summaryrefslogtreecommitdiffstats
path: root/encoder.cc
blob: 49a613e61cf883b772fa62a9dffb772a4e6cdf01 (plain)
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
/*  Lzip - LZMA lossless data compressor
    Copyright (C) 2008-2017 Antonio Diaz Diaz.

    This program is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation, either version 2 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#define _FILE_OFFSET_BITS 64

#include <algorithm>
#include <cerrno>
#include <cstdlib>
#include <cstring>
#include <string>
#include <vector>
#include <stdint.h>

#include "lzip.h"
#include "encoder_base.h"
#include "encoder.h"


const CRC32 crc32;


int LZ_encoder::get_match_pairs( Pair * pairs )
  {
  int len_limit = match_len_limit;
  if( len_limit > available_bytes() )
    {
    len_limit = available_bytes();
    if( len_limit < 4 ) return 0;
    }

  int maxlen = 0;
  int num_pairs = 0;
  const int pos1 = pos + 1;
  const int min_pos = ( pos > dictionary_size ) ? pos - dictionary_size : 0;
  const uint8_t * const data = ptr_to_current_pos();

  unsigned tmp = crc32[data[0]] ^ data[1];
  const int key2 = tmp & ( num_prev_positions2 - 1 );
  tmp ^= (unsigned)data[2] << 8;
  const int key3 = num_prev_positions2 + ( tmp & ( num_prev_positions3 - 1 ) );
  const int key4 = num_prev_positions23 +
                   ( ( tmp ^ ( crc32[data[3]] << 5 ) ) & key4_mask );

  if( pairs )
    {
    int np2 = prev_positions[key2];
    int np3 = prev_positions[key3];
    if( np2 > min_pos && buffer[np2-1] == data[0] )
      {
      pairs[0].dis = pos - np2;
      pairs[0].len = maxlen = 2;
      num_pairs = 1;
      }
    if( np2 != np3 && np3 > min_pos && buffer[np3-1] == data[0] )
      {
      maxlen = 3;
      np2 = np3;
      pairs[num_pairs].dis = pos - np2;
      ++num_pairs;
      }
    if( num_pairs > 0 )
      {
      const int delta = pos1 - np2;
      while( maxlen < len_limit && data[maxlen-delta] == data[maxlen] )
        ++maxlen;
      pairs[num_pairs-1].len = maxlen;
      if( maxlen >= len_limit ) pairs = 0;	// done. now just skip
      }
    if( maxlen < 3 ) maxlen = 3;
    }

  prev_positions[key2] = pos1;
  prev_positions[key3] = pos1;
  int newpos1 = prev_positions[key4];
  prev_positions[key4] = pos1;

  int32_t * ptr0 = pos_array + ( cyclic_pos << 1 );
  int32_t * ptr1 = ptr0 + 1;
  int len = 0, len0 = 0, len1 = 0;

  for( int count = cycles; ; )
    {
    if( newpos1 <= min_pos || --count < 0 ) { *ptr0 = *ptr1 = 0; break; }

    const int delta = pos1 - newpos1;
    int32_t * const newptr = pos_array +
      ( ( cyclic_pos - delta +
          ( ( cyclic_pos >= delta ) ? 0 : dictionary_size + 1 ) ) << 1 );
    if( data[len-delta] == data[len] )
      {
      while( ++len < len_limit && data[len-delta] == data[len] ) {}
      if( pairs && maxlen < len )
        {
        pairs[num_pairs].dis = delta - 1;
        pairs[num_pairs].len = maxlen = len;
        ++num_pairs;
        }
      if( len >= len_limit )
        {
        *ptr0 = newptr[0];
        *ptr1 = newptr[1];
        break;
        }
      }
    if( data[len-delta] < data[len] )
      {
      *ptr0 = newpos1;
      ptr0 = newptr + 1;
      newpos1 = *ptr0;
      len0 = len; if( len1 < len ) len = len1;
      }
    else
      {
      *ptr1 = newpos1;
      ptr1 = newptr;
      newpos1 = *ptr1;
      len1 = len; if( len0 < len ) len = len0;
      }
    }
  return num_pairs;
  }


void LZ_encoder::update_distance_prices()
  {
  for( int dis = start_dis_model; dis < modeled_distances; ++dis )
    {
    const int dis_slot = dis_slots[dis];
    const int direct_bits = ( dis_slot >> 1 ) - 1;
    const int base = ( 2 | ( dis_slot & 1 ) ) << direct_bits;
    const int price = price_symbol_reversed( bm_dis + ( base - dis_slot ),
                                             dis - base, direct_bits );
    for( int len_state = 0; len_state < len_states; ++len_state )
      dis_prices[len_state][dis] = price;
    }

  for( int len_state = 0; len_state < len_states; ++len_state )
    {
    int * const dsp = dis_slot_prices[len_state];
    const Bit_model * const bmds = bm_dis_slot[len_state];
    int slot = 0;
    for( ; slot < end_dis_model; ++slot )
      dsp[slot] = price_symbol6( bmds, slot );
    for( ; slot < num_dis_slots; ++slot )
      dsp[slot] = price_symbol6( bmds, slot ) +
                  (((( slot >> 1 ) - 1 ) - dis_align_bits ) << price_shift_bits );

    int * const dp = dis_prices[len_state];
    int dis = 0;
    for( ; dis < start_dis_model; ++dis )
      dp[dis] = dsp[dis];
    for( ; dis < modeled_distances; ++dis )
      dp[dis] += dsp[dis_slots[dis]];
    }
  }


/* Returns the number of bytes advanced (ahead).
   trials[0]..trials[ahead-1] contain the steps to encode.
   ( trials[0].dis4 == -1 ) means literal.
   A match/rep longer or equal than match_len_limit finishes the sequence.
*/
int LZ_encoder::sequence_optimizer( const int reps[num_rep_distances],
                                    const State state )
  {
  int num_pairs, num_trials;

  if( pending_num_pairs > 0 )			// from previous call
    {
    num_pairs = pending_num_pairs;
    pending_num_pairs = 0;
    }
  else
    num_pairs = read_match_distances();
  const int main_len = ( num_pairs > 0 ) ? pairs[num_pairs-1].len : 0;

  int replens[num_rep_distances];
  int rep_index = 0;
  for( int i = 0; i < num_rep_distances; ++i )
    {
    replens[i] = true_match_len( 0, reps[i] + 1 );
    if( replens[i] > replens[rep_index] ) rep_index = i;
    }
  if( replens[rep_index] >= match_len_limit )
    {
    trials[0].price = replens[rep_index];
    trials[0].dis4 = rep_index;
    move_and_update( replens[rep_index] );
    return replens[rep_index];
    }

  if( main_len >= match_len_limit )
    {
    trials[0].price = main_len;
    trials[0].dis4 = pairs[num_pairs-1].dis + num_rep_distances;
    move_and_update( main_len );
    return main_len;
    }

  const int pos_state = data_position() & pos_state_mask;
  const uint8_t prev_byte = peek( 1 );
  const uint8_t cur_byte = peek( 0 );
  const uint8_t match_byte = peek( reps[0] + 1 );

  trials[1].price = price0( bm_match[state()][pos_state] );
  if( state.is_char() )
    trials[1].price += price_literal( prev_byte, cur_byte );
  else
    trials[1].price += price_matched( prev_byte, cur_byte, match_byte );
  trials[1].dis4 = -1;					// literal

  const int match_price = price1( bm_match[state()][pos_state] );
  const int rep_match_price = match_price + price1( bm_rep[state()] );

  if( match_byte == cur_byte )
    trials[1].update( rep_match_price + price_shortrep( state, pos_state ), 0, 0 );

  num_trials = std::max( main_len, replens[rep_index] );

  if( num_trials < min_match_len )
    {
    trials[0].price = 1;
    trials[0].dis4 = trials[1].dis4;
    move_pos();
    return 1;
    }

  trials[0].state = state;
  for( int i = 0; i < num_rep_distances; ++i )
    trials[0].reps[i] = reps[i];

  for( int len = min_match_len; len <= num_trials; ++len )
    trials[len].price = infinite_price;

  for( int rep = 0; rep < num_rep_distances; ++rep )
    {
    if( replens[rep] < min_match_len ) continue;
    const int price = rep_match_price + price_rep( rep, state, pos_state );
    for( int len = min_match_len; len <= replens[rep]; ++len )
      trials[len].update( price + rep_len_prices.price( len, pos_state ),
                          rep, 0 );
    }

  if( main_len > replens[0] )
    {
    const int normal_match_price = match_price + price0( bm_rep[state()] );
    int i = 0, len = std::max( replens[0] + 1, (int)min_match_len );
    while( len > pairs[i].len ) ++i;
    while( true )
      {
      const int dis = pairs[i].dis;
      trials[len].update( normal_match_price + price_pair( dis, len, pos_state ),
                          dis + num_rep_distances, 0 );
      if( ++len > pairs[i].len && ++i >= num_pairs ) break;
      }
    }

  int cur = 0;
  while( true )				// price optimization loop
    {
    move_pos();
    if( ++cur >= num_trials )		// no more initialized trials
      {
      backward( cur );
      return cur;
      }

    const int num_pairs = read_match_distances();
    const int newlen = ( num_pairs > 0 ) ? pairs[num_pairs-1].len : 0;
    if( newlen >= match_len_limit )
      {
      pending_num_pairs = num_pairs;
      backward( cur );
      return cur;
      }

    // give final values to current trial
    Trial & cur_trial = trials[cur];
    State cur_state;
    {
    const int dis4 = cur_trial.dis4;
    int prev_index = cur_trial.prev_index;
    const int prev_index2 = cur_trial.prev_index2;

    if( prev_index2 == single_step_trial )
      {
      cur_state = trials[prev_index].state;
      if( prev_index + 1 == cur )			// len == 1
        {
        if( dis4 == 0 ) cur_state.set_short_rep();
        else cur_state.set_char();			// literal
        }
      else if( dis4 < num_rep_distances ) cur_state.set_rep();
      else cur_state.set_match();
      }
    else
      {
      if( prev_index2 == dual_step_trial )	// dis4 == 0 (rep0)
        --prev_index;
      else					// prev_index2 >= 0
        prev_index = prev_index2;
      cur_state.set_char_rep();
      }
    cur_trial.state = cur_state;
    for( int i = 0; i < num_rep_distances; ++i )
      cur_trial.reps[i] = trials[prev_index].reps[i];
    mtf_reps( dis4, cur_trial.reps );		// literal is ignored
    }

    const int pos_state = data_position() & pos_state_mask;
    const uint8_t prev_byte = peek( 1 );
    const uint8_t cur_byte = peek( 0 );
    const uint8_t match_byte = peek( cur_trial.reps[0] + 1 );

    int next_price = cur_trial.price +
                     price0( bm_match[cur_state()][pos_state] );
    if( cur_state.is_char() )
      next_price += price_literal( prev_byte, cur_byte );
    else
      next_price += price_matched( prev_byte, cur_byte, match_byte );

    // try last updates to next trial
    Trial & next_trial = trials[cur+1];

    next_trial.update( next_price, -1, cur );		// literal

    const int match_price = cur_trial.price + price1( bm_match[cur_state()][pos_state] );
    const int rep_match_price = match_price + price1( bm_rep[cur_state()] );

    if( match_byte == cur_byte && next_trial.dis4 != 0 &&
        next_trial.prev_index2 == single_step_trial )
      {
      const int price = rep_match_price + price_shortrep( cur_state, pos_state );
      if( price <= next_trial.price )
        {
        next_trial.price = price;
        next_trial.dis4 = 0;				// rep0
        next_trial.prev_index = cur;
        }
      }

    const int triable_bytes =
      std::min( available_bytes(), max_num_trials - 1 - cur );
    if( triable_bytes < min_match_len ) continue;

    const int len_limit = std::min( match_len_limit, triable_bytes );

    // try literal + rep0
    if( match_byte != cur_byte && next_trial.prev_index != cur )
      {
      const uint8_t * const data = ptr_to_current_pos();
      const int dis = cur_trial.reps[0] + 1;
      const int limit = std::min( match_len_limit + 1, triable_bytes );
      int len = 1;
      while( len < limit && data[len-dis] == data[len] ) ++len;
      if( --len >= min_match_len )
        {
        const int pos_state2 = ( pos_state + 1 ) & pos_state_mask;
        State state2 = cur_state; state2.set_char();
        const int price = next_price +
                          price1( bm_match[state2()][pos_state2] ) +
                          price1( bm_rep[state2()] ) +
                          price_rep0_len( len, state2, pos_state2 );
        while( num_trials < cur + 1 + len )
          trials[++num_trials].price = infinite_price;
        trials[cur+1+len].update2( price, cur + 1 );
        }
      }

    int start_len = min_match_len;

    // try rep distances
    for( int rep = 0; rep < num_rep_distances; ++rep )
      {
      const uint8_t * const data = ptr_to_current_pos();
      const int dis = cur_trial.reps[rep] + 1;
      int len;

      if( data[0-dis] != data[0] || data[1-dis] != data[1] ) continue;
      for( len = min_match_len; len < len_limit; ++len )
        if( data[len-dis] != data[len] ) break;
      while( num_trials < cur + len )
        trials[++num_trials].price = infinite_price;
      int price = rep_match_price + price_rep( rep, cur_state, pos_state );
      for( int i = min_match_len; i <= len; ++i )
        trials[cur+i].update( price + rep_len_prices.price( i, pos_state ),
                              rep, cur );

      if( rep == 0 ) start_len = len + 1;	// discard shorter matches

      // try rep + literal + rep0
      int len2 = len + 1;
      const int limit = std::min( match_len_limit + len2, triable_bytes );
      while( len2 < limit && data[len2-dis] == data[len2] ) ++len2;
      len2 -= len + 1;
      if( len2 < min_match_len ) continue;

      int pos_state2 = ( pos_state + len ) & pos_state_mask;
      State state2 = cur_state; state2.set_rep();
      price += rep_len_prices.price( len, pos_state ) +
               price0( bm_match[state2()][pos_state2] ) +
               price_matched( data[len-1], data[len], data[len-dis] );
      pos_state2 = ( pos_state2 + 1 ) & pos_state_mask;
      state2.set_char();
      price += price1( bm_match[state2()][pos_state2] ) +
               price1( bm_rep[state2()] ) +
               price_rep0_len( len2, state2, pos_state2 );
      while( num_trials < cur + len + 1 + len2 )
        trials[++num_trials].price = infinite_price;
      trials[cur+len+1+len2].update3( price, rep, cur + len + 1, cur );
      }

    // try matches
    if( newlen >= start_len && newlen <= len_limit )
      {
      const int normal_match_price = match_price +
                                     price0( bm_rep[cur_state()] );

      while( num_trials < cur + newlen )
        trials[++num_trials].price = infinite_price;

      int i = 0;
      while( pairs[i].len < start_len ) ++i;
      int dis = pairs[i].dis;
      for( int len = start_len; ; ++len )
        {
        int price = normal_match_price + price_pair( dis, len, pos_state );
        trials[cur+len].update( price, dis + num_rep_distances, cur );

        // try match + literal + rep0
        if( len == pairs[i].len )
          {
          const uint8_t * const data = ptr_to_current_pos();
          const int dis2 = dis + 1;
          int len2 = len + 1;
          const int limit = std::min( match_len_limit + len2, triable_bytes );
          while( len2 < limit && data[len2-dis2] == data[len2] ) ++len2;
          len2 -= len + 1;
          if( len2 >= min_match_len )
            {
            int pos_state2 = ( pos_state + len ) & pos_state_mask;
            State state2 = cur_state; state2.set_match();
            price += price0( bm_match[state2()][pos_state2] ) +
                     price_matched( data[len-1], data[len], data[len-dis2] );
            pos_state2 = ( pos_state2 + 1 ) & pos_state_mask;
            state2.set_char();
            price += price1( bm_match[state2()][pos_state2] ) +
                     price1( bm_rep[state2()] ) +
                     price_rep0_len( len2, state2, pos_state2 );

            while( num_trials < cur + len + 1 + len2 )
              trials[++num_trials].price = infinite_price;
            trials[cur+len+1+len2].update3( price, dis + num_rep_distances,
                                            cur + len + 1, cur );
            }
          if( ++i >= num_pairs ) break;
          dis = pairs[i].dis;
          }
        }
      }
    }
  }


bool LZ_encoder::encode_member( const unsigned long long member_size )
  {
  const unsigned long long member_size_limit =
    member_size - File_trailer::size - max_marker_size;
  const bool best = ( match_len_limit > 12 );
  const int dis_price_count = best ? 1 : 512;
  const int align_price_count = best ? 1 : dis_align_size;
  const int price_count = ( match_len_limit > 36 ) ? 1013 : 4093;
  int price_counter = 0;		// counters may decrement below 0
  int dis_price_counter = 0;
  int align_price_counter = 0;
  int reps[num_rep_distances];
  State state;
  for( int i = 0; i < num_rep_distances; ++i ) reps[i] = 0;

  if( data_position() != 0 || renc.member_position() != File_header::size )
    return false;				// can be called only once

  if( !data_finished() )			// encode first byte
    {
    const uint8_t prev_byte = 0;
    const uint8_t cur_byte = peek( 0 );
    renc.encode_bit( bm_match[state()][0], 0 );
    encode_literal( prev_byte, cur_byte );
    crc32.update_byte( crc_, cur_byte );
    get_match_pairs();
    move_pos();
    }

  while( !data_finished() )
    {
    if( price_counter <= 0 && pending_num_pairs == 0 )
      {
      price_counter = price_count;	// recalculate prices every these bytes
      if( dis_price_counter <= 0 )
        { dis_price_counter = dis_price_count; update_distance_prices(); }
      if( align_price_counter <= 0 )
        {
        align_price_counter = align_price_count;
        for( int i = 0; i < dis_align_size; ++i )
          align_prices[i] = price_symbol_reversed( bm_align, i, dis_align_bits );
        }
      match_len_prices.update_prices();
      rep_len_prices.update_prices();
      }

    int ahead = sequence_optimizer( reps, state );
    price_counter -= ahead;

    for( int i = 0; ahead > 0; )
      {
      const int pos_state = ( data_position() - ahead ) & pos_state_mask;
      const int len = trials[i].price;
      int dis = trials[i].dis4;

      bool bit = ( dis < 0 );
      renc.encode_bit( bm_match[state()][pos_state], !bit );
      if( bit )					// literal byte
        {
        const uint8_t prev_byte = peek( ahead + 1 );
        const uint8_t cur_byte = peek( ahead );
        crc32.update_byte( crc_, cur_byte );
        if( state.is_char_set_char() )
          encode_literal( prev_byte, cur_byte );
        else
          {
          const uint8_t match_byte = peek( ahead + reps[0] + 1 );
          encode_matched( prev_byte, cur_byte, match_byte );
          }
        }
      else					// match or repeated match
        {
        crc32.update_buf( crc_, ptr_to_current_pos() - ahead, len );
        mtf_reps( dis, reps );
        bit = ( dis < num_rep_distances );
        renc.encode_bit( bm_rep[state()], bit );
        if( bit )				// repeated match
          {
          bit = ( dis == 0 );
          renc.encode_bit( bm_rep0[state()], !bit );
          if( bit )
            renc.encode_bit( bm_len[state()][pos_state], len > 1 );
          else
            {
            renc.encode_bit( bm_rep1[state()], dis > 1 );
            if( dis > 1 )
              renc.encode_bit( bm_rep2[state()], dis > 2 );
            }
          if( len == 1 ) state.set_short_rep();
          else
            {
            renc.encode_len( rep_len_model, len, pos_state );
            rep_len_prices.decrement_counter( pos_state );
            state.set_rep();
            }
          }
        else					// match
          {
          dis -= num_rep_distances;
          encode_pair( dis, len, pos_state );
          if( dis >= modeled_distances ) --align_price_counter;
          --dis_price_counter;
          match_len_prices.decrement_counter( pos_state );
          state.set_match();
          }
        }
      ahead -= len; i += len;
      if( renc.member_position() >= member_size_limit )
        {
        if( !dec_pos( ahead ) ) return false;
        full_flush( state );
        return true;
        }
      }
    }
  full_flush( state );
  return true;
  }