wrest 0.5.5

Async HTTP client for Windows backed by WinHTTP, with a reqwest-compatible API
Documentation 
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

//! WHATWG Encoding Standard charset decoding.
//!
//! All 39 encodings mandated by the [WHATWG Encoding Standard][spec] are
//! supported:
//!
//! * **35 natively** -- UTF-8 takes a fast pure-Rust path; UTF-16LE/BE,
//!   x-user-defined, and the `replacement` pseudo-encoding are decoded in
//!   pure Rust as well; the remaining 30 single-byte and CJK encodings go
//!   through [`MultiByteToWideChar`].
//!
//! * **3 via ICU** -- ISO-8859-10 (Latin-6 / Nordic), ISO-8859-14
//!   (Latin-8 / Celtic), and EUC-JP (Extended Unix Code for Japanese)
//!   are absent from the Win32 NLS subsystem (see
//!   [below](#why-three-encodings-need-icu)).  On Windows 10 1903+ they
//!   are decoded at runtime through the system-bundled `icu.dll`.  On
//!   older builds, content labelled with these charsets will produce a
//!   decode error from [`Response::text()`](crate::Response::text).
//!
//! * **1 via lookup table** -- ISO-8859-16 (Latin-10 / South-Eastern
//!   European) is absent from both NLS *and* ICU, so it is decoded via a
//!   compile-time 128-entry Rust table (e.g. byte `0xAA` → `U+0218 Ș`).
//!   No runtime dependency required.
//!
//! # Why three encodings need ICU
//!
//! | WHATWG encoding | Windows CP | Why no NLS |
//! |-----------------|------------|------------|
//! | **ISO-8859-10** (Latin-6 / Nordic) | 28600 | No NLS file -- absent from `HKLM\...\Nls\CodePage`, not enumerated by `EnumSystemCodePages`. |
//! | **ISO-8859-14** (Latin-8 / Celtic) | 28604 | Same -- Microsoft never shipped an NLS implementation. |
//! | **EUC-JP** (Extended Unix Code for Japanese) | 51932 | Listed in the [Code Page Identifiers] table but implemented only in .NET managed code. `IsValidCodePage(51932)` returns `FALSE`. Windows ships CP 20932 (JIS X 0208-1990 + 0212-1990) at the native API level, but its repertoire differs from WHATWG's EUC-JP. |
//!
//! **ISO-8859-16** (Latin-10 / South-Eastern European, CP 28606) is also
//! absent from both NLS *and* ICU, but is decoded via a compile-time
//! pure-Rust lookup table -- no runtime dependency required.
//!
//! ## Why not language packs?
//!
//! Installing Windows language packs (Settings → Language) adds UI
//! translations, fonts, and input methods -- **not** NLS code-page conversion
//! tables.  The [`IsValidCodePage`] docs state:
//!
//! > *"Starting with Windows Vista, all code pages that can be installed are
//! > loaded by default."*
//!
//! Because these three code pages do not appear in `EnumSystemCodePages`
//! (neither `CP_INSTALLED` nor `CP_SUPPORTED`), they **cannot** be
//! installed on any Windows edition.  This is a platform-level omission, not
//! a per-machine configuration issue.
//!
//! [spec]: https://encoding.spec.whatwg.org/
//! [Code Page Identifiers]: https://learn.microsoft.com/windows/win32/intl/code-page-identifiers
//! [`IsValidCodePage`]: https://learn.microsoft.com/windows/win32/api/winnls/nf-winnls-isvalidcodepage
//! [`MultiByteToWideChar`]: https://learn.microsoft.com/windows/win32/api/stringapiset/nf-stringapiset-multibytetowidechar
//!
//! Label-to-code-page mapping follows the WHATWG Encoding Standard exactly,
//! including the deliberate aliasing of `ascii` / `iso-8859-1` → windows-1252.
//!
//! Reference: <https://encoding.spec.whatwg.org/>
//! Canonical label list: <https://encoding.spec.whatwg.org/encodings.json>

use crate::{Error, abi, util::string_from_utf16};

// ---------------------------------------------------------------------------
// Public entry point
// ---------------------------------------------------------------------------

/// Decode `data` according to the WHATWG charset `label`.
///
/// * **UTF-8** labels take a fast, pure-Rust path (`String::from_utf8` /
///   `from_utf8_lossy`), never calling into Win32.
/// * Unknown labels silently fall back to UTF-8 (matching reqwest behaviour).
/// * The `replacement` encoding (used by the spec to error-out certain
///   legacy labels) returns `U+FFFD` for any input.
pub(crate) fn decode_body(data: &[u8], charset: &str) -> Result<String, Error> {
    if data.is_empty() {
        return Ok(String::new());
    }

    let label = normalize_label(charset);

    // -- UTF-8 fast path ----------------------------------------------
    if is_utf8_label(&label) {
        trace!(label = charset, "charset: UTF-8 fast path");
        // Strip the UTF-8 BOM (EF BB BF) if present, matching
        // reqwest / encoding_rs behaviour.
        let data = data.strip_prefix(&[0xEF, 0xBB, 0xBF]).unwrap_or(data);
        return Ok(match String::from_utf8(data.to_vec()) {
            Ok(s) => s,
            Err(e) => String::from_utf8_lossy(e.as_bytes()).into_owned(),
        });
    }

    // -- Look up WHATWG label -> Windows code page ---------------------
    let Some(codepage) = whatwg_label_to_codepage(&label) else {
        // Unknown label -> fall back to UTF-8 lossy (matches reqwest)
        warn!(label = charset, "unknown charset label, falling back to UTF-8");
        return Ok(match String::from_utf8(data.to_vec()) {
            Ok(s) => s,
            Err(e) => String::from_utf8_lossy(e.as_bytes()).into_owned(),
        });
    };

    // -- Special decodings --------------------------------------------
    match codepage {
        CP_X_USER_DEFINED => return Ok(decode_x_user_defined(data)),
        CP_REPLACEMENT => return Ok(String::from('\u{FFFD}')),
        CP_UTF16_LE => return decode_utf16le(data),
        CP_UTF16_BE => return decode_utf16be(data),
        CP_ISO_8859_16 => return Ok(decode_iso_8859_16(data)),
        _ => {}
    }

    // -- ICU fallback for NLS-unsupported code pages -------------------
    if let Some(icu_name) = codepage_to_icu_name(codepage) {
        trace!(label = charset, codepage, icu = icu_name, "charset: ICU fallback decode");
        return abi::icu_decode(icu_name, data);
    }

    // -- Win32 MultiByteToWideChar ------------------------------------
    trace!(label = charset, codepage, "charset: Win32 codepage decode");
    abi::multi_byte_to_string(codepage, data)
}

// ---------------------------------------------------------------------------
// Sentinel code-page values (not real Windows CPs)
// ---------------------------------------------------------------------------

/// x-user-defined: 0x00-0x7F as ASCII, 0x80-0xFF -> U+F780-U+F7FF.
const CP_X_USER_DEFINED: u32 = 0;
/// "replacement" encoding: always yields a single U+FFFD.
const CP_REPLACEMENT: u32 = u32::MAX;
/// UTF-16LE: handled in Rust (MultiByteToWideChar does not support CP 1200).
const CP_UTF16_LE: u32 = 1200;
/// UTF-16BE: handled in Rust (MultiByteToWideChar does not support CP 1201).
const CP_UTF16_BE: u32 = 1201;
/// ISO-8859-16: absent from both NLS and ICU; decoded via pure-Rust table.
const CP_ISO_8859_16: u32 = 28606;

/// Returns the ICU converter name for code pages not supported by Win32
/// `MultiByteToWideChar` but available in the system-bundled `icu.dll`.
fn codepage_to_icu_name(codepage: u32) -> Option<&'static str> {
    match codepage {
        28600 => Some("ISO-8859-10"),
        28604 => Some("ISO-8859-14"),
        51932 => Some("EUC-JP"),
        _ => None,
    }
}

// ---------------------------------------------------------------------------
// Label helpers
// ---------------------------------------------------------------------------

/// Normalize a charset label per the WHATWG spec: strip ASCII whitespace,
/// lowercase.
fn normalize_label(label: &str) -> String {
    label
        .trim_matches(|c: char| c.is_ascii_whitespace())
        .to_ascii_lowercase()
}

/// True for labels that the WHATWG spec maps to UTF-8.
fn is_utf8_label(label: &str) -> bool {
    matches!(
        label,
        "utf-8"
            | "utf8"
            | "unicode-1-1-utf-8"
            | "unicode11utf8"
            | "unicode20utf8"
            | "x-unicode20utf8"
    )
}

// ---------------------------------------------------------------------------
// WHATWG label -> Windows code page
// ---------------------------------------------------------------------------

/// Maps every WHATWG Encoding Standard label to a Windows code page.
///
/// Labels are expected to be already ASCII-lowercased and trimmed.
/// Returns `None` for unrecognised labels.
fn whatwg_label_to_codepage(label: &str) -> Option<u32> {
    Some(match label {
        // -- UTF-8 ----------------------------------------------------
        "unicode-1-1-utf-8" | "unicode11utf8" | "unicode20utf8" | "utf-8" | "utf8"
        | "x-unicode20utf8" => 65001,

        // -- Legacy single-byte encodings -----------------------------

        // IBM866
        "866" | "cp866" | "csibm866" | "ibm866" => 866,

        // ISO-8859-2
        "csisolatin2" | "iso-8859-2" | "iso-ir-101" | "iso8859-2" | "iso88592" | "iso_8859-2"
        | "iso_8859-2:1987" | "l2" | "latin2" => 28592,

        // ISO-8859-3
        "csisolatin3" | "iso-8859-3" | "iso-ir-109" | "iso8859-3" | "iso88593" | "iso_8859-3"
        | "iso_8859-3:1988" | "l3" | "latin3" => 28593,

        // ISO-8859-4
        "csisolatin4" | "iso-8859-4" | "iso-ir-110" | "iso8859-4" | "iso88594" | "iso_8859-4"
        | "iso_8859-4:1988" | "l4" | "latin4" => 28594,

        // ISO-8859-5
        "csisolatincyrillic" | "cyrillic" | "iso-8859-5" | "iso-ir-144" | "iso8859-5"
        | "iso88595" | "iso_8859-5" | "iso_8859-5:1988" => 28595,

        // ISO-8859-6
        "arabic" | "asmo-708" | "csiso88596e" | "csiso88596i" | "csisolatinarabic" | "ecma-114"
        | "iso-8859-6" | "iso-8859-6-e" | "iso-8859-6-i" | "iso-ir-127" | "iso8859-6"
        | "iso88596" | "iso_8859-6" | "iso_8859-6:1987" => 28596,

        // ISO-8859-7
        "csisolatingreek" | "ecma-118" | "elot_928" | "greek" | "greek8" | "iso-8859-7"
        | "iso-ir-126" | "iso8859-7" | "iso88597" | "iso_8859-7" | "iso_8859-7:1987"
        | "sun_eu_greek" => 28597,

        // ISO-8859-8
        "csiso88598e" | "csisolatinhebrew" | "hebrew" | "iso-8859-8" | "iso-8859-8-e"
        | "iso-ir-138" | "iso8859-8" | "iso88598" | "iso_8859-8" | "iso_8859-8:1988" | "visual" => {
            28598
        }

        // ISO-8859-8-I
        "csiso88598i" | "iso-8859-8-i" | "logical" => 38598,

        // ISO-8859-10
        "csisolatin6" | "iso-8859-10" | "iso-ir-157" | "iso8859-10" | "iso885910" | "l6"
        | "latin6" => 28600,

        // ISO-8859-13
        "iso-8859-13" | "iso8859-13" | "iso885913" => 28603,

        // ISO-8859-14
        "iso-8859-14" | "iso8859-14" | "iso885914" => 28604,

        // ISO-8859-15
        "csisolatin9" | "iso-8859-15" | "iso8859-15" | "iso885915" | "iso_8859-15" | "l9" => 28605,

        // ISO-8859-16
        "iso-8859-16" => 28606,

        // KOI8-R
        "cskoi8r" | "koi" | "koi8" | "koi8-r" | "koi8_r" => 20866,

        // KOI8-U
        "koi8-ru" | "koi8-u" => 21866,

        // macintosh
        "csmacintosh" | "mac" | "macintosh" | "x-mac-roman" => 10000,

        // windows-874 (also TIS-620, ISO-8859-11)
        "dos-874" | "iso-8859-11" | "iso8859-11" | "iso885911" | "tis-620" | "windows-874" => 874,

        // windows-1250
        "cp1250" | "windows-1250" | "x-cp1250" => 1250,

        // windows-1251
        "cp1251" | "windows-1251" | "x-cp1251" => 1251,

        // windows-1252 -- WHATWG maps ascii / iso-8859-1 / latin1 here
        "ansi_x3.4-1968" | "ascii" | "cp1252" | "cp819" | "csisolatin1" | "ibm819"
        | "iso-8859-1" | "iso-ir-100" | "iso8859-1" | "iso88591" | "iso_8859-1"
        | "iso_8859-1:1987" | "l1" | "latin1" | "us-ascii" | "windows-1252" | "x-cp1252" => 1252,

        // windows-1253
        "cp1253" | "windows-1253" | "x-cp1253" => 1253,

        // windows-1254 (also ISO-8859-9)
        "cp1254" | "csisolatin5" | "iso-8859-9" | "iso-ir-148" | "iso8859-9" | "iso88599"
        | "iso_8859-9" | "iso_8859-9:1989" | "l5" | "latin5" | "windows-1254" | "x-cp1254" => 1254,

        // windows-1255
        "cp1255" | "windows-1255" | "x-cp1255" => 1255,

        // windows-1256
        "cp1256" | "windows-1256" | "x-cp1256" => 1256,

        // windows-1257
        "cp1257" | "windows-1257" | "x-cp1257" => 1257,

        // windows-1258
        "cp1258" | "windows-1258" | "x-cp1258" => 1258,

        // x-mac-cyrillic
        "x-mac-cyrillic" | "x-mac-ukrainian" => 10017,

        // -- Legacy multi-byte CJK ------------------------------------

        // GBK (includes gb2312)
        "chinese" | "csgb2312" | "csiso58gb231280" | "gb2312" | "gb_2312" | "gb_2312-80"
        | "gbk" | "iso-ir-58" | "x-gbk" => 936,

        // gb18030
        "gb18030" => 54936,

        // Big5
        "big5" | "big5-hkscs" | "cn-big5" | "csbig5" | "x-x-big5" => 950,

        // EUC-JP
        "cseucpkdfmtjapanese" | "euc-jp" | "x-euc-jp" => 51932,

        // ISO-2022-JP
        "csiso2022jp" | "iso-2022-jp" => 50220,

        // Shift_JIS
        "csshiftjis" | "ms932" | "ms_kanji" | "shift-jis" | "shift_jis" | "sjis"
        | "windows-31j" | "x-sjis" => 932,

        // EUC-KR
        "cseuckr" | "csksc56011987" | "euc-kr" | "iso-ir-149" | "korean" | "ks_c_5601-1987"
        | "ks_c_5601-1989" | "ksc5601" | "ksc_5601" | "windows-949" => 51949,

        // -- UTF-16 --------------------------------------------------
        "unicodefffe" | "utf-16be" => CP_UTF16_BE,
        "csunicode" | "iso-10646-ucs-2" | "ucs-2" | "unicode" | "unicodefeff" | "utf-16"
        | "utf-16le" => CP_UTF16_LE,

        // -- x-user-defined -------------------------------------------
        "x-user-defined" => CP_X_USER_DEFINED,

        // -- replacement (WHATWG: always produces U+FFFD) -------------
        "csiso2022kr" | "hz-gb-2312" | "iso-2022-cn" | "iso-2022-cn-ext" | "iso-2022-kr"
        | "replacement" => CP_REPLACEMENT,

        _ => return None,
    })
}

// ---------------------------------------------------------------------------
// Special decoders
// ---------------------------------------------------------------------------

/// x-user-defined: 0x00-0x7F -> identity, 0x80-0xFF -> U+F780-U+F7FF.
fn decode_x_user_defined(data: &[u8]) -> String {
    data.iter()
        .map(|&b| {
            if b < 0x80 {
                b as char
            } else {
                // SAFETY: 0xF780..=0xF7FF are valid Unicode code points (PUA).
                char::from_u32(0xF780 + u32::from(b) - 0x80).unwrap_or('\u{FFFD}')
            }
        })
        .collect()
}

/// WHATWG ISO-8859-16 (Latin-10) upper-half decode table.
///
/// Bytes 0x00..0x7F are ASCII (decoded as identity, not in this table).
/// Bytes 0x80..0x9F are C1 controls (identity).  The remaining 96
/// positions share most code points with ISO-8859-1 but override 40 of
/// them with Romanian, Polish, and other Central/Southeast-European
/// characters plus a handful of typographic symbols.
///
/// Source: <https://encoding.spec.whatwg.org/index-iso-8859-16.txt>
#[rustfmt::skip]
static ISO_8859_16_HIGH: [u16; 128] = [
    // 0x80..0x8F: C1 controls (identity)
    0x0080, 0x0081, 0x0082, 0x0083, 0x0084, 0x0085, 0x0086, 0x0087,
    0x0088, 0x0089, 0x008A, 0x008B, 0x008C, 0x008D, 0x008E, 0x008F,
    // 0x90..0x9F: C1 controls (identity)
    0x0090, 0x0091, 0x0092, 0x0093, 0x0094, 0x0095, 0x0096, 0x0097,
    0x0098, 0x0099, 0x009A, 0x009B, 0x009C, 0x009D, 0x009E, 0x009F,
    // 0xA0      0xA1   0xA2   0xA3   0xA4   0xA5   0xA6   0xA7
    0x00A0, 0x0104, 0x0105, 0x0141, 0x20AC, 0x201E, 0x0160, 0x00A7,
    // 0xA8      0xA9   0xAA   0xAB   0xAC   0xAD   0xAE   0xAF
    0x0161, 0x00A9, 0x0218, 0x00AB, 0x0179, 0x00AD, 0x017A, 0x017B,
    // 0xB0      0xB1   0xB2   0xB3   0xB4   0xB5   0xB6   0xB7
    0x00B0, 0x00B1, 0x010C, 0x0142, 0x017D, 0x201D, 0x00B6, 0x00B7,
    // 0xB8      0xB9   0xBA   0xBB   0xBC   0xBD   0xBE   0xBF
    0x017E, 0x010D, 0x0219, 0x00BB, 0x0152, 0x0153, 0x0178, 0x017C,
    // 0xC0      0xC1   0xC2   0xC3   0xC4   0xC5   0xC6   0xC7
    0x00C0, 0x00C1, 0x00C2, 0x0102, 0x00C4, 0x0106, 0x00C6, 0x00C7,
    // 0xC8      0xC9   0xCA   0xCB   0xCC   0xCD   0xCE   0xCF
    0x00C8, 0x00C9, 0x00CA, 0x00CB, 0x00CC, 0x00CD, 0x00CE, 0x00CF,
    // 0xD0      0xD1   0xD2   0xD3   0xD4   0xD5   0xD6   0xD7
    0x0110, 0x0143, 0x00D2, 0x00D3, 0x00D4, 0x0150, 0x00D6, 0x015A,
    // 0xD8      0xD9   0xDA   0xDB   0xDC   0xDD   0xDE   0xDF
    0x0170, 0x00D9, 0x00DA, 0x00DB, 0x00DC, 0x0118, 0x021A, 0x00DF,
    // 0xE0      0xE1   0xE2   0xE3   0xE4   0xE5   0xE6   0xE7
    0x00E0, 0x00E1, 0x00E2, 0x0103, 0x00E4, 0x0107, 0x00E6, 0x00E7,
    // 0xE8      0xE9   0xEA   0xEB   0xEC   0xED   0xEE   0xEF
    0x00E8, 0x00E9, 0x00EA, 0x00EB, 0x00EC, 0x00ED, 0x00EE, 0x00EF,
    // 0xF0      0xF1   0xF2   0xF3   0xF4   0xF5   0xF6   0xF7
    0x0111, 0x0144, 0x00F2, 0x00F3, 0x00F4, 0x0151, 0x00F6, 0x015B,
    // 0xF8      0xF9   0xFA   0xFB   0xFC   0xFD   0xFE   0xFF
    0x0171, 0x00F9, 0x00FA, 0x00FB, 0x00FC, 0x0119, 0x021B, 0x00FF,
];

/// Decode ISO-8859-16 via compile-time lookup table (no NLS / ICU needed).
fn decode_iso_8859_16(data: &[u8]) -> String {
    data.iter()
        .map(|&b| {
            if b < 0x80 {
                b as char
            } else {
                // Every entry in the table is a valid Unicode scalar value.
                char::from_u32(u32::from(ISO_8859_16_HIGH[(b - 0x80) as usize]))
                    .unwrap_or('\u{FFFD}')
            }
        })
        .collect()
}

/// Decode a UTF-16LE byte stream.
fn decode_utf16le(data: &[u8]) -> Result<String, Error> {
    // Strip BOM if present
    let data = data.strip_prefix(&[0xFF, 0xFE]).unwrap_or(data);
    if !data.len().is_multiple_of(2) {
        return Err(Error::decode("invalid UTF-16LE: odd byte count"));
    }
    let words: Vec<u16> = data
        .chunks_exact(2)
        .map(|c| u16::from_le_bytes([c[0], c[1]]))
        .collect();
    string_from_utf16(&words, "invalid UTF-16LE")
}

/// Decode a UTF-16BE byte stream.
fn decode_utf16be(data: &[u8]) -> Result<String, Error> {
    let data = data.strip_prefix(&[0xFE, 0xFF]).unwrap_or(data);
    if !data.len().is_multiple_of(2) {
        return Err(Error::decode("invalid UTF-16BE: odd byte count"));
    }
    let words: Vec<u16> = data
        .chunks_exact(2)
        .map(|c| u16::from_be_bytes([c[0], c[1]]))
        .collect();
    string_from_utf16(&words, "invalid UTF-16BE")
}

// ---------------------------------------------------------------------------
// Content-Type charset extraction
// ---------------------------------------------------------------------------

/// Extract the `charset` parameter from a `Content-Type` header value.
///
/// ```text
/// text/html; charset=utf-8       -> Some("utf-8")
/// application/json               -> None
/// text/html; charset="UTF-8"     -> Some("UTF-8")
/// ```
pub(crate) fn extract_charset_from_content_type(headers: &http::HeaderMap) -> Option<String> {
    let ct = headers.get(http::header::CONTENT_TYPE)?;
    let ct_str = ct.to_str().ok()?;
    // Split on ';' and search each parameter for `charset=…`.
    // Skipping the media-type segment avoids false positives from
    // substrings embedded in unrelated parameter values
    // (e.g. `x=charset=wrong; charset=right`).
    for param in ct_str.split(';').skip(1) {
        let trimmed = param.trim();
        let Some((key, value)) = trimmed.split_once('=') else {
            continue;
        };
        if !key.trim().eq_ignore_ascii_case("charset") {
            continue;
        }
        let value = value.trim_start_matches('"');
        let charset: String = value
            .chars()
            .take_while(|&c| c != '"' && c != ';' && !c.is_ascii_whitespace())
            .collect();
        if charset.is_empty() {
            return None;
        }
        return Some(charset);
    }
    None
}

// ---------------------------------------------------------------------------
// Tests
// ---------------------------------------------------------------------------

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

    // -- All 39 WHATWG encodings --------------------------------------

    /// Comprehensive test covering all 39 WHATWG-mandated encodings with
    /// non-trivial, non-ASCII text specific to each encoding.
    ///
    /// Byte sequences are derived from the WHATWG Encoding Standard
    /// single-byte indexes (<https://encoding.spec.whatwg.org/#indexes>)
    /// and verified against the authoritative index text files.
    #[test]
    fn all_39_whatwg_encodings() {
        // (name, label, data, expected)
        let cases: &[(&str, &str, &[u8], &str)] = &[
            // -- Pure-Rust paths ---------------------------------
            ("UTF-8", "utf-8", &[0x63, 0x61, 0x66, 0xC3, 0xA9], "caf\u{00E9}"),
            ("UTF-16BE", "utf-16be", &[0x00, 0x48, 0x00, 0x69], "Hi"),
            ("UTF-16LE", "utf-16le", &[0x48, 0x00, 0x69, 0x00], "Hi"),
            ("x-user-defined", "x-user-defined", &[0x48, 0x80, 0xFF], "H\u{F780}\u{F7FF}"),
            // -- NLS (MultiByteToWideChar) --------------------------
            (
                "IBM866",
                "ibm866",
                &[0x8F, 0xE0, 0xA8, 0xA2, 0xA5, 0xE2],
                "\u{041F}\u{0440}\u{0438}\u{0432}\u{0435}\u{0442}",
            ),
            ("ISO-8859-2", "iso-8859-2", &[0xA3, 0xF3, 0x64, 0xBC], "\u{0141}\u{00F3}d\u{017A}"),
            ("ISO-8859-3", "iso-8859-3", &[0xA1, 0xB1], "\u{0126}\u{0127}"),
            ("ISO-8859-4", "iso-8859-4", &[0xBD, 0xBF], "\u{014A}\u{014B}"),
            ("ISO-8859-5", "iso-8859-5", &[0xBC, 0xD8, 0xE0], "\u{041C}\u{0438}\u{0440}"),
            ("ISO-8859-6", "iso-8859-6", &[0xC7, 0xC8], "\u{0627}\u{0628}"),
            ("ISO-8859-7", "iso-8859-7", &[0xC1, 0xF9], "\u{0391}\u{03C9}"),
            ("ISO-8859-8", "iso-8859-8", &[0xE0, 0xE1], "\u{05D0}\u{05D1}"),
            ("ISO-8859-8-I", "iso-8859-8-i", &[0xF9, 0xFA], "\u{05E9}\u{05EA}"),
            ("ISO-8859-13", "iso-8859-13", &[0xD9, 0xF9], "\u{0141}\u{0142}"),
            ("ISO-8859-15", "iso-8859-15", &[0xA4, 0xBC], "\u{20AC}\u{0152}"),
            ("KOI8-R", "koi8-r", &[0xF2, 0xD5, 0xD3, 0xD8], "\u{0420}\u{0443}\u{0441}\u{044C}"),
            ("KOI8-U", "koi8-u", &[0xB4, 0xA4], "\u{0404}\u{0454}"),
            (
                "macintosh",
                "macintosh",
                &[0xC7, 0xC8, 0xD2, 0xD3],
                "\u{00AB}\u{00BB}\u{201C}\u{201D}",
            ),
            ("windows-874", "windows-874", &[0xA1, 0xA2], "\u{0E01}\u{0E02}"),
            ("windows-1250", "windows-1250", &[0x8A, 0x9A], "\u{0160}\u{0161}"),
            (
                "windows-1251",
                "windows-1251",
                &[0xCC, 0xEE, 0xF1, 0xEA, 0xE2, 0xE0],
                "\u{041C}\u{043E}\u{0441}\u{043A}\u{0432}\u{0430}",
            ),
            ("windows-1252", "windows-1252", &[0x80, 0x93, 0x94], "\u{20AC}\u{201C}\u{201D}"),
            ("windows-1253", "windows-1253", &[0xC1, 0xF9], "\u{0391}\u{03C9}"),
            ("windows-1254", "windows-1254", &[0xD0, 0xF0], "\u{011E}\u{011F}"),
            ("windows-1255", "windows-1255", &[0x80, 0xE0], "\u{20AC}\u{05D0}"),
            ("windows-1256", "windows-1256", &[0xC7, 0xC8], "\u{0627}\u{0628}"),
            ("windows-1257", "windows-1257", &[0xC0, 0xE0], "\u{0104}\u{0105}"),
            ("windows-1258", "windows-1258", &[0xD0, 0xF0], "\u{0110}\u{0111}"),
            ("x-mac-cyrillic", "x-mac-cyrillic", &[0x80, 0x81], "\u{0410}\u{0411}"),
            ("GBK", "gbk", &[0xC4, 0xE3, 0xBA, 0xC3], "\u{4F60}\u{597D}"),
            ("gb18030", "gb18030", &[0xC4, 0xE3, 0xBA, 0xC3], "\u{4F60}\u{597D}"),
            ("Big5", "big5", &[0xA4, 0xA4], "\u{4E2D}"),
            ("ISO-2022-JP", "iso-2022-jp", &[0x1B, 0x24, 0x42, 0x46, 0x7C], "\u{65E5}"),
            ("Shift_JIS", "shift_jis", &[0x82, 0xB1], "\u{3053}"),
            ("EUC-KR", "euc-kr", &[0xC7, 0xD1], "\u{D55C}"),
            // -- ICU fallback (Windows 10 1903+) -----------------
            ("ISO-8859-10", "iso-8859-10", &[0xA1, 0xA2], "\u{0104}\u{0112}"),
            ("ISO-8859-14", "iso-8859-14", &[0xA1, 0xD0], "\u{1E02}\u{0174}"),
            ("EUC-JP", "euc-jp", &[0xC6, 0xFC, 0xCB, 0xDC, 0xB8, 0xEC], "\u{65E5}\u{672C}\u{8A9E}"),
            // -- Compile-time lookup table --------------------------
            ("ISO-8859-16", "iso-8859-16", &[0xAA, 0xBA], "\u{0218}\u{0219}"),
        ];

        assert_eq!(cases.len(), 39, "must cover all 39 WHATWG encodings");
        for &(name, label, data, expected) in cases {
            let result =
                decode_body(data, label).unwrap_or_else(|e| panic!("{name} ({label}): {e}"));
            assert_eq!(result, expected, "{name} ({label})");
        }
    }

    // -- decode_body edge cases --------------------------------------

    #[test]
    fn decode_body_table() {
        let cases: &[(&[u8], &str, &str, &str)] = &[
            // (data, label, expected, description)

            // UTF-8 label variants
            (b"ok", "utf8", "ok", "utf8 alias"),
            (b"ok", "UTF-8", "ok", "uppercase label"),
            (b"ok", "  utf-8  ", "ok", "whitespace-trimmed label"),
            (b"ok", "unicode-1-1-utf-8", "ok", "utf-8 alias long"),
            (b"ok", "x-unicode20utf8", "ok", "utf-8 alias x-"),
            // UTF-8 BOM handling
            (&[0xEF, 0xBB, 0xBF, b'h', b'i'], "utf-8", "hi", "UTF-8 BOM stripped"),
            (&[0xEF, 0xBB, 0xBF], "utf-8", "", "UTF-8 BOM only"),
            // UTF-8 lossy fallback for invalid bytes
            (b"hi\xFFlo", "utf-8", "hi\u{FFFD}lo", "UTF-8 invalid byte → U+FFFD"),
            // Empty data always returns empty string
            (b"", "windows-1252", "", "empty data"),
            (b"", "utf-8", "", "empty UTF-8"),
            // Unknown label → UTF-8 fallback (matches reqwest)
            (b"hello", "totally-bogus", "hello", "unknown label → UTF-8"),
            // Unknown label with invalid UTF-8 → lossy fallback
            (b"hi\xFFlo", "totally-bogus", "hi\u{FFFD}lo", "unknown label → UTF-8 lossy"),
            // WHATWG aliasing: ascii / iso-8859-1 → windows-1252
            // 0x80 = € (U+20AC) in windows-1252
            (&[0x80], "ascii", "\u{20AC}", "ascii → windows-1252"),
            (&[0x80], "iso-8859-1", "\u{20AC}", "iso-8859-1 → windows-1252"),
            (&[0x80], "latin1", "\u{20AC}", "latin1 → windows-1252"),
            (&[0x80], "us-ascii", "\u{20AC}", "us-ascii → windows-1252"),
            // Case-insensitive label lookup
            (&[0x93], "WINDOWS-1252", "\u{201C}", "case: UPPER"),
            (&[0x93], "Windows-1252", "\u{201C}", "case: Mixed"),
            // replacement encoding: always yields U+FFFD
            (b"anything", "replacement", "\u{FFFD}", "replacement encoding"),
            (b"anything", "hz-gb-2312", "\u{FFFD}", "hz-gb-2312 → replacement"),
            (b"anything", "iso-2022-kr", "\u{FFFD}", "iso-2022-kr → replacement"),
            // x-user-defined: 0x80→U+F780, 0xFF→U+F7FF
            (&[0x80, 0xFF], "x-user-defined", "\u{F780}\u{F7FF}", "x-user-defined high bytes"),
            // ICU alias labels
            (&[0xA1], "latin6", "\u{0104}", "latin6 → ISO-8859-10"),
            (&[0xA4, 0xA2], "cseucpkdfmtjapanese", "\u{3042}", "cseucpkdfmtjapanese → EUC-JP"),
        ];

        for &(data, label, expected, desc) in cases {
            let result = decode_body(data, label).unwrap_or_else(|e| panic!("{desc}: {e}"));
            assert_eq!(result, expected, "{desc}");
        }
    }

    // -- UTF-16 edge cases ---------------------------------------------

    #[test]
    fn utf16_decode_table() {
        let cases: &[(&str, &[u8], &str, &str)] = &[
            // (charset, data, expected, description)

            // Basic
            ("utf-16le", &[0x41, 0x00, 0x42, 0x00], "AB", "LE basic"),
            ("utf-16be", &[0x00, 0x41, 0x00, 0x42], "AB", "BE basic"),
            // BOM stripping
            ("utf-16le", &[0xFF, 0xFE, 0x41, 0x00], "A", "LE BOM stripped"),
            ("utf-16be", &[0xFE, 0xFF, 0x00, 0x41], "A", "BE BOM stripped"),
        ];

        for &(charset, data, expected, desc) in cases {
            let result = decode_body(data, charset).unwrap_or_else(|e| panic!("{desc}: {e}"));
            assert_eq!(result, expected, "{desc}");
        }
    }

    #[test]
    fn utf16_errors_table() {
        // Lone surrogates → error (from_utf16 rejects them).
        // Odd byte count → error (structurally invalid UTF-16).
        let cases: &[(&str, &[u8], &str)] = &[
            ("utf-16le", &[0x00, 0xD8], "LE lone high surrogate"),
            ("utf-16be", &[0xD8, 0x00], "BE lone high surrogate"),
            ("utf-16le", &[0x41, 0x00, 0x42], "LE odd byte count"),
            ("utf-16be", &[0x00, 0x41, 0x42], "BE odd byte count"),
            // BOM stripping leaves odd remainder → still invalid.
            ("utf-16le", &[0xFF, 0xFE, 0x42], "LE BOM + odd remainder"),
            ("utf-16be", &[0xFE, 0xFF, 0x00], "BE BOM + odd remainder"),
        ];

        for &(charset, data, desc) in cases {
            assert!(decode_body(data, charset).is_err(), "{desc}: should fail");
        }
    }

    // -- ISO-8859-16 lookup table --------------------------------------

    /// Spot-check non-identity mappings against the WHATWG index:
    /// <https://encoding.spec.whatwg.org/index-iso-8859-16.txt>
    #[test]
    fn iso_8859_16_spot_checks() {
        let cases: &[(u8, char)] = &[
            (0xA1, '\u{0104}'), // Ą
            (0xA2, '\u{0105}'), // ą
            (0xA3, '\u{0141}'), // Ł
            (0xA4, '\u{20AC}'), // €
            (0xA5, '\u{201E}'), // „
            (0xA6, '\u{0160}'), // Š
            (0xA8, '\u{0161}'), // š
            (0xAC, '\u{0179}'), // Ź
            (0xAF, '\u{017B}'), // Ż
            (0xB2, '\u{010C}'), // Č
            (0xB3, '\u{0142}'), // ł
            (0xB4, '\u{017D}'), // Ž
            (0xB5, '\u{201D}'), // "
            (0xB8, '\u{017E}'), // ž
            (0xB9, '\u{010D}'), // č
            (0xBC, '\u{0152}'), // Œ
            (0xBD, '\u{0153}'), // œ
            (0xBE, '\u{0178}'), // Ÿ
            (0xBF, '\u{017C}'), // ż
            (0xC3, '\u{0102}'), // Ă
            (0xC5, '\u{0106}'), // Ć
            (0xD0, '\u{0110}'), // Đ
            (0xD1, '\u{0143}'), // Ń
            (0xD5, '\u{0150}'), // Ő
            (0xD7, '\u{015A}'), // Ś
            (0xD8, '\u{0170}'), // Ű
            (0xDD, '\u{0118}'), // Ę
            (0xE3, '\u{0103}'), // ă
            (0xE5, '\u{0107}'), // ć
            (0xF0, '\u{0111}'), // đ
            (0xF1, '\u{0144}'), // ń
            (0xF5, '\u{0151}'), // ő
            (0xF7, '\u{015B}'), // ś
            (0xF8, '\u{0171}'), // ű
            (0xFD, '\u{0119}'), // ę
            (0xFE, '\u{021B}'), // ț
        ];
        for &(byte, expected) in cases {
            let result =
                decode_body(&[byte], "iso-8859-16").unwrap_or_else(|e| panic!("0x{byte:02X}: {e}"));
            assert_eq!(result.chars().next().unwrap(), expected, "byte 0x{byte:02X}");
        }
    }

    /// Every byte 0x00..=0xFF decodes without panic and produces exactly
    /// one character.
    #[test]
    fn iso_8859_16_full_range() {
        let all_bytes: Vec<u8> = (0u8..=255).collect();
        let result = decode_body(&all_bytes, "iso-8859-16").expect("full range");
        assert_eq!(result.chars().count(), 256);
    }

    // -- Content-Type charset extraction -----------------------------

    #[test]
    fn extract_charset_table() {
        let cases: &[(&str, Option<&str>, &str)] = &[
            ("text/html; charset=utf-8", Some("utf-8"), "plain charset"),
            ("text/html; charset=\"UTF-8\"", Some("UTF-8"), "quoted charset"),
            ("application/json", None, "no charset param"),
            // Empty charset= value → None
            ("text/html; charset=", None, "empty charset value"),
            // Empty quoted charset= value → None
            ("text/html; charset=\"\"", None, "empty quoted charset value"),
            // Substring "charset=" inside another parameter's
            // value must not be treated as the charset parameter.
            (
                "text/html; x=charset=wrong; charset=right",
                Some("right"),
                "charset substring in other param value",
            ),
            // charset= as part of a longer parameter name must not match
            ("text/html; notcharset=oops", None, "charset prefix in param name"),
            // Case-insensitive parameter name
            ("text/html; Charset=Latin1", Some("Latin1"), "uppercase Charset"),
            ("text/html; CHARSET=Big5", Some("Big5"), "all-caps CHARSET"),
            // Parameter without '=' is skipped (covers the `else { continue }` branch).
            ("text/html; boundary; charset=UTF-8", Some("UTF-8"), "param without equals sign"),
        ];

        for &(content_type, expected, desc) in cases {
            let mut headers = http::HeaderMap::new();
            headers.insert(http::header::CONTENT_TYPE, content_type.parse().expect("valid"));
            assert_eq!(extract_charset_from_content_type(&headers).as_deref(), expected, "{desc}");
        }
    }

    #[test]
    fn extract_charset_no_content_type() {
        let headers = http::HeaderMap::new();
        assert_eq!(extract_charset_from_content_type(&headers), None);
    }
}