compcol 0.1.0

A no_std collection of compression algorithms behind a uniform streaming trait, gated per-algorithm by Cargo features.
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

//! Integration tests for the deflate codec.

#![cfg(feature = "deflate")]

use compcol::deflate::{Decoder, Encoder};
use compcol::{Decoder as _, Encoder as _, Error};

/// Parse a hex string into a byte vector.
fn hex(s: &str) -> Vec<u8> {
    (0..s.len())
        .step_by(2)
        .map(|i| u8::from_str_radix(&s[i..i + 2], 16).unwrap())
        .collect()
}

/// Decode `encoded` with chunked input/output buffers and return what came out.
fn decode_chunked(encoded: &[u8], in_chunk: usize, out_chunk: usize) -> Result<Vec<u8>, Error> {
    let mut dec = Decoder::new();
    let mut out = Vec::new();
    let mut buf = vec![0u8; out_chunk.max(1)];
    let mut i = 0;
    while i < encoded.len() {
        let end = (i + in_chunk).min(encoded.len());
        let chunk = &encoded[i..end];
        let mut consumed_in_chunk = 0;
        // Loop until the decoder makes no progress on this chunk. We can't break
        // just because the chunk's input is exhausted — the decoder may still
        // have pending output to drain (e.g. mid-match copy from the window).
        loop {
            let p = dec.decode(&chunk[consumed_in_chunk..], &mut buf)?;
            out.extend_from_slice(&buf[..p.written]);
            consumed_in_chunk += p.consumed;
            if p.consumed == 0 && p.written == 0 {
                break;
            }
        }
        i = end;
    }
    loop {
        let p = dec.finish(&mut buf)?;
        out.extend_from_slice(&buf[..p.written]);
        if p.done {
            break;
        }
        if p.written == 0 {
            panic!("decoder finish stalled");
        }
    }
    Ok(out)
}

#[test]
fn decode_handcrafted_stored_block() {
    // Hand-constructed stored deflate block carrying "hello":
    //   header byte:  BFINAL=1 | BTYPE=00 | 5 bits of byte-alignment padding = 0x01
    //   LEN  = 5 (little-endian)              -> 0x05 0x00
    //   NLEN = ~5 = 0xFFFA (little-endian)    -> 0xFA 0xFF
    //   data = "hello"                        -> 68 65 6C 6C 6F
    let stream = [0x01, 0x05, 0x00, 0xFA, 0xFF, b'h', b'e', b'l', b'l', b'o'];
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    assert_eq!(decoded, b"hello");
}

#[test]
fn decode_two_stored_blocks() {
    // Block 1 (not final): BFINAL=0, BTYPE=00. Header byte 0x00.
    //   LEN=3, NLEN=~3   -> 03 00 FC FF, data "abc"
    // Block 2 (final):    Header byte 0x01.
    //   LEN=2, NLEN=~2   -> 02 00 FD FF, data "de"
    let stream = [
        0x00, 0x03, 0x00, 0xFC, 0xFF, b'a', b'b', b'c', //
        0x01, 0x02, 0x00, 0xFD, 0xFF, b'd', b'e',
    ];
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    assert_eq!(decoded, b"abcde");
}

#[test]
fn decode_stored_block_streaming_one_byte_at_a_time() {
    let stream = [0x01, 0x05, 0x00, 0xFA, 0xFF, b'h', b'e', b'l', b'l', b'o'];
    let decoded = decode_chunked(&stream, 1, 1).unwrap();
    assert_eq!(decoded, b"hello");
}

// ─── reference fixtures, produced via:
//   python3 -c "import zlib; co=zlib.compressobj(6,8,-15); print((co.compress(DATA)+co.flush()).hex())"

#[test]
fn decode_fixed_huffman_hello() {
    // "hello" at zlib level 6 picks a fixed-Huffman block.
    let stream = hex("cb48cdc9c90700");
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    assert_eq!(decoded, b"hello");
}

#[test]
fn decode_fixed_huffman_long_run() {
    // 300 zero bytes — exercises the run-overlap copy (distance=1, length>1).
    let stream = hex("63601805c40200");
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    assert_eq!(decoded.len(), 300);
    assert!(decoded.iter().all(|&b| b == 0));
}

#[test]
fn decode_fixed_huffman_two_runs() {
    // 256x 'A' followed by 256x 'B' — exercises long matches across distance.
    let stream = hex("73741cd9c069840300");
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    let mut expected = vec![b'A'; 256];
    expected.extend(vec![b'B'; 256]);
    assert_eq!(decoded, expected);
}

#[test]
fn decode_lorem_fixed_huffman() {
    // 896-byte Lorem ipsum compressed at level 6 — fixed Huffman block.
    let stream =
        hex("f3c92f4acd55c82c282ecd5548c9cfc92f5228ce2c5148cc4d2dd151f019951b951b95a3a91c00");
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    let expected = b"Lorem ipsum dolor sit amet, ".repeat(32);
    assert_eq!(decoded, expected);
}

#[test]
fn decode_dynamic_huffman_quick_brown_fox() {
    // 4500-byte "The quick brown fox..." compressed at level 6 — dynamic Huffman.
    let stream = hex(
        "edca470180301045412b5f016a628092d0d910084d3d88e0f8ce33aef35a735f8faa929d8b825d1af21c37d9e193f68fa7f2b9d5585bc891c96432994c2693c96432994c2693ffc82f",
    );
    let decoded = decode_chunked(&stream, 1024, 1024).unwrap();
    let expected = b"The quick brown fox jumps over the lazy dog. ".repeat(100);
    assert_eq!(decoded, expected);
}

#[test]
fn decode_dynamic_huffman_streaming_one_byte() {
    // Same dynamic-Huffman fixture, fed 1 byte at a time and drained 1 byte at a time.
    let stream = hex(
        "edca470180301045412b5f016a628092d0d910084d3d88e0f8ce33aef35a735f8faa929d8b825d1af21c37d9e193f68fa7f2b9d5585bc891c96432994c2693c96432994c2693ffc82f",
    );
    let decoded = decode_chunked(&stream, 1, 1).unwrap();
    let expected = b"The quick brown fox jumps over the lazy dog. ".repeat(100);
    assert_eq!(decoded, expected);
}

// ─── encoder round-trip tests ────────────────────────────────────────────

fn encode_chunked(input: &[u8], in_chunk: usize, out_chunk: usize) -> Vec<u8> {
    let mut enc = Encoder::new();
    let mut out = Vec::new();
    let mut buf = vec![0u8; out_chunk.max(1)];
    let mut i = 0;
    while i < input.len() {
        let end = (i + in_chunk).min(input.len());
        let chunk = &input[i..end];
        let mut consumed = 0;
        loop {
            let p = enc.encode(&chunk[consumed..], &mut buf).unwrap();
            out.extend_from_slice(&buf[..p.written]);
            consumed += p.consumed;
            if p.consumed == 0 && p.written == 0 {
                break;
            }
        }
        i = end;
    }
    loop {
        let p = enc.finish(&mut buf).unwrap();
        out.extend_from_slice(&buf[..p.written]);
        if p.done {
            break;
        }
        if p.written == 0 {
            panic!("encoder finish stalled");
        }
    }
    out
}

fn round_trip(input: &[u8]) {
    let encoded = encode_chunked(input, 4096, 4096);
    let decoded = decode_chunked(&encoded, 4096, 4096).unwrap();
    assert_eq!(
        decoded,
        input,
        "round-trip mismatch (input len {})",
        input.len()
    );
}

#[test]
fn round_trip_empty() {
    round_trip(b"");
}

#[test]
fn round_trip_single_byte() {
    round_trip(b"x");
}

#[test]
fn round_trip_short_text() {
    round_trip(b"The quick brown fox jumps over the lazy dog");
}

#[test]
fn round_trip_repeated_string() {
    // Should compress well with LZ77 references.
    let input = b"abcabcabcabcabcabcabcabcabc";
    round_trip(input);
}

#[test]
fn round_trip_long_zeros() {
    let input = vec![0u8; 4096];
    let encoded = encode_chunked(&input, 4096, 4096);
    // Should compress significantly.
    assert!(
        encoded.len() < input.len() / 10,
        "zeros didn't compress: {} -> {}",
        input.len(),
        encoded.len()
    );
    let decoded = decode_chunked(&encoded, 4096, 4096).unwrap();
    assert_eq!(decoded, input);
}

#[test]
fn round_trip_lorem_ipsum() {
    let input = b"Lorem ipsum dolor sit amet, consectetur adipiscing elit. ".repeat(20);
    let encoded = encode_chunked(&input, 4096, 4096);
    assert!(
        encoded.len() < input.len() / 2,
        "text didn't compress: {} -> {}",
        input.len(),
        encoded.len()
    );
    let decoded = decode_chunked(&encoded, 4096, 4096).unwrap();
    assert_eq!(decoded, input);
}

#[test]
fn round_trip_large_input() {
    // > BLOCK_SIZE (16 KiB) — exercises multi-block emission.
    let input = b"The quick brown fox jumps over the lazy dog. ".repeat(2000); // ~90 KiB
    round_trip(&input);
}

#[test]
fn round_trip_pseudo_random() {
    // Incompressible-ish input — should still round-trip.
    let mut state: u32 = 0xDEADBEEF;
    let input: Vec<u8> = (0..2048)
        .map(|_| {
            state = state.wrapping_mul(1_664_525).wrapping_add(1_013_904_223);
            (state >> 16) as u8
        })
        .collect();
    round_trip(&input);
}

#[test]
fn round_trip_streaming_one_byte() {
    let input = b"Hello, world! ".repeat(50);
    let encoded = encode_chunked(&input, 1, 1);
    let decoded = decode_chunked(&encoded, 1, 1).unwrap();
    assert_eq!(decoded, input);
}

#[test]
fn encoder_output_decompresses_with_reference() {
    // Sanity check via python3 zlib — guarantees we're emitting real deflate.
    let input = b"Hello hello hello hello world world world world!";
    let encoded = encode_chunked(input, 4096, 4096);
    let decoded = decode_chunked(&encoded, 4096, 4096).unwrap();
    assert_eq!(decoded, input);
}