mdictlib 0.2.8

Library-first Rust parser for MDict .mdx and .mdd dictionaries
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
//! The version 1 keyword section.
//!
//! Version 1 carries a 16-byte keyword header of four 32-bit big-endian
//! fields, stores its keyword metadata **raw**, and writes summaries with a
//! one-byte length and no terminator. It has neither the version 2
//! decompressed-size field nor the version 2 keyword-header checksum, so there
//! is nothing to validate the header against before its own geometry is
//! reconciled against the file.
//!
//! Every field is widened at the read site, so no `u32` reaches the section
//! arithmetic below.

use std::mem;
use std::sync::Arc;

use crate::error::{Error, Result};
use crate::format::common::checked::{
    add_u64, checked_usize, ensure_u64_ceiling, ensure_u64_limit, ensure_usize_limit, mul_u64,
};
use crate::format::common::cursor::Cursor;
use crate::format::common::descriptors::{
    DecodedKeyRow, KeyBlockDescriptor, KeyRowContext, SectionRange, WireOperations,
};
use crate::format::common::encoding::TextEncoding;
use crate::format::common::source::FileSource;
use crate::limits::{MemoryBudget, try_reserve_vec};
use crate::types::{Header, Limits, OpenOptions};

/// Bytes in the version 1 keyword header: four `u32` fields.
const KEYWORD_HEADER_LEN: u64 = 16;

/// Smallest possible keyword metadata row: a `u32` entry count, two one-byte
/// summary lengths with empty summaries, and two `u32` block sizes.
const MIN_KEY_INFO_ROW_LEN: u64 = 4 + 1 + 1 + 4 + 4;

/// The version 1 lazy wire operations, selected once during open.
pub(super) const WIRE_OPERATIONS: WireOperations = WireOperations { decode_key_rows };

/// Exact validated ranges for the three keyword subsections.
pub(super) struct KeywordSectionRanges {
    pub(super) header: SectionRange,
    pub(super) index: SectionRange,
    pub(super) blocks: SectionRange,
}

/// The parsed version 1 keyword section.
pub(super) struct KeywordSection {
    pub(super) num_entries: u64,
    pub(super) blocks: Box<[KeyBlockDescriptor]>,
    pub(super) record_section_offset: u64,
    pub(super) retained_bytes: usize,
    pub(super) sections: KeywordSectionRanges,
}

/// Parses and validates the whole version 1 keyword section.
///
/// # Errors
///
/// Returns an error if the file declares encryption, a declared size exceeds a
/// limit, the raw metadata does not describe exactly the declared blocks, the
/// summed entry counts or block sizes disagree with the header, or any block
/// range falls outside the file.
pub(super) fn parse_keyword_section(
    source: &FileSource,
    header: &Header,
    key_encoding: TextEncoding,
    keyword_section_offset: u64,
    options: &OpenOptions,
    memory: &Arc<MemoryBudget>,
) -> Result<KeywordSection> {
    // No authorized version 1 artifact declares encryption, and no framing for
    // it has been established. Refuse rather than guess at a transformation.
    let encryption = header.encryption_mode();
    if encryption.has_keyword_header() || encryption.has_keyword_index() {
        return Err(Error::Unsupported(
            "encrypted MDict version 1 keyword sections",
        ));
    }

    let header_len = checked_usize(KEYWORD_HEADER_LEN, "keyword section header length")?;
    let _header_memory = memory.reserve(header_len, "keyword section header")?;
    let raw_header =
        source.read_exact_at(keyword_section_offset, header_len, "keyword section header")?;

    let mut cursor = Cursor::new(&raw_header);
    let num_blocks = cursor.read_u32_be_widened("keyword num_blocks")?;
    let num_entries = cursor.read_u32_be_widened("keyword num_entries")?;
    let key_info_len = cursor.read_u32_be_widened("keyword index length")?;
    let key_blocks_len = cursor.read_u32_be_widened("keyword blocks length")?;

    ensure_u64_limit(
        "key_index_bytes",
        key_info_len,
        options.limits.key_index_bytes,
    )?;
    let num_blocks =
        validate_keyword_header_counts(num_blocks, num_entries, key_info_len, &options.limits)?;

    let key_info_offset = add_u64(
        keyword_section_offset,
        KEYWORD_HEADER_LEN,
        "keyword index offset overflow",
    )?;
    let key_blocks_offset = add_u64(
        key_info_offset,
        key_info_len,
        "keyword blocks offset overflow",
    )?;
    let record_section_offset = add_u64(
        key_blocks_offset,
        key_blocks_len,
        "record section offset overflow",
    )?;
    source.ensure_range(key_info_offset, key_info_len, "keyword index")?;
    source.ensure_range(key_blocks_offset, key_blocks_len, "keyword block section")?;

    // Version 1 keyword metadata is stored raw: no envelope, no checksum, no
    // decompression step.
    let key_info_len_usize = checked_usize(key_info_len, "keyword index length")?;
    let _raw_memory = memory.reserve(key_info_len_usize, "raw keyword index")?;
    let key_info = source.read_exact_at(key_info_offset, key_info_len_usize, "keyword index")?;

    let decoded_summary_bytes = key_encoding.max_decoded_len(key_info_len_usize)?;
    let retained_bytes = decoded_summary_bytes
        .checked_add(
            num_blocks
                .checked_mul(mem::size_of::<KeyBlockDescriptor>())
                .ok_or(Error::InvalidFormat("keyword metadata size overflow"))?,
        )
        .ok_or(Error::InvalidFormat("keyword metadata size overflow"))?;
    let _metadata_memory = memory.reserve(retained_bytes, "keyword block metadata")?;

    let blocks = parse_keyword_index_entries(
        &key_info,
        num_blocks,
        key_encoding,
        key_blocks_offset,
        &options.limits,
    )?;

    let total_entries = blocks
        .last()
        .map(|block| {
            add_u64(
                block.entry_start_index,
                block.entry_count,
                "keyword entry count overflow",
            )
        })
        .transpose()?
        .unwrap_or(0);
    if total_entries != num_entries {
        return Err(Error::InvalidData(format!(
            "keyword entry count mismatch: header={num_entries}, summed={total_entries}"
        )));
    }

    let total_comp = blocks.iter().try_fold(0u64, |acc, block| {
        add_u64(acc, block.comp_size, "keyword blocks length overflow")
    })?;
    if total_comp != key_blocks_len {
        return Err(Error::InvalidData(format!(
            "keyword blocks length mismatch: header={key_blocks_len}, summed={total_comp}"
        )));
    }

    for block in &blocks {
        source.ensure_range(block.comp_offset, block.comp_size, "keyword block")?;
    }

    Ok(KeywordSection {
        num_entries,
        record_section_offset,
        blocks: blocks.into_boxed_slice(),
        retained_bytes,
        sections: KeywordSectionRanges {
            header: SectionRange::new(keyword_section_offset, KEYWORD_HEADER_LEN),
            index: SectionRange::new(key_info_offset, key_info_len),
            blocks: SectionRange::new(key_blocks_offset, key_blocks_len),
        },
    })
}

fn validate_keyword_header_counts(
    num_blocks: u64,
    num_entries: u64,
    key_info_len: u64,
    limits: &Limits,
) -> Result<usize> {
    if num_blocks > num_entries {
        return Err(Error::InvalidData(format!(
            "keyword block count {num_blocks} exceeds entry count {num_entries}"
        )));
    }
    let minimum_index_len = mul_u64(
        num_blocks,
        MIN_KEY_INFO_ROW_LEN,
        "minimum keyword index length overflow",
    )?;
    if minimum_index_len > key_info_len {
        return Err(Error::InvalidData(format!(
            "keyword index length {key_info_len} cannot contain {num_blocks} blocks"
        )));
    }

    let num_blocks = checked_usize(num_blocks, "keyword block count")?;
    let metadata_bytes = num_blocks
        .checked_mul(mem::size_of::<KeyBlockDescriptor>())
        .ok_or(Error::InvalidFormat("keyword block metadata size overflow"))?;
    ensure_usize_limit(
        "keyword_block_metadata_bytes",
        metadata_bytes,
        limits.block_metadata_bytes,
    )?;
    Ok(num_blocks)
}

fn parse_keyword_index_entries(
    bytes: &[u8],
    num_blocks: usize,
    encoding: TextEncoding,
    key_blocks_offset: u64,
    limits: &Limits,
) -> Result<Vec<KeyBlockDescriptor>> {
    let minimum_len = num_blocks
        .checked_mul(checked_usize(
            MIN_KEY_INFO_ROW_LEN,
            "minimum keyword index row length",
        )?)
        .ok_or(Error::InvalidFormat(
            "minimum keyword index length overflow",
        ))?;
    if minimum_len > bytes.len() {
        return Err(Error::InvalidData(format!(
            "keyword index has {} bytes but {num_blocks} blocks require at least {minimum_len}",
            bytes.len()
        )));
    }

    let mut cursor = Cursor::new(bytes);
    let mut blocks = Vec::new();
    try_reserve_vec(&mut blocks, num_blocks, "keyword block metadata")?;
    let mut comp_offset = key_blocks_offset;
    let mut entry_start_index = 0u64;

    for _ in 0..num_blocks {
        let entry_count = cursor.read_u32_be_widened("keyword block entry count")?;
        let first_key = read_summary(&mut cursor, encoding, "keyword block first key")?;
        let last_key = read_summary(&mut cursor, encoding, "keyword block last key")?;
        let comp_size = cursor.read_u32_be_widened("keyword block compressed size")?;
        let decomp_size = cursor.read_u32_be_widened("keyword block decompressed size")?;

        validate_key_block_sizes(entry_count, comp_size, decomp_size, encoding, limits)?;
        let next_comp_offset = add_u64(comp_offset, comp_size, "keyword block offset overflow")?;
        let next_entry_start_index = add_u64(
            entry_start_index,
            entry_count,
            "keyword entry index overflow",
        )?;

        blocks.push(KeyBlockDescriptor {
            entry_count,
            entry_start_index,
            first_key,
            last_key,
            comp_offset,
            comp_size,
            decomp_size,
        });

        comp_offset = next_comp_offset;
        entry_start_index = next_entry_start_index;
    }

    if !cursor.is_empty() {
        return Err(Error::InvalidData(format!(
            "keyword index has {} trailing bytes",
            cursor.remaining()
        )));
    }

    Ok(blocks)
}

/// Reads one version 1 summary: a one-byte length in encoding units followed
/// by exactly that many units, with no terminator.
fn read_summary(
    cursor: &mut Cursor<'_>,
    encoding: TextEncoding,
    context: &'static str,
) -> Result<String> {
    let units = usize::from(cursor.read_u8(context)?);
    let len = units
        .checked_mul(encoding.unit_size())
        .ok_or(Error::InvalidFormat("keyword summary length overflow"))?;
    let bytes = cursor.read_bytes(len, context)?;
    encoding.decode(bytes, context)
}

fn validate_key_block_sizes(
    entry_count: u64,
    comp_size: u64,
    decomp_size: u64,
    encoding: TextEncoding,
    limits: &Limits,
) -> Result<()> {
    if entry_count == 0 {
        return Err(Error::InvalidData(
            "keyword blocks must contain at least one entry".to_owned(),
        ));
    }
    ensure_u64_ceiling(
        "keyword_block_entries",
        entry_count,
        limits.key_block_entries,
    )?;
    ensure_u64_limit(
        "keyword_block_compressed_bytes",
        comp_size,
        limits.compressed_block_bytes,
    )?;
    ensure_u64_limit(
        "keyword_block_decompressed_bytes",
        decomp_size,
        limits.decompressed_block_bytes,
    )?;
    if comp_size < 8 {
        return Err(Error::InvalidData(format!(
            "keyword block is {comp_size} bytes; at least 8 are required"
        )));
    }

    // A version 1 row is a four-byte record offset plus a terminated key, so
    // the smallest possible entry is four bytes plus one terminator unit.
    let unit_size = u64::try_from(encoding.unit_size())
        .map_err(|_| Error::InvalidFormat("encoding unit size exceeds u64"))?;
    let minimum_entry_bytes = add_u64(4, unit_size, "minimum keyword entry size overflow")?;
    let minimum_decompressed_len = mul_u64(
        entry_count,
        minimum_entry_bytes,
        "minimum keyword block size overflow",
    )?;
    if minimum_decompressed_len > decomp_size {
        return Err(Error::InvalidData(format!(
            "keyword block declares {entry_count} entries in {decomp_size} decompressed bytes"
        )));
    }
    Ok(())
}

/// Decodes one whole decompressed version 1 key block into physical rows.
///
/// Each row is a **four-byte** big-endian record offset — widened here, so no
/// 32-bit value escapes into the shared descriptors — followed by an
/// encoding-terminated key.
///
/// # Errors
///
/// Returns an error if the block is truncated, declares more or fewer entries
/// than its metadata, contains a record offset beyond the decoded record
/// stream, decreases its record offsets, or holds undecodable key text.
fn decode_key_rows(bytes: &[u8], context: &KeyRowContext) -> Result<Vec<DecodedKeyRow>> {
    let mut cursor = Cursor::new(bytes);
    let expected_count = checked_usize(context.expected_entries, "decoded key entry count")?;
    let mut entries = Vec::new();
    try_reserve_vec(&mut entries, expected_count, "decoded key entries")?;
    let mut previous_record_start = None;
    while !cursor.is_empty() {
        if entries.len() == expected_count {
            return Err(Error::InvalidData(format!(
                "key block contains data after its declared {expected_count} entries"
            )));
        }
        let record_start = cursor.read_u32_be_widened("key block record offset")?;
        if record_start > context.total_decoded_record_len {
            return Err(Error::InvalidData(format!(
                "record start {record_start} exceeds total record bytes {}",
                context.total_decoded_record_len
            )));
        }
        if let Some(previous) = previous_record_start
            && previous > record_start
        {
            return Err(Error::InvalidData(format!(
                "record starts decrease inside key block from {previous} to {record_start}"
            )));
        }
        previous_record_start = Some(record_start);
        let start = cursor.offset();
        let (key_bytes, next_offset) =
            context
                .encoding
                .split_terminated(bytes, start, "key block entry key")?;
        let key = context.encoding.decode(key_bytes, "key block entry key")?;
        let key_len = next_offset
            .checked_sub(start)
            .ok_or(Error::InvalidFormat("key block cursor underflow"))?;
        cursor.read_bytes(key_len, "key block entry key bytes")?;
        entries.push(DecodedKeyRow { key, record_start });
    }
    if entries.len() != expected_count {
        return Err(Error::InvalidData(format!(
            "key block entry count mismatch: expected {expected_count}, decoded {}",
            entries.len()
        )));
    }
    Ok(entries)
}

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

    #[test]
    fn rejects_a_block_count_that_cannot_fit_the_raw_metadata() {
        // Fourteen bytes is the smallest possible row, so two blocks need 28.
        let error = validate_keyword_header_counts(2, 2, 27, &Limits::new()).unwrap_err();
        assert!(matches!(error, Error::InvalidData(_)));
        assert!(validate_keyword_header_counts(2, 2, 28, &Limits::new()).is_ok());
    }

    #[test]
    fn rejects_more_blocks_than_entries() {
        let error = validate_keyword_header_counts(3, 2, 4096, &Limits::new()).unwrap_err();
        assert!(matches!(error, Error::InvalidData(_)));
    }

    #[test]
    fn rejects_implausible_key_block_entry_count() {
        // Two entries need at least 2 * (4 + 1) = 10 decompressed bytes.
        let error =
            validate_key_block_sizes(2, 8, 9, TextEncoding::Utf8, &Limits::new()).unwrap_err();
        assert!(matches!(error, Error::InvalidData(_)));
        assert!(validate_key_block_sizes(2, 8, 10, TextEncoding::Utf8, &Limits::new()).is_ok());
    }

    #[test]
    fn rejects_oversized_key_blocks_before_they_can_be_read() {
        let limits = Limits::new();
        let oversized = u64::try_from(limits.compressed_block_bytes).unwrap() + 1;
        let error =
            validate_key_block_sizes(1, oversized, 9, TextEncoding::Utf8, &limits).unwrap_err();
        assert!(matches!(error, Error::LimitExceeded { .. }));
    }

    #[test]
    fn summary_lengths_count_encoding_units() {
        let mut cursor = Cursor::new(&[0x02, b'a', 0x00, b'b', 0x00]);
        let summary = read_summary(&mut cursor, TextEncoding::Utf16Le, "test summary").unwrap();
        assert_eq!(summary, "ab");
        assert!(cursor.is_empty());
    }

    #[test]
    fn empty_summaries_are_representable() {
        let mut cursor = Cursor::new(&[0x00]);
        let summary = read_summary(&mut cursor, TextEncoding::Utf8, "test summary").unwrap();
        assert!(summary.is_empty());
    }

    #[test]
    fn truncated_summaries_are_refused() {
        let mut cursor = Cursor::new(&[0x04, b'a']);
        let error = read_summary(&mut cursor, TextEncoding::Utf8, "test summary").unwrap_err();
        assert!(matches!(error, Error::Truncated { .. }));
    }
}