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
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
// Copyright 2015 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.

//! Implementation of [the WTF-8 encoding](https://simonsapin.github.io/wtf-8/).
//!
//! This library uses Rust’s type system to maintain
//! [well-formedness](https://simonsapin.github.io/wtf-8/#well-formed),
//! like the `String` and `&str` types do for UTF-8.
//!
//! Since [WTF-8 must not be used
//! for interchange](https://simonsapin.github.io/wtf-8/#intended-audience),
//! this library deliberately does not provide access to the underlying bytes
//! of WTF-8 strings,
//! nor can it decode WTF-8 from arbitrary bytes.
//! WTF-8 strings can be obtained from UTF-8, UTF-16, or code points.

// this module is imported from @SimonSapin's repo and has tons of dead code on
// unix (it's mostly used on windows), so don't worry about dead code here.
#![allow(dead_code)]

use core::str::next_code_point;

use borrow::Cow;
use char;
use fmt;
use hash::{Hash, Hasher};
use iter::FromIterator;
use mem;
use ops;
use rc::Rc;
use slice;
use str;
use sync::Arc;
use sys_common::AsInner;

const UTF8_REPLACEMENT_CHARACTER: &'static str = "\u{FFFD}";

/// A Unicode code point: from U+0000 to U+10FFFF.
///
/// Compare with the `char` type,
/// which represents a Unicode scalar value:
/// a code point that is not a surrogate (U+D800 to U+DFFF).
#[derive(Eq, PartialEq, Ord, PartialOrd, Clone, Copy)]
pub struct CodePoint {
    value: u32
}

/// Format the code point as `U+` followed by four to six hexadecimal digits.
/// Example: `U+1F4A9`
impl fmt::Debug for CodePoint {
    #[inline]
    fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
        write!(formatter, "U+{:04X}", self.value)
    }
}

impl CodePoint {
    /// Unsafely creates a new `CodePoint` without checking the value.
    ///
    /// Only use when `value` is known to be less than or equal to 0x10FFFF.
    #[inline]
    pub unsafe fn from_u32_unchecked(value: u32) -> CodePoint {
        CodePoint { value: value }
    }

    /// Creates a new `CodePoint` if the value is a valid code point.
    ///
    /// Returns `None` if `value` is above 0x10FFFF.
    #[inline]
    pub fn from_u32(value: u32) -> Option<CodePoint> {
        match value {
            0 ... 0x10FFFF => Some(CodePoint { value: value }),
            _ => None
        }
    }

    /// Creates a new `CodePoint` from a `char`.
    ///
    /// Since all Unicode scalar values are code points, this always succeeds.
    #[inline]
    pub fn from_char(value: char) -> CodePoint {
        CodePoint { value: value as u32 }
    }

    /// Returns the numeric value of the code point.
    #[inline]
    pub fn to_u32(&self) -> u32 {
        self.value
    }

    /// Optionally returns a Unicode scalar value for the code point.
    ///
    /// Returns `None` if the code point is a surrogate (from U+D800 to U+DFFF).
    #[inline]
    pub fn to_char(&self) -> Option<char> {
        match self.value {
            0xD800 ... 0xDFFF => None,
            _ => Some(unsafe { char::from_u32_unchecked(self.value) })
        }
    }

    /// Returns a Unicode scalar value for the code point.
    ///
    /// Returns `'\u{FFFD}'` (the replacement character “�”)
    /// if the code point is a surrogate (from U+D800 to U+DFFF).
    #[inline]
    pub fn to_char_lossy(&self) -> char {
        self.to_char().unwrap_or('\u{FFFD}')
    }
}

/// An owned, growable string of well-formed WTF-8 data.
///
/// Similar to `String`, but can additionally contain surrogate code points
/// if they’re not in a surrogate pair.
#[derive(Eq, PartialEq, Ord, PartialOrd, Clone)]
pub struct Wtf8Buf {
    bytes: Vec<u8>
}

impl ops::Deref for Wtf8Buf {
    type Target = Wtf8;

    fn deref(&self) -> &Wtf8 {
        self.as_slice()
    }
}

impl ops::DerefMut for Wtf8Buf {
    fn deref_mut(&mut self) -> &mut Wtf8 {
        self.as_mut_slice()
    }
}

/// Format the string with double quotes,
/// and surrogates as `\u` followed by four hexadecimal digits.
/// Example: `"a\u{D800}"` for a string with code points [U+0061, U+D800]
impl fmt::Debug for Wtf8Buf {
    #[inline]
    fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
        fmt::Debug::fmt(&**self, formatter)
    }
}

impl Wtf8Buf {
    /// Creates a new, empty WTF-8 string.
    #[inline]
    pub fn new() -> Wtf8Buf {
        Wtf8Buf { bytes: Vec::new() }
    }

    /// Creates a new, empty WTF-8 string with pre-allocated capacity for `n` bytes.
    #[inline]
    pub fn with_capacity(n: usize) -> Wtf8Buf {
        Wtf8Buf { bytes: Vec::with_capacity(n) }
    }

    /// Creates a WTF-8 string from a UTF-8 `String`.
    ///
    /// This takes ownership of the `String` and does not copy.
    ///
    /// Since WTF-8 is a superset of UTF-8, this always succeeds.
    #[inline]
    pub fn from_string(string: String) -> Wtf8Buf {
        Wtf8Buf { bytes: string.into_bytes() }
    }

    /// Creates a WTF-8 string from a UTF-8 `&str` slice.
    ///
    /// This copies the content of the slice.
    ///
    /// Since WTF-8 is a superset of UTF-8, this always succeeds.
    #[inline]
    pub fn from_str(str: &str) -> Wtf8Buf {
        Wtf8Buf { bytes: <[_]>::to_vec(str.as_bytes()) }
    }

    pub fn clear(&mut self) {
        self.bytes.clear()
    }

    /// Creates a WTF-8 string from a potentially ill-formed UTF-16 slice of 16-bit code units.
    ///
    /// This is lossless: calling `.encode_wide()` on the resulting string
    /// will always return the original code units.
    pub fn from_wide(v: &[u16]) -> Wtf8Buf {
        let mut string = Wtf8Buf::with_capacity(v.len());
        for item in char::decode_utf16(v.iter().cloned()) {
            match item {
                Ok(ch) => string.push_char(ch),
                Err(surrogate) => {
                    let surrogate = surrogate.unpaired_surrogate();
                    // Surrogates are known to be in the code point range.
                    let code_point = unsafe {
                        CodePoint::from_u32_unchecked(surrogate as u32)
                    };
                    // Skip the WTF-8 concatenation check,
                    // surrogate pairs are already decoded by decode_utf16
                    string.push_code_point_unchecked(code_point)
                }
            }
        }
        string
    }

    /// Copied from String::push
    /// This does **not** include the WTF-8 concatenation check.
    fn push_code_point_unchecked(&mut self, code_point: CodePoint) {
        let c = unsafe {
            char::from_u32_unchecked(code_point.value)
        };
        let mut bytes = [0; 4];
        let bytes = c.encode_utf8(&mut bytes).as_bytes();
        self.bytes.extend_from_slice(bytes)
    }

    #[inline]
    pub fn as_slice(&self) -> &Wtf8 {
        unsafe { Wtf8::from_bytes_unchecked(&self.bytes) }
    }

    #[inline]
    pub fn as_mut_slice(&mut self) -> &mut Wtf8 {
        unsafe { Wtf8::from_mut_bytes_unchecked(&mut self.bytes) }
    }

    /// Reserves capacity for at least `additional` more bytes to be inserted
    /// in the given `Wtf8Buf`.
    /// The collection may reserve more space to avoid frequent reallocations.
    ///
    /// # Panics
    ///
    /// Panics if the new capacity overflows `usize`.
    #[inline]
    pub fn reserve(&mut self, additional: usize) {
        self.bytes.reserve(additional)
    }

    #[inline]
    pub fn reserve_exact(&mut self, additional: usize) {
        self.bytes.reserve_exact(additional)
    }

    #[inline]
    pub fn shrink_to_fit(&mut self) {
        self.bytes.shrink_to_fit()
    }

    #[inline]
    pub fn shrink_to(&mut self, min_capacity: usize) {
        self.bytes.shrink_to(min_capacity)
    }

    /// Returns the number of bytes that this string buffer can hold without reallocating.
    #[inline]
    pub fn capacity(&self) -> usize {
        self.bytes.capacity()
    }

    /// Append a UTF-8 slice at the end of the string.
    #[inline]
    pub fn push_str(&mut self, other: &str) {
        self.bytes.extend_from_slice(other.as_bytes())
    }

    /// Append a WTF-8 slice at the end of the string.
    ///
    /// This replaces newly paired surrogates at the boundary
    /// with a supplementary code point,
    /// like concatenating ill-formed UTF-16 strings effectively would.
    #[inline]
    pub fn push_wtf8(&mut self, other: &Wtf8) {
        match ((&*self).final_lead_surrogate(), other.initial_trail_surrogate()) {
            // Replace newly paired surrogates by a supplementary code point.
            (Some(lead), Some(trail)) => {
                let len_without_lead_surrogate = self.len() - 3;
                self.bytes.truncate(len_without_lead_surrogate);
                let other_without_trail_surrogate = &other.bytes[3..];
                // 4 bytes for the supplementary code point
                self.bytes.reserve(4 + other_without_trail_surrogate.len());
                self.push_char(decode_surrogate_pair(lead, trail));
                self.bytes.extend_from_slice(other_without_trail_surrogate);
            }
            _ => self.bytes.extend_from_slice(&other.bytes)
        }
    }

    /// Append a Unicode scalar value at the end of the string.
    #[inline]
    pub fn push_char(&mut self, c: char) {
        self.push_code_point_unchecked(CodePoint::from_char(c))
    }

    /// Append a code point at the end of the string.
    ///
    /// This replaces newly paired surrogates at the boundary
    /// with a supplementary code point,
    /// like concatenating ill-formed UTF-16 strings effectively would.
    #[inline]
    pub fn push(&mut self, code_point: CodePoint) {
        if let trail @ 0xDC00...0xDFFF = code_point.to_u32() {
            if let Some(lead) = (&*self).final_lead_surrogate() {
                let len_without_lead_surrogate = self.len() - 3;
                self.bytes.truncate(len_without_lead_surrogate);
                self.push_char(decode_surrogate_pair(lead, trail as u16));
                return
            }
        }

        // No newly paired surrogates at the boundary.
        self.push_code_point_unchecked(code_point)
    }

    /// Shortens a string to the specified length.
    ///
    /// # Panics
    ///
    /// Panics if `new_len` > current length,
    /// or if `new_len` is not a code point boundary.
    #[inline]
    pub fn truncate(&mut self, new_len: usize) {
        assert!(is_code_point_boundary(self, new_len));
        self.bytes.truncate(new_len)
    }

    /// Consumes the WTF-8 string and tries to convert it to UTF-8.
    ///
    /// This does not copy the data.
    ///
    /// If the contents are not well-formed UTF-8
    /// (that is, if the string contains surrogates),
    /// the original WTF-8 string is returned instead.
    pub fn into_string(self) -> Result<String, Wtf8Buf> {
        match self.next_surrogate(0) {
            None => Ok(unsafe { String::from_utf8_unchecked(self.bytes) }),
            Some(_) => Err(self),
        }
    }

    /// Consumes the WTF-8 string and converts it lossily to UTF-8.
    ///
    /// This does not copy the data (but may overwrite parts of it in place).
    ///
    /// Surrogates are replaced with `"\u{FFFD}"` (the replacement character “�”)
    pub fn into_string_lossy(mut self) -> String {
        let mut pos = 0;
        loop {
            match self.next_surrogate(pos) {
                Some((surrogate_pos, _)) => {
                    pos = surrogate_pos + 3;
                    self.bytes[surrogate_pos..pos]
                        .copy_from_slice(UTF8_REPLACEMENT_CHARACTER.as_bytes());
                },
                None => return unsafe { String::from_utf8_unchecked(self.bytes) }
            }
        }
    }

    /// Converts this `Wtf8Buf` into a boxed `Wtf8`.
    #[inline]
    pub fn into_box(self) -> Box<Wtf8> {
        unsafe { mem::transmute(self.bytes.into_boxed_slice()) }
    }

    /// Converts a `Box<Wtf8>` into a `Wtf8Buf`.
    pub fn from_box(boxed: Box<Wtf8>) -> Wtf8Buf {
        let bytes: Box<[u8]> = unsafe { mem::transmute(boxed) };
        Wtf8Buf { bytes: bytes.into_vec() }
    }
}

/// Create a new WTF-8 string from an iterator of code points.
///
/// This replaces surrogate code point pairs with supplementary code points,
/// like concatenating ill-formed UTF-16 strings effectively would.
impl FromIterator<CodePoint> for Wtf8Buf {
    fn from_iter<T: IntoIterator<Item=CodePoint>>(iter: T) -> Wtf8Buf {
        let mut string = Wtf8Buf::new();
        string.extend(iter);
        string
    }
}

/// Append code points from an iterator to the string.
///
/// This replaces surrogate code point pairs with supplementary code points,
/// like concatenating ill-formed UTF-16 strings effectively would.
impl Extend<CodePoint> for Wtf8Buf {
    fn extend<T: IntoIterator<Item=CodePoint>>(&mut self, iter: T) {
        let iterator = iter.into_iter();
        let (low, _high) = iterator.size_hint();
        // Lower bound of one byte per code point (ASCII only)
        self.bytes.reserve(low);
        for code_point in iterator {
            self.push(code_point);
        }
    }
}

/// A borrowed slice of well-formed WTF-8 data.
///
/// Similar to `&str`, but can additionally contain surrogate code points
/// if they’re not in a surrogate pair.
#[derive(Eq, Ord, PartialEq, PartialOrd)]
pub struct Wtf8 {
    bytes: [u8]
}

impl AsInner<[u8]> for Wtf8 {
    fn as_inner(&self) -> &[u8] { &self.bytes }
}

/// Format the slice with double quotes,
/// and surrogates as `\u` followed by four hexadecimal digits.
/// Example: `"a\u{D800}"` for a slice with code points [U+0061, U+D800]
impl fmt::Debug for Wtf8 {
    fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
        fn write_str_escaped(f: &mut fmt::Formatter, s: &str) -> fmt::Result {
            use fmt::Write;
            for c in s.chars().flat_map(|c| c.escape_debug()) {
                f.write_char(c)?
            }
            Ok(())
        }

        formatter.write_str("\"")?;
        let mut pos = 0;
        while let Some((surrogate_pos, surrogate)) = self.next_surrogate(pos) {
            write_str_escaped(
                formatter,
                unsafe { str::from_utf8_unchecked(
                    &self.bytes[pos .. surrogate_pos]
                )},
            )?;
            write!(formatter, "\\u{{{:x}}}", surrogate)?;
            pos = surrogate_pos + 3;
        }
        write_str_escaped(
            formatter,
            unsafe { str::from_utf8_unchecked(&self.bytes[pos..]) },
        )?;
        formatter.write_str("\"")
    }
}

impl fmt::Display for Wtf8 {
    fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
        let wtf8_bytes = &self.bytes;
        let mut pos = 0;
        loop {
            match self.next_surrogate(pos) {
                Some((surrogate_pos, _)) => {
                    formatter.write_str(unsafe {
                        str::from_utf8_unchecked(&wtf8_bytes[pos .. surrogate_pos])
                    })?;
                    formatter.write_str(UTF8_REPLACEMENT_CHARACTER)?;
                    pos = surrogate_pos + 3;
                },
                None => {
                    let s = unsafe {
                        str::from_utf8_unchecked(&wtf8_bytes[pos..])
                    };
                    if pos == 0 {
                        return s.fmt(formatter)
                    } else {
                        return formatter.write_str(s)
                    }
                }
            }
        }
    }
}

impl Wtf8 {
    /// Creates a WTF-8 slice from a UTF-8 `&str` slice.
    ///
    /// Since WTF-8 is a superset of UTF-8, this always succeeds.
    #[inline]
    pub fn from_str(value: &str) -> &Wtf8 {
        unsafe { Wtf8::from_bytes_unchecked(value.as_bytes()) }
    }

    /// Creates a WTF-8 slice from a WTF-8 byte slice.
    ///
    /// Since the byte slice is not checked for valid WTF-8, this functions is
    /// marked unsafe.
    #[inline]
    unsafe fn from_bytes_unchecked(value: &[u8]) -> &Wtf8 {
        mem::transmute(value)
    }

    /// Creates a mutable WTF-8 slice from a mutable WTF-8 byte slice.
    ///
    /// Since the byte slice is not checked for valid WTF-8, this functions is
    /// marked unsafe.
    #[inline]
    unsafe fn from_mut_bytes_unchecked(value: &mut [u8]) -> &mut Wtf8 {
        mem::transmute(value)
    }

    /// Returns the length, in WTF-8 bytes.
    #[inline]
    pub fn len(&self) -> usize {
        self.bytes.len()
    }

    #[inline]
    pub fn is_empty(&self) -> bool {
        self.bytes.is_empty()
    }

    /// Returns the code point at `position` if it is in the ASCII range,
    /// or `b'\xFF' otherwise.
    ///
    /// # Panics
    ///
    /// Panics if `position` is beyond the end of the string.
    #[inline]
    pub fn ascii_byte_at(&self, position: usize) -> u8 {
        match self.bytes[position] {
            ascii_byte @ 0x00 ... 0x7F => ascii_byte,
            _ => 0xFF
        }
    }

    /// Returns an iterator for the string’s code points.
    #[inline]
    pub fn code_points(&self) -> Wtf8CodePoints {
        Wtf8CodePoints { bytes: self.bytes.iter() }
    }

    /// Tries to convert the string to UTF-8 and return a `&str` slice.
    ///
    /// Returns `None` if the string contains surrogates.
    ///
    /// This does not copy the data.
    #[inline]
    pub fn as_str(&self) -> Option<&str> {
        // Well-formed WTF-8 is also well-formed UTF-8
        // if and only if it contains no surrogate.
        match self.next_surrogate(0) {
            None => Some(unsafe { str::from_utf8_unchecked(&self.bytes) }),
            Some(_) => None,
        }
    }

    /// Lossily converts the string to UTF-8.
    /// Returns a UTF-8 `&str` slice if the contents are well-formed in UTF-8.
    ///
    /// Surrogates are replaced with `"\u{FFFD}"` (the replacement character “�”).
    ///
    /// This only copies the data if necessary (if it contains any surrogate).
    pub fn to_string_lossy(&self) -> Cow<str> {
        let surrogate_pos = match self.next_surrogate(0) {
            None => return Cow::Borrowed(unsafe { str::from_utf8_unchecked(&self.bytes) }),
            Some((pos, _)) => pos,
        };
        let wtf8_bytes = &self.bytes;
        let mut utf8_bytes = Vec::with_capacity(self.len());
        utf8_bytes.extend_from_slice(&wtf8_bytes[..surrogate_pos]);
        utf8_bytes.extend_from_slice(UTF8_REPLACEMENT_CHARACTER.as_bytes());
        let mut pos = surrogate_pos + 3;
        loop {
            match self.next_surrogate(pos) {
                Some((surrogate_pos, _)) => {
                    utf8_bytes.extend_from_slice(&wtf8_bytes[pos .. surrogate_pos]);
                    utf8_bytes.extend_from_slice(UTF8_REPLACEMENT_CHARACTER.as_bytes());
                    pos = surrogate_pos + 3;
                },
                None => {
                    utf8_bytes.extend_from_slice(&wtf8_bytes[pos..]);
                    return Cow::Owned(unsafe { String::from_utf8_unchecked(utf8_bytes) })
                }
            }
        }
    }

    /// Converts the WTF-8 string to potentially ill-formed UTF-16
    /// and return an iterator of 16-bit code units.
    ///
    /// This is lossless:
    /// calling `Wtf8Buf::from_ill_formed_utf16` on the resulting code units
    /// would always return the original WTF-8 string.
    #[inline]
    pub fn encode_wide(&self) -> EncodeWide {
        EncodeWide { code_points: self.code_points(), extra: 0 }
    }

    #[inline]
    fn next_surrogate(&self, mut pos: usize) -> Option<(usize, u16)> {
        let mut iter = self.bytes[pos..].iter();
        loop {
            let b = *iter.next()?;
            if b < 0x80 {
                pos += 1;
            } else if b < 0xE0 {
                iter.next();
                pos += 2;
            } else if b == 0xED {
                match (iter.next(), iter.next()) {
                    (Some(&b2), Some(&b3)) if b2 >= 0xA0 => {
                        return Some((pos, decode_surrogate(b2, b3)))
                    }
                    _ => pos += 3
                }
            } else if b < 0xF0 {
                iter.next();
                iter.next();
                pos += 3;
            } else {
                iter.next();
                iter.next();
                iter.next();
                pos += 4;
            }
        }
    }

    #[inline]
    fn final_lead_surrogate(&self) -> Option<u16> {
        let len = self.len();
        if len < 3 {
            return None
        }
        match &self.bytes[(len - 3)..] {
            &[0xED, b2 @ 0xA0...0xAF, b3] => Some(decode_surrogate(b2, b3)),
            _ => None
        }
    }

    #[inline]
    fn initial_trail_surrogate(&self) -> Option<u16> {
        let len = self.len();
        if len < 3 {
            return None
        }
        match &self.bytes[..3] {
            &[0xED, b2 @ 0xB0...0xBF, b3] => Some(decode_surrogate(b2, b3)),
            _ => None
        }
    }

    /// Boxes this `Wtf8`.
    #[inline]
    pub fn into_box(&self) -> Box<Wtf8> {
        let boxed: Box<[u8]> = self.bytes.into();
        unsafe { mem::transmute(boxed) }
    }

    /// Creates a boxed, empty `Wtf8`.
    pub fn empty_box() -> Box<Wtf8> {
        let boxed: Box<[u8]> = Default::default();
        unsafe { mem::transmute(boxed) }
    }

    #[inline]
    pub fn into_arc(&self) -> Arc<Wtf8> {
        let arc: Arc<[u8]> = Arc::from(&self.bytes);
        unsafe { Arc::from_raw(Arc::into_raw(arc) as *const Wtf8) }
    }

    #[inline]
    pub fn into_rc(&self) -> Rc<Wtf8> {
        let rc: Rc<[u8]> = Rc::from(&self.bytes);
        unsafe { Rc::from_raw(Rc::into_raw(rc) as *const Wtf8) }
    }
}


/// Return a slice of the given string for the byte range [`begin`..`end`).
///
/// # Panics
///
/// Panics when `begin` and `end` do not point to code point boundaries,
/// or point beyond the end of the string.
impl ops::Index<ops::Range<usize>> for Wtf8 {
    type Output = Wtf8;

    #[inline]
    fn index(&self, range: ops::Range<usize>) -> &Wtf8 {
        // is_code_point_boundary checks that the index is in [0, .len()]
        if range.start <= range.end &&
           is_code_point_boundary(self, range.start) &&
           is_code_point_boundary(self, range.end) {
            unsafe { slice_unchecked(self, range.start, range.end) }
        } else {
            slice_error_fail(self, range.start, range.end)
        }
    }
}

/// Return a slice of the given string from byte `begin` to its end.
///
/// # Panics
///
/// Panics when `begin` is not at a code point boundary,
/// or is beyond the end of the string.
impl ops::Index<ops::RangeFrom<usize>> for Wtf8 {
    type Output = Wtf8;

    #[inline]
    fn index(&self, range: ops::RangeFrom<usize>) -> &Wtf8 {
        // is_code_point_boundary checks that the index is in [0, .len()]
        if is_code_point_boundary(self, range.start) {
            unsafe { slice_unchecked(self, range.start, self.len()) }
        } else {
            slice_error_fail(self, range.start, self.len())
        }
    }
}

/// Return a slice of the given string from its beginning to byte `end`.
///
/// # Panics
///
/// Panics when `end` is not at a code point boundary,
/// or is beyond the end of the string.
impl ops::Index<ops::RangeTo<usize>> for Wtf8 {
    type Output = Wtf8;

    #[inline]
    fn index(&self, range: ops::RangeTo<usize>) -> &Wtf8 {
        // is_code_point_boundary checks that the index is in [0, .len()]
        if is_code_point_boundary(self, range.end) {
            unsafe { slice_unchecked(self, 0, range.end) }
        } else {
            slice_error_fail(self, 0, range.end)
        }
    }
}

impl ops::Index<ops::RangeFull> for Wtf8 {
    type Output = Wtf8;

    #[inline]
    fn index(&self, _range: ops::RangeFull) -> &Wtf8 {
        self
    }
}

#[inline]
fn decode_surrogate(second_byte: u8, third_byte: u8) -> u16 {
    // The first byte is assumed to be 0xED
    0xD800 | (second_byte as u16 & 0x3F) << 6 | third_byte as u16 & 0x3F
}

#[inline]
fn decode_surrogate_pair(lead: u16, trail: u16) -> char {
    let code_point = 0x10000 + ((((lead - 0xD800) as u32) << 10) | (trail - 0xDC00) as u32);
    unsafe { char::from_u32_unchecked(code_point) }
}

/// Copied from core::str::StrPrelude::is_char_boundary
#[inline]
pub fn is_code_point_boundary(slice: &Wtf8, index: usize) -> bool {
    if index == slice.len() { return true; }
    match slice.bytes.get(index) {
        None => false,
        Some(&b) => b < 128 || b >= 192,
    }
}

/// Copied from core::str::raw::slice_unchecked
#[inline]
pub unsafe fn slice_unchecked(s: &Wtf8, begin: usize, end: usize) -> &Wtf8 {
    // memory layout of an &[u8] and &Wtf8 are the same
    Wtf8::from_bytes_unchecked(slice::from_raw_parts(
        s.bytes.as_ptr().offset(begin as isize),
        end - begin
    ))
}

/// Copied from core::str::raw::slice_error_fail
#[inline(never)]
pub fn slice_error_fail(s: &Wtf8, begin: usize, end: usize) -> ! {
    assert!(begin <= end);
    panic!("index {} and/or {} in `{:?}` do not lie on character boundary",
          begin, end, s);
}

/// Iterator for the code points of a WTF-8 string.
///
/// Created with the method `.code_points()`.
#[derive(Clone)]
pub struct Wtf8CodePoints<'a> {
    bytes: slice::Iter<'a, u8>
}

impl<'a> Iterator for Wtf8CodePoints<'a> {
    type Item = CodePoint;

    #[inline]
    fn next(&mut self) -> Option<CodePoint> {
        next_code_point(&mut self.bytes).map(|c| CodePoint { value: c })
    }

    #[inline]
    fn size_hint(&self) -> (usize, Option<usize>) {
        let len = self.bytes.len();
        (len.saturating_add(3) / 4, Some(len))
    }
}

/// Generates a wide character sequence for potentially ill-formed UTF-16.
#[stable(feature = "rust1", since = "1.0.0")]
#[derive(Clone)]
pub struct EncodeWide<'a> {
    code_points: Wtf8CodePoints<'a>,
    extra: u16
}

// Copied from libunicode/u_str.rs
#[stable(feature = "rust1", since = "1.0.0")]
impl<'a> Iterator for EncodeWide<'a> {
    type Item = u16;

    #[inline]
    fn next(&mut self) -> Option<u16> {
        if self.extra != 0 {
            let tmp = self.extra;
            self.extra = 0;
            return Some(tmp);
        }

        let mut buf = [0; 2];
        self.code_points.next().map(|code_point| {
            let c = unsafe {
                char::from_u32_unchecked(code_point.value)
            };
            let n = c.encode_utf16(&mut buf).len();
            if n == 2 {
                self.extra = buf[1];
            }
            buf[0]
        })
    }

    #[inline]
    fn size_hint(&self) -> (usize, Option<usize>) {
        let (low, high) = self.code_points.size_hint();
        // every code point gets either one u16 or two u16,
        // so this iterator is between 1 or 2 times as
        // long as the underlying iterator.
        (low, high.and_then(|n| n.checked_mul(2)))
    }
}

impl Hash for CodePoint {
    #[inline]
    fn hash<H: Hasher>(&self, state: &mut H) {
        self.value.hash(state)
    }
}

impl Hash for Wtf8Buf {
    #[inline]
    fn hash<H: Hasher>(&self, state: &mut H) {
        state.write(&self.bytes);
        0xfeu8.hash(state)
    }
}

impl Hash for Wtf8 {
    #[inline]
    fn hash<H: Hasher>(&self, state: &mut H) {
        state.write(&self.bytes);
        0xfeu8.hash(state)
    }
}

impl Wtf8 {
    pub fn make_ascii_uppercase(&mut self) { self.bytes.make_ascii_uppercase() }
}

#[cfg(test)]
mod tests {
    use borrow::Cow;
    use super::*;

    #[test]
    fn code_point_from_u32() {
        assert!(CodePoint::from_u32(0).is_some());
        assert!(CodePoint::from_u32(0xD800).is_some());
        assert!(CodePoint::from_u32(0x10FFFF).is_some());
        assert!(CodePoint::from_u32(0x110000).is_none());
    }

    #[test]
    fn code_point_to_u32() {
        fn c(value: u32) -> CodePoint { CodePoint::from_u32(value).unwrap() }
        assert_eq!(c(0).to_u32(), 0);
        assert_eq!(c(0xD800).to_u32(), 0xD800);
        assert_eq!(c(0x10FFFF).to_u32(), 0x10FFFF);
    }

    #[test]
    fn code_point_from_char() {
        assert_eq!(CodePoint::from_char('a').to_u32(), 0x61);
        assert_eq!(CodePoint::from_char('💩').to_u32(), 0x1F4A9);
    }

    #[test]
    fn code_point_to_string() {
        assert_eq!(format!("{:?}", CodePoint::from_char('a')), "U+0061");
        assert_eq!(format!("{:?}", CodePoint::from_char('💩')), "U+1F4A9");
    }

    #[test]
    fn code_point_to_char() {
        fn c(value: u32) -> CodePoint { CodePoint::from_u32(value).unwrap() }
        assert_eq!(c(0x61).to_char(), Some('a'));
        assert_eq!(c(0x1F4A9).to_char(), Some('💩'));
        assert_eq!(c(0xD800).to_char(), None);
    }

    #[test]
    fn code_point_to_char_lossy() {
        fn c(value: u32) -> CodePoint { CodePoint::from_u32(value).unwrap() }
        assert_eq!(c(0x61).to_char_lossy(), 'a');
        assert_eq!(c(0x1F4A9).to_char_lossy(), '💩');
        assert_eq!(c(0xD800).to_char_lossy(), '\u{FFFD}');
    }

    #[test]
    fn wtf8buf_new() {
        assert_eq!(Wtf8Buf::new().bytes, b"");
    }

    #[test]
    fn wtf8buf_from_str() {
        assert_eq!(Wtf8Buf::from_str("").bytes, b"");
        assert_eq!(Wtf8Buf::from_str("aé 💩").bytes,
                   b"a\xC3\xA9 \xF0\x9F\x92\xA9");
    }

    #[test]
    fn wtf8buf_from_string() {
        assert_eq!(Wtf8Buf::from_string(String::from("")).bytes, b"");
        assert_eq!(Wtf8Buf::from_string(String::from("aé 💩")).bytes,
                   b"a\xC3\xA9 \xF0\x9F\x92\xA9");
    }

    #[test]
    fn wtf8buf_from_wide() {
        assert_eq!(Wtf8Buf::from_wide(&[]).bytes, b"");
        assert_eq!(Wtf8Buf::from_wide(
                      &[0x61, 0xE9, 0x20, 0xD83D, 0xD83D, 0xDCA9]).bytes,
                   b"a\xC3\xA9 \xED\xA0\xBD\xF0\x9F\x92\xA9");
    }

    #[test]
    fn wtf8buf_push_str() {
        let mut string = Wtf8Buf::new();
        assert_eq!(string.bytes, b"");
        string.push_str("aé 💩");
        assert_eq!(string.bytes, b"a\xC3\xA9 \xF0\x9F\x92\xA9");
    }

    #[test]
    fn wtf8buf_push_char() {
        let mut string = Wtf8Buf::from_str("aé ");
        assert_eq!(string.bytes, b"a\xC3\xA9 ");
        string.push_char('💩');
        assert_eq!(string.bytes, b"a\xC3\xA9 \xF0\x9F\x92\xA9");
    }

    #[test]
    fn wtf8buf_push() {
        let mut string = Wtf8Buf::from_str("aé ");
        assert_eq!(string.bytes, b"a\xC3\xA9 ");
        string.push(CodePoint::from_char('💩'));
        assert_eq!(string.bytes, b"a\xC3\xA9 \xF0\x9F\x92\xA9");

        fn c(value: u32) -> CodePoint { CodePoint::from_u32(value).unwrap() }

        let mut string = Wtf8Buf::new();
        string.push(c(0xD83D));  // lead
        string.push(c(0xDCA9));  // trail
        assert_eq!(string.bytes, b"\xF0\x9F\x92\xA9");  // Magic!

        let mut string = Wtf8Buf::new();
        string.push(c(0xD83D));  // lead
        string.push(c(0x20));  // not surrogate
        string.push(c(0xDCA9));  // trail
        assert_eq!(string.bytes, b"\xED\xA0\xBD \xED\xB2\xA9");

        let mut string = Wtf8Buf::new();
        string.push(c(0xD800));  // lead
        string.push(c(0xDBFF));  // lead
        assert_eq!(string.bytes, b"\xED\xA0\x80\xED\xAF\xBF");

        let mut string = Wtf8Buf::new();
        string.push(c(0xD800));  // lead
        string.push(c(0xE000));  // not surrogate
        assert_eq!(string.bytes, b"\xED\xA0\x80\xEE\x80\x80");

        let mut string = Wtf8Buf::new();
        string.push(c(0xD7FF));  // not surrogate
        string.push(c(0xDC00));  // trail
        assert_eq!(string.bytes, b"\xED\x9F\xBF\xED\xB0\x80");

        let mut string = Wtf8Buf::new();
        string.push(c(0x61));  // not surrogate, < 3 bytes
        string.push(c(0xDC00));  // trail
        assert_eq!(string.bytes, b"\x61\xED\xB0\x80");

        let mut string = Wtf8Buf::new();
        string.push(c(0xDC00));  // trail
        assert_eq!(string.bytes, b"\xED\xB0\x80");
    }

    #[test]
    fn wtf8buf_push_wtf8() {
        let mut string = Wtf8Buf::from_str("aé");
        assert_eq!(string.bytes, b"a\xC3\xA9");
        string.push_wtf8(Wtf8::from_str(" 💩"));
        assert_eq!(string.bytes, b"a\xC3\xA9 \xF0\x9F\x92\xA9");

        fn w(v: &[u8]) -> &Wtf8 { unsafe { Wtf8::from_bytes_unchecked(v) } }

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"\xED\xA0\xBD"));  // lead
        string.push_wtf8(w(b"\xED\xB2\xA9"));  // trail
        assert_eq!(string.bytes, b"\xF0\x9F\x92\xA9");  // Magic!

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"\xED\xA0\xBD"));  // lead
        string.push_wtf8(w(b" "));  // not surrogate
        string.push_wtf8(w(b"\xED\xB2\xA9"));  // trail
        assert_eq!(string.bytes, b"\xED\xA0\xBD \xED\xB2\xA9");

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"\xED\xA0\x80"));  // lead
        string.push_wtf8(w(b"\xED\xAF\xBF"));  // lead
        assert_eq!(string.bytes, b"\xED\xA0\x80\xED\xAF\xBF");

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"\xED\xA0\x80"));  // lead
        string.push_wtf8(w(b"\xEE\x80\x80"));  // not surrogate
        assert_eq!(string.bytes, b"\xED\xA0\x80\xEE\x80\x80");

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"\xED\x9F\xBF"));  // not surrogate
        string.push_wtf8(w(b"\xED\xB0\x80"));  // trail
        assert_eq!(string.bytes, b"\xED\x9F\xBF\xED\xB0\x80");

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"a"));  // not surrogate, < 3 bytes
        string.push_wtf8(w(b"\xED\xB0\x80"));  // trail
        assert_eq!(string.bytes, b"\x61\xED\xB0\x80");

        let mut string = Wtf8Buf::new();
        string.push_wtf8(w(b"\xED\xB0\x80"));  // trail
        assert_eq!(string.bytes, b"\xED\xB0\x80");
    }

    #[test]
    fn wtf8buf_truncate() {
        let mut string = Wtf8Buf::from_str("aé");
        string.truncate(1);
        assert_eq!(string.bytes, b"a");
    }

    #[test]
    #[should_panic]
    fn wtf8buf_truncate_fail_code_point_boundary() {
        let mut string = Wtf8Buf::from_str("aé");
        string.truncate(2);
    }

    #[test]
    #[should_panic]
    fn wtf8buf_truncate_fail_longer() {
        let mut string = Wtf8Buf::from_str("aé");
        string.truncate(4);
    }

    #[test]
    fn wtf8buf_into_string() {
        let mut string = Wtf8Buf::from_str("aé 💩");
        assert_eq!(string.clone().into_string(), Ok(String::from("aé 💩")));
        string.push(CodePoint::from_u32(0xD800).unwrap());
        assert_eq!(string.clone().into_string(), Err(string));
    }

    #[test]
    fn wtf8buf_into_string_lossy() {
        let mut string = Wtf8Buf::from_str("aé 💩");
        assert_eq!(string.clone().into_string_lossy(), String::from("aé 💩"));
        string.push(CodePoint::from_u32(0xD800).unwrap());
        assert_eq!(string.clone().into_string_lossy(), String::from("aé 💩�"));
    }

    #[test]
    fn wtf8buf_from_iterator() {
        fn f(values: &[u32]) -> Wtf8Buf {
            values.iter().map(|&c| CodePoint::from_u32(c).unwrap()).collect::<Wtf8Buf>()
        }
        assert_eq!(f(&[0x61, 0xE9, 0x20, 0x1F4A9]).bytes, b"a\xC3\xA9 \xF0\x9F\x92\xA9");

        assert_eq!(f(&[0xD83D, 0xDCA9]).bytes, b"\xF0\x9F\x92\xA9");  // Magic!
        assert_eq!(f(&[0xD83D, 0x20, 0xDCA9]).bytes, b"\xED\xA0\xBD \xED\xB2\xA9");
        assert_eq!(f(&[0xD800, 0xDBFF]).bytes, b"\xED\xA0\x80\xED\xAF\xBF");
        assert_eq!(f(&[0xD800, 0xE000]).bytes, b"\xED\xA0\x80\xEE\x80\x80");
        assert_eq!(f(&[0xD7FF, 0xDC00]).bytes, b"\xED\x9F\xBF\xED\xB0\x80");
        assert_eq!(f(&[0x61, 0xDC00]).bytes, b"\x61\xED\xB0\x80");
        assert_eq!(f(&[0xDC00]).bytes, b"\xED\xB0\x80");
    }

    #[test]
    fn wtf8buf_extend() {
        fn e(initial: &[u32], extended: &[u32]) -> Wtf8Buf {
            fn c(value: &u32) -> CodePoint { CodePoint::from_u32(*value).unwrap() }
            let mut string = initial.iter().map(c).collect::<Wtf8Buf>();
            string.extend(extended.iter().map(c));
            string
        }

        assert_eq!(e(&[0x61, 0xE9], &[0x20, 0x1F4A9]).bytes,
                   b"a\xC3\xA9 \xF0\x9F\x92\xA9");

        assert_eq!(e(&[0xD83D], &[0xDCA9]).bytes, b"\xF0\x9F\x92\xA9");  // Magic!
        assert_eq!(e(&[0xD83D, 0x20], &[0xDCA9]).bytes, b"\xED\xA0\xBD \xED\xB2\xA9");
        assert_eq!(e(&[0xD800], &[0xDBFF]).bytes, b"\xED\xA0\x80\xED\xAF\xBF");
        assert_eq!(e(&[0xD800], &[0xE000]).bytes, b"\xED\xA0\x80\xEE\x80\x80");
        assert_eq!(e(&[0xD7FF], &[0xDC00]).bytes, b"\xED\x9F\xBF\xED\xB0\x80");
        assert_eq!(e(&[0x61], &[0xDC00]).bytes, b"\x61\xED\xB0\x80");
        assert_eq!(e(&[], &[0xDC00]).bytes, b"\xED\xB0\x80");
    }

    #[test]
    fn wtf8buf_show() {
        let mut string = Wtf8Buf::from_str("a\té \u{7f}💩\r");
        string.push(CodePoint::from_u32(0xD800).unwrap());
        assert_eq!(format!("{:?}", string), "\"a\\té \\u{7f}\u{1f4a9}\\r\\u{d800}\"");
    }

    #[test]
    fn wtf8buf_as_slice() {
        assert_eq!(Wtf8Buf::from_str("aé").as_slice(), Wtf8::from_str("aé"));
    }

    #[test]
    fn wtf8buf_show_str() {
        let text = "a\té 💩\r";
        let string = Wtf8Buf::from_str(text);
        assert_eq!(format!("{:?}", text), format!("{:?}", string));
    }

    #[test]
    fn wtf8_from_str() {
        assert_eq!(&Wtf8::from_str("").bytes, b"");
        assert_eq!(&Wtf8::from_str("aé 💩").bytes, b"a\xC3\xA9 \xF0\x9F\x92\xA9");
    }

    #[test]
    fn wtf8_len() {
        assert_eq!(Wtf8::from_str("").len(), 0);
        assert_eq!(Wtf8::from_str("aé 💩").len(), 8);
    }

    #[test]
    fn wtf8_slice() {
        assert_eq!(&Wtf8::from_str("aé 💩")[1.. 4].bytes, b"\xC3\xA9 ");
    }

    #[test]
    #[should_panic]
    fn wtf8_slice_not_code_point_boundary() {
        &Wtf8::from_str("aé 💩")[2.. 4];
    }

    #[test]
    fn wtf8_slice_from() {
        assert_eq!(&Wtf8::from_str("aé 💩")[1..].bytes, b"\xC3\xA9 \xF0\x9F\x92\xA9");
    }

    #[test]
    #[should_panic]
    fn wtf8_slice_from_not_code_point_boundary() {
        &Wtf8::from_str("aé 💩")[2..];
    }

    #[test]
    fn wtf8_slice_to() {
        assert_eq!(&Wtf8::from_str("aé 💩")[..4].bytes, b"a\xC3\xA9 ");
    }

    #[test]
    #[should_panic]
    fn wtf8_slice_to_not_code_point_boundary() {
        &Wtf8::from_str("aé 💩")[5..];
    }

    #[test]
    fn wtf8_ascii_byte_at() {
        let slice = Wtf8::from_str("aé 💩");
        assert_eq!(slice.ascii_byte_at(0), b'a');
        assert_eq!(slice.ascii_byte_at(1), b'\xFF');
        assert_eq!(slice.ascii_byte_at(2), b'\xFF');
        assert_eq!(slice.ascii_byte_at(3), b' ');
        assert_eq!(slice.ascii_byte_at(4), b'\xFF');
    }

    #[test]
    fn wtf8_code_points() {
        fn c(value: u32) -> CodePoint { CodePoint::from_u32(value).unwrap() }
        fn cp(string: &Wtf8Buf) -> Vec<Option<char>> {
            string.code_points().map(|c| c.to_char()).collect::<Vec<_>>()
        }
        let mut string = Wtf8Buf::from_str("é ");
        assert_eq!(cp(&string), [Some('é'), Some(' ')]);
        string.push(c(0xD83D));
        assert_eq!(cp(&string), [Some('é'), Some(' '), None]);
        string.push(c(0xDCA9));
        assert_eq!(cp(&string), [Some('é'), Some(' '), Some('💩')]);
    }

    #[test]
    fn wtf8_as_str() {
        assert_eq!(Wtf8::from_str("").as_str(), Some(""));
        assert_eq!(Wtf8::from_str("aé 💩").as_str(), Some("aé 💩"));
        let mut string = Wtf8Buf::new();
        string.push(CodePoint::from_u32(0xD800).unwrap());
        assert_eq!(string.as_str(), None);
    }

    #[test]
    fn wtf8_to_string_lossy() {
        assert_eq!(Wtf8::from_str("").to_string_lossy(), Cow::Borrowed(""));
        assert_eq!(Wtf8::from_str("aé 💩").to_string_lossy(), Cow::Borrowed("aé 💩"));
        let mut string = Wtf8Buf::from_str("aé 💩");
        string.push(CodePoint::from_u32(0xD800).unwrap());
        let expected: Cow<str> = Cow::Owned(String::from("aé 💩�"));
        assert_eq!(string.to_string_lossy(), expected);
    }

    #[test]
    fn wtf8_display() {
        fn d(b: &[u8]) -> String {
            format!("{}", &unsafe { Wtf8::from_bytes_unchecked(b) })
        }

        assert_eq!("", d("".as_bytes()));
        assert_eq!("aé 💩", d("aé 💩".as_bytes()));

        let mut string = Wtf8Buf::from_str("aé 💩");
        string.push(CodePoint::from_u32(0xD800).unwrap());
        assert_eq!("aé 💩�", d(string.as_inner()));
    }

    #[test]
    fn wtf8_encode_wide() {
        let mut string = Wtf8Buf::from_str("aé ");
        string.push(CodePoint::from_u32(0xD83D).unwrap());
        string.push_char('💩');
        assert_eq!(string.encode_wide().collect::<Vec<_>>(),
                   vec![0x61, 0xE9, 0x20, 0xD83D, 0xD83D, 0xDCA9]);
    }
}