md-codec 0.35.0

Reference implementation of the Mnemonic Descriptor (MD) format for engravable BIP 388 wallet policy backups
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
//! Integration cells for `decode_with_correction` per plan §4.B.1.
//!
//! Covers the 6 cases:
//!  1. `zero_error_passthrough` — clean md1, no corrections.
//!  2. `one_error_at_position_0` — corrupt 1 char at position 0.
//!  3. `one_error_at_last_data_symbol` — corrupt 1 char at the last
//!     data-part position (just before the 13-symbol BCH checksum).
//!  4. `four_error_t_boundary` — BCH t=4 boundary.
//!  5. `five_error_too_many` — exceeds capacity → `TooManyErrors`.
//!  6. `multi_chunk_one_corrupted` — 1 chunk of a 3-chunk set corrupted.
//!
//! v0.35.0 (plan §2.D.1) adds 5 cells exercising single-string non-chunked
//! md1 auto-dispatch in `decode_with_correction` per SPEC v0.30 §2.3:
//!  7. `non_chunked_zero_error_passthrough` — valid non-chunked md1
//!     decodes successfully.
//!  8. `non_chunked_one_to_four_errors_corrected` — 1..=4 errors correct.
//!  9. `non_chunked_five_errors_too_many` — 5+ errors → `TooManyErrors`.
//! 10. `non_chunked_chunked_flag_corruption_yields_chunk_set_incomplete` —
//!     post-correction chunked-flag bit set with only 1 string supplied.
//! 11. `non_chunked_round_trip_parity_via_encode_md1_string` — full
//!     encode_md1_string ↔ decode_with_correction round-trip.

use md_codec::chunk::split;
use md_codec::encode::{Descriptor, encode_md1_string};
use md_codec::error::Error;
use md_codec::origin_path::{OriginPath, PathComponent, PathDecl, PathDeclPaths};
use md_codec::tag::Tag;
use md_codec::tlv::TlvSection;
use md_codec::tree::{Body, Node};
use md_codec::use_site_path::UseSitePath;
use md_codec::{CorrectionDetail, decode_with_correction};

/// Codex32 alphabet, mirroring `src/chunk.rs::CODEX32_ALPHABET`. Tests need
/// a known character → 5-bit-value mapping to construct deterministic
/// corruption patterns.
const CODEX32_ALPHABET: &[u8; 32] = b"qpzry9x8gf2tvdw0s3jn54khce6mua7l";

/// Helper: small single-chunk descriptor (bip84 wpkh). Round-trips through
/// `encode_md1_string` to produce one md1 chunk well under the 64-symbol
/// regular-form chunking threshold.
fn small_descriptor() -> Descriptor {
    Descriptor {
        n: 1,
        path_decl: PathDecl {
            n: 1,
            paths: PathDeclPaths::Shared(OriginPath {
                components: vec![
                    PathComponent {
                        hardened: true,
                        value: 84,
                    },
                    PathComponent {
                        hardened: true,
                        value: 0,
                    },
                    PathComponent {
                        hardened: true,
                        value: 0,
                    },
                ],
            }),
        },
        use_site_path: UseSitePath::standard_multipath(),
        tree: Node {
            tag: Tag::Wpkh,
            body: Body::KeyArg { index: 0 },
        },
        tlv: TlvSection::new_empty(),
    }
}

/// Helper: 4-cosigner divergent-path wsh sortedmulti template that's large
/// enough to force chunked encoding. Mirrors `tests/chunking.rs`'s
/// `multi_chunk_descriptor` — per-cosigner path body ~180 bits × 4 cosigners
/// ~720 bits, well above the 320-bit single-string limit so chunking is
/// guaranteed.
fn multi_chunk_descriptor() -> Descriptor {
    let mut paths = Vec::new();
    for cosigner in 0..4u32 {
        let mut components = Vec::new();
        for i in 0..15u32 {
            components.push(PathComponent {
                hardened: true,
                value: cosigner * 100 + i + 1,
            });
        }
        paths.push(OriginPath { components });
    }
    Descriptor {
        n: 4,
        path_decl: PathDecl {
            n: 4,
            paths: PathDeclPaths::Divergent(paths),
        },
        use_site_path: UseSitePath::standard_multipath(),
        tree: Node {
            tag: Tag::Wsh,
            body: Body::Children(vec![Node {
                tag: Tag::SortedMulti,
                body: Body::MultiKeys {
                    k: 2,
                    indices: (0..4).collect(),
                },
            }]),
        },
        tlv: TlvSection::new_empty(),
    }
}

/// Flip one character of an md1 chunk at the data-part position `pos`
/// (0-indexed, post-`md1` HRP). The flipped char is `original ^ mask`
/// in the 5-bit codex32 alphabet space.
fn corrupt_chunk_at(chunk: &str, pos: usize, xor_mask: u8) -> String {
    let hrp_len = 3; // "md1"
    let mut chars: Vec<char> = chunk.chars().collect();
    let abs_idx = hrp_len + pos;
    let original_char = chars[abs_idx];
    let original_sym = CODEX32_ALPHABET
        .iter()
        .position(|&b| b == original_char.to_ascii_lowercase() as u8)
        .expect("char in codex32 alphabet") as u8;
    let new_sym = (original_sym ^ (xor_mask & 0x1F)) & 0x1F;
    chars[abs_idx] = CODEX32_ALPHABET[new_sym as usize] as char;
    chars.iter().collect()
}

/// Extract the codex32 data-part (post-HRP) length.
fn data_part_len(chunk: &str) -> usize {
    chunk.len() - 3 // strip "md1"
}

// ---------------------------------------------------------------------------
// Cell 1: zero-error pass-through
// ---------------------------------------------------------------------------

#[test]
fn zero_error_passthrough() {
    let d = small_descriptor();
    let chunks = split(&d).unwrap();
    let refs: Vec<&str> = chunks.iter().map(|s| s.as_str()).collect();
    let (decoded, details) = decode_with_correction(&refs).expect("clean decode");
    assert_eq!(decoded, d, "round-trip preserves descriptor");
    assert!(details.is_empty(), "no corrections expected for clean input");
}

// ---------------------------------------------------------------------------
// Cell 2: 1 error at position 0
// ---------------------------------------------------------------------------

#[test]
fn one_error_at_position_0() {
    let d = small_descriptor();
    let chunks = split(&d).unwrap();
    let bad = corrupt_chunk_at(&chunks[0], 0, 0b10101);
    let (decoded, details) = decode_with_correction(&[bad.as_str()]).expect("1-error decode");
    assert_eq!(decoded, d, "corrected decode matches original");
    assert_eq!(details.len(), 1, "exactly 1 correction reported");
    assert_eq!(details[0].chunk_index, 0);
    assert_eq!(details[0].position, 0);
    // The original char at position 0 should be what we computed.
    let original_char = chunks[0].chars().nth(3).unwrap();
    assert_eq!(
        details[0].now, original_char,
        "correction restores the original char"
    );
    assert_ne!(details[0].was, details[0].now);
}

// ---------------------------------------------------------------------------
// Cell 3: 1 error at the last data-part position
// ---------------------------------------------------------------------------

#[test]
fn one_error_at_last_data_symbol() {
    let d = small_descriptor();
    let chunks = split(&d).unwrap();
    let last_pos = data_part_len(&chunks[0]) - 1;
    let bad = corrupt_chunk_at(&chunks[0], last_pos, 0b01110);
    let (decoded, details) =
        decode_with_correction(&[bad.as_str()]).expect("1-error at last position decodes");
    assert_eq!(decoded, d);
    assert_eq!(details.len(), 1);
    assert_eq!(details[0].position, last_pos);
    let original_char = chunks[0].chars().nth(3 + last_pos).unwrap();
    assert_eq!(details[0].now, original_char);
}

// ---------------------------------------------------------------------------
// Cell 4: 4-error t-boundary
// ---------------------------------------------------------------------------

#[test]
fn four_error_t_boundary() {
    let d = small_descriptor();
    let chunks = split(&d).unwrap();
    let dp_len = data_part_len(&chunks[0]);
    // 4 distinct, well-spaced positions across the data-part.
    let positions: [usize; 4] = [0, dp_len / 4, dp_len / 2, dp_len - 1];
    let masks: [u8; 4] = [0b00001, 0b10000, 0b11111, 0b01010];
    let mut bad = chunks[0].clone();
    for (&p, &m) in positions.iter().zip(&masks) {
        bad = corrupt_chunk_at(&bad, p, m);
    }
    let (decoded, details) =
        decode_with_correction(&[bad.as_str()]).expect("4-error t-boundary decodes");
    assert_eq!(decoded, d, "corrected decode matches original");
    assert_eq!(details.len(), 4, "exactly 4 corrections reported");
    // Positions should be reported in ascending order per decode_regular_errors's sort.
    let reported_positions: Vec<usize> = details.iter().map(|c| c.position).collect();
    let mut expected_positions: Vec<usize> = positions.to_vec();
    expected_positions.sort();
    assert_eq!(reported_positions, expected_positions);
    for det in &details {
        assert_eq!(det.chunk_index, 0);
        assert_ne!(det.was, det.now, "correction changes the character");
    }
}

// ---------------------------------------------------------------------------
// Cell 5: 5 errors — exceeds BCH t = 4 capacity → TooManyErrors
// ---------------------------------------------------------------------------

#[test]
fn five_error_too_many() {
    let d = small_descriptor();
    let chunks = split(&d).unwrap();
    let dp_len = data_part_len(&chunks[0]);
    let positions: [usize; 5] = [0, dp_len / 5, 2 * dp_len / 5, 3 * dp_len / 5, dp_len - 1];
    let masks: [u8; 5] = [0b00001, 0b00010, 0b00100, 0b01000, 0b10000];
    let mut bad = chunks[0].clone();
    for (&p, &m) in positions.iter().zip(&masks) {
        bad = corrupt_chunk_at(&bad, p, m);
    }
    let err = decode_with_correction(&[bad.as_str()])
        .expect_err("5-error pattern must not decode successfully");
    match err {
        Error::TooManyErrors { chunk_index, bound } => {
            assert_eq!(chunk_index, 0, "the only chunk is index 0");
            assert_eq!(bound, 8, "BCH(93,80,8) singleton bound is 8");
        }
        other => panic!("expected TooManyErrors, got {other:?}"),
    }
}

// ---------------------------------------------------------------------------
// Cell 6: multi-chunk set with one chunk corrupted (1 error)
// ---------------------------------------------------------------------------

#[test]
fn multi_chunk_one_corrupted() {
    let d = multi_chunk_descriptor();
    let chunks = split(&d).unwrap();
    assert!(
        chunks.len() >= 2,
        "multi-chunk descriptor must split into 2+ chunks; got {}",
        chunks.len()
    );
    // Pick the middle (or only-non-first) chunk to corrupt.
    let target_idx = chunks.len() / 2;
    let bad_chunk = corrupt_chunk_at(&chunks[target_idx], 4, 0b01101);
    let mut input: Vec<String> = chunks.to_vec();
    input[target_idx] = bad_chunk;
    let refs: Vec<&str> = input.iter().map(|s| s.as_str()).collect();
    let (decoded, details) = decode_with_correction(&refs).expect("multi-chunk decode");
    assert_eq!(decoded, d, "round-trip restores descriptor");
    assert_eq!(details.len(), 1, "exactly 1 correction across the chunk set");
    let det: &CorrectionDetail = &details[0];
    assert_eq!(
        det.chunk_index, target_idx,
        "correction reports the corrupted chunk's index"
    );
    assert_eq!(det.position, 4);
    let original_char = chunks[target_idx].chars().nth(3 + 4).unwrap();
    assert_eq!(det.now, original_char);
}

// ===========================================================================
// v0.35.0 — single-string non-chunked md1 auto-dispatch (plan §2.D.1).
// ===========================================================================

// ---------------------------------------------------------------------------
// Cell 7: non-chunked happy path — valid single-payload md1 decodes cleanly.
// ---------------------------------------------------------------------------

#[test]
fn non_chunked_zero_error_passthrough() {
    let d = small_descriptor();
    // `encode_md1_string` emits single-payload non-chunked form (no chunk
    // header). The first 5-bit symbol's bit 0 (chunked-flag per SPEC v0.30
    // §2.3) is 0 → `decode_with_correction` must auto-dispatch to the
    // non-chunked decode path instead of `reassemble`.
    let s = encode_md1_string(&d).expect("encode_md1_string");
    let (decoded, details) =
        decode_with_correction(&[s.as_str()]).expect("non-chunked decode succeeds");
    assert_eq!(decoded, d, "round-trip preserves descriptor");
    assert!(details.is_empty(), "no corrections expected for clean input");
}

// ---------------------------------------------------------------------------
// Cell 8: non-chunked 1..=4 error correction across the BCH t-boundary.
// ---------------------------------------------------------------------------

#[test]
fn non_chunked_one_to_four_errors_corrected() {
    let d = small_descriptor();
    let s = encode_md1_string(&d).expect("encode_md1_string");
    let dp_len = data_part_len(&s);

    // Walk 1..=4 errors; for each error budget, corrupt that many
    // well-spaced positions, expect a clean BCH correction back to the
    // original descriptor. NOTE: position 0 is excluded so we don't flip
    // bit 0 of the first 5-bit symbol (the chunked-flag); that case is
    // covered by Cell 10 below.
    for error_count in 1..=4usize {
        let positions: Vec<usize> = (0..error_count)
            .map(|i| 1 + ((dp_len - 2) * i) / error_count.max(1))
            .collect();
        let masks: [u8; 4] = [0b00001, 0b10000, 0b11111, 0b01010];
        let mut bad = s.clone();
        for (i, &p) in positions.iter().enumerate() {
            bad = corrupt_chunk_at(&bad, p, masks[i]);
        }
        let (decoded, details) = decode_with_correction(&[bad.as_str()])
            .unwrap_or_else(|e| panic!("{error_count}-error decode must succeed, got {e:?}"));
        assert_eq!(
            decoded, d,
            "{error_count}-error corrected decode matches original"
        );
        assert_eq!(
            details.len(),
            error_count,
            "{error_count}-error correction report length"
        );
        for det in &details {
            assert_eq!(det.chunk_index, 0);
            assert_ne!(det.was, det.now);
        }
    }
}

// ---------------------------------------------------------------------------
// Cell 9: non-chunked 5-error pattern → TooManyErrors (BCH t=4 boundary).
// ---------------------------------------------------------------------------

#[test]
fn non_chunked_five_errors_too_many() {
    let d = small_descriptor();
    let s = encode_md1_string(&d).expect("encode_md1_string");
    let dp_len = data_part_len(&s);
    // 5 distinct, well-spaced positions across the data-part (mirrors
    // Cell 5's chunked-case pattern). Position 0 is included; the
    // intent is to exercise the BCH-correction capacity exhaustion, not
    // the chunked-flag-bit branch.
    let positions: [usize; 5] = [0, dp_len / 5, 2 * dp_len / 5, 3 * dp_len / 5, dp_len - 1];
    let masks: [u8; 5] = [0b00001, 0b00010, 0b00100, 0b01000, 0b10000];
    let mut bad = s.clone();
    for (&p, &m) in positions.iter().zip(&masks) {
        bad = corrupt_chunk_at(&bad, p, m);
    }
    let err = decode_with_correction(&[bad.as_str()])
        .expect_err("5-error pattern must not decode successfully");
    match err {
        Error::TooManyErrors { chunk_index, bound } => {
            assert_eq!(chunk_index, 0, "the only chunk is index 0");
            assert_eq!(bound, 8, "BCH(93,80,8) singleton bound is 8");
        }
        other => panic!("expected TooManyErrors, got {other:?}"),
    }
}

// ---------------------------------------------------------------------------
// Cell 10: chunked-flag-set + only 1 string supplied → ChunkSetIncomplete.
//
// This is the SPEC v0.30 §2.3 ambiguity edge: post-correction, the first
// 5-bit symbol's bit 0 is set (== 1 → chunked) but the caller provided only
// 1 input string. Per plan §2.D.1, surface as the existing
// `ChunkSetIncomplete` variant (no new error kind).
// ---------------------------------------------------------------------------

#[test]
fn non_chunked_chunked_flag_corruption_yields_chunk_set_incomplete() {
    // Take a multi-chunk descriptor's chunked-form chunk[0] and supply it
    // alone — the chunk header's chunked-flag bit is set, but only 1 string
    // is provided, so `decode_with_correction` must reject with
    // `ChunkSetIncomplete`.
    let d = multi_chunk_descriptor();
    let chunks = split(&d).expect("split multi-chunk");
    assert!(
        chunks.len() >= 2,
        "multi-chunk descriptor must split into 2+ chunks; got {}",
        chunks.len()
    );
    let single = &chunks[0];
    let err = decode_with_correction(&[single.as_str()])
        .expect_err("chunked-form with only 1 string must not decode");
    match err {
        Error::ChunkSetIncomplete { got, expected } => {
            assert_eq!(got, 1, "exactly 1 string supplied");
            assert_eq!(
                expected,
                chunks.len(),
                "expected count matches chunk-set size"
            );
        }
        other => panic!("expected ChunkSetIncomplete, got {other:?}"),
    }
}

// ---------------------------------------------------------------------------
// Cell 11: full round-trip parity vs encode_md1_string (zero-error).
// ---------------------------------------------------------------------------

#[test]
fn non_chunked_round_trip_parity_via_encode_md1_string() {
    // Confirms `encode_md1_string` ↔ `decode_with_correction(&[s])` is a
    // bit-identity round-trip for non-chunked-form inputs.
    let d = small_descriptor();
    let s = encode_md1_string(&d).expect("encode_md1_string");
    let (decoded, details) =
        decode_with_correction(&[s.as_str()]).expect("non-chunked decode succeeds");
    assert_eq!(decoded, d, "round-trip preserves descriptor");
    assert!(details.is_empty(), "no corrections for clean round-trip");
    // Sanity: re-encode the decoded form and compare with the original
    // string to catch any subtle decode-side normalization drift.
    let s2 = encode_md1_string(&decoded).expect("re-encode_md1_string");
    assert_eq!(s, s2, "re-encoded string matches the original byte-for-byte");
}