compiler/libjpeg/main/jdhuff.c

   1 /*
   2     $Id$
   3 */
   4
   5 /*
   6  * jdhuff.c
   7  *
   8  * Copyright (C) 1991-1998, Thomas G. Lane.
   9  * This file is part of the Independent JPEG Group's software.
  10  * For conditions of distribution and use, see the accompanying README file.
  11  *
  12  * This file contains Huffman entropy decoding routines which are shared
  13  * by the sequential, progressive and lossless decoders.
  14  */
  15
  16 #define JPEG_INTERNALS
  17 #include "jinclude.h"
  18 #include "jpeglib.h"
  19 #include "jlossy.h"             /* Private declarations for lossy codec */
  20 #include "jlossls.h"            /* Private declarations for lossless codec */
  21 #include "jdhuff.h"             /* Declarations shared with jd*huff.c */
  22
  23
  24 /*
  25  * Compute the derived values for a Huffman table.
  26  * This routine also performs some validation checks on the table.
  27  */
  28
  29 JGLOBAL(void)
  30 jpeg_make_d_derived_tbl (j_decompress_ptr cinfo, boolean isDC, int tblno,
  31                          d_derived_tbl ** pdtbl)
  32 {
  33   JHUFF_TBL *htbl;
  34   d_derived_tbl *dtbl;
  35   int p, i, l, si, numsymbols;
  36   int lookbits, ctr;
  37   char huffsize[257];
  38   unsigned int huffcode[257];
  39   unsigned int code;
  40
  41   /* Note that huffsize[] and huffcode[] are filled in code-length order,
  42    * paralleling the order of the symbols themselves in htbl->huffval[].
  43    */
  44
  45   /* Find the input Huffman table */
  46   if (tblno < 0 || tblno >= NUM_HUFF_TBLS)
  47     ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno);
  48   htbl =
  49     isDC ? cinfo->dc_huff_tbl_ptrs[tblno] : cinfo->ac_huff_tbl_ptrs[tblno];
  50   if (htbl == NULL)
  51     ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno);
  52
  53   /* Allocate a workspace if we haven't already done so. */
  54   if (*pdtbl == NULL)
  55     *pdtbl = (d_derived_tbl *)
  56       (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
  57                                   SIZEOF(d_derived_tbl));
  58   dtbl = *pdtbl;
  59   dtbl->pub = htbl;             /* fill in back link */
  60
  61   /* Figure C.1: make table of Huffman code length for each symbol */
  62
  63   p = 0;
  64   for (l = 1; l <= 16; l++) {
  65     i = (int) htbl->bits[l];
  66     if (i < 0 || p + i > 256)   /* protect against table overrun */
  67       ERREXIT(cinfo, JERR_BAD_HUFF_TABLE);
  68     while (i--)
  69       huffsize[p++] = (char) l;
  70   }
  71   huffsize[p] = 0;
  72   numsymbols = p;
  73
  74   /* Figure C.2: generate the codes themselves */
  75   /* We also validate that the counts represent a legal Huffman code tree. */
  76
  77   code = 0;
  78   si = huffsize[0];
  79   p = 0;
  80   while (huffsize[p]) {
  81     while (((int) huffsize[p]) == si) {
  82       huffcode[p++] = code;
  83       code++;
  84     }
  85     /* code is now 1 more than the last code used for codelength si; but
  86      * it must still fit in si bits, since no code is allowed to be all ones.
  87      */
  88     if (((INT32) code) >= (((INT32) 1) << si))
  89       ERREXIT(cinfo, JERR_BAD_HUFF_TABLE);
  90     code <<= 1;
  91     si++;
  92   }
  93
  94   /* Figure F.15: generate decoding tables for bit-sequential decoding */
  95
  96   p = 0;
  97   for (l = 1; l <= 16; l++) {
  98     if (htbl->bits[l]) {
  99       /* valoffset[l] = huffval[] index of 1st symbol of code length l,
 100        * minus the minimum code of length l
 101        */
 102       dtbl->valoffset[l] = (INT32) p - (INT32) huffcode[p];
 103       p += htbl->bits[l];
 104       dtbl->maxcode[l] = huffcode[p-1]; /* maximum code of length l */
 105     } else {
 106       dtbl->maxcode[l] = -1;    /* -1 if no codes of this length */
 107     }
 108   }
 109   dtbl->maxcode[17] = 0xFFFFFL; /* ensures jpeg_huff_decode terminates */
 110
 111   /* Compute lookahead tables to speed up decoding.
 112    * First we set all the table entries to 0, indicating "too long";
 113    * then we iterate through the Huffman codes that are short enough and
 114    * fill in all the entries that correspond to bit sequences starting
 115    * with that code.
 116    */
 117
 118   MEMZERO(dtbl->look_nbits, SIZEOF(dtbl->look_nbits));
 119
 120   p = 0;
 121   for (l = 1; l <= HUFF_LOOKAHEAD; l++) {
 122     for (i = 1; i <= (int) htbl->bits[l]; i++, p++) {
 123       /* l = current code's length, p = its index in huffcode[] & huffval[]. */
 124       /* Generate left-justified code followed by all possible bit sequences */
 125       lookbits = huffcode[p] << (HUFF_LOOKAHEAD-l);
 126       for (ctr = 1 << (HUFF_LOOKAHEAD-l); ctr > 0; ctr--) {
 127         dtbl->look_nbits[lookbits] = l;
 128         dtbl->look_sym[lookbits] = htbl->huffval[p];
 129         lookbits++;
 130       }
 131     }
 132   }
 133
 134   /* Validate symbols as being reasonable.
 135    * For AC tables, we make no check, but accept all byte values 0..255.
 136    * For DC tables, we require the symbols to be in range 0..16.
 137    * (Tighter bounds could be applied depending on the data depth and mode,
 138    * but this is sufficient to ensure safe decoding.)
 139    */
 140   if (isDC) {
 141     for (i = 0; i < numsymbols; i++) {
 142       int sym = htbl->huffval[i];
 143       if (sym < 0 || sym > 16)
 144         ERREXIT(cinfo, JERR_BAD_HUFF_TABLE);
 145     }
 146   }
 147 }
 148
 149
 150 /*
 151  * Out-of-line code for bit fetching.
 152  * See jdhuff.h for info about usage.
 153  * Note: current values of get_buffer and bits_left are passed as parameters,
 154  * but are returned in the corresponding fields of the state struct.
 155  *
 156  * On most machines MIN_GET_BITS should be 25 to allow the full 32-bit width
 157  * of get_buffer to be used.  (On machines with wider words, an even larger
 158  * buffer could be used.)  However, on some machines 32-bit shifts are
 159  * quite slow and take time proportional to the number of places shifted.
 160  * (This is true with most PC compilers, for instance.)  In this case it may
 161  * be a win to set MIN_GET_BITS to the minimum value of 15.  This reduces the
 162  * average shift distance at the cost of more calls to jpeg_fill_bit_buffer.
 163  */
 164
 165 #ifdef SLOW_SHIFT_32
 166 #define MIN_GET_BITS  15        /* minimum allowable value */
 167 #else
 168 #define MIN_GET_BITS  (BIT_BUF_SIZE-7)
 169 #endif
 170
 171
 172 JGLOBAL(boolean)
 173 jpeg_fill_bit_buffer (bitread_working_state * state,
 174                       register bit_buf_type get_buffer, register int bits_left,
 175                       int nbits)
 176 /* Load up the bit buffer to a depth of at least nbits */
 177 {
 178   /* Copy heavily used state fields into locals (hopefully registers) */
 179   register const JOCTET * next_input_byte = state->next_input_byte;
 180   register size_t bytes_in_buffer = state->bytes_in_buffer;
 181   j_decompress_ptr cinfo = state->cinfo;
 182
 183   /* Attempt to load at least MIN_GET_BITS bits into get_buffer. */
 184   /* (It is assumed that no request will be for more than that many bits.) */
 185   /* We fail to do so only if we hit a marker or are forced to suspend. */
 186
 187   if (cinfo->unread_marker == 0) {      /* cannot advance past a marker */
 188     while (bits_left < MIN_GET_BITS) {
 189       register int c;
 190
 191       /* Attempt to read a byte */
 192       if (bytes_in_buffer == 0) {
 193         if (! (*cinfo->src->fill_input_buffer) (cinfo))
 194           return FALSE;
 195         next_input_byte = cinfo->src->next_input_byte;
 196         bytes_in_buffer = cinfo->src->bytes_in_buffer;
 197       }
 198       bytes_in_buffer--;
 199       c = GETJOCTET(*next_input_byte++);
 200
 201       /* If it's 0xFF, check and discard stuffed zero byte */
 202       if (c == 0xFF) {
 203         /* Loop here to discard any padding FF's on terminating marker,
 204          * so that we can save a valid unread_marker value.  NOTE: we will
 205          * accept multiple FF's followed by a 0 as meaning a single FF data
 206          * byte.  This data pattern is not valid according to the standard.
 207          */
 208         do {
 209           if (bytes_in_buffer == 0) {
 210             if (! (*cinfo->src->fill_input_buffer) (cinfo))
 211               return FALSE;
 212             next_input_byte = cinfo->src->next_input_byte;
 213             bytes_in_buffer = cinfo->src->bytes_in_buffer;
 214           }
 215           bytes_in_buffer--;
 216           c = GETJOCTET(*next_input_byte++);
 217         } while (c == 0xFF);
 218
 219         if (c == 0) {
 220           /* Found FF/00, which represents an FF data byte */
 221           c = 0xFF;
 222         } else {
 223           /* Oops, it's actually a marker indicating end of compressed data.
 224            * Save the marker code for later use.
 225            * Fine point: it might appear that we should save the marker into
 226            * bitread working state, not straight into permanent state.  But
 227            * once we have hit a marker, we cannot need to suspend within the
 228            * current MCU, because we will read no more bytes from the data
 229            * source.  So it is OK to update permanent state right away.
 230            */
 231           cinfo->unread_marker = c;
 232           /* See if we need to insert some fake zero bits. */
 233           goto no_more_bytes;
 234         }
 235       }
 236
 237       /* OK, load c into get_buffer */
 238       get_buffer = (get_buffer << 8) | c;
 239       bits_left += 8;
 240     } /* end while */
 241   } else {
 242   no_more_bytes:
 243     /* We get here if we've read the marker that terminates the compressed
 244      * data segment.  There should be enough bits in the buffer register
 245      * to satisfy the request; if so, no problem.
 246      */
 247     if (nbits > bits_left) {
 248       /* Uh-oh.  Report corrupted data to user and stuff zeroes into
 249        * the data stream, so that we can produce some kind of image.
 250        * We use a nonvolatile flag to ensure that only one warning message
 251        * appears per data segment.
 252        */
 253       huffd_common_ptr huffd;
 254       if (cinfo->process == JPROC_LOSSLESS)
 255         huffd = (huffd_common_ptr) ((j_lossless_d_ptr) cinfo->codec)->entropy_private;
 256       else
 257         huffd = (huffd_common_ptr) ((j_lossy_d_ptr) cinfo->codec)->entropy_private;
 258       if (! huffd->insufficient_data) {
 259         WARNMS(cinfo, JWRN_HIT_MARKER);
 260         huffd->insufficient_data = TRUE;
 261       }
 262       /* Fill the buffer with zero bits */
 263       get_buffer <<= MIN_GET_BITS - bits_left;
 264       bits_left = MIN_GET_BITS;
 265     }
 266   }
 267
 268   /* Unload the local registers */
 269   state->next_input_byte = next_input_byte;
 270   state->bytes_in_buffer = bytes_in_buffer;
 271   state->get_buffer = get_buffer;
 272   state->bits_left = bits_left;
 273
 274   return TRUE;
 275 }
 276
 277
 278 /*
 279  * Out-of-line code for Huffman code decoding.
 280  * See jdhuff.h for info about usage.
 281  */
 282
 283 JGLOBAL(int)
 284 jpeg_huff_decode (bitread_working_state * state,
 285                   register bit_buf_type get_buffer, register int bits_left,
 286                   d_derived_tbl * htbl, int min_bits)
 287 {
 288   register int l = min_bits;
 289   register INT32 code;
 290
 291   /* HUFF_DECODE has determined that the code is at least min_bits */
 292   /* bits long, so fetch that many bits in one swoop. */
 293
 294   CHECK_BIT_BUFFER(*state, l, return -1);
 295   code = GET_BITS(l);
 296
 297   /* Collect the rest of the Huffman code one bit at a time. */
 298   /* This is per Figure F.16 in the JPEG spec. */
 299
 300   while (code > htbl->maxcode[l]) {
 301     code <<= 1;
 302     CHECK_BIT_BUFFER(*state, 1, return -1);
 303     code |= GET_BITS(1);
 304     l++;
 305   }
 306
 307   /* Unload the local registers */
 308   state->get_buffer = get_buffer;
 309   state->bits_left = bits_left;
 310
 311   /* With garbage input we may reach the sentinel value l = 17. */
 312
 313   if (l > 16) {
 314     WARNMS(state->cinfo, JWRN_HUFF_BAD_CODE);
 315     return 0;                   /* fake a zero as the safest result */
 316   }
 317
 318   return htbl->pub->huffval[ (int) (code + htbl->valoffset[l]) ];
 319 }