sub_cc.c

   1 /*
   2  * decoder for Closed Captions
   3  *
   4  * This decoder relies on MPlayer's OSD to display subtitles.
   5  * Be warned that decoding is somewhat preliminary, though it basically works.
   6  *
   7  * Most notably, only the text information is decoded as of now, discarding
   8  * color, background and position info (see source below).
   9  *
  10  * uses source from the xine closed captions decoder
  11  *
  12  * Copyright (C) 2002 Matteo Giani
  13  *
  14  * This file is part of MPlayer.
  15  *
  16  * MPlayer is free software; you can redistribute it and/or modify
  17  * it under the terms of the GNU General Public License as published by
  18  * the Free Software Foundation; either version 2 of the License, or
  19  * (at your option) any later version.
  20  *
  21  * MPlayer is distributed in the hope that it will be useful,
  22  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  23  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  24  * GNU General Public License for more details.
  25  *
  26  * You should have received a copy of the GNU General Public License along
  27  * with MPlayer; if not, write to the Free Software Foundation, Inc.,
  28  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  29  */
  30
  31 #include <stdio.h>
  32 #include <stdlib.h>
  33 #include <string.h>
  34
  35 #include "config.h"
  36 #include "sub_cc.h"
  37
  38 #include "subreader.h"
  39
  40 #include "libvo/video_out.h"
  41 #include "libvo/sub.h"
  42
  43
  44 #define CC_MAX_LINE_LENGTH 64
  45
  46 static char chartbl[128];
  47
  48 static subtitle buf1,buf2;
  49 static subtitle *fb,*bb;
  50
  51 static unsigned int cursor_pos=0;
  52
  53 static int initialized=0;
  54
  55 #define CC_ROLLON 1
  56 #define CC_ROLLUP 2
  57
  58 static int cc_mode=CC_ROLLON;
  59 static int cc_lines=4; ///< number of visible rows in CC roll-up mode, not used in CC roll-on mode
  60
  61 static void build_char_table(void)
  62 {
  63   int i;
  64   /* first the normal ASCII codes */
  65   for (i = 0; i < 128; i++)
  66     chartbl[i] = (char) i;
  67   /* now the special codes */
  68   chartbl[0x2a] = 'á';
  69   chartbl[0x5c] = 'é';
  70   chartbl[0x5e] = 'í';
  71   chartbl[0x5f] = 'ó';
  72   chartbl[0x60] = 'ú';
  73   chartbl[0x7b] = 'ç';
  74   chartbl[0x7c] = '÷';
  75   chartbl[0x7d] = 'Ñ';
  76   chartbl[0x7e] = 'ñ';
  77   chartbl[0x7f] = '¤';    /* FIXME: this should be a solid block */
  78 }
  79
  80 static void clear_buffer(subtitle *buf)
  81 {
  82         int i;
  83         buf->lines=0;
  84         for(i=0;i<SUB_MAX_TEXT;i++) if(buf->text[i]) {free(buf->text[i]);buf->text[i]=NULL;}
  85 }
  86
  87
  88 /**
  89  \brief scroll buffer one line up
  90  \param buf buffer to scroll
  91 */
  92 static void scroll_buffer(subtitle* buf)
  93 {
  94         int i;
  95
  96         while(buf->lines > cc_lines)
  97         {
  98                 if(buf->text[0]) free(buf->text[0]);
  99
 100                 for(i = 0; i < (buf->lines - 1); i++) buf->text[i] = buf->text[i+1];
 101
 102                 buf->text[buf->lines-1] = NULL;
 103                 buf->lines--;
 104         }
 105 }
 106
 107
 108 void subcc_init(void)
 109 {
 110         int i;
 111         //printf("subcc_init(): initing...\n");
 112         build_char_table();
 113         for(i=0;i<SUB_MAX_TEXT;i++) {buf1.text[i]=buf2.text[i]=NULL;}
 114         buf1.lines=buf2.lines=0;
 115         fb=&buf1;
 116         bb=&buf2;
 117
 118         initialized=1;
 119 }
 120
 121
 122 static void display_buffer(subtitle *buf)
 123 {
 124         vo_sub = buf;
 125         vo_osd_changed(OSDTYPE_SUBTITLE);
 126 }
 127
 128
 129 static void append_char(char c)
 130 {
 131         if(!bb->lines) {bb->lines++; cursor_pos=0;}
 132         if(bb->text[bb->lines - 1]==NULL)
 133         {
 134                 bb->text[bb->lines - 1]=malloc(CC_MAX_LINE_LENGTH);
 135                 memset(bb->text[bb->lines - 1],0,CC_MAX_LINE_LENGTH);
 136                 cursor_pos=0;
 137         }
 138
 139         if(c=='\n')
 140         {
 141                 if(cursor_pos>0 && bb->lines < SUB_MAX_TEXT)
 142                 {
 143                         bb->lines++;cursor_pos=0;
 144                         if(cc_mode==CC_ROLLUP){ //Carriage return - scroll buffer one line up
 145                                 bb->text[bb->lines - 1]=calloc(1, CC_MAX_LINE_LENGTH);
 146                                 scroll_buffer(bb);
 147                         }
 148                 }
 149         }
 150         else
 151         {
 152                 if(cursor_pos==CC_MAX_LINE_LENGTH-1)
 153                 {
 154                         fprintf(stderr,"CC: append_char() reached CC_MAX_LINE_LENGTH!\n");
 155                         return;
 156                 }
 157                 bb->text[bb->lines - 1][cursor_pos++]=c;
 158         }
 159         //In CC roll-up mode data should be shown immediately
 160         if(cc_mode==CC_ROLLUP) display_buffer(bb);
 161 }
 162
 163
 164 static void swap_buffers(void)
 165 {
 166         subtitle *foo;
 167         foo=fb;
 168         fb=bb;
 169         bb=foo;
 170 }
 171
 172
 173 static void cc_decode_EIA608(unsigned short int data)
 174 {
 175
 176   static unsigned short int lastcode=0x0000;
 177   unsigned char c1 = data & 0x7f;
 178   unsigned char c2 = (data >> 8) & 0x7f;
 179
 180   if (c1 & 0x60) {              /* normal character, 0x20 <= c1 <= 0x7f */
 181            append_char(chartbl[c1]);
 182            if(c2 & 0x60)        /*c2 might not be a normal char even if c1 is*/
 183                    append_char(chartbl[c2]);
 184   }
 185   else if (c1 & 0x10)           // control code / special char
 186   {
 187 //        int channel= (c1 & 0x08) >> 3;
 188           c1&=~0x08;
 189           if(data!=lastcode)
 190           {
 191                 if(c2 & 0x40) { /*PAC, Preamble Address Code */
 192                         append_char('\n'); /*FIXME properly interpret PACs*/
 193                 }
 194                 else
 195                         switch(c1)
 196                         {
 197                                 case 0x10:      break; // ext attribute
 198                                 case 0x11:
 199                                         if((c2 & 0x30)==0x30)
 200                                         {
 201                                                 //printf("[debug]:Special char (ignored)\n");
 202                                                 /*cc_decode_special_char()*/;
 203                                         }
 204                                         else if (c2 & 0x20)
 205                                         {
 206                                                 //printf("[debug]: midrow_attr (ignored)\n");
 207                                                 /*cc_decode_midrow_attr()*/;
 208                                         }
 209                                         break;
 210                                 case 0x14:
 211                                         switch(c2)
 212                                         {
 213                                                 case 0x00: //CC roll-on mode
 214                                                            cc_mode=CC_ROLLON;
 215                                                            break;
 216                                                 case 0x25: //CC roll-up, 2 rows
 217                                                 case 0x26: //CC roll-up, 3 rows
 218                                                 case 0x27: //CC roll-up, 4 rows
 219                                                            cc_lines=c2-0x23;
 220                                                            cc_mode=CC_ROLLUP;
 221                                                            break;
 222                                                 case 0x2C: display_buffer(NULL); //EDM
 223                                                            clear_buffer(fb); break;
 224                                                 case 0x2d: append_char('\n');   //carriage return
 225                                                            break;
 226                                                 case 0x2e: clear_buffer(bb);    //ENM
 227                                                            break;
 228                                                 case 0x2f: swap_buffers();      //Swap buffers
 229                                                            display_buffer(fb);
 230                                                            clear_buffer(bb);
 231                                                            break;
 232                                         }
 233                                         break;
 234                                 case 0x17:
 235                                         if( c2>=0x21 && c2<=0x23) //TAB
 236                                         {
 237                                                 break;
 238                                         }
 239                         }
 240           }
 241   }
 242   lastcode=data;
 243 }
 244
 245 static void subcc_decode(unsigned char *inputbuffer, unsigned int inputlength)
 246 {
 247   /* The first number may denote a channel number. I don't have the
 248    * EIA-708 standard, so it is hard to say.
 249    * From what I could figure out so far, the general format seems to be:
 250    *
 251    * repeat
 252    *
 253    *   0xfe starts 2 byte sequence of unknown purpose. It might denote
 254    *        field #2 in line 21 of the VBI. We'll ignore it for the
 255    *        time being.
 256    *
 257    *   0xff starts 2 byte EIA-608 sequence, field #1 in line 21 of the VBI.
 258    *        Followed by a 3-code triplet that starts either with 0xff or
 259    *        0xfe. In either case, the following triplet needs to be ignored
 260    *        for line 21, field 1.
 261    *
 262    *   0x00 is padding, followed by 2 more 0x00.
 263    *
 264    *   0x01 always seems to appear at the beginning, always seems to
 265    *        be followed by 0xf8, 8-bit number.
 266    *        The lower 7 bits of this 8-bit number seem to denote the
 267    *        number of code triplets that follow.
 268    *        The most significant bit denotes whether the Line 21 field 1
 269    *        captioning information is at odd or even triplet offsets from this
 270    *        beginning triplet. 1 denotes odd offsets, 0 denotes even offsets.
 271    *
 272    *        Most captions are encoded with odd offsets, so this is what we
 273    *        will assume.
 274    *
 275    * until end of packet
 276    */
 277   unsigned char *current = inputbuffer;
 278   unsigned int curbytes = 0;
 279   unsigned char data1, data2;
 280   unsigned char cc_code;
 281   int odd_offset = 1;
 282
 283   while (curbytes < inputlength) {
 284     int skip = 2;
 285
 286     cc_code = *(current);
 287
 288     if (inputlength - curbytes < 2) {
 289 #ifdef LOG_DEBUG
 290       fprintf(stderr, "Not enough data for 2-byte CC encoding\n");
 291 #endif
 292       break;
 293     }
 294
 295     data1 = *(current+1);
 296     data2 = *(current + 2);
 297     current++; curbytes++;
 298
 299     switch (cc_code) {
 300     case 0xfe:
 301       /* expect 2 byte encoding (perhaps CC3, CC4?) */
 302       /* ignore for time being */
 303       break;
 304
 305     case 0xff:
 306       /* expect EIA-608 CC1/CC2 encoding */
 307       // FIXME check parity!
 308       // Parity check omitted assuming we are reading from a DVD and therefore
 309       // we should encounter no "transmission errors".
 310       cc_decode_EIA608(data1 | (data2 << 8));
 311       skip = 5;
 312       break;
 313
 314     case 0x00:
 315       /* This seems to be just padding */
 316       break;
 317
 318     case 0x01:
 319       odd_offset = data2 & 0x80;
 320       if (!odd_offset)
 321         skip = 5;
 322       break;
 323
 324     default:
 325 //#ifdef LOG_DEBUG
 326       fprintf(stderr, "Unknown CC encoding: %x\n", cc_code);
 327 //#endif
 328       break;
 329     }
 330     current += skip;
 331     curbytes += skip;
 332   }
 333 }
 334
 335
 336 void subcc_process_data(unsigned char *inputdata,unsigned int len)
 337 {
 338         if(!subcc_enabled) return;
 339         if(!initialized) subcc_init();
 340
 341         subcc_decode(inputdata, len);
 342 }