Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/libjpeg/jchuff.c @ 2:b50eed0cc0ef upstream

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
ADD: MuPDF v1.26.7: the MuPDF source as downloaded by a default build of PyMuPDF 1.26.4. The directory name has changed: no version number in the expanded directory now.
author Franz Glasner <fzglas.hg@dom66.de>
date Mon, 15 Sep 2025 11:43:07 +0200
parents
children
comparison
equal deleted inserted replaced
1:1d09e1dec1d9 2:b50eed0cc0ef
1 /*
2 * jchuff.c
3 *
4 * Copyright (C) 1991-1997, Thomas G. Lane.
5 * Modified 2006-2023 by Guido Vollbeding.
6 * This file is part of the Independent JPEG Group's software.
7 * For conditions of distribution and use, see the accompanying README file.
8 *
9 * This file contains Huffman entropy encoding routines.
10 * Both sequential and progressive modes are supported in this single module.
11 *
12 * Much of the complexity here has to do with supporting output suspension.
13 * If the data destination module demands suspension, we want to be able to
14 * back up to the start of the current MCU. To do this, we copy state
15 * variables into local working storage, and update them back to the
16 * permanent JPEG objects only upon successful completion of an MCU.
17 *
18 * We do not support output suspension for the progressive JPEG mode, since
19 * the library currently does not allow multiple-scan files to be written
20 * with output suspension.
21 */
22
23 #define JPEG_INTERNALS
24 #include "jinclude.h"
25 #include "jpeglib.h"
26
27
28 /* The legal range of a DCT coefficient is
29 * -1024 .. +1023 for 8-bit sample data precision;
30 * -16384 .. +16383 for 12-bit sample data precision.
31 * Hence the magnitude should always fit in sample data precision + 2 bits.
32 */
33
34 /* Derived data constructed for each Huffman table */
35
36 typedef struct {
37 unsigned int ehufco[256]; /* code for each symbol */
38 char ehufsi[256]; /* length of code for each symbol */
39 /* If no code has been allocated for a symbol S, ehufsi[S] contains 0 */
40 } c_derived_tbl;
41
42
43 /* Expanded entropy encoder object for Huffman encoding.
44 *
45 * The savable_state subrecord contains fields that change within an MCU,
46 * but must not be updated permanently until we complete the MCU.
47 */
48
49 typedef struct {
50 INT32 put_buffer; /* current bit-accumulation buffer */
51 int put_bits; /* # of bits now in it */
52 int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */
53 } savable_state;
54
55 /* This macro is to work around compilers with missing or broken
56 * structure assignment. You'll need to fix this code if you have
57 * such a compiler and you change MAX_COMPS_IN_SCAN.
58 */
59
60 #ifndef NO_STRUCT_ASSIGN
61 #define ASSIGN_STATE(dest,src) ((dest) = (src))
62 #else
63 #if MAX_COMPS_IN_SCAN == 4
64 #define ASSIGN_STATE(dest,src) \
65 ((dest).put_buffer = (src).put_buffer, \
66 (dest).put_bits = (src).put_bits, \
67 (dest).last_dc_val[0] = (src).last_dc_val[0], \
68 (dest).last_dc_val[1] = (src).last_dc_val[1], \
69 (dest).last_dc_val[2] = (src).last_dc_val[2], \
70 (dest).last_dc_val[3] = (src).last_dc_val[3])
71 #endif
72 #endif
73
74
75 typedef struct {
76 struct jpeg_entropy_encoder pub; /* public fields */
77
78 savable_state saved; /* Bit buffer & DC state at start of MCU */
79
80 /* These fields are NOT loaded into local working state. */
81 unsigned int restarts_to_go; /* MCUs left in this restart interval */
82 int next_restart_num; /* next restart number to write (0-7) */
83
84 /* Pointers to derived tables (these workspaces have image lifespan) */
85 c_derived_tbl * dc_derived_tbls[NUM_HUFF_TBLS];
86 c_derived_tbl * ac_derived_tbls[NUM_HUFF_TBLS];
87
88 /* Statistics tables for optimization */
89 long * dc_count_ptrs[NUM_HUFF_TBLS];
90 long * ac_count_ptrs[NUM_HUFF_TBLS];
91
92 /* Following fields used only in progressive mode */
93
94 /* Mode flag: TRUE for optimization, FALSE for actual data output */
95 boolean gather_statistics;
96
97 /* next_output_byte/free_in_buffer are local copies of cinfo->dest fields.
98 */
99 JOCTET * next_output_byte; /* => next byte to write in buffer */
100 size_t free_in_buffer; /* # of byte spaces remaining in buffer */
101 j_compress_ptr cinfo; /* link to cinfo (needed for dump_buffer) */
102
103 /* Coding status for AC components */
104 int ac_tbl_no; /* the table number of the single component */
105 unsigned int EOBRUN; /* run length of EOBs */
106 unsigned int BE; /* # of buffered correction bits before MCU */
107 char * bit_buffer; /* buffer for correction bits (1 per char) */
108 /* packing correction bits tightly would save some space but cost time... */
109 } huff_entropy_encoder;
110
111 typedef huff_entropy_encoder * huff_entropy_ptr;
112
113 /* Working state while writing an MCU (sequential mode).
114 * This struct contains all the fields that are needed by subroutines.
115 */
116
117 typedef struct {
118 JOCTET * next_output_byte; /* => next byte to write in buffer */
119 size_t free_in_buffer; /* # of byte spaces remaining in buffer */
120 savable_state cur; /* Current bit buffer & DC state */
121 j_compress_ptr cinfo; /* dump_buffer needs access to this */
122 } working_state;
123
124 /* MAX_CORR_BITS is the number of bits the AC refinement correction-bit
125 * buffer can hold. Larger sizes may slightly improve compression, but
126 * 1000 is already well into the realm of overkill.
127 * The minimum safe size is 64 bits.
128 */
129
130 #define MAX_CORR_BITS 1000 /* Max # of correction bits I can buffer */
131
132 /* IRIGHT_SHIFT is like RIGHT_SHIFT, but works on int rather than INT32.
133 * We assume that int right shift is unsigned if INT32 right shift is,
134 * which should be safe.
135 */
136
137 #ifdef RIGHT_SHIFT_IS_UNSIGNED
138 #define ISHIFT_TEMPS int ishift_temp;
139 #define IRIGHT_SHIFT(x,shft) \
140 ((ishift_temp = (x)) < 0 ? \
141 (ishift_temp >> (shft)) | ((~0) << (16-(shft))) : \
142 (ishift_temp >> (shft)))
143 #else
144 #define ISHIFT_TEMPS
145 #define IRIGHT_SHIFT(x,shft) ((x) >> (shft))
146 #endif
147
148
149 /*
150 * Compute the derived values for a Huffman table.
151 * This routine also performs some validation checks on the table.
152 */
153
154 LOCAL(void)
155 jpeg_make_c_derived_tbl (j_compress_ptr cinfo, boolean isDC, int tblno,
156 c_derived_tbl ** pdtbl)
157 {
158 JHUFF_TBL *htbl;
159 c_derived_tbl *dtbl;
160 int p, i, l, lastp, si, maxsymbol;
161 char huffsize[257];
162 unsigned int huffcode[257];
163 unsigned int code;
164
165 /* Note that huffsize[] and huffcode[] are filled in code-length order,
166 * paralleling the order of the symbols themselves in htbl->huffval[].
167 */
168
169 /* Find the input Huffman table */
170 if (tblno < 0 || tblno >= NUM_HUFF_TBLS)
171 ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno);
172 htbl =
173 isDC ? cinfo->dc_huff_tbl_ptrs[tblno] : cinfo->ac_huff_tbl_ptrs[tblno];
174 if (htbl == NULL)
175 htbl = jpeg_std_huff_table((j_common_ptr) cinfo, isDC, tblno);
176
177 /* Allocate a workspace if we haven't already done so. */
178 if (*pdtbl == NULL)
179 *pdtbl = (c_derived_tbl *) (*cinfo->mem->alloc_small)
180 ((j_common_ptr) cinfo, JPOOL_IMAGE, SIZEOF(c_derived_tbl));
181 dtbl = *pdtbl;
182
183 /* Figure C.1: make table of Huffman code length for each symbol */
184
185 p = 0;
186 for (l = 1; l <= 16; l++) {
187 i = (int) htbl->bits[l];
188 if (i < 0 || p + i > 256) /* protect against table overrun */
189 ERREXIT(cinfo, JERR_BAD_HUFF_TABLE);
190 while (i--)
191 huffsize[p++] = (char) l;
192 }
193 huffsize[p] = 0;
194 lastp = p;
195
196 /* Figure C.2: generate the codes themselves */
197 /* We also validate that the counts represent a legal Huffman code tree. */
198
199 code = 0;
200 si = huffsize[0];
201 p = 0;
202 while (huffsize[p]) {
203 while (((int) huffsize[p]) == si) {
204 huffcode[p++] = code;
205 code++;
206 }
207 /* code is now 1 more than the last code used for codelength si; but
208 * it must still fit in si bits, since no code is allowed to be all ones.
209 */
210 if (((INT32) code) >= (((INT32) 1) << si))
211 ERREXIT(cinfo, JERR_BAD_HUFF_TABLE);
212 code <<= 1;
213 si++;
214 }
215
216 /* Figure C.3: generate encoding tables */
217 /* These are code and size indexed by symbol value */
218
219 /* Set all codeless symbols to have code length 0;
220 * this lets us detect duplicate VAL entries here, and later
221 * allows emit_bits to detect any attempt to emit such symbols.
222 */
223 MEMZERO(dtbl->ehufsi, SIZEOF(dtbl->ehufsi));
224
225 /* This is also a convenient place to check for out-of-range
226 * and duplicated VAL entries. We allow 0..255 for AC symbols
227 * but only 0..15 for DC. (We could constrain them further
228 * based on data depth and mode, but this seems enough.)
229 */
230 maxsymbol = isDC ? 15 : 255;
231
232 for (p = 0; p < lastp; p++) {
233 i = htbl->huffval[p];
234 if (i < 0 || i > maxsymbol || dtbl->ehufsi[i])
235 ERREXIT(cinfo, JERR_BAD_HUFF_TABLE);
236 dtbl->ehufco[i] = huffcode[p];
237 dtbl->ehufsi[i] = huffsize[p];
238 }
239 }
240
241
242 /* Outputting bytes to the file.
243 * NB: these must be called only when actually outputting,
244 * that is, entropy->gather_statistics == FALSE.
245 */
246
247 /* Emit a byte, taking 'action' if must suspend. */
248 #define emit_byte_s(state,val,action) \
249 { *(state)->next_output_byte++ = (JOCTET) (val); \
250 if (--(state)->free_in_buffer == 0) \
251 if (! dump_buffer_s(state)) \
252 { action; } }
253
254 /* Emit a byte */
255 #define emit_byte_e(entropy,val) \
256 { *(entropy)->next_output_byte++ = (JOCTET) (val); \
257 if (--(entropy)->free_in_buffer == 0) \
258 dump_buffer_e(entropy); }
259
260
261 LOCAL(boolean)
262 dump_buffer_s (working_state * state)
263 /* Empty the output buffer; return TRUE if successful, FALSE if must suspend */
264 {
265 struct jpeg_destination_mgr * dest = state->cinfo->dest;
266
267 if (! (*dest->empty_output_buffer) (state->cinfo))
268 return FALSE;
269 /* After a successful buffer dump, must reset buffer pointers */
270 state->next_output_byte = dest->next_output_byte;
271 state->free_in_buffer = dest->free_in_buffer;
272 return TRUE;
273 }
274
275
276 LOCAL(void)
277 dump_buffer_e (huff_entropy_ptr entropy)
278 /* Empty the output buffer; we do not support suspension in this case. */
279 {
280 struct jpeg_destination_mgr * dest = entropy->cinfo->dest;
281
282 if (! (*dest->empty_output_buffer) (entropy->cinfo))
283 ERREXIT(entropy->cinfo, JERR_CANT_SUSPEND);
284 /* After a successful buffer dump, must reset buffer pointers */
285 entropy->next_output_byte = dest->next_output_byte;
286 entropy->free_in_buffer = dest->free_in_buffer;
287 }
288
289
290 /* Outputting bits to the file */
291
292 /* Only the right 24 bits of put_buffer are used; the valid bits are
293 * left-justified in this part. At most 16 bits can be passed to emit_bits
294 * in one call, and we never retain more than 7 bits in put_buffer
295 * between calls, so 24 bits are sufficient.
296 */
297
298 INLINE
299 LOCAL(boolean)
300 emit_bits_s (working_state * state, unsigned int code, int size)
301 /* Emit some bits; return TRUE if successful, FALSE if must suspend */
302 {
303 /* This routine is heavily used, so it's worth coding tightly. */
304 register INT32 put_buffer;
305 register int put_bits;
306
307 /* if size is 0, caller used an invalid Huffman table entry */
308 if (size == 0)
309 ERREXIT(state->cinfo, JERR_HUFF_MISSING_CODE);
310
311 /* mask off any extra bits in code */
312 put_buffer = ((INT32) code) & ((((INT32) 1) << size) - 1);
313
314 /* new number of bits in buffer */
315 put_bits = size + state->cur.put_bits;
316
317 put_buffer <<= 24 - put_bits; /* align incoming bits */
318
319 /* and merge with old buffer contents */
320 put_buffer |= state->cur.put_buffer;
321
322 while (put_bits >= 8) {
323 int c = (int) ((put_buffer >> 16) & 0xFF);
324
325 emit_byte_s(state, c, return FALSE);
326 if (c == 0xFF) { /* need to stuff a zero byte? */
327 emit_byte_s(state, 0, return FALSE);
328 }
329 put_buffer <<= 8;
330 put_bits -= 8;
331 }
332
333 state->cur.put_buffer = put_buffer; /* update state variables */
334 state->cur.put_bits = put_bits;
335
336 return TRUE;
337 }
338
339
340 INLINE
341 LOCAL(void)
342 emit_bits_e (huff_entropy_ptr entropy, unsigned int code, int size)
343 /* Emit some bits, unless we are in gather mode */
344 {
345 /* This routine is heavily used, so it's worth coding tightly. */
346 register INT32 put_buffer;
347 register int put_bits;
348
349 /* if size is 0, caller used an invalid Huffman table entry */
350 if (size == 0)
351 ERREXIT(entropy->cinfo, JERR_HUFF_MISSING_CODE);
352
353 if (entropy->gather_statistics)
354 return; /* do nothing if we're only getting stats */
355
356 /* mask off any extra bits in code */
357 put_buffer = ((INT32) code) & ((((INT32) 1) << size) - 1);
358
359 /* new number of bits in buffer */
360 put_bits = size + entropy->saved.put_bits;
361
362 put_buffer <<= 24 - put_bits; /* align incoming bits */
363
364 /* and merge with old buffer contents */
365 put_buffer |= entropy->saved.put_buffer;
366
367 while (put_bits >= 8) {
368 int c = (int) ((put_buffer >> 16) & 0xFF);
369
370 emit_byte_e(entropy, c);
371 if (c == 0xFF) { /* need to stuff a zero byte? */
372 emit_byte_e(entropy, 0);
373 }
374 put_buffer <<= 8;
375 put_bits -= 8;
376 }
377
378 entropy->saved.put_buffer = put_buffer; /* update variables */
379 entropy->saved.put_bits = put_bits;
380 }
381
382
383 LOCAL(boolean)
384 flush_bits_s (working_state * state)
385 {
386 if (! emit_bits_s(state, 0x7F, 7)) /* fill any partial byte with ones */
387 return FALSE;
388 state->cur.put_buffer = 0; /* and reset bit-buffer to empty */
389 state->cur.put_bits = 0;
390 return TRUE;
391 }
392
393
394 LOCAL(void)
395 flush_bits_e (huff_entropy_ptr entropy)
396 {
397 emit_bits_e(entropy, 0x7F, 7); /* fill any partial byte with ones */
398 entropy->saved.put_buffer = 0; /* and reset bit-buffer to empty */
399 entropy->saved.put_bits = 0;
400 }
401
402
403 /*
404 * Emit (or just count) a Huffman symbol.
405 */
406
407 INLINE
408 LOCAL(void)
409 emit_dc_symbol (huff_entropy_ptr entropy, int tbl_no, int symbol)
410 {
411 if (entropy->gather_statistics)
412 entropy->dc_count_ptrs[tbl_no][symbol]++;
413 else {
414 c_derived_tbl * tbl = entropy->dc_derived_tbls[tbl_no];
415 emit_bits_e(entropy, tbl->ehufco[symbol], tbl->ehufsi[symbol]);
416 }
417 }
418
419
420 INLINE
421 LOCAL(void)
422 emit_ac_symbol (huff_entropy_ptr entropy, int tbl_no, int symbol)
423 {
424 if (entropy->gather_statistics)
425 entropy->ac_count_ptrs[tbl_no][symbol]++;
426 else {
427 c_derived_tbl * tbl = entropy->ac_derived_tbls[tbl_no];
428 emit_bits_e(entropy, tbl->ehufco[symbol], tbl->ehufsi[symbol]);
429 }
430 }
431
432
433 /*
434 * Emit bits from a correction bit buffer.
435 */
436
437 LOCAL(void)
438 emit_buffered_bits (huff_entropy_ptr entropy, char * bufstart,
439 unsigned int nbits)
440 {
441 if (entropy->gather_statistics)
442 return; /* no real work */
443
444 while (nbits > 0) {
445 emit_bits_e(entropy, (unsigned int) (*bufstart), 1);
446 bufstart++;
447 nbits--;
448 }
449 }
450
451
452 /*
453 * Emit any pending EOBRUN symbol.
454 */
455
456 LOCAL(void)
457 emit_eobrun (huff_entropy_ptr entropy)
458 {
459 register int temp, nbits;
460
461 if (entropy->EOBRUN > 0) { /* if there is any pending EOBRUN */
462 temp = entropy->EOBRUN;
463 nbits = 0;
464 while ((temp >>= 1))
465 nbits++;
466 /* safety check: shouldn't happen given limited correction-bit buffer */
467 if (nbits > 14)
468 ERREXIT(entropy->cinfo, JERR_HUFF_MISSING_CODE);
469
470 emit_ac_symbol(entropy, entropy->ac_tbl_no, nbits << 4);
471 if (nbits)
472 emit_bits_e(entropy, entropy->EOBRUN, nbits);
473
474 entropy->EOBRUN = 0;
475
476 /* Emit any buffered correction bits */
477 emit_buffered_bits(entropy, entropy->bit_buffer, entropy->BE);
478 entropy->BE = 0;
479 }
480 }
481
482
483 /*
484 * Emit a restart marker & resynchronize predictions.
485 */
486
487 LOCAL(boolean)
488 emit_restart_s (working_state * state, int restart_num)
489 {
490 int ci;
491
492 if (! flush_bits_s(state))
493 return FALSE;
494
495 emit_byte_s(state, 0xFF, return FALSE);
496 emit_byte_s(state, JPEG_RST0 + restart_num, return FALSE);
497
498 /* Re-initialize DC predictions to 0 */
499 for (ci = 0; ci < state->cinfo->comps_in_scan; ci++)
500 state->cur.last_dc_val[ci] = 0;
501
502 /* The restart counter is not updated until we successfully write the MCU. */
503
504 return TRUE;
505 }
506
507
508 LOCAL(void)
509 emit_restart_e (huff_entropy_ptr entropy, int restart_num)
510 {
511 int ci;
512
513 emit_eobrun(entropy);
514
515 if (! entropy->gather_statistics) {
516 flush_bits_e(entropy);
517 emit_byte_e(entropy, 0xFF);
518 emit_byte_e(entropy, JPEG_RST0 + restart_num);
519 }
520
521 if (entropy->cinfo->Ss == 0) {
522 /* Re-initialize DC predictions to 0 */
523 for (ci = 0; ci < entropy->cinfo->comps_in_scan; ci++)
524 entropy->saved.last_dc_val[ci] = 0;
525 } else {
526 /* Re-initialize all AC-related fields to 0 */
527 entropy->EOBRUN = 0;
528 entropy->BE = 0;
529 }
530 }
531
532
533 /*
534 * MCU encoding for DC initial scan (either spectral selection,
535 * or first pass of successive approximation).
536 */
537
538 METHODDEF(boolean)
539 encode_mcu_DC_first (j_compress_ptr cinfo, JBLOCKARRAY MCU_data)
540 {
541 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
542 register int temp, temp2;
543 register int nbits;
544 int max_coef_bits;
545 int blkn, ci, tbl;
546 ISHIFT_TEMPS
547
548 entropy->next_output_byte = cinfo->dest->next_output_byte;
549 entropy->free_in_buffer = cinfo->dest->free_in_buffer;
550
551 /* Emit restart marker if needed */
552 if (cinfo->restart_interval)
553 if (entropy->restarts_to_go == 0)
554 emit_restart_e(entropy, entropy->next_restart_num);
555
556 /* Since we're encoding a difference, the range limit is twice as much. */
557 max_coef_bits = cinfo->data_precision + 3;
558
559 /* Encode the MCU data blocks */
560 for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
561 ci = cinfo->MCU_membership[blkn];
562 tbl = cinfo->cur_comp_info[ci]->dc_tbl_no;
563
564 /* Compute the DC value after the required point transform by Al.
565 * This is simply an arithmetic right shift.
566 */
567 temp = IRIGHT_SHIFT((int) (MCU_data[blkn][0][0]), cinfo->Al);
568
569 /* DC differences are figured on the point-transformed values. */
570 if ((temp2 = temp - entropy->saved.last_dc_val[ci]) == 0) {
571 /* Count/emit the Huffman-coded symbol for the number of bits */
572 emit_dc_symbol(entropy, tbl, 0);
573
574 continue;
575 }
576
577 entropy->saved.last_dc_val[ci] = temp;
578
579 /* Encode the DC coefficient difference per section G.1.2.1 */
580 if ((temp = temp2) < 0) {
581 temp = -temp; /* temp is abs value of input */
582 /* For a negative input, want temp2 = bitwise complement of abs(input) */
583 /* This code assumes we are on a two's complement machine */
584 temp2--;
585 }
586
587 /* Find the number of bits needed for the magnitude of the coefficient */
588 nbits = 0;
589 do nbits++; /* there must be at least one 1 bit */
590 while ((temp >>= 1));
591 /* Check for out-of-range coefficient values */
592 if (nbits > max_coef_bits)
593 ERREXIT(cinfo, JERR_BAD_DCT_COEF);
594
595 /* Count/emit the Huffman-coded symbol for the number of bits */
596 emit_dc_symbol(entropy, tbl, nbits);
597
598 /* Emit that number of bits of the value, if positive, */
599 /* or the complement of its magnitude, if negative. */
600 emit_bits_e(entropy, (unsigned int) temp2, nbits);
601 }
602
603 cinfo->dest->next_output_byte = entropy->next_output_byte;
604 cinfo->dest->free_in_buffer = entropy->free_in_buffer;
605
606 /* Update restart-interval state too */
607 if (cinfo->restart_interval) {
608 if (entropy->restarts_to_go == 0) {
609 entropy->restarts_to_go = cinfo->restart_interval;
610 entropy->next_restart_num++;
611 entropy->next_restart_num &= 7;
612 }
613 entropy->restarts_to_go--;
614 }
615
616 return TRUE;
617 }
618
619
620 /*
621 * MCU encoding for AC initial scan (either spectral selection,
622 * or first pass of successive approximation).
623 */
624
625 METHODDEF(boolean)
626 encode_mcu_AC_first (j_compress_ptr cinfo, JBLOCKARRAY MCU_data)
627 {
628 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
629 const int * natural_order;
630 JBLOCKROW block;
631 register int temp, temp2;
632 register int nbits;
633 register int r, k;
634 int Se, Al, max_coef_bits;
635
636 entropy->next_output_byte = cinfo->dest->next_output_byte;
637 entropy->free_in_buffer = cinfo->dest->free_in_buffer;
638
639 /* Emit restart marker if needed */
640 if (cinfo->restart_interval)
641 if (entropy->restarts_to_go == 0)
642 emit_restart_e(entropy, entropy->next_restart_num);
643
644 Se = cinfo->Se;
645 Al = cinfo->Al;
646 natural_order = cinfo->natural_order;
647 max_coef_bits = cinfo->data_precision + 2;
648
649 /* Encode the MCU data block */
650 block = MCU_data[0];
651
652 /* Encode the AC coefficients per section G.1.2.2, fig. G.3 */
653
654 r = 0; /* r = run length of zeros */
655
656 for (k = cinfo->Ss; k <= Se; k++) {
657 if ((temp = (*block)[natural_order[k]]) == 0) {
658 r++;
659 continue;
660 }
661 /* We must apply the point transform by Al. For AC coefficients this
662 * is an integer division with rounding towards 0. To do this portably
663 * in C, we shift after obtaining the absolute value; so the code is
664 * interwoven with finding the abs value (temp) and output bits (temp2).
665 */
666 if (temp < 0) {
667 temp = -temp; /* temp is abs value of input */
668 /* Apply the point transform, and watch out for case */
669 /* that nonzero coef is zero after point transform. */
670 if ((temp >>= Al) == 0) {
671 r++;
672 continue;
673 }
674 /* For a negative coef, want temp2 = bitwise complement of abs(coef) */
675 temp2 = ~temp;
676 } else {
677 /* Apply the point transform, and watch out for case */
678 /* that nonzero coef is zero after point transform. */
679 if ((temp >>= Al) == 0) {
680 r++;
681 continue;
682 }
683 temp2 = temp;
684 }
685
686 /* Emit any pending EOBRUN */
687 if (entropy->EOBRUN > 0)
688 emit_eobrun(entropy);
689 /* if run length > 15, must emit special run-length-16 codes (0xF0) */
690 while (r > 15) {
691 emit_ac_symbol(entropy, entropy->ac_tbl_no, 0xF0);
692 r -= 16;
693 }
694
695 /* Find the number of bits needed for the magnitude of the coefficient */
696 nbits = 0;
697 do nbits++; /* there must be at least one 1 bit */
698 while ((temp >>= 1));
699 /* Check for out-of-range coefficient values */
700 if (nbits > max_coef_bits)
701 ERREXIT(cinfo, JERR_BAD_DCT_COEF);
702
703 /* Count/emit Huffman symbol for run length / number of bits */
704 emit_ac_symbol(entropy, entropy->ac_tbl_no, (r << 4) + nbits);
705
706 /* Emit that number of bits of the value, if positive, */
707 /* or the complement of its magnitude, if negative. */
708 emit_bits_e(entropy, (unsigned int) temp2, nbits);
709
710 r = 0; /* reset zero run length */
711 }
712
713 if (r > 0) { /* If there are trailing zeroes, */
714 entropy->EOBRUN++; /* count an EOB */
715 if (entropy->EOBRUN == 0x7FFF)
716 emit_eobrun(entropy); /* force it out to avoid overflow */
717 }
718
719 cinfo->dest->next_output_byte = entropy->next_output_byte;
720 cinfo->dest->free_in_buffer = entropy->free_in_buffer;
721
722 /* Update restart-interval state too */
723 if (cinfo->restart_interval) {
724 if (entropy->restarts_to_go == 0) {
725 entropy->restarts_to_go = cinfo->restart_interval;
726 entropy->next_restart_num++;
727 entropy->next_restart_num &= 7;
728 }
729 entropy->restarts_to_go--;
730 }
731
732 return TRUE;
733 }
734
735
736 /*
737 * MCU encoding for DC successive approximation refinement scan.
738 * Note: we assume such scans can be multi-component,
739 * although the spec is not very clear on the point.
740 */
741
742 METHODDEF(boolean)
743 encode_mcu_DC_refine (j_compress_ptr cinfo, JBLOCKARRAY MCU_data)
744 {
745 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
746 int Al, blkn;
747
748 entropy->next_output_byte = cinfo->dest->next_output_byte;
749 entropy->free_in_buffer = cinfo->dest->free_in_buffer;
750
751 /* Emit restart marker if needed */
752 if (cinfo->restart_interval)
753 if (entropy->restarts_to_go == 0)
754 emit_restart_e(entropy, entropy->next_restart_num);
755
756 Al = cinfo->Al;
757
758 /* Encode the MCU data blocks */
759 for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
760 /* We simply emit the Al'th bit of the DC coefficient value. */
761 emit_bits_e(entropy, (unsigned int) (MCU_data[blkn][0][0] >> Al), 1);
762 }
763
764 cinfo->dest->next_output_byte = entropy->next_output_byte;
765 cinfo->dest->free_in_buffer = entropy->free_in_buffer;
766
767 /* Update restart-interval state too */
768 if (cinfo->restart_interval) {
769 if (entropy->restarts_to_go == 0) {
770 entropy->restarts_to_go = cinfo->restart_interval;
771 entropy->next_restart_num++;
772 entropy->next_restart_num &= 7;
773 }
774 entropy->restarts_to_go--;
775 }
776
777 return TRUE;
778 }
779
780
781 /*
782 * MCU encoding for AC successive approximation refinement scan.
783 */
784
785 METHODDEF(boolean)
786 encode_mcu_AC_refine (j_compress_ptr cinfo, JBLOCKARRAY MCU_data)
787 {
788 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
789 const int * natural_order;
790 JBLOCKROW block;
791 register int temp;
792 register int r, k;
793 int Se, Al;
794 int EOB;
795 char *BR_buffer;
796 unsigned int BR;
797 int absvalues[DCTSIZE2];
798
799 entropy->next_output_byte = cinfo->dest->next_output_byte;
800 entropy->free_in_buffer = cinfo->dest->free_in_buffer;
801
802 /* Emit restart marker if needed */
803 if (cinfo->restart_interval)
804 if (entropy->restarts_to_go == 0)
805 emit_restart_e(entropy, entropy->next_restart_num);
806
807 Se = cinfo->Se;
808 Al = cinfo->Al;
809 natural_order = cinfo->natural_order;
810
811 /* Encode the MCU data block */
812 block = MCU_data[0];
813
814 /* It is convenient to make a pre-pass to determine the transformed
815 * coefficients' absolute values and the EOB position.
816 */
817 EOB = 0;
818 for (k = cinfo->Ss; k <= Se; k++) {
819 temp = (*block)[natural_order[k]];
820 /* We must apply the point transform by Al. For AC coefficients this
821 * is an integer division with rounding towards 0. To do this portably
822 * in C, we shift after obtaining the absolute value.
823 */
824 if (temp < 0)
825 temp = -temp; /* temp is abs value of input */
826 temp >>= Al; /* apply the point transform */
827 absvalues[k] = temp; /* save abs value for main pass */
828 if (temp == 1)
829 EOB = k; /* EOB = index of last newly-nonzero coef */
830 }
831
832 /* Encode the AC coefficients per section G.1.2.3, fig. G.7 */
833
834 r = 0; /* r = run length of zeros */
835 BR = 0; /* BR = count of buffered bits added now */
836 BR_buffer = entropy->bit_buffer + entropy->BE; /* Append bits to buffer */
837
838 for (k = cinfo->Ss; k <= Se; k++) {
839 if ((temp = absvalues[k]) == 0) {
840 r++;
841 continue;
842 }
843
844 /* Emit any required ZRLs, but not if they can be folded into EOB */
845 while (r > 15 && k <= EOB) {
846 /* emit any pending EOBRUN and the BE correction bits */
847 emit_eobrun(entropy);
848 /* Emit ZRL */
849 emit_ac_symbol(entropy, entropy->ac_tbl_no, 0xF0);
850 r -= 16;
851 /* Emit buffered correction bits that must be associated with ZRL */
852 emit_buffered_bits(entropy, BR_buffer, BR);
853 BR_buffer = entropy->bit_buffer; /* BE bits are gone now */
854 BR = 0;
855 }
856
857 /* If the coef was previously nonzero, it only needs a correction bit.
858 * NOTE: a straight translation of the spec's figure G.7 would suggest
859 * that we also need to test r > 15. But if r > 15, we can only get here
860 * if k > EOB, which implies that this coefficient is not 1.
861 */
862 if (temp > 1) {
863 /* The correction bit is the next bit of the absolute value. */
864 BR_buffer[BR++] = (char) (temp & 1);
865 continue;
866 }
867
868 /* Emit any pending EOBRUN and the BE correction bits */
869 emit_eobrun(entropy);
870
871 /* Count/emit Huffman symbol for run length / number of bits */
872 emit_ac_symbol(entropy, entropy->ac_tbl_no, (r << 4) + 1);
873
874 /* Emit output bit for newly-nonzero coef */
875 temp = ((*block)[natural_order[k]] < 0) ? 0 : 1;
876 emit_bits_e(entropy, (unsigned int) temp, 1);
877
878 /* Emit buffered correction bits that must be associated with this code */
879 emit_buffered_bits(entropy, BR_buffer, BR);
880 BR_buffer = entropy->bit_buffer; /* BE bits are gone now */
881 BR = 0;
882 r = 0; /* reset zero run length */
883 }
884
885 if (r > 0 || BR > 0) { /* If there are trailing zeroes, */
886 entropy->EOBRUN++; /* count an EOB */
887 entropy->BE += BR; /* concat my correction bits to older ones */
888 /* We force out the EOB if we risk either:
889 * 1. overflow of the EOB counter;
890 * 2. overflow of the correction bit buffer during the next MCU.
891 */
892 if (entropy->EOBRUN == 0x7FFF || entropy->BE > (MAX_CORR_BITS-DCTSIZE2+1))
893 emit_eobrun(entropy);
894 }
895
896 cinfo->dest->next_output_byte = entropy->next_output_byte;
897 cinfo->dest->free_in_buffer = entropy->free_in_buffer;
898
899 /* Update restart-interval state too */
900 if (cinfo->restart_interval) {
901 if (entropy->restarts_to_go == 0) {
902 entropy->restarts_to_go = cinfo->restart_interval;
903 entropy->next_restart_num++;
904 entropy->next_restart_num &= 7;
905 }
906 entropy->restarts_to_go--;
907 }
908
909 return TRUE;
910 }
911
912
913 /* Encode a single block's worth of coefficients */
914
915 LOCAL(boolean)
916 encode_one_block (working_state * state, JCOEFPTR block, int last_dc_val,
917 c_derived_tbl *dctbl, c_derived_tbl *actbl)
918 {
919 register int temp, temp2;
920 register int nbits;
921 register int r, k;
922 int Se = state->cinfo->lim_Se;
923 int max_coef_bits = state->cinfo->data_precision + 3;
924 const int * natural_order = state->cinfo->natural_order;
925
926 /* Encode the DC coefficient difference per section F.1.2.1 */
927
928 if ((temp = block[0] - last_dc_val) == 0) {
929 /* Emit the Huffman-coded symbol for the number of bits */
930 if (! emit_bits_s(state, dctbl->ehufco[0], dctbl->ehufsi[0]))
931 return FALSE;
932 } else {
933 if ((temp2 = temp) < 0) {
934 temp = -temp; /* temp is abs value of input */
935 /* For a negative input, want temp2 = bitwise complement of abs(input) */
936 /* This code assumes we are on a two's complement machine */
937 temp2--;
938 }
939
940 /* Find the number of bits needed for the magnitude of the coefficient */
941 nbits = 0;
942 do nbits++; /* there must be at least one 1 bit */
943 while ((temp >>= 1));
944 /* Check for out-of-range coefficient values.
945 * Since we're encoding a difference, the range limit is twice as much.
946 */
947 if (nbits > max_coef_bits)
948 ERREXIT(state->cinfo, JERR_BAD_DCT_COEF);
949
950 /* Emit the Huffman-coded symbol for the number of bits */
951 if (! emit_bits_s(state, dctbl->ehufco[nbits], dctbl->ehufsi[nbits]))
952 return FALSE;
953
954 /* Emit that number of bits of the value, if positive, */
955 /* or the complement of its magnitude, if negative. */
956 if (! emit_bits_s(state, (unsigned int) temp2, nbits))
957 return FALSE;
958 }
959
960 /* Encode the AC coefficients per section F.1.2.2 */
961
962 r = 0; /* r = run length of zeros */
963
964 for (k = 1; k <= Se; k++) {
965 if ((temp = block[natural_order[k]]) == 0) {
966 r++;
967 continue;
968 }
969
970 /* if run length > 15, must emit special run-length-16 codes (0xF0) */
971 while (r > 15) {
972 if (! emit_bits_s(state, actbl->ehufco[0xF0], actbl->ehufsi[0xF0]))
973 return FALSE;
974 r -= 16;
975 }
976
977 if ((temp2 = temp) < 0) {
978 temp = -temp; /* temp is abs value of input */
979 /* For a negative coef, want temp2 = bitwise complement of abs(coef) */
980 /* This code assumes we are on a two's complement machine */
981 temp2--;
982 }
983
984 /* Find the number of bits needed for the magnitude of the coefficient */
985 nbits = 0;
986 do nbits++; /* there must be at least one 1 bit */
987 while ((temp >>= 1));
988 /* Check for out-of-range coefficient values.
989 * Use ">=" instead of ">" so can use the
990 * same one larger limit from DC check here.
991 */
992 if (nbits >= max_coef_bits)
993 ERREXIT(state->cinfo, JERR_BAD_DCT_COEF);
994
995 /* Emit Huffman symbol for run length / number of bits */
996 temp = (r << 4) + nbits;
997 if (! emit_bits_s(state, actbl->ehufco[temp], actbl->ehufsi[temp]))
998 return FALSE;
999
1000 /* Emit that number of bits of the value, if positive, */
1001 /* or the complement of its magnitude, if negative. */
1002 if (! emit_bits_s(state, (unsigned int) temp2, nbits))
1003 return FALSE;
1004
1005 r = 0; /* reset zero run length */
1006 }
1007
1008 /* If the last coef(s) were zero, emit an end-of-block code */
1009 if (r > 0)
1010 if (! emit_bits_s(state, actbl->ehufco[0], actbl->ehufsi[0]))
1011 return FALSE;
1012
1013 return TRUE;
1014 }
1015
1016
1017 /*
1018 * Encode and output one MCU's worth of Huffman-compressed coefficients.
1019 */
1020
1021 METHODDEF(boolean)
1022 encode_mcu_huff (j_compress_ptr cinfo, JBLOCKARRAY MCU_data)
1023 {
1024 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
1025 working_state state;
1026 int blkn, ci;
1027 jpeg_component_info * compptr;
1028
1029 /* Load up working state */
1030 state.next_output_byte = cinfo->dest->next_output_byte;
1031 state.free_in_buffer = cinfo->dest->free_in_buffer;
1032 ASSIGN_STATE(state.cur, entropy->saved);
1033 state.cinfo = cinfo;
1034
1035 /* Emit restart marker if needed */
1036 if (cinfo->restart_interval) {
1037 if (entropy->restarts_to_go == 0)
1038 if (! emit_restart_s(&state, entropy->next_restart_num))
1039 return FALSE;
1040 }
1041
1042 /* Encode the MCU data blocks */
1043 for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
1044 ci = cinfo->MCU_membership[blkn];
1045 compptr = cinfo->cur_comp_info[ci];
1046 if (! encode_one_block(&state,
1047 MCU_data[blkn][0], state.cur.last_dc_val[ci],
1048 entropy->dc_derived_tbls[compptr->dc_tbl_no],
1049 entropy->ac_derived_tbls[compptr->ac_tbl_no]))
1050 return FALSE;
1051 /* Update last_dc_val */
1052 state.cur.last_dc_val[ci] = MCU_data[blkn][0][0];
1053 }
1054
1055 /* Completed MCU, so update state */
1056 cinfo->dest->next_output_byte = state.next_output_byte;
1057 cinfo->dest->free_in_buffer = state.free_in_buffer;
1058 ASSIGN_STATE(entropy->saved, state.cur);
1059
1060 /* Update restart-interval state too */
1061 if (cinfo->restart_interval) {
1062 if (entropy->restarts_to_go == 0) {
1063 entropy->restarts_to_go = cinfo->restart_interval;
1064 entropy->next_restart_num++;
1065 entropy->next_restart_num &= 7;
1066 }
1067 entropy->restarts_to_go--;
1068 }
1069
1070 return TRUE;
1071 }
1072
1073
1074 /*
1075 * Finish up at the end of a Huffman-compressed scan.
1076 */
1077
1078 METHODDEF(void)
1079 finish_pass_huff (j_compress_ptr cinfo)
1080 {
1081 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
1082 working_state state;
1083
1084 if (cinfo->progressive_mode) {
1085 entropy->next_output_byte = cinfo->dest->next_output_byte;
1086 entropy->free_in_buffer = cinfo->dest->free_in_buffer;
1087
1088 /* Flush out any buffered data */
1089 emit_eobrun(entropy);
1090 flush_bits_e(entropy);
1091
1092 cinfo->dest->next_output_byte = entropy->next_output_byte;
1093 cinfo->dest->free_in_buffer = entropy->free_in_buffer;
1094 } else {
1095 /* Load up working state ... flush_bits needs it */
1096 state.next_output_byte = cinfo->dest->next_output_byte;
1097 state.free_in_buffer = cinfo->dest->free_in_buffer;
1098 ASSIGN_STATE(state.cur, entropy->saved);
1099 state.cinfo = cinfo;
1100
1101 /* Flush out the last data */
1102 if (! flush_bits_s(&state))
1103 ERREXIT(cinfo, JERR_CANT_SUSPEND);
1104
1105 /* Update state */
1106 cinfo->dest->next_output_byte = state.next_output_byte;
1107 cinfo->dest->free_in_buffer = state.free_in_buffer;
1108 ASSIGN_STATE(entropy->saved, state.cur);
1109 }
1110 }
1111
1112
1113 /*
1114 * Huffman coding optimization.
1115 *
1116 * We first scan the supplied data and count the number of uses of each symbol
1117 * that is to be Huffman-coded. (This process MUST agree with the code above.)
1118 * Then we build a Huffman coding tree for the observed counts.
1119 * Symbols which are not needed at all for the particular image are not
1120 * assigned any code, which saves space in the DHT marker as well as in
1121 * the compressed data.
1122 */
1123
1124
1125 /* Process a single block's worth of coefficients */
1126
1127 LOCAL(void)
1128 htest_one_block (j_compress_ptr cinfo, JCOEFPTR block, int last_dc_val,
1129 long dc_counts[], long ac_counts[])
1130 {
1131 register int temp;
1132 register int nbits;
1133 register int r, k;
1134 int Se = cinfo->lim_Se;
1135 int max_coef_bits = cinfo->data_precision + 3;
1136 const int * natural_order = cinfo->natural_order;
1137
1138 /* Encode the DC coefficient difference per section F.1.2.1 */
1139
1140 if ((temp = block[0] - last_dc_val) == 0) {
1141 /* Count the Huffman symbol for the number of bits */
1142 dc_counts[0]++;
1143 } else {
1144 if (temp < 0)
1145 temp = -temp; /* temp is abs value of input */
1146
1147 /* Find the number of bits needed for the magnitude of the coefficient */
1148 nbits = 0;
1149 do nbits++; /* there must be at least one 1 bit */
1150 while ((temp >>= 1));
1151 /* Check for out-of-range coefficient values.
1152 * Since we're encoding a difference, the range limit is twice as much.
1153 */
1154 if (nbits > max_coef_bits)
1155 ERREXIT(cinfo, JERR_BAD_DCT_COEF);
1156
1157 /* Count the Huffman symbol for the number of bits */
1158 dc_counts[nbits]++;
1159 }
1160
1161 /* Encode the AC coefficients per section F.1.2.2 */
1162
1163 r = 0; /* r = run length of zeros */
1164
1165 for (k = 1; k <= Se; k++) {
1166 if ((temp = block[natural_order[k]]) == 0) {
1167 r++;
1168 continue;
1169 }
1170
1171 /* if run length > 15, must emit special run-length-16 codes (0xF0) */
1172 while (r > 15) {
1173 ac_counts[0xF0]++;
1174 r -= 16;
1175 }
1176
1177 if (temp < 0)
1178 temp = -temp; /* temp is abs value of input */
1179
1180 /* Find the number of bits needed for the magnitude of the coefficient */
1181 nbits = 0;
1182 do nbits++; /* there must be at least one 1 bit */
1183 while ((temp >>= 1));
1184 /* Check for out-of-range coefficient values.
1185 * Use ">=" instead of ">" so can use the
1186 * same one larger limit from DC check here.
1187 */
1188 if (nbits >= max_coef_bits)
1189 ERREXIT(cinfo, JERR_BAD_DCT_COEF);
1190
1191 /* Count Huffman symbol for run length / number of bits */
1192 ac_counts[(r << 4) + nbits]++;
1193
1194 r = 0; /* reset zero run length */
1195 }
1196
1197 /* If the last coef(s) were zero, emit an end-of-block code */
1198 if (r > 0)
1199 ac_counts[0]++;
1200 }
1201
1202
1203 /*
1204 * Trial-encode one MCU's worth of Huffman-compressed coefficients.
1205 * No data is actually output, so no suspension return is possible.
1206 */
1207
1208 METHODDEF(boolean)
1209 encode_mcu_gather (j_compress_ptr cinfo, JBLOCKARRAY MCU_data)
1210 {
1211 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
1212 int blkn, ci;
1213 jpeg_component_info * compptr;
1214
1215 /* Take care of restart intervals if needed */
1216 if (cinfo->restart_interval) {
1217 if (entropy->restarts_to_go == 0) {
1218 /* Re-initialize DC predictions to 0 */
1219 for (ci = 0; ci < cinfo->comps_in_scan; ci++)
1220 entropy->saved.last_dc_val[ci] = 0;
1221 /* Update restart state */
1222 entropy->restarts_to_go = cinfo->restart_interval;
1223 }
1224 entropy->restarts_to_go--;
1225 }
1226
1227 for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
1228 ci = cinfo->MCU_membership[blkn];
1229 compptr = cinfo->cur_comp_info[ci];
1230 htest_one_block(cinfo, MCU_data[blkn][0], entropy->saved.last_dc_val[ci],
1231 entropy->dc_count_ptrs[compptr->dc_tbl_no],
1232 entropy->ac_count_ptrs[compptr->ac_tbl_no]);
1233 entropy->saved.last_dc_val[ci] = MCU_data[blkn][0][0];
1234 }
1235
1236 return TRUE;
1237 }
1238
1239
1240 /*
1241 * Generate the best Huffman code table for the given counts, fill htbl.
1242 *
1243 * The JPEG standard requires that no symbol be assigned a codeword of all
1244 * one bits (so that padding bits added at the end of a compressed segment
1245 * can't look like a valid code). Because of the canonical ordering of
1246 * codewords, this just means that there must be an unused slot in the
1247 * longest codeword length category. Section K.2 of the JPEG spec suggests
1248 * reserving such a slot by pretending that symbol 256 is a valid symbol
1249 * with count 1. In theory that's not optimal; giving it count zero but
1250 * including it in the symbol set anyway should give a better Huffman code.
1251 * But the theoretically better code actually seems to come out worse in
1252 * practice, because it produces more all-ones bytes (which incur stuffed
1253 * zero bytes in the final file). In any case the difference is tiny.
1254 *
1255 * The JPEG standard requires Huffman codes to be no more than 16 bits long.
1256 * If some symbols have a very small but nonzero probability, the Huffman tree
1257 * must be adjusted to meet the code length restriction. We currently use
1258 * the adjustment method suggested in JPEG section K.2. This method is *not*
1259 * optimal; it may not choose the best possible limited-length code. But
1260 * typically only very-low-frequency symbols will be given less-than-optimal
1261 * lengths, so the code is almost optimal. Experimental comparisons against
1262 * an optimal limited-length-code algorithm indicate that the difference is
1263 * microscopic --- usually less than a hundredth of a percent of total size.
1264 * So the extra complexity of an optimal algorithm doesn't seem worthwhile.
1265 */
1266
1267 LOCAL(void)
1268 jpeg_gen_optimal_table (j_compress_ptr cinfo, JHUFF_TBL * htbl, long freq[])
1269 {
1270 #define MAX_CLEN 32 /* assumed maximum initial code length */
1271 UINT8 bits[MAX_CLEN+1]; /* bits[k] = # of symbols with code length k */
1272 int codesize[257]; /* codesize[k] = code length of symbol k */
1273 int others[257]; /* next symbol in current branch of tree */
1274 int c1, c2, i, j;
1275 UINT8 *p;
1276 long v;
1277
1278 freq[256] = 1; /* make sure 256 has a nonzero count */
1279 /* Including the pseudo-symbol 256 in the Huffman procedure guarantees
1280 * that no real symbol is given code-value of all ones, because 256
1281 * will be placed last in the largest codeword category.
1282 * In the symbol list build procedure this element serves as sentinel
1283 * for the zero run loop.
1284 */
1285
1286 #ifndef DONT_USE_FANCY_HUFF_OPT
1287
1288 /* Build list of symbols sorted in order of descending frequency */
1289 /* This approach has several benefits (thank to John Korejwa for the idea):
1290 * 1.
1291 * If a codelength category is split during the length limiting procedure
1292 * below, the feature that more frequent symbols are assigned shorter
1293 * codewords remains valid for the adjusted code.
1294 * 2.
1295 * To reduce consecutive ones in a Huffman data stream (thus reducing the
1296 * number of stuff bytes in JPEG) it is preferable to follow 0 branches
1297 * (and avoid 1 branches) as much as possible. This is easily done by
1298 * assigning symbols to leaves of the Huffman tree in order of decreasing
1299 * frequency, with no secondary sort based on codelengths.
1300 * 3.
1301 * The symbol list can be built independently from the assignment of code
1302 * lengths by the Huffman procedure below.
1303 * Note: The symbol list build procedure must be performed first, because
1304 * the Huffman procedure assigning the codelengths clobbers the frequency
1305 * counts!
1306 */
1307
1308 /* Here we use the others array as a linked list of nonzero frequencies
1309 * to be sorted. Already sorted elements are removed from the list.
1310 */
1311
1312 /* Building list */
1313
1314 /* This item does not correspond to a valid symbol frequency and is used
1315 * as starting index.
1316 */
1317 j = 256;
1318
1319 for (i = 0;; i++) {
1320 if (freq[i] == 0) /* skip zero frequencies */
1321 continue;
1322 if (i > 255)
1323 break;
1324 others[j] = i; /* this symbol value */
1325 j = i; /* previous symbol value */
1326 }
1327 others[j] = -1; /* mark end of list */
1328
1329 /* Sorting list */
1330
1331 p = htbl->huffval;
1332 while ((c1 = others[256]) >= 0) {
1333 v = freq[c1];
1334 i = c1; /* first symbol value */
1335 j = 256; /* pseudo symbol value for starting index */
1336 while ((c2 = others[c1]) >= 0) {
1337 if (freq[c2] > v) {
1338 v = freq[c2];
1339 i = c2; /* this symbol value */
1340 j = c1; /* previous symbol value */
1341 }
1342 c1 = c2;
1343 }
1344 others[j] = others[i]; /* remove this symbol i from list */
1345 *p++ = (UINT8) i;
1346 }
1347
1348 #endif /* DONT_USE_FANCY_HUFF_OPT */
1349
1350 /* This algorithm is explained in section K.2 of the JPEG standard */
1351
1352 MEMZERO(bits, SIZEOF(bits));
1353 MEMZERO(codesize, SIZEOF(codesize));
1354 for (i = 0; i < 257; i++)
1355 others[i] = -1; /* init links to empty */
1356
1357 /* Huffman's basic algorithm to assign optimal code lengths to symbols */
1358
1359 for (;;) {
1360 /* Find the smallest nonzero frequency, set c1 = its symbol */
1361 /* In case of ties, take the larger symbol number */
1362 c1 = -1;
1363 v = 1000000000L;
1364 for (i = 0; i <= 256; i++) {
1365 if (freq[i] && freq[i] <= v) {
1366 v = freq[i];
1367 c1 = i;
1368 }
1369 }
1370
1371 /* Find the next smallest nonzero frequency, set c2 = its symbol */
1372 /* In case of ties, take the larger symbol number */
1373 c2 = -1;
1374 v = 1000000000L;
1375 for (i = 0; i <= 256; i++) {
1376 if (freq[i] && freq[i] <= v && i != c1) {
1377 v = freq[i];
1378 c2 = i;
1379 }
1380 }
1381
1382 /* Done if we've merged everything into one frequency */
1383 if (c2 < 0)
1384 break;
1385
1386 /* Else merge the two counts/trees */
1387 freq[c1] += freq[c2];
1388 freq[c2] = 0;
1389
1390 /* Increment the codesize of everything in c1's tree branch */
1391 codesize[c1]++;
1392 while (others[c1] >= 0) {
1393 c1 = others[c1];
1394 codesize[c1]++;
1395 }
1396
1397 others[c1] = c2; /* chain c2 onto c1's tree branch */
1398
1399 /* Increment the codesize of everything in c2's tree branch */
1400 codesize[c2]++;
1401 while (others[c2] >= 0) {
1402 c2 = others[c2];
1403 codesize[c2]++;
1404 }
1405 }
1406
1407 /* Now count the number of symbols of each code length */
1408 for (i = 0; i <= 256; i++) {
1409 if (codesize[i]) {
1410 /* The JPEG standard seems to think that this can't happen, */
1411 /* but I'm paranoid... */
1412 if (codesize[i] > MAX_CLEN)
1413 ERREXIT(cinfo, JERR_HUFF_CLEN_OUTOFBOUNDS);
1414
1415 bits[codesize[i]]++;
1416 }
1417 }
1418
1419 /* JPEG doesn't allow symbols with code lengths over 16 bits, so if the pure
1420 * Huffman procedure assigned any such lengths, we must adjust the coding.
1421 * Here is what the JPEG spec says about how this next bit works:
1422 * Since symbols are paired for the longest Huffman code, the symbols are
1423 * removed from this length category two at a time. The prefix for the pair
1424 * (which is one bit shorter) is allocated to one of the pair; then,
1425 * skipping the BITS entry for that prefix length, a code word from the next
1426 * shortest nonzero BITS entry is converted into a prefix for two code words
1427 * one bit longer.
1428 */
1429
1430 for (i = MAX_CLEN; i > 16; i--) {
1431 while (bits[i] > 0) {
1432 j = i - 2; /* find length of new prefix to be used */
1433 while (bits[j] == 0) {
1434 if (j == 0)
1435 ERREXIT(cinfo, JERR_HUFF_CLEN_OUTOFBOUNDS);
1436 j--;
1437 }
1438
1439 bits[i] -= 2; /* remove two symbols */
1440 bits[i-1]++; /* one goes in this length */
1441 bits[j+1] += 2; /* two new symbols in this length */
1442 bits[j]--; /* symbol of this length is now a prefix */
1443 }
1444 }
1445
1446 /* Remove the count for the pseudo-symbol 256 from the largest codelength */
1447 while (bits[i] == 0) /* find largest codelength still in use */
1448 i--;
1449 bits[i]--;
1450
1451 /* Return final symbol counts (only for lengths 0..16) */
1452 MEMCOPY(htbl->bits, bits, SIZEOF(htbl->bits));
1453
1454 #ifdef DONT_USE_FANCY_HUFF_OPT
1455
1456 /* Return a list of the symbols sorted by code length */
1457 /* Note: Due to the codelength changes made above, it can happen
1458 * that more frequent symbols are assigned longer codewords.
1459 */
1460 p = htbl->huffval;
1461 for (i = 1; i <= MAX_CLEN; i++) {
1462 for (j = 0; j <= 255; j++) {
1463 if (codesize[j] == i) {
1464 *p++ = (UINT8) j;
1465 }
1466 }
1467 }
1468
1469 #endif /* DONT_USE_FANCY_HUFF_OPT */
1470
1471 /* Set sent_table FALSE so updated table will be written to JPEG file. */
1472 htbl->sent_table = FALSE;
1473 }
1474
1475
1476 /*
1477 * Finish up a statistics-gathering pass and create the new Huffman tables.
1478 */
1479
1480 METHODDEF(void)
1481 finish_pass_gather (j_compress_ptr cinfo)
1482 {
1483 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
1484 int ci, tbl;
1485 jpeg_component_info * compptr;
1486 JHUFF_TBL **htblptr;
1487 boolean did_dc[NUM_HUFF_TBLS];
1488 boolean did_ac[NUM_HUFF_TBLS];
1489
1490 if (cinfo->progressive_mode)
1491 /* Flush out buffered data (all we care about is counting the EOB symbol) */
1492 emit_eobrun(entropy);
1493
1494 /* It's important not to apply jpeg_gen_optimal_table more than once
1495 * per table, because it clobbers the input frequency counts!
1496 */
1497 MEMZERO(did_dc, SIZEOF(did_dc));
1498 MEMZERO(did_ac, SIZEOF(did_ac));
1499
1500 for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
1501 compptr = cinfo->cur_comp_info[ci];
1502 /* DC needs no table for refinement scan */
1503 if (cinfo->Ss == 0 && cinfo->Ah == 0) {
1504 tbl = compptr->dc_tbl_no;
1505 if (! did_dc[tbl]) {
1506 htblptr = & cinfo->dc_huff_tbl_ptrs[tbl];
1507 if (*htblptr == NULL)
1508 *htblptr = jpeg_alloc_huff_table((j_common_ptr) cinfo);
1509 jpeg_gen_optimal_table(cinfo, *htblptr, entropy->dc_count_ptrs[tbl]);
1510 did_dc[tbl] = TRUE;
1511 }
1512 }
1513 /* AC needs no table when not present */
1514 if (cinfo->Se) {
1515 tbl = compptr->ac_tbl_no;
1516 if (! did_ac[tbl]) {
1517 htblptr = & cinfo->ac_huff_tbl_ptrs[tbl];
1518 if (*htblptr == NULL)
1519 *htblptr = jpeg_alloc_huff_table((j_common_ptr) cinfo);
1520 jpeg_gen_optimal_table(cinfo, *htblptr, entropy->ac_count_ptrs[tbl]);
1521 did_ac[tbl] = TRUE;
1522 }
1523 }
1524 }
1525 }
1526
1527
1528 /*
1529 * Initialize for a Huffman-compressed scan.
1530 * If gather_statistics is TRUE, we do not output anything during the scan,
1531 * just count the Huffman symbols used and generate Huffman code tables.
1532 */
1533
1534 METHODDEF(void)
1535 start_pass_huff (j_compress_ptr cinfo, boolean gather_statistics)
1536 {
1537 huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
1538 int ci, tbl;
1539 jpeg_component_info * compptr;
1540
1541 if (gather_statistics)
1542 entropy->pub.finish_pass = finish_pass_gather;
1543 else
1544 entropy->pub.finish_pass = finish_pass_huff;
1545
1546 if (cinfo->progressive_mode) {
1547 entropy->cinfo = cinfo;
1548 entropy->gather_statistics = gather_statistics;
1549
1550 /* We assume jcmaster.c already validated the scan parameters. */
1551
1552 /* Select execution routine */
1553 if (cinfo->Ah == 0) {
1554 if (cinfo->Ss == 0)
1555 entropy->pub.encode_mcu = encode_mcu_DC_first;
1556 else
1557 entropy->pub.encode_mcu = encode_mcu_AC_first;
1558 } else {
1559 if (cinfo->Ss == 0)
1560 entropy->pub.encode_mcu = encode_mcu_DC_refine;
1561 else {
1562 entropy->pub.encode_mcu = encode_mcu_AC_refine;
1563 /* AC refinement needs a correction bit buffer */
1564 if (entropy->bit_buffer == NULL)
1565 entropy->bit_buffer = (char *) (*cinfo->mem->alloc_small)
1566 ((j_common_ptr) cinfo, JPOOL_IMAGE, MAX_CORR_BITS * SIZEOF(char));
1567 }
1568 }
1569
1570 /* Initialize AC stuff */
1571 entropy->ac_tbl_no = cinfo->cur_comp_info[0]->ac_tbl_no;
1572 entropy->EOBRUN = 0;
1573 entropy->BE = 0;
1574 } else {
1575 if (gather_statistics)
1576 entropy->pub.encode_mcu = encode_mcu_gather;
1577 else
1578 entropy->pub.encode_mcu = encode_mcu_huff;
1579 }
1580
1581 for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
1582 compptr = cinfo->cur_comp_info[ci];
1583 /* DC needs no table for refinement scan */
1584 if (cinfo->Ss == 0 && cinfo->Ah == 0) {
1585 tbl = compptr->dc_tbl_no;
1586 if (gather_statistics) {
1587 /* Check for invalid table index */
1588 /* (make_c_derived_tbl does this in the other path) */
1589 if (tbl < 0 || tbl >= NUM_HUFF_TBLS)
1590 ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tbl);
1591 /* Allocate and zero the statistics tables */
1592 /* Note that jpeg_gen_optimal_table expects 257 entries in each table! */
1593 if (entropy->dc_count_ptrs[tbl] == NULL)
1594 entropy->dc_count_ptrs[tbl] = (long *) (*cinfo->mem->alloc_small)
1595 ((j_common_ptr) cinfo, JPOOL_IMAGE, 257 * SIZEOF(long));
1596 MEMZERO(entropy->dc_count_ptrs[tbl], 257 * SIZEOF(long));
1597 } else {
1598 /* Compute derived values for Huffman tables */
1599 /* We may do this more than once for a table, but it's not expensive */
1600 jpeg_make_c_derived_tbl(cinfo, TRUE, tbl,
1601 & entropy->dc_derived_tbls[tbl]);
1602 }
1603 /* Initialize DC predictions to 0 */
1604 entropy->saved.last_dc_val[ci] = 0;
1605 }
1606 /* AC needs no table when not present */
1607 if (cinfo->Se) {
1608 tbl = compptr->ac_tbl_no;
1609 if (gather_statistics) {
1610 if (tbl < 0 || tbl >= NUM_HUFF_TBLS)
1611 ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tbl);
1612 if (entropy->ac_count_ptrs[tbl] == NULL)
1613 entropy->ac_count_ptrs[tbl] = (long *) (*cinfo->mem->alloc_small)
1614 ((j_common_ptr) cinfo, JPOOL_IMAGE, 257 * SIZEOF(long));
1615 MEMZERO(entropy->ac_count_ptrs[tbl], 257 * SIZEOF(long));
1616 } else {
1617 jpeg_make_c_derived_tbl(cinfo, FALSE, tbl,
1618 & entropy->ac_derived_tbls[tbl]);
1619 }
1620 }
1621 }
1622
1623 /* Initialize bit buffer to empty */
1624 entropy->saved.put_buffer = 0;
1625 entropy->saved.put_bits = 0;
1626
1627 /* Initialize restart stuff */
1628 entropy->restarts_to_go = cinfo->restart_interval;
1629 entropy->next_restart_num = 0;
1630 }
1631
1632
1633 /*
1634 * Module initialization routine for Huffman entropy encoding.
1635 */
1636
1637 GLOBAL(void)
1638 jinit_huff_encoder (j_compress_ptr cinfo)
1639 {
1640 huff_entropy_ptr entropy;
1641 int i;
1642
1643 entropy = (huff_entropy_ptr) (*cinfo->mem->alloc_small)
1644 ((j_common_ptr) cinfo, JPOOL_IMAGE, SIZEOF(huff_entropy_encoder));
1645 cinfo->entropy = &entropy->pub;
1646 entropy->pub.start_pass = start_pass_huff;
1647
1648 /* Mark tables unallocated */
1649 for (i = 0; i < NUM_HUFF_TBLS; i++) {
1650 entropy->dc_derived_tbls[i] = entropy->ac_derived_tbls[i] = NULL;
1651 entropy->dc_count_ptrs[i] = entropy->ac_count_ptrs[i] = NULL;
1652 }
1653
1654 if (cinfo->progressive_mode)
1655 entropy->bit_buffer = NULL; /* needed only in AC refinement scan */
1656 }