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
use crate::error::{Error, Result};
use serde::{Deserialize, Serialize};
use serde_bytes::ByteBuf;
use std::{convert::TryFrom, fmt};
/// The compression algorithm identifier for `zstandard`.
pub const ALGORITHM_ZSTD: u8 = 0;
/// Defines the compression types supported by documents & entries. Format when encoded is a single
/// byte, with the lowest two bits indicating the actual compression type. The upper 6 bits are
/// reserved for possible future compression formats. For now, the only allowed compression is
/// zstd, where the upper 6 bits are 0.
#[derive(Clone, Copy, Debug, PartialEq, Eq)]
pub(crate) enum CompressType {
/// No compression
None,
/// Standard Compression
General,
/// Dictionary compression
Dict,
}
impl CompressType {
pub fn type_of(compress: &Compress) -> Self {
match compress {
Compress::None => CompressType::None,
Compress::General { .. } => CompressType::General,
Compress::Dict(_) => CompressType::Dict,
}
}
}
impl From<CompressType> for u8 {
fn from(val: CompressType) -> u8 {
match val {
CompressType::None => 0,
CompressType::General => 1,
CompressType::Dict => 2,
}
}
}
impl TryFrom<u8> for CompressType {
type Error = u8;
fn try_from(val: u8) -> Result<CompressType, u8> {
match val {
0 => Ok(CompressType::None),
1 => Ok(CompressType::General),
2 => Ok(CompressType::Dict),
_ => Err(val),
}
}
}
/// Compression settings for Documents and Entries.
#[derive(Clone, Debug, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
pub enum Compress {
/// Don't compress by default.
None,
/// Compress using the given algorithm identifier and compression level.
General {
/// The algorithm's identifier
algorithm: u8,
/// The compression level
level: u8,
},
/// Compress using the provided dictionary object
Dict(Dictionary),
}
impl Compress {
/// Create a new general Zstd Compression setting.
pub fn new_zstd_general(level: u8) -> Self {
Compress::General {
algorithm: ALGORITHM_ZSTD,
level,
}
}
/// Create a new ZStandard dictionary with the given compression level.
pub fn new_zstd_dict(level: u8, dict: Vec<u8>) -> Self {
Compress::Dict(Dictionary::new_zstd(level, dict))
}
/// Attempt to compress the data. Failure occurs if this shouldn't compress, compression fails,
/// or the result is longer than the original. On failure, the buffer is discarded.
pub(crate) fn compress(&self, mut dest: Vec<u8>, src: &[u8]) -> Result<Vec<u8>, ()> {
match self {
Compress::None => Err(()),
Compress::General { level, .. } => {
let dest_len = dest.len();
let max_len = zstd_safe::compress_bound(src.len());
dest.resize(dest_len + max_len, 0);
match zstd_safe::compress(&mut dest[dest_len..], src, *level as i32) {
Ok(len) if len < src.len() => {
dest.truncate(dest_len + len);
Ok(dest)
}
_ => Err(()),
}
}
Compress::Dict(dict) => {
let dest_len = dest.len();
let max_len = zstd_safe::compress_bound(src.len());
dest.resize(dest_len + max_len, 0u8);
match &dict.0 {
DictionaryPrivate::Unknown { level, .. } => {
match zstd_safe::compress(&mut dest[dest_len..], src, *level as i32) {
Ok(len) if len < src.len() => {
dest.truncate(dest_len + len);
Ok(dest)
}
_ => Err(()),
}
}
DictionaryPrivate::Zstd { cdict, .. } => {
let mut ctx = zstd_safe::CCtx::create();
match ctx.compress_using_cdict(&mut dest[dest_len..], src, cdict) {
Ok(len) if len < src.len() => {
dest.truncate(dest_len + len);
Ok(dest)
}
_ => Err(()),
}
}
}
}
}
}
/// Attempt to decompress the data. Fails if the result in `dest` would be greater than
/// `max_size`, or if decompression fails.
pub(crate) fn decompress(
&self,
mut dest: Vec<u8>,
src: &[u8],
marker: CompressType,
extra_size: usize,
max_size: usize,
) -> Result<Vec<u8>> {
match marker {
CompressType::None => {
if dest.len() + src.len() + extra_size > max_size {
Err(Error::FailDecompress(format!(
"Decompressed length {} would be larger than maximum of {}",
dest.len() + src.len() + extra_size,
max_size
)))
} else {
dest.reserve(src.len() + extra_size);
dest.extend_from_slice(src);
Ok(dest)
}
}
CompressType::General => {
// Prep for decompressed data
let header_len = dest.len();
let Ok(Some(expected_len)) = zstd_safe::get_frame_content_size(src) else {
return Err(Error::FailDecompress("Compression frame header is invalid".into()));
};
if expected_len > (max_size - header_len) as u64 {
return Err(Error::FailDecompress(format!(
"Decompressed length {} would be larger than maximum of {}",
dest.len() + src.len(),
max_size
)));
}
let expected_len = expected_len as usize;
dest.reserve(expected_len + extra_size);
dest.resize(header_len + expected_len, 0u8);
// Safety: Immediately before this, we reserve enough space for the header and the
// expected length, so setting the length is OK. The decompress function overwrites
// data and returns the new valid length, so no data is uninitialized after this
// block completes. In the event of a failure, the vec is freed, so it is never
// returned in an invalid state.
let len = zstd_safe::decompress(&mut dest[header_len..], src).map_err(|e| {
Error::FailDecompress(format!("Failed Decompression, zstd error = {}", e))
})?;
dest.truncate(header_len + len);
Ok(dest)
}
CompressType::Dict => {
// Fetch dictionary
let ddict = if let Compress::Dict(Dictionary(DictionaryPrivate::Zstd {
ddict,
..
})) = self
{
ddict
} else {
return Err(Error::BadHeader(
"Header uses dictionary compression, but this has no matching supported dictionary".into()));
};
// Prep for decompressed data
let header_len = dest.len();
let Ok(Some(expected_len)) = zstd_safe::get_frame_content_size(src) else {
return Err(Error::FailDecompress("Compression frame header is invalid".into()));
};
if expected_len > (max_size - header_len) as u64 {
return Err(Error::FailDecompress(format!(
"Decompressed length {} would be larger than maximum of {}",
dest.len() + src.len(),
max_size
)));
}
let expected_len = expected_len as usize;
dest.reserve(expected_len + extra_size);
dest.resize(header_len + expected_len, 0u8);
// Safety: Immediately before this, we reserve enough space for the header and the
// expected length, so setting the length is OK. The decompress function overwrites
// data and returns the new valid length, so no data is uninitialized after this
// block completes. In the event of a failure, the vec is freed, so it is never
// returned in an invalid state.
let mut dctx = zstd_safe::DCtx::create();
let len = dctx
.decompress_using_ddict(&mut dest[header_len..], src, ddict)
.map_err(|e| {
Error::FailDecompress(format!("Failed Decompression, zstd error = {}", e))
})?;
dest.truncate(header_len + len);
Ok(dest)
}
}
}
}
impl std::default::Default for Compress {
fn default() -> Self {
Compress::General {
algorithm: ALGORITHM_ZSTD,
level: 3,
}
}
}
/// A ZStandard Compression dictionary.
///
/// A new dictionary can be created by providing the desired compression level and the dictionary
/// as a byte vector.
#[derive(Clone, Debug, Serialize, Deserialize)]
pub struct Dictionary(DictionaryPrivate);
impl Dictionary {
/// Create a new ZStandard compression dictionary.
pub fn new_zstd(level: u8, dict: Vec<u8>) -> Self {
let cdict = zstd_safe::create_cdict(&dict, level as i32);
let ddict = zstd_safe::create_ddict(&dict);
Self(DictionaryPrivate::Zstd {
level,
dict,
cdict,
ddict,
})
}
}
#[derive(Serialize, Deserialize)]
#[serde(from = "DictionarySerde", into = "DictionarySerde")]
enum DictionaryPrivate {
Unknown {
algorithm: u8,
level: u8,
dict: Vec<u8>,
},
Zstd {
level: u8,
dict: Vec<u8>,
cdict: zstd_safe::CDict<'static>,
ddict: zstd_safe::DDict<'static>,
},
}
// Struct used solely for serialization/deserialization
#[derive(Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
struct DictionarySerde {
algorithm: u8,
level: u8,
dict: ByteBuf,
}
impl Clone for DictionaryPrivate {
fn clone(&self) -> Self {
match self {
DictionaryPrivate::Unknown {
algorithm,
level,
dict,
} => DictionaryPrivate::Unknown {
algorithm: *algorithm,
level: *level,
dict: dict.clone(),
},
DictionaryPrivate::Zstd { level, dict, .. } => DictionaryPrivate::Zstd {
level: *level,
dict: dict.clone(),
cdict: zstd_safe::create_cdict(dict, *level as i32),
ddict: zstd_safe::create_ddict(dict),
},
}
}
}
impl fmt::Debug for DictionaryPrivate {
fn fmt(&self, fmt: &mut fmt::Formatter) -> fmt::Result {
let (algorithm, level, dict) = match self {
DictionaryPrivate::Unknown {
algorithm,
level,
dict,
} => (algorithm, level, dict),
DictionaryPrivate::Zstd { level, dict, .. } => (&ALGORITHM_ZSTD, level, dict),
};
fmt.debug_struct("Dictionary")
.field("algorithm", algorithm)
.field("level", level)
.field("dict", dict)
.finish()
}
}
impl From<DictionarySerde> for DictionaryPrivate {
fn from(value: DictionarySerde) -> Self {
match value.algorithm {
ALGORITHM_ZSTD => {
let cdict = zstd_safe::create_cdict(&value.dict, value.level as i32);
let ddict = zstd_safe::create_ddict(&value.dict);
DictionaryPrivate::Zstd {
level: value.level,
dict: value.dict.into_vec(),
cdict,
ddict,
}
}
_ => DictionaryPrivate::Unknown {
algorithm: value.algorithm,
level: value.level,
dict: value.dict.into_vec(),
},
}
}
}
impl From<DictionaryPrivate> for DictionarySerde {
fn from(value: DictionaryPrivate) -> Self {
match value {
DictionaryPrivate::Unknown {
algorithm,
level,
dict,
} => Self {
algorithm,
level,
dict: ByteBuf::from(dict),
},
DictionaryPrivate::Zstd { level, dict, .. } => Self {
algorithm: ALGORITHM_ZSTD,
level,
dict: ByteBuf::from(dict),
},
}
}
}