libavcodec/mlz.c

   1 /*
   2  * Copyright (c) 2016 Umair Khan <omerjerk@gmail.com>
   3  *
   4  * This file is part of FFmpeg.
   5  *
   6  * FFmpeg is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * FFmpeg is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with FFmpeg; if not, write to the Free Software
  18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  19  */
  20
  21 #include "mlz.h"
  22
  23 av_cold int ff_mlz_init_dict(void *context, MLZ *mlz)
  24 {
  25     mlz->dict = av_mallocz(TABLE_SIZE * sizeof(*mlz->dict));
  26     if (!mlz->dict)
  27         return AVERROR(ENOMEM);
  28
  29     mlz->flush_code            = FLUSH_CODE;
  30     mlz->current_dic_index_max = DIC_INDEX_INIT;
  31     mlz->dic_code_bit          = CODE_BIT_INIT;
  32     mlz->bump_code             = (DIC_INDEX_INIT - 1);
  33     mlz->next_code             = FIRST_CODE;
  34     mlz->freeze_flag           = 0;
  35     mlz->context               = context;
  36
  37     return 0;
  38 }
  39
  40 av_cold void ff_mlz_flush_dict(MLZ *mlz) {
  41     MLZDict *dict = mlz->dict;
  42     int i;
  43     for ( i = 0; i < TABLE_SIZE; i++ ) {
  44         dict[i].string_code = CODE_UNSET;
  45         dict[i].parent_code = CODE_UNSET;
  46         dict[i].match_len = 0;
  47     }
  48     mlz->current_dic_index_max = DIC_INDEX_INIT;
  49     mlz->dic_code_bit          = CODE_BIT_INIT;  // DicCodeBitInit;
  50     mlz->bump_code             = mlz->current_dic_index_max - 1;
  51     mlz->next_code             = FIRST_CODE;
  52     mlz->freeze_flag           = 0;
  53 }
  54
  55 static void set_new_entry_dict(MLZDict* dict, int string_code, int parent_code, int char_code) {
  56     dict[string_code].parent_code = parent_code;
  57     dict[string_code].string_code = string_code;
  58     dict[string_code].char_code   = char_code;
  59     if (parent_code < FIRST_CODE) {
  60         dict[string_code].match_len = 2;
  61     } else {
  62         dict[string_code].match_len = (dict[parent_code].match_len) + 1;
  63     }
  64 }
  65
  66 static int decode_string(MLZ* mlz, unsigned char *buff, int string_code, int *first_char_code, unsigned long bufsize) {
  67     MLZDict* dict = mlz->dict;
  68     unsigned long count, offset;
  69     int current_code, parent_code, tmp_code;
  70
  71     count            = 0;
  72     current_code     = string_code;
  73     *first_char_code = CODE_UNSET;
  74
  75     while (count < bufsize) {
  76         switch (current_code) {
  77         case CODE_UNSET:
  78             return count;
  79             break;
  80         default:
  81             if (current_code < FIRST_CODE) {
  82                 *first_char_code = current_code;
  83                 buff[0] = current_code;
  84                 count++;
  85                 return count;
  86             } else {
  87                 offset  = dict[current_code].match_len - 1;
  88                 tmp_code = dict[current_code].char_code;
  89                 if (offset >= bufsize) {
  90                     av_log(mlz->context, AV_LOG_ERROR, "MLZ offset error.\n");
  91                     return count;
  92                 }
  93                 buff[offset] = tmp_code;
  94                 count++;
  95             }
  96             current_code = dict[current_code].parent_code;
  97             if ((current_code < 0) || (current_code > (DIC_INDEX_MAX - 1))) {
  98                 av_log(mlz->context, AV_LOG_ERROR, "MLZ dic index error.\n");
  99                 return count;
 100             }
 101             if (current_code > FIRST_CODE) {
 102                 parent_code = dict[current_code].parent_code;
 103                 offset = (dict[current_code].match_len) - 1;
 104                 if (parent_code < 0 || parent_code > DIC_INDEX_MAX-1) {
 105                     av_log(mlz->context, AV_LOG_ERROR, "MLZ dic index error.\n");
 106                     return count;
 107                 }
 108                 if (( offset > (DIC_INDEX_MAX - 1))) {
 109                     av_log(mlz->context, AV_LOG_ERROR, "MLZ dic offset error.\n");
 110                     return count;
 111                 }
 112             }
 113             break;
 114         }
 115     }
 116     return count;
 117 }
 118
 119 static int input_code(GetBitContext* gb, int len) {
 120     int tmp_code = 0;
 121     int i;
 122     for (i = 0; i < len; ++i) {
 123         tmp_code |= get_bits1(gb) << i;
 124     }
 125     return tmp_code;
 126 }
 127
 128 int ff_mlz_decompression(MLZ* mlz, GetBitContext* gb, int size, unsigned char *buff) {
 129     MLZDict *dict = mlz->dict;
 130     unsigned long output_chars;
 131     int string_code, last_string_code, char_code;
 132
 133     string_code = 0;
 134     char_code   = -1;
 135     last_string_code = -1;
 136     output_chars = 0;
 137
 138     while (output_chars < size) {
 139         string_code = input_code(gb, mlz->dic_code_bit);
 140         switch (string_code) {
 141             case FLUSH_CODE:
 142             case MAX_CODE:
 143                 ff_mlz_flush_dict(mlz);
 144                 char_code = -1;
 145                 last_string_code = -1;
 146                 break;
 147             case FREEZE_CODE:
 148                 mlz->freeze_flag = 1;
 149                 break;
 150             default:
 151                 if (string_code > mlz->current_dic_index_max) {
 152                     av_log(mlz->context, AV_LOG_ERROR, "String code %d exceeds maximum value of %d.\n", string_code, mlz->current_dic_index_max);
 153                     return output_chars;
 154                 }
 155                 if (string_code == (int) mlz->bump_code) {
 156                     ++mlz->dic_code_bit;
 157                     mlz->current_dic_index_max *= 2;
 158                     mlz->bump_code = mlz->current_dic_index_max - 1;
 159                 } else {
 160                     if (string_code >= mlz->next_code) {
 161                         int ret = decode_string(mlz, &buff[output_chars], last_string_code, &char_code, size - output_chars);
 162                         if (ret < 0 || ret > size - output_chars) {
 163                             av_log(mlz->context, AV_LOG_ERROR, "output chars overflow\n");
 164                             return output_chars;
 165                         }
 166                         output_chars += ret;
 167                         ret = decode_string(mlz, &buff[output_chars], char_code, &char_code, size - output_chars);
 168                         if (ret < 0 || ret > size - output_chars) {
 169                             av_log(mlz->context, AV_LOG_ERROR, "output chars overflow\n");
 170                             return output_chars;
 171                         }
 172                         output_chars += ret;
 173                         set_new_entry_dict(dict, mlz->next_code, last_string_code, char_code);
 174                         if (mlz->next_code >= TABLE_SIZE - 1) {
 175                             av_log(mlz->context, AV_LOG_ERROR, "Too many MLZ codes\n");
 176                             return output_chars;
 177                         }
 178                         mlz->next_code++;
 179                     } else {
 180                         int ret = decode_string(mlz, &buff[output_chars], string_code, &char_code, size - output_chars);
 181                         if (ret < 0 || ret > size - output_chars) {
 182                             av_log(mlz->context, AV_LOG_ERROR, "output chars overflow\n");
 183                             return output_chars;
 184                         }
 185                         output_chars += ret;
 186                         if (output_chars <= size && !mlz->freeze_flag) {
 187                             if (last_string_code != -1) {
 188                                 set_new_entry_dict(dict, mlz->next_code, last_string_code, char_code);
 189                                 if (mlz->next_code >= TABLE_SIZE - 1) {
 190                                     av_log(mlz->context, AV_LOG_ERROR, "Too many MLZ codes\n");
 191                                     return output_chars;
 192                                 }
 193                                 mlz->next_code++;
 194                             }
 195                         } else {
 196                             break;
 197                         }
 198                     }
 199                     last_string_code = string_code;
 200                 }
 201                 break;
 202         }
 203     }
 204     return output_chars;
 205 }