
        Self: Sized,
    {
        PathBuf::from(self.into_os_string_lossy())
    }

    /// Removes the last byte from this `Vec<u8>` and returns it.
    ///
    /// If this byte string is empty, then `None` is returned.
    ///
    /// If the last codepoint in this byte string is not ASCII, then removing
    /// the last byte could make this byte string contain invalid UTF-8.
    ///
    /// Note that this is equivalent to the generic `Vec::pop` method. This
    /// method is provided to permit callers to explicitly differentiate
    /// between popping bytes and codepoints.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foo");
    /// assert_eq!(s.pop_byte(), Some(b'o'));
    /// assert_eq!(s.pop_byte(), Some(b'o'));
    /// assert_eq!(s.pop_byte(), Some(b'f'));
    /// assert_eq!(s.pop_byte(), None);
    /// ```
    #[inline]
    fn pop_byte(&mut self) -> Option<u8> {
        self.as_vec_mut().pop()
    }

    /// Removes the last codepoint from this `Vec<u8>` and returns it.
    ///
    /// If this byte string is empty, then `None` is returned. If the last
    /// bytes of this byte string do not correspond to a valid UTF-8 code unit
    /// sequence, then the Unicode replacement codepoint is yielded instead in
    /// accordance with the
    /// [replacement codepoint substitution policy](index.html#handling-of-invalid-utf8-8).
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foo");
    /// assert_eq!(s.pop_char(), Some('o'));
    /// assert_eq!(s.pop_char(), Some('o'));
    /// assert_eq!(s.pop_char(), Some('f'));
    /// assert_eq!(s.pop_char(), None);
    /// ```
    ///
    /// This shows the replacement codepoint substitution policy. Note that
    /// the first pop yields a replacement codepoint but actually removes two
    /// bytes. This is in contrast with subsequent pops when encountering
    /// `\xFF` since `\xFF` is never a valid prefix for any valid UTF-8
    /// code unit sequence.
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from_slice(b"f\xFF\xFF\xFFoo\xE2\x98");
    /// assert_eq!(s.pop_char(), Some('\u{FFFD}'));
    /// assert_eq!(s.pop_char(), Some('o'));
    /// assert_eq!(s.pop_char(), Some('o'));
    /// assert_eq!(s.pop_char(), Some('\u{FFFD}'));
    /// assert_eq!(s.pop_char(), Some('\u{FFFD}'));
    /// assert_eq!(s.pop_char(), Some('\u{FFFD}'));
    /// assert_eq!(s.pop_char(), Some('f'));
    /// assert_eq!(s.pop_char(), None);
    /// ```
    #[inline]
    fn pop_char(&mut self) -> Option<char> {
        let (ch, size) = utf8::decode_last_lossy(self.as_vec());
        if size == 0 {
            return None;
        }
        let new_len = self.as_vec().len() - size;
        self.as_vec_mut().truncate(new_len);
        Some(ch)
    }

    /// Removes a `char` from this `Vec<u8>` at the given byte position and
    /// returns it.
    ///
    /// If the bytes at the given position do not lead to a valid UTF-8 code
    /// unit sequence, then a
    /// [replacement codepoint is returned instead](index.html#handling-of-invalid-utf8-8).
    ///
    /// # Panics
    ///
    /// Panics if `at` is larger than or equal to this byte string's length.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foo☃bar");
    /// assert_eq!(s.remove_char(3), '☃');
    /// assert_eq!(s, b"foobar");
    /// ```
    ///
    /// This example shows how the Unicode replacement codepoint policy is
    /// used:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from_slice(b"foo\xFFbar");
    /// assert_eq!(s.remove_char(3), '\u{FFFD}');
    /// assert_eq!(s, b"foobar");
    /// ```
    #[inline]
    fn remove_char(&mut self, at: usize) -> char {
        let (ch, size) = utf8::decode_lossy(&self.as_vec()[at..]);
        assert!(
            size > 0,
            "expected {} to be less than {}",
            at,
            self.as_vec().len(),
        );
        self.as_vec_mut().drain(at..at + size);
        ch
    }

    /// Inserts the given codepoint into this `Vec<u8>` at a particular byte
    /// position.
    ///
    /// This is an `O(n)` operation as it may copy a number of elements in this
    /// byte string proportional to its length.
    ///
    /// # Panics
    ///
    /// Panics if `at` is larger than the byte string's length.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foobar");
    /// s.insert_char(3, '☃');
    /// assert_eq!(s, "foo☃bar".as_bytes());
    /// ```
    #[inline]
    fn insert_char(&mut self, at: usize, ch: char) {
        self.insert_str(at, ch.encode_utf8(&mut [0; 4]).as_bytes());
    }

    /// Inserts the given byte string into this byte string at a particular
    /// byte position.
    ///
    /// This is an `O(n)` operation as it may copy a number of elements in this
    /// byte string proportional to its length.
    ///
    /// The given byte string may be any type that can be cheaply converted
    /// into a `&[u8]`. This includes, but is not limited to, `&str` and
    /// `&[u8]`.
    ///
    /// # Panics
    ///
    /// Panics if `at` is larger than the byte string's length.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foobar");
    /// s.insert_str(3, "☃☃☃");
    /// assert_eq!(s, "foo☃☃☃bar".as_bytes());
    /// ```
    #[inline]
    fn insert_str<B: AsRef<[u8]>>(&mut self, at: usize, bytes: B) {
        let bytes = bytes.as_ref();
        let len = self.as_vec().len();
        assert!(at <= len, "expected {} to be <= {}", at, len);

        // SAFETY: We'd like to efficiently splice in the given bytes into
        // this byte string. Since we are only working with `u8` elements here,
        // we only need to consider whether our bounds are correct and whether
        // our byte string has enough space.
        self.as_vec_mut().reserve(bytes.len());
        unsafe {
            // Shift bytes after `at` over by the length of `bytes` to make
            // room for it. This requires referencing two regions of memory
            // that may overlap, so we use ptr::copy.
            ptr::copy(
                self.as_vec().as_ptr().add(at),
                self.as_vec_mut().as_mut_ptr().add(at + bytes.len()),
                len - at,
            );
            // Now copy the bytes given into the room we made above. In this
            // case, we know that the given bytes cannot possibly overlap
            // with this byte string since we have a mutable borrow of the
            // latter. Thus, we can use a nonoverlapping copy.
            ptr::copy_nonoverlapping(
                bytes.as_ptr(),
                self.as_vec_mut().as_mut_ptr().add(at),
                bytes.len(),
            );
            self.as_vec_mut().set_len(len + bytes.len());
        }
    }

    /// Removes the specified range in this byte string and replaces it with
    /// the given bytes. The given bytes do not need to have the same length
    /// as the range provided.
    ///
    /// # Panics
    ///
    /// Panics if the given range is invalid.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foobar");
    /// s.replace_range(2..4, "xxxxx");
    /// assert_eq!(s, "foxxxxxar".as_bytes());
    /// ```
    #[inline]
    fn replace_range<R, B>(&mut self, range: R, replace_with: B)
    where
        R: ops::RangeBounds<usize>,
        B: AsRef<[u8]>,
    {
        self.as_vec_mut().splice(range, replace_with.as_ref().iter().cloned());
    }

    /// Creates a draining iterator that removes the specified range in this
    /// `Vec<u8>` and yields each of the removed bytes.
    ///
    /// Note that the elements specified by the given range are removed
    /// regardless of whether the returned iterator is fully exhausted.
    ///
    /// Also note that is is unspecified how many bytes are removed from the
    /// `Vec<u8>` if the `DrainBytes` iterator is leaked.
    ///
    /// # Panics
    ///
    /// Panics if the given range is not valid.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::ByteVec;
    ///
    /// let mut s = Vec::from("foobar");
    /// {
    ///     let mut drainer = s.drain_bytes(2..4);
    ///     assert_eq!(drainer.next(), Some(b'o'));
    ///     assert_eq!(drainer.next(), Some(b'b'));
    ///     assert_eq!(drainer.next(), None);
    /// }
    /// assert_eq!(s, "foar".as_bytes());
    /// ```
    #[inline]
    fn drain_bytes<R>(&mut self, range: R) -> DrainBytes<'_>
    where
        R: ops::RangeBounds<usize>,
    {
        DrainBytes { it: self.as_vec_mut().drain(range) }
    }
}

/// A draining byte oriented iterator for `Vec<u8>`.
///
/// This iterator is created by
/// [`ByteVec::drain_bytes`](trait.ByteVec.html#method.drain_bytes).
///
/// # Examples
///
/// Basic usage:
///
/// ```
/// use bstr::ByteVec;
///
/// let mut s = Vec::from("foobar");
/// {
///     let mut drainer = s.drain_bytes(2..4);
///     assert_eq!(drainer.next(), Some(b'o'));
///     assert_eq!(drainer.next(), Some(b'b'));
///     assert_eq!(drainer.next(), None);
/// }
/// assert_eq!(s, "foar".as_bytes());
/// ```
#[derive(Debug)]
pub struct DrainBytes<'a> {
    it: vec::Drain<'a, u8>,
}

impl<'a> iter::FusedIterator for DrainBytes<'a> {}

impl<'a> Iterator for DrainBytes<'a> {
    type Item = u8;

    #[inline]
    fn next(&mut self) -> Option<u8> {
        self.it.next()
    }
}

impl<'a> DoubleEndedIterator for DrainBytes<'a> {
    #[inline]
    fn next_back(&mut self) -> Option<u8> {
        self.it.next_back()
    }
}

impl<'a> ExactSizeIterator for DrainBytes<'a> {
    #[inline]
    fn len(&self) -> usize {
        self.it.len()
    }
}

/// An error that may occur when converting a `Vec<u8>` to a `String`.
///
/// This error includes the original `Vec<u8>` that failed to convert to a
/// `String`. This permits callers to recover the allocation used even if it
/// it not valid UTF-8.
///
/// # Examples
///
/// Basic usage:
///
/// ```
/// use bstr::{B, ByteVec};
///
/// let bytes = Vec::from_slice(b"foo\xFFbar");
/// let err = bytes.into_string().unwrap_err();
///
/// assert_eq!(err.utf8_error().valid_up_to(), 3);
/// assert_eq!(err.utf8_error().error_len(), Some(1));
///
/// // At no point in this example is an allocation performed.
/// let bytes = Vec::from(err.into_vec());
/// assert_eq!(bytes, B(b"foo\xFFbar"));
/// ```
#[derive(Debug, Eq, PartialEq)]
pub struct FromUtf8Error {
    original: Vec<u8>,
    err: Utf8Error,
}

impl FromUtf8Error {
    /// Return the original bytes as a slice that failed to convert to a
    /// `String`.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::{B, ByteVec};
    ///
    /// let bytes = Vec::from_slice(b"foo\xFFbar");
    /// let err = bytes.into_string().unwrap_err();
    ///
    /// // At no point in this example is an allocation performed.
    /// assert_eq!(err.as_bytes(), B(b"foo\xFFbar"));
    /// ```
    #[inline]
    pub fn as_bytes(&self) -> &[u8] {
        &self.original
    }

    /// Consume this error and return the original byte string that failed to
    /// convert to a `String`.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::{B, ByteVec};
    ///
    /// let bytes = Vec::from_slice(b"foo\xFFbar");
    /// let err = bytes.into_string().unwrap_err();
    /// let original = err.into_vec();
    ///
    /// // At no point in this example is an allocation performed.
    /// assert_eq!(original, B(b"foo\xFFbar"));
    /// ```
    #[inline]
    pub fn into_vec(self) -> Vec<u8> {
        self.original
    }

    /// Return the underlying UTF-8 error that occurred. This error provides
    /// information on the nature and location of the invalid UTF-8 detected.
    ///
    /// # Examples
    ///
    /// Basic usage:
    ///
    /// ```
    /// use bstr::{B, ByteVec};
    ///
    /// let bytes = Vec::from_slice(b"foo\xFFbar");
    /// let err = bytes.into_string().unwrap_err();
    ///
    /// assert_eq!(err.utf8_error().valid_up_to(), 3);
    /// assert_eq!(err.utf8_error().error_len(), Some(1));
    /// ```
    #[inline]
    pub fn utf8_error(&self) -> &Utf8Error {
        &self.err
    }
}

impl error::Error for FromUtf8Error {
    #[inline]
    fn description(&self) -> &str {
        "invalid UTF-8 vector"
    }
}

impl fmt::Display for FromUtf8Error {
    #[inline]
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        write!(f, "{}", self.err)
    }
}

#[cfg(test)]
mod tests {
    use crate::ext_vec::ByteVec;

    #[test]
    fn insert() {
        let mut s = vec![];
        s.insert_str(0, "foo");
        assert_eq!(s, "foo".as_bytes());

        let mut s = Vec::from("a");
        s.insert_str(0, "foo");
        assert_eq!(s, "fooa".as_bytes());

        let mut s = Vec::from("a");
        s.insert_str(1, "foo");
        assert_eq!(s, "afoo".as_bytes());

        let mut s = Vec::from("foobar");
        s.insert_str(3, "quux");
        assert_eq!(s, "fooquuxbar".as_bytes());

        let mut s = Vec::from("foobar");
        s.insert_str(3, "x");
        assert_eq!(s, "fooxbar".as_bytes());

        let mut s = Vec::from("foobar");
        s.insert_str(0, "x");
        assert_eq!(s, "xfoobar".as_bytes());

        let mut s = Vec::from("foobar");
        s.insert_str(6, "x");
        assert_eq!(s, "foobarx".as_bytes());

        let mut s = Vec::from("foobar");
        s.insert_str(3, "quuxbazquux");
        assert_eq!(s, "fooquuxbazquuxbar".as_bytes());
    }

    #[test]
    #[should_panic]
    fn insert_fail1() {
        let mut s = vec![];
        s.insert_str(1, "foo");
    }

    #[test]
    #[should_panic]
    fn insert_fail2() {
        let mut s = Vec::from("a");
        s.insert_str(2, "foo");
    }

    #[test]
    #[should_panic]
    fn insert_fail3() {
        let mut s = Vec::from("foobar");
        s.insert_str(7, "foo");
    }
}