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++) {
  85                 free(buf->text[i]);
  86                 buf->text[i] = NULL;
  87         }
  88 }
  89
  90
  91 /**
  92  \brief scroll buffer one line up
  93  \param buf buffer to scroll
  94 */
  95 static void scroll_buffer(subtitle* buf)
  96 {
  97         int i;
  98
  99         while(buf->lines > cc_lines)
 100         {
 101                 free(buf->text[0]);
 102
 103                 for(i = 0; i < buf->lines - 1; i++) buf->text[i] = buf->text[i+1];
 104
 105                 buf->text[buf->lines-1] = NULL;
 106                 buf->lines--;
 107         }
 108 }
 109
 110
 111 void subcc_init(void)
 112 {
 113         int i;
 114         //printf("subcc_init(): initing...\n");
 115         build_char_table();
 116         for(i=0;i<SUB_MAX_TEXT;i++) {buf1.text[i]=buf2.text[i]=NULL;}
 117         buf1.lines=buf2.lines=0;
 118         fb=&buf1;
 119         bb=&buf2;
 120
 121         initialized=1;
 122 }
 123
 124
 125 static void display_buffer(subtitle *buf)
 126 {
 127         vo_sub = buf;
 128         vo_osd_changed(OSDTYPE_SUBTITLE);
 129 }
 130
 131
 132 static void append_char(char c)
 133 {
 134         if(!bb->lines) {bb->lines++; cursor_pos=0;}
 135         if(bb->text[bb->lines - 1]==NULL)
 136         {
 137                 bb->text[bb->lines - 1] = calloc(1, CC_MAX_LINE_LENGTH);
 138                 cursor_pos=0;
 139         }
 140
 141         if(c=='\n')
 142         {
 143                 if(cursor_pos>0 && bb->lines < SUB_MAX_TEXT)
 144                 {
 145                         bb->lines++;cursor_pos=0;
 146                         if(cc_mode==CC_ROLLUP){ //Carriage return - scroll buffer one line up
 147                                 bb->text[bb->lines - 1]=calloc(1, CC_MAX_LINE_LENGTH);
 148                                 scroll_buffer(bb);
 149                         }
 150                 }
 151         }
 152         else
 153         {
 154                 if(cursor_pos==CC_MAX_LINE_LENGTH-1)
 155                 {
 156                         fprintf(stderr,"CC: append_char() reached CC_MAX_LINE_LENGTH!\n");
 157                         return;
 158                 }
 159                 bb->text[bb->lines - 1][cursor_pos++]=c;
 160         }
 161         //In CC roll-up mode data should be shown immediately
 162         if(cc_mode==CC_ROLLUP) display_buffer(bb);
 163 }
 164
 165
 166 static void swap_buffers(void)
 167 {
 168         subtitle *foo;
 169         foo=fb;
 170         fb=bb;
 171         bb=foo;
 172 }
 173
 174
 175 static void cc_decode_EIA608(unsigned short int data)
 176 {
 177
 178   static unsigned short int lastcode=0x0000;
 179   unsigned char c1 = data & 0x7f;
 180   unsigned char c2 = (data >> 8) & 0x7f;
 181
 182   if (c1 & 0x60) {              /* normal character, 0x20 <= c1 <= 0x7f */
 183            append_char(chartbl[c1]);
 184            if(c2 & 0x60)        /*c2 might not be a normal char even if c1 is*/
 185                    append_char(chartbl[c2]);
 186   }
 187   else if (c1 & 0x10)           // control code / special char
 188   {
 189 //        int channel= (c1 & 0x08) >> 3;
 190           c1&=~0x08;
 191           if(data!=lastcode)
 192           {
 193                 if(c2 & 0x40) { /*PAC, Preamble Address Code */
 194                         append_char('\n'); /*FIXME properly interpret PACs*/
 195                 }
 196                 else
 197                         switch(c1)
 198                         {
 199                                 case 0x10:      break; // ext attribute
 200                                 case 0x11:
 201                                         if((c2 & 0x30)==0x30)
 202                                         {
 203                                                 //printf("[debug]:Special char (ignored)\n");
 204                                                 /*cc_decode_special_char()*/;
 205                                         }
 206                                         else if (c2 & 0x20)
 207                                         {
 208                                                 //printf("[debug]: midrow_attr (ignored)\n");
 209                                                 /*cc_decode_midrow_attr()*/;
 210                                         }
 211                                         break;
 212                                 case 0x14:
 213                                         switch(c2)
 214                                         {
 215                                                 case 0x00: //CC roll-on mode
 216                                                            cc_mode=CC_ROLLON;
 217                                                            break;
 218                                                 case 0x25: //CC roll-up, 2 rows
 219                                                 case 0x26: //CC roll-up, 3 rows
 220                                                 case 0x27: //CC roll-up, 4 rows
 221                                                            cc_lines=c2-0x23;
 222                                                            cc_mode=CC_ROLLUP;
 223                                                            break;
 224                                                 case 0x2C: display_buffer(NULL); //EDM
 225                                                            clear_buffer(fb); break;
 226                                                 case 0x2d: append_char('\n');   //carriage return
 227                                                            break;
 228                                                 case 0x2e: clear_buffer(bb);    //ENM
 229                                                            break;
 230                                                 case 0x2f: swap_buffers();      //Swap buffers
 231                                                            display_buffer(fb);
 232                                                            clear_buffer(bb);
 233                                                            break;
 234                                         }
 235                                         break;
 236                                 case 0x17:
 237                                         if( c2>=0x21 && c2<=0x23) //TAB
 238                                         {
 239                                                 break;
 240                                         }
 241                         }
 242           }
 243   }
 244   lastcode=data;
 245 }
 246
 247 static void subcc_decode(unsigned char *inputbuffer, unsigned int inputlength)
 248 {
 249   /* The first number may denote a channel number. I don't have the
 250    * EIA-708 standard, so it is hard to say.
 251    * From what I could figure out so far, the general format seems to be:
 252    *
 253    * repeat
 254    *
 255    *   0xfe starts 2 byte sequence of unknown purpose. It might denote
 256    *        field #2 in line 21 of the VBI. We'll ignore it for the
 257    *        time being.
 258    *
 259    *   0xff starts 2 byte EIA-608 sequence, field #1 in line 21 of the VBI.
 260    *        Followed by a 3-code triplet that starts either with 0xff or
 261    *        0xfe. In either case, the following triplet needs to be ignored
 262    *        for line 21, field 1.
 263    *
 264    *   0x00 is padding, followed by 2 more 0x00.
 265    *
 266    *   0x01 always seems to appear at the beginning, always seems to
 267    *        be followed by 0xf8, 8-bit number.
 268    *        The lower 7 bits of this 8-bit number seem to denote the
 269    *        number of code triplets that follow.
 270    *        The most significant bit denotes whether the Line 21 field 1
 271    *        captioning information is at odd or even triplet offsets from this
 272    *        beginning triplet. 1 denotes odd offsets, 0 denotes even offsets.
 273    *
 274    *        Most captions are encoded with odd offsets, so this is what we
 275    *        will assume.
 276    *
 277    * until end of packet
 278    */
 279   unsigned char *current = inputbuffer;
 280   unsigned int curbytes = 0;
 281   unsigned char data1, data2;
 282   unsigned char cc_code;
 283   int odd_offset = 1;
 284
 285   while (curbytes < inputlength) {
 286     int skip = 2;
 287
 288     cc_code = current[0];
 289
 290     if (inputlength - curbytes < 2) {
 291 #ifdef LOG_DEBUG
 292       fprintf(stderr, "Not enough data for 2-byte CC encoding\n");
 293 #endif
 294       break;
 295     }
 296
 297     data1 = current[1];
 298     data2 = current[2];
 299     current++; curbytes++;
 300
 301     switch (cc_code) {
 302     case 0xfe:
 303       /* expect 2 byte encoding (perhaps CC3, CC4?) */
 304       /* ignore for time being */
 305       break;
 306
 307     case 0xff:
 308       /* expect EIA-608 CC1/CC2 encoding */
 309       // FIXME check parity!
 310       // Parity check omitted assuming we are reading from a DVD and therefore
 311       // we should encounter no "transmission errors".
 312       cc_decode_EIA608(data1 | (data2 << 8));
 313       skip = 5;
 314       break;
 315
 316     case 0x00:
 317       /* This seems to be just padding */
 318       break;
 319
 320     case 0x01:
 321       odd_offset = data2 & 0x80;
 322       if (!odd_offset)
 323         skip = 5;
 324       break;
 325
 326     default:
 327 //#ifdef LOG_DEBUG
 328       fprintf(stderr, "Unknown CC encoding: %x\n", cc_code);
 329 //#endif
 330       break;
 331     }
 332     current += skip;
 333     curbytes += skip;
 334   }
 335 }
 336
 337
 338 void subcc_process_data(unsigned char *inputdata,unsigned int len)
 339 {
 340         if(!subcc_enabled) return;
 341         if(!initialized) subcc_init();
 342
 343         subcc_decode(inputdata, len);
 344 }